From 92ac5891870614f35a8d106c90010a5c3e2daa65 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 2 May 2009 13:42:13 +0000 Subject: [PATCH 001/296] CDVDiso: Now uses automake, is C++, isn't split into two files, and doesn't change around the working directory on us. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1118 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Plugins.cpp | 10 +- plugins/CDVDiso/build.sh | 33 +- plugins/CDVDiso/src/CDVDiso.h | 57 +- .../CDVDiso/src/{CDVDisop.c => CDVDisop.cpp} | 40 +- plugins/CDVDiso/src/Linux/Config.c | 82 --- plugins/CDVDiso/src/Linux/Config.cpp | 326 +++++++++++ plugins/CDVDiso/src/Linux/Config.h | 61 ++ plugins/CDVDiso/src/Linux/Linux.c | 108 ---- .../CDVDiso/src/Linux/{conf.c => Linux.cpp} | 339 +----------- plugins/CDVDiso/src/Linux/interface.c | 25 + plugins/CDVDiso/src/Makefile.am | 31 ++ .../CDVDiso/src/Windows/CDVDiso_vs2008.vcproj | 8 +- .../src/Windows/{Config.c => Config.cpp} | 0 .../src/Windows/{Win32.c => Win32.cpp} | 4 +- plugins/CDVDiso/src/Windows/svnrev.h | 18 + plugins/CDVDiso/src/configure.ac | 83 +++ plugins/CDVDiso/src/install-sh | 519 ++++++++++++++++++ plugins/CDVDiso/src/{libiso.c => libiso.cpp} | 0 plugins/CDVDiso/src/missing | 367 +++++++++++++ plugins/CDVDiso/src/mkiso/Makefile | 27 - .../CDVDiso/src/mkiso/{mkiso.c => mkiso.cpp} | 5 +- 21 files changed, 1541 insertions(+), 602 deletions(-) rename plugins/CDVDiso/src/{CDVDisop.c => CDVDisop.cpp} (92%) mode change 100644 => 100755 delete mode 100644 plugins/CDVDiso/src/Linux/Config.c create mode 100755 plugins/CDVDiso/src/Linux/Config.cpp delete mode 100644 plugins/CDVDiso/src/Linux/Linux.c rename plugins/CDVDiso/src/Linux/{conf.c => Linux.cpp} (65%) mode change 100644 => 100755 create mode 100755 plugins/CDVDiso/src/Makefile.am rename plugins/CDVDiso/src/Windows/{Config.c => Config.cpp} (100%) mode change 100644 => 100755 rename plugins/CDVDiso/src/Windows/{Win32.c => Win32.cpp} (99%) mode change 100644 => 100755 create mode 100755 plugins/CDVDiso/src/Windows/svnrev.h create mode 100755 plugins/CDVDiso/src/configure.ac create mode 100755 plugins/CDVDiso/src/install-sh rename plugins/CDVDiso/src/{libiso.c => libiso.cpp} (100%) mode change 100644 => 100755 create mode 100755 plugins/CDVDiso/src/missing delete mode 100644 plugins/CDVDiso/src/mkiso/Makefile rename plugins/CDVDiso/src/mkiso/{mkiso.c => mkiso.cpp} (98%) mode change 100644 => 100755 diff --git a/pcsx2/Plugins.cpp b/pcsx2/Plugins.cpp index b1c64c3949..2a9fe78880 100644 --- a/pcsx2/Plugins.cpp +++ b/pcsx2/Plugins.cpp @@ -637,7 +637,7 @@ int InitPlugins() // throw Exception::InvalidOperation( "Bad coder mojo - InitPlugins called prior to plugins having been loaded." ); #ifndef _WIN32 - chdir(MAIN_DIR); + //chdir(MAIN_DIR); #endif int ret; @@ -712,7 +712,7 @@ int OpenPlugins(const char* pTitleFilename) #ifndef _WIN32 // change dir so that CDVD can find its config file - char file[255], pNewTitle[255]; + /*char file[255], pNewTitle[255]; chdir(MAIN_DIR); chdir(Config.PluginsDir); @@ -721,7 +721,7 @@ int OpenPlugins(const char* pTitleFilename) // because we are changing the dir, we have to set a new title if it is a relative dir sprintf(pNewTitle, "%s/%s", file, pTitleFilename); pTitleFilename = pNewTitle; - } + }*/ #endif if( !OpenStatus.CDVD ) @@ -808,14 +808,14 @@ int OpenPlugins(const char* pTitleFilename) } #ifndef _WIN32 - chdir(MAIN_DIR); + //chdir(MAIN_DIR); #endif return 0; OpenError: ClosePlugins( true ); #ifndef _WIN32 - chdir(MAIN_DIR); + //chdir(MAIN_DIR); #endif return -1; diff --git a/plugins/CDVDiso/build.sh b/plugins/CDVDiso/build.sh index 3b953bbd62..fbc599f31d 100644 --- a/plugins/CDVDiso/build.sh +++ b/plugins/CDVDiso/build.sh @@ -1,17 +1,34 @@ #!/bin/sh -curdir=`pwd` - echo ---------------- echo Building CDVDiso echo ---------------- -cd ${curdir}/src/Linux -make clean -make $@ +curdir=`pwd` -# copy the files -if [ -s cfgCDVDiso ] && [ -s libCDVDiso.so ] + +if test "${CDVDisoOPTIONS+set}" != set ; then +export CDVDisoOPTIONS="" +fi + +cd src + +if [ $# -gt 0 ] && [ $1 = "all" ] then -cp cfgCDVDiso libCDVDiso.so ${PCSX2PLUGINS} + +aclocal +automake -a +autoconf + +./configure ${CDVDisoOPTIONS} --prefix=${PCSX2PLUGINS} +make clean +make install + +else +make $@ +fi + +if [ $? -ne 0 ] +then +exit 1 fi diff --git a/plugins/CDVDiso/src/CDVDiso.h b/plugins/CDVDiso/src/CDVDiso.h index 0c4c173a0f..faddb1b616 100644 --- a/plugins/CDVDiso/src/CDVDiso.h +++ b/plugins/CDVDiso/src/CDVDiso.h @@ -16,6 +16,8 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#ifndef __CDVDISO_H__ +#define __CDVDISO_H__ #ifdef _MSC_VER #pragma warning(disable:4018) @@ -28,50 +30,50 @@ #ifdef __cplusplus #ifdef _MSC_VER -#define EXPORT_C(type) extern "C" __declspec(dllexport) type CALLBACK +#define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK #else -#define EXPORT_C(type) extern "C" type +#define EXPORT_C_(type) extern "C" type #endif #else #ifdef _MSC_VER -#define EXPORT_C(type) __declspec(dllexport) type __stdcall +#define EXPORT_C_(type) __declspec(dllexport) type __stdcall #else -#define EXPORT_C(type) type +#define EXPORT_C_(type) type #endif #endif -EXPORT_C(u32) PS2EgetLibType(); -EXPORT_C(u32) PS2EgetLibVersion2(u32 type); -EXPORT_C(char*) PS2EgetLibName(); +EXPORT_C_(u32) PS2EgetLibType(); +EXPORT_C_(u32) PS2EgetLibVersion2(u32 type); +EXPORT_C_(char*) PS2EgetLibName(); -EXPORT_C(s32) CDVDinit(); -EXPORT_C(s32) CDVDopen(const char* pTitleFilename); -EXPORT_C(void) CDVDclose(); -EXPORT_C(void) CDVDshutdown(); -EXPORT_C(s32) CDVDreadTrack(u32 lsn, int mode); +EXPORT_C_(s32) CDVDinit(); +EXPORT_C_(s32) CDVDopen(const char* pTitleFilename); +EXPORT_C_(void) CDVDclose(); +EXPORT_C_(void) CDVDshutdown(); +EXPORT_C_(s32) CDVDreadTrack(u32 lsn, int mode); // return can be NULL (for async modes) -EXPORT_C(u8*) CDVDgetBuffer(); +EXPORT_C_(u8*) CDVDgetBuffer(); -EXPORT_C(s32) CDVDreadSubQ(u32 lsn, cdvdSubQ* subq);//read subq from disc (only cds have subq data) -EXPORT_C(s32) CDVDgetTN(cdvdTN *Buffer); //disk information -EXPORT_C(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer); //track info: min,sec,frame,type -EXPORT_C(s32) CDVDgetTOC(void* toc); //gets ps2 style toc from disc -EXPORT_C(s32) CDVDgetDiskType(); //CDVD_TYPE_xxxx -EXPORT_C(s32) CDVDgetTrayStatus(); //CDVD_TRAY_xxxx -EXPORT_C(s32) CDVDctrlTrayOpen(); //open disc tray -EXPORT_C(s32) CDVDctrlTrayClose(); //close disc tray +EXPORT_C_(s32) CDVDreadSubQ(u32 lsn, cdvdSubQ* subq);//read subq from disc (only cds have subq data) +EXPORT_C_(s32) CDVDgetTN(cdvdTN *Buffer); //disk information +EXPORT_C_(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer); //track info: min,sec,frame,type +EXPORT_C_(s32) CDVDgetTOC(void* toc); //gets ps2 style toc from disc +EXPORT_C_(s32) CDVDgetDiskType(); //CDVD_TYPE_xxxx +EXPORT_C_(s32) CDVDgetTrayStatus(); //CDVD_TRAY_xxxx +EXPORT_C_(s32) CDVDctrlTrayOpen(); //open disc tray +EXPORT_C_(s32) CDVDctrlTrayClose(); //close disc tray // extended funcs -EXPORT_C(void) CDVDconfigure(); -EXPORT_C(void) CDVDabout(); -EXPORT_C(s32) CDVDtest(); -EXPORT_C(void) CDVDnewDiskCB(void (*callback)()); +EXPORT_C_(void) CDVDconfigure(); +EXPORT_C_(void) CDVDabout(); +EXPORT_C_(s32) CDVDtest(); +EXPORT_C_(void) CDVDnewDiskCB(void (*callback)()); #define CDVD_LOG __Log extern FILE *cdvdLog; @@ -104,6 +106,10 @@ extern _cdIso cdIso[8]; #define MSF2SECT(m,s,f) (((m)*60+(s)-2)*75+(f)) +extern const unsigned char version; +extern const unsigned char revision; +extern const unsigned char build; + extern char IsoFile[256]; extern char IsoCWD[256]; extern char CdDev[256]; @@ -130,3 +136,4 @@ void UpdateZmode(); void CfgOpenFile(); void SysMessage(char *fmt, ...); +#endif \ No newline at end of file diff --git a/plugins/CDVDiso/src/CDVDisop.c b/plugins/CDVDiso/src/CDVDisop.cpp old mode 100644 new mode 100755 similarity index 92% rename from plugins/CDVDiso/src/CDVDisop.c rename to plugins/CDVDiso/src/CDVDisop.cpp index 240b6a9edb..bae8f753c9 --- a/plugins/CDVDiso/src/CDVDisop.c +++ b/plugins/CDVDiso/src/CDVDisop.cpp @@ -55,7 +55,7 @@ char *LibName = "Linuz Iso CDVD "; const unsigned char version = PS2E_CDVD_VERSION; const unsigned char revision = 0; -const unsigned char build = 8; +const unsigned char build = 9; unsigned char cdbuffer[CD_FRAMESIZE_RAW * 10] = {0}; @@ -80,17 +80,17 @@ void lba_to_msf(s32 lba, u8* m, u8* s, u8* f) #define itob(i) ((i)/10*16 + (i)%10) /* u_char to BCD */ -EXPORT_C(char*) PS2EgetLibName() +EXPORT_C_(char*) PS2EgetLibName() { return LibName; } -EXPORT_C(u32) PS2EgetLibType() +EXPORT_C_(u32) PS2EgetLibType() { return PS2E_LT_CDVD; } -EXPORT_C(u32) PS2EgetLibVersion2(u32 type) +EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) { return (version << 16) | (revision << 8) | build; } @@ -111,7 +111,7 @@ void __Log(char *fmt, ...) #endif -EXPORT_C(s32) CDVDinit() +EXPORT_C_(s32) CDVDinit() { #ifdef _DEBUG cdvdLog = fopen("logs/cdvdLog.txt", "w"); @@ -133,7 +133,7 @@ EXPORT_C(s32) CDVDinit() return 0; } -EXPORT_C(void) CDVDshutdown() +EXPORT_C_(void) CDVDshutdown() { cdvdCurrentIso[0] = 0; #ifdef CDVD_LOG @@ -141,7 +141,7 @@ EXPORT_C(void) CDVDshutdown() #endif } -EXPORT_C(s32) CDVDopen(const char* pTitle) +EXPORT_C_(s32) CDVDopen(const char* pTitle) { LoadConf(); @@ -224,7 +224,7 @@ EXPORT_C(s32) CDVDopen(const char* pTitle) return 0; } -EXPORT_C(void) CDVDclose() +EXPORT_C_(void) CDVDclose() { strcpy(cdvdCurrentIso, IsoFile); @@ -233,7 +233,7 @@ EXPORT_C(void) CDVDclose() if (fdump != NULL) isoClose(fdump); } -EXPORT_C(s32) CDVDreadSubQ(u32 lsn, cdvdSubQ* subq) +EXPORT_C_(s32) CDVDreadSubQ(u32 lsn, cdvdSubQ* subq) { // fake it u8 min, sec, frm; @@ -256,7 +256,7 @@ EXPORT_C(s32) CDVDreadSubQ(u32 lsn, cdvdSubQ* subq) return 0; } -EXPORT_C(s32) CDVDgetTN(cdvdTN *Buffer) +EXPORT_C_(s32) CDVDgetTN(cdvdTN *Buffer) { Buffer->strack = 1; Buffer->etrack = 1; @@ -264,7 +264,7 @@ EXPORT_C(s32) CDVDgetTN(cdvdTN *Buffer) return 0; } -EXPORT_C(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer) +EXPORT_C_(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer) { if (Track == 0) { @@ -280,7 +280,7 @@ EXPORT_C(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer) } static int layer1start = -1; -EXPORT_C(s32) CDVDgetTOC(void* toc) +EXPORT_C_(s32) CDVDgetTOC(void* toc) { u8 type = CDVDgetDiskType(); u8* tocBuff = (u8*)toc; @@ -289,8 +289,6 @@ EXPORT_C(s32) CDVDgetTOC(void* toc) if (type == CDVD_TYPE_DVDV || type == CDVD_TYPE_PS2DVD) { - int i; - // get dvd structure format // scsi command 0x43 memset(tocBuff, 0, 2048); @@ -424,7 +422,7 @@ EXPORT_C(s32) CDVDgetTOC(void* toc) return 0; } -EXPORT_C(s32) CDVDreadTrack(u32 lsn, int mode) +EXPORT_C_(s32) CDVDreadTrack(u32 lsn, int mode) { int _lsn = lsn; @@ -461,32 +459,32 @@ EXPORT_C(s32) CDVDreadTrack(u32 lsn, int mode) return 0; } -EXPORT_C(u8*) CDVDgetBuffer() +EXPORT_C_(u8*) CDVDgetBuffer() { return pbuffer; } -EXPORT_C(s32) CDVDgetDiskType() +EXPORT_C_(s32) CDVDgetDiskType() { return cdtype; } -EXPORT_C(s32) CDVDgetTrayStatus() +EXPORT_C_(s32) CDVDgetTrayStatus() { return CDVD_TRAY_CLOSE; } -EXPORT_C(s32) CDVDctrlTrayOpen() +EXPORT_C_(s32) CDVDctrlTrayOpen() { return 0; } -EXPORT_C(s32) CDVDctrlTrayClose() +EXPORT_C_(s32) CDVDctrlTrayClose() { return 0; } -EXPORT_C(s32) CDVDtest() +EXPORT_C_(s32) CDVDtest() { if (*IsoFile == 0) return 0; diff --git a/plugins/CDVDiso/src/Linux/Config.c b/plugins/CDVDiso/src/Linux/Config.c deleted file mode 100644 index ae1aa1a4af..0000000000 --- a/plugins/CDVDiso/src/Linux/Config.c +++ /dev/null @@ -1,82 +0,0 @@ -/* CDVDiso - * Copyright (C) 2002-2004 CDVDiso Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include -#include -#include -#include -#include - -#include "CDVDiso.h" - - -const char *s_strIniPath="../inis/CDVDiso.ini"; - -void LoadConf() -{ - FILE *f; - char cfg[256]; - - //sprintf(cfg, "%s/.PS2E/CDVDiso.cfg", getenv("HOME")); - strcpy(cfg, s_strIniPath); - f = fopen(cfg, "r"); - - if (f == NULL) - { - printf("Unable to load %s\n", cfg); - strcpy(IsoFile, DEV_DEF); - strcpy(CdDev, CDDEV_DEF); - BlockDump = 0; - SaveConf(); - return; - } - - fscanf(f, "IsoFile = %[^\n]\n", IsoFile); - fscanf(f, "CdDev = %[^\n]\n", CdDev); - fscanf(f, "BlockDump = %d\n", &BlockDump); - - if (!strncmp(IsoFile, "CdDev =", 9)) *IsoFile = 0; // quick fix - if (*CdDev == 0) strcpy(CdDev, CDDEV_DEF); - - fclose(f); -} - -void SaveConf() -{ - FILE *f; - char cfg[256]; - - //sprintf(cfg, "%s/.PS2E", getenv("HOME")); - - //mkdir(cfg, 0755); - //sprintf(cfg, "%s/.PS2E/CDVDiso.cfg", getenv("HOME")); - strcpy(cfg, s_strIniPath); - - f = fopen(cfg, "w"); - if (f == NULL) - { - printf("Unable to save %s\n", cfg); - return; - } - - fprintf(f, "IsoFile = %s\n", IsoFile); - fprintf(f, "CdDev = %s\n", CdDev); - fprintf(f, "BlockDump = %d\n", BlockDump); - fclose(f); -} - diff --git a/plugins/CDVDiso/src/Linux/Config.cpp b/plugins/CDVDiso/src/Linux/Config.cpp new file mode 100755 index 0000000000..1f0b440231 --- /dev/null +++ b/plugins/CDVDiso/src/Linux/Config.cpp @@ -0,0 +1,326 @@ +/* CDVDiso + * Copyright (C) 2002-2004 CDVDiso Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "Config.h" + +const char *s_strIniPath="./inis/CDVDiso.ini"; +GtkWidget *AboutDlg, *ConfDlg, *MsgDlg, *FileSel; +GtkWidget *Edit, *CdEdit; + +void LoadConf() +{ + FILE *f; + char cfg[256]; + + //sprintf(cfg, "%s/.PS2E/CDVDiso.cfg", getenv("HOME")); + strcpy(cfg, s_strIniPath); + f = fopen(cfg, "r"); + + if (f == NULL) + { + printf("Unable to load %s\n", cfg); + strcpy(IsoFile, DEV_DEF); + strcpy(CdDev, CDDEV_DEF); + BlockDump = 0; + SaveConf(); + return; + } + + fscanf(f, "IsoFile = %[^\n]\n", IsoFile); + fscanf(f, "CdDev = %[^\n]\n", CdDev); + fscanf(f, "BlockDump = %d\n", &BlockDump); + + if (!strncmp(IsoFile, "CdDev =", 9)) *IsoFile = 0; // quick fix + if (*CdDev == 0) strcpy(CdDev, CDDEV_DEF); + + fclose(f); +} + +void SaveConf() +{ + FILE *f; + char cfg[256]; + + //sprintf(cfg, "%s/.PS2E", getenv("HOME")); + + //mkdir(cfg, 0755); + //sprintf(cfg, "%s/.PS2E/CDVDiso.cfg", getenv("HOME")); + strcpy(cfg, s_strIniPath); + + f = fopen(cfg, "w"); + if (f == NULL) + { + printf("Unable to save %s\n", cfg); + return; + } + + fprintf(f, "IsoFile = %s\n", IsoFile); + fprintf(f, "CdDev = %s\n", CdDev); + fprintf(f, "BlockDump = %d\n", BlockDump); + fclose(f); +} + +void SysMessage(char *fmt, ...) +{ + va_list list; + char tmp[256]; + char cmd[256]; + + va_start(list, fmt); + vsprintf(tmp, fmt, list); + va_end(list); + + sprintf(cmd, "message \"%s\"", tmp); + SysMessageLoc(tmp); +} + +void OnFile_Ok() +{ + gchar *File; + + gtk_widget_hide(FileSel); + File = gtk_file_selection_get_filename(GTK_FILE_SELECTION(FileSel)); + strcpy(IsoFile, File); + gtk_main_quit(); +} + +void OnFile_Cancel() +{ + gtk_widget_hide(FileSel); + gtk_main_quit(); +} + +void CfgOpenFile() +{ + GtkWidget *Ok, *Cancel; + + FileSel = gtk_file_selection_new("Select Iso File"); + + Ok = GTK_FILE_SELECTION(FileSel)->ok_button; + gtk_signal_connect(GTK_OBJECT(Ok), "clicked", + GTK_SIGNAL_FUNC(OnFile_Ok), NULL); + gtk_widget_show(Ok); + + Cancel = GTK_FILE_SELECTION(FileSel)->cancel_button; + gtk_signal_connect(GTK_OBJECT(Cancel), "clicked", + GTK_SIGNAL_FUNC(OnFile_Cancel), NULL); + gtk_widget_show(Cancel); + + gtk_widget_show(FileSel); + gdk_window_raise(FileSel->window); + + gtk_main(); + + SaveConf(); +} + +void OnMsg_Ok() +{ + gtk_widget_destroy(MsgDlg); + gtk_main_quit(); +} + +void SysMessageLoc(char *fmt, ...) +{ + GtkWidget *Ok, *Txt; + GtkWidget *Box, *Box1; + va_list list; + int w; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + if (msg[strlen(msg)-1] == '\n') msg[strlen(msg)-1] = 0; + + w = strlen(msg) * 6 + 20; + + MsgDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); + gtk_widget_set_usize(MsgDlg, w, 70); + gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "cdriso Msg"); + gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 0); + + Box = gtk_vbox_new(0, 0); + gtk_container_add(GTK_CONTAINER(MsgDlg), Box); + gtk_widget_show(Box); + + Txt = gtk_label_new(msg); + + gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); + gtk_widget_show(Txt); + + Box1 = gtk_hbutton_box_new(); + gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); + gtk_widget_show(Box1); + + Ok = gtk_button_new_with_label("Ok"); + gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); + gtk_container_add(GTK_CONTAINER(Box1), Ok); + GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); + gtk_widget_show(Ok); + + gtk_widget_show(MsgDlg); + + gtk_main(); +} + +void OnOk (GtkButton *button, gpointer user_data) +{ + char *tmp; + + stop = true; + tmp = gtk_entry_get_text(GTK_ENTRY(Edit)); + strcpy(IsoFile, tmp); + tmp = gtk_entry_get_text(GTK_ENTRY(CdEdit)); + strcpy(CdDev, tmp); + + if is_checked(ConfDlg, "checkBlockDump") + BlockDump = 1; + else + BlockDump = 0; + + SaveConf(); + gtk_widget_destroy(ConfDlg); + gtk_main_quit(); +} + +void OnCancel(GtkButton *button, gpointer user_data) +{ + stop = true; + gtk_widget_destroy(ConfDlg); + gtk_main_quit(); +} + +void OnFileSel_Ok() +{ + gchar *File; + + File = gtk_file_selection_get_filename(GTK_FILE_SELECTION(FileSel)); + gtk_entry_set_text(GTK_ENTRY(Edit), File); + gtk_widget_destroy(FileSel); +} + +void OnFileSel_Cancel() +{ + gtk_widget_destroy(FileSel); +} + +void OnFileSel(GtkButton *button, gpointer user_data) +{ + GtkWidget *Ok, *Cancel; + + FileSel = gtk_file_selection_new("Select Psx Iso File"); + gtk_file_selection_set_filename(GTK_FILE_SELECTION(FileSel), IsoFile); + + Ok = GTK_FILE_SELECTION(FileSel)->ok_button; + gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnFileSel_Ok), NULL); + gtk_widget_show(Ok); + + Cancel = GTK_FILE_SELECTION(FileSel)->cancel_button; + gtk_signal_connect(GTK_OBJECT(Cancel), "clicked", GTK_SIGNAL_FUNC(OnFileSel_Cancel), NULL); + gtk_widget_show(Cancel); + + gtk_widget_show(FileSel); + gdk_window_raise(FileSel->window); +} + + +EXPORT_C_(void) CDVDconfigure() +{ + int i; + + LoadConf(); + + ConfDlg = create_Config(); + + Edit = lookup_widget(ConfDlg, "GtkEntry_Iso"); + gtk_entry_set_text(GTK_ENTRY(Edit), IsoFile); + CdEdit = lookup_widget(ConfDlg, "GtkEntry_CdDev"); + gtk_entry_set_text(GTK_ENTRY(CdEdit), CdDev); + + Progress = lookup_widget(ConfDlg, "GtkProgressBar_Progress"); + + BtnCompress = lookup_widget(ConfDlg, "GtkButton_Compress"); + BtnDecompress = lookup_widget(ConfDlg, "GtkButton_Decompress"); + BtnCreate = lookup_widget(ConfDlg, "GtkButton_Create"); + BtnCreateZ = lookup_widget(ConfDlg, "GtkButton_CreateZ"); + + methodlist = NULL; + for (i = 0; i < 2; i++) + methodlist = g_list_append(methodlist, methods[i]); + + Method = lookup_widget(ConfDlg, "GtkCombo_Method"); + gtk_combo_set_popdown_strings(GTK_COMBO(Method), methodlist); + if (strstr(IsoFile, ".Z") != NULL) + gtk_entry_set_text(GTK_ENTRY(GTK_COMBO(Method)->entry), methods[0]); + else + gtk_entry_set_text(GTK_ENTRY(GTK_COMBO(Method)->entry), methods[1]); + + set_checked(ConfDlg, "checkBlockDump", (BlockDump == 1)); + + gtk_widget_show_all(ConfDlg); + gtk_main(); + + return 0; +} + +void OnAboutOk(GtkMenuItem * menuitem, gpointer userdata) +{ + gtk_widget_hide(AboutDlg); + gtk_main_quit(); +} + +EXPORT_C_(void) CDVDabout() +{ + GtkWidget *Label; + GtkWidget *Ok; + GtkWidget *Box, *BBox; + char AboutText[255]; + + sprintf(AboutText, "%s %d.%d\n", LibName, revision, build); + + AboutDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); + gtk_widget_set_usize(AboutDlg, 260, 80); + gtk_window_set_title(GTK_WINDOW(AboutDlg), "CDVD About Dialog"); + gtk_window_set_position(GTK_WINDOW(AboutDlg), GTK_WIN_POS_CENTER); + gtk_container_set_border_width(GTK_CONTAINER(AboutDlg), 10); + + Box = gtk_vbox_new(0, 0); + gtk_container_add(GTK_CONTAINER(AboutDlg), Box); + gtk_widget_show(Box); + + Label = gtk_label_new(AboutText); + gtk_box_pack_start(GTK_BOX(Box), Label, FALSE, FALSE, 0); + gtk_widget_show(Label); + + BBox = gtk_hbutton_box_new(); + gtk_box_pack_start(GTK_BOX(Box), BBox, FALSE, FALSE, 0); + gtk_widget_show(BBox); + + Ok = gtk_button_new_with_label("Ok"); + gtk_signal_connect(GTK_OBJECT(Ok), "clicked", + GTK_SIGNAL_FUNC(OnAboutOk), NULL); + gtk_container_add(GTK_CONTAINER(BBox), Ok); + GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); + gtk_widget_show(Ok); + + gtk_widget_show(AboutDlg); + gtk_main(); +} diff --git a/plugins/CDVDiso/src/Linux/Config.h b/plugins/CDVDiso/src/Linux/Config.h index ee6fad76c7..5009dd0f3d 100644 --- a/plugins/CDVDiso/src/Linux/Config.h +++ b/plugins/CDVDiso/src/Linux/Config.h @@ -16,5 +16,66 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + + #ifdef __cplusplus +extern "C" +{ +#endif + +#include "support.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus +} +#endif + +#include "CDVDiso.h" + void SaveConf(); void LoadConf(); +extern void SysMessageLoc(char *fmt, ...); + +extern char *LibName; + +extern const unsigned char revision; +extern const unsigned char build; + +extern GtkWidget *AboutDlg, *ConfDlg, *MsgDlg, *FileSel; +extern GtkWidget *Edit, *CdEdit; +extern bool stop; + +extern GtkWidget *Method,*Progress; +extern GtkWidget *BtnCompress, *BtnDecompress; +extern GtkWidget *BtnCreate, *BtnCreateZ; + +extern GList *methodlist; + +// Make it easier to check and set checkmarks in the gui +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) + + +extern void OnFileSel(GtkButton *button, gpointer user_data); +extern void OnStop(GtkButton *button, gpointer user_data); +extern void OnCompress(GtkButton *button, gpointer user_data); +extern void OnDecompress(GtkButton *button, gpointer user_data); +extern void OnCreate(GtkButton *button, gpointer user_data); +extern void OnCreateZ(GtkButton *button, gpointer user_data); +extern void OnOk(GtkButton *button, gpointer user_data); +extern void OnCancel(GtkButton *button, gpointer user_data); + diff --git a/plugins/CDVDiso/src/Linux/Linux.c b/plugins/CDVDiso/src/Linux/Linux.c deleted file mode 100644 index dc7072fe55..0000000000 --- a/plugins/CDVDiso/src/Linux/Linux.c +++ /dev/null @@ -1,108 +0,0 @@ -/* CDVDiso - * Copyright (C) 2002-2004 CDVDiso Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include -#include -#include -#include -#include -#include - -#include "Config.h" -#include "CDVDiso.h" - -void ExecCfg(char *arg) -{ - char cfg[256]; - struct stat buf; - - strcpy(cfg, "./cfgCDVDiso"); - if (stat(cfg, &buf) != -1) - { - sprintf(cfg, "%s %s", cfg, arg); - system(cfg); - return; - } - - strcpy(cfg, "./plugins/cfgCDVDiso"); - if (stat(cfg, &buf) != -1) - { - sprintf(cfg, "%s %s", cfg, arg); - system(cfg); - return; - } - - strcpy(cfg, "./cfg/cfgCDVDiso"); - if (stat(cfg, &buf) != -1) - { - sprintf(cfg, "%s %s", cfg, arg); - system(cfg); - return; - } - - sprintf(cfg, "%s/cfgCDVDiso", getenv("HOME")); - if (stat(cfg, &buf) != -1) - { - sprintf(cfg, "%s %s", cfg, arg); - system(cfg); - return; - } - - printf("cfgCDVDiso file not found!\n"); -} - -void CDVDconfigure() -{ - char *file; - getcwd(file, ArraySize(file)); - chdir("plugins"); - ExecCfg("configure"); - chdir(file); -} - -void CDVDabout() -{ - char *file; - getcwd(file, ArraySize(file)); - chdir("plugins"); - ExecCfg("about"); - chdir(file); -} - -void CfgOpenFile() -{ - char *file; - getcwd(file, ArraySize(file)); - chdir("plugins"); - ExecCfg("open"); - chdir(file); -} - -void SysMessage(char *fmt, ...) -{ - va_list list; - char tmp[256]; - char cmd[256]; - - va_start(list, fmt); - vsprintf(tmp, fmt, list); - va_end(list); - - sprintf(cmd, "message \"%s\"", tmp); - ExecCfg(cmd); -} diff --git a/plugins/CDVDiso/src/Linux/conf.c b/plugins/CDVDiso/src/Linux/Linux.cpp old mode 100644 new mode 100755 similarity index 65% rename from plugins/CDVDiso/src/Linux/conf.c rename to plugins/CDVDiso/src/Linux/Linux.cpp index 298c99962b..814e974714 --- a/plugins/CDVDiso/src/Linux/conf.c +++ b/plugins/CDVDiso/src/Linux/Linux.cpp @@ -16,211 +16,31 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include "interface.h" -#include "support.h" -#include "CDVDiso.h" #include "Config.h" - -// Make it easier to check and set checkmarks in the gui -#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) -#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) - unsigned char Zbuf[CD_FRAMESIZE_RAW * 10 * 2]; -extern char *LibName; - -extern const unsigned char revision; -extern const unsigned char build; - -GtkWidget *FileSel; - -void OnFile_Ok() -{ - gchar *File; - - gtk_widget_hide(FileSel); - File = gtk_file_selection_get_filename(GTK_FILE_SELECTION(FileSel)); - strcpy(IsoFile, File); - gtk_main_quit(); -} - -void OnFile_Cancel() -{ - gtk_widget_hide(FileSel); - gtk_main_quit(); -} - -void _CDRopen() -{ - GtkWidget *Ok, *Cancel; - - FileSel = gtk_file_selection_new("Select Iso File"); - - Ok = GTK_FILE_SELECTION(FileSel)->ok_button; - gtk_signal_connect(GTK_OBJECT(Ok), "clicked", - GTK_SIGNAL_FUNC(OnFile_Ok), NULL); - gtk_widget_show(Ok); - - Cancel = GTK_FILE_SELECTION(FileSel)->cancel_button; - gtk_signal_connect(GTK_OBJECT(Cancel), "clicked", - GTK_SIGNAL_FUNC(OnFile_Cancel), NULL); - gtk_widget_show(Cancel); - - gtk_widget_show(FileSel); - gdk_window_raise(FileSel->window); - - gtk_main(); - - SaveConf(); -} - -GtkWidget *MsgDlg; - -void OnMsg_Ok() -{ - gtk_widget_destroy(MsgDlg); - gtk_main_quit(); -} - -static void SysMessageLoc(char *fmt, ...) -{ - GtkWidget *Ok, *Txt; - GtkWidget *Box, *Box1; - va_list list; - int w; - char msg[512]; - - va_start(list, fmt); - vsprintf(msg, fmt, list); - va_end(list); - - if (msg[strlen(msg)-1] == '\n') msg[strlen(msg)-1] = 0; - - w = strlen(msg) * 6 + 20; - - MsgDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); - gtk_widget_set_usize(MsgDlg, w, 70); - gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); - gtk_window_set_title(GTK_WINDOW(MsgDlg), "cdriso Msg"); - gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 0); - - Box = gtk_vbox_new(0, 0); - gtk_container_add(GTK_CONTAINER(MsgDlg), Box); - gtk_widget_show(Box); - - Txt = gtk_label_new(msg); - - gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); - gtk_widget_show(Txt); - - Box1 = gtk_hbutton_box_new(); - gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); - gtk_widget_show(Box1); - - Ok = gtk_button_new_with_label("Ok"); - gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); - gtk_container_add(GTK_CONTAINER(Box1), Ok); - GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); - gtk_widget_show(Ok); - - gtk_widget_show(MsgDlg); - - gtk_main(); -} - -GtkWidget *ConfDlg; -GtkWidget *Edit, *CdEdit; -GtkWidget *FileSel; -GtkWidget *Progress; -GtkWidget *BtnCompress; -GtkWidget *BtnDecompress; -GtkWidget *BtnCreate; -GtkWidget *BtnCreateZ; -GtkWidget *Method; +GtkWidget *Method,*Progress; +GtkWidget *BtnCompress, *BtnDecompress; +GtkWidget *BtnCreate, *BtnCreateZ; GList *methodlist; extern char *methods[]; +unsigned char param[4]; +int cddev = -1; +bool stop; -int stop; +#define CD_LEADOUT (0xaa) -void OnOk(GtkMenuItem * menuitem, gpointer userdata) +union { - char *tmp; + struct cdrom_msf msf; + unsigned char buf[CD_FRAMESIZE_RAW]; +} cr; - stop = 1; - tmp = gtk_entry_get_text(GTK_ENTRY(Edit)); - strcpy(IsoFile, tmp); - tmp = gtk_entry_get_text(GTK_ENTRY(CdEdit)); - strcpy(CdDev, tmp); - - if is_checked(ConfDlg, "checkBlockDump") - BlockDump = 1; - else - BlockDump = 0; - - SaveConf(); - gtk_widget_destroy(ConfDlg); - gtk_main_quit(); -} - -void OnCancel(GtkMenuItem * menuitem, gpointer userdata) +void OnStop(GtkButton *button, gpointer user_data) { - stop = 1; - gtk_widget_destroy(ConfDlg); - gtk_main_quit(); -} - -void OnFileSel_Ok() -{ - gchar *File; - - File = gtk_file_selection_get_filename(GTK_FILE_SELECTION(FileSel)); - gtk_entry_set_text(GTK_ENTRY(Edit), File); - gtk_widget_destroy(FileSel); -} - -void OnFileSel_Cancel() -{ - gtk_widget_destroy(FileSel); -} - -void OnFileSel() -{ - GtkWidget *Ok, *Cancel; - - FileSel = gtk_file_selection_new("Select Psx Iso File"); - gtk_file_selection_set_filename(GTK_FILE_SELECTION(FileSel), IsoFile); - - Ok = GTK_FILE_SELECTION(FileSel)->ok_button; - gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnFileSel_Ok), NULL); - gtk_widget_show(Ok); - - Cancel = GTK_FILE_SELECTION(FileSel)->cancel_button; - gtk_signal_connect(GTK_OBJECT(Cancel), "clicked", GTK_SIGNAL_FUNC(OnFileSel_Cancel), NULL); - gtk_widget_show(Cancel); - - gtk_widget_show(FileSel); - gdk_window_raise(FileSel->window); -} - -void OnStop() -{ - stop = 1; + stop = true; } void UpdZmode() @@ -234,7 +54,7 @@ void UpdZmode() char buffer[2352 * 10]; -void OnCompress() +void OnCompress(GtkButton *button, gpointer user_data) { struct stat buf; u32 lsn; @@ -271,7 +91,7 @@ void OnCompress() gtk_widget_set_sensitive(BtnDecompress, FALSE); gtk_widget_set_sensitive(BtnCreate, FALSE); gtk_widget_set_sensitive(BtnCreateZ, FALSE); - stop = 0; + stop = false; for (lsn = 0; lsn < src->blocks; lsn++) { @@ -310,7 +130,7 @@ void OnCompress() } } -void OnDecompress() +void OnDecompress(GtkButton *button, gpointer user_data) { #if 0 struct stat buf; @@ -366,7 +186,7 @@ void OnDecompress() gtk_widget_set_sensitive(BtnDecompress, FALSE); gtk_widget_set_sensitive(BtnCreate, FALSE); gtk_widget_set_sensitive(BtnCreateZ, FALSE); - stop = 0; + stop = false; if (Zmode == 1) { @@ -433,16 +253,6 @@ void OnDecompress() #endif } -#define CD_LEADOUT (0xaa) -unsigned char param[4]; -int cddev = -1; - -union -{ - struct cdrom_msf msf; - unsigned char buf[CD_FRAMESIZE_RAW]; -} cr; - void incSector() { param[2]++; @@ -535,7 +345,7 @@ char *CDR_readTrack(unsigned char *time) } -void OnCreate() +void OnCreate(GtkButton *button, gpointer user_data) { FILE *f; struct stat buf; @@ -600,7 +410,7 @@ void OnCreate() time(&Ttime); - stop = 0; + stop = false; s = MSF2SECT(end[0], end[1], end[2]); gtk_widget_set_sensitive(BtnCompress, FALSE); gtk_widget_set_sensitive(BtnDecompress, FALSE); @@ -672,7 +482,7 @@ void OnCreate() if (!stop) SysMessageLoc("Iso Image Created OK"); } -void OnCreateZ() +void OnCreateZ(GtkButton *button, gpointer user_data) { FILE *f; FILE *t; @@ -758,7 +568,7 @@ void OnCreateZ() time(&Ttime); - stop = 0; + stop = false; s = MSF2SECT(end[0], end[1], end[2]) / blocks; gtk_widget_set_sensitive(BtnCompress, FALSE); gtk_widget_set_sensitive(BtnDecompress, FALSE); @@ -857,110 +667,3 @@ void OnCreateZ() if (!stop) SysMessageLoc("Compressed Iso Image Created OK"); } - -long CDRconfigure(void) -{ - int i; - - LoadConf(); - - ConfDlg = create_Config(); - - Edit = lookup_widget(ConfDlg, "GtkEntry_Iso"); - gtk_entry_set_text(GTK_ENTRY(Edit), IsoFile); - CdEdit = lookup_widget(ConfDlg, "GtkEntry_CdDev"); - gtk_entry_set_text(GTK_ENTRY(CdEdit), CdDev); - - Progress = lookup_widget(ConfDlg, "GtkProgressBar_Progress"); - - BtnCompress = lookup_widget(ConfDlg, "GtkButton_Compress"); - BtnDecompress = lookup_widget(ConfDlg, "GtkButton_Decompress"); - BtnCreate = lookup_widget(ConfDlg, "GtkButton_Create"); - BtnCreateZ = lookup_widget(ConfDlg, "GtkButton_CreateZ"); - - methodlist = NULL; - for (i = 0; i < 2; i++) - methodlist = g_list_append(methodlist, methods[i]); - - Method = lookup_widget(ConfDlg, "GtkCombo_Method"); - gtk_combo_set_popdown_strings(GTK_COMBO(Method), methodlist); - if (strstr(IsoFile, ".Z") != NULL) - gtk_entry_set_text(GTK_ENTRY(GTK_COMBO(Method)->entry), methods[0]); - else - gtk_entry_set_text(GTK_ENTRY(GTK_COMBO(Method)->entry), methods[1]); - - set_checked(ConfDlg, "checkBlockDump", (BlockDump == 1)); - - gtk_widget_show_all(ConfDlg); - gtk_main(); - - return 0; -} - -GtkWidget *AboutDlg; - -void OnAboutOk(GtkMenuItem * menuitem, gpointer userdata) -{ - gtk_widget_hide(AboutDlg); - gtk_main_quit(); -} - -void CDRabout(void) -{ - GtkWidget *Label; - GtkWidget *Ok; - GtkWidget *Box, *BBox; - char AboutText[255]; - - sprintf(AboutText, "%s %d.%d\n", LibName, revision, build); - - AboutDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); - gtk_widget_set_usize(AboutDlg, 260, 80); - gtk_window_set_title(GTK_WINDOW(AboutDlg), "CDVD About Dialog"); - gtk_window_set_position(GTK_WINDOW(AboutDlg), GTK_WIN_POS_CENTER); - gtk_container_set_border_width(GTK_CONTAINER(AboutDlg), 10); - - Box = gtk_vbox_new(0, 0); - gtk_container_add(GTK_CONTAINER(AboutDlg), Box); - gtk_widget_show(Box); - - Label = gtk_label_new(AboutText); - gtk_box_pack_start(GTK_BOX(Box), Label, FALSE, FALSE, 0); - gtk_widget_show(Label); - - BBox = gtk_hbutton_box_new(); - gtk_box_pack_start(GTK_BOX(Box), BBox, FALSE, FALSE, 0); - gtk_widget_show(BBox); - - Ok = gtk_button_new_with_label("Ok"); - gtk_signal_connect(GTK_OBJECT(Ok), "clicked", - GTK_SIGNAL_FUNC(OnAboutOk), NULL); - gtk_container_add(GTK_CONTAINER(BBox), Ok); - GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); - gtk_widget_show(Ok); - - gtk_widget_show(AboutDlg); - gtk_main(); -} - -int main(int argc, char *argv[]) -{ - if (argc < 2) return 0; - - gtk_init(NULL, NULL); - - if (!strcmp(argv[1], "open")) - _CDRopen(); - else if (!strcmp(argv[1], "configure")) - CDRconfigure(); - else if (!strcmp(argv[1], "message")) - { - if (argc > 2) SysMessageLoc(argv[2]); - } - else - CDRabout(); - - return 0; -} - - diff --git a/plugins/CDVDiso/src/Linux/interface.c b/plugins/CDVDiso/src/Linux/interface.c index bb7f7094a7..559fbed98d 100644 --- a/plugins/CDVDiso/src/Linux/interface.c +++ b/plugins/CDVDiso/src/Linux/interface.c @@ -57,43 +57,53 @@ create_Config (void) GtkWidget *button2; Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); gtk_container_set_border_width (GTK_CONTAINER (Config), 5); gtk_window_set_title (GTK_WINDOW (Config), _("CDVD Config Dialog")); vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); gtk_widget_show (vbox1); gtk_container_add (GTK_CONTAINER (Config), vbox1); gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); hbox1 = gtk_hbox_new (FALSE, 10); + gtk_widget_set_name (hbox1, "hbox1"); gtk_widget_show (hbox1); gtk_box_pack_start (GTK_BOX (vbox1), hbox1, TRUE, TRUE, 0); GtkEntry_Iso = gtk_entry_new (); + gtk_widget_set_name (GtkEntry_Iso, "GtkEntry_Iso"); gtk_widget_show (GtkEntry_Iso); gtk_box_pack_start (GTK_BOX (hbox1), GtkEntry_Iso, TRUE, TRUE, 0); button5 = gtk_button_new_with_mnemonic (_("Select Iso")); + gtk_widget_set_name (button5, "button5"); gtk_widget_show (button5); gtk_box_pack_start (GTK_BOX (hbox1), button5, FALSE, FALSE, 0); hbox2 = gtk_hbox_new (FALSE, 10); + gtk_widget_set_name (hbox2, "hbox2"); gtk_widget_show (hbox2); gtk_box_pack_start (GTK_BOX (vbox1), hbox2, FALSE, FALSE, 0); GtkProgressBar_Progress = gtk_progress_bar_new (); + gtk_widget_set_name (GtkProgressBar_Progress, "GtkProgressBar_Progress"); gtk_widget_show (GtkProgressBar_Progress); gtk_box_pack_start (GTK_BOX (hbox2), GtkProgressBar_Progress, TRUE, FALSE, 0); button6 = gtk_button_new_with_mnemonic (_("Stop")); + gtk_widget_set_name (button6, "button6"); gtk_widget_show (button6); gtk_box_pack_end (GTK_BOX (hbox2), button6, FALSE, FALSE, 0); hbox4 = gtk_hbox_new (FALSE, 5); + gtk_widget_set_name (hbox4, "hbox4"); gtk_widget_show (hbox4); gtk_box_pack_start (GTK_BOX (vbox1), hbox4, TRUE, TRUE, 0); label2 = gtk_label_new (_("Compression Method:")); + gtk_widget_set_name (label2, "label2"); gtk_widget_show (label2); gtk_box_pack_start (GTK_BOX (hbox4), label2, FALSE, FALSE, 0); gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); @@ -101,6 +111,7 @@ create_Config (void) GtkCombo_Method = gtk_combo_new (); g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Method)->popwin), "GladeParentKey", GtkCombo_Method); + gtk_widget_set_name (GtkCombo_Method, "GtkCombo_Method"); gtk_widget_show (GtkCombo_Method); gtk_box_pack_start (GTK_BOX (hbox4), GtkCombo_Method, TRUE, FALSE, 0); GtkCombo_Method_items = g_list_append (GtkCombo_Method_items, (gpointer) ""); @@ -108,63 +119,77 @@ create_Config (void) g_list_free (GtkCombo_Method_items); combo_entry1 = GTK_COMBO (GtkCombo_Method)->entry; + gtk_widget_set_name (combo_entry1, "combo_entry1"); gtk_widget_show (combo_entry1); hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); gtk_widget_show (hbuttonbox2); gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox2, TRUE, TRUE, 0); GtkButton_Compress = gtk_button_new_with_mnemonic (_("Compress Iso")); + gtk_widget_set_name (GtkButton_Compress, "GtkButton_Compress"); gtk_widget_show (GtkButton_Compress); gtk_container_add (GTK_CONTAINER (hbuttonbox2), GtkButton_Compress); GTK_WIDGET_SET_FLAGS (GtkButton_Compress, GTK_CAN_DEFAULT); GtkButton_Decompress = gtk_button_new_with_mnemonic (_("Decompress Iso")); + gtk_widget_set_name (GtkButton_Decompress, "GtkButton_Decompress"); gtk_widget_show (GtkButton_Decompress); gtk_container_add (GTK_CONTAINER (hbuttonbox2), GtkButton_Decompress); GTK_WIDGET_SET_FLAGS (GtkButton_Decompress, GTK_CAN_DEFAULT); hbox3 = gtk_hbox_new (FALSE, 5); + gtk_widget_set_name (hbox3, "hbox3"); gtk_widget_show (hbox3); gtk_box_pack_start (GTK_BOX (vbox1), hbox3, FALSE, FALSE, 0); label1 = gtk_label_new (_("Cdrom Device: ")); + gtk_widget_set_name (label1, "label1"); gtk_widget_show (label1); gtk_box_pack_start (GTK_BOX (hbox3), label1, FALSE, FALSE, 0); gtk_label_set_justify (GTK_LABEL (label1), GTK_JUSTIFY_CENTER); GtkEntry_CdDev = gtk_entry_new (); + gtk_widget_set_name (GtkEntry_CdDev, "GtkEntry_CdDev"); gtk_widget_show (GtkEntry_CdDev); gtk_box_pack_start (GTK_BOX (hbox3), GtkEntry_CdDev, TRUE, TRUE, 0); hbuttonbox3 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox3, "hbuttonbox3"); gtk_widget_show (hbuttonbox3); gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox3, TRUE, TRUE, 0); GtkButton_Create = gtk_button_new_with_mnemonic (_("Create Iso")); + gtk_widget_set_name (GtkButton_Create, "GtkButton_Create"); gtk_widget_show (GtkButton_Create); gtk_container_add (GTK_CONTAINER (hbuttonbox3), GtkButton_Create); GTK_WIDGET_SET_FLAGS (GtkButton_Create, GTK_CAN_DEFAULT); GtkButton_CreateZ = gtk_button_new_with_mnemonic (_("Create Compressed Iso")); + gtk_widget_set_name (GtkButton_CreateZ, "GtkButton_CreateZ"); gtk_widget_show (GtkButton_CreateZ); gtk_container_add (GTK_CONTAINER (hbuttonbox3), GtkButton_CreateZ); GTK_WIDGET_SET_FLAGS (GtkButton_CreateZ, GTK_CAN_DEFAULT); checkBlockDump = gtk_check_button_new_with_mnemonic (_("Create a dump of the running iso")); + gtk_widget_set_name (checkBlockDump, "checkBlockDump"); gtk_widget_show (checkBlockDump); gtk_box_pack_start (GTK_BOX (vbox1), checkBlockDump, FALSE, FALSE, 0); hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); gtk_widget_show (hbuttonbox1); gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); button1 = gtk_button_new_from_stock ("gtk-ok"); + gtk_widget_set_name (button1, "button1"); gtk_widget_show (button1); gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); button2 = gtk_button_new_from_stock ("gtk-cancel"); + gtk_widget_set_name (button2, "button2"); gtk_widget_show (button2); gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); diff --git a/plugins/CDVDiso/src/Makefile.am b/plugins/CDVDiso/src/Makefile.am new file mode 100755 index 0000000000..01b772801a --- /dev/null +++ b/plugins/CDVDiso/src/Makefile.am @@ -0,0 +1,31 @@ +# Create a shared library libCDVDiso +AUTOMAKE_OPTIONS = foreign +noinst_LIBRARIES = libCDVDiso.a +INCLUDES = $(shell pkg-config --cflags gtk+-2.0) -I@srcdir@/../../../common/include -I@srcdir@/../../../3rdparty -I@srcdir@/Linux -I../../../../3rdparty/zlib + +# Create a shared object by faking an exe (thanks to ODE makefiles) +traplibdir=$(prefix) + +if DEBUGBUILD +preext=d +endif + +EXEEXT=$(preext)@so_ext@ + +traplib_PROGRAMS=libCDVDiso +libCDVDiso_SOURCES= +libCDVDiso_DEPENDENCIES = libCDVDiso.a +libCDVDiso_LDFLAGS= @SHARED_LDFLAGS@ +libCDVDiso_LDFLAGS+=-Wl,-soname,@libCDVDiso_SONAME@ +libCDVDiso_LDADD=$(libCDVDiso_a_OBJECTS) + +libCDVDiso_a_SOURCES = CDVDisop.cpp CDVDiso.h libiso.cpp libiso.h mkiso/mkiso.cpp + +libCDVDiso_a_SOURCES += \ +Linux/Config.cpp Linux/Config.h Linux/Linux.cpp \ +Linux/interface.h Linux/support.c \ +Linux/interface.c Linux/support.h \ +Linux/callbacks.h +#Linux/callbacks.c + +#SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj b/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj index 652155f757..e99a890458 100644 --- a/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj +++ b/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj @@ -265,7 +265,7 @@ UniqueIdentifier="{4FC737F1-C7A5-4376-A066-2A32D752A2FF}" > svnrev.h +// +// This file acts as a template for the automatic SVN revision/version tag. +// It is used by the utility SubWCrev.exe to create an "svnrev.h" file for +// whichever project is being compiled (as indicated by command line options +// passed to SubWCRev.exe during the project's pre-build step). +// +// The SubWCRev.exe utility is part of TortoiseSVN and requires several DLLs +// installed by TortoiseSVN, so it will only be available if you have TortoiseSVN +// installed on your system. If you do not have it installed, a generic template +// is used instead (see svnrev_generic.h). Having TortoiseSVN is handy but not +// necessary. If you do not have it installed, everything will still compile +// fine except without the SVN revision tagged to the application/dll version. +// +// TortoiseSVN can be downloaded from http://tortoisesvn.tigris.org + +#define SVN_REV 798 +#define SVN_MODS 1 \ No newline at end of file diff --git a/plugins/CDVDiso/src/configure.ac b/plugins/CDVDiso/src/configure.ac new file mode 100755 index 0000000000..b181d19477 --- /dev/null +++ b/plugins/CDVDiso/src/configure.ac @@ -0,0 +1,83 @@ +AC_INIT(CDVDiso, 0.9,arcum42@gmail.com) + +AM_INIT_AUTOMAKE(CDVDiso,0.9) + +AC_PROG_CC([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CXX([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CPP([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) + +AC_PROG_INSTALL +AC_PROG_RANLIB + +dnl necessary for compiling assembly +AM_PROG_AS + +AC_SUBST(CDVDiso_CURRENT, 0) +AC_SUBST(CDVDiso_REVISION, 9) +AC_SUBST(CDVDiso_AGE, 0) +AC_SUBST(CDVDiso_RELEASE,[$CDVDiso_CURRENT].[$CDVDiso_REVISION].[$CDVDiso_AGE]) +AC_SUBST(CDVDiso_SONAME,libCDVDiso.so.[$CDVDiso_CURRENT].[$CDVDiso_REVISION].[$CDVDiso_AGE]) + +CFLAGS= +CPPFLAGS= +CXXFLAGS= + +dnl Check for debug build +AC_MSG_CHECKING(debug build) +AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), +debug=$enableval,debug=no) +if test "x$debug" == xyes +then + AC_DEFINE(_DEBUG,1,[_DEBUG]) + CFLAGS+=" -g -fPIC -Wall -Wno-unused-value " + CPPFLAGS+=" -g -fPIC -Wall -Wno-unused-value " + CXXFLAGS+=" -g -fPIC -Wall -Wno-unused-value -fpermissive " +else + AC_DEFINE(NDEBUG,1,[NDEBUG]) + CFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value " + CPPFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value " + CXXFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value -fpermissive " +fi +AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) +AC_MSG_RESULT($debug) + +AC_DEFINE(__LINUX__,1,[__LINUX__]) +AC_DEFINE(_FILE_OFFSET_BITS,64,[_FILE_OFFSET_BITS]) + +dnl Check for dev build +AC_MSG_CHECKING(for development build...) +AC_ARG_ENABLE(devbuild, AC_HELP_STRING([--enable-devbuild], [Special Build for developers that simplifies testing and adds extra checks]), + devbuild=$enableval,devbuild=no) +if test "x$devbuild" == xyes +then + AC_DEFINE(CDVDiso_DEVBUILD,1,[CDVDiso_DEVBUILD]) +fi +AC_MSG_RESULT($devbuild) +AM_CONDITIONAL(RELEASE_TO_PUBLIC, test x$devbuild = xno) + +AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) + +dnl gtk +AC_MSG_CHECKING(gtk2+) +AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) +LIBS+=$(pkg-config --libs gtk+-2.0) + +dnl bindir = pcsx2exe + +dnl assuming linux environment +so_ext=".so.$CDVDiso_RELEASE" +SHARED_LDFLAGS="-shared" +AC_SUBST(so_ext) +AC_SUBST(SHARED_LDFLAGS) + +AC_CHECK_LIB(stdc++,main,[LIBS="$LIBS -lstdc++"]) +AC_CHECK_LIB(z,main,[LIBS="$LIBS -lz"]) +AC_CHECK_LIB(bz2,main,[LIBS="$LIBS -lbz2"]) + +AC_OUTPUT([ + Makefile + ]) + +echo "Configuration:" +echo " Debug build? $debug" +echo " Dev build? $devbuild" \ No newline at end of file diff --git a/plugins/CDVDiso/src/install-sh b/plugins/CDVDiso/src/install-sh new file mode 100755 index 0000000000..a5897de6ea --- /dev/null +++ b/plugins/CDVDiso/src/install-sh @@ -0,0 +1,519 @@ +#!/bin/sh +# install - install a program, script, or datafile + +scriptversion=2006-12-25.00 + +# This originates from X11R5 (mit/util/scripts/install.sh), which was +# later released in X11R6 (xc/config/util/install.sh) with the +# following copyright and license. +# +# Copyright (C) 1994 X Consortium +# +# Permission is hereby granted, free of charge, to any person obtaining a copy +# of this software and associated documentation files (the "Software"), to +# deal in the Software without restriction, including without limitation the +# rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +# sell copies of the Software, and to permit persons to whom the Software is +# furnished to do so, subject to the following conditions: +# +# The above copyright notice and this permission notice shall be included in +# all copies or substantial portions of the Software. +# +# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +# X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN +# AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNEC- +# TION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +# +# Except as contained in this notice, the name of the X Consortium shall not +# be used in advertising or otherwise to promote the sale, use or other deal- +# ings in this Software without prior written authorization from the X Consor- +# tium. +# +# +# FSF changes to this file are in the public domain. +# +# Calling this script install-sh is preferred over install.sh, to prevent +# `make' implicit rules from creating a file called install from it +# when there is no Makefile. +# +# This script is compatible with the BSD install script, but was written +# from scratch. + +nl=' +' +IFS=" "" $nl" + +# set DOITPROG to echo to test this script + +# Don't use :- since 4.3BSD and earlier shells don't like it. +doit=${DOITPROG-} +if test -z "$doit"; then + doit_exec=exec +else + doit_exec=$doit +fi + +# Put in absolute file names if you don't have them in your path; +# or use environment vars. + +chgrpprog=${CHGRPPROG-chgrp} +chmodprog=${CHMODPROG-chmod} +chownprog=${CHOWNPROG-chown} +cmpprog=${CMPPROG-cmp} +cpprog=${CPPROG-cp} +mkdirprog=${MKDIRPROG-mkdir} +mvprog=${MVPROG-mv} +rmprog=${RMPROG-rm} +stripprog=${STRIPPROG-strip} + +posix_glob='?' +initialize_posix_glob=' + test "$posix_glob" != "?" || { + if (set -f) 2>/dev/null; then + posix_glob= + else + posix_glob=: + fi + } +' + +posix_mkdir= + +# Desired mode of installed file. +mode=0755 + +chgrpcmd= +chmodcmd=$chmodprog +chowncmd= +mvcmd=$mvprog +rmcmd="$rmprog -f" +stripcmd= + +src= +dst= +dir_arg= +dst_arg= + +copy_on_change=false +no_target_directory= + +usage="\ +Usage: $0 [OPTION]... [-T] SRCFILE DSTFILE + or: $0 [OPTION]... SRCFILES... DIRECTORY + or: $0 [OPTION]... -t DIRECTORY SRCFILES... + or: $0 [OPTION]... -d DIRECTORIES... + +In the 1st form, copy SRCFILE to DSTFILE. +In the 2nd and 3rd, copy all SRCFILES to DIRECTORY. +In the 4th, create DIRECTORIES. + +Options: + --help display this help and exit. + --version display version info and exit. + + -c (ignored) + -C install only if different (preserve the last data modification time) + -d create directories instead of installing files. + -g GROUP $chgrpprog installed files to GROUP. + -m MODE $chmodprog installed files to MODE. + -o USER $chownprog installed files to USER. + -s $stripprog installed files. + -t DIRECTORY install into DIRECTORY. + -T report an error if DSTFILE is a directory. + +Environment variables override the default commands: + CHGRPPROG CHMODPROG CHOWNPROG CMPPROG CPPROG MKDIRPROG MVPROG + RMPROG STRIPPROG +" + +while test $# -ne 0; do + case $1 in + -c) ;; + + -C) copy_on_change=true;; + + -d) dir_arg=true;; + + -g) chgrpcmd="$chgrpprog $2" + shift;; + + --help) echo "$usage"; exit $?;; + + -m) mode=$2 + case $mode in + *' '* | *' '* | *' +'* | *'*'* | *'?'* | *'['*) + echo "$0: invalid mode: $mode" >&2 + exit 1;; + esac + shift;; + + -o) chowncmd="$chownprog $2" + shift;; + + -s) stripcmd=$stripprog;; + + -t) dst_arg=$2 + shift;; + + -T) no_target_directory=true;; + + --version) echo "$0 $scriptversion"; exit $?;; + + --) shift + break;; + + -*) echo "$0: invalid option: $1" >&2 + exit 1;; + + *) break;; + esac + shift +done + +if test $# -ne 0 && test -z "$dir_arg$dst_arg"; then + # When -d is used, all remaining arguments are directories to create. + # When -t is used, the destination is already specified. + # Otherwise, the last argument is the destination. Remove it from $@. + for arg + do + if test -n "$dst_arg"; then + # $@ is not empty: it contains at least $arg. + set fnord "$@" "$dst_arg" + shift # fnord + fi + shift # arg + dst_arg=$arg + done +fi + +if test $# -eq 0; then + if test -z "$dir_arg"; then + echo "$0: no input file specified." >&2 + exit 1 + fi + # It's OK to call `install-sh -d' without argument. + # This can happen when creating conditional directories. + exit 0 +fi + +if test -z "$dir_arg"; then + trap '(exit $?); exit' 1 2 13 15 + + # Set umask so as not to create temps with too-generous modes. + # However, 'strip' requires both read and write access to temps. + case $mode in + # Optimize common cases. + *644) cp_umask=133;; + *755) cp_umask=22;; + + *[0-7]) + if test -z "$stripcmd"; then + u_plus_rw= + else + u_plus_rw='% 200' + fi + cp_umask=`expr '(' 777 - $mode % 1000 ')' $u_plus_rw`;; + *) + if test -z "$stripcmd"; then + u_plus_rw= + else + u_plus_rw=,u+rw + fi + cp_umask=$mode$u_plus_rw;; + esac +fi + +for src +do + # Protect names starting with `-'. + case $src in + -*) src=./$src;; + esac + + if test -n "$dir_arg"; then + dst=$src + dstdir=$dst + test -d "$dstdir" + dstdir_status=$? + else + + # Waiting for this to be detected by the "$cpprog $src $dsttmp" command + # might cause directories to be created, which would be especially bad + # if $src (and thus $dsttmp) contains '*'. + if test ! -f "$src" && test ! -d "$src"; then + echo "$0: $src does not exist." >&2 + exit 1 + fi + + if test -z "$dst_arg"; then + echo "$0: no destination specified." >&2 + exit 1 + fi + + dst=$dst_arg + # Protect names starting with `-'. + case $dst in + -*) dst=./$dst;; + esac + + # If destination is a directory, append the input filename; won't work + # if double slashes aren't ignored. + if test -d "$dst"; then + if test -n "$no_target_directory"; then + echo "$0: $dst_arg: Is a directory" >&2 + exit 1 + fi + dstdir=$dst + dst=$dstdir/`basename "$src"` + dstdir_status=0 + else + # Prefer dirname, but fall back on a substitute if dirname fails. + dstdir=` + (dirname "$dst") 2>/dev/null || + expr X"$dst" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ + X"$dst" : 'X\(//\)[^/]' \| \ + X"$dst" : 'X\(//\)$' \| \ + X"$dst" : 'X\(/\)' \| . 2>/dev/null || + echo X"$dst" | + sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ + s//\1/ + q + } + /^X\(\/\/\)[^/].*/{ + s//\1/ + q + } + /^X\(\/\/\)$/{ + s//\1/ + q + } + /^X\(\/\).*/{ + s//\1/ + q + } + s/.*/./; q' + ` + + test -d "$dstdir" + dstdir_status=$? + fi + fi + + obsolete_mkdir_used=false + + if test $dstdir_status != 0; then + case $posix_mkdir in + '') + # Create intermediate dirs using mode 755 as modified by the umask. + # This is like FreeBSD 'install' as of 1997-10-28. + umask=`umask` + case $stripcmd.$umask in + # Optimize common cases. + *[2367][2367]) mkdir_umask=$umask;; + .*0[02][02] | .[02][02] | .[02]) mkdir_umask=22;; + + *[0-7]) + mkdir_umask=`expr $umask + 22 \ + - $umask % 100 % 40 + $umask % 20 \ + - $umask % 10 % 4 + $umask % 2 + `;; + *) mkdir_umask=$umask,go-w;; + esac + + # With -d, create the new directory with the user-specified mode. + # Otherwise, rely on $mkdir_umask. + if test -n "$dir_arg"; then + mkdir_mode=-m$mode + else + mkdir_mode= + fi + + posix_mkdir=false + case $umask in + *[123567][0-7][0-7]) + # POSIX mkdir -p sets u+wx bits regardless of umask, which + # is incompatible with FreeBSD 'install' when (umask & 300) != 0. + ;; + *) + tmpdir=${TMPDIR-/tmp}/ins$RANDOM-$$ + trap 'ret=$?; rmdir "$tmpdir/d" "$tmpdir" 2>/dev/null; exit $ret' 0 + + if (umask $mkdir_umask && + exec $mkdirprog $mkdir_mode -p -- "$tmpdir/d") >/dev/null 2>&1 + then + if test -z "$dir_arg" || { + # Check for POSIX incompatibilities with -m. + # HP-UX 11.23 and IRIX 6.5 mkdir -m -p sets group- or + # other-writeable bit of parent directory when it shouldn't. + # FreeBSD 6.1 mkdir -m -p sets mode of existing directory. + ls_ld_tmpdir=`ls -ld "$tmpdir"` + case $ls_ld_tmpdir in + d????-?r-*) different_mode=700;; + d????-?--*) different_mode=755;; + *) false;; + esac && + $mkdirprog -m$different_mode -p -- "$tmpdir" && { + ls_ld_tmpdir_1=`ls -ld "$tmpdir"` + test "$ls_ld_tmpdir" = "$ls_ld_tmpdir_1" + } + } + then posix_mkdir=: + fi + rmdir "$tmpdir/d" "$tmpdir" + else + # Remove any dirs left behind by ancient mkdir implementations. + rmdir ./$mkdir_mode ./-p ./-- 2>/dev/null + fi + trap '' 0;; + esac;; + esac + + if + $posix_mkdir && ( + umask $mkdir_umask && + $doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir" + ) + then : + else + + # The umask is ridiculous, or mkdir does not conform to POSIX, + # or it failed possibly due to a race condition. Create the + # directory the slow way, step by step, checking for races as we go. + + case $dstdir in + /*) prefix='/';; + -*) prefix='./';; + *) prefix='';; + esac + + eval "$initialize_posix_glob" + + oIFS=$IFS + IFS=/ + $posix_glob set -f + set fnord $dstdir + shift + $posix_glob set +f + IFS=$oIFS + + prefixes= + + for d + do + test -z "$d" && continue + + prefix=$prefix$d + if test -d "$prefix"; then + prefixes= + else + if $posix_mkdir; then + (umask=$mkdir_umask && + $doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir") && break + # Don't fail if two instances are running concurrently. + test -d "$prefix" || exit 1 + else + case $prefix in + *\'*) qprefix=`echo "$prefix" | sed "s/'/'\\\\\\\\''/g"`;; + *) qprefix=$prefix;; + esac + prefixes="$prefixes '$qprefix'" + fi + fi + prefix=$prefix/ + done + + if test -n "$prefixes"; then + # Don't fail if two instances are running concurrently. + (umask $mkdir_umask && + eval "\$doit_exec \$mkdirprog $prefixes") || + test -d "$dstdir" || exit 1 + obsolete_mkdir_used=true + fi + fi + fi + + if test -n "$dir_arg"; then + { test -z "$chowncmd" || $doit $chowncmd "$dst"; } && + { test -z "$chgrpcmd" || $doit $chgrpcmd "$dst"; } && + { test "$obsolete_mkdir_used$chowncmd$chgrpcmd" = false || + test -z "$chmodcmd" || $doit $chmodcmd $mode "$dst"; } || exit 1 + else + + # Make a couple of temp file names in the proper directory. + dsttmp=$dstdir/_inst.$$_ + rmtmp=$dstdir/_rm.$$_ + + # Trap to clean up those temp files at exit. + trap 'ret=$?; rm -f "$dsttmp" "$rmtmp" && exit $ret' 0 + + # Copy the file name to the temp name. + (umask $cp_umask && $doit_exec $cpprog "$src" "$dsttmp") && + + # and set any options; do chmod last to preserve setuid bits. + # + # If any of these fail, we abort the whole thing. If we want to + # ignore errors from any of these, just make sure not to ignore + # errors from the above "$doit $cpprog $src $dsttmp" command. + # + { test -z "$chowncmd" || $doit $chowncmd "$dsttmp"; } && + { test -z "$chgrpcmd" || $doit $chgrpcmd "$dsttmp"; } && + { test -z "$stripcmd" || $doit $stripcmd "$dsttmp"; } && + { test -z "$chmodcmd" || $doit $chmodcmd $mode "$dsttmp"; } && + + # If -C, don't bother to copy if it wouldn't change the file. + if $copy_on_change && + old=`LC_ALL=C ls -dlL "$dst" 2>/dev/null` && + new=`LC_ALL=C ls -dlL "$dsttmp" 2>/dev/null` && + + eval "$initialize_posix_glob" && + $posix_glob set -f && + set X $old && old=:$2:$4:$5:$6 && + set X $new && new=:$2:$4:$5:$6 && + $posix_glob set +f && + + test "$old" = "$new" && + $cmpprog "$dst" "$dsttmp" >/dev/null 2>&1 + then + rm -f "$dsttmp" + else + # Rename the file to the real destination. + $doit $mvcmd -f "$dsttmp" "$dst" 2>/dev/null || + + # The rename failed, perhaps because mv can't rename something else + # to itself, or perhaps because mv is so ancient that it does not + # support -f. + { + # Now remove or move aside any old file at destination location. + # We try this two ways since rm can't unlink itself on some + # systems and the destination file might be busy for other + # reasons. In this case, the final cleanup might fail but the new + # file should still install successfully. + { + test ! -f "$dst" || + $doit $rmcmd -f "$dst" 2>/dev/null || + { $doit $mvcmd -f "$dst" "$rmtmp" 2>/dev/null && + { $doit $rmcmd -f "$rmtmp" 2>/dev/null; :; } + } || + { echo "$0: cannot unlink or rename $dst" >&2 + (exit 1); exit 1 + } + } && + + # Now rename the file to the real destination. + $doit $mvcmd "$dsttmp" "$dst" + } + fi || exit 1 + + trap '' 0 + fi +done + +# Local variables: +# eval: (add-hook 'write-file-hooks 'time-stamp) +# time-stamp-start: "scriptversion=" +# time-stamp-format: "%:y-%02m-%02d.%02H" +# time-stamp-end: "$" +# End: diff --git a/plugins/CDVDiso/src/libiso.c b/plugins/CDVDiso/src/libiso.cpp old mode 100644 new mode 100755 similarity index 100% rename from plugins/CDVDiso/src/libiso.c rename to plugins/CDVDiso/src/libiso.cpp diff --git a/plugins/CDVDiso/src/missing b/plugins/CDVDiso/src/missing new file mode 100755 index 0000000000..1c8ff7049d --- /dev/null +++ b/plugins/CDVDiso/src/missing @@ -0,0 +1,367 @@ +#! /bin/sh +# Common stub for a few missing GNU programs while installing. + +scriptversion=2006-05-10.23 + +# Copyright (C) 1996, 1997, 1999, 2000, 2002, 2003, 2004, 2005, 2006 +# Free Software Foundation, Inc. +# Originally by Fran,cois Pinard , 1996. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA +# 02110-1301, USA. + +# As a special exception to the GNU General Public License, if you +# distribute this file as part of a program that contains a +# configuration script generated by Autoconf, you may include it under +# the same distribution terms that you use for the rest of that program. + +if test $# -eq 0; then + echo 1>&2 "Try \`$0 --help' for more information" + exit 1 +fi + +run=: +sed_output='s/.* --output[ =]\([^ ]*\).*/\1/p' +sed_minuso='s/.* -o \([^ ]*\).*/\1/p' + +# In the cases where this matters, `missing' is being run in the +# srcdir already. +if test -f configure.ac; then + configure_ac=configure.ac +else + configure_ac=configure.in +fi + +msg="missing on your system" + +case $1 in +--run) + # Try to run requested program, and just exit if it succeeds. + run= + shift + "$@" && exit 0 + # Exit code 63 means version mismatch. This often happens + # when the user try to use an ancient version of a tool on + # a file that requires a minimum version. In this case we + # we should proceed has if the program had been absent, or + # if --run hadn't been passed. + if test $? = 63; then + run=: + msg="probably too old" + fi + ;; + + -h|--h|--he|--hel|--help) + echo "\ +$0 [OPTION]... PROGRAM [ARGUMENT]... + +Handle \`PROGRAM [ARGUMENT]...' for when PROGRAM is missing, or return an +error status if there is no known handling for PROGRAM. + +Options: + -h, --help display this help and exit + -v, --version output version information and exit + --run try to run the given command, and emulate it if it fails + +Supported PROGRAM values: + aclocal touch file \`aclocal.m4' + autoconf touch file \`configure' + autoheader touch file \`config.h.in' + autom4te touch the output file, or create a stub one + automake touch all \`Makefile.in' files + bison create \`y.tab.[ch]', if possible, from existing .[ch] + flex create \`lex.yy.c', if possible, from existing .c + help2man touch the output file + lex create \`lex.yy.c', if possible, from existing .c + makeinfo touch the output file + tar try tar, gnutar, gtar, then tar without non-portable flags + yacc create \`y.tab.[ch]', if possible, from existing .[ch] + +Send bug reports to ." + exit $? + ;; + + -v|--v|--ve|--ver|--vers|--versi|--versio|--version) + echo "missing $scriptversion (GNU Automake)" + exit $? + ;; + + -*) + echo 1>&2 "$0: Unknown \`$1' option" + echo 1>&2 "Try \`$0 --help' for more information" + exit 1 + ;; + +esac + +# Now exit if we have it, but it failed. Also exit now if we +# don't have it and --version was passed (most likely to detect +# the program). +case $1 in + lex|yacc) + # Not GNU programs, they don't have --version. + ;; + + tar) + if test -n "$run"; then + echo 1>&2 "ERROR: \`tar' requires --run" + exit 1 + elif test "x$2" = "x--version" || test "x$2" = "x--help"; then + exit 1 + fi + ;; + + *) + if test -z "$run" && ($1 --version) > /dev/null 2>&1; then + # We have it, but it failed. + exit 1 + elif test "x$2" = "x--version" || test "x$2" = "x--help"; then + # Could not run --version or --help. This is probably someone + # running `$TOOL --version' or `$TOOL --help' to check whether + # $TOOL exists and not knowing $TOOL uses missing. + exit 1 + fi + ;; +esac + +# If it does not exist, or fails to run (possibly an outdated version), +# try to emulate it. +case $1 in + aclocal*) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified \`acinclude.m4' or \`${configure_ac}'. You might want + to install the \`Automake' and \`Perl' packages. Grab them from + any GNU archive site." + touch aclocal.m4 + ;; + + autoconf) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified \`${configure_ac}'. You might want to install the + \`Autoconf' and \`GNU m4' packages. Grab them from any GNU + archive site." + touch configure + ;; + + autoheader) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified \`acconfig.h' or \`${configure_ac}'. You might want + to install the \`Autoconf' and \`GNU m4' packages. Grab them + from any GNU archive site." + files=`sed -n 's/^[ ]*A[CM]_CONFIG_HEADER(\([^)]*\)).*/\1/p' ${configure_ac}` + test -z "$files" && files="config.h" + touch_files= + for f in $files; do + case $f in + *:*) touch_files="$touch_files "`echo "$f" | + sed -e 's/^[^:]*://' -e 's/:.*//'`;; + *) touch_files="$touch_files $f.in";; + esac + done + touch $touch_files + ;; + + automake*) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified \`Makefile.am', \`acinclude.m4' or \`${configure_ac}'. + You might want to install the \`Automake' and \`Perl' packages. + Grab them from any GNU archive site." + find . -type f -name Makefile.am -print | + sed 's/\.am$/.in/' | + while read f; do touch "$f"; done + ;; + + autom4te) + echo 1>&2 "\ +WARNING: \`$1' is needed, but is $msg. + You might have modified some files without having the + proper tools for further handling them. + You can get \`$1' as part of \`Autoconf' from any GNU + archive site." + + file=`echo "$*" | sed -n "$sed_output"` + test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` + if test -f "$file"; then + touch $file + else + test -z "$file" || exec >$file + echo "#! /bin/sh" + echo "# Created by GNU Automake missing as a replacement of" + echo "# $ $@" + echo "exit 0" + chmod +x $file + exit 1 + fi + ;; + + bison|yacc) + echo 1>&2 "\ +WARNING: \`$1' $msg. You should only need it if + you modified a \`.y' file. You may need the \`Bison' package + in order for those modifications to take effect. You can get + \`Bison' from any GNU archive site." + rm -f y.tab.c y.tab.h + if test $# -ne 1; then + eval LASTARG="\${$#}" + case $LASTARG in + *.y) + SRCFILE=`echo "$LASTARG" | sed 's/y$/c/'` + if test -f "$SRCFILE"; then + cp "$SRCFILE" y.tab.c + fi + SRCFILE=`echo "$LASTARG" | sed 's/y$/h/'` + if test -f "$SRCFILE"; then + cp "$SRCFILE" y.tab.h + fi + ;; + esac + fi + if test ! -f y.tab.h; then + echo >y.tab.h + fi + if test ! -f y.tab.c; then + echo 'main() { return 0; }' >y.tab.c + fi + ;; + + lex|flex) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified a \`.l' file. You may need the \`Flex' package + in order for those modifications to take effect. You can get + \`Flex' from any GNU archive site." + rm -f lex.yy.c + if test $# -ne 1; then + eval LASTARG="\${$#}" + case $LASTARG in + *.l) + SRCFILE=`echo "$LASTARG" | sed 's/l$/c/'` + if test -f "$SRCFILE"; then + cp "$SRCFILE" lex.yy.c + fi + ;; + esac + fi + if test ! -f lex.yy.c; then + echo 'main() { return 0; }' >lex.yy.c + fi + ;; + + help2man) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified a dependency of a manual page. You may need the + \`Help2man' package in order for those modifications to take + effect. You can get \`Help2man' from any GNU archive site." + + file=`echo "$*" | sed -n "$sed_output"` + test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` + if test -f "$file"; then + touch $file + else + test -z "$file" || exec >$file + echo ".ab help2man is required to generate this page" + exit 1 + fi + ;; + + makeinfo) + echo 1>&2 "\ +WARNING: \`$1' is $msg. You should only need it if + you modified a \`.texi' or \`.texinfo' file, or any other file + indirectly affecting the aspect of the manual. The spurious + call might also be the consequence of using a buggy \`make' (AIX, + DU, IRIX). You might want to install the \`Texinfo' package or + the \`GNU make' package. Grab either from any GNU archive site." + # The file to touch is that specified with -o ... + file=`echo "$*" | sed -n "$sed_output"` + test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` + if test -z "$file"; then + # ... or it is the one specified with @setfilename ... + infile=`echo "$*" | sed 's/.* \([^ ]*\) *$/\1/'` + file=`sed -n ' + /^@setfilename/{ + s/.* \([^ ]*\) *$/\1/ + p + q + }' $infile` + # ... or it is derived from the source name (dir/f.texi becomes f.info) + test -z "$file" && file=`echo "$infile" | sed 's,.*/,,;s,.[^.]*$,,'`.info + fi + # If the file does not exist, the user really needs makeinfo; + # let's fail without touching anything. + test -f $file || exit 1 + touch $file + ;; + + tar) + shift + + # We have already tried tar in the generic part. + # Look for gnutar/gtar before invocation to avoid ugly error + # messages. + if (gnutar --version > /dev/null 2>&1); then + gnutar "$@" && exit 0 + fi + if (gtar --version > /dev/null 2>&1); then + gtar "$@" && exit 0 + fi + firstarg="$1" + if shift; then + case $firstarg in + *o*) + firstarg=`echo "$firstarg" | sed s/o//` + tar "$firstarg" "$@" && exit 0 + ;; + esac + case $firstarg in + *h*) + firstarg=`echo "$firstarg" | sed s/h//` + tar "$firstarg" "$@" && exit 0 + ;; + esac + fi + + echo 1>&2 "\ +WARNING: I can't seem to be able to run \`tar' with the given arguments. + You may want to install GNU tar or Free paxutils, or check the + command line arguments." + exit 1 + ;; + + *) + echo 1>&2 "\ +WARNING: \`$1' is needed, and is $msg. + You might have modified some files without having the + proper tools for further handling them. Check the \`README' file, + it often tells you about the needed prerequisites for installing + this package. You may also peek at any GNU archive site, in case + some other package would contain this missing \`$1' program." + exit 1 + ;; +esac + +exit 0 + +# Local variables: +# eval: (add-hook 'write-file-hooks 'time-stamp) +# time-stamp-start: "scriptversion=" +# time-stamp-format: "%:y-%02m-%02d.%02H" +# time-stamp-end: "$" +# End: diff --git a/plugins/CDVDiso/src/mkiso/Makefile b/plugins/CDVDiso/src/mkiso/Makefile deleted file mode 100644 index 90beb81202..0000000000 --- a/plugins/CDVDiso/src/mkiso/Makefile +++ /dev/null @@ -1,27 +0,0 @@ - -CC = gcc - -MKISO = mkiso -CFLAGS = -fPIC -Wall -O2 -fomit-frame-pointer -I.. -I. -D__LINUX__ -I../../../../3rdparty/zlib -D_FILE_OFFSET_BITS=64 -OBJS = mkiso.o ../libiso.o -LIBS = -OBJS+= ../../../../3rdparty/zlib/adler32.o ../../../../3rdparty/zlib/compress.o ../../../../3rdparty/zlib/crc32.o ../../../../3rdparty/zlib/gzio.o \ -../../../../3rdparty/zlib/uncompr.o ../../../../3rdparty/zlib/deflate.o ../../../../3rdparty/zlib/trees.o ../../../../3rdparty/zlib/zutil.o \ -../../../../3rdparty/zlib/inflate.o ../../../../3rdparty/zlib/infback.o ../../../../3rdparty/zlib/inftrees.o ../../../../3rdparty/zlib/inffast.o - -DEPS:= $(OBJS:.o=.d) - -all: mkiso - -mkiso: ${OBJS} - rm -f ${MKISO} - ${CC} ${CFLAGS} ${OBJS} -o ${MKISO} ${LIBS} - strip ${MKISO} - -clean: - rm -f ${OBJS} ${DEPS} ${MKISO} - -%.o: %.c - ${CC} ${CFLAGS} -c -o $@ $< -MD -MF $(patsubst %.o,%.d,$@) - --include ${DEPS} diff --git a/plugins/CDVDiso/src/mkiso/mkiso.c b/plugins/CDVDiso/src/mkiso/mkiso.cpp old mode 100644 new mode 100755 similarity index 98% rename from plugins/CDVDiso/src/mkiso/mkiso.c rename to plugins/CDVDiso/src/mkiso/mkiso.cpp index df31a048a3..6a25d2bb8b --- a/plugins/CDVDiso/src/mkiso/mkiso.c +++ b/plugins/CDVDiso/src/mkiso/mkiso.cpp @@ -20,6 +20,7 @@ #include #include #include +#include "string.h" #include "CDVDiso.h" @@ -150,7 +151,7 @@ void Decompress(char *filename) } -int main(int argc, char *argv[]) +/*int main(int argc, char *argv[]) { if (argc < 3) return 0; @@ -169,4 +170,4 @@ int main(int argc, char *argv[]) } return 0; -} +}*/ From 74ebe985f6fb2ca83fa657bc360f188cedd17668 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 2 May 2009 14:14:34 +0000 Subject: [PATCH 002/296] Plugins: Fix the dialog boxes on FWnull & USBnull so the buttons work. Fix inconsistant version numbers. Delete a stray file. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1121 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/CDVDiso/src/Windows/svnrev.h | 18 ------------------ plugins/FWnull/Linux/Config.cpp | 10 ++++++++-- plugins/FWnull/Makefile.am | 3 ++- plugins/FWnull/build.sh | 7 +------ plugins/SPU2null/configure.ac | 8 ++++---- plugins/USBnull/Linux/Config.cpp | 10 ++++++++-- plugins/USBnull/Makefile.am | 3 ++- plugins/USBnull/build.sh | 5 ----- plugins/USBnull/configure.ac | 2 +- plugins/dev9null/configure.ac | 2 +- 10 files changed, 27 insertions(+), 41 deletions(-) delete mode 100755 plugins/CDVDiso/src/Windows/svnrev.h diff --git a/plugins/CDVDiso/src/Windows/svnrev.h b/plugins/CDVDiso/src/Windows/svnrev.h deleted file mode 100755 index d0992dc800..0000000000 --- a/plugins/CDVDiso/src/Windows/svnrev.h +++ /dev/null @@ -1,18 +0,0 @@ -// svnrev_template.h --> svnrev.h -// -// This file acts as a template for the automatic SVN revision/version tag. -// It is used by the utility SubWCrev.exe to create an "svnrev.h" file for -// whichever project is being compiled (as indicated by command line options -// passed to SubWCRev.exe during the project's pre-build step). -// -// The SubWCRev.exe utility is part of TortoiseSVN and requires several DLLs -// installed by TortoiseSVN, so it will only be available if you have TortoiseSVN -// installed on your system. If you do not have it installed, a generic template -// is used instead (see svnrev_generic.h). Having TortoiseSVN is handy but not -// necessary. If you do not have it installed, everything will still compile -// fine except without the SVN revision tagged to the application/dll version. -// -// TortoiseSVN can be downloaded from http://tortoisesvn.tigris.org - -#define SVN_REV 798 -#define SVN_MODS 1 \ No newline at end of file diff --git a/plugins/FWnull/Linux/Config.cpp b/plugins/FWnull/Linux/Config.cpp index 515914ca1b..5cd7583574 100644 --- a/plugins/FWnull/Linux/Config.cpp +++ b/plugins/FWnull/Linux/Config.cpp @@ -29,12 +29,18 @@ using namespace std; #include "FW.h" #include "Config.h" + #ifdef __cplusplus extern "C" { -#include "interface.h" +#endif + #include "support.h" -//#include "callbacks.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus } +#endif GtkWidget *MsgDlg, *About, *Conf; extern string s_strIniPath; diff --git a/plugins/FWnull/Makefile.am b/plugins/FWnull/Makefile.am index 708c80c51b..60060ae259 100644 --- a/plugins/FWnull/Makefile.am +++ b/plugins/FWnull/Makefile.am @@ -28,6 +28,7 @@ FW.h Linux/Config.h Linux/Linux.h libFWnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ -Linux/callbacks.c Linux/callbacks.h +Linux/callbacks.h +#Linux/callbacks.c #SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/FWnull/build.sh b/plugins/FWnull/build.sh index e45a9a3793..c2df048482 100644 --- a/plugins/FWnull/build.sh +++ b/plugins/FWnull/build.sh @@ -29,9 +29,4 @@ fi if [ $? -ne 0 ] then exit 1 -fi - -if [ -s cfgFWnull ] && [ -s libFWnull.so ] -then -cp cfgFWnull libFWnull.so ${PCSX2PLUGINS} -fi +fi \ No newline at end of file diff --git a/plugins/SPU2null/configure.ac b/plugins/SPU2null/configure.ac index 59ffe18988..2cc0ffb87d 100644 --- a/plugins/SPU2null/configure.ac +++ b/plugins/SPU2null/configure.ac @@ -1,6 +1,6 @@ -AC_INIT(SPU2null, 0.5,arcum42@gmail.com) +AC_INIT(SPU2null, 0.71,arcum42@gmail.com) -AM_INIT_AUTOMAKE(SPU2null,0.8) +AM_INIT_AUTOMAKE(SPU2null,0.71) AC_PROG_CC([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) AC_PROG_CXX([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) @@ -13,8 +13,8 @@ dnl necessary for compiling assembly AM_PROG_AS AC_SUBST(SPU2null_CURRENT, 0) -AC_SUBST(SPU2null_REVISION, 8) -AC_SUBST(SPU2null_AGE, 0) +AC_SUBST(SPU2null_REVISION, 7) +AC_SUBST(SPU2null_AGE, 1) AC_SUBST(SPU2null_RELEASE,[$SPU2null_CURRENT].[$SPU2null_REVISION].[$SPU2null_AGE]) AC_SUBST(SPU2null_SONAME,libSPU2null.so.[$SPU2null_CURRENT].[$SPU2null_REVISION].[$SPU2null_AGE]) diff --git a/plugins/USBnull/Linux/Config.cpp b/plugins/USBnull/Linux/Config.cpp index 1456f2e2d6..f7b43f4852 100644 --- a/plugins/USBnull/Linux/Config.cpp +++ b/plugins/USBnull/Linux/Config.cpp @@ -26,12 +26,18 @@ #include using namespace std; + #ifdef __cplusplus extern "C" { -#include "interface.h" +#endif + #include "support.h" -//#include "callbacks.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus } +#endif #include "USB.h" diff --git a/plugins/USBnull/Makefile.am b/plugins/USBnull/Makefile.am index d4241d3ad0..838ab78919 100644 --- a/plugins/USBnull/Makefile.am +++ b/plugins/USBnull/Makefile.am @@ -28,6 +28,7 @@ USB.h Linux/Config.h Linux/Linux.h libUSBnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ -Linux/callbacks.c Linux/callbacks.h +Linux/callbacks.h +#Linux/callbacks.c #SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/USBnull/build.sh b/plugins/USBnull/build.sh index 65d48eb346..9965413fd5 100644 --- a/plugins/USBnull/build.sh +++ b/plugins/USBnull/build.sh @@ -30,8 +30,3 @@ if [ $? -ne 0 ] then exit 1 fi - -#if [ -s cfgFWnull ] && [ -s libUSBnull.so ] -#then -#cp cfgFWnull libFWnull.so ${PCSX2PLUGINS} -#fi diff --git a/plugins/USBnull/configure.ac b/plugins/USBnull/configure.ac index 226e7f2036..4a5de1dc8d 100644 --- a/plugins/USBnull/configure.ac +++ b/plugins/USBnull/configure.ac @@ -1,4 +1,4 @@ -AC_INIT(USBnull, 0.5,arcum42@gmail.com) +AC_INIT(USBnull, 0.6,arcum42@gmail.com) AM_INIT_AUTOMAKE(USBnull,0.6) diff --git a/plugins/dev9null/configure.ac b/plugins/dev9null/configure.ac index f2573aa6d2..a14ec42970 100644 --- a/plugins/dev9null/configure.ac +++ b/plugins/dev9null/configure.ac @@ -13,7 +13,7 @@ dnl necessary for compiling assembly AM_PROG_AS AC_SUBST(DEV9null_CURRENT, 0) -AC_SUBST(DEV9null_REVISION, 8) +AC_SUBST(DEV9null_REVISION, 4) AC_SUBST(DEV9null_AGE, 0) AC_SUBST(DEV9null_RELEASE,[$DEV9null_CURRENT].[$DEV9null_REVISION].[$DEV9null_AGE]) AC_SUBST(DEV9null_SONAME,libDEV9null.so.[$DEV9null_CURRENT].[$DEV9null_REVISION].[$DEV9null_AGE]) From a65543399ce8eb70811c7c1c9f211896b4f9c0d6 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 2 May 2009 14:36:46 +0000 Subject: [PATCH 003/296] Linux: Iron out a few issues with the configuration dialog in Linux. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1122 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/ConfigDlg.cpp | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/pcsx2/Linux/ConfigDlg.cpp b/pcsx2/Linux/ConfigDlg.cpp index df988d459a..8ba14b1a08 100644 --- a/pcsx2/Linux/ConfigDlg.cpp +++ b/pcsx2/Linux/ConfigDlg.cpp @@ -157,10 +157,9 @@ void OnConfButton(GtkButton *button, gpointer user_data) plugin_types type = strToPluginType(name); plugin_callback call = strToPluginCall(name); - // Don't uncomment till fixing CDVDIso's dialog box. - //gtk_widget_set_sensitive(ConfDlg, FALSE); - ConfPlugin(type, call, false); - //gtk_widget_set_sensitive(ConfDlg, TRUE); + gtk_widget_set_sensitive(ConfDlg, FALSE); + ConfPlugin(type, call, true); + gtk_widget_set_sensitive(ConfDlg, TRUE); } void SetComboToGList(GtkComboBox *widget, GList *list) From 8504f4e9bd4dca27327845e4594df9b20ad5c83e Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 3 May 2009 05:28:13 +0000 Subject: [PATCH 004/296] Atn Jake: Please review. Quick patch to counter problems reported with r1116. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1123 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Misc.cpp | 2 +- pcsx2/ps2/Iop/IopHwRead.cpp | 14 ++++++++++++-- 2 files changed, 13 insertions(+), 3 deletions(-) diff --git a/pcsx2/Misc.cpp b/pcsx2/Misc.cpp index d129f4ed9a..7899499157 100644 --- a/pcsx2/Misc.cpp +++ b/pcsx2/Misc.cpp @@ -437,7 +437,7 @@ void CycleFrameLimit(int dir) const char* limitMsg; u32 newOptions; u32 curFrameLimit = Config.Options & PCSX2_FRAMELIMIT_MASK; - u32 newFrameLimit; + u32 newFrameLimit = 0; static u32 oldFrameLimit = PCSX2_FRAMELIMIT_LIMIT; if( dir == 0 ) { diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index 85202e74c3..9263b71e02 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -399,7 +399,7 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) { switch( masked_addr ) { - mcase(0x1f801264): // unknown / reserved. + mcase(HW_SIO2_FIFO): // 0x1f801264 wouldn't make it here, so I assume you mean 01f808264, which is fifo, Jake... ret = psxHu32(addr); break; @@ -431,7 +431,17 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) ret = sio2_getIntr(); break; - jNO_DEFAULT; + mcase(0x1f80847c): // I'm adding this to try to counter reports of pcsx2 crashing when this address is called. + Console::Error("iopHwRead32_Page8: Calling 0x1f80847c for some unknown reason."); + ret = psxHu32(addr); + break; + + // Do we really want to assert here, if it's an unknown address? + //jNO_DEFAULT; + default: + Console::Error("iopHwRead32_Page8: Why are we calling 0x%x?", params addr); + ret = psxHu32(addr); + break; } } } From 94c006efd23b34fea37a4653c5b9ba58c69d6ec6 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 3 May 2009 11:09:15 +0000 Subject: [PATCH 005/296] ZeroPad: Apply the multiple key patch in issue 197 to ZeroPad. Currently Linux only. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1125 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/callbacks.h | 20 +++-- plugins/zeropad/Linux/interface.c | 90 +++++++++++++++++++ plugins/zeropad/Linux/linux.cpp | 129 ++++++++++++++-------------- plugins/zeropad/Linux/zeropad.glade | 43 ++++++++++ plugins/zeropad/Windows/win.cpp | 30 ++----- plugins/zeropad/zeropad.cpp | 47 +++++----- plugins/zeropad/zeropad.h | 12 ++- 7 files changed, 258 insertions(+), 113 deletions(-) diff --git a/plugins/zeropad/Linux/callbacks.h b/plugins/zeropad/Linux/callbacks.h index cb64e16292..d9c6d64ae1 100644 --- a/plugins/zeropad/Linux/callbacks.h +++ b/plugins/zeropad/Linux/callbacks.h @@ -10,21 +10,21 @@ OnConf_Pad2 (GtkButton *button, gpointer user_data); void -OnConf_Key (GtkButton *button, +OnConf_Pad3 (GtkButton *button, gpointer user_data); void -OnConf_Cancel (GtkButton *button, - gpointer user_data); - -void -OnConf_Ok (GtkButton *button, +OnConf_Pad4 (GtkButton *button, gpointer user_data); void on_joydevicescombo_changed (GtkComboBox *combobox, gpointer user_data); +void +OnConf_Key (GtkButton *button, + gpointer user_data); + void on_checkbutton_reverselx_toggled (GtkToggleButton *togglebutton, gpointer user_data); @@ -44,3 +44,11 @@ on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, void on_forcefeedback_toggled (GtkToggleButton *togglebutton, gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/zeropad/Linux/interface.c b/plugins/zeropad/Linux/interface.c index ffd4261e65..7b32fe1606 100644 --- a/plugins/zeropad/Linux/interface.c +++ b/plugins/zeropad/Linux/interface.c @@ -42,6 +42,8 @@ create_About (void) gchar *translators = _("translator-credits"); About = gtk_about_dialog_new (); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); gtk_about_dialog_set_version (GTK_ABOUT_DIALOG (About), VERSION); gtk_about_dialog_set_name (GTK_ABOUT_DIALOG (About), _("ZeroPAD")); gtk_about_dialog_set_authors (GTK_ABOUT_DIALOG (About), authors); @@ -65,6 +67,8 @@ create_Conf (void) GtkWidget *PAD1; GSList *PAD1_group = NULL; GtkWidget *PAD2; + GtkWidget *PAD3; + GtkWidget *PAD4; GtkWidget *label2; GtkWidget *vbox2; GtkWidget *label5; @@ -123,30 +127,37 @@ create_Conf (void) GtkWidget *okbutton1; Conf = gtk_dialog_new (); + gtk_widget_set_name (Conf, "Conf"); gtk_window_set_title (GTK_WINDOW (Conf), _("ZeroPAD Configuration Dialog")); gtk_window_set_type_hint (GTK_WINDOW (Conf), GDK_WINDOW_TYPE_HINT_DIALOG); dialog_vbox1 = GTK_DIALOG (Conf)->vbox; + gtk_widget_set_name (dialog_vbox1, "dialog_vbox1"); gtk_widget_show (dialog_vbox1); vbox1 = gtk_vbox_new (FALSE, 0); + gtk_widget_set_name (vbox1, "vbox1"); gtk_widget_show (vbox1); gtk_box_pack_start (GTK_BOX (dialog_vbox1), vbox1, TRUE, TRUE, 0); frame1 = gtk_frame_new (NULL); + gtk_widget_set_name (frame1, "frame1"); gtk_widget_show (frame1); gtk_box_pack_start (GTK_BOX (vbox1), frame1, FALSE, FALSE, 0); alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); gtk_widget_show (alignment1); gtk_container_add (GTK_CONTAINER (frame1), alignment1); gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); hbox1 = gtk_hbox_new (FALSE, 0); + gtk_widget_set_name (hbox1, "hbox1"); gtk_widget_show (hbox1); gtk_container_add (GTK_CONTAINER (alignment1), hbox1); PAD1 = gtk_radio_button_new_with_mnemonic (NULL, _("PAD1")); + gtk_widget_set_name (PAD1, "PAD1"); gtk_widget_show (PAD1); gtk_box_pack_start (GTK_BOX (hbox1), PAD1, FALSE, FALSE, 0); gtk_radio_button_set_group (GTK_RADIO_BUTTON (PAD1), PAD1_group); @@ -154,33 +165,54 @@ create_Conf (void) gtk_toggle_button_set_active (GTK_TOGGLE_BUTTON (PAD1), TRUE); PAD2 = gtk_radio_button_new_with_mnemonic (NULL, _("PAD2")); + gtk_widget_set_name (PAD2, "PAD2"); gtk_widget_show (PAD2); gtk_box_pack_start (GTK_BOX (hbox1), PAD2, FALSE, FALSE, 0); gtk_radio_button_set_group (GTK_RADIO_BUTTON (PAD2), PAD1_group); PAD1_group = gtk_radio_button_get_group (GTK_RADIO_BUTTON (PAD2)); + PAD3 = gtk_radio_button_new_with_mnemonic (NULL, _("PAD1 alt.")); + gtk_widget_set_name (PAD3, "PAD3"); + gtk_widget_show (PAD3); + gtk_box_pack_start (GTK_BOX (hbox1), PAD3, FALSE, FALSE, 0); + gtk_radio_button_set_group (GTK_RADIO_BUTTON (PAD3), PAD1_group); + PAD1_group = gtk_radio_button_get_group (GTK_RADIO_BUTTON (PAD3)); + + PAD4 = gtk_radio_button_new_with_mnemonic (NULL, _("PAD2 alt.")); + gtk_widget_set_name (PAD4, "PAD4"); + gtk_widget_show (PAD4); + gtk_box_pack_start (GTK_BOX (hbox1), PAD4, FALSE, FALSE, 0); + gtk_radio_button_set_group (GTK_RADIO_BUTTON (PAD4), PAD1_group); + PAD1_group = gtk_radio_button_get_group (GTK_RADIO_BUTTON (PAD4)); + label2 = gtk_label_new (_("Choose PAD to modify")); + gtk_widget_set_name (label2, "label2"); gtk_widget_show (label2); gtk_frame_set_label_widget (GTK_FRAME (frame1), label2); gtk_label_set_use_markup (GTK_LABEL (label2), TRUE); vbox2 = gtk_vbox_new (FALSE, 0); + gtk_widget_set_name (vbox2, "vbox2"); gtk_widget_show (vbox2); gtk_box_pack_start (GTK_BOX (vbox1), vbox2, TRUE, TRUE, 0); label5 = gtk_label_new (_("Joystick to use for this PAD")); + gtk_widget_set_name (label5, "label5"); gtk_widget_show (label5); gtk_box_pack_start (GTK_BOX (vbox2), label5, FALSE, FALSE, 0); joydevicescombo = gtk_combo_box_entry_new_text (); + gtk_widget_set_name (joydevicescombo, "joydevicescombo"); gtk_widget_show (joydevicescombo); gtk_box_pack_start (GTK_BOX (vbox2), joydevicescombo, TRUE, TRUE, 0); fixed1 = gtk_fixed_new (); + gtk_widget_set_name (fixed1, "fixed1"); gtk_widget_show (fixed1); gtk_box_pack_start (GTK_BOX (vbox1), fixed1, TRUE, TRUE, 0); eL2 = gtk_entry_new (); + gtk_widget_set_name (eL2, "eL2"); gtk_widget_show (eL2); gtk_fixed_put (GTK_FIXED (fixed1), eL2, 64, 40); gtk_widget_set_size_request (eL2, 64, 24); @@ -188,12 +220,14 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eL2), 8226); eL1 = gtk_entry_new (); + gtk_widget_set_name (eL1, "eL1"); gtk_widget_show (eL1); gtk_fixed_put (GTK_FIXED (fixed1), eL1, 64, 104); gtk_widget_set_size_request (eL1, 64, 24); gtk_entry_set_invisible_char (GTK_ENTRY (eL1), 8226); eSelect = gtk_entry_new (); + gtk_widget_set_name (eSelect, "eSelect"); gtk_widget_show (eSelect); gtk_fixed_put (GTK_FIXED (fixed1), eSelect, 168, 64); gtk_widget_set_size_request (eSelect, 64, 24); @@ -201,6 +235,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eSelect), 8226); eStart = gtk_entry_new (); + gtk_widget_set_name (eStart, "eStart"); gtk_widget_show (eStart); gtk_fixed_put (GTK_FIXED (fixed1), eStart, 280, 64); gtk_widget_set_size_request (eStart, 64, 24); @@ -208,12 +243,14 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eStart), 8226); eUp = gtk_entry_new (); + gtk_widget_set_name (eUp, "eUp"); gtk_widget_show (eUp); gtk_fixed_put (GTK_FIXED (fixed1), eUp, 64, 168); gtk_widget_set_size_request (eUp, 64, 24); gtk_entry_set_invisible_char (GTK_ENTRY (eUp), 8226); eRight = gtk_entry_new (); + gtk_widget_set_name (eRight, "eRight"); gtk_widget_show (eRight); gtk_fixed_put (GTK_FIXED (fixed1), eRight, 128, 208); gtk_widget_set_size_request (eRight, 64, 24); @@ -221,6 +258,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eRight), 8226); eLeft = gtk_entry_new (); + gtk_widget_set_name (eLeft, "eLeft"); gtk_widget_show (eLeft); gtk_fixed_put (GTK_FIXED (fixed1), eLeft, 0, 208); gtk_widget_set_size_request (eLeft, 64, 24); @@ -228,6 +266,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eLeft), 8226); eDown = gtk_entry_new (); + gtk_widget_set_name (eDown, "eDown"); gtk_widget_show (eDown); gtk_fixed_put (GTK_FIXED (fixed1), eDown, 64, 256); gtk_widget_set_size_request (eDown, 64, 24); @@ -235,6 +274,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eDown), 8226); eR2 = gtk_entry_new (); + gtk_widget_set_name (eR2, "eR2"); gtk_widget_show (eR2); gtk_fixed_put (GTK_FIXED (fixed1), eR2, 368, 32); gtk_widget_set_size_request (eR2, 64, 24); @@ -242,6 +282,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eR2), 8226); eR1 = gtk_entry_new (); + gtk_widget_set_name (eR1, "eR1"); gtk_widget_show (eR1); gtk_fixed_put (GTK_FIXED (fixed1), eR1, 368, 104); gtk_widget_set_size_request (eR1, 64, 24); @@ -249,6 +290,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eR1), 8226); eAnalog = gtk_entry_new (); + gtk_widget_set_name (eAnalog, "eAnalog"); gtk_widget_show (eAnalog); gtk_fixed_put (GTK_FIXED (fixed1), eAnalog, 224, 128); gtk_widget_set_size_request (eAnalog, 64, 24); @@ -256,6 +298,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eAnalog), 8226); eSquare = gtk_entry_new (); + gtk_widget_set_name (eSquare, "eSquare"); gtk_widget_show (eSquare); gtk_fixed_put (GTK_FIXED (fixed1), eSquare, 304, 208); gtk_widget_set_size_request (eSquare, 64, 24); @@ -263,6 +306,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eSquare), 8226); eTriangle = gtk_entry_new (); + gtk_widget_set_name (eTriangle, "eTriangle"); gtk_widget_show (eTriangle); gtk_fixed_put (GTK_FIXED (fixed1), eTriangle, 368, 168); gtk_widget_set_size_request (eTriangle, 64, 24); @@ -270,6 +314,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eTriangle), 8226); eCircle = gtk_entry_new (); + gtk_widget_set_name (eCircle, "eCircle"); gtk_widget_show (eCircle); gtk_fixed_put (GTK_FIXED (fixed1), eCircle, 432, 208); gtk_widget_set_size_request (eCircle, 64, 24); @@ -277,6 +322,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eCircle), 8226); eCross = gtk_entry_new (); + gtk_widget_set_name (eCross, "eCross"); gtk_widget_show (eCross); gtk_fixed_put (GTK_FIXED (fixed1), eCross, 368, 256); gtk_widget_set_size_request (eCross, 64, 24); @@ -284,6 +330,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eCross), 8226); eR3 = gtk_entry_new (); + gtk_widget_set_name (eR3, "eR3"); gtk_widget_show (eR3); gtk_fixed_put (GTK_FIXED (fixed1), eR3, 272, 272); gtk_widget_set_size_request (eR3, 64, 24); @@ -291,6 +338,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eR3), 8226); eL3 = gtk_entry_new (); + gtk_widget_set_name (eL3, "eL3"); gtk_widget_show (eL3); gtk_fixed_put (GTK_FIXED (fixed1), eL3, 176, 272); gtk_widget_set_size_request (eL3, 64, 24); @@ -298,6 +346,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eL3), 8226); eRx = gtk_entry_new (); + gtk_widget_set_name (eRx, "eRx"); gtk_widget_show (eRx); gtk_fixed_put (GTK_FIXED (fixed1), eRx, 272, 352); gtk_widget_set_size_request (eRx, 80, 24); @@ -305,6 +354,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eRx), 8226); eLx = gtk_entry_new (); + gtk_widget_set_name (eLx, "eLx"); gtk_widget_show (eLx); gtk_fixed_put (GTK_FIXED (fixed1), eLx, 160, 352); gtk_widget_set_size_request (eLx, 80, 24); @@ -312,6 +362,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eLx), 8226); eRy = gtk_entry_new (); + gtk_widget_set_name (eRy, "eRy"); gtk_widget_show (eRy); gtk_fixed_put (GTK_FIXED (fixed1), eRy, 272, 416); gtk_widget_set_size_request (eRy, 80, 24); @@ -319,6 +370,7 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eRy), 8226); eLy = gtk_entry_new (); + gtk_widget_set_name (eLy, "eLy"); gtk_widget_show (eLy); gtk_fixed_put (GTK_FIXED (fixed1), eLy, 160, 416); gtk_widget_set_size_request (eLy, 80, 24); @@ -326,151 +378,181 @@ create_Conf (void) gtk_entry_set_invisible_char (GTK_ENTRY (eLy), 8226); Select = gtk_button_new_with_mnemonic (_("Select")); + gtk_widget_set_name (Select, "Select"); gtk_widget_show (Select); gtk_fixed_put (GTK_FIXED (fixed1), Select, 168, 32); gtk_widget_set_size_request (Select, 64, 32); L1 = gtk_button_new_with_mnemonic (_("L1")); + gtk_widget_set_name (L1, "L1"); gtk_widget_show (L1); gtk_fixed_put (GTK_FIXED (fixed1), L1, 64, 72); gtk_widget_set_size_request (L1, 64, 32); Left = gtk_button_new_with_mnemonic (_("Left")); + gtk_widget_set_name (Left, "Left"); gtk_widget_show (Left); gtk_fixed_put (GTK_FIXED (fixed1), Left, 0, 176); gtk_widget_set_size_request (Left, 64, 32); Down = gtk_button_new_with_mnemonic (_("Down")); + gtk_widget_set_name (Down, "Down"); gtk_widget_show (Down); gtk_fixed_put (GTK_FIXED (fixed1), Down, 64, 224); gtk_widget_set_size_request (Down, 64, 32); Analog = gtk_button_new_with_mnemonic (_("Analog")); + gtk_widget_set_name (Analog, "Analog"); gtk_widget_show (Analog); gtk_fixed_put (GTK_FIXED (fixed1), Analog, 224, 96); gtk_widget_set_size_request (Analog, 64, 32); R2 = gtk_button_new_with_mnemonic (_("R2")); + gtk_widget_set_name (R2, "R2"); gtk_widget_show (R2); gtk_fixed_put (GTK_FIXED (fixed1), R2, 368, 0); gtk_widget_set_size_request (R2, 64, 32); Start = gtk_button_new_with_mnemonic (_("Start")); + gtk_widget_set_name (Start, "Start"); gtk_widget_show (Start); gtk_fixed_put (GTK_FIXED (fixed1), Start, 280, 32); gtk_widget_set_size_request (Start, 64, 32); R1 = gtk_button_new_with_mnemonic (_("R1")); + gtk_widget_set_name (R1, "R1"); gtk_widget_show (R1); gtk_fixed_put (GTK_FIXED (fixed1), R1, 368, 72); gtk_widget_set_size_request (R1, 64, 32); Cross = gtk_button_new_with_mnemonic (_("Cross")); + gtk_widget_set_name (Cross, "Cross"); gtk_widget_show (Cross); gtk_fixed_put (GTK_FIXED (fixed1), Cross, 368, 224); gtk_widget_set_size_request (Cross, 64, 32); Triangle = gtk_button_new_with_mnemonic (_("Triangle")); + gtk_widget_set_name (Triangle, "Triangle"); gtk_widget_show (Triangle); gtk_fixed_put (GTK_FIXED (fixed1), Triangle, 368, 136); gtk_widget_set_size_request (Triangle, 64, 32); Circle = gtk_button_new_with_mnemonic (_("Circle")); + gtk_widget_set_name (Circle, "Circle"); gtk_widget_show (Circle); gtk_fixed_put (GTK_FIXED (fixed1), Circle, 432, 176); gtk_widget_set_size_request (Circle, 64, 32); Square = gtk_button_new_with_mnemonic (_("Square")); + gtk_widget_set_name (Square, "Square"); gtk_widget_show (Square); gtk_fixed_put (GTK_FIXED (fixed1), Square, 304, 176); gtk_widget_set_size_request (Square, 64, 32); L3 = gtk_button_new_with_mnemonic (_("L3")); + gtk_widget_set_name (L3, "L3"); gtk_widget_show (L3); gtk_fixed_put (GTK_FIXED (fixed1), L3, 176, 240); gtk_widget_set_size_request (L3, 64, 32); Up = gtk_button_new_with_mnemonic (_("Up")); + gtk_widget_set_name (Up, "Up"); gtk_widget_show (Up); gtk_fixed_put (GTK_FIXED (fixed1), Up, 64, 136); gtk_widget_set_size_request (Up, 64, 32); L2 = gtk_button_new_with_mnemonic (_("L2")); + gtk_widget_set_name (L2, "L2"); gtk_widget_show (L2); gtk_fixed_put (GTK_FIXED (fixed1), L2, 64, 8); gtk_widget_set_size_request (L2, 64, 32); Lx = gtk_button_new_with_mnemonic (_("Lx")); + gtk_widget_set_name (Lx, "Lx"); gtk_widget_show (Lx); gtk_fixed_put (GTK_FIXED (fixed1), Lx, 160, 320); gtk_widget_set_size_request (Lx, 80, 32); Ly = gtk_button_new_with_mnemonic (_("Ly")); + gtk_widget_set_name (Ly, "Ly"); gtk_widget_show (Ly); gtk_fixed_put (GTK_FIXED (fixed1), Ly, 160, 384); gtk_widget_set_size_request (Ly, 80, 32); Rx = gtk_button_new_with_mnemonic (_("Rx")); + gtk_widget_set_name (Rx, "Rx"); gtk_widget_show (Rx); gtk_fixed_put (GTK_FIXED (fixed1), Rx, 272, 320); gtk_widget_set_size_request (Rx, 80, 32); Ry = gtk_button_new_with_mnemonic (_("Ry")); + gtk_widget_set_name (Ry, "Ry"); gtk_widget_show (Ry); gtk_fixed_put (GTK_FIXED (fixed1), Ry, 272, 384); gtk_widget_set_size_request (Ry, 80, 32); Right = gtk_button_new_with_mnemonic (_("Right")); + gtk_widget_set_name (Right, "Right"); gtk_widget_show (Right); gtk_fixed_put (GTK_FIXED (fixed1), Right, 128, 176); gtk_widget_set_size_request (Right, 64, 32); label3 = gtk_label_new (_("Analog Controls (move mouse or analog joystick to select)")); + gtk_widget_set_name (label3, "label3"); gtk_widget_show (label3); gtk_fixed_put (GTK_FIXED (fixed1), label3, 48, 296); gtk_widget_set_size_request (label3, 408, 16); gtk_label_set_single_line_mode (GTK_LABEL (label3), TRUE); R3 = gtk_button_new_with_mnemonic (_("R3")); + gtk_widget_set_name (R3, "R3"); gtk_widget_show (R3); gtk_fixed_put (GTK_FIXED (fixed1), R3, 272, 240); gtk_widget_set_size_request (R3, 64, 32); checkbutton_reverselx = gtk_check_button_new_with_mnemonic (_("Reverse LX")); + gtk_widget_set_name (checkbutton_reverselx, "checkbutton_reverselx"); gtk_widget_show (checkbutton_reverselx); gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverselx, 16, 328); gtk_widget_set_size_request (checkbutton_reverselx, 111, 22); checkbutton_reversely = gtk_check_button_new_with_mnemonic (_("Reverse LY")); + gtk_widget_set_name (checkbutton_reversely, "checkbutton_reversely"); gtk_widget_show (checkbutton_reversely); gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversely, 16, 352); gtk_widget_set_size_request (checkbutton_reversely, 111, 22); checkbutton_reverserx = gtk_check_button_new_with_mnemonic (_("Reverse RX")); + gtk_widget_set_name (checkbutton_reverserx, "checkbutton_reverserx"); gtk_widget_show (checkbutton_reverserx); gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverserx, 16, 376); gtk_widget_set_size_request (checkbutton_reverserx, 111, 22); checkbutton_reversery = gtk_check_button_new_with_mnemonic (_("Reverse RY")); + gtk_widget_set_name (checkbutton_reversery, "checkbutton_reversery"); gtk_widget_show (checkbutton_reversery); gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversery, 16, 400); gtk_widget_set_size_request (checkbutton_reversery, 111, 22); forcefeedback = gtk_check_button_new_with_mnemonic (_("Enable Force\nFeedback")); + gtk_widget_set_name (forcefeedback, "forcefeedback"); gtk_widget_show (forcefeedback); gtk_fixed_put (GTK_FIXED (fixed1), forcefeedback, 384, 320); gtk_widget_set_size_request (forcefeedback, 112, 48); dialog_action_area1 = GTK_DIALOG (Conf)->action_area; + gtk_widget_set_name (dialog_action_area1, "dialog_action_area1"); gtk_widget_show (dialog_action_area1); gtk_button_box_set_layout (GTK_BUTTON_BOX (dialog_action_area1), GTK_BUTTONBOX_END); cancelbutton1 = gtk_button_new_from_stock ("gtk-cancel"); + gtk_widget_set_name (cancelbutton1, "cancelbutton1"); gtk_widget_show (cancelbutton1); gtk_dialog_add_action_widget (GTK_DIALOG (Conf), cancelbutton1, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (cancelbutton1, GTK_CAN_DEFAULT); okbutton1 = gtk_button_new_from_stock ("gtk-ok"); + gtk_widget_set_name (okbutton1, "okbutton1"); gtk_widget_show (okbutton1); gtk_dialog_add_action_widget (GTK_DIALOG (Conf), okbutton1, GTK_RESPONSE_OK); GTK_WIDGET_SET_FLAGS (okbutton1, GTK_CAN_DEFAULT); @@ -481,6 +563,12 @@ create_Conf (void) g_signal_connect ((gpointer) PAD2, "clicked", G_CALLBACK (OnConf_Pad2), NULL); + g_signal_connect ((gpointer) PAD3, "clicked", + G_CALLBACK (OnConf_Pad3), + NULL); + g_signal_connect ((gpointer) PAD4, "clicked", + G_CALLBACK (OnConf_Pad4), + NULL); g_signal_connect ((gpointer) joydevicescombo, "changed", G_CALLBACK (on_joydevicescombo_changed), NULL); @@ -578,6 +666,8 @@ create_Conf (void) GLADE_HOOKUP_OBJECT (Conf, hbox1, "hbox1"); GLADE_HOOKUP_OBJECT (Conf, PAD1, "PAD1"); GLADE_HOOKUP_OBJECT (Conf, PAD2, "PAD2"); + GLADE_HOOKUP_OBJECT (Conf, PAD3, "PAD3"); + GLADE_HOOKUP_OBJECT (Conf, PAD4, "PAD4"); GLADE_HOOKUP_OBJECT (Conf, label2, "label2"); GLADE_HOOKUP_OBJECT (Conf, vbox2, "vbox2"); GLADE_HOOKUP_OBJECT (Conf, label5, "label5"); diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 410c236344..2c5707deba 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -103,7 +103,7 @@ void SaveConfig() return; } - for (j=0; j<2; j++) { + for (j=0; j<2 * PADSUBKEYS; j++) { for (i=0; i= 0 && joyid < (int)s_vjoysticks.size()) { - pjoy = s_vjoysticks[joyid]; - if( SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key)) ) { - status[(pjoy)->GetPAD()] &= ~(1<GetPAD()] |= (1<= 0 && joyid < (int)s_vjoysticks.size()) { + pjoy = s_vjoysticks[joyid]; + if( SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key)) ) { + status[(pjoy)->GetPAD()] &= ~(1<GetPAD()] |= (1<= 0 && joyid < (int)s_vjoysticks.size()) { pjoy = s_vjoysticks[joyid]; @@ -411,10 +402,10 @@ static int s_selectedpad = 0; void UpdateConf(int pad) { - s_selectedpad = pad; + s_selectedpad = pad; int i; - GtkWidget *Btn; + GtkWidget *Btn; for (i=0; i + 5 True False ZeroPAD @@ -151,6 +152,48 @@ and SSSPSXPad, TwinPad authors False + + + + True + True + PAD1 alt. + True + GTK_RELIEF_NORMAL + True + False + False + True + PAD1 + + + + 0 + False + False + + + + + + True + True + PAD2 alt. + True + GTK_RELIEF_NORMAL + True + False + False + True + PAD1 + + + + 0 + False + False + + diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index ad9e47e1d8..d89aadce3b 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -54,7 +54,7 @@ void SaveConfig() if(!szTemp) return; strcpy(szTemp, "\\inis\\zeropad.ini"); - for (j=0; j<2; j++) { + for (j=0; j<2 * PADSUBKEYS; j++) { for (i=0; i Date: Sun, 3 May 2009 11:49:25 +0000 Subject: [PATCH 006/296] ZeroPad: Commit bolche's Analog control code, ifdeffed out, so I don't forget to add a gui for it and enable it some day. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1126 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 83 +++++++++++++++++++++++++++++++++ plugins/zeropad/zeropad.h | 19 ++++++++ 2 files changed, 102 insertions(+) diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 2c5707deba..d14aebbe60 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -278,6 +278,59 @@ void CALLBACK PADupdate(int pad) case KeyPress: key = XLookupKeysym((XKeyEvent *)&E, 0); +#ifdef ANALOG_CONTROLS_HACK + switch(key) { + case KEY_PAD_LX_LEFT: + g_lanalog[pad].x = DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTLX ) + g_lanalog[pad].x = -g_lanalog[pad].x; + g_lanalog[pad].x += 0x80; + break; + case KEY_PAD_LX_RIGHT: + g_lanalog[pad].x = -DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTLX ) + g_lanalog[pad].x = -g_lanalog[pad].x; + g_lanalog[pad].x += 0x80; + break; + case KEY_PAD_LY_UP: + g_lanalog[pad].y = DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTLY ) + g_lanalog[pad].y = -g_lanalog[pad].y; + g_lanalog[pad].y += 0x80; + break; + case KEY_PAD_LY_DOWN: + g_lanalog[pad].y = -DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTLY ) + g_lanalog[pad].y = -g_lanalog[pad].y; + g_lanalog[pad].y += 0x80; + break; + case KEY_PAD_RX_LEFT: + g_ranalog[pad].x = DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTRX ) + g_ranalog[pad].x = -g_ranalog[pad].x; + g_ranalog[pad].x += 0x80; + break; + case KEY_PAD_RX_RIGHT: + g_ranalog[pad].x = -DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTRX ) + g_ranalog[pad].x = -g_ranalog[pad].x; + g_ranalog[pad].x += 0x80; + break; + case KEY_PAD_RY_UP: + g_ranalog[pad].y = DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTRY ) + g_ranalog[pad].y = -g_ranalog[pad].y; + g_ranalog[pad].y += 0x80; + break; + case KEY_PAD_RY_DOWN: + g_ranalog[pad].y = -DEF_VALUE/256; + if( conf.options&PADOPTION_REVERTRY ) + g_ranalog[pad].y = -g_ranalog[pad].y; + g_ranalog[pad].y += 0x80; + break; + } +#endif + i = FindKey(key, pad); keyPress |= (1 << i); keyRelease &= ~(1 << i); @@ -287,6 +340,36 @@ void CALLBACK PADupdate(int pad) break; case KeyRelease: key = XLookupKeysym((XKeyEvent *)&E, 0); + +#ifdef ANALOG_CONTROLS_HACK + switch(key) { + case KEY_PAD_LX_LEFT: + g_lanalog[pad].x = 0x80; + break; + case KEY_PAD_LY_UP: + g_lanalog[pad].y = 0x80; + break; + case KEY_PAD_RX_LEFT: + g_ranalog[pad].x = 0x80; + break; + case KEY_PAD_RY_UP: + g_ranalog[pad].y = 0x80; + break; + case KEY_PAD_LX_RIGHT: + g_lanalog[pad].x = 0x80; + break; + case KEY_PAD_LY_DOWN: + g_lanalog[pad].y = 0x80; + break; + case KEY_PAD_RX_RIGHT: + g_ranalog[pad].x = 0x80; + break; + case KEY_PAD_RY_DOWN: + g_ranalog[pad].y = 0x80; + break; + } +#endif + i = FindKey(key, pad); keyPress &= ~(1< Date: Sun, 3 May 2009 12:19:58 +0000 Subject: [PATCH 007/296] Zeropad: Clean up the code so that if I have to go back into it, it's easier to read. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1127 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 1268 +++++++++++++++------------- plugins/zeropad/Windows/resource.h | 2 +- plugins/zeropad/Windows/win.cpp | 398 +++++---- plugins/zeropad/zeropad.cpp | 467 +++++----- plugins/zeropad/zeropad.h | 94 ++- 5 files changed, 1222 insertions(+), 1007 deletions(-) diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index d14aebbe60..67afc24340 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -1,5 +1,5 @@ /* ZeroPAD - author: zerofrog(@gmail.com) - * Copyright (C) 2006-2007 + * Copyright (C) 2006-2007 * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -27,7 +27,8 @@ #include "zeropad.h" -extern "C" { +extern "C" +{ #include "interface.h" #include "support.h" #include "callbacks.h" @@ -41,48 +42,87 @@ static u32 s_bSDLInit = false; // holds all joystick info class JoystickInfo { -public: - JoystickInfo(); - ~JoystickInfo() { Destroy(); } - - void Destroy(); - // opens handles to all possible joysticks - static void EnumerateJoysticks(vector& vjoysticks); - - bool Init(int id, bool bStartThread=true); // opens a handle and gets information - void Assign(int pad); // assigns a joystick to a pad + public: + JoystickInfo(); + ~JoystickInfo() + { + Destroy(); + } - void TestForce(); + void Destroy(); + // opens handles to all possible joysticks + static void EnumerateJoysticks(vector& vjoysticks); - const string& GetName() { return devname; } - int GetNumButtons() { return numbuttons; } - int GetNumAxes() { return numaxes; } - int GetNumPOV() { return numpov; } - int GetId() { return _id; } - int GetPAD() { return pad; } - int GetDeadzone(int axis) { return deadzone; } + bool Init(int id, bool bStartThread = true); // opens a handle and gets information + void Assign(int pad); // assigns a joystick to a pad - void SaveState(); - int GetButtonState(int i) { return vbutstate[i]; } - int GetAxisState(int i) { return vaxisstate[i]; } - void SetButtonState(int i, int state) { vbutstate[i] = state; } - void SetAxisState(int i, int value) { vaxisstate[i] = value; } + void TestForce(); + + const string& GetName() + { + return devname; + } + int GetNumButtons() + { + return numbuttons; + } + int GetNumAxes() + { + return numaxes; + } + int GetNumPOV() + { + return numpov; + } + int GetId() + { + return _id; + } + int GetPAD() + { + return pad; + } + int GetDeadzone(int axis) + { + return deadzone; + } + + void SaveState(); + int GetButtonState(int i) + { + return vbutstate[i]; + } + int GetAxisState(int i) + { + return vaxisstate[i]; + } + void SetButtonState(int i, int state) + { + vbutstate[i] = state; + } + void SetAxisState(int i, int value) + { + vaxisstate[i] = value; + } #ifdef JOYSTICK_SUPPORT - SDL_Joystick* GetJoy() { return joy; } + SDL_Joystick* GetJoy() + { + return joy; + } #endif -private: + private: + + string devname; // pretty device name + int _id; + int numbuttons, numaxes, numpov; + int axisrange, deadzone; + int pad; + + vector vbutstate, vaxisstate; - string devname; // pretty device name - int _id; - int numbuttons, numaxes, numpov; - int axisrange, deadzone; - int pad; - - vector vbutstate, vaxisstate; - #ifdef JOYSTICK_SUPPORT - SDL_Joystick* joy; + SDL_Joystick* joy; #endif }; @@ -92,41 +132,48 @@ extern string s_strIniPath; void SaveConfig() { - int i, j; - FILE *f; - char cfg[255]; + int i, j; + FILE *f; + char cfg[255]; - strcpy(cfg, s_strIniPath.c_str()); - f = fopen(cfg,"w"); - if (f == NULL) { - printf("ZeroPAD: failed to save ini %s\n", s_strIniPath.c_str()); - return; - } - - for (j=0; j<2 * PADSUBKEYS; j++) { - for (i=0; i::iterator it; - FORIT(it, s_vjoysticks) delete *it; - s_vjoysticks.clear(); + vector::iterator it; + FORIT(it, s_vjoysticks) delete *it; + s_vjoysticks.clear(); } void _PADupdate(int pad) { - pthread_spin_lock(&s_mutexStatus); - status[pad] |= s_keyRelease[pad]; - status[pad] &= ~s_keyPress[pad]; - s_keyRelease[pad] = 0; - s_keyPress[pad] = 0; - pthread_spin_unlock(&s_mutexStatus); + pthread_spin_lock(&s_mutexStatus); + status[pad] |= s_keyRelease[pad]; + status[pad] &= ~s_keyPress[pad]; + s_keyRelease[pad] = 0; + s_keyPress[pad] = 0; + pthread_spin_unlock(&s_mutexStatus); } int _GetJoystickIdFromPAD(int pad) { // select the right joystick id int joyid = -1; - for (int p = 0; p < PADSUBKEYS; p++) { - for(int i = 0; i < PADKEYS; ++i) { - if (IS_JOYSTICK(conf.keys[(PadEnum[pad][p])][i]) || IS_JOYBUTTONS(conf.keys[(PadEnum[pad][p])][i])) { - joyid = PAD_GETJOYID(conf.keys[(PadEnum[pad][p])][i]); + + for (int p = 0; p < PADSUBKEYS; p++) + { + for (int i = 0; i < PADKEYS; ++i) + { + if (IS_JOYSTICK(conf.keys[(PadEnum[pad][p])][i]) || IS_JOYBUTTONS(conf.keys[(PadEnum[pad][p])][i])) + { + joyid = PAD_GETJOYID(conf.keys[(PadEnum[pad][p])][i]); return joyid; - } + } } } } void CALLBACK PADupdate(int pad) { - int i; - XEvent E; - int keyPress=0,keyRelease=0; - KeySym key; + int i; + XEvent E; + int keyPress = 0, keyRelease = 0; + KeySym key; - // keyboard input - while (XPending(GSdsp) > 0) { + // keyboard input + while (XPending(GSdsp) > 0) + { XNextEvent(GSdsp, &E); - switch (E.type) { - case KeyPress: - key = XLookupKeysym((XKeyEvent *)&E, 0); + switch (E.type) + { + case KeyPress: + key = XLookupKeysym((XKeyEvent *) & E, 0); -#ifdef ANALOG_CONTROLS_HACK - switch(key) { - case KEY_PAD_LX_LEFT: - g_lanalog[pad].x = DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTLX ) - g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; - break; - case KEY_PAD_LX_RIGHT: - g_lanalog[pad].x = -DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTLX ) - g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; - break; - case KEY_PAD_LY_UP: - g_lanalog[pad].y = DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTLY ) - g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - break; - case KEY_PAD_LY_DOWN: - g_lanalog[pad].y = -DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTLY ) - g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - break; - case KEY_PAD_RX_LEFT: - g_ranalog[pad].x = DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTRX ) - g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - break; - case KEY_PAD_RX_RIGHT: - g_ranalog[pad].x = -DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTRX ) - g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - break; - case KEY_PAD_RY_UP: - g_ranalog[pad].y = DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTRY ) - g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; - break; - case KEY_PAD_RY_DOWN: - g_ranalog[pad].y = -DEF_VALUE/256; - if( conf.options&PADOPTION_REVERTRY ) - g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; - break; - } -#endif - - i = FindKey(key, pad); - keyPress |= (1 << i); - keyRelease &= ~(1 << i); - - event.evt = KEYPRESS; - event.key = key; - break; - case KeyRelease: - key = XLookupKeysym((XKeyEvent *)&E, 0); - #ifdef ANALOG_CONTROLS_HACK - switch(key) { - case KEY_PAD_LX_LEFT: - g_lanalog[pad].x = 0x80; - break; - case KEY_PAD_LY_UP: - g_lanalog[pad].y = 0x80; - break; - case KEY_PAD_RX_LEFT: - g_ranalog[pad].x = 0x80; - break; - case KEY_PAD_RY_UP: - g_ranalog[pad].y = 0x80; - break; - case KEY_PAD_LX_RIGHT: - g_lanalog[pad].x = 0x80; - break; - case KEY_PAD_LY_DOWN: - g_lanalog[pad].y = 0x80; - break; - case KEY_PAD_RX_RIGHT: - g_ranalog[pad].x = 0x80; - break; - case KEY_PAD_RY_DOWN: - g_ranalog[pad].y = 0x80; - break; - } + switch (key) + { + case KEY_PAD_LX_LEFT: + g_lanalog[pad].x = DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; + g_lanalog[pad].x += 0x80; + break; + case KEY_PAD_LX_RIGHT: + g_lanalog[pad].x = -DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; + g_lanalog[pad].x += 0x80; + break; + case KEY_PAD_LY_UP: + g_lanalog[pad].y = DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; + g_lanalog[pad].y += 0x80; + break; + case KEY_PAD_LY_DOWN: + g_lanalog[pad].y = -DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; + g_lanalog[pad].y += 0x80; + break; + case KEY_PAD_RX_LEFT: + g_ranalog[pad].x = DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; + g_ranalog[pad].x += 0x80; + break; + case KEY_PAD_RX_RIGHT: + g_ranalog[pad].x = -DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; + g_ranalog[pad].x += 0x80; + break; + case KEY_PAD_RY_UP: + g_ranalog[pad].y = DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; + g_ranalog[pad].y += 0x80; + break; + case KEY_PAD_RY_DOWN: + g_ranalog[pad].y = -DEF_VALUE / 256; + if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; + g_ranalog[pad].y += 0x80; + break; + } #endif - - i = FindKey(key, pad); - keyPress &= ~(1<= 0 && joyid < (int)s_vjoysticks.size()) { - pjoy = s_vjoysticks[joyid]; - if( SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key)) ) { - status[(pjoy)->GetPAD()] &= ~(1<GetPAD()] |= (1<= 0 && joyid < (int)s_vjoysticks.size()) + { + pjoy = s_vjoysticks[joyid]; + + if (SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key))) + status[(pjoy)->GetPAD()] &= ~(1 << i); // pressed + else + status[(pjoy)->GetPAD()] |= (1 << i); // pressed } - } - else if( IS_JOYSTICK(key) ) { - int joyid = PAD_GETJOYID(key); - if( joyid >= 0 && joyid < (int)s_vjoysticks.size()) { + } + else if (IS_JOYSTICK(key)) + { + int joyid = PAD_GETJOYID(key); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + { - pjoy = s_vjoysticks[joyid]; - int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - int pad = (pjoy)->GetPAD(); - switch(i) { - case PAD_LX: - if( abs(value) > (pjoy)->GetDeadzone(value) ) { - g_lanalog[pad].x = value/256; - if( conf.options&PADOPTION_REVERTLX ) - g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; - } - else g_lanalog[pad].x = 0x80; - break; - case PAD_LY: - if( abs(value) > (pjoy)->GetDeadzone(value) ) { - g_lanalog[pad].y = value/256; - if( conf.options&PADOPTION_REVERTLX ) - g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - } - else g_lanalog[pad].y = 0x80; - break; - case PAD_RX: - if( abs(value) > (pjoy)->GetDeadzone(value) ) { - g_ranalog[pad].x = value/256; - if( conf.options&PADOPTION_REVERTLX ) - g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - } - else g_ranalog[pad].x = 0x80; - break; - case PAD_RY: - if( abs(value) > (pjoy)->GetDeadzone(value) ) { - g_ranalog[pad].y = value/256; - if( conf.options&PADOPTION_REVERTLX ) - g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; - } - else g_ranalog[pad].y = 0x80; - break; - } - } - } - else if( IS_POV(key) ) { - int joyid = PAD_GETJOYID(key); - if( joyid >= 0 && joyid < (int)s_vjoysticks.size()) { + pjoy = s_vjoysticks[joyid]; + int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); + int pad = (pjoy)->GetPAD(); + switch (i) + { + case PAD_LX: + if (abs(value) > (pjoy)->GetDeadzone(value)) + { + g_lanalog[pad].x = value / 256; + if (conf.options&PADOPTION_REVERTLX) + g_lanalog[pad].x = -g_lanalog[pad].x; + g_lanalog[pad].x += 0x80; + } + else + { + g_lanalog[pad].x = 0x80; + } + break; + case PAD_LY: + if (abs(value) > (pjoy)->GetDeadzone(value)) + { + g_lanalog[pad].y = value / 256; + if (conf.options&PADOPTION_REVERTLX) + g_lanalog[pad].y = -g_lanalog[pad].y; + g_lanalog[pad].y += 0x80; + } + else + { + g_lanalog[pad].y = 0x80; + } + break; + case PAD_RX: + if (abs(value) > (pjoy)->GetDeadzone(value)) + { + g_ranalog[pad].x = value / 256; + if (conf.options&PADOPTION_REVERTLX) + g_ranalog[pad].x = -g_ranalog[pad].x; + g_ranalog[pad].x += 0x80; + } + else + { + g_ranalog[pad].x = 0x80; + } + break; + case PAD_RY: + if (abs(value) > (pjoy)->GetDeadzone(value)) + { + g_ranalog[pad].y = value / 256; + if (conf.options&PADOPTION_REVERTLX) g_ranalog[pad].y = -g_ranalog[pad].y; + g_ranalog[pad].y += 0x80; + } + else + { + g_ranalog[pad].y = 0x80; + } + break; + } + } + } + else if (IS_POV(key)) + { + int joyid = PAD_GETJOYID(key); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + { + pjoy = s_vjoysticks[joyid]; - pjoy = s_vjoysticks[joyid]; - - int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - int pad = (pjoy)->GetPAD(); - if( PAD_GETPOVSIGN(key) && (value<-2048) ) - status[pad] &= ~(1<2048) ) - status[pad] &= ~(1<GetJoy(), PAD_GETJOYSTICK_AXIS(key)); + int pad = (pjoy)->GetPAD(); + if (PAD_GETPOVSIGN(key) && (value < -2048)) + status[pad] &= ~(1 << i); + else if (!PAD_GETPOVSIGN(key) && (value > 2048)) + status[pad] &= ~(1 << i); + else + status[pad] |= (1 << i); + } + } + } #endif - - pthread_spin_lock(&s_mutexStatus); - s_keyPress[pad] |= keyPress; - s_keyPress[pad] &= ~keyRelease; - s_keyRelease[pad] |= keyRelease; - s_keyRelease[pad] &= ~keyPress; - pthread_spin_unlock(&s_mutexStatus); + + pthread_spin_lock(&s_mutexStatus); + s_keyPress[pad] |= keyPress; + s_keyPress[pad] &= ~keyRelease; + s_keyRelease[pad] |= keyRelease; + s_keyRelease[pad] &= ~keyPress; + pthread_spin_unlock(&s_mutexStatus); } -static GtkWidget *Conf=NULL, *s_devicecombo=NULL; +static GtkWidget *Conf = NULL, *s_devicecombo = NULL; static int s_selectedpad = 0; void UpdateConf(int pad) @@ -489,186 +565,204 @@ void UpdateConf(int pad) int i; GtkWidget *Btn; - for (i=0; i 0) { - gtk_entry_set_text(GTK_ENTRY(Btn), tmp.c_str()); - } + if (tmp.size() > 0) + { + gtk_entry_set_text(GTK_ENTRY(Btn), tmp.c_str()); + } else gtk_entry_set_text(GTK_ENTRY(Btn), "Unknown"); - gtk_object_set_user_data(GTK_OBJECT(Btn), (void*)(PADKEYS*pad+i)); + gtk_object_set_user_data(GTK_OBJECT(Btn), (void*)(PADKEYS*pad + i)); } - // check bounds - int joyid = _GetJoystickIdFromPAD(pad); - if( joyid < 0 || joyid >= (int)s_vjoysticks.size() ) { - // get first unused joystick - for(joyid = 0; joyid < s_vjoysticks.size(); ++joyid) { - if( s_vjoysticks[joyid]->GetPAD() < 0 ) - break; - } - } - - if( joyid >= 0 && joyid < (int)s_vjoysticks.size() ) { - // select the combo - gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), joyid); - } - else gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), s_vjoysticks.size()); // no gamepad + // check bounds + int joyid = _GetJoystickIdFromPAD(pad); + + if (joyid < 0 || joyid >= (int)s_vjoysticks.size()) + { + // get first unused joystick + for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) + { + if (s_vjoysticks[joyid]->GetPAD() < 0) break; + } + } - int padopts = conf.options>>(16*pad); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverselx")), padopts&PADOPTION_REVERTLX); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversely")), padopts&PADOPTION_REVERTLY); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverserx")), padopts&PADOPTION_REVERTRX); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversery")), padopts&PADOPTION_REVERTRY); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "forcefeedback")), padopts&PADOPTION_FORCEFEEDBACK); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), joyid); // select the combo + else + gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), s_vjoysticks.size()); // no gamepad + + int padopts = conf.options >> (16 * pad); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverselx")), padopts&PADOPTION_REVERTLX); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversely")), padopts&PADOPTION_REVERTLY); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverserx")), padopts&PADOPTION_REVERTRX); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversery")), padopts&PADOPTION_REVERTRY); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "forcefeedback")), padopts&PADOPTION_FORCEFEEDBACK); } void OnConf_Key(GtkButton *button, gpointer user_data) { - GdkEvent *ev; - GtkWidget* label = lookup_widget(Conf, GetLabelFromButton(gtk_button_get_label(button)).c_str()); - if( label == NULL ) { - printf("couldn't find correct label\n"); - return; - } - - int id = (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); - int pad = id/PADKEYS; - int key = id%PADKEYS; - unsigned long *pkey = &conf.keys[pad][key]; + GdkEvent *ev; + GtkWidget* label = lookup_widget(Conf, GetLabelFromButton(gtk_button_get_label(button)).c_str()); + if (label == NULL) + { + printf("couldn't find correct label\n"); + return; + } - vector::iterator itjoy; + int id = (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); + int pad = id / PADKEYS; + int key = id % PADKEYS; + unsigned long *pkey = &conf.keys[pad][key]; - // save the states + vector::iterator itjoy; + + // save the states #ifdef JOYSTICK_SUPPORT - SDL_JoystickUpdate(); - FORIT(itjoy, s_vjoysticks) (*itjoy)->SaveState(); + SDL_JoystickUpdate(); + FORIT(itjoy, s_vjoysticks)(*itjoy)->SaveState(); #endif - - for (;;) { - ev = gdk_event_get(); - if (ev != NULL) { - if (ev->type == GDK_KEY_PRESS) { - *pkey = ev->key.keyval; - char* tmp = XKeysymToString(*pkey); - if (tmp != NULL) + for (;;) + { + ev = gdk_event_get(); + if (ev != NULL) + { + if (ev->type == GDK_KEY_PRESS) + { + *pkey = ev->key.keyval; + + char* tmp = XKeysymToString(*pkey); + if (tmp != NULL) gtk_entry_set_text(GTK_ENTRY(label), tmp); - else + else gtk_entry_set_text(GTK_ENTRY(label), "Unknown"); return; } - } + } #ifdef JOYSTICK_SUPPORT - SDL_JoystickUpdate(); - FORIT(itjoy, s_vjoysticks) { + SDL_JoystickUpdate(); + FORIT(itjoy, s_vjoysticks) + { + // MAKE sure to look for changes in the state!! + for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) + { + int but = SDL_JoystickGetButton((*itjoy)->GetJoy(), i); + if (but != (*itjoy)->GetButtonState(i)) + { + if (!but) // released, we don't really want this + { + (*itjoy)->SetButtonState(i, 0); + break; + } - // MAKE sure to look for changes in the state!! - - for(int i = 0; i < (*itjoy)->GetNumButtons(); ++i) { - int but = SDL_JoystickGetButton((*itjoy)->GetJoy(), i); - if( but != (*itjoy)->GetButtonState(i) ) { + *pkey = PAD_JOYBUTTON((*itjoy)->GetId(), i); + char str[32]; + sprintf(str, "JBut %d", i); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + } - if( !but ) { // released, we don't really want this - (*itjoy)->SetButtonState(i, 0); - break; - } + for (int i = 0; i < (*itjoy)->GetNumAxes(); ++i) + { + int value = SDL_JoystickGetAxis((*itjoy)->GetJoy(), i); - *pkey = PAD_JOYBUTTON((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JBut %d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } - } + if (value != (*itjoy)->GetAxisState(i)) + { - for(int i = 0; i < (*itjoy)->GetNumAxes(); ++i) { - int value = SDL_JoystickGetAxis((*itjoy)->GetJoy(), i); + if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this + { + // released, we don't really want this + (*itjoy)->SetButtonState(i, value); + break; + } - if( value != (*itjoy)->GetAxisState(i) ) { - if( abs(value) <= (*itjoy)->GetAxisState(i)) {// we don't want this - // released, we don't really want this - (*itjoy)->SetButtonState(i, value); - break; - } - - - if( abs(value) > 0x3fff ) { - if( key < 16 ) { // POV - *pkey = PAD_POV((*itjoy)->GetId(), value<0, i); - char str[32]; - sprintf(str, "JPOV %d%s", i, value<0?"-":"+"); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } - else { // axis - *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JAxis %d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } - } - } - } - } + if (abs(value) > 0x3fff) + { + if (key < 16) // POV + { + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + char str[32]; + sprintf(str, "JPOV %d%s", i, value < 0 ? "-" : "+"); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + else // axis + { + *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); + char str[32]; + sprintf(str, "JAxis %d", i); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + } + } + } + } #endif - } + } } void OnConf_Pad1(GtkButton *button, gpointer user_data) { - if( gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button)) ) - UpdateConf(0); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(0); } void OnConf_Pad2(GtkButton *button, gpointer user_data) { - if( gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button)) ) - UpdateConf(1); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(1); } void OnConf_Pad3(GtkButton *button, gpointer user_data) { - if( gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button)) ) - UpdateConf(2); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(2); } void OnConf_Pad4(GtkButton *button, gpointer user_data) { - if( gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button)) ) - UpdateConf(3); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(3); } void OnConf_Ok(GtkButton *button, gpointer user_data) @@ -684,7 +778,7 @@ void OnConf_Cancel(GtkButton *button, gpointer user_data) { gtk_widget_destroy(Conf); gtk_main_quit(); - LoadConfig(); // load previous config + LoadConfig(); // load previous config } void CALLBACK PADconfigure() @@ -693,86 +787,99 @@ void CALLBACK PADconfigure() getcwd(strcurdir, 256); s_strIniPath = strcurdir; s_strIniPath += "/inis/zeropad.ini"; - + LoadConfig(); Conf = create_Conf(); - // recreate - JoystickInfo::EnumerateJoysticks(s_vjoysticks); + // recreate + JoystickInfo::EnumerateJoysticks(s_vjoysticks); - s_devicecombo = lookup_widget(Conf, "joydevicescombo"); + s_devicecombo = lookup_widget(Conf, "joydevicescombo"); - // fill the combo - char str[255]; - vector::iterator it; - FORIT(it, s_vjoysticks) { - sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), - (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()); - gtk_combo_box_append_text (GTK_COMBO_BOX (s_devicecombo), str); - } - gtk_combo_box_append_text (GTK_COMBO_BOX (s_devicecombo), "No Gamepad"); + // fill the combo + char str[255]; + vector::iterator it; + + FORIT(it, s_vjoysticks) + { + sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), + (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()); + gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), str); + } + + gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), "No Gamepad"); UpdateConf(0); - gtk_widget_show_all(Conf); + gtk_widget_show_all(Conf); gtk_main(); } // GUI event handlers void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) { - int joyid = gtk_combo_box_get_active(combobox); + int joyid = gtk_combo_box_get_active(combobox); - // unassign every joystick with this pad - for(int i = 0; i < (int)s_vjoysticks.size(); ++i) { - if( s_vjoysticks[i]->GetPAD() == s_selectedpad ) - s_vjoysticks[i]->Assign(-1); - } - - if( joyid >= 0 && joyid < (int)s_vjoysticks.size() ) - s_vjoysticks[joyid]->Assign(s_selectedpad); + // unassign every joystick with this pad + for (int i = 0; i < (int)s_vjoysticks.size(); ++i) + { + if (s_vjoysticks[i]->GetPAD() == s_selectedpad) s_vjoysticks[i]->Assign(-1); + } + + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->Assign(s_selectedpad); } void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) { - int mask = PADOPTION_REVERTLX<<(16*s_selectedpad); - if( gtk_toggle_button_get_active(togglebutton) ) conf.options |= mask; - else conf.options &= ~mask; + int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; } void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) { - int mask = PADOPTION_REVERTLY<<(16*s_selectedpad); - if( gtk_toggle_button_get_active(togglebutton) ) conf.options |= mask; - else conf.options &= ~mask; + int mask = PADOPTION_REVERTLY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; } void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer user_data) { - int mask = PADOPTION_REVERTRX<<(16*s_selectedpad); - if( gtk_toggle_button_get_active(togglebutton) ) conf.options |= mask; - else conf.options &= ~mask; + int mask = PADOPTION_REVERTRX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; } void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) { - int mask = PADOPTION_REVERTRY<<(16*s_selectedpad); - if( gtk_toggle_button_get_active(togglebutton) ) conf.options |= mask; - else conf.options &= ~mask; + int mask = PADOPTION_REVERTRY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; } void on_forcefeedback_toggled(GtkToggleButton *togglebutton, gpointer user_data) { - int mask = PADOPTION_REVERTLX<<(16*s_selectedpad); - if( gtk_toggle_button_get_active(togglebutton) ) { - conf.options |= mask; + int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + { + conf.options |= mask; - int joyid = gtk_combo_box_get_active(GTK_COMBO_BOX(s_devicecombo)); - if( joyid >= 0 && joyid < (int)s_vjoysticks.size() ) - s_vjoysticks[joyid]->TestForce(); - } - else conf.options &= ~mask; + int joyid = gtk_combo_box_get_active(GTK_COMBO_BOX(s_devicecombo)); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->TestForce(); + } + else + { + conf.options &= ~mask; + } } GtkWidget *About = NULL; @@ -783,7 +890,8 @@ void OnAbout_Ok(GtkButton *button, gpointer user_data) gtk_main_quit(); } -void CALLBACK PADabout() { +void CALLBACK PADabout() +{ About = create_About(); @@ -791,7 +899,8 @@ void CALLBACK PADabout() { gtk_main(); } -s32 CALLBACK PADtest() { +s32 CALLBACK PADtest() +{ return 0; } @@ -804,36 +913,40 @@ void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) { #ifdef JOYSTICK_SUPPORT - if( !s_bSDLInit ) { - if( SDL_Init(SDL_INIT_JOYSTICK) < 0 ) - return; - SDL_JoystickEventState(SDL_QUERY); - s_bSDLInit = true; - } + if (!s_bSDLInit) + { + if (SDL_Init(SDL_INIT_JOYSTICK) < 0) return; + SDL_JoystickEventState(SDL_QUERY); + s_bSDLInit = true; + } - vector::iterator it; - FORIT(it, vjoysticks) delete *it; + vector::iterator it; + FORIT(it, vjoysticks) delete *it; - vjoysticks.resize(SDL_NumJoysticks()); - for(int i = 0; i < (int)vjoysticks.size(); ++i) { - vjoysticks[i] = new JoystickInfo(); - vjoysticks[i]->Init(i, true); - } + vjoysticks.resize(SDL_NumJoysticks()); + for (int i = 0; i < (int)vjoysticks.size(); ++i) + { + vjoysticks[i] = new JoystickInfo(); + vjoysticks[i]->Init(i, true); + } - // set the pads - for(int pad = 0; pad < 2; ++pad) { - // select the right joystick id - int joyid = -1; - for(int i = 0; i < PADKEYS; ++i) { - if( IS_JOYSTICK(conf.keys[pad][i]) || IS_JOYBUTTONS(conf.keys[pad][i]) ) { - joyid = PAD_GETJOYID(conf.keys[pad][i]); - break; - } - } + // set the pads + for (int pad = 0; pad < 2; ++pad) + { + // select the right joystick id + int joyid = -1; + for (int i = 0; i < PADKEYS; ++i) + { + if (IS_JOYSTICK(conf.keys[pad][i]) || IS_JOYBUTTONS(conf.keys[pad][i])) + { + joyid = PAD_GETJOYID(conf.keys[pad][i]); + break; + } + } - if( joyid >= 0 && joyid < (int)s_vjoysticks.size() ) - s_vjoysticks[joyid]->Assign(pad); - } + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + s_vjoysticks[joyid]->Assign(pad); + } #endif } @@ -841,77 +954,80 @@ void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) JoystickInfo::JoystickInfo() { #ifdef JOYSTICK_SUPPORT - joy = NULL; + joy = NULL; #endif - _id = -1; - pad = -1; - axisrange = 0x7fff; - deadzone = 2000; + _id = -1; + pad = -1; + axisrange = 0x7fff; + deadzone = 2000; } void JoystickInfo::Destroy() { #ifdef JOYSTICK_SUPPORT - if( joy != NULL ) { - if( SDL_JoystickOpened(_id) ) - SDL_JoystickClose(joy); - joy = NULL; - } + if (joy != NULL) + { + if (SDL_JoystickOpened(_id)) SDL_JoystickClose(joy); + joy = NULL; + } #endif } bool JoystickInfo::Init(int id, bool bStartThread) { #ifdef JOYSTICK_SUPPORT - Destroy(); - _id = id; + Destroy(); + _id = id; - joy = SDL_JoystickOpen(id); - if( joy == NULL ) { - printf("failed to open joystick %d\n", id); - return false; - } + joy = SDL_JoystickOpen(id); + if (joy == NULL) + { + printf("failed to open joystick %d\n", id); + return false; + } - numaxes = SDL_JoystickNumAxes(joy); - numbuttons = SDL_JoystickNumButtons(joy); - numpov = SDL_JoystickNumHats(joy); - devname = SDL_JoystickName(id); - vbutstate.resize(numbuttons); - vaxisstate.resize(numbuttons); - - return true; + numaxes = SDL_JoystickNumAxes(joy); + numbuttons = SDL_JoystickNumButtons(joy); + numpov = SDL_JoystickNumHats(joy); + devname = SDL_JoystickName(id); + vbutstate.resize(numbuttons); + vaxisstate.resize(numbuttons); + + return true; #else - return false; + return false; #endif } // assigns a joystick to a pad void JoystickInfo::Assign(int newpad) { - if( pad == newpad ) - return; + if (pad == newpad) return; + pad = newpad; - pad = newpad; - - if( pad >= 0 ) { - for(int i = 0; i < PADKEYS; ++i) { - if( IS_JOYBUTTONS(conf.keys[pad][i]) ) { - conf.keys[pad][i] = PAD_JOYBUTTON(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); - } - else if( IS_JOYSTICK(conf.keys[pad][i]) ) { - conf.keys[pad][i] = PAD_JOYSTICK(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); - } - } - } + if (pad >= 0) + { + for (int i = 0; i < PADKEYS; ++i) + { + if (IS_JOYBUTTONS(conf.keys[pad][i])) + { + conf.keys[pad][i] = PAD_JOYBUTTON(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); + } + else if (IS_JOYSTICK(conf.keys[pad][i])) + { + conf.keys[pad][i] = PAD_JOYSTICK(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); + } + } + } } void JoystickInfo::SaveState() { #ifdef JOYSTICK_SUPPORT - for(int i = 0; i < numbuttons; ++i) - vbutstate[i] = SDL_JoystickGetButton(joy, i); - for(int i = 0; i < numaxes; ++i) - vaxisstate[i] = SDL_JoystickGetAxis(joy, i); + for (int i = 0; i < numbuttons; ++i) + vbutstate[i] = SDL_JoystickGetButton(joy, i); + for (int i = 0; i < numaxes; ++i) + vaxisstate[i] = SDL_JoystickGetAxis(joy, i); #endif } diff --git a/plugins/zeropad/Windows/resource.h b/plugins/zeropad/Windows/resource.h index 5490aad8ee..029b8d33c1 100644 --- a/plugins/zeropad/Windows/resource.h +++ b/plugins/zeropad/Windows/resource.h @@ -59,7 +59,7 @@ #define IDC_SR2 1063 // Next default values for new objects -// +// #ifdef APSTUDIO_INVOKED #ifndef APSTUDIO_READONLY_SYMBOLS #define _APS_NEXT_RESOURCE_VALUE 114 diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index d89aadce3b..132ac8abb8 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -1,5 +1,5 @@ /* ZeroPAD - author: zerofrog(@gmail.com) - * Copyright (C) 2006-2007 + * Copyright (C) 2006-2007 * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -29,100 +29,106 @@ using namespace std; -HINSTANCE hInst=NULL; +HINSTANCE hInst = NULL; static pthread_spinlock_t s_mutexStatus; static u32 s_keyPress[2], s_keyRelease[2]; extern u16 status[2]; extern string s_strIniPath; LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); -WNDPROC GSwndProc=NULL; -HWND GShwnd=NULL; +WNDPROC GSwndProc = NULL; +HWND GShwnd = NULL; extern keyEvent event; void SaveConfig() { - char *szTemp; - char szIniFile[256], szValue[256], szProf[256]; - int i, j; - - GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); - szTemp = strrchr(szIniFile, '\\'); - - if(!szTemp) return; - strcpy(szTemp, "\\inis\\zeropad.ini"); - - for (j=0; j<2 * PADSUBKEYS; j++) { - for (i=0; i=0x60 && key<=0x69) { + strcpy(buff, "NONE"); + else + { + if (key >= 0x60 && key <= 0x69) sprintf(buff, "NumPad %c", '0' + key - 0x60); - } - else sprintf(buff, "%c", key); - } - } + else + sprintf(buff, "%c", key); + } + } else if (key >= 0x1000 && key < 0x2000) { - sprintf (buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); + sprintf(buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); } else if (key >= 0x2000 && key < 0x3000) { static const char name[][4] = { "MIN", "MAX" }; const int axis = (key & 0xff); - sprintf (buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); - if (index >= 17 && index <= 20) - buff[strlen (buff) -4] = '\0'; + + sprintf(buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); + if (index >= 17 && index <= 20) buff[strlen(buff) -4] = '\0'; } else if (key >= 0x3000 && key < 0x4000) { static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; const int pov = (key & 0xff); - sprintf (buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov /4, name[pov % 4]); + sprintf(buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov / 4, name[pov % 4]); } - return buff; + return buff; } -BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { +BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) +{ HWND hWC; - TCITEM tcI; - int i,key, numkeys; - u8* pkeyboard; - static int disabled=0; - static int padn=0; + TCITEM tcI; + int i, key, numkeys; + u8* pkeyboard; + static int disabled = 0; + static int padn = 0; - switch(uMsg) { + switch (uMsg) + { case WM_INITDIALOG: LoadConfig(); - padn = 0; - if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); + padn = 0; + if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); - for (i=0; i= 0) && (angle< 4500)) return 1;//forward - if ((direction==2) && (angle>= 4500) && (angle<13500)) return 1;//right - if ((direction==1) && (angle>=13500) && (angle<22500)) return 1;//backward - if ((direction==3) && (angle>=22500) && (angle<31500)) return 1;//left - if ((direction==0) && (angle>=31500) && (angle<36000)) return 1;//forward +int POV(u32 direction, u32 angle) +{ + if ((direction == 0) && (angle >= 0) && (angle < 4500)) return 1;//forward + if ((direction == 2) && (angle >= 4500) && (angle < 13500)) return 1;//right + if ((direction == 1) && (angle >= 13500) && (angle < 22500)) return 1;//backward + if ((direction == 3) && (angle >= 22500) && (angle < 31500)) return 1;//left + if ((direction == 0) && (angle >= 31500) && (angle < 36000)) return 1;//forward return 0; } @@ -111,12 +136,15 @@ void _KeyPress(int pad, u32 key) { int i; - for (int p=0; p > 8; stdpar[curPad][3] = status[curPad] & 0xff; stdpar[curPad][4] = g_ranalog[curPad].x; stdpar[curPad][5] = g_ranalog[curPad].y; stdpar[curPad][6] = g_lanalog[curPad].x; stdpar[curPad][7] = g_lanalog[curPad].y; - if (padMode[curPad] == 1) cmdLen = 20; - else cmdLen = 4; + + if (padMode[curPad] == 1) + cmdLen = 20; + else + cmdLen = 4; + button_check2 = stdpar[curPad][2] >> 4; - switch(stdpar[curPad][3]) + switch (stdpar[curPad][3]) { - case 0xBF: // X - stdpar[curPad][14] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][16])); - break; - case 0xDF: // Circle - stdpar[curPad][13] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][17])); - break; - case 0xEF: // Triangle - stdpar[curPad][12] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][19])); - break; - case 0x7F: // Square - stdpar[curPad][15] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][18])); - break; - case 0xFB: // L1 - stdpar[curPad][16] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][26])); - break; - case 0xF7: // R1 - stdpar[curPad][17] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][28])); - break; - case 0xFE: // L2 - stdpar[curPad][18] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][27])); - break; - case 0xFD: // R2 - stdpar[curPad][19] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][29])); - break; - default: - stdpar[curPad][14] = 0x00; // Not pressed - stdpar[curPad][13] = 0x00; // Not pressed - stdpar[curPad][12] = 0x00; // Not pressed - stdpar[curPad][15] = 0x00; // Not pressed - stdpar[curPad][16] = 0x00; // Not pressed - stdpar[curPad][17] = 0x00; // Not pressed - stdpar[curPad][18] = 0x00; // Not pressed - stdpar[curPad][19] = 0x00; // Not pressed - break; + case 0xBF: // X + stdpar[curPad][14] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][16])); + break; + case 0xDF: // Circle + stdpar[curPad][13] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][17])); + break; + case 0xEF: // Triangle + stdpar[curPad][12] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][19])); + break; + case 0x7F: // Square + stdpar[curPad][15] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][18])); + break; + case 0xFB: // L1 + stdpar[curPad][16] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][26])); + break; + case 0xF7: // R1 + stdpar[curPad][17] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][28])); + break; + case 0xFE: // L2 + stdpar[curPad][18] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][27])); + break; + case 0xFD: // R2 + stdpar[curPad][19] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][29])); + break; + default: + stdpar[curPad][14] = 0x00; // Not pressed + stdpar[curPad][13] = 0x00; // Not pressed + stdpar[curPad][12] = 0x00; // Not pressed + stdpar[curPad][15] = 0x00; // Not pressed + stdpar[curPad][16] = 0x00; // Not pressed + stdpar[curPad][17] = 0x00; // Not pressed + stdpar[curPad][18] = 0x00; // Not pressed + stdpar[curPad][19] = 0x00; // Not pressed + break; } - switch(button_check2) + switch (button_check2) { - case 0xE: // UP - stdpar[curPad][10] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][21])); - break; - case 0xB: // DOWN - stdpar[curPad][11] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][22])); - break; - case 0x7: // LEFT - stdpar[curPad][9] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][23])); - break; - case 0xD: // RIGHT - stdpar[curPad][8] = (pressure*255)/100; //0xff/(100/(100-conf.keys[curPad][24])); - break; - default: - stdpar[curPad][8] = 0x00; // Not pressed - stdpar[curPad][9] = 0x00; // Not pressed - stdpar[curPad][10] = 0x00; // Not pressed - stdpar[curPad][11] = 0x00; // Not pressed - break; + case 0xE: // UP + stdpar[curPad][10] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][21])); + break; + case 0xB: // DOWN + stdpar[curPad][11] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][22])); + break; + case 0x7: // LEFT + stdpar[curPad][9] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][23])); + break; + case 0xD: // RIGHT + stdpar[curPad][8] = (pressure * 255) / 100; //0xff/(100/(100-conf.keys[curPad][24])); + break; + default: + stdpar[curPad][8] = 0x00; // Not pressed + stdpar[curPad][9] = 0x00; // Not pressed + stdpar[curPad][10] = 0x00; // Not pressed + stdpar[curPad][11] = 0x00; // Not pressed + break; } buf = stdpar[curPad]; return padID[curPad]; @@ -435,7 +488,7 @@ u8 _PADpoll(u8 value) { buf = unk4c[curPad]; return 0xf3; - case 0x4d: + case 0x4d: cmdLen = 8; buf = unk4d[curPad]; return 0xf3; @@ -455,49 +508,56 @@ u8 _PADpoll(u8 value) { } } - switch (curCmd) { - case 0x43: - if(curByte == 2) + switch (curCmd) + { + case 0x43: + if (curByte == 2) { - switch(value){ - case 0: - buf[2] = 0; - buf[3] = 0; - break; - case 1: - buf[2] = 0xff; - buf[3] = 0xff; - break; + switch (value) + { + case 0: + buf[2] = 0; + buf[3] = 0; + break; + case 1: + buf[2] = 0xff; + buf[3] = 0xff; + break; } } break; case 0x44: - if (curByte == 2) { + if (curByte == 2) + { PADsetMode(curPad, value); } break; case 0x46: - if(curByte == 2) { - switch(value) { - case 0: // default - buf[5] = 0x2; - buf[6] = 0x0; - buf[7] = 0xA; - break; - case 1: // Param std conf change - buf[5] = 0x1; - buf[6] = 0x1; - buf[7] = 0x14; - break; + if (curByte == 2) + { + switch (value) + { + case 0: // default + buf[5] = 0x2; + buf[6] = 0x0; + buf[7] = 0xA; + break; + case 1: // Param std conf change + buf[5] = 0x1; + buf[6] = 0x1; + buf[7] = 0x14; + break; } } break; case 0x4c: - if (curByte == 2) { - switch (value) { + if (curByte == 2) + { + switch (value) + { case 0: // mode 0 - digital mode buf[5] = 0x4; break; @@ -517,10 +577,9 @@ u8 _PADpoll(u8 value) { u8 CALLBACK PADpoll(u8 value) { u8 ret; + ret = _PADpoll(value); -#ifdef PAD_LOG PAD_LOG("PADpoll: %x (%d: %x)\n", value, curByte, ret); -#endif return ret; } @@ -528,7 +587,7 @@ u8 CALLBACK PADpoll(u8 value) static keyEvent s_event; keyEvent* CALLBACK PADkeyEvent() { - s_event = event; - event.evt = 0; - return &s_event; + s_event = event; + event.evt = 0; + return &s_event; } diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index c43efb1794..eb7ac4e9ba 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -40,7 +40,8 @@ using namespace std; #define PADdefs -extern "C" { +extern "C" +{ #include "PS2Edefs.h" } @@ -48,20 +49,20 @@ extern char libraryName[256]; #define FORIT(it, v) for(it = (v).begin(); it != (v).end(); (it)++) -#define IS_KEYBOARD(key) (key<0x10000) -#define IS_JOYBUTTONS(key) (key>=0x10000 && key<0x20000) // buttons -#define IS_JOYSTICK(key) (key>=0x20000&&key<0x30000) // analog -#define IS_POV(key) (key>=0x30000&&key<0x40000) // uses analog as buttons (cares about sign) -#define IS_MOUSE(key) (key>=0x40000&&key<0x50000) // mouse +#define IS_KEYBOARD(key) (key < 0x10000) +#define IS_JOYBUTTONS(key) (key >= 0x10000 && key < 0x20000) // buttons +#define IS_JOYSTICK(key) (key >= 0x20000 && key < 0x30000) // analog +#define IS_POV(key) (key >= 0x30000 && key < 0x40000) // uses analog as buttons (cares about sign) +#define IS_MOUSE(key) (key >= 0x40000 && key < 0x50000) // mouse -#define PAD_GETKEY(key) ((key)&0xffff) -#define PAD_GETJOYID(key) (((key)&0xf000)>>12) -#define PAD_GETJOYBUTTON(key) ((key)&0xff) -#define PAD_GETJOYSTICK_AXIS(key) ((key)&0xff) -#define PAD_JOYBUTTON(joyid, buttonid) (0x10000|((joyid)<<12)|(buttonid)) -#define PAD_JOYSTICK(joyid, axisid) (0x20000|((joyid)<<12)|(axisid)) -#define PAD_POV(joyid, sign, axisid) (0x30000|((joyid)<<12)|((sign)<<8)|(axisid)) -#define PAD_GETPOVSIGN(key) (((key)&0x100)>>8) +#define PAD_GETKEY(key) ((key) & 0xffff) +#define PAD_GETJOYID(key) (((key) & 0xf000) >> 12) +#define PAD_GETJOYBUTTON(key) ((key) & 0xff) +#define PAD_GETJOYSTICK_AXIS(key) ((key) & 0xff) +#define PAD_JOYBUTTON(joyid, buttonid) (0x10000 | ((joyid) << 12) | (buttonid)) +#define PAD_JOYSTICK(joyid, axisid) (0x20000 | ((joyid) << 12) | (axisid)) +#define PAD_POV(joyid, sign, axisid) (0x30000 | ((joyid) << 12) | ((sign) << 8) | (axisid)) +#define PAD_GETPOVSIGN(key) (((key) & 0x100) >> 8) #define PADKEYS 20 @@ -74,14 +75,16 @@ extern char libraryName[256]; #define PADSUBKEYS 2 extern int PadEnum[2][2]; -typedef struct { +typedef struct +{ unsigned long keys[2 * PADSUBKEYS][PADKEYS]; int log; - int options; // upper 16 bits are for pad2 + int options; // upper 16 bits are for pad2 } PADconf; -typedef struct { - u8 x,y; +typedef struct +{ + u8 x, y; } PADAnalog; extern PADconf conf; @@ -90,28 +93,31 @@ extern PADAnalog g_lanalog[2], g_ranalog[2]; extern FILE *padLog; #define PAD_LOG __Log -#define PAD_RY 19 -#define PAD_LY 18 -#define PAD_RX 17 -#define PAD_LX 16 -#define PAD_LEFT 15 -#define PAD_DOWN 14 -#define PAD_RIGHT 13 -#define PAD_UP 12 -#define PAD_START 11 -#define PAD_R3 10 -#define PAD_L3 9 -#define PAD_SELECT 8 -#define PAD_SQUARE 7 -#define PAD_CROSS 6 -#define PAD_CIRCLE 5 -#define PAD_TRIANGLE 4 -#define PAD_R1 3 -#define PAD_L1 2 -#define PAD_R2 1 -#define PAD_L2 0 +enum gamePadValues +{ + PAD_RY = 19, + PAD_LY = 18, + PAD_RX = 17, + PAD_LX = 16, + PAD_LEFT = 15, + PAD_DOWN = 14, + PAD_RIGHT = 13, + PAD_UP = 12, + PAD_START = 11, + PAD_R3 = 10, + PAD_L3 = 9, + PAD_SELECT = 8, + PAD_SQUARE = 7, + PAD_CROSS = 6, + PAD_CIRCLE = 5, + PAD_TRIANGLE = 4, + PAD_R1 = 3, + PAD_L1 = 2, + PAD_R2 = 1, + PAD_L2 = 0 +}; -// Put in the code for bolche's analog contols hack, ifdeffed out, so I don't forget to +// Put in the code for bolche's analog contols hack, ifdeffed out, so I don't forget to // add a gui some day and activate it. //#define ANALOG_CONTROLS_HACK // The various KEY_PAD_xxx definitions are defined as the value of whatever key is pressed. @@ -151,11 +157,11 @@ void SaveConfig(); void SysMessage(char *fmt, ...); -inline int FindKey(int key, int pad) { - for (int p=0; p < PADSUBKEYS; p++) - for (int i=0; i Date: Sun, 3 May 2009 19:30:38 +0000 Subject: [PATCH 008/296] ZeroPad: Patch to fix a bug introduced in r1125. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1128 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 14 +++++++++----- plugins/zeropad/Windows/win.cpp | 16 ++++++++++------ plugins/zeropad/zeropad.h | 1 + 3 files changed, 20 insertions(+), 11 deletions(-) diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 67afc24340..6cea5ae5fe 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -317,6 +317,7 @@ int _GetJoystickIdFromPAD(int pad) } } } + return joyid; } void CALLBACK PADupdate(int pad) @@ -382,9 +383,10 @@ void CALLBACK PADupdate(int pad) #endif i = FindKey(key, pad); - keyPress |= (1 << i); - keyRelease &= ~(1 << i); - + if (i != -1) { + keyPress |= (1 << i); + keyRelease &= ~(1 << i); + } event.evt = KEYPRESS; event.key = key; break; @@ -423,8 +425,10 @@ void CALLBACK PADupdate(int pad) #endif i = FindKey(key, pad); - keyPress &= ~(1 << i); - keyRelease |= (1 << i); + if (i != -1) { + keyPress &= ~(1 << i); + keyRelease |= (1 << i); + } event.evt = KEYRELEASE; event.key = key; break; diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index 132ac8abb8..34a4eefc18 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -190,17 +190,21 @@ LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) if (lParam & 0x40000000) return TRUE; - i = FindKey(wparam, pad); - keyPress[pad] |= (1 << i); - keyRelease[pad] &= ~(1 << i); + i = FindKey(wParam, pad); + if (i != -1) { + keyPress[pad] |= (1 << i); + keyRelease[pad] &= ~(1 << i); + } event.evt = KEYPRESS; event.key = wParam; break; case WM_KEYUP: - i = FindKey(wparam, pad); - keyPress[pad] &= ~(1 << i); - keyRelease[pad] |= (1 << i); + i = FindKey(wParam, pad); + if (i != -1) { + keyPress[pad] &= ~(1 << i); + keyRelease[pad] |= (1 << i); + } event.evt = KEYRELEASE; event.key = wParam; break; diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index eb7ac4e9ba..43a422faff 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -162,6 +162,7 @@ inline int FindKey(int key, int pad) for (int p = 0; p < PADSUBKEYS; p++) for (int i = 0; i < PADKEYS; i++) if (key == conf.keys[(PadEnum[pad][p])][i]) return i; + return -1; } #endif From 66a8c03ded3bbe1a11c07beb9e4266534024ee2c Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Mon, 4 May 2009 02:06:35 +0000 Subject: [PATCH 009/296] Fix bug from r1116 [essentially a misplaced jNO_DEFAULT directive]. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1129 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/ps2/Iop/IopHwRead.cpp | 12 ------------ pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 8 ++++---- 2 files changed, 4 insertions(+), 16 deletions(-) diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index 9263b71e02..7ae0214f2b 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -399,10 +399,6 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) { switch( masked_addr ) { - mcase(HW_SIO2_FIFO): // 0x1f801264 wouldn't make it here, so I assume you mean 01f808264, which is fifo, Jake... - ret = psxHu32(addr); - break; - mcase(HW_SIO2_CTRL): ret = sio2_getCtrl(); break; @@ -431,15 +427,7 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) ret = sio2_getIntr(); break; - mcase(0x1f80847c): // I'm adding this to try to counter reports of pcsx2 crashing when this address is called. - Console::Error("iopHwRead32_Page8: Calling 0x1f80847c for some unknown reason."); - ret = psxHu32(addr); - break; - - // Do we really want to assert here, if it's an unknown address? - //jNO_DEFAULT; default: - Console::Error("iopHwRead32_Page8: Why are we calling 0x%x?", params addr); ret = psxHu32(addr); break; } diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 16743d3655..8b979de20d 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -1584,10 +1584,6 @@ - - @@ -2140,6 +2136,10 @@ + + From b02d6a19ddb5a9e210a70a90deb3cda568f7e343 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 4 May 2009 10:31:18 +0000 Subject: [PATCH 010/296] ZeroPad: Revert the changes to the Windows files in ZeroPad, and ifdef off the changes in the common areas, as the Windows version of ZeroPad is clearly only working by voodoo and bailing wire on Windows. (Note: unable to reproduce ZeroPad working properly in any revision [including playground days] on any version of Windows available. Likely because I don't have any poultry handy to sacrifice.) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1130 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Windows/win.cpp | 680 +++++++++++++++----------------- plugins/zeropad/zeropad.cpp | 19 +- plugins/zeropad/zeropad.h | 5 + 3 files changed, 352 insertions(+), 352 deletions(-) diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index 34a4eefc18..40e058721a 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -29,150 +29,142 @@ using namespace std; -HINSTANCE hInst = NULL; +HINSTANCE hInst=NULL; static pthread_spinlock_t s_mutexStatus; static u32 s_keyPress[2], s_keyRelease[2]; extern u16 status[2]; extern string s_strIniPath; LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); -WNDPROC GSwndProc = NULL; -HWND GShwnd = NULL; +WNDPROC GSwndProc=NULL; +HWND GShwnd=NULL; extern keyEvent event; void SaveConfig() { - char *szTemp; - char szIniFile[256], szValue[256], szProf[256]; - int i, j; + char *szTemp; + char szIniFile[256], szValue[256], szProf[256]; + int i, j; - GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); - szTemp = strrchr(szIniFile, '\\'); + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); - if (!szTemp) return; - strcpy(szTemp, "\\inis\\zeropad.ini"); + if(!szTemp) return; + strcpy(szTemp, "\\inis\\zeropad.ini"); - for (j = 0; j < 2 * PADSUBKEYS; j++) - { - for (i = 0; i < PADKEYS; i++) - { - sprintf(szProf, "%d_%d", j, i); - sprintf(szValue, "%d", conf.keys[j][i]); - WritePrivateProfileString("Interface", szProf, szValue, szIniFile); - } - } + for (j=0; j<2; j++) { + for (i=0; i= 0x60 && key <= 0x69) - sprintf(buff, "NumPad %c", '0' + key - 0x60); - else - sprintf(buff, "%c", key); - } - } - else if (key >= 0x1000 && key < 0x2000) - { - sprintf(buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); - } - else if (key >= 0x2000 && key < 0x3000) - { - static const char name[][4] = { "MIN", "MAX" }; - const int axis = (key & 0xff); - - sprintf(buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); - if (index >= 17 && index <= 20) buff[strlen(buff) -4] = '\0'; - } - else if (key >= 0x3000 && key < 0x4000) - { - static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; - const int pov = (key & 0xff); - sprintf(buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov / 4, name[pov % 4]); - } + const int key = conf.keys[pad][index]; + char buff[16]="NONE)"; + if (key < 0x100) + { + if (key == 0) + strcpy (buff, "NONE"); + else { + if(key>=0x60 && key<=0x69) { + sprintf(buff, "NumPad %c", '0' + key - 0x60); + } + else sprintf(buff, "%c", key); + } + } + else if (key >= 0x1000 && key < 0x2000) + { + sprintf (buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); + } + else if (key >= 0x2000 && key < 0x3000) + { + static const char name[][4] = { "MIN", "MAX" }; + const int axis = (key & 0xff); + sprintf (buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); + if (index >= 17 && index <= 20) + buff[strlen (buff) -4] = '\0'; + } + else if (key >= 0x3000 && key < 0x4000) + { + static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; + const int pov = (key & 0xff); + sprintf (buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov /4, name[pov % 4]); + } - return buff; + return buff; } -BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) -{ - HWND hWC; - TCITEM tcI; - int i, key, numkeys; - u8* pkeyboard; - static int disabled = 0; - static int padn = 0; +BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { + HWND hWC; + TCITEM tcI; + int i,key, numkeys; + u8* pkeyboard; + static int disabled=0; + static int padn=0; - switch (uMsg) - { - case WM_INITDIALOG: - LoadConfig(); - padn = 0; - if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); + switch(uMsg) { + case WM_INITDIALOG: + LoadConfig(); + padn = 0; + if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); - for (i = 0; i < PADKEYS; i++) - { - hWC = GetDlgItem(hW, IDC_L2 + i * 2); - Button_SetText(hWC, GetKeyLabel(padn, i).c_str()); - } + for (i=0; i Date: Mon, 4 May 2009 14:12:21 +0000 Subject: [PATCH 011/296] Improved handling of jNO_DEFAULT so that it performs a sanity check in Devel builds [currently applies to Pcsx2 only since we don't have a standard define macro for devbuild in the plugins]. Cleaned up and abbreviated logging for the new IOP HwRead/Write handlers. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1131 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Defs.h | 41 ++++++++++++++------- pcsx2/PrecompiledHeader.h | 25 +++++++++---- pcsx2/ps2/Iop/IopHwRead.cpp | 71 ++++++++++++++++-------------------- pcsx2/ps2/Iop/IopHwWrite.cpp | 53 ++++++++++----------------- 4 files changed, 96 insertions(+), 94 deletions(-) diff --git a/common/include/Pcsx2Defs.h b/common/include/Pcsx2Defs.h index b06ed47925..85c83d4a66 100644 --- a/common/include/Pcsx2Defs.h +++ b/common/include/Pcsx2Defs.h @@ -30,6 +30,8 @@ #include "Pcsx2Types.h" // Renamed ARRAYSIZE to ArraySize -- looks nice and gets rid of Windows.h conflicts (air) +// Notes: I'd have used ARRAY_SIZE instead but ran into cross-platform lib conflicts with +// that as well. >_< #ifndef ArraySize #define ArraySize(x) (sizeof(x)/sizeof((x)[0])) #endif @@ -39,29 +41,42 @@ // This is primarily useful for the default case switch optimizer, which enables VC to // generate more compact switches. -#ifdef NDEBUG -# define jBREAKPOINT() ((void) 0) -# ifdef _MSC_VER -# define jASSUME(exp) (__assume(exp)) +#ifndef jASSUME +# ifdef NDEBUG +# define jBREAKPOINT() ((void) 0) +# ifdef _MSC_VER +# define jASSUME(exp) (__assume(exp)) +# else +# define jASSUME(exp) ((void) sizeof(exp)) +# endif # else -# define jASSUME(exp) ((void) sizeof(exp)) +# if defined(_MSC_VER) +# define jBREAKPOINT() do { __asm int 3 } while(0) +# else +# define jBREAKPOINT() ((void) *(volatile char *) 0) +# endif +# define jASSUME(exp) if(exp) ; else jBREAKPOINT() # endif -#else -# if defined(_MSC_VER) -# define jBREAKPOINT() do { __asm int 3 } while(0) -# else -# define jBREAKPOINT() ((void) *(volatile char *) 0) -# endif -# define jASSUME(exp) if(exp) ; else jBREAKPOINT() #endif -// disable the default case in a switch +////////////////////////////////////////////////////////////////////////////////////////// +// jNO_DEFAULT -- disables the default case in a switch, which improves switch optimization +// under MSVC. +// +// How it Works: jASSUME turns into an __assume(0) under msvc compilers, which when specified +// in the default: case of a switch tells the compiler that the case is unreachable, and so +// the compiler will not generate any code, LUTs, or conditionals to handle it. In debug +// builds the default case will cause an assertion (meaning the jNO_DEFAULT has been used +// incorrectly, and that the default case is in fact used and needs to be handled). +// +#ifndef jNO_DEFAULT #define jNO_DEFAULT \ { \ default: \ jASSUME(0); \ break; \ } +#endif /* common defines */ #ifndef C_ASSERT diff --git a/pcsx2/PrecompiledHeader.h b/pcsx2/PrecompiledHeader.h index 3156cd4058..f9bbbe979a 100644 --- a/pcsx2/PrecompiledHeader.h +++ b/pcsx2/PrecompiledHeader.h @@ -12,6 +12,7 @@ #ifdef PCSX2_DEVBUILD # define _SECURE_SCL 1 # define _SECURE_SCL_THROWS 1 +# pragma warning(disable:4244) // disable warning C4244: '=' : conversion from 'big' to 'small', possible loss of data #else # define _SECURE_SCL 0 #endif @@ -19,15 +20,25 @@ #define NOMINMAX // Disables other libs inclusion of their own min/max macros (we use std instead) -#if defined (__linux__) // some distributions are lower case -# define __LINUX__ +#ifndef _WIN32 +# include // Non-Windows platforms need this #endif -#ifdef _WIN32 -// disable warning C4244: '=' : conversion from 'big' to 'small', possible loss of data -# pragma warning(disable:4244) -#else -# include // Non-Windows platforms need this +////////////////////////////////////////////////////////////////////////////////////////// +// Custom version of jNO_DEFAULT macro for devel builds. +// Raises an exception if the default case is reached. This notifies us that a jNO_DEFAULT +// directive has been used incorrectly. +// +// MSVC Note: To stacktrace LogicError exceptions, add Exception::LogicError to the C++ First- +// Chance Exception list (under Debug->Exceptions menu). +// +#ifdef PCSX2_DEVBUILD +#define jNO_DEFAULT \ +{ \ +default: \ + throw Exception::LogicError( "Incorrect usage of jNO_DEFAULT detected (default case is not unreachable!)" ); \ + break; \ +} #endif ////////////////////////////////////////////////////////////////////////////////////////// diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index 7ae0214f2b..f9ad47252a 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -50,28 +50,28 @@ u8 __fastcall iopHwRead8_Page1( u32 addr ) default: if( masked_addr >= 0x100 && masked_addr < 0x130 ) { - DevCon::Notice( "*Hardware Read8 from Counter16 [ignored] [addr=0x%02x]", params addr, psxHu8(addr) ); + DevCon::Notice( "HwRead8 from Counter16 [ignored], addr 0x%08x = 0x%02x", params addr, psxHu8(addr) ); ret = psxHu8( addr ); } else if( masked_addr >= 0x480 && masked_addr < 0x4a0 ) { - DevCon::Notice( "*Hardware Read8 from Counter32 [ignored] [addr=0x%02x]", params addr, psxHu8(addr) ); + DevCon::Notice( "HwRead8 from Counter32 [ignored], addr 0x%08x = 0x%02x", params addr, psxHu8(addr) ); ret = psxHu8( addr ); } else if( (masked_addr >= pgmsk(HW_USB_START)) && (masked_addr < pgmsk(HW_USB_END)) ) { ret = USBread8( addr ); - PSXHW_LOG( "Hardware Read8 from USB: addr 0x%08x = 0x%02x", addr, ret ); + PSXHW_LOG( "HwRead8 from USB, addr 0x%08x = 0x%02x", addr, ret ); } else { ret = psxHu8(addr); - PSXHW_LOG( "*Unknown Hardware Read8 from addr 0x%08x = 0x%02x", addr, ret ); + PSXHW_LOG( "HwRead8 from Unknown, addr 0x%08x = 0x%02x", addr, ret ); } return ret; } - PSXHW_LOG( "*Hardware Read8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); return ret; } @@ -88,7 +88,7 @@ u8 __fastcall iopHwRead8_Page3( u32 addr ) else ret = psxHu8( addr ); - PSXHW_LOG( "Hardware Read8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); return ret; } @@ -106,7 +106,7 @@ u8 __fastcall iopHwRead8_Page8( u32 addr ) else ret = psxHu8( addr ); - PSXHW_LOG( "Hardware Read8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); return ret; } @@ -215,7 +215,7 @@ static __forceinline T _HwRead_16or32_Page1( u32 addr ) ret = SPU2read( addr ); else { - DevCon::Notice( "*PCSX2* SPU2 Hardware Read32 (addr=0x%08X)? What manner of trickery is this?!", params addr ); + DevCon::Notice( "HwRead32 from SPU2? (addr=0x%08X) .. What manner of trickery is this?!", params addr ); ret = psxHu32(addr); } } @@ -311,7 +311,7 @@ static __forceinline T _HwRead_16or32_Page1( u32 addr ) } } - PSXHW_LOG( "Hardware Read%s from %s, addr 0x%08x = 0x%04x", + PSXHW_LOG( "HwRead%s from %s, addr 0x%08x = 0x%04x", (sizeof(T) == 2) ? "16" : "32", _log_GetIopHwName( addr ), addr, ret ); return ret; @@ -336,7 +336,7 @@ u16 __fastcall iopHwRead16_Page3( u32 addr ) jASSUME( (addr >> 12) == 0x1f803 ); u16 ret = psxHu16(addr); - PSXHW_LOG( "Hardware Read16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); return ret; } @@ -348,7 +348,7 @@ u16 __fastcall iopHwRead16_Page8( u32 addr ) jASSUME( (addr >> 12) == 0x1f808 ); u16 ret = psxHu16(addr); - PSXHW_LOG( "Hardware Read16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); return ret; } @@ -366,7 +366,7 @@ u32 __fastcall iopHwRead32_Page3( u32 addr ) // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); const u32 ret = psxHu32(addr); - PSXHW_LOG( "Hardware Read32 from %s, addr 0x%08x = 0x%08x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%08x", _log_GetIopHwName( addr ), addr, ret ); return ret; } @@ -399,34 +399,25 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) { switch( masked_addr ) { - mcase(HW_SIO2_CTRL): - ret = sio2_getCtrl(); - break; - - mcase(HW_SIO2_RECV1): - ret = sio2_getRecv1(); - break; - - mcase(HW_SIO2_RECV2): - ret = sio2_getRecv2(); - break; - - mcase(HW_SIO2_RECV3): - ret = sio2_getRecv3(); - break; - - mcase(0x1f808278): - ret = sio2_get8278(); - break; - - mcase(0x1f80827C): - ret = sio2_get827C(); - break; - - mcase(HW_SIO2_INTR): - ret = sio2_getIntr(); - break; + mcase(HW_SIO2_CTRL): ret = sio2_getCtrl(); break; + mcase(HW_SIO2_RECV1): ret = sio2_getRecv1(); break; + mcase(HW_SIO2_RECV2): ret = sio2_getRecv2(); break; + mcase(HW_SIO2_RECV3): ret = sio2_getRecv3(); break; + mcase(0x1f808278): ret = sio2_get8278(); break; + mcase(0x1f80827C): ret = sio2_get827C(); break; + mcase(HW_SIO2_INTR): ret = sio2_getIntr(); break; + // HW_SIO2_FIFO -- A yet unknown: Should this be ignored on 32 bit writes, or handled as a + // 4-byte FIFO input? + // The old IOP system just ignored it, so that's what we do here. I've included commented code + // for treating it as a 16/32 bit write though [which si what the SIO does, for example). + mcase(HW_SIO2_FIFO): + //ret = sio2_fifoOut(); + //ret |= sio2_fifoOut() << 8; + //ret |= sio2_fifoOut() << 16; + //ret |= sio2_fifoOut() << 24; + //break; + default: ret = psxHu32(addr); break; @@ -435,7 +426,7 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) } else ret = psxHu32(addr); - PSXHW_LOG( "Hardware Read32 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); return ret; } diff --git a/pcsx2/ps2/Iop/IopHwWrite.cpp b/pcsx2/ps2/Iop/IopHwWrite.cpp index 04c5e61ce0..c538ba52b4 100644 --- a/pcsx2/ps2/Iop/IopHwWrite.cpp +++ b/pcsx2/ps2/Iop/IopHwWrite.cpp @@ -49,12 +49,12 @@ void __fastcall iopHwWrite8_Page1( u32 addr, u8 val ) default: if( masked_addr >= 0x100 && masked_addr < 0x130 ) { - DevCon::Notice( "*Hardware Write8 to Counter16 [ignored] [addr=0x%02x]", params addr, psxHu8(addr) ); + DevCon::Notice( "HwWrite8 to Counter16 [ignored], addr 0x%08x = 0x%02x", params addr, psxHu8(addr) ); psxHu8( addr ) = val; } else if( masked_addr >= 0x480 && masked_addr < 0x4a0 ) { - DevCon::Notice( "*Hardware Write8 to Counter32 [ignored] [addr=0x%02x]", params addr, psxHu8(addr) ); + DevCon::Notice( "HwWrite8 to Counter32 [ignored], addr 0x%08x = 0x%02x", params addr, psxHu8(addr) ); psxHu8( addr ) = val; } else if( masked_addr >= pgmsk(HW_USB_START) && masked_addr < pgmsk(HW_USB_END) ) @@ -68,7 +68,7 @@ void __fastcall iopHwWrite8_Page1( u32 addr, u8 val ) break; } - PSXHW_LOG( "*Hardware Write8 to %s, addr 0x%08x = 0x%02x\n", _log_GetIopHwName(addr), addr, val ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x\n", _log_GetIopHwName(addr), addr, val ); } static char g_pbuf[1024]; @@ -98,7 +98,7 @@ void __fastcall iopHwWrite8_Page3( u32 addr, u8 val ) } } - PSXHW_LOG( "Hardware Write8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); psxHu8( addr ) = val; } @@ -112,7 +112,7 @@ void __fastcall iopHwWrite8_Page8( u32 addr, u8 val ) else psxHu8( addr ) = val; - PSXHW_LOG( "Hardware Write8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); } @@ -210,7 +210,7 @@ static __forceinline void _HwWrite_16or32_Page1( u32 addr, T val ) SPU2write( addr, val ); else { - DevCon::Notice( "*PCSX2* SPU2 Hardware Write32 (addr=0x%08X)? What manner of trickery is this?!", params addr ); + DevCon::Notice( "HwWrite32 to SPU2? (addr=0x%08X) .. What manner of trickery is this?!", params addr ); //psxHu(addr) = val; } } @@ -423,7 +423,7 @@ static __forceinline void _HwWrite_16or32_Page1( u32 addr, T val ) } } - PSXHW_LOG( "Hardware Write%s to %s, addr 0x%08x = 0x%04x", + PSXHW_LOG( "HwWrite%s to %s, addr 0x%08x = 0x%04x", sizeof(T) == 2 ? "16" : "32", _log_GetIopHwName( addr ), addr, val ); } @@ -441,7 +441,7 @@ void __fastcall iopHwWrite16_Page3( u32 addr, u16 val ) // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); psxHu16(addr) = val; - PSXHW_LOG( "Hardware Write16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } void __fastcall iopHwWrite16_Page8( u32 addr, u16 val ) @@ -449,7 +449,7 @@ void __fastcall iopHwWrite16_Page8( u32 addr, u16 val ) // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); psxHu16(addr) = val; - PSXHW_LOG( "Hardware Write16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } ////////////////////////////////////////////////////////////////////////////////////////// @@ -464,7 +464,7 @@ void __fastcall iopHwWrite32_Page3( u32 addr, u32 val ) // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); psxHu16(addr) = val; - PSXHW_LOG( "Hardware Write32 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } void __fastcall iopHwWrite32_Page8( u32 addr, u32 val ) @@ -493,36 +493,21 @@ void __fastcall iopHwWrite32_Page8( u32 addr, u32 val ) { switch( masked_addr ) { - case 0x264: // unknown / reserved. - case 0x26C: // recv1 [read-only] - case 0x270: // recv2 [read-only] - case 0x274: // recv3 [read-only] + mcase(HW_SIO2_CTRL): sio2_setCtrl( val ); break; + mcase(0x1f808278): sio2_set8278( val ); break; + mcase(0x1f80827C): sio2_set827C( val ); break; + mcase(HW_SIO2_INTR): sio2_setIntr( val ); break; + + // Other SIO2 registers are read-only, no-ops on write. + default: psxHu32(addr) = val; - break; - - case 0x268: - sio2_setCtrl( val ); - break; - - case 0x278: - sio2_set8278( val ); - break; - - case 0x27C: - sio2_set827C( val ); - break; - - case 0x280: - sio2_setIntr( val ); - break; - - jNO_DEFAULT; + break; } } } else psxHu32(addr) = val; - PSXHW_LOG( "Hardware Write32 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, val ); } From db41e52940ebfe056f29be2c0673fd80c0e502a6 Mon Sep 17 00:00:00 2001 From: sudonim1 Date: Mon, 4 May 2009 18:21:44 +0000 Subject: [PATCH 012/296] Apparently the MSB of integer register specifications in VU micro instructions is ignored, making the upper 16 registers mirror the lower 16. Changed all relevant instructions to reflect this. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1132 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/DebugTools/DisR5900.cpp | 4 +- pcsx2/DebugTools/DisVU0Micro.cpp | 3 + pcsx2/DebugTools/DisVU1Micro.cpp | 3 + pcsx2/DebugTools/DisVUops.h | 78 ++--- pcsx2/VUops.cpp | 265 ++++++++-------- pcsx2/x86/iVUmicro.cpp | 3 + pcsx2/x86/iVUmicroLower.cpp | 519 ++++++++++++++++--------------- pcsx2/x86/iVUzerorec.cpp | 169 +++++----- 8 files changed, 531 insertions(+), 513 deletions(-) diff --git a/pcsx2/DebugTools/DisR5900.cpp b/pcsx2/DebugTools/DisR5900.cpp index 5607247451..7c17ff7c8a 100644 --- a/pcsx2/DebugTools/DisR5900.cpp +++ b/pcsx2/DebugTools/DisR5900.cpp @@ -470,8 +470,8 @@ MakeDisF(disCVTSw, dName("CVTSw"); dCP132(_Fd_); dCP132(_Fs_);) * Format: OP rt, offset(base) * *********************************************************/ -MakeDisF(disLWC1, dName("LWC1"); dCP132(_Rt_); dOffset();) -MakeDisF(disSWC1, dName("SWC1"); dCP132(_Rt_); dOffset();) +MakeDisF(disLWC1, dName("LWC1"); dCP132(_Ft_); dOffset();) +MakeDisF(disSWC1, dName("SWC1"); dCP132(_Ft_); dOffset();) /********************************************************* * Conditional Move * diff --git a/pcsx2/DebugTools/DisVU0Micro.cpp b/pcsx2/DebugTools/DisVU0Micro.cpp index b1c994e2ac..62c42b97e1 100644 --- a/pcsx2/DebugTools/DisVU0Micro.cpp +++ b/pcsx2/DebugTools/DisVU0Micro.cpp @@ -40,6 +40,9 @@ typedef char* (*TdisR5900F)DisFInterface; #define _Ft_ ((code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ ((code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ ((code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) #define dName(i) sprintf(ostr, "%s %-7s,", ostr, i) #define dNameU(i) { char op[256]; sprintf(op, "%s.%s%s%s%s", i, _X ? "x" : "", _Y ? "y" : "", _Z ? "z" : "", _W ? "w" : ""); sprintf(ostr, "%s %-7s,", ostr, op); } diff --git a/pcsx2/DebugTools/DisVU1Micro.cpp b/pcsx2/DebugTools/DisVU1Micro.cpp index de50040b8c..02eb0fa9ee 100644 --- a/pcsx2/DebugTools/DisVU1Micro.cpp +++ b/pcsx2/DebugTools/DisVU1Micro.cpp @@ -42,6 +42,9 @@ typedef char* (*TdisR5900F)DisFInterface; #define _Ft_ ((code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ ((code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ ((code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) #define dName(i) sprintf(ostr, "%s %-12s", ostr, i); \ diff --git a/pcsx2/DebugTools/DisVUops.h b/pcsx2/DebugTools/DisVUops.h index 6ca88bcfb5..7316168242 100644 --- a/pcsx2/DebugTools/DisVUops.h +++ b/pcsx2/DebugTools/DisVUops.h @@ -26,55 +26,55 @@ MakeDisF(dis##VU##MI_DIV, dName("DIV"); dCP232f(_Fs_, _Fsf_); dCP232f(_Ft_, _Ftf_);) \ MakeDisF(dis##VU##MI_SQRT, dName("SQRT"); dCP232f(_Ft_, _Ftf_);) \ MakeDisF(dis##VU##MI_RSQRT, dName("RSQRT"); dCP232f(_Fs_, _Fsf_); dCP232f(_Ft_, _Ftf_);) \ -MakeDisF(dis##VU##MI_IADDI, dName("IADDI"); dCP232i(_Ft_); dCP232i(_Fs_); dImm5();) \ -MakeDisF(dis##VU##MI_IADDIU, dName("IADDIU"); dCP232i(_Ft_); dCP232i(_Fs_); dImm15();) \ -MakeDisF(dis##VU##MI_IADD, dName("IADD"); dCP232i(_Fd_); dCP232i(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_IAND, dName("IAND"); dCP232i(_Fd_); dCP232i(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_IOR, dName("IOR"); dCP232i(_Fd_); dCP232i(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_ISUB, dName("ISUB"); dCP232i(_Fd_); dCP232i(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_ISUBIU, dName("ISUBIU"); dCP232i(_Ft_); dCP232i(_Fs_); dImm15();) \ +MakeDisF(dis##VU##MI_IADDI, dName("IADDI"); dCP232i(_It_); dCP232i(_Is_); dImm5();) \ +MakeDisF(dis##VU##MI_IADDIU, dName("IADDIU"); dCP232i(_It_); dCP232i(_Is_); dImm15();) \ +MakeDisF(dis##VU##MI_IADD, dName("IADD"); dCP232i(_Id_); dCP232i(_Is_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_IAND, dName("IAND"); dCP232i(_Id_); dCP232i(_Is_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_IOR, dName("IOR"); dCP232i(_Id_); dCP232i(_Is_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_ISUB, dName("ISUB"); dCP232i(_Id_); dCP232i(_Is_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_ISUBIU, dName("ISUBIU"); dCP232i(_It_); dCP232i(_Is_); dImm15();) \ MakeDisF(dis##VU##MI_MOVE, if (_Fs_ == 0 && _Ft_ == 0) { dNameU("NOP"); } else { dNameU("MOVE"); dCP2128f(_Ft_); dCP2128f(_Fs_); }) \ -MakeDisF(dis##VU##MI_MFIR, dNameU("MFIR"); dCP2128f(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_MTIR, dNameU("MTIR"); dCP232i(_Ft_); dCP232f(_Fs_, _Fsf_);) \ +MakeDisF(dis##VU##MI_MFIR, dNameU("MFIR"); dCP2128f(_Ft_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_MTIR, dNameU("MTIR"); dCP232i(_It_); dCP232f(_Fs_, _Fsf_);) \ MakeDisF(dis##VU##MI_MR32, dNameU("MR32"); dCP2128f(_Ft_); dCP2128f(_Fs_);) \ -MakeDisF(dis##VU##MI_LQ, dNameU("LQ"); dCP2128f(_Ft_); dCP232i(_Fs_); dImm11();) \ -MakeDisF(dis##VU##MI_LQD, dNameU("LQD"); dCP2128f(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_LQI, dNameU("LQI"); dCP2128f(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_SQ, dNameU("SQ"); dCP2128f(_Fs_); dCP232i(_Ft_); dImm11(); ) \ -MakeDisF(dis##VU##MI_SQD, dNameU("SQD"); dCP2128f(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_SQI, dNameU("SQI"); dCP2128f(_Fs_); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_ILW, dNameU("ILW"); dCP232i(_Ft_); dImm11(); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_ISW, dNameU("ISW"); dCP232i(_Ft_); dImm11(); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_ILWR, dNameU("ILWR"); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_ISWR, dNameU("ISWR"); dCP232i(_Ft_); dCP232i(_Fs_);) \ +MakeDisF(dis##VU##MI_LQ, dNameU("LQ"); dCP2128f(_Ft_); dCP232i(_Is_); dImm11();) \ +MakeDisF(dis##VU##MI_LQD, dNameU("LQD"); dCP2128f(_Ft_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_LQI, dNameU("LQI"); dCP2128f(_Ft_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_SQ, dNameU("SQ"); dCP2128f(_Fs_); dCP232i(_It_); dImm11(); ) \ +MakeDisF(dis##VU##MI_SQD, dNameU("SQD"); dCP2128f(_Fs_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_SQI, dNameU("SQI"); dCP2128f(_Fs_); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_ILW, dNameU("ILW"); dCP232i(_It_); dImm11(); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_ISW, dNameU("ISW"); dCP232i(_It_); dImm11(); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_ILWR, dNameU("ILWR"); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_ISWR, dNameU("ISWR"); dCP232i(_It_); dCP232i(_Is_);) \ MakeDisF(dis##VU##MI_LOI, dName("LOI"); ) \ MakeDisF(dis##VU##MI_RINIT, dNameU("RINIT"); dCP232i(REG_R); dCP232f(_Fs_, _Fsf_);) \ MakeDisF(dis##VU##MI_RGET, dNameU("RGET"); dCP232i(REG_R); dCP2128f(_Ft_);) \ MakeDisF(dis##VU##MI_RNEXT, dNameU("RNEXT"); dCP232i(REG_R); dCP2128f(_Ft_);) \ MakeDisF(dis##VU##MI_RXOR, dNameU("RXOR"); dCP232i(REG_R); dCP232f(_Fs_, _Fsf_);) \ MakeDisF(dis##VU##MI_WAITQ, dName("WAITQ"); ) \ -MakeDisF(dis##VU##MI_FSAND, dName("FSAND"); dCP232i(_Ft_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff); ) \ -MakeDisF(dis##VU##MI_FSEQ, dName("FSEQ"); dCP232i(_Ft_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff);) \ -MakeDisF(dis##VU##MI_FSOR, dName("FSOR"); dCP232i(_Ft_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff);) \ +MakeDisF(dis##VU##MI_FSAND, dName("FSAND"); dCP232i(_It_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff); ) \ +MakeDisF(dis##VU##MI_FSEQ, dName("FSEQ"); dCP232i(_It_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff);) \ +MakeDisF(dis##VU##MI_FSOR, dName("FSOR"); dCP232i(_It_); dCP232i(REG_STATUS_FLAG); sprintf(ostr, "%s %.3x,", ostr, code&0xfff);) \ MakeDisF(dis##VU##MI_FSSET, dName("FSSET"); dCP232i(REG_STATUS_FLAG);) \ -MakeDisF(dis##VU##MI_FMAND, dName("FMAND"); dCP232i(_Ft_); dCP232i(REG_MAC_FLAG); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_FMEQ, dName("FMEQ"); dCP232i(_Ft_); dCP232i(REG_MAC_FLAG); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_FMOR, dName("FMOR"); dCP232i(_Ft_); dCP232i(REG_MAC_FLAG); dCP232i(_Fs_);) \ +MakeDisF(dis##VU##MI_FMAND, dName("FMAND"); dCP232i(_It_); dCP232i(REG_MAC_FLAG); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_FMEQ, dName("FMEQ"); dCP232i(_It_); dCP232i(REG_MAC_FLAG); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_FMOR, dName("FMOR"); dCP232i(_It_); dCP232i(REG_MAC_FLAG); dCP232i(_Is_);) \ MakeDisF(dis##VU##MI_FCAND, dName("FCAND"); dCP232i(1); sprintf(ostr, "%s %8.8x,", ostr, code&0xffffff); ) \ MakeDisF(dis##VU##MI_FCEQ, dName("FCEQ"); dCP232i(1); dCP232i(REG_CLIP_FLAG);) \ MakeDisF(dis##VU##MI_FCOR, dName("FCOR"); dCP232i(1); dCP232i(REG_CLIP_FLAG);) \ MakeDisF(dis##VU##MI_FCSET, dName("FCSET"); dCP232i(REG_CLIP_FLAG); sprintf(ostr, "%s %.6x,", ostr, code&0xffffff); ) \ -MakeDisF(dis##VU##MI_FCGET, dName("FCGET"); dCP232i(_Ft_); dCP232i(REG_CLIP_FLAG);) \ -MakeDisF(dis##VU##MI_IBEQ, dName("IBEQ"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_IBGEZ, dName("IBEZ"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_IBGTZ, dName("IBGTZ"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_IBLEZ, dName("IBLEZ"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_IBLTZ, dName("IBLTZ"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_IBNE, dName("IBNE"); dImm11(); dCP232i(_Ft_); dCP232i(_Fs_);) \ +MakeDisF(dis##VU##MI_FCGET, dName("FCGET"); dCP232i(_It_); dCP232i(REG_CLIP_FLAG);) \ +MakeDisF(dis##VU##MI_IBEQ, dName("IBEQ"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_IBGEZ, dName("IBEZ"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_IBGTZ, dName("IBGTZ"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_IBLEZ, dName("IBLEZ"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_IBLTZ, dName("IBLTZ"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_IBNE, dName("IBNE"); dImm11(); dCP232i(_It_); dCP232i(_Is_);) \ MakeDisF(dis##VU##MI_B, dName("B"); dImm11();) \ -MakeDisF(dis##VU##MI_BAL, dName("BAL"); dImm11(); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_JR, dName("JR"); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_JALR, dName("JALR"); dCP232i(_Ft_); dCP232i(_Fs_); ) \ +MakeDisF(dis##VU##MI_BAL, dName("BAL"); dImm11(); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_JR, dName("JR"); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_JALR, dName("JALR"); dCP232i(_It_); dCP232i(_Is_); ) \ MakeDisF(dis##VU##MI_MFP, dNameU("MFP"); dCP2128f(_Ft_); dCP232i(REG_P);) \ MakeDisF(dis##VU##MI_WAITP, dName("WAITP"); ) \ MakeDisF(dis##VU##MI_ESADD, dName("ESADD"); dCP2128f(_Fs_);) \ @@ -83,16 +83,16 @@ MakeDisF(dis##VU##MI_ELENG, dName("ELENG"); dCP2128f(_Fs_); ) \ MakeDisF(dis##VU##MI_ERLENG, dName("ERLENG"); dCP2128f(_Fs_); ) \ MakeDisF(dis##VU##MI_EATANxy, dName("EATANxy"); dCP2128f(_Fs_);) \ MakeDisF(dis##VU##MI_EATANxz, dName("EATANxz"); dCP2128f(_Fs_);) \ -MakeDisF(dis##VU##MI_ESUM, dName("ESUM"); dCP232i(_Fs_); ) \ +MakeDisF(dis##VU##MI_ESUM, dName("ESUM"); dCP2128f(_Fs_); ) \ MakeDisF(dis##VU##MI_ERCPR, dName("ERCPR"); dCP232f(_Fs_, _Fsf_); ) \ MakeDisF(dis##VU##MI_ESQRT, dName("ESQRT"); dCP232f(_Fs_, _Fsf_); ) \ MakeDisF(dis##VU##MI_ERSQRT, dName("ERSQRT"); dCP232f(_Fs_, _Fsf_); ) \ MakeDisF(dis##VU##MI_ESIN, dName("ESIN"); dCP232f(_Fs_, _Fsf_); ) \ MakeDisF(dis##VU##MI_EATAN, dName("EATAN"); dCP232f(_Fs_, _Fsf_); ) \ MakeDisF(dis##VU##MI_EEXP, dName("EEXP"); dCP232f(_Fs_, _Fsf_); ) \ -MakeDisF(dis##VU##MI_XITOP, dName("XITOP"); dCP232i(_Ft_);) \ -MakeDisF(dis##VU##MI_XGKICK, dName("XGKICK"); dCP232i(_Fs_);) \ -MakeDisF(dis##VU##MI_XTOP, dName("XTOP"); dCP232i(_Ft_);) \ +MakeDisF(dis##VU##MI_XITOP, dName("XITOP"); dCP232i(_It_);) \ +MakeDisF(dis##VU##MI_XGKICK, dName("XGKICK"); dCP232i(_Is_);) \ +MakeDisF(dis##VU##MI_XTOP, dName("XTOP"); dCP232i(_It_);) \ \ \ /*****************/ \ diff --git a/pcsx2/VUops.cpp b/pcsx2/VUops.cpp index c0341158b7..76a25daae6 100644 --- a/pcsx2/VUops.cpp +++ b/pcsx2/VUops.cpp @@ -30,6 +30,9 @@ #define _Ft_ ((VU->code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ ((VU->code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ ((VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 0xF) +#define _Is_ (_Fs_ & 0xF) +#define _Id_ (_Fd_ & 0xF) #define _X ((VU->code>>24) & 0x1) #define _Y ((VU->code>>23) & 0x1) @@ -1508,38 +1511,38 @@ void _vuRSQRT(VURegs * VU) { void _vuIADDI(VURegs * VU) { s16 imm = ((VU->code >> 6) & 0x1f); imm = ((imm & 0x10 ? 0xfff0 : 0) | (imm & 0xf)); - if(_Ft_ == 0) return; - VU->VI[_Ft_].SS[0] = VU->VI[_Fs_].SS[0] + imm; + if(_It_ == 0) return; + VU->VI[_It_].SS[0] = VU->VI[_Is_].SS[0] + imm; }//last checked 17/05/03 shadow NOTE: not quite sure about that void _vuIADDIU(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].SS[0] = VU->VI[_Fs_].SS[0] + (((VU->code >> 10) & 0x7800) | (VU->code & 0x7ff)); + if(_It_ == 0) return; + VU->VI[_It_].SS[0] = VU->VI[_Is_].SS[0] + (((VU->code >> 10) & 0x7800) | (VU->code & 0x7ff)); }//last checked 17/05/03 shadow void _vuIADD(VURegs * VU) { - if(_Fd_ == 0) return; - VU->VI[_Fd_].SS[0] = VU->VI[_Fs_].SS[0] + VU->VI[_Ft_].SS[0]; + if(_Id_ == 0) return; + VU->VI[_Id_].SS[0] = VU->VI[_Is_].SS[0] + VU->VI[_It_].SS[0]; }//last checked 17/05/03 shadow void _vuIAND(VURegs * VU) { - if(_Fd_ == 0) return; - VU->VI[_Fd_].US[0] = VU->VI[_Fs_].US[0] & VU->VI[_Ft_].US[0]; + if(_Id_ == 0) return; + VU->VI[_Id_].US[0] = VU->VI[_Is_].US[0] & VU->VI[_It_].US[0]; }//last checked 17/05/03 shadow void _vuIOR(VURegs * VU) { - if(_Fd_ == 0) return; - VU->VI[_Fd_].US[0] = VU->VI[_Fs_].US[0] | VU->VI[_Ft_].US[0]; + if(_Id_ == 0) return; + VU->VI[_Id_].US[0] = VU->VI[_Is_].US[0] | VU->VI[_It_].US[0]; } void _vuISUB(VURegs * VU) { - if(_Fd_ == 0) return; - VU->VI[_Fd_].SS[0] = VU->VI[_Fs_].SS[0] - VU->VI[_Ft_].SS[0]; + if(_Id_ == 0) return; + VU->VI[_Id_].SS[0] = VU->VI[_Is_].SS[0] - VU->VI[_It_].SS[0]; } void _vuISUBIU(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].SS[0] = VU->VI[_Fs_].SS[0] - (((VU->code >> 10) & 0x7800) | (VU->code & 0x7ff)); + if(_It_ == 0) return; + VU->VI[_It_].SS[0] = VU->VI[_Is_].SS[0] - (((VU->code >> 10) & 0x7800) | (VU->code & 0x7ff)); } void _vuMOVE(VURegs * VU) { @@ -1554,16 +1557,16 @@ void _vuMOVE(VURegs * VU) { void _vuMFIR(VURegs * VU) { if (_Ft_ == 0) return; - if (_X) VU->VF[_Ft_].SL[0] = (s32)VU->VI[_Fs_].SS[0]; - if (_Y) VU->VF[_Ft_].SL[1] = (s32)VU->VI[_Fs_].SS[0]; - if (_Z) VU->VF[_Ft_].SL[2] = (s32)VU->VI[_Fs_].SS[0]; - if (_W) VU->VF[_Ft_].SL[3] = (s32)VU->VI[_Fs_].SS[0]; + if (_X) VU->VF[_Ft_].SL[0] = (s32)VU->VI[_Is_].SS[0]; + if (_Y) VU->VF[_Ft_].SL[1] = (s32)VU->VI[_Is_].SS[0]; + if (_Z) VU->VF[_Ft_].SL[2] = (s32)VU->VI[_Is_].SS[0]; + if (_W) VU->VF[_Ft_].SL[3] = (s32)VU->VI[_Is_].SS[0]; } // Big bug!!! mov from fs to ft not ft to fs. asadr void _vuMTIR(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = *(u16*)&VU->VF[_Fs_].F[_Fsf_]; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = *(u16*)&VU->VF[_Fs_].F[_Fsf_]; } void _vuMR32(VURegs * VU) { @@ -1585,7 +1588,7 @@ void _vuLQ(VURegs * VU) { if (_Ft_ == 0) return; imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - addr = ((imm + VU->VI[_Fs_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = ((imm + VU->VI[_Is_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) VU->VF[_Ft_].UL[0] = ptr[0]; @@ -1598,10 +1601,10 @@ void _vuLQD( VURegs * VU ) { u32 addr; u32 *ptr; - if (_Fs_ != 0) VU->VI[_Fs_].US[0]--; + if (_Is_ != 0) VU->VI[_Is_].US[0]--; if (_Ft_ == 0) return; - addr = (VU->VI[_Fs_].US[0] * 16) & (VU == &VU1 ? 0x3fff : 0xfff); + addr = (VU->VI[_Is_].US[0] * 16) & (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) VU->VF[_Ft_].UL[0] = ptr[0]; if (_Y) VU->VF[_Ft_].UL[1] = ptr[1]; @@ -1614,14 +1617,14 @@ void _vuLQI(VURegs * VU) { u32 addr; u32 *ptr; - addr = (VU->VI[_Fs_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = (VU->VI[_Is_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) VU->VF[_Ft_].UL[0] = ptr[0]; if (_Y) VU->VF[_Ft_].UL[1] = ptr[1]; if (_Z) VU->VF[_Ft_].UL[2] = ptr[2]; if (_W) VU->VF[_Ft_].UL[3] = ptr[3]; } - if (_Fs_ != 0) VU->VI[_Fs_].US[0]++; + if (_Fs_ != 0) VU->VI[_Is_].US[0]++; } /* addr is now signed. Asadr */ @@ -1631,7 +1634,7 @@ void _vuSQ(VURegs * VU) { u32 *ptr; imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - addr = ((imm + VU->VI[_Ft_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = ((imm + VU->VI[_It_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) ptr[0] = VU->VF[_Fs_].UL[0]; if (_Y) ptr[1] = VU->VF[_Fs_].UL[1]; @@ -1643,8 +1646,8 @@ void _vuSQD(VURegs * VU) { u32 addr; u32 *ptr; - if(_Ft_ != 0) VU->VI[_Ft_].US[0]--; - addr = (VU->VI[_Ft_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + if(_Ft_ != 0) VU->VI[_It_].US[0]--; + addr = (VU->VI[_It_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) ptr[0] = VU->VF[_Fs_].UL[0]; if (_Y) ptr[1] = VU->VF[_Fs_].UL[1]; @@ -1656,13 +1659,13 @@ void _vuSQI(VURegs * VU) { u32 addr; u32 *ptr; - addr = (VU->VI[_Ft_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = (VU->VI[_It_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u32*)GET_VU_MEM(VU, addr); if (_X) ptr[0] = VU->VF[_Fs_].UL[0]; if (_Y) ptr[1] = VU->VF[_Fs_].UL[1]; if (_Z) ptr[2] = VU->VF[_Fs_].UL[2]; if (_W) ptr[3] = VU->VF[_Fs_].UL[3]; - if(_Ft_ != 0) VU->VI[_Ft_].US[0]++; + if(_Ft_ != 0) VU->VI[_It_].US[0]++; } /* addr now signed. asadr */ @@ -1670,15 +1673,15 @@ void _vuILW(VURegs * VU) { s16 imm; u16 addr; u16 *ptr; - if (_Ft_ == 0) return; + if (_It_ == 0) return; imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - addr = ((imm + VU->VI[_Fs_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = ((imm + VU->VI[_Is_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u16*)GET_VU_MEM(VU, addr); - if (_X) VU->VI[_Ft_].US[0] = ptr[0]; - if (_Y) VU->VI[_Ft_].US[0] = ptr[2]; - if (_Z) VU->VI[_Ft_].US[0] = ptr[4]; - if (_W) VU->VI[_Ft_].US[0] = ptr[6]; + if (_X) VU->VI[_It_].US[0] = ptr[0]; + if (_Y) VU->VI[_It_].US[0] = ptr[2]; + if (_Z) VU->VI[_It_].US[0] = ptr[4]; + if (_W) VU->VI[_It_].US[0] = ptr[6]; } void _vuISW(VURegs * VU) { @@ -1687,37 +1690,37 @@ void _vuISW(VURegs * VU) { u16 *ptr; imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - addr = ((imm + VU->VI[_Fs_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = ((imm + VU->VI[_Is_].SS[0]) * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u16*)GET_VU_MEM(VU, addr); - if (_X) { ptr[0] = VU->VI[_Ft_].US[0]; ptr[1] = 0; } - if (_Y) { ptr[2] = VU->VI[_Ft_].US[0]; ptr[3] = 0; } - if (_Z) { ptr[4] = VU->VI[_Ft_].US[0]; ptr[5] = 0; } - if (_W) { ptr[6] = VU->VI[_Ft_].US[0]; ptr[7] = 0; } + if (_X) { ptr[0] = VU->VI[_It_].US[0]; ptr[1] = 0; } + if (_Y) { ptr[2] = VU->VI[_It_].US[0]; ptr[3] = 0; } + if (_Z) { ptr[4] = VU->VI[_It_].US[0]; ptr[5] = 0; } + if (_W) { ptr[6] = VU->VI[_It_].US[0]; ptr[7] = 0; } } void _vuILWR(VURegs * VU) { u32 addr; u16 *ptr; - if (_Ft_ == 0) return; + if (_It_ == 0) return; - addr = (VU->VI[_Fs_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); + addr = (VU->VI[_Is_].US[0] * 16)& (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u16*)GET_VU_MEM(VU, addr); - if (_X) VU->VI[_Ft_].US[0] = ptr[0]; - if (_Y) VU->VI[_Ft_].US[0] = ptr[2]; - if (_Z) VU->VI[_Ft_].US[0] = ptr[4]; - if (_W) VU->VI[_Ft_].US[0] = ptr[6]; + if (_X) VU->VI[_It_].US[0] = ptr[0]; + if (_Y) VU->VI[_It_].US[0] = ptr[2]; + if (_Z) VU->VI[_It_].US[0] = ptr[4]; + if (_W) VU->VI[_It_].US[0] = ptr[6]; } void _vuISWR(VURegs * VU) { u32 addr; u16 *ptr; - addr = (VU->VI[_Fs_].US[0] * 16) & (VU == &VU1 ? 0x3fff : 0xfff); + addr = (VU->VI[_Is_].US[0] * 16) & (VU == &VU1 ? 0x3fff : 0xfff); ptr = (u16*)GET_VU_MEM(VU, addr); - if (_X) { ptr[0] = VU->VI[_Ft_].US[0]; ptr[1] = 0; } - if (_Y) { ptr[2] = VU->VI[_Ft_].US[0]; ptr[3] = 0; } - if (_Z) { ptr[4] = VU->VI[_Ft_].US[0]; ptr[5] = 0; } - if (_W) { ptr[6] = VU->VI[_Ft_].US[0]; ptr[7] = 0; } + if (_X) { ptr[0] = VU->VI[_It_].US[0]; ptr[1] = 0; } + if (_Y) { ptr[2] = VU->VI[_It_].US[0]; ptr[3] = 0; } + if (_Z) { ptr[4] = VU->VI[_It_].US[0]; ptr[5] = 0; } + if (_W) { ptr[6] = VU->VI[_It_].US[0]; ptr[7] = 0; } } /* code contributed by _Riff_ @@ -1780,23 +1783,23 @@ void _vuWAITQ(VURegs * VU) { void _vuFSAND(VURegs * VU) { u16 imm; imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = (VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) & imm; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = (VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) & imm; } void _vuFSEQ(VURegs * VU) { u16 imm; imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - if(_Ft_ == 0) return; - if((VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) == imm) VU->VI[_Ft_].US[0] = 1; - else VU->VI[_Ft_].US[0] = 0; + if(_It_ == 0) return; + if((VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) == imm) VU->VI[_It_].US[0] = 1; + else VU->VI[_It_].US[0] = 0; } void _vuFSOR(VURegs * VU) { u16 imm; imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = (VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) | imm; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = (VU->VI[REG_STATUS_FLAG].US[0] & 0xFFF) | imm; } void _vuFSSET(VURegs * VU) { @@ -1806,19 +1809,19 @@ void _vuFSSET(VURegs * VU) { } void _vuFMAND(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = VU->VI[_Fs_].US[0] & (VU->VI[REG_MAC_FLAG].UL & 0xFFFF); + if(_It_ == 0) return; + VU->VI[_It_].US[0] = VU->VI[_Is_].US[0] & (VU->VI[REG_MAC_FLAG].UL & 0xFFFF); } void _vuFMEQ(VURegs * VU) { - if(_Ft_ == 0) return; - if((VU->VI[REG_MAC_FLAG].UL & 0xFFFF) == VU->VI[_Fs_].US[0]){ - VU->VI[_Ft_].US[0] =1;} else { VU->VI[_Ft_].US[0] =0; } + if(_It_ == 0) return; + if((VU->VI[REG_MAC_FLAG].UL & 0xFFFF) == VU->VI[_Is_].US[0]){ + VU->VI[_It_].US[0] =1;} else { VU->VI[_It_].US[0] =0; } } void _vuFMOR(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = (VU->VI[REG_MAC_FLAG].UL & 0xFFFF) | VU->VI[_Fs_].US[0]; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = (VU->VI[REG_MAC_FLAG].UL & 0xFFFF) | VU->VI[_Is_].US[0]; } void _vuFCAND(VURegs * VU) { @@ -1843,8 +1846,8 @@ void _vuFCSET(VURegs * VU) { } void _vuFCGET(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = VU->VI[REG_CLIP_FLAG].UL & 0x0FFF; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = VU->VI[REG_CLIP_FLAG].UL & 0x0FFF; } s32 _branchAddr(VURegs * VU) { @@ -1859,42 +1862,42 @@ void _setBranch(VURegs * VU, u32 bpc) { } void _vuIBEQ(VURegs * VU) { - if (VU->VI[_Ft_].US[0] == VU->VI[_Fs_].US[0]) { + if (VU->VI[_It_].US[0] == VU->VI[_Is_].US[0]) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } } void _vuIBGEZ(VURegs * VU) { - if (VU->VI[_Fs_].SS[0] >= 0) { + if (VU->VI[_Is_].SS[0] >= 0) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } } void _vuIBGTZ(VURegs * VU) { - if (VU->VI[_Fs_].SS[0] > 0) { + if (VU->VI[_Is_].SS[0] > 0) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } } void _vuIBLEZ(VURegs * VU) { - if (VU->VI[_Fs_].SS[0] <= 0) { + if (VU->VI[_Is_].SS[0] <= 0) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } } void _vuIBLTZ(VURegs * VU) { - if (VU->VI[_Fs_].SS[0] < 0) { + if (VU->VI[_Is_].SS[0] < 0) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } } void _vuIBNE(VURegs * VU) { - if (VU->VI[_Ft_].US[0] != VU->VI[_Fs_].US[0]) { + if (VU->VI[_It_].US[0] != VU->VI[_Is_].US[0]) { s32 bpc = _branchAddr(VU); _setBranch(VU, bpc); } @@ -1908,19 +1911,19 @@ void _vuB(VURegs * VU) { void _vuBAL(VURegs * VU) { s32 bpc = _branchAddr(VU); - if (_Ft_) VU->VI[_Ft_].US[0] = (VU->VI[REG_TPC].UL + 8)/8; + if (_It_) VU->VI[_It_].US[0] = (VU->VI[REG_TPC].UL + 8)/8; _setBranch(VU, bpc); } void _vuJR(VURegs * VU) { - u32 bpc = VU->VI[_Fs_].US[0] * 8; + u32 bpc = VU->VI[_Is_].US[0] * 8; _setBranch(VU, bpc); } void _vuJALR(VURegs * VU) { - u32 bpc = VU->VI[_Fs_].US[0] * 8; - if (_Ft_) VU->VI[_Ft_].US[0] = (VU->VI[REG_TPC].UL + 8)/8; + u32 bpc = VU->VI[_Is_].US[0] * 8; + if (_It_) VU->VI[_It_].US[0] = (VU->VI[REG_TPC].UL + 8)/8; _setBranch(VU, bpc); } @@ -2039,14 +2042,14 @@ void _vuEEXP(VURegs * VU) { } void _vuXITOP(VURegs * VU) { - if (_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = VU->vifRegs->itop; + if (_It_ == 0) return; + VU->VI[_It_].US[0] = VU->vifRegs->itop; } void _vuXGKICK(VURegs * VU) { - u32* ptr = (u32*)GET_VU_MEM(VU, (VU->VI[_Fs_].US[0]*16) & (VU == &VU1 ? 0x3fff : 0xfff)); -// int temp = 0x4000 - ((VU->VI[_Fs_].US[0]*16) & 0x3fff); + u32* ptr = (u32*)GET_VU_MEM(VU, (VU->VI[_Is_].US[0]*16) & (VU == &VU1 ? 0x3fff : 0xfff)); +// int temp = 0x4000 - ((VU->VI[_Is_].US[0]*16) & 0x3fff); // u32 tempmem[0x8000]; // flush all pipelines first (in the right order) @@ -2056,14 +2059,14 @@ void _vuXGKICK(VURegs * VU) /* memset(tempmem, 0, sizeof(tempmem)); memcpy(tempmem, ptr, temp); ptr = (u32*)GET_VU_MEM(VU, 0); - memcpy(&tempmem[temp], ptr, ((VU->VI[_Fs_].US[0]*16) & 0x3fff)); + memcpy(&tempmem[temp], ptr, ((VU->VI[_Is_].US[0]*16) & 0x3fff)); GSGIFTRANSFER1((u32*)&tempmem[0], 0); - } else*/ GSGIFTRANSFER1((u32*)VU->Mem, (VU->VI[_Fs_].US[0]*16) & 0x3fff); + } else*/ GSGIFTRANSFER1((u32*)VU->Mem, (VU->VI[_Is_].US[0]*16) & 0x3fff); } void _vuXTOP(VURegs * VU) { - if(_Ft_ == 0) return; - VU->VI[_Ft_].US[0] = (u16)VU->vifRegs->top; + if(_It_ == 0) return; + VU->VI[_It_].US[0] = (u16)VU->vifRegs->top; } #define GET_VF0_FLAG(reg) (((reg)==0)?(1<VFwrite = 0; \ VUregsn->VFread0 = 0; \ VUregsn->VFread1 = 0; \ - VUregsn->VIwrite = 1 << _Fd_; \ - VUregsn->VIread = (1 << _Fs_) | (1 << _Ft_); \ + VUregsn->VIwrite = 1 << _Id_; \ + VUregsn->VIread = (1 << _Is_) | (1 << _It_); \ VUregsn->cycles = 0; \ } @@ -2209,8 +2212,8 @@ void _vuRegs##OP(VURegs * VU, _VURegsNum *VUregsn) { \ VUregsn->VFwrite = 0; \ VUregsn->VFread0 = 0; \ VUregsn->VFread1 = 0; \ - VUregsn->VIwrite = 1 << _Ft_; \ - VUregsn->VIread = 1 << _Fs_; \ + VUregsn->VIwrite = 1 << _It_; \ + VUregsn->VIread = 1 << _Is_; \ VUregsn->cycles = 0; \ } @@ -2512,7 +2515,7 @@ void _vuRegsMFIR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsMTIR(VURegs * VU, _VURegsNum *VUregsn) { @@ -2521,7 +2524,7 @@ void _vuRegsMTIR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = _Fs_; VUregsn->VFr0xyzw= 1 << (3-_Fsf_); VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = GET_VF0_FLAG(_Fs_); } @@ -2544,7 +2547,7 @@ void _vuRegsLQ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsLQD(VURegs * VU, _VURegsNum *VUregsn) { @@ -2553,8 +2556,8 @@ void _vuRegsLQD(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwxyzw = _XYZW; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Fs_; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIwrite = 1 << _Is_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsLQI(VURegs * VU, _VURegsNum *VUregsn) { @@ -2563,8 +2566,8 @@ void _vuRegsLQI(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwxyzw = _XYZW; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Fs_; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIwrite = 1 << _Is_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsSQ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2574,7 +2577,7 @@ void _vuRegsSQ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFr0xyzw= _XYZW; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = (1 << _Ft_)|GET_VF0_FLAG(_Fs_); + VUregsn->VIread = 1 << _It_; } void _vuRegsSQD(VURegs * VU, _VURegsNum *VUregsn) { @@ -2583,8 +2586,8 @@ void _vuRegsSQD(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = _Fs_; VUregsn->VFr0xyzw= _XYZW; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = (1 << _Ft_)|GET_VF0_FLAG(_Fs_); + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = 1 << _It_; } void _vuRegsSQI(VURegs * VU, _VURegsNum *VUregsn) { @@ -2593,8 +2596,8 @@ void _vuRegsSQI(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = _Fs_; VUregsn->VFr0xyzw= _XYZW; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = (1 << _Ft_)|GET_VF0_FLAG(_Fs_); + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = 1 << _It_; } void _vuRegsILW(VURegs * VU, _VURegsNum *VUregsn) { @@ -2602,8 +2605,8 @@ void _vuRegsILW(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = 1 << _Is_; VUregsn->cycles = 3; } @@ -2613,7 +2616,7 @@ void _vuRegsISW(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = (1 << _Fs_) | (1 << _Ft_); + VUregsn->VIread = (1 << _Is_) | (1 << _It_); } void _vuRegsILWR(VURegs * VU, _VURegsNum *VUregsn) { @@ -2621,8 +2624,8 @@ void _vuRegsILWR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = (1 << _Ft_); - VUregsn->VIread = (1 << _Fs_); + VUregsn->VIwrite = (1 << _It_); + VUregsn->VIread = (1 << _Is_); VUregsn->cycles = 3; } @@ -2632,7 +2635,7 @@ void _vuRegsISWR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = (1 << _Fs_) | (1 << _Ft_); + VUregsn->VIread = (1 << _Is_) | (1 << _It_); } void _vuRegsRINIT(VURegs * VU, _VURegsNum *VUregsn) { @@ -2689,7 +2692,7 @@ void _vuRegsFSAND(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 1 << REG_STATUS_FLAG; } @@ -2698,7 +2701,7 @@ void _vuRegsFSEQ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 1 << REG_STATUS_FLAG; } @@ -2707,7 +2710,7 @@ void _vuRegsFSOR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 1 << REG_STATUS_FLAG; } @@ -2725,8 +2728,8 @@ void _vuRegsFMAND(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Fs_); + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Is_); } void _vuRegsFMEQ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2734,8 +2737,8 @@ void _vuRegsFMEQ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Fs_); + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Is_); } void _vuRegsFMOR(VURegs * VU, _VURegsNum *VUregsn) { @@ -2743,8 +2746,8 @@ void _vuRegsFMOR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Fs_); + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = (1 << REG_MAC_FLAG) | (1 << _Is_); } void _vuRegsFCAND(VURegs * VU, _VURegsNum *VUregsn) { @@ -2788,7 +2791,7 @@ void _vuRegsFCGET(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 1 << REG_CLIP_FLAG; } @@ -2798,7 +2801,7 @@ void _vuRegsIBEQ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = (1 << _Fs_) | (1 << _Ft_); + VUregsn->VIread = (1 << _Is_) | (1 << _It_); } void _vuRegsIBGEZ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2807,7 +2810,7 @@ void _vuRegsIBGEZ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsIBGTZ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2816,7 +2819,7 @@ void _vuRegsIBGTZ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsIBLEZ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2825,7 +2828,7 @@ void _vuRegsIBLEZ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsIBLTZ(VURegs * VU, _VURegsNum *VUregsn) { @@ -2834,7 +2837,7 @@ void _vuRegsIBLTZ(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsIBNE(VURegs * VU, _VURegsNum *VUregsn) { @@ -2843,7 +2846,7 @@ void _vuRegsIBNE(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = (1 << _Fs_) | (1 << _Ft_); + VUregsn->VIread = (1 << _Is_) | (1 << _It_); } void _vuRegsB(VURegs * VU, _VURegsNum *VUregsn) { @@ -2860,7 +2863,7 @@ void _vuRegsBAL(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 0; } @@ -2870,7 +2873,7 @@ void _vuRegsJR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsJALR(VURegs * VU, _VURegsNum *VUregsn) { @@ -2878,8 +2881,8 @@ void _vuRegsJALR(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIwrite = 1 << _It_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsMFP(VURegs * VU, _VURegsNum *VUregsn) { @@ -2920,7 +2923,7 @@ void _vuRegsXITOP(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 0; VUregsn->cycles = 0; } @@ -2931,7 +2934,7 @@ void _vuRegsXGKICK(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; VUregsn->VIwrite = 0; - VUregsn->VIread = 1 << _Fs_; + VUregsn->VIread = 1 << _Is_; } void _vuRegsXTOP(VURegs * VU, _VURegsNum *VUregsn) { @@ -2939,7 +2942,7 @@ void _vuRegsXTOP(VURegs * VU, _VURegsNum *VUregsn) { VUregsn->VFwrite = 0; VUregsn->VFread0 = 0; VUregsn->VFread1 = 0; - VUregsn->VIwrite = 1 << _Ft_; + VUregsn->VIwrite = 1 << _It_; VUregsn->VIread = 0; VUregsn->cycles = 0; } diff --git a/pcsx2/x86/iVUmicro.cpp b/pcsx2/x86/iVUmicro.cpp index 9335151798..7738a52da9 100644 --- a/pcsx2/x86/iVUmicro.cpp +++ b/pcsx2/x86/iVUmicro.cpp @@ -46,6 +46,9 @@ #define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) #define _X (( VU->code>>24) & 0x1) #define _Y (( VU->code>>23) & 0x1) diff --git a/pcsx2/x86/iVUmicroLower.cpp b/pcsx2/x86/iVUmicroLower.cpp index f98ed4549c..14d3479b42 100644 --- a/pcsx2/x86/iVUmicroLower.cpp +++ b/pcsx2/x86/iVUmicroLower.cpp @@ -38,6 +38,9 @@ #define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) #define _X (( VU->code>>24) & 0x1) #define _Y (( VU->code>>23) & 0x1) @@ -257,28 +260,28 @@ void recVUMI_RSQRT(VURegs *VU, int info) //------------------------------------------------------------------ void _addISIMMtoIT(VURegs *VU, s16 imm, int info) { - int fsreg = -1, ftreg; - if (_Ft_ == 0) return; + int isreg = -1, itreg; + if (_It_ == 0) return; - if( _Fs_ == 0 ) { - ftreg = ALLOCVI(_Ft_, MODE_WRITE); - MOV32ItoR(ftreg, imm&0xffff); + if( _Is_ == 0 ) { + itreg = ALLOCVI(_It_, MODE_WRITE); + MOV32ItoR(itreg, imm&0xffff); return; } - ADD_VI_NEEDED(_Ft_); - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE); - if ( _Ft_ == _Fs_ ) { - if (imm != 0 ) ADD16ItoR(ftreg, imm); + if ( _It_ == _Is_ ) { + if (imm != 0 ) ADD16ItoR(itreg, imm); } else { if( imm ) { - LEA32RtoR(ftreg, fsreg, imm); - MOVZX32R16toR(ftreg, ftreg); + LEA32RtoR(itreg, isreg, imm); + MOVZX32R16toR(itreg, itreg); } - else MOV32RtoR(ftreg, fsreg); + else MOV32RtoR(itreg, isreg); } } //------------------------------------------------------------------ @@ -291,7 +294,7 @@ void recVUMI_IADDI(VURegs *VU, int info) { s16 imm; - if ( _Ft_ == 0 ) return; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_IADDI"); imm = ( VU->code >> 6 ) & 0x1f; imm = ( imm & 0x10 ? 0xfff0 : 0) | ( imm & 0xf ); @@ -307,7 +310,7 @@ void recVUMI_IADDIU(VURegs *VU, int info) { s16 imm; - if ( _Ft_ == 0 ) return; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_IADDIU"); imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); _addISIMMtoIT(VU, imm, info); @@ -320,42 +323,42 @@ void recVUMI_IADDIU(VURegs *VU, int info) //------------------------------------------------------------------ void recVUMI_IADD( VURegs *VU, int info ) { - int fdreg, fsreg = -1, ftreg = -1; - if ( _Fd_ == 0 ) return; + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; //Console::WriteLn("recVUMI_IADD"); - if ( ( _Ft_ == 0 ) && ( _Fs_ == 0 ) ) { - fdreg = ALLOCVI(_Fd_, MODE_WRITE); - XOR32RtoR(fdreg, fdreg); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); return; } - ADD_VI_NEEDED(_Fs_); - ADD_VI_NEEDED(_Ft_); - fdreg = ALLOCVI(_Fd_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); - if ( _Fs_ == 0 ) + if ( _Is_ == 0 ) { - if( (ftreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Ft_, MODE_READ)) >= 0 ) { - if( fdreg != ftreg ) MOV32RtoR(fdreg, ftreg); + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Ft_, 1)); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); } - else if ( _Ft_ == 0 ) + else if ( _It_ == 0 ) { - if( (fsreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Fs_, MODE_READ)) >= 0 ) { - if( fdreg != fsreg ) MOV32RtoR(fdreg, fsreg); + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Fs_, 1)); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); } else { - //ADD_VI_NEEDED(_Ft_); - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_READ); + //ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); - if( fdreg == fsreg ) ADD32RtoR(fdreg, ftreg); - else if( fdreg == ftreg ) ADD32RtoR(fdreg, fsreg); - else LEA32RRtoR(fdreg, fsreg, ftreg); - MOVZX32R16toR(fdreg, fdreg); // needed since don't know if fdreg's upper bits are 0 + if( idreg == isreg ) ADD32RtoR(idreg, itreg); + else if( idreg == itreg ) ADD32RtoR(idreg, isreg); + else LEA32RRtoR(idreg, isreg, itreg); + MOVZX32R16toR(idreg, idreg); // needed since don't know if idreg's upper bits are 0 } } //------------------------------------------------------------------ @@ -366,27 +369,27 @@ void recVUMI_IADD( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_IAND( VURegs *VU, int info ) { - int fdreg, fsreg = -1, ftreg = -1; - if ( _Fd_ == 0 ) return; + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; //Console::WriteLn("recVUMI_IAND"); - if ( ( _Fs_ == 0 ) || ( _Ft_ == 0 ) ) { - fdreg = ALLOCVI(_Fd_, MODE_WRITE); - XOR32RtoR(fdreg, fdreg); + if ( ( _Is_ == 0 ) || ( _It_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); return; } - ADD_VI_NEEDED(_Fs_); - ADD_VI_NEEDED(_Ft_); - fdreg = ALLOCVI(_Fd_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_READ); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); - if( fdreg == fsreg ) AND16RtoR(fdreg, ftreg); - else if( fdreg == ftreg ) AND16RtoR(fdreg, fsreg); + if( idreg == isreg ) AND16RtoR(idreg, itreg); + else if( idreg == itreg ) AND16RtoR(idreg, isreg); else { - MOV32RtoR(fdreg, ftreg); - AND32RtoR(fdreg, fsreg); + MOV32RtoR(idreg, itreg); + AND32RtoR(idreg, isreg); } } //------------------------------------------------------------------ @@ -397,43 +400,43 @@ void recVUMI_IAND( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_IOR( VURegs *VU, int info ) { - int fdreg, fsreg = -1, ftreg = -1; - if ( _Fd_ == 0 ) return; + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; //Console::WriteLn("recVUMI_IOR"); - if ( ( _Ft_ == 0 ) && ( _Fs_ == 0 ) ) { - fdreg = ALLOCVI(_Fd_, MODE_WRITE); - XOR32RtoR(fdreg, fdreg); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); return; } - ADD_VI_NEEDED(_Fs_); - ADD_VI_NEEDED(_Ft_); - fdreg = ALLOCVI(_Fd_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); - if ( _Fs_ == 0 ) + if ( _Is_ == 0 ) { - if( (ftreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Ft_, MODE_READ)) >= 0 ) { - if( fdreg != ftreg ) MOV32RtoR(fdreg, ftreg); + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Ft_, 1)); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); } - else if ( _Ft_ == 0 ) + else if ( _It_ == 0 ) { - if( (fsreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Fs_, MODE_READ)) >= 0 ) { - if( fdreg != fsreg ) MOV32RtoR(fdreg, fsreg); + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Fs_, 1)); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); } else { - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_READ); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); - if( fdreg == fsreg ) OR16RtoR(fdreg, ftreg); - else if( fdreg == ftreg ) OR16RtoR(fdreg, fsreg); + if( idreg == isreg ) OR16RtoR(idreg, itreg); + else if( idreg == itreg ) OR16RtoR(idreg, isreg); else { - MOV32RtoR(fdreg, fsreg); - OR32RtoR(fdreg, ftreg); + MOV32RtoR(idreg, isreg); + OR32RtoR(idreg, itreg); } } } @@ -445,47 +448,47 @@ void recVUMI_IOR( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_ISUB( VURegs *VU, int info ) { - int fdreg, fsreg = -1, ftreg = -1; - if ( _Fd_ == 0 ) return; + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; //Console::WriteLn("recVUMI_ISUB"); - if ( ( _Ft_ == 0 ) && ( _Fs_ == 0 ) ) { - fdreg = ALLOCVI(_Fd_, MODE_WRITE); - XOR32RtoR(fdreg, fdreg); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); return; } - ADD_VI_NEEDED(_Fs_); - ADD_VI_NEEDED(_Ft_); - fdreg = ALLOCVI(_Fd_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); - if ( _Fs_ == 0 ) + if ( _Is_ == 0 ) { - if( (ftreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Ft_, MODE_READ)) >= 0 ) { - if( fdreg != ftreg ) MOV32RtoR(fdreg, ftreg); + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Ft_, 1)); - NEG16R(fdreg); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); + NEG16R(idreg); } - else if ( _Ft_ == 0 ) + else if ( _It_ == 0 ) { - if( (fsreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Fs_, MODE_READ)) >= 0 ) { - if( fdreg != fsreg ) MOV32RtoR(fdreg, fsreg); + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); } - else MOVZX32M16toR(fdreg, VU_VI_ADDR(_Fs_, 1)); + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); } else { - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_READ); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); - if( fdreg == fsreg ) SUB16RtoR(fdreg, ftreg); - else if( fdreg == ftreg ) { - SUB16RtoR(fdreg, fsreg); - NEG16R(fdreg); + if( idreg == isreg ) SUB16RtoR(idreg, itreg); + else if( idreg == itreg ) { + SUB16RtoR(idreg, isreg); + NEG16R(idreg); } else { - MOV32RtoR(fdreg, fsreg); - SUB16RtoR(fdreg, ftreg); + MOV32RtoR(idreg, isreg); + SUB16RtoR(idreg, itreg); } } } @@ -498,7 +501,7 @@ void recVUMI_ISUBIU( VURegs *VU, int info ) { s16 imm; - if ( _Ft_ == 0 ) return; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_ISUBIU"); imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); imm = -imm; @@ -531,23 +534,23 @@ void recVUMI_MFIR( VURegs *VU, int info ) { if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; //Console::WriteLn("recVUMI_MFIR"); - _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Fs_, 1); + _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, 1); if( _XYZW_SS ) { - SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Fs_, 1)-2); + SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); _vuFlipRegSS(VU, EEREC_T); SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); _vuFlipRegSS(VU, EEREC_T); } else if (_X_Y_Z_W != 0xf) { - SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Fs_, 1)-2); + SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); VU_MERGE_REGS(EEREC_T, EEREC_TEMP); } else { - SSE2_MOVD_M32_to_XMM(EEREC_T, VU_VI_ADDR(_Fs_, 1)-2); + SSE2_MOVD_M32_to_XMM(EEREC_T, VU_VI_ADDR(_Is_, 1)-2); SSE2_PSRAD_I8_to_XMM(EEREC_T, 16); SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); } @@ -560,19 +563,19 @@ void recVUMI_MFIR( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_MTIR( VURegs *VU, int info ) { - if ( _Ft_ == 0 ) return; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_MTIR"); - _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Ft_, 2); + _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, 2); if( _Fsf_ == 0 ) { - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_Ft_, 0), EEREC_S); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_S); } else { _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_Ft_, 0), EEREC_TEMP); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_TEMP); } - AND32ItoM(VU_VI_ADDR(_Ft_, 0), 0xffff); + AND32ItoM(VU_VI_ADDR(_It_, 0), 0xffff); } //------------------------------------------------------------------ @@ -723,12 +726,12 @@ void recVUMI_LQ(VURegs *VU, int info) if ( _Ft_ == 0 ) return; //Console::WriteLn("recVUMI_LQ"); imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - if (_Fs_ == 0) { + if (_Is_ == 0) { _loadEAX(VU, -1, (uptr)GET_VU_MEM(VU, (u32)imm*16), info); } else { - int fsreg = ALLOCVI(_Fs_, MODE_READ); - _loadEAX(VU, recVUTransformAddr(fsreg, VU, _Fs_, imm), (uptr)VU->Mem, info); + int isreg = ALLOCVI(_Is_, MODE_READ); + _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem, info); } } //------------------------------------------------------------------ @@ -739,17 +742,17 @@ void recVUMI_LQ(VURegs *VU, int info) //------------------------------------------------------------------ void recVUMI_LQD( VURegs *VU, int info ) { - int fsreg; + int isreg; //Console::WriteLn("recVUMI_LQD"); - if ( _Fs_ != 0 ) { - fsreg = ALLOCVI(_Fs_, MODE_READ|MODE_WRITE); - SUB16ItoR( fsreg, 1 ); + if ( _Is_ != 0 ) { + isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); + SUB16ItoR( isreg, 1 ); } if ( _Ft_ == 0 ) return; - if ( _Fs_ == 0 ) _loadEAX(VU, -1, (uptr)VU->Mem, info); - else _loadEAX(VU, recVUTransformAddr(fsreg, VU, _Fs_, 0), (uptr)VU->Mem, info); + if ( _Is_ == 0 ) _loadEAX(VU, -1, (uptr)VU->Mem, info); + else _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); } //------------------------------------------------------------------ @@ -759,27 +762,27 @@ void recVUMI_LQD( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_LQI(VURegs *VU, int info) { - int fsreg; + int isreg; //Console::WriteLn("recVUMI_LQI"); if ( _Ft_ == 0 ) { - if( _Fs_ != 0 ) { - if( (fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_WRITE|MODE_READ)) >= 0 ) { - ADD16ItoR(fsreg, 1); + if( _Is_ != 0 ) { + if( (isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_WRITE|MODE_READ)) >= 0 ) { + ADD16ItoR(isreg, 1); } else { - ADD16ItoM( VU_VI_ADDR( _Fs_, 0 ), 1 ); + ADD16ItoM( VU_VI_ADDR( _Is_, 0 ), 1 ); } } return; } - if (_Fs_ == 0) { + if (_Is_ == 0) { _loadEAX(VU, -1, (uptr)VU->Mem, info); } else { - fsreg = ALLOCVI(_Fs_, MODE_READ|MODE_WRITE); - _loadEAX(VU, recVUTransformAddr(fsreg, VU, _Fs_, 0), (uptr)VU->Mem, info); - ADD16ItoR( fsreg, 1 ); + isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); + _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); + ADD16ItoR( isreg, 1 ); } } //------------------------------------------------------------------ @@ -948,10 +951,10 @@ void recVUMI_SQ(VURegs *VU, int info) s16 imm; //Console::WriteLn("recVUMI_SQ"); imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); - if ( _Ft_ == 0 ) _saveEAX(VU, -1, (uptr)GET_VU_MEM(VU, (int)imm * 16), info); + if ( _It_ == 0 ) _saveEAX(VU, -1, (uptr)GET_VU_MEM(VU, (int)imm * 16), info); else { - int ftreg = ALLOCVI(_Ft_, MODE_READ); - _saveEAX(VU, recVUTransformAddr(ftreg, VU, _Ft_, imm), (uptr)VU->Mem, info); + int itreg = ALLOCVI(_It_, MODE_READ); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, imm), (uptr)VU->Mem, info); } } //------------------------------------------------------------------ @@ -963,11 +966,11 @@ void recVUMI_SQ(VURegs *VU, int info) void recVUMI_SQD(VURegs *VU, int info) { //Console::WriteLn("recVUMI_SQD"); - if (_Ft_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); + if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); else { - int ftreg = ALLOCVI(_Ft_, MODE_READ|MODE_WRITE); - SUB16ItoR( ftreg, 1 ); - _saveEAX(VU, recVUTransformAddr(ftreg, VU, _Ft_, 0), (uptr)VU->Mem, info); + int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); + SUB16ItoR( itreg, 1 ); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); } } //------------------------------------------------------------------ @@ -979,11 +982,11 @@ void recVUMI_SQD(VURegs *VU, int info) void recVUMI_SQI(VURegs *VU, int info) { //Console::WriteLn("recVUMI_SQI"); - if (_Ft_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); + if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); else { - int ftreg = ALLOCVI(_Ft_, MODE_READ|MODE_WRITE); - _saveEAX(VU, recVUTransformAddr(ftreg, VU, _Ft_, 0), (uptr)VU->Mem, info); - ADD16ItoR( ftreg, 1 ); + int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); + ADD16ItoR( itreg, 1 ); } } //------------------------------------------------------------------ @@ -994,10 +997,10 @@ void recVUMI_SQI(VURegs *VU, int info) //------------------------------------------------------------------ void recVUMI_ILW(VURegs *VU, int info) { - int ftreg; + int itreg; s16 imm, off; - if ( ( _Ft_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; + if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; //Console::WriteLn("recVUMI_ILW"); imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); if (_X) off = 0; @@ -1005,15 +1008,15 @@ void recVUMI_ILW(VURegs *VU, int info) else if (_Z) off = 8; else if (_W) off = 12; - ADD_VI_NEEDED(_Fs_); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE); - if ( _Fs_ == 0 ) { - MOVZX32M16toR( ftreg, (uptr)GET_VU_MEM(VU, (int)imm * 16 + off) ); + if ( _Is_ == 0 ) { + MOVZX32M16toR( itreg, (uptr)GET_VU_MEM(VU, (int)imm * 16 + off) ); } else { - int fsreg = ALLOCVI(_Fs_, MODE_READ); - MOV32RmtoR(ftreg, recVUTransformAddr(fsreg, VU, _Fs_, imm), (uptr)VU->Mem + off); + int isreg = ALLOCVI(_Is_, MODE_READ); + MOV32RmtoR(itreg, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem + off); } } //------------------------------------------------------------------ @@ -1028,28 +1031,28 @@ void recVUMI_ISW( VURegs *VU, int info ) //Console::WriteLn("recVUMI_ISW"); imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); - if (_Fs_ == 0) { + if (_Is_ == 0) { uptr off = (uptr)GET_VU_MEM(VU, (int)imm * 16); - int ftreg = ALLOCVI(_Ft_, MODE_READ); + int itreg = ALLOCVI(_It_, MODE_READ); - if (_X) MOV32RtoM(off, ftreg); - if (_Y) MOV32RtoM(off+4, ftreg); - if (_Z) MOV32RtoM(off+8, ftreg); - if (_W) MOV32RtoM(off+12, ftreg); + if (_X) MOV32RtoM(off, itreg); + if (_Y) MOV32RtoM(off+4, itreg); + if (_Z) MOV32RtoM(off+8, itreg); + if (_W) MOV32RtoM(off+12, itreg); } else { - int x86reg, fsreg, ftreg; + int x86reg, isreg, itreg; - ADD_VI_NEEDED(_Ft_); - fsreg = ALLOCVI(_Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_READ); + ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); - x86reg = recVUTransformAddr(fsreg, VU, _Fs_, imm); + x86reg = recVUTransformAddr(isreg, VU, _Is_, imm); - if (_X) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem); - if (_Y) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+4); - if (_Z) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+8); - if (_W) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+12); + if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); + if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); + if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); + if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); } } //------------------------------------------------------------------ @@ -1060,24 +1063,24 @@ void recVUMI_ISW( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_ILWR( VURegs *VU, int info ) { - int off, ftreg; + int off, itreg; - if ( ( _Ft_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; + if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; //Console::WriteLn("recVUMI_ILWR"); if (_X) off = 0; else if (_Y) off = 4; else if (_Z) off = 8; else if (_W) off = 12; - ADD_VI_NEEDED(_Fs_); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE); - if ( _Fs_ == 0 ) { - MOVZX32M16toR( ftreg, (uptr)VU->Mem + off ); + if ( _Is_ == 0 ) { + MOVZX32M16toR( itreg, (uptr)VU->Mem + off ); } else { - int fsreg = ALLOCVI(_Fs_, MODE_READ); - MOVZX32Rm16toR(ftreg, recVUTransformAddr(fsreg, VU, _Fs_, 0), (uptr)VU->Mem + off); + int isreg = ALLOCVI(_Is_, MODE_READ); + MOVZX32Rm16toR(itreg, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem + off); } } //------------------------------------------------------------------ @@ -1088,26 +1091,26 @@ void recVUMI_ILWR( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_ISWR( VURegs *VU, int info ) { - int ftreg; + int itreg; //Console::WriteLn("recVUMI_ISWR"); - ADD_VI_NEEDED(_Fs_); - ftreg = ALLOCVI(_Ft_, MODE_READ); + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_READ); - if (_Fs_ == 0) { - if (_X) MOV32RtoM((uptr)VU->Mem, ftreg); - if (_Y) MOV32RtoM((uptr)VU->Mem+4, ftreg); - if (_Z) MOV32RtoM((uptr)VU->Mem+8, ftreg); - if (_W) MOV32RtoM((uptr)VU->Mem+12, ftreg); + if (_Is_ == 0) { + if (_X) MOV32RtoM((uptr)VU->Mem, itreg); + if (_Y) MOV32RtoM((uptr)VU->Mem+4, itreg); + if (_Z) MOV32RtoM((uptr)VU->Mem+8, itreg); + if (_W) MOV32RtoM((uptr)VU->Mem+12, itreg); } else { int x86reg; - int fsreg = ALLOCVI(_Fs_, MODE_READ); - x86reg = recVUTransformAddr(fsreg, VU, _Fs_, 0); + int isreg = ALLOCVI(_Is_, MODE_READ); + x86reg = recVUTransformAddr(isreg, VU, _Is_, 0); - if (_X) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem); - if (_Y) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+4); - if (_Z) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+8); - if (_W) MOV32RtoRm(x86reg, ftreg, (uptr)VU->Mem+12); + if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); + if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); + if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); + if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); } } //------------------------------------------------------------------ @@ -1262,15 +1265,15 @@ void recVUMI_WAITQ( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FSAND( VURegs *VU, int info ) { - int ftreg; + int itreg; u16 imm; //Console::WriteLn("recVUMI_FSAND"); imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - if(_Ft_ == 0) return; + if(_It_ == 0) return; - ftreg = ALLOCVI(_Ft_, MODE_WRITE); - MOV32MtoR( ftreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - AND32ItoR( ftreg, imm ); + itreg = ALLOCVI(_It_, MODE_WRITE); + MOV32MtoR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); + AND32ItoR( itreg, imm ); } //------------------------------------------------------------------ @@ -1280,18 +1283,18 @@ void recVUMI_FSAND( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FSEQ( VURegs *VU, int info ) { - int ftreg; + int itreg; u16 imm; - if ( _Ft_ == 0 ) return; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_FSEQ"); imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - ftreg = ALLOCVI(_Ft_, MODE_WRITE|MODE_8BITREG); + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); MOVZX32M16toR( EAX, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - XOR32RtoR(ftreg, ftreg); + XOR32RtoR(itreg, itreg); CMP16ItoR(EAX, imm); - SETE8R(ftreg); + SETE8R(itreg); } //------------------------------------------------------------------ @@ -1301,16 +1304,16 @@ void recVUMI_FSEQ( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FSOR( VURegs *VU, int info ) { - int ftreg; + int itreg; u32 imm; - if(_Ft_ == 0) return; + if(_It_ == 0) return; //Console::WriteLn("recVUMI_FSOR"); imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + itreg = ALLOCVI(_It_, MODE_WRITE); - MOVZX32M16toR( ftreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - OR32ItoR( ftreg, imm ); + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); + OR32ItoR( itreg, imm ); } //------------------------------------------------------------------ @@ -1346,18 +1349,18 @@ void recVUMI_FSSET(VURegs *VU, int info) //------------------------------------------------------------------ void recVUMI_FMAND( VURegs *VU, int info ) { - int fsreg, ftreg; - if ( _Ft_ == 0 ) return; + int isreg, itreg; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_FMAND"); - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_WRITE);//|MODE_8BITREG); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); - if( fsreg >= 0 ) { - if( ftreg != fsreg ) MOV32RtoR(ftreg, fsreg); + if( isreg >= 0 ) { + if( itreg != isreg ) MOV32RtoR(itreg, isreg); } - else MOVZX32M16toR(ftreg, VU_VI_ADDR(_Fs_, 1)); + else MOVZX32M16toR(itreg, VU_VI_ADDR(_Is_, 1)); - AND16MtoR( ftreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); + AND16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); } //------------------------------------------------------------------ @@ -1367,25 +1370,25 @@ void recVUMI_FMAND( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FMEQ( VURegs *VU, int info ) { - int ftreg, fsreg; - if ( _Ft_ == 0 ) return; + int itreg, isreg; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_FMEQ"); - if( _Ft_ == _Fs_ ) { - ftreg = ALLOCVI(_Ft_, MODE_WRITE|MODE_READ);//|MODE_8BITREG + if( _It_ == _Is_ ) { + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG - CMP16MtoR(ftreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); + CMP16MtoR(itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); SETE8R(EAX); - MOVZX32R8toR(ftreg, EAX); + MOVZX32R8toR(itreg, EAX); } else { - ADD_VI_NEEDED(_Fs_); - ftreg = ALLOCVI(_Ft_, MODE_WRITE|MODE_8BITREG); - fsreg = ALLOCVI(_Fs_, MODE_READ); + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); + isreg = ALLOCVI(_Is_, MODE_READ); - XOR32RtoR(ftreg, ftreg); + XOR32RtoR(itreg, itreg); - CMP16MtoR(fsreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); - SETE8R(ftreg); + CMP16MtoR(isreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); + SETE8R(itreg); } } //------------------------------------------------------------------ @@ -1396,27 +1399,27 @@ void recVUMI_FMEQ( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FMOR( VURegs *VU, int info ) { - int fsreg, ftreg; - if ( _Ft_ == 0 ) return; + int isreg, itreg; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_FMOR"); - if( _Fs_ == 0 ) { - ftreg = ALLOCVI(_Ft_, MODE_WRITE);//|MODE_8BITREG); - MOVZX32M16toR( ftreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + if( _Is_ == 0 ) { + itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); } - else if( _Ft_ == _Fs_ ) { - ftreg = ALLOCVI(_Ft_, MODE_WRITE|MODE_READ);//|MODE_8BITREG); - OR16MtoR( ftreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + else if( _It_ == _Is_ ) { + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG); + OR16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); } else { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE); - MOVZX32M16toR( ftreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); - if( fsreg >= 0 ) - OR16RtoR( ftreg, fsreg ); + if( isreg >= 0 ) + OR16RtoR( itreg, isreg ); else - OR16MtoR( ftreg, VU_VI_ADDR(_Fs_, 1) ); + OR16MtoR( itreg, VU_VI_ADDR(_Is_, 1) ); } } //------------------------------------------------------------------ @@ -1427,13 +1430,13 @@ void recVUMI_FMOR( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FCAND( VURegs *VU, int info ) { - int ftreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); + int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); //Console::WriteLn("recVUMI_FCAND"); MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); - XOR32RtoR( ftreg, ftreg ); + XOR32RtoR( itreg, itreg ); AND32ItoR( EAX, VU->code & 0xFFFFFF ); - SETNZ8R(ftreg); + SETNZ8R(itreg); } //------------------------------------------------------------------ @@ -1443,14 +1446,14 @@ void recVUMI_FCAND( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FCEQ( VURegs *VU, int info ) { - int ftreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); + int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); //Console::WriteLn("recVUMI_FCEQ"); MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); AND32ItoR( EAX, 0xffffff ); - XOR32RtoR( ftreg, ftreg ); + XOR32RtoR( itreg, itreg ); CMP32ItoR( EAX, VU->code&0xffffff ); - SETE8R(ftreg); + SETE8R(itreg); } //------------------------------------------------------------------ @@ -1460,14 +1463,14 @@ void recVUMI_FCEQ( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FCOR( VURegs *VU, int info ) { - int ftreg; + int itreg; //Console::WriteLn("recVUMI_FCOR"); - ftreg = ALLOCVI(1, MODE_WRITE); - MOV32MtoR( ftreg, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); - OR32ItoR ( ftreg, VU->code ); - AND32ItoR( ftreg, 0xffffff ); - ADD32ItoR( ftreg, 1 ); // If 24 1's will make 25th bit 1, else 0 - SHR32ItoR( ftreg, 24 ); // Get the 25th bit (also clears the rest of the garbage in the reg) + itreg = ALLOCVI(1, MODE_WRITE); + MOV32MtoR( itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); + OR32ItoR ( itreg, VU->code ); + AND32ItoR( itreg, 0xffffff ); + ADD32ItoR( itreg, 1 ); // If 24 1's will make 25th bit 1, else 0 + SHR32ItoR( itreg, 24 ); // Get the 25th bit (also clears the rest of the garbage in the reg) } //------------------------------------------------------------------ @@ -1492,13 +1495,13 @@ void recVUMI_FCSET( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_FCGET( VURegs *VU, int info ) { - int ftreg; - if(_Ft_ == 0) return; + int itreg; + if(_It_ == 0) return; //Console::WriteLn("recVUMI_FCGET"); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); + itreg = ALLOCVI(_It_, MODE_WRITE); - MOV32MtoR(ftreg, VU_VI_ADDR(REG_CLIP_FLAG, 1)); - AND32ItoR(ftreg, 0x0fff); + MOV32MtoR(itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1)); + AND32ItoR(itreg, 0x0fff); } //------------------------------------------------------------------ @@ -1944,11 +1947,11 @@ void recVUMI_EEXP( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_XITOP( VURegs *VU, int info ) { - int ftreg; - if (_Ft_ == 0) return; + int itreg; + if (_It_ == 0) return; //Console::WriteLn("recVUMI_XITOP"); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); - MOVZX32M16toR( ftreg, (uptr)&VU->vifRegs->itop ); + itreg = ALLOCVI(_It_, MODE_WRITE); + MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->itop ); } //------------------------------------------------------------------ @@ -1958,11 +1961,11 @@ void recVUMI_XITOP( VURegs *VU, int info ) //------------------------------------------------------------------ void recVUMI_XTOP( VURegs *VU, int info ) { - int ftreg; - if ( _Ft_ == 0 ) return; + int itreg; + if ( _It_ == 0 ) return; //Console::WriteLn("recVUMI_XTOP"); - ftreg = ALLOCVI(_Ft_, MODE_WRITE); - MOVZX32M16toR( ftreg, (uptr)&VU->vifRegs->top ); + itreg = ALLOCVI(_It_, MODE_WRITE); + MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->top ); } //------------------------------------------------------------------ diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index b5fdbc1c54..1284d945a3 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -71,6 +71,9 @@ extern void iDumpVU1Registers(); #define _Ft_ ((VU->code >> 16) & 0x1F) // The rt part of the instruction register #define _Fs_ ((VU->code >> 11) & 0x1F) // The rd part of the instruction register #define _Fd_ ((VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) static const u32 QWaitTimes[] = { 6, 12 }; static const u32 PWaitTimes[] = { 53, 43, 28, 23, 17, 11, 10 }; @@ -3600,92 +3603,92 @@ void recVUMI_BranchHandle() // supervu specific insts void recVUMI_IBQ_prep() { - int fsreg, ftreg; + int isreg, itreg; - if( _Fs_ == 0 ) { + if( _Is_ == 0 ) { #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Ft_ ) { - ftreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_ ) { + itreg = -1; } else #endif { - ftreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Ft_, MODE_READ); + itreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_, MODE_READ); } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( ftreg >= 0 ) { - CMP16ItoR( ftreg, 0 ); + if( itreg >= 0 ) { + CMP16ItoR(itreg, 0); } - else CMP16ItoM(SuperVUGetVIAddr(_Ft_, 1), 0); + else CMP16ItoM(SuperVUGetVIAddr(_It_, 1), 0); } - else if( _Ft_ == 0 ) { + else if( _It_ == 0 ) { #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( fsreg >= 0 ) { - CMP16ItoR( fsreg, 0 ); + if( isreg >= 0 ) { + CMP16RtoR(isreg, 0); } - else CMP16ItoM(SuperVUGetVIAddr(_Fs_, 1), 0); + else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); } else { - _addNeededX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Ft_); + _addNeededX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Ft_ ) { - ftreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_ ) { + itreg = -1; - if( fsreg <= 0 ) { + if( isreg <= 0 ) { // allocate fsreg - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ|MODE_WRITE); - MOV32MtoR(fsreg, SuperVUGetVIAddr(_Fs_, 1)); + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ|MODE_WRITE); + MOV32MtoR(isreg, SuperVUGetVIAddr(_Is_, 1)); } else - fsreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } } else #endif { - ftreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Ft_, MODE_READ); + itreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_, MODE_READ); } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( fsreg >= 0 ) { - if( ftreg >= 0 ) { - CMP16RtoR( fsreg, ftreg ); + if( isreg >= 0 ) { + if( itreg >= 0 ) { + CMP16RtoR( isreg, itreg ); } - else CMP16MtoR(fsreg, SuperVUGetVIAddr(_Ft_, 1)); + else CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); } - else if( ftreg >= 0 ) { - CMP16MtoR(ftreg, SuperVUGetVIAddr(_Fs_, 1)); + else if( itreg >= 0 ) { + CMP16MtoR(itreg, SuperVUGetVIAddr(_Is_, 1)); } else { - fsreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); - CMP16MtoR(fsreg, SuperVUGetVIAddr(_Ft_, 1)); + isreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); } } } @@ -3699,25 +3702,25 @@ void recVUMI_IBEQ( VURegs* vuu, s32 info ) void recVUMI_IBGEZ( VURegs* vuu, s32 info ) { - int fsreg; + int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } - if( fsreg >= 0 ) { - OR16RtoR(fsreg, fsreg); + if( isreg >= 0 ) { + TEST16RtoR(isreg, isreg); j8Ptr[ 0 ] = JS8( 0 ); } else { - CMP16ItoM( SuperVUGetVIAddr(_Fs_, 1), 0x0 ); + CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); j8Ptr[ 0 ] = JL8( 0 ); } @@ -3726,25 +3729,25 @@ void recVUMI_IBGEZ( VURegs* vuu, s32 info ) void recVUMI_IBGTZ( VURegs* vuu, s32 info ) { - int fsreg; + int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } - if( fsreg >= 0 ) { - CMP16ItoR(fsreg, 0); + if( isreg >= 0 ) { + CMP16ItoR(isreg, 0); j8Ptr[ 0 ] = JLE8( 0 ); } else { - CMP16ItoM( SuperVUGetVIAddr(_Fs_, 1), 0x0 ); + CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); j8Ptr[ 0 ] = JLE8( 0 ); } recVUMI_BranchHandle(); @@ -3752,25 +3755,25 @@ void recVUMI_IBGTZ( VURegs* vuu, s32 info ) void recVUMI_IBLEZ( VURegs* vuu, s32 info ) { - int fsreg; + int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } - if( fsreg >= 0 ) { - CMP16ItoR(fsreg, 0); + if( isreg >= 0 ) { + CMP16ItoR(isreg, 0); j8Ptr[ 0 ] = JG8( 0 ); } else { - CMP16ItoM( SuperVUGetVIAddr(_Fs_, 1), 0x0 ); + CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); j8Ptr[ 0 ] = JG8( 0 ); } recVUMI_BranchHandle(); @@ -3778,25 +3781,25 @@ void recVUMI_IBLEZ( VURegs* vuu, s32 info ) void recVUMI_IBLTZ( VURegs* vuu, s32 info ) { - int fsreg; + int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Fs_ ) { - fsreg = -1; + if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + isreg = -1; } else #endif { - fsreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Fs_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); } - if( fsreg >= 0 ) { - OR16RtoR(fsreg, fsreg); + if( isreg >= 0 ) { + TEST16RtoR(isreg, isreg); j8Ptr[ 0 ] = JNS8( 0 ); } else { - CMP16ItoM( SuperVUGetVIAddr(_Fs_, 1), 0x0 ); + CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); j8Ptr[ 0 ] = JGE8( 0 ); } recVUMI_BranchHandle(); @@ -3842,9 +3845,9 @@ void recVUMI_BAL( VURegs* vuu, s32 info ) SuperVUTestVU0Condition(0); } - if ( _Ft_ ) { - _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Ft_, 2); - MOV16ItoM( SuperVUGetVIAddr(_Ft_, 0), (pc+8)>>3 ); + if ( _It_ ) { + _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_, 2); + MOV16ItoM( SuperVUGetVIAddr(_It_, 0), (pc+8)>>3 ); } if( s_pCurBlock->blocks.size() > 1 ) { @@ -3859,8 +3862,8 @@ void recVUMI_BAL( VURegs* vuu, s32 info ) void recVUMI_JR( VURegs* vuu, s32 info ) { - int fsreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Fs_, MODE_READ); - LEA32RStoR(EAX, fsreg, 3); + int isreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + LEA32RStoR(EAX, isreg, 3); //Mask the address to something valid if(vuu == &VU0) @@ -3879,10 +3882,10 @@ void recVUMI_JR( VURegs* vuu, s32 info ) void recVUMI_JALR( VURegs* vuu, s32 info ) { - _addNeededX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Ft_); + _addNeededX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_); - int fsreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Fs_, MODE_READ); - LEA32RStoR(EAX, fsreg, 3); + int isreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + LEA32RStoR(EAX, isreg, 3); //Mask the address to something valid if(vuu == &VU0) @@ -3890,9 +3893,9 @@ void recVUMI_JALR( VURegs* vuu, s32 info ) else AND32ItoR(EAX, 0x3fff); - if ( _Ft_ ) { - _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Ft_, 2); - MOV16ItoM( SuperVUGetVIAddr(_Ft_, 0), (pc+8)>>3 ); + if ( _It_ ) { + _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_, 2); + MOV16ItoM( SuperVUGetVIAddr(_It_, 0), (pc+8)>>3 ); } if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 ) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); @@ -3957,15 +3960,15 @@ void recVUMI_XGKICK( VURegs *VU, int info ) recVUMI_XGKICK_(VU); } - int fsreg = _allocX86reg(X86ARG2, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Fs_, MODE_READ); - _freeX86reg(fsreg); // flush - x86regs[fsreg].inuse = 1; - x86regs[fsreg].type = X86TYPE_VITEMP; - x86regs[fsreg].needed = 1; - x86regs[fsreg].mode = MODE_WRITE|MODE_READ; - SHL32ItoR(fsreg, 4); - AND32ItoR(fsreg, 0x3fff); - s_XGKICKReg = fsreg; + int isreg = _allocX86reg(X86ARG2, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + _freeX86reg(isreg); // flush + x86regs[isreg].inuse = 1; + x86regs[isreg].type = X86TYPE_VITEMP; + x86regs[isreg].needed = 1; + x86regs[isreg].mode = MODE_WRITE|MODE_READ; + SHL32ItoR(isreg, 4); + AND32ItoR(isreg, 0x3fff); + s_XGKICKReg = isreg; if( !SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc ) { recVUMI_XGKICK_(VU); From 3647a635bc5c02bfdef642d20fcab5798a6d04b4 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Mon, 4 May 2009 19:04:21 +0000 Subject: [PATCH 013/296] Apply recent Iop HwRegister handler changes to the EE's mapping [EE has direct access to IOP memory and registers through a special tlb mapping, intended for ps2dev debugging, and generally used by bios only during boot-up). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1133 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IopHw.cpp | 21 ++++------- pcsx2/IopHw.h | 23 ++++-------- pcsx2/IopMem.h | 8 ++++ pcsx2/Memory.cpp | 69 ++++++++++++++++++++++------------ pcsx2/PrecompiledHeader.h | 12 +++--- pcsx2/ps2/Iop/IopHwWrite.cpp | 39 +++++++++++++++++-- pcsx2/ps2/Iop/IopHw_Internal.h | 2 +- 7 files changed, 110 insertions(+), 64 deletions(-) diff --git a/pcsx2/IopHw.cpp b/pcsx2/IopHw.cpp index 80c5108f9e..2b4635fb20 100644 --- a/pcsx2/IopHw.cpp +++ b/pcsx2/IopHw.cpp @@ -1313,26 +1313,19 @@ void psxHwWrite32(u32 add, u32 value) { PSXHW_LOG("*Known 32bit write at address %lx value %lx", add, value); } -u8 psxHw4Read8(u32 add) +__forceinline u8 psxHw4Read8(u32 add) { - //u8 hard; u16 mem = add & 0xFF; - - //Console::WriteLn("psxHw4Read8 0x%x, %x", params add, mem); - return cdvdRead(mem); - - //PSXHW_LOG( "Known 8bit read from addr 0x%x = 0x%x", add, hard ); - - //return hard; + u8 ret = cdvdRead(mem); + PSXHW_LOG("HwRead8 from Cdvd [segment 0x1f40], addr 0x%02x = 0x%02x", mem, ret); + return ret; } -void psxHw4Write8(u32 add, u8 value) +__forceinline void psxHw4Write8(u32 add, u8 value) { - - u16 mem = add & 0xFF; - //Console::WriteLn("psxHw4Write8 0x%x, %x", params add, mem); + u8 mem = (u8)add; // only lower 8 bits are relevant (cdvd regs mirror across the page) cdvdWrite(mem, value); - PSXHW_LOG("Known 8bit write to addr 0x%x = 0x%x", add, value); + PSXHW_LOG("HwWrite8 to Cdvd [segment 0x1f40], addr 0x%02x = 0x%02x", mem, value); } void psxDmaInterrupt(int n) diff --git a/pcsx2/IopHw.h b/pcsx2/IopHw.h index 4b31f5f2c0..898b2de489 100644 --- a/pcsx2/IopHw.h +++ b/pcsx2/IopHw.h @@ -220,28 +220,19 @@ extern void psxSetNextBranchDelta( s32 delta ); extern int iopTestCycle( u32 startCycle, s32 delta ); extern void _iopTestInterrupts(); -void psxHwReset(); +// Depreciated : Use iopHwRead* functions defined in IopMem.h instead. u8 psxHwRead8 (u32 add); u16 psxHwRead16(u32 add); u32 psxHwRead32(u32 add); +// Depreciated : Use iopHwWrite* functions defined in IopMem.h instead. void psxHwWrite8 (u32 add, u8 value); void psxHwWrite16(u32 add, u16 value); void psxHwWrite32(u32 add, u32 value); -u8 psxHw4Read8 (u32 add); -void psxHw4Write8(u32 add, u8 value); +extern void psxHwReset(); +extern u8 psxHw4Read8 (u32 add); +extern void psxHw4Write8(u32 add, u8 value); -void psxDmaInterrupt(int n); -void psxDmaInterrupt2(int n); - -int psxHwFreeze(gzFile f, int Mode); - -int psxHwConstRead8(u32 x86reg, u32 add, u32 sign); -int psxHwConstRead16(u32 x86reg, u32 add, u32 sign); -int psxHwConstRead32(u32 x86reg, u32 add); -void psxHwConstWrite8(u32 add, int mmreg); -void psxHwConstWrite16(u32 add, int mmreg); -void psxHwConstWrite32(u32 add, int mmreg); -int psxHw4ConstRead8 (u32 x86reg, u32 add, u32 sign); -void psxHw4ConstWrite8(u32 add, int mmreg); +extern void psxDmaInterrupt(int n); +extern void psxDmaInterrupt2(int n); diff --git a/pcsx2/IopMem.h b/pcsx2/IopMem.h index dc62657c45..ad618c6b7f 100644 --- a/pcsx2/IopMem.h +++ b/pcsx2/IopMem.h @@ -106,6 +106,14 @@ namespace IopMemory extern void __fastcall SifWrite16( u32 iopaddr, u16 data ); extern void __fastcall SifWrite32( u32 iopaddr, u32 data ); + extern u8 __fastcall iopHwRead8_generic( u32 addr ); + extern u16 __fastcall iopHwRead16_generic( u32 addr ); + extern u32 __fastcall iopHwRead32_generic( u32 addr ); + extern void __fastcall iopHwWrite8_generic( u32 addr, u8 val ); + extern void __fastcall iopHwWrite16_generic( u32 addr, u16 val ); + extern void __fastcall iopHwWrite32_generic( u32 addr, u32 val ); + + extern u8 __fastcall iopHwRead8_Page1( u32 iopaddr ); extern u8 __fastcall iopHwRead8_Page3( u32 iopaddr ); extern u8 __fastcall iopHwRead8_Page8( u32 iopaddr ); diff --git a/pcsx2/Memory.cpp b/pcsx2/Memory.cpp index a5eb775e2a..b791ce559a 100644 --- a/pcsx2/Memory.cpp +++ b/pcsx2/Memory.cpp @@ -155,6 +155,10 @@ vtlbHandler hw_by_page[0x10]; vtlbHandler gs_page_0; vtlbHandler gs_page_1; +vtlbHandler iopHw_by_page_01; +vtlbHandler iopHw_by_page_03; +vtlbHandler iopHw_by_page_08; + // Used to remap the VUmicro memory according to the VU0/VU1 dynarec setting. // (the VU memory operations are different for recs vs. interpreters) @@ -195,12 +199,6 @@ void memMapPhy() vtlb_MapHandler(tlb_fallback_2,0x1f800000,0x10000); vtlb_MapHandler(tlb_fallback_8,0x1f900000,0x10000); -#ifdef PCSX2_DEVBUILD - // Bind fallback handlers used for logging purposes only. - // In release mode the Vtlb will map these addresses directly instead of using - // the read/write handlers (which just issue logs and do normal memOps) -#endif - // map specific optimized page handlers for HW accesses vtlb_MapHandler(hw_by_page[0x0], 0x10000000, 0x01000); vtlb_MapHandler(hw_by_page[0x1], 0x10001000, 0x01000); @@ -216,6 +214,11 @@ void memMapPhy() vtlb_MapHandler(gs_page_0, 0x12000000, 0x01000); vtlb_MapHandler(gs_page_1, 0x12001000, 0x01000); + + vtlb_MapHandler(hw_by_page[0x1], 0x1f801000, 0x01000); + vtlb_MapHandler(hw_by_page[0x3], 0x1f803000, 0x01000); + vtlb_MapHandler(hw_by_page[0x8], 0x1f808000, 0x01000); + } //Why is this required ? @@ -245,8 +248,6 @@ mem8_t __fastcall _ext_memRead8 (u32 mem) { case 1: // hwm return hwRead8(mem); - case 2: // psh - return psxHwRead8(mem); case 3: // psh4 return psxHw4Read8(mem); case 6: // gsm @@ -301,8 +302,6 @@ mem32_t __fastcall _ext_memRead32(u32 mem) { switch (p) { - case 2: // psh - return psxHwRead32(mem); case 6: // gsm return gsRead32(mem); case 7: // dev9 @@ -354,8 +353,6 @@ void __fastcall _ext_memWrite8 (u32 mem, u8 value) case 1: // hwm hwWrite8(mem, value); return; - case 2: // psh - psxHwWrite8(mem, value); return; case 3: // psh4 psxHw4Write8(mem, value); return; case 6: // gsm @@ -398,8 +395,6 @@ template void __fastcall _ext_memWrite32(u32 mem, u32 value) { switch (p) { - case 2: // psh - psxHwWrite32(mem, value); return; case 6: // gsm gsWrite32(mem, value); return; case 7: // dev9 @@ -660,14 +655,13 @@ void memReset() vtlb_Init(); - tlb_fallback_0=vtlb_RegisterHandlerTempl1(_ext_mem,0); - tlb_fallback_2=vtlb_RegisterHandlerTempl1(_ext_mem,2); - tlb_fallback_3=vtlb_RegisterHandlerTempl1(_ext_mem,3); - tlb_fallback_4=vtlb_RegisterHandlerTempl1(_ext_mem,4); - tlb_fallback_5=vtlb_RegisterHandlerTempl1(_ext_mem,5); - //tlb_fallback_6=vtlb_RegisterHandlerTempl1(_ext_mem,6); - tlb_fallback_7=vtlb_RegisterHandlerTempl1(_ext_mem,7); - tlb_fallback_8=vtlb_RegisterHandlerTempl1(_ext_mem,8); + tlb_fallback_0 = vtlb_RegisterHandlerTempl1(_ext_mem,0); + tlb_fallback_3 = vtlb_RegisterHandlerTempl1(_ext_mem,3); + tlb_fallback_4 = vtlb_RegisterHandlerTempl1(_ext_mem,4); + tlb_fallback_5 = vtlb_RegisterHandlerTempl1(_ext_mem,5); + //tlb_fallback_6 = vtlb_RegisterHandlerTempl1(_ext_mem,6); + tlb_fallback_7 = vtlb_RegisterHandlerTempl1(_ext_mem,7); + tlb_fallback_8 = vtlb_RegisterHandlerTempl1(_ext_mem,8); // Dynarec versions of VUs vu0_micro_mem[0] = vtlb_RegisterHandlerTempl2(vuMicro,0,true); @@ -677,7 +671,36 @@ void memReset() vu0_micro_mem[1] = vtlb_RegisterHandlerTempl2(vuMicro,0,false); vu1_micro_mem[1] = vtlb_RegisterHandlerTempl2(vuMicro,1,false); - ////////////////////////////////////////////////////////////////////// + ////////////////////////////////////////////////////////////////////////////////////////// + // IOP's "secret" Hardware Register mapping, accessible from the EE (and meant for use + // by debugging or BIOS only). The IOP's hw regs are divided into three main pages in + // the 0x1f80 segment, and then another oddball page for CDVD in the 0x1f40 segment. + // + + using namespace IopMemory; + + tlb_fallback_2 = vtlb_RegisterHandler( + iopHwRead8_generic, iopHwRead16_generic, iopHwRead32_generic, _ext_memRead64<2>, _ext_memRead128<2>, + iopHwWrite8_generic, iopHwWrite16_generic, iopHwWrite32_generic, _ext_memWrite64<2>, _ext_memWrite128<2> + ); + + iopHw_by_page_01 = vtlb_RegisterHandler( + iopHwRead8_Page1, iopHwRead16_Page1, iopHwRead32_Page1, _ext_memRead64<2>, _ext_memRead128<2>, + iopHwWrite8_Page1, iopHwWrite16_Page1, iopHwWrite32_Page1, _ext_memWrite64<2>, _ext_memWrite128<2> + ); + + iopHw_by_page_03 = vtlb_RegisterHandler( + iopHwRead8_Page3, iopHwRead16_Page3, iopHwRead32_Page3, _ext_memRead64<2>, _ext_memRead128<2>, + iopHwWrite8_Page3, iopHwWrite16_Page3, iopHwWrite32_Page3, _ext_memWrite64<2>, _ext_memWrite128<2> + ); + + iopHw_by_page_08 = vtlb_RegisterHandler( + iopHwRead8_Page8, iopHwRead16_Page8, iopHwRead32_Page8, _ext_memRead64<2>, _ext_memRead128<2>, + iopHwWrite8_Page8, iopHwWrite16_Page8, iopHwWrite32_Page8, _ext_memWrite64<2>, _ext_memWrite128<2> + ); + + + ////////////////////////////////////////////////////////////////////////////////////////// // psHw Optimized Mappings // The HW Registers have been split into pages to improve optimization. // Anything not explicitly mapped into one of the hw_by_page handlers will be handled diff --git a/pcsx2/PrecompiledHeader.h b/pcsx2/PrecompiledHeader.h index f9bbbe979a..8e91ef6769 100644 --- a/pcsx2/PrecompiledHeader.h +++ b/pcsx2/PrecompiledHeader.h @@ -9,13 +9,13 @@ // disabled in release builds. :) #ifdef _MSC_VER -#ifdef PCSX2_DEVBUILD -# define _SECURE_SCL 1 -# define _SECURE_SCL_THROWS 1 # pragma warning(disable:4244) // disable warning C4244: '=' : conversion from 'big' to 'small', possible loss of data -#else -# define _SECURE_SCL 0 -#endif +# ifdef PCSX2_DEVBUILD +# define _SECURE_SCL 1 +# define _SECURE_SCL_THROWS 1 +# else +# define _SECURE_SCL 0 +# endif #endif #define NOMINMAX // Disables other libs inclusion of their own min/max macros (we use std instead) diff --git a/pcsx2/ps2/Iop/IopHwWrite.cpp b/pcsx2/ps2/Iop/IopHwWrite.cpp index c538ba52b4..24c5bb9ba8 100644 --- a/pcsx2/ps2/Iop/IopHwWrite.cpp +++ b/pcsx2/ps2/Iop/IopHwWrite.cpp @@ -23,6 +23,41 @@ namespace IopMemory { using namespace Internal; +// Template-compatible version of the psxHu macro. Used for writing. +#define psxHu(mem) (*(u32*)&psxH[(mem) & 0xffff]) + + +////////////////////////////////////////////////////////////////////////////////////////// +// +template< typename T > +static __forceinline void _generic_write( u32 addr, T val ) +{ + int bitsize = (sizeof(T) == 1) ? 8 : ( (sizeof(T) == 2) ? 16 : 32 ); + PSXHW_LOG( "HwWrite%d to %s, addr 0x%08x = 0x%08x\n", bitsize, _log_GetIopHwName(addr), addr, val ); + psxHu(addr) = val; +} + +void __fastcall iopHwWrite8_generic( u32 addr, u8 val ) { _generic_write( addr, val ); } +void __fastcall iopHwWrite16_generic( u32 addr, u16 val ) { _generic_write( addr, val ); } +void __fastcall iopHwWrite32_generic( u32 addr, u32 val ) { _generic_write( addr, val ); } + +////////////////////////////////////////////////////////////////////////////////////////// +// +template< typename T > +static __forceinline T _generic_read( u32 addr ) +{ + int bitsize = (sizeof(T) == 1) ? 8 : ( (sizeof(T) == 2) ? 16 : 32 ); + + T ret = psxHu(addr); + PSXHW_LOG( "HwRead%d from %s, addr 0x%08x = 0x%08x\n", bitsize, _log_GetIopHwName(addr), addr, ret ); + return ret; +} + +u8 __fastcall iopHwRead8_generic( u32 addr ) { return _generic_read( addr ); } +u16 __fastcall iopHwRead16_generic( u32 addr ) { return _generic_read( addr ); } +u32 __fastcall iopHwRead32_generic( u32 addr ) { return _generic_read( addr ); } + + ////////////////////////////////////////////////////////////////////////////////////////// // void __fastcall iopHwWrite8_Page1( u32 addr, u8 val ) @@ -116,9 +151,6 @@ void __fastcall iopHwWrite8_Page8( u32 addr, u8 val ) } -// Template-compatible version of the psxHu macro. Used for writing. -#define psxHu(mem) (*(u32*)&psxH[(mem) & 0xffff]) - ////////////////////////////////////////////////////////////////////////////////////////// // Templated handler for both 32 and 16 bit write operations, to Page 1 registers. // @@ -508,7 +540,6 @@ void __fastcall iopHwWrite32_Page8( u32 addr, u32 val ) else psxHu32(addr) = val; PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, val ); - } } diff --git a/pcsx2/ps2/Iop/IopHw_Internal.h b/pcsx2/ps2/Iop/IopHw_Internal.h index 1a78d64bec..c15e7d9a0b 100644 --- a/pcsx2/ps2/Iop/IopHw_Internal.h +++ b/pcsx2/ps2/Iop/IopHw_Internal.h @@ -201,4 +201,4 @@ static __forceinline const char* _log_GetIopHwName( u32 addr ) } } -} }; \ No newline at end of file +} }; From 002b9af00679e6ddfd8366b99a1d2236e0081ee2 Mon Sep 17 00:00:00 2001 From: sudonim1 Date: Mon, 4 May 2009 20:14:37 +0000 Subject: [PATCH 014/296] This is why a type safe emitter is important. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1134 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVUzerorec.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 1284d945a3..b3fcf2fee6 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -3637,7 +3637,7 @@ void recVUMI_IBQ_prep() s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); if( isreg >= 0 ) { - CMP16RtoR(isreg, 0); + CMP16ItoR(isreg, 0); } else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); From a7cbcae4bebc2c6fc6b06806c34228c345149be0 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Tue, 5 May 2009 01:26:18 +0000 Subject: [PATCH 015/296] Small speedup for the EE/IOP recompilers, by removing the legacy limit on the number of recompiled blocks (was causing recompilers to reset the code buffers way more than necessary). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1135 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/BaseblockEx.cpp | 3 --- pcsx2/x86/BaseblockEx.h | 6 ++---- pcsx2/x86/iR3000A.cpp | 13 ++----------- pcsx2/x86/ix86-32/iR5900-32.cpp | 10 +--------- 4 files changed, 5 insertions(+), 27 deletions(-) diff --git a/pcsx2/x86/BaseblockEx.cpp b/pcsx2/x86/BaseblockEx.cpp index bb933f301d..7d8e5730e2 100644 --- a/pcsx2/x86/BaseblockEx.cpp +++ b/pcsx2/x86/BaseblockEx.cpp @@ -21,9 +21,6 @@ BASEBLOCKEX* BaseBlocks::New(u32 startpc, uptr fnptr) { - if (blocks.size() == size) - return 0; - BASEBLOCKEX newblock; std::vector::iterator iter; memset(&newblock, 0, sizeof newblock); diff --git a/pcsx2/x86/BaseblockEx.h b/pcsx2/x86/BaseblockEx.h index fa38d61891..38f97954c5 100644 --- a/pcsx2/x86/BaseblockEx.h +++ b/pcsx2/x86/BaseblockEx.h @@ -53,17 +53,15 @@ private: // switch to a hash map later? std::multimap links; typedef std::multimap::iterator linkiter_t; - unsigned long size; uptr recompiler; std::vector blocks; public: - BaseBlocks(unsigned long size_, uptr recompiler_) : - size(size_), + BaseBlocks(uptr recompiler_) : recompiler(recompiler_), blocks(0) { - blocks.reserve(size); + blocks.reserve(0x4000); } BASEBLOCKEX* New(u32 startpc, uptr fnptr); diff --git a/pcsx2/x86/iR3000A.cpp b/pcsx2/x86/iR3000A.cpp index 2faa34ba69..2aa7351333 100644 --- a/pcsx2/x86/iR3000A.cpp +++ b/pcsx2/x86/iR3000A.cpp @@ -51,7 +51,6 @@ uptr psxhwLUT[0x10000]; #define HWADDR(mem) (psxhwLUT[mem >> 16] + (mem)) -#define PSX_NUMBLOCKS (1<<12) #define MAPBASE 0x48000000 #define RECMEM_SIZE (8*1024*1024) @@ -63,7 +62,7 @@ static BASEBLOCK *recRAM = NULL; // and the ptr to the blocks here static BASEBLOCK *recROM = NULL; // and here static BASEBLOCK *recROM1 = NULL; // also here void iopJITCompile(); -static BaseBlocks recBlocks(PSX_NUMBLOCKS, (uptr)iopJITCompile); +static BaseBlocks recBlocks((uptr)iopJITCompile); static u8 *recPtr = NULL; u32 psxpc; // recompiler psxpc int psxbranch; // set for branch @@ -986,16 +985,8 @@ void iopRecRecompile(u32 startpc) || s_pCurBlock->GetFnptr() == (uptr)iopJITCompileInBlock); s_pCurBlockEx = recBlocks.Get(HWADDR(startpc)); - if(!s_pCurBlockEx || s_pCurBlockEx->startpc != HWADDR(startpc)) { + if(!s_pCurBlockEx || s_pCurBlockEx->startpc != HWADDR(startpc)) s_pCurBlockEx = recBlocks.New(HWADDR(startpc), (uptr)recPtr); - - if( s_pCurBlockEx == NULL ) { - DevCon::WriteLn("IOP Recompiler data reset"); - recResetIOP(); - x86SetPtr( recPtr ); - s_pCurBlockEx = recBlocks.New(HWADDR(startpc), (uptr)recPtr); - } - } psxbranch = 0; diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index f3768ab66e..96575543f7 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -76,7 +76,6 @@ u32 g_cpuHasConstReg = 0, g_cpuFlushedConstReg = 0; #define X86 static const int RECSTACK_SIZE = 0x00010000; -static const int EE_NUMBLOCKS = (1<<15); static u8 *recMem = NULL; // the recompiled blocks will be here static u8* recStack = NULL; // stack mem @@ -85,7 +84,7 @@ static BASEBLOCK *recROM = NULL; // and here static BASEBLOCK *recROM1 = NULL; // also here static u32 *recRAMCopy = NULL; void JITCompile(); -static BaseBlocks recBlocks(EE_NUMBLOCKS, (uptr)JITCompile); +static BaseBlocks recBlocks((uptr)JITCompile); static u8* recPtr = NULL, *recStackPtr = NULL; EEINST* s_pInstCache = NULL; static u32 s_nInstCacheSize = 0; @@ -1283,13 +1282,6 @@ void recRecompile( const u32 startpc ) s_pCurBlockEx = recBlocks.New(HWADDR(startpc), (uptr)recPtr); - if( s_pCurBlockEx == NULL ) { - //Console::WriteLn("ee reset (blocks)"); - recResetEE(); - x86SetPtr( recPtr ); - s_pCurBlockEx = recBlocks.New(HWADDR(startpc), (uptr)recPtr); - } - assert(s_pCurBlockEx); branch = 0; From 6a075aca0394b3013fdca615982fec2e4a8b80bc Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 5 May 2009 02:25:48 +0000 Subject: [PATCH 016/296] pcsx2: - PCSX2 would repetitively call CDVDopen if loading an elf file and your cdvd plugin didn't have an ISO already pre-loaded. Basically I had to switch to CDVDnull whenever I wanted to load an elf... well not anymore xD microVU: - Implemented the VI regs thing where if an opcode tried to access vi regs 16~31, the VU's should return vi0~15 (viReg & 0xf); technically it should never happen, but from my tests the BIOS seems to do this... so maybe some rare games do it too... - Our VIF code apparently uses VU cycles for some stuff, so I now update the cycles variable VIF relies on. - Minor Optimizations git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1136 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Plugins.cpp | 4 +- pcsx2/System.cpp | 8 +- pcsx2/x86/iVU0micro.cpp | 4 +- pcsx2/x86/iVU1micro.cpp | 4 +- pcsx2/x86/microVU_Compile.inl | 25 +-- pcsx2/x86/microVU_Execute.inl | 1 + pcsx2/x86/microVU_Lower.inl | 278 ++++++++++++++++++---------------- pcsx2/x86/microVU_Misc.h | 10 +- 8 files changed, 180 insertions(+), 154 deletions(-) diff --git a/pcsx2/Plugins.cpp b/pcsx2/Plugins.cpp index 2a9fe78880..bafd5a6012 100644 --- a/pcsx2/Plugins.cpp +++ b/pcsx2/Plugins.cpp @@ -21,6 +21,7 @@ #include "IopCommon.h" #include "GS.h" +#include "HostGui.h" _GSinit GSinit; _GSopen GSopen; @@ -724,7 +725,8 @@ int OpenPlugins(const char* pTitleFilename) }*/ #endif - if( !OpenStatus.CDVD ) + // Don't Open CDVD plugin if directly loading an elf file + if( !OpenStatus.CDVD && g_Startup.BootMode != BootMode_Elf) { //first we need the data if (CDVDnewDiskCB) CDVDnewDiskCB(cdvdNewDiskCB); diff --git a/pcsx2/System.cpp b/pcsx2/System.cpp index 3ed6d4a102..eb06d7905d 100644 --- a/pcsx2/System.cpp +++ b/pcsx2/System.cpp @@ -400,8 +400,11 @@ void SysPrepareExecution( const char* elf_file, bool use_bios ) return; } - if (OpenPlugins(NULL) == -1) - return; + // Don't Open CDVD plugin if directly loading an elf file + if (elf_file) { g_Startup.BootMode = BootMode_Elf; } + if (OpenPlugins(NULL) == -1) { + return; + } if( elf_file == NULL ) { @@ -422,7 +425,6 @@ void SysPrepareExecution( const char* elf_file, bool use_bios ) { // Custom ELF specified (not using CDVD). // Run the BIOS and load the ELF. - loadElfFile( elf_file ); } } diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index 1a2ee09ea7..8500f60665 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -83,10 +83,10 @@ namespace VU0micro if((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); - FreezeMMXRegs(1); + //FreezeMMXRegs(1); runVUrec(VU0.VI[REG_TPC].UL, 5000, 0); FreezeXMMRegs(0); - FreezeMMXRegs(0); + //FreezeMMXRegs(0); } } diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index d282778f4c..4393cf9100 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -142,10 +142,10 @@ namespace VU1micro assert( (VU1.VI[REG_TPC].UL&7) == 0 ); FreezeXMMRegs(1); - FreezeMMXRegs(1); + //FreezeMMXRegs(1); runVUrec(VU1.VI[REG_TPC].UL, 5000, 1); FreezeXMMRegs(0); - FreezeMMXRegs(0); + //FreezeMMXRegs(0); } } #endif diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index b0cddab4ca..eb047e97da 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -138,6 +138,7 @@ microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { } incPC2(-2); } + iPC = endPC; } #define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) @@ -213,10 +214,6 @@ microVUt(void) mVUsetCycles() { incCycles(mVUstall); if (mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1] && mVUregsTemp.VFreg[0]) { // If upper Op && lower Op write to same VF reg mVUinfo |= (mVUregsTemp.r || mVUregsTemp.VI) ? _noWriteVF : _isNOP; // If lower Op doesn't modify anything else, then make it a NOP - tCycles(mVUregsTemp.VF[1].x, mVUregsTemp.VF[0].x) // Use max cycles from each vector - tCycles(mVUregsTemp.VF[1].y, mVUregsTemp.VF[0].y) - tCycles(mVUregsTemp.VF[1].z, mVUregsTemp.VF[0].z) - tCycles(mVUregsTemp.VF[1].w, mVUregsTemp.VF[0].w) } tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].x, mVUregsTemp.VF[0].x); tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].y, mVUregsTemp.VF[0].y); @@ -241,10 +238,12 @@ microVUt(void) mVUdivSet() { if (doDivFlag) { getFlagReg(flagReg1, fsInstance); if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } - MOV32RtoR(gprT1, flagReg1); - AND32ItoR(gprT1, 0xffff0fcf); - OR32MtoR (gprT1, (uptr)&mVU->divFlag); - MOV32RtoR(flagReg1, gprT1); + //MOV32RtoR(gprT1, flagReg1); + //AND32ItoR(gprT1, 0xffff0fcf); + //OR32MtoR (gprT1, (uptr)&mVU->divFlag); + //MOV32RtoR(flagReg1, gprT1); + AND16ItoR(flagReg1, 0x0fcf); + OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } } @@ -252,13 +251,17 @@ microVUt(void) mVUendProgram() { microVU* mVU = mVUx; incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; + if (mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_Q].UL, xmmPQ); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVU->p ? 3 : 2); - SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_P].UL, xmmPQ); + if (vuIndex) { + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVU->p ? 3 : 2); + SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_P].UL, xmmPQ); + } //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) + //AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)µVU0.regs->VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); @@ -270,7 +273,7 @@ microVUt(void) mVUtestCycles() { iPC = mVUstartPC; CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); - mVUendProgram(); + mVUendProgram(); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index a8765d4a8e..a8f4ab45af 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -151,6 +151,7 @@ microVUt(void) mVUcleanUp() { //mVUprint("microVU: VI0 = %x", params mVU->regs->VI[0].UL); mVUcurProg.x86ptr = x86Ptr; mVUcacheCheck(x86Ptr, mVUcurProg.x86start, (uptr)(mVUcurProg.x86end - mVUcurProg.x86start)); + mVU->regs->cycle += mVU->totalCycles - mVU->cycles; } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index dfe5fefb96..c37cb37c9c 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -458,7 +458,7 @@ microVUf(void) mVU_FCGET() { pass2 { mVUallocCFLAGa(gprT1, fvcInstance); AND32ItoR(gprT1, 0xfff); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FCGET vi%02d", _Ft_); } } @@ -491,38 +491,38 @@ microVUf(void) mVU_FCSET() { microVUf(void) mVU_FMAND() { microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Fs_, _Ft_); } + pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Fs_); + mVUallocVIa(gprT2, _Is_); AND16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMAND vi%02d, vi%02d", _Ft_, _Fs_); } } microVUf(void) mVU_FMEQ() { microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Fs_, _Ft_); } + pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Fs_); + mVUallocVIa(gprT2, _Is_); XOR32RtoR(gprT1, gprT2); SUB32ItoR(gprT1, 1); SHR32ItoR(gprT1, 31); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMEQ vi%02d, vi%02d", _Ft_, _Fs_); } } microVUf(void) mVU_FMOR() { microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Fs_, _Ft_); } + pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Fs_); + mVUallocVIa(gprT2, _Is_); OR16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMOR vi%02d, vi%02d", _Ft_, _Fs_); } } @@ -533,35 +533,35 @@ microVUf(void) mVU_FMOR() { microVUf(void) mVU_FSAND() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_Ft_); } + pass1 { mVUanalyzeSflag(_It_); } pass2 { mVUallocSFLAGa(gprT1, fvsInstance); AND16ItoR(gprT1, _Imm12_); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } } microVUf(void) mVU_FSEQ() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_Ft_); } + pass1 { mVUanalyzeSflag(_It_); } pass2 { mVUallocSFLAGa(gprT1, fvsInstance); XOR16ItoR(gprT1, _Imm12_); SUB16ItoR(gprT1, 1); SHR16ItoR(gprT1, 15); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } } microVUf(void) mVU_FSOR() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_Ft_); } + pass1 { mVUanalyzeSflag(_It_); } pass2 { mVUallocSFLAGa(gprT1, fvsInstance); OR16ItoR(gprT1, _Imm12_); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } } @@ -585,95 +585,95 @@ microVUf(void) mVU_FSSET() { microVUf(void) mVU_IADD() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Fd_, _Fs_, _Ft_); } + pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); - if (_Ft_ != _Fs_) { - mVUallocVIa(gprT2, _Ft_); + mVUallocVIa(gprT1, _Is_); + if (_It_ != _Is_) { + mVUallocVIa(gprT2, _It_); ADD16RtoR(gprT1, gprT2); } else ADD16RtoR(gprT1, gprT1); - mVUallocVIb(gprT1, _Fd_); + mVUallocVIb(gprT1, _Id_); } pass3 { mVUlog("IADD vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } microVUf(void) mVU_IADDI() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Fs_, _Ft_); } + pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); ADD16ItoR(gprT1, _Imm5_); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("IADDI vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm5_); } } microVUf(void) mVU_IADDIU() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Fs_, _Ft_); } + pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); ADD16ItoR(gprT1, _Imm15_); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("IADDIU vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm15_); } } microVUf(void) mVU_IAND() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Fd_, _Fs_, _Ft_); } + pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); - if (_Ft_ != _Fs_) { - mVUallocVIa(gprT2, _Ft_); + mVUallocVIa(gprT1, _Is_); + if (_It_ != _Is_) { + mVUallocVIa(gprT2, _It_); AND32RtoR(gprT1, gprT2); } - mVUallocVIb(gprT1, _Fd_); + mVUallocVIb(gprT1, _Id_); } pass3 { mVUlog("IAND vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } microVUf(void) mVU_IOR() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Fd_, _Fs_, _Ft_); } + pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); - if (_Ft_ != _Fs_) { - mVUallocVIa(gprT2, _Ft_); + mVUallocVIa(gprT1, _Is_); + if (_It_ != _Is_) { + mVUallocVIa(gprT2, _It_); OR32RtoR(gprT1, gprT2); } - mVUallocVIb(gprT1, _Fd_); + mVUallocVIb(gprT1, _Id_); } pass3 { mVUlog("IOR vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } microVUf(void) mVU_ISUB() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Fd_, _Fs_, _Ft_); } + pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { - if (_Ft_ != _Fs_) { - mVUallocVIa(gprT1, _Fs_); - mVUallocVIa(gprT2, _Ft_); + if (_It_ != _Is_) { + mVUallocVIa(gprT1, _Is_); + mVUallocVIa(gprT2, _It_); SUB16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _Fd_); + mVUallocVIb(gprT1, _Id_); } - else if (!isMMX(_Fd_)) { + else if (!isMMX(_Id_)) { XOR32RtoR(gprT1, gprT1); - mVUallocVIb(gprT1, _Fd_); + mVUallocVIb(gprT1, _Id_); } - else { PXORRtoR(mmVI(_Fd_), mmVI(_Fd_)); } + else { PXORRtoR(mmVI(_Id_), mmVI(_Id_)); } } pass3 { mVUlog("ISUB vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } microVUf(void) mVU_ISUBIU() { microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Fs_, _Ft_); } + pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); SUB16ItoR(gprT1, _Imm15_); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("ISUBIU vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm15_); } } @@ -684,9 +684,9 @@ microVUf(void) mVU_ISUBIU() { microVUf(void) mVU_MFIR() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Fs_); analyzeReg2(_Ft_); } + pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_); } pass2 { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); MOVSX32R16toR(gprT1, gprT1); SSE2_MOVD_R_to_XMM(xmmT1, gprT1); if (!_XYZW_SS) { mVUunpack_xyzw(xmmT1, xmmT1, 0); } @@ -728,10 +728,10 @@ microVUf(void) mVU_MR32() { microVUf(void) mVU_MTIR() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_Ft_, 1); } + pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->VF[_Fs_].UL[_Fsf_]); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("MTIR vi%02d, vf%02d%s", _Ft_, _Fs_, _Fsf_String); } } @@ -742,18 +742,18 @@ microVUf(void) mVU_MTIR() { microVUf(void) mVU_ILW() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Fs_); analyzeVIreg2(_Ft_, 4); } + pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { - if (!_Fs_) { + if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + getVUmem(_Imm11_) + offsetSS); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } else { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); ADD32ItoR(gprT1, _Imm11_); mVUaddrFix(gprT1); MOVZX32Rm16toR(gprT1, gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } } pass3 { mVUlog("ILW.%s vi%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } @@ -761,17 +761,17 @@ microVUf(void) mVU_ILW() { microVUf(void) mVU_ILWR() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Fs_); analyzeVIreg2(_Ft_, 4); } + pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { - if (!_Fs_) { + if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } else { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); mVUaddrFix(gprT1); MOVZX32Rm16toR(gprT1, gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } } pass3 { mVUlog("ILWR.%s vi%02d, vi%02d", _XYZW_String, _Ft_, _Fs_); } @@ -783,19 +783,19 @@ microVUf(void) mVU_ILWR() { microVUf(void) mVU_ISW() { microVU* mVU = mVUx; - pass1 { analyzeVIreg1(_Fs_); analyzeVIreg1(_Ft_); } + pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { - if (!_Fs_) { + if (!_Is_) { int imm = getVUmem(_Imm11_); - mVUallocVIa(gprT1, _Ft_); + mVUallocVIa(gprT1, _It_); if (_X) MOV32RtoM((uptr)mVU->regs->Mem + imm, gprT1); if (_Y) MOV32RtoM((uptr)mVU->regs->Mem + imm + 4, gprT1); if (_Z) MOV32RtoM((uptr)mVU->regs->Mem + imm + 8, gprT1); if (_W) MOV32RtoM((uptr)mVU->regs->Mem + imm + 12, gprT1); } else { - mVUallocVIa(gprT1, _Fs_); - mVUallocVIa(gprT2, _Ft_); + mVUallocVIa(gprT1, _Is_); + mVUallocVIa(gprT2, _It_); ADD32ItoR(gprT1, _Imm11_); mVUaddrFix(gprT1); if (_X) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem); @@ -809,18 +809,18 @@ microVUf(void) mVU_ISW() { microVUf(void) mVU_ISWR() { microVU* mVU = mVUx; - pass1 { analyzeVIreg1(_Fs_); analyzeVIreg1(_Ft_); } + pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { - if (!_Fs_) { - mVUallocVIa(gprT1, _Ft_); + if (!_Is_) { + mVUallocVIa(gprT1, _It_); if (_X) MOV32RtoM((uptr)mVU->regs->Mem, gprT1); if (_Y) MOV32RtoM((uptr)mVU->regs->Mem+4, gprT1); if (_Z) MOV32RtoM((uptr)mVU->regs->Mem+8, gprT1); if (_W) MOV32RtoM((uptr)mVU->regs->Mem+12, gprT1); } else { - mVUallocVIa(gprT1, _Fs_); - mVUallocVIa(gprT2, _Ft_); + mVUallocVIa(gprT1, _Is_); + mVUallocVIa(gprT2, _It_); mVUaddrFix(gprT1); if (_X) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem); if (_Y) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem+4); @@ -837,14 +837,14 @@ microVUf(void) mVU_ISWR() { microVUf(void) mVU_LQ() { microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Fs_, 0); } + pass1 { mVUanalyzeLQ(_Ft_, _Is_, 0); } pass2 { - if (!_Fs_) { + if (!_Is_) { mVUloadReg(xmmFt, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); ADD32ItoR(gprT1, _Imm11_); mVUaddrFix(gprT1); mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); @@ -856,16 +856,16 @@ microVUf(void) mVU_LQ() { microVUf(void) mVU_LQD() { microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Fs_, 1); } + pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } pass2 { - if (!_Fs_ && !noWriteVF) { + if (!_Is_ && !noWriteVF) { mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Fs_); + mVUallocVIa(gprT1, _Is_); SUB16ItoR(gprT1, 1); - mVUallocVIb(gprT1, _Fs_); + mVUallocVIb(gprT1, _Is_); if (!noWriteVF) { mVUaddrFix(gprT1); mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); @@ -873,19 +873,19 @@ microVUf(void) mVU_LQD() { } } } - pass3 { mVUlog("LQD.%s vf%02d, --vi%02d", _XYZW_String, _Ft_, _Fs_); } + pass3 { mVUlog("LQD.%s vf%02d, --vi%02d", _XYZW_String, _Ft_, _Is_); } } microVUf(void) mVU_LQI() { microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Fs_, 1); } + pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } pass2 { - if (!_Fs_ && !noWriteVF) { + if (!_Is_ && !noWriteVF) { mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa((!noWriteVF) ? gprT1 : gprT2, _Fs_); + mVUallocVIa((!noWriteVF) ? gprT1 : gprT2, _Is_); if (!noWriteVF) { MOV32RtoR(gprT2, gprT1); mVUaddrFix(gprT1); @@ -893,7 +893,7 @@ microVUf(void) mVU_LQI() { mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } ADD16ItoR(gprT2, 1); - mVUallocVIb(gprT2, _Fs_); + mVUallocVIb(gprT2, _Is_); } } pass3 { mVUlog("LQI.%s vf%02d, vi%02d++", _XYZW_String, _Ft_, _Fs_); } @@ -905,14 +905,14 @@ microVUf(void) mVU_LQI() { microVUf(void) mVU_SQ() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _Ft_, 0); } + pass1 { mVUanalyzeSQ(_Fs_, _It_, 0); } pass2 { - if (!_Ft_) { + if (!_It_) { getReg7(xmmFs, _Fs_); mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Ft_); + mVUallocVIa(gprT1, _It_); ADD32ItoR(gprT1, _Imm11_); mVUaddrFix(gprT1); getReg7(xmmFs, _Fs_); @@ -924,16 +924,16 @@ microVUf(void) mVU_SQ() { microVUf(void) mVU_SQD() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _Ft_, 1); } + pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } pass2 { - if (!_Ft_) { + if (!_It_) { getReg7(xmmFs, _Fs_); mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Ft_); + mVUallocVIa(gprT1, _It_); SUB16ItoR(gprT1, 1); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); mVUaddrFix(gprT1); getReg7(xmmFs, _Fs_); mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); @@ -944,20 +944,20 @@ microVUf(void) mVU_SQD() { microVUf(void) mVU_SQI() { microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _Ft_, 1); } + pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } pass2 { - if (!_Ft_) { + if (!_It_) { getReg7(xmmFs, _Fs_); mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Ft_); + mVUallocVIa(gprT1, _It_); MOV32RtoR(gprT2, gprT1); mVUaddrFix(gprT1); getReg7(xmmFs, _Fs_); mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); ADD16ItoR(gprT2, 1); - mVUallocVIb(gprT2, _Ft_); + mVUallocVIb(gprT2, _It_); } } pass3 { mVUlog("SQI.%s vf%02d, vi%02d++", _XYZW_String, _Fs_, _Ft_); } @@ -1056,20 +1056,20 @@ microVUf(void) mVU_WAITQ() { microVUf(void) mVU_XTOP() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg2(_Ft_, 1); } + pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->top); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("XTOP vi%02d", _Ft_); } } microVUf(void) mVU_XITOP() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg2(_Ft_, 1); } + pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->itop); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("XITOP vi%02d", _Ft_); } } @@ -1080,8 +1080,8 @@ microVUf(void) mVU_XITOP() { void __fastcall mVU_XGKICK_(u32 addr) { addr = (addr<<4) & 0x3fff; // Multiply addr by 16 to get real address - u32 *data = (u32*)(microVU1.regs->Mem + (addr&0x3fff)); - u32 size = mtgsThread->PrepDataPacket( GIF_PATH_1, data, (0x4000-(addr&0x3fff)) >> 4); + u32 *data = (u32*)(microVU1.regs->Mem + addr); + u32 size = mtgsThread->PrepDataPacket(GIF_PATH_1, data, (0x4000-addr) >> 4); u8 *pDest = mtgsThread->GetDataPacketPtr(); memcpy_aligned(pDest, microVU1.regs->Mem + addr, size<<4); mtgsThread->SendDataPacket(); @@ -1093,10 +1093,10 @@ void __fastcall mVU_XGKICK__(u32 addr) { microVUf(void) mVU_XGKICK() { microVU* mVU = mVUx; - pass1 { mVUanalyzeXGkick(_Fs_, 4); } + pass1 { mVUanalyzeXGkick(_Is_, 4); } pass2 { mVUprint("XGkick"); - mVUallocVIa(gprT2, _Fs_); // gprT2 = ECX for __fastcall + mVUallocVIa(gprT2, _Is_); // gprT2 = ECX for __fastcall mVUbackupRegs(); if (mtgsThread != NULL) CALLFunc((uptr)mVU_XGKICK_); else CALLFunc((uptr)mVU_XGKICK__); @@ -1109,116 +1109,130 @@ microVUf(void) mVU_XGKICK() { // Branches/Jumps //------------------------------------------------------------------ +#define setBranchA(x, _x_) { \ + pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; return; } mVUbranch = x; } \ + pass2 { mVUbranch = x; } \ +} + microVUf(void) mVU_B() { microVU* mVU = mVUx; - mVUbranch = 1; + setBranchA(1, 0); pass3 { mVUlog("B [%04x]", branchAddr, branchAddr); } } + microVUf(void) mVU_BAL() { microVU* mVU = mVUx; - mVUbranch = 2; - pass1 { analyzeVIreg2(_Ft_, 1); } + setBranchA(2, _It_); + pass1 { analyzeVIreg2(_It_, 1); } pass2 { MOV32ItoR(gprT1, bSaveAddr); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("BAL vi%02d [%04x]", _Ft_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBEQ() { microVU* mVU = mVUx; - mVUbranch = 3; - pass1 { mVUanalyzeBranch2(_Fs_, _Ft_); } + setBranchA(3, 0); + pass1 { mVUanalyzeBranch2(_Is_, _It_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else { mVUallocVIa(gprT2, _Ft_); XOR32RtoR(gprT1, gprT2); } + else { mVUallocVIa(gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBEQ vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBGEZ() { microVU* mVU = mVUx; - mVUbranch = 4; - pass1 { mVUanalyzeBranch1(_Fs_); } + setBranchA(4, 0); + pass1 { mVUanalyzeBranch1(_Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBGEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBGTZ() { microVU* mVU = mVUx; - mVUbranch = 5; - pass1 { mVUanalyzeBranch1(_Fs_); } + setBranchA(5, 0); + pass1 { mVUanalyzeBranch1(_Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBGTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBLEZ() { microVU* mVU = mVUx; - mVUbranch = 6; - pass1 { mVUanalyzeBranch1(_Fs_); } + setBranchA(6, 0); + pass1 { mVUanalyzeBranch1(_Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBLEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBLTZ() { microVU* mVU = mVUx; - mVUbranch = 7; - pass1 { mVUanalyzeBranch1(_Fs_); } + setBranchA(7, 0); + pass1 { mVUanalyzeBranch1(_Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBLTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_IBNE() { microVU* mVU = mVUx; - mVUbranch = 8; - pass1 { mVUanalyzeBranch2(_Fs_, _Ft_); } + setBranchA(8, 0); + pass1 { mVUanalyzeBranch2(_Is_, _It_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else { mVUallocVIa(gprT2, _Ft_); XOR32RtoR(gprT1, gprT2); } + else { mVUallocVIa(gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBNE vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } + microVUf(void) mVU_JR() { microVU* mVU = mVUx; mVUbranch = 9; - pass1 { mVUanalyzeBranch1(_Fs_); } + pass1 { mVUanalyzeBranch1(_Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); SHL32ItoR(gprT1, 3); AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("JR [vi%02d]", _Fs_); } } + microVUf(void) mVU_JALR() { microVU* mVU = mVUx; mVUbranch = 10; - pass1 { mVUanalyzeBranch1(_Fs_); analyzeVIreg2(_Ft_, 1); } + pass1 { mVUanalyzeBranch1(_Is_); analyzeVIreg2(_It_, 1); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Fs_); + else mVUallocVIa(gprT1, _Is_); SHL32ItoR(gprT1, 3); AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); MOV32ItoR(gprT1, bSaveAddr); - mVUallocVIb(gprT1, _Ft_); + mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("JALR vi%02d, [vi%02d]", _Ft_, _Fs_); } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 97980dc8a2..01ea81e287 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -66,9 +66,13 @@ declareAllVariables //------------------------------------------------------------------ // Helper Macros //------------------------------------------------------------------ -#define _Ft_ ((mVU->code >> 16) & 0x1F) // The ft/it part of the instruction register -#define _Fs_ ((mVU->code >> 11) & 0x1F) // The fs/is part of the instruction register -#define _Fd_ ((mVU->code >> 6) & 0x1F) // The fd/id part of the instruction register +#define _Ft_ ((mVU->code >> 16) & 0x1F) // The ft part of the instruction register +#define _Fs_ ((mVU->code >> 11) & 0x1F) // The fs part of the instruction register +#define _Fd_ ((mVU->code >> 6) & 0x1F) // The fd part of the instruction register + +#define _It_ ((mVU->code >> 16) & 0xF) // The it part of the instruction register +#define _Is_ ((mVU->code >> 11) & 0xF) // The is part of the instruction register +#define _Id_ ((mVU->code >> 6) & 0xF) // The id part of the instruction register #define _X ((mVU->code>>24) & 0x1) #define _Y ((mVU->code>>23) & 0x1) From 1cf028ccf50ebadc52f703ce3669a66b9b79a826 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 6 May 2009 02:15:43 +0000 Subject: [PATCH 017/296] Optimized vtlb direct path memory operations using some crafty asm code. Expect 2-3% speedups in most things, and perhaps more in some FMVs. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1137 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/COP0.cpp | 2 +- pcsx2/Hw.h | 20 +- pcsx2/HwRead.cpp | 4 +- pcsx2/IopMem.h | 60 ++-- pcsx2/Memory.cpp | 10 +- pcsx2/ps2/Iop/IopHwRead.cpp | 50 +-- pcsx2/ps2/Iop/IopHwWrite.cpp | 48 +-- pcsx2/vtlb.cpp | 58 ++-- pcsx2/vtlb.h | 8 +- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 - pcsx2/x86/ix86-32/iR5900-32.cpp | 12 +- pcsx2/x86/ix86-32/recVTLB.cpp | 335 +++++++++++++-------- pcsx2/x86/ix86/ix86_inlines.inl | 5 + pcsx2/x86/ix86/ix86_types.h | 2 + pcsx2_suite_2008.sln | 3 + 15 files changed, 344 insertions(+), 277 deletions(-) diff --git a/pcsx2/COP0.cpp b/pcsx2/COP0.cpp index a59f054525..6042250b18 100644 --- a/pcsx2/COP0.cpp +++ b/pcsx2/COP0.cpp @@ -56,7 +56,7 @@ void MapTLB(int i) if (tlb[i].S) { DevCon::WriteLn("OMG SPRAM MAPPING %08X %08X\n",params tlb[i].VPN2,tlb[i].Mask); - vtlb_VMapBuffer(tlb[i].VPN2,psS,0x4000); + vtlb_VMapBuffer(tlb[i].VPN2, psS, 0x4000); } if (tlb[i].VPN2 == 0x70000000) return; //uh uhh right ... diff --git a/pcsx2/Hw.h b/pcsx2/Hw.h index 996cbbc3a8..9a915ce5ba 100644 --- a/pcsx2/Hw.h +++ b/pcsx2/Hw.h @@ -384,8 +384,8 @@ void hwReset(); void hwShutdown(); // hw read functions -extern u8 hwRead8 (u32 mem); -extern u16 hwRead16(u32 mem); +extern mem8_t hwRead8 (u32 mem); +extern mem16_t hwRead16(u32 mem); extern mem32_t __fastcall hwRead32_page_00(u32 mem); extern mem32_t __fastcall hwRead32_page_01(u32 mem); @@ -409,14 +409,14 @@ extern void __fastcall hwRead128_generic(u32 mem, mem128_t *out); extern void hwWrite8 (u32 mem, u8 value); extern void hwWrite16(u32 mem, u16 value); -extern void __fastcall hwWrite32_page_00( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_01( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_02( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_03( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_0B( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_0E( u32 mem, u32 value ); -extern void __fastcall hwWrite32_page_0F( u32 mem, u32 value ); -extern void __fastcall hwWrite32_generic( u32 mem, u32 value ); +extern void __fastcall hwWrite32_page_00( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_01( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_02( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_03( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_0B( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_0E( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_page_0F( u32 mem, mem32_t value ); +extern void __fastcall hwWrite32_generic( u32 mem, mem32_t value ); extern void __fastcall hwWrite64_page_02( u32 mem, const mem64_t* srcval ); extern void __fastcall hwWrite64_page_03( u32 mem, const mem64_t* srcval ); diff --git a/pcsx2/HwRead.cpp b/pcsx2/HwRead.cpp index 2104c8acb8..60a9eb6ab5 100644 --- a/pcsx2/HwRead.cpp +++ b/pcsx2/HwRead.cpp @@ -52,7 +52,7 @@ static __forceinline void IntCHackCheck() ///////////////////////////////////////////////////////////////////////// // Hardware READ 8 bit -__forceinline u8 hwRead8(u32 mem) +__forceinline mem8_t hwRead8(u32 mem) { u8 ret; @@ -117,7 +117,7 @@ __forceinline u8 hwRead8(u32 mem) ///////////////////////////////////////////////////////////////////////// // Hardware READ 16 bit -__forceinline u16 hwRead16(u32 mem) +__forceinline mem16_t hwRead16(u32 mem) { u16 ret; diff --git a/pcsx2/IopMem.h b/pcsx2/IopMem.h index ad618c6b7f..b5a6e9e28e 100644 --- a/pcsx2/IopMem.h +++ b/pcsx2/IopMem.h @@ -98,39 +98,39 @@ void psxRecMemWrite32(); namespace IopMemory { // Sif functions not made yet (will for future Iop improvements): - extern u8 __fastcall SifRead8( u32 iopaddr ); - extern u16 __fastcall SifRead16( u32 iopaddr ); - extern u32 __fastcall SifRead32( u32 iopaddr ); + extern mem8_t __fastcall SifRead8( u32 iopaddr ); + extern mem16_t __fastcall SifRead16( u32 iopaddr ); + extern mem32_t __fastcall SifRead32( u32 iopaddr ); - extern void __fastcall SifWrite8( u32 iopaddr, u8 data ); - extern void __fastcall SifWrite16( u32 iopaddr, u16 data ); - extern void __fastcall SifWrite32( u32 iopaddr, u32 data ); + extern void __fastcall SifWrite8( u32 iopaddr, mem8_t data ); + extern void __fastcall SifWrite16( u32 iopaddr, mem16_t data ); + extern void __fastcall SifWrite32( u32 iopaddr, mem32_t data ); - extern u8 __fastcall iopHwRead8_generic( u32 addr ); - extern u16 __fastcall iopHwRead16_generic( u32 addr ); - extern u32 __fastcall iopHwRead32_generic( u32 addr ); - extern void __fastcall iopHwWrite8_generic( u32 addr, u8 val ); - extern void __fastcall iopHwWrite16_generic( u32 addr, u16 val ); - extern void __fastcall iopHwWrite32_generic( u32 addr, u32 val ); + extern mem8_t __fastcall iopHwRead8_generic( u32 addr ); + extern mem16_t __fastcall iopHwRead16_generic( u32 addr ); + extern mem32_t __fastcall iopHwRead32_generic( u32 addr ); + extern void __fastcall iopHwWrite8_generic( u32 addr, mem8_t val ); + extern void __fastcall iopHwWrite16_generic( u32 addr, mem16_t val ); + extern void __fastcall iopHwWrite32_generic( u32 addr, mem32_t val ); - extern u8 __fastcall iopHwRead8_Page1( u32 iopaddr ); - extern u8 __fastcall iopHwRead8_Page3( u32 iopaddr ); - extern u8 __fastcall iopHwRead8_Page8( u32 iopaddr ); - extern u16 __fastcall iopHwRead16_Page1( u32 iopaddr ); - extern u16 __fastcall iopHwRead16_Page3( u32 iopaddr ); - extern u16 __fastcall iopHwRead16_Page8( u32 iopaddr ); - extern u32 __fastcall iopHwRead32_Page1( u32 iopaddr ); - extern u32 __fastcall iopHwRead32_Page3( u32 iopaddr ); - extern u32 __fastcall iopHwRead32_Page8( u32 iopaddr ); + extern mem8_t __fastcall iopHwRead8_Page1( u32 iopaddr ); + extern mem8_t __fastcall iopHwRead8_Page3( u32 iopaddr ); + extern mem8_t __fastcall iopHwRead8_Page8( u32 iopaddr ); + extern mem16_t __fastcall iopHwRead16_Page1( u32 iopaddr ); + extern mem16_t __fastcall iopHwRead16_Page3( u32 iopaddr ); + extern mem16_t __fastcall iopHwRead16_Page8( u32 iopaddr ); + extern mem32_t __fastcall iopHwRead32_Page1( u32 iopaddr ); + extern mem32_t __fastcall iopHwRead32_Page3( u32 iopaddr ); + extern mem32_t __fastcall iopHwRead32_Page8( u32 iopaddr ); - extern void __fastcall iopHwWrite8_Page1( u32 iopaddr, u8 data ); - extern void __fastcall iopHwWrite8_Page3( u32 iopaddr, u8 data ); - extern void __fastcall iopHwWrite8_Page8( u32 iopaddr, u8 data ); - extern void __fastcall iopHwWrite16_Page1( u32 iopaddr, u16 data ); - extern void __fastcall iopHwWrite16_Page3( u32 iopaddr, u16 data ); - extern void __fastcall iopHwWrite16_Page8( u32 iopaddr, u16 data ); - extern void __fastcall iopHwWrite32_Page1( u32 iopaddr, u32 data ); - extern void __fastcall iopHwWrite32_Page3( u32 iopaddr, u32 data ); - extern void __fastcall iopHwWrite32_Page8( u32 iopaddr, u32 data ); + extern void __fastcall iopHwWrite8_Page1( u32 iopaddr, mem8_t data ); + extern void __fastcall iopHwWrite8_Page3( u32 iopaddr, mem8_t data ); + extern void __fastcall iopHwWrite8_Page8( u32 iopaddr, mem8_t data ); + extern void __fastcall iopHwWrite16_Page1( u32 iopaddr, mem16_t data ); + extern void __fastcall iopHwWrite16_Page3( u32 iopaddr, mem16_t data ); + extern void __fastcall iopHwWrite16_Page8( u32 iopaddr, mem16_t data ); + extern void __fastcall iopHwWrite32_Page1( u32 iopaddr, mem32_t data ); + extern void __fastcall iopHwWrite32_Page3( u32 iopaddr, mem32_t data ); + extern void __fastcall iopHwWrite32_Page8( u32 iopaddr, mem32_t data ); } \ No newline at end of file diff --git a/pcsx2/Memory.cpp b/pcsx2/Memory.cpp index b791ce559a..217e72bc0b 100644 --- a/pcsx2/Memory.cpp +++ b/pcsx2/Memory.cpp @@ -347,7 +347,7 @@ void __fastcall _ext_memRead128(u32 mem, mem128_t *out) } template -void __fastcall _ext_memWrite8 (u32 mem, u8 value) +void __fastcall _ext_memWrite8 (u32 mem, mem8_t value) { switch (p) { case 1: // hwm @@ -367,7 +367,7 @@ void __fastcall _ext_memWrite8 (u32 mem, u8 value) cpuTlbMissW(mem, cpuRegs.branch); } template -void __fastcall _ext_memWrite16(u32 mem, u16 value) +void __fastcall _ext_memWrite16(u32 mem, mem16_t value) { switch (p) { case 1: // hwm @@ -392,7 +392,7 @@ void __fastcall _ext_memWrite16(u32 mem, u16 value) } template -void __fastcall _ext_memWrite32(u32 mem, u32 value) +void __fastcall _ext_memWrite32(u32 mem, mem32_t value) { switch (p) { case 6: // gsm @@ -407,7 +407,7 @@ void __fastcall _ext_memWrite32(u32 mem, u32 value) } template -void __fastcall _ext_memWrite64(u32 mem, const u64* value) +void __fastcall _ext_memWrite64(u32 mem, const mem64_t* value) { /*switch (p) { @@ -423,7 +423,7 @@ void __fastcall _ext_memWrite64(u32 mem, const u64* value) } template -void __fastcall _ext_memWrite128(u32 mem, const u64 *value) +void __fastcall _ext_memWrite128(u32 mem, const mem128_t *value) { /*switch (p) { //case 1: // hwm diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index f9ad47252a..b89c3bfae1 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -25,14 +25,14 @@ using namespace Internal; ////////////////////////////////////////////////////////////////////////////////////////// // -u8 __fastcall iopHwRead8_Page1( u32 addr ) +mem8_t __fastcall iopHwRead8_Page1( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f801xxx: jASSUME( (addr >> 12) == 0x1f801 ); u32 masked_addr = addr & 0x0fff; - u8 ret; // using a return var can be helpful in debugging. + mem8_t ret; // using a return var can be helpful in debugging. switch( masked_addr ) { mcase(HW_SIO_DATA): ret = sioRead8(); break; @@ -71,42 +71,42 @@ u8 __fastcall iopHwRead8_Page1( u32 addr ) return ret; } - PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); return ret; } ////////////////////////////////////////////////////////////////////////////////////////// // -u8 __fastcall iopHwRead8_Page3( u32 addr ) +mem8_t __fastcall iopHwRead8_Page3( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); - u8 ret; + mem8_t ret; if( addr == 0x1f803100 ) // PS/EE/IOP conf related ret = 0x10; // Dram 2M else ret = psxHu8( addr ); - PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); return ret; } ////////////////////////////////////////////////////////////////////////////////////////// // -u8 __fastcall iopHwRead8_Page8( u32 addr ) +mem8_t __fastcall iopHwRead8_Page8( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); - u8 ret; + mem8_t ret; if( addr == HW_SIO2_FIFO ) ret = sio2_fifoOut();//sio2 serial data feed/fifo_out else ret = psxHu8( addr ); - PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); + PSXHW_LOG( "HwRead8 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, psxHu8(addr) ); return ret; } @@ -323,62 +323,62 @@ static __forceinline T _HwRead_16or32_Page1( u32 addr ) ////////////////////////////////////////////////////////////////////////////////////////// // -u16 __fastcall iopHwRead16_Page1( u32 addr ) +mem16_t __fastcall iopHwRead16_Page1( u32 addr ) { - return _HwRead_16or32_Page1( addr ); + return _HwRead_16or32_Page1( addr ); } ////////////////////////////////////////////////////////////////////////////////////////// // -u16 __fastcall iopHwRead16_Page3( u32 addr ) +mem16_t __fastcall iopHwRead16_Page3( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); - u16 ret = psxHu16(addr); - PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); + mem16_t ret = psxHu16(addr); + PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); return ret; } ////////////////////////////////////////////////////////////////////////////////////////// // -u16 __fastcall iopHwRead16_Page8( u32 addr ) +mem16_t __fastcall iopHwRead16_Page8( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); - u16 ret = psxHu16(addr); - PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); + mem16_t ret = psxHu16(addr); + PSXHW_LOG( "HwRead16 from %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, ret ); return ret; } ////////////////////////////////////////////////////////////////////////////////////////// // -u32 __fastcall iopHwRead32_Page1( u32 addr ) +mem32_t __fastcall iopHwRead32_Page1( u32 addr ) { - return _HwRead_16or32_Page1( addr ); + return _HwRead_16or32_Page1( addr ); } ////////////////////////////////////////////////////////////////////////////////////////// // -u32 __fastcall iopHwRead32_Page3( u32 addr ) +mem32_t __fastcall iopHwRead32_Page3( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); - const u32 ret = psxHu32(addr); - PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%08x", _log_GetIopHwName( addr ), addr, ret ); + const mem32_t ret = psxHu32(addr); + PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%08x", _log_GetIopHwName( addr ), addr, ret ); return ret; } ////////////////////////////////////////////////////////////////////////////////////////// // -u32 __fastcall iopHwRead32_Page8( u32 addr ) +mem32_t __fastcall iopHwRead32_Page8( u32 addr ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); u32 masked_addr = addr & 0x0fff; - u32 ret; + mem32_t ret; if( masked_addr >= 0x200 ) { @@ -426,7 +426,7 @@ u32 __fastcall iopHwRead32_Page8( u32 addr ) } else ret = psxHu32(addr); - PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); + PSXHW_LOG( "HwRead32 from %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, ret ); return ret; } diff --git a/pcsx2/ps2/Iop/IopHwWrite.cpp b/pcsx2/ps2/Iop/IopHwWrite.cpp index 24c5bb9ba8..d8e413375d 100644 --- a/pcsx2/ps2/Iop/IopHwWrite.cpp +++ b/pcsx2/ps2/Iop/IopHwWrite.cpp @@ -37,9 +37,9 @@ static __forceinline void _generic_write( u32 addr, T val ) psxHu(addr) = val; } -void __fastcall iopHwWrite8_generic( u32 addr, u8 val ) { _generic_write( addr, val ); } -void __fastcall iopHwWrite16_generic( u32 addr, u16 val ) { _generic_write( addr, val ); } -void __fastcall iopHwWrite32_generic( u32 addr, u32 val ) { _generic_write( addr, val ); } +void __fastcall iopHwWrite8_generic( u32 addr, mem8_t val ) { _generic_write( addr, val ); } +void __fastcall iopHwWrite16_generic( u32 addr, mem16_t val ) { _generic_write( addr, val ); } +void __fastcall iopHwWrite32_generic( u32 addr, mem32_t val ) { _generic_write( addr, val ); } ////////////////////////////////////////////////////////////////////////////////////////// // @@ -53,14 +53,14 @@ static __forceinline T _generic_read( u32 addr ) return ret; } -u8 __fastcall iopHwRead8_generic( u32 addr ) { return _generic_read( addr ); } -u16 __fastcall iopHwRead16_generic( u32 addr ) { return _generic_read( addr ); } -u32 __fastcall iopHwRead32_generic( u32 addr ) { return _generic_read( addr ); } +mem8_t __fastcall iopHwRead8_generic( u32 addr ) { return _generic_read( addr ); } +mem16_t __fastcall iopHwRead16_generic( u32 addr ) { return _generic_read( addr ); } +mem32_t __fastcall iopHwRead32_generic( u32 addr ) { return _generic_read( addr ); } ////////////////////////////////////////////////////////////////////////////////////////// // -void __fastcall iopHwWrite8_Page1( u32 addr, u8 val ) +void __fastcall iopHwWrite8_Page1( u32 addr, mem8_t val ) { // all addresses are assumed to be prefixed with 0x1f801xxx: jASSUME( (addr >> 12) == 0x1f801 ); @@ -103,13 +103,13 @@ void __fastcall iopHwWrite8_Page1( u32 addr, u8 val ) break; } - PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x\n", _log_GetIopHwName(addr), addr, val ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x\n", _log_GetIopHwName(addr), addr, val ); } static char g_pbuf[1024]; static int g_pbufi; -void __fastcall iopHwWrite8_Page3( u32 addr, u8 val ) +void __fastcall iopHwWrite8_Page3( u32 addr, mem8_t val ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); @@ -133,11 +133,11 @@ void __fastcall iopHwWrite8_Page3( u32 addr, u8 val ) } } - PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); psxHu8( addr ) = val; } -void __fastcall iopHwWrite8_Page8( u32 addr, u8 val ) +void __fastcall iopHwWrite8_Page8( u32 addr, mem8_t val ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); @@ -147,7 +147,7 @@ void __fastcall iopHwWrite8_Page8( u32 addr, u8 val ) else psxHu8( addr ) = val; - PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); + PSXHW_LOG( "HwWrite8 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName(addr), addr, psxHu8(addr) ); } @@ -463,43 +463,43 @@ static __forceinline void _HwWrite_16or32_Page1( u32 addr, T val ) ////////////////////////////////////////////////////////////////////////////////////////// // -void __fastcall iopHwWrite16_Page1( u32 addr, u16 val ) +void __fastcall iopHwWrite16_Page1( u32 addr, mem16_t val ) { - _HwWrite_16or32_Page1( addr, val ); + _HwWrite_16or32_Page1( addr, val ); } -void __fastcall iopHwWrite16_Page3( u32 addr, u16 val ) +void __fastcall iopHwWrite16_Page3( u32 addr, mem16_t val ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); psxHu16(addr) = val; - PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } -void __fastcall iopHwWrite16_Page8( u32 addr, u16 val ) +void __fastcall iopHwWrite16_Page8( u32 addr, mem16_t val ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); psxHu16(addr) = val; - PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite16 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } ////////////////////////////////////////////////////////////////////////////////////////// // -void __fastcall iopHwWrite32_Page1( u32 addr, u32 val ) +void __fastcall iopHwWrite32_Page1( u32 addr, mem32_t val ) { - _HwWrite_16or32_Page1( addr, val ); + _HwWrite_16or32_Page1( addr, val ); } -void __fastcall iopHwWrite32_Page3( u32 addr, u32 val ) +void __fastcall iopHwWrite32_Page3( u32 addr, mem32_t val ) { // all addresses are assumed to be prefixed with 0x1f803xxx: jASSUME( (addr >> 12) == 0x1f803 ); psxHu16(addr) = val; - PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%04x", _log_GetIopHwName( addr ), addr, val ); } -void __fastcall iopHwWrite32_Page8( u32 addr, u32 val ) +void __fastcall iopHwWrite32_Page8( u32 addr, mem32_t val ) { // all addresses are assumed to be prefixed with 0x1f808xxx: jASSUME( (addr >> 12) == 0x1f808 ); @@ -539,7 +539,7 @@ void __fastcall iopHwWrite32_Page8( u32 addr, u32 val ) } else psxHu32(addr) = val; - PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, val ); + PSXHW_LOG( "HwWrite32 to %s, addr 0x%08x = 0x%02x", _log_GetIopHwName( addr ), addr, val ); } } diff --git a/pcsx2/vtlb.cpp b/pcsx2/vtlb.cpp index 2a016c61bf..76edac5fb4 100644 --- a/pcsx2/vtlb.cpp +++ b/pcsx2/vtlb.cpp @@ -61,30 +61,6 @@ vtlbHandler UnmappedVirtHandler1; vtlbHandler UnmappedPhyHandler0; vtlbHandler UnmappedPhyHandler1; - /* - __asm - { - mov eax,ecx; - shr ecx,12; - mov ecx,[ecx*4+vmap]; //translate - add ecx,eax; //transform - - js callfunction; //if <0 its invalid ptr :) - - mov eax,[ecx]; - mov [edx],eax; - xor eax,eax; - ret; - -callfunction: - xchg eax,ecx; - shr eax,12; //get the 'ppn' - - //ecx = original addr - //eax = function entry + 0x800000 - //edx = data ptr - jmp [readfunctions8-0x800000+eax]; - }*/ ////////////////////////////////////////////////////////////////////////////////////////// // Interpreter Implementations of VTLB Memory Operations. @@ -516,22 +492,27 @@ void vtlb_Init() //the physical address space can be 'compressed' to just 29 bits.However, to properly handle exceptions //there must be a way to get the full address back.Thats why i use these 2 functions and encode the hi bit directly into em :) - UnmappedVirtHandler0=vtlb_RegisterHandler(vtlbUnmappedVRead8<0>,vtlbUnmappedVRead16<0>,vtlbUnmappedVRead32<0>,vtlbUnmappedVRead64<0>,vtlbUnmappedVRead128<0>, - vtlbUnmappedVWrite8<0>,vtlbUnmappedVWrite16<0>,vtlbUnmappedVWrite32<0>,vtlbUnmappedVWrite64<0>,vtlbUnmappedVWrite128<0>); + UnmappedVirtHandler0 = vtlb_RegisterHandler( + vtlbUnmappedVRead8<0>,vtlbUnmappedVRead16<0>,vtlbUnmappedVRead32<0>,vtlbUnmappedVRead64<0>,vtlbUnmappedVRead128<0>, + vtlbUnmappedVWrite8<0>,vtlbUnmappedVWrite16<0>,vtlbUnmappedVWrite32<0>,vtlbUnmappedVWrite64<0>,vtlbUnmappedVWrite128<0> + ); - UnmappedVirtHandler1=vtlb_RegisterHandler(vtlbUnmappedVRead8<0x80000000>,vtlbUnmappedVRead16<0x80000000>,vtlbUnmappedVRead32<0x80000000>, - vtlbUnmappedVRead64<0x80000000>,vtlbUnmappedVRead128<0x80000000>, - vtlbUnmappedVWrite8<0x80000000>,vtlbUnmappedVWrite16<0x80000000>,vtlbUnmappedVWrite32<0x80000000>, - vtlbUnmappedVWrite64<0x80000000>,vtlbUnmappedVWrite128<0x80000000>); + UnmappedVirtHandler1 = vtlb_RegisterHandler( + vtlbUnmappedVRead8<0x80000000>,vtlbUnmappedVRead16<0x80000000>,vtlbUnmappedVRead32<0x80000000>, vtlbUnmappedVRead64<0x80000000>,vtlbUnmappedVRead128<0x80000000>, + vtlbUnmappedVWrite8<0x80000000>,vtlbUnmappedVWrite16<0x80000000>,vtlbUnmappedVWrite32<0x80000000>, vtlbUnmappedVWrite64<0x80000000>,vtlbUnmappedVWrite128<0x80000000> + ); - UnmappedPhyHandler0=vtlb_RegisterHandler(vtlbUnmappedPRead8<0>,vtlbUnmappedPRead16<0>,vtlbUnmappedPRead32<0>,vtlbUnmappedPRead64<0>,vtlbUnmappedPRead128<0>, - vtlbUnmappedPWrite8<0>,vtlbUnmappedPWrite16<0>,vtlbUnmappedPWrite32<0>,vtlbUnmappedPWrite64<0>,vtlbUnmappedPWrite128<0>); + UnmappedPhyHandler0 = vtlb_RegisterHandler( + vtlbUnmappedPRead8<0>,vtlbUnmappedPRead16<0>,vtlbUnmappedPRead32<0>,vtlbUnmappedPRead64<0>,vtlbUnmappedPRead128<0>, + vtlbUnmappedPWrite8<0>,vtlbUnmappedPWrite16<0>,vtlbUnmappedPWrite32<0>,vtlbUnmappedPWrite64<0>,vtlbUnmappedPWrite128<0> + ); - UnmappedPhyHandler1=vtlb_RegisterHandler(vtlbUnmappedPRead8<0x80000000>,vtlbUnmappedPRead16<0x80000000>,vtlbUnmappedPRead32<0x80000000>, - vtlbUnmappedPRead64<0x80000000>,vtlbUnmappedPRead128<0x80000000>, - vtlbUnmappedPWrite8<0x80000000>,vtlbUnmappedPWrite16<0x80000000>,vtlbUnmappedPWrite32<0x80000000>, - vtlbUnmappedPWrite64<0x80000000>,vtlbUnmappedPWrite128<0x80000000>); - DefaultPhyHandler=vtlb_RegisterHandler(0,0,0,0,0,0,0,0,0,0); + UnmappedPhyHandler1 = vtlb_RegisterHandler( + vtlbUnmappedPRead8<0x80000000>,vtlbUnmappedPRead16<0x80000000>,vtlbUnmappedPRead32<0x80000000>, vtlbUnmappedPRead64<0x80000000>,vtlbUnmappedPRead128<0x80000000>, + vtlbUnmappedPWrite8<0x80000000>,vtlbUnmappedPWrite16<0x80000000>,vtlbUnmappedPWrite32<0x80000000>, vtlbUnmappedPWrite64<0x80000000>,vtlbUnmappedPWrite128<0x80000000> + ); + + DefaultPhyHandler = vtlb_RegisterHandler(0,0,0,0,0,0,0,0,0,0); //done ! @@ -542,6 +523,9 @@ void vtlb_Init() vtlb_VMapUnmap(0,(VTLB_VMAP_ITEMS-1)*VTLB_PAGE_SIZE); //yeah i know, its stupid .. but this code has to be here for now ;p vtlb_VMapUnmap((VTLB_VMAP_ITEMS-1)*VTLB_PAGE_SIZE,VTLB_PAGE_SIZE); + + extern void vtlb_dynarec_init(); + vtlb_dynarec_init(); } ////////////////////////////////////////////////////////////////////////////////////////// diff --git a/pcsx2/vtlb.h b/pcsx2/vtlb.h index b1674fc79d..2e37093489 100644 --- a/pcsx2/vtlb.h +++ b/pcsx2/vtlb.h @@ -48,13 +48,13 @@ extern void vtlb_VMapUnmap(u32 vaddr,u32 sz); //Memory functions -extern u8 __fastcall vtlb_memRead8(u32 mem); -extern u16 __fastcall vtlb_memRead16(u32 mem); +extern mem8_t __fastcall vtlb_memRead8(u32 mem); +extern mem16_t __fastcall vtlb_memRead16(u32 mem); extern u32 __fastcall vtlb_memRead32(u32 mem); extern void __fastcall vtlb_memRead64(u32 mem, u64 *out); extern void __fastcall vtlb_memRead128(u32 mem, u64 *out); -extern void __fastcall vtlb_memWrite8 (u32 mem, u8 value); -extern void __fastcall vtlb_memWrite16(u32 mem, u16 value); +extern void __fastcall vtlb_memWrite8 (u32 mem, mem8_t value); +extern void __fastcall vtlb_memWrite16(u32 mem, mem16_t value); extern void __fastcall vtlb_memWrite32(u32 mem, u32 value); extern void __fastcall vtlb_memWrite64(u32 mem, const u64* value); extern void __fastcall vtlb_memWrite128(u32 mem, const u64* value); diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 8b979de20d..88492c37a1 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -3028,10 +3028,6 @@ RelativePath="..\..\x86\ix86\ix86_types.h" > - - diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 96575543f7..e0e64c5fa7 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -340,14 +340,22 @@ int _flushUnusedConstReg() return 0; } +// ------------------------------------------------------------------------ +// recAllocStackMem -- an optimization trick to write data to a location so that +// recompiled code can reference it later on during execution. +// +// Intended use is for setting up 128 bit SSE immediates, and for compiling a +// struct-worth of const data for calling a function/handler (vtlb). +// u32* recAllocStackMem(int size, int align) { - // write to a temp loc, trick - if( (u32)recStackPtr % align ) recStackPtr += align - ((u32)recStackPtr%align); + recStackPtr += align - ((u32)recStackPtr % align); recStackPtr += size; return (u32*)(recStackPtr-size); } +////////////////////////////////////////////////////////////////////////////////////////// +// static const int REC_CACHEMEM = 0x01000000; static void __fastcall dyna_block_discard(u32 start,u32 sz); diff --git a/pcsx2/x86/ix86-32/recVTLB.cpp b/pcsx2/x86/ix86-32/recVTLB.cpp index b4ff537b44..03e9c06b65 100644 --- a/pcsx2/x86/ix86-32/recVTLB.cpp +++ b/pcsx2/x86/ix86-32/recVTLB.cpp @@ -157,87 +157,204 @@ void iMOV64_Smart( const ModSibBase& destRm, const ModSibBase& srcRm ) */ -////////////////////////////////////////////////////////////////////////////////////////// -// Dynarec Load Implementations - -static void _vtlb_DynGen_DirectRead( u32 bits, bool sign ) +/*template< u32 ReadOrWrite, u32 bitType > +static __naked void _indirectMagicNaked() { - switch( bits ) + enum { - case 8: - if( sign ) - MOVSX32Rm8toR(EAX,ECX); - else - MOVZX32Rm8toR(EAX,ECX); - break; + OffsetRWFT = (128*4*2*bitType) + (128*4*ReadOrWrite) + }; - case 16: - if( sign ) - MOVSX32Rm16toR(EAX,ECX); - else - MOVZX32Rm16toR(EAX,ECX); - break; + __asm + { + movzx eax, al + push ebx // return address - the indirect handler will return to this. + sub ecx, 0x80000000 + sub ecx, eax + + // jump to the indirect handler, which is a __fastcall C++ function. + // [edx is address, ecx is data, and the stack si the return address] + jmp [eax*4 + vtlbdata.RWFT + OffsetRWFT] + } +}*/ - case 32: - MOV32RmtoR(EAX,ECX); - break; +namespace vtlb_private +{ + static void DynGen_DirectRead( u32 bits, bool sign ) + { + switch( bits ) + { + case 8: + if( sign ) + MOVSX32Rm8toR(EAX,ECX); + else + MOVZX32Rm8toR(EAX,ECX); + break; - case 64: - iMOV64_Smart(ptr[edx],ptr[ecx]); - break; + case 16: + if( sign ) + MOVSX32Rm16toR(EAX,ECX); + else + MOVZX32Rm16toR(EAX,ECX); + break; - case 128: - iMOV128_SSE(ptr[edx],ptr[ecx]); - break; + case 32: + MOV32RmtoR(EAX,ECX); + break; - jNO_DEFAULT + case 64: + iMOV64_Smart(ptr[edx],ptr[ecx]); + break; + + case 128: + iMOV128_SSE(ptr[edx],ptr[ecx]); + break; + + jNO_DEFAULT + } + } + + // ------------------------------------------------------------------------ + static void DynGen_IndirectRead( u32 bits ) + { + int szidx; + + switch( bits ) + { + case 8: szidx=0; break; + case 16: szidx=1; break; + case 32: szidx=2; break; + case 64: szidx=3; break; + case 128: szidx=4; break; + jNO_DEFAULT + } + + MOVZX32R8toR(EAX,EAX); + SUB32ItoR(ECX,0x80000000); + SUB32RtoR(ECX,EAX); + xCALL( ptr32[(eax*4) + vtlbdata.RWFT[szidx][0]] ); + } + + // ------------------------------------------------------------------------ + static void DynGen_DirectWrite( u32 bits ) + { + switch(bits) + { + //8 , 16, 32 : data on EDX + case 8: + MOV8RtoRm(ECX,EDX); + break; + case 16: + MOV16RtoRm(ECX,EDX); + break; + case 32: + MOV32RtoRm(ECX,EDX); + break; + + case 64: + iMOV64_Smart(ptr[ecx],ptr[edx]); + break; + + case 128: + iMOV128_SSE(ptr[ecx],ptr[edx]); + break; + } } } // ------------------------------------------------------------------------ -static void _vtlb_DynGen_IndirectRead( u32 bits ) +// allocate one page for our naked indirect dispatcher function. +// this *must* be a full page, since we'll give it execution permission later. +// If it were smaller than a page we'd end up allowing execution rights on some +// other vars additionally (bad!). +// +PCSX2_ALIGNED( 0x1000, static u8 m_IndirectDispatchers[0x1000] ); + +// ------------------------------------------------------------------------ +// mode - 0 for read, 1 for write! +// operandsize - 0 thru 4 represents 8, 16, 32, 64, and 128 bits. +// +static u8* GetIndirectDispatcherPtr( int mode, int operandsize ) +{ + // Each dispatcher is aligned to 64 bytes. The actual dispatchers are only like + // 20-some bytes each, but 64 byte alignment on functions that are called + // more frequently than a hot sex hotline at 1:15am is probably a good thing. + + // 5*64? Because 5 operand types per each mode :D + + return &m_IndirectDispatchers[(mode*(5*64)) + (operandsize*64)]; +} + +// ------------------------------------------------------------------------ +// Generates a JS instruction that targets the appropriate templated instance of +// the vtlb Indirect Dispatcher. +// +static void DynGen_IndirectDispatch( int mode, int bits ) { int szidx; - switch( bits ) { - case 8: szidx=0; break; - case 16: szidx=1; break; - case 32: szidx=2; break; - case 64: szidx=3; break; - case 128: szidx=4; break; - jNO_DEFAULT + case 8: szidx=0; break; + case 16: szidx=1; break; + case 32: szidx=2; break; + case 64: szidx=3; break; + case 128: szidx=4; break; + jNO_DEFAULT; } - - MOVZX32R8toR(EAX,EAX); - SUB32RtoR(ECX,EAX); - //eax=[funct+eax] - MOV32RmSOffsettoR(EAX,EAX,(int)vtlbdata.RWFT[szidx][0],2); - SUB32ItoR(ECX,0x80000000); - CALL32R(EAX); + xJS( GetIndirectDispatcherPtr( mode, szidx ) ); } -// ------------------------------------------------------------------------ -// Recompiled input registers: -// ecx = source addr to read from -// edx = ptr to dest to write to +////////////////////////////////////////////////////////////////////////////////////////// +// One-time initialization procedure. Calling it multiple times shouldn't +// hurt anything tho. +// +void vtlb_dynarec_init() +{ + // In case init gets called multiple times: + HostSys::MemProtect( m_IndirectDispatchers, 0x1000, Protect_ReadWrite, false ); + + // clear the buffer to 0xcc (easier debugging). + memset_8<0xcc,0x1000>( m_IndirectDispatchers ); + + u8* baseptr = m_IndirectDispatchers; + + for( int mode=0; mode<2; ++mode ) + { + for( int bits=0; bits<5; ++bits ) + { + xSetPtr( GetIndirectDispatcherPtr( mode, bits ) ); + + xMOVZX( eax, al ); + xSUB( ecx, 0x80000000 ); + xSUB( ecx, eax ); + + // jump to the indirect handler, which is a __fastcall C++ function. + // [edx is address, ecx is data] + xCALL( ptr32[(eax*4) + vtlbdata.RWFT[bits][mode]] ); + xJMP( ebx ); + } + } + + HostSys::MemProtect( m_IndirectDispatchers, 0x1000, Protect_ReadOnly, true ); +} + +////////////////////////////////////////////////////////////////////////////////////////// +// Dynarec Load Implementations void vtlb_DynGenRead64(u32 bits) { jASSUME( bits == 64 || bits == 128 ); - MOV32RtoR(EAX,ECX); - SHR32ItoR(EAX,VTLB_PAGE_BITS); - MOV32RmSOffsettoR(EAX,EAX,(int)vtlbdata.vmap,2); - ADD32RtoR(ECX,EAX); - xForwardJS8 _fullread; + xMOV( eax, ecx ); + xSHR( eax, VTLB_PAGE_BITS ); + xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); + xMOV( ebx, 0xcdcdcdcd ); + uptr* writeback = ((uptr*)xGetPtr()) - 1; + xADD( ecx, eax ); - _vtlb_DynGen_DirectRead( bits, false ); - xForwardJump8 cont; - - _fullread.SetTarget(); + DynGen_IndirectDispatch( 0, bits ); + DynGen_DirectRead( bits, false ); - _vtlb_DynGen_IndirectRead( bits ); - cont.SetTarget(); + *writeback = (uptr)xGetPtr(); // return target for indirect's call/ret } // ------------------------------------------------------------------------ @@ -248,17 +365,17 @@ void vtlb_DynGenRead32(u32 bits, bool sign) { jASSUME( bits <= 32 ); - MOV32RtoR(EAX,ECX); - SHR32ItoR(EAX,VTLB_PAGE_BITS); - MOV32RmSOffsettoR(EAX,EAX,(int)vtlbdata.vmap,2); - ADD32RtoR(ECX,EAX); - xForwardJS8 _fullread; + xMOV( eax, ecx ); + xSHR( eax, VTLB_PAGE_BITS ); + xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); + xMOV( ebx, 0xcdcdcdcd ); + uptr* writeback = ((uptr*)xGetPtr()) - 1; + xADD( ecx, eax ); - _vtlb_DynGen_DirectRead( bits, sign ); - xForwardJump8 cont; - - _fullread.SetTarget(); - _vtlb_DynGen_IndirectRead( bits ); + DynGen_IndirectDispatch( 0, bits ); + DynGen_DirectRead( bits, sign ); + + *writeback = (uptr)xGetPtr(); // perform sign extension on the result: @@ -276,7 +393,6 @@ void vtlb_DynGenRead32(u32 bits, bool sign) else MOVZX32R16toR(EAX,EAX); } - cont.SetTarget(); } // ------------------------------------------------------------------------ @@ -326,6 +442,7 @@ void vtlb_DynGenRead64_Const( u32 bits, u32 addr_const ) // // TLB lookup is performed in const, with the assumption that the COP0/TLB will clear the // recompiler if the TLB is changed. +// void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) { u32 vmv_ptr = vtlbdata.vmap[addr_const>>VTLB_PAGE_BITS]; @@ -336,16 +453,16 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) { case 8: if( sign ) - MOVSX32M8toR(EAX,ppf); + xMOVSX( eax, ptr8[ppf] ); else - MOVZX32M8toR(EAX,ppf); + xMOVZX( eax, ptr8[ppf] ); break; case 16: if( sign ) - MOVSX32M16toR(EAX,ppf); + xMOVSX( eax, ptr16[ppf] ); else - MOVZX32M16toR(EAX,ppf); + xMOVZX( eax, ptr16[ppf] ); break; case 32: @@ -382,16 +499,16 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) if( bits==8 ) { if( sign ) - MOVSX32R8toR(EAX,EAX); + xMOVSX( eax, al ); else - MOVZX32R8toR(EAX,EAX); + xMOVZX( eax, al ); } else if( bits==16 ) { if( sign ) - MOVSX32R16toR(EAX,EAX); + xMOVSX( eax, ax ); else - MOVZX32R16toR(EAX,EAX); + xMOVZX( eax, ax ); } } } @@ -400,67 +517,19 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) ////////////////////////////////////////////////////////////////////////////////////////// // Dynarec Store Implementations -static void _vtlb_DynGen_DirectWrite( u32 bits ) -{ - switch(bits) - { - //8 , 16, 32 : data on EDX - case 8: - MOV8RtoRm(ECX,EDX); - break; - case 16: - MOV16RtoRm(ECX,EDX); - break; - case 32: - MOV32RtoRm(ECX,EDX); - break; - - case 64: - iMOV64_Smart(ptr[ecx],ptr[edx]); - break; - - case 128: - iMOV128_SSE(ptr[ecx],ptr[edx]); - break; - } -} - -// ------------------------------------------------------------------------ -static void _vtlb_DynGen_IndirectWrite( u32 bits ) -{ - int szidx=0; - switch( bits ) - { - case 8: szidx=0; break; - case 16: szidx=1; break; - case 32: szidx=2; break; - case 64: szidx=3; break; - case 128: szidx=4; break; - } - MOVZX32R8toR(EAX,EAX); - SUB32RtoR(ECX,EAX); - //eax=[funct+eax] - MOV32RmSOffsettoR(EAX,EAX,(int)vtlbdata.RWFT[szidx][1],2); - SUB32ItoR(ECX,0x80000000); - CALL32R(EAX); -} - -// ------------------------------------------------------------------------ void vtlb_DynGenWrite(u32 sz) { - MOV32RtoR(EAX,ECX); - SHR32ItoR(EAX,VTLB_PAGE_BITS); - MOV32RmSOffsettoR(EAX,EAX,(int)vtlbdata.vmap,2); - ADD32RtoR(ECX,EAX); - xForwardJS8 _full; + xMOV( eax, ecx ); + xSHR( eax, VTLB_PAGE_BITS ); + xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); + xMOV( ebx, 0xcdcdcdcd ); + uptr* writeback = ((uptr*)xGetPtr()) - 1; + xADD( ecx, eax ); - _vtlb_DynGen_DirectWrite( sz ); - xForwardJump8 cont; + DynGen_IndirectDispatch( 1, sz ); + DynGen_DirectWrite( sz ); - _full.SetTarget(); - _vtlb_DynGen_IndirectWrite( sz ); - - cont.SetTarget(); + *writeback = (uptr)xGetPtr(); } diff --git a/pcsx2/x86/ix86/ix86_inlines.inl b/pcsx2/x86/ix86/ix86_inlines.inl index 9eedcb0b1a..a88a7881b9 100644 --- a/pcsx2/x86/ix86/ix86_inlines.inl +++ b/pcsx2/x86/ix86/ix86_inlines.inl @@ -68,6 +68,11 @@ namespace x86Emitter return xAddressInfo( *this, right ); } + __forceinline xAddressInfo xAddressReg::operator+( const void* right ) const + { + return xAddressInfo( *this, (s32)right ); + } + __forceinline xAddressInfo xAddressReg::operator*( u32 right ) const { return xAddressInfo( Empty, *this, right ); diff --git a/pcsx2/x86/ix86/ix86_types.h b/pcsx2/x86/ix86/ix86_types.h index 95a6a3b3cd..3dc26befd5 100644 --- a/pcsx2/x86/ix86/ix86_types.h +++ b/pcsx2/x86/ix86/ix86_types.h @@ -317,6 +317,7 @@ __forceinline void xWrite( T val ) xAddressInfo operator+( const xAddressReg& right ) const; xAddressInfo operator+( const xAddressInfo& right ) const; xAddressInfo operator+( s32 right ) const; + xAddressInfo operator+( const void* right ) const; xAddressInfo operator*( u32 factor ) const; xAddressInfo operator<<( u32 shift ) const; @@ -381,6 +382,7 @@ __forceinline void xWrite( T val ) __forceinline xAddressInfo operator+( const xAddressInfo& right ) const { return xAddressInfo( *this ).Add( right ); } __forceinline xAddressInfo operator+( s32 imm ) const { return xAddressInfo( *this ).Add( imm ); } __forceinline xAddressInfo operator-( s32 imm ) const { return xAddressInfo( *this ).Add( -imm ); } + __forceinline xAddressInfo operator+( const void* addr ) const { return xAddressInfo( *this ).Add( (uptr)addr ); } }; extern const xRegisterSSE diff --git a/pcsx2_suite_2008.sln b/pcsx2_suite_2008.sln index fbaf270f39..57f5664f52 100644 --- a/pcsx2_suite_2008.sln +++ b/pcsx2_suite_2008.sln @@ -296,4 +296,7 @@ Global {F38D9DF0-F68D-49D9-B3A0-932E74FB74A0} = {E1828E40-2FBB-48FE-AE7F-5587755DCE0E} {0FADC26C-0E9D-4DD7-84B1-BF4F7754E90C} = {88F517F9-CE1C-4005-9BDF-4481FEB55053} EndGlobalSection + GlobalSection(ExtensibilityGlobals) = postSolution + AMDCaProjectFile = C:\pcsx2dev\pcsx2\CodeAnalyst\pcsx2_suite_2008.caw + EndGlobalSection EndGlobal From 2bf04882e33b6a869d92ac1ff617dfef64a67368 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 6 May 2009 12:37:37 +0000 Subject: [PATCH 018/296] Some code cleanups for the new vtlb optimization (no intentional functional changes -- just deleted unused code and converted all of vtlb to the new emitter syntax. Let's hope I didn't typo anything!) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1138 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IopMem.h | 2 + pcsx2/Memory.h | 47 +------ pcsx2/MemoryTypes.h | 24 ++++ pcsx2/vtlb.h | 10 +- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 + pcsx2/x86/ix86-32/iR5900-32.cpp | 3 +- pcsx2/x86/ix86-32/recVTLB.cpp | 152 ++++++++------------- 7 files changed, 91 insertions(+), 151 deletions(-) create mode 100644 pcsx2/MemoryTypes.h diff --git a/pcsx2/IopMem.h b/pcsx2/IopMem.h index b5a6e9e28e..ab7d9a4a5c 100644 --- a/pcsx2/IopMem.h +++ b/pcsx2/IopMem.h @@ -18,6 +18,8 @@ #pragma once +#include "MemoryTypes.h" + extern u8 *psxM; extern u8 *psxP; extern u8 *psxH; diff --git a/pcsx2/Memory.h b/pcsx2/Memory.h index 18ff8f20fc..3051b73bbb 100644 --- a/pcsx2/Memory.h +++ b/pcsx2/Memory.h @@ -23,24 +23,9 @@ #endif //#define ENABLECACHE +#include "MemoryTypes.h" #include "vtlb.h" -namespace Ps2MemSize -{ - static const uint Base = 0x02000000; // 32 MB main memory! - static const uint Rom = 0x00400000; // 4 MB main rom - static const uint Rom1 = 0x00040000; // DVD player - static const uint Rom2 = 0x00080000; // Chinese rom extension (?) - static const uint ERom = 0x001C0000; // DVD player extensions (?) - static const uint Hardware = 0x00010000; - static const uint Scratch = 0x00004000; - - static const uint IopRam = 0x00200000; // 2MB main ram on the IOP. - static const uint IopHardware = 0x00010000; - - static const uint GSregs = 0x00002000; // 8k for the GS registers and stuff. -} - extern u8 *psM; //32mb Main Ram extern u8 *psR; //4mb rom area extern u8 *psR1; //256kb rom1 area (actually 196kb, but can't mask this) @@ -56,7 +41,6 @@ extern u8 *psS; //0.015 mb, scratch pad #define PS2MEM_EROM psER #define PS2MEM_SCRATCH psS -extern u8 g_RealGSMem[Ps2MemSize::GSregs]; #define PS2MEM_GS g_RealGSMem #define PS2GS_BASE(mem) (g_RealGSMem+(mem&0x13ff)) @@ -174,34 +158,5 @@ extern void mmap_ClearCpuBlock( uint offset ); #define memWrite64 vtlb_memWrite64 #define memWrite128 vtlb_memWrite128 -#define _eeReadConstMem8 0&& -#define _eeReadConstMem16 0&& -#define _eeReadConstMem32 0&& -#define _eeReadConstMem128 0&& -#define _eeWriteConstMem8 0&& -#define _eeWriteConstMem16 0&& -#define _eeWriteConstMem32 0&& -#define _eeWriteConstMem64 0&& -#define _eeWriteConstMem128 0&& -#define _eeMoveMMREGtoR 0&& - -// extra ops -// These allow the old unused const versions of various HW accesses to continue to compile. -// (code left in for reference purposes, but is not needed by Vtlb) -#define _eeWriteConstMem16OP 0&& -#define _eeWriteConstMem32OP 0&& - -#define recMemConstRead8 0&& -#define recMemConstRead16 0&& -#define recMemConstRead32 0&& -#define recMemConstRead64 0&& -#define recMemConstRead128 0&& - -#define recMemConstWrite8 0&& -#define recMemConstWrite16 0&& -#define recMemConstWrite32 0&& -#define recMemConstWrite64 0&& -#define recMemConstWrite128 0&& - extern void loadBiosRom( const char *ext, u8 *dest, long maxSize ); extern u16 ba0R16(u32 mem); diff --git a/pcsx2/MemoryTypes.h b/pcsx2/MemoryTypes.h new file mode 100644 index 0000000000..deceae70f8 --- /dev/null +++ b/pcsx2/MemoryTypes.h @@ -0,0 +1,24 @@ + +#pragma once + +namespace Ps2MemSize +{ + static const uint Base = 0x02000000; // 32 MB main memory! + static const uint Rom = 0x00400000; // 4 MB main rom + static const uint Rom1 = 0x00040000; // DVD player + static const uint Rom2 = 0x00080000; // Chinese rom extension (?) + static const uint ERom = 0x001C0000; // DVD player extensions (?) + static const uint Hardware = 0x00010000; + static const uint Scratch = 0x00004000; + + static const uint IopRam = 0x00200000; // 2MB main ram on the IOP. + static const uint IopHardware = 0x00010000; + + static const uint GSregs = 0x00002000; // 8k for the GS registers and stuff. +} + +typedef u8 mem8_t; +typedef u16 mem16_t; +typedef u32 mem32_t; +typedef u64 mem64_t; +typedef u64 mem128_t; diff --git a/pcsx2/vtlb.h b/pcsx2/vtlb.h index 2e37093489..bcb0489ef5 100644 --- a/pcsx2/vtlb.h +++ b/pcsx2/vtlb.h @@ -1,11 +1,5 @@ -#ifndef _VTLB_H_ -#define _VTLB_H_ -typedef u8 mem8_t; -typedef u16 mem16_t; -typedef u32 mem32_t; -typedef u64 mem64_t; -typedef u64 mem128_t; +#pragma once // Specialized function pointers for each read type typedef mem8_t __fastcall vtlbMemR8FP(u32 addr); @@ -95,5 +89,3 @@ namespace vtlb_private PCSX2_ALIGNED_EXTERN( 64, MapData vtlbdata ); } - -#endif diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 88492c37a1..8f3bee102f 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -2088,6 +2088,10 @@ RelativePath="..\..\Memory.h" > + + diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index e0e64c5fa7..8f717c7fc1 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -344,8 +344,7 @@ int _flushUnusedConstReg() // recAllocStackMem -- an optimization trick to write data to a location so that // recompiled code can reference it later on during execution. // -// Intended use is for setting up 128 bit SSE immediates, and for compiling a -// struct-worth of const data for calling a function/handler (vtlb). +// Intended use is for setting up 64/128 bit SSE immediates, primarily. // u32* recAllocStackMem(int size, int align) { diff --git a/pcsx2/x86/ix86-32/recVTLB.cpp b/pcsx2/x86/ix86-32/recVTLB.cpp index 03e9c06b65..5ce11a9d58 100644 --- a/pcsx2/x86/ix86-32/recVTLB.cpp +++ b/pcsx2/x86/ix86-32/recVTLB.cpp @@ -155,86 +155,61 @@ void iMOV64_Smart( const ModSibBase& destRm, const ModSibBase& srcRm ) cont: ........ - */ - -/*template< u32 ReadOrWrite, u32 bitType > -static __naked void _indirectMagicNaked() -{ - enum - { - OffsetRWFT = (128*4*2*bitType) + (128*4*ReadOrWrite) - }; - - __asm - { - movzx eax, al - push ebx // return address - the indirect handler will return to this. - sub ecx, 0x80000000 - sub ecx, eax - - // jump to the indirect handler, which is a __fastcall C++ function. - // [edx is address, ecx is data, and the stack si the return address] - jmp [eax*4 + vtlbdata.RWFT + OffsetRWFT] - } -}*/ +*/ namespace vtlb_private { + // ------------------------------------------------------------------------ + // Prepares eax, ecx, and, ebx for Direct or Indirect operations. + // Returns the writeback pointer for ebx (return address from indirect handling) + // + static uptr* DynGen_PrepRegs() + { + xMOV( eax, ecx ); + xSHR( eax, VTLB_PAGE_BITS ); + xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); + xMOV( ebx, 0xcdcdcdcd ); + uptr* writeback = ((uptr*)xGetPtr()) - 1; + xADD( ecx, eax ); + + return writeback; + } + + // ------------------------------------------------------------------------ static void DynGen_DirectRead( u32 bits, bool sign ) { switch( bits ) { case 8: if( sign ) - MOVSX32Rm8toR(EAX,ECX); + xMOVSX( eax, ptr8[ecx] ); else - MOVZX32Rm8toR(EAX,ECX); + xMOVZX( eax, ptr8[ecx] ); break; case 16: if( sign ) - MOVSX32Rm16toR(EAX,ECX); + xMOVSX( eax, ptr16[ecx] ); else - MOVZX32Rm16toR(EAX,ECX); + xMOVZX( eax, ptr16[ecx] ); break; case 32: - MOV32RmtoR(EAX,ECX); + xMOV( eax, ptr[ecx] ); break; case 64: - iMOV64_Smart(ptr[edx],ptr[ecx]); + iMOV64_Smart( ptr[edx], ptr[ecx] ); break; case 128: - iMOV128_SSE(ptr[edx],ptr[ecx]); + iMOV128_SSE( ptr[edx], ptr[ecx] ); break; jNO_DEFAULT } } - // ------------------------------------------------------------------------ - static void DynGen_IndirectRead( u32 bits ) - { - int szidx; - - switch( bits ) - { - case 8: szidx=0; break; - case 16: szidx=1; break; - case 32: szidx=2; break; - case 64: szidx=3; break; - case 128: szidx=4; break; - jNO_DEFAULT - } - - MOVZX32R8toR(EAX,EAX); - SUB32ItoR(ECX,0x80000000); - SUB32RtoR(ECX,EAX); - xCALL( ptr32[(eax*4) + vtlbdata.RWFT[szidx][0]] ); - } - // ------------------------------------------------------------------------ static void DynGen_DirectWrite( u32 bits ) { @@ -242,21 +217,23 @@ namespace vtlb_private { //8 , 16, 32 : data on EDX case 8: - MOV8RtoRm(ECX,EDX); + xMOV( ptr[ecx], dl ); break; + case 16: - MOV16RtoRm(ECX,EDX); + xMOV( ptr[ecx], dx ); break; + case 32: - MOV32RtoRm(ECX,EDX); + xMOV( ptr[ecx], edx ); break; case 64: - iMOV64_Smart(ptr[ecx],ptr[edx]); + iMOV64_Smart( ptr[ecx], ptr[edx] ); break; case 128: - iMOV128_SSE(ptr[ecx],ptr[edx]); + iMOV128_SSE( ptr[ecx], ptr[edx] ); break; } } @@ -329,7 +306,7 @@ void vtlb_dynarec_init() xSUB( ecx, eax ); // jump to the indirect handler, which is a __fastcall C++ function. - // [edx is address, ecx is data] + // [ecx is address, edx is data] xCALL( ptr32[(eax*4) + vtlbdata.RWFT[bits][mode]] ); xJMP( ebx ); } @@ -344,12 +321,7 @@ void vtlb_DynGenRead64(u32 bits) { jASSUME( bits == 64 || bits == 128 ); - xMOV( eax, ecx ); - xSHR( eax, VTLB_PAGE_BITS ); - xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); - xMOV( ebx, 0xcdcdcdcd ); - uptr* writeback = ((uptr*)xGetPtr()) - 1; - xADD( ecx, eax ); + uptr* writeback = DynGen_PrepRegs(); DynGen_IndirectDispatch( 0, bits ); DynGen_DirectRead( bits, false ); @@ -365,33 +337,28 @@ void vtlb_DynGenRead32(u32 bits, bool sign) { jASSUME( bits <= 32 ); - xMOV( eax, ecx ); - xSHR( eax, VTLB_PAGE_BITS ); - xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); - xMOV( ebx, 0xcdcdcdcd ); - uptr* writeback = ((uptr*)xGetPtr()) - 1; - xADD( ecx, eax ); + uptr* writeback = DynGen_PrepRegs(); DynGen_IndirectDispatch( 0, bits ); DynGen_DirectRead( bits, sign ); - + *writeback = (uptr)xGetPtr(); // perform sign extension on the result: - if( bits==8 ) + if( bits == 8 ) { if( sign ) - MOVSX32R8toR(EAX,EAX); + xMOVSX( eax, al ); else - MOVZX32R8toR(EAX,EAX); + xMOVZX( eax, al ); } - else if( bits==16 ) + else if( bits == 16 ) { if( sign ) - MOVSX32R16toR(EAX,EAX); + xMOVSX( eax, ax ); else - MOVZX32R16toR(EAX,EAX); + xMOVZX( eax, ax ); } } @@ -407,11 +374,11 @@ void vtlb_DynGenRead64_Const( u32 bits, u32 addr_const ) switch( bits ) { case 64: - iMOV64_Smart(ptr[edx],ptr[ppf]); + iMOV64_Smart( ptr[edx], ptr[ppf] ); break; case 128: - iMOV128_SSE(ptr[edx],ptr[ppf]); + iMOV128_SSE( ptr[edx], ptr[ppf] ); break; jNO_DEFAULT @@ -430,8 +397,8 @@ void vtlb_DynGenRead64_Const( u32 bits, u32 addr_const ) case 128: szidx=4; break; } - MOV32ItoR( ECX, paddr ); - CALLFunc( (int)vtlbdata.RWFT[szidx][0][handler] ); + xMOV( ecx, paddr ); + xCALL( vtlbdata.RWFT[szidx][0][handler] ); } } @@ -466,7 +433,7 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) break; case 32: - MOV32MtoR(EAX,ppf); + xMOV( eax, ptr[ppf] ); break; } } @@ -487,12 +454,12 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) // Shortcut for the INTC_STAT register, which many games like to spin on heavily. if( (bits == 32) && !Config.Hacks.INTCSTATSlow && (paddr == INTC_STAT) ) { - MOV32MtoR( EAX, (uptr)&psHu32( INTC_STAT ) ); + xMOV( eax, &psHu32( INTC_STAT ) ); } else { - MOV32ItoR( ECX, paddr ); - CALLFunc( (int)vtlbdata.RWFT[szidx][0][handler] ); + xMOV( ecx, paddr ); + xCALL( vtlbdata.RWFT[szidx][0][handler] ); // perform sign extension on the result: @@ -519,12 +486,7 @@ void vtlb_DynGenRead32_Const( u32 bits, bool sign, u32 addr_const ) void vtlb_DynGenWrite(u32 sz) { - xMOV( eax, ecx ); - xSHR( eax, VTLB_PAGE_BITS ); - xMOV( eax, ptr[(eax*4) + vtlbdata.vmap] ); - xMOV( ebx, 0xcdcdcdcd ); - uptr* writeback = ((uptr*)xGetPtr()) - 1; - xADD( ecx, eax ); + uptr* writeback = DynGen_PrepRegs(); DynGen_IndirectDispatch( 1, sz ); DynGen_DirectWrite( sz ); @@ -547,13 +509,15 @@ void vtlb_DynGenWrite_Const( u32 bits, u32 addr_const ) { //8 , 16, 32 : data on EDX case 8: - MOV8RtoM(ppf,EDX); + xMOV( ptr[ppf], dl ); break; + case 16: - MOV16RtoM(ppf,EDX); + xMOV( ptr[ppf], dx ); break; + case 32: - MOV32RtoM(ppf,EDX); + xMOV( ptr[ppf], edx ); break; case 64: @@ -582,8 +546,8 @@ void vtlb_DynGenWrite_Const( u32 bits, u32 addr_const ) case 128: szidx=4; break; } - MOV32ItoR( ECX, paddr ); - CALLFunc( (int)vtlbdata.RWFT[szidx][1][handler] ); + xMOV( ecx, paddr ); + xCALL( vtlbdata.RWFT[szidx][1][handler] ); } } From 9541e502c2b6c189ee8c9e9b0e0aa5a60cd7229d Mon Sep 17 00:00:00 2001 From: refraction Date: Wed, 6 May 2009 23:51:27 +0000 Subject: [PATCH 019/296] A few minor counter changes to prevent scenarios where branch targets could be missed on counter resets/target changes, also a situation where the vsync counter could cause repetative branch tests to be triggered in quick succession on cycle counter overflow git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1140 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Counters.cpp | 50 +++++++++++++++++++++++++++++++++++++------ pcsx2/IopCounters.cpp | 35 ++++++++++++++++++++++++------ pcsx2/IopDma.cpp | 6 ++++++ pcsx2/R3000A.cpp | 7 ++++-- 4 files changed, 82 insertions(+), 16 deletions(-) diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index 89a5cf8bd6..c75d5f7e15 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -82,16 +82,39 @@ static __forceinline void _rcntSet( int cntidx ) c = ((0x10000 - counter.count) * counter.rate) - (cpuRegs.cycle - counter.sCycleT); c += cpuRegs.cycle - nextsCounter; // adjust for time passed since last rcntUpdate(); - if (c < nextCounter) nextCounter = c; - + if (c < nextCounter) + { + nextCounter = c; + + if((g_nextBranchCycle - nextsCounter) > (u32)nextCounter) //Need to update on counter resets/target changes + { + g_nextBranchCycle = nextsCounter + nextCounter; + } + } + // Ignore target diff if target is currently disabled. // (the overflow is all we care about since it goes first, and then the // target will be turned on afterward). - if( counter.target & EECNT_FUTURE_TARGET ) return; - c = ((counter.target - counter.count) * counter.rate) - (cpuRegs.cycle - counter.sCycleT); - c += cpuRegs.cycle - nextsCounter; // adjust for time passed since last rcntUpdate(); - if (c < nextCounter) nextCounter = c; + if( counter.target & EECNT_FUTURE_TARGET ) + { + return; + } + else + { + c = ((counter.target - counter.count) * counter.rate) - (cpuRegs.cycle - counter.sCycleT); + c += cpuRegs.cycle - nextsCounter; // adjust for time passed since last rcntUpdate(); + if (c < nextCounter) + { + nextCounter = c; + + if((g_nextBranchCycle - nextsCounter) > (u32)nextCounter) //Need to update on counter resets/target changes + { + g_nextBranchCycle = nextsCounter + nextCounter; + } + } + } + //cpuSetNextBranch( nextsCounter, nextCounter ); } @@ -100,7 +123,7 @@ static __forceinline void cpuRcntSet() int i; nextsCounter = cpuRegs.cycle; - nextCounter = (vsyncCounter.sCycle + vsyncCounter.CycleT) - cpuRegs.cycle; + nextCounter = vsyncCounter.CycleT - (cpuRegs.cycle - vsyncCounter.sCycle); for (i = 0; i < 4; i++) _rcntSet( i ); @@ -714,6 +737,19 @@ __forceinline void rcntWtarget(int index, u32 value) // If the target is behind the current count, set it up so that the counter must // overflow first before the target fires: + if(counters[index].mode.IsCounting) { + if(counters[index].mode.ClockSource != 0x3) { + + u32 change = cpuRegs.cycle - counters[index].sCycleT; + if( change > 0 ) + { + counters[index].count += change / counters[index].rate; + change -= (change / counters[index].rate) * counters[index].rate; + counters[index].sCycleT = cpuRegs.cycle - change; + } + } + } + if( counters[index].target <= rcntCycle(index) ) counters[index].target |= EECNT_FUTURE_TARGET; diff --git a/pcsx2/IopCounters.cpp b/pcsx2/IopCounters.cpp index 9e169100f7..c5f5034908 100644 --- a/pcsx2/IopCounters.cpp +++ b/pcsx2/IopCounters.cpp @@ -91,14 +91,32 @@ static void _rcntSet( int cntidx ) c = (u64)((overflowCap - counter.count) * counter.rate) - (psxRegs.cycle - counter.sCycleT); c += psxRegs.cycle - psxNextsCounter; // adjust for time passed since last rcntUpdate(); - if(c < (u64)psxNextCounter) psxNextCounter = (u32)c; + + if(c < (u64)psxNextCounter) + { + psxNextCounter = (u32)c; + + if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) //Need to update on counter resets/target changes + { + g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; + } + } //if((counter.mode & 0x10) == 0 || psxCounters[i].target > 0xffff) continue; if( counter.target & IOPCNT_FUTURE_TARGET ) return; - c = (s64)((counter.target - counter.count) * counter.rate) - (psxRegs.cycle - counter.sCycleT); - c += psxRegs.cycle - psxNextsCounter; // adjust for time passed since last rcntUpdate(); - if(c < (u64)psxNextCounter) psxNextCounter = (u32)c; + c = (s64)((counter.target - counter.count) * counter.rate) - (psxRegs.cycle - counter.sCycleT); + c += psxRegs.cycle - psxNextsCounter; // adjust for time passed since last rcntUpdate(); + + if(c < (u64)psxNextCounter) + { + psxNextCounter = (u32)c; + + if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) //Need to update on counter resets/target changes + { + g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; + } + } } @@ -383,6 +401,11 @@ void psxRcntUpdate() int i; //u32 change = 0; + g_psxNextBranchCycle = psxRegs.cycle + 32; + + psxNextCounter = 0x7fffffff; + psxNextsCounter = psxRegs.cycle; + for (i=0; i<=5; i++) { s32 change = psxRegs.cycle - psxCounters[i].sCycleT; @@ -424,9 +447,7 @@ void psxRcntUpdate() //if( psxCounters[i].count >= psxCounters[i].target ) _rcntTestTarget( i ); } - - psxNextCounter = 0xffffff; - psxNextsCounter = psxRegs.cycle; + if(SPU2async) { diff --git a/pcsx2/IopDma.cpp b/pcsx2/IopDma.cpp index 1af6cef622..7138ca8ccf 100644 --- a/pcsx2/IopDma.cpp +++ b/pcsx2/IopDma.cpp @@ -54,6 +54,12 @@ static void __fastcall psxDmaGeneric(u32 madr, u32 bcr, u32 chcr, u32 spuCore, _ psxNextsCounter = psxRegs.cycle; if (psxCounters[6].CycleT < psxNextCounter) psxNextCounter = psxCounters[6].CycleT; + + if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) + { + DevCon::Notice("SPU2async Setting new counter branch, old %x new %x ((%x - %x = %x) > %x delta)", params g_psxNextBranchCycle, psxNextsCounter + psxNextCounter, g_psxNextBranchCycle, psxNextsCounter, (g_psxNextBranchCycle - psxNextsCounter), psxNextCounter); + g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; + } } switch (chcr) diff --git a/pcsx2/R3000A.cpp b/pcsx2/R3000A.cpp index c821733549..368d5ad703 100644 --- a/pcsx2/R3000A.cpp +++ b/pcsx2/R3000A.cpp @@ -231,10 +231,13 @@ __releaseinline void psxBranchTest() psxRcntUpdate(); iopBranchAction = true; } - + else + { // start the next branch at the next counter event by default // the interrupt code below will assign nearer branches if needed. - g_psxNextBranchCycle = psxNextsCounter+psxNextCounter; + g_psxNextBranchCycle = psxNextsCounter+psxNextCounter; + } + if (psxRegs.interrupt) { From e6b9275c13c913fafc52a1b8c7ae16ab0e1b1959 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 7 May 2009 01:57:13 +0000 Subject: [PATCH 020/296] microVU: various minor fixes that effected big stuff... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1141 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 6 +++--- pcsx2/x86/microVU_Compile.inl | 6 +++--- pcsx2/x86/microVU_Execute.inl | 11 ++++------- pcsx2/x86/microVU_Misc.h | 5 +++-- 4 files changed, 13 insertions(+), 15 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 9dcbc0d1f5..ecf728a65e 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -17,8 +17,8 @@ */ #pragma once -#define mVUdebug // Prints Extra Info to Console -#define mVUlogProg // Dumps MicroPrograms into microVU0.txt/microVU1.txt +//#define mVUdebug // Prints Extra Info to Console +//#define mVUlogProg // Dumps MicroPrograms into microVU0.txt/microVU1.txt #include "Common.h" #include "VU.h" #include "GS.h" @@ -74,7 +74,7 @@ struct microProgram { microAllocInfo allocInfo; }; -#define mMaxProg 16 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) +#define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) template struct microProgManager { microProgram prog[mMaxProg]; // Store MicroPrograms in memory diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index eb047e97da..6bf8f65dc4 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -30,7 +30,7 @@ incPC2(1); \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ - if (bBlock) { nJMPcc((uptr)pBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ + if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ else { ajmp = JMPcc((uptr)0); } \ break @@ -347,12 +347,12 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { microBlock* bBlock = NULL; u32* ajmp = 0; switch (mVUbranch) { - case 3: branchCase(JZ32, JNZ32); // IBEQ + case 3: branchCase(JE32, JNE32); // IBEQ case 4: branchCase(JGE32, JNGE32); // IBGEZ case 5: branchCase(JG32, JNG32); // IBGTZ case 6: branchCase(JLE32, JNLE32); // IBLEQ case 7: branchCase(JL32, JNL32); // IBLTZ - case 8: branchCase(JNZ32, JZ32); // IBNEQ + case 8: branchCase(JNE32, JE32); // IBNEQ case 1: case 2: // B/BAL mVUprint("mVUcompile B/BAL"); diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index a8f4ab45af..a96647fbbd 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -82,10 +82,6 @@ microVUt(void) mVUdispatcherB() { microVU* mVU = mVUx; mVU->exitFunct = x86Ptr; - // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. - if (!vuIndex) { CALLFunc((uptr)mVUcleanUpVU0); } - else { CALLFunc((uptr)mVUcleanUpVU1); } - // Load EE's MXCSR state eeMXCSR = g_sseMXCSR; SSE_LDMXCSR((uptr)&eeMXCSR); @@ -98,15 +94,16 @@ microVUt(void) mVUdispatcherB() { MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprF0); - for (int i = 0; i < 8; i++) { - - } for (int i = 1; i < 16; i++) { if (isMMX(i)) { MOVDMMXtoM((uptr)&mVU->regs->VI[i].UL, mmVI(i)); } } SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); + // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. + if (!vuIndex) { CALLFunc((uptr)mVUcleanUpVU0); } + else { CALLFunc((uptr)mVUcleanUpVU1); } + // Restore cpu state POP32R(EDI); POP32R(ESI); diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 01ea81e287..07cef16dbc 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -195,7 +195,6 @@ declareAllVariables #define _fvmInstance (3<<16) // Mac Read Instance (at T-stage for lower instruction) #define _fvsInstance (3<<18) // Status Read Instance (at T-stage for lower instruction) #define _fvcInstance (3<<20) // Clip Read Instance (at T-stage for lower instruction) -#define _noWriteVF (1<<21) // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) #define _backupVI (1<<22) // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) #define _memReadIs (1<<23) // Read Is (VI reg) from memory (used by branches) #define _memReadIt (1<<24) // Read If (VI reg) from memory (used by branches) @@ -204,6 +203,8 @@ declareAllVariables #define _isFSSET (1<<27) // Cur Instruction is FSSET #define _doDivFlag (1<<28) // Transfer Div flag to Status Flag #define _doClip (1<<29) +#define _noWriteVF (1<<30) // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) + #define isNOP (mVUinfo & (1<<0)) #define isBranch (mVUinfo & (1<<1)) @@ -225,7 +226,6 @@ declareAllVariables #define fvmInstance ((mVUinfo >> 16) & 3) #define fvsInstance ((mVUinfo >> 18) & 3) #define fvcInstance ((mVUinfo >> 20) & 3) -#define noWriteVF (mVUinfo & (1<<21)) #define backupVI (mVUinfo & (1<<22)) #define memReadIs (mVUinfo & (1<<23)) #define memReadIt (mVUinfo & (1<<24)) @@ -234,6 +234,7 @@ declareAllVariables #define isFSSET (mVUinfo & (1<<27)) #define doDivFlag (mVUinfo & (1<<28)) #define doClip (mVUinfo & (1<<29)) +#define noWriteVF (mVUinfo & (1<<30)) #define isMMX(_VIreg_) 0//(_VIreg_ >= 1 && _VIreg_ <=8) #define mmVI(_VIreg_) (_VIreg_ - 1) From 3202a42578e26c24a2b4e79f345e3324923bf9d5 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 7 May 2009 06:32:10 +0000 Subject: [PATCH 021/296] GSdx: Reworked the sw renderer texture cache a bit, at page level its correctness was questionable if the base address was not page aligned, so now keeping track of blocks (32x more), may be slower or faster by a few percent (number of blocks vs. finer resolution). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1142 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSClut.cpp | 12 +- plugins/GSdx/GSLocalMemory.cpp | 429 ++++++++++++++++++++++-------- plugins/GSdx/GSLocalMemory.h | 184 ++++++------- plugins/GSdx/GSRendererSW.h | 41 ++- plugins/GSdx/GSTextureCacheSW.cpp | 273 +++++++++---------- plugins/GSdx/GSTextureCacheSW.h | 28 +- 6 files changed, 554 insertions(+), 413 deletions(-) diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index d850ac266f..ed8fad82b2 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -128,7 +128,7 @@ void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TE { ASSERT(TEX0.CSA == 0); - WriteCLUT_T32_I8_CSM1(&m_mem->m_vm32[m_mem->BlockAddress32(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T32_I8_CSM1((DWORD*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) @@ -137,31 +137,31 @@ void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TE GSVector4i dummy; // this just forces stack alignment and enables inlining the next call - WriteCLUT_T32_I4_CSM1(&m_mem->m_vm32[m_mem->BlockAddress32(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T32_I4_CSM1((DWORD*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { ASSERT(TEX0.CSA < 16); - WriteCLUT_T16_I8_CSM1(&m_mem->m_vm16[m_mem->BlockAddress16(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I8_CSM1((WORD*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { ASSERT(TEX0.CSA < 32); - WriteCLUT_T16_I4_CSM1(&m_mem->m_vm16[m_mem->BlockAddress16(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I4_CSM1((WORD*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16S_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WriteCLUT_T16_I8_CSM1(&m_mem->m_vm16[m_mem->BlockAddress16S(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I8_CSM1((WORD*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16S_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WriteCLUT_T16_I4_CSM1(&m_mem->m_vm16[m_mem->BlockAddress16S(0, 0, TEX0.CBP, 1)], m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I4_CSM1((WORD*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } template void GSClut::WriteCLUT32_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index b86c6ce8b8..1ef3bb5485 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -61,6 +61,15 @@ int GSLocalMemory::rowOffset16SZ[2048]; int GSLocalMemory::rowOffset8[2][2048]; int GSLocalMemory::rowOffset4[2][2048]; +int GSLocalMemory::blockOffset32[256]; +int GSLocalMemory::blockOffset32Z[256]; +int GSLocalMemory::blockOffset16[256]; +int GSLocalMemory::blockOffset16S[256]; +int GSLocalMemory::blockOffset16Z[256]; +int GSLocalMemory::blockOffset16SZ[256]; +int GSLocalMemory::blockOffset8[256]; +int GSLocalMemory::blockOffset4[256]; + // DWORD GSLocalMemory::m_xtbl[1024]; @@ -138,22 +147,60 @@ GSLocalMemory::GSLocalMemory() for(int x = 0; x < countof(rowOffset8[0]); x++) { - rowOffset8[0][x] = (int)PixelAddress8(x, 0, 0, 32) - (int)PixelAddress8(0, 0, 0, 32), + rowOffset8[0][x] = (int)PixelAddress8(x, 0, 0, 32) - (int)PixelAddress8(0, 0, 0, 32); rowOffset8[1][x] = (int)PixelAddress8(x, 2, 0, 32) - (int)PixelAddress8(0, 2, 0, 32); } for(int x = 0; x < countof(rowOffset4[0]); x++) { - rowOffset4[0][x] = (int)PixelAddress4(x, 0, 0, 32) - (int)PixelAddress4(0, 0, 0, 32), + rowOffset4[0][x] = (int)PixelAddress4(x, 0, 0, 32) - (int)PixelAddress4(0, 0, 0, 32); rowOffset4[1][x] = (int)PixelAddress4(x, 2, 0, 32) - (int)PixelAddress4(0, 2, 0, 32); } + for(int x = 0; x < countof(blockOffset32); x++) + { + blockOffset32[x] = (int)BlockNumber32(x << 3, 0, 0, 32) - (int)BlockNumber32(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset32Z); x++) + { + blockOffset32Z[x] = (int)BlockNumber32Z(x << 3, 0, 0, 32) - (int)BlockNumber32Z(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset16); x++) + { + blockOffset16[x] = (int)BlockNumber16(x << 3, 0, 0, 32) - (int)BlockNumber16(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset16S); x++) + { + blockOffset16S[x] = (int)BlockNumber16S(x << 3, 0, 0, 32) - (int)BlockNumber16S(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset16Z); x++) + { + blockOffset16Z[x] = (int)BlockNumber16Z(x << 3, 0, 0, 32) - (int)BlockNumber16Z(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset16SZ); x++) + { + blockOffset16SZ[x] = (int)BlockNumber16SZ(x << 3, 0, 0, 32) - (int)BlockNumber16SZ(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset8); x++) + { + blockOffset8[x] = (int)BlockNumber8(x << 3, 0, 0, 32) - (int)BlockNumber8(0, 0, 0, 32); + } + + for(int x = 0; x < countof(blockOffset4); x++) + { + blockOffset4[x] = (int)BlockNumber4(x << 3, 0, 0, 32) - (int)BlockNumber4(0, 0, 0, 32); + } + for(int i = 0; i < countof(m_psm); i++) { m_psm[i].pa = &GSLocalMemory::PixelAddress32; - m_psm[i].ba = &GSLocalMemory::BlockAddress32; - m_psm[i].pga = &GSLocalMemory::PageAddress32; - m_psm[i].pgn = &GSLocalMemory::PageNumber32; + m_psm[i].bn = &GSLocalMemory::BlockNumber32; m_psm[i].rp = &GSLocalMemory::ReadPixel32; m_psm[i].rpa = &GSLocalMemory::ReadPixel32; m_psm[i].wp = &GSLocalMemory::WritePixel32; @@ -167,11 +214,14 @@ GSLocalMemory::GSLocalMemory() m_psm[i].rtx = &GSLocalMemory::ReadTexture32; m_psm[i].rtxNP = &GSLocalMemory::ReadTexture32; m_psm[i].rtxP = &GSLocalMemory::ReadTexture32; + m_psm[i].rtxb = &GSLocalMemory::ReadTextureBlock32; + m_psm[i].rtxbP = &GSLocalMemory::ReadTextureBlock32; m_psm[i].bpp = m_psm[i].trbpp = 32; m_psm[i].pal = 0; m_psm[i].bs = CSize(8, 8); m_psm[i].pgs = CSize(64, 32); for(int j = 0; j < 8; j++) m_psm[i].rowOffset[j] = rowOffset32; + m_psm[i].blockOffset = blockOffset32; } m_psm[PSM_PSMCT16].pa = &GSLocalMemory::PixelAddress16; @@ -183,28 +233,14 @@ GSLocalMemory::GSLocalMemory() m_psm[PSM_PSMZ16].pa = &GSLocalMemory::PixelAddress16Z; m_psm[PSM_PSMZ16S].pa = &GSLocalMemory::PixelAddress16SZ; - m_psm[PSM_PSMCT16].ba = &GSLocalMemory::BlockAddress16; - m_psm[PSM_PSMCT16S].ba = &GSLocalMemory::BlockAddress16S; - m_psm[PSM_PSMT8].ba = &GSLocalMemory::BlockAddress8; - m_psm[PSM_PSMT4].ba = &GSLocalMemory::BlockAddress4; - m_psm[PSM_PSMZ32].ba = &GSLocalMemory::BlockAddress32Z; - m_psm[PSM_PSMZ24].ba = &GSLocalMemory::BlockAddress32Z; - m_psm[PSM_PSMZ16].ba = &GSLocalMemory::BlockAddress16Z; - m_psm[PSM_PSMZ16S].ba = &GSLocalMemory::BlockAddress16SZ; - - m_psm[PSM_PSMCT16].pga = &GSLocalMemory::PageAddress16; - m_psm[PSM_PSMCT16S].pga = &GSLocalMemory::PageAddress16; - m_psm[PSM_PSMZ16].pga = &GSLocalMemory::PageAddress16; - m_psm[PSM_PSMZ16S].pga = &GSLocalMemory::PageAddress16; - m_psm[PSM_PSMT8].pga = &GSLocalMemory::PageAddress8; - m_psm[PSM_PSMT4].pga = &GSLocalMemory::PageAddress4; - - m_psm[PSM_PSMCT16].pgn = &GSLocalMemory::PageNumber16; - m_psm[PSM_PSMCT16S].pgn = &GSLocalMemory::PageNumber16; - m_psm[PSM_PSMZ16].pgn = &GSLocalMemory::PageNumber16; - m_psm[PSM_PSMZ16S].pgn = &GSLocalMemory::PageNumber16; - m_psm[PSM_PSMT8].pgn = &GSLocalMemory::PageNumber8; - m_psm[PSM_PSMT4].pgn = &GSLocalMemory::PageNumber4; + m_psm[PSM_PSMCT16].bn = &GSLocalMemory::BlockNumber16; + m_psm[PSM_PSMCT16S].bn = &GSLocalMemory::BlockNumber16S; + m_psm[PSM_PSMT8].bn = &GSLocalMemory::BlockNumber8; + m_psm[PSM_PSMT4].bn = &GSLocalMemory::BlockNumber4; + m_psm[PSM_PSMZ32].bn = &GSLocalMemory::BlockNumber32Z; + m_psm[PSM_PSMZ24].bn = &GSLocalMemory::BlockNumber32Z; + m_psm[PSM_PSMZ16].bn = &GSLocalMemory::BlockNumber16Z; + m_psm[PSM_PSMZ16S].bn = &GSLocalMemory::BlockNumber16SZ; m_psm[PSM_PSMCT24].rp = &GSLocalMemory::ReadPixel24; m_psm[PSM_PSMCT16].rp = &GSLocalMemory::ReadPixel16; @@ -342,11 +378,44 @@ GSLocalMemory::GSLocalMemory() m_psm[PSM_PSMZ16].rtxNP = &GSLocalMemory::ReadTexture16ZNP; m_psm[PSM_PSMZ16S].rtxNP = &GSLocalMemory::ReadTexture16SZNP; + m_psm[PSM_PSMCT24].rtxP = &GSLocalMemory::ReadTexture24; + m_psm[PSM_PSMCT16].rtxP = &GSLocalMemory::ReadTexture16; + m_psm[PSM_PSMCT16S].rtxP = &GSLocalMemory::ReadTexture16S; m_psm[PSM_PSMT8].rtxP = &GSLocalMemory::ReadTexture8P; m_psm[PSM_PSMT4].rtxP = &GSLocalMemory::ReadTexture4P; m_psm[PSM_PSMT8H].rtxP = &GSLocalMemory::ReadTexture8HP; m_psm[PSM_PSMT4HL].rtxP = &GSLocalMemory::ReadTexture4HLP; m_psm[PSM_PSMT4HH].rtxP = &GSLocalMemory::ReadTexture4HHP; + m_psm[PSM_PSMZ32].rtxP = &GSLocalMemory::ReadTexture32Z; + m_psm[PSM_PSMZ24].rtxP = &GSLocalMemory::ReadTexture24Z; + m_psm[PSM_PSMZ16].rtxP = &GSLocalMemory::ReadTexture16Z; + m_psm[PSM_PSMZ16S].rtxP = &GSLocalMemory::ReadTexture16SZ; + + m_psm[PSM_PSMCT24].rtxb = &GSLocalMemory::ReadTextureBlock24; + m_psm[PSM_PSMCT16].rtxb = &GSLocalMemory::ReadTextureBlock16; + m_psm[PSM_PSMCT16S].rtxb = &GSLocalMemory::ReadTextureBlock16S; + m_psm[PSM_PSMT8].rtxb = &GSLocalMemory::ReadTextureBlock8; + m_psm[PSM_PSMT4].rtxb = &GSLocalMemory::ReadTextureBlock4; + m_psm[PSM_PSMT8H].rtxb = &GSLocalMemory::ReadTextureBlock8H; + m_psm[PSM_PSMT4HL].rtxb = &GSLocalMemory::ReadTextureBlock4HL; + m_psm[PSM_PSMT4HH].rtxb = &GSLocalMemory::ReadTextureBlock4HH; + m_psm[PSM_PSMZ32].rtxb = &GSLocalMemory::ReadTextureBlock32Z; + m_psm[PSM_PSMZ24].rtxb = &GSLocalMemory::ReadTextureBlock24Z; + m_psm[PSM_PSMZ16].rtxb = &GSLocalMemory::ReadTextureBlock16Z; + m_psm[PSM_PSMZ16S].rtxb = &GSLocalMemory::ReadTextureBlock16SZ; + + m_psm[PSM_PSMCT24].rtxbP = &GSLocalMemory::ReadTextureBlock24; + m_psm[PSM_PSMCT16].rtxbP = &GSLocalMemory::ReadTextureBlock16; + m_psm[PSM_PSMCT16S].rtxbP = &GSLocalMemory::ReadTextureBlock16S; + m_psm[PSM_PSMT8].rtxbP = &GSLocalMemory::ReadTextureBlock8P; + m_psm[PSM_PSMT4].rtxbP = &GSLocalMemory::ReadTextureBlock4P; + m_psm[PSM_PSMT8H].rtxbP = &GSLocalMemory::ReadTextureBlock8HP; + m_psm[PSM_PSMT4HL].rtxbP = &GSLocalMemory::ReadTextureBlock4HLP; + m_psm[PSM_PSMT4HH].rtxbP = &GSLocalMemory::ReadTextureBlock4HHP; + m_psm[PSM_PSMZ32].rtxbP = &GSLocalMemory::ReadTextureBlock32Z; + m_psm[PSM_PSMZ24].rtxbP = &GSLocalMemory::ReadTextureBlock24Z; + m_psm[PSM_PSMZ16].rtxbP = &GSLocalMemory::ReadTextureBlock16Z; + m_psm[PSM_PSMZ16S].rtxbP = &GSLocalMemory::ReadTextureBlock16SZ; m_psm[PSM_PSMT8].pal = m_psm[PSM_PSMT8H].pal = 256; m_psm[PSM_PSMT4].pal = m_psm[PSM_PSMT4HL].pal = m_psm[PSM_PSMT4HH].pal = 16; @@ -381,6 +450,15 @@ GSLocalMemory::GSLocalMemory() for(int i = 0; i < 8; i++) m_psm[PSM_PSMZ24].rowOffset[i] = rowOffset32Z; for(int i = 0; i < 8; i++) m_psm[PSM_PSMZ16].rowOffset[i] = rowOffset16Z; for(int i = 0; i < 8; i++) m_psm[PSM_PSMZ16S].rowOffset[i] = rowOffset16SZ; + + m_psm[PSM_PSMCT16].blockOffset = blockOffset16; + m_psm[PSM_PSMCT16S].blockOffset = blockOffset16S; + m_psm[PSM_PSMT8].blockOffset = blockOffset8; + m_psm[PSM_PSMT4].blockOffset = blockOffset4; + m_psm[PSM_PSMZ32].blockOffset = blockOffset32Z; + m_psm[PSM_PSMZ24].blockOffset = blockOffset32Z; + m_psm[PSM_PSMZ16].blockOffset = blockOffset16Z; + m_psm[PSM_PSMZ16S].blockOffset = blockOffset16SZ; } GSLocalMemory::~GSLocalMemory() @@ -505,20 +583,18 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, const psm_t& tbl = m_psm[psm]; writePixel wp = tbl.wp; - pixelAddress ba = tbl.ba; + pixelAddress bn = tbl.bn; int w = tbl.bs.cx; int h = tbl.bs.cy; int bpp = tbl.bpp; - int shift = 0; - switch(bpp) { - case 32: shift = 0; break; - case 16: shift = 1; c = (c & 0xffff) * 0x00010001; break; - case 8: shift = 2; c = (c & 0xff) * 0x01010101; break; - case 4: shift = 3; c = (c & 0xf) * 0x11111111; break; + case 32: break; + case 16: c = (c & 0xffff) * 0x00010001; break; + case 8: c = (c & 0xff) * 0x01010101; break; + case 4: c = (c & 0xf) * 0x11111111; break; } CRect clip; @@ -574,7 +650,7 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, { for(int x = clip.left; x < clip.right; x += w) { - GSVector4i* p = (GSVector4i*)&m_vm8[ba(x, y, bp, bw) << 2 >> shift]; + GSVector4i* p = (GSVector4i*)&m_vm8[bn(x, y, bp, bw) << 8]; for(int i = 0; i < 16; i += 4) { @@ -594,7 +670,7 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, { for(int x = clip.left; x < clip.right; x += w) { - DWORD* p = &m_vm32[ba(x, y, bp, bw)]; + DWORD* p = (DWORD*)&m_vm8[bn(x, y, bp, bw) << 8]; for(int i = 0; i < 64; i += 4) { @@ -618,7 +694,7 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, { for(int x = clip.left; x < clip.right; x += w) { - GSVector4i* p = (GSVector4i*)&m_vm8[ba(x, y, bp, bw) << 2 >> shift]; + GSVector4i* p = (GSVector4i*)&m_vm8[bn(x, y, bp, bw) << 8]; for(int i = 0; i < 16; i += 4) { @@ -636,7 +712,7 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, { for(int x = clip.left; x < clip.right; x += w) { - DWORD* p = (DWORD*)&m_vm8[ba(x, y, bp, bw) << 2 >> shift]; + DWORD* p = (DWORD*)&m_vm8[bn(x, y, bp, bw) << 8]; for(int i = 0; i < 64; i += 4) { @@ -670,14 +746,14 @@ void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, BYTE* src, int { switch(psm) { - case PSM_PSMCT32: WriteColumn32(y, (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], &src[x * 4], srcpitch); break; - case PSM_PSMCT16: WriteColumn16(y, (BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMCT16S: WriteColumn16(y, (BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMT8: WriteColumn8(y, (BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)], &src[x], srcpitch); break; - case PSM_PSMT4: WriteColumn4(y, (BYTE*)&m_vm8[BlockAddress4(x, y, bp, bw) >> 1], &src[x >> 1], srcpitch); break; - case PSM_PSMZ32: WriteColumn32(y, (BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)], &src[x * 4], srcpitch); break; - case PSM_PSMZ16: WriteColumn16(y, (BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMZ16S: WriteColumn16(y, (BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)], &src[x * 2], srcpitch); break; + case PSM_PSMCT32: WriteColumn32(y, BlockPtr32(x, y, bp, bw), &src[x * 4], srcpitch); break; + case PSM_PSMCT16: WriteColumn16(y, BlockPtr16(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMCT16S: WriteColumn16(y, BlockPtr16S(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMT8: WriteColumn8(y, BlockPtr8(x, y, bp, bw), &src[x], srcpitch); break; + case PSM_PSMT4: WriteColumn4(y, BlockPtr4(x, y, bp, bw), &src[x >> 1], srcpitch); break; + case PSM_PSMZ32: WriteColumn32(y, BlockPtr32Z(x, y, bp, bw), &src[x * 4], srcpitch); break; + case PSM_PSMZ16: WriteColumn16(y, BlockPtr16Z(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMZ16S: WriteColumn16(y, BlockPtr16SZ(x, y, bp, bw), &src[x * 2], srcpitch); break; // TODO default: __assume(0); } @@ -697,14 +773,14 @@ void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, BYTE* src, int s { switch(psm) { - case PSM_PSMCT32: WriteBlock32((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], &src[x * 4], srcpitch); break; - case PSM_PSMCT16: WriteBlock16((BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMCT16S: WriteBlock16((BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMT8: WriteBlock8((BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)], &src[x], srcpitch); break; - case PSM_PSMT4: WriteBlock4((BYTE*)&m_vm8[BlockAddress4(x, y, bp, bw) >> 1], &src[x >> 1], srcpitch); break; - case PSM_PSMZ32: WriteBlock32((BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)], &src[x * 4], srcpitch); break; - case PSM_PSMZ16: WriteBlock16((BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)], &src[x * 2], srcpitch); break; - case PSM_PSMZ16S: WriteBlock16((BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)], &src[x * 2], srcpitch); break; + case PSM_PSMCT32: WriteBlock32(BlockPtr32(x, y, bp, bw), &src[x * 4], srcpitch); break; + case PSM_PSMCT16: WriteBlock16(BlockPtr16(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMCT16S: WriteBlock16(BlockPtr16S(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMT8: WriteBlock8(BlockPtr8(x, y, bp, bw), &src[x], srcpitch); break; + case PSM_PSMT4: WriteBlock4(BlockPtr4(x, y, bp, bw), &src[x >> 1], srcpitch); break; + case PSM_PSMZ32: WriteBlock32(BlockPtr32Z(x, y, bp, bw), &src[x * 4], srcpitch); break; + case PSM_PSMZ16: WriteBlock16(BlockPtr16Z(x, y, bp, bw), &src[x * 2], srcpitch); break; + case PSM_PSMZ16S: WriteBlock16(BlockPtr16SZ(x, y, bp, bw), &src[x * 2], srcpitch); break; // TODO default: __assume(0); } @@ -763,14 +839,14 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, i switch(psm) { - case PSM_PSMCT32: dst = (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]; break; - case PSM_PSMCT16: dst = (BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)]; break; - case PSM_PSMCT16S: dst = (BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)]; break; - case PSM_PSMT8: dst = (BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)]; break; - case PSM_PSMT4: dst = (BYTE*)&m_vm8[BlockAddress4(x, y, bp, bw) >> 1]; break; - case PSM_PSMZ32: dst = (BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)]; break; - case PSM_PSMZ16: dst = (BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)]; break; - case PSM_PSMZ16S: dst = (BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)]; break; + case PSM_PSMCT32: dst = BlockPtr32(x, y, bp, bw); break; + case PSM_PSMCT16: dst = BlockPtr16(x, y, bp, bw); break; + case PSM_PSMCT16S: dst = BlockPtr16S(x, y, bp, bw); break; + case PSM_PSMT8: dst = BlockPtr8(x, y, bp, bw); break; + case PSM_PSMT4: dst = BlockPtr4(x, y, bp, bw); break; + case PSM_PSMZ32: dst = BlockPtr32Z(x, y, bp, bw); break; + case PSM_PSMZ16: dst = BlockPtr16Z(x, y, bp, bw); break; + case PSM_PSMZ16S: dst = BlockPtr16SZ(x, y, bp, bw); break; // TODO default: __assume(0); } @@ -844,14 +920,14 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, i switch(psm) { - case PSM_PSMCT32: dst = (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]; break; - case PSM_PSMCT16: dst = (BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)]; break; - case PSM_PSMCT16S: dst = (BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)]; break; - case PSM_PSMT8: dst = (BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)]; break; - case PSM_PSMT4: dst = (BYTE*)&m_vm8[BlockAddress4(x, y, bp, bw) >> 1]; break; - case PSM_PSMZ32: dst = (BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)]; break; - case PSM_PSMZ16: dst = (BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)]; break; - case PSM_PSMZ16S: dst = (BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)]; break; + case PSM_PSMCT32: dst = BlockPtr32(x, y, bp, bw); break; + case PSM_PSMCT16: dst = BlockPtr16(x, y, bp, bw); break; + case PSM_PSMCT16S: dst = BlockPtr16S(x, y, bp, bw); break; + case PSM_PSMT8: dst = BlockPtr8(x, y, bp, bw); break; + case PSM_PSMT4: dst = BlockPtr4(x, y, bp, bw); break; + case PSM_PSMZ32: dst = BlockPtr32Z(x, y, bp, bw); break; + case PSM_PSMZ16: dst = BlockPtr16Z(x, y, bp, bw); break; + case PSM_PSMZ16S: dst = BlockPtr16SZ(x, y, bp, bw); break; // TODO default: __assume(0); } @@ -1026,7 +1102,7 @@ void GSLocalMemory::WriteImage24(int& tx, int& ty, BYTE* src, int len, GIFRegBIT { for(int x = tx; x < tw; x += 8) { - UnpackAndWriteBlock24(src + (x - tx) * 3, srcpitch, (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]); + UnpackAndWriteBlock24(src + (x - tx) * 3, srcpitch, BlockPtr32(x, y, bp, bw)); } } @@ -1060,7 +1136,7 @@ void GSLocalMemory::WriteImage8H(int& tx, int& ty, BYTE* src, int len, GIFRegBIT { for(int x = tx; x < tw; x += 8) { - UnpackAndWriteBlock8H(src + (x - tx), srcpitch, (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]); + UnpackAndWriteBlock8H(src + (x - tx), srcpitch, BlockPtr32(x, y, bp, bw)); } } @@ -1094,7 +1170,7 @@ void GSLocalMemory::WriteImage4HL(int& tx, int& ty, BYTE* src, int len, GIFRegBI { for(int x = tx; x < tw; x += 8) { - UnpackAndWriteBlock4HL(src + (x - tx) / 2, srcpitch, (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]); + UnpackAndWriteBlock4HL(src + (x - tx) / 2, srcpitch, BlockPtr32(x, y, bp, bw)); } } @@ -1128,7 +1204,7 @@ void GSLocalMemory::WriteImage4HH(int& tx, int& ty, BYTE* src, int len, GIFRegBI { for(int x = tx; x < tw; x += 8) { - UnpackAndWriteBlock4HH(src + (x - tx) / 2, srcpitch, (BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)]); + UnpackAndWriteBlock4HH(src + (x - tx) / 2, srcpitch, BlockPtr32(x, y, bp, bw)); } } @@ -1161,7 +1237,7 @@ void GSLocalMemory::WriteImage24Z(int& tx, int& ty, BYTE* src, int len, GIFRegBI { for(int x = tx; x < tw; x += 8) { - UnpackAndWriteBlock24(src + (x - tx) * 3, srcpitch, (BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)]); + UnpackAndWriteBlock24(src + (x - tx) * 3, srcpitch, BlockPtr32Z(x, y, bp, bw)); } } @@ -1525,10 +1601,9 @@ void GSLocalMemory::ReadTexture32(const CRect& r, BYTE* dst, int dstpitch, const { FOREACH_BLOCK_START(8, 8, 32) { - ReadBlock32((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch); + ReadBlock32(BlockPtr32(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END - } void GSLocalMemory::ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const @@ -1537,7 +1612,7 @@ void GSLocalMemory::ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock24((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, TEXA); + ReadAndExpandBlock24(BlockPtr32(x, y, bp, bw), dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1545,7 +1620,7 @@ void GSLocalMemory::ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock24((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, TEXA); + ReadAndExpandBlock24(BlockPtr32(x, y, bp, bw), dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1557,7 +1632,7 @@ void GSLocalMemory::ReadTexture16(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1570,7 +1645,7 @@ void GSLocalMemory::ReadTexture16S(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16S(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1583,7 +1658,7 @@ void GSLocalMemory::ReadTexture8(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_START(16, 16, 32) { - ReadAndExpandBlock8_32((BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock8_32(BlockPtr8(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1594,7 +1669,7 @@ void GSLocalMemory::ReadTexture4(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_START(32, 16, 32) { - ReadAndExpandBlock4_32((BYTE*)&m_vm8[BlockAddress4(x, y, bp, bw) >> 1], dst, dstpitch, pal); + ReadAndExpandBlock4_32(BlockPtr4(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1605,7 +1680,7 @@ void GSLocalMemory::ReadTexture8H(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock8H_32((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock8H_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1616,7 +1691,7 @@ void GSLocalMemory::ReadTexture4HL(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock4HL_32((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock4HL_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1627,7 +1702,7 @@ void GSLocalMemory::ReadTexture4HH(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock4HH_32((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock4HH_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1636,7 +1711,7 @@ void GSLocalMemory::ReadTexture32Z(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(8, 8, 32) { - ReadBlock32((BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)], dst, dstpitch); + ReadBlock32(BlockPtr32Z(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -1647,7 +1722,7 @@ void GSLocalMemory::ReadTexture24Z(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock24((BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)], dst, dstpitch, TEXA); + ReadAndExpandBlock24(BlockPtr32Z(x, y, bp, bw), dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1655,7 +1730,7 @@ void GSLocalMemory::ReadTexture24Z(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock24((BYTE*)&m_vm32[BlockAddress32Z(x, y, bp, bw)], dst, dstpitch, TEXA); + ReadAndExpandBlock24(BlockPtr32Z(x, y, bp, bw), dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1667,7 +1742,7 @@ void GSLocalMemory::ReadTexture16Z(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16Z(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1680,7 +1755,7 @@ void GSLocalMemory::ReadTexture16SZ(const CRect& r, BYTE* dst, int dstpitch, con FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16SZ(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1689,6 +1764,113 @@ void GSLocalMemory::ReadTexture16SZ(const CRect& r, BYTE* dst, int dstpitch, con /////////////////// +void GSLocalMemory::ReadTextureBlock32(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock32(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock24(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + if(TEXA.AEM) + { + ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); + } + else + { + ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); + } +} + +void GSLocalMemory::ReadTextureBlock16(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + __declspec(align(16)) WORD block[16 * 8]; + + ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + + ExpandBlock16(block, dst, dstpitch, TEXA); +} + +void GSLocalMemory::ReadTextureBlock16S(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + __declspec(align(16)) WORD block[16 * 8]; + + ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + + ExpandBlock16(block, dst, dstpitch, TEXA); +} + +void GSLocalMemory::ReadTextureBlock8(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + const DWORD* pal = m_clut; + + ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, pal); +} + +void GSLocalMemory::ReadTextureBlock4(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + const UINT64* pal = m_clut; + + ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, pal); +} + +void GSLocalMemory::ReadTextureBlock8H(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + const DWORD* pal = m_clut; + + ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, pal); +} + +void GSLocalMemory::ReadTextureBlock4HL(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + const DWORD* pal = m_clut; + + ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, pal); +} + +void GSLocalMemory::ReadTextureBlock4HH(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + const DWORD* pal = m_clut; + + ReadAndExpandBlock4HH_32(BlockPtr(bp), dst, dstpitch, pal); +} + +void GSLocalMemory::ReadTextureBlock32Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock32(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock24Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + if(TEXA.AEM) + { + ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); + } + else + { + ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); + } +} + +void GSLocalMemory::ReadTextureBlock16Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + __declspec(align(16)) WORD block[16 * 8]; + + ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + + ExpandBlock16(block, dst, dstpitch, TEXA); +} + +void GSLocalMemory::ReadTextureBlock16SZ(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + __declspec(align(16)) WORD block[16 * 8]; + + ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + + ExpandBlock16(block, dst, dstpitch, TEXA); +} + +/////////////////// + void GSLocalMemory::ReadTexture(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) { readTexture rtx = m_psm[TEX0.PSM].rtx; @@ -1731,7 +1913,7 @@ void GSLocalMemory::ReadTexture16NP(const CRect& r, BYTE* dst, int dstpitch, con { FOREACH_BLOCK_START(16, 8, 16) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16(x, y, bp, bw)], dst, dstpitch); + ReadBlock16(BlockPtr16(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -1740,7 +1922,7 @@ void GSLocalMemory::ReadTexture16SNP(const CRect& r, BYTE* dst, int dstpitch, co { FOREACH_BLOCK_START(16, 8, 16) { - ReadBlock16((BYTE*)&m_vm16[BlockAddress16S(x, y, bp, bw)], dst, dstpitch); + ReadBlock16(BlockPtr16S(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -1753,7 +1935,7 @@ void GSLocalMemory::ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(16, 16, 32) { - ReadAndExpandBlock8_32((BYTE*)&m_vm8[BlockAddress8(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock8_32(BlockPtr8(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1765,7 +1947,7 @@ void GSLocalMemory::ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(16, 16, 16) { - ReadBlock8(&m_vm8[BlockAddress8(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 16); + ReadBlock8(BlockPtr8(x, y, bp, bw), (BYTE*)block, sizeof(block) / 16); ExpandBlock8_16(block, dst, dstpitch, pal); } @@ -1781,7 +1963,7 @@ void GSLocalMemory::ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(32, 16, 32) { - ReadAndExpandBlock4_32(&m_vm8[BlockAddress4(x, y, bp, bw) >> 1], dst, dstpitch, pal); + ReadAndExpandBlock4_32(BlockPtr4(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1793,7 +1975,7 @@ void GSLocalMemory::ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_START(32, 16, 16) { - ReadBlock4(&m_vm8[BlockAddress4(x, y, bp, bw)>>1], (BYTE*)block, sizeof(block) / 16); + ReadBlock4(BlockPtr4(x, y, bp, bw), (BYTE*)block, sizeof(block) / 16); ExpandBlock4_16(block, dst, dstpitch, pal); } @@ -1809,7 +1991,7 @@ void GSLocalMemory::ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, con { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock8H_32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock8H_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1821,7 +2003,7 @@ void GSLocalMemory::ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, con FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock8H_16(block, dst, dstpitch, pal); } @@ -1837,7 +2019,7 @@ void GSLocalMemory::ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, co { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock4HL_32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock4HL_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1849,7 +2031,7 @@ void GSLocalMemory::ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, co FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock4HL_16(block, dst, dstpitch, pal); } @@ -1865,7 +2047,7 @@ void GSLocalMemory::ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, co { FOREACH_BLOCK_START(8, 8, 32) { - ReadAndExpandBlock4HH_32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch, pal); + ReadAndExpandBlock4HH_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1877,7 +2059,7 @@ void GSLocalMemory::ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, co FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32((const BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); ExpandBlock4HH_16(block, dst, dstpitch, pal); } @@ -1889,7 +2071,7 @@ void GSLocalMemory::ReadTexture16ZNP(const CRect& r, BYTE* dst, int dstpitch, co { FOREACH_BLOCK_START(16, 8, 16) { - ReadBlock16((const BYTE*)&m_vm16[BlockAddress16Z(x, y, bp, bw)], dst, dstpitch); + ReadBlock16(BlockPtr16Z(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -1898,7 +2080,7 @@ void GSLocalMemory::ReadTexture16SZNP(const CRect& r, BYTE* dst, int dstpitch, c { FOREACH_BLOCK_START(16, 8, 16) { - ReadBlock16((const BYTE*)&m_vm16[BlockAddress16SZ(x, y, bp, bw)], dst, dstpitch); + ReadBlock16(BlockPtr16SZ(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -1996,7 +2178,7 @@ void GSLocalMemory::ReadTexture8P(const CRect& r, BYTE* dst, int dstpitch, const { FOREACH_BLOCK_START(16, 16, 8) { - ReadBlock8(&m_vm8[BlockAddress8(x, y, bp, bw)], dst, dstpitch); + ReadBlock8(BlockPtr8(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -2005,7 +2187,7 @@ void GSLocalMemory::ReadTexture4P(const CRect& r, BYTE* dst, int dstpitch, const { FOREACH_BLOCK_START(32, 16, 8) { - ReadBlock4P(&m_vm8[BlockAddress4(x, y, bp, bw) >> 1], dst, dstpitch); + ReadBlock4P(BlockPtr4(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -2014,7 +2196,7 @@ void GSLocalMemory::ReadTexture8HP(const CRect& r, BYTE* dst, int dstpitch, cons { FOREACH_BLOCK_START(8, 8, 8) { - ReadBlock8HP((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch); + ReadBlock8HP(BlockPtr32(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -2023,7 +2205,7 @@ void GSLocalMemory::ReadTexture4HLP(const CRect& r, BYTE* dst, int dstpitch, con { FOREACH_BLOCK_START(8, 8, 8) { - ReadBlock4HLP((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch); + ReadBlock4HLP(BlockPtr32(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } @@ -2032,13 +2214,40 @@ void GSLocalMemory::ReadTexture4HHP(const CRect& r, BYTE* dst, int dstpitch, con { FOREACH_BLOCK_START(8, 8, 8) { - ReadBlock4HHP((BYTE*)&m_vm32[BlockAddress32(x, y, bp, bw)], dst, dstpitch); + ReadBlock4HHP(BlockPtr32(x, y, bp, bw), dst, dstpitch); } FOREACH_BLOCK_END } // +void GSLocalMemory::ReadTextureBlock8P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock8(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock4P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock4P(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock8HP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock8HP(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock4HLP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock4HLP(BlockPtr(bp), dst, dstpitch); +} + +void GSLocalMemory::ReadTextureBlock4HHP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +{ + ReadBlock4HHP(BlockPtr(bp), dst, dstpitch); +} + +// + template void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx) { diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index bcc18e5a6f..b5063cb087 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -44,12 +44,13 @@ public: typedef void (GSLocalMemory::*writeImage)(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); typedef void (GSLocalMemory::*readImage)(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; typedef void (GSLocalMemory::*readTexture)(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + typedef void (GSLocalMemory::*readTextureBlock)(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; typedef union { struct { - pixelAddress pa, ba, pga, pgn; + pixelAddress pa, bn; readPixel rp; readPixelAddr rpa; writePixel wp; @@ -60,11 +61,16 @@ public: writeImage wi; readImage ri; readTexture rtx, rtxNP, rtxP; - DWORD bpp, pal, trbpp; + readTextureBlock rtxb, rtxbP; + WORD bpp, trbpp; + DWORD pal; CSize bs, pgs; int* rowOffset[8]; + int* blockOffset; }; + BYTE dummy[128]; + } psm_t; static psm_t m_psm[64]; @@ -110,6 +116,15 @@ protected: static int rowOffset8[2][2048]; static int rowOffset4[2][2048]; + static int blockOffset32[256]; + static int blockOffset32Z[256]; + static int blockOffset16[256]; + static int blockOffset16S[256]; + static int blockOffset16Z[256]; + static int blockOffset16SZ[256]; + static int blockOffset8[256]; + static int blockOffset4[256]; + __forceinline static DWORD Expand24To32(DWORD c, const GIFRegTEXA& TEXA) { return (((!TEXA.AEM | (c & 0xffffff)) ? TEXA.TA0 : 0) << 24) | (c & 0xffffff); @@ -138,180 +153,135 @@ public: // address - static DWORD PageNumber32(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber32(int x, int y, DWORD bp, DWORD bw) { - return (bp >> 5) + (y >> 5) * bw + (x >> 6); + return bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable32[(y >> 3) & 3][(x >> 3) & 7]; } - static DWORD PageNumber16(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber16(int x, int y, DWORD bp, DWORD bw) { - return (bp >> 5) + (y >> 6) * bw + (x >> 6); + return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD PageNumber8(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber16S(int x, int y, DWORD bp, DWORD bw) + { + return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16S[(y >> 3) & 7][(x >> 4) & 3]; + } + + static DWORD BlockNumber8(int x, int y, DWORD bp, DWORD bw) { ASSERT((bw & 1) == 0); - return (bp >> 5) + (y >> 6) * (bw >> 1) + (x >> 7); + return bp + ((y >> 1) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f) + blockTable8[(y >> 4) & 3][(x >> 4) & 7]; } - static DWORD PageNumber4(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber4(int x, int y, DWORD bp, DWORD bw) { ASSERT((bw & 1) == 0); - return (bp >> 5) + (y >> 7) * (bw >> 1) + (x >> 7); + return bp + ((y >> 2) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f) + blockTable4[(y >> 4) & 7][(x >> 5) & 3]; } - static DWORD PageAddress32(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber32Z(int x, int y, DWORD bp, DWORD bw) { - return PageNumber32(x, y, bp, bw) << 11; + return bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable32Z[(y >> 3) & 3][(x >> 3) & 7]; } - static DWORD PageAddress16(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber16Z(int x, int y, DWORD bp, DWORD bw) { - return PageNumber16(x, y, bp, bw) << 12; + return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16Z[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD PageAddress8(int x, int y, DWORD bp, DWORD bw) + static DWORD BlockNumber16SZ(int x, int y, DWORD bp, DWORD bw) { - return PageNumber8(x, y, bp, bw) << 13; + return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16SZ[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD PageAddress4(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr(DWORD bp) const { - return PageNumber4(x, y, bp, bw) << 14; + ASSERT(bp < 16384); + + return &m_vm8[bp << 8]; } - static DWORD BlockAddress32(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr32(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable32[(y >> 3) & 3][(x >> 3) & 7]; - return (page + block) << 6; + return &m_vm8[BlockNumber32(x, y, bp, bw) << 8]; } - static DWORD BlockAddress16(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr16(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16[(y >> 3) & 7][(x >> 4) & 3]; - return (page + block) << 7; + return &m_vm8[BlockNumber16(x, y, bp, bw) << 8]; } - static DWORD BlockAddress16S(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr16S(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16S[(y >> 3) & 7][(x >> 4) & 3]; - return (page + block) << 7; + return &m_vm8[BlockNumber16S(x, y, bp, bw) << 8]; } - static DWORD BlockAddress8(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr8(int x, int y, DWORD bp, DWORD bw) const { - ASSERT((bw & 1) == 0); - DWORD page = bp + ((y >> 1) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f); - DWORD block = blockTable8[(y >> 4) & 3][(x >> 4) & 7]; - return (page + block) << 8; + return &m_vm8[BlockNumber8(x, y, bp, bw) << 8]; } - static DWORD BlockAddress4(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr4(int x, int y, DWORD bp, DWORD bw) const { - ASSERT((bw & 1) == 0); - DWORD page = bp + ((y >> 2) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f); - DWORD block = blockTable4[(y >> 4) & 7][(x >> 5) & 3]; - return (page + block) << 9; + return &m_vm8[BlockNumber4(x, y, bp, bw) << 8]; } - static DWORD BlockAddress32Z(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr32Z(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable32Z[(y >> 3) & 3][(x >> 3) & 7]; - return (page + block) << 6; + return &m_vm8[BlockNumber32Z(x, y, bp, bw) << 8]; } - static DWORD BlockAddress16Z(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr16Z(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16Z[(y >> 3) & 7][(x >> 4) & 3]; - return (page + block) << 7; + return &m_vm8[BlockNumber16Z(x, y, bp, bw) << 8]; } - static DWORD BlockAddress16SZ(int x, int y, DWORD bp, DWORD bw) + BYTE* BlockPtr16SZ(int x, int y, DWORD bp, DWORD bw) const { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16SZ[(y >> 3) & 7][(x >> 4) & 3]; - return (page + block) << 7; + return &m_vm8[BlockNumber16SZ(x, y, bp, bw) << 8]; } static DWORD PixelAddressOrg32(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable32[(y >> 3) & 3][(x >> 3) & 7]; - DWORD word = ((page + block) << 6) + columnTable32[y & 7][x & 7]; - ASSERT(word < 1024*1024); - return word; + return (BlockNumber32(x, y, bp, bw) << 6) + columnTable32[y & 7][x & 7]; } static DWORD PixelAddressOrg16(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16[(y >> 3) & 7][(x >> 4) & 3]; - DWORD word = ((page + block) << 7) + columnTable16[y & 7][x & 15]; - ASSERT(word < 1024*1024*2); - return word; + return (BlockNumber16(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } static DWORD PixelAddressOrg16S(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16S[(y >> 3) & 7][(x >> 4) & 3]; - DWORD word = ((page + block) << 7) + columnTable16[y & 7][x & 15]; - ASSERT(word < 1024*1024*2); - return word; + return (BlockNumber16S(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } static DWORD PixelAddressOrg8(int x, int y, DWORD bp, DWORD bw) { - ASSERT((bw & 1) == 0); - DWORD page = bp + ((y >> 1) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f); - DWORD block = blockTable8[(y >> 4) & 3][(x >> 4) & 7]; - DWORD word = ((page + block) << 8) + columnTable8[y & 15][x & 15]; - ASSERT(word < 1024*1024*4); - return word; + return (BlockNumber8(x, y, bp, bw) << 8) + columnTable8[y & 15][x & 15]; } static DWORD PixelAddressOrg4(int x, int y, DWORD bp, DWORD bw) { - ASSERT((bw & 1) == 0); - DWORD page = bp + ((y >> 2) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f); - DWORD block = blockTable4[(y >> 4) & 7][(x >> 5) & 3]; - DWORD word = ((page + block) << 9) + columnTable4[y & 15][x & 31]; - ASSERT(word < 1024*1024*8); - return word; + return (BlockNumber4(x, y, bp, bw) << 9) + columnTable4[y & 15][x & 31]; } static DWORD PixelAddressOrg32Z(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable32Z[(y >> 3) & 3][(x >> 3) & 7]; - DWORD word = ((page + block) << 6) + columnTable32[y & 7][x & 7]; - ASSERT(word < 1024*1024); - return word; + return (BlockNumber32Z(x, y, bp, bw) << 6) + columnTable32[y & 7][x & 7]; } static DWORD PixelAddressOrg16Z(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16Z[(y >> 3) & 7][(x >> 4) & 3]; - DWORD word = ((page + block) << 7) + columnTable16[y & 7][x & 15]; - ASSERT(word < 1024*1024*2); - return word; + return (BlockNumber16Z(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } static DWORD PixelAddressOrg16SZ(int x, int y, DWORD bp, DWORD bw) { - DWORD page = bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f); - DWORD block = blockTable16SZ[(y >> 3) & 7][(x >> 4) & 3]; - DWORD word = ((page + block) << 7) + columnTable16[y & 7][x & 15]; - ASSERT(word < 1024*1024*2); - return word; + return (BlockNumber16SZ(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } static __forceinline DWORD PixelAddress32(int x, int y, DWORD bp, DWORD bw) @@ -952,7 +922,7 @@ public: void ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; - // + // * => 32 void ReadTexture32(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; @@ -971,7 +941,21 @@ public: void ReadTexture(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); void ReadTextureNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - // 32/16 + void ReadTextureBlock32(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock24(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16S(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8H(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HL(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HH(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock32Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock24Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16SZ(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + + // * => 32/16 void ReadTexture16NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture16SNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; @@ -986,7 +970,7 @@ public: void ReadTextureNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); void ReadTextureNPNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - // 32/8 + // pal ? 8 : 32 void ReadTexture8P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture4P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; @@ -994,6 +978,12 @@ public: void ReadTexture4HLP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture4HHP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8HP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HLP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HHP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + // static DWORD m_xtbl[1024], m_ytbl[1024]; diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 178800736d..3189f49a44 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -654,28 +654,29 @@ protected: if(fst) { GSVector4i uv = GSVector4i(m_vtrace.m_min.t.xyxy(m_vtrace.m_max.t)).sra32(16); -/* - int tw = context->TEX0.TW; - int th = context->TEX0.TH; - GSVector4i u = uv & GSVector4i::xffffffff().srl32(32 - tw); - GSVector4i v = uv & GSVector4i::xffffffff().srl32(32 - th); + GSVector4i u, v; - GSVector4i uu = uv.sra32(tw); - GSVector4i vv = uv.sra32(th); + int mask; + + if(wms == CLAMP_REPEAT || wmt == CLAMP_REPEAT) + { + int tw = context->TEX0.TW; + int th = context->TEX0.TH; + + u = uv & GSVector4i::xffffffff().srl32(32 - tw); + v = uv & GSVector4i::xffffffff().srl32(32 - th); + + GSVector4i uu = uv.sra32(tw); + GSVector4i vv = uv.sra32(th); + + mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); + } - int mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); -*/ switch(wms) { case CLAMP_REPEAT: -/* - if(mask & 0x000f) - { - if(vr.x < u.x) vr.x = u.x; - if(vr.z > u.z + 1) vr.z = u.z + 1; - } -*/ + if(mask & 0x000f) {if(vr.x < u.x) vr.x = u.x; if(vr.z > u.z + 1) vr.z = u.z + 1;} break; case CLAMP_CLAMP: case CLAMP_REGION_CLAMP: @@ -691,13 +692,7 @@ protected: switch(wmt) { case CLAMP_REPEAT: -/* - if(mask & 0xf000) - { - if(vr.y < v.y) vr.y = v.y; - if(vr.w > v.w + 1) vr.w = v.w + 1; - } -*/ + if(mask & 0xf000) {if(vr.y < v.y) vr.y = v.y; if(vr.w > v.w + 1) vr.w = v.w + 1;} break; case CLAMP_CLAMP: case CLAMP_REGION_CLAMP: diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index e1b6e66aa2..e22bfb0e06 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -22,18 +22,13 @@ #include "StdAfx.h" #include "GSTextureCacheSW.h" -// static FILE* m_log = NULL; - GSTextureCacheSW::GSTextureCacheSW(GSState* state) : m_state(state) { - // m_log = _tfopen(_T("c:\\log.txt"), _T("w")); } GSTextureCacheSW::~GSTextureCacheSW() { - // fclose(m_log); - RemoveAll(); } @@ -43,20 +38,16 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - const CAtlList& t2p = m_p2t[TEX0.TBP0 >> 5]; - - // fprintf(m_log, "lu %05x %d %d (%d) ", TEX0.TBP0, TEX0.TBW, TEX0.PSM, t2p.GetCount()); - - // if(r) fprintf(m_log, "(%d %d %d %d) ", r->left, r->top, r->right, r->bottom); + const CAtlMap& map = m_map[TEX0.TBP0 >> 5]; GSTexture* t = NULL; - POSITION pos = t2p.GetHeadPosition(); + POSITION pos = map.GetStartPosition(); while(pos) { - GSTexture* t2 = t2p.GetNext(pos)->t; - + GSTexture* t2 = map.GetNextKey(pos); + // if(t2->m_TEX0.TBP0 != TEX0.TBP0 || t2->m_TEX0.TBW != TEX0.TBW || t2->m_TEX0.PSM != TEX0.PSM || t2->m_TEX0.TW != TEX0.TW || t2->m_TEX0.TH != TEX0.TH) if(((t2->m_TEX0.ai32[0] ^ TEX0.ai32[0]) | ((t2->m_TEX0.ai32[1] ^ TEX0.ai32[1]) & 3)) != 0) { @@ -68,8 +59,6 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE continue; } - // fprintf(m_log, "cache hit\n"); - t = t2; t->m_age = 0; @@ -79,8 +68,6 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE if(t == NULL) { - // fprintf(m_log, "cache miss\n"); - t = new GSTexture(m_state); t->m_pos = m_textures.AddTail(t); @@ -91,37 +78,38 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE DWORD bp = TEX0.TBP0; DWORD bw = TEX0.TBW; - for(int j = 0, y = 0; y < th; j++, y += psm.pgs.cy) + CSize s = (bp & 31) == 0 ? psm.pgs : psm.bs; + + for(int y = 0; y < th; y += s.cy) { - DWORD page = psm.pgn(0, y, bp, bw); + DWORD base = psm.bn(0, y, bp, bw); - for(int i = 0, x = 0; x < tw && page < MAX_PAGES; i++, x += psm.pgs.cx, page++) + for(int x = 0; x < tw; x += s.cx) { - GSTexturePage* p = new GSTexturePage(); - - p->t = t; - p->row = j; - p->col = i; + DWORD page = (base + psm.blockOffset[x >> 3]) >> 5; - GSTexturePageEntry* p2te = new GSTexturePageEntry(); + if(page >= MAX_PAGES) + { + continue; + } - p2te->p2t = &m_p2t[page]; - p2te->pos = m_p2t[page].AddHead(p); - - t->m_p2te.AddTail(p2te); - - t->m_maxpages++; + m_map[page].SetAt(t, true); } } } if(!t->Update(TEX0, TEXA, r)) { + printf("!@#$%\n"); // memory allocation may fail if the game is too hungry + m_textures.RemoveAt(t->m_pos); - delete t; + for(int i = 0; i < MAX_PAGES; i++) + { + m_map[i].RemoveKey(t); + } - printf("!@#$%\n"); // memory allocation may fail if the game is too hungry + delete t; return NULL; } @@ -142,18 +130,7 @@ void GSTextureCacheSW::RemoveAll() for(int i = 0; i < MAX_PAGES; i++) { - CAtlList& t2p = m_p2t[i]; - - ASSERT(t2p.IsEmpty()); - - POSITION pos = t2p.GetHeadPosition(); - - while(pos) - { - delete t2p.GetNext(pos); - } - - t2p.RemoveAll(); + m_map[i].RemoveAll(); } } @@ -167,61 +144,60 @@ void GSTextureCacheSW::IncAge() GSTexture* t = m_textures.GetNext(pos); - if(++t->m_age > 3) + if(++t->m_age > 30) { m_textures.RemoveAt(cur); + for(int i = 0; i < MAX_PAGES; i++) + { + m_map[i].RemoveKey(t); + } + delete t; } } } -void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& r) +void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& rect) { const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[BITBLTBUF.DPSM]; - CRect r2; - - r2.left = r.left & ~(psm.pgs.cx - 1); - r2.top = r.top & ~(psm.pgs.cy - 1); - r2.right = (r.right + (psm.pgs.cx - 1)) & ~(psm.pgs.cx - 1); - r2.bottom = (r.bottom + (psm.pgs.cy - 1)) & ~(psm.pgs.cy - 1); - DWORD bp = BITBLTBUF.DBP; DWORD bw = BITBLTBUF.DBW; - // fprintf(m_log, "ivm %05x %d %d (%d %d %d %d)\n", bp, bw, BITBLTBUF.DPSM, r2.left, r2.top, r2.right, r2.bottom); + CSize s = (bp & 31) == 0 ? psm.pgs : psm.bs; - for(int y = r2.top; y < r2.bottom; y += psm.pgs.cy) + CRect r; + + r.left = rect.left & ~(s.cx - 1); + r.top = rect.top & ~(s.cy - 1); + r.right = (rect.right + (s.cx - 1)) & ~(s.cx - 1); + r.bottom = (rect.bottom + (s.cy - 1)) & ~(s.cy - 1); + + for(int y = r.top; y < r.bottom; y += s.cy) { - DWORD page = psm.pgn(r2.left, y, bp, bw); + DWORD base = psm.bn(0, y, bp, bw); - for(int x = r2.left; x < r2.right && page < MAX_PAGES; x += psm.pgs.cx, page++) + for(int x = r.left; x < r.right; x += s.cx) { - const CAtlList& t2p = m_p2t[page]; + DWORD page = (base + psm.blockOffset[x >> 3]) >> 5; - POSITION pos = t2p.GetHeadPosition(); + if(page >= MAX_PAGES) + { + continue; + } + + const CAtlMap& map = m_map[page]; + + POSITION pos = map.GetStartPosition(); while(pos) { - GSTexturePage* p = t2p.GetNext(pos); + GSTexture* t = map.GetNextKey(pos); - DWORD flag = 1 << p->col; + t->m_valid[page] = 0; - if((p->t->m_valid[p->row] & flag) == 0) - { - continue; - } - - if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, p->t->m_TEX0.PSM)) - { - p->t->m_valid[p->row] &= ~flag; - p->t->m_pages--; - - // fprintf(m_log, "ivm hit %05x %d %d (%d %d) (%d)", p->t->m_TEX0.TBP0, p->t->m_TEX0.TBW, p->t->m_TEX0.PSM, p->row, p->col, p->t->m_pages); - // if(p->t->m_pages == 0) fprintf(m_log, " *"); - // fprintf(m_log, "\n"); - } + t->m_complete = false; } } } @@ -233,10 +209,9 @@ GSTextureCacheSW::GSTexture::GSTexture(GSState* state) : m_state(state) , m_buff(NULL) , m_tw(0) - , m_maxpages(0) - , m_pages(0) - , m_pos(NULL) , m_age(0) + , m_pos(NULL) + , m_complete(false) { memset(m_valid, 0, sizeof(m_valid)); } @@ -247,30 +222,11 @@ GSTextureCacheSW::GSTexture::~GSTexture() { _aligned_free(m_buff); } - - POSITION pos = m_p2te.GetHeadPosition(); - - while(pos) - { - GSTexturePageEntry* p2te = m_p2te.GetNext(pos); - - GSTexturePage* p = p2te->p2t->GetAt(p2te->pos); - - ASSERT(p->t == this); - - delete p; - - p2te->p2t->RemoveAt(p2te->pos); - - delete p2te; - } - - m_p2te.RemoveAll(); } -bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* r) +bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* rect) { - if(m_pages == m_maxpages) + if(m_complete) { return true; } @@ -282,16 +238,16 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - int tw = 1 << TEX0.TW; - int th = 1 << TEX0.TH; + DWORD bp = TEX0.TBP0; + DWORD bw = TEX0.TBW; - if(tw < psm.bs.cx) tw = psm.bs.cx; - if(th < psm.bs.cy) th = psm.bs.cy; + CSize s = psm.bs; + + int tw = max(1 << TEX0.TW, s.cx); + int th = max(1 << TEX0.TH, s.cy); if(m_buff == NULL) { - // fprintf(m_log, "up new (%d %d)\n", tw, th); - m_buff = _aligned_malloc(tw * th * sizeof(DWORD), 16); if(m_buff == NULL) @@ -302,75 +258,84 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX m_tw = max(psm.pal > 0 ? 5 : 3, TEX0.TW); // makes one row 32 bytes at least, matches the smallest block size that is allocated above for m_buff } - CRect r2; + CRect r(0, 0, tw, th); - if(r) + if(rect) { - r2.left = r->left & ~(psm.pgs.cx - 1); - r2.top = r->top & ~(psm.pgs.cy - 1); - r2.right = (r->right + (psm.pgs.cx - 1)) & ~(psm.pgs.cx - 1); - r2.bottom = (r->bottom + (psm.pgs.cy - 1)) & ~(psm.pgs.cy - 1); + r.left = rect->left & ~(s.cx - 1); + r.top = rect->top & ~(s.cy - 1); + r.right = (rect->right + (s.cx - 1)) & ~(s.cx - 1); + r.bottom = (rect->bottom + (s.cy - 1)) & ~(s.cy - 1); } - // TODO + if(r.left == 0 && r.top == 0 && r.right == tw && r.bottom == th) + { + m_complete = true; // lame, but better than nothing + } - GSLocalMemory::readTexture rt = psm.pal > 0 ? psm.rtxP : psm.rtx; + GSLocalMemory::readTextureBlock rtxb = psm.rtxbP; + int bytes = psm.pal > 0 ? 1 : 4; - BYTE* dst = (BYTE*)m_buff; - DWORD pitch = (1 << m_tw) * bytes; - DWORD mask = pitch - 1; - for(int j = 0, y = 0; y < th; j++, y += psm.pgs.cy, dst += pitch * psm.pgs.cy) + BYTE* dst = (BYTE*)m_buff + pitch * r.top; + + DWORD blocks = 0; + + for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) { - if(m_valid[j] == mask) + DWORD base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += s.cx) { - continue; - } + DWORD block = base + psm.blockOffset[x >> 3]; - if(r) - { - if(y < r2.top) continue; - if(y >= r2.bottom) break; - } - - DWORD page = psm.pgn(0, y, TEX0.TBP0, TEX0.TBW); - - for(int i = 0, x = 0; x < tw && page < MAX_PAGES; i++, x += psm.pgs.cx, page++) - { - if(r) - { - if(x < r2.left) continue; - if(x >= r2.right) break; - } - - DWORD flag = 1 << i; - - if(m_valid[j] & flag) + if(block >= MAX_BLOCKS) { continue; } - m_valid[j] |= flag; - m_pages++; + DWORD row = block >> 5; + DWORD col = 1 << (block & 31); - ASSERT(m_pages <= m_maxpages); + if(m_valid[row] & col) + { + continue; + } - CRect r; - - r.left = x; - r.top = y; - r.right = min(x + psm.pgs.cx, tw); - r.bottom = min(y + psm.pgs.cy, th); + // unfortunatelly a block may be part of the same texture multiple times at different places (when (1 << tw) > (tbw << 6), ex. 1024 > 640), + // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side - // fprintf(m_log, "up fetch (%d %d) (%d %d %d %d)\n", j, i, r.left, r.top, r.right, r.bottom); + // m_valid[row] |= col; - (mem.*rt)(r, &dst[x * bytes], pitch, TEX0, TEXA); + (mem.*rtxb)(block, &dst[x * bytes], pitch, TEXA); - m_state->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * bytes); + blocks++; + } + } + + m_state->m_perfmon.Put(GSPerfMon::Unswizzle, s.cx * s.cy * bytes * blocks); + + for(int y = r.top; y < r.bottom; y += s.cy) + { + DWORD base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += s.cx) + { + DWORD block = base + psm.blockOffset[x >> 3]; + + if(block >= MAX_BLOCKS) + { + continue; + } + + DWORD row = block >> 5; + DWORD col = 1 << (block & 31); + + m_valid[row] |= col; } } return true; -} \ No newline at end of file +} diff --git a/plugins/GSdx/GSTextureCacheSW.h b/plugins/GSdx/GSTextureCacheSW.h index 8f123300a6..1214c9e059 100644 --- a/plugins/GSdx/GSTextureCacheSW.h +++ b/plugins/GSdx/GSTextureCacheSW.h @@ -24,27 +24,11 @@ #include "GSRenderer.h" #define MAX_PAGES 512 +#define MAX_BLOCKS 16384 class GSTextureCacheSW { public: - class GSTexture; - class GSTexturePage; - - class GSTexturePage - { - public: - GSTexture* t; - DWORD row, col; - }; - - class GSTexturePageEntry - { - public: - CAtlList* p2t; - POSITION pos; - }; - class GSTexture { public: @@ -53,12 +37,10 @@ public: GIFRegTEXA m_TEXA; void* m_buff; DWORD m_tw; - DWORD m_valid[32]; - DWORD m_maxpages; - DWORD m_pages; - CAtlList m_p2te; - POSITION m_pos; + DWORD m_valid[MAX_PAGES]; // each DWORD bits map to the 32 blocks of that page DWORD m_age; + POSITION m_pos; + bool m_complete; explicit GSTexture(GSState* state); virtual ~GSTexture(); @@ -69,7 +51,7 @@ public: protected: GSState* m_state; CAtlList m_textures; - CAtlList m_p2t[MAX_PAGES]; + CAtlMap m_map[MAX_PAGES]; public: GSTextureCacheSW(GSState* state); From 91a4ee79cc03c6faf61673cb77092ff57c688f43 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 7 May 2009 08:21:57 +0000 Subject: [PATCH 022/296] Made some tweaks to the EErec block manager in hopes of resolving Issue 208 [Soul Calibur 3 missing gfx], involving the willbranch3 logic path. Emitter: Added xJcc8 / xJcc32 functions for doing modified jump targets. Also: minor cleanup to recent counters fixes (nothing relevant, just reduced out some copy-paste jobs for clarity). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1143 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Counters.cpp | 15 +---- pcsx2/IopCounters.cpp | 12 +--- pcsx2/NakedAsm.h | 52 ++++++++-------- pcsx2/x86/BaseblockEx.cpp | 11 ++-- pcsx2/x86/BaseblockEx.h | 2 +- pcsx2/x86/iR3000A.cpp | 97 ++++++++++++++---------------- pcsx2/x86/ix86-32/iR5900-32.cpp | 69 +++++++++------------ pcsx2/x86/ix86/ix86_instructions.h | 2 + pcsx2/x86/ix86/ix86_jmp.cpp | 46 +++++++++----- 9 files changed, 143 insertions(+), 163 deletions(-) diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index c75d5f7e15..5dc192f0ec 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -85,16 +85,12 @@ static __forceinline void _rcntSet( int cntidx ) if (c < nextCounter) { nextCounter = c; - - if((g_nextBranchCycle - nextsCounter) > (u32)nextCounter) //Need to update on counter resets/target changes - { - g_nextBranchCycle = nextsCounter + nextCounter; - } + cpuSetNextBranch( nextsCounter, nextCounter ); //Need to update on counter resets/target changes } // Ignore target diff if target is currently disabled. // (the overflow is all we care about since it goes first, and then the - // target will be turned on afterward). + // target will be turned on afterward, and handled in the next event test). if( counter.target & EECNT_FUTURE_TARGET ) { @@ -107,14 +103,9 @@ static __forceinline void _rcntSet( int cntidx ) if (c < nextCounter) { nextCounter = c; - - if((g_nextBranchCycle - nextsCounter) > (u32)nextCounter) //Need to update on counter resets/target changes - { - g_nextBranchCycle = nextsCounter + nextCounter; - } + cpuSetNextBranch( nextsCounter, nextCounter ); //Need to update on counter resets/target changes } } - //cpuSetNextBranch( nextsCounter, nextCounter ); } diff --git a/pcsx2/IopCounters.cpp b/pcsx2/IopCounters.cpp index c5f5034908..fb38ff3a94 100644 --- a/pcsx2/IopCounters.cpp +++ b/pcsx2/IopCounters.cpp @@ -95,11 +95,7 @@ static void _rcntSet( int cntidx ) if(c < (u64)psxNextCounter) { psxNextCounter = (u32)c; - - if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) //Need to update on counter resets/target changes - { - g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; - } + psxSetNextBranch( psxNextsCounter, psxNextCounter ); //Need to update on counter resets/target changes } //if((counter.mode & 0x10) == 0 || psxCounters[i].target > 0xffff) continue; @@ -111,11 +107,7 @@ static void _rcntSet( int cntidx ) if(c < (u64)psxNextCounter) { psxNextCounter = (u32)c; - - if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) //Need to update on counter resets/target changes - { - g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; - } + psxSetNextBranch( psxNextsCounter, psxNextCounter ); //Need to update on counter resets/target changes } } diff --git a/pcsx2/NakedAsm.h b/pcsx2/NakedAsm.h index 589bdc2957..e9685e9fa6 100644 --- a/pcsx2/NakedAsm.h +++ b/pcsx2/NakedAsm.h @@ -25,43 +25,39 @@ // Common to Windows and Linux extern "C" { -// acoroutine.S -void so_call(coroutine_t coro); -void so_resume(void); -void so_exit(void); + // acoroutine.S + void so_call(coroutine_t coro); + void so_resume(void); + void so_exit(void); -// I can't find where the Linux recRecompile is defined. Is it used anymore? -// If so, namespacing might break it. :/ (air) -void recRecompile( u32 startpc ); + void recRecompile( u32 startpc ); -// aR3000A.S -void iopRecRecompile(u32 startpc); + // aR3000A.S + void iopRecRecompile(u32 startpc); } -// Linux specific #ifdef __LINUX__ -PCSX2_ALIGNED16( u8 _xmm_backup[16*2] ); -PCSX2_ALIGNED16( u8 _mmx_backup[8*4] ); +PCSX2_ALIGNED16_EXTERN( u8 _xmm_backup[16*2] ); +PCSX2_ALIGNED16_EXTERN( u8 _mmx_backup[8*4] ); extern "C" { -// aVUzerorec.S -void* SuperVUGetProgram(u32 startpc, int vuindex); -void SuperVUCleanupProgram(u32 startpc, int vuindex); -void svudispfn(); - -// aR3000A.S -void iopJITCompile(); -void iopJITCompileInBlock(); -void iopDispatcherReg(); - -// aR5900-32.S -void JITCompile(); -void JITCompileInBlock(); -void DispatcherReg(); - + // aVUzerorec.S + void* SuperVUGetProgram(u32 startpc, int vuindex); + void SuperVUCleanupProgram(u32 startpc, int vuindex); + void svudispfn(); + + // aR3000A.S + void iopJITCompile(); + void iopJITCompileInBlock(); + void iopDispatcherReg(); + + // aR5900-32.S + void JITCompile(); + void JITCompileInBlock(); + void DispatcherReg(); } #endif -#endif +#endif diff --git a/pcsx2/x86/BaseblockEx.cpp b/pcsx2/x86/BaseblockEx.cpp index 7d8e5730e2..f660e19944 100644 --- a/pcsx2/x86/BaseblockEx.cpp +++ b/pcsx2/x86/BaseblockEx.cpp @@ -90,12 +90,13 @@ BASEBLOCKEX* BaseBlocks::GetByX86(uptr ip) return &blocks[imin]; } -void BaseBlocks::Link(u32 pc, uptr jumpptr) +void BaseBlocks::Link(u32 pc, s32* jumpptr) { BASEBLOCKEX *targetblock = Get(pc); if (targetblock && targetblock->startpc == pc) - *(u32*)jumpptr = targetblock->fnptr - (jumpptr + 4); + *jumpptr = (s32)(targetblock->fnptr - (sptr)(jumpptr + 1)); else - *(u32*)jumpptr = recompiler - (jumpptr + 4); - links.insert(std::pair(pc, jumpptr)); -} \ No newline at end of file + *jumpptr = (s32)(recompiler - (sptr)(jumpptr + 1)); + links.insert(std::pair(pc, (uptr)jumpptr)); +} + diff --git a/pcsx2/x86/BaseblockEx.h b/pcsx2/x86/BaseblockEx.h index 38f97954c5..db8466b44b 100644 --- a/pcsx2/x86/BaseblockEx.h +++ b/pcsx2/x86/BaseblockEx.h @@ -115,7 +115,7 @@ public: blocks.erase(blocks.begin() + idx); } - void Link(u32 pc, uptr jumpptr); + void Link(u32 pc, s32* jumpptr); __forceinline void Reset() { diff --git a/pcsx2/x86/iR3000A.cpp b/pcsx2/x86/iR3000A.cpp index 2aa7351333..074222ea11 100644 --- a/pcsx2/x86/iR3000A.cpp +++ b/pcsx2/x86/iR3000A.cpp @@ -39,6 +39,8 @@ #include "SamplProf.h" #include "NakedAsm.h" +using namespace x86Emitter; + extern u32 g_psxNextBranchCycle; extern void psxBREAK(); extern void zeroEx(); @@ -57,11 +59,51 @@ uptr psxhwLUT[0x10000]; // R3000A statics int psxreclog = 0; +#ifdef _MSC_VER + +static u32 g_temp; + +// The address for all cleared blocks. It recompiles the current pc and then +// dispatches to the recompiled block address. +static __declspec(naked) void iopJITCompile() +{ + __asm { + mov esi, dword ptr [psxRegs.pc] + push esi + call iopRecRecompile + add esp, 4 + mov ebx, esi + shr esi, 16 + mov ecx, dword ptr [psxRecLUT+esi*4] + jmp dword ptr [ecx+ebx] + } +} + +static __declspec(naked) void iopJITCompileInBlock() +{ + __asm { + jmp iopJITCompile + } +} + +// called when jumping to variable psxpc address +static __declspec(naked) void iopDispatcherReg() +{ + __asm { + mov eax, dword ptr [psxRegs.pc] + mov ebx, eax + shr eax, 16 + mov ecx, dword ptr [psxRecLUT+eax*4] + jmp dword ptr [ecx+ebx] + } +} +#endif // _MSC_VER + + static u8 *recMem = NULL; // the recompiled blocks will be here static BASEBLOCK *recRAM = NULL; // and the ptr to the blocks here static BASEBLOCK *recROM = NULL; // and here static BASEBLOCK *recROM1 = NULL; // also here -void iopJITCompile(); static BaseBlocks recBlocks((uptr)iopJITCompile); static u8 *recPtr = NULL; u32 psxpc; // recompiler psxpc @@ -596,46 +638,6 @@ static void recShutdown() u32 g_psxlastpc = 0; -#ifdef _MSC_VER - -static u32 g_temp; - -// The address for all cleared blocks. It recompiles the current pc and then -// dispatches to the recompiled block address. -static __declspec(naked) void iopJITCompile() -{ - __asm { - mov esi, dword ptr [psxRegs.pc] - push esi - call iopRecRecompile - add esp, 4 - mov ebx, esi - shr esi, 16 - mov ecx, dword ptr [psxRecLUT+esi*4] - jmp dword ptr [ecx+ebx] - } -} - -static __declspec(naked) void iopJITCompileInBlock() -{ - __asm { - jmp iopJITCompile - } -} - -// called when jumping to variable psxpc address -static __declspec(naked) void iopDispatcherReg() -{ - __asm { - mov eax, dword ptr [psxRegs.pc] - mov ebx, eax - shr eax, 16 - mov ecx, dword ptr [psxRecLUT+eax*4] - jmp dword ptr [ecx+ebx] - } -} -#endif // _MSC_VER - static void iopClearRecLUT(BASEBLOCK* base, int count) { for (int i = 0; i < count; i++) @@ -778,7 +780,6 @@ void psxSetBranchReg(u32 reg) void psxSetBranchImm( u32 imm ) { - u32* ptr; psxbranch = 1; assert( imm ); @@ -787,16 +788,9 @@ void psxSetBranchImm( u32 imm ) _psxFlushCall(FLUSH_EVERYTHING); iPsxBranchTest(imm, imm <= psxpc); - ptr = JMP32(0); - recBlocks.Link(HWADDR(imm), (uptr)ptr); + recBlocks.Link(HWADDR(imm), xJcc32()); } -//fixme : this is all a huge hack, we base the counter advancements on the average an opcode should take (wtf?) -// If that wasn't bad enough we have default values like 9/8 which will get cast to int later -// (yeah, that means all sync code couldn't have worked to begin with) -// So for now these are new settings that work. -// (rama) - static __forceinline u32 psxScaleBlockCycles() { return s_psxBlockCycles * (Config.Hacks.IOPCycleDouble ? 2 : 1); @@ -1139,8 +1133,7 @@ StartRecomp: assert( psxpc == s_nEndBlock ); _psxFlushCall(FLUSH_EVERYTHING); MOV32ItoM((uptr)&psxRegs.pc, psxpc); - u32 *ptr = JMP32(0); - recBlocks.Link(HWADDR(s_nEndBlock), (uptr)ptr); + recBlocks.Link(HWADDR(s_nEndBlock), xJcc32() ); psxbranch = 3; } } diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 8f717c7fc1..629198bff4 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -484,15 +484,6 @@ void recResetEE( void ) recLUT_SetPage(recLUT, hwLUT, recROM1, 0xa000, i, i - 0x1e00); } - // drk||Raziel says this is useful but I'm not sure why. Something to do with forward jumps. - // Anyways, it causes random crashing for some reasom, possibly because of memory - // corrupition elsewhere in the recs. I can't reproduce the problem here though, - // so a fix will have to wait until later. -_- (air) - - //x86SetPtr(recMem+REC_CACHEMEM); - //dyna_block_discard_recmem=(u8*)x86Ptr; - //JMP32( (uptr)&dyna_block_discard - ( (u32)x86Ptr + 5 )); - x86SetPtr(recMem); recPtr = recMem; @@ -725,7 +716,7 @@ void recBREAK( void ) { } } } // end namespace R5900::Dynarec::OpcodeImpl // Clears the recLUT table so that all blocks are mapped to the JIT recompiler by default. -static void ClearRecLUT(BASEBLOCK* base, int count) +static __releaseinline void ClearRecLUT(BASEBLOCK* base, int count) { for (int i = 0; i < count; i++) base[i].SetFnptr((uptr)JITCompile); @@ -967,7 +958,7 @@ void iFlushCall(int flushtype) //} -u32 eeScaleBlockCycles() +static u32 scaleBlockCycles_helper() { // Note: s_nBlockCycles is 3 bit fixed point. Divide by 8 when done! @@ -999,12 +990,6 @@ u32 eeScaleBlockCycles() scalarHigh = 7; break; - case 3: // Sync hack x3 - scalarLow = 10; - scalarMid = 19; - scalarHigh = 10; - break; - jNO_DEFAULT } @@ -1016,19 +1001,14 @@ u32 eeScaleBlockCycles() return temp >> (3+2); } -static void iBranch(u32 newpc, int type) +static u32 eeScaleBlockCycles() { - u32* ptr; - - MOV32ItoM((uptr)&cpuRegs.pc, newpc); - if (type == 0) - ptr = JMP32(0); - else if (type == 1) - ptr = JS32(0); - - recBlocks.Link(HWADDR(newpc), (uptr)ptr); + // Ensures block cycles count is never less than 1: + u32 retval = scaleBlockCycles_helper(); + return (retval < 1) ? 1 : retval; } + // Generates dynarec code for Event tests followed by a block dispatch (branch). // Parameters: // newpc - address to jump to at the end of the block. If newpc == 0xffffffff then @@ -1058,20 +1038,23 @@ static void iBranchTest(u32 newpc, bool noDispatch) xCMP(eax, ptr32[&cpuRegs.cycle]); xCMOVL(eax, ptr32[&cpuRegs.cycle]); xMOV(ptr32[&cpuRegs.cycle], eax); - RET(); } else { - MOV32MtoR(EAX, (uptr)&cpuRegs.cycle); - ADD32ItoR(EAX, eeScaleBlockCycles()); - MOV32RtoM((uptr)&cpuRegs.cycle, EAX); // update cycles - SUB32MtoR(EAX, (uptr)&g_nextBranchCycle); - if (!noDispatch) { + xMOV(eax, &cpuRegs.cycle); + xADD(eax, eeScaleBlockCycles()); + xMOV(&cpuRegs.cycle, eax); // update cycles + xSUB(eax, &g_nextBranchCycle); + if (!noDispatch) + { if (newpc == 0xffffffff) - JS32((uptr)DispatcherReg - ( (uptr)x86Ptr + 6 )); + xJS( DispatcherReg ); else - iBranch(newpc, 1); + { + xMOV( ptr32[&cpuRegs.pc], newpc ); + recBlocks.Link( HWADDR(newpc), xJcc32( Jcc_Signed ) ); + } } - RET(); } + xRET(); } static void checkcodefn() @@ -1170,7 +1153,7 @@ void recompileNextInstruction(int delayslot) return; } } - //If thh COP0 DIE bit is disabled, double the cycles. Happens rarely. + //If the COP0 DIE bit is disabled, double the cycles. Happens rarely. s_nBlockCycles += opcode.cycles * (2 - ((cpuRegs.CP0.n.Config >> 18) & 0x1)); opcode.recompile(); @@ -1235,14 +1218,14 @@ void badespfn() { void __fastcall dyna_block_discard(u32 start,u32 sz) { DevCon::WriteLn("dyna_block_discard .. start: %08X count=%d", params start,sz); - Cpu->Clear(start, sz); + recClear(start, sz); } void __fastcall dyna_page_reset(u32 start,u32 sz) { DevCon::WriteLn("dyna_page_reset .. start=%08X size=%d", params start,sz*4); - Cpu->Clear(start & ~0xfffUL, 0x400); + recClear(start & ~0xfffUL, 0x400); manual_counter[start >> 12]++; mmap_MarkCountedRamPage(PSM(start), start & ~0xfffUL); } @@ -1657,10 +1640,14 @@ StartRecomp: // performance reasons. int numinsts = (pc - startpc) / 4; - if( numinsts > 12 ) + if( numinsts > 6 ) iBranchTest(pc); else - iBranch(pc,0); // unconditional static link + { + xMOV( ptr32[&cpuRegs.pc], pc ); + xADD( ptr32[&cpuRegs.cycle], eeScaleBlockCycles() ); + recBlocks.Link( HWADDR(pc), xJcc32() ); + } } } diff --git a/pcsx2/x86/ix86/ix86_instructions.h b/pcsx2/x86/ix86/ix86_instructions.h index 230ca93a5e..889848d4f8 100644 --- a/pcsx2/x86/ix86/ix86_instructions.h +++ b/pcsx2/x86/ix86/ix86_instructions.h @@ -212,6 +212,8 @@ namespace x86Emitter // JMP / Jcc Instructions! extern void xJcc( JccComparisonType comparison, const void* target ); + extern s8* xJcc8( JccComparisonType comparison=Jcc_Unconditional, s8 displacement=0 ); + extern s32* xJcc32( JccComparisonType comparison=Jcc_Unconditional, s32 displacement=0 ); // ------------------------------------------------------------------------ // Conditional jumps to fixed targets. diff --git a/pcsx2/x86/ix86/ix86_jmp.cpp b/pcsx2/x86/ix86/ix86_jmp.cpp index ab5611ce9e..fb304af18b 100644 --- a/pcsx2/x86/ix86/ix86_jmp.cpp +++ b/pcsx2/x86/ix86/ix86_jmp.cpp @@ -73,6 +73,34 @@ xSmartJump::~xSmartJump() m_baseptr = NULL; // just in case (sometimes helps in debugging too) } +// ------------------------------------------------------------------------ +// Emits a 32 bit jump, and returns a pointer to the 32 bit displacement. +// (displacements should be assigned relative to the end of the jump instruction, +// or in other words *(retval+1) ) +__emitinline s32* xJcc32( JccComparisonType comparison, s32 displacement ) +{ + if( comparison == Jcc_Unconditional ) + xWrite8( 0xe9 ); + else + { + xWrite8( 0x0f ); + xWrite8( 0x80 | comparison ); + } + xWrite( displacement ); + + return ((s32*)xGetPtr()) - 1; +} + +// ------------------------------------------------------------------------ +// Emits a 32 bit jump, and returns a pointer to the 8 bit displacement. +// (displacements should be assigned relative to the end of the jump instruction, +// or in other words *(retval+1) ) +__emitinline s8* xJcc8( JccComparisonType comparison, s8 displacement ) +{ + xWrite8( (comparison == Jcc_Unconditional) ? 0xeb : (0x70 | comparison) ); + xWrite( displacement ); + return (s8*)xGetPtr() - 1; +} // ------------------------------------------------------------------------ // Writes a jump at the current x86Ptr, which targets a pre-established target address. @@ -84,7 +112,7 @@ xSmartJump::~xSmartJump() __emitinline void Internal::xJccKnownTarget( JccComparisonType comparison, const void* target, bool slideForward ) { // Calculate the potential j8 displacement first, assuming an instruction length of 2: - sptr displacement8 = (sptr)target - ((sptr)xGetPtr() + 2); + sptr displacement8 = (sptr)target - (sptr)(xGetPtr() + 2); const int slideVal = slideForward ? ((comparison == Jcc_Unconditional) ? 3 : 4) : 0; displacement8 -= slideVal; @@ -94,22 +122,12 @@ __emitinline void Internal::xJccKnownTarget( JccComparisonType comparison, const if( slideForward ) jASSUME( displacement8 >= 0 ); if( is_s8( displacement8 ) ) - { - xWrite8( (comparison == Jcc_Unconditional) ? 0xeb : (0x70 | comparison) ); - xWrite( displacement8 ); - } + xJcc8( comparison, displacement8 ); else { // Perform a 32 bit jump instead. :( - - if( comparison == Jcc_Unconditional ) - xWrite8( 0xe9 ); - else - { - xWrite8( 0x0f ); - xWrite8( 0x80 | comparison ); - } - xWrite( (sptr)target - ((sptr)xGetPtr() + 4) ); + s32* bah = xJcc32( comparison ); + *bah = (s32)target - (s32)xGetPtr(); } } From 6dfe946e7f31f3c10acd000d500d3f70a443fed3 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 7 May 2009 11:15:08 +0000 Subject: [PATCH 023/296] Linux: These weren't intended as externs... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1144 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Makefile.am | 2 +- pcsx2/NakedAsm.h | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/pcsx2/Makefile.am b/pcsx2/Makefile.am index f62070e274..eb5f646a7a 100644 --- a/pcsx2/Makefile.am +++ b/pcsx2/Makefile.am @@ -19,7 +19,7 @@ ps2/Iop/IopHwRead.cpp ps2/Iop/IopHwWrite.cpp ps2/Iop/IopHw_Internal.h libpcsx2_a_SOURCES += \ CDVD.h CDVDiso.h CDVDisodrv.h CDVDlib.h COP0.h Cache.h CdRom.h Common.h Counters.h Decode_XA.h \ Elfheader.h Exceptions.h GS.h Hw.h IopBios.h IopBios2.h IopCounters.h IopDma.h IopHw.h IopMem.h IopSio2.h Memcpyfast.h \ -Memory.h MemoryCard.h Misc.h Patch.h Paths.h Plugins.h PrecompiledHeader.h IopCommon.h R3000A.h R5900.h R5900OpcodeTables.h \ +Memory.h MemoryTypes.h MemoryCard.h Misc.h Patch.h Paths.h Plugins.h PrecompiledHeader.h IopCommon.h R3000A.h R5900.h R5900OpcodeTables.h \ SPR.h SamplProf.h SaveState.h Sif.h Sifcmd.h Sio.h SafeArray.h Stats.h StringUtils.h System.h Threading.h \ VU.h VUflags.h VUmicro.h VUops.h Vif.h VifDma.h cheatscpp.h vtlb.h NakedAsm.h R5900Exceptions.h HostGui.h Pcsx2Config.h \ Dump.h diff --git a/pcsx2/NakedAsm.h b/pcsx2/NakedAsm.h index e9685e9fa6..2ed4621f65 100644 --- a/pcsx2/NakedAsm.h +++ b/pcsx2/NakedAsm.h @@ -38,8 +38,8 @@ extern "C" #ifdef __LINUX__ -PCSX2_ALIGNED16_EXTERN( u8 _xmm_backup[16*2] ); -PCSX2_ALIGNED16_EXTERN( u8 _mmx_backup[8*4] ); +PCSX2_ALIGNED16( u8 _xmm_backup[16*2] ); +PCSX2_ALIGNED16( u8 _mmx_backup[8*4] ); extern "C" { From c64691811acc72ca2e2b1a94468605837cf5e33c Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 7 May 2009 12:34:58 +0000 Subject: [PATCH 024/296] GSdx: 3.4.9. Texture Function => Decal/TCC=RGB: Av = At (gs_user v3 and v5) or Av = Af (v6)? Seems it was Af... Sonic unleashed was one of the games relying on this setting, could not find any other yet. Also removed the green stripes, shadow sw-mode only. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1145 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSCrc.h | 3 ++- plugins/GSdx/GSDrawScanline.cpp | 1 + plugins/GSdx/GSDrawScanlineCodeGenerator.cpp | 22 ++++++++++++++++++-- plugins/GSdx/GSRendererSW.h | 17 ++++++--------- plugins/GSdx/GSSetupPrimCodeGenerator.cpp | 2 +- plugins/GSdx/GSState.cpp | 20 ++++++++++++++++++ plugins/GSdx/GSTextureCacheSW.cpp | 4 +++- plugins/GSdx/GSVertexSW.cpp | 3 +-- plugins/GSdx/GSVertexSW.h | 9 +++++--- plugins/GSdx/res/tfx10.fx | 9 +++++++- plugins/GSdx/res/tfx9.fx | 9 +++++++- 12 files changed, 77 insertions(+), 23 deletions(-) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index 34307df48b..aec9656c20 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -128,6 +128,7 @@ CRC::Game CRC::m_games[] = {0x14FE77F7, TalesOfAbyss, US}, {0x045D77E9, TalesOfAbyss, US}, // undub {0xAA5EC3A3, TalesOfAbyss, JP}, + {0xFB236A46, SonicUnleashed, US}, }; CAtlMap CRC::m_map; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index 12f41bd93f..083b69df1a 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -63,7 +63,8 @@ public: Onimusha3, MajokkoALaMode2, TalesOfAbyss, - TitleCount + SonicUnleashed, + TitleCount, }; enum Region diff --git a/plugins/GSdx/GSDrawScanline.cpp b/plugins/GSdx/GSDrawScanline.cpp index ccdd34f4b7..4813bc0440 100644 --- a/plugins/GSdx/GSDrawScanline.cpp +++ b/plugins/GSdx/GSDrawScanline.cpp @@ -189,6 +189,7 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) sel.iip = m_sel.iip; sel.tfx = m_sel.tfx; + sel.tcc = m_sel.tcc; sel.fst = m_sel.fst; sel.fge = m_sel.fge; sel.sprite = m_sel.sprite; diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp index 5b0c12411c..a6eb172b3e 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp @@ -35,6 +35,11 @@ GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(GSScanlineEnvironment& m_sel.key = key; + if(m_sel.tfx == TFX_DECAL && m_sel.tcc == 0) + { + printf("*** TFX_DECAL && !TCC *** \n"); + } + Generate(); } @@ -449,7 +454,7 @@ void GSDrawScanlineCodeGenerator::Init(int params) } } - if(m_sel.tfx != TFX_DECAL) + if(!(m_sel.tfx == TFX_DECAL && m_sel.tcc)) { if(m_sel.iip) { @@ -582,7 +587,7 @@ void GSDrawScanlineCodeGenerator::Step() } } - if(m_sel.tfx != TFX_DECAL) + if(!(m_sel.tfx == TFX_DECAL && m_sel.tcc)) { if(m_sel.iip) { @@ -1202,6 +1207,19 @@ void GSDrawScanlineCodeGenerator::AlphaTFX() case TFX_DECAL: + // if(!tcc) gat = gat.mix16(ga.srl16(7)); + + if(!m_sel.tcc) + { + // GSVector4i ga = iip ? gaf : m_env.c.ga; + + movdqa(xmm4, xmmword[m_sel.iip ? &m_env.temp.ga : &m_env.c.ga]); + + psrlw(xmm4, 7); + + mix16(xmm6, xmm4, xmm3); + } + break; case TFX_HIGHLIGHT: diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 3189f49a44..c0d2cb0889 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -134,7 +134,7 @@ protected: int amin, amax; - if(PRIM->TME && (context->TEX0.TCC || context->TEX0.TFX == TFX_DECAL)) + if(PRIM->TME && context->TEX0.TCC) { DWORD bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; DWORD cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; @@ -325,11 +325,6 @@ protected: } } - if(p.sel.tfx == TFX_DECAL) - { - p.sel.tcc = 1; - } - if(p.sel.fst == 0) { // skip per pixel division if q is constant @@ -500,7 +495,7 @@ protected: { GS_PRIM_CLASS primclass = GSUtil::GetPrimClass(PRIM->PRIM); - m_vtrace.Update(m_vertices, m_count, primclass, PRIM->IIP, PRIM->TME, m_context->TEX0.TFX); + m_vtrace.Update(m_vertices, m_count, primclass, PRIM->IIP, PRIM->TME, m_context->TEX0.TFX, m_context->TEX0.TCC); if(m_dump) { @@ -522,9 +517,9 @@ protected: str.Format(_T("c:\\temp1\\_%05d_f%I64d_tex_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), (int)m_context->TEX0.TBP0, (int)m_context->TEX0.PSM); if(PRIM->TME) if(s_save) {m_mem.SaveBMP(str, m_context->TEX0.TBP0, m_context->TEX0.TBW, m_context->TEX0.PSM, 1 << m_context->TEX0.TW, 1 << m_context->TEX0.TH);} str.Format(_T("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), m_context->FRAME.Block(), m_context->FRAME.PSM); - if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize(1).cx, 512);}//GetFrameSize(1).cy); + if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);}//GetFrameSize(1).cy); str.Format(_T("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), m_context->ZBUF.Block(), m_context->ZBUF.PSM); - if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize(1).cx, 512);} + if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512);} } GSRasterizerData data; @@ -581,9 +576,9 @@ protected: { CString str; str.Format(_T("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), m_context->FRAME.Block(), m_context->FRAME.PSM); - if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize(1).cx, 512);}//GetFrameSize(1).cy); + if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);}//GetFrameSize(1).cy); str.Format(_T("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), m_context->ZBUF.Block(), m_context->ZBUF.PSM); - if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize(1).cx, 512);} + if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512);} } if(0)//stats.ticks > 1000000) diff --git a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp index d44768d68f..4fa97f8ab6 100644 --- a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp +++ b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp @@ -33,7 +33,7 @@ GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, U m_en.z = m_sel.zb ? 1 : 0; m_en.f = m_sel.fb && m_sel.fge ? 1 : 0; m_en.t = m_sel.fb && m_sel.tfx != TFX_NONE ? 1 : 0; - m_en.c = m_sel.fb && m_sel.tfx != TFX_DECAL ? 1 : 0; + m_en.c = m_sel.fb && !(m_sel.tfx == TFX_DECAL && m_sel.tcc) ? 1 : 0; #if _M_AMD64 #error TODO diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 11d681a337..d6a9a07a30 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -2130,6 +2130,25 @@ bool GSC_TalesOfAbyss(const GSFrameInfo& fi, int& skip) return true; } +bool GSC_SonicUnleashed(const GSFrameInfo& fi, int& skip) +{ + if(skip == 0) + { + if(fi.TME && fi.FBP == 0x02200 && fi.FPSM == PSM_PSMCT16S && fi.TBP0 == 0x00000 && fi.TPSM == PSM_PSMCT16) + { + skip = 1000; // shadow + } + } + else + { + if(fi.TME && fi.FBP == 0x00000 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x02200 && fi.TPSM == PSM_PSMCT16S) + { + skip = 2; + } + } + + return true; +} bool GSState::IsBadFrame(int& skip) { GSFrameInfo fi; @@ -2175,6 +2194,7 @@ bool GSState::IsBadFrame(int& skip) map[CRC::GiTS] = GSC_GiTS; map[CRC::Onimusha3] = GSC_Onimusha3; map[CRC::TalesOfAbyss] = GSC_TalesOfAbyss; + map[CRC::SonicUnleashed] = GSC_SonicUnleashed; } // TODO: just set gsc in SetGameCRC once diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index e22bfb0e06..0344e0c2a7 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -54,7 +54,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE continue; } - if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC == 1 && !(t2->m_TEXA == (GSVector4i)TEXA).alltrue())) + if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && !(t2->m_TEXA == (GSVector4i)TEXA).alltrue())) { continue; } @@ -306,6 +306,8 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX // unfortunatelly a block may be part of the same texture multiple times at different places (when (1 << tw) > (tbw << 6), ex. 1024 > 640), // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side + + // TODO: still bogus if those repeated parts aren't fetched together // m_valid[row] |= col; diff --git a/plugins/GSdx/GSVertexSW.cpp b/plugins/GSdx/GSVertexSW.cpp index a8fbefd0c5..d037c1611f 100644 --- a/plugins/GSdx/GSVertexSW.cpp +++ b/plugins/GSdx/GSVertexSW.cpp @@ -38,8 +38,7 @@ GSVertexTrace::GSVertexTraceCodeGenerator::GSVertexTraceCodeGenerator(DWORD key, DWORD primclass = (key >> 0) & 3; DWORD iip = (key >> 2) & 1; DWORD tme = (key >> 3) & 1; - DWORD tfx = (key >> 4) & 3; - DWORD color = !(tme && tfx == TFX_DECAL); + DWORD color = (key >> 4) & 1; int n = 1; diff --git a/plugins/GSdx/GSVertexSW.h b/plugins/GSdx/GSVertexSW.h index 44e0d0ca25..ac21c0bd6d 100644 --- a/plugins/GSdx/GSVertexSW.h +++ b/plugins/GSdx/GSVertexSW.h @@ -244,11 +244,14 @@ public: struct {DWORD xyzf:4, stq:4, rgba:4;}; } m_eq; - void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, DWORD iip, DWORD tme, DWORD tfx) + void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, DWORD iip, DWORD tme, DWORD tfx, DWORD tcc) { - if(!tme) tfx = 0; + DWORD key = primclass | (iip << 2) | (tme << 3); - DWORD key = primclass | (iip << 2) | (tme << 3) | (tfx << 4); + if(!(tme && tfx == TFX_DECAL && tcc)) + { + key |= 1 << 4; + } m_map.Lookup(key)(v, count, m_min, m_max); diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index fe101d1674..8f0fe77a9b 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -423,7 +423,14 @@ PS_OUTPUT ps_main(PS_INPUT input) } else if(TFX == 1) { - c = t; + if(TCC == 0) + { + c.rgb = t.rgb; + } + else + { + c = t; + } } else if(TFX == 2) { diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index d3c9d05d85..d825cf187b 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -250,7 +250,14 @@ float4 ps_main(PS_INPUT input) : COLOR } else if(TFX == 1) { - c = t; + if(TCC == 0) + { + c.rgb = t.rgb; + } + else + { + c = t; + } } else if(TFX == 2) { From 9c0a978d5c1ebc652e8ed2d30ad32aed88a7e1e7 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 7 May 2009 15:56:10 +0000 Subject: [PATCH 025/296] IopMem: Added FireWire port handling to the new IopHardware handlers (oops!). EErec: Removed an instance of duplicated cycle counting for split blocks. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1146 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/ps2/Iop/IopHwRead.cpp | 6 +++++- pcsx2/ps2/Iop/IopHwWrite.cpp | 6 +++++- pcsx2/ps2/Iop/IopHw_Internal.h | 10 +++++----- pcsx2/x86/ix86-32/iR5900-32.cpp | 3 --- 4 files changed, 15 insertions(+), 10 deletions(-) diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index b89c3bfae1..372d67683d 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -395,7 +395,7 @@ mem32_t __fastcall iopHwRead32_Page8( u32 addr ) const int parm = (masked_addr-0x240) / 8; ret = (masked_addr & 4) ? sio2_getSend2( parm ) : sio2_getSend1( parm ); } - else + else if( masked_addr <= 0x280 ) { switch( masked_addr ) { @@ -423,6 +423,10 @@ mem32_t __fastcall iopHwRead32_Page8( u32 addr ) break; } } + else if( addr >= pgmsk(HW_FW_START) && addr <= pgmsk(HW_FW_END) ) + { + ret = FWread32( addr ); + } } else ret = psxHu32(addr); diff --git a/pcsx2/ps2/Iop/IopHwWrite.cpp b/pcsx2/ps2/Iop/IopHwWrite.cpp index d8e413375d..9b6bbb80e9 100644 --- a/pcsx2/ps2/Iop/IopHwWrite.cpp +++ b/pcsx2/ps2/Iop/IopHwWrite.cpp @@ -521,7 +521,7 @@ void __fastcall iopHwWrite32_Page8( u32 addr, mem32_t val ) const int parm = (masked_addr-0x240) / 8; if(masked_addr & 4) sio2_setSend2( parm, val ); else sio2_setSend1( parm, val ); } - else + else if( masked_addr <= 0x280 ) { switch( masked_addr ) { @@ -536,6 +536,10 @@ void __fastcall iopHwWrite32_Page8( u32 addr, mem32_t val ) break; } } + else if( addr >= pgmsk(HW_FW_START) && addr <= pgmsk(HW_FW_END) ) + { + FWwrite32( addr, val ); + } } else psxHu32(addr) = val; diff --git a/pcsx2/ps2/Iop/IopHw_Internal.h b/pcsx2/ps2/Iop/IopHw_Internal.h index c15e7d9a0b..a4eaab9252 100644 --- a/pcsx2/ps2/Iop/IopHw_Internal.h +++ b/pcsx2/ps2/Iop/IopHw_Internal.h @@ -190,13 +190,13 @@ static __forceinline const char* _log_GetIopHwName( u32 addr ) { return "SPU2"; } - else if( addr >= 0x1f808200 ) + else if( addr >= pgmsk(HW_FW_START) && addr <= pgmsk(HW_FW_END) ) { - if( addr < 0x1f808240 ) - return "SIO2 param"; - else if( addr < 0x260 ) - return "SIO2 send"; + return "FW"; } + else if( addr >= 0x1f808200 && addr < 0x1f808240 ) { return "SIO2 param"; } + else if( addr >= 0x1f808240 && addr < 0x1f808260 ) { return "SIO2 send"; } + return "Unknown"; } } diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 629198bff4..17f820ff57 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -1623,11 +1623,8 @@ StartRecomp: } else { - assert( branch != 3 ); if( branch ) assert( !willbranch3 ); - else - ADD32ItoM((int)&cpuRegs.cycle, eeScaleBlockCycles() ); if( willbranch3 || !branch) { From 8ff0e69f0c7d5e948281a6392d908e970b1b5721 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 7 May 2009 15:58:51 +0000 Subject: [PATCH 026/296] Gah... dumb mistake. >_< git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1147 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/ps2/Iop/IopHwRead.cpp | 2 +- pcsx2/ps2/Iop/IopHwWrite.cpp | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/pcsx2/ps2/Iop/IopHwRead.cpp b/pcsx2/ps2/Iop/IopHwRead.cpp index 372d67683d..455691fe24 100644 --- a/pcsx2/ps2/Iop/IopHwRead.cpp +++ b/pcsx2/ps2/Iop/IopHwRead.cpp @@ -423,7 +423,7 @@ mem32_t __fastcall iopHwRead32_Page8( u32 addr ) break; } } - else if( addr >= pgmsk(HW_FW_START) && addr <= pgmsk(HW_FW_END) ) + else if( masked_addr >= pgmsk(HW_FW_START) && masked_addr <= pgmsk(HW_FW_END) ) { ret = FWread32( addr ); } diff --git a/pcsx2/ps2/Iop/IopHwWrite.cpp b/pcsx2/ps2/Iop/IopHwWrite.cpp index 9b6bbb80e9..0a05edf598 100644 --- a/pcsx2/ps2/Iop/IopHwWrite.cpp +++ b/pcsx2/ps2/Iop/IopHwWrite.cpp @@ -536,7 +536,7 @@ void __fastcall iopHwWrite32_Page8( u32 addr, mem32_t val ) break; } } - else if( addr >= pgmsk(HW_FW_START) && addr <= pgmsk(HW_FW_END) ) + else if( masked_addr >= pgmsk(HW_FW_START) && masked_addr <= pgmsk(HW_FW_END) ) { FWwrite32( addr, val ); } From a0a4abdcc1b13cf108dcf0860e46ef599d685168 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 7 May 2009 23:37:57 +0000 Subject: [PATCH 027/296] Zeropad: Break up linux.cpp into multiple files. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1149 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/gui.cpp | 289 +++++++++++++++ plugins/zeropad/Linux/joystick.cpp | 154 ++++++++ plugins/zeropad/Linux/joystick.h | 126 +++++++ plugins/zeropad/Linux/linux.cpp | 549 +++-------------------------- plugins/zeropad/Makefile.am | 2 +- 5 files changed, 626 insertions(+), 494 deletions(-) create mode 100644 plugins/zeropad/Linux/gui.cpp create mode 100644 plugins/zeropad/Linux/joystick.cpp create mode 100644 plugins/zeropad/Linux/joystick.h diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp new file mode 100644 index 0000000000..8812cabc0b --- /dev/null +++ b/plugins/zeropad/Linux/gui.cpp @@ -0,0 +1,289 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include +#include +#include + +#define JOYSTICK_SUPPORT +#ifdef JOYSTICK_SUPPORT +#include "joystick.h" +#endif + +#include "zeropad.h" + +extern "C" +{ +#include "interface.h" +#include "support.h" +#include "callbacks.h" +} + +extern string s_strIniPath; +GtkWidget *Conf = NULL, *s_devicecombo = NULL; +extern void UpdateConf(int pad); + +void SaveConfig() +{ + int i, j; + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "w"); + if (f == NULL) + { + printf("ZeroPAD: failed to save ini %s\n", s_strIniPath.c_str()); + return; + } + + for (j = 0; j < 2 * PADSUBKEYS; j++) + { + for (i = 0; i < PADKEYS; i++) + { + fprintf(f, "[%d][%d] = 0x%lx\n", j, i, conf.keys[j][i]); + } + } + fprintf(f, "log = %d\n", conf.log); + fprintf(f, "options = %d\n", conf.options); + fclose(f); +} + +void LoadConfig() +{ + FILE *f; + char str[256]; + char cfg[255]; + int i, j; + + memset(&conf, 0, sizeof(conf)); + conf.keys[0][0] = XK_a; // L2 + conf.keys[0][1] = XK_semicolon; // R2 + conf.keys[0][2] = XK_w; // L1 + conf.keys[0][3] = XK_p; // R1 + conf.keys[0][4] = XK_i; // TRIANGLE + conf.keys[0][5] = XK_l; // CIRCLE + conf.keys[0][6] = XK_k; // CROSS + conf.keys[0][7] = XK_j; // SQUARE + conf.keys[0][8] = XK_v; // SELECT + conf.keys[0][11] = XK_n; // START + conf.keys[0][12] = XK_e; // UP + conf.keys[0][13] = XK_f; // RIGHT + conf.keys[0][14] = XK_d; // DOWN + conf.keys[0][15] = XK_s; // LEFT + conf.log = 0; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "r"); + if (f == NULL) + { + printf("ZeroPAD: failed to load ini %s\n", s_strIniPath.c_str()); + SaveConfig(); //save and return + return; + } + + for (j = 0; j < 2 * PADSUBKEYS; j++) + { + for (i = 0; i < PADKEYS; i++) + { + sprintf(str, "[%d][%d] = 0x%%x\n", j, i); + if (fscanf(f, str, &conf.keys[j][i]) == 0) conf.keys[j][i] = 0; + } + } + fscanf(f, "log = %d\n", &conf.log); + fscanf(f, "options = %d\n", &conf.options); + fclose(f); +} + +GtkWidget *MsgDlg; + +void OnMsg_Ok() +{ + gtk_widget_destroy(MsgDlg); + gtk_main_quit(); +} + +// GUI event handlers +void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) +{ + int joyid = gtk_combo_box_get_active(combobox); + + // unassign every joystick with this pad + for (int i = 0; i < (int)s_vjoysticks.size(); ++i) + { + if (s_vjoysticks[i]->GetPAD() == s_selectedpad) s_vjoysticks[i]->Assign(-1); + } + + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->Assign(s_selectedpad); +} + +void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; +} + +void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + int mask = PADOPTION_REVERTLY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; +} + +void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + int mask = PADOPTION_REVERTRX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; +} + +void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + int mask = PADOPTION_REVERTRY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + conf.options |= mask; + else + conf.options &= ~mask; +} + +void on_forcefeedback_toggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) + { + conf.options |= mask; + + int joyid = gtk_combo_box_get_active(GTK_COMBO_BOX(s_devicecombo)); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->TestForce(); + } + else + { + conf.options &= ~mask; + } +} + +void SysMessage(char *fmt, ...) +{ + GtkWidget *Ok, *Txt; + GtkWidget *Box, *Box1; + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + if (msg[strlen(msg)-1] == '\n') msg[strlen(msg)-1] = 0; + + MsgDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); + gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "GSsoft Msg"); + gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); + + Box = gtk_vbox_new(5, 0); + gtk_container_add(GTK_CONTAINER(MsgDlg), Box); + gtk_widget_show(Box); + + Txt = gtk_label_new(msg); + + gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); + gtk_widget_show(Txt); + + Box1 = gtk_hbutton_box_new(); + gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); + gtk_widget_show(Box1); + + Ok = gtk_button_new_with_label("Ok"); + gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); + gtk_container_add(GTK_CONTAINER(Box1), Ok); + GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); + gtk_widget_show(Ok); + + gtk_widget_show(MsgDlg); + + gtk_main(); +} + +void OnConf_Pad1(GtkButton *button, gpointer user_data) +{ + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(0); +} + +void OnConf_Pad2(GtkButton *button, gpointer user_data) +{ + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(1); +} + +void OnConf_Pad3(GtkButton *button, gpointer user_data) +{ + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(2); +} + +void OnConf_Pad4(GtkButton *button, gpointer user_data) +{ + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) + UpdateConf(3); +} + +void OnConf_Ok(GtkButton *button, gpointer user_data) +{ +// conf.analog = gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(Analog)); + SaveConfig(); + + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void OnConf_Cancel(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(Conf); + gtk_main_quit(); + LoadConfig(); // load previous config +} + +GtkWidget *About = NULL; + +void OnAbout_Ok(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(About); + gtk_main_quit(); +} + +void CALLBACK PADabout() +{ + + About = create_About(); + + gtk_widget_show_all(About); + gtk_main(); +} + +s32 CALLBACK PADtest() +{ + return 0; +} \ No newline at end of file diff --git a/plugins/zeropad/Linux/joystick.cpp b/plugins/zeropad/Linux/joystick.cpp new file mode 100644 index 0000000000..35d28c7c75 --- /dev/null +++ b/plugins/zeropad/Linux/joystick.cpp @@ -0,0 +1,154 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "joystick.h" + +////////////////////////// +// Joystick definitions // +////////////////////////// + +int s_selectedpad = 0; +vector s_vjoysticks; +static u32 s_bSDLInit = false; + +// opens handles to all possible joysticks +void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) +{ + + if (!s_bSDLInit) + { + if (SDL_Init(SDL_INIT_JOYSTICK) < 0) return; + SDL_JoystickEventState(SDL_QUERY); + s_bSDLInit = true; + } + + vector::iterator it = vjoysticks.begin(); + + // Delete everything in the vector vjoysticks. + while (it < vjoysticks.end()) + { + delete *it; + it ++; + } + + vjoysticks.resize(SDL_NumJoysticks()); + + for (int i = 0; i < (int)vjoysticks.size(); ++i) + { + vjoysticks[i] = new JoystickInfo(); + vjoysticks[i]->Init(i, true); + } + + // set the pads + for (int pad = 0; pad < 2; ++pad) + { + // select the right joystick id + int joyid = -1; + + for (int i = 0; i < PADKEYS; ++i) + { + if (IS_JOYSTICK(conf.keys[pad][i]) || IS_JOYBUTTONS(conf.keys[pad][i])) + { + joyid = PAD_GETJOYID(conf.keys[pad][i]); + break; + } + } + + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) s_vjoysticks[joyid]->Assign(pad); + } + +} + +JoystickInfo::JoystickInfo() +{ + joy = NULL; + + _id = -1; + pad = -1; + axisrange = 0x7fff; + deadzone = 2000; +} + +void JoystickInfo::Destroy() +{ + if (joy != NULL) + { + if (SDL_JoystickOpened(_id)) SDL_JoystickClose(joy); + joy = NULL; + } +} + +bool JoystickInfo::Init(int id, bool bStartThread) +{ + Destroy(); + _id = id; + + joy = SDL_JoystickOpen(id); + if (joy == NULL) + { + printf("failed to open joystick %d\n", id); + return false; + } + + numaxes = SDL_JoystickNumAxes(joy); + numbuttons = SDL_JoystickNumButtons(joy); + numpov = SDL_JoystickNumHats(joy); + numhats = SDL_JoystickNumHats(joy); + devname = SDL_JoystickName(id); + vbutstate.resize(numbuttons); + vaxisstate.resize(numbuttons); + + return true; +} + +// assigns a joystick to a pad +void JoystickInfo::Assign(int newpad) +{ + if (pad == newpad) return; + pad = newpad; + + if (pad >= 0) + { + for (int i = 0; i < PADKEYS; ++i) + { + if (IS_JOYBUTTONS(conf.keys[pad][i])) + { + conf.keys[pad][i] = PAD_JOYBUTTON(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); + } + else if (IS_JOYSTICK(conf.keys[pad][i])) + { + conf.keys[pad][i] = PAD_JOYSTICK(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); + } + } + } +} + +void JoystickInfo::SaveState() +{ + for (int i = 0; i < numbuttons; ++i) + vbutstate[i] = SDL_JoystickGetButton(joy, i); + for (int i = 0; i < numaxes; ++i) + vaxisstate[i] = SDL_JoystickGetAxis(joy, i); + + //for (int i = 0; i < numhats; ++i) + // vhatstate[i] = SDL_JoystickGetHat(joy, i); +} + +void JoystickInfo::TestForce() +{ +} diff --git a/plugins/zeropad/Linux/joystick.h b/plugins/zeropad/Linux/joystick.h new file mode 100644 index 0000000000..b7e53ac37b --- /dev/null +++ b/plugins/zeropad/Linux/joystick.h @@ -0,0 +1,126 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include "zeropad.h" + +// holds all joystick info +class JoystickInfo +{ + public: + JoystickInfo(); + ~JoystickInfo() + { + Destroy(); + } + + void Destroy(); + // opens handles to all possible joysticks + static void EnumerateJoysticks(vector& vjoysticks); + + bool Init(int id, bool bStartThread = true); // opens a handle and gets information + void Assign(int pad); // assigns a joystick to a pad + + void TestForce(); + + const string& GetName() + { + return devname; + } + int GetNumButtons() + { + return numbuttons; + } + int GetNumAxes() + { + return numaxes; + } + int GetNumPOV() + { + return numpov; + } + /*int GetNumHats() + { + return numhats; + }*/ + int GetId() + { + return _id; + } + int GetPAD() + { + return pad; + } + int GetDeadzone(int axis) + { + return deadzone; + } + + void SaveState(); + int GetButtonState(int i) + { + return vbutstate[i]; + } + int GetAxisState(int i) + { + return vaxisstate[i]; + } + /*int GetHatState(int i) + { + return vhatstate[i]; + }*/ + void SetButtonState(int i, int state) + { + vbutstate[i] = state; + } + void SetAxisState(int i, int value) + { + vaxisstate[i] = value; + } + /*void SetHatState(int i, int value) + { + vhatstate[i] = value; + }*/ + SDL_Joystick* GetJoy() + { + return joy; + } + + private: + + string devname; // pretty device name + int _id; + int numbuttons, numaxes, numpov, numhats; + int axisrange, deadzone; + int pad; + + vector vbutstate, vaxisstate; + + SDL_Joystick* joy; +}; + + +extern int s_selectedpad; +extern vector s_vjoysticks; +//extern void JoystickInfo::EnumerateJoysticks(vector& vjoysticks); +/*extern JoystickInfo::JoystickInfo(); +extern void JoystickInfo::Destroy(); +extern bool JoystickInfo::Init(int id, bool bStartThread); +extern void JoystickInfo::Assign(int newpad); +extern void JoystickInfo::SaveState(); +extern void JoystickInfo::TestForce();*/ \ No newline at end of file diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 6cea5ae5fe..0418741d43 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -22,7 +22,7 @@ #define JOYSTICK_SUPPORT #ifdef JOYSTICK_SUPPORT -#include +#include "joystick.h" #endif #include "zeropad.h" @@ -37,126 +37,11 @@ extern "C" Display *GSdsp; static pthread_spinlock_t s_mutexStatus; static u32 s_keyPress[2], s_keyRelease[2]; // thread safe -static u32 s_bSDLInit = false; - -// holds all joystick info -class JoystickInfo -{ - public: - JoystickInfo(); - ~JoystickInfo() - { - Destroy(); - } - - void Destroy(); - // opens handles to all possible joysticks - static void EnumerateJoysticks(vector& vjoysticks); - - bool Init(int id, bool bStartThread = true); // opens a handle and gets information - void Assign(int pad); // assigns a joystick to a pad - - void TestForce(); - - const string& GetName() - { - return devname; - } - int GetNumButtons() - { - return numbuttons; - } - int GetNumAxes() - { - return numaxes; - } - int GetNumPOV() - { - return numpov; - } - int GetId() - { - return _id; - } - int GetPAD() - { - return pad; - } - int GetDeadzone(int axis) - { - return deadzone; - } - - void SaveState(); - int GetButtonState(int i) - { - return vbutstate[i]; - } - int GetAxisState(int i) - { - return vaxisstate[i]; - } - void SetButtonState(int i, int state) - { - vbutstate[i] = state; - } - void SetAxisState(int i, int value) - { - vaxisstate[i] = value; - } -#ifdef JOYSTICK_SUPPORT - SDL_Joystick* GetJoy() - { - return joy; - } -#endif - - private: - - string devname; // pretty device name - int _id; - int numbuttons, numaxes, numpov; - int axisrange, deadzone; - int pad; - - vector vbutstate, vaxisstate; - -#ifdef JOYSTICK_SUPPORT - SDL_Joystick* joy; -#endif -}; - -static vector s_vjoysticks; +extern GtkWidget *Conf, *s_devicecombo; extern string s_strIniPath; -void SaveConfig() -{ - int i, j; - FILE *f; - char cfg[255]; - - strcpy(cfg, s_strIniPath.c_str()); - f = fopen(cfg, "w"); - if (f == NULL) - { - printf("ZeroPAD: failed to save ini %s\n", s_strIniPath.c_str()); - return; - } - - for (j = 0; j < 2 * PADSUBKEYS; j++) - { - for (i = 0; i < PADKEYS; i++) - { - fprintf(f, "[%d][%d] = 0x%lx\n", j, i, conf.keys[j][i]); - } - } - fprintf(f, "log = %d\n", conf.log); - fprintf(f, "options = %d\n", conf.options); - fclose(f); -} - -static char* s_pGuiKeyMap[] = +static const char* s_pGuiKeyMap[] = { "L2", "R2", "L1", "R1", "Triangle", "Circle", "Cross", "Square", @@ -172,102 +57,6 @@ string GetLabelFromButton(const char* buttonname) return label; } -void LoadConfig() -{ - FILE *f; - char str[256]; - char cfg[255]; - int i, j; - - memset(&conf, 0, sizeof(conf)); - conf.keys[0][0] = XK_a; // L2 - conf.keys[0][1] = XK_semicolon; // R2 - conf.keys[0][2] = XK_w; // L1 - conf.keys[0][3] = XK_p; // R1 - conf.keys[0][4] = XK_i; // TRIANGLE - conf.keys[0][5] = XK_l; // CIRCLE - conf.keys[0][6] = XK_k; // CROSS - conf.keys[0][7] = XK_j; // SQUARE - conf.keys[0][8] = XK_v; // SELECT - conf.keys[0][11] = XK_n; // START - conf.keys[0][12] = XK_e; // UP - conf.keys[0][13] = XK_f; // RIGHT - conf.keys[0][14] = XK_d; // DOWN - conf.keys[0][15] = XK_s; // LEFT - conf.log = 0; - - strcpy(cfg, s_strIniPath.c_str()); - f = fopen(cfg, "r"); - if (f == NULL) - { - printf("ZeroPAD: failed to load ini %s\n", s_strIniPath.c_str()); - SaveConfig(); //save and return - return; - } - - for (j = 0; j < 2 * PADSUBKEYS; j++) - { - for (i = 0; i < PADKEYS; i++) - { - sprintf(str, "[%d][%d] = 0x%%x\n", j, i); - if (fscanf(f, str, &conf.keys[j][i]) == 0) conf.keys[j][i] = 0; - } - } - fscanf(f, "log = %d\n", &conf.log); - fscanf(f, "options = %d\n", &conf.options); - fclose(f); -} - -GtkWidget *MsgDlg; - -void OnMsg_Ok() -{ - gtk_widget_destroy(MsgDlg); - gtk_main_quit(); -} - -void SysMessage(char *fmt, ...) -{ - GtkWidget *Ok, *Txt; - GtkWidget *Box, *Box1; - va_list list; - char msg[512]; - - va_start(list, fmt); - vsprintf(msg, fmt, list); - va_end(list); - - if (msg[strlen(msg)-1] == '\n') msg[strlen(msg)-1] = 0; - - MsgDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); - gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); - gtk_window_set_title(GTK_WINDOW(MsgDlg), "GSsoft Msg"); - gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); - - Box = gtk_vbox_new(5, 0); - gtk_container_add(GTK_CONTAINER(MsgDlg), Box); - gtk_widget_show(Box); - - Txt = gtk_label_new(msg); - - gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); - gtk_widget_show(Txt); - - Box1 = gtk_hbutton_box_new(); - gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); - gtk_widget_show(Box1); - - Ok = gtk_button_new_with_label("Ok"); - gtk_signal_connect(GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); - gtk_container_add(GTK_CONTAINER(Box1), Ok); - GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); - gtk_widget_show(Ok); - - gtk_widget_show(MsgDlg); - - gtk_main(); -} - s32 _PADopen(void *pDsp) { GSdsp = *(Display**)pDsp; @@ -276,7 +65,9 @@ s32 _PADopen(void *pDsp) s_keyRelease[0] = s_keyRelease[1] = 0; XAutoRepeatOff(GSdsp); +#ifdef JOYSTICK_SUPPORT JoystickInfo::EnumerateJoysticks(s_vjoysticks); +#endif return 0; } @@ -286,8 +77,15 @@ void _PADclose() pthread_spin_destroy(&s_mutexStatus); XAutoRepeatOn(GSdsp); - vector::iterator it; - FORIT(it, s_vjoysticks) delete *it; + vector::iterator it = s_vjoysticks.begin(); + + // Delete everything in the vector vjoysticks. + while (it < s_vjoysticks.end()) + { + delete *it; + it ++; + } + s_vjoysticks.clear(); } @@ -482,8 +280,7 @@ void CALLBACK PADupdate(int pad) if (abs(value) > (pjoy)->GetDeadzone(value)) { g_lanalog[pad].x = value / 256; - if (conf.options&PADOPTION_REVERTLX) - g_lanalog[pad].x = -g_lanalog[pad].x; + if (conf.options & PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; g_lanalog[pad].x += 0x80; } else @@ -495,8 +292,7 @@ void CALLBACK PADupdate(int pad) if (abs(value) > (pjoy)->GetDeadzone(value)) { g_lanalog[pad].y = value / 256; - if (conf.options&PADOPTION_REVERTLX) - g_lanalog[pad].y = -g_lanalog[pad].y; + if (conf.options & PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; g_lanalog[pad].y += 0x80; } else @@ -508,8 +304,7 @@ void CALLBACK PADupdate(int pad) if (abs(value) > (pjoy)->GetDeadzone(value)) { g_ranalog[pad].x = value / 256; - if (conf.options&PADOPTION_REVERTLX) - g_ranalog[pad].x = -g_ranalog[pad].x; + if (conf.options & PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; g_ranalog[pad].x += 0x80; } else @@ -521,7 +316,7 @@ void CALLBACK PADupdate(int pad) if (abs(value) > (pjoy)->GetDeadzone(value)) { g_ranalog[pad].y = value / 256; - if (conf.options&PADOPTION_REVERTLX) g_ranalog[pad].y = -g_ranalog[pad].y; + if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; g_ranalog[pad].y += 0x80; } else @@ -560,9 +355,6 @@ void CALLBACK PADupdate(int pad) pthread_spin_unlock(&s_mutexStatus); } -static GtkWidget *Conf = NULL, *s_devicecombo = NULL; -static int s_selectedpad = 0; - void UpdateConf(int pad) { s_selectedpad = pad; @@ -654,12 +446,18 @@ void OnConf_Key(GtkButton *button, gpointer user_data) int key = id % PADKEYS; unsigned long *pkey = &conf.keys[pad][key]; - vector::iterator itjoy; - // save the states #ifdef JOYSTICK_SUPPORT + vector::iterator itjoy = s_vjoysticks.begin(); + SDL_JoystickUpdate(); - FORIT(itjoy, s_vjoysticks)(*itjoy)->SaveState(); + + // Save everything in the vector s_vjoysticks. + while (itjoy < s_vjoysticks.end()) + { + (*itjoy)->SaveState(); + itjoy++; + } #endif for (;;) @@ -681,8 +479,11 @@ void OnConf_Key(GtkButton *button, gpointer user_data) } #ifdef JOYSTICK_SUPPORT + itjoy = s_vjoysticks.begin(); + SDL_JoystickUpdate(); - FORIT(itjoy, s_vjoysticks) + + while (itjoy < s_vjoysticks.end()) { // MAKE sure to look for changes in the state!! for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) @@ -740,51 +541,28 @@ void OnConf_Key(GtkButton *button, gpointer user_data) } } } + + + /*for (int i = 0; i < (*itjoy)->GetNumHats(); ++i) + { + int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); + + if (value != (*itjoy)->GetAxisState(i)) + { + *pkey = PAD_HAT((*itjoy)->GetId(), i); + char str[32]; + sprintf(str, "JHat %d", i); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + }*/ + + itjoy++; } #endif } } -void OnConf_Pad1(GtkButton *button, gpointer user_data) -{ - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(0); -} - -void OnConf_Pad2(GtkButton *button, gpointer user_data) -{ - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(1); -} - -void OnConf_Pad3(GtkButton *button, gpointer user_data) -{ - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(2); -} - -void OnConf_Pad4(GtkButton *button, gpointer user_data) -{ - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(3); -} - -void OnConf_Ok(GtkButton *button, gpointer user_data) -{ -// conf.analog = gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(Analog)); - SaveConfig(); - - gtk_widget_destroy(Conf); - gtk_main_quit(); -} - -void OnConf_Cancel(GtkButton *button, gpointer user_data) -{ - gtk_widget_destroy(Conf); - gtk_main_quit(); - LoadConfig(); // load previous config -} - void CALLBACK PADconfigure() { char strcurdir[256]; @@ -797,19 +575,23 @@ void CALLBACK PADconfigure() Conf = create_Conf(); // recreate +#ifdef JOYSTICK_SUPPORT JoystickInfo::EnumerateJoysticks(s_vjoysticks); +#endif - s_devicecombo = lookup_widget(Conf, "joydevicescombo"); + s_devicecombo = lookup_widget(Conf, "joydevicescombo"); // fill the combo char str[255]; - vector::iterator it; + vector::iterator it = s_vjoysticks.begin(); - FORIT(it, s_vjoysticks) + // Delete everything in the vector vjoysticks. + while (it < s_vjoysticks.end()) { sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), - (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()); + (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()/*, (*it)->GetNumHats()*/); // ,hats: %d gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), str); + it++; } gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), "No Gamepad"); @@ -819,222 +601,3 @@ void CALLBACK PADconfigure() gtk_widget_show_all(Conf); gtk_main(); } - -// GUI event handlers -void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) -{ - int joyid = gtk_combo_box_get_active(combobox); - - // unassign every joystick with this pad - for (int i = 0; i < (int)s_vjoysticks.size(); ++i) - { - if (s_vjoysticks[i]->GetPAD() == s_selectedpad) s_vjoysticks[i]->Assign(-1); - } - - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->Assign(s_selectedpad); -} - -void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) -{ - int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); - if (gtk_toggle_button_get_active(togglebutton)) - conf.options |= mask; - else - conf.options &= ~mask; -} - -void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) -{ - int mask = PADOPTION_REVERTLY << (16 * s_selectedpad); - if (gtk_toggle_button_get_active(togglebutton)) - conf.options |= mask; - else - conf.options &= ~mask; -} - -void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer user_data) -{ - int mask = PADOPTION_REVERTRX << (16 * s_selectedpad); - if (gtk_toggle_button_get_active(togglebutton)) - conf.options |= mask; - else - conf.options &= ~mask; -} - -void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) -{ - int mask = PADOPTION_REVERTRY << (16 * s_selectedpad); - if (gtk_toggle_button_get_active(togglebutton)) - conf.options |= mask; - else - conf.options &= ~mask; -} - -void on_forcefeedback_toggled(GtkToggleButton *togglebutton, gpointer user_data) -{ - int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); - if (gtk_toggle_button_get_active(togglebutton)) - { - conf.options |= mask; - - int joyid = gtk_combo_box_get_active(GTK_COMBO_BOX(s_devicecombo)); - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->TestForce(); - } - else - { - conf.options &= ~mask; - } -} - -GtkWidget *About = NULL; - -void OnAbout_Ok(GtkButton *button, gpointer user_data) -{ - gtk_widget_destroy(About); - gtk_main_quit(); -} - -void CALLBACK PADabout() -{ - - About = create_About(); - - gtk_widget_show_all(About); - gtk_main(); -} - -s32 CALLBACK PADtest() -{ - return 0; -} - -////////////////////////// -// Joystick definitions // -////////////////////////// - -// opens handles to all possible joysticks -void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) -{ -#ifdef JOYSTICK_SUPPORT - - if (!s_bSDLInit) - { - if (SDL_Init(SDL_INIT_JOYSTICK) < 0) return; - SDL_JoystickEventState(SDL_QUERY); - s_bSDLInit = true; - } - - vector::iterator it; - FORIT(it, vjoysticks) delete *it; - - vjoysticks.resize(SDL_NumJoysticks()); - for (int i = 0; i < (int)vjoysticks.size(); ++i) - { - vjoysticks[i] = new JoystickInfo(); - vjoysticks[i]->Init(i, true); - } - - // set the pads - for (int pad = 0; pad < 2; ++pad) - { - // select the right joystick id - int joyid = -1; - for (int i = 0; i < PADKEYS; ++i) - { - if (IS_JOYSTICK(conf.keys[pad][i]) || IS_JOYBUTTONS(conf.keys[pad][i])) - { - joyid = PAD_GETJOYID(conf.keys[pad][i]); - break; - } - } - - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) - s_vjoysticks[joyid]->Assign(pad); - } - -#endif -} - -JoystickInfo::JoystickInfo() -{ -#ifdef JOYSTICK_SUPPORT - joy = NULL; -#endif - _id = -1; - pad = -1; - axisrange = 0x7fff; - deadzone = 2000; -} - -void JoystickInfo::Destroy() -{ -#ifdef JOYSTICK_SUPPORT - if (joy != NULL) - { - if (SDL_JoystickOpened(_id)) SDL_JoystickClose(joy); - joy = NULL; - } -#endif -} - -bool JoystickInfo::Init(int id, bool bStartThread) -{ -#ifdef JOYSTICK_SUPPORT - Destroy(); - _id = id; - - joy = SDL_JoystickOpen(id); - if (joy == NULL) - { - printf("failed to open joystick %d\n", id); - return false; - } - - numaxes = SDL_JoystickNumAxes(joy); - numbuttons = SDL_JoystickNumButtons(joy); - numpov = SDL_JoystickNumHats(joy); - devname = SDL_JoystickName(id); - vbutstate.resize(numbuttons); - vaxisstate.resize(numbuttons); - - return true; -#else - return false; -#endif -} - -// assigns a joystick to a pad -void JoystickInfo::Assign(int newpad) -{ - if (pad == newpad) return; - pad = newpad; - - if (pad >= 0) - { - for (int i = 0; i < PADKEYS; ++i) - { - if (IS_JOYBUTTONS(conf.keys[pad][i])) - { - conf.keys[pad][i] = PAD_JOYBUTTON(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); - } - else if (IS_JOYSTICK(conf.keys[pad][i])) - { - conf.keys[pad][i] = PAD_JOYSTICK(_id, PAD_GETJOYBUTTON(conf.keys[pad][i])); - } - } - } -} - -void JoystickInfo::SaveState() -{ -#ifdef JOYSTICK_SUPPORT - for (int i = 0; i < numbuttons; ++i) - vbutstate[i] = SDL_JoystickGetButton(joy, i); - for (int i = 0; i < numaxes; ++i) - vaxisstate[i] = SDL_JoystickGetAxis(joy, i); -#endif -} - -void JoystickInfo::TestForce() -{ -} diff --git a/plugins/zeropad/Makefile.am b/plugins/zeropad/Makefile.am index 08047050a6..81e048bcc9 100644 --- a/plugins/zeropad/Makefile.am +++ b/plugins/zeropad/Makefile.am @@ -27,4 +27,4 @@ libZeroPAD_LDFLAGS= @SHARED_LDFLAGS@ libZeroPAD_LDFLAGS+=-Wl,-soname,@ZEROPAD_SONAME@ libZeroPAD_LDADD=$(libZeroPAD_a_OBJECTS) -libZeroPAD_a_SOURCES = zeropad.cpp Linux/linux.cpp Linux/support.c Linux/interface.c +libZeroPAD_a_SOURCES = zeropad.cpp Linux/gui.cpp Linux/linux.cpp Linux/joystick.cpp Linux/support.c Linux/interface.c From 0005b749b595fea62e335ca6bac34b88445bc5ab Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 7 May 2009 23:49:12 +0000 Subject: [PATCH 028/296] Zeropad: fix minor typo in previous commit that got copied and pasted several times. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1150 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/joystick.cpp | 2 +- plugins/zeropad/Linux/linux.cpp | 9 +++++---- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/plugins/zeropad/Linux/joystick.cpp b/plugins/zeropad/Linux/joystick.cpp index 35d28c7c75..a4af64108d 100644 --- a/plugins/zeropad/Linux/joystick.cpp +++ b/plugins/zeropad/Linux/joystick.cpp @@ -40,7 +40,7 @@ void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) vector::iterator it = vjoysticks.begin(); // Delete everything in the vector vjoysticks. - while (it < vjoysticks.end()) + while (it != vjoysticks.end()) { delete *it; it ++; diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 0418741d43..dcf346c382 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -80,7 +80,7 @@ void _PADclose() vector::iterator it = s_vjoysticks.begin(); // Delete everything in the vector vjoysticks. - while (it < s_vjoysticks.end()) + while (it != s_vjoysticks.end()) { delete *it; it ++; @@ -336,6 +336,7 @@ void CALLBACK PADupdate(int pad) int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); + if (PAD_GETPOVSIGN(key) && (value < -2048)) status[pad] &= ~(1 << i); else if (!PAD_GETPOVSIGN(key) && (value > 2048)) @@ -453,7 +454,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) SDL_JoystickUpdate(); // Save everything in the vector s_vjoysticks. - while (itjoy < s_vjoysticks.end()) + while (itjoy != s_vjoysticks.end()) { (*itjoy)->SaveState(); itjoy++; @@ -483,7 +484,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) SDL_JoystickUpdate(); - while (itjoy < s_vjoysticks.end()) + while (itjoy != s_vjoysticks.end()) { // MAKE sure to look for changes in the state!! for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) @@ -586,7 +587,7 @@ void CALLBACK PADconfigure() vector::iterator it = s_vjoysticks.begin(); // Delete everything in the vector vjoysticks. - while (it < s_vjoysticks.end()) + while (it != s_vjoysticks.end()) { sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()/*, (*it)->GetNumHats()*/); // ,hats: %d From 912185348e93b2ff1c2143c001b28695fdcd684b Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 8 May 2009 02:12:02 +0000 Subject: [PATCH 029/296] microVU: - Fixed up some code so blocks are recompiled less often. - Tweaked flag code. - Minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1151 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 4 ---- pcsx2/x86/microVU.h | 11 ++++++----- pcsx2/x86/microVU_Alloc.h | 10 ++++++++++ pcsx2/x86/microVU_Alloc.inl | 18 +++++++++++------- pcsx2/x86/microVU_Analyze.inl | 13 +++++++++++-- pcsx2/x86/microVU_Compile.inl | 30 +++++++++++++++++++++++------- pcsx2/x86/microVU_Execute.inl | 16 ++++------------ pcsx2/x86/microVU_Lower.inl | 4 ++++ pcsx2/x86/microVU_Misc.h | 3 ++- 9 files changed, 71 insertions(+), 38 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 72e8aa9cf1..bc01d468a4 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -117,10 +117,6 @@ microVUt(void) mVUclear(u32 addr, u32 size) { microVU* mVU = mVUx; memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); mVU->prog.cleared = 1; // Next execution searches/creates a new microprogram - // Note: It might be better to copy old recompiled blocks to the new microprogram rec data - // however, if games primarily do big writes, its probably not worth it. - // The cost of invalidating bad blocks is also kind of expensive, which is another reason - // that its probably not worth it... } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index ecf728a65e..9e1c483b62 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -49,16 +49,17 @@ public: } microBlock* search(microRegInfo* pState) { if (listSize < 0) return NULL; - //if (blockList[0].pState.needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) + if (blockList[0].pState.needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) for (int i = 0; i <= listSize; i++) { if (!memcmp(pState, &blockList[i].pState, sizeof(microRegInfo) - 1)) return &blockList[i]; } - /*} + } else { // Can do Simple Search (Only Matches the Important Pipeline Stuff) for (int i = 0; i <= listSize; i++) { - if ((blockList[i].pState.q == pState->q) && (blockList[i].pState.p == pState->p)) { return &blockList[i]; } + if ((blockList[i].pState.q == pState->q) && (blockList[i].pState.p == pState->p) + && (blockList[i].pState.clip == pState->clip)) { return &blockList[i]; } } - }*/ + } return NULL; } }; @@ -90,7 +91,7 @@ struct microVU { u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size u32 progSize; // VU Micro Program Size (microSize/4) - static const u32 cacheSize = 0x800000; // VU Cache Size + static const u32 cacheSize = 0xb00000; // VU Cache Size microProgManager<0x4000> prog; // Micro Program Data diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_Alloc.h index a16c3f4396..0cd2f4b48e 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_Alloc.h @@ -29,6 +29,10 @@ union regInfo { }; }; +#if defined(_MSC_VER) +#pragma pack(push, 1) +#pragma warning(disable:4996) +#endif struct microRegInfo { regInfo VF[32]; u8 VI[32]; @@ -38,7 +42,12 @@ struct microRegInfo { u8 xgkick; u8 clip; u8 needExactMatch; // If set, block needs an exact match of pipeline state (needs to be last byte in struct) +#if defined(_MSC_VER) }; +#pragma pack(pop) +#else +} __attribute__((packed)); +#endif struct microTempRegInfo { regInfo VF[2]; // Holds cycle info for Fd, VF[0] = Upper Instruction, VF[1] = Lower Instruction @@ -59,6 +68,7 @@ struct microBlock { template struct microAllocInfo { + microBlock* pBlock; // Pointer to a block in mVUblocks microBlock block; // Block/Pipeline info microTempRegInfo regsTemp; // Temp Pipeline info (used so that new pipeline info isn't conflicting between upper and lower instructions in the same cycle) u8 branch; // 0 = No Branch, 1 = B. 2 = BAL, 3~8 = Conditional Branches, 9 = JALR, 10 = JR diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index 1d972814ee..6c9c610f19 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -665,13 +665,17 @@ microVUt(void) mVUallocFMAC26b(int& ACCw, int& ACCr) { // Flag Allocators //------------------------------------------------------------------ -#define getFlagReg(regX, fInst) { \ - switch (fInst) { \ - case 0: regX = gprF0; break; \ - case 1: regX = gprF1; break; \ - case 2: regX = gprF2; break; \ - case 3: regX = gprF3; break; \ - } \ +#define getFlagReg(regX, fInst) { \ + switch (fInst) { \ + case 0: regX = gprF0; break; \ + case 1: regX = gprF1; break; \ + case 2: regX = gprF2; break; \ + case 3: regX = gprF3; break; \ + default: \ + Console::Error("microVU: Flag Instance Error (fInst = %d)", params fInst); \ + regX = gprF0; \ + break; \ + } \ } microVUt(void) mVUallocSFLAGa(int reg, int fInstance) { diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 81512ba1f9..ec91ca1fd1 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -257,7 +257,7 @@ microVUt(void) mVUanalyzeSflag(int It) { if (!It) { mVUinfo |= _isNOP; } else { // Sets _isSflag at instruction that FSxxx opcode reads it's status flag from mVUinfo |= _swapOps; - if (mVUcount < 4) { mVUregs.needExactMatch = 1; } + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } if (mVUcount >= 1) { incPC2(-2); mVUinfo |= _isSflag; incPC2(2); } // Note: _isSflag is used for status flag optimizations. // Do to stalls, it can only be set one instruction prior to the status flag read instruction @@ -285,7 +285,7 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { if (!It) { mVUinfo |= _isNOP; } else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo |= _swapOps; - if (mVUcount < 4) { mVUregs.needExactMatch = 1; } + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } int curPC = iPC; for (int i = mVUcount, j = 0; i > 1; i--, j++) { incPC2(-2); @@ -297,6 +297,15 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { analyzeVIreg2(It, 1); } +//------------------------------------------------------------------ +// Cflag - Clip Flag Opcodes +//------------------------------------------------------------------ + +microVUt(void) mVUanalyzeCflag() { + microVU* mVU = mVUx; + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } +} + //------------------------------------------------------------------ // XGkick //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 6bf8f65dc4..aeb6660033 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -127,6 +127,12 @@ microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { mVUcount = xCount; // Restore count mVUregs.clip = xClip&3; // Note: Clip timing isn't cycle-accurate between block linking; but hopefully doesn't matter + // Temp Hack-fix until flag-algorithm rewrite + for (int i = 0; i < 4; i++) { + bStatus[i] = 0; + bMac[i] = 0; + } + // Setup Last 4 instances of Status/Mac flags (needed for accurate block linking) iPC = endPC; for (int i = 3, j = 3, ii = 1, jj = 1; aCount > 0; ii++, jj++, aCount--) { @@ -238,20 +244,17 @@ microVUt(void) mVUdivSet() { if (doDivFlag) { getFlagReg(flagReg1, fsInstance); if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } - //MOV32RtoR(gprT1, flagReg1); - //AND32ItoR(gprT1, 0xffff0fcf); - //OR32MtoR (gprT1, (uptr)&mVU->divFlag); - //MOV32RtoR(flagReg1, gprT1); AND16ItoR(flagReg1, 0x0fcf); OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } } -microVUt(void) mVUendProgram() { +microVUt(void) mVUendProgram(int fStatus, int fMac) { microVU* mVU = mVUx; incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; + // Save P/Q Regs if (mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_Q].UL, xmmPQ); if (vuIndex) { @@ -259,6 +262,15 @@ microVUt(void) mVUendProgram() { SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_P].UL, xmmPQ); } + // Save Flag Instances + getFlagReg(fStatus, fStatus); + getFlagReg(fMac, fMac); + MOV32RtoR(gprT1, fStatus); + AND32ItoR(gprT1, 0xffff); + SHR32ItoR(fMac, 16); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); + MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, fMac); + //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) //AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag @@ -273,7 +285,7 @@ microVUt(void) mVUtestCycles() { iPC = mVUstartPC; CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); - mVUendProgram(); + mVUendProgram(0, 0); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } @@ -304,6 +316,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info mVUblock.x86ptrStart = thisPtr; pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager + mVUpBlock = pBlock; for (int branch = 0;; ) { incPC(1); @@ -330,6 +343,9 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { int bStatus[4]; int bMac[4]; mVUsetFlags(bStatus, bMac); mVUtestCycles(); + //SysPrintf("bS[0] = %08x, bS[1] = %08x, bS[2] = %08x, bS[3] = %08x\n", bStatus[0], bStatus[1], bStatus[2], bStatus[3]); + //SysPrintf("bM[0] = %08x, bM[1] = %08x, bM[2] = %08x, bM[3] = %08x\n", bMac[0], bMac[1], bMac[2], bMac[3]); + //SysPrintf("mVUcount = %d\n", mVUcount); // Second Pass iPC = mVUstartPC; @@ -414,7 +430,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { mVUprint("microVU: Possible infinite compiling loop!"); } // Do E-bit end stuff here - mVUendProgram(); + mVUendProgram(bStatus[3], bMac[3]); return thisPtr; //ToDo: Save pipeline state? } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index a96647fbbd..b89cf2a29b 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -86,20 +86,14 @@ microVUt(void) mVUdispatcherB() { eeMXCSR = g_sseMXCSR; SSE_LDMXCSR((uptr)&eeMXCSR); - // Save Regs - MOV32RtoR(gprT1, gprF0); // ToDo: Ensure Correct Flag instances - AND32ItoR(gprT1, 0xffff); - SHR32ItoR(gprF0, 16); - MOV32RtoM((uptr)&mVU->regs->VI[REG_R].UL, gprR); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); - MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprF0); - + // Save Regs (Other Regs Saved in mVUcompile) + MOV32RtoM((uptr)&mVU->regs->VI[REG_R].UL, gprR); + SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); + for (int i = 1; i < 16; i++) { if (isMMX(i)) { MOVDMMXtoM((uptr)&mVU->regs->VI[i].UL, mmVI(i)); } } - SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); - // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. if (!vuIndex) { CALLFunc((uptr)mVUcleanUpVU0); } else { CALLFunc((uptr)mVUcleanUpVU1); } @@ -110,8 +104,6 @@ microVUt(void) mVUdispatcherB() { POP32R(EBP); POP32R(EBX); - //write8(0xcc); - if (isMMX(1)) EMMS(); RET(); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index c37cb37c9c..59cb754961 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -431,6 +431,7 @@ microVUf(void) mVU_ESUM() { microVUf(void) mVU_FCAND() { microVU* mVU = mVUx; + pass1 { mVUanalyzeCflag(); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); AND32ItoR(gprT1, _Imm24_); @@ -443,6 +444,7 @@ microVUf(void) mVU_FCAND() { microVUf(void) mVU_FCEQ() { microVU* mVU = mVUx; + pass1 { mVUanalyzeCflag(); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); XOR32ItoR(gprT1, _Imm24_); @@ -455,6 +457,7 @@ microVUf(void) mVU_FCEQ() { microVUf(void) mVU_FCGET() { microVU* mVU = mVUx; + pass1 { mVUanalyzeCflag(); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); AND32ItoR(gprT1, 0xfff); @@ -465,6 +468,7 @@ microVUf(void) mVU_FCGET() { microVUf(void) mVU_FCOR() { microVU* mVU = mVUx; + pass1 { mVUanalyzeCflag(); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); OR32ItoR(gprT1, _Imm24_); diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 07cef16dbc..ac79714c6e 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -152,6 +152,7 @@ declareAllVariables #define mVUbranch mVUallocInfo.branch #define mVUcycles mVUallocInfo.cycles #define mVUcount mVUallocInfo.count +#define mVUpBlock mVUallocInfo.pBlock #define mVUblock mVUallocInfo.block #define mVUregs mVUallocInfo.block.pState #define mVUregsTemp mVUallocInfo.regsTemp @@ -263,5 +264,5 @@ declareAllVariables #define mVUcacheCheck(ptr, start, limit) { \ uptr diff = ptr - start; \ - if (diff >= limit) { Console::Error("microVU Error: Program went over it's cache limit. Size = %x", params diff); } \ + if (diff >= limit) { Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); } \ } From afc5760e6c69a75c9eab814a73cfc2011de05bdf Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 8 May 2009 04:16:25 +0000 Subject: [PATCH 030/296] microVU: Cleanup and added microVU_Flags.inl to the project. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1152 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 + pcsx2/x86/microVU.cpp | 32 ++-- pcsx2/x86/microVU.h | 3 +- pcsx2/x86/microVU_Alloc.h | 33 ++-- pcsx2/x86/microVU_Alloc.inl | 34 ++-- pcsx2/x86/microVU_Analyze.inl | 40 ++--- pcsx2/x86/microVU_Compile.inl | 183 +++------------------ pcsx2/x86/microVU_Execute.inl | 32 ++-- pcsx2/x86/microVU_Flags.inl | 175 ++++++++++++++++++++ pcsx2/x86/microVU_Log.inl | 32 ++-- pcsx2/x86/microVU_Lower.inl | 32 ++-- pcsx2/x86/microVU_Misc.h | 34 +++- pcsx2/x86/microVU_Misc.inl | 2 +- pcsx2/x86/microVU_Tables.inl | 32 ++-- pcsx2/x86/microVU_Upper.inl | 12 +- 15 files changed, 359 insertions(+), 321 deletions(-) create mode 100644 pcsx2/x86/microVU_Flags.inl diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 8f3bee102f..589cc0a396 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -2050,6 +2050,10 @@ RelativePath="..\..\x86\microVU_Execute.inl" > + + diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index bc01d468a4..eda6882dd4 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ // Micro VU recompiler! - author: cottonvibes(@gmail.com) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 9e1c483b62..1a3610136d 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -1,5 +1,5 @@ /* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2009 Pcsx2-Playground Team + * Copyright (C) 2009 Pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -151,5 +151,6 @@ typedef void (*mVUrecCall)(u32, u32) __attribute__((__fastcall)); // Not sure if #include "microVU_Upper.inl" #include "microVU_Lower.inl" #include "microVU_Tables.inl" +#include "microVU_Flags.inl" #include "microVU_Compile.inl" #include "microVU_Execute.inl" diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_Alloc.h index 0cd2f4b48e..9dbb5521d1 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_Alloc.h @@ -1,24 +1,23 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once - union regInfo { u32 reg; struct { diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index 6c9c610f19..cc177f8c4b 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -1,26 +1,26 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU //------------------------------------------------------------------ -// Micro VU - recPass 1 Functions +// Micro VU - Pass 2 Functions //------------------------------------------------------------------ //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index ec91ca1fd1..96943fb0e6 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -1,26 +1,26 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU //------------------------------------------------------------------ -// Micro VU - recPass 0 Functions +// Micro VU - Pass 1 Functions //------------------------------------------------------------------ //------------------------------------------------------------------ @@ -51,7 +51,6 @@ microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { microVU* mVU = mVUx; - //mVUprint("microVU: FMAC1 Opcode"); mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg1(Ft); @@ -64,7 +63,6 @@ microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { microVUt(void) mVUanalyzeFMAC2(int Fs, int Ft) { microVU* mVU = mVUx; - //mVUprint("microVU: FMAC2 Opcode"); analyzeReg1(Fs); analyzeReg2(Ft); } @@ -84,7 +82,6 @@ microVUt(void) mVUanalyzeFMAC2(int Fs, int Ft) { microVUt(void) mVUanalyzeFMAC3(int Fd, int Fs, int Ft) { microVU* mVU = mVUx; - //mVUprint("microVU: FMAC3 Opcode"); mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg3(Ft); @@ -115,7 +112,6 @@ microVUt(void) mVUanalyzeFMAC4(int Fs, int Ft) { microVUt(void) mVUanalyzeIALU1(int Id, int Is, int It) { microVU* mVU = mVUx; - //mVUprint("microVU: IALU1 Opcode"); if (!Id) { mVUinfo |= _isNOP; } analyzeVIreg1(Is); analyzeVIreg1(It); @@ -124,7 +120,6 @@ microVUt(void) mVUanalyzeIALU1(int Id, int Is, int It) { microVUt(void) mVUanalyzeIALU2(int Is, int It) { microVU* mVU = mVUx; - //mVUprint("microVU: IALU2 Opcode"); if (!It) { mVUinfo |= _isNOP; } analyzeVIreg1(Is); analyzeVIreg2(It, 1); @@ -211,7 +206,6 @@ microVUt(void) mVUanalyzeMFP(int Ft) { microVUt(void) mVUanalyzeLQ(int Ft, int Is, bool writeIs) { microVU* mVU = mVUx; - //mVUprint("microVU: LQ Opcode"); analyzeVIreg1(Is); analyzeReg2(Ft); if (!Ft) { mVUinfo |= (writeIs && Is) ? _noWriteVF : _isNOP; } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index aeb6660033..0cdfd8e36d 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU @@ -34,20 +34,6 @@ else { ajmp = JMPcc((uptr)0); } \ break -// ToDo: Fix this properly. -#define flagSetMacro(xFlag, pFlag, xF, yF, zF) { \ - yF += (mVUstall > 3) ? 3 : mVUstall; \ - if (yF > zF) { \ - pFlag += (yF-zF); \ - if (pFlag >= xFlag) pFlag = (xFlag-1); \ - zF++; \ - xF = (yF-zF); \ - zF = yF; \ - yF -= xF; \ - } \ - yF++; \ -} - #define startLoop() { mVUdebug1(); mVUstall = 0; memset(&mVUregsTemp, 0, sizeof(mVUregsTemp)); } #define calcCycles(reg, x) { reg = ((reg > x) ? (reg - x) : 0); } #define tCycles(dest, src) { dest = aMax(dest, src); } @@ -59,132 +45,13 @@ // Helper Functions //------------------------------------------------------------------ -// Optimizes out unneeded status flag updates -microVUt(void) mVUstatusFlagOp() { - microVU* mVU = mVUx; - int curPC = iPC; - int i = mVUcount; - bool runLoop = 1; - if (doStatus) { mVUinfo |= _isSflag; } - else { - for (; i > 0; i--) { - incPC2(-2); - if (isSflag) { runLoop = 0; break; } - if (doStatus) { mVUinfo |= _isSflag; break; } - } - } - if (runLoop) { - for (; i > 0; i--) { - incPC2(-2); - if (isSflag) break; - mVUinfo &= ~_doStatus; - } - } - iPC = curPC; -} - -// Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! -microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { - microVU* mVU = mVUx; - - // Ensure last ~4+ instructions update mac flags - int endPC = iPC; - u32 aCount = 1; // Amount of instructions needed to get 4 valid status/mac flag instances - for (int i = mVUcount, iX = 0; i > 0; i--, aCount++) { - if (doStatus) { mVUinfo |= _doMac; iX++; if ((iX >= 4) || (aCount > 4)) { break; } } - incPC2(-2); - } - - // Status/Mac Flags Setup Code - int xStatus = 8, xMac = 8; // Flag Instances start at #0 on every block ((8&3) == 0) - int pStatus = 3, pMac = 3; - int xClip = mVUregs.clip + 8, pClip = mVUregs.clip + 7; // Clip Instance starts from where it left off - int xS = 0, yS = 1, zS = 0; - int xM = 0, yM = 1, zM = 0; - int xC = 0, yC = 1, zC = 0; - u32 xCount = mVUcount; // Backup count - iPC = mVUstartPC; - for (mVUcount = 0; mVUcount < xCount; mVUcount++) { - if (((xCount - mVUcount) > aCount) && isFSSET) mVUstatusFlagOp(); // Don't Optimize out on the last ~4+ instructions - - flagSetMacro(xStatus, pStatus, xS, yS, zS); // Handles _fvsinstances - flagSetMacro(xMac, pMac, xM, yM, zM); // Handles _fvminstances - flagSetMacro(xClip, pClip, xC, yC, zC); // Handles _fvcinstances - - mVUinfo |= (xStatus&3) << 12; // _fsInstance - mVUinfo |= (xMac&3) << 10; // _fmInstance - mVUinfo |= (xClip&3) << 14; // _fcInstance - - mVUinfo |= (pStatus&3) << 18; // _fvsInstance - mVUinfo |= (pMac&3) << 16; // _fvmInstance - mVUinfo |= (pClip&3) << 20; // _fvcInstance - - if (doStatus||isFSSET||doDivFlag) { xStatus = (xStatus+1); } - if (doMac) { xMac = (xMac+1); } - if (doClip) { xClip = (xClip+1); } - incPC2(2); - } - mVUcount = xCount; // Restore count - mVUregs.clip = xClip&3; // Note: Clip timing isn't cycle-accurate between block linking; but hopefully doesn't matter - - // Temp Hack-fix until flag-algorithm rewrite - for (int i = 0; i < 4; i++) { - bStatus[i] = 0; - bMac[i] = 0; - } - - // Setup Last 4 instances of Status/Mac flags (needed for accurate block linking) - iPC = endPC; - for (int i = 3, j = 3, ii = 1, jj = 1; aCount > 0; ii++, jj++, aCount--) { - if ((doStatus||isFSSET||doDivFlag) && (i >= 0)) { - for (; (ii > 0 && i >= 0); i--, ii--) { xStatus = (xStatus-1) & 3; bStatus[i] = xStatus; } - } - if (doMac && (j >= 0)) { - for (; (jj > 0 && j >= 0); j--, jj--) { xMac = (xMac-1) & 3; bMac[j] = xMac; } - } - incPC2(-2); - } - iPC = endPC; -} - -#define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) -#define getFlagReg2(x) ((x == bStatus[3]) ? gprESP : ((x == bStatus[2]) ? gprR : ((x == bStatus[1]) ? gprT2 : gprT1))) - // Recompiles Code for Proper Flags and Q/P regs on Block Linkings microVUt(void) mVUsetupBranch(int* bStatus, int* bMac) { microVU* mVU = mVUx; mVUprint("mVUsetupBranch"); - PUSH32R(gprR); // Backup gprR - MOV32RtoM((uptr)&mVU->espBackup, gprESP); - - MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); - MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); - MOV32RtoR(gprR, getFlagReg1(bStatus[2])); - MOV32RtoR(gprESP, getFlagReg1(bStatus[3])); - - MOV32RtoR(gprF0, gprT1); - MOV32RtoR(gprF1, gprT2); - MOV32RtoR(gprF2, gprR); - MOV32RtoR(gprF3, gprESP); - - AND32ItoR(gprT1, 0xffff0000); - AND32ItoR(gprT2, 0xffff0000); - AND32ItoR(gprR, 0xffff0000); - AND32ItoR(gprESP, 0xffff0000); - - AND32ItoR(gprF0, 0x0000ffff); - AND32ItoR(gprF1, 0x0000ffff); - AND32ItoR(gprF2, 0x0000ffff); - AND32ItoR(gprF3, 0x0000ffff); - - OR32RtoR(gprF0, getFlagReg2(bMac[0])); - OR32RtoR(gprF1, getFlagReg2(bMac[1])); - OR32RtoR(gprF2, getFlagReg2(bMac[2])); - OR32RtoR(gprF3, getFlagReg2(bMac[3])); - - MOV32MtoR(gprESP, (uptr)&mVU->espBackup); - POP32R(gprR); // Restore gprR + // Shuffle Flag Instances + mVUsetupFlags(bStatus, bMac); // Shuffle P/Q regs since every block starts at instance #0 if (mVU->p || mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, shufflePQ); } @@ -238,17 +105,6 @@ microVUt(void) mVUsetCycles() { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUdivSet() { - microVU* mVU = mVUx; - int flagReg1, flagReg2; - if (doDivFlag) { - getFlagReg(flagReg1, fsInstance); - if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } - AND16ItoR(flagReg1, 0x0fcf); - OR32MtoR (flagReg1, (uptr)&mVU->divFlag); - } -} - microVUt(void) mVUendProgram(int fStatus, int fMac) { microVU* mVU = mVUx; incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) @@ -289,6 +145,7 @@ microVUt(void) mVUtestCycles() { x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } + //------------------------------------------------------------------ // Recompiler //------------------------------------------------------------------ @@ -427,7 +284,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { } } mVUprint("mVUcompile ebit"); - if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { mVUprint("microVU: Possible infinite compiling loop!"); } + if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } // Do E-bit end stuff here mVUendProgram(bStatus[3], bMac[3]); diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index b89cf2a29b..aebc97e94a 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl new file mode 100644 index 0000000000..b42f9cb2c3 --- /dev/null +++ b/pcsx2/x86/microVU_Flags.inl @@ -0,0 +1,175 @@ +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +#pragma once +#ifdef PCSX2_MICROVU + +// ToDo: Fix this properly. +#define flagSetMacro(xFlag, pFlag, xF, yF, zF) { \ + yF += (mVUstall > 3) ? 3 : mVUstall; \ + if (yF > zF) { \ + pFlag += (yF-zF); \ + if (pFlag >= xFlag) pFlag = (xFlag-1); \ + zF++; \ + xF = (yF-zF); \ + zF = yF; \ + yF -= xF; \ + } \ + yF++; \ +} + +// Sets FDIV Flags at the proper time +microVUt(void) mVUdivSet() { + microVU* mVU = mVUx; + int flagReg1, flagReg2; + if (doDivFlag) { + getFlagReg(flagReg1, fsInstance); + if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } + AND16ItoR(flagReg1, 0x0fcf); + OR32MtoR (flagReg1, (uptr)&mVU->divFlag); + } +} + +// Optimizes out unneeded status flag updates +microVUt(void) mVUstatusFlagOp() { + microVU* mVU = mVUx; + int curPC = iPC; + int i = mVUcount; + bool runLoop = 1; + if (doStatus) { mVUinfo |= _isSflag; } + else { + for (; i > 0; i--) { + incPC2(-2); + if (isSflag) { runLoop = 0; break; } + if (doStatus) { mVUinfo |= _isSflag; break; } + } + } + if (runLoop) { + for (; i > 0; i--) { + incPC2(-2); + if (isSflag) break; + mVUinfo &= ~_doStatus; + } + } + iPC = curPC; +} + +// Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! +microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { + microVU* mVU = mVUx; + + // Ensure last ~4+ instructions update mac flags + int endPC = iPC; + u32 aCount = 1; // Amount of instructions needed to get 4 valid status/mac flag instances + for (int i = mVUcount, iX = 0; i > 0; i--, aCount++) { + if (doStatus) { mVUinfo |= _doMac; iX++; if ((iX >= 4) || (aCount > 4)) { break; } } + incPC2(-2); + } + + // Status/Mac Flags Setup Code + int xStatus = 8, xMac = 8; // Flag Instances start at #0 on every block ((8&3) == 0) + int pStatus = 3, pMac = 3; + int xClip = mVUregs.clip + 8, pClip = mVUregs.clip + 7; // Clip Instance starts from where it left off + int xS = 0, yS = 1, zS = 0; + int xM = 0, yM = 1, zM = 0; + int xC = 0, yC = 1, zC = 0; + u32 xCount = mVUcount; // Backup count + iPC = mVUstartPC; + for (mVUcount = 0; mVUcount < xCount; mVUcount++) { + if (((xCount - mVUcount) > aCount) && isFSSET) mVUstatusFlagOp(); // Don't Optimize out on the last ~4+ instructions + + flagSetMacro(xStatus, pStatus, xS, yS, zS); // Handles _fvsinstances + flagSetMacro(xMac, pMac, xM, yM, zM); // Handles _fvminstances + flagSetMacro(xClip, pClip, xC, yC, zC); // Handles _fvcinstances + + mVUinfo |= (xStatus&3) << 12; // _fsInstance + mVUinfo |= (xMac&3) << 10; // _fmInstance + mVUinfo |= (xClip&3) << 14; // _fcInstance + + mVUinfo |= (pStatus&3) << 18; // _fvsInstance + mVUinfo |= (pMac&3) << 16; // _fvmInstance + mVUinfo |= (pClip&3) << 20; // _fvcInstance + + if (doStatus||isFSSET||doDivFlag) { xStatus = (xStatus+1); } + if (doMac) { xMac = (xMac+1); } + if (doClip) { xClip = (xClip+1); } + incPC2(2); + } + mVUcount = xCount; // Restore count + mVUregs.clip = xClip&3; // Note: Clip timing isn't cycle-accurate between block linking; but hopefully doesn't matter + + // Temp Hack-fix until flag-algorithm rewrite + for (int i = 0; i < 4; i++) { + bStatus[i] = 0; + bMac[i] = 0; + } + + // Setup Last 4 instances of Status/Mac flags (needed for accurate block linking) + iPC = endPC; + for (int i = 3, j = 3, ii = 1, jj = 1; aCount > 0; ii++, jj++, aCount--) { + if ((doStatus||isFSSET||doDivFlag) && (i >= 0)) { + for (; (ii > 0 && i >= 0); i--, ii--) { xStatus = (xStatus-1) & 3; bStatus[i] = xStatus; } + } + if (doMac && (j >= 0)) { + for (; (jj > 0 && j >= 0); j--, jj--) { xMac = (xMac-1) & 3; bMac[j] = xMac; } + } + incPC2(-2); + } + iPC = endPC; +} + +#define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) +#define getFlagReg2(x) ((x == bStatus[3]) ? gprESP : ((x == bStatus[2]) ? gprR : ((x == bStatus[1]) ? gprT2 : gprT1))) + +// Recompiles Code for Proper Flags on Block Linkings +microVUt(void) mVUsetupFlags(int* bStatus, int* bMac) { + microVU* mVU = mVUx; + + PUSH32R(gprR); // Backup gprR + MOV32RtoM((uptr)&mVU->espBackup, gprESP); + + MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); + MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); + MOV32RtoR(gprR, getFlagReg1(bStatus[2])); + MOV32RtoR(gprESP, getFlagReg1(bStatus[3])); + + MOV32RtoR(gprF0, gprT1); + MOV32RtoR(gprF1, gprT2); + MOV32RtoR(gprF2, gprR); + MOV32RtoR(gprF3, gprESP); + + AND32ItoR(gprT1, 0xffff0000); + AND32ItoR(gprT2, 0xffff0000); + AND32ItoR(gprR, 0xffff0000); + AND32ItoR(gprESP, 0xffff0000); + + AND32ItoR(gprF0, 0x0000ffff); + AND32ItoR(gprF1, 0x0000ffff); + AND32ItoR(gprF2, 0x0000ffff); + AND32ItoR(gprF3, 0x0000ffff); + + OR32RtoR(gprF0, getFlagReg2(bMac[0])); + OR32RtoR(gprF1, getFlagReg2(bMac[1])); + OR32RtoR(gprF2, getFlagReg2(bMac[2])); + OR32RtoR(gprF3, getFlagReg2(bMac[3])); + + MOV32MtoR(gprESP, (uptr)&mVU->espBackup); + POP32R(gprR); // Restore gprR +} + +#endif //PCSX2_MICROVU \ No newline at end of file diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 9c05a82317..925001fb48 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 59cb754961..a995bb9ddd 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index ac79714c6e..e28852b638 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -1,5 +1,5 @@ /* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2009 Pcsx2-Playground Team + * Copyright (C) 2009 Pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -142,10 +142,12 @@ declareAllVariables #define microVUf(aType) template aType #define microVUq(aType) template __forceinline aType +// Define Passes #define pass1 if (recPass == 0) #define pass2 if (recPass == 1) #define pass3 if (recPass == 2) +// Misc Macros... #define mVUcurProg mVU->prog.prog[mVU->prog.cur] #define mVUblocks mVU->prog.prog[mVU->prog.cur].block #define mVUallocInfo mVU->prog.prog[mVU->prog.cur].allocInfo @@ -169,13 +171,8 @@ declareAllVariables #define bSaveAddr (((xPC + (2 * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) -#define _Fsf_String ((_Fsf_ == 3) ? "w" : ((_Fsf_ == 2) ? "z" : ((_Fsf_ == 1) ? "y" : "x"))) -#define _Ftf_String ((_Ftf_ == 3) ? "w" : ((_Ftf_ == 2) ? "z" : ((_Ftf_ == 1) ? "y" : "x"))) -#define xyzwStr(x,s) (_X_Y_Z_W == x) ? s : -#define _XYZW_String (xyzwStr(1, "w") (xyzwStr(2, "z") (xyzwStr(3, "zw") (xyzwStr(4, "y") (xyzwStr(5, "yw") (xyzwStr(6, "yz") (xyzwStr(7, "yzw") (xyzwStr(8, "x") (xyzwStr(9, "xw") (xyzwStr(10, "xz") (xyzwStr(11, "xzw") (xyzwStr(12, "xy") (xyzwStr(13, "xyw") (xyzwStr(14, "xyz") "xyzw")))))))))))))) -#define _BC_String (_bc_x ? "x" : (_bc_y ? "y" : (_bc_z ? "z" : "w"))) - +// Pass 1 uses these to set mVUinfo #define _isNOP (1<<0) // Skip Lower Instruction #define _isBranch (1<<1) // Cur Instruction is a Branch #define _isEOB (1<<2) // End of Block @@ -206,7 +203,7 @@ declareAllVariables #define _doClip (1<<29) #define _noWriteVF (1<<30) // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) - +// Pass 2 uses these to read mVUinfo #define isNOP (mVUinfo & (1<<0)) #define isBranch (mVUinfo & (1<<1)) #define isEOB (mVUinfo & (1<<2)) @@ -237,9 +234,26 @@ declareAllVariables #define doClip (mVUinfo & (1<<29)) #define noWriteVF (mVUinfo & (1<<30)) -#define isMMX(_VIreg_) 0//(_VIreg_ >= 1 && _VIreg_ <=8) +// Pass 3 Helper Macros +#define _Fsf_String ((_Fsf_ == 3) ? "w" : ((_Fsf_ == 2) ? "z" : ((_Fsf_ == 1) ? "y" : "x"))) +#define _Ftf_String ((_Ftf_ == 3) ? "w" : ((_Ftf_ == 2) ? "z" : ((_Ftf_ == 1) ? "y" : "x"))) +#define xyzwStr(x,s) (_X_Y_Z_W == x) ? s : +#define _XYZW_String (xyzwStr(1, "w") (xyzwStr(2, "z") (xyzwStr(3, "zw") (xyzwStr(4, "y") (xyzwStr(5, "yw") (xyzwStr(6, "yz") (xyzwStr(7, "yzw") (xyzwStr(8, "x") (xyzwStr(9, "xw") (xyzwStr(10, "xz") (xyzwStr(11, "xzw") (xyzwStr(12, "xy") (xyzwStr(13, "xyw") (xyzwStr(14, "xyz") "xyzw")))))))))))))) +#define _BC_String (_bc_x ? "x" : (_bc_y ? "y" : (_bc_z ? "z" : "w"))) +#define mVUlogFtFs() { mVUlog(".%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } +#define mVUlogFd() { mVUlog(".%s vf%02d, vf%02d", _XYZW_String, _Fd_, _Fs_); } +#define mVUlogACC() { mVUlog(".%s ACC, vf%02d", _XYZW_String, _Fs_); } +#define mVUlogFt() { mVUlog(", vf%02d", _Ft_); } +#define mVUlogBC() { mVUlog(", vf%02d%s", _Ft_, _BC_String); } +#define mVUlogI() { mVUlog(", I"); } +#define mVUlogQ() { mVUlog(", Q"); } +#define mVUlogCLIP() { mVUlog("w.xyz vf%02d, vf%02dw", _Fs_, _Ft_); } + +// Store VI regs in mmx regs? +#define isMMX(_VIreg_) 0 //(_VIreg_ >= 1 && _VIreg_ <=8) #define mmVI(_VIreg_) (_VIreg_ - 1) +// Debug Stuff... #ifdef mVUdebug #define mVUprint Console::Status #define mVUdebug1() { \ @@ -254,6 +268,7 @@ declareAllVariables #define mVUdebug1() {} #endif +// Program Logging... #ifdef mVUlogProg #define mVUlog __mVULog #define mVUdumpProg __mVUdumpProgram @@ -262,6 +277,7 @@ declareAllVariables #define mVUdumpProg 0&& #endif +// Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ uptr diff = ptr - start; \ if (diff >= limit) { Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); } \ diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 48c91520e7..cdd08254e9 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -1,5 +1,5 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team +* Copyright (C) 2009 Pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by diff --git a/pcsx2/x86/microVU_Tables.inl b/pcsx2/x86/microVU_Tables.inl index 26dc6ea8cd..fc4b0b8852 100644 --- a/pcsx2/x86/microVU_Tables.inl +++ b/pcsx2/x86/microVU_Tables.inl @@ -1,20 +1,20 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team -* -* This program is free software; you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation; either version 2 of the License, or -* (at your option) any later version. -* -* This program is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have meived a copy of the GNU General Public License -* along with this program; if not, write to the Free Software -* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA -*/ + * Copyright (C) 2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ #pragma once #ifdef PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index f265db53a9..f22a08cf8f 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -1,5 +1,5 @@ /* Pcsx2 - Pc Ps2 Emulator -* Copyright (C) 2009 Pcsx2-Playground Team +* Copyright (C) 2009 Pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -15,6 +15,7 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ + #pragma once #ifdef PCSX2_MICROVU @@ -78,15 +79,6 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX // Helper Macros //------------------------------------------------------------------ -#define mVUlogFtFs() { mVUlog(".%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } -#define mVUlogFd() { mVUlog(".%s vf%02d, vf%02d", _XYZW_String, _Fd_, _Fs_); } -#define mVUlogACC() { mVUlog(".%s ACC, vf%02d", _XYZW_String, _Fs_); } -#define mVUlogFt() { mVUlog(", vf%02d", _Ft_); } -#define mVUlogBC() { mVUlog(", vf%02d%s", _Ft_, _BC_String); } -#define mVUlogI() { mVUlog(", I"); } -#define mVUlogQ() { mVUlog(", Q"); } -#define mVUlogCLIP() { mVUlog("w.xyz vf%02d, vf%02dw", _Fs_, _Ft_); } - // FMAC1 - Normal FMAC Opcodes #define mVU_FMAC1(operation, OPname) { \ microVU* mVU = mVUx; \ From e330c57556ffb83e03810a56aa5e59b5e66c59e3 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 8 May 2009 04:44:32 +0000 Subject: [PATCH 031/296] Minor bug fixes. Reformat iVUzerorec.cpp & get rid of all occurrances of the FORIT macro. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1153 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iR3000Atables.cpp | 5 +- pcsx2/x86/iVUzerorec.cpp | 3521 +++++++++++++---------- pcsx2/x86/ix86/implement/xmm/moremovs.h | 23 +- pcsx2/x86/ix86/ix86_inlines.inl | 5 + pcsx2/x86/ix86/ix86_types.h | 2 +- 5 files changed, 2022 insertions(+), 1534 deletions(-) diff --git a/pcsx2/x86/iR3000Atables.cpp b/pcsx2/x86/iR3000Atables.cpp index fddd6f9690..75b3732054 100644 --- a/pcsx2/x86/iR3000Atables.cpp +++ b/pcsx2/x86/iR3000Atables.cpp @@ -1567,9 +1567,10 @@ void rpsxBGEZAL() _psxFlushAllUnused(); if( PSX_IS_CONST1(_Rs_) ) { - if( g_psxConstRegs[_Rs_] < 0 ) + if( (int)g_psxConstRegs[_Rs_] < 0 ) branchTo = psxpc+4; - else MOV32ItoM((uptr)&psxRegs.GPR.r[31], psxpc+4); + else + MOV32ItoM((uptr)&psxRegs.GPR.r[31], psxpc+4); psxRecompileNextInstruction(1); psxSetBranchImm( branchTo ); diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index b3fcf2fee6..4cfa941f58 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -5,12 +5,12 @@ * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. - * + * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. - * + * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA @@ -49,16 +49,16 @@ extern void iDumpVU1Registers(); // SuperVURec optimization options, uncomment only for debugging purposes #define SUPERVU_CACHING // vu programs are saved and queried via memcompare (should be no reason to disable this) -#define SUPERVU_WRITEBACKS // don't flush the writebacks after every block -#define SUPERVU_X86CACHING // use x86reg caching (faster) (not really. rather lots slower :p (rama) ) -#define SUPERVU_VIBRANCHDELAY // when integers are modified right before a branch that uses the integer, - // the old integer value is used in the branch - // fixes kh2 +#define SUPERVU_WRITEBACKS // don't flush the writebacks after every block +#define SUPERVU_X86CACHING // use x86reg caching (faster) (not really. rather lots slower :p (rama) ) +#define SUPERVU_VIBRANCHDELAY // when integers are modified right before a branch that uses the integer, + // the old integer value is used in the branch, fixes kh2 #define SUPERVU_PROPAGATEFLAGS // the correct behavior of VUs, for some reason superman breaks gfx with it on... +// registers won't be flushed at block boundaries (faster) (nothing noticable speed-wise, causes SPS in Ratchet and clank (Nneeve) ) #ifndef _DEBUG -//#define SUPERVU_INTERCACHING // registers won't be flushed at block boundaries (faster) (nothing noticable speed-wise, causes SPS in Ratchet and clank (Nneeve) ) +//#define SUPERVU_INTERCACHING #endif #define SUPERVU_CHECKCONDITION 0 // has to be 0!! @@ -84,8 +84,8 @@ static const u32 s_MemSize[2] = {VU0_MEMSIZE, VU1_MEMSIZE}; static u8* s_recVUMem = NULL, *s_recVUPtr = NULL; // tables which are defined at the bottom of this massive file. -extern void (*recVU_UPPER_OPCODE[64])( VURegs* VU, s32 info ); -extern void (*recVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); +extern void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info); +extern void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info); #define INST_Q_READ 0x0001 // flush Q #define INST_P_READ 0x0002 // flush P @@ -102,19 +102,20 @@ extern void (*recVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); #define VFFREE_INVALID0 0x80000000 // (vffree[i]&0xf) is invalid -#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); ++(it)) +//#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); ++(it)) #ifdef _DEBUG -u32 s_vucount=0; +u32 s_vucount = 0; static u32 g_vu1lastrec = 0, skipparent = -1; static u32 s_svulast = 0, s_vufnheader; -static u32 badaddrs[][2] = {0,0xffff}; +static u32 badaddrs[][2] = {0, 0xffff}; #endif union VURecRegs { - struct { + struct + { u16 reg; u16 type; }; @@ -130,14 +131,16 @@ struct VuFunctionHeader struct RANGE { RANGE() : pmem(NULL) {} - + u16 start, size; void* pmem; // all the mem }; VuFunctionHeader() : pprogfunc(NULL), startpc(0xffffffff) {} - ~VuFunctionHeader() { - for(vector::iterator it = ranges.begin(); it != ranges.end(); ++it) { + ~VuFunctionHeader() + { + for (vector::iterator it = ranges.begin(); it != ranges.end(); ++it) + { free(it->pmem); } } @@ -160,103 +163,113 @@ struct VuBlockHeader // one vu inst (lower and upper) class VuInstruction { -public: - VuInstruction() { memzero_obj(*this); nParentPc = -1; vicached = -1; } + public: + VuInstruction() + { + memzero_obj(*this); + nParentPc = -1; + vicached = -1; + } - int nParentPc; // used for syncing with flag writes, -1 for no parent + int nParentPc; // used for syncing with flag writes, -1 for no parent - _vuopinfo info; + _vuopinfo info; - _VURegsNum regs[2]; // [0] - lower, [1] - upper - u32 livevars[2]; // live variables right before this inst, [0] - inst, [1] - float - u32 addvars[2]; // live variables to add - u32 usedvars[2]; // set if var is used in the future including vars used in this inst - u32 keepvars[2]; - u16 pqcycles; // the number of cycles to stall if function writes to the regs - u16 type; // INST_ + _VURegsNum regs[2]; // [0] - lower, [1] - upper + u32 livevars[2]; // live variables right before this inst, [0] - inst, [1] - float + u32 addvars[2]; // live variables to add + u32 usedvars[2]; // set if var is used in the future including vars used in this inst + u32 keepvars[2]; + u16 pqcycles; // the number of cycles to stall if function writes to the regs + u16 type; // INST_ - u32 pClipWrite, pMACWrite, pStatusWrite; // addrs to write the flags - u32 vffree[2]; - s8 vfwrite[2], vfread0[2], vfread1[2], vfacc[2]; - s8 vfflush[2]; // extra flush regs - s8 vicached; // if >= 0, then use the cached integer s_VIBranchDelay - VuInstruction *pPrevInst; + u32 pClipWrite, pMACWrite, pStatusWrite; // addrs to write the flags + u32 vffree[2]; + s8 vfwrite[2], vfread0[2], vfread1[2], vfacc[2]; + s8 vfflush[2]; // extra flush regs + s8 vicached; // if >= 0, then use the cached integer s_VIBranchDelay + VuInstruction *pPrevInst; - int SetCachedRegs(int upper, u32 vuxyz); - void Recompile(list::iterator& itinst, u32 vuxyz); + int SetCachedRegs(int upper, u32 vuxyz); + void Recompile(list::iterator& itinst, u32 vuxyz); }; -#define BLOCKTYPE_EOP 0x01 // at least one of the children of the block contains eop (or the block itself) -#define BLOCKTYPE_FUNCTION 0x02 -#define BLOCKTYPE_HASEOP 0x04 // last inst of block is an eop -#define BLOCKTYPE_MACFLAGS 0x08 -#define BLOCKTYPE_ANALYZED 0x40 -#define BLOCKTYPE_IGNORE 0x80 // special for recursive fns -#define BLOCKTYPE_ANALYZEDPARENT 0x100 +enum BlockType +{ + BLOCKTYPE_EOP = 0x01, // at least one of the children of the block contains eop (or the block itself) + BLOCKTYPE_FUNCTION = 0x02, + BLOCKTYPE_HASEOP = 0x04, // last inst of block is an eop + BLOCKTYPE_MACFLAGS = 0x08, + BLOCKTYPE_ANALYZED = 0x40, + BLOCKTYPE_IGNORE = 0x80, // special for recursive fns + BLOCKTYPE_ANALYZEDPARENT = 0x100 +}; // base block used when recompiling class VuBaseBlock { -public: - typedef list LISTBLOCKS; + public: + typedef list LISTBLOCKS; - VuBaseBlock(); - - // returns true if the leads to a EOP (ALL VU blocks must ret true) - void AssignVFRegs(); - void AssignVIRegs(int parent); + VuBaseBlock(); - list::iterator GetInstIterAtPc(int instpc); - void GetInstsAtPc(int instpc, list& listinsts); + // returns true if the leads to a EOP (ALL VU blocks must ret true) + void AssignVFRegs(); + void AssignVIRegs(int parent); - void Recompile(); + list::iterator GetInstIterAtPc(int instpc); + void GetInstsAtPc(int instpc, list& listinsts); - u16 type; // BLOCKTYPE_ - u16 id; - u16 startpc; - u16 endpc; // first inst not in block - void* pcode; // x86 code pointer - void* pendcode; // end of the x86 code pointer - int cycles; - list insts; - list parents; - LISTBLOCKS blocks; // blocks branches to - u32* pChildJumps[4]; // addrs that need to be filled with the children's start addrs - // if highest bit is set, addr needs to be relational - u32 vuxyz; // corresponding bit is set if reg's xyz channels are used only - u32 vuxy; // corresponding bit is set if reg's xyz channels are used only + void Recompile(); - _xmmregs startregs[iREGCNT_XMM], endregs[iREGCNT_XMM]; - int nStartx86, nEndx86; // indices into s_vecRegArray + u16 type; // BLOCKTYPE_ + u16 id; + u16 startpc; + u16 endpc; // first inst not in block + void* pcode; // x86 code pointer + void* pendcode; // end of the x86 code pointer + int cycles; + list insts; + list parents; + LISTBLOCKS blocks; // blocks branches to + u32* pChildJumps[4]; // addrs that need to be filled with the children's start addrs + // if highest bit is set, addr needs to be relational + u32 vuxyz; // corresponding bit is set if reg's xyz channels are used only + u32 vuxy; // corresponding bit is set if reg's xyz channels are used only - int allocX86Regs; - int prevFlagsOutOfBlock; + _xmmregs startregs[iREGCNT_XMM], endregs[iREGCNT_XMM]; + int nStartx86, nEndx86; // indices into s_vecRegArray + + int allocX86Regs; + int prevFlagsOutOfBlock; }; struct WRITEBACK { - WRITEBACK() : nParentPc(0), cycle(0) //, pStatusWrite(NULL), pMACWrite(NULL) - { - viwrite[0] = viwrite[1] = 0; - viread[0] = viread[1] = 0; - } + WRITEBACK() : nParentPc(0), cycle(0) //, pStatusWrite(NULL), pMACWrite(NULL) + { + viwrite[0] = viwrite[1] = 0; + viread[0] = viread[1] = 0; + } void InitInst(VuInstruction* pinst, int cycle) const { - u32 write = viwrite[0]|viwrite[1]; - pinst->type = ((write&(1<type = ((write & (1 << REG_CLIP_FLAG)) ? INST_CLIP_WRITE : 0) | + ((write & (1 << REG_MAC_FLAG)) ? INST_MAC_WRITE : 0) | + ((write & (1 << REG_STATUS_FLAG)) ? INST_STATUS_WRITE : 0) | + ((write & (1 << REG_Q)) ? INST_Q_WRITE : 0); pinst->nParentPc = nParentPc; pinst->info.cycle = cycle; - for(int i = 0; i < 2; ++i) { + for (int i = 0; i < 2; ++i) + { pinst->regs[i].VIwrite = viwrite[i]; pinst->regs[i].VIread = viread[i]; } } - static int SortWritebacks(const WRITEBACK& w1, const WRITEBACK& w2) { + static int SortWritebacks(const WRITEBACK& w1, const WRITEBACK& w2) + { return w1.cycle < w2.cycle; } @@ -277,20 +290,24 @@ struct VUPIPELINES VuBaseBlock::VuBaseBlock() { - type = 0; endpc = 0; cycles = 0; pcode = NULL; id = 0; + type = 0; + endpc = 0; + cycles = 0; + pcode = NULL; + id = 0; memzero_obj(pChildJumps); memzero_obj(startregs); memzero_obj(endregs); allocX86Regs = nStartx86 = nEndx86 = -1; - prevFlagsOutOfBlock = 0; + prevFlagsOutOfBlock = 0; } #define SUPERVU_STACKSIZE 0x1000 static list s_listVUHeaders[2]; static list* s_plistCachedHeaders[2] = {NULL, NULL}; -static VuFunctionHeader** recVUHeaders[2] = {NULL,NULL}; -static VuBlockHeader* recVUBlocks[2] = {NULL,NULL}; +static VuFunctionHeader** recVUHeaders[2] = {NULL, NULL}; +static VuBlockHeader* recVUBlocks[2] = {NULL, NULL}; static u8* recVUStack = NULL, *recVUStackPtr = NULL; static vector<_x86regs> s_vecRegArray(128); @@ -319,68 +336,93 @@ void SuperVUAlloc(int vuindex) { // The old -1 crap has been depreciated on this function. Please // specify either 0 or 1, thanks. - jASSUME( vuindex >= 0 ); - + jASSUME(vuindex >= 0); + // upper 4 bits must be zero! - if( s_recVUMem == NULL ) + if (s_recVUMem == NULL) { // upper 4 bits must be zero! // Changed "first try base" to 0xb800000, since 0x0c000000 liked to fail a lot. (air) s_recVUMem = SysMmapEx(0x0e000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); - if( s_recVUMem == NULL ) + if (s_recVUMem == NULL) { throw Exception::OutOfMemory( - fmt_string( "SuperVU Error > failed to allocate recompiler memory (addr: 0x%x)", (u32)s_recVUMem ) + fmt_string("SuperVU Error > failed to allocate recompiler memory (addr: 0x%x)", (u32)s_recVUMem) ); } - ProfilerRegisterSource( "VURec", s_recVUMem, VU_EXESIZE); + ProfilerRegisterSource("VURec", s_recVUMem, VU_EXESIZE); - if( recVUStack == NULL ) recVUStack = new u8[SUPERVU_STACKSIZE * 4]; + if (recVUStack == NULL) recVUStack = new u8[SUPERVU_STACKSIZE * 4]; } - if( vuindex >= 0 ) + if (vuindex >= 0) { - jASSUME( s_recVUMem != NULL ); + jASSUME(s_recVUMem != NULL); - if( recVUHeaders[vuindex] == NULL ) - recVUHeaders[vuindex] = new VuFunctionHeader* [s_MemSize[vuindex]/8]; - if( recVUBlocks[vuindex] == NULL ) - recVUBlocks[vuindex] = new VuBlockHeader[s_MemSize[vuindex]/8]; - if( s_plistCachedHeaders[vuindex] == NULL ) - s_plistCachedHeaders[vuindex] = new list[s_MemSize[vuindex]/8]; + if (recVUHeaders[vuindex] == NULL) + recVUHeaders[vuindex] = new VuFunctionHeader* [s_MemSize[vuindex] / 8]; + if (recVUBlocks[vuindex] == NULL) + recVUBlocks[vuindex] = new VuBlockHeader[s_MemSize[vuindex] / 8]; + if (s_plistCachedHeaders[vuindex] == NULL) + s_plistCachedHeaders[vuindex] = new list[s_MemSize[vuindex] / 8]; } } +void DestroyCachedHeaders(int vuindex, int j) +{ + list::iterator it = s_plistCachedHeaders[vuindex][j].begin(); + + while (it != s_plistCachedHeaders[vuindex][j].end()) + { + delete *it; + it++; + } + + s_plistCachedHeaders[vuindex][j].clear(); +} + +void DestroyVUHeaders(int vuindex) +{ + list::iterator it = s_listVUHeaders[vuindex].begin(); + + while (it != s_listVUHeaders[vuindex].end()) + { + delete *it; + it++; + } + + s_listVUHeaders[vuindex].clear(); +} + // destroy VU resources void SuperVUDestroy(int vuindex) { list::iterator it; - if( vuindex < 0 ) + if (vuindex < 0) { SuperVUDestroy(0); SuperVUDestroy(1); - ProfilerTerminateSource( "VURec" ); + ProfilerTerminateSource("VURec"); SafeSysMunmap(s_recVUMem, VU_EXESIZE); - safe_delete_array( recVUStack ); + safe_delete_array(recVUStack); } else { - safe_delete_array( recVUHeaders[vuindex] ); - safe_delete_array( recVUBlocks[vuindex] ); + safe_delete_array(recVUHeaders[vuindex]); + safe_delete_array(recVUBlocks[vuindex]); - if( s_plistCachedHeaders[vuindex] != NULL ) { - for(u32 j = 0; j < s_MemSize[vuindex]/8; ++j) { - FORIT(it, s_plistCachedHeaders[vuindex][j]) delete *it; - s_plistCachedHeaders[vuindex][j].clear(); + if (s_plistCachedHeaders[vuindex] != NULL) + { + for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) + { + DestroyCachedHeaders(vuindex, j); } - safe_delete_array( s_plistCachedHeaders[vuindex] ); + safe_delete_array(s_plistCachedHeaders[vuindex]); } - - FORIT(it, s_listVUHeaders[vuindex]) delete *it; - s_listVUHeaders[vuindex].clear(); + DestroyVUHeaders(vuindex); } } @@ -391,14 +433,14 @@ void SuperVUReset(int vuindex) s_vucount = 0; #endif - if( s_recVUMem == NULL ) + if (s_recVUMem == NULL) return; //jASSUME( s_recVUMem != NULL ); - if( vuindex < 0 ) + if (vuindex < 0) { - DbgCon::Status( "SuperVU reset > Resetting recompiler memory and structures." ); + DbgCon::Status("SuperVU reset > Resetting recompiler memory and structures."); // Does this cause problems on VU recompiler resets? It could, if the VU works like // the EE used to, and actually tries to re-enter the recBlock after issuing a clear. (air) @@ -410,20 +452,19 @@ void SuperVUReset(int vuindex) } else { - DbgCon::Status( "SuperVU reset [VU%d] > Resetting the recs and junk", params vuindex ); + DbgCon::Status("SuperVU reset [VU%d] > Resetting the recs and junk", params vuindex); list::iterator it; - if( recVUHeaders[vuindex] ) memset( recVUHeaders[vuindex], 0, sizeof(VuFunctionHeader*) * (s_MemSize[vuindex]/8) ); - if( recVUBlocks[vuindex] ) memset( recVUBlocks[vuindex], 0, sizeof(VuBlockHeader) * (s_MemSize[vuindex]/8) ); - - if( s_plistCachedHeaders[vuindex] != NULL ) { - for(u32 j = 0; j < s_MemSize[vuindex]/8; ++j) { - FORIT(it, s_plistCachedHeaders[vuindex][j]) delete *it; - s_plistCachedHeaders[vuindex][j].clear(); + if (recVUHeaders[vuindex]) memset(recVUHeaders[vuindex], 0, sizeof(VuFunctionHeader*) * (s_MemSize[vuindex] / 8)); + if (recVUBlocks[vuindex]) memset(recVUBlocks[vuindex], 0, sizeof(VuBlockHeader) * (s_MemSize[vuindex] / 8)); + + if (s_plistCachedHeaders[vuindex] != NULL) + { + for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) + { + DestroyCachedHeaders(vuindex, j); } } - - FORIT(it, s_listVUHeaders[vuindex]) delete *it; - s_listVUHeaders[vuindex].clear(); + DestroyVUHeaders(vuindex); } } @@ -432,21 +473,25 @@ void __fastcall SuperVUClear(u32 startpc, u32 size, int vuindex) { vector::iterator itrange; list::iterator it = s_listVUHeaders[vuindex].begin(); - u32 endpc = startpc+(size+(8-(size&7))); // Adding this code to ensure size is always a multiple of 8, it can be simplified to startpc+size if size is always a multiple of 8 (cottonvibes) - while( it != s_listVUHeaders[vuindex].end() ) { + u32 endpc = startpc + (size + (8 - (size & 7))); // Adding this code to ensure size is always a multiple of 8, it can be simplified to startpc+size if size is always a multiple of 8 (cottonvibes) + while (it != s_listVUHeaders[vuindex].end()) + { // for every fn, check if it has code in the range - FORIT(itrange, (*it)->ranges) { - if( startpc < (u32)itrange->start+itrange->size && itrange->start < endpc ) + for(itrange = (*it)->ranges.begin(); itrange != (*it)->ranges.end(); itrange++) + { + if (startpc < (u32)itrange->start + itrange->size && itrange->start < endpc) break; } - if( itrange != (*it)->ranges.end() ) { + if (itrange != (*it)->ranges.end()) + { recVUHeaders[vuindex][(*it)->startpc/8] = NULL; #ifdef SUPERVU_CACHING list* plist = &s_plistCachedHeaders[vuindex][(*it)->startpc/8]; plist->push_back(*it); - if( plist->size() > 30 ) { + if (plist->size() > 30) + { // list is too big, delete //Console::Notice("Performance warning: deleting cached VU program!"); delete plist->front(); @@ -477,33 +522,38 @@ u32 s_TotalVUCycles; // total cycles since start of program execution u32 SuperVUGetVIAddr(int reg, int read) { - assert( s_pCurInst != NULL ); + assert(s_pCurInst != NULL); - switch(reg) { + switch (reg) + { case REG_STATUS_FLAG: - { - u32 addr = (read==2) ? s_PrevStatusWrite : (read ? s_StatusRead : s_pCurInst->pStatusWrite); - assert(!read || addr != 0); - return addr; - } + { + u32 addr = (read == 2) ? s_PrevStatusWrite : (read ? s_StatusRead : s_pCurInst->pStatusWrite); + assert(!read || addr != 0); + return addr; + } case REG_MAC_FLAG: - { - return (read==2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); - } + { + return (read == 2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); + } case REG_CLIP_FLAG: - { - u32 addr = (read==2) ? s_PrevClipWrite : (read ? s_ClipRead : s_pCurInst->pClipWrite); - assert( !read || addr != 0 ); - return addr; - } - case REG_Q: return (read || s_WriteToReadQ) ? (uptr)&VU->VI[REG_Q] : (uptr)&VU->q; - case REG_P: return read ? (uptr)&VU->VI[REG_P] : (uptr)&VU->p; - case REG_I: return s_PrevIWrite; + { + u32 addr = (read == 2) ? s_PrevClipWrite : (read ? s_ClipRead : s_pCurInst->pClipWrite); + assert(!read || addr != 0); + return addr; + } + case REG_Q: + return (read || s_WriteToReadQ) ? (uptr)&VU->VI[REG_Q] : (uptr)&VU->q; + case REG_P: + return read ? (uptr)&VU->VI[REG_P] : (uptr)&VU->p; + case REG_I: + return s_PrevIWrite; } #ifdef SUPERVU_VIBRANCHDELAY - if( (read != 0) && (s_pCurInst->regs[0].pipe == VUPIPE_BRANCH) && (s_pCurInst->vicached >= 0) && (s_pCurInst->vicached == reg) ) { - return (uptr)&s_VIBranchDelay; // test for branch delays + if ((read != 0) && (s_pCurInst->regs[0].pipe == VUPIPE_BRANCH) && (s_pCurInst->vicached >= 0) && (s_pCurInst->vicached == reg)) + { + return (uptr)&s_VIBranchDelay; // test for branch delays } #endif @@ -517,47 +567,54 @@ void SuperVUDumpBlock(list& blocks, int vuindex) u32 *mem; u32 i; - Path::CreateDirectory( "dumps" ); - string filename( Path::Combine( "dumps", fmt_string( "svu%cdump%.4X.txt", s_vu?'0':'1', s_pFnHeader->startpc ) ) ); + Path::CreateDirectory("dumps"); + string filename(Path::Combine("dumps", fmt_string("svu%cdump%.4X.txt", s_vu ? '0' : '1', s_pFnHeader->startpc))); //Console::WriteLn( "dump1 %x => %s", params s_pFnHeader->startpc, filename ); - f = fopen( filename.c_str(), "w" ); + f = fopen(filename.c_str(), "w"); fprintf(f, "Format: upper_inst lower_inst\ntype f:vf_live_vars vf_used_vars i:vi_live_vars vi_used_vars inst_cycle pq_inst\n"); fprintf(f, "Type: %.2x - qread, %.2x - pread, %.2x - clip_write, %.2x - status_write\n" - "%.2x - mac_write, %.2x -qflush\n", - INST_Q_READ, INST_P_READ, INST_CLIP_WRITE, INST_STATUS_WRITE, INST_MAC_WRITE, INST_Q_WRITE); + "%.2x - mac_write, %.2x -qflush\n", + INST_Q_READ, INST_P_READ, INST_CLIP_WRITE, INST_STATUS_WRITE, INST_MAC_WRITE, INST_Q_WRITE); fprintf(f, "XMM: Upper: read0 read1 write acc temp; Lower: read0 read1 write acc temp\n\n"); list::iterator itblock; list::iterator itinst; VuBaseBlock::LISTBLOCKS::iterator itchild; - FORIT(itblock, blocks) { - fprintf(f, "block:%c %x-%x; children: ", ((*itblock)->type&BLOCKTYPE_HASEOP)?'*':' ', - (*itblock)->startpc, (*itblock)->endpc-8); - FORIT(itchild, (*itblock)->blocks) { + for(itblock = blocks.begin(); itblock != blocks.end(); itblock++) + { + fprintf(f, "block:%c %x-%x; children: ", ((*itblock)->type&BLOCKTYPE_HASEOP) ? '*' : ' ', + (*itblock)->startpc, (*itblock)->endpc - 8); + + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { fprintf(f, "%x ", (*itchild)->startpc); } fprintf(f, "; vuxyz = %x, vuxy = %x\n", (*itblock)->vuxyz&(*itblock)->insts.front().usedvars[1], - (*itblock)->vuxy&(*itblock)->insts.front().usedvars[1]); + (*itblock)->vuxy&(*itblock)->insts.front().usedvars[1]); itinst = (*itblock)->insts.begin(); i = (*itblock)->startpc; - while(itinst != (*itblock)->insts.end() ) { - assert( i <= (*itblock)->endpc ); - if( itinst->type & INST_DUMMY ) { - if( itinst->nParentPc >= 0 && !(itinst->type&INST_DUMMY_)) { + while (itinst != (*itblock)->insts.end()) + { + assert(i <= (*itblock)->endpc); + if (itinst->type & INST_DUMMY) + { + if (itinst->nParentPc >= 0 && !(itinst->type&INST_DUMMY_)) + { // search for the parent fprintf(f, "writeback 0x%x (%x)\n", itinst->type, itinst->nParentPc); } } - else { - mem = (u32*)&VU->Micro[i]; - char* pstr = disVU1MicroUF( mem[1], i+4 ); + else + { + mem = (u32*) & VU->Micro[i]; + char* pstr = disVU1MicroUF(mem[1], i + 4); fprintf(f, "%.4x: %-40s", i, pstr); - if( mem[1] & 0x80000000 ) fprintf(f, " I=%f(%.8x)\n", *(float*)mem, mem[0]); - else fprintf(f, "%s\n", disVU1MicroLF( mem[0], i )); + if (mem[1] & 0x80000000) fprintf(f, " I=%f(%.8x)\n", *(float*)mem, mem[0]); + else fprintf(f, "%s\n", disVU1MicroLF(mem[0], i)); i += 8; } @@ -567,75 +624,88 @@ void SuperVUDumpBlock(list& blocks, int vuindex) fprintf(f, "\n"); _x86regs* pregs; - if( (*itblock)->nStartx86 >= 0 || (*itblock)->nEndx86 >= 0 ) { + if ((*itblock)->nStartx86 >= 0 || (*itblock)->nEndx86 >= 0) + { fprintf(f, "X86: AX CX DX BX SP BP SI DI\n"); } - if( (*itblock)->nStartx86 >= 0 ) { + if ((*itblock)->nStartx86 >= 0) + { pregs = &s_vecRegArray[(*itblock)->nStartx86]; fprintf(f, "STR: "); - for(i = 0; i < iREGCNT_GPR; ++i) { - if( pregs[i].inuse ) fprintf(f, "%.2d ", pregs[i].reg); - else fprintf(f, "-1 "); + for (i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse) + fprintf(f, "%.2d ", pregs[i].reg); + else + fprintf(f, "-1 "); } fprintf(f, "\n"); } - if( (*itblock)->nEndx86 >= 0 ) { + if ((*itblock)->nEndx86 >= 0) + { fprintf(f, "END: "); pregs = &s_vecRegArray[(*itblock)->nEndx86]; - for(i = 0; i < iREGCNT_GPR; ++i) { - if( pregs[i].inuse ) fprintf(f, "%.2d ", pregs[i].reg); - else fprintf(f, "-1 "); + for (i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse) + fprintf(f, "%.2d ", pregs[i].reg); + else + fprintf(f, "-1 "); } fprintf(f, "\n"); } itinst = (*itblock)->insts.begin(); - for ( i = (*itblock)->startpc; i < (*itblock)->endpc; ++itinst ) { + for (i = (*itblock)->startpc; i < (*itblock)->endpc; ++itinst) + { - if( itinst->type & INST_DUMMY ) { + if (itinst->type & INST_DUMMY) + { } - else { + else + { sprintf(str, "%.4x:%x f:%.8x_%.8x", i, itinst->type, itinst->livevars[1], itinst->usedvars[1]); fprintf(f, "%-46s i:%.8x_%.8x c:%d pq:%d\n", str, - itinst->livevars[0], itinst->usedvars[0], (int)itinst->info.cycle, (int)itinst->pqcycles ); + itinst->livevars[0], itinst->usedvars[0], (int)itinst->info.cycle, (int)itinst->pqcycles); sprintf(str, "XMM r0:%d r1:%d w:%d a:%d t:%x;", - itinst->vfread0[1], itinst->vfread1[1], itinst->vfwrite[1], itinst->vfacc[1], itinst->vffree[1]); + itinst->vfread0[1], itinst->vfread1[1], itinst->vfwrite[1], itinst->vfacc[1], itinst->vffree[1]); fprintf(f, "%-46s r0:%d r1:%d w:%d a:%d t:%x\n", str, - itinst->vfread0[0], itinst->vfread1[0], itinst->vfwrite[0], itinst->vfacc[0], itinst->vffree[0]); + itinst->vfread0[0], itinst->vfread1[0], itinst->vfwrite[0], itinst->vfacc[0], itinst->vffree[0]); i += 8; } } #ifdef __LINUX__ - // dump the asm - if( (*itblock)->pcode != NULL ) { - char command[255]; - FILE* fasm = fopen( "mydump1", "wb" ); - //Console::WriteLn("writing: %x, %x", params (*itblock)->startpc, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode); - fwrite( (*itblock)->pcode, 1, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode, fasm ); - fclose( fasm ); - sprintf( command, "objdump -D --target=binary --architecture=i386 -M intel mydump1 > tempdump"); - system( command ); - fasm = fopen("tempdump", "r"); - // read all of it and write it to f - fseek(fasm, 0, SEEK_END); - vector vbuffer(ftell(fasm)); - fseek(fasm, 0, SEEK_SET); - fread(&vbuffer[0], vbuffer.size(), 1, fasm); - - fprintf(f, "\n\n"); - fwrite(&vbuffer[0], vbuffer.size(), 1, f); - fclose(fasm); - } + // dump the asm + if ((*itblock)->pcode != NULL) + { + char command[255]; + FILE* fasm = fopen("mydump1", "wb"); + //Console::WriteLn("writing: %x, %x", params (*itblock)->startpc, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode); + fwrite((*itblock)->pcode, 1, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode, fasm); + fclose(fasm); + sprintf(command, "objdump -D --target=binary --architecture=i386 -M intel mydump1 > tempdump"); + system(command); + fasm = fopen("tempdump", "r"); + // read all of it and write it to f + fseek(fasm, 0, SEEK_END); + vector vbuffer(ftell(fasm)); + fseek(fasm, 0, SEEK_SET); + fread(&vbuffer[0], vbuffer.size(), 1, fasm); + + fprintf(f, "\n\n"); + fwrite(&vbuffer[0], vbuffer.size(), 1, f); + fclose(fasm); + } #endif fprintf(f, "\n---------------\n"); } - fclose( f ); + fclose(f); } // uncomment to count svu exec time @@ -644,12 +714,13 @@ void SuperVUDumpBlock(list& blocks, int vuindex) // Private methods void* SuperVUGetProgram(u32 startpc, int vuindex) { - assert( startpc < s_MemSize[vuindex] ); - assert( (startpc%8) == 0 ); - assert( recVUHeaders[vuindex] != NULL ); + assert(startpc < s_MemSize[vuindex]); + assert((startpc % 8) == 0); + assert(recVUHeaders[vuindex] != NULL); VuFunctionHeader** pheader = &recVUHeaders[vuindex][startpc/8]; - if( *pheader == NULL ) { + if (*pheader == NULL) + { #ifdef _DEBUG // if( vuindex ) VU1.VI[REG_TPC].UL = startpc; // else VU0.VI[REG_TPC].UL = startpc; @@ -662,11 +733,13 @@ void* SuperVUGetProgram(u32 startpc, int vuindex) //QueryPerformanceCounter(&svubase); #ifdef SUPERVU_CACHING - void* pmem = (vuindex&1) ? VU1.Micro : VU0.Micro; + void* pmem = (vuindex & 1) ? VU1.Micro : VU0.Micro; // check if program exists in cache list::iterator it; - FORIT(it, s_plistCachedHeaders[vuindex][startpc/8]) { - if( (*it)->IsSame(pmem) ) { + for(it = s_plistCachedHeaders[vuindex][startpc/8].begin(); it != s_plistCachedHeaders[vuindex][startpc/8].end(); it++) + { + if ((*it)->IsSame(pmem)) + { // found, transfer to regular lists void* pfn = (*it)->pprogfunc; recVUHeaders[vuindex][startpc/8] = *it; @@ -679,21 +752,25 @@ void* SuperVUGetProgram(u32 startpc, int vuindex) *pheader = SuperVURecompileProgram(startpc, vuindex); - if( *pheader == NULL ) { - assert( s_TotalVUCycles > 0 ); - if( vuindex ) VU1.VI[REG_TPC].UL = startpc; - else VU0.VI[REG_TPC].UL = startpc; + if (*pheader == NULL) + { + assert(s_TotalVUCycles > 0); + if (vuindex) + VU1.VI[REG_TPC].UL = startpc; + else + VU0.VI[REG_TPC].UL = startpc; + return (void*)SuperVUEndProgram; } //QueryPerformanceCounter(&svufinal); //svutime += (u32)(svufinal.QuadPart-svubase.QuadPart); - assert( (*pheader)->pprogfunc != NULL ); + assert((*pheader)->pprogfunc != NULL); } - //else assert( (*pheader)->IsSame((vuindex&1) ? VU1.Micro : VU0.Micro) ); + //else assert( (*pheader)->IsSame((vuindex&1) ? VU1.Micro : VU0.Micro) ); - assert( (*pheader)->startpc == startpc ); + assert((*pheader)->startpc == startpc); return (*pheader)->pprogfunc; } @@ -702,9 +779,9 @@ bool VuFunctionHeader::IsSame(void* pmem) { #ifdef SUPERVU_CACHING vector::iterator it; - FORIT(it, ranges) + for(it = ranges.begin(); it != ranges.end(); it++) { - if( memcmp_mmx((u8*)pmem+it->start, it->pmem, it->size) ) + if (memcmp_mmx((u8*)pmem + it->start, it->pmem, it->size)) return false; } #endif @@ -713,74 +790,81 @@ bool VuFunctionHeader::IsSame(void* pmem) list::iterator VuBaseBlock::GetInstIterAtPc(int instpc) { - assert( instpc >= 0 ); + assert(instpc >= 0); u32 curpc = startpc; list::iterator it; - for(it = insts.begin(); it != insts.end(); ++it) { - if( it->type & INST_DUMMY ) + for (it = insts.begin(); it != insts.end(); ++it) + { + if (it->type & INST_DUMMY) continue; - if( curpc == instpc ) + if (curpc == instpc) break; curpc += 8; } - if( it != insts.end() ) + if (it != insts.end()) return it; - assert( 0 ); + assert(0); return insts.begin(); } void VuBaseBlock::GetInstsAtPc(int instpc, list& listinsts) { - assert( instpc >= 0 ); + assert(instpc >= 0); listinsts.clear(); u32 curpc = startpc; list::iterator it; - for(it = insts.begin(); it != insts.end(); ++it) { - if( it->type & INST_DUMMY ) + for (it = insts.begin(); it != insts.end(); ++it) + { + if (it->type & INST_DUMMY) continue; - if( curpc == instpc ) + if (curpc == instpc) break; curpc += 8; } - if( it != insts.end() ) { + if (it != insts.end()) + { listinsts.push_back(&(*it)); return; } // look for the pc in other blocks - for(list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) { - if( *itblock == this ) + for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) + { + if (*itblock == this) continue; - if( instpc >= (*itblock)->startpc && instpc < (*itblock)->endpc ) { + if (instpc >= (*itblock)->startpc && instpc < (*itblock)->endpc) + { listinsts.push_back(&(*(*itblock)->GetInstIterAtPc(instpc))); } } - assert(listinsts.size()>0); + assert(listinsts.size() > 0); } static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) { - assert( vuindex < 2 ); - assert( s_recVUPtr != NULL ); + assert(vuindex < 2); + assert(s_recVUPtr != NULL); //Console::WriteLn("svu%c rec: %x", params '0'+vuindex, startpc); // if recPtr reached the mem limit reset whole mem - if ( ( (uptr)s_recVUPtr - (uptr)s_recVUMem ) >= VU_EXESIZE-0x40000 ) { + if (((uptr)s_recVUPtr - (uptr)s_recVUMem) >= VU_EXESIZE - 0x40000) + { //Console::WriteLn("SuperVU reset mem"); - SuperVUReset(0); - SuperVUReset(1); - SuperVUReset(-1); - if( s_TotalVUCycles > 0 ) { + SuperVUReset(0); + SuperVUReset(1); + SuperVUReset(-1); + if (s_TotalVUCycles > 0) + { // already executing, so return NULL return NULL; } @@ -794,7 +878,7 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) s_listVUHeaders[vuindex].push_back(s_pFnHeader); s_pFnHeader->startpc = startpc; - memset( recVUBlocks[s_vu], 0, sizeof(VuBlockHeader) * (s_MemSize[s_vu]/8) ); + memset(recVUBlocks[s_vu], 0, sizeof(VuBlockHeader) * (s_MemSize[s_vu] / 8)); // analyze the global graph s_listBlocks.clear(); @@ -807,17 +891,21 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) // fill parents VuBaseBlock::LISTBLOCKS::iterator itchild; - FORIT(itblock, s_listBlocks) { - FORIT(itchild, (*itblock)->blocks) + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { (*itchild)->parents.push_back(*itblock); + } //(*itblock)->type &= ~(BLOCKTYPE_IGNORE|BLOCKTYPE_ANALYZED); } - assert( s_listBlocks.front()->startpc == startpc ); + assert(s_listBlocks.front()->startpc == startpc); s_listBlocks.front()->type |= BLOCKTYPE_FUNCTION; - FORIT(itblock, s_listBlocks) { + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { SuperVUInitLiveness(*itblock); } @@ -826,8 +914,7 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) SuperVUAssignRegs(); #ifdef _DEBUG - if( (s_vu && (vudump&1)) || (!s_vu && (vudump&16)) ) - SuperVUDumpBlock(s_listBlocks, s_vu); + if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); #endif // code generation @@ -842,9 +929,10 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) VuFunctionHeader::RANGE r; s_pFnHeader->ranges.reserve(s_listBlocks.size()); - FORIT(itblock, s_listBlocks) { + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { r.start = (*itblock)->startpc; - r.size = (*itblock)->endpc-(*itblock)->startpc; + r.size = (*itblock)->endpc - (*itblock)->startpc; #ifdef SUPERVU_CACHING //memxor_mmx(r.checksum, &VU->Micro[r.start], r.size); r.pmem = malloc(r.size); @@ -854,30 +942,31 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) } #if defined(_DEBUG) && defined(__LINUX__) - // dump at the end to capture the actual code - if( (s_vu && (vudump&1)) || (!s_vu && (vudump&16)) ) - SuperVUDumpBlock(s_listBlocks, s_vu); + // dump at the end to capture the actual code + if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); #endif // destroy - for(list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) { + for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) + { delete *itblock; } s_listBlocks.clear(); - assert( s_recVUPtr < s_recVUMem+VU_EXESIZE ); + assert(s_recVUPtr < s_recVUMem + VU_EXESIZE); return s_pFnHeader; } -static int _recbranchAddr(u32 vucode) { +static int _recbranchAddr(u32 vucode) +{ s32 bpc = pc + (_Imm11_ << 3); -/* - if ( bpc < 0 ) { - Console::WriteLn("zerorec branch warning: bpc < 0 ( %x ); Using unsigned imm11", params bpc); - bpc = pc + (_UImm11_ << 3); - }*/ - bpc &= (s_MemSize[s_vu]-1); + /* + if ( bpc < 0 ) { + Console::WriteLn("zerorec branch warning: bpc < 0 ( %x ); Using unsigned imm11", params bpc); + bpc = pc + (_UImm11_ << 3); + }*/ + bpc &= (s_MemSize[s_vu] - 1); return bpc; } @@ -894,20 +983,22 @@ static VuInstruction SuperVUFlushInst() void SuperVUAddWritebacks(VuBaseBlock* pblock, const list& listWritebacks) { #ifdef SUPERVU_WRITEBACKS - // regardless of repetition, add the pipes (for selfloops) - list::const_iterator itwriteback = listWritebacks.begin(); - list::iterator itinst = pblock->insts.begin(), itinst2; + // regardless of repetition, add the pipes (for selfloops) + list::const_iterator itwriteback = listWritebacks.begin(); + list::iterator itinst = pblock->insts.begin(), itinst2; - while(itwriteback != listWritebacks.end()) { - if( itinst != pblock->insts.end() && (itinst->info.cycle < itwriteback->cycle || (itinst->type&INST_DUMMY) ) ) { - ++itinst; - continue; - } - - itinst2 = pblock->insts.insert(itinst, VuInstruction()); - itwriteback->InitInst(&(*itinst2), vucycle); - ++itwriteback; + while (itwriteback != listWritebacks.end()) + { + if (itinst != pblock->insts.end() && (itinst->info.cycle < itwriteback->cycle || (itinst->type&INST_DUMMY))) + { + ++itinst; + continue; } + + itinst2 = pblock->insts.insert(itinst, VuInstruction()); + itwriteback->InitInst(&(*itinst2), vucycle); + ++itwriteback; + } #endif } @@ -938,13 +1029,14 @@ static VuInstruction* getDelayInst(VuInstruction* pInst) int delay = 1; VuInstruction* pDelayInst = NULL; VuInstruction* pTargetInst = pInst->pPrevInst; - while( 1 ) { - if( pTargetInst != NULL - && pTargetInst->info.cycle+delay==pInst->info.cycle - && (pTargetInst->regs[0].pipe == VUPIPE_IALU||pTargetInst->regs[0].pipe == VUPIPE_FMAC) - && ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) - && (delay == 1 || ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) == ((pTargetInst->regs[0].VIwrite & pInst->pPrevInst->regs[0].VIread) & 0xffff)) - && !(pTargetInst->regs[0].VIread&((1<info.cycle + delay == pInst->info.cycle + && (pTargetInst->regs[0].pipe == VUPIPE_IALU || pTargetInst->regs[0].pipe == VUPIPE_FMAC) + && ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) + && (delay == 1 || ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) == ((pTargetInst->regs[0].VIwrite & pInst->pPrevInst->regs[0].VIread) & 0xffff)) + && !(pTargetInst->regs[0].VIread&((1 << REG_STATUS_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_CLIP_FLAG)))) { pDelayInst = pTargetInst; pTargetInst = pTargetInst->pPrevInst; @@ -957,7 +1049,7 @@ static VuInstruction* getDelayInst(VuInstruction* pInst) } else break; } - if( delay > 1 ) DevCon::WriteLn("supervu: %d cycle branch delay detected: %x %x", params delay-1, pc, s_pFnHeader->startpc); + if (delay > 1) DevCon::WriteLn("supervu: %d cycle branch delay detected: %x %x", params delay - 1, pc, s_pFnHeader->startpc); return pDelayInst; } #endif @@ -969,44 +1061,48 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V startpc &= (s_vu ? 0x3fff : 0xfff); VuBlockHeader* pbh = &recVUBlocks[s_vu][startpc/8]; - if ( pbh->pblock != NULL ) { + if (pbh->pblock != NULL) + { VuBaseBlock* pblock = pbh->pblock; list::iterator itinst; - if( pblock->startpc == startpc ) { - SuperVUAddWritebacks(pblock, pipes.listWritebacks); + if (pblock->startpc == startpc) + { + SuperVUAddWritebacks(pblock, pipes.listWritebacks); return pblock; } // have to divide the blocks, pnewblock is first block - assert( startpc > pblock->startpc ); - assert( startpc < pblock->endpc ); + assert(startpc > pblock->startpc); + assert(startpc < pblock->endpc); - u32 dummyinst = (startpc-pblock->startpc)>>3; + u32 dummyinst = (startpc - pblock->startpc) >> 3; // count inst non-dummy insts itinst = pblock->insts.begin(); int cycleoff = 0; - while(dummyinst > 0) { - if( itinst->type & INST_DUMMY ) + while (dummyinst > 0) + { + if (itinst->type & INST_DUMMY) ++itinst; - else { - cycleoff = itinst->info.cycle; + else + { + cycleoff = itinst->info.cycle; ++itinst; --dummyinst; } } // NOTE: still leaves insts with their writebacks in different blocks - while( itinst->type & INST_DUMMY ) + while (itinst->type & INST_DUMMY) ++itinst; - // the difference in cycles between dummy insts (naruto utlimate ninja) - int cyclediff = 0; - if( parent == pblock ) - cyclediff = itinst->info.cycle-cycleoff; + // the difference in cycles between dummy insts (naruto utlimate ninja) + int cyclediff = 0; + if (parent == pblock) + cyclediff = itinst->info.cycle - cycleoff; cycleoff = itinst->info.cycle; // new block @@ -1015,12 +1111,12 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V pnewblock->startpc = startpc; pnewblock->endpc = pblock->endpc; - pnewblock->cycles = pblock->cycles-cycleoff+cyclediff; - + pnewblock->cycles = pblock->cycles - cycleoff + cyclediff; + pnewblock->blocks.splice(pnewblock->blocks.end(), pblock->blocks); pnewblock->insts.splice(pnewblock->insts.end(), pblock->insts, itinst, pblock->insts.end()); pnewblock->type = pblock->type; - + // any writebacks in the next 3 cycles also belong to original block // for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); ) { // if( (itinst->type & INST_DUMMY) && itinst->nParentPc >= 0 && itinst->nParentPc < (int)startpc ) { @@ -1035,14 +1131,17 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V // } pbh = &recVUBlocks[s_vu][startpc/8]; - for(u32 inst = startpc; inst < pblock->endpc; inst += 8) { - if( pbh->pblock == pblock ) + for (u32 inst = startpc; inst < pblock->endpc; inst += 8) + { + if (pbh->pblock == pblock) pbh->pblock = pnewblock; ++pbh; } - FORIT(itinst, pnewblock->insts) + for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); itinst++) + { itinst->info.cycle -= cycleoff; + } SuperVUAddWritebacks(pnewblock, pipes.listWritebacks); @@ -1058,7 +1157,7 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V VuBaseBlock* pblock = new VuBaseBlock(); s_listBlocks.push_back(pblock); - + int i = 0; branch = 0; pc = startpc; @@ -1088,16 +1187,19 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V #endif // first analysis pass for status flags - while(1) { - u32* ptr = (u32*)&VU->Micro[pc]; + while (1) + { + u32* ptr = (u32*) & VU->Micro[pc]; pc += 8; int prevbranch = branch; - if( ptr[1] & 0x40000000 ) + if (ptr[1] & 0x40000000) branch = 1; - if( !(ptr[1] & 0x80000000) ) { // not I - switch( ptr[0]>>25 ) { + if (!(ptr[1] & 0x80000000)) // not I + { + switch (ptr[0] >> 25) + { case 0x24: // jr case 0x25: // jalr case 0x20: // B @@ -1117,7 +1219,8 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V break; case 0x16: // fsand - if( (ptr[0]&0xc0) ) { + if ((ptr[0]&0xc0)) + { // sometimes full sticky bits are needed (simple series 2000 - oane chapara) //Console::WriteLn("needSticky: %x-%x", params s_pFnHeader->startpc, startpc); needFullStatusFlag = 2; @@ -1126,38 +1229,43 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V } } - if( prevbranch ) + if (prevbranch) break; - if( pc >= s_MemSize[s_vu] ) { - Console::Error( "inf vu0 prog %x",params startpc); - break; - } + if (pc >= s_MemSize[s_vu]) + { + Console::Error("inf vu0 prog %x", params startpc); + break; + } } // second full pass pc = startpc; branch = 0; - VuInstruction* pprevinst=NULL, *pinst = NULL; + VuInstruction* pprevinst = NULL, *pinst = NULL; - while(1) { + while (1) + { - if( pc == s_MemSize[s_vu] ) { - branch |= 8; - break; - } + if (pc == s_MemSize[s_vu]) + { + branch |= 8; + break; + } - if( !branch && pbh->pblock != NULL ) { + if (!branch && pbh->pblock != NULL) + { pblock->blocks.push_back(pbh->pblock); break; } int prevbranch = branch; - if( !prevbranch ) { + if (!prevbranch) + { pbh->pblock = pblock; } - else assert( prevbranch || pbh->pblock == NULL); + else assert(prevbranch || pbh->pblock == NULL); pblock->insts.push_back(VuInstruction()); @@ -1167,35 +1275,42 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V SuperVUAnalyzeOp(VU, &pinst->info, pinst->regs); #ifdef SUPERVU_VIBRANCHDELAY - if( pinst->regs[0].pipe == VUPIPE_BRANCH && pblock->insts.size() > 1 ) { - + if (pinst->regs[0].pipe == VUPIPE_BRANCH && pblock->insts.size() > 1) + { + VuInstruction* pdelayinst = getDelayInst(pinst); - if( pdelayinst ) { + if (pdelayinst) + { pdelayinst->type |= INST_CACHE_VI; // find the correct register u32 mask = pdelayinst->regs[0].VIwrite & pinst->regs[0].VIread; - for(int i = 0; i < 16; ++i) { - if( mask & (1<vicached = i; break; - } + } } pinst->vicached = pdelayinst->vicached; } - } + } #endif - if( prevbranch ) { - if( pinst->regs[0].pipe == VUPIPE_BRANCH ) + if (prevbranch) + { + if (pinst->regs[0].pipe == VUPIPE_BRANCH) hasSecondBranch = 1; pinst->type |= INST_BRANCH_DELAY; } // check write back - for(itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ) { - if( pinst->info.cycle >= itwriteback->cycle ) { + for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end();) + { + if (pinst->info.cycle >= itwriteback->cycle) + { itinst = pblock->insts.insert(--pblock->insts.end(), VuInstruction()); itwriteback->InitInst(&(*itinst), pinst->info.cycle); itwriteback = listWritebacks.erase(itwriteback); @@ -1205,78 +1320,86 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V // add new writebacks WRITEBACK w; - const u32 allflags = (1<regs[j].VIwrite & allflags; + const u32 allflags = (1 << REG_CLIP_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG); + for (int j = 0; j < 2; ++j) w.viwrite[j] = pinst->regs[j].VIwrite & allflags; - if( pinst->info.macflag & VUOP_WRITE ) w.viwrite[1] |= (1<info.statusflag & VUOP_WRITE ) w.viwrite[1] |= (1<info.macflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_MAC_FLAG); + if (pinst->info.statusflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_STATUS_FLAG); - if( (pinst->info.macflag|pinst->info.statusflag) & VUOP_READ ) + if ((pinst->info.macflag | pinst->info.statusflag) & VUOP_READ) macflags = 1; - if( pinst->regs[0].VIread & ((1<regs[0].VIread & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) macflags = 1; // if( pinst->regs[1].pipe == VUPIPE_FMAC && (pinst->regs[1].VFwrite==0&&!(pinst->regs[1].VIwrite&(1<regs[0].VIread |= (1<VIwrite |= lregs->VIwrite & (1<info.statusflag&VUOP_WRITE)&&!(pinst->regs[0].VIwrite&(1<info.statusflag&VUOP_WRITE) && !(pinst->regs[0].VIwrite&(1 << REG_STATUS_FLAG))) && needFullStatusFlag) + { // don't read if first inst - if( needFullStatusFlag == 1 ) - w.viread[1] |= (1<regs[j].VIread & allflags; - if( (pinst->regs[j].VIread&(1<regs[j].VIwrite&(1<regs[j].VIread&(1 << REG_STATUS_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_STATUS_FLAG))) + { // don't need the read anymore - pinst->regs[j].VIread &= ~(1<regs[j].VIread &= ~(1 << REG_STATUS_FLAG); } - if( (pinst->regs[j].VIread&(1<regs[j].VIwrite&(1<regs[j].VIread&(1 << REG_MAC_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_MAC_FLAG))) + { // don't need the read anymore - pinst->regs[j].VIread &= ~(1<regs[j].VIread &= ~(1 << REG_MAC_FLAG); } pinst->regs[j].VIwrite &= ~allflags; } - if( pinst->info.macflag & VUOP_READ) w.viread[1] |= 1<info.statusflag & VUOP_READ) w.viread[1] |= 1<info.macflag & VUOP_READ) w.viread[1] |= 1 << REG_MAC_FLAG; + if (pinst->info.statusflag & VUOP_READ) w.viread[1] |= 1 << REG_STATUS_FLAG; - w.nParentPc = pc-8; - w.cycle = pinst->info.cycle+4; + w.nParentPc = pc - 8; + w.cycle = pinst->info.cycle + 4; listWritebacks.push_back(w); } - if( pinst->info.q&VUOP_READ ) pinst->type |= INST_Q_READ; - if( pinst->info.p&VUOP_READ ) pinst->type |= INST_P_READ; + if (pinst->info.q&VUOP_READ) pinst->type |= INST_Q_READ; + if (pinst->info.p&VUOP_READ) pinst->type |= INST_P_READ; - if( pinst->info.q&VUOP_WRITE ) { - pinst->pqcycles = QWaitTimes[pinst->info.pqinst]+1; + if (pinst->info.q&VUOP_WRITE) + { + pinst->pqcycles = QWaitTimes[pinst->info.pqinst] + 1; memset(&w, 0, sizeof(w)); - w.nParentPc = pc-8; - w.cycle = pinst->info.cycle+pinst->pqcycles; - w.viwrite[0] = 1<info.cycle + pinst->pqcycles; + w.viwrite[0] = 1 << REG_Q; listWritebacks.push_back(w); } - if( pinst->info.p&VUOP_WRITE ) - pinst->pqcycles = PWaitTimes[pinst->info.pqinst]+1; + if (pinst->info.p&VUOP_WRITE) + pinst->pqcycles = PWaitTimes[pinst->info.pqinst] + 1; - if( prevbranch ) { + if (prevbranch) + { break; } // make sure there is always a branch - // sensible soccer overflows on vu0, so increase the limit... - if( (s_vu==1 && i >= 0x799) || (s_vu==0 && i >= 0x201) ) { + // sensible soccer overflows on vu0, so increase the limit... + if ((s_vu == 1 && i >= 0x799) || (s_vu == 0 && i >= 0x201)) + { Console::Error("VuRec base block doesn't terminate!"); - assert(0); + assert(0); break; } @@ -1284,7 +1407,7 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V pbh++; } - if( macflags ) + if (macflags) pblock->type |= BLOCKTYPE_MACFLAGS; pblock->endpc = pc; @@ -1293,14 +1416,17 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V pblock->cycles = vucycle; #ifdef SUPERVU_WRITEBACKS - if( !branch || (branch&8) ) + if (!branch || (branch&8)) #endif { // flush writebacks - if( listWritebacks.size() > 0 ) { + if (listWritebacks.size() > 0) + { listWritebacks.sort(WRITEBACK::SortWritebacks); - for(itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) { - if( itwriteback->viwrite[0] & (1<viwrite[0] & (1 << REG_Q)) + { // ignore all Q writebacks continue; } @@ -1313,12 +1439,13 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V } } - if( !branch ) + if (!branch) return pblock; - if( branch & 8 ) { + if (branch & 8) + { // what if also a jump? - pblock->type |= BLOCKTYPE_EOP|BLOCKTYPE_HASEOP; + pblock->type |= BLOCKTYPE_EOP | BLOCKTYPE_HASEOP; // add an instruction to flush p and q (if written) pblock->insts.push_back(SuperVUFlushInst()); @@ -1326,8 +1453,8 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V } // it is a (cond) branch or a jump - u32 vucode = *(u32*)(VU->Micro+lastpc-16); - int bpc = _recbranchAddr(vucode)-8; + u32 vucode = *(u32*)(VU->Micro + lastpc - 16); + int bpc = _recbranchAddr(vucode) - 8; VUPIPELINES newpipes; memcpy(newpipes.fmac, VU->fmac, sizeof(newpipes.fmac)); @@ -1335,38 +1462,44 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V memcpy(&newpipes.efu, &VU->efu, sizeof(newpipes.efu)); memcpy(newpipes.ialu, VU->ialu, sizeof(newpipes.ialu)); - for(i = 0; i < 8; ++i) newpipes.fmac[i].sCycle -= vucycle; + for (i = 0; i < 8; ++i) newpipes.fmac[i].sCycle -= vucycle; newpipes.fdiv.sCycle -= vucycle; newpipes.efu.sCycle -= vucycle; - for(i = 0; i < 8; ++i) newpipes.ialu[i].sCycle -= vucycle; + for (i = 0; i < 8; ++i) newpipes.ialu[i].sCycle -= vucycle; - if( listWritebacks.size() > 0 ) { - // flush all when jumping, send down the pipe when in branching - bool bFlushWritebacks = (vucode>>25)==0x24||(vucode>>25)==0x25;//||(vucode>>25)==0x20||(vucode>>25)==0x21; + if (listWritebacks.size() > 0) + { + // flush all when jumping, send down the pipe when in branching + bool bFlushWritebacks = (vucode >> 25) == 0x24 || (vucode >> 25) == 0x25;//||(vucode>>25)==0x20||(vucode>>25)==0x21; listWritebacks.sort(WRITEBACK::SortWritebacks); - for(itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) { - if( itwriteback->viwrite[0] & (1<viwrite[0] & (1 << REG_Q)) + { // ignore all Q writebacks continue; } - if( itwriteback->cycle < vucycle || bFlushWritebacks ) { + if (itwriteback->cycle < vucycle || bFlushWritebacks) + { pblock->insts.push_back(VuInstruction()); itwriteback->InitInst(&pblock->insts.back(), vucycle); } - else { + else + { newpipes.listWritebacks.push_back(*itwriteback); newpipes.listWritebacks.back().cycle -= vucycle; } } } - if( newpipes.listWritebacks.size() > 0 ) // other blocks might read the mac flags + if (newpipes.listWritebacks.size() > 0) // other blocks might read the mac flags pblock->type |= BLOCKTYPE_MACFLAGS; - u32 firstbranch = vucode>>25; - switch(firstbranch) { + u32 firstbranch = vucode >> 25; + switch (firstbranch) + { case 0x24: // jr pblock->type |= BLOCKTYPE_EOP; // jump out of procedure, since not returning, set EOP pblock->insts.push_back(SuperVUFlushInst()); @@ -1374,68 +1507,69 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V break; case 0x25: // jalr - { - // linking, so will return to procedure - pblock->insts.push_back(SuperVUFlushInst()); + { + // linking, so will return to procedure + pblock->insts.push_back(SuperVUFlushInst()); - VuBaseBlock* pjumpblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + VuBaseBlock* pjumpblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert( pblock != NULL ); + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); - pblock->blocks.push_back(pjumpblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } + pblock->blocks.push_back(pjumpblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } case 0x20: // B - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert( pblock != NULL ); + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); - pblock->blocks.push_back(pbranchblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } + pblock->blocks.push_back(pbranchblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } case 0x21: // BAL - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert( pblock != NULL ); - pblock->blocks.push_back(pbranchblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); + pblock->blocks.push_back(pbranchblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } case 0x28: // IBEQ case 0x2f: // IBGEZ case 0x2d: // IBGTZ case 0x2e: // IBLEZ case 0x2c: // IBLTZ case 0x29: // IBNE - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert( pblock != NULL ); - pblock->blocks.push_back(pbranchblock); - - // if has a second branch that is B or BAL, skip this - u32 secondbranch = (*(u32*)(VU->Micro+lastpc-8))>>25; - if( !hasSecondBranch || (secondbranch != 0x21 && secondbranch != 0x20) ) { - pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + // update pblock since could have changed pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); pblock->blocks.push_back(pbranchblock); - } - break; - } + // if has a second branch that is B or BAL, skip this + u32 secondbranch = (*(u32*)(VU->Micro + lastpc - 8)) >> 25; + if (!hasSecondBranch || (secondbranch != 0x21 && secondbranch != 0x20)) + { + pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + } + + break; + } default: assert(pblock->blocks.size() == 1); break; @@ -1448,71 +1582,75 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V ///// NOTE! This could still be a hack for KH2/GoW, but until we know how it properly works, this will do for now./// ///////////////////////////////////////////////////////////////////////////////////////////////////////////////////// - if( hasSecondBranch && firstbranch != 0xff ) { //check the previous jump was conditional and there is a second branch + if (hasSecondBranch && firstbranch != 0xff) //check the previous jump was conditional and there is a second branch + { #else - if( hasSecondBranch) { + if (hasSecondBranch) + { #endif - u32 vucode = *(u32*)(VU->Micro+lastpc-8); + u32 vucode = *(u32*)(VU->Micro + lastpc - 8); pc = lastpc; int bpc = _recbranchAddr(vucode); - switch(vucode>>25) { + switch (vucode >> 25) + { case 0x24: // jr Console::Error("svurec bad jr jump!"); assert(0); break; case 0x25: // jalr - { - Console::Error("svurec bad jalr jump!"); - assert(0); - break; - } + { + Console::Error("svurec bad jalr jump!"); + assert(0); + break; + } case 0x20: // B - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - break; - } + pblock->blocks.push_back(pbranchblock); + break; + } case 0x21: // BAL - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - // replace instead of pushing a new block - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - break; - } + // replace instead of pushing a new block + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + break; + } case 0x28: // IBEQ case 0x2f: // IBGEZ case 0x2d: // IBGTZ case 0x2e: // IBLEZ case 0x2c: // IBLTZ case 0x29: // IBNE - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); - // only add the block if the previous branch doesn't include the next instruction (ie, if a direct jump) - if( firstbranch == 0x24 || firstbranch == 0x25 || firstbranch == 0x20 || firstbranch == 0x21 ) { - pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + // only add the block if the previous branch doesn't include the next instruction (ie, if a direct jump) + if (firstbranch == 0x24 || firstbranch == 0x25 || firstbranch == 0x20 || firstbranch == 0x21) + { + pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - } + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + } - break; - } + break; + } - jNO_DEFAULT; + jNO_DEFAULT; } } @@ -1523,46 +1661,51 @@ static void SuperVUInitLiveness(VuBaseBlock* pblock) { list::iterator itinst, itnext; - assert( pblock->insts.size() > 0 ); + assert(pblock->insts.size() > 0); - for(itinst = pblock->insts.begin(); itinst != pblock->insts.end(); ++itinst) { + for (itinst = pblock->insts.begin(); itinst != pblock->insts.end(); ++itinst) + { - if( itinst->type & INST_DUMMY_ ) { + if (itinst->type & INST_DUMMY_) + { itinst->addvars[0] = itinst->addvars[1] = 0xffffffff; itinst->livevars[0] = itinst->livevars[1] = 0xffffffff; itinst->keepvars[0] = itinst->keepvars[1] = 0xffffffff; itinst->usedvars[0] = itinst->usedvars[1] = 0; } - else { + else + { itinst->addvars[0] = itinst->regs[0].VIread | itinst->regs[1].VIread; - itinst->addvars[1] = (itinst->regs[0].VFread0 ? (1 << itinst->regs[0].VFread0) : 0) | - (itinst->regs[0].VFread1 ? (1 << itinst->regs[0].VFread1) : 0) | - (itinst->regs[1].VFread0 ? (1 << itinst->regs[1].VFread0) : 0) | - (itinst->regs[1].VFread1 ? (1 << itinst->regs[1].VFread1) : 0); + itinst->addvars[1] = (itinst->regs[0].VFread0 ? (1 << itinst->regs[0].VFread0) : 0) | + (itinst->regs[0].VFread1 ? (1 << itinst->regs[0].VFread1) : 0) | + (itinst->regs[1].VFread0 ? (1 << itinst->regs[1].VFread0) : 0) | + (itinst->regs[1].VFread1 ? (1 << itinst->regs[1].VFread1) : 0); // vf0 is not handled by VFread - if( !itinst->regs[0].VFread0 && (itinst->regs[0].VIread & (1<addvars[1] |= 1; - if( !itinst->regs[1].VFread0 && (itinst->regs[1].VIread & (1<addvars[1] |= 1; - if( !itinst->regs[0].VFread1 && (itinst->regs[0].VIread & (1<regs[0].VFr1xyzw != 0xff ) itinst->addvars[1] |= 1; - if( !itinst->regs[1].VFread1 && (itinst->regs[1].VIread & (1<regs[1].VFr1xyzw != 0xff ) itinst->addvars[1] |= 1; - - + if (!itinst->regs[0].VFread0 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; + if (!itinst->regs[1].VFread0 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; + if (!itinst->regs[0].VFread1 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[0].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; + if (!itinst->regs[1].VFread1 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[1].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; + + u32 vfwrite = 0; - if( itinst->regs[0].VFwrite != 0 ) { - if( itinst->regs[0].VFwxyzw != 0xf ) itinst->addvars[1] |= 1<regs[0].VFwrite; - else vfwrite |= 1<regs[0].VFwrite; + if (itinst->regs[0].VFwrite != 0) + { + if (itinst->regs[0].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[0].VFwrite; + else vfwrite |= 1 << itinst->regs[0].VFwrite; } - if( itinst->regs[1].VFwrite != 0 ) { - if( itinst->regs[1].VFwxyzw != 0xf ) itinst->addvars[1] |= 1<regs[1].VFwrite; - else vfwrite |= 1<regs[1].VFwrite; + if (itinst->regs[1].VFwrite != 0) + { + if (itinst->regs[1].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[1].VFwrite; + else vfwrite |= 1 << itinst->regs[1].VFwrite; } - if( (itinst->regs[1].VIwrite & (1<regs[1].VFwxyzw != 0xf ) - itinst->addvars[1] |= 1<regs[1].VIwrite & (1 << REG_ACC_FLAG)) && itinst->regs[1].VFwxyzw != 0xf) + itinst->addvars[1] |= 1 << REG_ACC_FLAG; - u32 viwrite = (itinst->regs[0].VIwrite|itinst->regs[1].VIwrite); + u32 viwrite = (itinst->regs[0].VIwrite | itinst->regs[1].VIwrite); - itinst->usedvars[0] = itinst->addvars[0]|viwrite; - itinst->usedvars[1] = itinst->addvars[1]|vfwrite; + itinst->usedvars[0] = itinst->addvars[0] | viwrite; + itinst->usedvars[1] = itinst->addvars[1] | vfwrite; // itinst->addvars[0] &= ~viwrite; // itinst->addvars[1] &= ~vfwrite; @@ -1572,8 +1715,10 @@ static void SuperVUInitLiveness(VuBaseBlock* pblock) } itinst = --pblock->insts.end(); - while( itinst != pblock->insts.begin() ) { - itnext = itinst; --itnext; + while (itinst != pblock->insts.begin()) + { + itnext = itinst; + --itnext; itnext->usedvars[0] |= itinst->usedvars[0]; itnext->usedvars[1] |= itinst->usedvars[1]; @@ -1584,13 +1729,13 @@ static void SuperVUInitLiveness(VuBaseBlock* pblock) u32 COMPUTE_LIVE(u32 R, u32 K, u32 L) { - u32 live = R | ((L)&(K)); + u32 live = R | ((L) & (K)); // special process mac and status flags // only propagate liveness if doesn't write to the flag - if( !(L&(1<insts.end(); - - if( pb->blocks.size() > 0 ) { - livevars[0] = 0; livevars[1] = 0; - for( itchild = pb->blocks.begin(); itchild != pb->blocks.end(); ++itchild) { + + if (pb->blocks.size() > 0) + { + livevars[0] = 0; + livevars[1] = 0; + for (itchild = pb->blocks.begin(); itchild != pb->blocks.end(); ++itchild) + { VuInstruction& front = (*itchild)->insts.front(); livevars[0] |= front.livevars[0]; livevars[1] |= front.livevars[1]; @@ -1623,40 +1773,47 @@ static void SuperVULivenessAnalysis() newlive = COMPUTE_LIVE(itinst->addvars[0], itinst->keepvars[0], livevars[0]); - // should propagate status flags whose parent insts are not in this block + // should propagate status flags whose parent insts are not in this block // if( itinst->nParentPc >= 0 && (itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) // newlive |= livevars[0]&((1<livevars[0] != newlive ) { + if (itinst->livevars[0] != newlive) + { changed = TRUE; itinst->livevars[0] = newlive; } newlive = COMPUTE_LIVE(itinst->addvars[1], itinst->keepvars[1], livevars[1]); - if( itinst->livevars[1] != newlive ) { + if (itinst->livevars[1] != newlive) + { changed = TRUE; itinst->livevars[1] = newlive; } } - while( itinst != pb->insts.begin() ) { + while (itinst != pb->insts.begin()) + { - itnext = itinst; --itnext; + itnext = itinst; + --itnext; newlive = COMPUTE_LIVE(itnext->addvars[0], itnext->keepvars[0], itinst->livevars[0]); - // should propagate status flags whose parent insts are not in this block + // should propagate status flags whose parent insts are not in this block // if( itnext->nParentPc >= 0 && (itnext->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) && !(itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) // newlive |= itinst->livevars[0]&((1<livevars[0] != newlive ) { + if (itnext->livevars[0] != newlive) + { changed = TRUE; itnext->livevars[0] = newlive; itnext->livevars[1] = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); } - else { + else + { newlive = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); - if( itnext->livevars[1] != newlive ) { + if (itnext->livevars[1] != newlive) + { changed = TRUE; itnext->livevars[1] = newlive; } @@ -1691,7 +1848,8 @@ static void SuperVULivenessAnalysis() // } } - } while(changed); + } + while (changed); } static void SuperVUEliminateDeadCode() @@ -1702,7 +1860,8 @@ static void SuperVUEliminateDeadCode() list listParents; list::iterator itparent; - FORIT(itblock, s_listBlocks) { + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { #ifdef _DEBUG u32 startpc = (*itblock)->startpc; @@ -1711,88 +1870,104 @@ static void SuperVUEliminateDeadCode() itnext = (*itblock)->insts.begin(); itinst = itnext++; - while(itnext != (*itblock)->insts.end() ) { - if( itinst->type & (INST_CLIP_WRITE|INST_MAC_WRITE|INST_STATUS_WRITE) ) { - u32 live0 = itnext->livevars[0]; - if( itinst->nParentPc >= 0 && itnext->nParentPc >= 0 && itinst->nParentPc != itnext->nParentPc ) { // superman returns - // take the live vars from the next next inst - list::iterator itnextnext = itnext; ++itnextnext; - if( itnextnext != (*itblock)->insts.end() ) { - live0 = itnextnext->livevars[0]; - } - } - - itinst->regs[0].VIwrite &= live0; + while (itnext != (*itblock)->insts.end()) + { + if (itinst->type & (INST_CLIP_WRITE | INST_MAC_WRITE | INST_STATUS_WRITE)) + { + u32 live0 = itnext->livevars[0]; + if (itinst->nParentPc >= 0 && itnext->nParentPc >= 0 && itinst->nParentPc != itnext->nParentPc) // superman returns + { + // take the live vars from the next next inst + list::iterator itnextnext = itnext; + ++itnextnext; + if (itnextnext != (*itblock)->insts.end()) + { + live0 = itnextnext->livevars[0]; + } + } + + itinst->regs[0].VIwrite &= live0; itinst->regs[1].VIwrite &= live0; - - u32 viwrite = itinst->regs[0].VIwrite|itinst->regs[1].VIwrite; + + u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); int removetype = 0; - FORIT(itparent, listParents) { + for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) + { VuInstruction* parent = *itparent; - if( viwrite & (1<regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); } else removetype |= INST_CLIP_WRITE; - - if( parent->info.macflag && (itinst->type & INST_MAC_WRITE) ) { - if( !(viwrite&(1<info.macflag && (itinst->type & INST_MAC_WRITE)) + { + if (!(viwrite&(1 << REG_MAC_FLAG))) + { //parent->info.macflag = 0; - // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1<regs[1].VIwrite & (1<regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); #endif // if VUPIPE_FMAC and destination is vf00, probably need to keep the mac flag - if( parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0&&!(parent->regs[1].VIwrite&(1<regs[0].VIwrite |= ((1<regs[1].VIwrite |= ((1<regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) + { + parent->regs[0].VIwrite |= ((1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= ((1 << REG_MAC_FLAG)); + } + else + removetype |= INST_MAC_WRITE; } - else { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); } } else removetype |= INST_MAC_WRITE; - if( parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) { - if( !(viwrite&(1<info.statusflag && (itinst->type & INST_STATUS_WRITE)) + { + if (!(viwrite&(1 << REG_STATUS_FLAG))) + { //parent->info.statusflag = 0; - // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1<regs[1].VIwrite & (1<regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); #endif - if( parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0&&!(parent->regs[1].VIwrite&(1<regs[0].VIwrite |= ((1<regs[1].VIwrite |= ((1<regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) + { + parent->regs[0].VIwrite |= ((1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= ((1 << REG_STATUS_FLAG)); + } + else + removetype |= INST_STATUS_WRITE; } - else { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); } } else removetype |= INST_STATUS_WRITE; } itinst->type &= ~removetype; - if( itinst->type == 0 ) { + if (itinst->type == 0) + { itnext = (*itblock)->insts.erase(itinst); itinst = itnext++; continue; @@ -1805,63 +1980,75 @@ static void SuperVUEliminateDeadCode() ++itnext; } - if( itinst->type & INST_DUMMY ) { + if (itinst->type & INST_DUMMY) + { // last inst with the children u32 mask = 0; - for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); ++itchild) { + for (itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); ++itchild) + { mask |= (*itchild)->insts.front().livevars[0]; } itinst->regs[0].VIwrite &= mask; itinst->regs[1].VIwrite &= mask; - u32 viwrite = itinst->regs[0].VIwrite|itinst->regs[1].VIwrite; + u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; - if( itinst->nParentPc >= 0 ) { + if (itinst->nParentPc >= 0) + { (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); int removetype = 0; - FORIT(itparent, listParents) { + for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) + { VuInstruction* parent = *itparent; - if( viwrite & (1<regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); } else removetype |= INST_CLIP_WRITE; - - if( parent->info.macflag && (itinst->type & INST_MAC_WRITE) ) { - if( !(viwrite&(1<info.macflag && (itinst->type & INST_MAC_WRITE)) + { + if (!(viwrite&(1 << REG_MAC_FLAG))) + { //parent->info.macflag = 0; #ifndef SUPERVU_WRITEBACKS - assert( !(parent->regs[0].VIwrite & (1<regs[1].VIwrite & (1<regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); #endif removetype |= INST_MAC_WRITE; } - else { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); } } else removetype |= INST_MAC_WRITE; - if( parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) { - if( !(viwrite&(1<info.statusflag && (itinst->type & INST_STATUS_WRITE)) + { + if (!(viwrite&(1 << REG_STATUS_FLAG))) + { //parent->info.statusflag = 0; #ifndef SUPERVU_WRITEBACKS - assert( !(parent->regs[0].VIwrite & (1<regs[1].VIwrite & (1<regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); #endif removetype |= INST_STATUS_WRITE; } - else { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite&(1<regs[1].VIwrite |= (itinst->regs[1].VIwrite&(1<regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); } } else removetype |= INST_STATUS_WRITE; } itinst->type &= ~removetype; - if( itinst->type == 0 ) { + if (itinst->type == 0) + { (*itblock)->insts.erase(itinst); } } @@ -1877,219 +2064,250 @@ void VuBaseBlock::AssignVFRegs() list::iterator itinst, itnext, itinst2; // init the start regs - if( type & BLOCKTYPE_ANALYZED ) return; // nothing changed + if (type & BLOCKTYPE_ANALYZED) return; // nothing changed memcpy(xmmregs, startregs, sizeof(xmmregs)); - if( type & BLOCKTYPE_ANALYZED ) { + if (type & BLOCKTYPE_ANALYZED) + { // check if changed - for(i = 0; i < iREGCNT_XMM; ++i) { - if( xmmregs[i].inuse != startregs[i].inuse ) + for (i = 0; i < iREGCNT_XMM; ++i) + { + if (xmmregs[i].inuse != startregs[i].inuse) break; - if( xmmregs[i].inuse && (xmmregs[i].reg != startregs[i].reg || xmmregs[i].type != startregs[i].type) ) + if (xmmregs[i].inuse && (xmmregs[i].reg != startregs[i].reg || xmmregs[i].type != startregs[i].type)) break; } - if( i == iREGCNT_XMM ) return; // nothing changed + if (i == iREGCNT_XMM) return; // nothing changed } u8* oldX86 = x86Ptr; - FORIT(itinst, insts) { + for(itinst = insts.begin(); itinst != insts.end(); itinst++) + { - if( itinst->type & INST_DUMMY ) - continue; + if (itinst->type & INST_DUMMY) continue; // reserve, go from upper to lower int lastwrite = -1; - for(i = 1; i >= 0; --i) { - _VURegsNum* regs = itinst->regs+i; + for (i = 1; i >= 0; --i) + { + _VURegsNum* regs = itinst->regs + i; // redo the counters so that the proper regs are released - for(int j = 0; j < iREGCNT_XMM; ++j) { - if( xmmregs[j].inuse ) { - if( xmmregs[j].type == XMMTYPE_VFREG ) { + for (int j = 0; j < iREGCNT_XMM; ++j) + { + if (xmmregs[j].inuse) + { + if (xmmregs[j].type == XMMTYPE_VFREG) + { int count = 0; itinst2 = itinst; - if( i ) { - if( itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg ) { + if (i) + { + if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg) + { itinst2 = insts.end(); break; } - else { + else + { ++count; ++itinst2; } } - while(itinst2 != insts.end() ) { - if( itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg || - itinst2->regs[1].VFread0 == xmmregs[j].reg || itinst2->regs[1].VFread1 == xmmregs[j].reg || itinst2->regs[1].VFwrite == xmmregs[j].reg) + while (itinst2 != insts.end()) + { + if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg || + itinst2->regs[1].VFread0 == xmmregs[j].reg || itinst2->regs[1].VFread1 == xmmregs[j].reg || itinst2->regs[1].VFwrite == xmmregs[j].reg) break; - + ++count; ++itinst2; } - xmmregs[j].counter = 1000-count; + xmmregs[j].counter = 1000 - count; } - else { - assert( xmmregs[j].type == XMMTYPE_ACC ); + else + { + assert(xmmregs[j].type == XMMTYPE_ACC); int count = 0; itinst2 = itinst; - if( i ) ++itinst2; // acc isn't used in lower insts + if (i) ++itinst2; // acc isn't used in lower insts - while(itinst2 != insts.end() ) { - assert( !((itinst2->regs[0].VIread|itinst2->regs[0].VIwrite) & (1<regs[0].VIread | itinst2->regs[0].VIwrite) & (1 << REG_ACC_FLAG))); - if( (itinst2->regs[1].VIread|itinst2->regs[1].VIwrite) & (1<regs[1].VIread | itinst2->regs[1].VIwrite) & (1 << REG_ACC_FLAG)) break; ++count; ++itinst2; } - xmmregs[j].counter = 1000-count; + xmmregs[j].counter = 1000 - count; } } } - if( regs->VFread0 ) _addNeededVFtoXMMreg(regs->VFread0); - if( regs->VFread1 ) _addNeededVFtoXMMreg(regs->VFread1); - if( regs->VFwrite ) _addNeededVFtoXMMreg(regs->VFwrite); - if( regs->VIread & (1<VIread & (1<VFread0) _addNeededVFtoXMMreg(regs->VFread0); + if (regs->VFread1) _addNeededVFtoXMMreg(regs->VFread1); + if (regs->VFwrite) _addNeededVFtoXMMreg(regs->VFwrite); + if (regs->VIread & (1 << REG_ACC_FLAG)) _addNeededACCtoXMMreg(); + if (regs->VIread & (1 << REG_VF0_FLAG)) _addNeededVFtoXMMreg(0); // alloc itinst->vfread0[i] = itinst->vfread1[i] = itinst->vfwrite[i] = itinst->vfacc[i] = -1; itinst->vfflush[i] = -1; - if( regs->VFread0 ) itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); - else if( regs->VIread & (1<vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); - if( regs->VFread1 ) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); - else if( (regs->VIread & (1<VFr1xyzw != 0xff ) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); - if( regs->VIread & (1<vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + if (regs->VFread0) itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); + else if (regs->VIread & (1 << REG_VF0_FLAG)) itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + if (regs->VFread1) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); + else if ((regs->VIread & (1 << REG_VF0_FLAG)) && regs->VFr1xyzw != 0xff) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + if (regs->VIread & (1 << REG_ACC_FLAG)) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); int reusereg = -1; // 0 - VFwrite, 1 - VFAcc - if( regs->VFwrite ) { - assert( !(regs->VIwrite&(1<VFwrite) + { + assert(!(regs->VIwrite&(1 << REG_ACC_FLAG))); - if( regs->VFwxyzw == 0xf ) { + if (regs->VFwxyzw == 0xf) + { itinst->vfwrite[i] = _checkXMMreg(XMMTYPE_VFREG, regs->VFwrite, 0); - if( itinst->vfwrite[i] < 0 ) reusereg = 0; + if (itinst->vfwrite[i] < 0) reusereg = 0; } - else { + else + { itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); } } - else if( regs->VIwrite & (1<VIwrite & (1 << REG_ACC_FLAG)) + { - if( regs->VFwxyzw == 0xf ) { + if (regs->VFwxyzw == 0xf) + { itinst->vfacc[i] = _checkXMMreg(XMMTYPE_ACC, 0, 0); - if( itinst->vfacc[i] < 0 ) reusereg = 1; + if (itinst->vfacc[i] < 0) reusereg = 1; } - else { + else + { itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); } } - if( reusereg >= 0 ) { + if (reusereg >= 0) + { // reuse - itnext = itinst; itnext++; + itnext = itinst; + itnext++; u8 type = reusereg ? XMMTYPE_ACC : XMMTYPE_VFREG; u8 reg = reusereg ? 0 : regs->VFwrite; - if( itinst->vfacc[i] >= 0 && lastwrite != itinst->vfacc[i] && - (itnext == insts.end() || ((regs->VIread&(1<usedvars[0]&(1<livevars[0]&(1<vfacc[i] >= 0 && lastwrite != itinst->vfacc[i] && + (itnext == insts.end() || ((regs->VIread&(1 << REG_ACC_FLAG)) && (!(itnext->usedvars[0]&(1 << REG_ACC_FLAG)) || !(itnext->livevars[0]&(1 << REG_ACC_FLAG)))))) + { - assert( reusereg == 0 ); - if(itnext == insts.end() || (itnext->livevars[0]&(1<vfacc[i]); + assert(reusereg == 0); + if (itnext == insts.end() || (itnext->livevars[0]&(1 << REG_ACC_FLAG))) _freeXMMreg(itinst->vfacc[i]); xmmregs[itinst->vfacc[i]].inuse = 1; xmmregs[itinst->vfacc[i]].reg = reg; xmmregs[itinst->vfacc[i]].type = type; xmmregs[itinst->vfacc[i]].mode = 0; itinst->vfwrite[i] = itinst->vfacc[i]; } - else if( itinst->vfread0[i] >= 0 && lastwrite != itinst->vfread0[i] && - (itnext == insts.end() || (regs->VFread0 > 0 && (!(itnext->usedvars[1]&(1<VFread0)) || !(itnext->livevars[1]&(1<VFread0))))) ) { + else if (itinst->vfread0[i] >= 0 && lastwrite != itinst->vfread0[i] && + (itnext == insts.end() || (regs->VFread0 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread0)) || !(itnext->livevars[1]&(1 << regs->VFread0)))))) + { - if(itnext == insts.end() || (itnext->livevars[1]®s->VFread0)) _freeXMMreg(itinst->vfread0[i]); + if (itnext == insts.end() || (itnext->livevars[1]®s->VFread0)) _freeXMMreg(itinst->vfread0[i]); xmmregs[itinst->vfread0[i]].inuse = 1; xmmregs[itinst->vfread0[i]].reg = reg; xmmregs[itinst->vfread0[i]].type = type; xmmregs[itinst->vfread0[i]].mode = 0; - if( reusereg ) itinst->vfacc[i] = itinst->vfread0[i]; + if (reusereg) itinst->vfacc[i] = itinst->vfread0[i]; else itinst->vfwrite[i] = itinst->vfread0[i]; } - else if( itinst->vfread1[i] >= 0 && lastwrite != itinst->vfread1[i] && - (itnext == insts.end() || (regs->VFread1 > 0 && (!(itnext->usedvars[1]&(1<VFread1)) || !(itnext->livevars[1]&(1<VFread1))))) ) { + else if (itinst->vfread1[i] >= 0 && lastwrite != itinst->vfread1[i] && + (itnext == insts.end() || (regs->VFread1 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread1)) || !(itnext->livevars[1]&(1 << regs->VFread1)))))) + { - if(itnext == insts.end() || (itnext->livevars[1]®s->VFread1)) _freeXMMreg(itinst->vfread1[i]); + if (itnext == insts.end() || (itnext->livevars[1]®s->VFread1)) _freeXMMreg(itinst->vfread1[i]); xmmregs[itinst->vfread1[i]].inuse = 1; xmmregs[itinst->vfread1[i]].reg = reg; xmmregs[itinst->vfread1[i]].type = type; xmmregs[itinst->vfread1[i]].mode = 0; - if( reusereg ) itinst->vfacc[i] = itinst->vfread1[i]; + if (reusereg) itinst->vfacc[i] = itinst->vfread1[i]; else itinst->vfwrite[i] = itinst->vfread1[i]; } - else { - if( reusereg ) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + else + { + if (reusereg) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); else itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); } } - if( itinst->vfwrite[i] >= 0 ) lastwrite = itinst->vfwrite[i]; - else if( itinst->vfacc[i] >= 0 ) lastwrite = itinst->vfacc[i]; + if (itinst->vfwrite[i] >= 0) lastwrite = itinst->vfwrite[i]; + else if (itinst->vfacc[i] >= 0) lastwrite = itinst->vfacc[i]; // always alloc at least 1 temp reg - int free0 = (i||regs->VFwrite||regs->VFread0||regs->VFread1||(regs->VIwrite&(1<VIread&(1<VFwrite || regs->VFread0 || regs->VFread1 || (regs->VIwrite & (1 << REG_ACC_FLAG)) || (regs->VIread & (1 << REG_VF0_FLAG))) + ? _allocTempXMMreg(XMMT_FPS, -1) : -1; + int free1 = 0, free2 = 0; - if( i==0 && itinst->vfwrite[1] >= 0 && (itinst->vfread0[0]==itinst->vfwrite[1]||itinst->vfread1[0]==itinst->vfwrite[1]) ) { + if (i == 0 && itinst->vfwrite[1] >= 0 && (itinst->vfread0[0] == itinst->vfwrite[1] || itinst->vfread1[0] == itinst->vfwrite[1])) + { itinst->vfflush[i] = _allocTempXMMreg(XMMT_FPS, -1); } - if( i == 1 && (regs->VIwrite & (1<VIwrite & (1 << REG_CLIP_FLAG))) + { // CLIP inst, need two extra regs - if( free0 < 0 ) + if (free0 < 0) free0 = _allocTempXMMreg(XMMT_FPS, -1); free1 = _allocTempXMMreg(XMMT_FPS, -1); free2 = _allocTempXMMreg(XMMT_FPS, -1); _freeXMMreg(free1); _freeXMMreg(free2); } - else if( regs->VIwrite & (1<VIwrite & (1 << REG_P)) + { // EFU inst, need extra reg - free1 = _allocTempXMMreg(XMMT_FPS, -1); - if( free0 == -1 ) + free1 = _allocTempXMMreg(XMMT_FPS, -1); + if (free0 == -1) free0 = free1; _freeXMMreg(free1); } - if( itinst->vfflush[i] >= 0 ) _freeXMMreg(itinst->vfflush[i]); - if( free0 >= 0 ) _freeXMMreg(free0); + if (itinst->vfflush[i] >= 0) _freeXMMreg(itinst->vfflush[i]); + if (free0 >= 0) _freeXMMreg(free0); - itinst->vffree[i] = (free0&0xf)|(free1<<8)|(free2<<16); - if( free0 == -1 ) itinst->vffree[i] |= VFFREE_INVALID0; + itinst->vffree[i] = (free0 & 0xf) | (free1 << 8) | (free2 << 16); + if (free0 == -1) itinst->vffree[i] |= VFFREE_INVALID0; _clearNeededXMMregs(); } } - assert( x86Ptr == oldX86 ); - u32 analyzechildren = !(type&BLOCKTYPE_ANALYZED); + assert(x86Ptr == oldX86); + u32 analyzechildren = !(type & BLOCKTYPE_ANALYZED); type |= BLOCKTYPE_ANALYZED; //memset(endregs, 0, sizeof(endregs)); - if( analyzechildren ) { - FORIT(itchild, blocks) (*itchild)->AssignVFRegs(); + if (analyzechildren) + { + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { + (*itchild)->AssignVFRegs(); + } } } @@ -2105,38 +2323,46 @@ void VuBaseBlock::AssignVIRegs(int parent) { const int maxregs = 6; - if( parent ) { - if( (type&BLOCKTYPE_ANALYZEDPARENT) ) + if (parent) + { + if ((type&BLOCKTYPE_ANALYZEDPARENT)) return; type |= BLOCKTYPE_ANALYZEDPARENT; s_markov.parents.push_back(this); - for(LISTBLOCKS::iterator it = blocks.begin(); it != blocks.end(); ++it) { + for (LISTBLOCKS::iterator it = blocks.begin(); it != blocks.end(); ++it) + { (*it)->AssignVIRegs(0); } return; } - if( (type&BLOCKTYPE_ANALYZED) ) + if ((type&BLOCKTYPE_ANALYZED)) return; // child - assert( allocX86Regs == -1 ); + assert(allocX86Regs == -1); allocX86Regs = s_vecRegArray.size(); - s_vecRegArray.resize(allocX86Regs+iREGCNT_GPR); + s_vecRegArray.resize(allocX86Regs + iREGCNT_GPR); _x86regs* pregs = &s_vecRegArray[allocX86Regs]; memset(pregs, 0, sizeof(_x86regs)*iREGCNT_GPR); - assert( parents.size() > 0 ); + assert(parents.size() > 0); list::iterator itparent; u32 usedvars = insts.front().usedvars[0]; u32 livevars = insts.front().livevars[0]; - if( parents.size() > 0 ) { + if (parents.size() > 0) + { u32 usedvars2 = 0xffffffff; - FORIT(itparent, parents) usedvars2 &= (*itparent)->insts.front().usedvars[0]; + + for(itparent = parents.begin(); itparent != parents.end(); itparent++) + { + usedvars2 &= (*itparent)->insts.front().usedvars[0]; + } + usedvars |= usedvars2; } @@ -2145,30 +2371,37 @@ void VuBaseBlock::AssignVIRegs(int parent) // currently order doesn't matter int num = 0; - if( usedvars ) { - for(int i = 1; i < 16; ++i) { - if( usedvars & (1<= maxregs ) break; + if (++num >= maxregs) break; } } } - if( num < maxregs) { + if (num < maxregs) + { livevars &= ~usedvars; livevars &= insts.back().usedvars[0]; - if( livevars ) { - for(int i = 1; i < 16; ++i) { - if( livevars & (1<= maxregs) break; + if (++num >= maxregs) break; } } } @@ -2176,7 +2409,9 @@ void VuBaseBlock::AssignVIRegs(int parent) s_markov.children.push_back(this); type |= BLOCKTYPE_ANALYZED; - FORIT(itparent, parents) { + + for(itparent = parents.begin(); itparent != parents.end(); itparent++) + { (*itparent)->AssignVIRegs(1); } } @@ -2185,24 +2420,33 @@ static void SuperVUAssignRegs() { list::iterator itblock, itblock2; - FORIT(itblock, s_listBlocks) (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + } s_listBlocks.front()->AssignVFRegs(); // VI assignments, find markov blanket for each node in the graph // then allocate regs based on the commonly used ones #ifdef SUPERVU_X86CACHING - FORIT(itblock, s_listBlocks) (*itblock)->type &= ~(BLOCKTYPE_ANALYZED|BLOCKTYPE_ANALYZEDPARENT); + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~(BLOCKTYPE_ANALYZED | BLOCKTYPE_ANALYZEDPARENT); + } s_vecRegArray.resize(0); u8 usedregs[16]; // note: first block always has to start with no alloc regs bool bfirst = true; - FORIT(itblock, s_listBlocks) { + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { - if( !((*itblock)->type & BLOCKTYPE_ANALYZED) ) { + if (!((*itblock)->type & BLOCKTYPE_ANALYZED)) + { - if( (*itblock)->parents.size() == 0 ) { + if ((*itblock)->parents.size() == 0) + { (*itblock)->type |= BLOCKTYPE_ANALYZED; bfirst = false; continue; @@ -2214,19 +2458,23 @@ static void SuperVUAssignRegs() // assign the regs int regid = s_vecRegArray.size(); - s_vecRegArray.resize(regid+iREGCNT_GPR); + s_vecRegArray.resize(regid + iREGCNT_GPR); _x86regs* mergedx86 = &s_vecRegArray[regid]; memset(mergedx86, 0, sizeof(_x86regs)*iREGCNT_GPR); - if( !bfirst ) { - *(u32*)usedregs = *((u32*)usedregs+1) = *((u32*)usedregs+2) = *((u32*)usedregs+3) = 0; + if (!bfirst) + { + *(u32*)usedregs = *((u32*)usedregs + 1) = *((u32*)usedregs + 2) = *((u32*)usedregs + 3) = 0; - FORIT(itblock2, s_markov.children) { - assert( (*itblock2)->allocX86Regs >= 0 ); + for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) + { + assert((*itblock2)->allocX86Regs >= 0); _x86regs* pregs = &s_vecRegArray[(*itblock2)->allocX86Regs]; - for(int i = 0; i < iREGCNT_GPR; ++i) { - if( pregs[i].inuse && pregs[i].reg < 16) { + for (int i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse && pregs[i].reg < 16) + { //assert( pregs[i].reg < 16); usedregs[pregs[i].reg]++; } @@ -2234,27 +2482,31 @@ static void SuperVUAssignRegs() } int num = 1; - for(int i = 0; i < 16; ++i) { - if( usedregs[i] == s_markov.children.size() ) { + for (int i = 0; i < 16; ++i) + { + if (usedregs[i] == s_markov.children.size()) + { // use mergedx86[num].inuse = 1; mergedx86[num].reg = i; - mergedx86[num].type = (s_vu?X86TYPE_VU1:0)|X86TYPE_VI; + mergedx86[num].type = (s_vu ? X86TYPE_VU1 : 0) | X86TYPE_VI; mergedx86[num].mode = MODE_READ; - if( ++num >= iREGCNT_GPR ) + if (++num >= iREGCNT_GPR) break; - if( num == ESP ) + if (num == ESP) ++num; } } - FORIT(itblock2, s_markov.children) { - assert( (*itblock2)->nStartx86 == -1 ); + for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) + { + assert((*itblock2)->nStartx86 == -1); (*itblock2)->nStartx86 = regid; } - FORIT(itblock2, s_markov.parents) { - assert( (*itblock2)->nEndx86 == -1 ); + for(itblock2 = s_markov.parents.begin(); itblock2 != s_markov.parents.end(); itblock2++) + { + assert((*itblock2)->nEndx86 == -1); (*itblock2)->nEndx86 = regid; } } @@ -2285,35 +2537,36 @@ static int s_needFlush; // first bit - Q, second bit - P, third bit - Q has been static int s_JumpX86; static int s_ScheduleXGKICK = 0, s_XGKICKReg = -1; -void recVUMI_XGKICK_( VURegs *VU ); +void recVUMI_XGKICK_(VURegs *VU); void SuperVUCleanupProgram(u32 startpc, int vuindex) { #ifdef SUPERVU_COUNT QueryPerformanceCounter(&svufinal); - svutime += (u32)(svufinal.QuadPart-svubase.QuadPart); + svutime += (u32)(svufinal.QuadPart - svubase.QuadPart); #endif - assert( s_vu1esp == 0 ); + assert(s_vu1esp == 0); VU = vuindex ? &VU1 : &VU0; VU->cycle += s_TotalVUCycles; - + //VU cycle stealing hack, 3000 cycle maximum so it doesn't get out of hand if (s_TotalVUCycles < 3000) cpuRegs.cycle += s_TotalVUCycles * Config.Hacks.VUCycleSteal; else cpuRegs.cycle += 3000 * Config.Hacks.VUCycleSteal; - if( (int)s_writeQ > 0 ) VU->VI[REG_Q] = VU->q; - if( (int)s_writeP > 0 ) { + if ((int)s_writeQ > 0) VU->VI[REG_Q] = VU->q; + if ((int)s_writeP > 0) + { assert(VU == &VU1); VU1.VI[REG_P] = VU1.p; // only VU1 } //memset(recVUStack, 0, SUPERVU_STACKSIZE * 4); - // Could clear allocation info to prevent possibly bad data being used in other parts of pcsx2; + // Could clear allocation info to prevent possibly bad data being used in other parts of pcsx2; // not doing this because it's slow and not needed (rama) // _initXMMregs(); // _initMMXregs(); @@ -2325,7 +2578,8 @@ void SuperVUCleanupProgram(u32 startpc, int vuindex) // entry point of all vu programs from emulator calls __declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) { - __asm { + __asm + { mov eax, dword ptr [esp] mov s_TotalVUCycles, 0 // necessary to be here! add esp, 4 @@ -2339,12 +2593,14 @@ __declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) mov s_vuebx, ebx } #ifdef _DEBUG - __asm { + __asm + { mov s_vu1esp, esp } #endif - __asm { + __asm + { //stmxcsr s_ssecsr ldmxcsr g_sseVUMXCSR @@ -2359,7 +2615,8 @@ __declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) // exit point of all vu programs __declspec(naked) static void SuperVUEndProgram() { - __asm { + __asm + { // restore cpu state ldmxcsr g_sseMXCSR @@ -2370,12 +2627,14 @@ __declspec(naked) static void SuperVUEndProgram() } #ifdef _DEBUG - __asm { + __asm + { sub s_vu1esp, esp } #endif - __asm { + __asm + { call SuperVUCleanupProgram jmp s_callstack // so returns correctly } @@ -2387,25 +2646,28 @@ __declspec(naked) static void SuperVUEndProgram() void SuperVUFlush(int p, int wait) { u8* pjmp[3]; - if( !(s_needFlush&(1<info.cycle < recwait ) return; + if (!wait && s_pCurInst->info.cycle < recwait) return; - if( recwait == 0 ) { + if (recwait == 0) + { // write didn't happen this block MOV32MtoR(EAX, p ? (uptr)&s_writeP : (uptr)&s_writeQ); OR32RtoR(EAX, EAX); pjmp[0] = JS8(0); - if( s_pCurInst->info.cycle ) SUB32ItoR(EAX, s_pCurInst->info.cycle); + if (s_pCurInst->info.cycle) SUB32ItoR(EAX, s_pCurInst->info.cycle); // if writeQ <= total+offset - if( !wait ) { // only write back if time is up + if (!wait) // only write back if time is up + { CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); pjmp[1] = JG8(0); } - else { + else + { // add (writeQ-total-offset) to s_TotalVUCycles // necessary? CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); @@ -2414,32 +2676,34 @@ void SuperVUFlush(int p, int wait) x86SetJ8(pjmp[2]); } } - else if( wait && s_pCurInst->info.cycle < recwait ) { + else if (wait && s_pCurInst->info.cycle < recwait) + { ADD32ItoM((uptr)&s_TotalVUCycles, recwait); } - MOV32MtoR(EAX, SuperVUGetVIAddr(p?REG_P:REG_Q, 0)); + MOV32MtoR(EAX, SuperVUGetVIAddr(p ? REG_P : REG_Q, 0)); MOV32ItoM(p ? (uptr)&s_writeP : (uptr)&s_writeQ, 0x80000000); - MOV32RtoM(SuperVUGetVIAddr(p?REG_P:REG_Q, 1), EAX); + MOV32RtoM(SuperVUGetVIAddr(p ? REG_P : REG_Q, 1), EAX); - if( recwait == 0 ) { - if( !wait ) x86SetJ8(pjmp[1]); + if (recwait == 0) + { + if (!wait) x86SetJ8(pjmp[1]); x86SetJ8(pjmp[0]); } - if( wait || (!p && recwait == 0 && s_pCurInst->info.cycle >= 12) || (!p && recwait > 0 && s_pCurInst->info.cycle >= recwait ) ) - s_needFlush &= ~(1<info.cycle >= 12) || (!p && recwait > 0 && s_pCurInst->info.cycle >= recwait)) + s_needFlush &= ~(1 << p); } // executed only once per program static u32* SuperVUStaticAlloc(u32 size) { - assert( recVUStackPtr+size <= recVUStack+SUPERVU_STACKSIZE ); + assert(recVUStackPtr + size <= recVUStack + SUPERVU_STACKSIZE); // always zero - if( size == 4 ) *(u32*)recVUStackPtr = 0; + if (size == 4) *(u32*)recVUStackPtr = 0; else memset(recVUStackPtr, 0, size); recVUStackPtr += size; - return (u32*)(recVUStackPtr-size); + return (u32*)(recVUStackPtr - size); } static void SuperVURecompile() @@ -2451,48 +2715,70 @@ static void SuperVURecompile() list::iterator itblock; - FORIT(itblock, s_listBlocks) (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + } + s_listBlocks.front()->Recompile(); + // make sure everything compiled - FORIT(itblock, s_listBlocks) assert( ((*itblock)->type & BLOCKTYPE_ANALYZED) && (*itblock)->pcode != NULL ); + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + } + { + assert(((*itblock)->type & BLOCKTYPE_ANALYZED) && (*itblock)->pcode != NULL); + } // link all blocks - FORIT(itblock, s_listBlocks) { + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { VuBaseBlock::LISTBLOCKS::iterator itchild; - assert( (*itblock)->blocks.size() <= ArraySize((*itblock)->pChildJumps) ); + assert((*itblock)->blocks.size() <= ArraySize((*itblock)->pChildJumps)); - int i = 0; - FORIT(itchild, (*itblock)->blocks) { + int i = 0; + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { - if( (u32)(uptr)(*itblock)->pChildJumps[i] == 0xffffffff ) + if ((u32)(uptr)(*itblock)->pChildJumps[i] == 0xffffffff) continue; - if( (*itblock)->pChildJumps[i] == NULL ) { + if ((*itblock)->pChildJumps[i] == NULL) + { VuBaseBlock* pchild = *itchild; - if( pchild->type & BLOCKTYPE_HASEOP) { - assert( pchild->blocks.size() == 0); - - AND32ItoM( (uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu?~0x100:~0x001 ); // E flag - AND32ItoM( (uptr)&VU->vifRegs->stat, ~0x4 ); + if (pchild->type & BLOCKTYPE_HASEOP) + { + assert(pchild->blocks.size() == 0); + + AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag + AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); MOV32ItoM((uptr)&VU->VI[REG_TPC], pchild->endpc); - JMP32( (uptr)SuperVUEndProgram - ( (uptr)x86Ptr + 5 )); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); } // only other case is when there are two branches - else assert( (*itblock)->insts.back().regs[0].pipe == VUPIPE_BRANCH ); + else + { + assert((*itblock)->insts.back().regs[0].pipe == VUPIPE_BRANCH); + } continue; } - if( (u32)(uptr)(*itblock)->pChildJumps[i] & 0x80000000 ) { + if ((u32)(uptr)(*itblock)->pChildJumps[i] & 0x80000000) + { // relative - assert( (uptr)(*itblock)->pChildJumps[i] <= 0xffffffff); - (*itblock)->pChildJumps[i] = (u32*)((uptr)(*itblock)->pChildJumps[i] & 0x7fffffff); + assert((uptr)(*itblock)->pChildJumps[i] <= 0xffffffff); + (*itblock)->pChildJumps[i] = (u32*)((uptr)(*itblock)->pChildJumps[i] & 0x7fffffff); *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode - ((uptr)(*itblock)->pChildJumps[i] + 4); } - else *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode; + else + { + *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode; + } ++i; } @@ -2506,13 +2792,14 @@ static void SuperVURecompile() u32 s_saveecx, s_saveedx, s_saveebx, s_saveesi, s_saveedi, s_saveebp; u32 g_curdebugvu; - + //float vuDouble(u32 f); #if defined(_MSC_VER) __declspec(naked) static void svudispfn() { - __asm { + __asm + { mov g_curdebugvu, eax mov s_saveecx, ecx mov s_saveedx, edx @@ -2529,21 +2816,24 @@ void svudispfntemp() #ifdef _DEBUG static u32 i; - - if( ((vudump&8) && g_curdebugvu) || ((vudump&0x80) && !g_curdebugvu) ) { //&& g_vu1lastrec != g_vu1last ) { - if( skipparent != g_vu1lastrec ) { - for(i = 0; i < ArraySize(badaddrs); ++i) { - if( s_svulast == badaddrs[i][1] && g_vu1lastrec == badaddrs[i][0] ) + if (((vudump&8) && g_curdebugvu) || ((vudump&0x80) && !g_curdebugvu)) //&& g_vu1lastrec != g_vu1last ) { + { + + if (skipparent != g_vu1lastrec) + { + for (i = 0; i < ArraySize(badaddrs); ++i) + { + if (s_svulast == badaddrs[i][1] && g_vu1lastrec == badaddrs[i][0]) break; } - - if( i == ArraySize(badaddrs) ) + + if (i == ArraySize(badaddrs)) { - //static int curesp; - //__asm mov curesp, esp + //static int curesp; + //__asm mov curesp, esp //Console::WriteLn("tVU: %x %x %x", s_svulast, s_vucount, s_vufnheader); - if( g_curdebugvu ) iDumpVU1Registers(); + if (g_curdebugvu) iDumpVU1Registers(); else iDumpVU0Registers(); s_vucount++; } @@ -2554,14 +2844,15 @@ void svudispfntemp() #endif #if defined(_MSC_VER) - __asm { - mov ecx, s_saveecx - mov edx, s_saveedx - mov ebx, s_saveebx - mov esi, s_saveesi - mov edi, s_saveedi - mov ebp, s_saveebp - ret + __asm + { + mov ecx, s_saveecx + mov edx, s_saveedx + mov ebx, s_saveebx + mov esi, s_saveesi + mov edi, s_saveedi + mov ebp, s_saveebp + ret } #endif } @@ -2569,35 +2860,43 @@ void svudispfntemp() // frees all regs taking into account the livevars void SuperVUFreeXMMregs(u32* livevars) { - for(int i = 0; i < iREGCNT_XMM; ++i) { - if( xmmregs[i].inuse ) { + for (int i = 0; i < iREGCNT_XMM; ++i) + { + if (xmmregs[i].inuse) + { // same reg - if( (xmmregs[i].mode & MODE_WRITE) ) { + if ((xmmregs[i].mode & MODE_WRITE)) + { #ifdef SUPERVU_INTERCACHING - if( xmmregs[i].type == XMMTYPE_VFREG ) { - if( !(livevars[1] & (1<VF[xmmregs[i].reg] : (uptr)&VU->ACC; + u32 addr = xmmregs[i].type == XMMTYPE_VFREG ? (uptr) & VU->VF[xmmregs[i].reg] : (uptr) & VU->ACC; - if( xmmregs[i].mode & MODE_VUZ ) { + if (xmmregs[i].mode & MODE_VUZ) + { SSE_MOVHPS_XMM_to_M64(addr, (x86SSERegType)i); SSE_SHUFPS_M128_to_XMM((x86SSERegType)i, addr, 0xc4); } - else SSE_MOVHPS_M64_to_XMM((x86SSERegType)i, addr+8); + else + { + SSE_MOVHPS_M64_to_XMM((x86SSERegType)i, addr + 8); + } xmmregs[i].mode &= ~MODE_VUXYZ; } - + _freeXMMreg(i); } } @@ -2608,27 +2907,28 @@ void SuperVUFreeXMMregs(u32* livevars) void SuperVUTestVU0Condition(u32 incstack) { - if( s_vu && !SUPERVU_CHECKCONDITION ) return; // vu0 only + if (s_vu && !SUPERVU_CHECKCONDITION) return; // vu0 only CMP32ItoM((uptr)&s_TotalVUCycles, 512); // sometimes games spin on vu0, so be careful with this value - // woody hangs if too high + // woody hangs if too high - if( incstack ) { + if (incstack) + { u8* ptr = JB8(0); ADD32ItoR(ESP, incstack); //CALLFunc((u32)timeout); - JMP32( (uptr)SuperVUEndProgram - ( (uptr)x86Ptr + 5 )); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); x86SetJ8(ptr); } - else JAE32( (uptr)SuperVUEndProgram - ( (uptr)x86Ptr + 6 ) ); + else JAE32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 6)); } void VuBaseBlock::Recompile() { - if( type & BLOCKTYPE_ANALYZED ) return; - + if (type & BLOCKTYPE_ANALYZED) return; + x86Align(16); pcode = x86Ptr; @@ -2636,16 +2936,19 @@ void VuBaseBlock::Recompile() MOV32ItoM((uptr)&s_vufnheader, s_pFnHeader->startpc); MOV32ItoM((uptr)&VU->VI[REG_TPC], startpc); MOV32ItoM((uptr)&s_svulast, startpc); + list::iterator itparent; - for(itparent = parents.begin(); itparent != parents.end(); ++itparent) { - if( (*itparent)->blocks.size()==1 && (*itparent)->blocks.front()->startpc == startpc && - ((*itparent)->insts.size() < 2 || (----(*itparent)->insts.end())->regs[0].pipe != VUPIPE_BRANCH) ) { + for (itparent = parents.begin(); itparent != parents.end(); ++itparent) + { + if ((*itparent)->blocks.size() == 1 && (*itparent)->blocks.front()->startpc == startpc && + ((*itparent)->insts.size() < 2 || (----(*itparent)->insts.end())->regs[0].pipe != VUPIPE_BRANCH)) + { MOV32ItoM((uptr)&skipparent, (*itparent)->startpc); break; } } - if( itparent == parents.end() ) MOV32ItoM((uptr)&skipparent, -1); + if (itparent == parents.end()) MOV32ItoM((uptr)&skipparent, -1); MOV32ItoR(EAX, s_vu); CALLFunc((uptr)svudispfn); @@ -2659,16 +2962,16 @@ void VuBaseBlock::Recompile() s_XGKICKReg = -1; s_ScheduleXGKICK = 0; - s_ClipRead = s_PrevClipWrite = (uptr)&VU->VI[REG_CLIP_FLAG]; - s_StatusRead = s_PrevStatusWrite = (uptr)&VU->VI[REG_STATUS_FLAG]; - s_MACRead = s_PrevMACWrite = (uptr)&VU->VI[REG_MAC_FLAG]; - s_PrevIWrite = (uptr)&VU->VI[REG_I]; + s_ClipRead = s_PrevClipWrite = (uptr) & VU->VI[REG_CLIP_FLAG]; + s_StatusRead = s_PrevStatusWrite = (uptr) & VU->VI[REG_STATUS_FLAG]; + s_MACRead = s_PrevMACWrite = (uptr) & VU->VI[REG_MAC_FLAG]; + s_PrevIWrite = (uptr) & VU->VI[REG_I]; s_JumpX86 = 0; - s_UnconditionalDelay = 0; + s_UnconditionalDelay = 0; memcpy(xmmregs, startregs, sizeof(xmmregs)); #ifdef SUPERVU_X86CACHING - if( nStartx86 >= 0 ) + if (nStartx86 >= 0) memcpy(x86regs, &s_vecRegArray[nStartx86], sizeof(x86regs)); else _initX86regs(); #else @@ -2676,41 +2979,50 @@ void VuBaseBlock::Recompile() #endif list::iterator itinst; - FORIT(itinst, insts) { + for(itinst = insts.begin(); itinst != insts.end(); itinst++) + { s_pCurInst = &(*itinst); - if( s_JumpX86 > 0 ) { - if( !x86regs[s_JumpX86].inuse ) { + if (s_JumpX86 > 0) + { + if (!x86regs[s_JumpX86].inuse) + { // load s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_READ); } - x86regs[s_JumpX86].needed = 1; + x86regs[s_JumpX86].needed = 1; } - if( s_ScheduleXGKICK && s_XGKICKReg > 0 ) { - assert( x86regs[s_XGKICKReg].inuse ); + if (s_ScheduleXGKICK && s_XGKICKReg > 0) + { + assert(x86regs[s_XGKICKReg].inuse); x86regs[s_XGKICKReg].needed = 1; } itinst->Recompile(itinst, vuxyz); - if( s_ScheduleXGKICK > 0 ) { - if( s_ScheduleXGKICK-- == 1 ) { + if (s_ScheduleXGKICK > 0) + { + if (s_ScheduleXGKICK-- == 1) + { recVUMI_XGKICK_(VU); } } } - assert( pc == endpc ); - assert( s_ScheduleXGKICK == 0 ); + assert(pc == endpc); + assert(s_ScheduleXGKICK == 0); // flush flags - if( s_PrevClipWrite != (uptr)&VU->VI[REG_CLIP_FLAG] ) { + if (s_PrevClipWrite != (uptr)&VU->VI[REG_CLIP_FLAG]) + { MOV32MtoR(EAX, s_PrevClipWrite); MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); } - if( s_PrevStatusWrite != (uptr)&VU->VI[REG_STATUS_FLAG] ) { + if (s_PrevStatusWrite != (uptr)&VU->VI[REG_STATUS_FLAG]) + { MOV32MtoR(EAX, s_PrevStatusWrite); MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); } - if( s_PrevMACWrite != (uptr)&VU->VI[REG_MAC_FLAG] ) { + if (s_PrevMACWrite != (uptr)&VU->VI[REG_MAC_FLAG]) + { MOV32MtoR(EAX, s_PrevMACWrite); MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); } @@ -2724,43 +3036,51 @@ void VuBaseBlock::Recompile() // MOVZX32M8toR(EAX, s_MACRead); // MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); // } - if( s_PrevIWrite != (uptr)&VU->VI[REG_I] ) { + if (s_PrevIWrite != (uptr)&VU->VI[REG_I]) + { MOV32ItoM((uptr)&VU->VI[REG_I], *(u32*)s_PrevIWrite); // never changes } ADD32ItoM((uptr)&s_TotalVUCycles, cycles); // compute branches, jumps, eop - if( type & BLOCKTYPE_HASEOP ) { + if (type & BLOCKTYPE_HASEOP) + { // end _freeXMMregs(); _freeX86regs(); - AND32ItoM( (uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu?~0x100:~0x001 ); // E flag - AND32ItoM( (uptr)&VU->vifRegs->stat, ~0x4 ); - if( !branch ) MOV32ItoM((uptr)&VU->VI[REG_TPC], endpc); - JMP32( (uptr)SuperVUEndProgram - ( (uptr)x86Ptr + 5 )); + AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag + AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); + if (!branch) MOV32ItoM((uptr)&VU->VI[REG_TPC], endpc); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); } - else { + else + { u32 livevars[2] = {0}; - list::iterator lastinst = GetInstIterAtPc(endpc-8); + list::iterator lastinst = GetInstIterAtPc(endpc - 8); lastinst++; - if( lastinst != insts.end() ) { + if (lastinst != insts.end()) + { livevars[0] = lastinst->livevars[0]; livevars[1] = lastinst->livevars[1]; } - else { + else + { // take from children - if( blocks.size() > 0 ) { + if (blocks.size() > 0) + { LISTBLOCKS::iterator itchild; - FORIT(itchild, blocks) { + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { livevars[0] |= (*itchild)->insts.front().livevars[0]; livevars[1] |= (*itchild)->insts.front().livevars[1]; } } - else { + else + { livevars[0] = ~0; livevars[1] = ~0; } @@ -2771,37 +3091,46 @@ void VuBaseBlock::Recompile() // get rid of any writes, otherwise _freeX86regs will write x86regs[s_JumpX86].mode &= ~MODE_WRITE; - if( branch == 1 ) { - if( !x86regs[s_JumpX86].inuse ) { - assert( x86regs[s_JumpX86].type == X86TYPE_VUJUMP ); + if (branch == 1) + { + if (!x86regs[s_JumpX86].inuse) + { + assert(x86regs[s_JumpX86].type == X86TYPE_VUJUMP); s_JumpX86 = 0xffffffff; // notify to jump from g_recWriteback } } // align VI regs #ifdef SUPERVU_X86CACHING - if( nEndx86 >= 0 ) { + if (nEndx86 >= 0) + { _x86regs* endx86 = &s_vecRegArray[nEndx86]; - for(int i = 0; i < iREGCNT_GPR; ++i) { - if( endx86[i].inuse ) { + for (int i = 0; i < iREGCNT_GPR; ++i) + { + if (endx86[i].inuse) + { - if( s_JumpX86 == i && x86regs[s_JumpX86].inuse ) { + if (s_JumpX86 == i && x86regs[s_JumpX86].inuse) + { x86regs[s_JumpX86].inuse = 0; x86regs[EAX].inuse = 1; MOV32RtoR(EAX, s_JumpX86); s_JumpX86 = EAX; } - if( x86regs[i].inuse ) { - if( x86regs[i].type == endx86[i].type && x86regs[i].reg == endx86[i].reg ) { + if (x86regs[i].inuse) + { + if (x86regs[i].type == endx86[i].type && x86regs[i].reg == endx86[i].reg) + { _freeX86reg(i); // will continue to use it continue; } - if( x86regs[i].type == (X86TYPE_VI|(s_vu?X86TYPE_VU1:0)) ) { + if (x86regs[i].type == (X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0))) + { #ifdef SUPERVU_INTERCACHING - if( livevars[0] & (1<startpc); + // store the last block executed + MOV32ItoM((uptr)&g_nLastBlockExecuted, s_pCurBlock->startpc); - switch(branch) { + switch (branch) + { case 1: // branch, esi has new prog SuperVUTestVU0Condition(0); - if( s_JumpX86 == 0xffffffff ) + if (s_JumpX86 == 0xffffffff) JMP32M((uptr)&g_recWriteback); else JMPR(s_JumpX86); @@ -2846,31 +3177,31 @@ void VuBaseBlock::Recompile() // fall through case 0x10: // jump, esi has new vupc - { - _freeXMMregs(); - _freeX86regs(); - - SuperVUTestVU0Condition(8); + { + _freeXMMregs(); + _freeX86regs(); - // already onto stack - CALLFunc((uptr)SuperVUGetProgram); - ADD32ItoR(ESP, 8); - JMPR(EAX); - break; - } - - case 0x13: // jr with uncon branch, uncond branch takes precendence (dropship) - { + SuperVUTestVU0Condition(8); + + // already onto stack + CALLFunc((uptr)SuperVUGetProgram); + ADD32ItoR(ESP, 8); + JMPR(EAX); + break; + } + + case 0x13: // jr with uncon branch, uncond branch takes precendence (dropship) + { // s32 delta = (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) << 3; // ADD32ItoRmOffset(ESP, delta, 0); - ADD32ItoR(ESP, 8); // restore - pChildJumps[0] = (u32*)((uptr)JMP32(0)|0x80000000); - break; - } + ADD32ItoR(ESP, 8); // restore + pChildJumps[0] = (u32*)((uptr)JMP32(0) | 0x80000000); + break; + } case 0: - case 3: // unconditional branch - pChildJumps[s_UnconditionalDelay] = (u32*)((uptr)JMP32(0)|0x80000000); + case 3: // unconditional branch + pChildJumps[s_UnconditionalDelay] = (u32*)((uptr)JMP32(0) | 0x80000000); break; default: @@ -2880,11 +3211,12 @@ void VuBaseBlock::Recompile() } } - pendcode = x86Ptr; + pendcode = x86Ptr; type |= BLOCKTYPE_ANALYZED; LISTBLOCKS::iterator itchild; - FORIT(itchild, blocks) { + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { (*itchild)->Recompile(); } } @@ -2893,80 +3225,92 @@ void VuBaseBlock::Recompile() int VuInstruction::SetCachedRegs(int upper, u32 vuxyz) { - if( vfread0[upper] >= 0 ) { + if (vfread0[upper] >= 0) + { SuperVUFreeXMMreg(vfread0[upper], XMMTYPE_VFREG, regs[upper].VFread0); - _allocVFtoXMMreg(VU, vfread0[upper], regs[upper].VFread0, MODE_READ|GET_VUXYZMODE(regs[upper].VFread0)); + _allocVFtoXMMreg(VU, vfread0[upper], regs[upper].VFread0, MODE_READ | GET_VUXYZMODE(regs[upper].VFread0)); } - if( vfread1[upper] >= 0 ) { + if (vfread1[upper] >= 0) + { SuperVUFreeXMMreg(vfread1[upper], XMMTYPE_VFREG, regs[upper].VFread1); - _allocVFtoXMMreg(VU, vfread1[upper], regs[upper].VFread1, MODE_READ|GET_VUXYZMODE(regs[upper].VFread1)); + _allocVFtoXMMreg(VU, vfread1[upper], regs[upper].VFread1, MODE_READ | GET_VUXYZMODE(regs[upper].VFread1)); } - if( vfacc[upper] >= 0 && (regs[upper].VIread&(1<= 0 && (regs[upper].VIread&(1 << REG_ACC_FLAG))) + { SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); _allocACCtoXMMreg(VU, vfacc[upper], MODE_READ); } - if( vfwrite[upper] >= 0 ) { - assert( regs[upper].VFwrite > 0); + if (vfwrite[upper] >= 0) + { + assert(regs[upper].VFwrite > 0); SuperVUFreeXMMreg(vfwrite[upper], XMMTYPE_VFREG, regs[upper].VFwrite); _allocVFtoXMMreg(VU, vfwrite[upper], regs[upper].VFwrite, - MODE_WRITE|(regs[upper].VFwxyzw != 0xf?MODE_READ:0)|GET_VUXYZMODE(regs[upper].VFwrite)); + MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0) | GET_VUXYZMODE(regs[upper].VFwrite)); } - if( vfacc[upper] >= 0 && (regs[upper].VIwrite&(1<= 0 && (regs[upper].VIwrite&(1 << REG_ACC_FLAG))) + { SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); - _allocACCtoXMMreg(VU, vfacc[upper], MODE_WRITE|(regs[upper].VFwxyzw != 0xf?MODE_READ:0)); + _allocACCtoXMMreg(VU, vfacc[upper], MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0)); } int info = PROCESS_VU_SUPER; - if( vfread0[upper] >= 0 ) info |= PROCESS_EE_SET_S(vfread0[upper]); - if( vfread1[upper] >= 0 ) info |= PROCESS_EE_SET_T(vfread1[upper]); - if( vfacc[upper] >= 0 ) info |= PROCESS_VU_SET_ACC(vfacc[upper]); - if( vfwrite[upper] >= 0 ) { - if( regs[upper].VFwrite == _Ft_ && vfread1[upper] < 0 ) { + if (vfread0[upper] >= 0) info |= PROCESS_EE_SET_S(vfread0[upper]); + if (vfread1[upper] >= 0) info |= PROCESS_EE_SET_T(vfread1[upper]); + if (vfacc[upper] >= 0) info |= PROCESS_VU_SET_ACC(vfacc[upper]); + if (vfwrite[upper] >= 0) + { + if (regs[upper].VFwrite == _Ft_ && vfread1[upper] < 0) + { info |= PROCESS_EE_SET_T(vfwrite[upper]); } - else { - assert( regs[upper].VFwrite == _Fd_ ); + else + { + assert(regs[upper].VFwrite == _Fd_); info |= PROCESS_EE_SET_D(vfwrite[upper]); } } - if( !(vffree[upper]&VFFREE_INVALID0) ) { + if (!(vffree[upper]&VFFREE_INVALID0)) + { SuperVUFreeXMMreg(vffree[upper]&0xf, XMMTYPE_TEMP, 0); _allocTempXMMreg(XMMT_FPS, vffree[upper]&0xf); } - info |= PROCESS_VU_SET_TEMP(vffree[upper]&0xf); + info |= PROCESS_VU_SET_TEMP(vffree[upper] & 0xf); - if( vfflush[upper] >= 0 ) { + if (vfflush[upper] >= 0) + { SuperVUFreeXMMreg(vfflush[upper], XMMTYPE_TEMP, 0); _allocTempXMMreg(XMMT_FPS, vfflush[upper]); } - if( upper && (regs[upper].VIwrite & (1 << REG_CLIP_FLAG)) ) { + if (upper && (regs[upper].VIwrite & (1 << REG_CLIP_FLAG))) + { // CLIP inst, need two extra temp registers, put it EEREC_D and EEREC_ACC - assert( vfwrite[upper] == -1 ); - SuperVUFreeXMMreg((vffree[upper]>>8)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper]>>8)&0xf); - info |= PROCESS_EE_SET_D((vffree[upper]>>8)&0xf); + assert(vfwrite[upper] == -1); + SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); + info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); - SuperVUFreeXMMreg((vffree[upper]>>16)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper]>>16)&0xf); - info |= PROCESS_EE_SET_ACC((vffree[upper]>>16)&0xf); + SuperVUFreeXMMreg((vffree[upper] >> 16)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 16)&0xf); + info |= PROCESS_EE_SET_ACC((vffree[upper] >> 16) & 0xf); - _freeXMMreg((vffree[upper]>>8)&0xf); // don't need anymore - _freeXMMreg((vffree[upper]>>16)&0xf); // don't need anymore + _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore + _freeXMMreg((vffree[upper] >> 16)&0xf); // don't need anymore } - else if( regs[upper].VIwrite & (1<>8)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper]>>8)&0xf); - info |= PROCESS_EE_SET_D((vffree[upper]>>8)&0xf); - _freeXMMreg((vffree[upper]>>8)&0xf); // don't need anymore + else if (regs[upper].VIwrite & (1 << REG_P)) + { + SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); + info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); + _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore } - if( vfflush[upper] >= 0 ) _freeXMMreg(vfflush[upper]); - if( !(vffree[upper]&VFFREE_INVALID0) ) - _freeXMMreg(vffree[upper]&0xf); // don't need anymore + if (vfflush[upper] >= 0) _freeXMMreg(vfflush[upper]); + if (!(vffree[upper]&VFFREE_INVALID0)) + _freeXMMreg(vffree[upper]&0xf); // don't need anymore - if( (regs[0].VIwrite|regs[1].VIwrite) & ((1<::iterator& itinst, u32 vuxyz) //static PCSX2_ALIGNED16(VECTOR _VFc); u32 *ptr; u8* pjmp; - int vfregstore=0; + int vfregstore = 0; - assert( s_pCurInst == this); + assert(s_pCurInst == this); s_WriteToReadQ = 0; - ptr = (u32*)&VU->Micro[ pc ]; + ptr = (u32*) & VU->Micro[ pc ]; - if( type & INST_Q_READ ) - SuperVUFlush(0, (ptr[0] == 0x800003bf)||!!(regs[0].VIwrite & (1<::iterator& itinst, u32 vuxyz) // }; #ifdef SUPERVU_PROPAGATEFLAGS - if( nParentPc != -1 && (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) ) { + if (nParentPc != -1 && (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc)) + { // if( !s_vu ) { // for(int j = 0; j < ARRAYSIZE(badaddrs); ++j) { @@ -3013,32 +3359,38 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // } // } - list::iterator itblock; - FORIT(itblock, s_listBlocks) { - if( nParentPc >= (*itblock)->startpc && nParentPc < (*itblock)->endpc ) { - pparentinst = &(*(*itblock)->GetInstIterAtPc(nParentPc)); - //if( !s_vu ) SysPrintf("%x ", nParentPc); - if( find(s_pCurBlock->parents.begin(), s_pCurBlock->parents.end(), *itblock) != s_pCurBlock->parents.end() ) - nParentCheckForExecution = (*itblock)->startpc; - break; - } - } + list::iterator itblock; + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + if (nParentPc >= (*itblock)->startpc && nParentPc < (*itblock)->endpc) + { + pparentinst = &(*(*itblock)->GetInstIterAtPc(nParentPc)); + //if( !s_vu ) SysPrintf("%x ", nParentPc); + if (find(s_pCurBlock->parents.begin(), s_pCurBlock->parents.end(), *itblock) != s_pCurBlock->parents.end()) + nParentCheckForExecution = (*itblock)->startpc; + break; + } + } - assert( pparentinst != NULL ); - } + assert(pparentinst != NULL); + } #endif - if( type & INST_CLIP_WRITE ) { - if( nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc ) { + if (type & INST_CLIP_WRITE) + { + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { - if( !CHECK_VUCLIPFLAGHACK && pparentinst != NULL ) { + if (!CHECK_VUCLIPFLAGHACK && pparentinst != NULL) + { - if( nParentCheckForExecution >= 0 ) { - if( pparentinst->pClipWrite == 0 ) + if (nParentCheckForExecution >= 0) + { + if (pparentinst->pClipWrite == 0) pparentinst->pClipWrite = (uptr)SuperVUStaticAlloc(4); - if( s_ClipRead == 0 ) - s_ClipRead = (uptr)&VU->VI[REG_CLIP_FLAG]; + if (s_ClipRead == 0) + s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); u8* jptr = JNE8(0); @@ -3050,71 +3402,80 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) x86SetJ8(jptr2); } } - else s_ClipRead = (uptr)&VU->VI[REG_CLIP_FLAG]; + else s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; } - else { + else + { s_ClipRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pClipWrite; if (s_ClipRead == 0) Console::WriteLn("super ClipRead allocation error!"); } } // before modifying, check if they will ever be read - if( s_pCurBlock->type & BLOCKTYPE_MACFLAGS ) { + if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) + { - u8 outofblock=0; - if( type & INST_STATUS_WRITE ) { - - if( nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc ) { + u8 outofblock = 0; + if (type & INST_STATUS_WRITE) + { - // reading from out of this block, so already flushed to mem - if( pparentinst != NULL ) { //&& pparentinst->pStatusWrite != NULL ) { + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { - // might not have processed it yet, so reserve a mem loc - if( pparentinst->pStatusWrite == 0 ) { - pparentinst->pStatusWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pparentinst->pStatusWrite, 0); - } + // reading from out of this block, so already flushed to mem + if (pparentinst != NULL) //&& pparentinst->pStatusWrite != NULL ) { + { + + // might not have processed it yet, so reserve a mem loc + if (pparentinst->pStatusWrite == 0) + { + pparentinst->pStatusWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pparentinst->pStatusWrite, 0); + } // if( s_pCurBlock->prevFlagsOutOfBlock && s_StatusRead != NULL ) { -// // or instead since don't now which parent we came from +// // or instead since don't now which parent we came from // MOV32MtoR(EAX, pparentinst->pStatusWrite); // OR32RtoM(s_StatusRead, EAX); // MOV32ItoM(pparentinst->pStatusWrite, 0); // } - if( nParentCheckForExecution >= 0 ) { + if (nParentCheckForExecution >= 0) + { - // don't now which parent we came from, so have to check + // don't now which parent we came from, so have to check // uptr tempstatus = (uptr)SuperVUStaticAlloc(4); // if( s_StatusRead != NULL ) // MOV32MtoR(EAX, s_StatusRead); // else // MOV32MtoR(EAX, (uptr)&VU->VI[REG_STATUS_FLAG]); // s_StatusRead = tempstatus; - if( s_StatusRead == 0 ) - s_StatusRead = (uptr)&VU->VI[REG_STATUS_FLAG]; + if (s_StatusRead == 0) + s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; - CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); - u8* jptr = JNE8(0); - MOV32MtoR(EAX, pparentinst->pStatusWrite); - MOV32ItoM(pparentinst->pStatusWrite, 0); - MOV32RtoM(s_StatusRead, EAX); - x86SetJ8(jptr); - } - else { - uptr tempstatus = (uptr)SuperVUStaticAlloc(4); - MOV32MtoR(EAX, pparentinst->pStatusWrite); - MOV32RtoM(tempstatus, EAX); - MOV32ItoM(pparentinst->pStatusWrite, 0); - s_StatusRead = tempstatus; - } + CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); + u8* jptr = JNE8(0); + MOV32MtoR(EAX, pparentinst->pStatusWrite); + MOV32ItoM(pparentinst->pStatusWrite, 0); + MOV32RtoM(s_StatusRead, EAX); + x86SetJ8(jptr); + } + else + { + uptr tempstatus = (uptr)SuperVUStaticAlloc(4); + MOV32MtoR(EAX, pparentinst->pStatusWrite); + MOV32RtoM(tempstatus, EAX); + MOV32ItoM(pparentinst->pStatusWrite, 0); + s_StatusRead = tempstatus; + } - outofblock = 2; - } - else - s_StatusRead = (uptr)&VU->VI[REG_STATUS_FLAG]; - } - else { + outofblock = 2; + } + else + s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; + } + else + { s_StatusRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pStatusWrite; if (s_StatusRead == 0) Console::WriteLn("super StatusRead allocation error!"); // if( pc >= (u32)s_pCurBlock->endpc-8 ) { @@ -3127,20 +3488,24 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // } } } - if( type & INST_MAC_WRITE ) { - - if( nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc ) { + if (type & INST_MAC_WRITE) + { + + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { // reading from out of this block, so already flushed to mem - if( pparentinst != NULL ) {//&& pparentinst->pMACWrite != NULL ) { - // necessary for (katamari) - // towards the end, so variable might be leaked to another block (silent hill 4) - - // might not have processed it yet, so reserve a mem loc - if( pparentinst->pMACWrite == 0 ) { - pparentinst->pMACWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pparentinst->pMACWrite, 0); - } + if (pparentinst != NULL) //&& pparentinst->pMACWrite != NULL ) { + { + // necessary for (katamari) + // towards the end, so variable might be leaked to another block (silent hill 4) + + // might not have processed it yet, so reserve a mem loc + if (pparentinst->pMACWrite == 0) + { + pparentinst->pMACWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pparentinst->pMACWrite, 0); + } // if( s_pCurBlock->prevFlagsOutOfBlock && s_MACRead != NULL ) { // // or instead since don't now which parent we came from @@ -3148,9 +3513,10 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // OR32RtoM(s_MACRead, EAX); // MOV32ItoM(pparentinst->pMACWrite, 0); // } - if( nParentCheckForExecution >= 0 ) { + if (nParentCheckForExecution >= 0) + { - // don't now which parent we came from, so have to check + // don't now which parent we came from, so have to check // uptr tempmac = (uptr)SuperVUStaticAlloc(4); // if( s_MACRead != NULL ) // MOV32MtoR(EAX, s_MACRead); @@ -3158,28 +3524,29 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // MOV32MtoR(EAX, (uptr)&VU->VI[REG_MAC_FLAG]); // s_MACRead = tempmac; - if( s_MACRead == 0 ) - s_MACRead = (uptr)&VU->VI[REG_MAC_FLAG]; + if (s_MACRead == 0) + s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; - CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); - u8* jptr = JNE8(0); - MOV32MtoR(EAX, pparentinst->pMACWrite); - MOV32ItoM(pparentinst->pMACWrite, 0); - MOV32RtoM(s_MACRead, EAX); - x86SetJ8(jptr); - } - else { - uptr tempMAC = (uptr)SuperVUStaticAlloc(4); - MOV32MtoR(EAX, pparentinst->pMACWrite); - MOV32RtoM(tempMAC, EAX); - MOV32ItoM(pparentinst->pMACWrite, 0); - s_MACRead = tempMAC; - } + CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); + u8* jptr = JNE8(0); + MOV32MtoR(EAX, pparentinst->pMACWrite); + MOV32ItoM(pparentinst->pMACWrite, 0); + MOV32RtoM(s_MACRead, EAX); + x86SetJ8(jptr); + } + else + { + uptr tempMAC = (uptr)SuperVUStaticAlloc(4); + MOV32MtoR(EAX, pparentinst->pMACWrite); + MOV32RtoM(tempMAC, EAX); + MOV32ItoM(pparentinst->pMACWrite, 0); + s_MACRead = tempMAC; + } - outofblock = 2; - } - else - s_MACRead = (uptr)&VU->VI[REG_MAC_FLAG]; + outofblock = 2; + } + else + s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; // if( pc >= (u32)s_pCurBlock->endpc-8 ) { // // towards the end, so variable might be leaked to another block (silent hill 4) @@ -3189,139 +3556,161 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // MOV32ItoM(s_MACRead, 0); // s_MACRead = tempMAC; // } - } - else { + } + else + { s_MACRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pMACWrite; } } - s_pCurBlock->prevFlagsOutOfBlock = outofblock; + s_pCurBlock->prevFlagsOutOfBlock = outofblock; + } + else if (pparentinst != NULL) + { + // make sure to reset the mac and status flags! (katamari) + if (pparentinst->pStatusWrite) + MOV32ItoM(pparentinst->pStatusWrite, 0); + if (pparentinst->pMACWrite) + MOV32ItoM(pparentinst->pMACWrite, 0); } - else if( pparentinst != NULL) { - // make sure to reset the mac and status flags! (katamari) - if( pparentinst->pStatusWrite) - MOV32ItoM(pparentinst->pStatusWrite, 0); - if( pparentinst->pMACWrite) - MOV32ItoM(pparentinst->pMACWrite, 0); - } - assert( s_ClipRead != 0 ); - assert( s_MACRead != 0 ); - assert( s_StatusRead != 0 ); + assert(s_ClipRead != 0); + assert(s_MACRead != 0); + assert(s_StatusRead != 0); - return; + return; } - s_pCurBlock->prevFlagsOutOfBlock = 0; + s_pCurBlock->prevFlagsOutOfBlock = 0; #ifdef _DEBUG MOV32ItoR(EAX, pc); #endif - assert( !(type & (INST_CLIP_WRITE|INST_STATUS_WRITE|INST_MAC_WRITE)) ); + assert(!(type & (INST_CLIP_WRITE | INST_STATUS_WRITE | INST_MAC_WRITE))); pc += 8; - list::const_iterator itinst2; + list::const_iterator itinst2; - if( (regs[0].VIwrite|regs[1].VIwrite) & ((1<type & BLOCKTYPE_MACFLAGS ) { - if( pMACWrite == 0 ) { - pMACWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pMACWrite, 0); - } - if( pStatusWrite == 0 ) { + if ((regs[0].VIwrite | regs[1].VIwrite) & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) + { + if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) + { + if (pMACWrite == 0) + { + pMACWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pMACWrite, 0); + } + if (pStatusWrite == 0) + { pStatusWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pStatusWrite, 0); - } + //MOV32ItoM(pStatusWrite, 0); + } } - else { - assert( s_StatusRead == (uptr)&VU->VI[REG_STATUS_FLAG] ); - assert( s_MACRead == (uptr)&VU->VI[REG_MAC_FLAG] ); + else + { + assert(s_StatusRead == (uptr)&VU->VI[REG_STATUS_FLAG]); + assert(s_MACRead == (uptr)&VU->VI[REG_MAC_FLAG]); pMACWrite = s_MACRead; pStatusWrite = s_StatusRead; } } - if( (pClipWrite == 0) && ((regs[0].VIwrite|regs[1].VIwrite) & (1<insts.end() ) { - if( (itinst2->regs[0].VIread|itinst2->regs[0].VIwrite|itinst2->regs[1].VIread|itinst2->regs[1].VIwrite) && (1<insts.end()) + { + if ((itinst2->regs[0].VIread | itinst2->regs[0].VIwrite | itinst2->regs[1].VIread | itinst2->regs[1].VIwrite) && (1 << x86regs[j].reg)) break; - + ++count; ++itinst2; } - x86regs[j].counter = 1000-count; + x86regs[j].counter = 1000 - count; } } #endif - if (s_vu == 0 && (ptr[1] & 0x20000000)) { // M flag + if (s_vu == 0 && (ptr[1] & 0x20000000)) // M flag + { OR8ItoM((uptr)&VU->flags, VUFLAG_MFLAGSET); } - if (ptr[1] & 0x10000000) { // D flag - TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu?0x400:0x004); + if (ptr[1] & 0x10000000) // D flag + { + TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x400 : 0x004); u8* ptr = JZ8(0); - OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu?0x200:0x002); - _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu?INTC_VU1:INTC_VU0); + OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x200 : 0x002); + _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); x86SetJ8(ptr); } - if (ptr[1] & 0x08000000) { // T flag - TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu?0x800:0x008); + if (ptr[1] & 0x08000000) // T flag + { + TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x800 : 0x008); u8* ptr = JZ8(0); - OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu?0x400:0x004); - _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu?INTC_VU1:INTC_VU0); + OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x400 : 0x004); + _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); x86SetJ8(ptr); } - // check upper flags - if (ptr[1] & 0x80000000) { // I flag + // check upper flags + if (ptr[1] & 0x80000000) // I flag + { - assert( !(regs[0].VIwrite & ((1<code = ptr[1]; + VU->code = ptr[1]; s_vuInfo = SetCachedRegs(1, vuxyz); - if( s_JumpX86 > 0 ) x86regs[s_JumpX86].needed = 1; - if( s_ScheduleXGKICK && s_XGKICKReg > 0 ) x86regs[s_XGKICKReg].needed = 1; + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; + + recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); - recVU_UPPER_OPCODE[ VU->code & 0x3f ]( VU, s_vuInfo ); - s_PrevIWrite = (uptr)ptr; _clearNeededXMMregs(); _clearNeededX86regs(); } - else { - if( regs[0].VIwrite & (1<insts.end()); - u32* codeptr2 = ptr+2; + u32* codeptr2 = ptr + 2; - while(itinst2 != s_pCurBlock->insts.end() ) { - if( !(itinst2->type & INST_DUMMY) && ((itinst2->regs[0].VIwrite&(1<insts.end()) + { + if (!(itinst2->type & INST_DUMMY) && ((itinst2->regs[0].VIwrite&(1 << REG_Q)) || codeptr2[0] == 0x800003bf)) // waitq, or fdiv inst + { break; } - if( (itinst2->type & INST_Q_WRITE) && itinst2->nParentPc == pc-8 ) { + if ((itinst2->type & INST_Q_WRITE) && itinst2->nParentPc == pc - 8) + { break; } - if( itinst2->type & INST_Q_READ ) { + if (itinst2->type & INST_Q_READ) + { cacheq = 1; break; } - if( itinst2->type & INST_DUMMY ) { + if (itinst2->type & INST_DUMMY) + { ++itinst2; continue; } @@ -3329,29 +3718,32 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) ++itinst2; } - if( itinst2 == s_pCurBlock->insts.end() ) + if (itinst2 == s_pCurBlock->insts.end()) cacheq = 1; int x86temp = -1; - if( cacheq ) + if (cacheq) x86temp = _allocX86reg(-1, X86TYPE_TEMP, 0, 0); // new is written so flush old // if type & INST_Q_READ, already flushed - if( !(type & INST_Q_READ) && s_recWriteQ == 0 ) MOV32MtoR(EAX, (uptr)&s_writeQ); - - if( cacheq ) + if (!(type & INST_Q_READ) && s_recWriteQ == 0) MOV32MtoR(EAX, (uptr)&s_writeQ); + + if (cacheq) MOV32MtoR(x86temp, (uptr)&s_TotalVUCycles); - if( !(type & INST_Q_READ) ) { - if( s_recWriteQ == 0 ) { + if (!(type & INST_Q_READ)) + { + if (s_recWriteQ == 0) + { OR32RtoR(EAX, EAX); pjmp = JS8(0); MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); x86SetJ8(pjmp); } - else if( s_needFlush & 1 ) { + else if (s_needFlush & 1) + { MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); s_needFlush &= ~1; @@ -3359,42 +3751,48 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) } // write new Q - if( cacheq ) { - assert(s_pCurInst->pqcycles>1); - ADD32ItoR(x86temp, s_pCurInst->info.cycle+s_pCurInst->pqcycles); + if (cacheq) + { + assert(s_pCurInst->pqcycles > 1); + ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); MOV32RtoM((uptr)&s_writeQ, x86temp); s_needFlush |= 1; } - else { + else + { // won't be writing back s_WriteToReadQ = 1; s_needFlush &= ~1; MOV32ItoM((uptr)&s_writeQ, 0x80000001); } - s_recWriteQ = s_pCurInst->info.cycle+s_pCurInst->pqcycles; + s_recWriteQ = s_pCurInst->info.cycle + s_pCurInst->pqcycles; - if( x86temp >= 0 ) + if (x86temp >= 0) _freeX86reg(x86temp); } - if( regs[0].VIwrite & (1<::iterator& itinst, u32 vuxyz) } // write new P - assert(s_pCurInst->pqcycles>1); - ADD32ItoR(x86temp, s_pCurInst->info.cycle+s_pCurInst->pqcycles); + assert(s_pCurInst->pqcycles > 1); + ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); MOV32RtoM((uptr)&s_writeP, x86temp); s_needFlush |= 2; - s_recWriteP = s_pCurInst->info.cycle+s_pCurInst->pqcycles; + s_recWriteP = s_pCurInst->info.cycle + s_pCurInst->pqcycles; _freeX86reg(x86temp); } - if( ptr[0] == 0x800003bf ) // waitq + if (ptr[0] == 0x800003bf) // waitq SuperVUFlush(0, 1); - if( ptr[0] == 0x800007bf ) // waitp + if (ptr[0] == 0x800007bf) // waitp SuperVUFlush(1, 1); #ifdef PCSX2_DEVBUILD - if ( regs[1].VIread & regs[0].VIwrite & ~((1<startpc); } #endif u32 modewrite = 0; - if( vfwrite[1] >= 0 && xmmregs[vfwrite[1]].inuse && xmmregs[vfwrite[1]].type == XMMTYPE_VFREG && xmmregs[vfwrite[1]].reg == regs[1].VFwrite ) + if (vfwrite[1] >= 0 && xmmregs[vfwrite[1]].inuse && xmmregs[vfwrite[1]].type == XMMTYPE_VFREG && xmmregs[vfwrite[1]].reg == regs[1].VFwrite) modewrite = xmmregs[vfwrite[1]].mode & MODE_WRITE; - VU->code = ptr[1]; + VU->code = ptr[1]; s_vuInfo = SetCachedRegs(1, vuxyz); - if (vfwrite[1] >= 0) { - assert( regs[1].VFwrite > 0 ); + if (vfwrite[1] >= 0) + { + assert(regs[1].VFwrite > 0); - if (vfwrite[0] == vfwrite[1]) { + if (vfwrite[0] == vfwrite[1]) + { //Console::WriteLn("*PCSX2*: Warning, VF write to the same reg in both lower/upper cycle %x", params s_pCurBlock->startpc); } - if (vfread0[0] == vfwrite[1] || vfread1[0] == vfwrite[1] ) { - assert( regs[0].VFread0 == regs[1].VFwrite || regs[0].VFread1 == regs[1].VFwrite ); - assert( vfflush[0] >= 0 ); - if( modewrite ) { + if (vfread0[0] == vfwrite[1] || vfread1[0] == vfwrite[1]) + { + assert(regs[0].VFread0 == regs[1].VFwrite || regs[0].VFread1 == regs[1].VFwrite); + assert(vfflush[0] >= 0); + if (modewrite) + { SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[regs[1].VFwrite], (x86SSERegType)vfwrite[1]); } vfregstore = 1; } - } + } - if( s_JumpX86 > 0 ) x86regs[s_JumpX86].needed = 1; - if( s_ScheduleXGKICK && s_XGKICKReg > 0 ) x86regs[s_XGKICKReg].needed = 1; + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; - recVU_UPPER_OPCODE[ VU->code & 0x3f ]( VU, s_vuInfo ); + recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); _clearNeededXMMregs(); _clearNeededX86regs(); - + // necessary because status can be set by both upper and lower - if( regs[1].VIwrite & (1<code = ptr[0]; + VU->code = ptr[0]; s_vuInfo = SetCachedRegs(0, vuxyz); - if( vfregstore ) { + if (vfregstore) + { // load SSE_MOVAPS_M128_to_XMM(vfflush[0], (uptr)&VU->VF[regs[1].VFwrite]); - assert( xmmregs[vfwrite[1]].mode & MODE_WRITE ); + assert(xmmregs[vfwrite[1]].mode & MODE_WRITE); // replace with vfflush - if( _Fs_ == regs[1].VFwrite ) { + if (_Fs_ == regs[1].VFwrite) + { s_vuInfo &= ~PROCESS_EE_SET_S(0xf); s_vuInfo |= PROCESS_EE_SET_S(vfflush[0]); } - if( _Ft_ == regs[1].VFwrite ) { + if (_Ft_ == regs[1].VFwrite) + { s_vuInfo &= ~PROCESS_EE_SET_T(0xf); s_vuInfo |= PROCESS_EE_SET_T(vfflush[0]); } - xmmregs[vfflush[0]].mode |= MODE_NOFLUSH|MODE_WRITE; // so that lower inst doesn't flush + xmmregs[vfflush[0]].mode |= MODE_NOFLUSH | MODE_WRITE; // so that lower inst doesn't flush } - // notify vuinsts that upper inst is a fmac - if( regs[1].pipe == VUPIPE_FMAC ) - s_vuInfo |= PROCESS_VU_SET_FMAC(); + // notify vuinsts that upper inst is a fmac + if (regs[1].pipe == VUPIPE_FMAC) + s_vuInfo |= PROCESS_VU_SET_FMAC(); - if( s_JumpX86 > 0 ) x86regs[s_JumpX86].needed = 1; - if( s_ScheduleXGKICK && s_XGKICKReg > 0 ) x86regs[s_XGKICKReg].needed = 1; + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; #ifdef SUPERVU_VIBRANCHDELAY - if ( type & INST_CACHE_VI ) { - assert( vicached >= 0 ); - int cachedreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), vicached, MODE_READ); + if (type & INST_CACHE_VI) + { + assert(vicached >= 0); + int cachedreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), vicached, MODE_READ); MOV32RtoM((uptr)&s_VIBranchDelay, cachedreg); } #endif - // check if inst before branch and the write is the same as the read in the branch (wipeout) + // check if inst before branch and the write is the same as the read in the branch (wipeout) // int oldreg=0; // if( pc == s_pCurBlock->endpc-16 ) { // itinst2 = itinst; ++itinst2; // if( itinst2->regs[0].pipe == VUPIPE_BRANCH && (itinst->regs[0].VIwrite&itinst2->regs[0].VIread) ) { -// +// // CALLFunc((u32)branchfn); // assert( itinst->regs[0].VIwrite & 0xffff ); // Console::WriteLn("vi write before branch"); @@ -3511,7 +3919,7 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // if( itinst->regs[0].VIwrite & (1<::iterator& itinst, u32 vuxyz) // x86regs[oldreg].type = X86TYPE_VITEMP; // } - recVU_LOWER_OPCODE[ VU->code >> 25 ]( VU, s_vuInfo ); + recVU_LOWER_OPCODE[ VU->code >> 25 ](VU, s_vuInfo); // if( pc == s_pCurBlock->endpc-8 && s_CacheVIReg >= 0 ) { // // revert @@ -3541,18 +3949,21 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) } // clip is always written so ok - if( (regs[0].VIwrite|regs[1].VIwrite) & (1<::iterator& itinst, u32 vuxyz) void recVUMI_BranchHandle() { - int bpc = _recbranchAddr(VU->code); + int bpc = _recbranchAddr(VU->code); int curjump = 0; - if( s_pCurInst->type & INST_BRANCH_DELAY ) { - assert( (branch&0x17)!=0x10 && (branch&0x17)!=4 ); // no jump handlig for now + if (s_pCurInst->type & INST_BRANCH_DELAY) + { + assert((branch&0x17) != 0x10 && (branch&0x17) != 4); // no jump handlig for now - if( (branch & 0x7) == 3 ) { + if ((branch & 0x7) == 3) + { // previous was a direct jump curjump = 1; } - else if( branch & 1 ) curjump = 2; + else if (branch & 1) curjump = 2; } - assert( s_JumpX86 > 0 ); + assert(s_JumpX86 > 0); - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR - s_pCurBlock->pChildJumps[curjump] = (u32*)x86Ptr-1; + s_pCurBlock->pChildJumps[curjump] = (u32*)x86Ptr - 1; - if( !(s_pCurInst->type & INST_BRANCH_DELAY) ) { + if (!(s_pCurInst->type & INST_BRANCH_DELAY)) + { j8Ptr[1] = JMP8(0); - x86SetJ8( j8Ptr[ 0 ] ); + x86SetJ8(j8Ptr[ 0 ]); - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION ) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), pc+8); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), pc + 8); MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR - s_pCurBlock->pChildJumps[curjump+1] = (u32*)x86Ptr-1; + s_pCurBlock->pChildJumps[curjump+1] = (u32*)x86Ptr - 1; - x86SetJ8( j8Ptr[ 1 ] ); + x86SetJ8(j8Ptr[ 1 ]); } else - x86SetJ8( j8Ptr[ 0 ] ); + x86SetJ8(j8Ptr[ 0 ]); branch |= 1; } @@ -3605,302 +4019,337 @@ void recVUMI_IBQ_prep() { int isreg, itreg; - if( _Is_ == 0 ) { + if (_Is_ == 0) + { #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_ ) { - itreg = -1; - } - else + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) + { + itreg = -1; + } + else #endif - { - itreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_, MODE_READ); - } + { + itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); + } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( itreg >= 0 ) { + if (itreg >= 0) + { CMP16ItoR(itreg, 0); } else CMP16ItoM(SuperVUGetVIAddr(_It_, 1), 0); } - else if( _It_ == 0 ) { + else if (_It_ == 0) + { #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { - isreg = -1; - } - else + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else #endif - { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); - } + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( isreg >= 0 ) { + if (isreg >= 0) + { CMP16ItoR(isreg, 0); } - else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); + else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); } - else { - _addNeededX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_); + else + { + _addNeededX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { - isreg = -1; - } - else + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else #endif - { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); - } + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_ ) { - itreg = -1; + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) + { + itreg = -1; - if( isreg <= 0 ) { - // allocate fsreg - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { - isreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ|MODE_WRITE); - MOV32MtoR(isreg, SuperVUGetVIAddr(_Is_, 1)); - } - else - isreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); - } - } - else + if (isreg <= 0) + { + // allocate fsreg + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ | MODE_WRITE); + MOV32MtoR(isreg, SuperVUGetVIAddr(_Is_, 1)); + } + else + isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + } + else #endif - { - itreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _It_, MODE_READ); - } + { + itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); + } s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - if( isreg >= 0 ) { - if( itreg >= 0 ) { - CMP16RtoR( isreg, itreg ); + if (isreg >= 0) + { + if (itreg >= 0) + { + CMP16RtoR(isreg, itreg); } else CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); } - else if( itreg >= 0 ) { + else if (itreg >= 0) + { CMP16MtoR(itreg, SuperVUGetVIAddr(_Is_, 1)); } - else { - isreg = _allocX86reg(-1, X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + else + { + isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); } } } -void recVUMI_IBEQ( VURegs* vuu, s32 info ) +void recVUMI_IBEQ(VURegs* vuu, s32 info) { recVUMI_IBQ_prep(); - j8Ptr[ 0 ] = JNE8( 0 ); + j8Ptr[ 0 ] = JNE8(0); recVUMI_BranchHandle(); } -void recVUMI_IBGEZ( VURegs* vuu, s32 info ) +void recVUMI_IBGEZ(VURegs* vuu, s32 info) { int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { isreg = -1; } - else + else #endif { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); } - if( isreg >= 0 ) { + if (isreg >= 0) + { TEST16RtoR(isreg, isreg); - j8Ptr[ 0 ] = JS8( 0 ); + j8Ptr[ 0 ] = JS8(0); } - else { - CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); - j8Ptr[ 0 ] = JL8( 0 ); + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JL8(0); } recVUMI_BranchHandle(); } -void recVUMI_IBGTZ( VURegs* vuu, s32 info ) +void recVUMI_IBGTZ(VURegs* vuu, s32 info) { int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { isreg = -1; } - else + else #endif { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); } - if( isreg >= 0 ) { + if (isreg >= 0) + { CMP16ItoR(isreg, 0); - j8Ptr[ 0 ] = JLE8( 0 ); + j8Ptr[ 0 ] = JLE8(0); } - else { - CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); - j8Ptr[ 0 ] = JLE8( 0 ); + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JLE8(0); } recVUMI_BranchHandle(); } -void recVUMI_IBLEZ( VURegs* vuu, s32 info ) +void recVUMI_IBLEZ(VURegs* vuu, s32 info) { int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { isreg = -1; } - else + else #endif { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); } - if( isreg >= 0 ) { + if (isreg >= 0) + { CMP16ItoR(isreg, 0); - j8Ptr[ 0 ] = JG8( 0 ); + j8Ptr[ 0 ] = JG8(0); } - else { - CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); - j8Ptr[ 0 ] = JG8( 0 ); + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JG8(0); } recVUMI_BranchHandle(); } -void recVUMI_IBLTZ( VURegs* vuu, s32 info ) +void recVUMI_IBLTZ(VURegs* vuu, s32 info) { int isreg; s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); #ifdef SUPERVU_VIBRANCHDELAY - if( s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_ ) { + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { isreg = -1; } - else + else #endif { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); } - if( isreg >= 0 ) { + if (isreg >= 0) + { TEST16RtoR(isreg, isreg); - j8Ptr[ 0 ] = JNS8( 0 ); + j8Ptr[ 0 ] = JNS8(0); } - else { - CMP16ItoM( SuperVUGetVIAddr(_Is_, 1), 0x0 ); - j8Ptr[ 0 ] = JGE8( 0 ); + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JGE8(0); } recVUMI_BranchHandle(); } -void recVUMI_IBNE( VURegs* vuu, s32 info ) +void recVUMI_IBNE(VURegs* vuu, s32 info) { recVUMI_IBQ_prep(); - j8Ptr[ 0 ] = JE8( 0 ); + j8Ptr[ 0 ] = JE8(0); recVUMI_BranchHandle(); } -void recVUMI_B( VURegs* vuu, s32 info ) +void recVUMI_B(VURegs* vuu, s32 info) { // supervu will take care of the rest - int bpc = _recbranchAddr(VU->code); - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + int bpc = _recbranchAddr(VU->code); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); // loops to self, so check condition - if( bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION) ) { + if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) + { SuperVUTestVU0Condition(0); } - if( s_pCurBlock->blocks.size() > 1 ) { + if (s_pCurBlock->blocks.size() > 1) + { s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); MOV32ItoR(s_JumpX86, 1); - s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr-1; - s_UnconditionalDelay = 1; + s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; + s_UnconditionalDelay = 1; } branch |= 3; } -void recVUMI_BAL( VURegs* vuu, s32 info ) +void recVUMI_BAL(VURegs* vuu, s32 info) { - int bpc = _recbranchAddr(VU->code); - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION ) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + int bpc = _recbranchAddr(VU->code); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); // loops to self, so check condition - if( bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION) ) { + if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) + { SuperVUTestVU0Condition(0); } - if ( _It_ ) { - _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_, 2); - MOV16ItoM( SuperVUGetVIAddr(_It_, 0), (pc+8)>>3 ); + if (_It_) + { + _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); + MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); } - if( s_pCurBlock->blocks.size() > 1 ) { + if (s_pCurBlock->blocks.size() > 1) + { s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); MOV32ItoR(s_JumpX86, 1); - s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr-1; - s_UnconditionalDelay = 1; + s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; + s_UnconditionalDelay = 1; } branch |= 3; } -void recVUMI_JR( VURegs* vuu, s32 info ) +void recVUMI_JR(VURegs* vuu, s32 info) { - int isreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); LEA32RStoR(EAX, isreg, 3); //Mask the address to something valid - if(vuu == &VU0) + if (vuu == &VU0) AND32ItoR(EAX, 0xfff); else AND32ItoR(EAX, 0x3fff); - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 ) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); - - if( !(s_pCurBlock->type & BLOCKTYPE_HASEOP) ) { + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); + + if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) + { PUSH32I(s_vu); PUSH32R(EAX); } branch |= 0x10; // 0x08 is reserved } -void recVUMI_JALR( VURegs* vuu, s32 info ) +void recVUMI_JALR(VURegs* vuu, s32 info) { - _addNeededX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_); + _addNeededX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_); - int isreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); LEA32RStoR(EAX, isreg, 3); //Mask the address to something valid - if(vuu == &VU0) + if (vuu == &VU0) AND32ItoR(EAX, 0xfff); else AND32ItoR(EAX, 0x3fff); - if ( _It_ ) { - _deleteX86reg(X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _It_, 2); - MOV16ItoM( SuperVUGetVIAddr(_It_, 0), (pc+8)>>3 ); + if (_It_) + { + _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); + MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); } - if( (s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 ) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); - - if( !(s_pCurBlock->type & BLOCKTYPE_HASEOP) ) { + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); + + if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) + { PUSH32I(s_vu); PUSH32R(EAX); } @@ -3911,7 +4360,7 @@ void recVUMI_JALR( VURegs* vuu, s32 info ) #ifdef PCSX2_DEVBUILD void vu1xgkick(u32* pMem, u32 addr) { - assert( addr < 0x4000 ); + assert(addr < 0x4000); #ifdef SUPERVU_COUNT StopSVUCounter(); #endif @@ -3924,9 +4373,9 @@ void vu1xgkick(u32* pMem, u32 addr) } #endif -void recVUMI_XGKICK_( VURegs *VU ) +void recVUMI_XGKICK_(VURegs *VU) { - assert( s_XGKICKReg > 0 && x86regs[s_XGKICKReg].inuse && x86regs[s_XGKICKReg].type == X86TYPE_VITEMP); + assert(s_XGKICKReg > 0 && x86regs[s_XGKICKReg].inuse && x86regs[s_XGKICKReg].type == X86TYPE_VITEMP); x86regs[s_XGKICKReg].inuse = 0; // so free doesn't flush _freeX86regs(); @@ -3937,268 +4386,284 @@ void recVUMI_XGKICK_( VURegs *VU ) //CALLFunc((u32)countfn); - if( mtgsThread != NULL ) { + if (mtgsThread != NULL) + { CALLFunc((uptr)VU1XGKICK_MTGSTransfer); ADD32ItoR(ESP, 8); } - else { + else + { #ifdef PCSX2_DEVBUILD CALLFunc((uptr)vu1xgkick); ADD32ItoR(ESP, 8); #else - CALLFunc((uptr)GSgifTransfer1); + CALLFunc((uptr)GSgifTransfer1); #endif } s_ScheduleXGKICK = 0; } -void recVUMI_XGKICK( VURegs *VU, int info ) +void recVUMI_XGKICK(VURegs *VU, int info) { - if( s_ScheduleXGKICK ) { + if (s_ScheduleXGKICK) + { // second xgkick, so launch the first recVUMI_XGKICK_(VU); } - int isreg = _allocX86reg(X86ARG2, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), _Is_, MODE_READ); + int isreg = _allocX86reg(X86ARG2, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); _freeX86reg(isreg); // flush x86regs[isreg].inuse = 1; x86regs[isreg].type = X86TYPE_VITEMP; x86regs[isreg].needed = 1; - x86regs[isreg].mode = MODE_WRITE|MODE_READ; + x86regs[isreg].mode = MODE_WRITE | MODE_READ; SHL32ItoR(isreg, 4); AND32ItoR(isreg, 0x3fff); s_XGKICKReg = isreg; - if( !SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc ) { + if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) + { recVUMI_XGKICK_(VU); } - else { - if( g_VUGameFixes & VUFIX_XGKICKDELAY2 ) - s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc-pc)/8); - else - s_ScheduleXGKICK = 2; + else + { + if (g_VUGameFixes & VUFIX_XGKICKDELAY2) + s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc - pc) / 8); + else + s_ScheduleXGKICK = 2; } } -void recVU_UPPER_FD_00( VURegs* VU, s32 info ); -void recVU_UPPER_FD_01( VURegs* VU, s32 info ); -void recVU_UPPER_FD_10( VURegs* VU, s32 info ); -void recVU_UPPER_FD_11( VURegs* VU, s32 info ); -void recVULowerOP( VURegs* VU, s32 info ); -void recVULowerOP_T3_00( VURegs* VU, s32 info ); -void recVULowerOP_T3_01( VURegs* VU, s32 info ); -void recVULowerOP_T3_10( VURegs* VU, s32 info ); -void recVULowerOP_T3_11( VURegs* VU, s32 info ); -void recVUunknown( VURegs* VU, s32 info ); +void recVU_UPPER_FD_00(VURegs* VU, s32 info); +void recVU_UPPER_FD_01(VURegs* VU, s32 info); +void recVU_UPPER_FD_10(VURegs* VU, s32 info); +void recVU_UPPER_FD_11(VURegs* VU, s32 info); +void recVULowerOP(VURegs* VU, s32 info); +void recVULowerOP_T3_00(VURegs* VU, s32 info); +void recVULowerOP_T3_01(VURegs* VU, s32 info); +void recVULowerOP_T3_10(VURegs* VU, s32 info); +void recVULowerOP_T3_11(VURegs* VU, s32 info); +void recVUunknown(VURegs* VU, s32 info); -void (*recVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ) = { - recVUMI_LQ , recVUMI_SQ , recVUunknown , recVUunknown, - recVUMI_ILW , recVUMI_ISW , recVUunknown , recVUunknown, - recVUMI_IADDIU, recVUMI_ISUBIU, recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_FCEQ , recVUMI_FCSET , recVUMI_FCAND, recVUMI_FCOR, /* 0x10 */ - recVUMI_FSEQ , recVUMI_FSSET , recVUMI_FSAND, recVUMI_FSOR, - recVUMI_FMEQ , recVUunknown , recVUMI_FMAND, recVUMI_FMOR, - recVUMI_FCGET , recVUunknown , recVUunknown , recVUunknown, - recVUMI_B , recVUMI_BAL , recVUunknown , recVUunknown, /* 0x20 */ - recVUMI_JR , recVUMI_JALR , recVUunknown , recVUunknown, - recVUMI_IBEQ , recVUMI_IBNE , recVUunknown , recVUunknown, - recVUMI_IBLTZ , recVUMI_IBGTZ , recVUMI_IBLEZ, recVUMI_IBGEZ, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVULowerOP , recVUunknown , recVUunknown , recVUunknown, /* 0x40*/ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x50 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x60 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x70 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, -}; - -void (*recVULowerOP_T3_00_OPCODE[32])(VURegs* VU, s32 info) = { - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_MOVE , recVUMI_LQI , recVUMI_DIV , recVUMI_MTIR, - recVUMI_RNEXT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_MFP , recVUMI_XTOP , recVUMI_XGKICK, - recVUMI_ESADD , recVUMI_EATANxy, recVUMI_ESQRT, recVUMI_ESIN, -}; - -void (*recVULowerOP_T3_01_OPCODE[32])(VURegs* VU, s32 info) = { - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_MR32 , recVUMI_SQI , recVUMI_SQRT , recVUMI_MFIR, - recVUMI_RGET , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUMI_XITOP, recVUunknown, - recVUMI_ERSADD, recVUMI_EATANxz, recVUMI_ERSQRT, recVUMI_EATAN, -}; - -void (*recVULowerOP_T3_10_OPCODE[32])(VURegs* VU, s32 info) = { - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_LQD , recVUMI_RSQRT, recVUMI_ILWR, - recVUMI_RINIT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_ELENG , recVUMI_ESUM , recVUMI_ERCPR, recVUMI_EEXP, -}; - -void (*recVULowerOP_T3_11_OPCODE[32])(VURegs* VU, s32 info) = { - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_SQD , recVUMI_WAITQ, recVUMI_ISWR, - recVUMI_RXOR , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_ERLENG, recVUunknown , recVUMI_WAITP, recVUunknown, -}; - -void (*recVULowerOP_OPCODE[64])(VURegs* VU, s32 info) = { - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x20 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_IADD , recVUMI_ISUB , recVUMI_IADDI, recVUunknown, /* 0x30 */ - recVUMI_IAND , recVUMI_IOR , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVULowerOP_T3_00, recVULowerOP_T3_01, recVULowerOP_T3_10, recVULowerOP_T3_11, -}; - -void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info) = { - recVUMI_ADDx , recVUMI_ADDy , recVUMI_ADDz , recVUMI_ADDw, - recVUMI_SUBx , recVUMI_SUBy , recVUMI_SUBz , recVUMI_SUBw, - recVUMI_MADDx , recVUMI_MADDy , recVUMI_MADDz , recVUMI_MADDw, - recVUMI_MSUBx , recVUMI_MSUBy , recVUMI_MSUBz , recVUMI_MSUBw, - recVUMI_MAXx , recVUMI_MAXy , recVUMI_MAXz , recVUMI_MAXw, /* 0x10 */ - recVUMI_MINIx , recVUMI_MINIy , recVUMI_MINIz , recVUMI_MINIw, - recVUMI_MULx , recVUMI_MULy , recVUMI_MULz , recVUMI_MULw, - recVUMI_MULq , recVUMI_MAXi , recVUMI_MULi , recVUMI_MINIi, - recVUMI_ADDq , recVUMI_MADDq , recVUMI_ADDi , recVUMI_MADDi, /* 0x20 */ - recVUMI_SUBq , recVUMI_MSUBq , recVUMI_SUBi , recVUMI_MSUBi, - recVUMI_ADD , recVUMI_MADD , recVUMI_MUL , recVUMI_MAX, - recVUMI_SUB , recVUMI_MSUB , recVUMI_OPMSUB, recVUMI_MINI, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVU_UPPER_FD_00, recVU_UPPER_FD_01, recVU_UPPER_FD_10, recVU_UPPER_FD_11, -}; - -void (*recVU_UPPER_FD_00_TABLE[32])(VURegs* VU, s32 info) = { - recVUMI_ADDAx, recVUMI_SUBAx , recVUMI_MADDAx, recVUMI_MSUBAx, - recVUMI_ITOF0, recVUMI_FTOI0, recVUMI_MULAx , recVUMI_MULAq , - recVUMI_ADDAq, recVUMI_SUBAq, recVUMI_ADDA , recVUMI_SUBA , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_01_TABLE[32])(VURegs* VU, s32 info) = { - recVUMI_ADDAy , recVUMI_SUBAy , recVUMI_MADDAy, recVUMI_MSUBAy, - recVUMI_ITOF4 , recVUMI_FTOI4 , recVUMI_MULAy , recVUMI_ABS , - recVUMI_MADDAq, recVUMI_MSUBAq, recVUMI_MADDA , recVUMI_MSUBA , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_10_TABLE[32])(VURegs* VU, s32 info) = { - recVUMI_ADDAz , recVUMI_SUBAz , recVUMI_MADDAz, recVUMI_MSUBAz, - recVUMI_ITOF12, recVUMI_FTOI12, recVUMI_MULAz , recVUMI_MULAi , - recVUMI_ADDAi, recVUMI_SUBAi , recVUMI_MULA , recVUMI_OPMULA, - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_11_TABLE[32])(VURegs* VU, s32 info) = { - recVUMI_ADDAw , recVUMI_SUBAw , recVUMI_MADDAw, recVUMI_MSUBAw, - recVUMI_ITOF15, recVUMI_FTOI15, recVUMI_MULAw , recVUMI_CLIP , - recVUMI_MADDAi, recVUMI_MSUBAi, recVUunknown , recVUMI_NOP , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; +void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info) = +{ + recVUMI_LQ , recVUMI_SQ , recVUunknown , recVUunknown, + recVUMI_ILW , recVUMI_ISW , recVUunknown , recVUunknown, + recVUMI_IADDIU, recVUMI_ISUBIU, recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_FCEQ , recVUMI_FCSET , recVUMI_FCAND, recVUMI_FCOR, /* 0x10 */ + recVUMI_FSEQ , recVUMI_FSSET , recVUMI_FSAND, recVUMI_FSOR, + recVUMI_FMEQ , recVUunknown , recVUMI_FMAND, recVUMI_FMOR, + recVUMI_FCGET , recVUunknown , recVUunknown , recVUunknown, + recVUMI_B , recVUMI_BAL , recVUunknown , recVUunknown, /* 0x20 */ + recVUMI_JR , recVUMI_JALR , recVUunknown , recVUunknown, + recVUMI_IBEQ , recVUMI_IBNE , recVUunknown , recVUunknown, + recVUMI_IBLTZ , recVUMI_IBGTZ , recVUMI_IBLEZ, recVUMI_IBGEZ, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVULowerOP , recVUunknown , recVUunknown , recVUunknown, /* 0x40*/ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x50 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x60 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x70 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, +}; -void recVU_UPPER_FD_00( VURegs* VU, s32 info ) -{ - recVU_UPPER_FD_00_TABLE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVU_UPPER_FD_01( VURegs* VU, s32 info ) -{ - recVU_UPPER_FD_01_TABLE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVU_UPPER_FD_10( VURegs* VU, s32 info ) -{ - recVU_UPPER_FD_10_TABLE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVU_UPPER_FD_11( VURegs* VU, s32 info ) -{ - recVU_UPPER_FD_11_TABLE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVULowerOP( VURegs* VU, s32 info ) -{ - recVULowerOP_OPCODE[ VU->code & 0x3f ]( VU, info ); -} - -void recVULowerOP_T3_00( VURegs* VU, s32 info ) -{ - recVULowerOP_T3_00_OPCODE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVULowerOP_T3_01( VURegs* VU, s32 info ) -{ - recVULowerOP_T3_01_OPCODE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVULowerOP_T3_10( VURegs* VU, s32 info ) -{ - recVULowerOP_T3_10_OPCODE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); -} - -void recVULowerOP_T3_11( VURegs* VU, s32 info ) -{ - recVULowerOP_T3_11_OPCODE[ ( VU->code >> 6 ) & 0x1f ]( VU, info ); +void (*recVULowerOP_T3_00_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_MOVE , recVUMI_LQI , recVUMI_DIV , recVUMI_MTIR, + recVUMI_RNEXT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_MFP , recVUMI_XTOP , recVUMI_XGKICK, + recVUMI_ESADD , recVUMI_EATANxy, recVUMI_ESQRT, recVUMI_ESIN, +}; + +void (*recVULowerOP_T3_01_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_MR32 , recVUMI_SQI , recVUMI_SQRT , recVUMI_MFIR, + recVUMI_RGET , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUMI_XITOP, recVUunknown, + recVUMI_ERSADD, recVUMI_EATANxz, recVUMI_ERSQRT, recVUMI_EATAN, +}; + +void (*recVULowerOP_T3_10_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_LQD , recVUMI_RSQRT, recVUMI_ILWR, + recVUMI_RINIT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_ELENG , recVUMI_ESUM , recVUMI_ERCPR, recVUMI_EEXP, +}; + +void (*recVULowerOP_T3_11_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_SQD , recVUMI_WAITQ, recVUMI_ISWR, + recVUMI_RXOR , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_ERLENG, recVUunknown , recVUMI_WAITP, recVUunknown, +}; + +void (*recVULowerOP_OPCODE[64])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x20 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_IADD , recVUMI_ISUB , recVUMI_IADDI, recVUunknown, /* 0x30 */ + recVUMI_IAND , recVUMI_IOR , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVULowerOP_T3_00, recVULowerOP_T3_01, recVULowerOP_T3_10, recVULowerOP_T3_11, +}; + +void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info) = +{ + recVUMI_ADDx , recVUMI_ADDy , recVUMI_ADDz , recVUMI_ADDw, + recVUMI_SUBx , recVUMI_SUBy , recVUMI_SUBz , recVUMI_SUBw, + recVUMI_MADDx , recVUMI_MADDy , recVUMI_MADDz , recVUMI_MADDw, + recVUMI_MSUBx , recVUMI_MSUBy , recVUMI_MSUBz , recVUMI_MSUBw, + recVUMI_MAXx , recVUMI_MAXy , recVUMI_MAXz , recVUMI_MAXw, /* 0x10 */ + recVUMI_MINIx , recVUMI_MINIy , recVUMI_MINIz , recVUMI_MINIw, + recVUMI_MULx , recVUMI_MULy , recVUMI_MULz , recVUMI_MULw, + recVUMI_MULq , recVUMI_MAXi , recVUMI_MULi , recVUMI_MINIi, + recVUMI_ADDq , recVUMI_MADDq , recVUMI_ADDi , recVUMI_MADDi, /* 0x20 */ + recVUMI_SUBq , recVUMI_MSUBq , recVUMI_SUBi , recVUMI_MSUBi, + recVUMI_ADD , recVUMI_MADD , recVUMI_MUL , recVUMI_MAX, + recVUMI_SUB , recVUMI_MSUB , recVUMI_OPMSUB, recVUMI_MINI, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVU_UPPER_FD_00, recVU_UPPER_FD_01, recVU_UPPER_FD_10, recVU_UPPER_FD_11, +}; + +void (*recVU_UPPER_FD_00_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAx, recVUMI_SUBAx , recVUMI_MADDAx, recVUMI_MSUBAx, + recVUMI_ITOF0, recVUMI_FTOI0, recVUMI_MULAx , recVUMI_MULAq , + recVUMI_ADDAq, recVUMI_SUBAq, recVUMI_ADDA , recVUMI_SUBA , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_01_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAy , recVUMI_SUBAy , recVUMI_MADDAy, recVUMI_MSUBAy, + recVUMI_ITOF4 , recVUMI_FTOI4 , recVUMI_MULAy , recVUMI_ABS , + recVUMI_MADDAq, recVUMI_MSUBAq, recVUMI_MADDA , recVUMI_MSUBA , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_10_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAz , recVUMI_SUBAz , recVUMI_MADDAz, recVUMI_MSUBAz, + recVUMI_ITOF12, recVUMI_FTOI12, recVUMI_MULAz , recVUMI_MULAi , + recVUMI_ADDAi, recVUMI_SUBAi , recVUMI_MULA , recVUMI_OPMULA, + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_11_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAw , recVUMI_SUBAw , recVUMI_MADDAw, recVUMI_MSUBAw, + recVUMI_ITOF15, recVUMI_FTOI15, recVUMI_MULAw , recVUMI_CLIP , + recVUMI_MADDAi, recVUMI_MSUBAi, recVUunknown , recVUMI_NOP , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void recVU_UPPER_FD_00(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_00_TABLE[(VU->code >> 6) & 0x1f ](VU, info); } -void recVUunknown( VURegs* VU, s32 info ) -{ - Console::Notice("Unknown SVU micromode opcode called"); +void recVU_UPPER_FD_01(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_01_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVU_UPPER_FD_10(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_10_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVU_UPPER_FD_11(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_11_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP(VURegs* VU, s32 info) +{ + recVULowerOP_OPCODE[ VU->code & 0x3f ](VU, info); +} + +void recVULowerOP_T3_00(VURegs* VU, s32 info) +{ + recVULowerOP_T3_00_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_01(VURegs* VU, s32 info) +{ + recVULowerOP_T3_01_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_10(VURegs* VU, s32 info) +{ + recVULowerOP_T3_10_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_11(VURegs* VU, s32 info) +{ + recVULowerOP_T3_11_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVUunknown(VURegs* VU, s32 info) +{ + Console::Notice("Unknown SVU micromode opcode called"); } diff --git a/pcsx2/x86/ix86/implement/xmm/moremovs.h b/pcsx2/x86/ix86/implement/xmm/moremovs.h index 98cfafc230..fa30d4f441 100644 --- a/pcsx2/x86/ix86/implement/xmm/moremovs.h +++ b/pcsx2/x86/ix86/implement/xmm/moremovs.h @@ -64,6 +64,7 @@ public: // This happens when using Mem,Reg or Reg,Mem forms where the address is simple displacement // which can be checked for alignment at runtime. // + template< u8 Prefix, bool isAligned > class SimdImpl_MoveSSE { @@ -78,9 +79,17 @@ public: if( to != from ) xOpWrite0F( Prefix, OpcodeA, to, from ); } - __forceinline void operator()( const xRegisterSSE& to, const void* from ) const + __forceinline void operator()( const xRegisterSSE& to, void* from ) const { - xOpWrite0F( Prefix, (isAligned || ((uptr)from & 0x0f) == 0) ? OpcodeA : OpcodeU, to, from ); + u16 opcode; + + // @$@$@!$#@! GCC & Debug builds. + if (isAligned || ((uptr)from & 0x0f) == 0) + opcode = OpcodeA; + else + opcode = OpcodeU; + + xOpWrite0F( Prefix, opcode, to, from ); } __forceinline void operator()( void* to, const xRegisterSSE& from ) const @@ -92,7 +101,15 @@ public: { // ModSib form is aligned if it's displacement-only and the displacement is aligned: bool isReallyAligned = isAligned || ( ((from.Displacement & 0x0f) == 0) && from.Index.IsEmpty() && from.Base.IsEmpty() ); - xOpWrite0F( Prefix, isReallyAligned ? OpcodeA : OpcodeU, to, from ); + u16 opcode; + + // See previous comment. + if (isReallyAligned) + opcode = OpcodeA; + else + opcode = OpcodeU; + + xOpWrite0F( Prefix, opcode, to, from ); } __forceinline void operator()( const ModSibBase& to, const xRegisterSSE& from ) const diff --git a/pcsx2/x86/ix86/ix86_inlines.inl b/pcsx2/x86/ix86/ix86_inlines.inl index a88a7881b9..8e2dad8daf 100644 --- a/pcsx2/x86/ix86/ix86_inlines.inl +++ b/pcsx2/x86/ix86/ix86_inlines.inl @@ -216,7 +216,12 @@ namespace x86Emitter if( !is_s8( displacement ) ) { assert( false ); +// Don't ask. --arcum42 +#ifdef __LINUX__ +#ifndef DEBUG Console::Error( "Emitter Error: Invalid short jump displacement = 0x%x", params (int)displacement ); +#endif +#endif } BasePtr[-1] = (s8)displacement; } diff --git a/pcsx2/x86/ix86/ix86_types.h b/pcsx2/x86/ix86/ix86_types.h index 3dc26befd5..47d9337bd0 100644 --- a/pcsx2/x86/ix86/ix86_types.h +++ b/pcsx2/x86/ix86/ix86_types.h @@ -635,7 +635,7 @@ __forceinline void xWrite( T val ) JccComparisonType m_cc; // comparison type of the instruction public: - const int GetMaxInstructionSize() const + int GetMaxInstructionSize() const { jASSUME( m_cc != Jcc_Unknown ); return ( m_cc == Jcc_Unconditional ) ? 5 : 6; From f63bea379a2106e9aa0e528080caa0a4f3f486c7 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 8 May 2009 05:44:02 +0000 Subject: [PATCH 032/296] better fix for my elf-fix a while back. basically if you run an elf file now, the cdvd plugin will attempt to open once, and if it has an error (like loading no-game), it'll let you continue running the elf file, and it won't bug you again. but if it did open successfully, it'll initiate the cdvd plugin normally (allowing elf files to read the cdvd contents, like the Snes emu for ps2...) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1154 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/HostGui.h | 3 ++- pcsx2/Plugins.cpp | 15 +++++++++++---- pcsx2/System.cpp | 4 ++-- 3 files changed, 15 insertions(+), 7 deletions(-) diff --git a/pcsx2/HostGui.h b/pcsx2/HostGui.h index 3620d36cd8..d411f95acb 100644 --- a/pcsx2/HostGui.h +++ b/pcsx2/HostGui.h @@ -25,7 +25,8 @@ enum StartupMode { BootMode_Bios, BootMode_Quick, - BootMode_Elf + BootMode_Elf, + BootMode_Normal }; class StartupParams diff --git a/pcsx2/Plugins.cpp b/pcsx2/Plugins.cpp index bafd5a6012..dcd9709a4b 100644 --- a/pcsx2/Plugins.cpp +++ b/pcsx2/Plugins.cpp @@ -607,6 +607,7 @@ static PluginOpenStatusFlags OpenStatus = {0}; static bool loadp = false; static bool initp = false; +static bool cdvdElf = false; int LoadPlugins() { @@ -659,6 +660,7 @@ int InitPlugins() ret = FWinit(); if (ret != 0) { Msgbox::Alert("FWinit error: %d", params ret); return -1; } + cdvdElf = false; initp = true; return 0; } @@ -725,19 +727,24 @@ int OpenPlugins(const char* pTitleFilename) }*/ #endif - // Don't Open CDVD plugin if directly loading an elf file - if( !OpenStatus.CDVD && g_Startup.BootMode != BootMode_Elf) + // Don't Repetitively Open CDVD plugin if directly loading an elf file and Open failed once already + if( !OpenStatus.CDVD && !cdvdElf) { //first we need the data if (CDVDnewDiskCB) CDVDnewDiskCB(cdvdNewDiskCB); ret = CDVDopen(pTitleFilename); - if (ret != 0) { Msgbox::Alert("Error Opening CDVD Plugin"); goto OpenError; } + if (ret != 0) { + if (g_Startup.BootMode != BootMode_Elf) { Msgbox::Alert("Error Opening CDVD Plugin"); goto OpenError; } + else { Console::Notice("Running ELF File Without CDVD Plugin Support!"); cdvdElf = 1; goto skipOpenCDVD; } + } OpenStatus.CDVD = true; cdvdNewDiskCB(); } +skipOpenCDVD: + if( !OpenStatus.GS ) { ret = gsOpen(); if (ret != 0) { Msgbox::Alert("Error Opening GS Plugin"); goto OpenError; } @@ -833,7 +840,7 @@ OpenError: void ClosePlugins( bool closegs ) { - // Close pads first since they attatch to the GS's window. + // Close pads first since they attach to the GS's window. CLOSE_PLUGIN( PAD1 ); CLOSE_PLUGIN( PAD2 ); diff --git a/pcsx2/System.cpp b/pcsx2/System.cpp index eb06d7905d..aa1170d3b2 100644 --- a/pcsx2/System.cpp +++ b/pcsx2/System.cpp @@ -400,8 +400,8 @@ void SysPrepareExecution( const char* elf_file, bool use_bios ) return; } - // Don't Open CDVD plugin if directly loading an elf file - if (elf_file) { g_Startup.BootMode = BootMode_Elf; } + g_Startup.BootMode = (elf_file) ? BootMode_Elf : BootMode_Normal; + if (OpenPlugins(NULL) == -1) { return; } From b57e5f59892dc2086b942f75d19a56542e55d2aa Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 8 May 2009 05:46:53 +0000 Subject: [PATCH 033/296] More cleanup on iVUzerorec.cpp. Fix a few things from the last commit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1155 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVUzerorec.cpp | 235 +++++++++++++++++--------------- pcsx2/x86/ix86/ix86_inlines.inl | 5 +- 2 files changed, 125 insertions(+), 115 deletions(-) diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 4cfa941f58..39b0930833 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -534,7 +534,8 @@ u32 SuperVUGetVIAddr(int reg, int read) } case REG_MAC_FLAG: { - return (read == 2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); + u32 addr = (read == 2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); + return addr; } case REG_CLIP_FLAG: { @@ -796,16 +797,12 @@ list::iterator VuBaseBlock::GetInstIterAtPc(int instpc) list::iterator it; for (it = insts.begin(); it != insts.end(); ++it) { - if (it->type & INST_DUMMY) - continue; - - if (curpc == instpc) - break; + if (it->type & INST_DUMMY) continue; + if (curpc == instpc) break; curpc += 8; } - if (it != insts.end()) - return it; + if (it != insts.end()) return it; assert(0); return insts.begin(); @@ -821,11 +818,8 @@ void VuBaseBlock::GetInstsAtPc(int instpc, list& listinsts) list::iterator it; for (it = insts.begin(); it != insts.end(); ++it) { - if (it->type & INST_DUMMY) - continue; - - if (curpc == instpc) - break; + if (it->type & INST_DUMMY) continue; + if (curpc == instpc) break; curpc += 8; } @@ -838,8 +832,7 @@ void VuBaseBlock::GetInstsAtPc(int instpc, list& listinsts) // look for the pc in other blocks for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) { - if (*itblock == this) - continue; + if (*itblock == this) continue; if (instpc >= (*itblock)->startpc && instpc < (*itblock)->endpc) { @@ -1331,9 +1324,9 @@ static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const V if (pinst->regs[0].VIread & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) macflags = 1; -// if( pinst->regs[1].pipe == VUPIPE_FMAC && (pinst->regs[1].VFwrite==0&&!(pinst->regs[1].VIwrite&(1<regs[0].VIread |= (1<VIwrite |= lregs->VIwrite & (1<regs[1].pipe == VUPIPE_FMAC && (pinst->regs[1].VFwrite==0&&!(pinst->regs[1].VIwrite&(1<regs[0].VIread |= (1<VIwrite |= lregs->VIwrite & (1<addvars[0], itinst->keepvars[0], livevars[0]); // should propagate status flags whose parent insts are not in this block -// if( itinst->nParentPc >= 0 && (itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) -// newlive |= livevars[0]&((1<nParentPc >= 0 && (itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) +// newlive |= livevars[0]&((1<livevars[0] != newlive) { @@ -1800,8 +1792,8 @@ static void SuperVULivenessAnalysis() newlive = COMPUTE_LIVE(itnext->addvars[0], itnext->keepvars[0], itinst->livevars[0]); // should propagate status flags whose parent insts are not in this block -// if( itnext->nParentPc >= 0 && (itnext->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) && !(itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) -// newlive |= itinst->livevars[0]&((1<nParentPc >= 0 && (itnext->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) && !(itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) +// newlive |= itinst->livevars[0]&((1<livevars[0] != newlive) { @@ -1974,7 +1966,10 @@ static void SuperVUEliminateDeadCode() } } #ifdef _DEBUG - else curpc += 8; + else + { + curpc += 8; + } #endif itinst = itnext; ++itnext; @@ -2166,10 +2161,16 @@ void VuBaseBlock::AssignVFRegs() itinst->vfread0[i] = itinst->vfread1[i] = itinst->vfwrite[i] = itinst->vfacc[i] = -1; itinst->vfflush[i] = -1; - if (regs->VFread0) itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); - else if (regs->VIread & (1 << REG_VF0_FLAG)) itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); - if (regs->VFread1) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); - else if ((regs->VIread & (1 << REG_VF0_FLAG)) && regs->VFr1xyzw != 0xff) itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + if (regs->VFread0) + itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); + else if (regs->VIread & (1 << REG_VF0_FLAG)) + itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + + if (regs->VFread1) + itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); + else if ((regs->VIread & (1 << REG_VF0_FLAG)) && regs->VFr1xyzw != 0xff) + itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + if (regs->VIread & (1 << REG_ACC_FLAG)) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); int reusereg = -1; // 0 - VFwrite, 1 - VFAcc @@ -2228,29 +2229,38 @@ void VuBaseBlock::AssignVFRegs() { if (itnext == insts.end() || (itnext->livevars[1]®s->VFread0)) _freeXMMreg(itinst->vfread0[i]); + xmmregs[itinst->vfread0[i]].inuse = 1; xmmregs[itinst->vfread0[i]].reg = reg; xmmregs[itinst->vfread0[i]].type = type; xmmregs[itinst->vfread0[i]].mode = 0; - if (reusereg) itinst->vfacc[i] = itinst->vfread0[i]; - else itinst->vfwrite[i] = itinst->vfread0[i]; + + if (reusereg) + itinst->vfacc[i] = itinst->vfread0[i]; + else + itinst->vfwrite[i] = itinst->vfread0[i]; } else if (itinst->vfread1[i] >= 0 && lastwrite != itinst->vfread1[i] && (itnext == insts.end() || (regs->VFread1 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread1)) || !(itnext->livevars[1]&(1 << regs->VFread1)))))) { if (itnext == insts.end() || (itnext->livevars[1]®s->VFread1)) _freeXMMreg(itinst->vfread1[i]); + xmmregs[itinst->vfread1[i]].inuse = 1; xmmregs[itinst->vfread1[i]].reg = reg; xmmregs[itinst->vfread1[i]].type = type; xmmregs[itinst->vfread1[i]].mode = 0; - if (reusereg) itinst->vfacc[i] = itinst->vfread1[i]; - else itinst->vfwrite[i] = itinst->vfread1[i]; + if (reusereg) + itinst->vfacc[i] = itinst->vfread1[i]; + else + itinst->vfwrite[i] = itinst->vfread1[i]; } else { - if (reusereg) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); - else itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); + if (reusereg) + itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + else + itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); } } @@ -2270,8 +2280,8 @@ void VuBaseBlock::AssignVFRegs() if (i == 1 && (regs->VIwrite & (1 << REG_CLIP_FLAG))) { // CLIP inst, need two extra regs - if (free0 < 0) - free0 = _allocTempXMMreg(XMMT_FPS, -1); + if (free0 < 0) free0 = _allocTempXMMreg(XMMT_FPS, -1); + free1 = _allocTempXMMreg(XMMT_FPS, -1); free2 = _allocTempXMMreg(XMMT_FPS, -1); _freeXMMreg(free1); @@ -2281,8 +2291,7 @@ void VuBaseBlock::AssignVFRegs() { // EFU inst, need extra reg free1 = _allocTempXMMreg(XMMT_FPS, -1); - if (free0 == -1) - free0 = free1; + if (free0 == -1) free0 = free1; _freeXMMreg(free1); } @@ -2724,9 +2733,6 @@ static void SuperVURecompile() // make sure everything compiled for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - (*itblock)->type &= ~BLOCKTYPE_ANALYZED; - } { assert(((*itblock)->type & BLOCKTYPE_ANALYZED) && (*itblock)->pcode != NULL); } @@ -2973,7 +2979,8 @@ void VuBaseBlock::Recompile() #ifdef SUPERVU_X86CACHING if (nStartx86 >= 0) memcpy(x86regs, &s_vecRegArray[nStartx86], sizeof(x86regs)); - else _initX86regs(); + else + _initX86regs(); #else _initX86regs(); #endif @@ -3026,13 +3033,13 @@ void VuBaseBlock::Recompile() MOV32MtoR(EAX, s_PrevMACWrite); MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); } -// if( s_StatusRead != (uptr)&VU->VI[REG_STATUS_FLAG] ) { -// // only lower 8 bits valid! +// if( s_StatusRead != (uptr)&VU->VI[REG_STATUS_FLAG] ) { +// // only lower 8 bits valid! // MOVZX32M8toR(EAX, s_StatusRead); // MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); // } // if( s_MACRead != (uptr)&VU->VI[REG_MAC_FLAG] ) { -// // only lower 8 bits valid! +// // only lower 8 bits valid! // MOVZX32M8toR(EAX, s_MACRead); // MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); // } @@ -3051,7 +3058,9 @@ void VuBaseBlock::Recompile() _freeX86regs(); AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); + if (!branch) MOV32ItoM((uptr)&VU->VI[REG_TPC], endpc); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); } else @@ -3127,18 +3136,19 @@ void VuBaseBlock::Recompile() continue; } +#ifdef SUPERVU_INTERCACHING if (x86regs[i].type == (X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0))) { -#ifdef SUPERVU_INTERCACHING if (livevars[0] & (1 << x86regs[i].reg)) _freeX86reg(i); else x86regs[i].inuse = 0; -#else - _freeX86reg(i); -#endif } - else _freeX86reg(i); + else +#endif + { + _freeX86reg(i); + } } // realloc @@ -3192,8 +3202,8 @@ void VuBaseBlock::Recompile() case 0x13: // jr with uncon branch, uncond branch takes precendence (dropship) { -// s32 delta = (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) << 3; -// ADD32ItoRmOffset(ESP, delta, 0); +// s32 delta = (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) << 3; +// ADD32ItoRmOffset(ESP, delta, 0); ADD32ItoR(ESP, 8); // restore pChildJumps[0] = (u32*)((uptr)JMP32(0) | 0x80000000); @@ -3343,21 +3353,21 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) // if true, will check if parent block was executed before getting the results of the flags (superman returns) int nParentCheckForExecution = -1; -// int badaddrs[] = { -// 0x60,0x68,0x70,0x60,0x68,0x70,0x88,0x90,0x98,0x98,0xa8,0xb8,0x88,0x90, -// 0x4a8,0x4a8,0x398,0x3a0,0x3a8,0xa0 -// }; +// int badaddrs[] = { +// 0x60,0x68,0x70,0x60,0x68,0x70,0x88,0x90,0x98,0x98,0xa8,0xb8,0x88,0x90, +// 0x4a8,0x4a8,0x398,0x3a0,0x3a8,0xa0 +// }; #ifdef SUPERVU_PROPAGATEFLAGS if (nParentPc != -1 && (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc)) { -// if( !s_vu ) { -// for(int j = 0; j < ARRAYSIZE(badaddrs); ++j) { -// if( badaddrs[j] == nParentPc ) -// goto NoParent; -// } -// } +// if( !s_vu ) { +// for(int j = 0; j < ARRAYSIZE(badaddrs); ++j) { +// if( badaddrs[j] == nParentPc ) +// goto NoParent; +// } +// } list::iterator itblock; for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) @@ -3433,23 +3443,24 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) //MOV32ItoM(pparentinst->pStatusWrite, 0); } -// if( s_pCurBlock->prevFlagsOutOfBlock && s_StatusRead != NULL ) { -// // or instead since don't now which parent we came from -// MOV32MtoR(EAX, pparentinst->pStatusWrite); -// OR32RtoM(s_StatusRead, EAX); -// MOV32ItoM(pparentinst->pStatusWrite, 0); -// } +// if( s_pCurBlock->prevFlagsOutOfBlock && s_StatusRead != NULL ) { +// // or instead since don't now which parent we came from +// MOV32MtoR(EAX, pparentinst->pStatusWrite); +// OR32RtoM(s_StatusRead, EAX); +// MOV32ItoM(pparentinst->pStatusWrite, 0); +// } if (nParentCheckForExecution >= 0) { // don't now which parent we came from, so have to check -// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); -// if( s_StatusRead != NULL ) -// MOV32MtoR(EAX, s_StatusRead); -// else -// MOV32MtoR(EAX, (uptr)&VU->VI[REG_STATUS_FLAG]); -// s_StatusRead = tempstatus; +// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); +// if( s_StatusRead != NULL ) +// MOV32MtoR(EAX, s_StatusRead); +// else +// MOV32MtoR(EAX, (uptr)&VU->VI[REG_STATUS_FLAG]); +// s_StatusRead = tempstatus; + if (s_StatusRead == 0) s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; @@ -3478,14 +3489,14 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) { s_StatusRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pStatusWrite; if (s_StatusRead == 0) Console::WriteLn("super StatusRead allocation error!"); -// if( pc >= (u32)s_pCurBlock->endpc-8 ) { -// // towards the end, so variable might be leaded to another block (silent hill 4) -// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); -// MOV32MtoR(EAX, s_StatusRead); -// MOV32RtoM(tempstatus, EAX); -// MOV32ItoM(s_StatusRead, 0); -// s_StatusRead = tempstatus; -// } +// if( pc >= (u32)s_pCurBlock->endpc-8 ) { +// // towards the end, so variable might be leaded to another block (silent hill 4) +// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); +// MOV32MtoR(EAX, s_StatusRead); +// MOV32RtoM(tempstatus, EAX); +// MOV32ItoM(s_StatusRead, 0); +// s_StatusRead = tempstatus; +// } } } if (type & INST_MAC_WRITE) @@ -3507,25 +3518,24 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) //MOV32ItoM(pparentinst->pMACWrite, 0); } -// if( s_pCurBlock->prevFlagsOutOfBlock && s_MACRead != NULL ) { -// // or instead since don't now which parent we came from -// MOV32MtoR(EAX, pparentinst->pMACWrite); -// OR32RtoM(s_MACRead, EAX); -// MOV32ItoM(pparentinst->pMACWrite, 0); -// } +// if( s_pCurBlock->prevFlagsOutOfBlock && s_MACRead != NULL ) { +// // or instead since don't now which parent we came from +// MOV32MtoR(EAX, pparentinst->pMACWrite); +// OR32RtoM(s_MACRead, EAX); +// MOV32ItoM(pparentinst->pMACWrite, 0); +// } if (nParentCheckForExecution >= 0) { // don't now which parent we came from, so have to check -// uptr tempmac = (uptr)SuperVUStaticAlloc(4); -// if( s_MACRead != NULL ) -// MOV32MtoR(EAX, s_MACRead); -// else -// MOV32MtoR(EAX, (uptr)&VU->VI[REG_MAC_FLAG]); -// s_MACRead = tempmac; +// uptr tempmac = (uptr)SuperVUStaticAlloc(4); +// if( s_MACRead != NULL ) +// MOV32MtoR(EAX, s_MACRead); +// else +// MOV32MtoR(EAX, (uptr)&VU->VI[REG_MAC_FLAG]); +// s_MACRead = tempmac; - if (s_MACRead == 0) - s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; + if (s_MACRead == 0) s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); u8* jptr = JNE8(0); @@ -3548,14 +3558,14 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) else s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; -// if( pc >= (u32)s_pCurBlock->endpc-8 ) { -// // towards the end, so variable might be leaked to another block (silent hill 4) -// uptr tempMAC = (uptr)SuperVUStaticAlloc(4); -// MOV32MtoR(EAX, s_MACRead); -// MOV32RtoM(tempMAC, EAX); -// MOV32ItoM(s_MACRead, 0); -// s_MACRead = tempMAC; -// } +// if( pc >= (u32)s_pCurBlock->endpc-8 ) { +// // towards the end, so variable might be leaked to another block (silent hill 4) +// uptr tempMAC = (uptr)SuperVUStaticAlloc(4); +// MOV32MtoR(EAX, s_MACRead); +// MOV32RtoM(tempMAC, EAX); +// MOV32ItoM(s_MACRead, 0); +// s_MACRead = tempMAC; +// } } else { @@ -3809,12 +3819,11 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) _freeX86reg(x86temp); } - - if (ptr[0] == 0x800003bf) // waitq - SuperVUFlush(0, 1); - - if (ptr[0] == 0x800007bf) // waitp - SuperVUFlush(1, 1); + + // waitq + if (ptr[0] == 0x800003bf) SuperVUFlush(0, 1); + // waitp + if (ptr[0] == 0x800007bf) SuperVUFlush(1, 1); #ifdef PCSX2_DEVBUILD if (regs[1].VIread & regs[0].VIwrite & ~((1 << REG_Q) | (1 << REG_P) | (1 << REG_VF0_FLAG) | (1 << REG_ACC_FLAG))) @@ -4428,10 +4437,12 @@ void recVUMI_XGKICK(VURegs *VU, int info) } else { + // Erementar Gerad hack. + // Corrects the color of some graphics on a game that has somewhat scrambled graphics either way, and only works with ZeroGS. Not even ZZOgl. :) if (g_VUGameFixes & VUFIX_XGKICKDELAY2) s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc - pc) / 8); else - s_ScheduleXGKICK = 2; + s_ScheduleXGKICK = 2; } } diff --git a/pcsx2/x86/ix86/ix86_inlines.inl b/pcsx2/x86/ix86/ix86_inlines.inl index 8e2dad8daf..4462aecd0b 100644 --- a/pcsx2/x86/ix86/ix86_inlines.inl +++ b/pcsx2/x86/ix86/ix86_inlines.inl @@ -217,10 +217,9 @@ namespace x86Emitter { assert( false ); // Don't ask. --arcum42 -#ifdef __LINUX__ -#ifndef DEBUG +#if !defined(__LINUX__) || !defined(DEBUG) + Console::Error( "Emitter Error: Invalid short jump displacement = 0x%x", params (int)displacement ); -#endif #endif } BasePtr[-1] = (s8)displacement; From a80dfabca87ef09d34ef28671ebf8a14fae72115 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 8 May 2009 12:58:57 +0000 Subject: [PATCH 034/296] Initial commit of a new null plugin, GSnull (Linux only until I get around to porting it). Yes, it's just what it sounds like; running the pcsx2 without any graphics. May be useful under somewhat bizarre circumstances. :) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1156 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 272 +++++++++++ plugins/GSnull/GS.h | 77 +++ plugins/GSnull/License.txt | 342 ++++++++++++++ plugins/GSnull/Linux/Config.cpp | 171 +++++++ plugins/GSnull/Linux/Config.h | 25 + plugins/GSnull/Linux/Linux.cpp | 53 +++ plugins/GSnull/Linux/Linux.h | 17 + plugins/GSnull/Linux/callbacks.c | 34 ++ plugins/GSnull/Linux/callbacks.h | 14 + plugins/GSnull/Linux/gsnull.glade | 443 ++++++++++++++++++ plugins/GSnull/Linux/interface.c | 240 ++++++++++ plugins/GSnull/Linux/interface.h | 6 + plugins/GSnull/Linux/support.c | 144 ++++++ plugins/GSnull/Linux/support.h | 69 +++ plugins/GSnull/Makefile.am | 34 ++ plugins/GSnull/ReadMe.txt | 28 ++ plugins/GSnull/Windows/Config.cpp | 51 ++ plugins/GSnull/Windows/GS.def | 19 + plugins/GSnull/Windows/GS.rc | 116 +++++ plugins/GSnull/Windows/GSnull_vc2008.vcproj | 201 ++++++++ plugins/GSnull/Windows/Makefile | 29 ++ plugins/GSnull/Windows/ProjectRootDir.vsprops | 11 + plugins/GSnull/Windows/Win32.cpp | 81 ++++ plugins/GSnull/Windows/resource.h | 21 + plugins/GSnull/build.sh | 32 ++ plugins/GSnull/configure.ac | 80 ++++ plugins/GSnull/missing | 1 + 27 files changed, 2611 insertions(+) create mode 100644 plugins/GSnull/GS.cpp create mode 100644 plugins/GSnull/GS.h create mode 100644 plugins/GSnull/License.txt create mode 100644 plugins/GSnull/Linux/Config.cpp create mode 100644 plugins/GSnull/Linux/Config.h create mode 100644 plugins/GSnull/Linux/Linux.cpp create mode 100644 plugins/GSnull/Linux/Linux.h create mode 100644 plugins/GSnull/Linux/callbacks.c create mode 100644 plugins/GSnull/Linux/callbacks.h create mode 100644 plugins/GSnull/Linux/gsnull.glade create mode 100644 plugins/GSnull/Linux/interface.c create mode 100644 plugins/GSnull/Linux/interface.h create mode 100644 plugins/GSnull/Linux/support.c create mode 100644 plugins/GSnull/Linux/support.h create mode 100644 plugins/GSnull/Makefile.am create mode 100644 plugins/GSnull/ReadMe.txt create mode 100644 plugins/GSnull/Windows/Config.cpp create mode 100644 plugins/GSnull/Windows/GS.def create mode 100644 plugins/GSnull/Windows/GS.rc create mode 100644 plugins/GSnull/Windows/GSnull_vc2008.vcproj create mode 100644 plugins/GSnull/Windows/Makefile create mode 100644 plugins/GSnull/Windows/ProjectRootDir.vsprops create mode 100644 plugins/GSnull/Windows/Win32.cpp create mode 100644 plugins/GSnull/Windows/resource.h create mode 100644 plugins/GSnull/build.sh create mode 100644 plugins/GSnull/configure.ac create mode 120000 plugins/GSnull/missing diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp new file mode 100644 index 0000000000..96b566706a --- /dev/null +++ b/plugins/GSnull/GS.cpp @@ -0,0 +1,272 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include + +#include +#include + +using namespace std; + +#include "GS.h" +#ifdef __LINUX__ +Display *display; +int screen; +#endif + +const unsigned char version = PS2E_GS_VERSION; +const unsigned char revision = 0; +const unsigned char build = 1; // increase that with each version + +static char *libraryName = "GSnull Driver"; +FILE *gsLog; +Config conf; +string s_strIniPath="inis/GSnull.ini"; +void (*GSirq)(); + +EXPORT_C_(u32) PS2EgetLibType() +{ + return PS2E_LT_GS; +} + +EXPORT_C_(char*) PS2EgetLibName() +{ + return libraryName; +} + +EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) +{ + return (version<<16) | (revision<<8) | build; +} + +void __Log(char *fmt, ...) +{ + va_list list; + + if (!conf.Log || gsLog == NULL) return; + + va_start(list, fmt); + vfprintf(gsLog, fmt, list); + va_end(list); +} + +EXPORT_C_(void) GSprintf(int timeout, char *fmt, ...) +{ + va_list list; + char msg[512]; + char cmd[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + GS_LOG("GSprintf:%s", msg); + printf("GSprintf:%s", msg); +} + +void SysPrintf(const char *fmt, ...) +{ + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + GS_LOG(msg); + printf("GSnull:%s", msg); +} + +// basic funcs + +EXPORT_C_(s32) GSinit() +{ + LoadConfig(); + +#ifdef GS_LOG + gsLog = fopen("logs/gsLog.txt", "w"); + if (gsLog) setvbuf(gsLog, NULL, _IONBF, 0); + GS_LOG("GSnull plugin version %d,%d\n",revision,build); + GS_LOG("GS init\n"); +#endif + + SysPrintf("Initializing GSnull\n"); + return 0; +} + +EXPORT_C_(void) GSshutdown() +{ +#ifdef GS_LOG + if (gsLog) fclose(gsLog); +#endif + + SysPrintf("Shutting down GSnull\n"); +} + +EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) +{ +#ifdef GS_LOG + GS_LOG("GS open\n"); +#endif + assert( GSirq != NULL ); + +#ifdef __LINUX__ + display = XOpenDisplay(0); + screen = DefaultScreen(display); + + if( pDsp != NULL ) *(Display**)pDsp = display; +#else + // Haven't implemented anything for Windows yet. + pDsp = NULL; +#endif + SysPrintf("Opening GSnull\n"); + return 0; +} + +EXPORT_C_(void) GSclose() +{ + SysPrintf("Closing GSnull\n"); +#ifdef __LINUX__ + XCloseDisplay(display); +#endif +} + +EXPORT_C_(void) GSirqCallback(void (*callback)()) +{ + GSirq = callback; +} + +EXPORT_C_(s32) GSfreeze(int mode, freezeData *data) +{ + return 0; +} + +EXPORT_C_(s32) GStest() +{ + SysPrintf("Testing GSnull\n"); + return 0; +} + + +EXPORT_C_(void) GSvsync(int field) +{ +} + +EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) +{ +} + +EXPORT_C_(void) GSgifTransfer2(u32 *pMem, u32 size) +{ +} + +EXPORT_C_(void) GSgifTransfer3(u32 *pMem, u32 size) +{ +} + + // returns the last tag processed (64 bits) +EXPORT_C_(void) GSgetLastTag(u64* ptag) +{ +} + +EXPORT_C_(void) GSgifSoftReset(u32 mask) +{ + SysPrintf("Doing a soft reset of the GS plugin.\n"); +} + +EXPORT_C_(void) GSreadFIFO(u64 *mem) +{ +} + +EXPORT_C_(void) GSreadFIFO2(u64 *mem, int qwc) +{ +} + +// extended funcs + +// GSkeyEvent gets called when there is a keyEvent from the PAD plugin +EXPORT_C_(void) GSkeyEvent(keyEvent *ev) +{ + //SysPrintf("There is a key event.\n"); +} + +EXPORT_C_(void) GSchangeSaveState(int, const char* filename) +{ +} + +EXPORT_C_(void) GSmakeSnapshot(char *path) +{ + + SysPrintf("Taking a snapshot.\n"); +} + +EXPORT_C_(void) GSmakeSnapshot2(char *pathname, int* snapdone, int savejpg) +{ + SysPrintf("Taking a snapshot to %s.\n", pathname); +} + +EXPORT_C_(void) GSsetBaseMem(void*) +{ +} + +EXPORT_C_(void) GSsetGameCRC(int crc, int gameoptions) +{ + SysPrintf("Setting the crc to '%x' with 0x%x for options.\n", crc, gameoptions); +} + +// controls frame skipping in the GS, if this routine isn't present, frame skipping won't be done +EXPORT_C_(void) GSsetFrameSkip(int frameskip) +{ + SysPrintf("Frameskip set to %d.\n", frameskip); +} + +// if start is 1, starts recording spu2 data, else stops +// returns a non zero value if successful +// for now, pData is not used +EXPORT_C_(int) GSsetupRecording(int start, void* pData) +{ + if (start) + SysPrintf("Pretending to record.\n"); + else + SysPrintf("Pretending to stop recording.\n"); + + return 1; +} + +EXPORT_C_(void) GSreset() +{ + SysPrintf("Doing a reset of the GS plugin."); +} + +EXPORT_C_(void) GSwriteCSR(u32 value) +{ +} + +EXPORT_C_(void) GSgetDriverInfo(GSdriverInfo *info) +{ +} + +#ifdef _WIN32 +EXPORT_C_(s32) GSsetWindowInfo(winInfo *info) +{ + return 0; +} +#endif \ No newline at end of file diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h new file mode 100644 index 0000000000..5cee36c9af --- /dev/null +++ b/plugins/GSnull/GS.h @@ -0,0 +1,77 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __GS_H__ +#define __GS_H__ + +#include + +#ifdef __cplusplus +extern "C" +{ +#endif +#define GSdefs +#include "PS2Edefs.h" +#ifdef __cplusplus +} +#endif + +#ifdef _WIN32 + +#include +#include + +#else + +#include +#include +#include + +#endif + +/*#ifdef _MSC_VER +#define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK +#else +#define EXPORT_C_(type) extern "C" type +#endif*/ + +#ifdef _MSC_VER +#define EXPORT_C_(type) extern "C" type CALLBACK +#else +#define EXPORT_C_(type) extern "C" type +#endif + +#define GS_LOG __Log + +typedef struct +{ + int Log; +} Config; + +extern Config conf; +extern FILE *gsLog; + +extern void (*GSirq)(); + +extern void __Log(char *fmt, ...); +extern void SysMessage(char *fmt, ...); +extern void SysPrintf(const char *fmt, ...); +extern void SaveConfig(); +extern void LoadConfig(); + +#endif \ No newline at end of file diff --git a/plugins/GSnull/License.txt b/plugins/GSnull/License.txt new file mode 100644 index 0000000000..7d1f8605f8 --- /dev/null +++ b/plugins/GSnull/License.txt @@ -0,0 +1,342 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc. + 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +License is intended to guarantee your freedom to share and change free +software--to make sure the software is free for all its users. This +General Public License applies to most of the Free Software +Foundation's software and to any other program whose authors commit to +using it. (Some other Free Software Foundation software is covered by +the GNU Library General Public License instead.) You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +this service if you wish), that you receive source code or can get it +if you want it, that you can change the software or use pieces of it +in new free programs; and that you know you can do these things. + + To protect your rights, we need to make restrictions that forbid +anyone to deny you these rights or to ask you to surrender the rights. +These restrictions translate to certain responsibilities for you if you +distribute copies of the software, or if you modify it. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must give the recipients all the rights that +you have. You must make sure that they, too, receive or can get the +source code. And you must show them these terms so they know their +rights. + + We protect your rights with two steps: (1) copyright the software, and +(2) offer you this license which gives you legal permission to copy, +distribute and/or modify the software. + + Also, for each author's protection and ours, we want to make certain +that everyone understands that there is no warranty for this free +software. If the software is modified by someone else and passed on, we +want its recipients to know that what they have is not the original, so +that any problems introduced by others will not reflect on the original +authors' reputations. + + Finally, any free program is threatened constantly by software +patents. We wish to avoid the danger that redistributors of a free +program will individually obtain patent licenses, in effect making the +program proprietary. To prevent this, we have made it clear that any +patent must be licensed for everyone's free use or not licensed at all. + + The precise terms and conditions for copying, distribution and +modification follow. + + GNU GENERAL PUBLIC LICENSE + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License applies to any program or other work which contains +a notice placed by the copyright holder saying it may be distributed +under the terms of this General Public License. The "Program", below, +refers to any such program or work, and a "work based on the Program" +means either the Program or any derivative work under copyright law: +that is to say, a work containing the Program or a portion of it, +either verbatim or with modifications and/or translated into another +language. (Hereinafter, translation is included without limitation in +the term "modification".) Each licensee is addressed as "you". + +Activities other than copying, distribution and modification are not +covered by this License; they are outside its scope. The act of +running the Program is not restricted, and the output from the Program +is covered only if its contents constitute a work based on the +Program (independent of having been made by running the Program). +Whether that is true depends on what the Program does. + + 1. You may copy and distribute verbatim copies of the Program's +source code as you receive it, in any medium, provided that you +conspicuously and appropriately publish on each copy an appropriate +copyright notice and disclaimer of warranty; keep intact all the +notices that refer to this License and to the absence of any warranty; +and give any other recipients of the Program a copy of this License +along with the Program. + +You may charge a fee for the physical act of transferring a copy, and +you may at your option offer warranty protection in exchange for a fee. + + 2. You may modify your copy or copies of the Program or any portion +of it, thus forming a work based on the Program, and copy and +distribute such modifications or work under the terms of Section 1 +above, provided that you also meet all of these conditions: + + a) You must cause the modified files to carry prominent notices + stating that you changed the files and the date of any change. + + b) You must cause any work that you distribute or publish, that in + whole or in part contains or is derived from the Program or any + part thereof, to be licensed as a whole at no charge to all third + parties under the terms of this License. + + c) If the modified program normally reads commands interactively + when run, you must cause it, when started running for such + interactive use in the most ordinary way, to print or display an + announcement including an appropriate copyright notice and a + notice that there is no warranty (or else, saying that you provide + a warranty) and that users may redistribute the program under + these conditions, and telling the user how to view a copy of this + License. (Exception: if the Program itself is interactive but + does not normally print such an announcement, your work based on + the Program is not required to print an announcement.) + +These requirements apply to the modified work as a whole. If +identifiable sections of that work are not derived from the Program, +and can be reasonably considered independent and separate works in +themselves, then this License, and its terms, do not apply to those +sections when you distribute them as separate works. But when you +distribute the same sections as part of a whole which is a work based +on the Program, the distribution of the whole must be on the terms of +this License, whose permissions for other licensees extend to the +entire whole, and thus to each and every part regardless of who wrote it. + +Thus, it is not the intent of this section to claim rights or contest +your rights to work written entirely by you; rather, the intent is to +exercise the right to control the distribution of derivative or +collective works based on the Program. + +In addition, mere aggregation of another work not based on the Program +with the Program (or with a work based on the Program) on a volume of +a storage or distribution medium does not bring the other work under +the scope of this License. + + 3. You may copy and distribute the Program (or a work based on it, +under Section 2) in object code or executable form under the terms of +Sections 1 and 2 above provided that you also do one of the following: + + a) Accompany it with the complete corresponding machine-readable + source code, which must be distributed under the terms of Sections + 1 and 2 above on a medium customarily used for software interchange; or, + + b) Accompany it with a written offer, valid for at least three + years, to give any third party, for a charge no more than your + cost of physically performing source distribution, a complete + machine-readable copy of the corresponding source code, to be + distributed under the terms of Sections 1 and 2 above on a medium + customarily used for software interchange; or, + + c) Accompany it with the information you received as to the offer + to distribute corresponding source code. (This alternative is + allowed only for noncommercial distribution and only if you + received the program in object code or executable form with such + an offer, in accord with Subsection b above.) + +The source code for a work means the preferred form of the work for +making modifications to it. For an executable work, complete source +code means all the source code for all modules it contains, plus any +associated interface definition files, plus the scripts used to +control compilation and installation of the executable. However, as a +special exception, the source code distributed need not include +anything that is normally distributed (in either source or binary +form) with the major components (compiler, kernel, and so on) of the +operating system on which the executable runs, unless that component +itself accompanies the executable. + +If distribution of executable or object code is made by offering +access to copy from a designated place, then offering equivalent +access to copy the source code from the same place counts as +distribution of the source code, even though third parties are not +compelled to copy the source along with the object code. + + 4. You may not copy, modify, sublicense, or distribute the Program +except as expressly provided under this License. Any attempt +otherwise to copy, modify, sublicense or distribute the Program is +void, and will automatically terminate your rights under this License. +However, parties who have received copies, or rights, from you under +this License will not have their licenses terminated so long as such +parties remain in full compliance. + + 5. You are not required to accept this License, since you have not +signed it. However, nothing else grants you permission to modify or +distribute the Program or its derivative works. These actions are +prohibited by law if you do not accept this License. Therefore, by +modifying or distributing the Program (or any work based on the +Program), you indicate your acceptance of this License to do so, and +all its terms and conditions for copying, distributing or modifying +the Program or works based on it. + + 6. Each time you redistribute the Program (or any work based on the +Program), the recipient automatically receives a license from the +original licensor to copy, distribute or modify the Program subject to +these terms and conditions. You may not impose any further +restrictions on the recipients' exercise of the rights granted herein. +You are not responsible for enforcing compliance by third parties to +this License. + + 7. If, as a consequence of a court judgment or allegation of patent +infringement or for any other reason (not limited to patent issues), +conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot +distribute so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you +may not distribute the Program at all. For example, if a patent +license would not permit royalty-free redistribution of the Program by +all those who receive copies directly or indirectly through you, then +the only way you could satisfy both it and this License would be to +refrain entirely from distribution of the Program. + +If any portion of this section is held invalid or unenforceable under +any particular circumstance, the balance of the section is intended to +apply and the section as a whole is intended to apply in other +circumstances. + +It is not the purpose of this section to induce you to infringe any +patents or other property right claims or to contest validity of any +such claims; this section has the sole purpose of protecting the +integrity of the free software distribution system, which is +implemented by public license practices. Many people have made +generous contributions to the wide range of software distributed +through that system in reliance on consistent application of that +system; it is up to the author/donor to decide if he or she is willing +to distribute software through any other system and a licensee cannot +impose that choice. + +This section is intended to make thoroughly clear what is believed to +be a consequence of the rest of this License. + + 8. If the distribution and/or use of the Program is restricted in +certain countries either by patents or by copyrighted interfaces, the +original copyright holder who places the Program under this License +may add an explicit geographical distribution limitation excluding +those countries, so that distribution is permitted only in or among +countries not thus excluded. In such case, this License incorporates +the limitation as if written in the body of this License. + + 9. The Free Software Foundation may publish revised and/or new versions +of the General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + +Each version is given a distinguishing version number. If the Program +specifies a version number of this License which applies to it and "any +later version", you have the option of following the terms and conditions +either of that version or of any later version published by the Free +Software Foundation. If the Program does not specify a version number of +this License, you may choose any version ever published by the Free Software +Foundation. + + 10. If you wish to incorporate parts of the Program into other free +programs whose distribution conditions are different, write to the author +to ask for permission. For software which is copyrighted by the Free +Software Foundation, write to the Free Software Foundation; we sometimes +make exceptions for this. Our decision will be guided by the two goals +of preserving the free status of all derivatives of our free software and +of promoting the sharing and reuse of software generally. + + NO WARRANTY + + 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY +FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN +OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES +PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED +OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS +TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE +PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, +REPAIR OR CORRECTION. + + 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR +REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, +INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING +OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED +TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY +YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER +PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE +POSSIBILITY OF SUCH DAMAGES. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +convey the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + + +Also add information on how to contact you by electronic and paper mail. + +If the program is interactive, make it output a short notice like this +when it starts in an interactive mode: + + Gnomovision version 69, Copyright (C) year name of author + Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, the commands you use may +be called something other than `show w' and `show c'; they could even be +mouse-clicks or menu items--whatever suits your program. + +You should also get your employer (if you work as a programmer) or your +school, if any, to sign a "copyright disclaimer" for the program, if +necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the program + `Gnomovision' (which makes passes at compilers) written by James Hacker. + + , 1 April 1989 + Ty Coon, President of Vice + +This General Public License does not permit incorporating your program into +proprietary programs. If your program is a subroutine library, you may +consider it more useful to permit linking proprietary applications with the +library. If this is what you want to do, use the GNU Library General +Public License instead of this License. + + diff --git a/plugins/GSnull/Linux/Config.cpp b/plugins/GSnull/Linux/Config.cpp new file mode 100644 index 0000000000..d292f5748f --- /dev/null +++ b/plugins/GSnull/Linux/Config.cpp @@ -0,0 +1,171 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include +#include +#include +#include +#include +using namespace std; + +#include "GS.h" +#include "Config.h" + + #ifdef __cplusplus +extern "C" +{ +#endif + +#include "support.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus +} +#endif + +GtkWidget *MsgDlg, *About, *Conf; +extern string s_strIniPath; + +void OnMsg_Ok() +{ + gtk_widget_destroy(MsgDlg); + gtk_main_quit(); +} + +void cfgSysMessage(char *fmt, ...) +{ + GtkWidget *Ok,*Txt; + GtkWidget *Box,*Box1; + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + if (msg[strlen(msg) - 1] == '\n') msg[strlen(msg)-1] = 0; + + MsgDlg = gtk_window_new (GTK_WINDOW_POPUP); + gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "FireWire Msg"); + gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); + + Box = gtk_vbox_new(5, 0); + gtk_container_add(GTK_CONTAINER(MsgDlg), Box); + gtk_widget_show(Box); + + Txt = gtk_label_new(msg); + + gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); + gtk_widget_show(Txt); + + Box1 = gtk_hbutton_box_new(); + gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); + gtk_widget_show(Box1); + + Ok = gtk_button_new_with_label("Ok"); + gtk_signal_connect (GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); + gtk_container_add(GTK_CONTAINER(Box1), Ok); + GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); + gtk_widget_show(Ok); + + gtk_widget_show(MsgDlg); + + gtk_main(); +} + +void OnAbout_Ok(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(About); + gtk_main_quit(); +} + +void CFGabout() +{ + About = create_About(); + gtk_widget_show_all(About); + gtk_main(); +} + +void OnConf_Ok(GtkButton *button, gpointer user_data) { + SaveConfig(); + + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void OnConf_Cancel(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void CFGconfigure() +{ + Conf = create_Config(); + + LoadConfig(); + + gtk_widget_show_all(Conf); + gtk_main(); +} + +long CFGmessage(char *msg) { + cfgSysMessage(msg); + + return 0; +} + +void LoadConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "r"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + SaveConfig();//save and return + return; + } + //fscanf(f, "options = %hhx\n", &confOptions); + fclose(f); +} + +void SaveConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg,"w"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + return; + } + + //fprintf(f, "options = %hhx\n", confOptions); + fclose(f); +} + diff --git a/plugins/GSnull/Linux/Config.h b/plugins/GSnull/Linux/Config.h new file mode 100644 index 0000000000..fca002d5be --- /dev/null +++ b/plugins/GSnull/Linux/Config.h @@ -0,0 +1,25 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +void SaveConf(); +void LoadConf(); + +extern long CFGmessage(char *msg); +extern void CFGconfigure(); +extern void cfgSysMessage(char *fmt, ...); +extern void CFGabout(); diff --git a/plugins/GSnull/Linux/Linux.cpp b/plugins/GSnull/Linux/Linux.cpp new file mode 100644 index 0000000000..7bbd963ccf --- /dev/null +++ b/plugins/GSnull/Linux/Linux.cpp @@ -0,0 +1,53 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include "GS.h" +#include "Config.h" + +void SysMessage(char *fmt, ...) +{ + va_list list; + char msg[512]; + char cmd[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + cfgSysMessage(msg); +} + +void GSconfigure() +{ + CFGconfigure(); +} + +void GSabout() +{ + CFGabout(); +} + diff --git a/plugins/GSnull/Linux/Linux.h b/plugins/GSnull/Linux/Linux.h new file mode 100644 index 0000000000..65cf4eac58 --- /dev/null +++ b/plugins/GSnull/Linux/Linux.h @@ -0,0 +1,17 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ \ No newline at end of file diff --git a/plugins/GSnull/Linux/callbacks.c b/plugins/GSnull/Linux/callbacks.c new file mode 100644 index 0000000000..e87ed8be2d --- /dev/null +++ b/plugins/GSnull/Linux/callbacks.c @@ -0,0 +1,34 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/GSnull/Linux/callbacks.h b/plugins/GSnull/Linux/callbacks.h new file mode 100644 index 0000000000..e606c242f0 --- /dev/null +++ b/plugins/GSnull/Linux/callbacks.h @@ -0,0 +1,14 @@ +#include + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/GSnull/Linux/gsnull.glade b/plugins/GSnull/Linux/gsnull.glade new file mode 100644 index 0000000000..cef4eab65e --- /dev/null +++ b/plugins/GSnull/Linux/gsnull.glade @@ -0,0 +1,443 @@ + + + + + + + 5 + True + DEV9config + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + 5 + True + True + 5 + + + + True + Device: + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + False + True + False + True + False + + + + True + True + True + True + 0 + + True + * + False + + + + + + True + GTK_SELECTION_BROWSE + + + + True + + + + True + + False + False + GTK_JUSTIFY_LEFT + False + False + 0 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + + + + + + + 0 + False + False + + + + + + + + True + Ethernet + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + 5 + True + True + 5 + + + + True + Device: + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + False + True + False + True + False + + + + True + True + True + True + 0 + + True + * + False + + + + + + True + GTK_SELECTION_BROWSE + + + + True + + + + True + + False + False + GTK_JUSTIFY_LEFT + False + False + 0 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + + + + + + + 0 + False + False + + + + + + + + True + Hdd + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + + True + True + True + Cancel + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + + 5 + True + DEV9about + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + FireWire Driver + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + Author: linuzappz <linuzappz@hotmail.com> + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + diff --git a/plugins/GSnull/Linux/interface.c b/plugins/GSnull/Linux/interface.c new file mode 100644 index 0000000000..5c23358a21 --- /dev/null +++ b/plugins/GSnull/Linux/interface.c @@ -0,0 +1,240 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + +#define GLADE_HOOKUP_OBJECT(component,widget,name) \ + g_object_set_data_full (G_OBJECT (component), name, \ + gtk_widget_ref (widget), (GDestroyNotify) gtk_widget_unref) + +#define GLADE_HOOKUP_OBJECT_NO_REF(component,widget,name) \ + g_object_set_data (G_OBJECT (component), name, widget) + +GtkWidget* +create_Config (void) +{ + GtkWidget *Config; + GtkWidget *vbox1; + GtkWidget *frame2; + GtkWidget *hbox1; + GtkWidget *label4; + GtkWidget *GtkCombo_Eth; + GList *GtkCombo_Eth_items = NULL; + GtkWidget *combo_entry1; + GtkWidget *label1; + GtkWidget *frame3; + GtkWidget *hbox2; + GtkWidget *label5; + GtkWidget *GtkCombo_Hdd; + GList *GtkCombo_Hdd_items = NULL; + GtkWidget *entry1; + GtkWidget *label15; + GtkWidget *hbuttonbox1; + GtkWidget *button1; + GtkWidget *button2; + + Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); + gtk_container_set_border_width (GTK_CONTAINER (Config), 5); + gtk_window_set_title (GTK_WINDOW (Config), _("DEV9config")); + + vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); + gtk_widget_show (vbox1); + gtk_container_add (GTK_CONTAINER (Config), vbox1); + gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); + + frame2 = gtk_frame_new (NULL); + gtk_widget_set_name (frame2, "frame2"); + gtk_widget_show (frame2); + gtk_box_pack_start (GTK_BOX (vbox1), frame2, TRUE, TRUE, 0); + + hbox1 = gtk_hbox_new (TRUE, 5); + gtk_widget_set_name (hbox1, "hbox1"); + gtk_widget_show (hbox1); + gtk_container_add (GTK_CONTAINER (frame2), hbox1); + gtk_container_set_border_width (GTK_CONTAINER (hbox1), 5); + + label4 = gtk_label_new (_("Device:")); + gtk_widget_set_name (label4, "label4"); + gtk_widget_show (label4); + gtk_box_pack_start (GTK_BOX (hbox1), label4, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label4), GTK_JUSTIFY_CENTER); + + GtkCombo_Eth = gtk_combo_new (); + g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Eth)->popwin), + "GladeParentKey", GtkCombo_Eth); + gtk_widget_set_name (GtkCombo_Eth, "GtkCombo_Eth"); + gtk_widget_show (GtkCombo_Eth); + gtk_box_pack_start (GTK_BOX (hbox1), GtkCombo_Eth, FALSE, FALSE, 0); + GtkCombo_Eth_items = g_list_append (GtkCombo_Eth_items, (gpointer) ""); + gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Eth), GtkCombo_Eth_items); + g_list_free (GtkCombo_Eth_items); + + combo_entry1 = GTK_COMBO (GtkCombo_Eth)->entry; + gtk_widget_set_name (combo_entry1, "combo_entry1"); + gtk_widget_show (combo_entry1); + + label1 = gtk_label_new (_("Ethernet")); + gtk_widget_set_name (label1, "label1"); + gtk_widget_show (label1); + gtk_frame_set_label_widget (GTK_FRAME (frame2), label1); + + frame3 = gtk_frame_new (NULL); + gtk_widget_set_name (frame3, "frame3"); + gtk_widget_show (frame3); + gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + + hbox2 = gtk_hbox_new (TRUE, 5); + gtk_widget_set_name (hbox2, "hbox2"); + gtk_widget_show (hbox2); + gtk_container_add (GTK_CONTAINER (frame3), hbox2); + gtk_container_set_border_width (GTK_CONTAINER (hbox2), 5); + + label5 = gtk_label_new (_("Device:")); + gtk_widget_set_name (label5, "label5"); + gtk_widget_show (label5); + gtk_box_pack_start (GTK_BOX (hbox2), label5, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label5), GTK_JUSTIFY_CENTER); + + GtkCombo_Hdd = gtk_combo_new (); + g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Hdd)->popwin), + "GladeParentKey", GtkCombo_Hdd); + gtk_widget_set_name (GtkCombo_Hdd, "GtkCombo_Hdd"); + gtk_widget_show (GtkCombo_Hdd); + gtk_box_pack_start (GTK_BOX (hbox2), GtkCombo_Hdd, FALSE, FALSE, 0); + GtkCombo_Hdd_items = g_list_append (GtkCombo_Hdd_items, (gpointer) ""); + gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Hdd), GtkCombo_Hdd_items); + g_list_free (GtkCombo_Hdd_items); + + entry1 = GTK_COMBO (GtkCombo_Hdd)->entry; + gtk_widget_set_name (entry1, "entry1"); + gtk_widget_show (entry1); + + label15 = gtk_label_new (_("Hdd")); + gtk_widget_set_name (label15, "label15"); + gtk_widget_show (label15); + gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + + hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); + gtk_widget_show (hbuttonbox1); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + + button1 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button1, "button1"); + gtk_widget_show (button1); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); + GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + + button2 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button2, "button2"); + gtk_widget_show (button2); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); + GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button1, "clicked", + G_CALLBACK (OnConf_Ok), + NULL); + g_signal_connect ((gpointer) button2, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); + GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); + GLADE_HOOKUP_OBJECT (Config, frame2, "frame2"); + GLADE_HOOKUP_OBJECT (Config, hbox1, "hbox1"); + GLADE_HOOKUP_OBJECT (Config, label4, "label4"); + GLADE_HOOKUP_OBJECT (Config, GtkCombo_Eth, "GtkCombo_Eth"); + GLADE_HOOKUP_OBJECT (Config, combo_entry1, "combo_entry1"); + GLADE_HOOKUP_OBJECT (Config, label1, "label1"); + GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); + GLADE_HOOKUP_OBJECT (Config, hbox2, "hbox2"); + GLADE_HOOKUP_OBJECT (Config, label5, "label5"); + GLADE_HOOKUP_OBJECT (Config, GtkCombo_Hdd, "GtkCombo_Hdd"); + GLADE_HOOKUP_OBJECT (Config, entry1, "entry1"); + GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); + GLADE_HOOKUP_OBJECT (Config, button1, "button1"); + GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + + return Config; +} + +GtkWidget* +create_About (void) +{ + GtkWidget *About; + GtkWidget *vbox2; + GtkWidget *label2; + GtkWidget *label3; + GtkWidget *hbuttonbox2; + GtkWidget *button3; + + About = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_title (GTK_WINDOW (About), _("DEV9about")); + + vbox2 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox2, "vbox2"); + gtk_widget_show (vbox2); + gtk_container_add (GTK_CONTAINER (About), vbox2); + gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); + + label2 = gtk_label_new (_("FireWire Driver")); + gtk_widget_set_name (label2, "label2"); + gtk_widget_show (label2); + gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); + + label3 = gtk_label_new (_("Author: linuzappz ")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); + + hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); + gtk_widget_show (hbuttonbox2); + gtk_box_pack_start (GTK_BOX (vbox2), hbuttonbox2, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox2), 30); + + button3 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button3, "button3"); + gtk_widget_show (button3); + gtk_container_add (GTK_CONTAINER (hbuttonbox2), button3); + GTK_WIDGET_SET_FLAGS (button3, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button3, "clicked", + G_CALLBACK (OnAbout_Ok), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); + GLADE_HOOKUP_OBJECT (About, vbox2, "vbox2"); + GLADE_HOOKUP_OBJECT (About, label2, "label2"); + GLADE_HOOKUP_OBJECT (About, label3, "label3"); + GLADE_HOOKUP_OBJECT (About, hbuttonbox2, "hbuttonbox2"); + GLADE_HOOKUP_OBJECT (About, button3, "button3"); + + return About; +} + diff --git a/plugins/GSnull/Linux/interface.h b/plugins/GSnull/Linux/interface.h new file mode 100644 index 0000000000..69c303ba3b --- /dev/null +++ b/plugins/GSnull/Linux/interface.h @@ -0,0 +1,6 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +GtkWidget* create_Config (void); +GtkWidget* create_About (void); diff --git a/plugins/GSnull/Linux/support.c b/plugins/GSnull/Linux/support.c new file mode 100644 index 0000000000..00aff29822 --- /dev/null +++ b/plugins/GSnull/Linux/support.c @@ -0,0 +1,144 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include + +#include "support.h" + +GtkWidget* +lookup_widget (GtkWidget *widget, + const gchar *widget_name) +{ + GtkWidget *parent, *found_widget; + + for (;;) + { + if (GTK_IS_MENU (widget)) + parent = gtk_menu_get_attach_widget (GTK_MENU (widget)); + else + parent = widget->parent; + if (!parent) + parent = (GtkWidget*) g_object_get_data (G_OBJECT (widget), "GladeParentKey"); + if (parent == NULL) + break; + widget = parent; + } + + found_widget = (GtkWidget*) g_object_get_data (G_OBJECT (widget), + widget_name); + if (!found_widget) + g_warning ("Widget not found: %s", widget_name); + return found_widget; +} + +static GList *pixmaps_directories = NULL; + +/* Use this function to set the directory containing installed pixmaps. */ +void +add_pixmap_directory (const gchar *directory) +{ + pixmaps_directories = g_list_prepend (pixmaps_directories, + g_strdup (directory)); +} + +/* This is an internally used function to find pixmap files. */ +static gchar* +find_pixmap_file (const gchar *filename) +{ + GList *elem; + + /* We step through each of the pixmaps directory to find it. */ + elem = pixmaps_directories; + while (elem) + { + gchar *pathname = g_strdup_printf ("%s%s%s", (gchar*)elem->data, + G_DIR_SEPARATOR_S, filename); + if (g_file_test (pathname, G_FILE_TEST_EXISTS)) + return pathname; + g_free (pathname); + elem = elem->next; + } + return NULL; +} + +/* This is an internally used function to create pixmaps. */ +GtkWidget* +create_pixmap (GtkWidget *widget, + const gchar *filename) +{ + gchar *pathname = NULL; + GtkWidget *pixmap; + + if (!filename || !filename[0]) + return gtk_image_new (); + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return gtk_image_new (); + } + + pixmap = gtk_image_new_from_file (pathname); + g_free (pathname); + return pixmap; +} + +/* This is an internally used function to create pixmaps. */ +GdkPixbuf* +create_pixbuf (const gchar *filename) +{ + gchar *pathname = NULL; + GdkPixbuf *pixbuf; + GError *error = NULL; + + if (!filename || !filename[0]) + return NULL; + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return NULL; + } + + pixbuf = gdk_pixbuf_new_from_file (pathname, &error); + if (!pixbuf) + { + fprintf (stderr, "Failed to load pixbuf file: %s: %s\n", + pathname, error->message); + g_error_free (error); + } + g_free (pathname); + return pixbuf; +} + +/* This is used to set ATK action descriptions. */ +void +glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description) +{ + gint n_actions, i; + + n_actions = atk_action_get_n_actions (action); + for (i = 0; i < n_actions; i++) + { + if (!strcmp (atk_action_get_name (action, i), action_name)) + atk_action_set_description (action, i, description); + } +} + diff --git a/plugins/GSnull/Linux/support.h b/plugins/GSnull/Linux/support.h new file mode 100644 index 0000000000..a32649e53c --- /dev/null +++ b/plugins/GSnull/Linux/support.h @@ -0,0 +1,69 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* + * Standard gettext macros. + */ +#ifdef ENABLE_NLS +# include +# undef _ +# define _(String) dgettext (PACKAGE, String) +# define Q_(String) g_strip_context ((String), gettext (String)) +# ifdef gettext_noop +# define N_(String) gettext_noop (String) +# else +# define N_(String) (String) +# endif +#else +# define textdomain(String) (String) +# define gettext(String) (String) +# define dgettext(Domain,Message) (Message) +# define dcgettext(Domain,Message,Type) (Message) +# define bindtextdomain(Domain,Directory) (Domain) +# define _(String) (String) +# define Q_(String) g_strip_context ((String), (String)) +# define N_(String) (String) +#endif + + +/* + * Public Functions. + */ + +/* + * This function returns a widget in a component created by Glade. + * Call it with the toplevel widget in the component (i.e. a window/dialog), + * or alternatively any widget in the component, and the name of the widget + * you want returned. + */ +GtkWidget* lookup_widget (GtkWidget *widget, + const gchar *widget_name); + + +/* Use this function to set the directory containing installed pixmaps. */ +void add_pixmap_directory (const gchar *directory); + + +/* + * Private Functions. + */ + +/* This is used to create the pixmaps used in the interface. */ +GtkWidget* create_pixmap (GtkWidget *widget, + const gchar *filename); + +/* This is used to create the pixbufs used in the interface. */ +GdkPixbuf* create_pixbuf (const gchar *filename); + +/* This is used to set ATK action descriptions. */ +void glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description); + diff --git a/plugins/GSnull/Makefile.am b/plugins/GSnull/Makefile.am new file mode 100644 index 0000000000..d976e49677 --- /dev/null +++ b/plugins/GSnull/Makefile.am @@ -0,0 +1,34 @@ +# Create a shared library libGSnull +AUTOMAKE_OPTIONS = foreign +noinst_LIBRARIES = libGSnull.a +INCLUDES = -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty -I@srcdir@/Linux + +libGSnull_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) +libGSnull_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) + +# Create a shared object by faking an exe (thanks to ODE makefiles) +traplibdir=$(prefix) + +if DEBUGBUILD +preext=d +endif + +EXEEXT=$(preext)@so_ext@ + +traplib_PROGRAMS=libGSnull +libGSnull_SOURCES= +libGSnull_DEPENDENCIES = libGSnull.a +libGSnull_LDFLAGS= @SHARED_LDFLAGS@ +libGSnull_LDFLAGS+=-Wl,-soname,@libGSnull_SONAME@ +libGSnull_LDADD=$(libGSnull_a_OBJECTS) + +libGSnull_a_SOURCES = GS.cpp Linux/Config.cpp Linux/Linux.cpp \ +FW.h Linux/Config.h Linux/Linux.h + +libGSnull_a_SOURCES += \ +Linux/interface.h Linux/support.c \ +Linux/interface.c Linux/support.h \ +Linux/callbacks.h +#Linux/callbacks.c + +#SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/GSnull/ReadMe.txt b/plugins/GSnull/ReadMe.txt new file mode 100644 index 0000000000..28bf61f48a --- /dev/null +++ b/plugins/GSnull/ReadMe.txt @@ -0,0 +1,28 @@ +GSnull v0.1 +------------- + + This is an extension to use with play station2 emulators + as PCSX2 (only one right now). + The plugin is free open source code. + + Usage: + ----- + Place the file "GSnull.so.x.x.x" (linux) or "GSnull.dll" (win32) + at the Plugin directory of the Emulator to use it. + + Changes: + ------- + + v0.1: +* First Release +* Tested with Pcsx2 +* Based off of FWnull. +* Currently Linux only. This may change. + + Authors: + ------- +arcum42 + +based off work by: + shadow + linuzappz diff --git a/plugins/GSnull/Windows/Config.cpp b/plugins/GSnull/Windows/Config.cpp new file mode 100644 index 0000000000..62db91dd61 --- /dev/null +++ b/plugins/GSnull/Windows/Config.cpp @@ -0,0 +1,51 @@ +#include + +#include "../FW.h" + +extern HINSTANCE hInst; +void SaveConfig() +{ + + Config *Conf1 = &conf; + char *szTemp; + char szIniFile[256], szValue[256]; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if(!szTemp) return; + strcpy(szTemp, "\\inis\\fwnull.ini"); + sprintf(szValue,"%u",Conf1->Log); + WritePrivateProfileString("Interface", "Logging",szValue,szIniFile); + +} + +void LoadConfig() { + FILE *fp; + + + Config *Conf1 = &conf; + char *szTemp; + char szIniFile[256], szValue[256]; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if(!szTemp) return ; + strcpy(szTemp, "\\inis\\fwnull.ini"); + fp=fopen("inis\\fwnull.ini","rt");//check if firewirenull.ini really exists + if (!fp) + { + CreateDirectory("inis",NULL); + memset(&conf, 0, sizeof(conf)); + conf.Log = 0;//default value + SaveConfig();//save and return + return ; + } + fclose(fp); + GetPrivateProfileString("Interface", "Logging", NULL, szValue, 20, szIniFile); + Conf1->Log = strtoul(szValue, NULL, 10); + return ; + +} + diff --git a/plugins/GSnull/Windows/GS.def b/plugins/GSnull/Windows/GS.def new file mode 100644 index 0000000000..51e786a1fd --- /dev/null +++ b/plugins/GSnull/Windows/GS.def @@ -0,0 +1,19 @@ +; FireWire.def : Declares the module parameters for the DLL. + +LIBRARY "FWnull" + +EXPORTS + ; Explicit exports can go here + PS2EgetLibType @2 + PS2EgetLibName @3 + PS2EgetLibVersion2 @4 + FWinit @5 + FWshutdown @6 + FWopen @7 + FWclose @8 + FWread32 @11 + FWwrite32 @14 + FWirqCallback @15 + FWconfigure @17 + FWtest @18 + FWabout @19 diff --git a/plugins/GSnull/Windows/GS.rc b/plugins/GSnull/Windows/GS.rc new file mode 100644 index 0000000000..a505f6108d --- /dev/null +++ b/plugins/GSnull/Windows/GS.rc @@ -0,0 +1,116 @@ +// Microsoft Visual C++ generated resource script. +// +#include "resource.h" + +#define APSTUDIO_READONLY_SYMBOLS +///////////////////////////////////////////////////////////////////////////// +// +// Generated from the TEXTINCLUDE 2 resource. +// +#include "afxresmw.h" + +///////////////////////////////////////////////////////////////////////////// +#undef APSTUDIO_READONLY_SYMBOLS + +///////////////////////////////////////////////////////////////////////////// +// Spanish (Argentina) resources + +#if !defined(AFX_RESOURCE_DLL) || defined(AFX_TARG_ESS) +#ifdef _WIN32 +LANGUAGE LANG_SPANISH, SUBLANG_SPANISH_ARGENTINA +#pragma code_page(1252) +#endif //_WIN32 + +#ifdef APSTUDIO_INVOKED +///////////////////////////////////////////////////////////////////////////// +// +// TEXTINCLUDE +// + +1 TEXTINCLUDE +BEGIN + "resource.h\0" +END + +2 TEXTINCLUDE +BEGIN + "#include ""afxresmw.h""\r\0" +END + +3 TEXTINCLUDE +BEGIN + "\r\0" +END + +#endif // APSTUDIO_INVOKED + + +///////////////////////////////////////////////////////////////////////////// +// +// Dialog +// + +IDD_CONFIG DIALOGEX 0, 0, 212, 121 +STYLE DS_SETFONT | DS_MODALFRAME | WS_POPUP | WS_CAPTION | WS_SYSMENU +CAPTION "Firewireconfigure" +FONT 8, "MS Sans Serif", 0, 0, 0x0 +BEGIN + DEFPUSHBUTTON "OK",IDOK,48,100,50,14 + PUSHBUTTON "Cancel",IDCANCEL,113,100,50,14 + CONTROL "Enable Logging (for develop use only)",IDC_LOGGING, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,7,187,13 +END + +IDD_ABOUT DIALOGEX 0, 0, 177, 106 +STYLE DS_SETFONT | DS_MODALFRAME | DS_CENTER | WS_POPUP | WS_CAPTION | WS_SYSMENU +CAPTION "FireWire About" +FONT 8, "MS Sans Serif", 0, 0, 0x0 +BEGIN + DEFPUSHBUTTON "OK",IDOK,65,85,50,14 + LTEXT "FireWire Driver",IDC_NAME,70,10,48,8 + GROUPBOX "",IDC_STATIC,5,35,170,40 + LTEXT "Author: Shadow and linuzappz",IDC_STATIC,29,19,141,10 +END + + +///////////////////////////////////////////////////////////////////////////// +// +// DESIGNINFO +// + +#ifdef APSTUDIO_INVOKED +GUIDELINES DESIGNINFO +BEGIN + IDD_CONFIG, DIALOG + BEGIN + LEFTMARGIN, 7 + RIGHTMARGIN, 205 + TOPMARGIN, 7 + BOTTOMMARGIN, 114 + END + + IDD_ABOUT, DIALOG + BEGIN + LEFTMARGIN, 7 + RIGHTMARGIN, 170 + TOPMARGIN, 7 + BOTTOMMARGIN, 99 + END +END +#endif // APSTUDIO_INVOKED + +#endif // Spanish (Argentina) resources +///////////////////////////////////////////////////////////////////////////// + + + +#ifndef APSTUDIO_INVOKED +///////////////////////////////////////////////////////////////////////////// +// +// Generated from the TEXTINCLUDE 3 resource. +// + + +///////////////////////////////////////////////////////////////////////////// +#endif // not APSTUDIO_INVOKED + diff --git a/plugins/GSnull/Windows/GSnull_vc2008.vcproj b/plugins/GSnull/Windows/GSnull_vc2008.vcproj new file mode 100644 index 0000000000..7ca94e36c7 --- /dev/null +++ b/plugins/GSnull/Windows/GSnull_vc2008.vcproj @@ -0,0 +1,201 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/plugins/GSnull/Windows/Makefile b/plugins/GSnull/Windows/Makefile new file mode 100644 index 0000000000..afa9da3337 --- /dev/null +++ b/plugins/GSnull/Windows/Makefile @@ -0,0 +1,29 @@ + +RC = windres +STRIP = strip + +PLUGIN = FWnull.dll +CFLAGS = -Wall -O2 -fomit-frame-pointer -I.. -D__WIN32__ -D__MINGW32__ +LIBS = -lcomctl32 -lwsock32 -lwinmm -lgdi32 -lcomdlg32 +RESOBJ = FireWireNull.o +OBJS = ../FW.o Config.o Win32.o ${RESOBJ} + +DEPS:= $(OBJS:.o=.d) + +all: plugin + +plugin: ${OBJS} + dllwrap --def plugin.def -o ${PLUGIN} ${OBJS} ${LIBS} +# ${CC} -shared -Wl,--kill-at,--output-def,plugin.def ${CFLAGS} ${OBJS} -o ${PLUGIN} ${LIBS} + ${STRIP} ${PLUGIN} + +.PHONY: clear plugin + +clean: + rm -f ${OBJS} ${DEPS} ${PLUGIN} + +${RESOBJ}: FireWireNull.rc + ${RC} -D__MINGW32__ -I rc -O coff -o $@ -i $< + +-include ${DEPS} + diff --git a/plugins/GSnull/Windows/ProjectRootDir.vsprops b/plugins/GSnull/Windows/ProjectRootDir.vsprops new file mode 100644 index 0000000000..b8400f4c9c --- /dev/null +++ b/plugins/GSnull/Windows/ProjectRootDir.vsprops @@ -0,0 +1,11 @@ + + + + diff --git a/plugins/GSnull/Windows/Win32.cpp b/plugins/GSnull/Windows/Win32.cpp new file mode 100644 index 0000000000..c09edf89f0 --- /dev/null +++ b/plugins/GSnull/Windows/Win32.cpp @@ -0,0 +1,81 @@ +#include +#include +#include + + +#include "resource.h" +#include "../FW.h" + +HINSTANCE hInst; + +void SysMessage(char *fmt, ...) { + va_list list; + char tmp[512]; + + va_start(list,fmt); + vsprintf(tmp,fmt,list); + va_end(list); + MessageBox(0, tmp, "FW Plugin Msg", 0); +} + +BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { + + switch(uMsg) { + case WM_INITDIALOG: + LoadConfig(); + if (conf.Log) CheckDlgButton(hW, IDC_LOGGING, TRUE); + return TRUE; + + case WM_COMMAND: + switch(LOWORD(wParam)) { + case IDCANCEL: + EndDialog(hW, TRUE); + return TRUE; + case IDOK: + if (IsDlgButtonChecked(hW, IDC_LOGGING)) + conf.Log = 1; + else conf.Log = 0; + SaveConfig(); + EndDialog(hW, FALSE); + return TRUE; + } + } + return FALSE; +} + +BOOL CALLBACK AboutDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { + switch(uMsg) { + case WM_INITDIALOG: + return TRUE; + + case WM_COMMAND: + switch(LOWORD(wParam)) { + case IDOK: + EndDialog(hW, FALSE); + return TRUE; + } + } + return FALSE; +} + +void CALLBACK FWconfigure() { + DialogBox(hInst, + MAKEINTRESOURCE(IDD_CONFIG), + GetActiveWindow(), + (DLGPROC)ConfigureDlgProc); +} + +void CALLBACK FWabout() { + DialogBox(hInst, + MAKEINTRESOURCE(IDD_ABOUT), + GetActiveWindow(), + (DLGPROC)AboutDlgProc); +} + +BOOL APIENTRY DllMain(HANDLE hModule, // DLL INIT + DWORD dwReason, + LPVOID lpReserved) { + hInst = (HINSTANCE)hModule; + return TRUE; // very quick :) +} + diff --git a/plugins/GSnull/Windows/resource.h b/plugins/GSnull/Windows/resource.h new file mode 100644 index 0000000000..6ea8fead11 --- /dev/null +++ b/plugins/GSnull/Windows/resource.h @@ -0,0 +1,21 @@ +//{{NO_DEPENDENCIES}} +// Microsoft Visual C++ generated include file. +// Used by FireWireNull.rc +// +#define IDD_CONFDLG 101 +#define IDD_CONFIG 101 +#define IDD_ABOUT 103 +#define IDC_NAME 1000 +#define IDC_CHECK1 1007 +#define IDC_LOGGING 1007 + +// Next default values for new objects +// +#ifdef APSTUDIO_INVOKED +#ifndef APSTUDIO_READONLY_SYMBOLS +#define _APS_NEXT_RESOURCE_VALUE 105 +#define _APS_NEXT_COMMAND_VALUE 40001 +#define _APS_NEXT_CONTROL_VALUE 1008 +#define _APS_NEXT_SYMED_VALUE 101 +#endif +#endif diff --git a/plugins/GSnull/build.sh b/plugins/GSnull/build.sh new file mode 100644 index 0000000000..e4118a87d9 --- /dev/null +++ b/plugins/GSnull/build.sh @@ -0,0 +1,32 @@ +#!/bin/sh + +echo --------------- +echo Building GSnull +echo --------------- + +curdir=`pwd` + + +if test "${GSnullOPTIONS+set}" != set ; then +export GSnullOPTIONS="" +fi + +if [ $# -gt 0 ] && [ $1 = "all" ] +then + +aclocal +automake -a +autoconf + +./configure ${GSnullOPTIONS} --prefix=${PCSX2PLUGINS} +make clean +make install + +else +make $@ +fi + +if [ $? -ne 0 ] +then +exit 1 +fi \ No newline at end of file diff --git a/plugins/GSnull/configure.ac b/plugins/GSnull/configure.ac new file mode 100644 index 0000000000..09f8e74eed --- /dev/null +++ b/plugins/GSnull/configure.ac @@ -0,0 +1,80 @@ +AC_INIT(GSnull, 0.1,arcum42@gmail.com) + +AM_INIT_AUTOMAKE(GSnull,0.1) + +AC_PROG_CC([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CXX([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CPP([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) + +AC_PROG_INSTALL +AC_PROG_RANLIB + +dnl necessary for compiling assembly +AM_PROG_AS + +AC_SUBST(GSnull_CURRENT, 0) +AC_SUBST(GSnull_REVISION, 1) +AC_SUBST(GSnull_AGE, 0) +AC_SUBST(GSnull_RELEASE,[$GSnull_CURRENT].[$GSnull_REVISION].[$GSnull_AGE]) +AC_SUBST(GSnull_SONAME,libGSnull.so.[$GSnull_CURRENT].[$GSnull_REVISION].[$GSnull_AGE]) + +CFLAGS= +CPPFLAGS= +CXXFLAGS= + +dnl Check for debug build +AC_MSG_CHECKING(debug build) +AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), +debug=$enableval,debug=no) +if test "x$debug" == xyes +then + AC_DEFINE(_DEBUG,1,[_DEBUG]) + CFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " +else + AC_DEFINE(NDEBUG,1,[NDEBUG]) + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " +fi +AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) +AC_MSG_RESULT($debug) + +AC_DEFINE(__LINUX__,1,[__LINUX__]) + +dnl Check for dev build +AC_MSG_CHECKING(for development build...) +AC_ARG_ENABLE(devbuild, AC_HELP_STRING([--enable-devbuild], [Special Build for developers that simplifies testing and adds extra checks]), + devbuild=$enableval,devbuild=no) +if test "x$devbuild" == xyes +then + AC_DEFINE(GSnull_DEVBUILD,1,[GSnull_DEVBUILD]) +fi +AC_MSG_RESULT($devbuild) +AM_CONDITIONAL(RELEASE_TO_PUBLIC, test x$devbuild = xno) + +AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) + +dnl gtk +AC_MSG_CHECKING(gtk2+) +AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) +LIBS+=$(pkg-config --libs gtk+-2.0) + +dnl bindir = pcsx2exe + +dnl assuming linux environment +so_ext=".so.$GSnull_RELEASE" +SHARED_LDFLAGS="-shared" +AC_SUBST(so_ext) +AC_SUBST(SHARED_LDFLAGS) + +AC_CHECK_LIB(stdc++,main,[LIBS="$LIBS -lstdc++"]) + +AC_OUTPUT([ + Makefile + ]) + +echo "Configuration:" +echo " Debug build? $debug" +echo " Dev build? $devbuild" \ No newline at end of file diff --git a/plugins/GSnull/missing b/plugins/GSnull/missing new file mode 120000 index 0000000000..1f0fe88e65 --- /dev/null +++ b/plugins/GSnull/missing @@ -0,0 +1 @@ +/usr/share/automake-1.10/missing \ No newline at end of file From 85c8e4c4afe7038fff9056ac4076e0a1f2fb67eb Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 8 May 2009 22:40:56 +0000 Subject: [PATCH 035/296] Mostly port the GSnull plugin to Windows. Currently asserts on startup in Windows about GSirq not being null. (Though it works if you hit ignore.) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1158 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 1 - plugins/GSnull/GS.h | 21 ++++++----- plugins/GSnull/Windows/Config.cpp | 8 ++-- plugins/GSnull/Windows/GS.def | 42 ++++++++++++++++----- plugins/GSnull/Windows/GSnull_vc2008.sln | 20 ++++++++++ plugins/GSnull/Windows/GSnull_vc2008.vcproj | 16 ++++---- plugins/GSnull/Windows/Win32.cpp | 8 ++-- 7 files changed, 79 insertions(+), 37 deletions(-) create mode 100644 plugins/GSnull/Windows/GSnull_vc2008.sln diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index 96b566706a..e32ad8a4e2 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -72,7 +72,6 @@ EXPORT_C_(void) GSprintf(int timeout, char *fmt, ...) { va_list list; char msg[512]; - char cmd[512]; va_start(list, fmt); vsprintf(msg, fmt, list); diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h index 5cee36c9af..56b6ff8417 100644 --- a/plugins/GSnull/GS.h +++ b/plugins/GSnull/GS.h @@ -21,16 +21,6 @@ #include -#ifdef __cplusplus -extern "C" -{ -#endif -#define GSdefs -#include "PS2Edefs.h" -#ifdef __cplusplus -} -#endif - #ifdef _WIN32 #include @@ -44,6 +34,16 @@ extern "C" #endif +#ifdef __cplusplus +extern "C" +{ +#endif +#define GSdefs +#include "PS2Edefs.h" +#ifdef __cplusplus +} +#endif + /*#ifdef _MSC_VER #define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK #else @@ -63,6 +63,7 @@ typedef struct int Log; } Config; + extern Config conf; extern FILE *gsLog; diff --git a/plugins/GSnull/Windows/Config.cpp b/plugins/GSnull/Windows/Config.cpp index 62db91dd61..593766f613 100644 --- a/plugins/GSnull/Windows/Config.cpp +++ b/plugins/GSnull/Windows/Config.cpp @@ -1,6 +1,6 @@ #include -#include "../FW.h" +#include "../GS.h" extern HINSTANCE hInst; void SaveConfig() @@ -14,7 +14,7 @@ void SaveConfig() szTemp = strrchr(szIniFile, '\\'); if(!szTemp) return; - strcpy(szTemp, "\\inis\\fwnull.ini"); + strcpy(szTemp, "\\inis\\gsnull.ini"); sprintf(szValue,"%u",Conf1->Log); WritePrivateProfileString("Interface", "Logging",szValue,szIniFile); @@ -32,8 +32,8 @@ void LoadConfig() { szTemp = strrchr(szIniFile, '\\'); if(!szTemp) return ; - strcpy(szTemp, "\\inis\\fwnull.ini"); - fp=fopen("inis\\fwnull.ini","rt");//check if firewirenull.ini really exists + strcpy(szTemp, "\\inis\\gsnull.ini"); + fp=fopen("inis\\gsnull.ini","rt");//check if gsnull.ini really exists if (!fp) { CreateDirectory("inis",NULL); diff --git a/plugins/GSnull/Windows/GS.def b/plugins/GSnull/Windows/GS.def index 51e786a1fd..d71a22d69d 100644 --- a/plugins/GSnull/Windows/GS.def +++ b/plugins/GSnull/Windows/GS.def @@ -7,13 +7,35 @@ EXPORTS PS2EgetLibType @2 PS2EgetLibName @3 PS2EgetLibVersion2 @4 - FWinit @5 - FWshutdown @6 - FWopen @7 - FWclose @8 - FWread32 @11 - FWwrite32 @14 - FWirqCallback @15 - FWconfigure @17 - FWtest @18 - FWabout @19 + GSinit @5 + GSopen @6 + GSclose @7 + GSshutdown @8 + GSvsync @9 + GSgifTransfer1 @10 + GSgifTransfer2 @11 + GSgifTransfer3 @12 + GSgetLastTag @13 + GSgifSoftReset @14 + GSreadFIFO @15 + GSreadFIFO2 @16 + GSkeyEvent @17 + GSchangeSaveState @18 + GSmakeSnapshot @19 + GSmakeSnapshot2 @20 + GSirqCallback @21 + GSprintf @22 + GSsetBaseMem @23 + GSsetGameCRC @24 + GSsetFrameSkip @25 + GSsetupRecording @26 + GSreset @27 + GSwriteCSR @28 + GSgetDriverInfo @29 + GSsetWindowInfo @30 + GSfreeze @31 + GSconfigure @32 + GSabout @33 + GStest @34 + + diff --git a/plugins/GSnull/Windows/GSnull_vc2008.sln b/plugins/GSnull/Windows/GSnull_vc2008.sln new file mode 100644 index 0000000000..cbdf517657 --- /dev/null +++ b/plugins/GSnull/Windows/GSnull_vc2008.sln @@ -0,0 +1,20 @@ + +Microsoft Visual Studio Solution File, Format Version 10.00 +# Visual C++ Express 2008 +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "GSnull", "GSnull_vc2008.vcproj", "{3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}" +EndProject +Global + GlobalSection(SolutionConfigurationPlatforms) = preSolution + Debug|Win32 = Debug|Win32 + Release|Win32 = Release|Win32 + EndGlobalSection + GlobalSection(ProjectConfigurationPlatforms) = postSolution + {3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}.Debug|Win32.ActiveCfg = Debug|Win32 + {3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}.Debug|Win32.Build.0 = Debug|Win32 + {3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}.Release|Win32.ActiveCfg = Release|Win32 + {3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}.Release|Win32.Build.0 = Release|Win32 + EndGlobalSection + GlobalSection(SolutionProperties) = preSolution + HideSolutionNode = FALSE + EndGlobalSection +EndGlobal diff --git a/plugins/GSnull/Windows/GSnull_vc2008.vcproj b/plugins/GSnull/Windows/GSnull_vc2008.vcproj index 7ca94e36c7..e08a6c645c 100644 --- a/plugins/GSnull/Windows/GSnull_vc2008.vcproj +++ b/plugins/GSnull/Windows/GSnull_vc2008.vcproj @@ -2,9 +2,9 @@ @@ -55,7 +55,7 @@ /> @@ -122,7 +122,7 @@ /> @@ -161,7 +161,7 @@ > diff --git a/plugins/GSnull/Windows/Win32.cpp b/plugins/GSnull/Windows/Win32.cpp index c09edf89f0..0071924a65 100644 --- a/plugins/GSnull/Windows/Win32.cpp +++ b/plugins/GSnull/Windows/Win32.cpp @@ -4,7 +4,7 @@ #include "resource.h" -#include "../FW.h" +#include "../GS.h" HINSTANCE hInst; @@ -15,7 +15,7 @@ void SysMessage(char *fmt, ...) { va_start(list,fmt); vsprintf(tmp,fmt,list); va_end(list); - MessageBox(0, tmp, "FW Plugin Msg", 0); + MessageBox(0, tmp, "GS Plugin Msg", 0); } BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { @@ -58,14 +58,14 @@ BOOL CALLBACK AboutDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { return FALSE; } -void CALLBACK FWconfigure() { +EXPORT_C_(void) GSconfigure() { DialogBox(hInst, MAKEINTRESOURCE(IDD_CONFIG), GetActiveWindow(), (DLGPROC)ConfigureDlgProc); } -void CALLBACK FWabout() { +EXPORT_C_(void) GSabout() { DialogBox(hInst, MAKEINTRESOURCE(IDD_ABOUT), GetActiveWindow(), From f4c3560c754c51f26ca9ae98f9508a49d8cf26ab Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 8 May 2009 23:05:43 +0000 Subject: [PATCH 036/296] microVU: - mac flags now stored memory instead of in upper 16 bits of GPRs. - minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1159 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 5 +++-- pcsx2/x86/microVU_Alloc.inl | 16 +++++++------- pcsx2/x86/microVU_Compile.inl | 9 +++----- pcsx2/x86/microVU_Execute.inl | 21 +++++++++++-------- pcsx2/x86/microVU_Flags.inl | 39 ++++++++++------------------------- pcsx2/x86/microVU_Lower.inl | 2 +- pcsx2/x86/microVU_Misc.h | 8 +++---- pcsx2/x86/microVU_Misc.inl | 24 ++++++++++----------- pcsx2/x86/microVU_Upper.inl | 7 ++++--- 9 files changed, 57 insertions(+), 74 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 1a3610136d..7a6c1d3809 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -102,7 +102,6 @@ struct microVU { u8* exitFunct; // Ptr Function to the Exit code for recompiled programs u32 code; // Contains the current Instruction u32 iReg; // iReg (only used in recompilation, not execution) - u32 clipFlag[4]; // 4 instances of clip flag (used in execution) u32 divFlag; // 1 instance of I/D flags u32 VIbackup[2]; // Holds a backup of a VI reg if modified before a branch u32 branch; // Holds branch compare result (IBxx) OR Holds address to Jump to (JALR/JR) @@ -112,7 +111,9 @@ struct microVU { u32 totalCycles; u32 cycles; - PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ + PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) + PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) + PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ }; // microVU rec structs diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index cc177f8c4b..b33fdb7bb0 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -679,27 +679,25 @@ microVUt(void) mVUallocFMAC26b(int& ACCw, int& ACCr) { } microVUt(void) mVUallocSFLAGa(int reg, int fInstance) { - microVU* mVU = mVUx; getFlagReg(fInstance, fInstance); MOVZX32R16toR(reg, fInstance); } microVUt(void) mVUallocSFLAGb(int reg, int fInstance) { getFlagReg(fInstance, fInstance); - MOV16RtoR(fInstance, reg); + //AND32ItoR(reg, 0xffff); + MOV32RtoR(fInstance, reg); } microVUt(void) mVUallocMFLAGa(int reg, int fInstance) { - getFlagReg(fInstance, fInstance); - MOV32RtoR(reg, fInstance); - SHR32ItoR(reg, 16); + microVU* mVU = mVUx; + MOVZX32M16toR(reg, (uptr)&mVU->macFlag[fInstance]); } microVUt(void) mVUallocMFLAGb(int reg, int fInstance) { - getFlagReg(fInstance, fInstance); - AND32ItoR(fInstance, 0xffff); - SHL32ItoR(reg, 16); - OR32RtoR(fInstance, reg); + microVU* mVU = mVUx; + //AND32ItoR(reg, 0xffff); + MOV32RtoM((uptr)&mVU->macFlag[fInstance], reg); } microVUt(void) mVUallocCFLAGa(int reg, int fInstance) { diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 0cdfd8e36d..e11280b892 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -120,12 +120,9 @@ microVUt(void) mVUendProgram(int fStatus, int fMac) { // Save Flag Instances getFlagReg(fStatus, fStatus); - getFlagReg(fMac, fMac); - MOV32RtoR(gprT1, fStatus); - AND32ItoR(gprT1, 0xffff); - SHR32ItoR(fMac, 16); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); - MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, fMac); + mVUallocMFLAGa(gprT1, fMac); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index aebc97e94a..edfd6130a2 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -48,17 +48,14 @@ microVUt(void) mVUdispatcherA() { // Load Regs MOV32MtoR(gprR, (uptr)&mVU->regs->VI[REG_R].UL); MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); - MOV32MtoR(gprF1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); - SHL32ItoR(gprF0, 16); - AND32ItoR(gprF1, 0xffff); - OR32RtoR (gprF0, gprF1); + AND32ItoR(gprF0, 0xffff); MOV32RtoR(gprF1, gprF0); MOV32RtoR(gprF2, gprF0); MOV32RtoR(gprF3, gprF0); - for (int i = 1; i < 16; i++) { - if (isMMX(i)) { MOVQMtoR(mmVI(i), (uptr)&mVU->regs->VI[i].UL); } - } + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); + SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, xmmT1); SSE_MOVAPS_M128_to_XMM(xmmACC, (uptr)&mVU->regs->ACC.UL[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); @@ -67,10 +64,14 @@ microVUt(void) mVUdispatcherA() { SSE_MOVAPS_M128_to_XMM(xmmPQ, (uptr)&mVU->regs->VI[REG_Q].UL); SSE_SHUFPS_XMM_to_XMM(xmmPQ, xmmT1, 0); // wzyx = PPQQ + for (int i = 1; i < 16; i++) { + if (isMMX(i)) { MOVQMtoR(mmVI(i), (uptr)&mVU->regs->VI[i].UL); } + } + //PUSH32R(EAX); //CALLFunc((uptr)testFunction); - //POP32R(EAX); //write8(0xcc); + //POP32R(EAX); // Jump to Recompiled Code Block JMPR(EAX); @@ -140,7 +141,9 @@ microVUt(void) mVUcleanUp() { //mVUprint("microVU: VI0 = %x", params mVU->regs->VI[0].UL); mVUcurProg.x86ptr = x86Ptr; mVUcacheCheck(x86Ptr, mVUcurProg.x86start, (uptr)(mVUcurProg.x86end - mVUcurProg.x86start)); - mVU->regs->cycle += mVU->totalCycles - mVU->cycles; + mVU->cycles = mVU->totalCycles - mVU->cycles; + mVU->regs->cycle += mVU->cycles; + cpuRegs.cycle += ((mVU->cycles < 3000) ? mVU->cycles : 3000) * Config.Hacks.VUCycleSteal; } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index b42f9cb2c3..b359f37fc5 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -39,7 +39,7 @@ microVUt(void) mVUdivSet() { int flagReg1, flagReg2; if (doDivFlag) { getFlagReg(flagReg1, fsInstance); - if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } + if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } AND16ItoR(flagReg1, 0x0fcf); OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } @@ -134,42 +134,25 @@ microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { } #define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) -#define getFlagReg2(x) ((x == bStatus[3]) ? gprESP : ((x == bStatus[2]) ? gprR : ((x == bStatus[1]) ? gprT2 : gprT1))) +#define shuffleMac ((bMac[3]<<6)|(bMac[2]<<4)|(bMac[1]<<2)|bMac[0]) // Recompiles Code for Proper Flags on Block Linkings microVUt(void) mVUsetupFlags(int* bStatus, int* bMac) { microVU* mVU = mVUx; - PUSH32R(gprR); // Backup gprR - MOV32RtoM((uptr)&mVU->espBackup, gprESP); - + PUSH32R(gprR); // Backup gprR MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); MOV32RtoR(gprR, getFlagReg1(bStatus[2])); - MOV32RtoR(gprESP, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, gprT1); + MOV32RtoR(gprF1, gprT2); + MOV32RtoR(gprF2, gprR); + POP32R(gprR); // Restore gprR - MOV32RtoR(gprF0, gprT1); - MOV32RtoR(gprF1, gprT2); - MOV32RtoR(gprF2, gprR); - MOV32RtoR(gprF3, gprESP); - - AND32ItoR(gprT1, 0xffff0000); - AND32ItoR(gprT2, 0xffff0000); - AND32ItoR(gprR, 0xffff0000); - AND32ItoR(gprESP, 0xffff0000); - - AND32ItoR(gprF0, 0x0000ffff); - AND32ItoR(gprF1, 0x0000ffff); - AND32ItoR(gprF2, 0x0000ffff); - AND32ItoR(gprF3, 0x0000ffff); - - OR32RtoR(gprF0, getFlagReg2(bMac[0])); - OR32RtoR(gprF1, getFlagReg2(bMac[1])); - OR32RtoR(gprF2, getFlagReg2(bMac[2])); - OR32RtoR(gprF3, getFlagReg2(bMac[3])); - - MOV32MtoR(gprESP, (uptr)&mVU->espBackup); - POP32R(gprR); // Restore gprR + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->macFlag); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleMac); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); } #endif //PCSX2_MICROVU \ No newline at end of file diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index a995bb9ddd..781c3a5193 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -576,7 +576,7 @@ microVUf(void) mVU_FSSET() { pass2 { int flagReg1, flagReg2; getFlagReg(flagReg1, fsInstance); - if (!(doStatus||doDivFlag)) { getFlagReg(flagReg2, fpsInstance); MOV16RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction + if (!(doStatus||doDivFlag)) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction AND16ItoR(flagReg1, 0x03f); // Remember not to modify upper 16 bits because of mac flag OR16ItoR (flagReg1, (_Imm12_ & 0xfc0)); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index e28852b638..55418ae81c 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -129,11 +129,11 @@ declareAllVariables #define gprT1 0 // Temp Reg #define gprT2 1 // Temp Reg #define gprR 2 // R Reg -#define gprF0 3 // MAC Flag::Status Flag 0 +#define gprF0 3 // Status Flag 0 #define gprESP 4 // Don't use? -#define gprF1 5 // MAC Flag::Status Flag 1 -#define gprF2 6 // MAC Flag::Status Flag 2 -#define gprF3 7 // MAC Flag::Status Flag 3 +#define gprF1 5 // Status Flag 1 +#define gprF2 6 // Status Flag 2 +#define gprF3 7 // Status Flag 3 // Template Stuff #define mVUx (vuIndex ? µVU1 : µVU0) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index cdd08254e9..bae60ade93 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -105,20 +105,20 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { return;*/ switch ( xyzw ) { - case 5: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xe1); //WZXY + case 5: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xe1); //WZXY SSE_MOVSS_XMM_to_M32(offset+4, reg); - SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // YW - case 6: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xc9); + case 6: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc9); SSE_MOVLPS_XMM_to_M64(offset+4, reg); break; // YZ - case 7: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x93); //ZYXW + case 7: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x93); //ZYXW SSE_MOVHPS_XMM_to_M64(offset+4, reg); SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // YZW case 9: SSE_MOVSS_XMM_to_M32(offset, reg); - SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // XW case 10: SSE_MOVSS_XMM_to_M32(offset, reg); @@ -128,7 +128,7 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { case 11: SSE_MOVSS_XMM_to_M32(offset, reg); SSE_MOVHPS_XMM_to_M64(offset+8, reg); break; //XZW - case 13: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x4b); //YXZW + case 13: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x4b); //YXZW SSE_MOVHPS_XMM_to_M64(offset, reg); SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // XYW @@ -163,20 +163,20 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { return;*/ switch ( xyzw ) { - case 5: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xe1); //WZXY + case 5: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xe1); //WZXY SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+4); - SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // YW - case 6: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xc9); + case 6: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc9); SSE_MOVLPS_XMM_to_Rm(gprReg, reg, offset+4); break; // YZ - case 7: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x93); //ZYXW + case 7: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x93); //ZYXW SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset+4); SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // YZW case 9: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); - SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // XW case 10: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); @@ -186,7 +186,7 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { case 11: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset+8); break; //XZW - case 13: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x4b); //YXZW + case 13: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x4b); //YXZW SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset); SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // XYW diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index f22a08cf8f..a6fe9f0f3b 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -34,13 +34,14 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; + //SysPrintf("doStatus = %d; doMac = %d\n", doStatus>>9, doMac>>8); if (!doFlags) return; if (!doMac) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw if (doStatus) { getFlagReg(sReg, fsInstance); // Set sReg to valid GPR by Cur Flag Instance mVUallocSFLAGa(sReg, fpsInstance); // Get Prev Status Flag - AND16ItoR(sReg, 0xff0); // Keep Sticky and D/I flags + AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags } //-------------------------Check for Signed flags------------------------------ @@ -53,7 +54,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg - AND16ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation + AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation pjmp = JZ8(0); // Skip if none are if (doMac) SHL16ItoR(mReg, 4 + ADD_XYZW); if (doStatus) OR16ItoR(sReg, 0x82); // SS, S flags @@ -62,7 +63,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX //-------------------------Check for Zero flags------------------------------ - AND16ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation + AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation pjmp = JZ8(0); // Skip if none are if (doMac) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } if (doStatus) { OR16ItoR(sReg, 0x41); } // ZS, Z flags From e10ca2ba49950c892a25a61584546dd0143216d6 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 9 May 2009 00:03:27 +0000 Subject: [PATCH 037/296] All right, this version of GSnull doesn't assert, and if you finagle enough with LilyPad's settings, you can get the keyboard to work (but not a gamepad). Occasionally causes the pad plugin to close on starting emulation. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1160 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 9 ++++++--- plugins/GSnull/GS.h | 1 - 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index e32ad8a4e2..d3cc3bbe55 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -31,6 +31,9 @@ using namespace std; Display *display; int screen; #endif +#ifdef _WIN32 +HINSTANCE HInst; +#endif const unsigned char version = PS2E_GS_VERSION; const unsigned char revision = 0; @@ -125,7 +128,7 @@ EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) #ifdef GS_LOG GS_LOG("GS open\n"); #endif - assert( GSirq != NULL ); + //assert( GSirq != NULL ); #ifdef __LINUX__ display = XOpenDisplay(0); @@ -133,8 +136,8 @@ EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) if( pDsp != NULL ) *(Display**)pDsp = display; #else - // Haven't implemented anything for Windows yet. - pDsp = NULL; + if( pDsp != NULL ) *(int*)pDsp = (int)GetNextWindow(GetForegroundWindow(),GW_HWNDNEXT); +(NULL); #endif SysPrintf("Opening GSnull\n"); return 0; diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h index 56b6ff8417..9cd5c0287f 100644 --- a/plugins/GSnull/GS.h +++ b/plugins/GSnull/GS.h @@ -63,7 +63,6 @@ typedef struct int Log; } Config; - extern Config conf; extern FILE *gsLog; From 6ea395be4b7bfe3a20acf4db5d608ada586b643c Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 9 May 2009 08:37:02 +0000 Subject: [PATCH 038/296] GSdx: optimizations here and there git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1161 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GS.h | 32 +++++- plugins/GSdx/GSRendererSW.h | 155 ++++++++++++++++++------------ plugins/GSdx/GSState.cpp | 109 +++++++++------------ plugins/GSdx/GSState.h | 1 - plugins/GSdx/GSTextureCacheSW.cpp | 98 +++++++++++-------- plugins/GSdx/GSVertexSW.h | 8 +- 6 files changed, 238 insertions(+), 165 deletions(-) diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index f670531aa4..43585a0aab 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -510,7 +510,9 @@ REG64_(GIFReg, ALPHA) UINT32 FIX:8; UINT32 _PAD2:24; REG_END2 - __forceinline bool IsOpaque() const {return (A == B || C == 2 && FIX == 0) && D == 0 || (A == 0 && B == 2 && C == 2 && D == 2 && FIX == 0x80);} // output will be Cs/As + // opaque => output will be Cs/As + __forceinline bool IsOpaque() const {return (A == B || C == 2 && FIX == 0) && D == 0 || (A == 0 && B == D && C == 2 && FIX == 0x80);} + __forceinline bool IsOpaque(int amin, int amax) const {return (A == B || amax == 0) && D == 0 || A == 0 && B == D && amin == 0x80 && amax == 0x80;} REG_END2 REG64_(GIFReg, BITBLTBUF) @@ -1061,21 +1063,41 @@ REG_SET_END __declspec(align(16)) struct GIFPath { GIFTag tag; + UINT32 reg; UINT32 nreg; - UINT32 _pad[3]; + UINT32 nloop; + UINT32 adonly; GSVector4i regs; void SetTag(const void* mem) { GSVector4i v = GSVector4i::load(mem); GSVector4i::store(&tag, v); - nreg = 0; + reg = 0; regs = v.uph8(v >> 4) & 0x0f0f0f0f; + nreg = tag.NREG; + nloop = tag.NLOOP; + adonly = nreg == 1 && regs.u8[0] == GIF_REG_A_D; } - DWORD GetReg() + __forceinline DWORD GetReg() { - return regs.u8[nreg]; // (DWORD)GET_GIF_REG(tag, nreg); + return regs.u8[reg]; // (DWORD)GET_GIF_REG(tag, nreg); + } + + __forceinline bool StepReg() + { + if((++reg & 0xf) == nreg) + { + reg = 0; + + if(--nloop == 0) + { + return false; + } + } + + return true; } }; diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index c0d2cb0889..2f540cacc6 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -117,9 +117,74 @@ protected: return true; } + void GetAlphaMinMax() + { + if(m_vtrace.m_alpha.valid) + { + return; + } + + const GSDrawingEnvironment& env = m_env; + const GSDrawingContext* context = m_context; + + GSVector4i a = GSVector4i(m_vtrace.m_min.c.wwww(m_vtrace.m_max.c)) >> 7; + + if(PRIM->TME && context->TEX0.TCC) + { + DWORD bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; + DWORD cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; + DWORD pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; + + if(bpp == 32) + { + a.y = 0; + a.w = 0xff; + } + else if(bpp == 24) + { + a.y = env.TEXA.AEM ? 0 : env.TEXA.TA0; + a.w = env.TEXA.TA0; + } + else if(bpp == 16) + { + a.y = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); + a.w = max(env.TEXA.TA0, env.TEXA.TA1); + } + else + { + m_mem.m_clut.GetAlphaMinMax32(a.y, a.w); + } + + switch(context->TEX0.TFX) + { + case TFX_MODULATE: + a.x = (a.x * a.y) >> 7; + a.z = (a.z * a.w) >> 7; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_DECAL: + break; + case TFX_HIGHLIGHT: + a.x = a.x + a.y; + a.z = a.z + a.w; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_HIGHLIGHT2: + break; + default: + __assume(0); + } + } + + m_vtrace.m_alpha.min = a.x; + m_vtrace.m_alpha.max = a.z; + m_vtrace.m_alpha.valid = true; + } + bool TryAlphaTest(DWORD& fm, DWORD& zm) { - const GSDrawingEnvironment& env = m_env; const GSDrawingContext* context = m_context; bool pass = true; @@ -130,62 +195,10 @@ protected: } else if(context->TEST.ATST != ATST_ALWAYS) { - GSVector4i af = GSVector4i(m_vtrace.m_min.c.wwww(m_vtrace.m_max.c)) >> 7; + GetAlphaMinMax(); - int amin, amax; - - if(PRIM->TME && context->TEX0.TCC) - { - DWORD bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; - DWORD cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; - DWORD pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; - - if(bpp == 32) - { - return false; - } - else if(bpp == 24) - { - amin = env.TEXA.AEM ? 0 : env.TEXA.TA0; - amax = env.TEXA.TA0; - } - else if(bpp == 16) - { - amin = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); - amax = max(env.TEXA.TA0, env.TEXA.TA1); - } - else - { - m_mem.m_clut.GetAlphaMinMax32(amin, amax); - } - - switch(context->TEX0.TFX) - { - case TFX_MODULATE: - amin = (amin * af.x) >> 7; - amax = (amax * af.z) >> 7; - if(amin > 255) amin = 255; - if(amax > 255) amax = 255; - break; - case TFX_DECAL: - break; - case TFX_HIGHLIGHT: - amin = amin + af.x; - amax = amax + af.z; - if(amin > 255) amin = 255; - if(amax > 255) amax = 255; - break; - case TFX_HIGHLIGHT2: - break; - default: - __assume(0); - } - } - else - { - amin = af.x; - amax = af.z; - } + int amin = m_vtrace.m_alpha.min; + int amax = m_vtrace.m_alpha.max; int aref = context->TEST.AREF; @@ -252,7 +265,7 @@ protected: const GSDrawingEnvironment& env = m_env; const GSDrawingContext* context = m_context; - p.vm = m_mem.m_vm32; + p.vm = m_mem.m_vm8; p.fbo = m_mem.GetOffset(context->FRAME.Block(), context->FRAME.FBW, context->FRAME.PSM); p.zbo = m_mem.GetOffset(context->ZBUF.Block(), context->FRAME.FBW, context->ZBUF.PSM); @@ -446,7 +459,31 @@ protected: p.sel.datm = context->TEST.DATM; } - if(PRIM->ABE && !context->ALPHA.IsOpaque() || PRIM->AA1) + int amin = 0, amax = 0xff; + + if(PRIM->ABE && context->ALPHA.A != context->ALPHA.B && !PRIM->AA1) + { + if(context->ALPHA.C == 0) + { + GetAlphaMinMax(); + + amin = m_vtrace.m_alpha.min; + amax = m_vtrace.m_alpha.max; + } + else if(context->ALPHA.C == 1) + { + if(p.sel.fpsm == 1) + { + amin = amax = 0x80; + } + } + else if(context->ALPHA.C == 1) + { + amin = amax = context->ALPHA.FIX; + } + } + + if(PRIM->ABE && !context->ALPHA.IsOpaque(amin, amax) || PRIM->AA1) { p.sel.abe = PRIM->ABE; p.sel.ababcd = context->ALPHA.ai32[0]; @@ -581,7 +618,7 @@ protected: if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512);} } - if(0)//stats.ticks > 1000000) + if(0)//stats.ticks > 5000000) { printf("* [%I64d | %012I64x] ticks %I64d prims %d (%d) pixels %d (%d)\n", m_perfmon.GetFrame(), p.sel.key, diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index d6a9a07a30..81bba7ec3d 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -80,7 +80,7 @@ GSState::GSState(BYTE* base, bool mt, void (*irq)()) m_sssize += sizeof(m_tr.x); m_sssize += sizeof(m_tr.y); m_sssize += m_mem.m_vmsize; - m_sssize += (sizeof(m_path[0].tag) + sizeof(m_path[0].nreg)) * 3; + m_sssize += (sizeof(m_path[0].tag) + sizeof(m_path[0].reg)) * 3; m_sssize += sizeof(m_q); ASSERT(base); @@ -453,14 +453,6 @@ void GSState::GIFPackedRegHandlerA_D(GIFPackedReg* r) (this->*m_fpGIFRegHandlers[(BYTE)r->A_D.ADDR])(&r->r); } -void GSState::GIFPackedRegHandlerA_D(GIFPackedReg* r, int size) -{ - for(int i = 0; i < size; i++) - { - (this->*m_fpGIFRegHandlers[(BYTE)r[i].A_D.ADDR])(&r[i].r); - } -} - void GSState::GIFPackedRegHandlerNOP(GIFPackedReg* r) { } @@ -1181,7 +1173,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) while(size > 0) { - if(path.tag.NLOOP == 0) + if(path.nloop == 0) { path.SetTag(mem); @@ -1193,20 +1185,15 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) m_path3hack = 1; } - if(path.tag.NLOOP > 0) // eeuser 7.2.2. GIFtag: "... when NLOOP is 0, the GIF does not output anything, and values other than the EOP field are disregarded." + if(path.nloop > 0) // eeuser 7.2.2. GIFtag: "... when NLOOP is 0, the GIF does not output anything, and values other than the EOP field are disregarded." { m_q = 1.0f; - if(path.tag.PRE) + if(path.tag.PRE && (path.tag.FLG & 2) == 0) { - ASSERT(path.tag.FLG != GIF_FLG_IMAGE); // kingdom hearts, ffxii, tales of abyss, berserk - - if((path.tag.FLG & 2) == 0) - { - GIFReg r; - r.i64 = path.tag.PRIM; - (this->*m_fpGIFRegHandlers[GIF_A_D_REG_PRIM])(&r); - } + GIFReg r; + r.i64 = path.tag.PRIM; + (this->*m_fpGIFRegHandlers[GIF_A_D_REG_PRIM])(&r); } } } @@ -1218,37 +1205,44 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) // first try a shortcut for a very common case - if(path.nreg == 0 && path.tag.NREG == 1 && size >= path.tag.NLOOP && path.GetReg() == GIF_REG_A_D) + if(path.adonly && size >= path.nloop) { - int n = path.tag.NLOOP; + size -= path.nloop; - GIFPackedRegHandlerA_D((GIFPackedReg*)mem, n); + do + { + (this->*m_fpGIFRegHandlers[(BYTE)((GIFPackedReg*)mem)->A_D.ADDR])(&((GIFPackedReg*)mem)->r); - mem += n * sizeof(GIFPackedReg); - size -= n; - - path.tag.NLOOP = 0; + mem += sizeof(GIFPackedReg); + } + while(--path.nloop > 0); } else { - while(size > 0) + do { - (this->*m_fpGIFPackedRegHandlers[path.GetReg()])((GIFPackedReg*)mem); + DWORD reg = path.GetReg(); - size--; - mem += sizeof(GIFPackedReg); - - if((++path.nreg & 0xf) == path.tag.NREG) + switch(reg) { - path.nreg = 0; - path.tag.NLOOP--; - - if(path.tag.NLOOP == 0) - { - break; - } + case GIF_REG_RGBA: + GIFPackedRegHandlerRGBA((GIFPackedReg*)mem); + break; + case GIF_REG_STQ: + GIFPackedRegHandlerSTQ((GIFPackedReg*)mem); + break; + case GIF_REG_UV: + GIFPackedRegHandlerUV((GIFPackedReg*)mem); + break; + default: + (this->*m_fpGIFPackedRegHandlers[reg])((GIFPackedReg*)mem); + break; } + + mem += sizeof(GIFPackedReg); + size--; } + while(path.StepReg() && size > 0); } break; @@ -1257,24 +1251,14 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) size *= 2; - while(size > 0) + do { (this->*m_fpGIFRegHandlers[path.GetReg()])((GIFReg*)mem); - size--; mem += sizeof(GIFReg); - - if((++path.nreg & 0xf) == path.tag.NREG) - { - path.nreg = 0; - path.tag.NLOOP--; - - if(path.tag.NLOOP == 0) - { - break; - } - } + size--; } + while(path.StepReg() && size > 0); if(size & 1) mem += sizeof(GIFReg); @@ -1286,13 +1270,13 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) ASSERT(0); - path.tag.NLOOP = 0; + path.nloop = 0; break; case GIF_FLG_IMAGE: { - int len = (int)min(size, path.tag.NLOOP); + int len = (int)min(size, path.nloop); //ASSERT(!(len&3)); @@ -1315,7 +1299,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) } mem += len * 16; - path.tag.NLOOP -= len; + path.nloop -= len; size -= len; } @@ -1328,7 +1312,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) if(index == 0) { - if(path.tag.EOP && path.tag.NLOOP == 0) + if(path.tag.EOP && path.nloop == 0) { break; } @@ -1342,13 +1326,13 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) if(index == 0) { - if(size == 0 && path.tag.NLOOP > 0) + if(size == 0 && path.nloop > 0) { if(m_mt) { // TODO - path.tag.NLOOP = 0; + path.nloop = 0; } else { @@ -1433,8 +1417,11 @@ int GSState::Freeze(GSFreezeData* fd, bool sizeonly) for(int i = 0; i < 3; i++) { + m_path[i].tag.NREG = m_path[i].nreg; + m_path[i].tag.NLOOP = m_path[i].nloop; + WriteState(data, &m_path[i].tag); - WriteState(data, &m_path[i].nreg); + WriteState(data, &m_path[i].reg); } WriteState(data, &m_q); @@ -1525,7 +1512,7 @@ int GSState::Defrost(const GSFreezeData* fd) for(int i = 0; i < 3; i++) { ReadState(&m_path[i].tag, data); - ReadState(&m_path[i].nreg, data); + ReadState(&m_path[i].reg, data); m_path[i].SetTag(&m_path[i].tag); // expand regs } diff --git a/plugins/GSdx/GSState.h b/plugins/GSdx/GSState.h index 5aeb4fe202..d22a814890 100644 --- a/plugins/GSdx/GSState.h +++ b/plugins/GSdx/GSState.h @@ -55,7 +55,6 @@ class GSState : public GSAlignedClass<16> void GIFPackedRegHandlerXYZF3(GIFPackedReg* r); void GIFPackedRegHandlerXYZ3(GIFPackedReg* r); void GIFPackedRegHandlerA_D(GIFPackedReg* r); - void GIFPackedRegHandlerA_D(GIFPackedReg* r, int size); void GIFPackedRegHandlerNOP(GIFPackedReg* r); typedef void (GSState::*GIFRegHandler)(GIFReg* r); diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index 0344e0c2a7..9342c20dda 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -283,61 +283,83 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX DWORD blocks = 0; - for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) + if(tw <= (bw << 6)) { - DWORD base = psm.bn(0, y, bp, bw); - - for(int x = r.left; x < r.right; x += s.cx) + for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) { - DWORD block = base + psm.blockOffset[x >> 3]; + DWORD base = psm.bn(0, y, bp, bw); - if(block >= MAX_BLOCKS) + for(int x = r.left; x < r.right; x += s.cx) { - continue; + DWORD block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + DWORD row = block >> 5; + DWORD col = 1 << (block & 31); + + if((m_valid[row] & col) == 0) + { + m_valid[row] |= col; + + (mem.*rtxb)(block, &dst[x * bytes], pitch, TEXA); + + blocks++; + } + } } + } + } + else + { + // unfortunatelly a block may be part of the same texture multiple times at different places (tw 1024 > tbw 640, between 640 -> 1024 it is repeated from the next row), + // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side + + // TODO: still bogus if those repeated parts aren't fetched together - DWORD row = block >> 5; - DWORD col = 1 << (block & 31); + for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) + { + DWORD base = psm.bn(0, y, bp, bw); - if(m_valid[row] & col) + for(int x = r.left; x < r.right; x += s.cx) { - continue; + DWORD block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + DWORD row = block >> 5; + DWORD col = 1 << (block & 31); + + if((m_valid[row] & col) == 0) + { + (mem.*rtxb)(block, &dst[x * bytes], pitch, TEXA); + + blocks++; + } + } } + } - // unfortunatelly a block may be part of the same texture multiple times at different places (when (1 << tw) > (tbw << 6), ex. 1024 > 640), - // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side - - // TODO: still bogus if those repeated parts aren't fetched together + for(int y = r.top; y < r.bottom; y += s.cy) + { + DWORD base = psm.bn(0, y, bp, bw); - // m_valid[row] |= col; + for(int x = r.left; x < r.right; x += s.cx) + { + DWORD block = base + psm.blockOffset[x >> 3]; - (mem.*rtxb)(block, &dst[x * bytes], pitch, TEXA); + if(block < MAX_BLOCKS) + { + DWORD row = block >> 5; + DWORD col = 1 << (block & 31); - blocks++; + m_valid[row] |= col; + } + } } } m_state->m_perfmon.Put(GSPerfMon::Unswizzle, s.cx * s.cy * bytes * blocks); - for(int y = r.top; y < r.bottom; y += s.cy) - { - DWORD base = psm.bn(0, y, bp, bw); - - for(int x = r.left; x < r.right; x += s.cx) - { - DWORD block = base + psm.blockOffset[x >> 3]; - - if(block >= MAX_BLOCKS) - { - continue; - } - - DWORD row = block >> 5; - DWORD col = 1 << (block & 31); - - m_valid[row] |= col; - } - } - return true; } diff --git a/plugins/GSdx/GSVertexSW.h b/plugins/GSdx/GSVertexSW.h index ac21c0bd6d..146aa1531b 100644 --- a/plugins/GSdx/GSVertexSW.h +++ b/plugins/GSdx/GSVertexSW.h @@ -237,6 +237,8 @@ __declspec(align(16)) class GSVertexTrace public: GSVertexSW m_min, m_max; + struct {int min, max; bool valid;} m_alpha; // source alpha range after tfx + union { DWORD value; @@ -256,9 +258,10 @@ public: m_map.Lookup(key)(v, count, m_min, m_max); m_eq.value = (m_min.p == m_max.p).mask() | ((m_min.t == m_max.t).mask() << 4) | ((m_min.c == m_max.c).mask() << 8); + + m_alpha.valid = false; } /* -*/ void Update(const GSVertexSW* v, int count) { GSVertexSW min, max; @@ -284,5 +287,8 @@ public: m_max = max; m_eq.value = (min.p == max.p).mask() | ((min.t == max.t).mask() << 4) | ((min.c == max.c).mask() << 8); + + m_alpha.valid = false; } +*/ }; From 0b23b770513f7218d71acf69c020b2c76b8995a9 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 9 May 2009 20:06:57 +0000 Subject: [PATCH 039/296] Zeropad: Continuing to do some cleanup. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1162 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/joystick.cpp | 6 +- plugins/zeropad/Linux/joystick.h | 42 +++--- plugins/zeropad/Linux/linux.cpp | 142 ++++++-------------- plugins/zeropad/Makefile.am | 3 +- plugins/zeropad/Windows/ZeroPAD_2008.vcproj | 8 ++ plugins/zeropad/configure.ac | 8 +- plugins/zeropad/zeropad.cpp | 49 +++---- plugins/zeropad/zeropad.h | 17 ++- 8 files changed, 114 insertions(+), 161 deletions(-) diff --git a/plugins/zeropad/Linux/joystick.cpp b/plugins/zeropad/Linux/joystick.cpp index a4af64108d..514de20e5a 100644 --- a/plugins/zeropad/Linux/joystick.cpp +++ b/plugins/zeropad/Linux/joystick.cpp @@ -108,7 +108,6 @@ bool JoystickInfo::Init(int id, bool bStartThread) numaxes = SDL_JoystickNumAxes(joy); numbuttons = SDL_JoystickNumButtons(joy); numpov = SDL_JoystickNumHats(joy); - numhats = SDL_JoystickNumHats(joy); devname = SDL_JoystickName(id); vbutstate.resize(numbuttons); vaxisstate.resize(numbuttons); @@ -144,9 +143,8 @@ void JoystickInfo::SaveState() vbutstate[i] = SDL_JoystickGetButton(joy, i); for (int i = 0; i < numaxes; ++i) vaxisstate[i] = SDL_JoystickGetAxis(joy, i); - - //for (int i = 0; i < numhats; ++i) - // vhatstate[i] = SDL_JoystickGetHat(joy, i); + /*for (int i = 0; i < numpov; ++i) + vpovstate[i] = SDL_JoystickGetHat(joy, i);*/ } void JoystickInfo::TestForce() diff --git a/plugins/zeropad/Linux/joystick.h b/plugins/zeropad/Linux/joystick.h index b7e53ac37b..fd4d2266d8 100644 --- a/plugins/zeropad/Linux/joystick.h +++ b/plugins/zeropad/Linux/joystick.h @@ -42,60 +42,71 @@ class JoystickInfo { return devname; } + int GetNumButtons() { return numbuttons; } + int GetNumAxes() { return numaxes; } + int GetNumPOV() { return numpov; } - /*int GetNumHats() - { - return numhats; - }*/ + int GetId() { return _id; } + int GetPAD() { return pad; } + int GetDeadzone(int axis) { return deadzone; } void SaveState(); + int GetButtonState(int i) { return vbutstate[i]; } + int GetAxisState(int i) { return vaxisstate[i]; } - /*int GetHatState(int i) + + int GetPOVState(int i) { - return vhatstate[i]; - }*/ + //printf("Getting POV State of %d.\n", i); + return vpovstate[i]; + } + void SetButtonState(int i, int state) { vbutstate[i] = state; } + void SetAxisState(int i, int value) { vaxisstate[i] = value; } - /*void SetHatState(int i, int value) + + void SetPOVState(int i, int value) { - vhatstate[i] = value; - }*/ + //printf("We should set %d to %d.\n", i, value); + vpovstate[i] = value; + } + SDL_Joystick* GetJoy() { return joy; @@ -105,11 +116,11 @@ class JoystickInfo string devname; // pretty device name int _id; - int numbuttons, numaxes, numpov, numhats; + int numbuttons, numaxes, numpov; int axisrange, deadzone; int pad; - vector vbutstate, vaxisstate; + vector vbutstate, vaxisstate, vpovstate; SDL_Joystick* joy; }; @@ -117,10 +128,3 @@ class JoystickInfo extern int s_selectedpad; extern vector s_vjoysticks; -//extern void JoystickInfo::EnumerateJoysticks(vector& vjoysticks); -/*extern JoystickInfo::JoystickInfo(); -extern void JoystickInfo::Destroy(); -extern bool JoystickInfo::Init(int id, bool bStartThread); -extern void JoystickInfo::Assign(int newpad); -extern void JoystickInfo::SaveState(); -extern void JoystickInfo::TestForce();*/ \ No newline at end of file diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index dcf346c382..e61b1f5792 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -138,50 +138,23 @@ void CALLBACK PADupdate(int pad) switch (key) { case KEY_PAD_LX_LEFT: - g_lanalog[pad].x = DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; + case KEY_PAD_LY_UP: + case KEY_PAD_RX_LEFT: + case KEY_PAD_RY_UP: + Analog::ConfigurePad(KeypadToPad(key), pad, DEF_VALUE); break; case KEY_PAD_LX_RIGHT: - g_lanalog[pad].x = -DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; - break; - case KEY_PAD_LY_UP: - g_lanalog[pad].y = DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - break; case KEY_PAD_LY_DOWN: - g_lanalog[pad].y = -DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - break; - case KEY_PAD_RX_LEFT: - g_ranalog[pad].x = DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - break; case KEY_PAD_RX_RIGHT: - g_ranalog[pad].x = -DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - break; - case KEY_PAD_RY_UP: - g_ranalog[pad].y = DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; - break; case KEY_PAD_RY_DOWN: - g_ranalog[pad].y = -DEF_VALUE / 256; - if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; + Analog::ConfigurePad(KeypadToPad(key), pad, -DEF_VALUE); break; } #endif i = FindKey(key, pad); - if (i != -1) { + if (i != -1) + { keyPress |= (1 << i); keyRelease &= ~(1 << i); } @@ -196,34 +169,21 @@ void CALLBACK PADupdate(int pad) switch (key) { case KEY_PAD_LX_LEFT: - g_lanalog[pad].x = 0x80; - break; - case KEY_PAD_LY_UP: - g_lanalog[pad].y = 0x80; - break; - case KEY_PAD_RX_LEFT: - g_ranalog[pad].x = 0x80; - break; - case KEY_PAD_RY_UP: - g_ranalog[pad].y = 0x80; - break; case KEY_PAD_LX_RIGHT: - g_lanalog[pad].x = 0x80; - break; + case KEY_PAD_LY_UP: case KEY_PAD_LY_DOWN: - g_lanalog[pad].y = 0x80; - break; + case KEY_PAD_RX_LEFT: case KEY_PAD_RX_RIGHT: - g_ranalog[pad].x = 0x80; - break; + case KEY_PAD_RY_UP: case KEY_PAD_RY_DOWN: - g_ranalog[pad].y = 0x80; + Analog::ResetPad(KeypadToPad(key)); break; } #endif i = FindKey(key, pad); - if (i != -1) { + if (i != -1) + { keyPress &= ~(1 << i); keyRelease |= (1 << i); } @@ -277,52 +237,13 @@ void CALLBACK PADupdate(int pad) switch (i) { case PAD_LX: - if (abs(value) > (pjoy)->GetDeadzone(value)) - { - g_lanalog[pad].x = value / 256; - if (conf.options & PADOPTION_REVERTLX) g_lanalog[pad].x = -g_lanalog[pad].x; - g_lanalog[pad].x += 0x80; - } - else - { - g_lanalog[pad].x = 0x80; - } - break; case PAD_LY: - if (abs(value) > (pjoy)->GetDeadzone(value)) - { - g_lanalog[pad].y = value / 256; - if (conf.options & PADOPTION_REVERTLY) g_lanalog[pad].y = -g_lanalog[pad].y; - g_lanalog[pad].y += 0x80; - } - else - { - g_lanalog[pad].y = 0x80; - } - break; case PAD_RX: - if (abs(value) > (pjoy)->GetDeadzone(value)) - { - g_ranalog[pad].x = value / 256; - if (conf.options & PADOPTION_REVERTRX) g_ranalog[pad].x = -g_ranalog[pad].x; - g_ranalog[pad].x += 0x80; - } - else - { - g_ranalog[pad].x = 0x80; - } - break; case PAD_RY: if (abs(value) > (pjoy)->GetDeadzone(value)) - { - g_ranalog[pad].y = value / 256; - if (conf.options&PADOPTION_REVERTRY) g_ranalog[pad].y = -g_ranalog[pad].y; - g_ranalog[pad].y += 0x80; - } + Analog::ConfigurePad(i, pad, value); else - { - g_ranalog[pad].y = 0x80; - } + Analog::ResetPad(i, pad); break; } } @@ -516,6 +437,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this { // released, we don't really want this + // SetButtonState? Shouldn't this be SetAxisState? (*itjoy)->SetButtonState(i, value); break; } @@ -543,21 +465,33 @@ void OnConf_Key(GtkButton *button, gpointer user_data) } } - - /*for (int i = 0; i < (*itjoy)->GetNumHats(); ++i) + /*for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) { int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); - if (value != (*itjoy)->GetAxisState(i)) + if (value != (*itjoy)->GetPOVState(i)) { - *pkey = PAD_HAT((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JHat %d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; + if (abs(value) > 0x3fff) + { + if (key < 16) // POV + { + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + char str[32]; + sprintf(str, "JPOV %d%s", i, value < 0 ? "-" : "+"); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + else // axis + { + *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); + char str[32]; + sprintf(str, "JAxis %d", i); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + } + } } }*/ - itjoy++; } #endif @@ -590,7 +524,7 @@ void CALLBACK PADconfigure() while (it != s_vjoysticks.end()) { sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), - (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()/*, (*it)->GetNumHats()*/); // ,hats: %d + (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()); gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), str); it++; } diff --git a/plugins/zeropad/Makefile.am b/plugins/zeropad/Makefile.am index 81e048bcc9..9376619c69 100644 --- a/plugins/zeropad/Makefile.am +++ b/plugins/zeropad/Makefile.am @@ -27,4 +27,5 @@ libZeroPAD_LDFLAGS= @SHARED_LDFLAGS@ libZeroPAD_LDFLAGS+=-Wl,-soname,@ZEROPAD_SONAME@ libZeroPAD_LDADD=$(libZeroPAD_a_OBJECTS) -libZeroPAD_a_SOURCES = zeropad.cpp Linux/gui.cpp Linux/linux.cpp Linux/joystick.cpp Linux/support.c Linux/interface.c +libZeroPAD_a_SOURCES = analog.cpp analog.h zeropad.cpp zeropad.h \ +Linux/gui.cpp Linux/linux.cpp Linux/joystick.cpp Linux/support.c Linux/interface.c diff --git a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj index f1e2d0552f..957161cb48 100644 --- a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj +++ b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj @@ -250,6 +250,10 @@ RelativePath="..\zeropad.cpp" > + + + + > 8; stdpar[curPad][3] = status[curPad] & 0xff; - stdpar[curPad][4] = g_ranalog[curPad].x; - stdpar[curPad][5] = g_ranalog[curPad].y; - stdpar[curPad][6] = g_lanalog[curPad].x; - stdpar[curPad][7] = g_lanalog[curPad].y; + stdpar[curPad][4] = Analog::Pad(PAD_RX, curPad); + stdpar[curPad][5] = Analog::Pad(PAD_RY, curPad); + stdpar[curPad][6] = Analog::Pad(PAD_LX, curPad); + stdpar[curPad][7] = Analog::Pad(PAD_LY, curPad); if (padMode[curPad] == 1) cmdLen = 20; @@ -473,44 +466,44 @@ u8 _PADpoll(u8 value) buf = stdpar[curPad]; return padID[curPad]; - case 0x43: // CONFIG_MODE + case CMD_CONFIG_MODE: // CONFIG_MODE cmdLen = 8; buf = stdcfg[curPad]; if (stdcfg[curPad][3] == 0xff) return 0xf3; else return padID[curPad]; - case 0x44: // SET_MODE_AND_LOCK + case CMD_SET_MODE_AND_LOCK: // SET_MODE_AND_LOCK cmdLen = 8; buf = stdmode[curPad]; return 0xf3; - case 0x45: // QUERY_MODEL_AND_MODE + case CMD_QUERY_MODEL_AND_MODE: // QUERY_MODEL_AND_MODE cmdLen = 8; buf = stdmodel[curPad]; buf[4] = padMode[curPad]; return 0xf3; - case 0x46: // ?? + case CMD_QUERY_ACT: // ?? cmdLen = 8; buf = unk46[curPad]; return 0xf3; - case 0x47: // ?? + case CMD_QUERY_COMB: // ?? cmdLen = 8; buf = unk47[curPad]; return 0xf3; - case 0x4c: // QUERY_MODE ?? + case CMD_QUERY_MODE: // QUERY_MODE ?? cmdLen = 8; buf = unk4c[curPad]; return 0xf3; - case 0x4d: + case CMD_VIBRATION_TOGGLE: cmdLen = 8; buf = unk4d[curPad]; return 0xf3; - case 0x4f: // SET_DS2_NATIVE_MODE + case CMD_SET_DS2_NATIVE_MODE: // SET_DS2_NATIVE_MODE cmdLen = 8; padID[curPad] = 0x79; // setting ds2 mode ds2mode = 1; // Set DS2 Mode @@ -527,7 +520,7 @@ u8 _PADpoll(u8 value) switch (curCmd) { - case 0x43: + case CMD_CONFIG_MODE: if (curByte == 2) { switch (value) @@ -544,14 +537,14 @@ u8 _PADpoll(u8 value) } break; - case 0x44: + case CMD_SET_MODE_AND_LOCK: if (curByte == 2) { PADsetMode(curPad, value); } break; - case 0x46: + case CMD_QUERY_ACT: if (curByte == 2) { switch (value) @@ -570,7 +563,7 @@ u8 _PADpoll(u8 value) } break; - case 0x4c: + case CMD_QUERY_MODE: if (curByte == 2) { switch (value) diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index 8d633ec780..81bd31fe4b 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -44,6 +44,7 @@ extern "C" { #include "PS2Edefs.h" } +#include "analog.h" extern char libraryName[256]; @@ -93,11 +94,25 @@ typedef struct } PADAnalog; extern PADconf conf; - extern PADAnalog g_lanalog[2], g_ranalog[2]; extern FILE *padLog; #define PAD_LOG __Log +enum PadCommands +{ + CMD_SET_VREF_PARAM = 0x40, + CMD_QUERY_DS2_ANALOG_MODE = 0x41, + CMD_READ_DATA_AND_VIBRATE = 0x42, + CMD_CONFIG_MODE = 0x43, + CMD_SET_MODE_AND_LOCK = 0x44, + CMD_QUERY_MODEL_AND_MODE = 0x45, + CMD_QUERY_ACT = 0x46, // ?? + CMD_QUERY_COMB = 0x47, // ?? + CMD_QUERY_MODE = 0x4C, // QUERY_MODE ?? + CMD_VIBRATION_TOGGLE = 0x4D, + CMD_SET_DS2_NATIVE_MODE = 0x4F // SET_DS2_NATIVE_MODE +}; + enum gamePadValues { PAD_RY = 19, From 05867d69cb4b9148af4ffb090da312f1741bf418 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 9 May 2009 20:07:54 +0000 Subject: [PATCH 040/296] Zeropad: Add analog.cpp & analog.h. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1163 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/analog.cpp | 153 +++++++++++++++++++++++++++++++++++++ plugins/zeropad/analog.h | 32 ++++++++ 2 files changed, 185 insertions(+) create mode 100644 plugins/zeropad/analog.cpp create mode 100644 plugins/zeropad/analog.h diff --git a/plugins/zeropad/analog.cpp b/plugins/zeropad/analog.cpp new file mode 100644 index 0000000000..50f3eb6734 --- /dev/null +++ b/plugins/zeropad/analog.cpp @@ -0,0 +1,153 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include "analog.h" +PADAnalog g_lanalog[NUM_OF_PADS], g_ranalog[NUM_OF_PADS]; + +namespace Analog +{ + u8 Pad(int padvalue, u8 i) + { + switch (padvalue) + { + case PAD_LX: + return g_lanalog[i].x; + break; + + case PAD_RX: + return g_ranalog[i].x; + break; + + case PAD_LY: + return g_lanalog[i].y; + break; + + case PAD_RY: + return g_ranalog[i].y; + break; + + default: + return 0; + break; + } + } + + void SetPad(int padvalue, u8 i, u8 value) + { + switch (padvalue) + { + case PAD_LX: + g_lanalog[i].x = value; + break; + + case PAD_RX: + g_ranalog[i].x = value; + break; + + case PAD_LY: + g_lanalog[i].y = value; + break; + + case PAD_RY: + g_ranalog[i].y = value; + break; + + default: + break; + } + } + + void InvertPad(int padvalue, u8 i) + { + SetPad(padvalue, i, -Pad(padvalue, i)); + } + + void ResetPad(int padvalue, u8 i) + { + SetPad(padvalue, i, 0x80); + } + + void Init() + { + for (int i = 0; i < 2; ++i) + { + ResetPad(PAD_LX, i); + ResetPad(PAD_LY, i); + ResetPad(PAD_RX, i); + ResetPad(PAD_RY, i); + } + } + + bool RevertPad(u8 padvalue) + { + switch (padvalue) + { + case PAD_LX: + return (conf.options & PADOPTION_REVERTLX); + break; + + case PAD_RX: + return (conf.options & PADOPTION_REVERTRX); + break; + + case PAD_LY: + return (conf.options & PADOPTION_REVERTLY); + break; + + case PAD_RY: + return (conf.options & PADOPTION_REVERTRY); + break; + + default: + return false; + break; + } + } + + void ConfigurePad(int padvalue, u8 i, int value) + { + SetPad(padvalue, i, value / 256); + if (RevertPad(padvalue)) InvertPad(padvalue, i); + SetPad(padvalue, i, Pad(padvalue, i) + 0x80); + } + +#ifdef ANALOG_CONTROLS_HACK + int KeypadToPad(u8 keypress) + { + switch (keypress) + { + case KEY_PAD_LX_LEFT: + case KEY_PAD_LX_RIGHT: + return PAD_LX; + break; + case KEY_PAD_LY_UP: + case KEY_PAD_LY_DOWN: + return PAD_LY; + break; + case KEY_PAD_RX_LEFT: + case KEY_PAD_RX_RIGHT: + return PAD_RX; + break; + case KEY_PAD_RY_UP: + case KEY_PAD_RY_DOWN: + return PAD_RY; + break; + } + } +#endif +} \ No newline at end of file diff --git a/plugins/zeropad/analog.h b/plugins/zeropad/analog.h new file mode 100644 index 0000000000..2de476df81 --- /dev/null +++ b/plugins/zeropad/analog.h @@ -0,0 +1,32 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #define NUM_OF_PADS 2 + + #include "zeropad.h" + namespace Analog +{ + extern void Init(); + extern u8 Pad(int padvalue, u8 i); + extern void SetPad(int padvalue, u8 i, u8 value); + extern void InvertPad(int padvalue, u8 i); + extern bool RevertPad(u8 padvalue); + extern void ResetPad(int padvalue, u8 i); + extern void ConfigurePad(int padvalue, u8 i, int value); + extern int KeypadToPad(u8 keypress); +} \ No newline at end of file From ae975a354e997621ddf70e68e6a802fbb917c7d9 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 9 May 2009 22:44:06 +0000 Subject: [PATCH 041/296] Zeropad: More cleanup, and some incomplete work on the POV code, which looks like it was only halfway implemented. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1164 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/gui.cpp | 2 +- plugins/zeropad/Linux/joystick.cpp | 15 +++++---- plugins/zeropad/Linux/joystick.h | 4 +-- plugins/zeropad/Linux/linux.cpp | 50 ++++++++++++++++++------------ plugins/zeropad/zeropad.cpp | 24 ++++++++++++-- plugins/zeropad/zeropad.h | 5 ++- 6 files changed, 68 insertions(+), 32 deletions(-) diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index 8812cabc0b..06a8c06bc2 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -199,7 +199,7 @@ void SysMessage(char *fmt, ...) MsgDlg = gtk_window_new(GTK_WINDOW_TOPLEVEL); gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); - gtk_window_set_title(GTK_WINDOW(MsgDlg), "GSsoft Msg"); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "ZeroPad"); gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); Box = gtk_vbox_new(5, 0); diff --git a/plugins/zeropad/Linux/joystick.cpp b/plugins/zeropad/Linux/joystick.cpp index 514de20e5a..f1d05204e7 100644 --- a/plugins/zeropad/Linux/joystick.cpp +++ b/plugins/zeropad/Linux/joystick.cpp @@ -101,7 +101,7 @@ bool JoystickInfo::Init(int id, bool bStartThread) joy = SDL_JoystickOpen(id); if (joy == NULL) { - printf("failed to open joystick %d\n", id); + PAD_LOG("failed to open joystick %d\n", id); return false; } @@ -109,9 +109,12 @@ bool JoystickInfo::Init(int id, bool bStartThread) numbuttons = SDL_JoystickNumButtons(joy); numpov = SDL_JoystickNumHats(joy); devname = SDL_JoystickName(id); + + vaxisstate.resize(numaxes); vbutstate.resize(numbuttons); - vaxisstate.resize(numbuttons); + vpovstate.resize(numpov); + //PAD_LOG("There are %d buttons, %d axises, and %d povs.\n", numbuttons, numaxes, numpov); return true; } @@ -140,11 +143,11 @@ void JoystickInfo::Assign(int newpad) void JoystickInfo::SaveState() { for (int i = 0; i < numbuttons; ++i) - vbutstate[i] = SDL_JoystickGetButton(joy, i); + SetButtonState(i, SDL_JoystickGetButton(joy, i)); for (int i = 0; i < numaxes; ++i) - vaxisstate[i] = SDL_JoystickGetAxis(joy, i); - /*for (int i = 0; i < numpov; ++i) - vpovstate[i] = SDL_JoystickGetHat(joy, i);*/ + SetAxisState(i, SDL_JoystickGetAxis(joy, i)); + for (int i = 0; i < numpov; ++i) + SetPOVState(i, SDL_JoystickGetHat(joy, i)); } void JoystickInfo::TestForce() diff --git a/plugins/zeropad/Linux/joystick.h b/plugins/zeropad/Linux/joystick.h index fd4d2266d8..35a2be6057 100644 --- a/plugins/zeropad/Linux/joystick.h +++ b/plugins/zeropad/Linux/joystick.h @@ -87,7 +87,7 @@ class JoystickInfo int GetPOVState(int i) { - //printf("Getting POV State of %d.\n", i); + //PAD_LOG("Getting POV State of %d.\n", i); return vpovstate[i]; } @@ -103,7 +103,7 @@ class JoystickInfo void SetPOVState(int i, int value) { - //printf("We should set %d to %d.\n", i, value); + //PAD_LOG("We should set %d to %d.\n", i, value); vpovstate[i] = value; } diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index e61b1f5792..a49d99c1af 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -266,6 +266,7 @@ void CALLBACK PADupdate(int pad) status[pad] |= (1 << i); } } + // Need to add in new POV code here, to match the new configuration pov code. } #endif @@ -279,6 +280,7 @@ void CALLBACK PADupdate(int pad) void UpdateConf(int pad) { + initLogging(); s_selectedpad = pad; int i; @@ -292,7 +294,7 @@ void UpdateConf(int pad) Btn = lookup_widget(Conf, GetLabelFromButton(s_pGuiKeyMap[i]).c_str()); if (Btn == NULL) { - printf("ZeroPAD: cannot find key %s\n", s_pGuiKeyMap[i]); + PAD_LOG("ZeroPAD: cannot find key %s\n", s_pGuiKeyMap[i]); continue; } @@ -359,7 +361,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) GtkWidget* label = lookup_widget(Conf, GetLabelFromButton(gtk_button_get_label(button)).c_str()); if (label == NULL) { - printf("couldn't find correct label\n"); + PAD_LOG("couldn't find correct label\n"); return; } @@ -433,16 +435,15 @@ void OnConf_Key(GtkButton *button, gpointer user_data) if (value != (*itjoy)->GetAxisState(i)) { + PAD_LOG("Change in joystick %d: %d.\n", i, value); if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this { // released, we don't really want this - // SetButtonState? Shouldn't this be SetAxisState? - (*itjoy)->SetButtonState(i, value); + (*itjoy)->SetAxisState(i, value); break; } - if (abs(value) > 0x3fff) { if (key < 16) // POV @@ -465,33 +466,44 @@ void OnConf_Key(GtkButton *button, gpointer user_data) } } - /*for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) +#ifdef EXPERAMENTAL_POV_CODE + for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) { int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); if (value != (*itjoy)->GetPOVState(i)) { - if (abs(value) > 0x3fff) + switch (value) { - if (key < 16) // POV - { + char str[32]; + + case SDL_HAT_UP: *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - char str[32]; - sprintf(str, "JPOV %d%s", i, value < 0 ? "-" : "+"); + sprintf(str, "JPOV%d%s", i, "U"); gtk_entry_set_text(GTK_ENTRY(label), str); return; - } - else // axis - { - *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JAxis %d", i); + case SDL_HAT_RIGHT: + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + sprintf(str, "JPOV%d%s", i, "R"); gtk_entry_set_text(GTK_ENTRY(label), str); return; - } + case SDL_HAT_DOWN: + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + sprintf(str, "JPOV%d%s", i, "D"); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + case SDL_HAT_LEFT: + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + sprintf(str, "JPOV%d%s", i, "L"); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; + // Not handling SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, + // SDL_HAT_LEFTUP, or SDL_HAT_LEFTDOWN here. They should be + // handled in the PADUpdate code, though. } } - }*/ + } +#endif itjoy++; } #endif diff --git a/plugins/zeropad/zeropad.cpp b/plugins/zeropad/zeropad.cpp index e736428be1..454ae40495 100644 --- a/plugins/zeropad/zeropad.cpp +++ b/plugins/zeropad/zeropad.cpp @@ -251,18 +251,31 @@ u32 CALLBACK PS2EgetLibVersion2(u32 type) return (version << 16) | (revision << 8) | build; } -void __Log(char *fmt, ...) +void __Log(const char *fmt, ...) { va_list list; - if (!conf.log || padLog == NULL) return; + //if (padLog == NULL || !conf.log) return; + if (padLog == NULL) return; va_start(list, fmt); vfprintf(padLog, fmt, list); va_end(list); } +void __LogToConsole(const char *fmt, ...) +{ + va_list list; -s32 CALLBACK PADinit(u32 flags) + va_start(list, fmt); + + if (padLog != NULL) vfprintf(padLog, fmt, list); + + printf("ZeroPad: "); + vprintf(fmt, list); + va_end(list); +} + +void initLogging() { #ifdef PAD_LOG if (padLog == NULL) @@ -272,6 +285,11 @@ s32 CALLBACK PADinit(u32 flags) } PAD_LOG("PADinit\n"); #endif +} + +s32 CALLBACK PADinit(u32 flags) +{ + initLogging(); pads |= flags; status[0] = 0xffff; diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index 81bd31fe4b..07e989d155 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -96,7 +96,9 @@ typedef struct extern PADconf conf; extern PADAnalog g_lanalog[2], g_ranalog[2]; extern FILE *padLog; +extern void initLogging(); #define PAD_LOG __Log +//#define PAD_LOG __LogToConsole enum PadCommands { @@ -171,7 +173,8 @@ void _KeyRelease(int pad, u32 key); void PADsetMode(int pad, int mode); void _PADupdate(int pad); -void __Log(char *fmt, ...); +void __Log(const char *fmt, ...); +void __LogToConsole(const char *fmt, ...); void LoadConfig(); void SaveConfig(); From 25090899cecef9700f17de209528570f28473bf7 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 10 May 2009 05:57:57 +0000 Subject: [PATCH 042/296] Zeropad: Experimental. Enable analog key controls in Linux. Currently does not work well if you also have actual joysticks set up as analog controls. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1165 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/callbacks.c | 114 +++ plugins/zeropad/Linux/gui.cpp | 23 +- plugins/zeropad/Linux/interface.c | 645 ++++++++----- plugins/zeropad/Linux/joystick.h | 4 + plugins/zeropad/Linux/linux.cpp | 231 +++-- plugins/zeropad/Linux/linux.h | 42 + plugins/zeropad/Linux/zeropad.glade | 994 +++++++++++++------- plugins/zeropad/Windows/ZeroPAD_2008.vcproj | 4 + plugins/zeropad/analog.cpp | 30 +- plugins/zeropad/bitwise.h | 41 + plugins/zeropad/zeropad.cpp | 83 +- plugins/zeropad/zeropad.h | 31 +- 12 files changed, 1460 insertions(+), 782 deletions(-) create mode 100644 plugins/zeropad/Linux/callbacks.c create mode 100644 plugins/zeropad/Linux/linux.h create mode 100644 plugins/zeropad/bitwise.h diff --git a/plugins/zeropad/Linux/callbacks.c b/plugins/zeropad/Linux/callbacks.c new file mode 100644 index 0000000000..e3a92ed51e --- /dev/null +++ b/plugins/zeropad/Linux/callbacks.c @@ -0,0 +1,114 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Pad1 (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Pad2 (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Pad3 (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Pad4 (GtkButton *button, + gpointer user_data) +{ + +} + + +void +on_joydevicescombo_changed (GtkComboBox *combobox, + gpointer user_data) +{ + +} + + +void +OnConf_Key (GtkButton *button, + gpointer user_data) +{ + +} + + +void +on_checkbutton_reverselx_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_checkbutton_reversely_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_checkbutton_reverserx_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_forcefeedback_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index 06a8c06bc2..3c075e778a 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -26,6 +26,7 @@ #endif #include "zeropad.h" +#include "linux.h" extern "C" { @@ -119,7 +120,7 @@ void OnMsg_Ok() } // GUI event handlers -void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) +void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) { int joyid = gtk_combo_box_get_active(combobox); @@ -132,7 +133,7 @@ void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->Assign(s_selectedpad); } -void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) +void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); if (gtk_toggle_button_get_active(togglebutton)) @@ -141,7 +142,7 @@ void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer conf.options &= ~mask; } -void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) +void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTLY << (16 * s_selectedpad); if (gtk_toggle_button_get_active(togglebutton)) @@ -150,7 +151,7 @@ void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer conf.options &= ~mask; } -void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer user_data) +void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTRX << (16 * s_selectedpad); if (gtk_toggle_button_get_active(togglebutton)) @@ -159,7 +160,7 @@ void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer conf.options &= ~mask; } -void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) +void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTRY << (16 * s_selectedpad); if (gtk_toggle_button_get_active(togglebutton)) @@ -228,26 +229,22 @@ void SysMessage(char *fmt, ...) void OnConf_Pad1(GtkButton *button, gpointer user_data) { - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(0); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) UpdateConf(0); } void OnConf_Pad2(GtkButton *button, gpointer user_data) { - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(1); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) UpdateConf(1); } void OnConf_Pad3(GtkButton *button, gpointer user_data) { - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(2); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) UpdateConf(2); } void OnConf_Pad4(GtkButton *button, gpointer user_data) { - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) - UpdateConf(3); + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(button))) UpdateConf(3); } void OnConf_Ok(GtkButton *button, gpointer user_data) diff --git a/plugins/zeropad/Linux/interface.c b/plugins/zeropad/Linux/interface.c index 7b32fe1606..808afbc3e6 100644 --- a/plugins/zeropad/Linux/interface.c +++ b/plugins/zeropad/Linux/interface.c @@ -95,33 +95,50 @@ create_Conf (void) GtkWidget *eLx; GtkWidget *eRy; GtkWidget *eLy; - GtkWidget *Select; - GtkWidget *L1; - GtkWidget *Left; - GtkWidget *Down; - GtkWidget *Analog; - GtkWidget *R2; - GtkWidget *Start; - GtkWidget *R1; - GtkWidget *Cross; - GtkWidget *Triangle; - GtkWidget *Circle; - GtkWidget *Square; - GtkWidget *L3; - GtkWidget *Up; - GtkWidget *L2; - GtkWidget *Lx; - GtkWidget *Ly; - GtkWidget *Rx; - GtkWidget *Ry; - GtkWidget *Right; - GtkWidget *label3; + GtkWidget *eL_Up; + GtkWidget *eL_Left; + GtkWidget *eL_Right; + GtkWidget *eL_Down; + GtkWidget *eR_Up; + GtkWidget *eR_Left; + GtkWidget *eR_Right; + GtkWidget *eR_Down; GtkWidget *R3; + GtkWidget *Select; + GtkWidget *Start; + GtkWidget *Square; + GtkWidget *Circle; + GtkWidget *L2; + GtkWidget *L1; + GtkWidget *Down; + GtkWidget *Up; + GtkWidget *Left; + GtkWidget *Right; + GtkWidget *R1; + GtkWidget *R2; + GtkWidget *Triangle; + GtkWidget *Cross; + GtkWidget *L3; + GtkWidget *checkbutton_reversery; + GtkWidget *checkbutton_reverserx; + GtkWidget *R_Down; + GtkWidget *Ry; + GtkWidget *Rx; + GtkWidget *R_Up; + GtkWidget *R_Right; + GtkWidget *R_Left; + GtkWidget *forcefeedback; GtkWidget *checkbutton_reverselx; GtkWidget *checkbutton_reversely; - GtkWidget *checkbutton_reverserx; - GtkWidget *checkbutton_reversery; - GtkWidget *forcefeedback; + GtkWidget *L_Down; + GtkWidget *Ly; + GtkWidget *Lx; + GtkWidget *L_Up; + GtkWidget *L_Left; + GtkWidget *L_Right; + GtkWidget *label3; + GtkWidget *Analog; + GtkWidget *label6; GtkWidget *dialog_action_area1; GtkWidget *cancelbutton1; GtkWidget *okbutton1; @@ -194,7 +211,7 @@ create_Conf (void) vbox2 = gtk_vbox_new (FALSE, 0); gtk_widget_set_name (vbox2, "vbox2"); gtk_widget_show (vbox2); - gtk_box_pack_start (GTK_BOX (vbox1), vbox2, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (vbox1), vbox2, FALSE, TRUE, 0); label5 = gtk_label_new (_("Joystick to use for this PAD")); gtk_widget_set_name (label5, "label5"); @@ -204,7 +221,7 @@ create_Conf (void) joydevicescombo = gtk_combo_box_entry_new_text (); gtk_widget_set_name (joydevicescombo, "joydevicescombo"); gtk_widget_show (joydevicescombo); - gtk_box_pack_start (GTK_BOX (vbox2), joydevicescombo, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (vbox2), joydevicescombo, FALSE, TRUE, 0); fixed1 = gtk_fixed_new (); gtk_widget_set_name (fixed1, "fixed1"); @@ -214,7 +231,7 @@ create_Conf (void) eL2 = gtk_entry_new (); gtk_widget_set_name (eL2, "eL2"); gtk_widget_show (eL2); - gtk_fixed_put (GTK_FIXED (fixed1), eL2, 64, 40); + gtk_fixed_put (GTK_FIXED (fixed1), eL2, 64, 32); gtk_widget_set_size_request (eL2, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eL2), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eL2), 8226); @@ -222,14 +239,14 @@ create_Conf (void) eL1 = gtk_entry_new (); gtk_widget_set_name (eL1, "eL1"); gtk_widget_show (eL1); - gtk_fixed_put (GTK_FIXED (fixed1), eL1, 64, 104); + gtk_fixed_put (GTK_FIXED (fixed1), eL1, 64, 80); gtk_widget_set_size_request (eL1, 64, 24); gtk_entry_set_invisible_char (GTK_ENTRY (eL1), 8226); eSelect = gtk_entry_new (); gtk_widget_set_name (eSelect, "eSelect"); gtk_widget_show (eSelect); - gtk_fixed_put (GTK_FIXED (fixed1), eSelect, 168, 64); + gtk_fixed_put (GTK_FIXED (fixed1), eSelect, 200, 184); gtk_widget_set_size_request (eSelect, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eSelect), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eSelect), 8226); @@ -237,7 +254,7 @@ create_Conf (void) eStart = gtk_entry_new (); gtk_widget_set_name (eStart, "eStart"); gtk_widget_show (eStart); - gtk_fixed_put (GTK_FIXED (fixed1), eStart, 280, 64); + gtk_fixed_put (GTK_FIXED (fixed1), eStart, 272, 184); gtk_widget_set_size_request (eStart, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eStart), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eStart), 8226); @@ -245,14 +262,14 @@ create_Conf (void) eUp = gtk_entry_new (); gtk_widget_set_name (eUp, "eUp"); gtk_widget_show (eUp); - gtk_fixed_put (GTK_FIXED (fixed1), eUp, 64, 168); + gtk_fixed_put (GTK_FIXED (fixed1), eUp, 64, 136); gtk_widget_set_size_request (eUp, 64, 24); gtk_entry_set_invisible_char (GTK_ENTRY (eUp), 8226); eRight = gtk_entry_new (); gtk_widget_set_name (eRight, "eRight"); gtk_widget_show (eRight); - gtk_fixed_put (GTK_FIXED (fixed1), eRight, 128, 208); + gtk_fixed_put (GTK_FIXED (fixed1), eRight, 128, 184); gtk_widget_set_size_request (eRight, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eRight), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eRight), 8226); @@ -260,7 +277,7 @@ create_Conf (void) eLeft = gtk_entry_new (); gtk_widget_set_name (eLeft, "eLeft"); gtk_widget_show (eLeft); - gtk_fixed_put (GTK_FIXED (fixed1), eLeft, 0, 208); + gtk_fixed_put (GTK_FIXED (fixed1), eLeft, 0, 184); gtk_widget_set_size_request (eLeft, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eLeft), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eLeft), 8226); @@ -268,7 +285,7 @@ create_Conf (void) eDown = gtk_entry_new (); gtk_widget_set_name (eDown, "eDown"); gtk_widget_show (eDown); - gtk_fixed_put (GTK_FIXED (fixed1), eDown, 64, 256); + gtk_fixed_put (GTK_FIXED (fixed1), eDown, 64, 232); gtk_widget_set_size_request (eDown, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eDown), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eDown), 8226); @@ -276,7 +293,7 @@ create_Conf (void) eR2 = gtk_entry_new (); gtk_widget_set_name (eR2, "eR2"); gtk_widget_show (eR2); - gtk_fixed_put (GTK_FIXED (fixed1), eR2, 368, 32); + gtk_fixed_put (GTK_FIXED (fixed1), eR2, 408, 32); gtk_widget_set_size_request (eR2, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eR2), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eR2), 8226); @@ -284,7 +301,7 @@ create_Conf (void) eR1 = gtk_entry_new (); gtk_widget_set_name (eR1, "eR1"); gtk_widget_show (eR1); - gtk_fixed_put (GTK_FIXED (fixed1), eR1, 368, 104); + gtk_fixed_put (GTK_FIXED (fixed1), eR1, 408, 80); gtk_widget_set_size_request (eR1, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eR1), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eR1), 8226); @@ -292,7 +309,7 @@ create_Conf (void) eAnalog = gtk_entry_new (); gtk_widget_set_name (eAnalog, "eAnalog"); gtk_widget_show (eAnalog); - gtk_fixed_put (GTK_FIXED (fixed1), eAnalog, 224, 128); + gtk_fixed_put (GTK_FIXED (fixed1), eAnalog, 240, 232); gtk_widget_set_size_request (eAnalog, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eAnalog), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eAnalog), 8226); @@ -300,7 +317,7 @@ create_Conf (void) eSquare = gtk_entry_new (); gtk_widget_set_name (eSquare, "eSquare"); gtk_widget_show (eSquare); - gtk_fixed_put (GTK_FIXED (fixed1), eSquare, 304, 208); + gtk_fixed_put (GTK_FIXED (fixed1), eSquare, 344, 184); gtk_widget_set_size_request (eSquare, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eSquare), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eSquare), 8226); @@ -308,7 +325,7 @@ create_Conf (void) eTriangle = gtk_entry_new (); gtk_widget_set_name (eTriangle, "eTriangle"); gtk_widget_show (eTriangle); - gtk_fixed_put (GTK_FIXED (fixed1), eTriangle, 368, 168); + gtk_fixed_put (GTK_FIXED (fixed1), eTriangle, 408, 136); gtk_widget_set_size_request (eTriangle, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eTriangle), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eTriangle), 8226); @@ -316,7 +333,7 @@ create_Conf (void) eCircle = gtk_entry_new (); gtk_widget_set_name (eCircle, "eCircle"); gtk_widget_show (eCircle); - gtk_fixed_put (GTK_FIXED (fixed1), eCircle, 432, 208); + gtk_fixed_put (GTK_FIXED (fixed1), eCircle, 472, 184); gtk_widget_set_size_request (eCircle, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eCircle), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eCircle), 8226); @@ -324,7 +341,7 @@ create_Conf (void) eCross = gtk_entry_new (); gtk_widget_set_name (eCross, "eCross"); gtk_widget_show (eCross); - gtk_fixed_put (GTK_FIXED (fixed1), eCross, 368, 256); + gtk_fixed_put (GTK_FIXED (fixed1), eCross, 408, 232); gtk_widget_set_size_request (eCross, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eCross), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eCross), 8226); @@ -332,7 +349,7 @@ create_Conf (void) eR3 = gtk_entry_new (); gtk_widget_set_name (eR3, "eR3"); gtk_widget_show (eR3); - gtk_fixed_put (GTK_FIXED (fixed1), eR3, 272, 272); + gtk_fixed_put (GTK_FIXED (fixed1), eR3, 272, 56); gtk_widget_set_size_request (eR3, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eR3), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eR3), 8226); @@ -340,7 +357,7 @@ create_Conf (void) eL3 = gtk_entry_new (); gtk_widget_set_name (eL3, "eL3"); gtk_widget_show (eL3); - gtk_fixed_put (GTK_FIXED (fixed1), eL3, 176, 272); + gtk_fixed_put (GTK_FIXED (fixed1), eL3, 200, 56); gtk_widget_set_size_request (eL3, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eL3), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eL3), 8226); @@ -348,197 +365,302 @@ create_Conf (void) eRx = gtk_entry_new (); gtk_widget_set_name (eRx, "eRx"); gtk_widget_show (eRx); - gtk_fixed_put (GTK_FIXED (fixed1), eRx, 272, 352); - gtk_widget_set_size_request (eRx, 80, 24); + gtk_fixed_put (GTK_FIXED (fixed1), eRx, 408, 384); + gtk_widget_set_size_request (eRx, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eRx), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eRx), 8226); eLx = gtk_entry_new (); gtk_widget_set_name (eLx, "eLx"); gtk_widget_show (eLx); - gtk_fixed_put (GTK_FIXED (fixed1), eLx, 160, 352); - gtk_widget_set_size_request (eLx, 80, 24); + gtk_fixed_put (GTK_FIXED (fixed1), eLx, 64, 384); + gtk_widget_set_size_request (eLx, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eLx), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eLx), 8226); eRy = gtk_entry_new (); gtk_widget_set_name (eRy, "eRy"); gtk_widget_show (eRy); - gtk_fixed_put (GTK_FIXED (fixed1), eRy, 272, 416); - gtk_widget_set_size_request (eRy, 80, 24); + gtk_fixed_put (GTK_FIXED (fixed1), eRy, 408, 432); + gtk_widget_set_size_request (eRy, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eRy), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eRy), 8226); eLy = gtk_entry_new (); gtk_widget_set_name (eLy, "eLy"); gtk_widget_show (eLy); - gtk_fixed_put (GTK_FIXED (fixed1), eLy, 160, 416); - gtk_widget_set_size_request (eLy, 80, 24); + gtk_fixed_put (GTK_FIXED (fixed1), eLy, 64, 432); + gtk_widget_set_size_request (eLy, 64, 24); gtk_editable_set_editable (GTK_EDITABLE (eLy), FALSE); gtk_entry_set_invisible_char (GTK_ENTRY (eLy), 8226); + eL_Up = gtk_entry_new (); + gtk_widget_set_name (eL_Up, "eL_Up"); + gtk_widget_show (eL_Up); + gtk_fixed_put (GTK_FIXED (fixed1), eL_Up, 64, 336); + gtk_widget_set_size_request (eL_Up, 64, 24); + + eL_Left = gtk_entry_new (); + gtk_widget_set_name (eL_Left, "eL_Left"); + gtk_widget_show (eL_Left); + gtk_fixed_put (GTK_FIXED (fixed1), eL_Left, 0, 408); + gtk_widget_set_size_request (eL_Left, 64, 24); + + eL_Right = gtk_entry_new (); + gtk_widget_set_name (eL_Right, "eL_Right"); + gtk_widget_show (eL_Right); + gtk_fixed_put (GTK_FIXED (fixed1), eL_Right, 128, 408); + gtk_widget_set_size_request (eL_Right, 64, 24); + + eL_Down = gtk_entry_new (); + gtk_widget_set_name (eL_Down, "eL_Down"); + gtk_widget_show (eL_Down); + gtk_fixed_put (GTK_FIXED (fixed1), eL_Down, 64, 480); + gtk_widget_set_size_request (eL_Down, 64, 24); + + eR_Up = gtk_entry_new (); + gtk_widget_set_name (eR_Up, "eR_Up"); + gtk_widget_show (eR_Up); + gtk_fixed_put (GTK_FIXED (fixed1), eR_Up, 408, 336); + gtk_widget_set_size_request (eR_Up, 64, 24); + + eR_Left = gtk_entry_new (); + gtk_widget_set_name (eR_Left, "eR_Left"); + gtk_widget_show (eR_Left); + gtk_fixed_put (GTK_FIXED (fixed1), eR_Left, 344, 408); + gtk_widget_set_size_request (eR_Left, 64, 24); + + eR_Right = gtk_entry_new (); + gtk_widget_set_name (eR_Right, "eR_Right"); + gtk_widget_show (eR_Right); + gtk_fixed_put (GTK_FIXED (fixed1), eR_Right, 472, 408); + gtk_widget_set_size_request (eR_Right, 64, 24); + + eR_Down = gtk_entry_new (); + gtk_widget_set_name (eR_Down, "eR_Down"); + gtk_widget_show (eR_Down); + gtk_fixed_put (GTK_FIXED (fixed1), eR_Down, 408, 480); + gtk_widget_set_size_request (eR_Down, 64, 24); + + R3 = gtk_button_new_with_mnemonic (_("R3")); + gtk_widget_set_name (R3, "R3"); + gtk_widget_show (R3); + gtk_fixed_put (GTK_FIXED (fixed1), R3, 272, 32); + gtk_widget_set_size_request (R3, 64, 24); + Select = gtk_button_new_with_mnemonic (_("Select")); gtk_widget_set_name (Select, "Select"); gtk_widget_show (Select); - gtk_fixed_put (GTK_FIXED (fixed1), Select, 168, 32); - gtk_widget_set_size_request (Select, 64, 32); - - L1 = gtk_button_new_with_mnemonic (_("L1")); - gtk_widget_set_name (L1, "L1"); - gtk_widget_show (L1); - gtk_fixed_put (GTK_FIXED (fixed1), L1, 64, 72); - gtk_widget_set_size_request (L1, 64, 32); - - Left = gtk_button_new_with_mnemonic (_("Left")); - gtk_widget_set_name (Left, "Left"); - gtk_widget_show (Left); - gtk_fixed_put (GTK_FIXED (fixed1), Left, 0, 176); - gtk_widget_set_size_request (Left, 64, 32); - - Down = gtk_button_new_with_mnemonic (_("Down")); - gtk_widget_set_name (Down, "Down"); - gtk_widget_show (Down); - gtk_fixed_put (GTK_FIXED (fixed1), Down, 64, 224); - gtk_widget_set_size_request (Down, 64, 32); - - Analog = gtk_button_new_with_mnemonic (_("Analog")); - gtk_widget_set_name (Analog, "Analog"); - gtk_widget_show (Analog); - gtk_fixed_put (GTK_FIXED (fixed1), Analog, 224, 96); - gtk_widget_set_size_request (Analog, 64, 32); - - R2 = gtk_button_new_with_mnemonic (_("R2")); - gtk_widget_set_name (R2, "R2"); - gtk_widget_show (R2); - gtk_fixed_put (GTK_FIXED (fixed1), R2, 368, 0); - gtk_widget_set_size_request (R2, 64, 32); + gtk_fixed_put (GTK_FIXED (fixed1), Select, 200, 160); + gtk_widget_set_size_request (Select, 64, 24); Start = gtk_button_new_with_mnemonic (_("Start")); gtk_widget_set_name (Start, "Start"); gtk_widget_show (Start); - gtk_fixed_put (GTK_FIXED (fixed1), Start, 280, 32); - gtk_widget_set_size_request (Start, 64, 32); - - R1 = gtk_button_new_with_mnemonic (_("R1")); - gtk_widget_set_name (R1, "R1"); - gtk_widget_show (R1); - gtk_fixed_put (GTK_FIXED (fixed1), R1, 368, 72); - gtk_widget_set_size_request (R1, 64, 32); - - Cross = gtk_button_new_with_mnemonic (_("Cross")); - gtk_widget_set_name (Cross, "Cross"); - gtk_widget_show (Cross); - gtk_fixed_put (GTK_FIXED (fixed1), Cross, 368, 224); - gtk_widget_set_size_request (Cross, 64, 32); - - Triangle = gtk_button_new_with_mnemonic (_("Triangle")); - gtk_widget_set_name (Triangle, "Triangle"); - gtk_widget_show (Triangle); - gtk_fixed_put (GTK_FIXED (fixed1), Triangle, 368, 136); - gtk_widget_set_size_request (Triangle, 64, 32); - - Circle = gtk_button_new_with_mnemonic (_("Circle")); - gtk_widget_set_name (Circle, "Circle"); - gtk_widget_show (Circle); - gtk_fixed_put (GTK_FIXED (fixed1), Circle, 432, 176); - gtk_widget_set_size_request (Circle, 64, 32); + gtk_fixed_put (GTK_FIXED (fixed1), Start, 272, 160); + gtk_widget_set_size_request (Start, 64, 24); Square = gtk_button_new_with_mnemonic (_("Square")); gtk_widget_set_name (Square, "Square"); gtk_widget_show (Square); - gtk_fixed_put (GTK_FIXED (fixed1), Square, 304, 176); - gtk_widget_set_size_request (Square, 64, 32); + gtk_fixed_put (GTK_FIXED (fixed1), Square, 344, 160); + gtk_widget_set_size_request (Square, 64, 24); - L3 = gtk_button_new_with_mnemonic (_("L3")); - gtk_widget_set_name (L3, "L3"); - gtk_widget_show (L3); - gtk_fixed_put (GTK_FIXED (fixed1), L3, 176, 240); - gtk_widget_set_size_request (L3, 64, 32); - - Up = gtk_button_new_with_mnemonic (_("Up")); - gtk_widget_set_name (Up, "Up"); - gtk_widget_show (Up); - gtk_fixed_put (GTK_FIXED (fixed1), Up, 64, 136); - gtk_widget_set_size_request (Up, 64, 32); + Circle = gtk_button_new_with_mnemonic (_("Circle")); + gtk_widget_set_name (Circle, "Circle"); + gtk_widget_show (Circle); + gtk_fixed_put (GTK_FIXED (fixed1), Circle, 472, 160); + gtk_widget_set_size_request (Circle, 64, 24); L2 = gtk_button_new_with_mnemonic (_("L2")); gtk_widget_set_name (L2, "L2"); gtk_widget_show (L2); gtk_fixed_put (GTK_FIXED (fixed1), L2, 64, 8); - gtk_widget_set_size_request (L2, 64, 32); + gtk_widget_set_size_request (L2, 64, 24); - Lx = gtk_button_new_with_mnemonic (_("Lx")); - gtk_widget_set_name (Lx, "Lx"); - gtk_widget_show (Lx); - gtk_fixed_put (GTK_FIXED (fixed1), Lx, 160, 320); - gtk_widget_set_size_request (Lx, 80, 32); + L1 = gtk_button_new_with_mnemonic (_("L1")); + gtk_widget_set_name (L1, "L1"); + gtk_widget_show (L1); + gtk_fixed_put (GTK_FIXED (fixed1), L1, 64, 56); + gtk_widget_set_size_request (L1, 64, 24); - Ly = gtk_button_new_with_mnemonic (_("Ly")); - gtk_widget_set_name (Ly, "Ly"); - gtk_widget_show (Ly); - gtk_fixed_put (GTK_FIXED (fixed1), Ly, 160, 384); - gtk_widget_set_size_request (Ly, 80, 32); + Down = gtk_button_new_with_mnemonic (_("Down")); + gtk_widget_set_name (Down, "Down"); + gtk_widget_show (Down); + gtk_fixed_put (GTK_FIXED (fixed1), Down, 64, 208); + gtk_widget_set_size_request (Down, 64, 24); - Rx = gtk_button_new_with_mnemonic (_("Rx")); - gtk_widget_set_name (Rx, "Rx"); - gtk_widget_show (Rx); - gtk_fixed_put (GTK_FIXED (fixed1), Rx, 272, 320); - gtk_widget_set_size_request (Rx, 80, 32); + Up = gtk_button_new_with_mnemonic (_("Up")); + gtk_widget_set_name (Up, "Up"); + gtk_widget_show (Up); + gtk_fixed_put (GTK_FIXED (fixed1), Up, 64, 112); + gtk_widget_set_size_request (Up, 64, 24); - Ry = gtk_button_new_with_mnemonic (_("Ry")); - gtk_widget_set_name (Ry, "Ry"); - gtk_widget_show (Ry); - gtk_fixed_put (GTK_FIXED (fixed1), Ry, 272, 384); - gtk_widget_set_size_request (Ry, 80, 32); + Left = gtk_button_new_with_mnemonic (_("Left")); + gtk_widget_set_name (Left, "Left"); + gtk_widget_show (Left); + gtk_fixed_put (GTK_FIXED (fixed1), Left, 0, 160); + gtk_widget_set_size_request (Left, 64, 24); Right = gtk_button_new_with_mnemonic (_("Right")); gtk_widget_set_name (Right, "Right"); gtk_widget_show (Right); - gtk_fixed_put (GTK_FIXED (fixed1), Right, 128, 176); - gtk_widget_set_size_request (Right, 64, 32); + gtk_fixed_put (GTK_FIXED (fixed1), Right, 128, 160); + gtk_widget_set_size_request (Right, 64, 24); - label3 = gtk_label_new (_("Analog Controls (move mouse or analog joystick to select)")); - gtk_widget_set_name (label3, "label3"); - gtk_widget_show (label3); - gtk_fixed_put (GTK_FIXED (fixed1), label3, 48, 296); - gtk_widget_set_size_request (label3, 408, 16); - gtk_label_set_single_line_mode (GTK_LABEL (label3), TRUE); + R1 = gtk_button_new_with_mnemonic (_("R1")); + gtk_widget_set_name (R1, "R1"); + gtk_widget_show (R1); + gtk_fixed_put (GTK_FIXED (fixed1), R1, 408, 56); + gtk_widget_set_size_request (R1, 64, 24); - R3 = gtk_button_new_with_mnemonic (_("R3")); - gtk_widget_set_name (R3, "R3"); - gtk_widget_show (R3); - gtk_fixed_put (GTK_FIXED (fixed1), R3, 272, 240); - gtk_widget_set_size_request (R3, 64, 32); + R2 = gtk_button_new_with_mnemonic (_("R2")); + gtk_widget_set_name (R2, "R2"); + gtk_widget_show (R2); + gtk_fixed_put (GTK_FIXED (fixed1), R2, 408, 8); + gtk_widget_set_size_request (R2, 64, 24); + + Triangle = gtk_button_new_with_mnemonic (_("Triangle")); + gtk_widget_set_name (Triangle, "Triangle"); + gtk_widget_show (Triangle); + gtk_fixed_put (GTK_FIXED (fixed1), Triangle, 408, 112); + gtk_widget_set_size_request (Triangle, 64, 24); + + Cross = gtk_button_new_with_mnemonic (_("Cross")); + gtk_widget_set_name (Cross, "Cross"); + gtk_widget_show (Cross); + gtk_fixed_put (GTK_FIXED (fixed1), Cross, 408, 208); + gtk_widget_set_size_request (Cross, 64, 24); + + L3 = gtk_button_new_with_mnemonic (_("L3")); + gtk_widget_set_name (L3, "L3"); + gtk_widget_show (L3); + gtk_fixed_put (GTK_FIXED (fixed1), L3, 200, 32); + gtk_widget_set_size_request (L3, 64, 24); + + checkbutton_reversery = gtk_check_button_new_with_mnemonic (_("Reverse RY")); + gtk_widget_set_name (checkbutton_reversery, "checkbutton_reversery"); + gtk_widget_show (checkbutton_reversery); + gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversery, 384, 536); + gtk_widget_set_size_request (checkbutton_reversery, 111, 22); + + checkbutton_reverserx = gtk_check_button_new_with_mnemonic (_("Reverse RX")); + gtk_widget_set_name (checkbutton_reverserx, "checkbutton_reverserx"); + gtk_widget_show (checkbutton_reverserx); + gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverserx, 384, 512); + gtk_widget_set_size_request (checkbutton_reverserx, 111, 22); + + R_Down = gtk_button_new_with_mnemonic (_("Down")); + gtk_widget_set_name (R_Down, "R_Down"); + gtk_widget_show (R_Down); + gtk_fixed_put (GTK_FIXED (fixed1), R_Down, 408, 456); + gtk_widget_set_size_request (R_Down, 64, 24); + + Ry = gtk_button_new_with_mnemonic (_("Ry")); + gtk_widget_set_name (Ry, "Ry"); + gtk_widget_show (Ry); + gtk_fixed_put (GTK_FIXED (fixed1), Ry, 408, 408); + gtk_widget_set_size_request (Ry, 64, 24); + + Rx = gtk_button_new_with_mnemonic (_("Rx")); + gtk_widget_set_name (Rx, "Rx"); + gtk_widget_show (Rx); + gtk_fixed_put (GTK_FIXED (fixed1), Rx, 408, 360); + gtk_widget_set_size_request (Rx, 64, 24); + + R_Up = gtk_button_new_with_mnemonic (_("Up")); + gtk_widget_set_name (R_Up, "R_Up"); + gtk_widget_show (R_Up); + gtk_fixed_put (GTK_FIXED (fixed1), R_Up, 408, 312); + gtk_widget_set_size_request (R_Up, 64, 24); + + R_Right = gtk_button_new_with_mnemonic (_("Right")); + gtk_widget_set_name (R_Right, "R_Right"); + gtk_widget_show (R_Right); + gtk_fixed_put (GTK_FIXED (fixed1), R_Right, 472, 384); + gtk_widget_set_size_request (R_Right, 64, 24); + + R_Left = gtk_button_new_with_mnemonic (_("Left")); + gtk_widget_set_name (R_Left, "R_Left"); + gtk_widget_show (R_Left); + gtk_fixed_put (GTK_FIXED (fixed1), R_Left, 344, 384); + gtk_widget_set_size_request (R_Left, 64, 24); + + forcefeedback = gtk_check_button_new_with_mnemonic (_("Enable Force\nFeedback")); + gtk_widget_set_name (forcefeedback, "forcefeedback"); + gtk_widget_show (forcefeedback); + gtk_fixed_put (GTK_FIXED (fixed1), forcefeedback, 216, 512); + gtk_widget_set_size_request (forcefeedback, 112, 48); checkbutton_reverselx = gtk_check_button_new_with_mnemonic (_("Reverse LX")); gtk_widget_set_name (checkbutton_reverselx, "checkbutton_reverselx"); gtk_widget_show (checkbutton_reverselx); - gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverselx, 16, 328); + gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverselx, 48, 512); gtk_widget_set_size_request (checkbutton_reverselx, 111, 22); checkbutton_reversely = gtk_check_button_new_with_mnemonic (_("Reverse LY")); gtk_widget_set_name (checkbutton_reversely, "checkbutton_reversely"); gtk_widget_show (checkbutton_reversely); - gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversely, 16, 352); + gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversely, 48, 536); gtk_widget_set_size_request (checkbutton_reversely, 111, 22); - checkbutton_reverserx = gtk_check_button_new_with_mnemonic (_("Reverse RX")); - gtk_widget_set_name (checkbutton_reverserx, "checkbutton_reverserx"); - gtk_widget_show (checkbutton_reverserx); - gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reverserx, 16, 376); - gtk_widget_set_size_request (checkbutton_reverserx, 111, 22); + L_Down = gtk_button_new_with_mnemonic (_("Down")); + gtk_widget_set_name (L_Down, "L_Down"); + gtk_widget_show (L_Down); + gtk_fixed_put (GTK_FIXED (fixed1), L_Down, 64, 456); + gtk_widget_set_size_request (L_Down, 64, 24); - checkbutton_reversery = gtk_check_button_new_with_mnemonic (_("Reverse RY")); - gtk_widget_set_name (checkbutton_reversery, "checkbutton_reversery"); - gtk_widget_show (checkbutton_reversery); - gtk_fixed_put (GTK_FIXED (fixed1), checkbutton_reversery, 16, 400); - gtk_widget_set_size_request (checkbutton_reversery, 111, 22); + Ly = gtk_button_new_with_mnemonic (_("Ly")); + gtk_widget_set_name (Ly, "Ly"); + gtk_widget_show (Ly); + gtk_fixed_put (GTK_FIXED (fixed1), Ly, 64, 408); + gtk_widget_set_size_request (Ly, 64, 24); - forcefeedback = gtk_check_button_new_with_mnemonic (_("Enable Force\nFeedback")); - gtk_widget_set_name (forcefeedback, "forcefeedback"); - gtk_widget_show (forcefeedback); - gtk_fixed_put (GTK_FIXED (fixed1), forcefeedback, 384, 320); - gtk_widget_set_size_request (forcefeedback, 112, 48); + Lx = gtk_button_new_with_mnemonic (_("Lx")); + gtk_widget_set_name (Lx, "Lx"); + gtk_widget_show (Lx); + gtk_fixed_put (GTK_FIXED (fixed1), Lx, 64, 360); + gtk_widget_set_size_request (Lx, 64, 24); + + L_Up = gtk_button_new_with_mnemonic (_("Up")); + gtk_widget_set_name (L_Up, "L_Up"); + gtk_widget_show (L_Up); + gtk_fixed_put (GTK_FIXED (fixed1), L_Up, 64, 312); + gtk_widget_set_size_request (L_Up, 64, 24); + + L_Left = gtk_button_new_with_mnemonic (_("Left")); + gtk_widget_set_name (L_Left, "L_Left"); + gtk_widget_show (L_Left); + gtk_fixed_put (GTK_FIXED (fixed1), L_Left, 0, 384); + gtk_widget_set_size_request (L_Left, 64, 24); + + L_Right = gtk_button_new_with_mnemonic (_("Right")); + gtk_widget_set_name (L_Right, "L_Right"); + gtk_widget_show (L_Right); + gtk_fixed_put (GTK_FIXED (fixed1), L_Right, 128, 384); + gtk_widget_set_size_request (L_Right, 64, 24); + + label3 = gtk_label_new (_("Analog Controls (move mouse or analog joystick to select)")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_fixed_put (GTK_FIXED (fixed1), label3, 72, 256); + gtk_widget_set_size_request (label3, 400, 24); + gtk_label_set_line_wrap (GTK_LABEL (label3), TRUE); + gtk_label_set_single_line_mode (GTK_LABEL (label3), TRUE); + + Analog = gtk_button_new_with_mnemonic (_("Analog")); + gtk_widget_set_name (Analog, "Analog"); + gtk_widget_show (Analog); + gtk_fixed_put (GTK_FIXED (fixed1), Analog, 240, 208); + gtk_widget_set_size_request (Analog, 64, 24); + + label6 = gtk_label_new (_("(Note: The analog key controls and joystick controls do not work well together currently, and should not be mixed. )")); + gtk_widget_set_name (label6, "label6"); + gtk_widget_show (label6); + gtk_fixed_put (GTK_FIXED (fixed1), label6, 56, 280); + gtk_widget_set_size_request (label6, 448, 32); + gtk_label_set_justify (GTK_LABEL (label6), GTK_JUSTIFY_CENTER); + gtk_label_set_line_wrap (GTK_LABEL (label6), TRUE); dialog_action_area1 = GTK_DIALOG (Conf)->action_area; gtk_widget_set_name (dialog_action_area1, "dialog_action_area1"); @@ -572,83 +694,107 @@ create_Conf (void) g_signal_connect ((gpointer) joydevicescombo, "changed", G_CALLBACK (on_joydevicescombo_changed), NULL); + g_signal_connect ((gpointer) R3, "clicked", + G_CALLBACK (OnConf_Key), + NULL); g_signal_connect ((gpointer) Select, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) L1, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Left, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Down, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Analog, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) R2, "clicked", - G_CALLBACK (OnConf_Key), - NULL); g_signal_connect ((gpointer) Start, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) R1, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Cross, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Triangle, "clicked", + g_signal_connect ((gpointer) Square, "clicked", G_CALLBACK (OnConf_Key), NULL); g_signal_connect ((gpointer) Circle, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) Square, "clicked", + g_signal_connect ((gpointer) L2, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) L3, "clicked", + g_signal_connect ((gpointer) L1, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Down, "clicked", G_CALLBACK (OnConf_Key), NULL); g_signal_connect ((gpointer) Up, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) L2, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Lx, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Ly, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Rx, "clicked", - G_CALLBACK (OnConf_Key), - NULL); - g_signal_connect ((gpointer) Ry, "clicked", + g_signal_connect ((gpointer) Left, "clicked", G_CALLBACK (OnConf_Key), NULL); g_signal_connect ((gpointer) Right, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) R3, "clicked", + g_signal_connect ((gpointer) R1, "clicked", G_CALLBACK (OnConf_Key), NULL); + g_signal_connect ((gpointer) R2, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Triangle, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Cross, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) L3, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) checkbutton_reversery, "toggled", + G_CALLBACK (on_checkbutton_reversery_toggled), + NULL); + g_signal_connect ((gpointer) checkbutton_reverserx, "toggled", + G_CALLBACK (on_checkbutton_reverserx_toggled), + NULL); + g_signal_connect ((gpointer) R_Down, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Ry, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Rx, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) R_Up, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) R_Right, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) R_Left, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) forcefeedback, "toggled", + G_CALLBACK (on_forcefeedback_toggled), + NULL); g_signal_connect ((gpointer) checkbutton_reverselx, "toggled", G_CALLBACK (on_checkbutton_reverselx_toggled), NULL); g_signal_connect ((gpointer) checkbutton_reversely, "toggled", G_CALLBACK (on_checkbutton_reversely_toggled), NULL); - g_signal_connect ((gpointer) checkbutton_reverserx, "toggled", - G_CALLBACK (on_checkbutton_reverserx_toggled), + g_signal_connect ((gpointer) L_Down, "clicked", + G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) checkbutton_reversery, "toggled", - G_CALLBACK (on_checkbutton_reversery_toggled), + g_signal_connect ((gpointer) Ly, "clicked", + G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) forcefeedback, "toggled", - G_CALLBACK (on_forcefeedback_toggled), + g_signal_connect ((gpointer) Lx, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) L_Up, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) L_Left, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) L_Right, "clicked", + G_CALLBACK (OnConf_Key), + NULL); + g_signal_connect ((gpointer) Analog, "clicked", + G_CALLBACK (OnConf_Key), NULL); g_signal_connect ((gpointer) cancelbutton1, "clicked", G_CALLBACK (OnConf_Cancel), @@ -694,33 +840,50 @@ create_Conf (void) GLADE_HOOKUP_OBJECT (Conf, eLx, "eLx"); GLADE_HOOKUP_OBJECT (Conf, eRy, "eRy"); GLADE_HOOKUP_OBJECT (Conf, eLy, "eLy"); - GLADE_HOOKUP_OBJECT (Conf, Select, "Select"); - GLADE_HOOKUP_OBJECT (Conf, L1, "L1"); - GLADE_HOOKUP_OBJECT (Conf, Left, "Left"); - GLADE_HOOKUP_OBJECT (Conf, Down, "Down"); - GLADE_HOOKUP_OBJECT (Conf, Analog, "Analog"); - GLADE_HOOKUP_OBJECT (Conf, R2, "R2"); - GLADE_HOOKUP_OBJECT (Conf, Start, "Start"); - GLADE_HOOKUP_OBJECT (Conf, R1, "R1"); - GLADE_HOOKUP_OBJECT (Conf, Cross, "Cross"); - GLADE_HOOKUP_OBJECT (Conf, Triangle, "Triangle"); - GLADE_HOOKUP_OBJECT (Conf, Circle, "Circle"); - GLADE_HOOKUP_OBJECT (Conf, Square, "Square"); - GLADE_HOOKUP_OBJECT (Conf, L3, "L3"); - GLADE_HOOKUP_OBJECT (Conf, Up, "Up"); - GLADE_HOOKUP_OBJECT (Conf, L2, "L2"); - GLADE_HOOKUP_OBJECT (Conf, Lx, "Lx"); - GLADE_HOOKUP_OBJECT (Conf, Ly, "Ly"); - GLADE_HOOKUP_OBJECT (Conf, Rx, "Rx"); - GLADE_HOOKUP_OBJECT (Conf, Ry, "Ry"); - GLADE_HOOKUP_OBJECT (Conf, Right, "Right"); - GLADE_HOOKUP_OBJECT (Conf, label3, "label3"); + GLADE_HOOKUP_OBJECT (Conf, eL_Up, "eL_Up"); + GLADE_HOOKUP_OBJECT (Conf, eL_Left, "eL_Left"); + GLADE_HOOKUP_OBJECT (Conf, eL_Right, "eL_Right"); + GLADE_HOOKUP_OBJECT (Conf, eL_Down, "eL_Down"); + GLADE_HOOKUP_OBJECT (Conf, eR_Up, "eR_Up"); + GLADE_HOOKUP_OBJECT (Conf, eR_Left, "eR_Left"); + GLADE_HOOKUP_OBJECT (Conf, eR_Right, "eR_Right"); + GLADE_HOOKUP_OBJECT (Conf, eR_Down, "eR_Down"); GLADE_HOOKUP_OBJECT (Conf, R3, "R3"); + GLADE_HOOKUP_OBJECT (Conf, Select, "Select"); + GLADE_HOOKUP_OBJECT (Conf, Start, "Start"); + GLADE_HOOKUP_OBJECT (Conf, Square, "Square"); + GLADE_HOOKUP_OBJECT (Conf, Circle, "Circle"); + GLADE_HOOKUP_OBJECT (Conf, L2, "L2"); + GLADE_HOOKUP_OBJECT (Conf, L1, "L1"); + GLADE_HOOKUP_OBJECT (Conf, Down, "Down"); + GLADE_HOOKUP_OBJECT (Conf, Up, "Up"); + GLADE_HOOKUP_OBJECT (Conf, Left, "Left"); + GLADE_HOOKUP_OBJECT (Conf, Right, "Right"); + GLADE_HOOKUP_OBJECT (Conf, R1, "R1"); + GLADE_HOOKUP_OBJECT (Conf, R2, "R2"); + GLADE_HOOKUP_OBJECT (Conf, Triangle, "Triangle"); + GLADE_HOOKUP_OBJECT (Conf, Cross, "Cross"); + GLADE_HOOKUP_OBJECT (Conf, L3, "L3"); + GLADE_HOOKUP_OBJECT (Conf, checkbutton_reversery, "checkbutton_reversery"); + GLADE_HOOKUP_OBJECT (Conf, checkbutton_reverserx, "checkbutton_reverserx"); + GLADE_HOOKUP_OBJECT (Conf, R_Down, "R_Down"); + GLADE_HOOKUP_OBJECT (Conf, Ry, "Ry"); + GLADE_HOOKUP_OBJECT (Conf, Rx, "Rx"); + GLADE_HOOKUP_OBJECT (Conf, R_Up, "R_Up"); + GLADE_HOOKUP_OBJECT (Conf, R_Right, "R_Right"); + GLADE_HOOKUP_OBJECT (Conf, R_Left, "R_Left"); + GLADE_HOOKUP_OBJECT (Conf, forcefeedback, "forcefeedback"); GLADE_HOOKUP_OBJECT (Conf, checkbutton_reverselx, "checkbutton_reverselx"); GLADE_HOOKUP_OBJECT (Conf, checkbutton_reversely, "checkbutton_reversely"); - GLADE_HOOKUP_OBJECT (Conf, checkbutton_reverserx, "checkbutton_reverserx"); - GLADE_HOOKUP_OBJECT (Conf, checkbutton_reversery, "checkbutton_reversery"); - GLADE_HOOKUP_OBJECT (Conf, forcefeedback, "forcefeedback"); + GLADE_HOOKUP_OBJECT (Conf, L_Down, "L_Down"); + GLADE_HOOKUP_OBJECT (Conf, Ly, "Ly"); + GLADE_HOOKUP_OBJECT (Conf, Lx, "Lx"); + GLADE_HOOKUP_OBJECT (Conf, L_Up, "L_Up"); + GLADE_HOOKUP_OBJECT (Conf, L_Left, "L_Left"); + GLADE_HOOKUP_OBJECT (Conf, L_Right, "L_Right"); + GLADE_HOOKUP_OBJECT (Conf, label3, "label3"); + GLADE_HOOKUP_OBJECT (Conf, Analog, "Analog"); + GLADE_HOOKUP_OBJECT (Conf, label6, "label6"); GLADE_HOOKUP_OBJECT_NO_REF (Conf, dialog_action_area1, "dialog_action_area1"); GLADE_HOOKUP_OBJECT (Conf, cancelbutton1, "cancelbutton1"); GLADE_HOOKUP_OBJECT (Conf, okbutton1, "okbutton1"); diff --git a/plugins/zeropad/Linux/joystick.h b/plugins/zeropad/Linux/joystick.h index 35a2be6057..43cbce9f90 100644 --- a/plugins/zeropad/Linux/joystick.h +++ b/plugins/zeropad/Linux/joystick.h @@ -16,6 +16,9 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + #ifndef __JOYSTICK_H__ + #define __JOYSTICK_H__ + #include #include "zeropad.h" @@ -128,3 +131,4 @@ class JoystickInfo extern int s_selectedpad; extern vector s_vjoysticks; +#endif diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index a49d99c1af..4bbaaf62c2 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -16,38 +16,21 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include -#include -#include - -#define JOYSTICK_SUPPORT -#ifdef JOYSTICK_SUPPORT -#include "joystick.h" -#endif - -#include "zeropad.h" - -extern "C" -{ -#include "interface.h" -#include "support.h" -#include "callbacks.h" -} +#include "linux.h" Display *GSdsp; static pthread_spinlock_t s_mutexStatus; static u32 s_keyPress[2], s_keyRelease[2]; // thread safe -extern GtkWidget *Conf, *s_devicecombo; -extern string s_strIniPath; - static const char* s_pGuiKeyMap[] = { "L2", "R2", "L1", "R1", "Triangle", "Circle", "Cross", "Square", "Select", "L3", "R3", "Start", "Up", "Right", "Down", "Left", - "Lx", "Rx", "Ly", "Ry" + "Lx", "Rx", "Ly", "Ry", + "L_Up", "L_Right", "L_Down", "L_Left", + "R_Up", "R_Right", "R_Down", "R_Left" }; string GetLabelFromButton(const char* buttonname) @@ -108,9 +91,11 @@ int _GetJoystickIdFromPAD(int pad) { for (int i = 0; i < PADKEYS; ++i) { - if (IS_JOYSTICK(conf.keys[(PadEnum[pad][p])][i]) || IS_JOYBUTTONS(conf.keys[(PadEnum[pad][p])][i])) + u32 temp = conf.keys[(PadEnum[pad][p])][i]; + + if (IS_JOYSTICK( temp) || IS_JOYBUTTONS(temp)) { - joyid = PAD_GETJOYID(conf.keys[(PadEnum[pad][p])][i]); + joyid = PAD_GETJOYID(temp); return joyid; } } @@ -134,30 +119,47 @@ void CALLBACK PADupdate(int pad) case KeyPress: key = XLookupKeysym((XKeyEvent *) & E, 0); + i = FindKey(key, pad); #ifdef ANALOG_CONTROLS_HACK - switch (key) + if ((i > PAD_RY) && (i <= PAD_R_LEFT)) { - case KEY_PAD_LX_LEFT: - case KEY_PAD_LY_UP: - case KEY_PAD_RX_LEFT: - case KEY_PAD_RY_UP: - Analog::ConfigurePad(KeypadToPad(key), pad, DEF_VALUE); + switch (i) + { + case PAD_R_LEFT: + Analog::ConfigurePad(PAD_RX, pad, DEF_VALUE); break; - case KEY_PAD_LX_RIGHT: - case KEY_PAD_LY_DOWN: - case KEY_PAD_RX_RIGHT: - case KEY_PAD_RY_DOWN: - Analog::ConfigurePad(KeypadToPad(key), pad, -DEF_VALUE); + case PAD_R_UP: + Analog::ConfigurePad(PAD_RY, pad, DEF_VALUE); + break; + case PAD_L_LEFT: + Analog::ConfigurePad(PAD_LX, pad, DEF_VALUE); + break; + case PAD_L_UP: + Analog::ConfigurePad(PAD_LY, pad, DEF_VALUE); + break; + case PAD_R_DOWN: + Analog::ConfigurePad(PAD_RY, pad, -DEF_VALUE); + break; + case PAD_R_RIGHT: + Analog::ConfigurePad(PAD_RX, pad, -DEF_VALUE); + break; + case PAD_L_DOWN: + Analog::ConfigurePad(PAD_LY, pad, -DEF_VALUE); + break; + case PAD_L_RIGHT: + Analog::ConfigurePad(PAD_LX, pad, -DEF_VALUE); break; } + i += 0xff00; + } #endif - - i = FindKey(key, pad); if (i != -1) { keyPress |= (1 << i); keyRelease &= ~(1 << i); } + //PAD_LOG("Key pressed:%d\n", i); + event.evt = KEYPRESS; event.key = key; break; @@ -165,28 +167,47 @@ void CALLBACK PADupdate(int pad) case KeyRelease: key = XLookupKeysym((XKeyEvent *) & E, 0); + i = FindKey(key, pad); #ifdef ANALOG_CONTROLS_HACK - switch (key) + + if ((i > PAD_RY) && (i <= PAD_R_LEFT)) { - case KEY_PAD_LX_LEFT: - case KEY_PAD_LX_RIGHT: - case KEY_PAD_LY_UP: - case KEY_PAD_LY_DOWN: - case KEY_PAD_RX_LEFT: - case KEY_PAD_RX_RIGHT: - case KEY_PAD_RY_UP: - case KEY_PAD_RY_DOWN: - Analog::ResetPad(KeypadToPad(key)); + switch (i) + { + case PAD_R_LEFT: + Analog::ResetPad(PAD_RX, pad); + break; + case PAD_R_UP: + Analog::ResetPad(PAD_RY, pad); + break; + case PAD_L_LEFT: + Analog::ResetPad(PAD_LX, pad); + break; + case PAD_L_UP: + Analog::ResetPad(PAD_LY, pad); + break; + case PAD_R_DOWN: + Analog::ResetPad(PAD_RY, pad); + break; + case PAD_R_RIGHT: + Analog::ResetPad(PAD_RX, pad); + break; + case PAD_L_DOWN: + Analog::ResetPad(PAD_LY, pad); + break; + case PAD_L_RIGHT: + Analog::ResetPad(PAD_LX, pad); break; } + i += 0xff00; + } #endif - - i = FindKey(key, pad); if (i != -1) { - keyPress &= ~(1 << i); - keyRelease |= (1 << i); + clear_bit(keyPress, i); + set_bit(keyRelease, i); } + event.evt = KEYRELEASE; event.key = key; break; @@ -215,20 +236,20 @@ void CALLBACK PADupdate(int pad) { int joyid = PAD_GETJOYID(key); - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) { pjoy = s_vjoysticks[joyid]; if (SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key))) - status[(pjoy)->GetPAD()] &= ~(1 << i); // pressed + clear_bit(status[(pjoy)->GetPAD()], i); // pressed else - status[(pjoy)->GetPAD()] |= (1 << i); // pressed + set_bit(status[(pjoy)->GetPAD()], i); // pressed } } else if (IS_JOYSTICK(key)) { int joyid = PAD_GETJOYID(key); - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) { pjoy = s_vjoysticks[joyid]; @@ -248,10 +269,29 @@ void CALLBACK PADupdate(int pad) } } } +#ifdef EXPERAMENTAL_POV_CODE else if (IS_POV(key)) { int joyid = PAD_GETJOYID(key); - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) + { + pjoy = s_vjoysticks[joyid]; + + int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); + int pad = (pjoy)->GetPAD(); + + //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVSIGN(key), key); + if PAD_GETPOVSIGN(key) + set_bit(status[pad], i); + else + clear_bit(status[pad], i); + } + } +#else + else if (IS_POV(key)) + { + int joyid = PAD_GETJOYID(key); + if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) { pjoy = s_vjoysticks[joyid]; @@ -259,20 +299,20 @@ void CALLBACK PADupdate(int pad) int pad = (pjoy)->GetPAD(); if (PAD_GETPOVSIGN(key) && (value < -2048)) - status[pad] &= ~(1 << i); + clear_bit(status[pad], i); else if (!PAD_GETPOVSIGN(key) && (value > 2048)) - status[pad] &= ~(1 << i); + clear_bit(status[pad], i); else - status[pad] |= (1 << i); + set_bit(status[pad], i); } } - // Need to add in new POV code here, to match the new configuration pov code. +#endif } #endif pthread_spin_lock(&s_mutexStatus); s_keyPress[pad] |= keyPress; - s_keyPress[pad] &= ~keyRelease; + s_keyPress[pad] &= ~keyRelease; s_keyRelease[pad] |= keyRelease; s_keyRelease[pad] &= ~keyPress; pthread_spin_unlock(&s_mutexStatus); @@ -288,8 +328,7 @@ void UpdateConf(int pad) for (i = 0; i < ArraySize(s_pGuiKeyMap); i++) { - if (s_pGuiKeyMap[i] == NULL) - continue; + if (s_pGuiKeyMap[i] == NULL) continue; Btn = lookup_widget(Conf, GetLabelFromButton(s_pGuiKeyMap[i]).c_str()); if (Btn == NULL) @@ -306,19 +345,43 @@ void UpdateConf(int pad) } else if (IS_JOYBUTTONS(conf.keys[pad][i])) { - tmp.resize(20); + tmp.resize(28); sprintf(&tmp[0], "JBut %d", PAD_GETJOYBUTTON(conf.keys[pad][i])); } else if (IS_JOYSTICK(conf.keys[pad][i])) { - tmp.resize(20); + tmp.resize(28); sprintf(&tmp[0], "JAxis %d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); } else if (IS_POV(conf.keys[pad][i])) +#ifdef EXPERAMENTAL_POV_CODE { - tmp.resize(20); + tmp.resize(28); + switch(PAD_GETPOVSIGN(conf.keys[pad][i])) + { + case SDL_HAT_UP: + sprintf(&tmp[0], "JPOVU-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + break; + + case SDL_HAT_RIGHT: + sprintf(&tmp[0], "JPOVR-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + break; + + case SDL_HAT_DOWN: + sprintf(&tmp[0], "JPOVD-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + break; + + case SDL_HAT_LEFT: + sprintf(&tmp[0], "JPOVL-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + break; + } + } +#else + { + tmp.resize(28); sprintf(&tmp[0], "JPOV %d%s", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i]), PAD_GETPOVSIGN(conf.keys[pad][i]) ? "-" : "+"); } +#endif if (tmp.size() > 0) { @@ -327,13 +390,13 @@ void UpdateConf(int pad) else gtk_entry_set_text(GTK_ENTRY(Btn), "Unknown"); - gtk_object_set_user_data(GTK_OBJECT(Btn), (void*)(PADKEYS*pad + i)); + gtk_object_set_user_data(GTK_OBJECT(Btn), (void*)(PADKEYS * pad + i)); } // check bounds int joyid = _GetJoystickIdFromPAD(pad); - if (joyid < 0 || joyid >= (int)s_vjoysticks.size()) + if ((joyid < 0) || (joyid >= (int)s_vjoysticks.size())) { // get first unused joystick for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) @@ -348,17 +411,20 @@ void UpdateConf(int pad) gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), s_vjoysticks.size()); // no gamepad int padopts = conf.options >> (16 * pad); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverselx")), padopts&PADOPTION_REVERTLX); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversely")), padopts&PADOPTION_REVERTLY); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reverserx")), padopts&PADOPTION_REVERTRX); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "checkbutton_reversery")), padopts&PADOPTION_REVERTRY); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(Conf, "forcefeedback")), padopts&PADOPTION_FORCEFEEDBACK); + + set_checked(Conf, "checkbutton_reverselx", padopts & PADOPTION_REVERTLX); + set_checked(Conf, "checkbutton_reversely", padopts & PADOPTION_REVERTLY); + set_checked(Conf, "checkbutton_reverserx", padopts & PADOPTION_REVERTRX); + set_checked(Conf, "checkbutton_reversery", padopts & PADOPTION_REVERTRY); + set_checked(Conf, "forcefeedback", padopts & PADOPTION_FORCEFEEDBACK); } void OnConf_Key(GtkButton *button, gpointer user_data) { GdkEvent *ev; - GtkWidget* label = lookup_widget(Conf, GetLabelFromButton(gtk_button_get_label(button)).c_str()); + const char* buttonname = gtk_widget_get_name(GTK_WIDGET(button));//gtk_button_get_label(button); + const char* labelname = GetLabelFromButton(buttonname).c_str(); + GtkWidget* label = lookup_widget(Conf, labelname); if (label == NULL) { PAD_LOG("couldn't find correct label\n"); @@ -368,7 +434,8 @@ void OnConf_Key(GtkButton *button, gpointer user_data) int id = (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); int pad = id / PADKEYS; int key = id % PADKEYS; - unsigned long *pkey = &conf.keys[pad][key]; + PAD_LOG("Button = '%s', Label = '%s', id = %d, pad = %d, key = %d\n", buttonname, labelname, id, pad, key); + u32 *pkey = &conf.keys[pad][key]; // save the states #ifdef JOYSTICK_SUPPORT @@ -478,23 +545,23 @@ void OnConf_Key(GtkButton *button, gpointer user_data) char str[32]; case SDL_HAT_UP: - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - sprintf(str, "JPOV%d%s", i, "U"); + *pkey = PAD_POV((*itjoy)->GetId(), value, i); + sprintf(str, "JPOVU-%d", i); gtk_entry_set_text(GTK_ENTRY(label), str); return; case SDL_HAT_RIGHT: - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - sprintf(str, "JPOV%d%s", i, "R"); + *pkey = PAD_POV((*itjoy)->GetId(), value, i); + sprintf(str, "JPOVR-%d", i); gtk_entry_set_text(GTK_ENTRY(label), str); return; case SDL_HAT_DOWN: - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - sprintf(str, "JPOV%d%s", i, "D"); + *pkey = PAD_POV((*itjoy)->GetId(), value, i); + sprintf(str, "JPOVD-%d", i); gtk_entry_set_text(GTK_ENTRY(label), str); return; case SDL_HAT_LEFT: - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - sprintf(str, "JPOV%d%s", i, "L"); + *pkey = PAD_POV((*itjoy)->GetId(), value, i); + sprintf(str, "JPOVL-%d", i); gtk_entry_set_text(GTK_ENTRY(label), str); return; // Not handling SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, diff --git a/plugins/zeropad/Linux/linux.h b/plugins/zeropad/Linux/linux.h new file mode 100644 index 0000000000..e121503a1d --- /dev/null +++ b/plugins/zeropad/Linux/linux.h @@ -0,0 +1,42 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include + +#define JOYSTICK_SUPPORT +#ifdef JOYSTICK_SUPPORT +#include "joystick.h" +#endif + +#include "zeropad.h" + +extern "C" +{ +#include "interface.h" +#include "support.h" +#include "callbacks.h" +} + +extern GtkWidget *Conf, *s_devicecombo; +extern string s_strIniPath; + + +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) diff --git a/plugins/zeropad/Linux/zeropad.glade b/plugins/zeropad/Linux/zeropad.glade index 4b1b936296..c8fc069ca7 100644 --- a/plugins/zeropad/Linux/zeropad.glade +++ b/plugins/zeropad/Linux/zeropad.glade @@ -271,14 +271,14 @@ and SSSPSXPad, TwinPad authors 0 - True + False True 0 - True + False True @@ -303,7 +303,7 @@ and SSSPSXPad, TwinPad authors 64 - 40 + 32 @@ -323,7 +323,7 @@ and SSSPSXPad, TwinPad authors 64 - 104 + 80 @@ -342,8 +342,8 @@ and SSSPSXPad, TwinPad authors False - 168 - 64 + 200 + 184 @@ -362,8 +362,8 @@ and SSSPSXPad, TwinPad authors False - 280 - 64 + 272 + 184 @@ -383,7 +383,7 @@ and SSSPSXPad, TwinPad authors 64 - 168 + 136 @@ -403,7 +403,7 @@ and SSSPSXPad, TwinPad authors 128 - 208 + 184 @@ -423,7 +423,7 @@ and SSSPSXPad, TwinPad authors 0 - 208 + 184 @@ -443,7 +443,7 @@ and SSSPSXPad, TwinPad authors 64 - 256 + 232 @@ -462,7 +462,7 @@ and SSSPSXPad, TwinPad authors False - 368 + 408 32 @@ -482,8 +482,8 @@ and SSSPSXPad, TwinPad authors False - 368 - 104 + 408 + 80 @@ -502,8 +502,8 @@ and SSSPSXPad, TwinPad authors False - 224 - 128 + 240 + 232 @@ -522,8 +522,8 @@ and SSSPSXPad, TwinPad authors False - 304 - 208 + 344 + 184 @@ -542,8 +542,8 @@ and SSSPSXPad, TwinPad authors False - 368 - 168 + 408 + 136 @@ -562,8 +562,8 @@ and SSSPSXPad, TwinPad authors False - 432 - 208 + 472 + 184 @@ -582,8 +582,8 @@ and SSSPSXPad, TwinPad authors False - 368 - 256 + 408 + 232 @@ -603,7 +603,7 @@ and SSSPSXPad, TwinPad authors 272 - 272 + 56 @@ -622,14 +622,14 @@ and SSSPSXPad, TwinPad authors False - 176 - 272 + 200 + 56 - 80 + 64 24 True True @@ -642,14 +642,14 @@ and SSSPSXPad, TwinPad authors False - 272 - 352 + 408 + 384 - 80 + 64 24 True True @@ -662,14 +662,14 @@ and SSSPSXPad, TwinPad authors False - 160 - 352 + 64 + 384 - 80 + 64 24 True True @@ -682,14 +682,14 @@ and SSSPSXPad, TwinPad authors False - 272 - 416 + 408 + 432 - 80 + 64 24 True True @@ -702,15 +702,193 @@ and SSSPSXPad, TwinPad authors False - 160 - 416 + 64 + 432 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 64 + 336 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 0 + 408 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 128 + 408 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 64 + 480 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 408 + 336 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 344 + 408 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 472 + 408 + + + + + + 64 + 24 + True + True + True + True + 0 + + True + * + False + + + 408 + 480 + + + + + + 64 + 24 + True + True + R3 + True + GTK_RELIEF_NORMAL + True + + + + 272 + 32 64 - 32 + 24 True True Select @@ -720,105 +898,15 @@ and SSSPSXPad, TwinPad authors - 168 - 32 - - - - - - 64 - 32 - True - True - L1 - True - GTK_RELIEF_NORMAL - True - - - - 64 - 72 - - - - - - 64 - 32 - True - True - Left - True - GTK_RELIEF_NORMAL - True - - - - 0 - 176 - - - - - - 64 - 32 - True - True - Down - True - GTK_RELIEF_NORMAL - True - - - - 64 - 224 - - - - - - 64 - 32 - True - True - Analog - True - GTK_RELIEF_NORMAL - True - - - - 224 - 96 - - - - - - 64 - 32 - True - True - R2 - True - GTK_RELIEF_NORMAL - True - - - - 368 - 0 + 200 + 160 64 - 32 + 24 True True Start @@ -828,87 +916,15 @@ and SSSPSXPad, TwinPad authors - 280 - 32 - - - - - - 64 - 32 - True - True - R1 - True - GTK_RELIEF_NORMAL - True - - - - 368 - 72 - - - - - - 64 - 32 - True - True - Cross - True - GTK_RELIEF_NORMAL - True - - - - 368 - 224 - - - - - - 64 - 32 - True - True - Triangle - True - GTK_RELIEF_NORMAL - True - - - - 368 - 136 - - - - - - 64 - 32 - True - True - Circle - True - GTK_RELIEF_NORMAL - True - - - - 432 - 176 + 272 + 160 64 - 32 + 24 True True Square @@ -918,51 +934,33 @@ and SSSPSXPad, TwinPad authors - 304 - 176 + 344 + 160 - + 64 - 32 + 24 True True - L3 + Circle True GTK_RELIEF_NORMAL True - + - 176 - 240 - - - - - - 64 - 32 - True - True - Up - True - GTK_RELIEF_NORMAL - True - - - - 64 - 136 + 472 + 160 64 - 32 + 24 True True L2 @@ -978,81 +976,81 @@ and SSSPSXPad, TwinPad authors - - 80 - 32 + + 64 + 24 True True - Lx + L1 True GTK_RELIEF_NORMAL True - + - 160 - 320 + 64 + 56 - - 80 - 32 + + 64 + 24 True True - Ly + Down True GTK_RELIEF_NORMAL True - + - 160 - 384 + 64 + 208 - - 80 - 32 + + 64 + 24 True True - Rx + Up True GTK_RELIEF_NORMAL True - + - 272 - 320 + 64 + 112 - - 80 - 32 + + 64 + 24 True True - Ry + Left True GTK_RELIEF_NORMAL True - + - 272 - 384 + 0 + 160 64 - 32 + 24 True True Right @@ -1063,114 +1061,97 @@ and SSSPSXPad, TwinPad authors 128 - 176 + 160 - - 408 - 16 - True - Analog Controls (move mouse or analog joystick to select) - False - False - GTK_JUSTIFY_LEFT - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - True - 0 - - - 48 - 296 - - - - - + 64 - 32 + 24 True True - R3 + R1 True GTK_RELIEF_NORMAL True - + - 272 - 240 + 408 + 56 - - 111 - 22 + + 64 + 24 True True - Reverse LX + R2 True GTK_RELIEF_NORMAL True - False - False - True - + - 16 - 328 + 408 + 8 - - 111 - 22 + + 64 + 24 True True - Reverse LY + Triangle True GTK_RELIEF_NORMAL True - False - False - True - + - 16 - 352 + 408 + 112 - - 111 - 22 + + 64 + 24 True True - Reverse RX + Cross True GTK_RELIEF_NORMAL True - False - False - True - + - 16 - 376 + 408 + 208 + + + + + + 64 + 24 + True + True + L3 + True + GTK_RELIEF_NORMAL + True + + + + 200 + 32 @@ -1190,8 +1171,137 @@ and SSSPSXPad, TwinPad authors - 16 - 400 + 384 + 536 + + + + + + 111 + 22 + True + True + Reverse RX + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + 384 + 512 + + + + + + 64 + 24 + True + True + Down + True + GTK_RELIEF_NORMAL + True + + + + 408 + 456 + + + + + + 64 + 24 + True + True + Ry + True + GTK_RELIEF_NORMAL + True + + + + 408 + 408 + + + + + + 64 + 24 + True + True + Rx + True + GTK_RELIEF_NORMAL + True + + + + 408 + 360 + + + + + + 64 + 24 + True + True + Up + True + GTK_RELIEF_NORMAL + True + + + + 408 + 312 + + + + + + 64 + 24 + True + True + Right + True + GTK_RELIEF_NORMAL + True + + + + 472 + 384 + + + + + + 64 + 24 + True + True + Left + True + GTK_RELIEF_NORMAL + True + + + + 344 + 384 @@ -1212,8 +1322,228 @@ Feedback - 384 - 320 + 216 + 512 + + + + + + 111 + 22 + True + True + Reverse LX + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + 48 + 512 + + + + + + 111 + 22 + True + True + Reverse LY + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + 48 + 536 + + + + + + 64 + 24 + True + True + Down + True + GTK_RELIEF_NORMAL + True + + + + 64 + 456 + + + + + + 64 + 24 + True + True + Ly + True + GTK_RELIEF_NORMAL + True + + + + 64 + 408 + + + + + + 64 + 24 + True + True + Lx + True + GTK_RELIEF_NORMAL + True + + + + 64 + 360 + + + + + + 64 + 24 + True + True + Up + True + GTK_RELIEF_NORMAL + True + + + + 64 + 312 + + + + + + 64 + 24 + True + True + Left + True + GTK_RELIEF_NORMAL + True + + + + 0 + 384 + + + + + + 64 + 24 + True + True + Right + True + GTK_RELIEF_NORMAL + True + + + + 128 + 384 + + + + + + 400 + 24 + True + Analog Controls (move mouse or analog joystick to select) + False + False + GTK_JUSTIFY_LEFT + True + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + True + 0 + + + 72 + 256 + + + + + + 64 + 24 + True + True + Analog + True + GTK_RELIEF_NORMAL + True + + + + 240 + 208 + + + + + + 448 + 32 + True + (Note: The analog key controls and joystick controls do not work well together currently, and should not be mixed. ) + False + False + GTK_JUSTIFY_CENTER + True + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 56 + 280 diff --git a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj index 957161cb48..d5927633ca 100644 --- a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj +++ b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj @@ -280,6 +280,10 @@ RelativePath="..\analog.h" > + + + static void __forceinline set_bit(T &value, int bit) + { + value |= (1 << bit); + } + + template + static void __forceinline clear_bit(T &value, int bit) + { + value &= ~(1 << bit); + } + + template + static void __forceinline toggle_bit(T &value, int bit) + { + value ^= (1 << bit); + } + + template + static bool __forceinline test_bit(T &value, int bit) + { + return (value & (1 << bit)); + } \ No newline at end of file diff --git a/plugins/zeropad/zeropad.cpp b/plugins/zeropad/zeropad.cpp index 454ae40495..b67549fc54 100644 --- a/plugins/zeropad/zeropad.cpp +++ b/plugins/zeropad/zeropad.cpp @@ -39,11 +39,12 @@ keyEvent event; u16 status[2]; int pressure; +static keyEvent s_event; string s_strIniPath = "inis/zeropad.ini"; -const unsigned char version = PS2E_PAD_VERSION; -const unsigned char revision = 0; -const unsigned char build = 3; // increase that with each version +const u32 version = PS2E_PAD_VERSION; +const u32 revision = 0; +const u32 build = 3; // increase that with each version int PadEnum[2][2] = {{0, 2}, {1, 3}}; @@ -121,73 +122,6 @@ int cmdLen; int ds2mode = 0; // DS Mode at start FILE *padLog = NULL; -int POV(u32 direction, u32 angle) -{ - if ((direction == 0) && (angle >= 0) && (angle < 4500)) return 1;//forward - if ((direction == 2) && (angle >= 4500) && (angle < 13500)) return 1;//right - if ((direction == 1) && (angle >= 13500) && (angle < 22500)) return 1;//backward - if ((direction == 3) && (angle >= 22500) && (angle < 31500)) return 1;//left - if ((direction == 0) && (angle >= 31500) && (angle < 36000)) return 1;//forward - return 0; -} - -void _KeyPress(int pad, u32 key) -{ - int i; - -#ifdef _WIN32 - for (i=0; i> 8) +#ifdef __LINUX__ +#define PADKEYS 28 +#else #define PADKEYS 20 +#endif #define PADOPTION_FORCEFEEDBACK 1 #define PADOPTION_REVERTLX 0x2 @@ -79,13 +84,14 @@ extern char libraryName[256]; #define PADSUBKEYS 2 #endif +//#define EXPERAMENTAL_POV_CODE extern int PadEnum[2][2]; typedef struct { - unsigned long keys[2 * PADSUBKEYS][PADKEYS]; - int log; - int options; // upper 16 bits are for pad2 + u32 keys[2 * PADSUBKEYS][PADKEYS]; + u32 log; + u32 options; // upper 16 bits are for pad2 } PADconf; typedef struct @@ -117,6 +123,14 @@ enum PadCommands enum gamePadValues { + PAD_R_LEFT = 27, + PAD_R_DOWN = 26, + PAD_R_RIGHT = 25, + PAD_R_UP = 24, + PAD_L_LEFT = 23, + PAD_L_DOWN = 22, + PAD_L_RIGHT = 21, + PAD_L_UP = 20, PAD_RY = 19, PAD_LY = 18, PAD_RX = 17, @@ -141,20 +155,11 @@ enum gamePadValues // Put in the code for bolche's analog contols hack, ifdeffed out, so I don't forget to // add a gui some day and activate it. -//#define ANALOG_CONTROLS_HACK -// The various KEY_PAD_xxx definitions are defined as the value of whatever key is pressed. +#define ANALOG_CONTROLS_HACK // DEF_VALUE is the strength you press the control. // Code taken from http://forums.pcsx2.net/thread-4699.html #ifdef ANALOG_CONTROLS_HACK -#define KEY_PAD_RY_UP 0xff50 -#define KEY_PAD_RY_DOWN 0xff57 -#define KEY_PAD_LY_UP 0xff52 -#define KEY_PAD_LY_DOWN 0xff54 -#define KEY_PAD_RX_LEFT 0xffff -#define KEY_PAD_RX_RIGHT 0xff56 -#define KEY_PAD_LX_LEFT 0xff51 -#define KEY_PAD_LX_RIGHT 0xff53 #define DEF_VALUE 32766 #endif From d19229d334df5f01b48394887415829901a1af84 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 11 May 2009 01:39:33 +0000 Subject: [PATCH 043/296] microVU: - Major flag algorithm rewrites (work in progress) - Abused macros to create my tables, knocking off 800 lines of code. - VU0 and VU1 have different dynarec cache sizes now (vu0's is 1/4 the size of vu1's) - Minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1166 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 1 + pcsx2/x86/microVU.h | 9 +- pcsx2/x86/microVU_Alloc.h | 5 +- pcsx2/x86/microVU_Analyze.inl | 9 +- pcsx2/x86/microVU_Compile.inl | 39 +- pcsx2/x86/microVU_Execute.inl | 4 + pcsx2/x86/microVU_Flags.inl | 219 ++++--- pcsx2/x86/microVU_Lower.inl | 12 +- pcsx2/x86/microVU_Misc.h | 9 + pcsx2/x86/microVU_Tables.inl | 1152 +++++---------------------------- 10 files changed, 376 insertions(+), 1083 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index eda6882dd4..f53d540dd4 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -43,6 +43,7 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr) { mVU->microSize = (vuIndex ? 0x4000 : 0x1000); mVU->progSize = (vuIndex ? 0x4000 : 0x1000) / 4; mVU->cache = NULL; + mVU->cacheSize = mVUcacheSize; memset(&mVU->prog, 0, sizeof(mVU->prog)); mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7a6c1d3809..19e23990fe 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -49,15 +49,15 @@ public: } microBlock* search(microRegInfo* pState) { if (listSize < 0) return NULL; - if (blockList[0].pState.needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) + if (pState->needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) for (int i = 0; i <= listSize; i++) { - if (!memcmp(pState, &blockList[i].pState, sizeof(microRegInfo) - 1)) return &blockList[i]; + if (!memcmp(pState, &blockList[i].pState, sizeof(microRegInfo)/* - 4*/)) return &blockList[i]; } } else { // Can do Simple Search (Only Matches the Important Pipeline Stuff) for (int i = 0; i <= listSize; i++) { if ((blockList[i].pState.q == pState->q) && (blockList[i].pState.p == pState->p) - && (blockList[i].pState.clip == pState->clip)) { return &blockList[i]; } + && (blockList[i].pState.flags == pState->flags)) { return &blockList[i]; } } } return NULL; @@ -87,11 +87,12 @@ struct microProgManager { microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; +#define mVUcacheSize (0x1f00000 / ((vuIndex) ? 1 : 4)) struct microVU { u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size u32 progSize; // VU Micro Program Size (microSize/4) - static const u32 cacheSize = 0xb00000; // VU Cache Size + u32 cacheSize; // VU Cache Size microProgManager<0x4000> prog; // Micro Program Data diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_Alloc.h index 9dbb5521d1..8e50e24f73 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_Alloc.h @@ -39,8 +39,8 @@ struct microRegInfo { u8 p; u8 r; u8 xgkick; - u8 clip; - u8 needExactMatch; // If set, block needs an exact match of pipeline state (needs to be last byte in struct) + u8 flags; // clip x2 :: status x2 + u32 needExactMatch; // If set, block needs an exact match of pipeline state (needs to be last 2 bytes in struct) #if defined(_MSC_VER) }; #pragma pack(pop) @@ -75,6 +75,7 @@ struct microAllocInfo { u32 count; // Number of VU 64bit instructions ran (starts at 0 for each block) u32 curPC; // Current PC u32 startPC; // Start PC for Cur Block + u32 flagInfo; // Holds information to help with flag instances on block linking u32 info[pSize/8]; // Info for Instructions in current block u8 stall[pSize/8]; // Info on how much each instruction stalled (stores the max amount of cycles to stall for the current opcodes) }; diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 96943fb0e6..cf136c9a0e 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -249,9 +249,9 @@ microVUt(void) mVUanalyzeR2(int Ft, bool canBeNOP) { microVUt(void) mVUanalyzeSflag(int It) { microVU* mVU = mVUx; if (!It) { mVUinfo |= _isNOP; } - else { // Sets _isSflag at instruction that FSxxx opcode reads it's status flag from + else { mVUinfo |= _swapOps; - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf /*<< mVUcount*/; } if (mVUcount >= 1) { incPC2(-2); mVUinfo |= _isSflag; incPC2(2); } // Note: _isSflag is used for status flag optimizations. // Do to stalls, it can only be set one instruction prior to the status flag read instruction @@ -279,7 +279,7 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { if (!It) { mVUinfo |= _isNOP; } else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo |= _swapOps; - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } int curPC = iPC; for (int i = mVUcount, j = 0; i > 1; i--, j++) { incPC2(-2); @@ -297,7 +297,8 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { microVUt(void) mVUanalyzeCflag() { microVU* mVU = mVUx; - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch = 1; } + mVUinfo |= _swapOps; + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index e11280b892..ace1301b7e 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -24,8 +24,7 @@ //------------------------------------------------------------------ #define branchCase(JMPcc, nJMPcc) \ - mVUsetupBranch(bStatus, bMac); \ - mVUprint("mVUcompile branchCase"); \ + mVUsetupBranch(xStatus, xMac, xClip, xCycles); \ CMP16ItoM((uptr)&mVU->branch, 0); \ incPC2(1); \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ @@ -46,12 +45,12 @@ //------------------------------------------------------------------ // Recompiles Code for Proper Flags and Q/P regs on Block Linkings -microVUt(void) mVUsetupBranch(int* bStatus, int* bMac) { +microVUt(void) mVUsetupBranch(int* xStatus, int* xMac, int* xClip, int xCycles) { microVU* mVU = mVUx; mVUprint("mVUsetupBranch"); // Shuffle Flag Instances - mVUsetupFlags(bStatus, bMac); + mVUsetupFlags(xStatus, xMac, xClip, xCycles); // Shuffle P/Q regs since every block starts at instance #0 if (mVU->p || mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, shufflePQ); } @@ -105,7 +104,7 @@ microVUt(void) mVUsetCycles() { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUendProgram(int fStatus, int fMac) { +microVUt(void) mVUendProgram(int fStatus, int fMac, int fClip) { microVU* mVU = mVUx; incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; @@ -121,8 +120,10 @@ microVUt(void) mVUendProgram(int fStatus, int fMac) { // Save Flag Instances getFlagReg(fStatus, fStatus); mVUallocMFLAGa(gprT1, fMac); + mVUallocCFLAGa(gprT2, fClip); MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); + MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) @@ -133,12 +134,15 @@ microVUt(void) mVUendProgram(int fStatus, int fMac) { JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); } +#define sI ((mVUpBlock->pState.needExactMatch & 0x000f) ? 0 : ((mVUpBlock->pState.flags >> 0) & 3)) +#define cI ((mVUpBlock->pState.needExactMatch & 0x0f00) ? 0 : ((mVUpBlock->pState.flags >> 2) & 3)) + microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; iPC = mVUstartPC; CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); - mVUendProgram(0, 0); + mVUendProgram(sI, 0, cI); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } @@ -171,8 +175,10 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUblock.x86ptrStart = thisPtr; pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager mVUpBlock = pBlock; + mVUregs.flags = 0; + mVUflagInfo = 0; - for (int branch = 0;; ) { + for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); mVUinfo = 0; incCycles(1); @@ -186,20 +192,21 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (mVU->p) { mVUinfo |= _readP; } if (mVU->q) { mVUinfo |= _readQ; } else { mVUinfo |= _writeQ; } - if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); if (mVUbranch) { Console::Error("microVU Warning: Branch in E-bit/Branch delay slot!"); mVUinfo |= _isNOP; } break; } + if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; if (mVUbranch) { Console::Error("microVU Warning: Branch in E-bit/Branch delay slot!"); mVUinfo |= _isNOP; } break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } + if (mVUbranch) { mVUsetFlagInfo(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } incPC(1); mVUcount++; } // Sets Up Flag instances - int bStatus[4]; int bMac[4]; - mVUsetFlags(bStatus, bMac); + int xStatus[4], xMac[4], xClip[4]; + int xCycles = mVUsetFlags(xStatus, xMac, xClip); mVUtestCycles(); //SysPrintf("bS[0] = %08x, bS[1] = %08x, bS[2] = %08x, bS[3] = %08x\n", bStatus[0], bStatus[1], bStatus[2], bStatus[3]); //SysPrintf("bM[0] = %08x, bM[1] = %08x, bM[2] = %08x, bM[3] = %08x\n", bMac[0], bMac[1], bMac[2], bMac[3]); //SysPrintf("mVUcount = %d\n", mVUcount); + //SysPrintf("mVUflagInfo = %d\n", mVUflagInfo); // Second Pass iPC = mVUstartPC; @@ -227,7 +234,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile B/BAL"); incPC(-3); // Go back to branch opcode (to get branch imm addr) - mVUsetupBranch(bStatus, bMac); + mVUsetupBranch(xStatus, xMac, xClip, xCycles); // Check if branch-block has already been compiled pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); @@ -239,7 +246,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile JR/JALR"); memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - mVUsetupBranch(bStatus, bMac); + mVUsetupBranch(xStatus, xMac, xClip, xCycles); mVUbackupRegs(); MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) @@ -264,7 +271,8 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { } else { uptr jumpAddr; - u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs, so back them up + u32 bPC = iPC; // mVUcompile can modify iPC, mVUregs, and mVUflagInfo, so back them up + u32 bFlagInfo = mVUflagInfo; memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); incPC2(1); // Get PC for branch not-taken @@ -272,6 +280,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { else mVUcompileVU1(xPC, (uptr)&mVUregs); iPC = bPC; + mVUflagInfo = bFlagInfo; incPC(-3); // Go back to branch opcode (to get branch imm addr) if (!vuIndex) jumpAddr = (uptr)mVUcompileVU0(branchAddr, (uptr)&pBlock->pStateEnd); else jumpAddr = (uptr)mVUcompileVU1(branchAddr, (uptr)&pBlock->pStateEnd); @@ -284,7 +293,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } // Do E-bit end stuff here - mVUendProgram(bStatus[3], bMac[3]); + mVUendProgram(findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); return thisPtr; //ToDo: Save pipeline state? } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index edfd6130a2..d207f157fa 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -57,6 +57,10 @@ microVUt(void) mVUdispatcherA() { SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, xmmT1); + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); + SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, xmmT1); + SSE_MOVAPS_M128_to_XMM(xmmACC, (uptr)&mVU->regs->ACC.UL[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); SSE_MOVAPS_M128_to_XMM(xmmMin, (uptr)mVU_minvals); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index b359f37fc5..20d1e8a81b 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -19,20 +19,6 @@ #pragma once #ifdef PCSX2_MICROVU -// ToDo: Fix this properly. -#define flagSetMacro(xFlag, pFlag, xF, yF, zF) { \ - yF += (mVUstall > 3) ? 3 : mVUstall; \ - if (yF > zF) { \ - pFlag += (yF-zF); \ - if (pFlag >= xFlag) pFlag = (xFlag-1); \ - zF++; \ - xF = (yF-zF); \ - zF = yF; \ - yF -= xF; \ - } \ - yF++; \ -} - // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet() { microVU* mVU = mVUx; @@ -69,90 +55,177 @@ microVUt(void) mVUstatusFlagOp() { iPC = curPC; } +int findFlagInst(int* fFlag, int cycles) { + int j = 0, jValue = -1; + for (int i = 0; i < 4; i++) { + if ((fFlag[i] <= cycles) && (fFlag[i] > jValue)) { j = i; jValue = fFlag[i]; } + } + return j; +} + +// Setup Last 4 instances of Status/Mac/Clip flags (needed for accurate block linking) +void sortFlag(int* fFlag, int* bFlag, int cycles) { + for (int i = 0; i < 4; i++) { + bFlag[i] = findFlagInst(fFlag, cycles); + cycles++; + } +} + // Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! -microVUt(void) mVUsetFlags(int* bStatus, int* bMac) { +microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { microVU* mVU = mVUx; - // Ensure last ~4+ instructions update mac flags int endPC = iPC; - u32 aCount = 1; // Amount of instructions needed to get 4 valid status/mac flag instances - for (int i = mVUcount, iX = 0; i > 0; i--, aCount++) { - if (doStatus) { mVUinfo |= _doMac; iX++; if ((iX >= 4) || (aCount > 4)) { break; } } + u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking + + // Ensure last ~4+ instructions update mac flags + for (int i = mVUcount; i > 0; i--, aCount++) { + if (doStatus) { if (__Mac) { mVUinfo |= _doMac; } if (aCount > 4) { break; } } incPC2(-2); } // Status/Mac Flags Setup Code - int xStatus = 8, xMac = 8; // Flag Instances start at #0 on every block ((8&3) == 0) - int pStatus = 3, pMac = 3; - int xClip = mVUregs.clip + 8, pClip = mVUregs.clip + 7; // Clip Instance starts from where it left off - int xS = 0, yS = 1, zS = 0; - int xM = 0, yM = 1, zM = 0; - int xC = 0, yC = 1, zC = 0; + int xS = 0, xM = 0, xC = 0; + for (int i = 0; i < 4; i++) { + xStatus[i] = i; + xMac [i] = i; + xClip [i] = i; + } + + if (!(mVUpBlock->pState.needExactMatch & 0x00f)) { + xS = (mVUpBlock->pState.flags >> 0) & 3; + xStatus[0] = -1; xStatus[1] = -1; + xStatus[2] = -1; xStatus[3] = -1; + xStatus[(xS-1)&3] = 0; + } + + if (!(mVUpBlock->pState.needExactMatch & 0xf00)) { + xC = (mVUpBlock->pState.flags >> 2) & 3; + xClip[0] = -1; xClip[1] = -1; + xClip[2] = -1; xClip[3] = -1; + xClip[(xC-1)&3] = 0; + } + + if (!(mVUpBlock->pState.needExactMatch & 0x0f0)) { + xMac[0] = -1; xMac[1] = -1; + xMac[2] = -1; xMac[3] = -1; + } + + int cycles = 0; u32 xCount = mVUcount; // Backup count iPC = mVUstartPC; for (mVUcount = 0; mVUcount < xCount; mVUcount++) { - if (((xCount - mVUcount) > aCount) && isFSSET) mVUstatusFlagOp(); // Don't Optimize out on the last ~4+ instructions + if (isFSSET) { + if (__Status) { // Don't Optimize out on the last ~4+ instructions + if ((xCount - mVUcount) > aCount) { mVUstatusFlagOp(); } + } + else mVUstatusFlagOp(); + } + cycles += mVUstall; - flagSetMacro(xStatus, pStatus, xS, yS, zS); // Handles _fvsinstances - flagSetMacro(xMac, pMac, xM, yM, zM); // Handles _fvminstances - flagSetMacro(xClip, pClip, xC, yC, zC); // Handles _fvcinstances + mVUinfo |= findFlagInst(xStatus, cycles) << 18; // _fvsInstance + mVUinfo |= findFlagInst(xMac, cycles) << 16; // _fvmInstance + mVUinfo |= findFlagInst(xClip, cycles) << 20; // _fvcInstance - mVUinfo |= (xStatus&3) << 12; // _fsInstance - mVUinfo |= (xMac&3) << 10; // _fmInstance - mVUinfo |= (xClip&3) << 14; // _fcInstance + mVUinfo |= (xS & 3) << 12; // _fsInstance + mVUinfo |= (xM & 3) << 10; // _fmInstance + mVUinfo |= (xC & 3) << 14; // _fcInstance - mVUinfo |= (pStatus&3) << 18; // _fvsInstance - mVUinfo |= (pMac&3) << 16; // _fvmInstance - mVUinfo |= (pClip&3) << 20; // _fvcInstance + if (doStatus||isFSSET||doDivFlag) { xStatus [xS] = cycles + 4; xS = (xS+1)&3; } + if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1)&3; } + if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1)&3; } - if (doStatus||isFSSET||doDivFlag) { xStatus = (xStatus+1); } - if (doMac) { xMac = (xMac+1); } - if (doClip) { xClip = (xClip+1); } + cycles++; incPC2(2); } - mVUcount = xCount; // Restore count - mVUregs.clip = xClip&3; // Note: Clip timing isn't cycle-accurate between block linking; but hopefully doesn't matter - // Temp Hack-fix until flag-algorithm rewrite - for (int i = 0; i < 4; i++) { - bStatus[i] = 0; - bMac[i] = 0; - } - - // Setup Last 4 instances of Status/Mac flags (needed for accurate block linking) - iPC = endPC; - for (int i = 3, j = 3, ii = 1, jj = 1; aCount > 0; ii++, jj++, aCount--) { - if ((doStatus||isFSSET||doDivFlag) && (i >= 0)) { - for (; (ii > 0 && i >= 0); i--, ii--) { xStatus = (xStatus-1) & 3; bStatus[i] = xStatus; } - } - if (doMac && (j >= 0)) { - for (; (jj > 0 && j >= 0); j--, jj--) { xMac = (xMac-1) & 3; bMac[j] = xMac; } - } - incPC2(-2); - } - iPC = endPC; + mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) | ((__Status) ? 0 : xS); + return cycles; } #define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) -#define shuffleMac ((bMac[3]<<6)|(bMac[2]<<4)|(bMac[1]<<2)|bMac[0]) +#define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) +#define shuffleClip ((bClip[3]<<6)|(bClip[2]<<4)|(bClip[1]<<2)|bClip[0]) // Recompiles Code for Proper Flags on Block Linkings -microVUt(void) mVUsetupFlags(int* bStatus, int* bMac) { +microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { microVU* mVU = mVUx; - PUSH32R(gprR); // Backup gprR - MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); - MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); - MOV32RtoR(gprR, getFlagReg1(bStatus[2])); - MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); - MOV32RtoR(gprF0, gprT1); - MOV32RtoR(gprF1, gprT2); - MOV32RtoR(gprF2, gprR); - POP32R(gprR); // Restore gprR + if (__Status) { + int bStatus[4]; + sortFlag(xStatus, bStatus, cycles); + PUSH32R(gprR); // Backup gprR + MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); + MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); + MOV32RtoR(gprR, getFlagReg1(bStatus[2])); + MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, gprT1); + MOV32RtoR(gprF1, gprT2); + MOV32RtoR(gprF2, gprR); + POP32R(gprR); // Restore gprR + } - SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->macFlag); - SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleMac); - SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); + if (__Mac) { + int bMac[4]; + sortFlag(xMac, bMac, cycles); + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->macFlag); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleMac); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); + } + + if (__Clip) { + int bClip[4]; + sortFlag(xClip, bClip, cycles); + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->clipFlag); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleClip); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->clipFlag, xmmT1); + } } -#endif //PCSX2_MICROVU \ No newline at end of file +microVUt(void) mVUpass4(int startPC) { + + microVU* mVU = mVUx; + int oldPC = iPC; + int oldCount = mVUcount; + int oldBranch = mVUbranch; + iPC = startPC / 4; + mVUcount = 0; + mVUbranch = 0; + for (int branch = 0; mVUcount < 4; mVUcount++) { + incPC(1); + if ( curI & _Ebit_ ) { branch = 1; } + if ( curI & _MDTbit_ ) { branch = 2; } + if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(); incPC(1); } + if (branch >= 2) { break; } + else if (branch == 1) { branch = 2; } + if (mVUbranch) { branch = 3; mVUbranch = 0; } + incPC(1); + } + iPC = oldPC; + mVUcount = oldCount; + mVUbranch = oldBranch; + setCode(); +} + +#define branchType1 if (mVUbranch <= 2) // B/BAL +#define branchType2 else if (mVUbranch >= 9) // JR/JALR +#define branchType3 else // Conditional Branch + +microVUt(void) mVUsetFlagInfo() { + microVU* mVU = mVUx; + branchType1 { incPC(-1); mVUpass4(branchAddr); incPC(1); } + branchType2 { mVUflagInfo |= 0xffffffff; } + branchType3 { + incPC(-1); + mVUpass4(branchAddr); + int backupFlagInfo = mVUflagInfo; + mVUflagInfo = 0; + incPC(4); // Branch Not Taken + mVUpass4(xPC); + incPC(-3); + //if (mVUflagInfo != backupFlagInfo) { mVUflagInfo |= __NeedExact; } + mVUflagInfo |= backupFlagInfo; + } +} + +#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 781c3a5193..615afeeb11 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -440,6 +440,7 @@ microVUf(void) mVU_FCAND() { mVUallocVIb(gprT1, 1); } pass3 { mVUlog("FCAND vi01, $%x", _Imm24_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } microVUf(void) mVU_FCEQ() { @@ -453,6 +454,7 @@ microVUf(void) mVU_FCEQ() { mVUallocVIb(gprT1, 1); } pass3 { mVUlog("FCEQ vi01, $%x", _Imm24_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } microVUf(void) mVU_FCGET() { @@ -463,7 +465,8 @@ microVUf(void) mVU_FCGET() { AND32ItoR(gprT1, 0xfff); mVUallocVIb(gprT1, _It_); } - pass3 { mVUlog("FCGET vi%02d", _Ft_); } + pass3 { mVUlog("FCGET vi%02d", _Ft_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } microVUf(void) mVU_FCOR() { @@ -477,6 +480,7 @@ microVUf(void) mVU_FCOR() { mVUallocVIb(gprT1, 1); } pass3 { mVUlog("FCOR vi01, $%x", _Imm24_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } microVUf(void) mVU_FCSET() { @@ -503,6 +507,7 @@ microVUf(void) mVU_FMAND() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMAND vi%02d, vi%02d", _Ft_, _Fs_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } microVUf(void) mVU_FMEQ() { @@ -517,6 +522,7 @@ microVUf(void) mVU_FMEQ() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMEQ vi%02d, vi%02d", _Ft_, _Fs_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } microVUf(void) mVU_FMOR() { @@ -529,6 +535,7 @@ microVUf(void) mVU_FMOR() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FMOR vi%02d, vi%02d", _Ft_, _Fs_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } //------------------------------------------------------------------ @@ -544,6 +551,7 @@ microVUf(void) mVU_FSAND() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); } } microVUf(void) mVU_FSEQ() { @@ -557,6 +565,7 @@ microVUf(void) mVU_FSEQ() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); } } microVUf(void) mVU_FSOR() { @@ -568,6 +577,7 @@ microVUf(void) mVU_FSOR() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); SysPrintf("b\n"); } } microVUf(void) mVU_FSSET() { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 55418ae81c..4e5df160a1 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -146,6 +146,7 @@ declareAllVariables #define pass1 if (recPass == 0) #define pass2 if (recPass == 1) #define pass3 if (recPass == 2) +#define pass4 if (recPass == 3) // Misc Macros... #define mVUcurProg mVU->prog.prog[mVU->prog.cur] @@ -159,6 +160,7 @@ declareAllVariables #define mVUregs mVUallocInfo.block.pState #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC +#define mVUflagInfo mVUregs.needExactMatch //mVUallocInfo.flagInfo #define mVUinfo mVUallocInfo.info[iPC / 2] #define mVUstall mVUallocInfo.stall[iPC / 2] #define mVUstartPC mVUallocInfo.startPC @@ -249,6 +251,13 @@ declareAllVariables #define mVUlogQ() { mVUlog(", Q"); } #define mVUlogCLIP() { mVUlog("w.xyz vf%02d, vf%02dw", _Fs_, _Ft_); } +// Flag Info +//#define __NeedExact (1<<12) +//#define __ExactMatch (mVUregs.needExactMatch & (1<<12)) +#define __Status (mVUflagInfo & (0xf<<0)) +#define __Mac (mVUflagInfo & (0xf<<4)) +#define __Clip (mVUflagInfo & (0xf<<8)) + // Store VI regs in mmx regs? #define isMMX(_VIreg_) 0 //(_VIreg_ >= 1 && _VIreg_ <=8) #define mmVI(_VIreg_) (_VIreg_ - 1) diff --git a/pcsx2/x86/microVU_Tables.inl b/pcsx2/x86/microVU_Tables.inl index fc4b0b8852..8d54233993 100644 --- a/pcsx2/x86/microVU_Tables.inl +++ b/pcsx2/x86/microVU_Tables.inl @@ -37,1015 +37,195 @@ microVUf(void) mVUunknown(); //------------------------------------------------------------------ //------------------------------------------------------------------ -// mVULOWER_OPCODE +// Opcode Tables //------------------------------------------------------------------ -void (* mVULOWER_OPCODE00 [128])() = { - mVU_LQ<0,0> , mVU_SQ<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_ILW<0,0> , mVU_ISW<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_IADDIU<0,0> , mVU_ISUBIU<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_FCEQ<0,0> , mVU_FCSET<0,0> , mVU_FCAND<0,0> , mVU_FCOR<0,0>, /* 0x10 */ - mVU_FSEQ<0,0> , mVU_FSSET<0,0> , mVU_FSAND<0,0> , mVU_FSOR<0,0>, - mVU_FMEQ<0,0> , mVUunknown<0,0> , mVU_FMAND<0,0> , mVU_FMOR<0,0>, - mVU_FCGET<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_B<0,0> , mVU_BAL<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x20 */ - mVU_JR<0,0> , mVU_JALR<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_IBEQ<0,0> , mVU_IBNE<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_IBLTZ<0,0> , mVU_IBGTZ<0,0> , mVU_IBLEZ<0,0> , mVU_IBGEZ<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x30 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVULowerOP<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x40*/ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x50 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x60 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x70 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, +#define microVU_LOWER_OPCODE(x, y) void (* mVULOWER_OPCODE##x##y [128])() = { \ + mVU_LQ , mVU_SQ , mVUunknown , mVUunknown, \ + mVU_ILW , mVU_ISW , mVUunknown , mVUunknown, \ + mVU_IADDIU , mVU_ISUBIU , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_FCEQ , mVU_FCSET , mVU_FCAND , mVU_FCOR, /* 0x10 */ \ + mVU_FSEQ , mVU_FSSET , mVU_FSAND , mVU_FSOR, \ + mVU_FMEQ , mVUunknown , mVU_FMAND , mVU_FMOR, \ + mVU_FCGET , mVUunknown , mVUunknown , mVUunknown, \ + mVU_B , mVU_BAL , mVUunknown , mVUunknown, /* 0x20 */ \ + mVU_JR , mVU_JALR , mVUunknown , mVUunknown, \ + mVU_IBEQ , mVU_IBNE , mVUunknown , mVUunknown, \ + mVU_IBLTZ , mVU_IBGTZ , mVU_IBLEZ , mVU_IBGEZ, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP , mVUunknown , mVUunknown , mVUunknown, /* 0x40*/ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x50 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x60 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x70 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -void (* mVULOWER_OPCODE01 [128])() = { - mVU_LQ<0,1> , mVU_SQ<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_ILW<0,1> , mVU_ISW<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_IADDIU<0,1> , mVU_ISUBIU<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_FCEQ<0,1> , mVU_FCSET<0,1> , mVU_FCAND<0,1> , mVU_FCOR<0,1>, /* 0x10 */ - mVU_FSEQ<0,1> , mVU_FSSET<0,1> , mVU_FSAND<0,1> , mVU_FSOR<0,1>, - mVU_FMEQ<0,1> , mVUunknown<0,1> , mVU_FMAND<0,1> , mVU_FMOR<0,1>, - mVU_FCGET<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_B<0,1> , mVU_BAL<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x20 */ - mVU_JR<0,1> , mVU_JALR<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_IBEQ<0,1> , mVU_IBNE<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_IBLTZ<0,1> , mVU_IBGTZ<0,1> , mVU_IBLEZ<0,1> , mVU_IBGEZ<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x30 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVULowerOP<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x40*/ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x50 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x60 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x70 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, +#define microVU_LowerOP_T3_00_OPCODE(x, y) void (* mVULowerOP_T3_00_OPCODE##x##y [32])() = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MOVE , mVU_LQI , mVU_DIV , mVU_MTIR, \ + mVU_RNEXT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_MFP , mVU_XTOP , mVU_XGKICK, \ + mVU_ESADD , mVU_EATANxy , mVU_ESQRT , mVU_ESIN, \ }; -void (* mVULOWER_OPCODE10 [128])() = { - mVU_LQ<1,0> , mVU_SQ<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_ILW<1,0> , mVU_ISW<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_IADDIU<1,0> , mVU_ISUBIU<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_FCEQ<1,0> , mVU_FCSET<1,0> , mVU_FCAND<1,0> , mVU_FCOR<1,0>, /* 0x10 */ - mVU_FSEQ<1,0> , mVU_FSSET<1,0> , mVU_FSAND<1,0> , mVU_FSOR<1,0>, - mVU_FMEQ<1,0> , mVUunknown<1,0> , mVU_FMAND<1,0> , mVU_FMOR<1,0>, - mVU_FCGET<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_B<1,0> , mVU_BAL<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x20 */ - mVU_JR<1,0> , mVU_JALR<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_IBEQ<1,0> , mVU_IBNE<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_IBLTZ<1,0> , mVU_IBGTZ<1,0> , mVU_IBLEZ<1,0> , mVU_IBGEZ<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x30 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVULowerOP<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x40*/ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x50 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x60 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x70 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, +#define microVU_LowerOP_T3_01_OPCODE(x, y) void (* mVULowerOP_T3_01_OPCODE##x##y [32])() = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MR32 , mVU_SQI , mVU_SQRT , mVU_MFIR, \ + mVU_RGET , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVU_XITOP , mVUunknown, \ + mVU_ERSADD , mVU_EATANxz , mVU_ERSQRT , mVU_EATAN, \ }; -void (* mVULOWER_OPCODE11 [128])() = { - mVU_LQ<1,1> , mVU_SQ<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_ILW<1,1> , mVU_ISW<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_IADDIU<1,1> , mVU_ISUBIU<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_FCEQ<1,1> , mVU_FCSET<1,1> , mVU_FCAND<1,1> , mVU_FCOR<1,1>, /* 0x10 */ - mVU_FSEQ<1,1> , mVU_FSSET<1,1> , mVU_FSAND<1,1> , mVU_FSOR<1,1>, - mVU_FMEQ<1,1> , mVUunknown<1,1> , mVU_FMAND<1,1> , mVU_FMOR<1,1>, - mVU_FCGET<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_B<1,1> , mVU_BAL<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x20 */ - mVU_JR<1,1> , mVU_JALR<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_IBEQ<1,1> , mVU_IBNE<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_IBLTZ<1,1> , mVU_IBGTZ<1,1> , mVU_IBLEZ<1,1> , mVU_IBGEZ<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x30 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVULowerOP<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x40*/ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x50 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x60 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x70 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, +#define microVU_LowerOP_T3_10_OPCODE(x, y) void (* mVULowerOP_T3_10_OPCODE##x##y [32])() = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_LQD , mVU_RSQRT , mVU_ILWR, \ + mVU_RINIT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ELENG , mVU_ESUM , mVU_ERCPR , mVU_EEXP, \ }; -void (* mVULOWER_OPCODE02 [128])() = { - mVU_LQ<0,2> , mVU_SQ<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_ILW<0,2> , mVU_ISW<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_IADDIU<0,2> , mVU_ISUBIU<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_FCEQ<0,2> , mVU_FCSET<0,2> , mVU_FCAND<0,2> , mVU_FCOR<0,2>, /* 0x10 */ - mVU_FSEQ<0,2> , mVU_FSSET<0,2> , mVU_FSAND<0,2> , mVU_FSOR<0,2>, - mVU_FMEQ<0,2> , mVUunknown<0,2> , mVU_FMAND<0,2> , mVU_FMOR<0,2>, - mVU_FCGET<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_B<0,2> , mVU_BAL<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x20 */ - mVU_JR<0,2> , mVU_JALR<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_IBEQ<0,2> , mVU_IBNE<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_IBLTZ<0,2> , mVU_IBGTZ<0,2> , mVU_IBLEZ<0,2> , mVU_IBGEZ<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x30 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVULowerOP<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x40*/ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x50 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x60 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x70 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, +#define microVU_LowerOP_T3_11_OPCODE(x, y) void (* mVULowerOP_T3_11_OPCODE##x##y [32])() = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_SQD , mVU_WAITQ , mVU_ISWR, \ + mVU_RXOR , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ERLENG , mVUunknown , mVU_WAITP , mVUunknown, \ }; -void (* mVULOWER_OPCODE12 [128])() = { - mVU_LQ<1,2> , mVU_SQ<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_ILW<1,2> , mVU_ISW<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_IADDIU<1,2> , mVU_ISUBIU<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_FCEQ<1,2> , mVU_FCSET<1,2> , mVU_FCAND<1,2> , mVU_FCOR<1,2>, /* 0x10 */ - mVU_FSEQ<1,2> , mVU_FSSET<1,2> , mVU_FSAND<1,2> , mVU_FSOR<1,2>, - mVU_FMEQ<1,2> , mVUunknown<1,2> , mVU_FMAND<1,2> , mVU_FMOR<1,2>, - mVU_FCGET<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_B<1,2> , mVU_BAL<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x20 */ - mVU_JR<1,2> , mVU_JALR<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_IBEQ<1,2> , mVU_IBNE<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_IBLTZ<1,2> , mVU_IBGTZ<1,2> , mVU_IBLEZ<1,2> , mVU_IBGEZ<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x30 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVULowerOP<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x40*/ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x50 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x60 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x70 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVULowerOP_T3_00_OPCODE -//------------------------------------------------------------------ -void (* mVULowerOP_T3_00_OPCODE00 [32])() = { - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_MOVE<0,0> , mVU_LQI<0,0> , mVU_DIV<0,0> , mVU_MTIR<0,0>, - mVU_RNEXT<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x10 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVU_MFP<0,0> , mVU_XTOP<0,0> , mVU_XGKICK<0,0>, - mVU_ESADD<0,0> , mVU_EATANxy<0,0> , mVU_ESQRT<0,0> , mVU_ESIN<0,0>, +#define microVU_LowerOP_OPCODE(x, y) void (* mVULowerOP_OPCODE##x##y [64])() = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x20 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_IADD , mVU_ISUB , mVU_IADDI , mVUunknown, /* 0x30 */ \ + mVU_IAND , mVU_IOR , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP_T3_00, mVULowerOP_T3_01, mVULowerOP_T3_10, mVULowerOP_T3_11, \ }; -void (* mVULowerOP_T3_00_OPCODE01 [32])() = { - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_MOVE<0,1> , mVU_LQI<0,1> , mVU_DIV<0,1> , mVU_MTIR<0,1>, - mVU_RNEXT<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x10 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVU_MFP<0,1> , mVU_XTOP<0,1> , mVU_XGKICK<0,1>, - mVU_ESADD<0,1> , mVU_EATANxy<0,1> , mVU_ESQRT<0,1> , mVU_ESIN<0,1>, +#define microVU_UPPER_OPCODE(x, y) void (* mVU_UPPER_OPCODE##x##y [64])() = { \ + mVU_ADDx , mVU_ADDy , mVU_ADDz , mVU_ADDw, \ + mVU_SUBx , mVU_SUBy , mVU_SUBz , mVU_SUBw, \ + mVU_MADDx , mVU_MADDy , mVU_MADDz , mVU_MADDw, \ + mVU_MSUBx , mVU_MSUBy , mVU_MSUBz , mVU_MSUBw, \ + mVU_MAXx , mVU_MAXy , mVU_MAXz , mVU_MAXw, /* 0x10 */ \ + mVU_MINIx , mVU_MINIy , mVU_MINIz , mVU_MINIw, \ + mVU_MULx , mVU_MULy , mVU_MULz , mVU_MULw, \ + mVU_MULq , mVU_MAXi , mVU_MULi , mVU_MINIi, \ + mVU_ADDq , mVU_MADDq , mVU_ADDi , mVU_MADDi, /* 0x20 */ \ + mVU_SUBq , mVU_MSUBq , mVU_SUBi , mVU_MSUBi, \ + mVU_ADD , mVU_MADD , mVU_MUL , mVU_MAX, \ + mVU_SUB , mVU_MSUB , mVU_OPMSUB , mVU_MINI, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_UPPER_FD_00, mVU_UPPER_FD_01, mVU_UPPER_FD_10, mVU_UPPER_FD_11, \ }; -void (* mVULowerOP_T3_00_OPCODE10 [32])() = { - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_MOVE<1,0> , mVU_LQI<1,0> , mVU_DIV<1,0> , mVU_MTIR<1,0>, - mVU_RNEXT<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x10 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVU_MFP<1,0> , mVU_XTOP<1,0> , mVU_XGKICK<1,0>, - mVU_ESADD<1,0> , mVU_EATANxy<1,0> , mVU_ESQRT<1,0> , mVU_ESIN<1,0>, +#define microVU_UPPER_FD_00_TABLE(x, y) void (* mVU_UPPER_FD_00_TABLE##x##y [32])() = { \ + mVU_ADDAx , mVU_SUBAx , mVU_MADDAx , mVU_MSUBAx, \ + mVU_ITOF0 , mVU_FTOI0 , mVU_MULAx , mVU_MULAq, \ + mVU_ADDAq , mVU_SUBAq , mVU_ADDA , mVU_SUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -void (* mVULowerOP_T3_00_OPCODE11 [32])() = { - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_MOVE<1,1> , mVU_LQI<1,1> , mVU_DIV<1,1> , mVU_MTIR<1,1>, - mVU_RNEXT<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x10 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVU_MFP<1,1> , mVU_XTOP<1,1> , mVU_XGKICK<1,1>, - mVU_ESADD<1,1> , mVU_EATANxy<1,1> , mVU_ESQRT<1,1> , mVU_ESIN<1,1>, +#define microVU_UPPER_FD_01_TABLE(x, y) void (* mVU_UPPER_FD_01_TABLE##x##y [32])() = { \ + mVU_ADDAy , mVU_SUBAy , mVU_MADDAy , mVU_MSUBAy, \ + mVU_ITOF4 , mVU_FTOI4 , mVU_MULAy , mVU_ABS, \ + mVU_MADDAq , mVU_MSUBAq , mVU_MADDA , mVU_MSUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -void (* mVULowerOP_T3_00_OPCODE02 [32])() = { - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_MOVE<0,2> , mVU_LQI<0,2> , mVU_DIV<0,2> , mVU_MTIR<0,2>, - mVU_RNEXT<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x10 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVU_MFP<0,2> , mVU_XTOP<0,2> , mVU_XGKICK<0,2>, - mVU_ESADD<0,2> , mVU_EATANxy<0,2> , mVU_ESQRT<0,2> , mVU_ESIN<0,2>, +#define microVU_UPPER_FD_10_TABLE(x, y) void (* mVU_UPPER_FD_10_TABLE##x##y [32])() = { \ + mVU_ADDAz , mVU_SUBAz , mVU_MADDAz , mVU_MSUBAz, \ + mVU_ITOF12 , mVU_FTOI12 , mVU_MULAz , mVU_MULAi, \ + mVU_ADDAi , mVU_SUBAi , mVU_MULA , mVU_OPMULA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -void (* mVULowerOP_T3_00_OPCODE12 [32])() = { - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_MOVE<1,2> , mVU_LQI<1,2> , mVU_DIV<1,2> , mVU_MTIR<1,2>, - mVU_RNEXT<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x10 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVU_MFP<1,2> , mVU_XTOP<1,2> , mVU_XGKICK<1,2>, - mVU_ESADD<1,2> , mVU_EATANxy<1,2> , mVU_ESQRT<1,2> , mVU_ESIN<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVULowerOP_T3_01_OPCODE -//------------------------------------------------------------------ -void (* mVULowerOP_T3_01_OPCODE00 [32])() = { - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_MR32<0,0> , mVU_SQI<0,0> , mVU_SQRT<0,0> , mVU_MFIR<0,0>, - mVU_RGET<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x10 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVU_XITOP<0,0> , mVUunknown<0,0>, - mVU_ERSADD<0,0> , mVU_EATANxz<0,0> , mVU_ERSQRT<0,0> , mVU_EATAN<0,0>, -}; - -void (* mVULowerOP_T3_01_OPCODE01 [32])() = { - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_MR32<0,1> , mVU_SQI<0,1> , mVU_SQRT<0,1> , mVU_MFIR<0,1>, - mVU_RGET<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x10 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVU_XITOP<0,1> , mVUunknown<0,1>, - mVU_ERSADD<0,1> , mVU_EATANxz<0,1> , mVU_ERSQRT<0,1> , mVU_EATAN<0,1>, -}; - -void (* mVULowerOP_T3_01_OPCODE10 [32])() = { - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_MR32<1,0> , mVU_SQI<1,0> , mVU_SQRT<1,0> , mVU_MFIR<1,0>, - mVU_RGET<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x10 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVU_XITOP<1,0> , mVUunknown<1,0>, - mVU_ERSADD<1,0> , mVU_EATANxz<1,0> , mVU_ERSQRT<1,0> , mVU_EATAN<1,0>, -}; - -void (* mVULowerOP_T3_01_OPCODE11 [32])() = { - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_MR32<1,1> , mVU_SQI<1,1> , mVU_SQRT<1,1> , mVU_MFIR<1,1>, - mVU_RGET<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x10 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVU_XITOP<1,1> , mVUunknown<1,1>, - mVU_ERSADD<1,1> , mVU_EATANxz<1,1> , mVU_ERSQRT<1,1> , mVU_EATAN<1,1>, -}; - -void (* mVULowerOP_T3_01_OPCODE02 [32])() = { - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_MR32<0,2> , mVU_SQI<0,2> , mVU_SQRT<0,2> , mVU_MFIR<0,2>, - mVU_RGET<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x10 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVU_XITOP<0,2> , mVUunknown<0,2>, - mVU_ERSADD<0,2> , mVU_EATANxz<0,2> , mVU_ERSQRT<0,2> , mVU_EATAN<0,2>, -}; - -void (* mVULowerOP_T3_01_OPCODE12 [32])() = { - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_MR32<1,2> , mVU_SQI<1,2> , mVU_SQRT<1,2> , mVU_MFIR<1,2>, - mVU_RGET<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x10 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVU_XITOP<1,2> , mVUunknown<1,2>, - mVU_ERSADD<1,2> , mVU_EATANxz<1,2> , mVU_ERSQRT<1,2> , mVU_EATAN<1,2>, +#define microVU_UPPER_FD_11_TABLE(x, y) void (* mVU_UPPER_FD_11_TABLE##x##y [32])() = { \ + mVU_ADDAw , mVU_SUBAw , mVU_MADDAw , mVU_MSUBAw, \ + mVU_ITOF15 , mVU_FTOI15 , mVU_MULAw , mVU_CLIP, \ + mVU_MADDAi , mVU_MSUBAi , mVUunknown , mVU_NOP, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; //------------------------------------------------------------------ //------------------------------------------------------------------ -// mVULowerOP_T3_10_OPCODE +// Create Table Instances //------------------------------------------------------------------ -void (* mVULowerOP_T3_10_OPCODE00 [32])() = { - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVU_LQD<0,0> , mVU_RSQRT<0,0> , mVU_ILWR<0,0>, - mVU_RINIT<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x10 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_ELENG<0,0> , mVU_ESUM<0,0> , mVU_ERCPR<0,0> , mVU_EEXP<0,0>, -}; +#define mVUcreateTable(x,y) \ +microVU_LOWER_OPCODE(x,y) \ +microVU_LowerOP_T3_00_OPCODE(x,y) \ +microVU_LowerOP_T3_01_OPCODE(x,y) \ +microVU_LowerOP_T3_10_OPCODE(x,y) \ +microVU_LowerOP_T3_11_OPCODE(x,y) \ +microVU_LowerOP_OPCODE(x,y) \ +microVU_UPPER_OPCODE(x,y) \ +microVU_UPPER_FD_00_TABLE(x,y) \ +microVU_UPPER_FD_01_TABLE(x,y) \ +microVU_UPPER_FD_10_TABLE(x,y) \ +microVU_UPPER_FD_11_TABLE(x,y) -void (* mVULowerOP_T3_10_OPCODE01 [32])() = { - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVU_LQD<0,1> , mVU_RSQRT<0,1> , mVU_ILWR<0,1>, - mVU_RINIT<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x10 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_ELENG<0,1> , mVU_ESUM<0,1> , mVU_ERCPR<0,1> , mVU_EEXP<0,1>, -}; - -void (* mVULowerOP_T3_10_OPCODE10 [32])() = { - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVU_LQD<1,0> , mVU_RSQRT<1,0> , mVU_ILWR<1,0>, - mVU_RINIT<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x10 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_ELENG<1,0> , mVU_ESUM<1,0> , mVU_ERCPR<1,0> , mVU_EEXP<1,0>, -}; - -void (* mVULowerOP_T3_10_OPCODE11 [32])() = { - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVU_LQD<1,1> , mVU_RSQRT<1,1> , mVU_ILWR<1,1>, - mVU_RINIT<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x10 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_ELENG<1,1> , mVU_ESUM<1,1> , mVU_ERCPR<1,1> , mVU_EEXP<1,1>, -}; - -void (* mVULowerOP_T3_10_OPCODE02 [32])() = { - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVU_LQD<0,2> , mVU_RSQRT<0,2> , mVU_ILWR<0,2>, - mVU_RINIT<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x10 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_ELENG<0,2> , mVU_ESUM<0,2> , mVU_ERCPR<0,2> , mVU_EEXP<0,2>, -}; - -void (* mVULowerOP_T3_10_OPCODE12 [32])() = { - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVU_LQD<1,2> , mVU_RSQRT<1,2> , mVU_ILWR<1,2>, - mVU_RINIT<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x10 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_ELENG<1,2> , mVU_ESUM<1,2> , mVU_ERCPR<1,2> , mVU_EEXP<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVULowerOP_T3_11_OPCODE -//------------------------------------------------------------------ -void (* mVULowerOP_T3_11_OPCODE00 [32])() = { - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVU_SQD<0,0> , mVU_WAITQ<0,0> , mVU_ISWR<0,0>, - mVU_RXOR<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x10 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_ERLENG<0,0> , mVUunknown<0,0> , mVU_WAITP<0,0> , mVUunknown<0,0>, -}; - -void (* mVULowerOP_T3_11_OPCODE01 [32])() = { - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVU_SQD<0,1> , mVU_WAITQ<0,1> , mVU_ISWR<0,1>, - mVU_RXOR<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x10 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_ERLENG<0,1> , mVUunknown<0,1> , mVU_WAITP<0,1> , mVUunknown<0,1>, -}; - -void (* mVULowerOP_T3_11_OPCODE10 [32])() = { - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVU_SQD<1,0> , mVU_WAITQ<1,0> , mVU_ISWR<1,0>, - mVU_RXOR<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x10 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_ERLENG<1,0> , mVUunknown<1,0> , mVU_WAITP<1,0> , mVUunknown<1,0>, -}; - -void (* mVULowerOP_T3_11_OPCODE11 [32])() = { - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVU_SQD<1,1> , mVU_WAITQ<1,1> , mVU_ISWR<1,1>, - mVU_RXOR<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x10 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_ERLENG<1,1> , mVUunknown<1,1> , mVU_WAITP<1,1> , mVUunknown<1,1>, -}; - -void (* mVULowerOP_T3_11_OPCODE02 [32])() = { - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVU_SQD<0,2> , mVU_WAITQ<0,2> , mVU_ISWR<0,2>, - mVU_RXOR<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x10 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_ERLENG<0,2> , mVUunknown<0,2> , mVU_WAITP<0,2> , mVUunknown<0,2>, -}; - -void (* mVULowerOP_T3_11_OPCODE12 [32])() = { - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVU_SQD<1,2> , mVU_WAITQ<1,2> , mVU_ISWR<1,2>, - mVU_RXOR<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x10 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_ERLENG<1,2> , mVUunknown<1,2> , mVU_WAITP<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVULowerOP_OPCODE -//------------------------------------------------------------------ -void (* mVULowerOP_OPCODE00 [64])() = { - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x10 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x20 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_IADD<0,0> , mVU_ISUB<0,0> , mVU_IADDI<0,0> , mVUunknown<0,0>, /* 0x30 */ - mVU_IAND<0,0> , mVU_IOR<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVULowerOP_T3_00<0,0>, mVULowerOP_T3_01<0,0>, mVULowerOP_T3_10<0,0>, mVULowerOP_T3_11<0,0>, -}; - -void (* mVULowerOP_OPCODE01 [64])() = { - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x10 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x20 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_IADD<0,1> , mVU_ISUB<0,1> , mVU_IADDI<0,1> , mVUunknown<0,1>, /* 0x30 */ - mVU_IAND<0,1> , mVU_IOR<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVULowerOP_T3_00<0,1>, mVULowerOP_T3_01<0,1>, mVULowerOP_T3_10<0,1>, mVULowerOP_T3_11<0,1>, -}; - -void (* mVULowerOP_OPCODE10 [64])() = { - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x10 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x20 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_IADD<1,0> , mVU_ISUB<1,0> , mVU_IADDI<1,0> , mVUunknown<1,0>, /* 0x30 */ - mVU_IAND<1,0> , mVU_IOR<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVULowerOP_T3_00<1,0>, mVULowerOP_T3_01<1,0>, mVULowerOP_T3_10<1,0>, mVULowerOP_T3_11<1,0>, -}; - -void (* mVULowerOP_OPCODE11 [64])() = { - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x10 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x20 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_IADD<1,1> , mVU_ISUB<1,1> , mVU_IADDI<1,1> , mVUunknown<1,1>, /* 0x30 */ - mVU_IAND<1,1> , mVU_IOR<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVULowerOP_T3_00<1,1>, mVULowerOP_T3_01<1,1>, mVULowerOP_T3_10<1,1>, mVULowerOP_T3_11<1,1>, -}; - -void (* mVULowerOP_OPCODE02 [64])() = { - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x10 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x20 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_IADD<0,2> , mVU_ISUB<0,2> , mVU_IADDI<0,2> , mVUunknown<0,2>, /* 0x30 */ - mVU_IAND<0,2> , mVU_IOR<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVULowerOP_T3_00<0,2>, mVULowerOP_T3_01<0,2>, mVULowerOP_T3_10<0,2>, mVULowerOP_T3_11<0,2>, -}; - -void (* mVULowerOP_OPCODE12 [64])() = { - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x10 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x20 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_IADD<1,2> , mVU_ISUB<1,2> , mVU_IADDI<1,2> , mVUunknown<1,2>, /* 0x30 */ - mVU_IAND<1,2> , mVU_IOR<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVULowerOP_T3_00<1,2>, mVULowerOP_T3_01<1,2>, mVULowerOP_T3_10<1,2>, mVULowerOP_T3_11<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVU_UPPER_OPCODE -//------------------------------------------------------------------ -void (* mVU_UPPER_OPCODE00 [64])() = { - mVU_ADDx<0,0> , mVU_ADDy<0,0> , mVU_ADDz<0,0> , mVU_ADDw<0,0>, - mVU_SUBx<0,0> , mVU_SUBy<0,0> , mVU_SUBz<0,0> , mVU_SUBw<0,0>, - mVU_MADDx<0,0> , mVU_MADDy<0,0> , mVU_MADDz<0,0> , mVU_MADDw<0,0>, - mVU_MSUBx<0,0> , mVU_MSUBy<0,0> , mVU_MSUBz<0,0> , mVU_MSUBw<0,0>, - mVU_MAXx<0,0> , mVU_MAXy<0,0> , mVU_MAXz<0,0> , mVU_MAXw<0,0>, /* 0x10 */ - mVU_MINIx<0,0> , mVU_MINIy<0,0> , mVU_MINIz<0,0> , mVU_MINIw<0,0>, - mVU_MULx<0,0> , mVU_MULy<0,0> , mVU_MULz<0,0> , mVU_MULw<0,0>, - mVU_MULq<0,0> , mVU_MAXi<0,0> , mVU_MULi<0,0> , mVU_MINIi<0,0>, - mVU_ADDq<0,0> , mVU_MADDq<0,0> , mVU_ADDi<0,0> , mVU_MADDi<0,0>, /* 0x20 */ - mVU_SUBq<0,0> , mVU_MSUBq<0,0> , mVU_SUBi<0,0> , mVU_MSUBi<0,0>, - mVU_ADD<0,0> , mVU_MADD<0,0> , mVU_MUL<0,0> , mVU_MAX<0,0>, - mVU_SUB<0,0> , mVU_MSUB<0,0> , mVU_OPMSUB<0,0> , mVU_MINI<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, /* 0x30 */ - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVU_UPPER_FD_00<0,0>, mVU_UPPER_FD_01<0,0>, mVU_UPPER_FD_10<0,0>, mVU_UPPER_FD_11<0,0>, -}; - -void (* mVU_UPPER_OPCODE01 [64])() = { - mVU_ADDx<0,1> , mVU_ADDy<0,1> , mVU_ADDz<0,1> , mVU_ADDw<0,1>, - mVU_SUBx<0,1> , mVU_SUBy<0,1> , mVU_SUBz<0,1> , mVU_SUBw<0,1>, - mVU_MADDx<0,1> , mVU_MADDy<0,1> , mVU_MADDz<0,1> , mVU_MADDw<0,1>, - mVU_MSUBx<0,1> , mVU_MSUBy<0,1> , mVU_MSUBz<0,1> , mVU_MSUBw<0,1>, - mVU_MAXx<0,1> , mVU_MAXy<0,1> , mVU_MAXz<0,1> , mVU_MAXw<0,1>, /* 0x10 */ - mVU_MINIx<0,1> , mVU_MINIy<0,1> , mVU_MINIz<0,1> , mVU_MINIw<0,1>, - mVU_MULx<0,1> , mVU_MULy<0,1> , mVU_MULz<0,1> , mVU_MULw<0,1>, - mVU_MULq<0,1> , mVU_MAXi<0,1> , mVU_MULi<0,1> , mVU_MINIi<0,1>, - mVU_ADDq<0,1> , mVU_MADDq<0,1> , mVU_ADDi<0,1> , mVU_MADDi<0,1>, /* 0x20 */ - mVU_SUBq<0,1> , mVU_MSUBq<0,1> , mVU_SUBi<0,1> , mVU_MSUBi<0,1>, - mVU_ADD<0,1> , mVU_MADD<0,1> , mVU_MUL<0,1> , mVU_MAX<0,1>, - mVU_SUB<0,1> , mVU_MSUB<0,1> , mVU_OPMSUB<0,1> , mVU_MINI<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, /* 0x30 */ - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVU_UPPER_FD_00<0,1>, mVU_UPPER_FD_01<0,1>, mVU_UPPER_FD_10<0,1>, mVU_UPPER_FD_11<0,1>, -}; - -void (* mVU_UPPER_OPCODE10 [64])() = { - mVU_ADDx<1,0> , mVU_ADDy<1,0> , mVU_ADDz<1,0> , mVU_ADDw<1,0>, - mVU_SUBx<1,0> , mVU_SUBy<1,0> , mVU_SUBz<1,0> , mVU_SUBw<1,0>, - mVU_MADDx<1,0> , mVU_MADDy<1,0> , mVU_MADDz<1,0> , mVU_MADDw<1,0>, - mVU_MSUBx<1,0> , mVU_MSUBy<1,0> , mVU_MSUBz<1,0> , mVU_MSUBw<1,0>, - mVU_MAXx<1,0> , mVU_MAXy<1,0> , mVU_MAXz<1,0> , mVU_MAXw<1,0>, /* 0x10 */ - mVU_MINIx<1,0> , mVU_MINIy<1,0> , mVU_MINIz<1,0> , mVU_MINIw<1,0>, - mVU_MULx<1,0> , mVU_MULy<1,0> , mVU_MULz<1,0> , mVU_MULw<1,0>, - mVU_MULq<1,0> , mVU_MAXi<1,0> , mVU_MULi<1,0> , mVU_MINIi<1,0>, - mVU_ADDq<1,0> , mVU_MADDq<1,0> , mVU_ADDi<1,0> , mVU_MADDi<1,0>, /* 0x20 */ - mVU_SUBq<1,0> , mVU_MSUBq<1,0> , mVU_SUBi<1,0> , mVU_MSUBi<1,0>, - mVU_ADD<1,0> , mVU_MADD<1,0> , mVU_MUL<1,0> , mVU_MAX<1,0>, - mVU_SUB<1,0> , mVU_MSUB<1,0> , mVU_OPMSUB<1,0> , mVU_MINI<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, /* 0x30 */ - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVU_UPPER_FD_00<1,0>, mVU_UPPER_FD_01<1,0>, mVU_UPPER_FD_10<1,0>, mVU_UPPER_FD_11<1,0>, -}; - -void (* mVU_UPPER_OPCODE11 [64])() = { - mVU_ADDx<1,1> , mVU_ADDy<1,1> , mVU_ADDz<1,1> , mVU_ADDw<1,1>, - mVU_SUBx<1,1> , mVU_SUBy<1,1> , mVU_SUBz<1,1> , mVU_SUBw<1,1>, - mVU_MADDx<1,1> , mVU_MADDy<1,1> , mVU_MADDz<1,1> , mVU_MADDw<1,1>, - mVU_MSUBx<1,1> , mVU_MSUBy<1,1> , mVU_MSUBz<1,1> , mVU_MSUBw<1,1>, - mVU_MAXx<1,1> , mVU_MAXy<1,1> , mVU_MAXz<1,1> , mVU_MAXw<1,1>, /* 0x10 */ - mVU_MINIx<1,1> , mVU_MINIy<1,1> , mVU_MINIz<1,1> , mVU_MINIw<1,1>, - mVU_MULx<1,1> , mVU_MULy<1,1> , mVU_MULz<1,1> , mVU_MULw<1,1>, - mVU_MULq<1,1> , mVU_MAXi<1,1> , mVU_MULi<1,1> , mVU_MINIi<1,1>, - mVU_ADDq<1,1> , mVU_MADDq<1,1> , mVU_ADDi<1,1> , mVU_MADDi<1,1>, /* 0x20 */ - mVU_SUBq<1,1> , mVU_MSUBq<1,1> , mVU_SUBi<1,1> , mVU_MSUBi<1,1>, - mVU_ADD<1,1> , mVU_MADD<1,1> , mVU_MUL<1,1> , mVU_MAX<1,1>, - mVU_SUB<1,1> , mVU_MSUB<1,1> , mVU_OPMSUB<1,1> , mVU_MINI<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, /* 0x30 */ - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVU_UPPER_FD_00<1,1>, mVU_UPPER_FD_01<1,1>, mVU_UPPER_FD_10<1,1>, mVU_UPPER_FD_11<1,1>, -}; - -void (* mVU_UPPER_OPCODE02 [64])() = { - mVU_ADDx<0,2> , mVU_ADDy<0,2> , mVU_ADDz<0,2> , mVU_ADDw<0,2>, - mVU_SUBx<0,2> , mVU_SUBy<0,2> , mVU_SUBz<0,2> , mVU_SUBw<0,2>, - mVU_MADDx<0,2> , mVU_MADDy<0,2> , mVU_MADDz<0,2> , mVU_MADDw<0,2>, - mVU_MSUBx<0,2> , mVU_MSUBy<0,2> , mVU_MSUBz<0,2> , mVU_MSUBw<0,2>, - mVU_MAXx<0,2> , mVU_MAXy<0,2> , mVU_MAXz<0,2> , mVU_MAXw<0,2>, /* 0x10 */ - mVU_MINIx<0,2> , mVU_MINIy<0,2> , mVU_MINIz<0,2> , mVU_MINIw<0,2>, - mVU_MULx<0,2> , mVU_MULy<0,2> , mVU_MULz<0,2> , mVU_MULw<0,2>, - mVU_MULq<0,2> , mVU_MAXi<0,2> , mVU_MULi<0,2> , mVU_MINIi<0,2>, - mVU_ADDq<0,2> , mVU_MADDq<0,2> , mVU_ADDi<0,2> , mVU_MADDi<0,2>, /* 0x20 */ - mVU_SUBq<0,2> , mVU_MSUBq<0,2> , mVU_SUBi<0,2> , mVU_MSUBi<0,2>, - mVU_ADD<0,2> , mVU_MADD<0,2> , mVU_MUL<0,2> , mVU_MAX<0,2>, - mVU_SUB<0,2> , mVU_MSUB<0,2> , mVU_OPMSUB<0,2> , mVU_MINI<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, /* 0x30 */ - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVU_UPPER_FD_00<0,2>, mVU_UPPER_FD_01<0,2>, mVU_UPPER_FD_10<0,2>, mVU_UPPER_FD_11<0,2>, -}; - -void (* mVU_UPPER_OPCODE12 [64])() = { - mVU_ADDx<1,2> , mVU_ADDy<1,2> , mVU_ADDz<1,2> , mVU_ADDw<1,2>, - mVU_SUBx<1,2> , mVU_SUBy<1,2> , mVU_SUBz<1,2> , mVU_SUBw<1,2>, - mVU_MADDx<1,2> , mVU_MADDy<1,2> , mVU_MADDz<1,2> , mVU_MADDw<1,2>, - mVU_MSUBx<1,2> , mVU_MSUBy<1,2> , mVU_MSUBz<1,2> , mVU_MSUBw<1,2>, - mVU_MAXx<1,2> , mVU_MAXy<1,2> , mVU_MAXz<1,2> , mVU_MAXw<1,2>, /* 0x10 */ - mVU_MINIx<1,2> , mVU_MINIy<1,2> , mVU_MINIz<1,2> , mVU_MINIw<1,2>, - mVU_MULx<1,2> , mVU_MULy<1,2> , mVU_MULz<1,2> , mVU_MULw<1,2>, - mVU_MULq<1,2> , mVU_MAXi<1,2> , mVU_MULi<1,2> , mVU_MINIi<1,2>, - mVU_ADDq<1,2> , mVU_MADDq<1,2> , mVU_ADDi<1,2> , mVU_MADDi<1,2>, /* 0x20 */ - mVU_SUBq<1,2> , mVU_MSUBq<1,2> , mVU_SUBi<1,2> , mVU_MSUBi<1,2>, - mVU_ADD<1,2> , mVU_MADD<1,2> , mVU_MUL<1,2> , mVU_MAX<1,2>, - mVU_SUB<1,2> , mVU_MSUB<1,2> , mVU_OPMSUB<1,2> , mVU_MINI<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, /* 0x30 */ - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVU_UPPER_FD_00<1,2>, mVU_UPPER_FD_01<1,2>, mVU_UPPER_FD_10<1,2>, mVU_UPPER_FD_11<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVU_UPPER_FD_00_TABLE -//------------------------------------------------------------------ -void (* mVU_UPPER_FD_00_TABLE00 [32])() = { - mVU_ADDAx<0,0> , mVU_SUBAx<0,0> , mVU_MADDAx<0,0> , mVU_MSUBAx<0,0>, - mVU_ITOF0<0,0> , mVU_FTOI0<0,0> , mVU_MULAx<0,0> , mVU_MULAq<0,0>, - mVU_ADDAq<0,0> , mVU_SUBAq<0,0> , mVU_ADDA<0,0> , mVU_SUBA<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, -}; - -void (* mVU_UPPER_FD_00_TABLE01 [32])() = { - mVU_ADDAx<0,1> , mVU_SUBAx<0,1> , mVU_MADDAx<0,1> , mVU_MSUBAx<0,1>, - mVU_ITOF0<0,1> , mVU_FTOI0<0,1> , mVU_MULAx<0,1> , mVU_MULAq<0,1>, - mVU_ADDAq<0,1> , mVU_SUBAq<0,1> , mVU_ADDA<0,1> , mVU_SUBA<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, -}; - -void (* mVU_UPPER_FD_00_TABLE10 [32])() = { - mVU_ADDAx<1,0> , mVU_SUBAx<1,0> , mVU_MADDAx<1,0> , mVU_MSUBAx<1,0>, - mVU_ITOF0<1,0> , mVU_FTOI0<1,0> , mVU_MULAx<1,0> , mVU_MULAq<1,0>, - mVU_ADDAq<1,0> , mVU_SUBAq<1,0> , mVU_ADDA<1,0> , mVU_SUBA<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, -}; - -void (* mVU_UPPER_FD_00_TABLE11 [32])() = { - mVU_ADDAx<1,1> , mVU_SUBAx<1,1> , mVU_MADDAx<1,1> , mVU_MSUBAx<1,1>, - mVU_ITOF0<1,1> , mVU_FTOI0<1,1> , mVU_MULAx<1,1> , mVU_MULAq<1,1>, - mVU_ADDAq<1,1> , mVU_SUBAq<1,1> , mVU_ADDA<1,1> , mVU_SUBA<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, -}; - -void (* mVU_UPPER_FD_00_TABLE02 [32])() = { - mVU_ADDAx<0,2> , mVU_SUBAx<0,2> , mVU_MADDAx<0,2> , mVU_MSUBAx<0,2>, - mVU_ITOF0<0,2> , mVU_FTOI0<0,2> , mVU_MULAx<0,2> , mVU_MULAq<0,2>, - mVU_ADDAq<0,2> , mVU_SUBAq<0,2> , mVU_ADDA<0,2> , mVU_SUBA<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, -}; - -void (* mVU_UPPER_FD_00_TABLE12 [32])() = { - mVU_ADDAx<1,2> , mVU_SUBAx<1,2> , mVU_MADDAx<1,2> , mVU_MSUBAx<1,2>, - mVU_ITOF0<1,2> , mVU_FTOI0<1,2> , mVU_MULAx<1,2> , mVU_MULAq<1,2>, - mVU_ADDAq<1,2> , mVU_SUBAq<1,2> , mVU_ADDA<1,2> , mVU_SUBA<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVU_UPPER_FD_01_TABLE -//------------------------------------------------------------------ -void (* mVU_UPPER_FD_01_TABLE00 [32])() = { - mVU_ADDAy<0,0> , mVU_SUBAy<0,0> , mVU_MADDAy<0,0> , mVU_MSUBAy<0,0>, - mVU_ITOF4<0,0> , mVU_FTOI4<0,0> , mVU_MULAy<0,0> , mVU_ABS<0,0>, - mVU_MADDAq<0,0> , mVU_MSUBAq<0,0> , mVU_MADDA<0,0> , mVU_MSUBA<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, -}; - -void (* mVU_UPPER_FD_01_TABLE01 [32])() = { - mVU_ADDAy<0,1> , mVU_SUBAy<0,1> , mVU_MADDAy<0,1> , mVU_MSUBAy<0,1>, - mVU_ITOF4<0,1> , mVU_FTOI4<0,1> , mVU_MULAy<0,1> , mVU_ABS<0,1>, - mVU_MADDAq<0,1> , mVU_MSUBAq<0,1> , mVU_MADDA<0,1> , mVU_MSUBA<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, -}; - -void (* mVU_UPPER_FD_01_TABLE10 [32])() = { - mVU_ADDAy<1,0> , mVU_SUBAy<1,0> , mVU_MADDAy<1,0> , mVU_MSUBAy<1,0>, - mVU_ITOF4<1,0> , mVU_FTOI4<1,0> , mVU_MULAy<1,0> , mVU_ABS<1,0>, - mVU_MADDAq<1,0> , mVU_MSUBAq<1,0> , mVU_MADDA<1,0> , mVU_MSUBA<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, -}; - -void (* mVU_UPPER_FD_01_TABLE11 [32])() = { - mVU_ADDAy<1,1> , mVU_SUBAy<1,1> , mVU_MADDAy<1,1> , mVU_MSUBAy<1,1>, - mVU_ITOF4<1,1> , mVU_FTOI4<1,1> , mVU_MULAy<1,1> , mVU_ABS<1,1>, - mVU_MADDAq<1,1> , mVU_MSUBAq<1,1> , mVU_MADDA<1,1> , mVU_MSUBA<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, -}; - -void (* mVU_UPPER_FD_01_TABLE02 [32])() = { - mVU_ADDAy<0,2> , mVU_SUBAy<0,2> , mVU_MADDAy<0,2> , mVU_MSUBAy<0,2>, - mVU_ITOF4<0,2> , mVU_FTOI4<0,2> , mVU_MULAy<0,2> , mVU_ABS<0,2>, - mVU_MADDAq<0,2> , mVU_MSUBAq<0,2> , mVU_MADDA<0,2> , mVU_MSUBA<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, -}; - -void (* mVU_UPPER_FD_01_TABLE12 [32])() = { - mVU_ADDAy<1,2> , mVU_SUBAy<1,2> , mVU_MADDAy<1,2> , mVU_MSUBAy<1,2>, - mVU_ITOF4<1,2> , mVU_FTOI4<1,2> , mVU_MULAy<1,2> , mVU_ABS<1,2>, - mVU_MADDAq<1,2> , mVU_MSUBAq<1,2> , mVU_MADDA<1,2> , mVU_MSUBA<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVU_UPPER_FD_10_TABLE -//------------------------------------------------------------------ -void (* mVU_UPPER_FD_10_TABLE00 [32])() = { - mVU_ADDAz<0,0> , mVU_SUBAz<0,0> , mVU_MADDAz<0,0> , mVU_MSUBAz<0,0>, - mVU_ITOF12<0,0> , mVU_FTOI12<0,0> , mVU_MULAz<0,0> , mVU_MULAi<0,0>, - mVU_ADDAi<0,0> , mVU_SUBAi<0,0> , mVU_MULA<0,0> , mVU_OPMULA<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, -}; - -void (* mVU_UPPER_FD_10_TABLE01 [32])() = { - mVU_ADDAz<0,1> , mVU_SUBAz<0,1> , mVU_MADDAz<0,1> , mVU_MSUBAz<0,1>, - mVU_ITOF12<0,1> , mVU_FTOI12<0,1> , mVU_MULAz<0,1> , mVU_MULAi<0,1>, - mVU_ADDAi<0,1> , mVU_SUBAi<0,1> , mVU_MULA<0,1> , mVU_OPMULA<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, -}; - -void (* mVU_UPPER_FD_10_TABLE10 [32])() = { - mVU_ADDAz<1,0> , mVU_SUBAz<1,0> , mVU_MADDAz<1,0> , mVU_MSUBAz<1,0>, - mVU_ITOF12<1,0> , mVU_FTOI12<1,0> , mVU_MULAz<1,0> , mVU_MULAi<1,0>, - mVU_ADDAi<1,0> , mVU_SUBAi<1,0> , mVU_MULA<1,0> , mVU_OPMULA<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, -}; - -void (* mVU_UPPER_FD_10_TABLE11 [32])() = { - mVU_ADDAz<1,1> , mVU_SUBAz<1,1> , mVU_MADDAz<1,1> , mVU_MSUBAz<1,1>, - mVU_ITOF12<1,1> , mVU_FTOI12<1,1> , mVU_MULAz<1,1> , mVU_MULAi<1,1>, - mVU_ADDAi<1,1> , mVU_SUBAi<1,1> , mVU_MULA<1,1> , mVU_OPMULA<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, -}; - -void (* mVU_UPPER_FD_10_TABLE02 [32])() = { - mVU_ADDAz<0,2> , mVU_SUBAz<0,2> , mVU_MADDAz<0,2> , mVU_MSUBAz<0,2>, - mVU_ITOF12<0,2> , mVU_FTOI12<0,2> , mVU_MULAz<0,2> , mVU_MULAi<0,2>, - mVU_ADDAi<0,2> , mVU_SUBAi<0,2> , mVU_MULA<0,2> , mVU_OPMULA<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, -}; - -void (* mVU_UPPER_FD_10_TABLE12 [32])() = { - mVU_ADDAz<1,2> , mVU_SUBAz<1,2> , mVU_MADDAz<1,2> , mVU_MSUBAz<1,2>, - mVU_ITOF12<1,2> , mVU_FTOI12<1,2> , mVU_MULAz<1,2> , mVU_MULAi<1,2>, - mVU_ADDAi<1,2> , mVU_SUBAi<1,2> , mVU_MULA<1,2> , mVU_OPMULA<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// mVU_UPPER_FD_11_TABLE -//------------------------------------------------------------------ -void (* mVU_UPPER_FD_11_TABLE00 [32])() = { - mVU_ADDAw<0,0> , mVU_SUBAw<0,0> , mVU_MADDAw<0,0> , mVU_MSUBAw<0,0>, - mVU_ITOF15<0,0> , mVU_FTOI15<0,0> , mVU_MULAw<0,0> , mVU_CLIP<0,0>, - mVU_MADDAi<0,0> , mVU_MSUBAi<0,0> , mVUunknown<0,0> , mVU_NOP<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, - mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0> , mVUunknown<0,0>, -}; - -void (* mVU_UPPER_FD_11_TABLE01 [32])() = { - mVU_ADDAw<0,1> , mVU_SUBAw<0,1> , mVU_MADDAw<0,1> , mVU_MSUBAw<0,1>, - mVU_ITOF15<0,1> , mVU_FTOI15<0,1> , mVU_MULAw<0,1> , mVU_CLIP<0,1>, - mVU_MADDAi<0,1> , mVU_MSUBAi<0,1> , mVUunknown<0,1> , mVU_NOP<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, - mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1> , mVUunknown<0,1>, -}; - -void (* mVU_UPPER_FD_11_TABLE10 [32])() = { - mVU_ADDAw<1,0> , mVU_SUBAw<1,0> , mVU_MADDAw<1,0> , mVU_MSUBAw<1,0>, - mVU_ITOF15<1,0> , mVU_FTOI15<1,0> , mVU_MULAw<1,0> , mVU_CLIP<1,0>, - mVU_MADDAi<1,0> , mVU_MSUBAi<1,0> , mVUunknown<1,0> , mVU_NOP<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, - mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0> , mVUunknown<1,0>, -}; - -void (* mVU_UPPER_FD_11_TABLE11 [32])() = { - mVU_ADDAw<1,1> , mVU_SUBAw<1,1> , mVU_MADDAw<1,1> , mVU_MSUBAw<1,1>, - mVU_ITOF15<1,1> , mVU_FTOI15<1,1> , mVU_MULAw<1,1> , mVU_CLIP<1,1>, - mVU_MADDAi<1,1> , mVU_MSUBAi<1,1> , mVUunknown<1,1> , mVU_NOP<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, - mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1> , mVUunknown<1,1>, -}; - -void (* mVU_UPPER_FD_11_TABLE02 [32])() = { - mVU_ADDAw<0,2> , mVU_SUBAw<0,2> , mVU_MADDAw<0,2> , mVU_MSUBAw<0,2>, - mVU_ITOF15<0,2> , mVU_FTOI15<0,2> , mVU_MULAw<0,2> , mVU_CLIP<0,2>, - mVU_MADDAi<0,2> , mVU_MSUBAi<0,2> , mVUunknown<0,2> , mVU_NOP<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, - mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2> , mVUunknown<0,2>, -}; - -void (* mVU_UPPER_FD_11_TABLE12 [32])() = { - mVU_ADDAw<1,2> , mVU_SUBAw<1,2> , mVU_MADDAw<1,2> , mVU_MSUBAw<1,2>, - mVU_ITOF15<1,2> , mVU_FTOI15<1,2> , mVU_MULAw<1,2> , mVU_CLIP<1,2>, - mVU_MADDAi<1,2> , mVU_MSUBAi<1,2> , mVUunknown<1,2> , mVU_NOP<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, - mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2> , mVUunknown<1,2>, -}; -//------------------------------------------------------------------ +mVUcreateTable(0,0) +mVUcreateTable(0,1) +mVUcreateTable(0,2) +mVUcreateTable(0,3) +mVUcreateTable(1,0) +mVUcreateTable(1,1) +mVUcreateTable(1,2) +mVUcreateTable(1,3) //------------------------------------------------------------------ // Table Functions @@ -1053,15 +233,19 @@ void (* mVU_UPPER_FD_11_TABLE12 [32])() = { #define doTableStuff(tableName, args) { \ pass1 { \ if (vuIndex) tableName##10[ args ](); \ - else tableName##00[ args ](); \ + else tableName##00[ args ](); \ } \ pass2 { \ if (vuIndex) tableName##11[ args ](); \ - else tableName##01[ args ](); \ + else tableName##01[ args ](); \ } \ pass3 { \ if (vuIndex) tableName##12[ args ](); \ - else tableName##02[ args ](); \ + else tableName##02[ args ](); \ + } \ + pass4 { \ + if (vuIndex) tableName##13[ args ](); \ + else tableName##03[ args ](); \ } \ } From ef01d581b007e258efbb8c99750ba143971887d5 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 11 May 2009 06:15:16 +0000 Subject: [PATCH 044/296] microVU: coded a speedhack that optimizes out all status flag updates in a microprogram if the entire program doesn't read the status flags. FFX's intro got a 8~10% speedup with the hack, and it should be safe for 90~95% of games. It can be enabled in microVU_Misc.h git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1167 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 2 ++ pcsx2/x86/microVU.h | 1 + pcsx2/x86/microVU_Alloc.h | 2 +- pcsx2/x86/microVU_Compile.inl | 6 ++++-- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 7 ++++--- pcsx2/x86/microVU_Misc.h | 7 ++++++- pcsx2/x86/microVU_Misc.inl | 17 +++++++++++++++++ pcsx2/x86/microVU_Upper.inl | 1 + 9 files changed, 37 insertions(+), 8 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index f53d540dd4..cc836a3a70 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -128,6 +128,7 @@ microVUt(void) mVUclear(u32 addr, u32 size) { microVUt(void) mVUclearProg(int progIndex) { microVU* mVU = mVUx; mVU->prog.prog[progIndex].used = 1; + mVU->prog.prog[progIndex].sFlagHack = 0; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; for (u32 i = 0; i < (mVU->progSize / 2); i++) { mVU->prog.prog[progIndex].block[i]->reset(); @@ -139,6 +140,7 @@ microVUt(void) mVUcacheProg(int progIndex) { microVU* mVU = mVUx; memcpy_fast(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize); mVUdumpProg(progIndex); + mVUcheckSflag(progIndex); } // Finds the least used program, (if program list full clears and returns an old program; if not-full, returns free program) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 19e23990fe..2137e9518c 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -68,6 +68,7 @@ template struct microProgram { u32 data[progSize/4]; u32 used; // Number of times its been used + u32 sFlagHack; // Optimize out Status Flag Updates if Program doesn't use Status Flags u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_Alloc.h index 8e50e24f73..dc4e81d867 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_Alloc.h @@ -75,7 +75,7 @@ struct microAllocInfo { u32 count; // Number of VU 64bit instructions ran (starts at 0 for each block) u32 curPC; // Current PC u32 startPC; // Start PC for Cur Block - u32 flagInfo; // Holds information to help with flag instances on block linking + u32 sFlagHack; // Optimize out all Status flag updates if microProgram doesn't use Status flags u32 info[pSize/8]; // Info for Instructions in current block u8 stall[pSize/8]; // Info on how much each instruction stalled (stores the max amount of cycles to stall for the current opcodes) }; diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index ace1301b7e..4d30137cf5 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -118,10 +118,12 @@ microVUt(void) mVUendProgram(int fStatus, int fMac, int fClip) { } // Save Flag Instances - getFlagReg(fStatus, fStatus); + if (!mVUflagHack) { + getFlagReg(fStatus, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); + } mVUallocMFLAGa(gprT1, fMac); mVUallocCFLAGa(gprT2, fClip); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 20d1e8a81b..37a469a272 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -151,7 +151,7 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { microVU* mVU = mVUx; - if (__Status) { + if (__Status && !mVUflagHack) { int bStatus[4]; sortFlag(xStatus, bStatus, cycles); PUSH32R(gprR); // Backup gprR diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 615afeeb11..c829a0b964 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -551,7 +551,7 @@ microVUf(void) mVU_FSAND() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } microVUf(void) mVU_FSEQ() { @@ -565,7 +565,7 @@ microVUf(void) mVU_FSEQ() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } microVUf(void) mVU_FSOR() { @@ -577,7 +577,7 @@ microVUf(void) mVU_FSOR() { mVUallocVIb(gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); SysPrintf("b\n"); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } microVUf(void) mVU_FSSET() { @@ -591,6 +591,7 @@ microVUf(void) mVU_FSSET() { OR16ItoR (flagReg1, (_Imm12_ & 0xfc0)); } pass3 { mVUlog("FSSET $%x", _Imm12_); } + pass4 { mVUsFlagHack = 0; } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 4e5df160a1..42f39247ee 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -160,7 +160,8 @@ declareAllVariables #define mVUregs mVUallocInfo.block.pState #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC -#define mVUflagInfo mVUregs.needExactMatch //mVUallocInfo.flagInfo +#define mVUflagInfo mVUregs.needExactMatch +#define mVUsFlagHack mVUallocInfo.sFlagHack #define mVUinfo mVUallocInfo.info[iPC / 2] #define mVUstall mVUallocInfo.stall[iPC / 2] #define mVUstartPC mVUallocInfo.startPC @@ -286,6 +287,10 @@ declareAllVariables #define mVUdumpProg 0&& #endif +// Status Flag Speed Hack +#define CHECK_VU_FLAGHACK 0 // Set to 1 to turn hack on +#define mVUflagHack (mVUcurProg.sFlagHack) + // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ uptr diff = ptr - start; \ diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index bae60ade93..8f0e53850e 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -299,4 +299,21 @@ microVUt(void) mVUrestoreRegs() { POP32R(gprR); // Restore EDX } + +microVUt(void) mVUcheckSflag(int progIndex) { + if (CHECK_VU_FLAGHACK) { + + microVU* mVU = mVUx; + mVUsFlagHack = 1; + for (u32 i = 0; i < mVU->progSize; i+=2) { + mVU->code = mVU->prog.prog[progIndex].data[i+1]; + mVUopU(); + mVU->code = mVU->prog.prog[progIndex].data[i]; + mVUopL(); + } + mVUflagInfo = 0; + mVU->prog.prog[progIndex].sFlagHack = mVUsFlagHack; + } +} + #endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index a6fe9f0f3b..c80718581e 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -35,6 +35,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; //SysPrintf("doStatus = %d; doMac = %d\n", doStatus>>9, doMac>>8); + if (mVUflagHack) { mVUinfo &= ~_doStatus; } if (!doFlags) return; if (!doMac) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw From d3547e066201648a68ccfd7805901af20a85a845 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Mon, 11 May 2009 08:18:00 +0000 Subject: [PATCH 045/296] GSdx: trying to get rid of windows dependencies, still many classes to replace :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1168 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 4 +- plugins/GSdx/GPULocalMemory.cpp | 4 +- plugins/GSdx/GPULocalMemory.h | 2 +- plugins/GSdx/GPURenderer.cpp | 2 +- plugins/GSdx/GPURenderer.h | 41 ++++---- plugins/GSdx/GPUSettingsDlg.cpp | 19 ++-- plugins/GSdx/GPUSettingsDlg.h | 2 +- plugins/GSdx/GPUState.cpp | 8 +- plugins/GSdx/GPUState.h | 8 +- plugins/GSdx/GS.cpp | 60 ++++++------ plugins/GSdx/GSCapture.cpp | 10 +- plugins/GSdx/GSCaptureDlg.cpp | 19 ++-- plugins/GSdx/GSCaptureDlg.h | 2 +- plugins/GSdx/GSCodeBuffer.cpp | 6 +- plugins/GSdx/GSCodeBuffer.h | 2 +- plugins/GSdx/GSCrc.cpp | 10 +- plugins/GSdx/GSCrc.h | 2 +- plugins/GSdx/GSDevice.h | 20 ++-- plugins/GSdx/GSDevice10.cpp | 26 ++--- plugins/GSdx/GSDevice10.h | 11 +-- plugins/GSdx/GSDevice7.h | 2 +- plugins/GSdx/GSDevice9.cpp | 30 +++--- plugins/GSdx/GSDevice9.h | 9 +- plugins/GSdx/GSDeviceNull.h | 2 +- plugins/GSdx/GSDirtyRect.cpp | 14 ++- plugins/GSdx/GSDirtyRect.h | 2 +- plugins/GSdx/GSDump.cpp | 16 +-- plugins/GSdx/GSDump.h | 2 +- plugins/GSdx/GSFunctionMap.h | 80 +++++++-------- plugins/GSdx/GSLocalMemory.cpp | 36 +++---- plugins/GSdx/GSLocalMemory.h | 6 +- plugins/GSdx/GSPerfMon.cpp | 1 - plugins/GSdx/GSPerfMon.h | 4 - plugins/GSdx/GSRasterizer.cpp | 22 ++--- plugins/GSdx/GSRasterizer.h | 4 +- plugins/GSdx/GSRenderer.h | 46 ++++----- plugins/GSdx/GSRendererHW.h | 86 +++++++++++----- plugins/GSdx/GSRendererHW10.cpp | 2 +- plugins/GSdx/GSRendererHW10.h | 2 +- plugins/GSdx/GSRendererHW9.cpp | 2 +- plugins/GSdx/GSRendererHW9.h | 2 +- plugins/GSdx/GSRendererSW.h | 68 ++++++++++--- plugins/GSdx/GSSettingsDlg.cpp | 24 ++--- plugins/GSdx/GSSettingsDlg.h | 2 +- plugins/GSdx/GSState.cpp | 7 +- plugins/GSdx/GSTexture.h | 2 +- plugins/GSdx/GSTexture10.cpp | 4 +- plugins/GSdx/GSTexture10.h | 2 +- plugins/GSdx/GSTexture7.cpp | 2 +- plugins/GSdx/GSTexture7.h | 2 +- plugins/GSdx/GSTexture9.cpp | 6 +- plugins/GSdx/GSTexture9.h | 2 +- plugins/GSdx/GSTextureCache.h | 156 +++++++++++++++--------------- plugins/GSdx/GSTextureCache10.cpp | 4 +- plugins/GSdx/GSTextureCache9.cpp | 4 +- plugins/GSdx/GSTextureCacheSW.cpp | 49 ++++------ plugins/GSdx/GSTextureCacheSW.h | 5 +- plugins/GSdx/GSTextureFX10.cpp | 134 +++++++++++++------------ plugins/GSdx/GSTextureFX10.h | 12 +-- plugins/GSdx/GSTextureFX9.cpp | 116 ++++++++++++---------- plugins/GSdx/GSTextureFX9.h | 12 +-- plugins/GSdx/GSTextureNull.h | 2 +- plugins/GSdx/GSUtil.cpp | 54 ++++------- plugins/GSdx/GSVector.h | 4 +- plugins/GSdx/GSWnd.cpp | 4 +- plugins/GSdx/GSWnd.h | 2 +- plugins/GSdx/GSdx.cpp | 2 + plugins/GSdx/stdafx.cpp | 31 ++++++ plugins/GSdx/stdafx.h | 46 +++------ 69 files changed, 727 insertions(+), 659 deletions(-) diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index d9ea74a71c..63ab5b5309 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -239,9 +239,7 @@ EXPORT_C GPUupdateLace() EXPORT_C GPUmakeSnapshot() { - LPCTSTR path = _T("C:\\"); // TODO - - s_gpu->MakeSnapshot(path); + s_gpu->MakeSnapshot("c:\\"); // TODO } EXPORT_C GPUdisplayText(char* text) diff --git a/plugins/GSdx/GPULocalMemory.cpp b/plugins/GSdx/GPULocalMemory.cpp index f5a1b8a6ec..4dcdc97975 100644 --- a/plugins/GSdx/GPULocalMemory.cpp +++ b/plugins/GSdx/GPULocalMemory.cpp @@ -568,7 +568,7 @@ void GPULocalMemory::Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int } } -void GPULocalMemory::SaveBMP(LPCTSTR path, CRect r, int tp, int cx, int cy) +void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy) { r.left <<= m_scale.cx; r.top <<= m_scale.cy; @@ -578,7 +578,7 @@ void GPULocalMemory::SaveBMP(LPCTSTR path, CRect r, int tp, int cx, int cy) r.left &= ~1; r.right &= ~1; - if(FILE* fp = _tfopen(path, _T("wb"))) + if(FILE* fp = fopen(path.c_str(), "wb")) { BITMAPINFOHEADER bih; memset(&bih, 0, sizeof(bih)); diff --git a/plugins/GSdx/GPULocalMemory.h b/plugins/GSdx/GPULocalMemory.h index 4d7111e7d4..a4adfb4855 100644 --- a/plugins/GSdx/GPULocalMemory.h +++ b/plugins/GSdx/GPULocalMemory.h @@ -79,7 +79,7 @@ public: void Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels); void Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels); - void SaveBMP(LPCTSTR path, CRect r, int tp, int cx, int cy); + void SaveBMP(const string& path, CRect r, int tp, int cx, int cy); }; #pragma warning(default: 4244) \ No newline at end of file diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index 8680530c07..e6cd2ade78 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -22,4 +22,4 @@ #include "StdAfx.h" #include "GPURenderer.h" -CAtlMap GPURendererBase::m_wnd2gpu; \ No newline at end of file +map GPURendererBase::m_wnd2gpu; \ No newline at end of file diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index 17df468a44..9adaa4968a 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -38,13 +38,15 @@ class GPURendererBase : public GPUState, protected GPURendererSettings protected: HWND m_hWnd; WNDPROC m_wndproc; - static CAtlMap m_wnd2gpu; + static map m_wnd2gpu; static LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) { - if(CAtlMap::CPair* pair = m_wnd2gpu.Lookup(hWnd)) + map::iterator i = m_wnd2gpu.find(hWnd); + + if(i != m_wnd2gpu.end()) { - return pair->m_value->OnMessage(message, wParam, lParam); + return (*i).second->OnMessage(message, wParam, lParam); } ASSERT(0); @@ -92,7 +94,7 @@ public: { SetWindowLongPtr(m_hWnd, GWLP_WNDPROC, (LONG_PTR)m_wndproc); - m_wnd2gpu.RemoveKey(m_hWnd); + m_wnd2gpu.erase(m_hWnd); } } @@ -102,7 +104,8 @@ public: m_wndproc = (WNDPROC)GetWindowLongPtr(hWnd, GWLP_WNDPROC); SetWindowLongPtr(hWnd, GWLP_WNDPROC, (LONG_PTR)WndProc); - m_wnd2gpu.SetAt(hWnd, this); + + m_wnd2gpu[hWnd] = this; DWORD style = GetWindowLong(hWnd, GWL_STYLE); style |= WS_OVERLAPPEDWINDOW; @@ -115,7 +118,7 @@ public: } virtual void VSync() = 0; - virtual bool MakeSnapshot(LPCTSTR path) = 0; + virtual bool MakeSnapshot(const string& path) = 0; }; template @@ -207,7 +210,7 @@ protected: if(m_count > 0) { /* - Dump(_T("db")); + Dump("db"); if(m_env.PRIM.TME) { @@ -218,9 +221,7 @@ protected: r.right = r.left + 256; r.bottom = r.top + 256; - CString str; - str.Format(_T("da_%d_%d_%d_%d_%d"), m_env.STATUS.TP, r); - Dump(str, m_env.STATUS.TP, r, false); + Dump(format("da_%d_%d_%d_%d_%d", m_env.STATUS.TP, r).c_str(), m_env.STATUS.TP, r, false); } */ @@ -228,7 +229,7 @@ protected: m_count = 0; - //Dump(_T("dc"), false); + //Dump("dc", false); } } @@ -334,7 +335,7 @@ public: // osd static UINT64 s_frame = 0; - static CString s_stats; + static string s_stats; if(m_perfmon.GetFrame() - s_frame >= 30) { @@ -349,8 +350,8 @@ public: int w = r.Width() << m_scale.cx; int h = r.Height() << m_scale.cy; - s_stats.Format( - _T("%I64d | %d x %d | %.2f fps (%d%%) | %d/%d | %d%% CPU | %.2f | %.2f"), + s_stats = format( + "%I64d | %d x %d | %.2f fps (%d%%) | %d/%d | %d%% CPU | %.2f | %.2f", m_perfmon.GetFrame(), w, h, fps, (int)(100.0 * fps / m_env.GetFPS()), (int)m_perfmon.Get(GSPerfMon::Prim), (int)m_perfmon.Get(GSPerfMon::Draw), @@ -363,10 +364,10 @@ public: if(fillrate > 0) { - s_stats.Format(_T("%s | %.2f mpps"), CString(s_stats), fps * fillrate / (1024 * 1024)); + s_stats = format("%s | %.2f mpps", s_stats.c_str(), fps * fillrate / (1024 * 1024)); } - SetWindowText(m_hWnd, s_stats); + SetWindowText(m_hWnd, s_stats.c_str()); } if(m_dev.IsLost()) @@ -383,12 +384,8 @@ public: m_dev.Present(r); } - virtual bool MakeSnapshot(LPCTSTR path) + virtual bool MakeSnapshot(const string& path) { - CString fn; - - fn.Format(_T("%s_%s"), path, CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))); - - return m_dev.SaveCurrent(fn + _T(".bmp")); + return m_dev.SaveCurrent(format("%s_%s", path.c_str(), CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))) + ".bmp"); } }; diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 0d937ad0f2..0bccfa1d53 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -136,17 +136,17 @@ BOOL GPUSettingsDlg::OnInitDialog() memset(&caps, 0, sizeof(caps)); caps.PixelShaderVersion = D3DPS_VERSION(0, 0); - m_modes.RemoveAll(); + m_modes.clear(); // windowed { D3DDISPLAYMODE mode; memset(&mode, 0, sizeof(mode)); - m_modes.AddTail(mode); + m_modes.push_back(mode); int iItem = m_resolution.AddString(_T("Windowed")); - m_resolution.SetItemDataPtr(iItem, m_modes.GetTailPosition()); + m_resolution.SetItemDataPtr(iItem, &m_modes.back()); m_resolution.SetCurSel(iItem); } @@ -170,8 +170,8 @@ BOOL GPUSettingsDlg::OnInitDialog() str.Format(_T("%dx%d %dHz"), mode.Width, mode.Height, mode.RefreshRate); int iItem = m_resolution.AddString(str); - m_modes.AddTail(mode); - m_resolution.SetItemDataPtr(iItem, m_modes.GetTailPosition()); + m_modes.push_back(mode); + m_resolution.SetItemDataPtr(iItem, &m_modes.back()); if(ModeWidth == mode.Width && ModeHeight == mode.Height && ModeRefreshRate == mode.RefreshRate) { @@ -185,16 +185,16 @@ BOOL GPUSettingsDlg::OnInitDialog() bool isdx10avail = GSUtil::IsDirect3D10Available(); - CAtlArray renderers; + vector renderers; for(size_t i = 0; i < countof(g_renderers); i++) { if(i == 2 && !isdx10avail) continue; - renderers.Add(g_renderers[i]); + renderers.push_back(g_renderers[i]); } - GSSetting::InitComboBox(renderers.GetData(), renderers.GetCount(), m_renderer, pApp->GetProfileInt(_T("GPUSettings"), _T("Renderer"), 1)); + GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, pApp->GetProfileInt(_T("GPUSettings"), _T("Renderer"), 1)); GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, pApp->GetProfileInt(_T("Settings"), _T("PixelShaderVersion2"), D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); GSSetting::InitComboBox(g_filter, countof(g_filter), m_filter, pApp->GetProfileInt(_T("GPUSettings"), _T("filter"), 0)); GSSetting::InitComboBox(g_dithering, countof(g_dithering), m_dithering, pApp->GetProfileInt(_T("GPUSettings"), _T("dithering"), 1)); @@ -225,7 +225,8 @@ void GPUSettingsDlg::OnOK() if(m_resolution.GetCurSel() >= 0) { - D3DDISPLAYMODE& mode = m_modes.GetAt((POSITION)m_resolution.GetItemData(m_resolution.GetCurSel())); + D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); + pApp->WriteProfileInt(_T("Settings"), _T("ModeWidth"), mode.Width); pApp->WriteProfileInt(_T("Settings"), _T("ModeHeight"), mode.Height); pApp->WriteProfileInt(_T("Settings"), _T("ModeRefreshRate"), mode.RefreshRate); diff --git a/plugins/GSdx/GPUSettingsDlg.h b/plugins/GSdx/GPUSettingsDlg.h index 4602ef685b..04828cf892 100644 --- a/plugins/GSdx/GPUSettingsDlg.h +++ b/plugins/GSdx/GPUSettingsDlg.h @@ -29,7 +29,7 @@ class GPUSettingsDlg : public CDialog DECLARE_DYNAMIC(GPUSettingsDlg) private: - CAtlList m_modes; + list m_modes; public: GPUSettingsDlg(CWnd* pParent = NULL); // standard constructor diff --git a/plugins/GSdx/GPUState.cpp b/plugins/GSdx/GPUState.cpp index 8aa4d30ea3..a8e7630f22 100644 --- a/plugins/GSdx/GPUState.cpp +++ b/plugins/GSdx/GPUState.cpp @@ -356,7 +356,7 @@ int GPUState::PH_Command(GPUReg* r, int size) Invalidate(r2); - Dump(_T("f")); + Dump("f"); return 3; } @@ -568,7 +568,7 @@ int GPUState::PH_Move(GPUReg* r, int size) Invalidate(CRect(dst, CSize(w, h))); - // Dump(_T("m")); + // Dump("m"); return 4; } @@ -597,7 +597,7 @@ int GPUState::PH_Write(GPUReg* r, int size) Invalidate(r2); - Dump(_T("w")); + Dump("w"); m_perfmon.Put(GSPerfMon::Swizzle, w * h * 2); @@ -626,7 +626,7 @@ int GPUState::PH_Read(GPUReg* r, int size) m_mem.ReadRect(r2, (WORD*)m_read.buff); - Dump(_T("r")); + Dump("r"); m_env.STATUS.IMG = 1; diff --git a/plugins/GSdx/GPUState.h b/plugins/GSdx/GPUState.h index 8888dd7ce3..82d93692fd 100644 --- a/plugins/GSdx/GPUState.h +++ b/plugins/GSdx/GPUState.h @@ -88,7 +88,7 @@ protected: int s_n; - void Dump(LPCTSTR s, UINT32 TP, const CRect& r, int inc = true) + void Dump(const string& s, UINT32 TP, const CRect& r, int inc = true) { //if(m_perfmon.GetFrame() < 1000) //if((m_env.TWIN.ai32 & 0xfffff) == 0) @@ -103,12 +103,10 @@ protected: #ifdef DEBUG dir = 2; #endif - CString str; - str.Format(_T("c:\\temp%d\\%04d_%s.bmp"), dir, s_n, s); - m_mem.SaveBMP(str, r, TP, m_env.CLUT.X, m_env.CLUT.Y); + m_mem.SaveBMP(format("c:\\temp%d\\%04d_%s.bmp", dir, s_n, s), r, TP, m_env.CLUT.X, m_env.CLUT.Y); } - void Dump(LPCTSTR s, int inc = true) + void Dump(const string& s, int inc = true) { Dump(s, 2, CRect(0, 0, 1024, 512), inc); } diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 3dd274f709..c4fce6e68a 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -133,7 +133,7 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); - if(!s_gs->Create(CString(title))) + if(!s_gs->Create(title)) { GSclose(); @@ -152,15 +152,21 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) EXPORT_C_(INT32) GSopen(void* dsp, char* title, int mt) { AFX_MANAGE_STATE(AfxGetStaticModuleState()); + int renderer; - if (mt == 2){ //pcsx2 sent a switch renderer request + if(mt == 2) + { + // pcsx2 sent a switch renderer request renderer = 1; //DX9 sw mt = 1; } - else { //normal init + else + { + // normal init renderer = AfxGetApp()->GetProfileInt(_T("Settings"), _T("renderer"), 0); } + return GSopen(dsp, title, mt, renderer); } @@ -211,7 +217,7 @@ EXPORT_C GSvsync(int field) EXPORT_C_(UINT32) GSmakeSnapshot(char* path) { - return s_gs->MakeSnapshot(CString(path) + _T("gsdx")); + return s_gs->MakeSnapshot(string(path) + "gsdx"); } EXPORT_C GSkeyEvent(keyEvent* ev) @@ -303,7 +309,7 @@ EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) ::SetPriorityClass(::GetCurrentProcess(), HIGH_PRIORITY_CLASS); - CAtlArray buff; + vector buff; if(FILE* fp = fopen(lpszCmdLine, "rb")) { @@ -313,7 +319,7 @@ EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) GSsetBaseMem(regs); HWND hWnd = NULL; - GSopen(&hWnd, _T(""), true, renderer); + GSopen(&hWnd, "", true, renderer); DWORD crc; fread(&crc, 4, 1, fp); @@ -348,20 +354,20 @@ EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) switch(index) { case 0: - if(buff.GetCount() < 0x4000) buff.SetCount(0x4000); + if(buff.size() < 0x4000) buff.resize(0x4000); addr = 0x4000 - size; - fread(buff.GetData() + addr, size, 1, fp); - GSgifTransfer1(buff.GetData(), addr); + fread(&buff[0] + addr, size, 1, fp); + GSgifTransfer1(&buff[0], addr); break; case 1: - if(buff.GetCount() < size) buff.SetCount(size); - fread(buff.GetData(), size, 1, fp); - GSgifTransfer2(buff.GetData(), size / 16); + if(buff.size() < size) buff.resize(size); + fread(&buff[0], size, 1, fp); + GSgifTransfer2(&buff[0], size / 16); break; case 2: - if(buff.GetCount() < size) buff.SetCount(size); - fread(buff.GetData(), size, 1, fp); - GSgifTransfer3(buff.GetData(), size / 16); + if(buff.size() < size) buff.resize(size); + fread(&buff[0], size, 1, fp); + GSgifTransfer3(&buff[0], size / 16); break; } break; @@ -371,8 +377,8 @@ EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) break; case 2: fread(&size, 4, 1, fp); - if(buff.GetCount() < size) buff.SetCount(size); - GSreadFIFO2(buff.GetData(), size / 16); + if(buff.size() < size) buff.resize(size); + GSreadFIFO2(&buff[0], size / 16); break; case 3: fread(regs, 0x2000, 1, fp); @@ -394,15 +400,15 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow { ::SetPriorityClass(::GetCurrentProcess(), HIGH_PRIORITY_CLASS); - FILE* file = _tfopen(_T("c:\\log.txt"), _T("a")); + FILE* file = fopen("c:\\log.txt", "a"); - _ftprintf(file, _T("-------------------------\n\n")); + fprintf(file, "-------------------------\n\n"); if(1) { GSLocalMemory mem; - static struct {int psm; LPCSTR name;} s_format[] = + static struct {int psm; const char* name;} s_format[] = { {PSM_PSMCT32, "32"}, {PSM_PSMCT24, "24"}, @@ -432,7 +438,7 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow int w = 1 << tbw; int h = 1 << tbw; - _ftprintf(file, _T("%d x %d\n\n"), w, h); + fprintf(file, "%d x %d\n\n", w, h); for(int i = 0; i < countof(s_format); i++) { @@ -496,7 +502,7 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow end = clock(); - _ftprintf(file, _T("%6d %6d | "), (int)((float)trlen * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); + fprintf(file, "%6d %6d | ", (int)((float)trlen * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); start = clock(); @@ -510,7 +516,7 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow end = clock(); - _ftprintf(file, _T("%6d %6d | "), (int)((float)trlen * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); + fprintf(file, "%6d %6d | ", (int)((float)trlen * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); start = clock(); @@ -521,7 +527,7 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow end = clock(); - _ftprintf(file, _T("%6d %6d "), (int)((float)len * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); + fprintf(file, "%6d %6d ", (int)((float)len * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); if(psm.pal > 0) { @@ -534,15 +540,15 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow end = clock(); - _ftprintf(file, _T("| %6d %6d "), (int)((float)len * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); + fprintf(file, "| %6d %6d ", (int)((float)len * n / (end - start) / 1000), (int)((float)(w * h) * n / (end - start) / 1000)); } - _ftprintf(file, _T("\n")); + fprintf(file, "\n"); fflush(file); } - _ftprintf(file, _T("\n")); + fprintf(file, "\n"); } _aligned_free(ptr); diff --git a/plugins/GSdx/GSCapture.cpp b/plugins/GSdx/GSCapture.cpp index bd1ccca41e..069e71841b 100644 --- a/plugins/GSdx/GSCapture.cpp +++ b/plugins/GSdx/GSCapture.cpp @@ -60,7 +60,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource class GSSourceOutputPin : public CBaseOutputPin { CSize m_size; - CAtlArray m_mts; + vector m_mts; public: GSSourceOutputPin(CSize size, REFERENCE_TIME atpf, CBaseFilter* pFilter, CCritSec* pLock, HRESULT& hr) @@ -91,7 +91,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource vih.bmiHeader.biSizeImage = m_size.cx * m_size.cy * 2; mt.SetFormat((BYTE*)&vih, sizeof(vih)); - m_mts.Add(mt); + m_mts.push_back(mt); #endif @@ -106,7 +106,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource vih.bmiHeader.biSizeImage = m_size.cx * m_size.cy * 4; mt.SetFormat((BYTE*)&vih, sizeof(vih)); - m_mts.Add(mt); + m_mts.push_back(mt); } HRESULT GSSourceOutputPin::DecideBufferSize(IMemAllocator* pAlloc, ALLOCATOR_PROPERTIES* pProperties) @@ -137,9 +137,9 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource HRESULT CheckMediaType(const CMediaType* pmt) { - for(int i = 0, j = m_mts.GetCount(); i < j; i++) + for(vector::iterator i = m_mts.begin(); i != m_mts.end(); i++) { - if(m_mts[i].majortype == pmt->majortype && m_mts[i].subtype == pmt->subtype) + if(i->majortype == pmt->majortype && i->subtype == pmt->subtype) { return S_OK; } diff --git a/plugins/GSdx/GSCaptureDlg.cpp b/plugins/GSdx/GSCaptureDlg.cpp index 8dcd094875..32432a5df4 100644 --- a/plugins/GSdx/GSCaptureDlg.cpp +++ b/plugins/GSdx/GSCaptureDlg.cpp @@ -44,11 +44,11 @@ int GSCaptureDlg::GetSelCodec(Codec& c) if(iSel < 0) return 0; - POSITION pos = (POSITION)m_codeclist.GetItemDataPtr(iSel); + Codec* codec = (Codec*)m_codeclist.GetItemDataPtr(iSel); - if(pos == NULL) return 2; + if(codec == NULL) return 2; - c = m_codecs.GetAt(pos); + c = *codec; if(!c.filter) { @@ -83,7 +83,7 @@ BOOL GSCaptureDlg::OnInitDialog() { __super::OnInitDialog(); - m_codecs.RemoveAll(); + m_codecs.clear(); m_codeclist.ResetContent(); m_codeclist.SetItemDataPtr(m_codeclist.AddString(_T("Uncompressed")), NULL); @@ -116,7 +116,8 @@ BOOL GSCaptureDlg::OnInitDialog() else if(str.Find(L"@device:cm:") == 0) prefix = _T("(VfW) "); c.FriendlyName = prefix + c.FriendlyName; - m_codeclist.SetItemDataPtr(m_codeclist.AddString(c.FriendlyName), m_codecs.AddTail(c)); + m_codecs.push_back(c); + m_codeclist.SetItemDataPtr(m_codeclist.AddString(c.FriendlyName), &m_codecs.back()); } EndEnumSysDev @@ -128,13 +129,17 @@ BOOL GSCaptureDlg::OnInitDialog() { CString DisplayName; - POSITION pos = (POSITION)m_codeclist.GetItemDataPtr(i); + Codec* codec = (Codec*)m_codeclist.GetItemDataPtr(i); - if(pos) DisplayName = m_codecs.GetAt(pos).DisplayName; + if(codec) + { + DisplayName = codec->DisplayName; + } if(DisplayName == DisplayNameToFind) { m_codeclist.SetCurSel(i); + break; } } diff --git a/plugins/GSdx/GSCaptureDlg.h b/plugins/GSdx/GSCaptureDlg.h index 419b0cd62a..04d7215bd1 100644 --- a/plugins/GSdx/GSCaptureDlg.h +++ b/plugins/GSdx/GSCaptureDlg.h @@ -39,7 +39,7 @@ private: CComBSTR DisplayName; }; - CAtlList m_codecs; + list m_codecs; int GetSelCodec(Codec& c); diff --git a/plugins/GSdx/GSCodeBuffer.cpp b/plugins/GSdx/GSCodeBuffer.cpp index fe8214b97a..7f90f53569 100644 --- a/plugins/GSdx/GSCodeBuffer.cpp +++ b/plugins/GSdx/GSCodeBuffer.cpp @@ -32,9 +32,9 @@ GSCodeBuffer::GSCodeBuffer(size_t blocksize) GSCodeBuffer::~GSCodeBuffer() { - while(!m_buffers.IsEmpty()) + for(list::iterator i = m_buffers.begin(); i != m_buffers.end(); i++) { - VirtualFree(m_buffers.RemoveHead(), 0, MEM_RELEASE); + VirtualFree(*i, 0, MEM_RELEASE); } } @@ -51,7 +51,7 @@ void* GSCodeBuffer::GetBuffer(size_t size) m_pos = 0; - m_buffers.AddTail(m_ptr); + m_buffers.push_back(m_ptr); } BYTE* ptr = &m_ptr[m_pos]; diff --git a/plugins/GSdx/GSCodeBuffer.h b/plugins/GSdx/GSCodeBuffer.h index 5aaed5bc38..7011d00950 100644 --- a/plugins/GSdx/GSCodeBuffer.h +++ b/plugins/GSdx/GSCodeBuffer.h @@ -23,7 +23,7 @@ class GSCodeBuffer { - CAtlList m_buffers; + list m_buffers; size_t m_blocksize; size_t m_pos, m_reserved; BYTE* m_ptr; diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index aec9656c20..85abd05cb5 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -131,11 +131,11 @@ CRC::Game CRC::m_games[] = {0xFB236A46, SonicUnleashed, US}, }; -CAtlMap CRC::m_map; +hash_map CRC::m_map; CRC::Game CRC::Lookup(DWORD crc) { - if(m_map.IsEmpty()) + if(m_map.empty()) { for(int i = 0; i < countof(m_games); i++) { @@ -143,9 +143,11 @@ CRC::Game CRC::Lookup(DWORD crc) } } - if(CAtlMap::CPair* pair = m_map.Lookup(crc)) + hash_map::iterator i = m_map.find(crc); + + if(i != m_map.end()) { - return *pair->m_value; + return *(*i).second; } return m_games[0]; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index 083b69df1a..c5187f3b87 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -91,7 +91,7 @@ public: private: static Game m_games[]; - static CAtlMap m_map; + static hash_map m_map; public: static Game Lookup(DWORD crc); diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index e21f5ffdc4..01d06f42cc 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -46,7 +46,7 @@ struct InterlaceConstantBuffer template class GSDevice { - CAtlList m_pool; + list m_pool; protected: HWND m_hWnd; @@ -62,15 +62,15 @@ protected: { Recycle(t); - for(POSITION pos = m_pool.GetHeadPosition(); pos; m_pool.GetNext(pos)) + for(list::iterator i = m_pool.begin(); i != m_pool.end(); i++) { - const Texture& t2 = m_pool.GetAt(pos); + const Texture& t2 = *i; if(t2.GetType() == type && t2.GetWidth() == w && t2.GetHeight() == h && t2.GetFormat() == format) { t = t2; - m_pool.RemoveAt(pos); + m_pool.erase(i); return true; } @@ -102,7 +102,7 @@ public: virtual bool Reset(int w, int h, bool fs) { - m_pool.RemoveAll(); + m_pool.clear(); m_backbuffer = Texture(); m_merge = Texture(); m_weavebob = Texture(); @@ -121,7 +121,7 @@ public: virtual void EndScene() = 0; - virtual void Draw(LPCTSTR str) = 0; + virtual void Draw(const string& s) = 0; virtual bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) = 0; @@ -157,18 +157,18 @@ public: { if(t) { - m_pool.AddHead(t); + m_pool.push_front(t); - while(m_pool.GetCount() > 200) + while(m_pool.size() > 200) { - m_pool.RemoveTail(); + m_pool.pop_back(); } t = Texture(); } } - bool SaveCurrent(LPCTSTR fn) + bool SaveCurrent(const string& fn) { return m_current.Save(fn); } diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 5cd874a69d..97465aa013 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -112,9 +112,7 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_convert.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - hr = CompileShader(IDR_CONVERT10_FX, main, NULL, &m_convert.ps[i]); + hr = CompileShader(IDR_CONVERT10_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } memset(&bd, 0, sizeof(bd)); @@ -155,9 +153,7 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_merge.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - hr = CompileShader(IDR_MERGE10_FX, main, NULL, &m_merge.ps[i]); + hr = CompileShader(IDR_MERGE10_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); } memset(&bsd, 0, sizeof(bsd)); @@ -187,9 +183,7 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_interlace.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - hr = CompileShader(IDR_INTERLACE10_FX, main, NULL, &m_interlace.ps[i]); + hr = CompileShader(IDR_INTERLACE10_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); } // @@ -298,7 +292,7 @@ void GSDevice10::EndScene() // OMSetRenderTargets(NULL, NULL); } -void GSDevice10::Draw(LPCTSTR str) +void GSDevice10::Draw(const string& s) { /* BOOL fs; @@ -741,13 +735,13 @@ void GSDevice10::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, cons EndScene(); } -HRESULT GSDevice10::CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) +HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) { HRESULT hr; CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry, "vs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "vs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -776,13 +770,13 @@ HRESULT GSDevice10::CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* mac return hr; } -HRESULT GSDevice10::CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs) +HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs) { HRESULT hr; CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry, "gs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "gs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -804,13 +798,13 @@ HRESULT GSDevice10::CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* mac return hr; } -HRESULT GSDevice10::CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps) +HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps) { HRESULT hr; CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry, "ps_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "ps_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index 55124884de..a3ad0c1a75 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -103,7 +103,7 @@ public: void Present(const CRect& r); void BeginScene(); void EndScene(); - void Draw(LPCTSTR str); + void Draw(const string& s); bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); void ClearRenderTarget(Texture& t, const GSVector4& c); @@ -138,10 +138,7 @@ public: void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); - HRESULT CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il); - HRESULT CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs); - HRESULT CompileShader(UINT id, LPCSTR entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps); - - // TODO - bool SaveToFileD32S8X24(ID3D10Texture2D* ds, LPCTSTR fn); + HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il); + HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs); + HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps); }; diff --git a/plugins/GSdx/GSDevice7.h b/plugins/GSdx/GSDevice7.h index 94823dbf88..42d2853cf7 100644 --- a/plugins/GSdx/GSDevice7.h +++ b/plugins/GSdx/GSDevice7.h @@ -48,7 +48,7 @@ public: void Present(const CRect& r); void BeginScene() {} void EndScene() {} - void Draw(LPCTSTR str) {} + void Draw(const string& s) {} bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} void ClearRenderTarget(Texture& t, const GSVector4& c) {} diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 03984fc848..73f46d7482 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -141,9 +141,7 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_convert.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - CompileShader(IDR_CONVERT9_FX, main, NULL, &m_convert.ps[i]); + CompileShader(IDR_CONVERT9_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } m_convert.dss.DepthEnable = false; @@ -170,9 +168,7 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_merge.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - CompileShader(IDR_MERGE9_FX, main, NULL, &m_merge.ps[i]); + CompileShader(IDR_MERGE9_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); } m_merge.bs.BlendEnable = true; @@ -188,9 +184,7 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) for(int i = 0; i < countof(m_interlace.ps); i++) { - CStringA main; - main.Format("ps_main%d", i); - CompileShader(IDR_INTERLACE9_FX, main, NULL, &m_interlace.ps[i]); + CompileShader(IDR_INTERLACE9_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); } // @@ -393,7 +387,7 @@ void GSDevice9::EndScene() m_dev->EndScene(); } -void GSDevice9::Draw(LPCTSTR str) +void GSDevice9::Draw(const string& s) { /* if(!m_pp.Windowed) @@ -941,9 +935,9 @@ static HRESULT LoadShader(UINT id, LPCSTR& data, DWORD& size) return S_OK; } -HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il) +HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il) { - LPCSTR target; + const char* target; if(m_d3dcaps.VertexShaderVersion >= D3DVS_VERSION(3, 0)) { @@ -962,7 +956,7 @@ HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry, target, 0, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); LPCSTR data; DWORD size; @@ -971,7 +965,7 @@ HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, if(FAILED(hr)) return E_FAIL; - hr = D3DXCompileShader(data, size, macro, NULL, entry, target, 0, &shader, &error, NULL); + hr = D3DXCompileShader(data, size, macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); if(SUCCEEDED(hr)) { @@ -1001,9 +995,9 @@ HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, return S_OK; } -HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps) +HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps) { - LPCSTR target = NULL; + const char* target = NULL; UINT flags = 0; if(m_d3dcaps.PixelShaderVersion >= D3DPS_VERSION(3, 0)) @@ -1024,7 +1018,7 @@ HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry, target, flags, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, flags, &shader, &error, NULL); LPCSTR data; DWORD size; @@ -1033,7 +1027,7 @@ HRESULT GSDevice9::CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, if(FAILED(hr)) return E_FAIL; - hr = D3DXCompileShader(data, size, macro, NULL, entry, target, 0, &shader, &error, NULL); + hr = D3DXCompileShader(data, size, macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); if(SUCCEEDED(hr)) { diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 5a16aa0106..105b3f72f9 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -139,7 +139,7 @@ public: void Present(const CRect& r); void BeginScene(); void EndScene(); - void Draw(LPCTSTR str); + void Draw(const string& s); bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); void ClearRenderTarget(Texture& t, const GSVector4& c); @@ -179,14 +179,11 @@ public: void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); - HRESULT CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il); - HRESULT CompileShader(UINT id, LPCSTR entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps); + HRESULT CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il); + HRESULT CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps); virtual bool IsCurrentRGBA() { return false; } - - // TODO - bool SaveToFileD24S8(IDirect3DSurface9* ds, LPCTSTR fn); }; diff --git a/plugins/GSdx/GSDeviceNull.h b/plugins/GSdx/GSDeviceNull.h index e2782f1228..acd471f71d 100644 --- a/plugins/GSdx/GSDeviceNull.h +++ b/plugins/GSdx/GSDeviceNull.h @@ -43,7 +43,7 @@ public: void Present(const CRect& r) {} void BeginScene() {} void EndScene() {} - void Draw(LPCTSTR str) {} + void Draw(const string& s) {} bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} void ClearRenderTarget(Texture& t, const GSVector4& c) {} diff --git a/plugins/GSdx/GSDirtyRect.cpp b/plugins/GSdx/GSDirtyRect.cpp index 1d8b7ff123..f1d12497bb 100644 --- a/plugins/GSdx/GSDirtyRect.cpp +++ b/plugins/GSdx/GSDirtyRect.cpp @@ -62,9 +62,17 @@ CRect GSDirtyRect::GetDirtyRect(const GIFRegTEX0& TEX0) CRect GSDirtyRectList::GetDirtyRect(const GIFRegTEX0& TEX0, CSize size) { - if(IsEmpty()) return CRect(0, 0, 0, 0); + if(empty()) + { + return CRect(0, 0, 0, 0); + } + CRect r(INT_MAX, INT_MAX, 0, 0); - POSITION pos = GetHeadPosition(); - while(pos) r |= GetNext(pos).GetDirtyRect(TEX0); + + for(list::iterator i = begin(); i != end(); i++) + { + r |= i->GetDirtyRect(TEX0); + } + return r & CRect(0, 0, size.cx, size.cy); } diff --git a/plugins/GSdx/GSDirtyRect.h b/plugins/GSdx/GSDirtyRect.h index 8e6f5734df..a573d79ab2 100644 --- a/plugins/GSdx/GSDirtyRect.h +++ b/plugins/GSdx/GSDirtyRect.h @@ -34,7 +34,7 @@ public: CRect GetDirtyRect(const GIFRegTEX0& TEX0); }; -class GSDirtyRectList : public CAtlList +class GSDirtyRectList : public list { public: GSDirtyRectList() {} diff --git a/plugins/GSdx/GSDump.cpp b/plugins/GSdx/GSDump.cpp index 70421b35db..7f1095a5b3 100644 --- a/plugins/GSdx/GSDump.cpp +++ b/plugins/GSdx/GSDump.cpp @@ -36,10 +36,10 @@ GSDump::~GSDump() Close(); } -void GSDump::Open(const CString& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs) +void GSDump::Open(const string& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs) { - m_gs = _tfopen(fn + _T(".gs"), _T("wb")); - m_obj = _tfopen(fn + _T(".obj"), _T("wt")); + m_gs = fopen((fn + ".gs").c_str(), "wb"); + m_obj = fopen((fn + ".obj").c_str(), "wt"); m_frames = 0; m_objects = 0; @@ -123,20 +123,20 @@ void GSDump::Object(GSVertexSW* vertices, int count, GS_PRIM_CLASS primclass) float y = vertices[i].p.y; float z = vertices[i].p.z; - _ftprintf(m_obj, _T("v %f %f %f\n"), x, y, z); + fprintf(m_obj, "v %f %f %f\n", x, y, z); } for(int i = 0; i < count; i++) { - _ftprintf(m_obj, _T("vt %f %f %f\n"), vertices[i].t.x, vertices[i].t.y, vertices[i].t.z); + fprintf(m_obj, "vt %f %f %f\n", vertices[i].t.x, vertices[i].t.y, vertices[i].t.z); } for(int i = 0; i < count; i++) { - _ftprintf(m_obj, _T("vn %f %f %f\n"), 0.0f, 0.0f, 0.0f); + fprintf(m_obj, "vn %f %f %f\n", 0.0f, 0.0f, 0.0f); } - _ftprintf(m_obj, _T("g f%d_o%d_p%d_v%d\n"), m_frames, m_objects, primclass, count); + fprintf(m_obj, "g f%d_o%d_p%d_v%d\n", m_frames, m_objects, primclass, count); for(int i = 0; i < count; i += 3) { @@ -144,7 +144,7 @@ void GSDump::Object(GSVertexSW* vertices, int count, GS_PRIM_CLASS primclass) int b = m_vertices + i + 2; int c = m_vertices + i + 3; - _ftprintf(m_obj, _T("f %d/%d/%d %d/%d/%d %d/%d/%d \n"), a, a, a, b, b, b, c, c, c); + fprintf(m_obj, "f %d/%d/%d %d/%d/%d %d/%d/%d\n", a, a, a, b, b, b, c, c, c); } m_vertices += count; diff --git a/plugins/GSdx/GSDump.h b/plugins/GSdx/GSDump.h index bfa2d0682b..1ffc8167ec 100644 --- a/plugins/GSdx/GSDump.h +++ b/plugins/GSdx/GSDump.h @@ -55,7 +55,7 @@ public: GSDump(); virtual ~GSDump(); - void Open(const CString& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs); + void Open(const string& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs); void Close(); void ReadFIFO(UINT32 size); void Transfer(int index, BYTE* mem, size_t size); diff --git a/plugins/GSdx/GSFunctionMap.h b/plugins/GSdx/GSFunctionMap.h index a6fb5127d7..34074703d0 100644 --- a/plugins/GSdx/GSFunctionMap.h +++ b/plugins/GSdx/GSFunctionMap.h @@ -22,6 +22,7 @@ #pragma once #include "GS.h" +#include "GSCodeBuffer.h" struct GSRasterizerStats { @@ -50,8 +51,9 @@ protected: VALUE f; }; - CRBMap m_map; - CRBMap m_map_active; + hash_map m_map; + hash_map m_map_active; + ActivePtr* m_active; virtual VALUE GetDefaultFunction(KEY key) = 0; @@ -64,28 +66,25 @@ public: virtual ~GSFunctionMap() { - POSITION pos = m_map_active.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - delete m_map_active.GetNextValue(pos); + delete (*i).second; } - - m_map_active.RemoveAll(); - } - - void SetAt(KEY key, VALUE f) - { - m_map.SetAt(key, f); } VALUE Lookup(KEY key) { m_active = NULL; - if(!m_map_active.Lookup(key, m_active)) + hash_map::iterator i = m_map_active.find(key); + + if(i != m_map_active.end()) { - CRBMap::CPair* pair = m_map.Lookup(key); + m_active = (*i).second; + } + else + { + hash_map::iterator i = m_map.find(key); ActivePtr* p = new ActivePtr(); @@ -93,9 +92,9 @@ public: p->frame = (UINT64)-1; - p->f = pair ? pair->m_value : GetDefaultFunction(key); + p->f = i != m_map.end() ? (*i).second : GetDefaultFunction(key); - m_map_active.SetAt(key, p); + m_map_active[key] = p; m_active = p; } @@ -122,11 +121,9 @@ public: { __int64 ttpf = 0; - POSITION pos = m_map_active.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - ActivePtr* p = m_map_active.GetNextValue(pos); + ActivePtr* p = (*i).second; if(p->frames) { @@ -134,14 +131,10 @@ public: } } - pos = m_map_active.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - KEY key; - ActivePtr* p; - - m_map_active.GetNextAssoc(pos, key, p); + KEY key = (*i).first; + ActivePtr* p = (*i).second; if(p->frames > 0) { @@ -150,7 +143,7 @@ public: __int64 ppf = p->frames > 0 ? p->pixels / p->frames : 0; printf("[%012I64x]%c %6.2f%% | %5.2f%% | f %4I64d | p %10I64d | tpp %4I64d | tpf %9I64d | ppf %7I64d\n", - (UINT64)key, !m_map.Lookup(key) ? '*' : ' ', + (UINT64)key, m_map.find(key) == m_map.end() ? '*' : ' ', (float)(tpf * 10000 / 50000000) / 100, (float)(tpf * 10000 / ttpf) / 100, p->frames, p->pixels, @@ -160,15 +153,14 @@ public: } }; -#include "GSCodeBuffer.h" #include "vtune/JITProfiling.h" template class GSCodeGeneratorFunctionMap : public GSFunctionMap { DWORD m_id; - CStringA m_name; - CRBMap m_cgmap; + string m_name; + hash_map m_cgmap; GSCodeBuffer m_cb; enum {MAX_SIZE = 4096}; @@ -177,7 +169,7 @@ protected: virtual CG* Create(KEY key, void* ptr, size_t maxsize = MAX_SIZE) = 0; public: - GSCodeGeneratorFunctionMap(LPCSTR name) + GSCodeGeneratorFunctionMap(const char* name) : m_id(0x100000) , m_name(name) { @@ -185,11 +177,9 @@ public: virtual ~GSCodeGeneratorFunctionMap() { - POSITION pos = m_cgmap.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_cgmap.begin(); i != m_cgmap.end(); i++) { - delete m_cgmap.GetNextValue(pos); + delete (*i).second; } } @@ -197,7 +187,13 @@ public: { CG* cg = NULL; - if(!m_cgmap.Lookup(key, cg)) + hash_map::iterator i = m_cgmap.find(key); + + if(i != m_cgmap.end()) + { + cg = (*i).second; + } + else { void* ptr = m_cb.GetBuffer(MAX_SIZE); @@ -207,20 +203,18 @@ public: m_cb.ReleaseBuffer(cg->getSize()); - m_cgmap.SetAt(key, cg); + m_cgmap[key] = cg; // vtune method registration - CStringA name; - - name.Format("%s<%016I64x>()", m_name, (UINT64)key); + string name = format("%s<%016I64x>()", m_name.c_str(), (UINT64)key); iJIT_Method_Load ml; memset(&ml, 0, sizeof(ml)); ml.method_id = m_id++; - ml.method_name = (LPSTR)(LPCSTR)name; + ml.method_name = (char*)name.c_str(); ml.method_load_address = (void*)cg->getCode(); ml.method_size = cg->getSize(); diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 1ef3bb5485..4a8a8ebff7 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -465,11 +465,9 @@ GSLocalMemory::~GSLocalMemory() { VirtualFree(m_vm8, 0, MEM_RELEASE); - POSITION pos = m_omap.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_omap.begin(); i != m_omap.end(); i++) { - Offset* o = m_omap.GetNextValue(pos); + Offset* o = (*i).second; for(int i = 0; i < countof(o->col); i++) { @@ -479,16 +477,10 @@ GSLocalMemory::~GSLocalMemory() _aligned_free(o); } - m_omap.RemoveAll(); - - pos = m_o4map.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_o4map.begin(); i != m_o4map.end(); i++) { - _aligned_free(m_o4map.GetNextValue(pos)); + _aligned_free((*i).second); } - - m_o4map.RemoveAll(); } GSLocalMemory::Offset* GSLocalMemory::GetOffset(DWORD bp, DWORD bw, DWORD psm) @@ -499,9 +491,11 @@ GSLocalMemory::Offset* GSLocalMemory::GetOffset(DWORD bp, DWORD bw, DWORD psm) DWORD hash = bp | (bw << 14) | (psm << 20); - if(CRBMap::CPair* pair = m_omap.Lookup(hash)) + hash_map::iterator i = m_omap.find(hash); + + if(i != m_omap.end()) { - return pair->m_value; + return (*i).second; } Offset* o = (Offset*)_aligned_malloc(sizeof(Offset), 16); @@ -524,7 +518,7 @@ GSLocalMemory::Offset* GSLocalMemory::GetOffset(DWORD bp, DWORD bw, DWORD psm) memcpy(o->col[i], m_psm[psm].rowOffset[0], sizeof(int) * 2048); } - m_omap.SetAt(hash, o); + m_omap[hash] = o; return o; } @@ -546,9 +540,11 @@ GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, cons DWORD hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28); - if(CRBMap::CPair* pair = m_o4map.Lookup(hash)) + hash_map::iterator i = m_o4map.find(hash); + + if(i != m_o4map.end()) { - return pair->m_value; + return (*i).second; } Offset4* o = (Offset4*)_aligned_malloc(sizeof(Offset4), 16); @@ -573,7 +569,7 @@ GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, cons o->col[i].y = m_psm[zpsm].rowOffset[0][i * 4] << zs; } - m_o4map.SetAt(hash, o); + m_o4map[hash] = o; return o; } @@ -2543,7 +2539,7 @@ if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); } } -HRESULT GSLocalMemory::SaveBMP(LPCTSTR fn, DWORD bp, DWORD bw, DWORD psm, int w, int h) +HRESULT GSLocalMemory::SaveBMP(const string& fn, DWORD bp, DWORD bw, DWORD psm, int w, int h) { int pitch = w * 4; int size = pitch * h; @@ -2571,7 +2567,7 @@ HRESULT GSLocalMemory::SaveBMP(LPCTSTR fn, DWORD bp, DWORD bw, DWORD psm, int w, for(int i = 0; i < w; i++) ((DWORD*)p)[i] = (this->*rp)(i, j, TEX0.TBP0, TEX0.TBW); - if(FILE* fp = _tfopen(fn, _T("wb"))) + if(FILE* fp = fopen(fn.c_str(), "wb")) { BITMAPINFOHEADER bih; memset(&bih, 0, sizeof(bih)); diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index b5063cb087..fadd7ecee0 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -141,8 +141,8 @@ protected: // - CRBMapC m_omap; - CRBMapC m_o4map; + hash_map m_omap; + hash_map m_o4map; public: GSLocalMemory(); @@ -993,7 +993,7 @@ public: // - HRESULT SaveBMP(LPCTSTR fn, DWORD bp, DWORD bw, DWORD psm, int w, int h); + HRESULT SaveBMP(const string& fn, DWORD bp, DWORD bw, DWORD psm, int w, int h); }; #pragma warning(default: 4244) \ No newline at end of file diff --git a/plugins/GSdx/GSPerfMon.cpp b/plugins/GSdx/GSPerfMon.cpp index ba3aa1a809..3a5f2d297e 100644 --- a/plugins/GSdx/GSPerfMon.cpp +++ b/plugins/GSdx/GSPerfMon.cpp @@ -33,7 +33,6 @@ GSPerfMon::GSPerfMon() { memset(m_counters, 0, sizeof(m_counters)); memset(m_stats, 0, sizeof(m_stats)); - memset(m_warnings, 0, sizeof(m_warnings)); } void GSPerfMon::Put(counter_t c, double val) diff --git a/plugins/GSdx/GSPerfMon.h b/plugins/GSdx/GSPerfMon.h index 1c9b94921d..42463551c5 100644 --- a/plugins/GSdx/GSPerfMon.h +++ b/plugins/GSdx/GSPerfMon.h @@ -25,12 +25,10 @@ class GSPerfMon { public: enum counter_t {Frame, Prim, Draw, Swizzle, Unswizzle, Fillrate, Quad, CounterLast}; - enum warning_t {DATE, PABE, ABE, COLCLAMP, DepthTexture, WarningLast}; protected: double m_counters[CounterLast]; double m_stats[CounterLast]; - bool m_warnings[WarningLast]; UINT64 m_begin, m_total, m_start, m_frame; clock_t m_lastframe; int m_count; @@ -47,8 +45,6 @@ public: UINT64 GetFrame() {return m_frame;} void Put(counter_t c, double val = 0); double Get(counter_t c) {return m_stats[c];} - void Put(warning_t c) {m_warnings[c] = true;} - bool Get(warning_t c) {bool b = m_warnings[c]; m_warnings[c] = false; return b;} void Update(); int CPU(); }; diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index 053e789084..17795fd9c5 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -922,10 +922,12 @@ GSRasterizerList::~GSRasterizerList() void GSRasterizerList::FreeRasterizers() { - while(!IsEmpty()) + for(list::iterator i = begin(); i != end(); i++) { - delete RemoveHead(); + delete *i; } + + clear(); } void GSRasterizerList::Draw(const GSRasterizerData* data) @@ -936,11 +938,9 @@ void GSRasterizerList::Draw(const GSRasterizerData* data) __int64 start = __rdtsc(); - POSITION pos = GetTailPosition(); - - while(pos) + for(list::reverse_iterator i = rbegin(); i != rend(); i++) { - GetPrev(pos)->Draw(data); + (*i)->Draw(data); } while(*m_sync) @@ -950,13 +950,11 @@ void GSRasterizerList::Draw(const GSRasterizerData* data) m_stats.ticks = __rdtsc() - start; - pos = GetHeadPosition(); - - while(pos) + for(list::iterator i = begin(); i != end(); i++) { GSRasterizerStats s; - GetNext(pos)->GetStats(s); + (*i)->GetStats(s); m_stats.pixels += s.pixels; m_stats.prims = max(m_stats.prims, s.prims); @@ -970,8 +968,8 @@ void GSRasterizerList::GetStats(GSRasterizerStats& stats) void GSRasterizerList::PrintStats() { - if(!IsEmpty()) + if(!empty()) { - GetHead()->PrintStats(); + front()->PrintStats(); } } diff --git a/plugins/GSdx/GSRasterizer.h b/plugins/GSdx/GSRasterizer.h index 0f24c794f2..651899b91a 100644 --- a/plugins/GSdx/GSRasterizer.h +++ b/plugins/GSdx/GSRasterizer.h @@ -123,7 +123,7 @@ public: void Draw(const GSRasterizerData* data); }; -class GSRasterizerList : protected CAtlList, public IRasterizer +class GSRasterizerList : protected list, public IRasterizer { long* m_sync; GSRasterizerStats m_stats; @@ -142,7 +142,7 @@ public: for(int i = 0; i < threads; i++) { - AddTail(new GSRasterizerMT(new DS(parent, i), i, threads, m_sync)); + push_back(new GSRasterizerMT(new DS(parent, i), i, threads, m_sync)); } } diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 56bd999fb7..e25a0fe6f4 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -108,9 +108,9 @@ public: m_blur = rs.m_blur; }; - virtual bool Create(LPCTSTR title) = 0; + virtual bool Create(const string& title) = 0; virtual void VSync(int field) = 0; - virtual bool MakeSnapshot(LPCTSTR path) = 0; + virtual bool MakeSnapshot(const string& path) = 0; }; template class GSRenderer : public GSRendererBase @@ -323,7 +323,7 @@ protected: void DoSnapshot(int field) { - if(!m_snapshot.IsEmpty()) + if(!m_snapshot.empty()) { if(!m_dump && (::GetAsyncKeyState(VK_SHIFT) & 0x8000)) { @@ -339,9 +339,9 @@ protected: delete [] fd.data; } - m_dev.SaveCurrent(m_snapshot + _T(".bmp")); + m_dev.SaveCurrent(m_snapshot + ".bmp"); - m_snapshot.Empty(); + m_snapshot.clear(); } else { @@ -408,7 +408,7 @@ public: bool s_save; bool s_savez; - CString m_snapshot; + string m_snapshot; GSCapture m_capture; public: @@ -422,9 +422,9 @@ public: s_savez = !!AfxGetApp()->GetProfileInt(_T("Debug"), _T("savez"), 0); } - bool Create(LPCTSTR title) + bool Create(const string& title) { - if(!m_wnd.Create(title)) + if(!m_wnd.Create(title.c_str())) { return false; } @@ -456,7 +456,7 @@ public: // osd static UINT64 s_frame = 0; - static CString s_stats; + static string s_stats; if(m_perfmon.GetFrame() - s_frame >= 30) { @@ -465,11 +465,13 @@ public: s_frame = m_perfmon.GetFrame(); double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); + + string interlace = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; - s_stats.Format( - _T("%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f"), + s_stats = format( + "%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f", m_perfmon.GetFrame(), GetDisplaySize().cx, GetDisplaySize().cy, fps, (int)(100.0 * fps / GetFPS()), - m_regs->SMODE2.INT ? (CString(_T("Interlaced ")) + (m_regs->SMODE2.FFMD ? _T("(frame)") : _T("(field)"))) : _T("Progressive"), + interlace.c_str(), GSSettingsDlg::g_interlace[m_interlace].name, GSSettingsDlg::g_aspectratio[m_aspectratio].name, (int)m_perfmon.Get(GSPerfMon::Quad), @@ -484,26 +486,20 @@ public: if(fillrate > 0) { - s_stats.Format(_T("%s | %.2f mpps"), CString(s_stats), fps * fillrate / (1024 * 1024)); + s_stats += format(" | %.2f mpps", fps * fillrate / (1024 * 1024)); } if(m_capture.IsCapturing()) { - s_stats += _T(" | Recording..."); + s_stats += " | Recording..."; } - if(m_perfmon.Get(GSPerfMon::COLCLAMP)) _tprintf(_T("*** NOT SUPPORTED: color wrap ***\n")); - if(m_perfmon.Get(GSPerfMon::PABE)) _tprintf(_T("*** NOT SUPPORTED: per pixel alpha blend ***\n")); - if(m_perfmon.Get(GSPerfMon::DATE)) _tprintf(_T("*** PERFORMANCE WARNING: destination alpha test used ***\n")); - if(m_perfmon.Get(GSPerfMon::ABE)) _tprintf(_T("*** NOT SUPPORTED: alpha blending mode ***\n")); - if(m_perfmon.Get(GSPerfMon::DepthTexture)) _tprintf(_T("*** NOT SUPPORTED: depth texture ***\n")); - - m_wnd.SetWindowText(s_stats); + m_wnd.SetWindowText(s_stats.c_str()); } if(m_osd) { - m_dev.Draw(s_stats + _T("\n\nF5: interlace mode\nF6: aspect ratio\nF7: OSD")); + m_dev.Draw(s_stats + "\n\nF5: interlace mode\nF6: aspect ratio\nF7: OSD"); } if(m_frameskip) @@ -535,11 +531,11 @@ public: DoCapture(); } - bool MakeSnapshot(LPCTSTR path) + bool MakeSnapshot(const string& path) { - if(m_snapshot.IsEmpty()) + if(m_snapshot.empty()) { - m_snapshot.Format(_T("%s_%s"), path, CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))); + m_snapshot = format("%s_%s", path.c_str(), CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))); } return true; diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index e0d1818d25..1fb8a3b597 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -261,9 +261,12 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp2\\_%05d_f%I64d_fr%d_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM); - if(s_save) rt->m_texture.Save(str); + if(s_save) + { + rt->m_texture.Save(format("c:\\temp2\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); + } + + s_n++; } return true; @@ -321,19 +324,45 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp2\\_%05d_f%I64d_tex_%05x_%d_%d%d_%02x_%02x_%02x_%02x.dds"), - s_n++, m_perfmon.GetFrame(), (int)context->TEX0.TBP0, (int)context->TEX0.PSM, - (int)context->CLAMP.WMS, (int)context->CLAMP.WMT, - (int)context->CLAMP.MINU, (int)context->CLAMP.MAXU, - (int)context->CLAMP.MINV, (int)context->CLAMP.MAXV); - if(PRIM->TME) if(s_save) tex->m_texture.Save(str, true); - str.Format(_T("c:\\temp2\\_%05d_f%I64d_tpx_%05x_%d.dds"), s_n-1, m_perfmon.GetFrame(), context->TEX0.CBP, context->TEX0.CPSM); - if(PRIM->TME && tex->m_palette) if(s_save) tex->m_palette.Save(str, true); - str.Format(_T("c:\\temp2\\_%05d_f%I64d_rt0_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), context->FRAME.Block(), context->FRAME.PSM); - if(s_save) rt->m_texture.Save(str); - str.Format(_T("c:\\temp2\\_%05d_f%I64d_rz0_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), context->ZBUF.Block(), context->ZBUF.PSM); - if(s_savez) ds->m_texture.Save(str); + UINT64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save && PRIM->TME) + { + s = format("c:\\temp2\\_%05d_f%I64d_tex_%05x_%d_%d%d_%02x_%02x_%02x_%02x.dds", + s_n, frame, (int)context->TEX0.TBP0, (int)context->TEX0.PSM, + (int)context->CLAMP.WMS, (int)context->CLAMP.WMT, + (int)context->CLAMP.MINU, (int)context->CLAMP.MAXU, + (int)context->CLAMP.MINV, (int)context->CLAMP.MAXV); + + tex->m_texture.Save(s, true); + + if(tex->m_palette) + { + s = format("c:\\temp2\\_%05d_f%I64d_tpx_%05x_%d.dds", s_n, frame, context->TEX0.CBP, context->TEX0.CPSM); + + tex->m_palette.Save(s, true); + } + } + + s_n++; + + if(s_save) + { + s = format("c:\\temp2\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, context->FRAME.Block(), context->FRAME.PSM); + + rt->m_texture.Save(s); + } + + if(s_savez) + { + s = format("c:\\temp2\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, context->ZBUF.Block(), context->ZBUF.PSM); + + ds->m_texture.Save(s); + } + + s_n++; } int prim = PRIM->PRIM; @@ -349,12 +378,25 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp2\\_%05d_f%I64d_rt1_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), context->FRAME.Block(), context->FRAME.PSM); - if(s_save) rt->m_texture.Save(str); - str.Format(_T("c:\\temp2\\_%05d_f%I64d_rz1_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), context->ZBUF.Block(), context->ZBUF.PSM); - if(s_savez) ds->m_texture.Save(str); - // if(s_savez) m_dev.SaveToFileD32S8X24(ds->m_texture, str); // TODO + UINT64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save) + { + s = format("c:\\temp2\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, context->FRAME.Block(), context->FRAME.PSM); + + rt->m_texture.Save(s); + } + + if(s_savez) + { + s = format("c:\\temp2\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, context->ZBUF.Block(), context->ZBUF.PSM); + + ds->m_texture.Save(s); + } + + s_n++; } m_tc->InvalidateTextures(context->FRAME, context->ZBUF); diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 0265f582d8..4069eb711c 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -30,7 +30,7 @@ GSRendererHW10::GSRendererHW10(BYTE* base, bool mt, void (*irq)(), const GSRende InitVertexKick(); } -bool GSRendererHW10::Create(LPCTSTR title) +bool GSRendererHW10::Create(const string& title) { if(!__super::Create(title)) return false; diff --git a/plugins/GSdx/GSRendererHW10.h b/plugins/GSdx/GSRendererHW10.h index a9985dfc0f..23a1dbfadf 100644 --- a/plugins/GSdx/GSRendererHW10.h +++ b/plugins/GSdx/GSRendererHW10.h @@ -50,7 +50,7 @@ protected: public: GSRendererHW10(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs); - bool Create(LPCTSTR title); + bool Create(const string& title); template void VertexKick(bool skip); }; \ No newline at end of file diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 47acb327a3..48bff43558 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -33,7 +33,7 @@ GSRendererHW9::GSRendererHW9(BYTE* base, bool mt, void (*irq)(), const GSRendere InitVertexKick(); } -bool GSRendererHW9::Create(LPCTSTR title) +bool GSRendererHW9::Create(const string& title) { if(!__super::Create(title)) return false; diff --git a/plugins/GSdx/GSRendererHW9.h b/plugins/GSdx/GSRendererHW9.h index c91a0cb42d..073e0c552e 100644 --- a/plugins/GSdx/GSRendererHW9.h +++ b/plugins/GSdx/GSRendererHW9.h @@ -59,7 +59,7 @@ protected: public: GSRendererHW9(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs); - bool Create(LPCTSTR title); + bool Create(const string& title); template void VertexKick(bool skip); }; diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 2f540cacc6..86de0df48f 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -109,9 +109,12 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp1\\_%05d_f%I64d_fr%d_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM); - if(s_save) t.Save(str); + if(s_save) + { + t.Save(format("c:\\temp1\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); + } + + s_n++; } return true; @@ -550,13 +553,34 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp1\\_%05d_f%I64d_tex_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), (int)m_context->TEX0.TBP0, (int)m_context->TEX0.PSM); - if(PRIM->TME) if(s_save) {m_mem.SaveBMP(str, m_context->TEX0.TBP0, m_context->TEX0.TBW, m_context->TEX0.PSM, 1 << m_context->TEX0.TW, 1 << m_context->TEX0.TH);} - str.Format(_T("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), m_context->FRAME.Block(), m_context->FRAME.PSM); - if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);}//GetFrameSize(1).cy); - str.Format(_T("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), m_context->ZBUF.Block(), m_context->ZBUF.PSM); - if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512);} + UINT64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save && PRIM->TME) + { + s = format("c:\\temp1\\_%05d_f%I64d_tex_%05x_%d.bmp", s_n, frame, (int)m_context->TEX0.TBP0, (int)m_context->TEX0.PSM); + + m_mem.SaveBMP(s, m_context->TEX0.TBP0, m_context->TEX0.TBW, m_context->TEX0.PSM, 1 << m_context->TEX0.TW, 1 << m_context->TEX0.TH); + } + + s_n++; + + if(s_save) + { + s = format("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); + + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);//GetFrameSize(1).cy); + } + + if(s_savez) + { + s = format("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); + + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512); + } + + s_n++; } GSRasterizerData data; @@ -611,11 +635,25 @@ protected: if(s_dump) { - CString str; - str.Format(_T("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp"), s_n++, m_perfmon.GetFrame(), m_context->FRAME.Block(), m_context->FRAME.PSM); - if(s_save) {m_mem.SaveBMP(str, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);}//GetFrameSize(1).cy); - str.Format(_T("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp"), s_n-1, m_perfmon.GetFrame(), m_context->ZBUF.Block(), m_context->ZBUF.PSM); - if(s_savez) {m_mem.SaveBMP(str, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512);} + UINT64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save) + { + s = format("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); + + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);//GetFrameSize(1).cy); + } + + if(s_savez) + { + s = format("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); + + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512); + } + + s_n++; } if(0)//stats.ticks > 5000000) diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 333744a560..133cc864d6 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -147,17 +147,17 @@ BOOL GSSettingsDlg::OnInitDialog() memset(&caps, 0, sizeof(caps)); caps.PixelShaderVersion = D3DPS_VERSION(0, 0); - m_modes.RemoveAll(); + m_modes.clear(); // windowed { D3DDISPLAYMODE mode; memset(&mode, 0, sizeof(mode)); - m_modes.AddTail(mode); + m_modes.push_back(mode); int iItem = m_resolution.AddString(_T("Windowed")); - m_resolution.SetItemDataPtr(iItem, m_modes.GetTailPosition()); + m_resolution.SetItemDataPtr(iItem, &m_modes.back()); m_resolution.SetCurSel(iItem); } @@ -177,12 +177,11 @@ BOOL GSSettingsDlg::OnInitDialog() if(S_OK == d3d->EnumAdapterModes(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8, i, &mode)) { - CString str; - str.Format(_T("%dx%d %dHz"), mode.Width, mode.Height, mode.RefreshRate); - int iItem = m_resolution.AddString(str); + string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); + int iItem = m_resolution.AddString(str.c_str()); - m_modes.AddTail(mode); - m_resolution.SetItemDataPtr(iItem, m_modes.GetTailPosition()); + m_modes.push_back(mode); + m_resolution.SetItemDataPtr(iItem, &m_modes.back()); if(ModeWidth == mode.Width && ModeHeight == mode.Height && ModeRefreshRate == mode.RefreshRate) { @@ -196,16 +195,16 @@ BOOL GSSettingsDlg::OnInitDialog() bool isdx10avail = GSUtil::IsDirect3D10Available(); - CAtlArray renderers; + vector renderers; for(size_t i = 0; i < countof(g_renderers); i++) { if(i >= 3 && i <= 5 && !isdx10avail) continue; - renderers.Add(g_renderers[i]); + renderers.push_back(g_renderers[i]); } - GSSetting::InitComboBox(renderers.GetData(), renderers.GetCount(), m_renderer, pApp->GetProfileInt(_T("Settings"), _T("Renderer"), 0)); + GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, pApp->GetProfileInt(_T("Settings"), _T("Renderer"), 0)); GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, pApp->GetProfileInt(_T("Settings"), _T("PixelShaderVersion2"), D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); GSSetting::InitComboBox(g_interlace, countof(g_interlace), m_interlace, pApp->GetProfileInt(_T("Settings"), _T("Interlace"), 0)); GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, pApp->GetProfileInt(_T("Settings"), _T("AspectRatio"), 1)); @@ -251,7 +250,8 @@ void GSSettingsDlg::OnOK() if(m_resolution.GetCurSel() >= 0) { - D3DDISPLAYMODE& mode = m_modes.GetAt((POSITION)m_resolution.GetItemData(m_resolution.GetCurSel())); + D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); + pApp->WriteProfileInt(_T("Settings"), _T("ModeWidth"), mode.Width); pApp->WriteProfileInt(_T("Settings"), _T("ModeHeight"), mode.Height); pApp->WriteProfileInt(_T("Settings"), _T("ModeRefreshRate"), mode.RefreshRate); diff --git a/plugins/GSdx/GSSettingsDlg.h b/plugins/GSdx/GSSettingsDlg.h index e2b68c41bf..1f48ce710d 100644 --- a/plugins/GSdx/GSSettingsDlg.h +++ b/plugins/GSdx/GSSettingsDlg.h @@ -29,7 +29,7 @@ class GSSettingsDlg : public CDialog DECLARE_DYNAMIC(GSSettingsDlg) private: - CAtlList m_modes; + list m_modes; public: GSSettingsDlg(CWnd* pParent = NULL); // standard constructor diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 81bba7ec3d..10aa63ff4d 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -1000,11 +1000,11 @@ void GSState::FlushWrite() InvalidateVideoMem(m_env.BITBLTBUF, r); /* static int n = 0; - CString str; - str.Format(_T("c:\\temp1\\[%04d]_%05x_%d_%d_%d_%d_%d_%d.bmp"), + string s; + s = format("c:\\temp1\\[%04d]_%05x_%d_%d_%d_%d_%d_%d.bmp", n++, (int)m_env.BITBLTBUF.DBP, (int)m_env.BITBLTBUF.DBW, (int)m_env.BITBLTBUF.DPSM, r.left, r.top, r.right, r.bottom); - m_mem.SaveBMP(str, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, r.right, r.bottom); + m_mem.SaveBMP(s, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, r.right, r.bottom); */ } @@ -1275,6 +1275,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) break; case GIF_FLG_IMAGE: + { int len = (int)min(size, path.nloop); diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index fbf204ada8..39d8af3a27 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -42,7 +42,7 @@ public: virtual bool Update(const CRect& r, const void* data, int pitch) = 0; virtual bool Map(BYTE** bits, int& pitch, const RECT* r = NULL) = 0; virtual void Unmap() = 0; - virtual bool Save(CString fn, bool dds = false) = 0; + virtual bool Save(const string& fn, bool dds = false) = 0; CSize GetSize() {return CSize(GetWidth(), GetHeight());} }; diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 615e8c25bc..4b14de0539 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -109,7 +109,7 @@ void GSTexture10::Unmap() } } -bool GSTexture10::Save(CString fn, bool dds) +bool GSTexture10::Save(const string& fn, bool dds) { CComPtr res; @@ -164,7 +164,7 @@ bool GSTexture10::Save(CString fn, bool dds) res = m_texture; } - return SUCCEEDED(D3DX10SaveTextureToFile(res, dds ? D3DX10_IFF_DDS : D3DX10_IFF_BMP, fn)); + return SUCCEEDED(D3DX10SaveTextureToFile(res, dds ? D3DX10_IFF_DDS : D3DX10_IFF_BMP, fn.c_str())); } ID3D10Texture2D* GSTexture10::operator->() diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index 9c1eb30964..3609224696 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -46,7 +46,7 @@ public: bool Update(const CRect& r, const void* data, int pitch); bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); void Unmap(); - bool Save(CString fn, bool dds = false); + bool Save(const string& fn, bool dds = false); ID3D10Texture2D* operator->(); // TODO: remove direct access diff --git a/plugins/GSdx/GSTexture7.cpp b/plugins/GSdx/GSTexture7.cpp index e69406e928..c817e2130d 100644 --- a/plugins/GSdx/GSTexture7.cpp +++ b/plugins/GSdx/GSTexture7.cpp @@ -166,7 +166,7 @@ void GSTexture7::Unmap() } } -bool GSTexture7::Save(CString fn, bool dds) +bool GSTexture7::Save(const string& fn, bool dds) { // TODO diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index 62eef7ea28..b2bdc0656a 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -47,7 +47,7 @@ public: bool Update(const CRect& r, const void* data, int pitch); bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); void Unmap(); - bool Save(CString fn, bool dds = false); + bool Save(const string& fn, bool dds = false); IDirectDrawSurface7* operator->(); // TODO: remove direct access diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index bff9df141c..b12fc2f684 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -138,7 +138,7 @@ void GSTexture9::Unmap() } } -bool GSTexture9::Save(CString fn, bool dds) +bool GSTexture9::Save(const string& fn, bool dds) { CComPtr surface; @@ -181,12 +181,12 @@ bool GSTexture9::Save(CString fn, bool dds) if(surface != NULL) { - return SUCCEEDED(D3DXSaveSurfaceToFile(fn, dds ? D3DXIFF_DDS : D3DXIFF_BMP, surface, NULL, NULL)); + return SUCCEEDED(D3DXSaveSurfaceToFile(fn.c_str(), dds ? D3DXIFF_DDS : D3DXIFF_BMP, surface, NULL, NULL)); } /* if(CComQIPtr texture = surface) { - return SUCCEEDED(D3DXSaveTextureToFile(fn, dds ? D3DXIFF_DDS : D3DXIFF_BMP, texture, NULL)); + return SUCCEEDED(D3DXSaveTextureToFile(fn.c_str(), dds ? D3DXIFF_DDS : D3DXIFF_BMP, texture, NULL)); } */ return false; diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index c952cda68e..fe46918e34 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -45,7 +45,7 @@ public: bool Update(const CRect& r, const void* data, int pitch); bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); void Unmap(); - bool Save(CString fn, bool dds = false); + bool Save(const string& fn, bool dds = false); IDirect3DTexture9* operator->(); // TODO: remove direct access diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index bc84dad34c..6168bc88a4 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -114,11 +114,9 @@ public: CRect r(0, 0, w, h); - POSITION pos = m_dirty.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) { - const CRect& dirty = m_dirty.GetNext(pos).GetDirtyRect(m_TEX0) & r; + const CRect& dirty = i->GetDirtyRect(m_TEX0) & r; if(!(m_valid & dirty).IsRectEmpty()) { @@ -145,7 +143,7 @@ public: } } - m_dirty.RemoveAll(); + m_dirty.clear(); m_renderer->MinMaxUV(w, h, r); @@ -214,23 +212,21 @@ public: protected: GSRenderer* m_renderer; - CAtlList m_rt; - CAtlList m_ds; - CAtlList m_tex; + list m_rt; + list m_ds; + list m_tex; - template void RecycleByAge(CAtlList& l, int maxage = 60) + template void RecycleByAge(list& l, int maxage = 60) { - POSITION pos = l.GetHeadPosition(); - - while(pos) + for(list::iterator i = l.begin(); i != l.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - T* t = l.GetNext(pos); + T* t = *j; if(++t->m_age > maxage) { - l.RemoveAt(cur); + l.erase(j); delete t; } @@ -254,9 +250,26 @@ public: void RemoveAll() { - while(m_rt.GetCount()) delete m_rt.RemoveHead(); - while(m_ds.GetCount()) delete m_ds.RemoveHead(); - while(m_tex.GetCount()) delete m_tex.RemoveHead(); + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + { + delete *i; + } + + m_rt.clear(); + + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) + { + delete *i; + } + + m_ds.clear(); + + for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) + { + delete *i; + } + + m_tex.clear(); } GSRenderTarget* GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb = false) @@ -265,13 +278,13 @@ public: if(rt == NULL) { - for(POSITION pos = m_rt.GetHeadPosition(); pos; m_rt.GetNext(pos)) + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) { - GSRenderTarget* rt2 = m_rt.GetAt(pos); + GSRenderTarget* rt2 = *i; if(rt2->m_TEX0.TBP0 == TEX0.TBP0) { - m_rt.MoveToHead(pos); + m_rt.splice(m_rt.begin(), m_rt, i); rt = rt2; @@ -288,9 +301,9 @@ public: { // HACK: try to find something close to the base pointer - for(POSITION pos = m_rt.GetHeadPosition(); pos; m_rt.GetNext(pos)) + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) { - GSRenderTarget* rt2 = m_rt.GetAt(pos); + GSRenderTarget* rt2 = *i; if(rt2->m_TEX0.TBP0 <= TEX0.TBP0 && TEX0.TBP0 < rt2->m_TEX0.TBP0 + 0x700 && (!rt || rt2->m_TEX0.TBP0 >= rt->m_TEX0.TBP0)) { @@ -317,7 +330,7 @@ public: return NULL; } - m_rt.AddHead(rt); + m_rt.push_front(rt); } if(m_renderer->CanUpscale()) @@ -351,13 +364,13 @@ public: if(ds == NULL) { - for(POSITION pos = m_ds.GetHeadPosition(); pos; m_ds.GetNext(pos)) + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) { - GSDepthStencil* ds2 = m_ds.GetAt(pos); + GSDepthStencil* ds2 = *i; if(ds2->m_TEX0.TBP0 == TEX0.TBP0) { - m_ds.MoveToHead(pos); + m_ds.splice(m_ds.begin(), m_ds, i); ds = ds2; @@ -383,7 +396,7 @@ public: return NULL; } - m_ds.AddHead(ds); + m_ds.push_front(ds); } if(m_renderer->m_context->DepthWrite()) @@ -459,9 +472,9 @@ public: GSTexture* t = NULL; - for(POSITION pos = m_tex.GetHeadPosition(); pos; m_tex.GetNext(pos)) + for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) { - t = m_tex.GetAt(pos); + t = *i; if(GSUtil::HasSharedBits(t->m_TEX0.TBP0, t->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) { @@ -470,7 +483,7 @@ public: && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.i64 == t->m_CLAMP.i64)) && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) { - m_tex.MoveToHead(pos); + m_tex.splice(m_tex.begin(), m_tex, i); break; } @@ -481,11 +494,11 @@ public: if(t == NULL) { - for(POSITION pos = m_rt.GetHeadPosition(); pos; m_rt.GetNext(pos)) + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) { - GSRenderTarget* rt = m_rt.GetAt(pos); + GSRenderTarget* rt = *i; - if(rt->m_dirty.IsEmpty() && GSUtil::HasSharedBits(rt->m_TEX0.TBP0, rt->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + if(rt->m_dirty.empty() && GSUtil::HasSharedBits(rt->m_TEX0.TBP0, rt->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) { t = CreateTexture(); @@ -496,7 +509,7 @@ public: return NULL; } - m_tex.AddHead(t); + m_tex.push_front(t); break; } @@ -505,11 +518,11 @@ public: if(t == NULL) { - for(POSITION pos = m_ds.GetHeadPosition(); pos; m_ds.GetNext(pos)) + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) { - GSDepthStencil* ds = m_ds.GetAt(pos); + GSDepthStencil* ds = *i; - if(ds->m_dirty.IsEmpty() && ds->m_used && GSUtil::HasSharedBits(ds->m_TEX0.TBP0, ds->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + if(ds->m_dirty.empty() && ds->m_used && GSUtil::HasSharedBits(ds->m_TEX0.TBP0, ds->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) { t = CreateTexture(); @@ -520,7 +533,7 @@ public: return NULL; } - m_tex.AddHead(t); + m_tex.push_front(t); break; } @@ -538,7 +551,7 @@ public: return NULL; } - m_tex.AddHead(t); + m_tex.push_front(t); } if(pal > 0) @@ -574,18 +587,16 @@ public: void InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) { - POSITION pos = m_tex.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - GSTexture* t = m_tex.GetNext(pos); + GSTexture* t = *j; if(GSUtil::HasSharedBits(FRAME.Block(), FRAME.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM) || GSUtil::HasSharedBits(ZBUF.Block(), ZBUF.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) { - m_tex.RemoveAt(cur); + m_tex.erase(j); delete t; } @@ -596,26 +607,23 @@ public: { bool found = false; - POSITION pos = m_tex.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - GSTexture* t = m_tex.GetNext(pos); + GSTexture* t = *j; if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) { if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) { - t->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, r)); + t->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); found = true; } else { - m_tex.RemoveAt(cur); - + m_tex.erase(j); delete t; } } @@ -637,31 +645,29 @@ public: if(r2.bottom > 0 && r2.top < h && r2.right > 0 && r2.left < w) { - t->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, r2)); + t->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r2)); } } } } } - pos = m_rt.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - GSRenderTarget* rt = m_rt.GetNext(pos); + GSRenderTarget* rt = *j; if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) { if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM)) { - rt->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, r)); + rt->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); rt->m_TEX0.TBW = BITBLTBUF.DBW; } else { - m_rt.RemoveAt(cur); + m_rt.erase(j); delete rt; continue; } @@ -679,7 +685,7 @@ public: if(r.bottom > y) { // TODO: do not add this rect above too - rt->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); + rt->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); rt->m_TEX0.TBW = BITBLTBUF.DBW; continue; } @@ -689,24 +695,22 @@ public: // copypaste for ds - pos = m_ds.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_ds.begin(); i != m_ds.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - GSDepthStencil* ds = m_ds.GetNext(pos); + GSDepthStencil* ds = *j; if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, ds->m_TEX0.TBP0, ds->m_TEX0.PSM)) { if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM)) { - ds->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, r)); + ds->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); ds->m_TEX0.TBW = BITBLTBUF.DBW; } else { - m_ds.RemoveAt(cur); + m_ds.erase(j); delete ds; continue; } @@ -724,7 +728,7 @@ public: if(r.bottom > y) { // TODO: do not add this rect above too - ds->m_dirty.AddTail(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); + ds->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); ds->m_TEX0.TBW = BITBLTBUF.DBW; continue; } @@ -736,13 +740,11 @@ public: void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& r) { - POSITION pos = m_rt.GetHeadPosition(); - - while(pos) + for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) { - POSITION cur = pos; + list::iterator j = i++; - GSRenderTarget* rt = m_rt.GetNext(pos); + GSRenderTarget* rt = *j; if(GSUtil::HasSharedBits(BITBLTBUF.SBP, BITBLTBUF.SPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) { @@ -760,7 +762,7 @@ public: } else { - m_rt.RemoveAt(cur); + m_rt.erase(j); delete rt; continue; } diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 68c989325f..0ab8210bdc 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -39,7 +39,7 @@ void GSTextureCache10::GSRenderTargetHW10::Update() CRect r = m_dirty.GetDirtyRect(m_TEX0, m_texture.GetSize()); - m_dirty.RemoveAll(); + m_dirty.clear(); if(r.IsRectEmpty()) return; @@ -93,7 +93,7 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) return; } - if(!m_dirty.IsEmpty()) + if(!m_dirty.empty()) { return; } diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index 31fe12f84a..52f733b4de 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -39,7 +39,7 @@ void GSTextureCache9::GSRenderTarget9::Update() CRect r = m_dirty.GetDirtyRect(m_TEX0, m_texture.GetSize()); - m_dirty.RemoveAll(); + m_dirty.clear(); if(r.IsRectEmpty()) return; @@ -96,7 +96,7 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) return; } - if(!m_dirty.IsEmpty()) + if(!m_dirty.empty()) { return; } diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index 9342c20dda..2aed3b6b64 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -38,16 +38,14 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - const CAtlMap& map = m_map[TEX0.TBP0 >> 5]; + const hash_map& map = m_map[TEX0.TBP0 >> 5]; GSTexture* t = NULL; - POSITION pos = map.GetStartPosition(); - - while(pos) + for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) { - GSTexture* t2 = map.GetNextKey(pos); - + GSTexture* t2 = (*i).first; + // if(t2->m_TEX0.TBP0 != TEX0.TBP0 || t2->m_TEX0.TBW != TEX0.TBW || t2->m_TEX0.PSM != TEX0.PSM || t2->m_TEX0.TW != TEX0.TW || t2->m_TEX0.TH != TEX0.TH) if(((t2->m_TEX0.ai32[0] ^ TEX0.ai32[0]) | ((t2->m_TEX0.ai32[1] ^ TEX0.ai32[1]) & 3)) != 0) { @@ -70,7 +68,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE { t = new GSTexture(m_state); - t->m_pos = m_textures.AddTail(t); + m_textures[t] = true; int tw = 1 << TEX0.TW; int th = 1 << TEX0.TH; @@ -93,7 +91,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE continue; } - m_map[page].SetAt(t, true); + m_map[page][t] = true; } } } @@ -102,11 +100,11 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE { printf("!@#$%\n"); // memory allocation may fail if the game is too hungry - m_textures.RemoveAt(t->m_pos); + m_textures.erase(t); for(int i = 0; i < MAX_PAGES; i++) { - m_map[i].RemoveKey(t); + m_map[i].erase(t); } delete t; @@ -119,38 +117,34 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE void GSTextureCacheSW::RemoveAll() { - POSITION pos = m_textures.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_textures.begin(); i != m_textures.end(); i++) { - delete m_textures.GetNext(pos); + delete i->first; } - m_textures.RemoveAll(); + m_textures.clear(); for(int i = 0; i < MAX_PAGES; i++) { - m_map[i].RemoveAll(); + m_map[i].clear(); } } void GSTextureCacheSW::IncAge() { - POSITION pos = m_textures.GetHeadPosition(); - - while(pos) + for(hash_map::iterator i = m_textures.begin(); i != m_textures.end(); ) { - POSITION cur = pos; + hash_map::iterator j = i++; - GSTexture* t = m_textures.GetNext(pos); + GSTexture* t = j->first; if(++t->m_age > 30) { - m_textures.RemoveAt(cur); + m_textures.erase(j); for(int i = 0; i < MAX_PAGES; i++) { - m_map[i].RemoveKey(t); + m_map[i].erase(t); } delete t; @@ -187,13 +181,11 @@ void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, cons continue; } - const CAtlMap& map = m_map[page]; + const hash_map& map = m_map[page]; - POSITION pos = map.GetStartPosition(); - - while(pos) + for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) { - GSTexture* t = map.GetNextKey(pos); + GSTexture* t = (*i).first; t->m_valid[page] = 0; @@ -210,7 +202,6 @@ GSTextureCacheSW::GSTexture::GSTexture(GSState* state) , m_buff(NULL) , m_tw(0) , m_age(0) - , m_pos(NULL) , m_complete(false) { memset(m_valid, 0, sizeof(m_valid)); diff --git a/plugins/GSdx/GSTextureCacheSW.h b/plugins/GSdx/GSTextureCacheSW.h index 1214c9e059..540cdd6e1b 100644 --- a/plugins/GSdx/GSTextureCacheSW.h +++ b/plugins/GSdx/GSTextureCacheSW.h @@ -39,7 +39,6 @@ public: DWORD m_tw; DWORD m_valid[MAX_PAGES]; // each DWORD bits map to the 32 blocks of that page DWORD m_age; - POSITION m_pos; bool m_complete; explicit GSTexture(GSState* state); @@ -50,8 +49,8 @@ public: protected: GSState* m_state; - CAtlList m_textures; - CAtlMap m_map[MAX_PAGES]; + hash_map m_textures; + hash_map m_map[MAX_PAGES]; public: GSTextureCacheSW(GSState* state); diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 0f3073d9cf..80167bcc3a 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -163,27 +163,29 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { CComPtr vs; - if(CRBMap >::CPair* pair = m_vs.Lookup(sel)) + hash_map >::iterator i = m_vs.find(sel); + + if(i != m_vs.end()) { - vs = pair->m_value; + vs = (*i).second; } else { - CStringA str[5]; + string str[5]; - str[0].Format("%d", sel.bpp); - str[1].Format("%d", sel.bppz); - str[2].Format("%d", sel.tme); - str[3].Format("%d", sel.fst); - str[4].Format("%d", sel.prim); + str[0] = format("%d", sel.bpp); + str[1] = format("%d", sel.bppz); + str[2] = format("%d", sel.tme); + str[3] = format("%d", sel.fst); + str[4] = format("%d", sel.prim); D3D10_SHADER_MACRO macro[] = { - {"VS_BPP", str[0]}, - {"VS_BPPZ", str[1]}, - {"VS_TME", str[2]}, - {"VS_FST", str[3]}, - {"VS_PRIM", str[4]}, + {"VS_BPP", str[0].c_str()}, + {"VS_BPPZ", str[1].c_str()}, + {"VS_TME", str[2].c_str()}, + {"VS_FST", str[3].c_str()}, + {"VS_PRIM", str[4].c_str()}, {NULL, NULL}, }; @@ -206,7 +208,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) m_il = il; } - m_vs.SetAt(sel, vs); + m_vs[sel] = vs; } if(m_vs_cb_cache.Update(cb)) @@ -227,27 +229,29 @@ bool GSTextureFX10::SetupGS(GSSelector sel) if(sel.prim > 0 && (sel.iip == 0 || sel.prim == 3)) // geometry shader works in every case, but not needed { - if(CRBMap >::CPair* pair = m_gs.Lookup(sel)) + hash_map >::iterator i = m_gs.find(sel); + + if(i != m_gs.end()) { - gs = pair->m_value; + gs = (*i).second; } else { - CStringA str[2]; + string str[2]; - str[0].Format("%d", sel.iip); - str[1].Format("%d", sel.prim); + str[0] = format("%d", sel.iip); + str[1] = format("%d", sel.prim); D3D10_SHADER_MACRO macro[] = { - {"IIP", str[0]}, - {"PRIM", str[1]}, + {"IIP", str[0].c_str()}, + {"PRIM", str[1].c_str()}, {NULL, NULL}, }; hr = m_dev->CompileShader(IDR_TFX10_FX, "gs_main", macro, &gs); - m_gs.SetAt(sel, gs); + m_gs[sel] = gs; } } @@ -271,49 +275,51 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl CComPtr ps; - if(CRBMap >::CPair* pair = m_ps.Lookup(sel)) + hash_map >::iterator i = m_ps.find(sel); + + if(i != m_ps.end()) { - ps = pair->m_value; + ps = (*i).second; } else { - CStringA str[13]; + string str[13]; - str[0].Format("%d", sel.fst); - str[1].Format("%d", sel.wms); - str[2].Format("%d", sel.wmt); - str[3].Format("%d", sel.bpp); - str[4].Format("%d", sel.aem); - str[5].Format("%d", sel.tfx); - str[6].Format("%d", sel.tcc); - str[7].Format("%d", sel.ate); - str[8].Format("%d", sel.atst); - str[9].Format("%d", sel.fog); - str[10].Format("%d", sel.clr1); - str[11].Format("%d", sel.fba); - str[12].Format("%d", sel.aout); + str[0] = format("%d", sel.fst); + str[1] = format("%d", sel.wms); + str[2] = format("%d", sel.wmt); + str[3] = format("%d", sel.bpp); + str[4] = format("%d", sel.aem); + str[5] = format("%d", sel.tfx); + str[6] = format("%d", sel.tcc); + str[7] = format("%d", sel.ate); + str[8] = format("%d", sel.atst); + str[9] = format("%d", sel.fog); + str[10] = format("%d", sel.clr1); + str[11] = format("%d", sel.fba); + str[12] = format("%d", sel.aout); D3D10_SHADER_MACRO macro[] = { - {"FST", str[0]}, - {"WMS", str[1]}, - {"WMT", str[2]}, - {"BPP", str[3]}, - {"AEM", str[4]}, - {"TFX", str[5]}, - {"TCC", str[6]}, - {"ATE", str[7]}, - {"ATST", str[8]}, - {"FOG", str[9]}, - {"CLR1", str[10]}, - {"FBA", str[11]}, - {"AOUT", str[12]}, + {"FST", str[0].c_str()}, + {"WMS", str[1].c_str()}, + {"WMT", str[2].c_str()}, + {"BPP", str[3].c_str()}, + {"AEM", str[4].c_str()}, + {"TFX", str[5].c_str()}, + {"TCC", str[6].c_str()}, + {"ATE", str[7].c_str()}, + {"ATST", str[8].c_str()}, + {"FOG", str[9].c_str()}, + {"CLR1", str[10].c_str()}, + {"FBA", str[11].c_str()}, + {"AOUT", str[12].c_str()}, {NULL, NULL}, }; hr = m_dev->CompileShader(IDR_TFX10_FX, "ps_main", macro, &ps); - m_ps.SetAt(sel, ps); + m_ps[sel] = ps; } if(m_ps_cb_cache.Update(cb)) @@ -332,9 +338,11 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl ssel.min = ssel.mag = 0; } - if(CRBMap >::CPair* pair = m_ps_ss.Lookup(ssel)) + hash_map >::iterator i = m_ps_ss.find(ssel); + + if(i != m_ps_ss.end()) { - ss0 = pair->m_value; + ss0 = (*i).second; } else { @@ -358,7 +366,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl hr = (*m_dev)->CreateSamplerState(&sd, &ss0); - m_ps_ss.SetAt(ssel, ss0); + m_ps_ss[ssel] = ss0; } if(sel.bpp == 3) @@ -388,9 +396,11 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr dss; - if(CRBMap >::CPair* pair = m_om_dss.Lookup(dssel)) + hash_map >::iterator i = m_om_dss.find(dssel); + + if(i != m_om_dss.end()) { - dss = pair->m_value; + dss = (*i).second; } else { @@ -430,16 +440,18 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); - m_om_dss.SetAt(dssel, dss); + m_om_dss[dssel] = dss; } m_dev->OMSetDepthStencilState(dss, 1); CComPtr bs; - if(CRBMap >::CPair* pair = m_om_bs.Lookup(bsel)) + hash_map >::iterator j = m_om_bs.find(bsel); + + if(j != m_om_bs.end()) { - bs = pair->m_value; + bs = (*j).second; } else { @@ -571,7 +583,7 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, hr = (*m_dev)->CreateBlendState(&bd, &bs); - m_om_bs.SetAt(bsel, bs); + m_om_bs[bsel] = bs; } m_dev->OMSetBlendState(bs, bf); diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 6dcaedc29e..7c554e66c1 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -213,15 +213,15 @@ public: private: GSDevice10* m_dev; CComPtr m_il; - CRBMapC > m_vs; + hash_map > m_vs; CComPtr m_vs_cb; - CRBMapC > m_gs; - CRBMapC > m_ps; + hash_map > m_gs; + hash_map > m_ps; CComPtr m_ps_cb; - CRBMapC > m_ps_ss; + hash_map > m_ps_ss; CComPtr m_palette_ss; - CRBMapC > m_om_dss; - CRBMapC > m_om_bs; + hash_map > m_om_dss; + hash_map > m_om_bs; CComPtr m_vb, m_vb_old; int m_vb_max; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 82c213d165..c6709d9e1f 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -50,9 +50,11 @@ bool GSTextureFX9::CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix) { DWORD hash = (size << 20) | (msk << 10) | fix; - if(CRBMap::CPair* pair = m_mskfix.Lookup(hash)) + hash_map::iterator i = m_mskfix.find(hash); + + if(i != m_mskfix.end()) { - t = pair->m_value; + t = (*i).second; } else { @@ -74,7 +76,7 @@ bool GSTextureFX9::CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix) t.Unmap(); } - m_mskfix.SetAt(hash, t); + m_mskfix[hash] = t; } return true; @@ -93,25 +95,27 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { CComPtr vs; - if(CRBMap >::CPair* pair = m_vs.Lookup(sel)) + hash_map >::iterator i = m_vs.find(sel); + + if(i != m_vs.end()) { - vs = pair->m_value; + vs = (*i).second; } else { - CStringA str[4]; + string str[4]; - str[0].Format("%d", sel.bppz); - str[1].Format("%d", sel.tme); - str[2].Format("%d", sel.fst); - str[3].Format("%d", sel.logz); + str[0] = format("%d", sel.bppz); + str[1] = format("%d", sel.tme); + str[2] = format("%d", sel.fst); + str[3] = format("%d", sel.logz); D3DXMACRO macro[] = { - {"VS_BPPZ", str[0]}, - {"VS_TME", str[1]}, - {"VS_FST", str[2]}, - {"VS_LOGZ", str[3]}, + {"VS_BPPZ", str[0].c_str()}, + {"VS_TME", str[1].c_str()}, + {"VS_FST", str[2].c_str()}, + {"VS_LOGZ", str[3].c_str()}, {NULL, NULL}, }; @@ -133,7 +137,7 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) m_il = il; } - m_vs.SetAt(sel, vs); + m_vs[sel] = vs; } m_dev->VSSetShader(vs, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); @@ -187,47 +191,49 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample CComPtr ps; - if(CRBMap >::CPair* pair = m_ps.Lookup(sel)) + hash_map >::iterator i = m_ps.find(sel); + + if(i != m_ps.end()) { - ps = pair->m_value; + ps = (*i).second; } else { - CStringA str[12]; + string str[12]; - str[0].Format("%d", sel.fst); - str[1].Format("%d", sel.wms); - str[2].Format("%d", sel.wmt); - str[3].Format("%d", sel.bpp); - str[4].Format("%d", sel.aem); - str[5].Format("%d", sel.tfx); - str[6].Format("%d", sel.tcc); - str[7].Format("%d", sel.ate); - str[8].Format("%d", sel.atst); - str[9].Format("%d", sel.fog); - str[10].Format("%d", sel.clr1); - str[11].Format("%d", sel.rt); + str[0] = format("%d", sel.fst); + str[1] = format("%d", sel.wms); + str[2] = format("%d", sel.wmt); + str[3] = format("%d", sel.bpp); + str[4] = format("%d", sel.aem); + str[5] = format("%d", sel.tfx); + str[6] = format("%d", sel.tcc); + str[7] = format("%d", sel.ate); + str[8] = format("%d", sel.atst); + str[9] = format("%d", sel.fog); + str[10] = format("%d", sel.clr1); + str[11] = format("%d", sel.rt); D3DXMACRO macro[] = { - {"FST", str[0]}, - {"WMS", str[1]}, - {"WMT", str[2]}, - {"BPP", str[3]}, - {"AEM", str[4]}, - {"TFX", str[5]}, - {"TCC", str[6]}, - {"ATE", str[7]}, - {"ATST", str[8]}, - {"FOG", str[9]}, - {"CLR1", str[10]}, - {"RT", str[11]}, + {"FST", str[0].c_str()}, + {"WMS", str[1].c_str()}, + {"WMT", str[2].c_str()}, + {"BPP", str[3].c_str()}, + {"AEM", str[4].c_str()}, + {"TFX", str[5].c_str()}, + {"TCC", str[6].c_str()}, + {"ATE", str[7].c_str()}, + {"ATST", str[8].c_str()}, + {"FOG", str[9].c_str()}, + {"CLR1", str[10].c_str()}, + {"RT", str[11].c_str()}, {NULL, NULL}, }; hr = m_dev->CompileShader(IDR_TFX9_FX, "ps_main", macro, &ps); - m_ps.SetAt(sel, ps); + m_ps[sel] = ps; } m_dev->PSSetShader(ps, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); @@ -241,9 +247,11 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample ssel.min = ssel.mag = 0; } - if(CRBMap::CPair* pair = m_ps_ss.Lookup(ssel)) + hash_map::iterator i = m_ps_ss.find(ssel); + + if(i != m_ps_ss.end()) { - ss = pair->m_value; + ss = (*i).second; } else { @@ -259,7 +267,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample ss->AddressU = ssel.tau ? D3DTADDRESS_WRAP : D3DTADDRESS_CLAMP; ss->AddressV = ssel.tav ? D3DTADDRESS_WRAP : D3DTADDRESS_CLAMP; - m_ps_ss.SetAt(ssel, ss); + m_ps_ss[ssel] = ss; } } @@ -282,9 +290,11 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, { Direct3DDepthStencilState9* dss = NULL; - if(CRBMap::CPair* pair = m_om_dss.Lookup(dssel)) + hash_map::iterator i = m_om_dss.find(dssel); + + if(i != m_om_dss.end()) { - dss = pair->m_value; + dss = (*i).second; } else { @@ -318,16 +328,18 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, dss->DepthFunc = ztst[dssel.ztst]; } - m_om_dss.SetAt(dssel, dss); + m_om_dss[dssel] = dss; } m_dev->OMSetDepthStencilState(dss, 3); Direct3DBlendState9* bs = NULL; - if(CRBMap::CPair* pair = m_om_bs.Lookup(bsel)) + hash_map::iterator j = m_om_bs.find(bsel); + + if(j != m_om_bs.end()) { - bs = pair->m_value; + bs = (*j).second; } else { @@ -457,7 +469,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(bsel.wb) bs->RenderTargetWriteMask |= D3DCOLORWRITEENABLE_BLUE; if(bsel.wa) bs->RenderTargetWriteMask |= D3DCOLORWRITEENABLE_ALPHA; - m_om_bs.SetAt(bsel, bs); + m_om_bs[bsel] = bs; } m_dev->OMSetBlendState(bs, 0x010101 * bf); diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index e2811059c9..a8c1ab9a3f 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -149,13 +149,13 @@ public: private: GSDevice9* m_dev; CComPtr m_il; - CRBMapC > m_vs; + hash_map > m_vs; D3DXHANDLE m_vs_params; - CRBMapC > m_ps; - CRBMapC m_ps_ss; - CRBMapC m_om_dss; - CRBMapC m_om_bs; - CRBMapC m_mskfix; + hash_map > m_ps; + hash_map m_ps_ss; + hash_map m_om_dss; + hash_map m_om_bs; + hash_map m_mskfix; public: GSTextureFX9(); diff --git a/plugins/GSdx/GSTextureNull.h b/plugins/GSdx/GSTextureNull.h index 38f21990fe..1d6e07868b 100644 --- a/plugins/GSdx/GSTextureNull.h +++ b/plugins/GSdx/GSTextureNull.h @@ -40,5 +40,5 @@ public: bool Update(const CRect& r, const void* data, int pitch) {return true;} bool Map(BYTE** bits, int& pitch, const RECT* r = NULL) {return true;} void Unmap() {} - bool Save(CString fn, bool dds = false) {return false;} + bool Save(const string& fn, bool dds = false) {return false;} }; diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index 4f06fe7fe9..b6f391ddf3 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -148,11 +148,7 @@ void GSUtil::FitRect(CRect& r, int aspectratio) bool GSUtil::CheckDirectX() { - CString str; - - str.Format(_T("d3dx9_%d.dll"), D3DX_SDK_VERSION); - - if(HINSTANCE hDll = LoadLibrary(str)) + if(HINSTANCE hDll = LoadLibrary(format("d3dx9_%d.dll", D3DX_SDK_VERSION).c_str())) { FreeLibrary(hDll); } @@ -199,9 +195,9 @@ bool GSUtil::CheckSSE() { if(!_CheckSSE()) { - CString str; - str.Format(_T("This CPU does not support SSE %d.%02d"), _M_SSE >> 8, _M_SSE & 0xff); - AfxMessageBox(str, MB_OK); + string s = format("This CPU does not support SSE %d.%02d", _M_SSE >> 8, _M_SSE & 0xff); + + AfxMessageBox(s.c_str(), MB_OK); return false; } @@ -223,50 +219,40 @@ bool GSUtil::IsDirect3D10Available() char* GSUtil::GetLibName() { - CString str; + static string str = format("GSdx %d", SVN_REV); - str.Format(_T("GSdx %d"), SVN_REV); - - if(SVN_MODS) str += _T("m"); + if(SVN_MODS) str += "m"; #if _M_AMD64 - str += _T(" 64-bit"); + str += " 64-bit"; #endif - CAtlList sl; + list sl; #ifdef __INTEL_COMPILER - CString s; - s.Format(_T("Intel C++ %d.%02d"), __INTEL_COMPILER/100, __INTEL_COMPILER%100); - sl.AddTail(s); + sl.push_back(format("Intel C++ %d.%02d", __INTEL_COMPILER / 100, __INTEL_COMPILER % 100)); #elif _MSC_VER - CString s; - s.Format(_T("MSVC %d.%02d"), _MSC_VER/100, _MSC_VER%100); - sl.AddTail(s); + sl.push_back(format("MSVC %d.%02d", _MSC_VER / 100, _MSC_VER % 100)); #endif #if _M_SSE >= 0x402 - sl.AddTail(_T("SSE42")); + sl.push_back("SSE42"); #elif _M_SSE >= 0x401 - sl.AddTail(_T("SSE41")); + sl.push_back("SSE41"); #elif _M_SSE >= 0x301 - sl.AddTail(_T("SSSE3")); + sl.push_back("SSSE3"); #elif _M_SSE >= 0x200 - sl.AddTail(_T("SSE2")); + sl.push_back("SSE2"); #elif _M_SSE >= 0x100 - sl.AddTail(_T("SSE")); + sl.push_back("SSE"); #endif - POSITION pos = sl.GetHeadPosition(); - - while(pos) + for(list::iterator i = sl.begin(); i != sl.end(); ) { - if(pos == sl.GetHeadPosition()) str += _T(" ("); - str += sl.GetNext(pos); - str += pos ? _T(", ") : _T(")"); + if(i == sl.begin()) str += " ("; + str += *i; + str += ++i != sl.end() ? ", " : ")"; } - static char buff[256]; - strncpy(buff, CStringA(str), min(countof(buff)-1, str.GetLength())); - return buff; + return (char*)str.c_str(); } \ No newline at end of file diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 1d98b97822..2f413e3a0a 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -35,8 +35,8 @@ public: } }; -typedef __declspec(align(8)) GSVector2T GSVector2; -typedef __declspec(align(8)) GSVector2T GSVector2i; +typedef GSVector2T GSVector2; +typedef GSVector2T GSVector2i; class GSVector4; diff --git a/plugins/GSdx/GSWnd.cpp b/plugins/GSdx/GSWnd.cpp index f2d195a99c..c01f3390bb 100644 --- a/plugins/GSdx/GSWnd.cpp +++ b/plugins/GSdx/GSWnd.cpp @@ -35,7 +35,7 @@ GSWnd::~GSWnd() DestroyWindow(); } -bool GSWnd::Create(LPCTSTR title) +bool GSWnd::Create(const string& title) { CRect r; @@ -53,7 +53,7 @@ bool GSWnd::Create(LPCTSTR title) LPCTSTR wc = AfxRegisterWndClass(CS_VREDRAW|CS_HREDRAW|CS_DBLCLKS, AfxGetApp()->LoadStandardCursor(IDC_ARROW), 0, 0); - return !!CreateEx(0, wc, title, WS_OVERLAPPEDWINDOW, r, NULL, 0); + return !!CreateEx(0, wc, title.c_str(), WS_OVERLAPPEDWINDOW, r, NULL, 0); } void GSWnd::Show() diff --git a/plugins/GSdx/GSWnd.h b/plugins/GSdx/GSWnd.h index aa76393b6f..4de785f724 100644 --- a/plugins/GSdx/GSWnd.h +++ b/plugins/GSdx/GSWnd.h @@ -29,7 +29,7 @@ public: GSWnd(); virtual ~GSWnd(); - virtual bool Create(LPCTSTR title); + virtual bool Create(const string& title); void Show(); void Hide(); diff --git a/plugins/GSdx/GSdx.cpp b/plugins/GSdx/GSdx.cpp index 799d55ea5c..48a30d5e92 100644 --- a/plugins/GSdx/GSdx.cpp +++ b/plugins/GSdx/GSdx.cpp @@ -21,6 +21,8 @@ #include "stdafx.h" #include "GSdx.h" +#include +#include // // Note! diff --git a/plugins/GSdx/stdafx.cpp b/plugins/GSdx/stdafx.cpp index bda8426831..7811c93b2e 100644 --- a/plugins/GSdx/stdafx.cpp +++ b/plugins/GSdx/stdafx.cpp @@ -6,3 +6,34 @@ // TODO: reference any additional headers you need in STDAFX.H // and not in this file + +string format(const char* fmt, ...) +{ + va_list args; + va_start(args, fmt); + + int result = -1, length = 256; + + char* buffer = NULL; + + while(result == -1) + { + if(buffer) delete [] buffer; + + buffer = new char[length + 1]; + + memset(buffer, 0, length + 1); + + result = _vsnprintf(buffer, length, fmt, args); + + length *= 2; + } + + va_end(args); + + string s(buffer); + + delete [] buffer; + + return s; +} diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index 826f5a6029..8b3acecd95 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -36,9 +36,6 @@ #include // MFC support for Windows Common Controls #endif // _AFX_NO_AFXCMN_SUPPORT //#include -#include -#include -#include #include #include #include @@ -46,6 +43,17 @@ #include #include +#include +#include +#include +#include +#include + +using namespace std; +using namespace stdext; + +extern string format(const char* fmt, ...); + #if !defined(_M_SSE) && (defined(_M_AMD64) || defined(_M_IX86_FP) && _M_IX86_FP >= 2) #define _M_SSE 0x200 #endif @@ -73,35 +81,3 @@ #define D3DCOLORWRITEENABLE_RGBA (D3DCOLORWRITEENABLE_RGB|D3DCOLORWRITEENABLE_ALPHA) #define QI(i) (riid == __uuidof(i)) ? GetInterface((i*)this, ppv) : - -template class CRBMapC : public CRBMap -{ - // CRBMap + a cache for the last value (simple, but already a lot better) - - CPair* m_pair; - -public: - CRBMapC() : m_pair(NULL) {} - - CPair* Lookup(KINARGTYPE key) - { - if(m_pair && key == m_pair->m_key) - { - return m_pair; - } - - m_pair = __super::Lookup(key); - - return m_pair; - } - - POSITION SetAt(KINARGTYPE key, VINARGTYPE value) - { - POSITION pos = __super::SetAt(key, value); - - m_pair = __super::GetAt(pos); - - return pos; - } -}; - From 6083e3a64335d6b9f5064658c50b29095a2eb24a Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 11 May 2009 10:44:12 +0000 Subject: [PATCH 046/296] Linux: Add -m32 to the compiler flags of pcsx2 and all plugins. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1169 96395faa-99c1-11dd-bbfe-3dabce05a288 --- 3rdparty/SoundTouch/configure.ac | 10 ++++++++++ build.sh | 4 ++-- pcsx2/configure.ac | 19 ++++++++++++++----- plugins/CDVDiso/src/configure.ac | 15 +++++++++------ plugins/CDVDisoEFP/src/Linux/Makefile | 4 ++-- plugins/CDVDlinuz/Src/Linux/Makefile | 4 ++-- plugins/CDVDnull/configure.ac | 15 +++++++++------ plugins/FWnull/configure.ac | 15 +++++++++------ plugins/GSnull/configure.ac | 15 +++++++++------ plugins/PeopsSPU2/Makefile | 8 ++++---- plugins/SPU2null/configure.ac | 15 +++++++++------ plugins/USBnull/configure.ac | 15 +++++++++------ plugins/dev9null/configure.ac | 15 +++++++++------ plugins/spu2-x/src/configure.ac | 15 +++++++++------ plugins/zerogs/opengl/configure | 15 +++++++++------ plugins/zerogs/opengl/configure.ac | 15 +++++++++------ plugins/zeropad/configure.ac | 11 +++++++---- plugins/zerospu2/configure.ac | 15 +++++++++------ 18 files changed, 140 insertions(+), 85 deletions(-) diff --git a/3rdparty/SoundTouch/configure.ac b/3rdparty/SoundTouch/configure.ac index 6ee3f30f64..d7ed878820 100644 --- a/3rdparty/SoundTouch/configure.ac +++ b/3rdparty/SoundTouch/configure.ac @@ -11,6 +11,16 @@ AC_PROG_CXX AC_PROG_CC AC_PROG_RANLIB +CFLAGS= +CPPFLAGS= +CXXFLAGS= +CCASFLAGS= + +CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " +CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " +CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " +CCASFLAGS+=" -m32 " + dnl Check for 64bit CPU AC_MSG_CHECKING(for a x86-64 CPU) dnl if test "$build_os" == "target_os" diff --git a/build.sh b/build.sh index a6aaa8710d..94bc4c8acd 100644 --- a/build.sh +++ b/build.sh @@ -28,7 +28,7 @@ export PCSX2PLUGINS="`pwd`/bin/plugins" curdir=`pwd` echo "Building the Pcsx2 Suite." -echo "Note: will not compile on Linux x64." +echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies." cd ${curdir}/plugins sh build.sh $option @@ -39,7 +39,7 @@ exit 1 fi echo "Building Pcsx2." -echo "Note: will not compile on Linux x64." +echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies." cd ${curdir}/pcsx2 sh build.sh $option diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index 449e023f95..ca5fb7d265 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -7,6 +7,10 @@ AC_PROG_CPP([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) AC_PROG_RANLIB +#AS="$CC" +#AC_SUBST(AS) +#AC_SUBST(ASFLAGS) + dnl necessary for compiling assembly AM_PROG_AS @@ -50,14 +54,19 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+=" ${DEBUG_FLAGS} -Xlinker -zmuldefs " - CXXFLAGS+=" ${DEBUG_FLAGS} -fpermissive -Xlinker -zmuldefs " - CCASFLAGS+=" -D_DEBUG ${DEBUG_FLAGS} " + CFLAGS+=" ${DEBUG_FLAGS} -Xlinker -zmuldefs -m32 " + CXXFLAGS+=" ${DEBUG_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " + CCASFLAGS+=" -D_DEBUG ${DEBUG_FLAGS} -m32 " + #LDFLAGS+=" -m32 " + #ASFLAGS+=" -m32 " MYOBJDIR="Debug" else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+=" ${NORMAL_FLAGS}" - CXXFLAGS+=" ${NORMAL_FLAGS} -fpermissive -Xlinker -zmuldefs " + CFLAGS+=" ${NORMAL_FLAGS} -m32 " + CXXFLAGS+=" ${NORMAL_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " + CCASFLAGS+=" ${NORMAL_FLAGS} -m32 " + #LDFLAGS+=" -m32 " + #ASFLAGS+=" -m32 " MYOBJDIR="Release" fi AC_MSG_RESULT($debug) diff --git a/plugins/CDVDiso/src/configure.ac b/plugins/CDVDiso/src/configure.ac index b181d19477..d42c8b2865 100755 --- a/plugins/CDVDiso/src/configure.ac +++ b/plugins/CDVDiso/src/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(CDVDiso_SONAME,libCDVDiso.so.[$CDVDiso_CURRENT].[$CDVDiso_REVISION].[$C CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+=" -g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+=" -g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+=" -g -fPIC -Wall -Wno-unused-value -fpermissive " + CFLAGS+=" -g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+=" -g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+=" -g -fPIC -Wall -Wno-unused-value -fpermissive -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value " - CPPFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value " - CXXFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value -fpermissive " + CFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value -m32 " + CPPFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value -m32 " + CXXFLAGS+=" -O3 -fomit-frame-pointer -fPIC -Wno-unused-value -fpermissive -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/CDVDisoEFP/src/Linux/Makefile b/plugins/CDVDisoEFP/src/Linux/Makefile index 1fee325f34..4f1bf81b83 100644 --- a/plugins/CDVDisoEFP/src/Linux/Makefile +++ b/plugins/CDVDisoEFP/src/Linux/Makefile @@ -3,7 +3,7 @@ PLUGIN = libCDVDisoEFP.so PLUGINOBJS = CDVDiso.o PLUGINHEADERS = CDVDiso.h PLUGINFLAGS = -fPIC -Wall -O2 -fomit-frame-pointer -D_LARGEFILE64_SOURCE \ - -I.. -I. -I./Linux + -I.. -I. -I./Linux -m32 PLUGINLIBS = -lz -lbz2 CFG = cfgCDVDisoEFP @@ -14,7 +14,7 @@ CFGHEADERS = interface.h aboutbox.h mainbox.h selectionbox.h \ devicebox.h conversionbox.h progressbox.h messagebox.h \ tablerebuild.h device.h CD.h DVD.h CFGFLAGS = -fPIC -Wall -O2 -fomit-frame-pointer -D_LARGEFILE64_SOURCE \ - -I.. -I. -I./Linux + -I.. -I. -I./Linux -m32 CFGLIBS = -lz -lbz2 COMP = compCDVDisoEFP diff --git a/plugins/CDVDlinuz/Src/Linux/Makefile b/plugins/CDVDlinuz/Src/Linux/Makefile index be45567df9..a3525a8be4 100644 --- a/plugins/CDVDlinuz/Src/Linux/Makefile +++ b/plugins/CDVDlinuz/Src/Linux/Makefile @@ -3,14 +3,14 @@ PLUGIN = libCDVDlinuz.so PLUGINOBJS = CDVDlinuz.o ../buffer.o PLUGINHEADERS = CDVDlinuz.h ../buffer.h PLUGINFLAGS = -fPIC -Wall -O2 -fomit-frame-pointer -D_LARGEFILE64_SOURCE \ - -I.. -I. -I./Linux + -I.. -I. -I./Linux -m32 PLUGINLIBS = CFG = cfgCDVDlinuz CFGOBJS = aboutbox.o mainbox.o interface.o CFGHEADERS = aboutbox.h mainbox.h interface.h CFGFLAGS = -fPIC -Wall -O2 -fomit-frame-pointer -D_LARGEFILE64_SOURCE \ - -I.. -I. -I./Linux + -I.. -I. -I./Linux -m32 CFGLIBS = SHAREDOBJS = actualfile.o conf.o logfile.o \ diff --git a/plugins/CDVDnull/configure.ac b/plugins/CDVDnull/configure.ac index 0ef1c2ae4c..9075e3be14 100644 --- a/plugins/CDVDnull/configure.ac +++ b/plugins/CDVDnull/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(CDVDnull_SONAME,libCDVDnull.so.[$CDVDnull_CURRENT].[$CDVDnull_REVISION] CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/FWnull/configure.ac b/plugins/FWnull/configure.ac index c0f2fe3358..8171e574a4 100644 --- a/plugins/FWnull/configure.ac +++ b/plugins/FWnull/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(FWnull_SONAME,libFWnull.so.[$FWnull_CURRENT].[$FWnull_REVISION].[$FWnul CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/GSnull/configure.ac b/plugins/GSnull/configure.ac index 09f8e74eed..8973091c0e 100644 --- a/plugins/GSnull/configure.ac +++ b/plugins/GSnull/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(GSnull_SONAME,libGSnull.so.[$GSnull_CURRENT].[$GSnull_REVISION].[$GSnul CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/PeopsSPU2/Makefile b/plugins/PeopsSPU2/Makefile index 685c267710..c620587c0c 100644 --- a/plugins/PeopsSPU2/Makefile +++ b/plugins/PeopsSPU2/Makefile @@ -16,9 +16,9 @@ NOTHREADLIB = FALSE ############################################################################## CC = gcc -CCFLAGS1 = -fPIC -c -Wall -O3 -CCFLAGS2 = -fPIC -c -Wall -O2 -ffast-math -CCFLAGS3 = -fPIC -c -Wall -O3 -ffast-math -fomit-frame-pointer +CCFLAGS1 = -fPIC -c -Wall -O3 -m32 +CCFLAGS2 = -fPIC -c -Wall -O2 -ffast-math -m32 +CCFLAGS3 = -fPIC -c -Wall -O3 -ffast-math -fomit-frame-pointer -m32 INCLUDE = LINK = gcc @@ -32,7 +32,7 @@ ifeq ($(USEALSA), TRUE) CCFLAGS3+= -DUSEALSA else OBJ+= oss.o - LINKFLAGS = -shared -Wl,-soname,libspu2PeopsOSS.so.1.6 -fPIC -fomit-frame-pointer -o libspu2PeopsOSS.so.1.6 + LINKFLAGS = -shared -Wl,-soname,libspu2PeopsOSS.so.1.6 -fPIC -fomit-frame-pointer -o libspu2PeopsOSS.so.1.6 -m32 endif ifeq ($(NOTHREADLIB), TRUE) diff --git a/plugins/SPU2null/configure.ac b/plugins/SPU2null/configure.ac index 2cc0ffb87d..cee047ebe0 100644 --- a/plugins/SPU2null/configure.ac +++ b/plugins/SPU2null/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(SPU2null_SONAME,libSPU2null.so.[$SPU2null_CURRENT].[$SPU2null_REVISION] CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/USBnull/configure.ac b/plugins/USBnull/configure.ac index 4a5de1dc8d..a2177049ab 100644 --- a/plugins/USBnull/configure.ac +++ b/plugins/USBnull/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(USBnull_SONAME,libUSBnull.so.[$USBnull_CURRENT].[$USBnull_REVISION].[$U CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/dev9null/configure.ac b/plugins/dev9null/configure.ac index a14ec42970..7b9d8dbaef 100644 --- a/plugins/dev9null/configure.ac +++ b/plugins/dev9null/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(DEV9null_SONAME,libDEV9null.so.[$DEV9null_CURRENT].[$DEV9null_REVISION] CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/spu2-x/src/configure.ac b/plugins/spu2-x/src/configure.ac index eb2596fcef..37d1dd2f14 100644 --- a/plugins/spu2-x/src/configure.ac +++ b/plugins/spu2-x/src/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(SPU2X_SONAME,libSPU2X.so.[$SPU2X_CURRENT].[$SPU2X_REVISION].[$SPU2X_AGE CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/zerogs/opengl/configure b/plugins/zerogs/opengl/configure index 8650685df9..c0c7bb22fe 100644 --- a/plugins/zerogs/opengl/configure +++ b/plugins/zerogs/opengl/configure @@ -4428,6 +4428,7 @@ ZEROGS_SONAME=libZeroGSogl.so.$ZEROGS_CURRENT.$ZEROGS_REVISION.$ZEROGS_AGE CFLAGS=" -I/opt/cg/include -L/opt/cg/lib " CPPFLAGS=" -I/opt/cg/include -L/opt/cg/lib " CXXFLAGS=" -I/opt/cg/include -L/opt/cg/lib " +CCASFLAGS= { $as_echo "$as_me:$LINENO: checking debug build" >&5 $as_echo_n "checking debug build... " >&6; } @@ -4445,18 +4446,20 @@ cat >>confdefs.h <<\_ACEOF #define _DEBUG 1 _ACEOF - CFLAGS+="-g " - CPPFLAGS+="-g " - CXXFLAGS+="-g -fpermissive " + CFLAGS+="-g -m32 " + CPPFLAGS+="-g -m32 " + CXXFLAGS+="-g -fpermissive -m32 " + CCASFLAGS+=" -m32 " else cat >>confdefs.h <<\_ACEOF #define NDEBUG 1 _ACEOF - CFLAGS+="-O3 -fomit-frame-pointer " - CPPFLAGS+="-O3 -fomit-frame-pointer " - CXXFLAGS+="-O3 -fomit-frame-pointer -fpermissive " + CFLAGS+="-O3 -fomit-frame-pointer -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fpermissive -m32 " + CCASFLAGS+=" -m32 " fi { $as_echo "$as_me:$LINENO: result: $debug" >&5 $as_echo "$debug" >&6; } diff --git a/plugins/zerogs/opengl/configure.ac b/plugins/zerogs/opengl/configure.ac index b602230a08..040f6f7de5 100644 --- a/plugins/zerogs/opengl/configure.ac +++ b/plugins/zerogs/opengl/configure.ac @@ -22,6 +22,7 @@ AC_SUBST(ZEROGS_SONAME, libZeroGSogl.so.[$ZEROGS_CURRENT].[$ZEROGS_REVISION].[$Z CFLAGS=" -I/opt/cg/include -L/opt/cg/lib " CPPFLAGS=" -I/opt/cg/include -L/opt/cg/lib " CXXFLAGS=" -I/opt/cg/include -L/opt/cg/lib " +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -30,14 +31,16 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g " - CPPFLAGS+="-g " - CXXFLAGS+="-g -fpermissive " + CFLAGS+="-g -m32 " + CPPFLAGS+="-g -m32 " + CXXFLAGS+="-g -fpermissive -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer " - CPPFLAGS+="-O3 -fomit-frame-pointer " - CXXFLAGS+="-O3 -fomit-frame-pointer -fpermissive " + CFLAGS+="-O3 -fomit-frame-pointer -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fpermissive -m32 " + CCASFLAGS+=" -m32 " fi AC_MSG_RESULT($debug) diff --git a/plugins/zeropad/configure.ac b/plugins/zeropad/configure.ac index f8e9a86fea..95b38496c2 100644 --- a/plugins/zeropad/configure.ac +++ b/plugins/zeropad/configure.ac @@ -20,6 +20,7 @@ AC_SUBST(ZEROPAD_SONAME, libZeroPAD.so.[$ZEROPAD_CURRENT].[$ZEROPAD_REVISION].[$ CFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -28,12 +29,14 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g " - CXXFLAGS+="-g " + CFLAGS+="-g -m32 " + CXXFLAGS+="-g -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O2 -fomit-frame-pointer " - CXXFLAGS+="-O2 -fomit-frame-pointer " + CFLAGS+="-O2 -fomit-frame-pointer -m32 " + CXXFLAGS+="-O2 -fomit-frame-pointer -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/zerospu2/configure.ac b/plugins/zerospu2/configure.ac index bebc95f237..b0b2ce69cb 100644 --- a/plugins/zerospu2/configure.ac +++ b/plugins/zerospu2/configure.ac @@ -21,6 +21,7 @@ AC_SUBST(ZEROSPU2_SONAME, libZeroSPU2.so.[$ZEROSPU2_CURRENT].[$ZEROSPU2_REVISION CFLAGS= CPPFLAGS= CXXFLAGS= +CCASFLAGS= dnl Check for debug build AC_MSG_CHECKING(debug build) @@ -29,14 +30,16 @@ debug=$enableval,debug=no) if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value " + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " - CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value " + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) From a17abfa76e9928bbe1c0c5b5bf42f96e95050520 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 11 May 2009 12:12:56 +0000 Subject: [PATCH 047/296] Fix up r1169 a bit, and a few minor changes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1170 96395faa-99c1-11dd-bbfe-3dabce05a288 --- 3rdparty/SoundTouch/configure.ac | 6 +++--- pcsx2/Makefile.am | 37 ++++++++++++++++++-------------- pcsx2/Mdec.cpp | 4 +++- pcsx2/configure.ac | 17 --------------- 4 files changed, 27 insertions(+), 37 deletions(-) diff --git a/3rdparty/SoundTouch/configure.ac b/3rdparty/SoundTouch/configure.ac index d7ed878820..73f71cf3d0 100644 --- a/3rdparty/SoundTouch/configure.ac +++ b/3rdparty/SoundTouch/configure.ac @@ -16,9 +16,9 @@ CPPFLAGS= CXXFLAGS= CCASFLAGS= -CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " -CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " -CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " +CFLAGS+=" -m32 " +CPPFLAGS+=" -m32 " +CXXFLAGS+=" -m32 " CCASFLAGS+=" -m32 " dnl Check for 64bit CPU diff --git a/pcsx2/Makefile.am b/pcsx2/Makefile.am index eb5f646a7a..4a8ef1cf76 100644 --- a/pcsx2/Makefile.am +++ b/pcsx2/Makefile.am @@ -3,25 +3,30 @@ INCLUDES = -I@srcdir@/x86/ -I@srcdir@/../common/include -I@srcdir@/../3rdparty noinst_LIBRARIES = libpcsx2.a libpcsx2_a_SOURCES = \ -CDVD.cpp CDVDiso.cpp CDVDisodrv.cpp COP0.cpp COP2.cpp Cache.cpp CdRom.cpp Console.cpp HwRead.cpp HwWrite.cpp \ -Counters.cpp Decode_XA.cpp Elfheader.cpp FPU.cpp FPU2.cpp FiFo.cpp Gif.cpp GS.cpp Hw.cpp Interpreter.cpp \ -IopBios.cpp IopCounters.cpp IopDma.cpp IopHw.cpp IopMem.cpp IopSio2.cpp MMI.cpp MTGS.cpp \ -Memory.cpp MemoryCard.cpp Misc.cpp Patch.cpp PathUtils.cpp Plugins.cpp PrecompiledHeader.cpp R3000A.cpp \ -R3000AInterpreter.cpp R3000AOpcodeTables.cpp R5900.cpp R5900OpcodeImpl.cpp R5900OpcodeTables.cpp \ -SPR.cpp SaveState.cpp Sif.cpp Sio.cpp SourceLog.cpp Stats.cpp System.cpp ThreadTools.cpp \ -VU0.cpp VU0micro.cpp VU0microInterp.cpp VU1micro.cpp VU1microInterp.cpp VUflags.cpp VUmicroMem.cpp VUops.cpp \ -Vif.cpp VifDma.cpp vssprintf.cpp vtlb.cpp xmlpatchloader.cpp AlignedMalloc.cpp \ -RecoverySystem.cpp Saveslots.cpp Dump.cpp +AlignedMalloc.cpp Decode_XA.cpp HwRead.cpp MTGS.cpp R3000AInterpreter.cpp Sio.cpp VUflags.cpp \ +CDVD.cpp Dump.cpp HwWrite.cpp R3000AOpcodeTables.cpp SourceLog.cpp VUmicroMem.cpp \ +CDVDiso.cpp Elfheader.cpp Interpreter.cpp Memory.cpp R5900.cpp Stats.cpp VUops.cpp \ +CDVDisodrv.cpp FPU.cpp IopBios.cpp MemoryCard.cpp R5900OpcodeImpl.cpp System.cpp Vif.cpp \ +COP0.cpp FPU2.cpp IopCounters.cpp Misc.cpp R5900OpcodeTables.cpp ThreadTools.cpp VifDma.cpp \ +COP2.cpp FiFo.cpp IopDma.cpp Patch.cpp RecoverySystem.cpp VU0.cpp vssprintf.cpp \ +Cache.cpp GS.cpp IopHw.cpp PathUtils.cpp SPR.cpp VU0micro.cpp vtlb.cpp \ +CdRom.cpp Gif.cpp IopMem.cpp Plugins.cpp SaveState.cpp VU0microInterp.cpp xmlpatchloader.cpp \ +Console.cpp IopSio2.cpp PrecompiledHeader.cpp Saveslots.cpp VU1micro.cpp \ +Counters.cpp Hw.cpp MMI.cpp R3000A.cpp Sif.cpp VU1microInterp.cpp +#Mdec.cpp HashTools.cpp -libpcsx2_a_SOURCES += \ + +libpcsx2_a_SOURCES += \ ps2/Iop/IopHwRead.cpp ps2/Iop/IopHwWrite.cpp ps2/Iop/IopHw_Internal.h libpcsx2_a_SOURCES += \ -CDVD.h CDVDiso.h CDVDisodrv.h CDVDlib.h COP0.h Cache.h CdRom.h Common.h Counters.h Decode_XA.h \ -Elfheader.h Exceptions.h GS.h Hw.h IopBios.h IopBios2.h IopCounters.h IopDma.h IopHw.h IopMem.h IopSio2.h Memcpyfast.h \ -Memory.h MemoryTypes.h MemoryCard.h Misc.h Patch.h Paths.h Plugins.h PrecompiledHeader.h IopCommon.h R3000A.h R5900.h R5900OpcodeTables.h \ -SPR.h SamplProf.h SaveState.h Sif.h Sifcmd.h Sio.h SafeArray.h Stats.h StringUtils.h System.h Threading.h \ -VU.h VUflags.h VUmicro.h VUops.h Vif.h VifDma.h cheatscpp.h vtlb.h NakedAsm.h R5900Exceptions.h HostGui.h Pcsx2Config.h \ -Dump.h +CDVD.h Common.h HashMap.h IopDma.h MemoryCard.h PrecompiledHeader.h SafeArray.h StringUtils.h Vif.h \ +CDVDiso.h Counters.h HostGui.h IopHw.h MemoryTypes.h R3000A.h SamplProf.h System.h VifDma.h \ +CDVDisodrv.h Decode_XA.h Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ +CDVDlib.h Dump.h IopBios.h IopSio2.h NakedAsm.h R5900Exceptions.h Sif.h VU.h vtlb.h \ +COP0.h Elfheader.h IopBios2.h Mdec.h Patch.h R5900OpcodeTables.h Sifcmd.h VUflags.h \ +Cache.h Exceptions.h IopCommon.h MemcpyFast.h Paths.h Sio.h VUmicro.h \ +CdRom.h GS.h IopCounters.h Memory.h Plugins.h SPR.h Stats.h VUops.h +#RedtapeWindows.h SUBDIRS = x86 . DebugTools IPU RDebug tinyxml Linux \ No newline at end of file diff --git a/pcsx2/Mdec.cpp b/pcsx2/Mdec.cpp index 8a8f84ddf9..c6b5da470b 100644 --- a/pcsx2/Mdec.cpp +++ b/pcsx2/Mdec.cpp @@ -18,6 +18,8 @@ /* This code was based on the FPSE v0.08 Mdec decoder*/ +#ifdef 0 + #include #include @@ -413,7 +415,7 @@ int SaveState::mdecFreeze() { return 0; } - +#endif diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index ca5fb7d265..33ef288569 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -7,10 +7,6 @@ AC_PROG_CPP([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) AC_PROG_RANLIB -#AS="$CC" -#AC_SUBST(AS) -#AC_SUBST(ASFLAGS) - dnl necessary for compiling assembly AM_PROG_AS @@ -57,31 +53,18 @@ then CFLAGS+=" ${DEBUG_FLAGS} -Xlinker -zmuldefs -m32 " CXXFLAGS+=" ${DEBUG_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " CCASFLAGS+=" -D_DEBUG ${DEBUG_FLAGS} -m32 " - #LDFLAGS+=" -m32 " - #ASFLAGS+=" -m32 " MYOBJDIR="Debug" else AC_DEFINE(NDEBUG,1,[NDEBUG]) CFLAGS+=" ${NORMAL_FLAGS} -m32 " CXXFLAGS+=" ${NORMAL_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " CCASFLAGS+=" ${NORMAL_FLAGS} -m32 " - #LDFLAGS+=" -m32 " - #ASFLAGS+=" -m32 " MYOBJDIR="Release" fi AC_MSG_RESULT($debug) AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) -#AC_MSG_CHECKING(turn on memcpy_fast_) -#AC_ARG_ENABLE(memcpyfast, AC_HELP_STRING([--enable-memcpyfast], [Turns on memcpy_fast - EXPERIMENTAL]), -#memcpyfast=$enableval,memcpyfast=no) -#if test "x$memcpyfast" == xyes -#then -# AC_DEFINE(LINUX_USE_FAST_MEMORY,1,[LINUX_USE_FAST_MEMORY]) -#fi -#AC_MSG_RESULT($memcpyfast) - #AC_MSG_CHECKING(turn on microVU) #AC_ARG_ENABLE(microVU, AC_HELP_STRING([--enable-microVU], [Turns on the currently incomplete microVU files - Not a good idea]), # microVU=$enableval,microVU=no) From 29d11c81379716f096d4dfefdc4a7971fc928ba6 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Mon, 11 May 2009 21:10:08 +0000 Subject: [PATCH 048/296] GSdx: small fixes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1171 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRendererSW.h | 2 ++ plugins/GSdx/GSState.cpp | 21 ++++++++++++++------- 2 files changed, 16 insertions(+), 7 deletions(-) diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 86de0df48f..127e94ecd0 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -167,6 +167,8 @@ protected: if(a.z > 0xff) a.z = 0xff; break; case TFX_DECAL: + a.x = a.y; + a.z = a.w; break; case TFX_HIGHLIGHT: a.x = a.x + a.y; diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 10aa63ff4d..620f41959b 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -2051,14 +2051,21 @@ bool GSC_GodOfWar2(const GSFrameInfo& fi, int& skip) { if(skip == 0) { - if(fi.TME && fi.FBP == 0x00100 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x00100 && fi.TPSM == PSM_PSMCT16 // ntsc - || fi.TME && fi.FBP == 0x02100 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x02100 && fi.TPSM == PSM_PSMCT16) // pal + if(fi.TME) { - skip = 29; // shadows - } - else if(fi.TME && fi.FBP == 0x00500 && fi.FPSM == PSM_PSMCT24 && fi.TBP0 == 0x02100 && fi.TPSM == PSM_PSMCT32) // pal - { - // skip = 17; // only looks correct at native resolution + if(fi.FBP == 0x00100 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x00100 && fi.TPSM == PSM_PSMCT16 // ntsc + || fi.FBP == 0x02100 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x02100 && fi.TPSM == PSM_PSMCT16) // pal + { + skip = 29; // shadows + } + else if(fi.FBP == 0x00500 && fi.FPSM == PSM_PSMCT24 && fi.TBP0 == 0x02100 && fi.TPSM == PSM_PSMCT32) // pal + { + // skip = 17; // only looks correct at native resolution + } + else if(fi.FBP == 0x01300 && fi.FPSM == PSM_PSMCT24 && fi.TBP0 == 0x00100 && fi.TPSM == PSM_PSMCT32) // ntsc + { + // skip = 15; // only looks correct at native resolution + } } } else From 9e1b3fb33e60db929c92fa8ee7e534f385f8662a Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 12 May 2009 08:21:44 +0000 Subject: [PATCH 049/296] microVU: fixed a few problems... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1172 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 43 +++++++++++++++++++++++------------ pcsx2/x86/microVU_Execute.inl | 8 ++----- pcsx2/x86/microVU_Lower.inl | 4 ++-- pcsx2/x86/microVU_Upper.inl | 4 ++-- 4 files changed, 34 insertions(+), 25 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index cf136c9a0e..5c8bbd117f 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -28,7 +28,7 @@ //------------------------------------------------------------------ #define aReg(x) mVUregs.VF[x] -#define bReg(x) mVUregsTemp.VFreg[0] = x; mVUregsTemp.VF[0] +#define bReg(x, y) mVUregsTemp.VFreg[y] = x; mVUregsTemp.VF[0] #define aMax(x, y) ((x > y) ? x : y) #define analyzeReg1(reg) { \ @@ -40,13 +40,13 @@ } \ } -#define analyzeReg2(reg) { \ - if (reg) { \ - if (_X) { bReg(reg).x = 4; } \ - if (_Y) { bReg(reg).y = 4; } \ - if (_Z) { bReg(reg).z = 4; } \ - if (_W) { bReg(reg).w = 4; } \ - } \ +#define analyzeReg2(reg, isLowOp) { \ + if (reg) { \ + if (_X) { bReg(reg, isLowOp).x = 4; } \ + if (_Y) { bReg(reg, isLowOp).y = 4; } \ + if (_Z) { bReg(reg, isLowOp).z = 4; } \ + if (_W) { bReg(reg, isLowOp).w = 4; } \ + } \ } microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { @@ -54,7 +54,7 @@ microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg1(Ft); - analyzeReg2(Fd); + analyzeReg2(Fd, 0); } //------------------------------------------------------------------ @@ -64,7 +64,7 @@ microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { microVUt(void) mVUanalyzeFMAC2(int Fs, int Ft) { microVU* mVU = mVUx; analyzeReg1(Fs); - analyzeReg2(Ft); + analyzeReg2(Ft, 0); } //------------------------------------------------------------------ @@ -85,7 +85,7 @@ microVUt(void) mVUanalyzeFMAC3(int Fd, int Fs, int Ft) { mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg3(Ft); - analyzeReg2(Fd); + analyzeReg2(Fd, 0); } //------------------------------------------------------------------ @@ -143,7 +143,7 @@ microVUt(void) mVUanalyzeMR32(int Fs, int Ft) { microVU* mVU = mVUx; if (!Ft) { mVUinfo |= _isNOP; } analyzeReg6(Fs); - analyzeReg2(Ft); + analyzeReg2(Ft, 1); } //------------------------------------------------------------------ @@ -197,9 +197,22 @@ microVUt(void) mVUanalyzeEFU2(int Fs, u8 xCycles) { microVUt(void) mVUanalyzeMFP(int Ft) { microVU* mVU = mVUx; if (!Ft) { mVUinfo |= _isNOP; } - analyzeReg2(Ft); + analyzeReg2(Ft, 1); } +//------------------------------------------------------------------ +// MOVE - MOVE Opcode +//------------------------------------------------------------------ + +microVUt(void) mVUanalyzeMOVE(int Fs, int Ft) { + microVU* mVU = mVUx; + if (!Ft || (Ft == Fs)) { mVUinfo |= _isNOP; } + if (mVUregsTemp.VFreg[0] == Fs) { mVUinfo |= _swapOps; } + analyzeReg1(Fs); + analyzeReg2(Ft, 1); +} + + //------------------------------------------------------------------ // LQx - LQ/LQD/LQI Opcodes //------------------------------------------------------------------ @@ -207,7 +220,7 @@ microVUt(void) mVUanalyzeMFP(int Ft) { microVUt(void) mVUanalyzeLQ(int Ft, int Is, bool writeIs) { microVU* mVU = mVUx; analyzeVIreg1(Is); - analyzeReg2(Ft); + analyzeReg2(Ft, 1); if (!Ft) { mVUinfo |= (writeIs && Is) ? _noWriteVF : _isNOP; } if (writeIs) { analyzeVIreg2(Is, 1); } } @@ -238,7 +251,7 @@ microVUt(void) mVUanalyzeR1(int Fs, int Fsf) { microVUt(void) mVUanalyzeR2(int Ft, bool canBeNOP) { microVU* mVU = mVUx; if (!Ft) { mVUinfo |= ((canBeNOP) ? _isNOP : _noWriteVF); } - analyzeReg2(Ft); + analyzeReg2(Ft, 1); analyzeRreg(); } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index d207f157fa..a4ab238cd3 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -27,7 +27,6 @@ void testFunction() { mVUprint("microVU: Entered Execution Mode"); } // Generates the code for entering recompiled blocks microVUt(void) mVUdispatcherA() { - static u32 PCSX2_ALIGNED16(vuMXCSR); microVU* mVU = mVUx; mVU->startFunct = x86Ptr; @@ -42,8 +41,7 @@ microVUt(void) mVUdispatcherA() { PUSH32R(EDI); // Load VU's MXCSR state - vuMXCSR = g_sseVUMXCSR; - SSE_LDMXCSR((uptr)&vuMXCSR); + SSE_LDMXCSR((uptr)&g_sseVUMXCSR); // Load Regs MOV32MtoR(gprR, (uptr)&mVU->regs->VI[REG_R].UL); @@ -83,13 +81,11 @@ microVUt(void) mVUdispatcherA() { // Generates the code to exit from recompiled blocks microVUt(void) mVUdispatcherB() { - static u32 PCSX2_ALIGNED16(eeMXCSR); microVU* mVU = mVUx; mVU->exitFunct = x86Ptr; // Load EE's MXCSR state - eeMXCSR = g_sseMXCSR; - SSE_LDMXCSR((uptr)&eeMXCSR); + SSE_LDMXCSR((uptr)&g_sseMXCSR); // Save Regs (Other Regs Saved in mVUcompile) MOV32RtoM((uptr)&mVU->regs->VI[REG_R].UL, gprR); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index c829a0b964..493e2c9da3 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -699,7 +699,7 @@ microVUf(void) mVU_ISUBIU() { microVUf(void) mVU_MFIR() { microVU* mVU = mVUx; - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_); } + pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } pass2 { mVUallocVIa(gprT1, _Is_); MOVSX32R16toR(gprT1, gprT1); @@ -722,7 +722,7 @@ microVUf(void) mVU_MFP() { microVUf(void) mVU_MOVE() { microVU* mVU = mVUx; - pass1 { if (!_Ft_ || (_Ft_ == _Fs_)) { mVUinfo |= _isNOP; } analyzeReg1(_Fs_); analyzeReg2(_Ft_); } + pass1 { mVUanalyzeMOVE(_Fs_, _Ft_); } pass2 { mVUloadReg(xmmT1, (uptr)&mVU->regs->VF[_Fs_].UL[0], _X_Y_Z_W); mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index c80718581e..3102b6c75e 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -37,8 +37,8 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX //SysPrintf("doStatus = %d; doMac = %d\n", doStatus>>9, doMac>>8); if (mVUflagHack) { mVUinfo &= ~_doStatus; } if (!doFlags) return; - if (!doMac) { regT1 = reg; } - else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw + if (!doMac || (_XYZW_SS && modXYZW)) { regT1 = reg; } + else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw if (doStatus) { getFlagReg(sReg, fsInstance); // Set sReg to valid GPR by Cur Flag Instance mVUallocSFLAGa(sReg, fpsInstance); // Get Prev Status Flag From d363c21e8ffa77d0b37c96e505106c82eaa5e322 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 13 May 2009 05:43:40 +0000 Subject: [PATCH 050/296] microVU: -increased rec-cache so GOW2 should be happy (untested) -hacked-fixed some problems for now that I'll be working on properly fixing over the next few days. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1173 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 2 +- pcsx2/x86/microVU.h | 2 +- pcsx2/x86/microVU_Analyze.inl | 1 + pcsx2/x86/microVU_Execute.inl | 4 ++-- pcsx2/x86/microVU_Flags.inl | 8 +++++--- pcsx2/x86/microVU_Misc.h | 6 +++--- pcsx2/x86/microVU_Misc.inl | 2 +- 7 files changed, 14 insertions(+), 11 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index cc836a3a70..bcf478c048 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -163,7 +163,7 @@ microVUt(int) mVUfindLeastUsedProg() { }*/ mVUclearProg(j); // Clear old data if overwriting old program mVUcacheProg(j); // Cache Micro Program - Console::Notice("microVU%d: MicroProgram Cache Full!", params vuIndex); + //Console::Notice("microVU%d: MicroProgram Cache Full!", params vuIndex); return j; } } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 2137e9518c..f29f200b25 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -88,7 +88,7 @@ struct microProgManager { microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; -#define mVUcacheSize (0x1f00000 / ((vuIndex) ? 1 : 4)) +#define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) struct microVU { u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 5c8bbd117f..cc6d0d4cd7 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -234,6 +234,7 @@ microVUt(void) mVUanalyzeSQ(int Fs, int It, bool writeIt) { analyzeReg1(Fs); analyzeVIreg1(It); if (writeIt) { analyzeVIreg2(It, 1); } + if (mVUregsTemp.VFreg[0] == Fs) { mVUinfo |= _swapOps; } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index a4ab238cd3..85285c3c6b 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -53,11 +53,11 @@ microVUt(void) mVUdispatcherA() { SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); - SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, xmmT1); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); - SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, xmmT1); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->clipFlag, xmmT1); SSE_MOVAPS_M128_to_XMM(xmmACC, (uptr)&mVU->regs->ACC.UL[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 37a469a272..8a1a228059 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -53,6 +53,7 @@ microVUt(void) mVUstatusFlagOp() { } } iPC = curPC; + Console::Status("microVU%d: FSSET Optimization", params vuIndex); } int findFlagInst(int* fFlag, int cycles) { @@ -172,14 +173,15 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleMac); SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); } - +/* if (__Clip) { int bClip[4]; - sortFlag(xClip, bClip, cycles); + //sortFlag(xClip, bClip, cycles); + SysPrintf("__Clip\n"); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->clipFlag); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleClip); SSE_MOVAPS_XMM_to_M128((uptr)mVU->clipFlag, xmmT1); - } + }*/ } microVUt(void) mVUpass4(int startPC) { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 42f39247ee..c02f4a675c 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -222,11 +222,11 @@ declareAllVariables #define fmInstance ((mVUinfo >> 10) & 3) #define fsInstance ((mVUinfo >> 12) & 3) #define fpsInstance ((((mVUinfo>>12) & 3) - 1) & 0x3) -#define fcInstance ((mVUinfo >> 14) & 3) -#define fpcInstance ((((mVUinfo>>14) & 3) - 1) & 0x3) +#define fcInstance 0//((mVUinfo >> 14) & 3) +#define fpcInstance 0//((((mVUinfo>>14) & 3) - 1) & 0x3) #define fvmInstance ((mVUinfo >> 16) & 3) #define fvsInstance ((mVUinfo >> 18) & 3) -#define fvcInstance ((mVUinfo >> 20) & 3) +#define fvcInstance 0//((mVUinfo >> 20) & 3) #define backupVI (mVUinfo & (1<<22)) #define memReadIs (mVUinfo & (1<<23)) #define memReadIt (mVUinfo & (1<<24)) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 8f0e53850e..0299609bd6 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -299,7 +299,7 @@ microVUt(void) mVUrestoreRegs() { POP32R(gprR); // Restore EDX } - +// Reads entire microProgram and finds out if Status Flag is Used microVUt(void) mVUcheckSflag(int progIndex) { if (CHECK_VU_FLAGHACK) { From b1ddb5990cef12e0ee3c1182fb81c26d98b9bf2f Mon Sep 17 00:00:00 2001 From: sudonim1 Date: Wed, 13 May 2009 11:00:03 +0000 Subject: [PATCH 051/296] Found the typo responsible for the clipping bug. Changed a few lines from cotton's idea of readability to mine, at least until cotton reverts it. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1174 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 2 +- pcsx2/x86/microVU_Flags.inl | 9 ++++++--- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index cc6d0d4cd7..e10803938d 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -98,7 +98,7 @@ microVUt(void) mVUanalyzeFMAC3(int Fd, int Fs, int Ft) { microVUt(void) mVUanalyzeFMAC4(int Fs, int Ft) { microVU* mVU = mVUx; - mVUinfo |= doClip; + mVUinfo |= _doClip; analyzeReg1(Fs); analyzeReg4(Ft); } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 8a1a228059..b6608e41a8 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -132,9 +132,12 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { mVUinfo |= (xM & 3) << 10; // _fmInstance mVUinfo |= (xC & 3) << 14; // _fcInstance - if (doStatus||isFSSET||doDivFlag) { xStatus [xS] = cycles + 4; xS = (xS+1)&3; } - if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1)&3; } - if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1)&3; } + if (doStatus || isFSSET || doDivFlag) + xStatus[xS++ & 3] = cycles + 4; + if (doMac) + xMac[xM++ & 3] = cycles + 4; + if (doClip) + xClip[xC++ & 3] = cycles + 4; cycles++; incPC2(2); From 8385a8166f4dee2c4b057096e857534536da7b30 Mon Sep 17 00:00:00 2001 From: sudonim1 Date: Wed, 13 May 2009 11:30:02 +0000 Subject: [PATCH 052/296] And the same typo elsewhere. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1175 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Lower.inl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 493e2c9da3..5214fc0f85 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -485,7 +485,7 @@ microVUf(void) mVU_FCOR() { microVUf(void) mVU_FCSET() { microVU* mVU = mVUx; - pass1 { mVUinfo |= doClip; } + pass1 { mVUinfo |= _doClip; } pass2 { MOV32ItoR(gprT1, _Imm24_); mVUallocCFLAGb(gprT1, fcInstance); From 3c715556e9c4decc2aa064b33a77f406fb876f31 Mon Sep 17 00:00:00 2001 From: ramapcsx2 Date: Wed, 13 May 2009 12:15:41 +0000 Subject: [PATCH 053/296] microVU: due to the 2 flag fixes the hacks can go now :p git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1176 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Flags.inl | 10 +++++----- pcsx2/x86/microVU_Misc.h | 6 +++--- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index b6608e41a8..a448e0f3a2 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -143,7 +143,7 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { incPC2(2); } - mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) | ((__Status) ? 0 : xS); + mVUregs.flags = ((__Clip) ? 0 : ((xC & 3) << 2)) | ((__Status) ? 0 : (xS & 3)); return cycles; } @@ -176,15 +176,15 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleMac); SSE_MOVAPS_XMM_to_M128((uptr)mVU->macFlag, xmmT1); } -/* + if (__Clip) { int bClip[4]; - //sortFlag(xClip, bClip, cycles); - SysPrintf("__Clip\n"); + sortFlag(xClip, bClip, cycles); + //SysPrintf("__Clip\n"); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->clipFlag); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleClip); SSE_MOVAPS_XMM_to_M128((uptr)mVU->clipFlag, xmmT1); - }*/ + } } microVUt(void) mVUpass4(int startPC) { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index c02f4a675c..42f39247ee 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -222,11 +222,11 @@ declareAllVariables #define fmInstance ((mVUinfo >> 10) & 3) #define fsInstance ((mVUinfo >> 12) & 3) #define fpsInstance ((((mVUinfo>>12) & 3) - 1) & 0x3) -#define fcInstance 0//((mVUinfo >> 14) & 3) -#define fpcInstance 0//((((mVUinfo>>14) & 3) - 1) & 0x3) +#define fcInstance ((mVUinfo >> 14) & 3) +#define fpcInstance ((((mVUinfo>>14) & 3) - 1) & 0x3) #define fvmInstance ((mVUinfo >> 16) & 3) #define fvsInstance ((mVUinfo >> 18) & 3) -#define fvcInstance 0//((mVUinfo >> 20) & 3) +#define fvcInstance ((mVUinfo >> 20) & 3) #define backupVI (mVUinfo & (1<<22)) #define memReadIs (mVUinfo & (1<<23)) #define memReadIt (mVUinfo & (1<<24)) From b3bdaab4f07cbaaa12584b1381152a5157251b8c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 14 May 2009 07:36:10 +0000 Subject: [PATCH 054/296] microVU: - Implemented Nneeve's logical min/max algorithm. This should fix the problems with DaZ and mVU. - Applied a patch by Gigaherz that more clearly distinguishes Immediate values in microProgram log files. - Added a speedhack to disable the logical min/max code. (see below) Note: From my testing, using DaZ on mVU doesn't do much. However I have an AMD cpu, and they don't benefit as much as Intel C2D's from DaZ. So this could be effecting results. The logical min/max code is SLOW, and the little-benefit I get with DaZ means I get better performance with DaZ off and the min/max speedhack (which disables the extra min/max code). It would be nice is someone with an Intel C2D can compare the speed of: -mVU normal without DaZ -mVU normal with DaZ -mVU min/max speedhack without DaZ -mVU min/max speedhack with DaZ git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1177 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Alloc.inl | 4 +- pcsx2/x86/microVU_Flags.inl | 18 ++-- pcsx2/x86/microVU_Log.inl | 22 +++-- pcsx2/x86/microVU_Lower.inl | 6 +- pcsx2/x86/microVU_Misc.h | 8 +- pcsx2/x86/microVU_Misc.inl | 63 ++++++++++++- pcsx2/x86/microVU_Upper.inl | 176 ++++++++++++++++++------------------ 7 files changed, 180 insertions(+), 117 deletions(-) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index b33fdb7bb0..f9e7111a07 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -166,7 +166,7 @@ microVUt(void) mVUallocFMAC4a(int& ACC, int& Fs, int& Ft) { microVUt(void) mVUallocFMAC4b(int& ACC, int& Fs) { microVU* mVU = mVUx; if (CHECK_VU_OVERFLOW) mVUclamp1(Fs, xmmT1, _xyzw_ACC); - mVUmergeRegs(ACC, Fs, _X_Y_Z_W); + mVUmergeRegs(ACC, Fs, _X_Y_Z_W); } //------------------------------------------------------------------ @@ -435,7 +435,7 @@ microVUt(void) mVUallocFMAC14a(int& ACCw, int& ACCr, int& Fs, int& Ft) { microVUt(void) mVUallocFMAC14b(int& ACCw, int& ACCr) { microVU* mVU = mVUx; if (CHECK_VU_OVERFLOW) mVUclamp1(ACCr, xmmFt, _xyzw_ACC); - mVUmergeRegs(ACCw, ACCr, _X_Y_Z_W); + mVUmergeRegs(ACCw, ACCr, _X_Y_Z_W); } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index a448e0f3a2..7091fc711a 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -128,22 +128,19 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { mVUinfo |= findFlagInst(xMac, cycles) << 16; // _fvmInstance mVUinfo |= findFlagInst(xClip, cycles) << 20; // _fvcInstance - mVUinfo |= (xS & 3) << 12; // _fsInstance - mVUinfo |= (xM & 3) << 10; // _fmInstance - mVUinfo |= (xC & 3) << 14; // _fcInstance + mVUinfo |= xS << 12; // _fsInstance + mVUinfo |= xM << 10; // _fmInstance + mVUinfo |= xC << 14; // _fcInstance - if (doStatus || isFSSET || doDivFlag) - xStatus[xS++ & 3] = cycles + 4; - if (doMac) - xMac[xM++ & 3] = cycles + 4; - if (doClip) - xClip[xC++ & 3] = cycles + 4; + if (doStatus || isFSSET || doDivFlag) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } + if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } + if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } cycles++; incPC2(2); } - mVUregs.flags = ((__Clip) ? 0 : ((xC & 3) << 2)) | ((__Status) ? 0 : (xS & 3)); + mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) | ((__Status) ? 0 : xS); return cycles; } @@ -180,7 +177,6 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { if (__Clip) { int bClip[4]; sortFlag(xClip, bClip, cycles); - //SysPrintf("__Clip\n"); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->clipFlag); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleClip); SSE_MOVAPS_XMM_to_M128((uptr)mVU->clipFlag, xmmT1); diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 925001fb48..0e86c23953 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -41,7 +41,7 @@ microVUx(void) __mVULog(const char* fmt, ...) { microVUt(void) __mVUdumpProgram(int progIndex) { microVU* mVU = mVUx; - bool bitX[9]; + bool bitX[7]; char str[30]; int delay = 0; mVUbranch = 0; @@ -73,10 +73,8 @@ microVUt(void) __mVUdumpProgram(int progIndex) { bitX[4] = 0; bitX[5] = 0; bitX[6] = 0; - bitX[7] = 0; - bitX[8] = 0; - if (mVU->code & _Ibit_) { bitX[0] = 1; bitX[5] = 1; bitX[7] = 1; } + if (mVU->code & _Ibit_) { bitX[0] = 1; bitX[5] = 1; } if (mVU->code & _Ebit_) { bitX[1] = 1; bitX[5] = 1; delay = 2; } if (mVU->code & _Mbit_) { bitX[2] = 1; bitX[5] = 1; } if (mVU->code & _Dbit_) { bitX[3] = 1; bitX[5] = 1; } @@ -101,12 +99,18 @@ microVUt(void) __mVUdumpProgram(int progIndex) { } iPC = i; - if (bitX[7]) { mVUlog(""); } mVU->code = mVU->prog.prog[progIndex].data[i]; - mVUlog("
\n[%04x] (%08x) ", i*4, mVU->code); - mVUopL(); - mVUlog("\n\n

"); - if (bitX[7]) { mVUlog(""); } + + if(bitX[0]) { + mVUlog("
\n"); + mVUlog("[%04x] (%08x) %f", i*4, mVU->code, *(float*)&mVU->code); + mVUlog("\n\n

"); + } + else { + mVUlog("
\n[%04x] (%08x) ", i*4, mVU->code); + mVUopL(); + mVUlog("\n\n

"); + } } mVUlog("\n"); mVUlog("\n"); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 5214fc0f85..9919bb83bb 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -74,7 +74,7 @@ microVUf(void) mVU_DIV() { x86SetJ8(djmp); mVUunpack_xyzw(xmmFs, xmmFs, 0); - mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); + mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); } pass3 { mVUlog("DIV Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } @@ -92,7 +92,7 @@ microVUf(void) mVU_SQRT() { if (CHECK_VU_OVERFLOW) SSE_MINSS_XMM_to_XMM(xmmFt, xmmMax); // Clamp infinities (only need to do positive clamp since xmmFt is positive) SSE_SQRTSS_XMM_to_XMM(xmmFt, xmmFt); mVUunpack_xyzw(xmmFt, xmmFt, 0); - mVUmergeRegs(xmmPQ, xmmFt, writeQ ? 4 : 8); + mVUmergeRegs(xmmPQ, xmmFt, writeQ ? 4 : 8); } pass3 { mVUlog("SQRT Q, vf%02d%s", _Ft_, _Ftf_String); } } @@ -130,7 +130,7 @@ microVUf(void) mVU_RSQRT() { x86SetJ8(djmp); mVUunpack_xyzw(xmmFs, xmmFs, 0); - mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); + mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); } pass3 { mVUlog("RSQRT Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 42f39247ee..b995fc9c0e 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -107,6 +107,7 @@ declareAllVariables #define getVUmem(x) (((vuIndex == 1) ? (x & 0x3ff) : ((x >= 0x400) ? (x & 0x43f) : (x & 0xff))) * 16) #define offsetSS ((_X) ? (0) : ((_Y) ? (4) : ((_Z) ? 8: 12))) +#define offsetReg ((_X) ? (0) : ((_Y) ? (1) : ((_Z) ? 2: 3))) #define xmmT1 0 // Temp Reg #define xmmFs 1 // Holds the Value of Fs (writes back result Fd) @@ -174,6 +175,7 @@ declareAllVariables #define bSaveAddr (((xPC + (2 * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) +#define mVUflagHack (mVUcurProg.sFlagHack) // Pass 1 uses these to set mVUinfo #define _isNOP (1<<0) // Skip Lower Instruction @@ -287,9 +289,9 @@ declareAllVariables #define mVUdumpProg 0&& #endif -// Status Flag Speed Hack -#define CHECK_VU_FLAGHACK 0 // Set to 1 to turn hack on -#define mVUflagHack (mVUcurProg.sFlagHack) +// Speed Hacks (Set to 1 to turn On) +#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack +#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 0299609bd6..a88093e03b 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -205,7 +205,7 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { } // Modifies the Source Reg! -microVUx(void) mVUmergeRegs(int dest, int src, int xyzw) { +void mVUmergeRegs(int dest, int src, int xyzw) { xyzw &= 0xf; if ( (dest != src) && (xyzw != 0) ) { if ( cpucaps.hasStreamingSIMD4Extensions && (xyzw != 0x8) && (xyzw != 0xf) ) { @@ -316,4 +316,65 @@ microVUt(void) mVUcheckSflag(int progIndex) { } } +static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; +static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK2[4]) = {0x00000000, 0x40000000, 0x00000000, 0x40000000}; + +// Warning: Modifies xmmT1 and xmmT2 +void MIN_MAX_(x86SSERegType to, x86SSERegType from, bool min) { + + // XY + SSE2_PSHUFD_XMM_to_XMM(xmmT1, to, 0x50); + SSE2_PAND_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0x50); + SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); + if (min) SSE2_MINPD_XMM_to_XMM(xmmT1, xmmT2); + else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); + SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x88); + mVUmergeRegs(to, xmmT1, 0xc); + + // ZW + SSE2_PSHUFD_XMM_to_XMM(xmmT1, to, 0xfa); + SSE2_PAND_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0xfa); + SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); + if (min) SSE2_MINPD_XMM_to_XMM(xmmT1, xmmT2); + else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); + SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x88); + mVUmergeRegs(to, xmmT1, 0x3); +} + +// Warning: Modifies from and to's upper 3 vectors +void MIN_MAX_SS(x86SSERegType to, x86SSERegType from, bool min) { + SSE2_PSHUFD_XMM_to_XMM(to, to, 0x50); + SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(from, from, 0x50); + SSE2_PAND_M128_to_XMM (from, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (from, (uptr)MIN_MAX_MASK2); + if (min) SSE2_MINPD_XMM_to_XMM(to, from); + else SSE2_MAXPD_XMM_to_XMM(to, from); +} + +void SSE_MAX2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + if (CHECK_VU_MINMAXHACK) { SSE_MAXPS_XMM_to_XMM(to, from); } + else { MIN_MAX_(to, from, 0); } +} +void SSE_MIN2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + if (CHECK_VU_MINMAXHACK) { SSE_MINPS_XMM_to_XMM(to, from); } + else { MIN_MAX_(to, from, 1); } +} +void SSE_MAX2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + if (CHECK_VU_MINMAXHACK) { SSE_MAXSS_XMM_to_XMM(to, from); } + else { MIN_MAX_SS(to, from, 0); } +} +void SSE_MIN2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + if (CHECK_VU_MINMAXHACK) { SSE_MINSS_XMM_to_XMM(to, from); } + else { MIN_MAX_SS(to, from, 1); } +} + + #endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 3102b6c75e..eb51cc2b42 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -25,7 +25,7 @@ #define AND_XYZW ((_XYZW_SS && modXYZW) ? (1) : (doMac ? (_X_Y_Z_W) : (flipMask[_X_Y_Z_W]))) #define ADD_XYZW ((_XYZW_SS && modXYZW) ? (_X ? 3 : (_Y ? 2 : (_Z ? 1 : 0))) : 0) -#define SHIFT_XYZW(gprReg) { if (_XYZW_SS && modXYZW && !_W) { SHL16ItoR(gprReg, ADD_XYZW); } } +#define SHIFT_XYZW(gprReg) { if (_XYZW_SS && modXYZW && !_W) { SHL32ItoR(gprReg, ADD_XYZW); } } // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modXYZW) { @@ -57,8 +57,8 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation pjmp = JZ8(0); // Skip if none are - if (doMac) SHL16ItoR(mReg, 4 + ADD_XYZW); - if (doStatus) OR16ItoR(sReg, 0x82); // SS, S flags + if (doMac) SHL32ItoR(mReg, 4 + ADD_XYZW); + if (doStatus) OR32ItoR(sReg, 0x82); // SS, S flags if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking x86SetJ8(pjmp); @@ -67,7 +67,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation pjmp = JZ8(0); // Skip if none are if (doMac) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (doStatus) { OR16ItoR(sReg, 0x41); } // ZS, Z flags + if (doStatus) { OR32ItoR(sReg, 0x41); } // ZS, Z flags x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ @@ -476,90 +476,90 @@ microVUf(void) mVU_ABS() { } pass3 { mVUlog("ABS"); mVUlogFtFs(); } } -microVUf(void) mVU_ADD() { mVU_FMAC1 (ADD, "ADD"); } -microVUf(void) mVU_ADDi() { mVU_FMAC6 (ADD, "ADDi"); } -microVUf(void) mVU_ADDq() { mVU_FMAC22(ADD, "ADDq"); } -microVUf(void) mVU_ADDx() { mVU_FMAC3 (ADD, "ADDx"); } -microVUf(void) mVU_ADDy() { mVU_FMAC3 (ADD, "ADDy"); } -microVUf(void) mVU_ADDz() { mVU_FMAC3 (ADD, "ADDz"); } -microVUf(void) mVU_ADDw() { mVU_FMAC3 (ADD, "ADDw"); } -microVUf(void) mVU_ADDA() { mVU_FMAC4 (ADD, "ADDA"); } -microVUf(void) mVU_ADDAi() { mVU_FMAC7 (ADD, "ADDAi"); } -microVUf(void) mVU_ADDAq() { mVU_FMAC23(ADD, "ADDAq"); } -microVUf(void) mVU_ADDAx() { mVU_FMAC5 (ADD, "ADDAx"); } -microVUf(void) mVU_ADDAy() { mVU_FMAC5 (ADD, "ADDAy"); } -microVUf(void) mVU_ADDAz() { mVU_FMAC5 (ADD, "ADDAz"); } -microVUf(void) mVU_ADDAw() { mVU_FMAC5 (ADD, "ADDAw"); } -microVUf(void) mVU_SUB() { mVU_FMAC1 (SUB, "SUB"); } -microVUf(void) mVU_SUBi() { mVU_FMAC6 (SUB, "SUBi"); } -microVUf(void) mVU_SUBq() { mVU_FMAC22(SUB, "SUBq"); } -microVUf(void) mVU_SUBx() { mVU_FMAC3 (SUB, "SUBx"); } -microVUf(void) mVU_SUBy() { mVU_FMAC3 (SUB, "SUBy"); } -microVUf(void) mVU_SUBz() { mVU_FMAC3 (SUB, "SUBz"); } -microVUf(void) mVU_SUBw() { mVU_FMAC3 (SUB, "SUBw"); } -microVUf(void) mVU_SUBA() { mVU_FMAC4 (SUB, "SUBA"); } -microVUf(void) mVU_SUBAi() { mVU_FMAC7 (SUB, "SUBAi"); } -microVUf(void) mVU_SUBAq() { mVU_FMAC23(SUB, "SUBAq"); } -microVUf(void) mVU_SUBAx() { mVU_FMAC5 (SUB, "SUBAx"); } -microVUf(void) mVU_SUBAy() { mVU_FMAC5 (SUB, "SUBAy"); } -microVUf(void) mVU_SUBAz() { mVU_FMAC5 (SUB, "SUBAz"); } -microVUf(void) mVU_SUBAw() { mVU_FMAC5 (SUB, "SUBAw"); } -microVUf(void) mVU_MUL() { mVU_FMAC1 (MUL, "MUL"); } -microVUf(void) mVU_MULi() { mVU_FMAC6 (MUL, "MULi"); } -microVUf(void) mVU_MULq() { mVU_FMAC22(MUL, "MULq"); } -microVUf(void) mVU_MULx() { mVU_FMAC3 (MUL, "MULx"); } -microVUf(void) mVU_MULy() { mVU_FMAC3 (MUL, "MULy"); } -microVUf(void) mVU_MULz() { mVU_FMAC3 (MUL, "MULz"); } -microVUf(void) mVU_MULw() { mVU_FMAC3 (MUL, "MULw"); } -microVUf(void) mVU_MULA() { mVU_FMAC4 (MUL, "MULA"); } -microVUf(void) mVU_MULAi() { mVU_FMAC7 (MUL, "MULAi"); } -microVUf(void) mVU_MULAq() { mVU_FMAC23(MUL, "MULAq"); } -microVUf(void) mVU_MULAx() { mVU_FMAC5 (MUL, "MULAx"); } -microVUf(void) mVU_MULAy() { mVU_FMAC5 (MUL, "MULAy"); } -microVUf(void) mVU_MULAz() { mVU_FMAC5 (MUL, "MULAz"); } -microVUf(void) mVU_MULAw() { mVU_FMAC5 (MUL, "MULAw"); } -microVUf(void) mVU_MADD() { mVU_FMAC8 (ADD, "MADD"); } -microVUf(void) mVU_MADDi() { mVU_FMAC12(ADD, "MADDi"); } -microVUf(void) mVU_MADDq() { mVU_FMAC24(ADD, "MADDq"); } -microVUf(void) mVU_MADDx() { mVU_FMAC10(ADD, "MADDx"); } -microVUf(void) mVU_MADDy() { mVU_FMAC10(ADD, "MADDy"); } -microVUf(void) mVU_MADDz() { mVU_FMAC10(ADD, "MADDz"); } -microVUf(void) mVU_MADDw() { mVU_FMAC10(ADD, "MADDw"); } -microVUf(void) mVU_MADDA() { mVU_FMAC14(ADD, "MADDA"); } -microVUf(void) mVU_MADDAi() { mVU_FMAC16(ADD, "MADDAi"); } -microVUf(void) mVU_MADDAq() { mVU_FMAC26(ADD, "MADDAq"); } -microVUf(void) mVU_MADDAx() { mVU_FMAC15(ADD, "MADDAx"); } -microVUf(void) mVU_MADDAy() { mVU_FMAC15(ADD, "MADDAy"); } -microVUf(void) mVU_MADDAz() { mVU_FMAC15(ADD, "MADDAz"); } -microVUf(void) mVU_MADDAw() { mVU_FMAC15(ADD, "MADDAw"); } -microVUf(void) mVU_MSUB() { mVU_FMAC9 (SUB, "MSUB"); } -microVUf(void) mVU_MSUBi() { mVU_FMAC13(SUB, "MSUBi"); } -microVUf(void) mVU_MSUBq() { mVU_FMAC25(SUB, "MSUBq"); } -microVUf(void) mVU_MSUBx() { mVU_FMAC11(SUB, "MSUBx"); } -microVUf(void) mVU_MSUBy() { mVU_FMAC11(SUB, "MSUBy"); } -microVUf(void) mVU_MSUBz() { mVU_FMAC11(SUB, "MSUBz"); } -microVUf(void) mVU_MSUBw() { mVU_FMAC11(SUB, "MSUBw"); } -microVUf(void) mVU_MSUBA() { mVU_FMAC14(SUB, "MSUBA"); } -microVUf(void) mVU_MSUBAi() { mVU_FMAC16(SUB, "MSUBAi"); } -microVUf(void) mVU_MSUBAq() { mVU_FMAC26(SUB, "MSUBAq"); } -microVUf(void) mVU_MSUBAx() { mVU_FMAC15(SUB, "MSUBAx"); } -microVUf(void) mVU_MSUBAy() { mVU_FMAC15(SUB, "MSUBAy"); } -microVUf(void) mVU_MSUBAz() { mVU_FMAC15(SUB, "MSUBAz"); } -microVUf(void) mVU_MSUBAw() { mVU_FMAC15(SUB, "MSUBAw"); } -microVUf(void) mVU_MAX() { mVU_FMAC1 (MAX, "MAX"); } -microVUf(void) mVU_MAXi() { mVU_FMAC6 (MAX, "MAXi"); } -microVUf(void) mVU_MAXx() { mVU_FMAC3 (MAX, "MAXx"); } -microVUf(void) mVU_MAXy() { mVU_FMAC3 (MAX, "MAXy"); } -microVUf(void) mVU_MAXz() { mVU_FMAC3 (MAX, "MAXz"); } -microVUf(void) mVU_MAXw() { mVU_FMAC3 (MAX, "MAXw"); } -microVUf(void) mVU_MINI() { mVU_FMAC1 (MIN, "MINI"); } -microVUf(void) mVU_MINIi() { mVU_FMAC6 (MIN, "MINIi"); } -microVUf(void) mVU_MINIx() { mVU_FMAC3 (MIN, "MINIx"); } -microVUf(void) mVU_MINIy() { mVU_FMAC3 (MIN, "MINIy"); } -microVUf(void) mVU_MINIz() { mVU_FMAC3 (MIN, "MINIz"); } -microVUf(void) mVU_MINIw() { mVU_FMAC3 (MIN, "MINIw"); } -microVUf(void) mVU_OPMULA() { mVU_FMAC18(MUL, "OPMULA"); } -microVUf(void) mVU_OPMSUB() { mVU_FMAC19(SUB, "OPMSUB"); } +microVUf(void) mVU_ADD() { mVU_FMAC1 (ADD, "ADD"); } +microVUf(void) mVU_ADDi() { mVU_FMAC6 (ADD, "ADDi"); } +microVUf(void) mVU_ADDq() { mVU_FMAC22(ADD, "ADDq"); } +microVUf(void) mVU_ADDx() { mVU_FMAC3 (ADD, "ADDx"); } +microVUf(void) mVU_ADDy() { mVU_FMAC3 (ADD, "ADDy"); } +microVUf(void) mVU_ADDz() { mVU_FMAC3 (ADD, "ADDz"); } +microVUf(void) mVU_ADDw() { mVU_FMAC3 (ADD, "ADDw"); } +microVUf(void) mVU_ADDA() { mVU_FMAC4 (ADD, "ADDA"); } +microVUf(void) mVU_ADDAi() { mVU_FMAC7 (ADD, "ADDAi"); } +microVUf(void) mVU_ADDAq() { mVU_FMAC23(ADD, "ADDAq"); } +microVUf(void) mVU_ADDAx() { mVU_FMAC5 (ADD, "ADDAx"); } +microVUf(void) mVU_ADDAy() { mVU_FMAC5 (ADD, "ADDAy"); } +microVUf(void) mVU_ADDAz() { mVU_FMAC5 (ADD, "ADDAz"); } +microVUf(void) mVU_ADDAw() { mVU_FMAC5 (ADD, "ADDAw"); } +microVUf(void) mVU_SUB() { mVU_FMAC1 (SUB, "SUB"); } +microVUf(void) mVU_SUBi() { mVU_FMAC6 (SUB, "SUBi"); } +microVUf(void) mVU_SUBq() { mVU_FMAC22(SUB, "SUBq"); } +microVUf(void) mVU_SUBx() { mVU_FMAC3 (SUB, "SUBx"); } +microVUf(void) mVU_SUBy() { mVU_FMAC3 (SUB, "SUBy"); } +microVUf(void) mVU_SUBz() { mVU_FMAC3 (SUB, "SUBz"); } +microVUf(void) mVU_SUBw() { mVU_FMAC3 (SUB, "SUBw"); } +microVUf(void) mVU_SUBA() { mVU_FMAC4 (SUB, "SUBA"); } +microVUf(void) mVU_SUBAi() { mVU_FMAC7 (SUB, "SUBAi"); } +microVUf(void) mVU_SUBAq() { mVU_FMAC23(SUB, "SUBAq"); } +microVUf(void) mVU_SUBAx() { mVU_FMAC5 (SUB, "SUBAx"); } +microVUf(void) mVU_SUBAy() { mVU_FMAC5 (SUB, "SUBAy"); } +microVUf(void) mVU_SUBAz() { mVU_FMAC5 (SUB, "SUBAz"); } +microVUf(void) mVU_SUBAw() { mVU_FMAC5 (SUB, "SUBAw"); } +microVUf(void) mVU_MUL() { mVU_FMAC1 (MUL, "MUL"); } +microVUf(void) mVU_MULi() { mVU_FMAC6 (MUL, "MULi"); } +microVUf(void) mVU_MULq() { mVU_FMAC22(MUL, "MULq"); } +microVUf(void) mVU_MULx() { mVU_FMAC3 (MUL, "MULx"); } +microVUf(void) mVU_MULy() { mVU_FMAC3 (MUL, "MULy"); } +microVUf(void) mVU_MULz() { mVU_FMAC3 (MUL, "MULz"); } +microVUf(void) mVU_MULw() { mVU_FMAC3 (MUL, "MULw"); } +microVUf(void) mVU_MULA() { mVU_FMAC4 (MUL, "MULA"); } +microVUf(void) mVU_MULAi() { mVU_FMAC7 (MUL, "MULAi"); } +microVUf(void) mVU_MULAq() { mVU_FMAC23(MUL, "MULAq"); } +microVUf(void) mVU_MULAx() { mVU_FMAC5 (MUL, "MULAx"); } +microVUf(void) mVU_MULAy() { mVU_FMAC5 (MUL, "MULAy"); } +microVUf(void) mVU_MULAz() { mVU_FMAC5 (MUL, "MULAz"); } +microVUf(void) mVU_MULAw() { mVU_FMAC5 (MUL, "MULAw"); } +microVUf(void) mVU_MADD() { mVU_FMAC8 (ADD, "MADD"); } +microVUf(void) mVU_MADDi() { mVU_FMAC12(ADD, "MADDi"); } +microVUf(void) mVU_MADDq() { mVU_FMAC24(ADD, "MADDq"); } +microVUf(void) mVU_MADDx() { mVU_FMAC10(ADD, "MADDx"); } +microVUf(void) mVU_MADDy() { mVU_FMAC10(ADD, "MADDy"); } +microVUf(void) mVU_MADDz() { mVU_FMAC10(ADD, "MADDz"); } +microVUf(void) mVU_MADDw() { mVU_FMAC10(ADD, "MADDw"); } +microVUf(void) mVU_MADDA() { mVU_FMAC14(ADD, "MADDA"); } +microVUf(void) mVU_MADDAi() { mVU_FMAC16(ADD, "MADDAi"); } +microVUf(void) mVU_MADDAq() { mVU_FMAC26(ADD, "MADDAq"); } +microVUf(void) mVU_MADDAx() { mVU_FMAC15(ADD, "MADDAx"); } +microVUf(void) mVU_MADDAy() { mVU_FMAC15(ADD, "MADDAy"); } +microVUf(void) mVU_MADDAz() { mVU_FMAC15(ADD, "MADDAz"); } +microVUf(void) mVU_MADDAw() { mVU_FMAC15(ADD, "MADDAw"); } +microVUf(void) mVU_MSUB() { mVU_FMAC9 (SUB, "MSUB"); } +microVUf(void) mVU_MSUBi() { mVU_FMAC13(SUB, "MSUBi"); } +microVUf(void) mVU_MSUBq() { mVU_FMAC25(SUB, "MSUBq"); } +microVUf(void) mVU_MSUBx() { mVU_FMAC11(SUB, "MSUBx"); } +microVUf(void) mVU_MSUBy() { mVU_FMAC11(SUB, "MSUBy"); } +microVUf(void) mVU_MSUBz() { mVU_FMAC11(SUB, "MSUBz"); } +microVUf(void) mVU_MSUBw() { mVU_FMAC11(SUB, "MSUBw"); } +microVUf(void) mVU_MSUBA() { mVU_FMAC14(SUB, "MSUBA"); } +microVUf(void) mVU_MSUBAi() { mVU_FMAC16(SUB, "MSUBAi"); } +microVUf(void) mVU_MSUBAq() { mVU_FMAC26(SUB, "MSUBAq"); } +microVUf(void) mVU_MSUBAx() { mVU_FMAC15(SUB, "MSUBAx"); } +microVUf(void) mVU_MSUBAy() { mVU_FMAC15(SUB, "MSUBAy"); } +microVUf(void) mVU_MSUBAz() { mVU_FMAC15(SUB, "MSUBAz"); } +microVUf(void) mVU_MSUBAw() { mVU_FMAC15(SUB, "MSUBAw"); } +microVUf(void) mVU_MAX() { mVU_FMAC1 (MAX2, "MAX"); } +microVUf(void) mVU_MAXi() { mVU_FMAC6 (MAX2, "MAXi"); } +microVUf(void) mVU_MAXx() { mVU_FMAC3 (MAX2, "MAXx"); } +microVUf(void) mVU_MAXy() { mVU_FMAC3 (MAX2, "MAXy"); } +microVUf(void) mVU_MAXz() { mVU_FMAC3 (MAX2, "MAXz"); } +microVUf(void) mVU_MAXw() { mVU_FMAC3 (MAX2, "MAXw"); } +microVUf(void) mVU_MINI() { mVU_FMAC1 (MIN2, "MINI"); } +microVUf(void) mVU_MINIi() { mVU_FMAC6 (MIN2, "MINIi"); } +microVUf(void) mVU_MINIx() { mVU_FMAC3 (MIN2, "MINIx"); } +microVUf(void) mVU_MINIy() { mVU_FMAC3 (MIN2, "MINIy"); } +microVUf(void) mVU_MINIz() { mVU_FMAC3 (MIN2, "MINIz"); } +microVUf(void) mVU_MINIw() { mVU_FMAC3 (MIN2, "MINIw"); } +microVUf(void) mVU_OPMULA() { mVU_FMAC18(MUL, "OPMULA"); } +microVUf(void) mVU_OPMSUB() { mVU_FMAC19(SUB, "OPMSUB"); } microVUf(void) mVU_NOP() { pass3 { mVUlog("NOP"); } } microVUq(void) mVU_FTOIx(uptr addr) { microVU* mVU = mVUx; From 40447aabc14a61b11883b484f1c1f7fd3c56b302 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 14 May 2009 08:56:50 +0000 Subject: [PATCH 055/296] microVU: fixed a bug in the Jump opcodes. this probably fixes a lot of problems in games (grandia 3 intro works properly now, ffxii shows more stuff in-game, etc...) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1178 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 5 ++--- pcsx2/x86/microVU_Lower.inl | 10 ++++------ 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index f29f200b25..55c773c780 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -109,9 +109,8 @@ struct microVU { u32 branch; // Holds branch compare result (IBxx) OR Holds address to Jump to (JALR/JR) u32 p; // Holds current P instance index u32 q; // Holds current Q instance index - u32 espBackup; // Temp Backup for ESP - u32 totalCycles; - u32 cycles; + u32 totalCycles; // Total Cycles that mVU is expected to run for + u32 cycles; // Cycles Counter PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 9919bb83bb..c874cdebe5 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1225,10 +1225,9 @@ microVUf(void) mVU_IBNE() { microVUf(void) mVU_JR() { microVU* mVU = mVUx; mVUbranch = 9; - pass1 { mVUanalyzeBranch1(_Is_); } + pass1 { analyzeVIreg1(_Is_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + mVUallocVIa(gprT1, _Is_); SHL32ItoR(gprT1, 3); AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); @@ -1239,10 +1238,9 @@ microVUf(void) mVU_JR() { microVUf(void) mVU_JALR() { microVU* mVU = mVUx; mVUbranch = 10; - pass1 { mVUanalyzeBranch1(_Is_); analyzeVIreg2(_It_, 1); } + pass1 { analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 1); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + mVUallocVIa(gprT1, _Is_); SHL32ItoR(gprT1, 3); AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); From 5648d07b96f396882d380e38afefae9428cb9e95 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 14 May 2009 16:41:52 +0000 Subject: [PATCH 056/296] GSdx: more refactoring (aka search and replace) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1179 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 73 +- plugins/GSdx/GPU.h | 278 +- plugins/GSdx/GPUDrawScanline.cpp | 6 +- plugins/GSdx/GPUDrawScanline.h | 8 +- plugins/GSdx/GPUDrawScanlineCodeGenerator.cpp | 6 +- plugins/GSdx/GPUDrawScanlineCodeGenerator.h | 2 +- plugins/GSdx/GPUDrawingEnvironment.h | 8 +- plugins/GSdx/GPULocalMemory.cpp | 202 +- plugins/GSdx/GPULocalMemory.h | 49 +- plugins/GSdx/GPURenderer.h | 29 +- plugins/GSdx/GPURendererSW.h | 55 +- plugins/GSdx/GPUScanlineEnvironment.h | 43 +- plugins/GSdx/GPUSettingsDlg.cpp | 10 +- plugins/GSdx/GPUState.cpp | 96 +- plugins/GSdx/GPUState.h | 24 +- plugins/GSdx/GS.cpp | 56 +- plugins/GSdx/GS.h | 962 ++--- plugins/GSdx/GSBlock.h | 806 +--- plugins/GSdx/GSCapture.cpp | 74 +- plugins/GSdx/GSCapture.h | 5 +- plugins/GSdx/GSClut.cpp | 283 +- plugins/GSdx/GSClut.h | 40 +- plugins/GSdx/GSCodeBuffer.cpp | 4 +- plugins/GSdx/GSCodeBuffer.h | 2 +- plugins/GSdx/GSCrc.cpp | 6 +- plugins/GSdx/GSCrc.h | 6 +- plugins/GSdx/GSDevice.h | 22 +- plugins/GSdx/GSDevice10.cpp | 54 +- plugins/GSdx/GSDevice10.h | 30 +- plugins/GSdx/GSDevice7.cpp | 19 +- plugins/GSdx/GSDevice7.h | 6 +- plugins/GSdx/GSDevice9.cpp | 54 +- plugins/GSdx/GSDevice9.h | 32 +- plugins/GSdx/GSDeviceNull.h | 6 +- plugins/GSdx/GSDirtyRect.cpp | 65 +- plugins/GSdx/GSDirtyRect.h | 14 +- plugins/GSdx/GSDrawScanline.cpp | 40 +- plugins/GSdx/GSDrawScanline.h | 12 +- plugins/GSdx/GSDrawScanlineCodeGenerator.cpp | 10 +- plugins/GSdx/GSDump.cpp | 6 +- plugins/GSdx/GSDump.h | 6 +- plugins/GSdx/GSFunctionMap.h | 8 +- plugins/GSdx/GSLocalMemory.cpp | 591 ++- plugins/GSdx/GSLocalMemory.h | 486 +-- plugins/GSdx/GSRasterizer.cpp | 79 +- plugins/GSdx/GSRasterizer.h | 9 +- plugins/GSdx/GSRenderer.h | 67 +- plugins/GSdx/GSRendererHW.h | 71 +- plugins/GSdx/GSRendererHW10.cpp | 12 +- plugins/GSdx/GSRendererHW10.h | 6 +- plugins/GSdx/GSRendererHW9.cpp | 16 +- plugins/GSdx/GSRendererHW9.h | 4 +- plugins/GSdx/GSRendererNull.h | 4 +- plugins/GSdx/GSRendererSW.h | 65 +- plugins/GSdx/GSScanlineEnvironment.h | 100 +- plugins/GSdx/GSSettingsDlg.cpp | 18 +- plugins/GSdx/GSSetupPrimCodeGenerator.cpp | 2 +- plugins/GSdx/GSSetupPrimCodeGenerator.h | 4 +- plugins/GSdx/GSState.cpp | 271 +- plugins/GSdx/GSState.h | 47 +- plugins/GSdx/GSTables.cpp | 32 +- plugins/GSdx/GSTables.h | 32 +- plugins/GSdx/GSTexture.h | 6 +- plugins/GSdx/GSTexture10.cpp | 16 +- plugins/GSdx/GSTexture10.h | 4 +- plugins/GSdx/GSTexture7.cpp | 28 +- plugins/GSdx/GSTexture7.h | 5 +- plugins/GSdx/GSTexture9.cpp | 22 +- plugins/GSdx/GSTexture9.h | 4 +- plugins/GSdx/GSTextureCache.h | 135 +- plugins/GSdx/GSTextureCache10.cpp | 77 +- plugins/GSdx/GSTextureCache10.h | 3 +- plugins/GSdx/GSTextureCache9.cpp | 78 +- plugins/GSdx/GSTextureCache9.h | 3 +- plugins/GSdx/GSTextureCacheSW.cpp | 114 +- plugins/GSdx/GSTextureCacheSW.h | 12 +- plugins/GSdx/GSTextureFX10.cpp | 14 +- plugins/GSdx/GSTextureFX10.h | 120 +- plugins/GSdx/GSTextureFX9.cpp | 28 +- plugins/GSdx/GSTextureFX9.h | 118 +- plugins/GSdx/GSTextureNull.h | 4 +- plugins/GSdx/GSThread.cpp | 63 + plugins/GSdx/GSThread.h | 42 + plugins/GSdx/GSUtil.cpp | 119 +- plugins/GSdx/GSUtil.h | 18 +- plugins/GSdx/GSVector.cpp | 52 + plugins/GSdx/GSVector.h | 162 +- plugins/GSdx/GSVertex.h | 2 +- plugins/GSdx/GSVertexHW.h | 22 +- plugins/GSdx/GSVertexList.h | 4 +- plugins/GSdx/GSVertexSW.cpp | 10 +- plugins/GSdx/GSVertexSW.h | 16 +- plugins/GSdx/GSdx_vs2008.vcproj | 3620 +++++++++-------- plugins/GSdx/sse.h | 80 - plugins/GSdx/stdafx.h | 101 +- plugins/GSdx/xbyak/xbyak.h | 4 +- 96 files changed, 4863 insertions(+), 5676 deletions(-) create mode 100644 plugins/GSdx/GSThread.cpp create mode 100644 plugins/GSdx/GSThread.h delete mode 100644 plugins/GSdx/sse.h diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index 63ab5b5309..4583b48b90 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -32,7 +32,7 @@ static HRESULT s_hr = E_FAIL; static GPURendererBase* s_gpu = NULL; -EXPORT_C_(UINT32) PSEgetLibType() +EXPORT_C_(uint32) PSEgetLibType() { return PSE_LT_GPU; } @@ -42,15 +42,15 @@ EXPORT_C_(char*) PSEgetLibName() return GSUtil::GetLibName(); } -EXPORT_C_(UINT32) PSEgetLibVersion() +EXPORT_C_(uint32) PSEgetLibVersion() { - static const UINT32 version = 1; - static const UINT32 revision = 1; + static const uint32 version = 1; + static const uint32 revision = 1; return version << 16 | revision << 8 | PLUGIN_VERSION; } -EXPORT_C_(INT32) GPUinit() +EXPORT_C_(int32) GPUinit() { AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -59,7 +59,7 @@ EXPORT_C_(INT32) GPUinit() return 0; } -EXPORT_C_(INT32) GPUshutdown() +EXPORT_C_(int32) GPUshutdown() { AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -68,7 +68,7 @@ EXPORT_C_(INT32) GPUshutdown() return 0; } -EXPORT_C_(INT32) GPUclose() +EXPORT_C_(int32) GPUclose() { AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -86,7 +86,7 @@ EXPORT_C_(INT32) GPUclose() return 0; } -EXPORT_C_(INT32) GPUopen(HWND hWnd) +EXPORT_C_(int32) GPUopen(HWND hWnd) { AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -103,8 +103,8 @@ EXPORT_C_(INT32) GPUopen(HWND hWnd) rs.m_dither = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("dithering"), 1); rs.m_aspectratio = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("AspectRatio"), 1); rs.m_vsync = !!AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("vsync"), FALSE); - rs.m_scale.cx = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_x"), 0); - rs.m_scale.cy = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_y"), 0); + rs.m_scale.x = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_x"), 0); + rs.m_scale.y = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_y"), 0); int threads = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("swthreads"), 1); @@ -131,7 +131,7 @@ EXPORT_C_(INT32) GPUopen(HWND hWnd) return 0; } -EXPORT_C_(INT32) GPUconfigure() +EXPORT_C_(int32) GPUconfigure() { AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -146,7 +146,7 @@ EXPORT_C_(INT32) GPUconfigure() return 0; } -EXPORT_C_(INT32) GPUtest() +EXPORT_C_(int32) GPUtest() { // TODO @@ -158,76 +158,79 @@ EXPORT_C GPUabout() // TODO } -EXPORT_C GPUwriteDataMem(const BYTE* mem, UINT32 size) +EXPORT_C GPUwriteDataMem(const uint8* mem, uint32 size) { s_gpu->WriteData(mem, size); } -EXPORT_C GPUwriteData(UINT32 data) +EXPORT_C GPUwriteData(uint32 data) { - s_gpu->WriteData((BYTE*)&data, 1); + s_gpu->WriteData((uint8*)&data, 1); } -EXPORT_C GPUreadDataMem(BYTE* mem, UINT32 size) +EXPORT_C GPUreadDataMem(uint8* mem, uint32 size) { s_gpu->ReadData(mem, size); } -EXPORT_C_(UINT32) GPUreadData() +EXPORT_C_(uint32) GPUreadData() { - UINT32 data = 0; + uint32 data = 0; - s_gpu->ReadData((BYTE*)&data, 1); + s_gpu->ReadData((uint8*)&data, 1); return data; } -EXPORT_C GPUwriteStatus(UINT32 status) +EXPORT_C GPUwriteStatus(uint32 status) { s_gpu->WriteStatus(status); } -EXPORT_C_(UINT32) GPUreadStatus() +EXPORT_C_(uint32) GPUreadStatus() { return s_gpu->ReadStatus(); } -EXPORT_C_(UINT32) GPUdmaChain(const BYTE* mem, UINT32 addr) +EXPORT_C_(uint32) GPUdmaChain(const uint8* mem, uint32 addr) { - // TODO - - UINT32 last[3]; + uint32 last[3]; memset(last, 0xff, sizeof(last)); do { - if(addr == last[1] || addr == last[2]) break; + if(addr == last[1] || addr == last[2]) + { + break; + } + (addr < last[0] ? last[1] : last[2]) = addr; + last[0] = addr; - BYTE size = mem[addr + 3]; + uint8 size = mem[addr + 3]; if(size > 0) { s_gpu->WriteData(&mem[addr + 4], size); } - addr = *(UINT32*)&mem[addr] & 0xffffff; + addr = *(uint32*)&mem[addr] & 0xffffff; } while(addr != 0xffffff); return 0; } -EXPORT_C_(UINT32) GPUgetMode() +EXPORT_C_(uint32) GPUgetMode() { // TODO return 0; } -EXPORT_C GPUsetMode(UINT32) +EXPORT_C GPUsetMode(uint32 mode) { // TODO } @@ -239,7 +242,7 @@ EXPORT_C GPUupdateLace() EXPORT_C GPUmakeSnapshot() { - s_gpu->MakeSnapshot("c:\\"); // TODO + s_gpu->MakeSnapshot("c:/"); // TODO } EXPORT_C GPUdisplayText(char* text) @@ -247,12 +250,12 @@ EXPORT_C GPUdisplayText(char* text) // TODO } -EXPORT_C GPUdisplayFlags(UINT32 flags) +EXPORT_C GPUdisplayFlags(uint32 flags) { // TODO } -EXPORT_C_(INT32) GPUfreeze(UINT32 type, GPUFreezeData* data) +EXPORT_C_(int32) GPUfreeze(uint32 type, GPUFreezeData* data) { if(!data || data->version != 1) { @@ -288,12 +291,12 @@ EXPORT_C_(INT32) GPUfreeze(UINT32 type, GPUFreezeData* data) return 0; } -EXPORT_C GPUgetScreenPic(BYTE* mem) +EXPORT_C GPUgetScreenPic(uint8* mem) { // TODO } -EXPORT_C GPUshowScreenPic(BYTE* mem) +EXPORT_C GPUshowScreenPic(uint8* mem) { // TODO } diff --git a/plugins/GSdx/GPU.h b/plugins/GSdx/GPU.h index d341870565..d1ec2d6a3e 100644 --- a/plugins/GSdx/GPU.h +++ b/plugins/GSdx/GPU.h @@ -33,209 +33,209 @@ enum }; REG32_(GPUReg, STATUS) - UINT32 TX:4; - UINT32 TY:1; - UINT32 ABR:2; - UINT32 TP:2; - UINT32 DTD:1; - UINT32 DFE:1; - UINT32 MD:1; - UINT32 ME:1; - UINT32 _PAD0:3; - UINT32 WIDTH1:1; - UINT32 WIDTH0:2; - UINT32 HEIGHT:1; - UINT32 ISPAL:1; - UINT32 ISRGB24:1; - UINT32 ISINTER:1; - UINT32 DEN:1; - UINT32 _PAD1:2; - UINT32 IDLE:1; - UINT32 IMG:1; - UINT32 COM:1; - UINT32 DMA:2; - UINT32 LCF:1; + uint32 TX:4; + uint32 TY:1; + uint32 ABR:2; + uint32 TP:2; + uint32 DTD:1; + uint32 DFE:1; + uint32 MD:1; + uint32 ME:1; + uint32 _PAD0:3; + uint32 WIDTH1:1; + uint32 WIDTH0:2; + uint32 HEIGHT:1; + uint32 ISPAL:1; + uint32 ISRGB24:1; + uint32 ISINTER:1; + uint32 DEN:1; + uint32 _PAD1:2; + uint32 IDLE:1; + uint32 IMG:1; + uint32 COM:1; + uint32 DMA:2; + uint32 LCF:1; /* - UINT32 TX:4; - UINT32 TY:1; - UINT32 ABR:2; - UINT32 TP:2; - UINT32 DTD:1; - UINT32 DFE:1; - UINT32 PBW:1; - UINT32 PBC:1; - UINT32 _PAD0:3; - UINT32 HRES2:1; - UINT32 HRES1:2; - UINT32 VRES:1; - UINT32 ISPAL:1; - UINT32 ISRGB24:1; - UINT32 ISINTER:1; - UINT32 ISSTOP:1; - UINT32 _PAD1:1; - UINT32 DMARDY:1; - UINT32 IDIDLE:1; - UINT32 DATARDY:1; - UINT32 ISEMPTY:1; - UINT32 TMODE:2; - UINT32 ODE:1; + uint32 TX:4; + uint32 TY:1; + uint32 ABR:2; + uint32 TP:2; + uint32 DTD:1; + uint32 DFE:1; + uint32 PBW:1; + uint32 PBC:1; + uint32 _PAD0:3; + uint32 HRES2:1; + uint32 HRES1:2; + uint32 VRES:1; + uint32 ISPAL:1; + uint32 ISRGB24:1; + uint32 ISINTER:1; + uint32 ISSTOP:1; + uint32 _PAD1:1; + uint32 DMARDY:1; + uint32 IDIDLE:1; + uint32 DATARDY:1; + uint32 ISEMPTY:1; + uint32 TMODE:2; + uint32 ODE:1; */ REG_END REG32_(GPUReg, PACKET) - UINT32 _PAD:24; - UINT32 OPTION:5; - UINT32 TYPE:3; + uint32 _PAD:24; + uint32 OPTION:5; + uint32 TYPE:3; REG_END REG32_(GPUReg, PRIM) - UINT32 VTX:24; - UINT32 TGE:1; - UINT32 ABE:1; - UINT32 TME:1; - UINT32 _PAD2:1; - UINT32 IIP:1; - UINT32 TYPE:3; + uint32 VTX:24; + uint32 TGE:1; + uint32 ABE:1; + uint32 TME:1; + uint32 _PAD2:1; + uint32 IIP:1; + uint32 TYPE:3; REG_END REG32_(GPUReg, POLYGON) - UINT32 _PAD:24; - UINT32 TGE:1; - UINT32 ABE:1; - UINT32 TME:1; - UINT32 VTX:1; - UINT32 IIP:1; - UINT32 TYPE:3; + uint32 _PAD:24; + uint32 TGE:1; + uint32 ABE:1; + uint32 TME:1; + uint32 VTX:1; + uint32 IIP:1; + uint32 TYPE:3; REG_END REG32_(GPUReg, LINE) - UINT32 _PAD:24; - UINT32 ZERO1:1; - UINT32 ABE:1; - UINT32 ZERO2:1; - UINT32 PLL:1; - UINT32 IIP:1; - UINT32 TYPE:3; + uint32 _PAD:24; + uint32 ZERO1:1; + uint32 ABE:1; + uint32 ZERO2:1; + uint32 PLL:1; + uint32 IIP:1; + uint32 TYPE:3; REG_END REG32_(GPUReg, SPRITE) - UINT32 _PAD:24; - UINT32 ZERO:1; - UINT32 ABE:1; - UINT32 TME:1; - UINT32 SIZE:2; - UINT32 TYPE:3; + uint32 _PAD:24; + uint32 ZERO:1; + uint32 ABE:1; + uint32 TME:1; + uint32 SIZE:2; + uint32 TYPE:3; REG_END REG32_(GPUReg, RESET) - UINT32 _PAD:32; + uint32 _PAD:32; REG_END REG32_(GPUReg, DEN) - UINT32 DEN:1; - UINT32 _PAD:31; + uint32 DEN:1; + uint32 _PAD:31; REG_END REG32_(GPUReg, DMA) - UINT32 DMA:2; - UINT32 _PAD:30; + uint32 DMA:2; + uint32 _PAD:30; REG_END REG32_(GPUReg, DAREA) - UINT32 X:10; - UINT32 Y:9; - UINT32 _PAD:13; + uint32 X:10; + uint32 Y:9; + uint32 _PAD:13; REG_END REG32_(GPUReg, DHRANGE) - UINT32 X1:12; - UINT32 X2:12; - UINT32 _PAD:8; + uint32 X1:12; + uint32 X2:12; + uint32 _PAD:8; REG_END REG32_(GPUReg, DVRANGE) - UINT32 Y1:10; - UINT32 Y2:11; - UINT32 _PAD:11; + uint32 Y1:10; + uint32 Y2:11; + uint32 _PAD:11; REG_END REG32_(GPUReg, DMODE) - UINT32 WIDTH0:2; - UINT32 HEIGHT:1; - UINT32 ISPAL:1; - UINT32 ISRGB24:1; - UINT32 ISINTER:1; - UINT32 WIDTH1:1; - UINT32 REVERSE:1; - UINT32 _PAD:24; + uint32 WIDTH0:2; + uint32 HEIGHT:1; + uint32 ISPAL:1; + uint32 ISRGB24:1; + uint32 ISINTER:1; + uint32 WIDTH1:1; + uint32 REVERSE:1; + uint32 _PAD:24; REG_END REG32_(GPUReg, GPUINFO) - UINT32 PARAM:24; - UINT32 _PAD:8; + uint32 PARAM:24; + uint32 _PAD:8; REG_END REG32_(GPUReg, MODE) - UINT32 TX:4; - UINT32 TY:1; - UINT32 ABR:2; - UINT32 TP:2; - UINT32 DTD:1; - UINT32 DFE:1; - UINT32 _PAD:21; + uint32 TX:4; + uint32 TY:1; + uint32 ABR:2; + uint32 TP:2; + uint32 DTD:1; + uint32 DFE:1; + uint32 _PAD:21; REG_END REG32_(GPUReg, MASK) - UINT32 MD:1; - UINT32 ME:1; - UINT32 _PAD:30; + uint32 MD:1; + uint32 ME:1; + uint32 _PAD:30; REG_END REG32_(GPUReg, DRAREA) - UINT32 X:10; - UINT32 Y:10; - UINT32 _PAD:12; + uint32 X:10; + uint32 Y:10; + uint32 _PAD:12; REG_END REG32_(GPUReg, DROFF) - INT32 X:11; - INT32 Y:11; - INT32 _PAD:10; + int32 X:11; + int32 Y:11; + int32 _PAD:10; REG_END REG32_(GPUReg, RGB) - UINT32 R:8; - UINT32 G:8; - UINT32 B:8; - UINT32 _PAD:8; + uint32 R:8; + uint32 G:8; + uint32 B:8; + uint32 _PAD:8; REG_END REG32_(GPUReg, XY) - INT32 X:11; - INT32 _PAD1:5; - INT32 Y:11; - INT32 _PAD2:5; + int32 X:11; + int32 _PAD1:5; + int32 Y:11; + int32 _PAD2:5; REG_END REG32_(GPUReg, UV) - UINT32 U:8; - UINT32 V:8; - UINT32 _PAD:16; + uint32 U:8; + uint32 V:8; + uint32 _PAD:16; REG_END REG32_(GPUReg, TWIN) - UINT32 TWW:5; - UINT32 TWH:5; - UINT32 TWX:5; - UINT32 TWY:5; - UINT32 _PAD:12; + uint32 TWW:5; + uint32 TWH:5; + uint32 TWX:5; + uint32 TWY:5; + uint32 _PAD:12; REG_END REG32_(GPUReg, CLUT) - UINT32 _PAD1:16; - UINT32 X:6; - UINT32 Y:9; - UINT32 _PAD2:1; + uint32 _PAD1:16; + uint32 X:6; + uint32 Y:9; + uint32 _PAD2:1; REG_END REG32_SET(GPUReg) @@ -266,10 +266,10 @@ REG_SET_END struct GPUFreezeData { - UINT32 version; // == 1 - UINT32 status; - UINT32 control[256]; - UINT16 vram[1024 * 1024]; + uint32 version; // == 1 + uint32 status; + uint32 control[256]; + uint16 vram[1024 * 1024]; }; #pragma pack(pop) diff --git a/plugins/GSdx/GPUDrawScanline.cpp b/plugins/GSdx/GPUDrawScanline.cpp index 9fc3c64e01..e067278dd0 100644 --- a/plugins/GSdx/GPUDrawScanline.cpp +++ b/plugins/GSdx/GPUDrawScanline.cpp @@ -51,7 +51,7 @@ void GPUDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) if(m_env.sel.twin) { - DWORD u, v; + uint32 u, v; u = ~(env.TWIN.TWW << 3) & 0xff; v = ~(env.TWIN.TWH << 3) & 0xff; @@ -100,7 +100,7 @@ GPUDrawScanline::GPUSetupPrimMap::GPUSetupPrimMap(GPUScanlineEnvironment& env) { } -GPUSetupPrimCodeGenerator* GPUDrawScanline::GPUSetupPrimMap::Create(DWORD key, void* ptr, size_t maxsize) +GPUSetupPrimCodeGenerator* GPUDrawScanline::GPUSetupPrimMap::Create(uint32 key, void* ptr, size_t maxsize) { return new GPUSetupPrimCodeGenerator(m_env, ptr, maxsize); } @@ -113,7 +113,7 @@ GPUDrawScanline::GPUDrawScanlineMap::GPUDrawScanlineMap(GPUScanlineEnvironment& { } -GPUDrawScanlineCodeGenerator* GPUDrawScanline::GPUDrawScanlineMap::Create(DWORD key, void* ptr, size_t maxsize) +GPUDrawScanlineCodeGenerator* GPUDrawScanline::GPUDrawScanlineMap::Create(uint32 key, void* ptr, size_t maxsize) { return new GPUDrawScanlineCodeGenerator(m_env, ptr, maxsize); } diff --git a/plugins/GSdx/GPUDrawScanline.h b/plugins/GSdx/GPUDrawScanline.h index 9a39d87875..9d6f01fda3 100644 --- a/plugins/GSdx/GPUDrawScanline.h +++ b/plugins/GSdx/GPUDrawScanline.h @@ -34,24 +34,24 @@ class GPUDrawScanline : public GSAlignedClass<16>, public IDrawScanline // - class GPUSetupPrimMap : public GSCodeGeneratorFunctionMap + class GPUSetupPrimMap : public GSCodeGeneratorFunctionMap { GPUScanlineEnvironment& m_env; public: GPUSetupPrimMap(GPUScanlineEnvironment& env); - GPUSetupPrimCodeGenerator* Create(DWORD key, void* ptr, size_t maxsize); + GPUSetupPrimCodeGenerator* Create(uint32 key, void* ptr, size_t maxsize); } m_sp; // - class GPUDrawScanlineMap : public GSCodeGeneratorFunctionMap + class GPUDrawScanlineMap : public GSCodeGeneratorFunctionMap { GPUScanlineEnvironment& m_env; public: GPUDrawScanlineMap(GPUScanlineEnvironment& env); - GPUDrawScanlineCodeGenerator* Create(DWORD key, void* ptr, size_t maxsize); + GPUDrawScanlineCodeGenerator* Create(uint32 key, void* ptr, size_t maxsize); } m_ds; protected: diff --git a/plugins/GSdx/GPUDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GPUDrawScanlineCodeGenerator.cpp index 523b434f11..1a7699658d 100644 --- a/plugins/GSdx/GPUDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GPUDrawScanlineCodeGenerator.cpp @@ -119,7 +119,7 @@ void GPUDrawScanlineCodeGenerator::Init(int params) mov(eax, dword[esp + _top]); - // WORD* fb = &m_env.vm[(top << (10 + m_env.sel.scalex)) + left]; + // uint16* fb = &m_env.vm[(top << (10 + m_env.sel.scalex)) + left]; mov(edi, eax); shl(edi, 10 + m_env.sel.scalex); @@ -222,7 +222,7 @@ void GPUDrawScanlineCodeGenerator::Step() // fb += 8; - add(edi, 8 * sizeof(WORD)); + add(edi, 8 * sizeof(uint16)); if(m_env.sel.tme) { @@ -1020,7 +1020,7 @@ const GSVector4i GPUDrawScanlineCodeGenerator::m_test[8] = GSVector4i::zero(), }; -__declspec(align(16)) const WORD GPUDrawScanlineCodeGenerator::m_dither[4][16] = +__declspec(align(16)) const uint16 GPUDrawScanlineCodeGenerator::m_dither[4][16] = { {7, 0, 6, 1, 7, 0, 6, 1, 7, 0, 6, 1, 7, 0, 6, 1}, {2, 5, 3, 4, 2, 5, 3, 4, 2, 5, 3, 4, 2, 5, 3, 4}, diff --git a/plugins/GSdx/GPUDrawScanlineCodeGenerator.h b/plugins/GSdx/GPUDrawScanlineCodeGenerator.h index 6e1d7f4256..423074b8bf 100644 --- a/plugins/GSdx/GPUDrawScanlineCodeGenerator.h +++ b/plugins/GSdx/GPUDrawScanlineCodeGenerator.h @@ -32,7 +32,7 @@ class GPUDrawScanlineCodeGenerator : public CodeGenerator void operator = (const GPUDrawScanlineCodeGenerator&); static const GSVector4i m_test[8]; - static const WORD m_dither[4][16]; + static const uint16 m_dither[4][16]; util::Cpu m_cpu; diff --git a/plugins/GSdx/GPUDrawingEnvironment.h b/plugins/GSdx/GPUDrawingEnvironment.h index e42064a450..c89ea4920a 100644 --- a/plugins/GSdx/GPUDrawingEnvironment.h +++ b/plugins/GSdx/GPUDrawingEnvironment.h @@ -55,21 +55,19 @@ public: DVRANGE.Y2 = 256; } - CRect GetDisplayRect() + GSVector4i GetDisplayRect() { static int s_width[] = {256, 320, 512, 640, 368, 384, 512, 640}; static int s_height[] = {240, 480}; - CRect r; + GSVector4i r; r.left = DAREA.X & ~7; // FIXME r.top = DAREA.Y; r.right = r.left + s_width[(STATUS.WIDTH1 << 2) | STATUS.WIDTH0]; r.bottom = r.top + (DVRANGE.Y2 - DVRANGE.Y1) * s_height[STATUS.HEIGHT] / 240; - r &= CRect(0, 0, 1024, 512); - - return r; + return r.rintersect(GSVector4i(0, 0, 1024, 512)); } int GetFPS() diff --git a/plugins/GSdx/GPULocalMemory.cpp b/plugins/GSdx/GPULocalMemory.cpp index 4dcdc97975..c0bfefb076 100644 --- a/plugins/GSdx/GPULocalMemory.cpp +++ b/plugins/GSdx/GPULocalMemory.cpp @@ -27,16 +27,16 @@ const GSVector4i GPULocalMemory::m_xxbx(0x00007c00); const GSVector4i GPULocalMemory::m_xgxx(0x000003e0); const GSVector4i GPULocalMemory::m_rxxx(0x0000001f); -GPULocalMemory::GPULocalMemory(const CSize& scale) +GPULocalMemory::GPULocalMemory(const GSVector2i& scale) { - m_scale.cx = min(max(scale.cx, 0), 2); - m_scale.cy = min(max(scale.cy, 0), 2); + m_scale.x = min(max(scale.x, 0), 2); + m_scale.y = min(max(scale.y, 0), 2); // - int size = (1 << (12 + 11)) * sizeof(WORD); + int size = (1 << (12 + 11)) * sizeof(uint16); - m_vm = (WORD*)VirtualAlloc(NULL, size * 2, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); + m_vm = (uint16*)VirtualAlloc(NULL, size * 2, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); memset(m_vm, 0, size); @@ -49,7 +49,7 @@ GPULocalMemory::GPULocalMemory(const CSize& scale) size = 256 * 256 * (1 + 1 + 4) * 32; - m_texture.buff[0] = (BYTE*)VirtualAlloc(NULL, size, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); + m_texture.buff[0] = (uint8*)VirtualAlloc(NULL, size, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); m_texture.buff[1] = m_texture.buff[0] + 256 * 256 * 32; m_texture.buff[2] = m_texture.buff[1] + 256 * 256 * 32; @@ -61,8 +61,8 @@ GPULocalMemory::GPULocalMemory(const CSize& scale) { for(int x = 0; x < 16; x++, offset += 256 * 256) { - m_texture.page[0][y][x] = &((BYTE*)m_texture.buff[0])[offset]; - m_texture.page[1][y][x] = &((BYTE*)m_texture.buff[1])[offset]; + m_texture.page[0][y][x] = &((uint8*)m_texture.buff[0])[offset]; + m_texture.page[1][y][x] = &((uint8*)m_texture.buff[1])[offset]; } } @@ -70,7 +70,7 @@ GPULocalMemory::GPULocalMemory(const CSize& scale) { for(int x = 0; x < 16; x++, offset += 256 * 256) { - m_texture.page[2][y][x] = &((DWORD*)m_texture.buff[2])[offset]; + m_texture.page[2][y][x] = &((uint32*)m_texture.buff[2])[offset]; } } } @@ -82,18 +82,18 @@ GPULocalMemory::~GPULocalMemory() VirtualFree(m_texture.buff[0], 0, MEM_RELEASE); } -const WORD* GPULocalMemory::GetCLUT(int tp, int cx, int cy) +const uint16* GPULocalMemory::GetCLUT(int tp, int cx, int cy) { if(m_clut.dirty || m_clut.tp != tp || m_clut.cx != cx || m_clut.cy != cy) { - WORD* src = GetPixelAddressScaled(cx << 4, cy); - WORD* dst = m_clut.buff; + uint16* src = GetPixelAddressScaled(cx << 4, cy); + uint16* dst = m_clut.buff; - if(m_scale.cx == 0) + if(m_scale.x == 0) { memcpy(dst, src, (tp == 0 ? 16 : 256) * 2); } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { if(tp == 0) { @@ -110,7 +110,7 @@ const WORD* GPULocalMemory::GetCLUT(int tp, int cx, int cy) } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { if(tp == 0) { @@ -152,7 +152,7 @@ const void* GPULocalMemory::GetTexture(int tp, int tx, int ty) void* buff = m_texture.page[tp][ty][tx]; - UINT32 flag = 1 << tx; + uint32 flag = 1 << tx; if((m_texture.valid[tp][ty] & flag) == 0) { @@ -161,16 +161,16 @@ const void* GPULocalMemory::GetTexture(int tp, int tx, int ty) switch(tp) { case 0: - ReadPage4(tx, ty, (BYTE*)buff); + ReadPage4(tx, ty, (uint8*)buff); bpp = 4; break; case 1: - ReadPage8(tx, ty, (BYTE*)buff); + ReadPage8(tx, ty, (uint8*)buff); bpp = 8; break; case 2: case 3: - ReadPage16(tx, ty, (WORD*)buff); + ReadPage16(tx, ty, (uint16*)buff); bpp = 16; default: // FIXME: __assume(0); // vc9 generates bogus code in release mode @@ -185,7 +185,7 @@ const void* GPULocalMemory::GetTexture(int tp, int tx, int ty) return buff; } -void GPULocalMemory::Invalidate(const CRect& r) +void GPULocalMemory::Invalidate(const GSVector4i& r) { if(!m_clut.dirty) { @@ -207,7 +207,7 @@ void GPULocalMemory::Invalidate(const CRect& r) for(int x = 0, xe = min(r.right, 1024), i = 0; x < xe; x += 64, i++) { - DWORD flag = 1 << i; + uint32 flag = 1 << i; if(r.left >= x + 256) continue; @@ -224,14 +224,14 @@ void GPULocalMemory::Invalidate(const CRect& r) } } -void GPULocalMemory::FillRect(const CRect& r, WORD c) +void GPULocalMemory::FillRect(const GSVector4i& r, uint16 c) { Invalidate(r); - WORD* RESTRICT dst = GetPixelAddressScaled(r.left, r.top); + uint16* RESTRICT dst = GetPixelAddressScaled(r.left, r.top); - int w = r.Width() << m_scale.cx; - int h = r.Height() << m_scale.cy; + int w = r.width() << m_scale.x; + int h = r.height() << m_scale.y; int pitch = GetWidth(); @@ -244,32 +244,32 @@ void GPULocalMemory::FillRect(const CRect& r, WORD c) } } -void GPULocalMemory::WriteRect(const CRect& r, const WORD* RESTRICT src) +void GPULocalMemory::WriteRect(const GSVector4i& r, const uint16* RESTRICT src) { Invalidate(r); - WORD* RESTRICT dst = GetPixelAddressScaled(r.left, r.top); + uint16* RESTRICT dst = GetPixelAddressScaled(r.left, r.top); - int w = r.Width(); - int h = r.Height(); + int w = r.width(); + int h = r.height(); int pitch = GetWidth(); - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int j = 0; j < h; j++, src += w) { - for(int k = 1 << m_scale.cy; k >= 1; k--, dst += pitch) + for(int k = 1 << m_scale.y; k >= 1; k--, dst += pitch) { memcpy(dst, src, w * 2); } } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int j = 0; j < h; j++, src += w) { - for(int k = 1 << m_scale.cy; k >= 1; k--, dst += pitch) + for(int k = 1 << m_scale.y; k >= 1; k--, dst += pitch) { for(int i = 0; i < w; i++) { @@ -279,11 +279,11 @@ void GPULocalMemory::WriteRect(const CRect& r, const WORD* RESTRICT src) } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int j = 0; j < h; j++, src += w) { - for(int k = 1 << m_scale.cy; k >= 1; k--, dst += pitch) + for(int k = 1 << m_scale.y; k >= 1; k--, dst += pitch) { for(int i = 0; i < w; i++) { @@ -301,23 +301,23 @@ void GPULocalMemory::WriteRect(const CRect& r, const WORD* RESTRICT src) } } -void GPULocalMemory::ReadRect(const CRect& r, WORD* RESTRICT dst) +void GPULocalMemory::ReadRect(const GSVector4i& r, uint16* RESTRICT dst) { - WORD* RESTRICT src = GetPixelAddressScaled(r.left, r.top); + uint16* RESTRICT src = GetPixelAddressScaled(r.left, r.top); - int w = r.Width(); - int h = r.Height(); + int w = r.width(); + int h = r.height(); - int pitch = GetWidth() << m_scale.cy; + int pitch = GetWidth() << m_scale.y; - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int j = 0; j < h; j++, src += pitch, dst += w) { memcpy(dst, src, w * 2); } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int j = 0; j < h; j++, src += pitch, dst += w) { @@ -327,7 +327,7 @@ void GPULocalMemory::ReadRect(const CRect& r, WORD* RESTRICT dst) } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int j = 0; j < h; j++, src += pitch, dst += w) { @@ -343,31 +343,31 @@ void GPULocalMemory::ReadRect(const CRect& r, WORD* RESTRICT dst) } } -void GPULocalMemory::MoveRect(const CPoint& src, const CPoint& dst, int w, int h) +void GPULocalMemory::MoveRect(int sx, int sy, int dx, int dy, int w, int h) { - Invalidate(CRect(dst, CSize(w, h))); + Invalidate(GSVector4i(dx, dy, dx + w, dy + h)); - WORD* s = GetPixelAddressScaled(src.x, src.y); - WORD* d = GetPixelAddressScaled(dst.x, dst.y); + uint16* s = GetPixelAddressScaled(sx, sy); + uint16* d = GetPixelAddressScaled(dx, dy); - w <<= m_scale.cx; - h <<= m_scale.cy; + w <<= m_scale.x; + h <<= m_scale.y; int pitch = GetWidth(); for(int i = 0; i < h; i++, s += pitch, d += pitch) { - memcpy(d, s, w * sizeof(WORD)); + memcpy(d, s, w * sizeof(uint16)); } } -void GPULocalMemory::ReadPage4(int tx, int ty, BYTE* RESTRICT dst) +void GPULocalMemory::ReadPage4(int tx, int ty, uint8* RESTRICT dst) { - WORD* src = GetPixelAddressScaled(tx << 6, ty << 8); + uint16* src = GetPixelAddressScaled(tx << 6, ty << 8); - int pitch = GetWidth() << m_scale.cy; + int pitch = GetWidth() << m_scale.y; - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { @@ -380,7 +380,7 @@ void GPULocalMemory::ReadPage4(int tx, int ty, BYTE* RESTRICT dst) } } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { @@ -393,7 +393,7 @@ void GPULocalMemory::ReadPage4(int tx, int ty, BYTE* RESTRICT dst) } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { @@ -412,36 +412,36 @@ void GPULocalMemory::ReadPage4(int tx, int ty, BYTE* RESTRICT dst) } } -void GPULocalMemory::ReadPage8(int tx, int ty, BYTE* RESTRICT dst) +void GPULocalMemory::ReadPage8(int tx, int ty, uint8* RESTRICT dst) { - WORD* src = GetPixelAddressScaled(tx << 6, ty << 8); + uint16* src = GetPixelAddressScaled(tx << 6, ty << 8); - int pitch = GetWidth() << m_scale.cy; + int pitch = GetWidth() << m_scale.y; - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { memcpy(dst, src, 256); } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { for(int i = 0; i < 128; i++) { - ((WORD*)dst)[i] = src[i * 2]; + ((uint16*)dst)[i] = src[i * 2]; } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { for(int i = 0; i < 128; i++) { - ((WORD*)dst)[i] = src[i * 4]; + ((uint16*)dst)[i] = src[i * 4]; } } } @@ -451,20 +451,20 @@ void GPULocalMemory::ReadPage8(int tx, int ty, BYTE* RESTRICT dst) } } -void GPULocalMemory::ReadPage16(int tx, int ty, WORD* RESTRICT dst) +void GPULocalMemory::ReadPage16(int tx, int ty, uint16* RESTRICT dst) { - WORD* src = GetPixelAddressScaled(tx << 6, ty << 8); + uint16* src = GetPixelAddressScaled(tx << 6, ty << 8); - int pitch = GetWidth() << m_scale.cy; + int pitch = GetWidth() << m_scale.y; - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { memcpy(dst, src, 512); } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { @@ -474,7 +474,7 @@ void GPULocalMemory::ReadPage16(int tx, int ty, WORD* RESTRICT dst) } } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int j = 0; j < 256; j++, src += pitch, dst += 256) { @@ -490,9 +490,9 @@ void GPULocalMemory::ReadPage16(int tx, int ty, WORD* RESTRICT dst) } } -void GPULocalMemory::ReadFrame32(const CRect& r, DWORD* RESTRICT dst, bool rgb24) +void GPULocalMemory::ReadFrame32(const GSVector4i& r, uint32* RESTRICT dst, bool rgb24) { - WORD* src = GetPixelAddress(r.left, r.top); + uint16* src = GetPixelAddress(r.left, r.top); int pitch = GetWidth(); @@ -500,19 +500,19 @@ void GPULocalMemory::ReadFrame32(const CRect& r, DWORD* RESTRICT dst, bool rgb24 { for(int i = r.top; i < r.bottom; i++, src += pitch, dst += pitch) { - Expand24(src, dst, r.Width()); + Expand24(src, dst, r.width()); } } else { for(int i = r.top; i < r.bottom; i++, src += pitch, dst += pitch) { - Expand16(src, dst, r.Width()); + Expand16(src, dst, r.width()); } } } -void GPULocalMemory::Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels) +void GPULocalMemory::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int pixels) { GSVector4i rm = m_rxxx; GSVector4i gm = m_xgxx; @@ -534,11 +534,11 @@ void GPULocalMemory::Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int } } -void GPULocalMemory::Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels) +void GPULocalMemory::Expand24(const uint16* RESTRICT src, uint32* RESTRICT dst, int pixels) { - BYTE* s = (BYTE*)src; + uint8* s = (uint8*)src; - if(m_scale.cx == 0) + if(m_scale.x == 0) { for(int i = 0; i < pixels; i += 2, s += 6) { @@ -546,7 +546,7 @@ void GPULocalMemory::Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int dst[i + 1] = (s[5] << 16) | (s[4] << 8) | s[3]; } } - else if(m_scale.cx == 1) + else if(m_scale.x == 1) { for(int i = 0; i < pixels; i += 4, s += 12) { @@ -554,7 +554,7 @@ void GPULocalMemory::Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int dst[i + 2] = dst[i + 3] = (s[9] << 16) | (s[8] << 8) | s[5]; } } - else if(m_scale.cx == 2) + else if(m_scale.x == 2) { for(int i = 0; i < pixels; i += 8, s += 24) { @@ -568,12 +568,14 @@ void GPULocalMemory::Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int } } -void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy) +void GPULocalMemory::SaveBMP(const string& path, const GSVector4i& r2, int tp, int cx, int cy) { - r.left <<= m_scale.cx; - r.top <<= m_scale.cy; - r.right <<= m_scale.cx; - r.bottom <<= m_scale.cy; + GSVector4i r; + + r.left = r2.left << m_scale.x; + r.top = r2.top << m_scale.y; + r.right = r2.right << m_scale.x; + r.bottom = r2.bottom << m_scale.y; r.left &= ~1; r.right &= ~1; @@ -583,8 +585,8 @@ void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy BITMAPINFOHEADER bih; memset(&bih, 0, sizeof(bih)); bih.biSize = sizeof(bih); - bih.biWidth = r.Width(); - bih.biHeight = r.Height(); + bih.biWidth = r.width(); + bih.biHeight = r.height(); bih.biPlanes = 1; bih.biBitCount = 32; bih.biCompression = BI_RGB; @@ -602,10 +604,10 @@ void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy int pitch = GetWidth(); - WORD* buff = (WORD*)_aligned_malloc(pitch * sizeof(WORD), 16); - DWORD* buff32 = (DWORD*)_aligned_malloc(pitch * sizeof(DWORD), 16); - WORD* src = GetPixelAddress(r.left, r.bottom - 1); - const WORD* clut = GetCLUT(tp, cx, cy); + uint16* buff = (uint16*)_aligned_malloc(pitch * sizeof(WORD), 16); + uint32* buff32 = (uint32*)_aligned_malloc(pitch * sizeof(uint32), 16); + uint16* src = GetPixelAddress(r.left, r.bottom - 1); + const uint16* clut = GetCLUT(tp, cx, cy); for(int j = r.bottom - 1; j >= r.top; j--, src -= pitch) { @@ -613,26 +615,26 @@ void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy { case 0: // 4 bpp - for(int i = 0, k = r.Width() / 2; i < k; i++) + for(int i = 0, k = r.width() / 2; i < k; i++) { - buff[i * 2 + 0] = clut[((BYTE*)src)[i] & 0xf]; - buff[i * 2 + 1] = clut[((BYTE*)src)[i] >> 4]; + buff[i * 2 + 0] = clut[((uint8*)src)[i] & 0xf]; + buff[i * 2 + 1] = clut[((uint8*)src)[i] >> 4]; } break; case 1: // 8 bpp - for(int i = 0, k = r.Width(); i < k; i++) + for(int i = 0, k = r.width(); i < k; i++) { - buff[i] = clut[((BYTE*)src)[i]]; + buff[i] = clut[((uint8*)src)[i]]; } break; case 2: // 16 bpp; - for(int i = 0, k = r.Width(); i < k; i++) + for(int i = 0, k = r.width(); i < k; i++) { buff[i] = src[i]; } @@ -646,14 +648,14 @@ void GPULocalMemory::SaveBMP(const string& path, CRect r, int tp, int cx, int cy break; } - Expand16(buff, buff32, r.Width()); + Expand16(buff, buff32, r.width()); - for(int i = 0, k = r.Width(); i < k; i++) + for(int i = 0, k = r.width(); i < k; i++) { buff32[i] = (buff32[i] & 0xff00ff00) | ((buff32[i] & 0x00ff0000) >> 16) | ((buff32[i] & 0x000000ff) << 16); } - fwrite(buff32, 1, r.Width() * 4, fp); + fwrite(buff32, 1, r.width() * 4, fp); } _aligned_free(buff); diff --git a/plugins/GSdx/GPULocalMemory.h b/plugins/GSdx/GPULocalMemory.h index a4adfb4855..5759779a7e 100644 --- a/plugins/GSdx/GPULocalMemory.h +++ b/plugins/GSdx/GPULocalMemory.h @@ -31,55 +31,56 @@ class GPULocalMemory static const GSVector4i m_xgxx; static const GSVector4i m_rxxx; - WORD* m_vm; + uint16* m_vm; struct { - WORD* buff; + uint16* buff; int tp, cx, cy; bool dirty; } m_clut; struct { - BYTE* buff[3]; + uint8* buff[3]; void* page[3][2][16]; - WORD valid[3][2]; + uint16 valid[3][2]; } m_texture; - CSize m_scale; + GSVector2i m_scale; public: - GPULocalMemory(const CSize& scale); + GPULocalMemory(const GSVector2i& scale); virtual ~GPULocalMemory(); - CSize GetScale() {return m_scale;} - int GetWidth() {return 1 << (10 + m_scale.cx);} - int GetHeight() {return 1 << (9 + m_scale.cy);} + GSVector2i GetScale() {return m_scale;} - WORD* GetPixelAddress(int x, int y) const {return &m_vm[(y << (10 + m_scale.cx)) + x];} - WORD* GetPixelAddressScaled(int x, int y) const {return &m_vm[((y << m_scale.cy) << (10 + m_scale.cx)) + (x << m_scale.cx)];} + int GetWidth() {return 1 << (10 + m_scale.x);} + int GetHeight() {return 1 << (9 + m_scale.y);} - const WORD* GetCLUT(int tp, int cx, int cy); + uint16* GetPixelAddress(int x, int y) const {return &m_vm[(y << (10 + m_scale.x)) + x];} + uint16* GetPixelAddressScaled(int x, int y) const {return &m_vm[((y << m_scale.y) << (10 + m_scale.x)) + (x << m_scale.x)];} + + const uint16* GetCLUT(int tp, int cx, int cy); const void* GetTexture(int tp, int tx, int ty); - void Invalidate(const CRect& r); + void Invalidate(const GSVector4i& r); - void FillRect(const CRect& r, WORD c); - void WriteRect(const CRect& r, const WORD* RESTRICT src); - void ReadRect(const CRect& r, WORD* RESTRICT dst); - void MoveRect(const CPoint& src, const CPoint& dst, int w, int h); + void FillRect(const GSVector4i& r, uint16 c); + void WriteRect(const GSVector4i& r, const uint16* RESTRICT src); + void ReadRect(const GSVector4i& r, uint16* RESTRICT dst); + void MoveRect(int sx, int sy, int dx, int dy, int w, int h); - void ReadPage4(int tx, int ty, BYTE* RESTRICT dst); - void ReadPage8(int tx, int ty, BYTE* RESTRICT dst); - void ReadPage16(int tx, int ty, WORD* RESTRICT dst); + void ReadPage4(int tx, int ty, uint8* RESTRICT dst); + void ReadPage8(int tx, int ty, uint8* RESTRICT dst); + void ReadPage16(int tx, int ty, uint16* RESTRICT dst); - void ReadFrame32(const CRect& r, DWORD* RESTRICT dst, bool rgb24); + void ReadFrame32(const GSVector4i& r, uint32* RESTRICT dst, bool rgb24); - void Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels); - void Expand24(const WORD* RESTRICT src, DWORD* RESTRICT dst, int pixels); + void Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int pixels); + void Expand24(const uint16* RESTRICT src, uint32* RESTRICT dst, int pixels); - void SaveBMP(const string& path, CRect r, int tp, int cx, int cy); + void SaveBMP(const string& path, const GSVector4i& r, int tp, int cx, int cy); }; #pragma warning(default: 4244) \ No newline at end of file diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index 9adaa4968a..bd683a67c5 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -30,7 +30,7 @@ struct GPURendererSettings int m_dither; int m_aspectratio; bool m_vsync; - CSize m_scale; + GSVector2i m_scale; }; class GPURendererBase : public GPUState, protected GPURendererSettings @@ -142,7 +142,7 @@ protected: void VertexKick() { - if(m_vl.GetCount() < m_env.PRIM.VTX) + if(m_vl.GetCount() < (int)m_env.PRIM.VTX) { return; } @@ -214,7 +214,7 @@ protected: if(m_env.PRIM.TME) { - CRect r; + GSVector4i r; r.left = m_env.STATUS.TX << 6; r.top = m_env.STATUS.TY << 8; @@ -246,11 +246,8 @@ protected: return false; } - CSize s; + GSVector2i s = st[0].GetSize(); - s.cx = st[0].GetWidth(); - s.cy = st[0].GetHeight(); - GSVector4 sr[2]; sr[0].x = 0; @@ -262,8 +259,8 @@ protected: dr[0].x = 0; dr[0].y = 0; - dr[0].z = (float)s.cx; - dr[0].w = (float)s.cy; + dr[0].z = (float)s.x; + dr[0].w = (float)s.y; GSVector4 c(0, 0, 0, 1); @@ -345,10 +342,10 @@ public: double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); - CRect r = m_env.GetDisplayRect(); + GSVector4i r = m_env.GetDisplayRect(); - int w = r.Width() << m_scale.cx; - int h = r.Height() << m_scale.cy; + int w = r.width() << m_scale.x; + int h = r.height() << m_scale.y; s_stats = format( "%I64d | %d x %d | %.2f fps (%d%%) | %d/%d | %d%% CPU | %.2f | %.2f", @@ -375,13 +372,11 @@ public: ResetDevice(); } - CRect r; + GSVector4i r; - GetClientRect(m_hWnd, &r); + GetClientRect(m_hWnd, r); - GSUtil::FitRect(r, m_aspectratio); - - m_dev.Present(r); + m_dev.Present(r.fit(m_aspectratio)); } virtual bool MakeSnapshot(const string& path) diff --git a/plugins/GSdx/GPURendererSW.h b/plugins/GSdx/GPURendererSW.h index 2ee2599b29..4ea9d29082 100644 --- a/plugins/GSdx/GPURendererSW.h +++ b/plugins/GSdx/GPURendererSW.h @@ -38,31 +38,32 @@ protected: bool GetOutput(Texture& t) { - CRect r = m_env.GetDisplayRect(); + GSVector4i r = m_env.GetDisplayRect(); - r.left <<= m_scale.cx; - r.top <<= m_scale.cy; - r.right <<= m_scale.cx; - r.bottom <<= m_scale.cy; + r.left <<= m_scale.x; + r.top <<= m_scale.y; + r.right <<= m_scale.x; + r.bottom <<= m_scale.y; // TODO - static DWORD* buff = (DWORD*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(DWORD), 16); + static uint32* buff = (uint32*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(uint32), 16); m_mem.ReadFrame32(r, buff, !!m_env.STATUS.ISRGB24); - r.OffsetRect(-r.TopLeft()); + int w = r.width(); + int h = r.height(); - if(m_texture.GetWidth() != r.Width() || m_texture.GetHeight() != r.Height()) + if(m_texture.GetWidth() != w || m_texture.GetHeight() != h) { m_texture = Texture(); } - if(!m_texture && !m_dev.CreateTexture(m_texture, r.Width(), r.Height())) + if(!m_texture && !m_dev.CreateTexture(m_texture, w, h)) { return false; } - m_texture.Update(r, buff, m_mem.GetWidth() * sizeof(DWORD)); + m_texture.Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); t = m_texture; @@ -75,8 +76,8 @@ protected: // TODO: x/y + off.x/y should wrap around at +/-1024 - int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.cx; - int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.cy; + int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.x; + int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.y; int s = m_v.UV.X; int t = m_v.UV.Y; @@ -85,7 +86,7 @@ protected: v.p = pt.xyxy(GSVector4::zero()); v.t = (pt.zwzw(GSVector4::zero()) + GSVector4(0.125f)) * 256.0f; - v.c = GSVector4((DWORD)m_v.RGB.ai32) * 128.0f; + v.c = GSVector4(m_v.RGB.u32) * 128.0f; __super::VertexKick(); } @@ -107,14 +108,14 @@ protected: GSVector4i GetScissor() { - GSVector4i v; + GSVector4i r; - v.x = (int)m_env.DRAREATL.X << m_scale.cx; - v.y = (int)m_env.DRAREATL.Y << m_scale.cy; - v.z = min((int)(m_env.DRAREABR.X + 1) << m_scale.cx, m_mem.GetWidth()); - v.w = min((int)(m_env.DRAREABR.Y + 1) << m_scale.cy, m_mem.GetHeight()); + r.left = (int)m_env.DRAREATL.X << m_scale.x; + r.top = (int)m_env.DRAREATL.Y << m_scale.y; + r.right = min((int)(m_env.DRAREABR.X + 1) << m_scale.x, m_mem.GetWidth()); + r.bottom = min((int)(m_env.DRAREABR.Y + 1) << m_scale.y, m_mem.GetHeight()); - return v; + return r; } void Draw() @@ -141,7 +142,7 @@ protected: { p.sel.tme = env.PRIM.TME; p.sel.tlu = env.STATUS.TP < 2; - p.sel.twin = (env.TWIN.ai32 & 0xfffff) != 0; + p.sel.twin = (env.TWIN.u32 & 0xfffff) != 0; p.sel.ltf = m_filter == 1 && env.PRIM.TYPE == GPU_POLYGON || m_filter == 2 ? 1 : 0; const void* t = m_mem.GetTexture(env.STATUS.TP, env.STATUS.TX, env.STATUS.TY); @@ -155,7 +156,7 @@ protected: p.sel.dtd = m_dither ? env.STATUS.DTD : 0; p.sel.md = env.STATUS.MD; p.sel.sprite = env.PRIM.TYPE == GPU_SPRITE; - p.sel.scalex = m_mem.GetScale().cx; + p.sel.scalex = m_mem.GetScale().x; // @@ -198,14 +199,12 @@ protected: br = br.maxv(p); } - GSVector4i scissor = data.scissor; + GSVector4i r = GSVector4i(tl.xyxy(br)).rintersect(data.scissor); - CRect r; - - r.left = max(scissor.x, min(scissor.z, (int)tl.x)) >> m_scale.cx; - r.top = max(scissor.y, min(scissor.w, (int)tl.y)) >> m_scale.cy; - r.right = max(scissor.x, min(scissor.z, (int)br.x)) >> m_scale.cx; - r.bottom = max(scissor.y, min(scissor.w, (int)br.y)) >> m_scale.cy; + r.left >>= m_scale.x; + r.top >>= m_scale.y; + r.right >>= m_scale.x; + r.bottom >>= m_scale.y; Invalidate(r); } diff --git a/plugins/GSdx/GPUScanlineEnvironment.h b/plugins/GSdx/GPUScanlineEnvironment.h index 4c7d9e66da..189341e7f8 100644 --- a/plugins/GSdx/GPUScanlineEnvironment.h +++ b/plugins/GSdx/GPUScanlineEnvironment.h @@ -28,32 +28,32 @@ union GPUScanlineSelector { struct { - DWORD iip:1; // 0 - DWORD me:1; // 1 - DWORD abe:1; // 2 - DWORD abr:2; // 3 - DWORD tge:1; // 5 - DWORD tme:1; // 6 - DWORD twin:1; // 7 - DWORD tlu:1; // 8 - DWORD dtd:1; // 9 - DWORD ltf:1; // 10 - DWORD md:1; // 11 - DWORD sprite:1; // 12 - DWORD scalex:2; // 13 + uint32 iip:1; // 0 + uint32 me:1; // 1 + uint32 abe:1; // 2 + uint32 abr:2; // 3 + uint32 tge:1; // 5 + uint32 tme:1; // 6 + uint32 twin:1; // 7 + uint32 tlu:1; // 8 + uint32 dtd:1; // 9 + uint32 ltf:1; // 10 + uint32 md:1; // 11 + uint32 sprite:1; // 12 + uint32 scalex:2; // 13 }; struct { - DWORD _pad1:1; // 0 - DWORD rfb:2; // 1 - DWORD _pad2:2; // 3 - DWORD tfx:2; // 5 + uint32 _pad1:1; // 0 + uint32 rfb:2; // 1 + uint32 _pad2:2; // 3 + uint32 tfx:2; // 5 }; - DWORD key; + uint32 key; - operator DWORD() {return key;} + operator uint32() {return key;} }; __declspec(align(16)) struct GPUScanlineParam @@ -61,17 +61,16 @@ __declspec(align(16)) struct GPUScanlineParam GPUScanlineSelector sel; const void* tex; - const WORD* clut; + const uint16* clut; }; __declspec(align(16)) struct GPUScanlineEnvironment { GPUScanlineSelector sel; - // GPULocalMemory* mem; // TODO: obsolite void* vm; const void* tex; - const WORD* clut; + const uint16* clut; // GSVector4i md; // similar to gs fba diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 0bccfa1d53..b1bbc9aa17 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -154,13 +154,13 @@ BOOL GPUSettingsDlg::OnInitDialog() if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - UINT ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); - UINT ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); - UINT ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); + uint32 ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); + uint32 ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); + uint32 ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); - UINT nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); + uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); - for(UINT i = 0; i < nModes; i++) + for(uint32 i = 0; i < nModes; i++) { D3DDISPLAYMODE mode; diff --git a/plugins/GSdx/GPUState.cpp b/plugins/GSdx/GPUState.cpp index a8e7630f22..dae0e4cc1a 100644 --- a/plugins/GSdx/GPUState.cpp +++ b/plugins/GSdx/GPUState.cpp @@ -22,7 +22,7 @@ #include "stdafx.h" #include "GPUState.h" -GPUState::GPUState(const CSize& scale) +GPUState::GPUState(const GSVector2i& scale) : m_mem(scale) , s_n(0) { @@ -64,7 +64,7 @@ void GPUState::Reset() { m_env.Reset(); - m_mem.Invalidate(CRect(0, 0, 1024, 512)); + m_mem.Invalidate(GSVector4i(0, 0, 1024, 512)); memset(&m_v, 0, sizeof(m_v)); } @@ -88,18 +88,18 @@ void GPUState::SetPrim(GPUReg* r) switch(r->PRIM.TYPE) { case GPU_POLYGON: - PRIM.ai32 = (r->PRIM.ai32 & 0xF7000000) | 3; // TYPE IIP TME ABE TGE + PRIM.u32 = (r->PRIM.u32 & 0xF7000000) | 3; // TYPE IIP TME ABE TGE break; case GPU_LINE: - PRIM.ai32 = (r->PRIM.ai32 & 0xF2000000) | 2; // TYPE IIP ABE + PRIM.u32 = (r->PRIM.u32 & 0xF2000000) | 2; // TYPE IIP ABE PRIM.TGE = 1; // ? break; case GPU_SPRITE: - PRIM.ai32 = (r->PRIM.ai32 & 0xE7000000) | 2; // TYPE TME ABE TGE + PRIM.u32 = (r->PRIM.u32 & 0xE7000000) | 2; // TYPE TME ABE TGE break; } - if(m_env.PRIM.ai32 != PRIM.ai32) + if(m_env.PRIM.u32 != PRIM.u32) { Flush(); @@ -109,38 +109,38 @@ void GPUState::SetPrim(GPUReg* r) void GPUState::SetCLUT(GPUReg* r) { - UINT32 mask = 0xFFFF0000; // X Y + uint32 mask = 0xFFFF0000; // X Y - UINT32 value = (m_env.CLUT.ai32 & ~mask) | (r->ai32 & mask); + uint32 value = (m_env.CLUT.u32 & ~mask) | (r->u32 & mask); - if(m_env.CLUT.ai32 != value) + if(m_env.CLUT.u32 != value) { Flush(); - m_env.CLUT.ai32 = value; + m_env.CLUT.u32 = value; } } void GPUState::SetTPAGE(GPUReg* r) { - UINT32 mask = 0x000001FF; // TP ABR TY TX + uint32 mask = 0x000001FF; // TP ABR TY TX - UINT32 value = (m_env.STATUS.ai32 & ~mask) | ((r->ai32 >> 16) & mask); + uint32 value = (m_env.STATUS.u32 & ~mask) | ((r->u32 >> 16) & mask); - if(m_env.STATUS.ai32 != value) + if(m_env.STATUS.u32 != value) { Flush(); - m_env.STATUS.ai32 = value; + m_env.STATUS.u32 = value; } } -void GPUState::Invalidate(const CRect& r) +void GPUState::Invalidate(const GSVector4i& r) { m_mem.Invalidate(r); } -void GPUState::WriteData(const BYTE* mem, UINT32 size) +void GPUState::WriteData(const uint8* mem, uint32 size) { GSPerfMonAutoTimer pmat(m_perfmon); @@ -164,7 +164,7 @@ void GPUState::WriteData(const BYTE* mem, UINT32 size) m_write.Remove(i); } -void GPUState::ReadData(BYTE* mem, UINT32 size) +void GPUState::ReadData(uint8* mem, uint32 size) { GSPerfMonAutoTimer pmat(m_perfmon); @@ -193,38 +193,38 @@ void GPUState::ReadData(BYTE* mem, UINT32 size) } } -void GPUState::WriteStatus(UINT32 status) +void GPUState::WriteStatus(uint32 status) { GSPerfMonAutoTimer pmat(m_perfmon); - UINT32 b = status >> 24; + uint32 b = status >> 24; m_status[b] = status; (this->*m_fpGPUStatusCommandHandlers[b])((GPUReg*)&status); } -UINT32 GPUState::ReadStatus() +uint32 GPUState::ReadStatus() { GSPerfMonAutoTimer pmat(m_perfmon); m_env.STATUS.LCF = ~m_env.STATUS.LCF; // ? - return m_env.STATUS.ai32; + return m_env.STATUS.u32; } void GPUState::Freeze(GPUFreezeData* data) { - data->status = m_env.STATUS.ai32; + data->status = m_env.STATUS.u32; memcpy(data->control, m_status, 256 * 4); - m_mem.ReadRect(CRect(0, 0, 1024, 512), data->vram); + m_mem.ReadRect(GSVector4i(0, 0, 1024, 512), data->vram); } void GPUState::Defrost(const GPUFreezeData* data) { - m_env.STATUS.ai32 = data->status; + m_env.STATUS.u32 = data->status; memcpy(m_status, data->control, 256 * 4); - m_mem.WriteRect(CRect(0, 0, 1024, 512), data->vram); + m_mem.WriteRect(GSVector4i(0, 0, 1024, 512), data->vram); for(int i = 0; i <= 8; i++) { @@ -289,24 +289,24 @@ void GPUState::SCH_DisplayMode(GPUReg* r) void GPUState::SCH_GPUInfo(GPUReg* r) { - UINT32 value = 0; + uint32 value = 0; switch(r->GPUINFO.PARAM) { case 0x2: - value = m_env.TWIN.ai32; + value = m_env.TWIN.u32; break; case 0x0: case 0x1: case 0x3: - value = m_env.DRAREATL.ai32; + value = m_env.DRAREATL.u32; break; case 0x4: - value = m_env.DRAREABR.ai32; + value = m_env.DRAREABR.u32; break; case 0x5: case 0x6: - value = m_env.DROFF.ai32; + value = m_env.DROFF.u32; break; case 0x7: value = 2; @@ -321,7 +321,7 @@ void GPUState::SCH_GPUInfo(GPUReg* r) } m_read.RemoveAll(); - m_read.Append((BYTE*)&value, 4); + m_read.Append((uint8*)&value, 4); m_read.cur = 0; } @@ -343,14 +343,14 @@ int GPUState::PH_Command(GPUReg* r, int size) Flush(); - CRect r2; + GSVector4i r2; r2.left = r[1].XY.X; r2.top = r[1].XY.Y; r2.right = r2.left + r[2].XY.X; r2.bottom = r2.top + r[2].XY.Y; - WORD c = (WORD)(((r[0].RGB.R >> 3) << 10) | ((r[0].RGB.R >> 3) << 5) | (r[0].RGB.R >> 3)); + uint16 c = (uint16)(((r[0].RGB.R >> 3) << 10) | ((r[0].RGB.R >> 3) << 5) | (r[0].RGB.R >> 3)); m_mem.FillRect(r2, c); @@ -439,7 +439,7 @@ int GPUState::PH_Line(GPUReg* r, int size) for(int i = 1; i < size; i++) { - if(r[i].ai32 == 0x55555555) + if(r[i].u32 == 0x55555555) { vertices = i - 1; } @@ -553,20 +553,18 @@ int GPUState::PH_Move(GPUReg* r, int size) Flush(); - CPoint src, dst; + int sx = r[1].XY.X; + int sy = r[1].XY.Y; - src.x = r[1].XY.X; - src.y = r[1].XY.Y; - - dst.x = r[2].XY.X; - dst.y = r[2].XY.Y; + int dx = r[2].XY.X; + int dy = r[2].XY.Y; int w = r[3].XY.X; int h = r[3].XY.Y; - m_mem.MoveRect(src, dst, w, h); + m_mem.MoveRect(sx, sy, dx, dy, w, h); - Invalidate(CRect(dst, CSize(w, h))); + Invalidate(GSVector4i(dx, dy, dx + w, dy + h)); // Dump("m"); @@ -586,14 +584,14 @@ int GPUState::PH_Write(GPUReg* r, int size) Flush(); - CRect r2; + GSVector4i r2; r2.left = r[1].XY.X; r2.top = r[1].XY.Y; r2.right = r2.left + w; r2.bottom = r2.top + h; - m_mem.WriteRect(r2, (const WORD*)&r[3]); + m_mem.WriteRect(r2, (const uint16*)&r[3]); Invalidate(r2); @@ -613,7 +611,7 @@ int GPUState::PH_Read(GPUReg* r, int size) int w = r[2].XY.X; int h = r[2].XY.Y; - CRect r2; + GSVector4i r2; r2.left = r[1].XY.X; r2.top = r[1].XY.Y; @@ -624,7 +622,7 @@ int GPUState::PH_Read(GPUReg* r, int size) m_read.cur = 0; m_read.Reserve(m_read.bytes); - m_mem.ReadRect(r2, (WORD*)m_read.buff); + m_mem.ReadRect(r2, (uint16*)m_read.buff); Dump("r"); @@ -693,7 +691,7 @@ GPUState::Buffer::Buffer() { bytes = 0; maxbytes = 4096; - buff = (BYTE*)_aligned_malloc(maxbytes, 16); + buff = (uint8*)_aligned_malloc(maxbytes, 16); cur = 0; } @@ -708,11 +706,11 @@ void GPUState::Buffer::Reserve(int size) { maxbytes = (maxbytes + size + 1023) & ~1023; - buff = (BYTE*)_aligned_realloc(buff, maxbytes, 16); + buff = (uint8*)_aligned_realloc(buff, maxbytes, 16); } } -void GPUState::Buffer::Append(const BYTE* src, int size) +void GPUState::Buffer::Append(const uint8* src, int size) { Reserve(bytes + (int)size); diff --git a/plugins/GSdx/GPUState.h b/plugins/GSdx/GPUState.h index 82d93692fd..db55189d3f 100644 --- a/plugins/GSdx/GPUState.h +++ b/plugins/GSdx/GPUState.h @@ -65,14 +65,14 @@ class GPUState : public GSAlignedClass<16> public: int bytes; int maxbytes; - BYTE* buff; + uint8* buff; int cur; public: Buffer(); ~Buffer(); void Reserve(int size); - void Append(const BYTE* src, int size); + void Append(const uint8* src, int size); void Remove(int size); void RemoveAll(); }; @@ -88,10 +88,10 @@ protected: int s_n; - void Dump(const string& s, UINT32 TP, const CRect& r, int inc = true) + void Dump(const string& s, uint32 TP, const GSVector4i& r, int inc = true) { //if(m_perfmon.GetFrame() < 1000) - //if((m_env.TWIN.ai32 & 0xfffff) == 0) + //if((m_env.TWIN.u32 & 0xfffff) == 0) //if(!m_env.STATUS.ME && !m_env.STATUS.MD) return; @@ -108,7 +108,7 @@ protected: void Dump(const string& s, int inc = true) { - Dump(s, 2, CRect(0, 0, 1024, 512), inc); + Dump(s, 2, GSVector4i(0, 0, 1024, 512), inc); } public: @@ -116,10 +116,10 @@ public: GPULocalMemory m_mem; GPUVertex m_v; GSPerfMon m_perfmon; - UINT32 m_status[256]; + uint32 m_status[256]; public: - GPUState(const CSize& scale); + GPUState(const GSVector2i& scale); virtual ~GPUState(); virtual void Reset(); @@ -127,13 +127,13 @@ public: virtual void FlushPrim() = 0; virtual void ResetPrim() = 0; virtual void VertexKick() = 0; - virtual void Invalidate(const CRect& r); + virtual void Invalidate(const GSVector4i& r); - void WriteData(const BYTE* mem, UINT32 size); - void ReadData(BYTE* mem, UINT32 size); + void WriteData(const uint8* mem, uint32 size); + void ReadData(uint8* mem, uint32 size); - void WriteStatus(UINT32 status); - UINT32 ReadStatus(); + void WriteStatus(uint32 status); + uint32 ReadStatus(); void Freeze(GPUFreezeData* data); void Defrost(const GPUFreezeData* data); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index c4fce6e68a..a06decfe00 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -35,9 +35,9 @@ static HRESULT s_hr = E_FAIL; static GSRendererBase* s_gs = NULL; static void (*s_irq)() = NULL; -static BYTE* s_basemem = NULL; +static uint8* s_basemem = NULL; -EXPORT_C_(UINT32) PS2EgetLibType() +EXPORT_C_(uint32) PS2EgetLibType() { return PS2E_LT_GS; } @@ -47,15 +47,15 @@ EXPORT_C_(char*) PS2EgetLibName() return GSUtil::GetLibName(); } -EXPORT_C_(UINT32) PS2EgetLibVersion2(UINT32 type) +EXPORT_C_(uint32) PS2EgetLibVersion2(uint32 type) { - const UINT32 revision = 0; - const UINT32 build = 1; + const uint32 revision = 0; + const uint32 build = 1; return (build << 0) | (revision << 8) | (PS2E_GS_VERSION << 16) | (PLUGIN_VERSION << 24); } -EXPORT_C_(UINT32) PS2EgetCpuPlatform() +EXPORT_C_(uint32) PS2EgetCpuPlatform() { #if _M_AMD64 return PS2E_X86_64; @@ -64,7 +64,7 @@ EXPORT_C_(UINT32) PS2EgetCpuPlatform() #endif } -EXPORT_C GSsetBaseMem(BYTE* mem) +EXPORT_C GSsetBaseMem(uint8* mem) { s_basemem = mem - 0x12000000; } @@ -175,37 +175,37 @@ EXPORT_C GSreset() s_gs->Reset(); } -EXPORT_C GSgifSoftReset(int mask) +EXPORT_C GSgifSoftReset(uint32 mask) { - s_gs->SoftReset((BYTE)mask); + s_gs->SoftReset(mask); } -EXPORT_C GSwriteCSR(UINT32 csr) +EXPORT_C GSwriteCSR(uint32 csr) { s_gs->WriteCSR(csr); } -EXPORT_C GSreadFIFO(BYTE* mem) +EXPORT_C GSreadFIFO(uint8* mem) { s_gs->ReadFIFO(mem, 1); } -EXPORT_C GSreadFIFO2(BYTE* mem, UINT32 size) +EXPORT_C GSreadFIFO2(uint8* mem, uint32 size) { s_gs->ReadFIFO(mem, size); } -EXPORT_C GSgifTransfer1(BYTE* mem, UINT32 addr) +EXPORT_C GSgifTransfer1(uint8* mem, uint32 addr) { s_gs->Transfer<0>(mem + addr, (0x4000 - addr) / 16); } -EXPORT_C GSgifTransfer2(BYTE* mem, UINT32 size) +EXPORT_C GSgifTransfer2(uint8* mem, uint32 size) { s_gs->Transfer<1>(mem, size); } -EXPORT_C GSgifTransfer3(BYTE* mem, UINT32 size) +EXPORT_C GSgifTransfer3(uint8* mem, uint32 size) { s_gs->Transfer<2>(mem, size); } @@ -215,7 +215,7 @@ EXPORT_C GSvsync(int field) s_gs->VSync(field); } -EXPORT_C_(UINT32) GSmakeSnapshot(char* path) +EXPORT_C_(uint32) GSmakeSnapshot(char* path) { return s_gs->MakeSnapshot(string(path) + "gsdx"); } @@ -224,7 +224,7 @@ EXPORT_C GSkeyEvent(keyEvent* ev) { } -EXPORT_C_(INT32) GSfreeze(int mode, GSFreezeData* data) +EXPORT_C_(int) GSfreeze(int mode, GSFreezeData* data) { if(mode == FREEZE_SAVE) { @@ -279,12 +279,12 @@ EXPORT_C GSirqCallback(void (*irq)()) s_irq = irq; } -EXPORT_C GSsetGameCRC(DWORD crc, int options) +EXPORT_C GSsetGameCRC(uint32 crc, int options) { s_gs->SetGameCRC(crc, options); } -EXPORT_C GSgetLastTag(UINT32* tag) +EXPORT_C GSgetLastTag(uint32* tag) { s_gs->GetLastTag(tag); } @@ -309,25 +309,25 @@ EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) ::SetPriorityClass(::GetCurrentProcess(), HIGH_PRIORITY_CLASS); - vector buff; + vector buff; if(FILE* fp = fopen(lpszCmdLine, "rb")) { GSinit(); - BYTE regs[0x2000]; + uint8 regs[0x2000]; GSsetBaseMem(regs); HWND hWnd = NULL; GSopen(&hWnd, "", true, renderer); - DWORD crc; + uint32 crc; fread(&crc, 4, 1, fp); GSsetGameCRC(crc, 0); GSFreezeData fd; fread(&fd.size, 4, 1, fp); - fd.data = new BYTE[fd.size]; + fd.data = new uint8[fd.size]; fread(fd.data, fd.size, 1, fp); GSfreeze(FREEZE_LOAD, &fd); delete [] fd.data; @@ -425,9 +425,9 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow {PSM_PSMZ16S, "16ZS"}, }; - BYTE* ptr = (BYTE*)_aligned_malloc(1024 * 1024 * 4, 16); + uint8* ptr = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); - for(int i = 0; i < 1024 * 1024 * 4; i++) ptr[i] = (BYTE)i; + for(int i = 0; i < 1024 * 1024 * 4; i++) ptr[i] = (uint8)i; // @@ -470,7 +470,7 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow TRXREG.RRW = w; TRXREG.RRH = h; - CRect r(0, 0, w, h); + GSVector4i r(0, 0, w, h); GIFRegTEX0 TEX0; @@ -560,9 +560,9 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow { GSLocalMemory mem; - BYTE* ptr = (BYTE*)_aligned_malloc(1024 * 1024 * 4, 16); + uint8* ptr = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); - for(int i = 0; i < 1024 * 1024 * 4; i++) ptr[i] = (BYTE)i; + for(int i = 0; i < 1024 * 1024 * 4; i++) ptr[i] = (uint8)i; const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[PSM_PSMCT32]; diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index 43585a0aab..90ddc3f126 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -212,14 +212,14 @@ enum GS_AFAIL #define REG32(name) \ union name \ { \ - UINT32 ai32; \ + uint32 u32; \ struct { \ #define REG64(name) \ union name \ { \ - UINT64 i64; \ - UINT32 ai32[2]; \ + uint64 u64; \ + uint32 u32[2]; \ void operator = (const GSVector4i& v) {GSVector4i::storel(this, v);} \ operator GSVector4i() const {return GSVector4i::loadl(this);} \ struct { \ @@ -227,8 +227,8 @@ union name \ #define REG128(name)\ union name \ { \ - UINT64 ai64[2]; \ - UINT32 ai32[4]; \ + uint64 u64[2]; \ + uint32 u32[4]; \ struct { \ #define REG32_(prefix, name) REG32(prefix##name) @@ -241,176 +241,176 @@ union name \ #define REG32_SET(name) \ union name \ { \ - UINT32 ai32; \ + uint32 u32; \ #define REG64_SET(name) \ union name \ { \ - UINT64 i64; \ - UINT32 ai32[2]; \ + uint64 u64; \ + uint32 u32[2]; \ #define REG128_SET(name)\ union name \ { \ - __m128i ai128; \ - UINT64 ai64[2]; \ - UINT32 ai32[4]; \ + __m128i m128; \ + uint64 u64[2]; \ + uint32 u32[4]; \ #define REG_SET_END }; REG64_(GSReg, BGCOLOR) - UINT32 R:8; - UINT32 G:8; - UINT32 B:8; - UINT32 _PAD1:8; - UINT32 _PAD2:32; + uint32 R:8; + uint32 G:8; + uint32 B:8; + uint32 _PAD1:8; + uint32 _PAD2:32; REG_END REG64_(GSReg, BUSDIR) - UINT32 DIR:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 DIR:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GSReg, CSR) - UINT32 rSIGNAL:1; - UINT32 rFINISH:1; - UINT32 rHSINT:1; - UINT32 rVSINT:1; - UINT32 rEDWINT:1; - UINT32 rZERO1:1; - UINT32 rZERO2:1; - UINT32 r_PAD1:1; - UINT32 rFLUSH:1; - UINT32 rRESET:1; - UINT32 r_PAD2:2; - UINT32 rNFIELD:1; - UINT32 rFIELD:1; - UINT32 rFIFO:2; - UINT32 rREV:8; - UINT32 rID:8; - UINT32 wSIGNAL:1; - UINT32 wFINISH:1; - UINT32 wHSINT:1; - UINT32 wVSINT:1; - UINT32 wEDWINT:1; - UINT32 wZERO1:1; - UINT32 wZERO2:1; - UINT32 w_PAD1:1; - UINT32 wFLUSH:1; - UINT32 wRESET:1; - UINT32 w_PAD2:2; - UINT32 wNFIELD:1; - UINT32 wFIELD:1; - UINT32 wFIFO:2; - UINT32 wREV:8; - UINT32 wID:8; + uint32 rSIGNAL:1; + uint32 rFINISH:1; + uint32 rHSINT:1; + uint32 rVSINT:1; + uint32 rEDWINT:1; + uint32 rZERO1:1; + uint32 rZERO2:1; + uint32 r_PAD1:1; + uint32 rFLUSH:1; + uint32 rRESET:1; + uint32 r_PAD2:2; + uint32 rNFIELD:1; + uint32 rFIELD:1; + uint32 rFIFO:2; + uint32 rREV:8; + uint32 rID:8; + uint32 wSIGNAL:1; + uint32 wFINISH:1; + uint32 wHSINT:1; + uint32 wVSINT:1; + uint32 wEDWINT:1; + uint32 wZERO1:1; + uint32 wZERO2:1; + uint32 w_PAD1:1; + uint32 wFLUSH:1; + uint32 wRESET:1; + uint32 w_PAD2:2; + uint32 wNFIELD:1; + uint32 wFIELD:1; + uint32 wFIFO:2; + uint32 wREV:8; + uint32 wID:8; REG_END REG64_(GSReg, DISPFB) // (-1/2) - UINT32 FBP:9; - UINT32 FBW:6; - UINT32 PSM:5; - UINT32 _PAD:12; - UINT32 DBX:11; - UINT32 DBY:11; - UINT32 _PAD2:10; + uint32 FBP:9; + uint32 FBW:6; + uint32 PSM:5; + uint32 _PAD:12; + uint32 DBX:11; + uint32 DBY:11; + uint32 _PAD2:10; REG_END2 - UINT32 Block() const {return FBP << 5;} + uint32 Block() const {return FBP << 5;} REG_END2 REG64_(GSReg, DISPLAY) // (-1/2) - UINT32 DX:12; - UINT32 DY:11; - UINT32 MAGH:4; - UINT32 MAGV:2; - UINT32 _PAD:3; - UINT32 DW:12; - UINT32 DH:11; - UINT32 _PAD2:9; + uint32 DX:12; + uint32 DY:11; + uint32 MAGH:4; + uint32 MAGV:2; + uint32 _PAD:3; + uint32 DW:12; + uint32 DH:11; + uint32 _PAD2:9; REG_END REG64_(GSReg, EXTBUF) - UINT32 EXBP:14; - UINT32 EXBW:6; - UINT32 FBIN:2; - UINT32 WFFMD:1; - UINT32 EMODA:2; - UINT32 EMODC:2; - UINT32 _PAD1:5; - UINT32 WDX:11; - UINT32 WDY:11; - UINT32 _PAD2:10; + uint32 EXBP:14; + uint32 EXBW:6; + uint32 FBIN:2; + uint32 WFFMD:1; + uint32 EMODA:2; + uint32 EMODC:2; + uint32 _PAD1:5; + uint32 WDX:11; + uint32 WDY:11; + uint32 _PAD2:10; REG_END REG64_(GSReg, EXTDATA) - UINT32 SX:12; - UINT32 SY:11; - UINT32 SMPH:4; - UINT32 SMPV:2; - UINT32 _PAD1:3; - UINT32 WW:12; - UINT32 WH:11; - UINT32 _PAD2:9; + uint32 SX:12; + uint32 SY:11; + uint32 SMPH:4; + uint32 SMPV:2; + uint32 _PAD1:3; + uint32 WW:12; + uint32 WH:11; + uint32 _PAD2:9; REG_END REG64_(GSReg, EXTWRITE) - UINT32 WRITE:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 WRITE:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GSReg, IMR) - UINT32 _PAD1:8; - UINT32 SIGMSK:1; - UINT32 FINISHMSK:1; - UINT32 HSMSK:1; - UINT32 VSMSK:1; - UINT32 EDWMSK:1; - UINT32 _PAD2:19; - UINT32 _PAD3:32; + uint32 _PAD1:8; + uint32 SIGMSK:1; + uint32 FINISHMSK:1; + uint32 HSMSK:1; + uint32 VSMSK:1; + uint32 EDWMSK:1; + uint32 _PAD2:19; + uint32 _PAD3:32; REG_END REG64_(GSReg, PMODE) - UINT32 EN1:1; - UINT32 EN2:1; - UINT32 CRTMD:3; - UINT32 MMOD:1; - UINT32 AMOD:1; - UINT32 SLBG:1; - UINT32 ALP:8; - UINT32 _PAD:16; - UINT32 _PAD1:32; + uint32 EN1:1; + uint32 EN2:1; + uint32 CRTMD:3; + uint32 MMOD:1; + uint32 AMOD:1; + uint32 SLBG:1; + uint32 ALP:8; + uint32 _PAD:16; + uint32 _PAD1:32; REG_END REG64_(GSReg, SIGLBLID) - UINT32 SIGID:32; - UINT32 LBLID:32; + uint32 SIGID:32; + uint32 LBLID:32; REG_END REG64_(GSReg, SMODE1) - UINT32 RC:3; - UINT32 LC:7; - UINT32 T1248:2; - UINT32 SLCK:1; - UINT32 CMOD:2; - UINT32 EX:1; - UINT32 PRST:1; - UINT32 SINT:1; - UINT32 XPCK:1; - UINT32 PCK2:2; - UINT32 SPML:4; - UINT32 GCONT:1; // YCrCb - UINT32 PHS:1; - UINT32 PVS:1; - UINT32 PEHS:1; - UINT32 PEVS:1; - UINT32 CLKSEL:2; - UINT32 NVCK:1; - UINT32 SLCK2:1; - UINT32 VCKSEL:2; - UINT32 VHP:1; - UINT32 _PAD1:27; + uint32 RC:3; + uint32 LC:7; + uint32 T1248:2; + uint32 SLCK:1; + uint32 CMOD:2; + uint32 EX:1; + uint32 PRST:1; + uint32 SINT:1; + uint32 XPCK:1; + uint32 PCK2:2; + uint32 SPML:4; + uint32 GCONT:1; // YCrCb + uint32 PHS:1; + uint32 PVS:1; + uint32 PEHS:1; + uint32 PEVS:1; + uint32 CLKSEL:2; + uint32 NVCK:1; + uint32 SLCK2:1; + uint32 VCKSEL:2; + uint32 VHP:1; + uint32 _PAD1:27; REG_END /* @@ -430,11 +430,11 @@ CLKSEL=1 CMOD=0 EX=0 GCONT=0 LC=32 NVCK=1 PCK2=0 PEHS=0 PEVS=0 PHS=0 PRST=1 PVS= */ REG64_(GSReg, SMODE2) - UINT32 INT:1; - UINT32 FFMD:1; - UINT32 DPMS:2; - UINT32 _PAD2:28; - UINT32 _PAD3:32; + uint32 INT:1; + uint32 FFMD:1; + uint32 DPMS:2; + uint32 _PAD2:28; + uint32 _PAD3:32; REG_END REG64_(GSReg, SRFSH) @@ -474,41 +474,41 @@ REG_SET_END // #define SET_GIF_REG(gifTag, iRegNo, uiValue) \ - {((GIFTag*)&gifTag)->ai64[1] |= (((uiValue) & 0xf) << ((iRegNo) << 2));} + {((GIFTag*)&gifTag)->u64[1] |= (((uiValue) & 0xf) << ((iRegNo) << 2));} #ifdef _M_AMD64 #define GET_GIF_REG(tag, reg) \ - (((tag).ai64[1] >> ((reg) << 2)) & 0xf) + (((tag).u64[1] >> ((reg) << 2)) & 0xf) #else #define GET_GIF_REG(tag, reg) \ - (((tag).ai32[2 + ((reg) >> 3)] >> (((reg) & 7) << 2)) & 0xf) + (((tag).u32[2 + ((reg) >> 3)] >> (((reg) & 7) << 2)) & 0xf) #endif // // GIFTag REG128(GIFTag) - UINT32 NLOOP:15; - UINT32 EOP:1; - UINT32 _PAD1:16; - UINT32 _PAD2:14; - UINT32 PRE:1; - UINT32 PRIM:11; - UINT32 FLG:2; // enum GIF_FLG - UINT32 NREG:4; - UINT64 REGS:64; + uint32 NLOOP:15; + uint32 EOP:1; + uint32 _PAD1:16; + uint32 _PAD2:14; + uint32 PRE:1; + uint32 PRIM:11; + uint32 FLG:2; // enum GIF_FLG + uint32 NREG:4; + uint64 REGS:64; REG_END // GIFReg REG64_(GIFReg, ALPHA) - UINT32 A:2; - UINT32 B:2; - UINT32 C:2; - UINT32 D:2; - UINT32 _PAD1:24; - UINT32 FIX:8; - UINT32 _PAD2:24; + uint32 A:2; + uint32 B:2; + uint32 C:2; + uint32 D:2; + uint32 _PAD1:24; + uint32 FIX:8; + uint32 _PAD2:24; REG_END2 // opaque => output will be Cs/As __forceinline bool IsOpaque() const {return (A == B || C == 2 && FIX == 0) && D == 0 || (A == 0 && B == D && C == 2 && FIX == 0x80);} @@ -516,18 +516,18 @@ REG_END2 REG_END2 REG64_(GIFReg, BITBLTBUF) - UINT32 SBP:14; - UINT32 _PAD1:2; - UINT32 SBW:6; - UINT32 _PAD2:2; - UINT32 SPSM:6; - UINT32 _PAD3:2; - UINT32 DBP:14; - UINT32 _PAD4:2; - UINT32 DBW:6; - UINT32 _PAD5:2; - UINT32 DPSM:6; - UINT32 _PAD6:2; + uint32 SBP:14; + uint32 _PAD1:2; + uint32 SBW:6; + uint32 _PAD2:2; + uint32 SPSM:6; + uint32 _PAD3:2; + uint32 DBP:14; + uint32 _PAD4:2; + uint32 DBW:6; + uint32 _PAD5:2; + uint32 DPSM:6; + uint32 _PAD6:2; REG_END REG64_(GIFReg, CLAMP) @@ -535,212 +535,212 @@ union { struct { - UINT32 WMS:2; - UINT32 WMT:2; - UINT32 MINU:10; - UINT32 MAXU:10; - UINT32 _PAD1:8; - UINT32 _PAD2:2; - UINT32 MAXV:10; - UINT32 _PAD3:20; + uint32 WMS:2; + uint32 WMT:2; + uint32 MINU:10; + uint32 MAXU:10; + uint32 _PAD1:8; + uint32 _PAD2:2; + uint32 MAXV:10; + uint32 _PAD3:20; }; struct { - UINT64 _PAD4:24; - UINT64 MINV:10; - UINT64 _PAD5:30; + uint64 _PAD4:24; + uint64 MINV:10; + uint64 _PAD5:30; }; }; REG_END REG64_(GIFReg, COLCLAMP) - UINT32 CLAMP:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 CLAMP:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GIFReg, DIMX) - INT32 DM00:3; - INT32 _PAD00:1; - INT32 DM01:3; - INT32 _PAD01:1; - INT32 DM02:3; - INT32 _PAD02:1; - INT32 DM03:3; - INT32 _PAD03:1; - INT32 DM10:3; - INT32 _PAD10:1; - INT32 DM11:3; - INT32 _PAD11:1; - INT32 DM12:3; - INT32 _PAD12:1; - INT32 DM13:3; - INT32 _PAD13:1; - INT32 DM20:3; - INT32 _PAD20:1; - INT32 DM21:3; - INT32 _PAD21:1; - INT32 DM22:3; - INT32 _PAD22:1; - INT32 DM23:3; - INT32 _PAD23:1; - INT32 DM30:3; - INT32 _PAD30:1; - INT32 DM31:3; - INT32 _PAD31:1; - INT32 DM32:3; - INT32 _PAD32:1; - INT32 DM33:3; - INT32 _PAD33:1; + int32 DM00:3; + int32 _PAD00:1; + int32 DM01:3; + int32 _PAD01:1; + int32 DM02:3; + int32 _PAD02:1; + int32 DM03:3; + int32 _PAD03:1; + int32 DM10:3; + int32 _PAD10:1; + int32 DM11:3; + int32 _PAD11:1; + int32 DM12:3; + int32 _PAD12:1; + int32 DM13:3; + int32 _PAD13:1; + int32 DM20:3; + int32 _PAD20:1; + int32 DM21:3; + int32 _PAD21:1; + int32 DM22:3; + int32 _PAD22:1; + int32 DM23:3; + int32 _PAD23:1; + int32 DM30:3; + int32 _PAD30:1; + int32 DM31:3; + int32 _PAD31:1; + int32 DM32:3; + int32 _PAD32:1; + int32 DM33:3; + int32 _PAD33:1; REG_END REG64_(GIFReg, DTHE) - UINT32 DTHE:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 DTHE:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GIFReg, FBA) - UINT32 FBA:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 FBA:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GIFReg, FINISH) - UINT32 _PAD1:32; - UINT32 _PAD2:32; + uint32 _PAD1:32; + uint32 _PAD2:32; REG_END REG64_(GIFReg, FOG) - UINT32 _PAD1:32; - UINT32 _PAD2:24; - UINT32 F:8; + uint32 _PAD1:32; + uint32 _PAD2:24; + uint32 F:8; REG_END REG64_(GIFReg, FOGCOL) - UINT32 FCR:8; - UINT32 FCG:8; - UINT32 FCB:8; - UINT32 _PAD1:8; - UINT32 _PAD2:32; + uint32 FCR:8; + uint32 FCG:8; + uint32 FCB:8; + uint32 _PAD1:8; + uint32 _PAD2:32; REG_END REG64_(GIFReg, FRAME) - UINT32 FBP:9; - UINT32 _PAD1:7; - UINT32 FBW:6; - UINT32 _PAD2:2; - UINT32 PSM:6; - UINT32 _PAD3:2; - UINT32 FBMSK:32; + uint32 FBP:9; + uint32 _PAD1:7; + uint32 FBW:6; + uint32 _PAD2:2; + uint32 PSM:6; + uint32 _PAD3:2; + uint32 FBMSK:32; REG_END2 - UINT32 Block() const {return FBP << 5;} + uint32 Block() const {return FBP << 5;} REG_END2 REG64_(GIFReg, HWREG) - UINT32 DATA_LOWER:32; - UINT32 DATA_UPPER:32; + uint32 DATA_LOWER:32; + uint32 DATA_UPPER:32; REG_END REG64_(GIFReg, LABEL) - UINT32 ID:32; - UINT32 IDMSK:32; + uint32 ID:32; + uint32 IDMSK:32; REG_END REG64_(GIFReg, MIPTBP1) - UINT64 TBP1:14; - UINT64 TBW1:6; - UINT64 TBP2:14; - UINT64 TBW2:6; - UINT64 TBP3:14; - UINT64 TBW3:6; - UINT64 _PAD:4; + uint64 TBP1:14; + uint64 TBW1:6; + uint64 TBP2:14; + uint64 TBW2:6; + uint64 TBP3:14; + uint64 TBW3:6; + uint64 _PAD:4; REG_END REG64_(GIFReg, MIPTBP2) - UINT64 TBP4:14; - UINT64 TBW4:6; - UINT64 TBP5:14; - UINT64 TBW5:6; - UINT64 TBP6:14; - UINT64 TBW6:6; - UINT64 _PAD:4; + uint64 TBP4:14; + uint64 TBW4:6; + uint64 TBP5:14; + uint64 TBW5:6; + uint64 TBP6:14; + uint64 TBW6:6; + uint64 _PAD:4; REG_END REG64_(GIFReg, NOP) - UINT32 _PAD1:32; - UINT32 _PAD2:32; + uint32 _PAD1:32; + uint32 _PAD2:32; REG_END REG64_(GIFReg, PABE) - UINT32 PABE:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 PABE:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GIFReg, PRIM) - UINT32 PRIM:3; - UINT32 IIP:1; - UINT32 TME:1; - UINT32 FGE:1; - UINT32 ABE:1; - UINT32 AA1:1; - UINT32 FST:1; - UINT32 CTXT:1; - UINT32 FIX:1; - UINT32 _PAD1:21; - UINT32 _PAD2:32; + uint32 PRIM:3; + uint32 IIP:1; + uint32 TME:1; + uint32 FGE:1; + uint32 ABE:1; + uint32 AA1:1; + uint32 FST:1; + uint32 CTXT:1; + uint32 FIX:1; + uint32 _PAD1:21; + uint32 _PAD2:32; REG_END REG64_(GIFReg, PRMODE) - UINT32 _PRIM:3; - UINT32 IIP:1; - UINT32 TME:1; - UINT32 FGE:1; - UINT32 ABE:1; - UINT32 AA1:1; - UINT32 FST:1; - UINT32 CTXT:1; - UINT32 FIX:1; - UINT32 _PAD2:21; - UINT32 _PAD3:32; + uint32 _PRIM:3; + uint32 IIP:1; + uint32 TME:1; + uint32 FGE:1; + uint32 ABE:1; + uint32 AA1:1; + uint32 FST:1; + uint32 CTXT:1; + uint32 FIX:1; + uint32 _PAD2:21; + uint32 _PAD3:32; REG_END REG64_(GIFReg, PRMODECONT) - UINT32 AC:1; - UINT32 _PAD1:31; - UINT32 _PAD2:32; + uint32 AC:1; + uint32 _PAD1:31; + uint32 _PAD2:32; REG_END REG64_(GIFReg, RGBAQ) - UINT32 R:8; - UINT32 G:8; - UINT32 B:8; - UINT32 A:8; + uint32 R:8; + uint32 G:8; + uint32 B:8; + uint32 A:8; float Q; REG_END REG64_(GIFReg, SCANMSK) - UINT32 MSK:2; - UINT32 _PAD1:30; - UINT32 _PAD2:32; + uint32 MSK:2; + uint32 _PAD1:30; + uint32 _PAD2:32; REG_END REG64_(GIFReg, SCISSOR) - UINT32 SCAX0:11; - UINT32 _PAD1:5; - UINT32 SCAX1:11; - UINT32 _PAD2:5; - UINT32 SCAY0:11; - UINT32 _PAD3:5; - UINT32 SCAY1:11; - UINT32 _PAD4:5; + uint32 SCAX0:11; + uint32 _PAD1:5; + uint32 SCAX1:11; + uint32 _PAD2:5; + uint32 SCAY0:11; + uint32 _PAD3:5; + uint32 SCAY1:11; + uint32 _PAD4:5; REG_END REG64_(GIFReg, SIGNAL) - UINT32 ID:32; - UINT32 IDMSK:32; + uint32 ID:32; + uint32 IDMSK:32; REG_END REG64_(GIFReg, ST) @@ -749,16 +749,16 @@ REG64_(GIFReg, ST) REG_END REG64_(GIFReg, TEST) - UINT32 ATE:1; - UINT32 ATST:3; - UINT32 AREF:8; - UINT32 AFAIL:2; - UINT32 DATE:1; - UINT32 DATM:1; - UINT32 ZTE:1; - UINT32 ZTST:2; - UINT32 _PAD1:13; - UINT32 _PAD2:32; + uint32 ATE:1; + uint32 ATST:3; + uint32 AREF:8; + uint32 AFAIL:2; + uint32 DATE:1; + uint32 DATM:1; + uint32 ZTE:1; + uint32 ZTST:2; + uint32 _PAD1:13; + uint32 _PAD2:32; REG_END2 __forceinline bool DoFirstPass() {return !ATE || ATST != 0;} // not all pixels fail automatically __forceinline bool DoSecondPass() {return ATE && ATST != 1 && AFAIL != 0;} // pixels may fail, write fb/z @@ -770,42 +770,42 @@ union { struct { - UINT32 TBP0:14; - UINT32 TBW:6; - UINT32 PSM:6; - UINT32 TW:4; - UINT32 _PAD1:2; - UINT32 _PAD2:2; - UINT32 TCC:1; - UINT32 TFX:2; - UINT32 CBP:14; - UINT32 CPSM:4; - UINT32 CSM:1; - UINT32 CSA:5; - UINT32 CLD:3; + uint32 TBP0:14; + uint32 TBW:6; + uint32 PSM:6; + uint32 TW:4; + uint32 _PAD1:2; + uint32 _PAD2:2; + uint32 TCC:1; + uint32 TFX:2; + uint32 CBP:14; + uint32 CPSM:4; + uint32 CSM:1; + uint32 CSA:5; + uint32 CLD:3; }; struct { - UINT64 _PAD3:30; - UINT64 TH:4; - UINT64 _PAD4:30; + uint64 _PAD3:30; + uint64 TH:4; + uint64 _PAD4:30; }; }; REG_END REG64_(GIFReg, TEX1) - UINT32 LCM:1; - UINT32 _PAD1:1; - UINT32 MXL:3; - UINT32 MMAG:1; - UINT32 MMIN:3; - UINT32 MTBA:1; - UINT32 _PAD2:9; - UINT32 L:2; - UINT32 _PAD3:11; - UINT32 K:12; - UINT32 _PAD4:20; + uint32 LCM:1; + uint32 _PAD1:1; + uint32 MXL:3; + uint32 MMAG:1; + uint32 MMIN:3; + uint32 MTBA:1; + uint32 _PAD2:9; + uint32 L:2; + uint32 _PAD3:11; + uint32 K:12; + uint32 _PAD4:20; REG_END2 bool IsLinear() const { @@ -817,105 +817,105 @@ REG_END2 REG_END2 REG64_(GIFReg, TEX2) - UINT32 _PAD1:20; - UINT32 PSM:6; - UINT32 _PAD2:6; - UINT32 _PAD3:5; - UINT32 CBP:14; - UINT32 CPSM:4; - UINT32 CSM:1; - UINT32 CSA:5; - UINT32 CLD:3; + uint32 _PAD1:20; + uint32 PSM:6; + uint32 _PAD2:6; + uint32 _PAD3:5; + uint32 CBP:14; + uint32 CPSM:4; + uint32 CSM:1; + uint32 CSA:5; + uint32 CLD:3; REG_END REG64_(GIFReg, TEXA) - UINT32 TA0:8; - UINT32 _PAD1:7; - UINT32 AEM:1; - UINT32 _PAD2:16; - UINT32 TA1:8; - UINT32 _PAD3:24; + uint32 TA0:8; + uint32 _PAD1:7; + uint32 AEM:1; + uint32 _PAD2:16; + uint32 TA1:8; + uint32 _PAD3:24; REG_END REG64_(GIFReg, TEXCLUT) - UINT32 CBW:6; - UINT32 COU:6; - UINT32 COV:10; - UINT32 _PAD1:10; - UINT32 _PAD2:32; + uint32 CBW:6; + uint32 COU:6; + uint32 COV:10; + uint32 _PAD1:10; + uint32 _PAD2:32; REG_END REG64_(GIFReg, TEXFLUSH) - UINT32 _PAD1:32; - UINT32 _PAD2:32; + uint32 _PAD1:32; + uint32 _PAD2:32; REG_END REG64_(GIFReg, TRXDIR) - UINT32 XDIR:2; - UINT32 _PAD1:30; - UINT32 _PAD2:32; + uint32 XDIR:2; + uint32 _PAD1:30; + uint32 _PAD2:32; REG_END REG64_(GIFReg, TRXPOS) - UINT32 SSAX:11; - UINT32 _PAD1:5; - UINT32 SSAY:11; - UINT32 _PAD2:5; - UINT32 DSAX:11; - UINT32 _PAD3:5; - UINT32 DSAY:11; - UINT32 DIR:2; - UINT32 _PAD4:3; + uint32 SSAX:11; + uint32 _PAD1:5; + uint32 SSAY:11; + uint32 _PAD2:5; + uint32 DSAX:11; + uint32 _PAD3:5; + uint32 DSAY:11; + uint32 DIR:2; + uint32 _PAD4:3; REG_END REG64_(GIFReg, TRXREG) - UINT32 RRW:12; - UINT32 _PAD1:20; - UINT32 RRH:12; - UINT32 _PAD2:20; + uint32 RRW:12; + uint32 _PAD1:20; + uint32 RRH:12; + uint32 _PAD2:20; REG_END // GSState::GIFPackedRegHandlerUV and GSState::GIFRegHandlerUV will make sure that the _PAD1/2 bits are set to zero REG64_(GIFReg, UV) - UINT32 U:16; -// UINT32 _PAD1:2; - UINT32 V:16; -// UINT32 _PAD2:2; - UINT32 _PAD3:32; + uint32 U:16; +// uint32 _PAD1:2; + uint32 V:16; +// uint32 _PAD2:2; + uint32 _PAD3:32; REG_END // GSState::GIFRegHandlerXYOFFSET will make sure that the _PAD1/2 bits are set to zero REG64_(GIFReg, XYOFFSET) - UINT32 OFX; // :16; UINT32 _PAD1:16; - UINT32 OFY; // :16; UINT32 _PAD2:16; + uint32 OFX; // :16; uint32 _PAD1:16; + uint32 OFY; // :16; uint32 _PAD2:16; REG_END REG64_(GIFReg, XYZ) - UINT32 X:16; - UINT32 Y:16; - UINT32 Z:32; + uint32 X:16; + uint32 Y:16; + uint32 Z:32; REG_END REG64_(GIFReg, XYZF) - UINT32 X:16; - UINT32 Y:16; - UINT32 Z:24; - UINT32 F:8; + uint32 X:16; + uint32 Y:16; + uint32 Z:24; + uint32 F:8; REG_END REG64_(GIFReg, ZBUF) - UINT32 ZBP:9; - UINT32 _PAD1:15; - // UINT32 PSM:4; - // UINT32 _PAD2:4; - UINT32 PSM:6; - UINT32 _PAD2:2; - UINT32 ZMSK:1; - UINT32 _PAD3:31; + uint32 ZBP:9; + uint32 _PAD1:15; + // uint32 PSM:4; + // uint32 _PAD2:4; + uint32 PSM:6; + uint32 _PAD2:2; + uint32 ZMSK:1; + uint32 _PAD3:31; REG_END2 - UINT32 Block() const {return ZBP << 5;} + uint32 Block() const {return ZBP << 5;} REG_END2 REG64_SET(GIFReg) @@ -964,87 +964,87 @@ REG_SET_END // GIFPacked REG128_(GIFPacked, PRIM) - UINT32 PRIM:11; - UINT32 _PAD1:21; - UINT32 _PAD2:32; - UINT32 _PAD3:32; - UINT32 _PAD4:32; + uint32 PRIM:11; + uint32 _PAD1:21; + uint32 _PAD2:32; + uint32 _PAD3:32; + uint32 _PAD4:32; REG_END REG128_(GIFPacked, RGBA) - UINT32 R:8; - UINT32 _PAD1:24; - UINT32 G:8; - UINT32 _PAD2:24; - UINT32 B:8; - UINT32 _PAD3:24; - UINT32 A:8; - UINT32 _PAD4:24; + uint32 R:8; + uint32 _PAD1:24; + uint32 G:8; + uint32 _PAD2:24; + uint32 B:8; + uint32 _PAD3:24; + uint32 A:8; + uint32 _PAD4:24; REG_END REG128_(GIFPacked, STQ) float S; float T; float Q; - UINT32 _PAD1:32; + uint32 _PAD1:32; REG_END REG128_(GIFPacked, UV) - UINT32 U:14; - UINT32 _PAD1:18; - UINT32 V:14; - UINT32 _PAD2:18; - UINT32 _PAD3:32; - UINT32 _PAD4:32; + uint32 U:14; + uint32 _PAD1:18; + uint32 V:14; + uint32 _PAD2:18; + uint32 _PAD3:32; + uint32 _PAD4:32; REG_END REG128_(GIFPacked, XYZF2) - UINT32 X:16; - UINT32 _PAD1:16; - UINT32 Y:16; - UINT32 _PAD2:16; - UINT32 _PAD3:4; - UINT32 Z:24; - UINT32 _PAD4:4; - UINT32 _PAD5:4; - UINT32 F:8; - UINT32 _PAD6:3; - UINT32 ADC:1; - UINT32 _PAD7:16; + uint32 X:16; + uint32 _PAD1:16; + uint32 Y:16; + uint32 _PAD2:16; + uint32 _PAD3:4; + uint32 Z:24; + uint32 _PAD4:4; + uint32 _PAD5:4; + uint32 F:8; + uint32 _PAD6:3; + uint32 ADC:1; + uint32 _PAD7:16; REG_END REG128_(GIFPacked, XYZ2) - UINT32 X:16; - UINT32 _PAD1:16; - UINT32 Y:16; - UINT32 _PAD2:16; - UINT32 Z:32; - UINT32 _PAD3:15; - UINT32 ADC:1; - UINT32 _PAD4:16; + uint32 X:16; + uint32 _PAD1:16; + uint32 Y:16; + uint32 _PAD2:16; + uint32 Z:32; + uint32 _PAD3:15; + uint32 ADC:1; + uint32 _PAD4:16; REG_END REG128_(GIFPacked, FOG) - UINT32 _PAD1:32; - UINT32 _PAD2:32; - UINT32 _PAD3:32; - UINT32 _PAD4:4; - UINT32 F:8; - UINT32 _PAD5:20; + uint32 _PAD1:32; + uint32 _PAD2:32; + uint32 _PAD3:32; + uint32 _PAD4:4; + uint32 F:8; + uint32 _PAD5:20; REG_END REG128_(GIFPacked, A_D) - UINT64 DATA:64; - UINT32 ADDR:8; // enum GIF_A_D_REG - UINT32 _PAD1:24; - UINT32 _PAD2:32; + uint64 DATA:64; + uint32 ADDR:8; // enum GIF_A_D_REG + uint32 _PAD1:24; + uint32 _PAD2:32; REG_END REG128_(GIFPacked, NOP) - UINT32 _PAD1:32; - UINT32 _PAD2:32; - UINT32 _PAD3:32; - UINT32 _PAD4:32; + uint32 _PAD1:32; + uint32 _PAD2:32; + uint32 _PAD3:32; + uint32 _PAD4:32; REG_END REG128_SET(GIFPackedReg) @@ -1063,10 +1063,10 @@ REG_SET_END __declspec(align(16)) struct GIFPath { GIFTag tag; - UINT32 reg; - UINT32 nreg; - UINT32 nloop; - UINT32 adonly; + uint32 reg; + uint32 nreg; + uint32 nloop; + uint32 adonly; GSVector4i regs; void SetTag(const void* mem) @@ -1080,9 +1080,9 @@ __declspec(align(16)) struct GIFPath adonly = nreg == 1 && regs.u8[0] == GIF_REG_A_D; } - __forceinline DWORD GetReg() + __forceinline uint8 GetReg() { - return regs.u8[reg]; // (DWORD)GET_GIF_REG(tag, nreg); + return regs.u8[reg]; // GET_GIF_REG(tag, reg); } __forceinline bool StepReg() @@ -1108,36 +1108,36 @@ struct GSPrivRegSet struct { GSRegPMODE PMODE; - UINT64 _pad1; + uint64 _pad1; GSRegSMODE1 SMODE1; - UINT64 _pad2; + uint64 _pad2; GSRegSMODE2 SMODE2; - UINT64 _pad3; + uint64 _pad3; GSRegSRFSH SRFSH; - UINT64 _pad4; + uint64 _pad4; GSRegSYNCH1 SYNCH1; - UINT64 _pad5; + uint64 _pad5; GSRegSYNCH2 SYNCH2; - UINT64 _pad6; + uint64 _pad6; GSRegSYNCV SYNCV; - UINT64 _pad7; + uint64 _pad7; struct { GSRegDISPFB DISPFB; - UINT64 _pad1; + uint64 _pad1; GSRegDISPLAY DISPLAY; - UINT64 _pad2; + uint64 _pad2; } DISP[2]; GSRegEXTBUF EXTBUF; - UINT64 _pad8; + uint64 _pad8; GSRegEXTDATA EXTDATA; - UINT64 _pad9; + uint64 _pad9; GSRegEXTWRITE EXTWRITE; - UINT64 _pad10; + uint64 _pad10; GSRegBGCOLOR BGCOLOR; - UINT64 _pad11; + uint64 _pad11; }; - BYTE _pad12[0x1000]; + uint8 _pad12[0x1000]; }; union @@ -1145,27 +1145,27 @@ struct GSPrivRegSet struct { GSRegCSR CSR; - UINT64 _pad13; + uint64 _pad13; GSRegIMR IMR; - UINT64 _pad14; - UINT64 _unk1[4]; + uint64 _pad14; + uint64 _unk1[4]; GSRegBUSDIR BUSDIR; - UINT64 _pad15; - UINT64 _unk2[6]; + uint64 _pad15; + uint64 _unk2[6]; GSRegSIGLBLID SIGLBLID; - UINT64 _pad16; + uint64 _pad16; }; - BYTE _pad17[0x1000]; + uint8 _pad17[0x1000]; }; }; #pragma pack(pop) enum {KEYPRESS=1, KEYRELEASE=2}; -struct keyEvent {UINT32 key, event;}; +struct keyEvent {uint32 key, event;}; enum {FREEZE_LOAD=0, FREEZE_SAVE=1, FREEZE_SIZE=2}; -struct GSFreezeData {int size; BYTE* data;}; +struct GSFreezeData {int size; uint8* data;}; enum stateType {ST_WRITE, ST_TRANSFER, ST_VSYNC}; diff --git a/plugins/GSdx/GSBlock.h b/plugins/GSdx/GSBlock.h index e244a2b240..2ac27a0f12 100644 --- a/plugins/GSdx/GSBlock.h +++ b/plugins/GSdx/GSBlock.h @@ -42,10 +42,8 @@ class GSBlock static const GSVector4i m_uw8hmask3; public: - template __forceinline static void WriteColumn32(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template __forceinline static void WriteColumn32(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s0 = (const GSVector4i*)&src[srcpitch * 0]; const GSVector4i* s1 = (const GSVector4i*)&src[srcpitch * 1]; @@ -92,33 +90,10 @@ public: #endif } - - #else - - const BYTE* d = &columnTable32[(i & 3) << 1][0]; - - for(int j = 0; j < 2; j++, d += 8, src += srcpitch) - { - for(int i = 0; i < 8; i++) - { - if(mask == 0xffffffff) - { - ((DWORD*)dst)[d[i]] = ((DWORD*)src)[i]; - } - else - { - ((DWORD*)dst)[d[i]] = (((DWORD*)dst)[d[i]] & ~mask) | (((DWORD*)src)[i] & mask); - } - } - } - - #endif } - template __forceinline static void WriteColumn16(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template __forceinline static void WriteColumn16(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s0 = (const GSVector4i*)&src[srcpitch * 0]; const GSVector4i* s1 = (const GSVector4i*)&src[srcpitch * 1]; @@ -134,26 +109,10 @@ public: ((GSVector4i*)dst)[i * 4 + 1] = v2; ((GSVector4i*)dst)[i * 4 + 2] = v1; ((GSVector4i*)dst)[i * 4 + 3] = v3; - - #else - - const BYTE* d = &columnTable16[(i & 3) << 1][0]; - - for(int j = 0; j < 2; j++, d += 16, src += srcpitch) - { - for(int i = 0; i < 16; i++) - { - ((WORD*)dst)[d[i]] = ((WORD*)src)[i]; - } - } - - #endif } - template __forceinline static void WriteColumn8(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template __forceinline static void WriteColumn8(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - GSVector4i v0 = GSVector4i::load(&src[srcpitch * 0]); GSVector4i v1 = GSVector4i::load(&src[srcpitch * 1]); GSVector4i v2 = GSVector4i::load(&src[srcpitch * 2]); @@ -178,28 +137,12 @@ public: ((GSVector4i*)dst)[i * 4 + 1] = v2; ((GSVector4i*)dst)[i * 4 + 2] = v1; ((GSVector4i*)dst)[i * 4 + 3] = v3; - - #else - - const BYTE* d = &columnTable8[(i & 3) << 2][0]; - - for(int j = 0; j < 4; j++, d += 16, src += srcpitch) - { - for(int i = 0; i < 16; i++) - { - dst[d[i]] = src[i]; - } - } - - #endif } - template __forceinline static void WriteColumn4(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template __forceinline static void WriteColumn4(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { // TODO: pshufb - #if _M_SSE >= 0x200 - GSVector4i v0 = GSVector4i::load(&src[srcpitch * 0]); GSVector4i v1 = GSVector4i::load(&src[srcpitch * 1]); GSVector4i v2 = GSVector4i::load(&src[srcpitch * 2]); @@ -225,26 +168,9 @@ public: ((GSVector4i*)dst)[i * 4 + 1] = v1; ((GSVector4i*)dst)[i * 4 + 2] = v2; ((GSVector4i*)dst)[i * 4 + 3] = v3; - - #else - - const WORD* d = &columnTable4[(i & 3) << 2][0]; - - for(int j = 0; j < 4; j++, d += 32, src += srcpitch) - { - for(int i = 0; i < 32; i++) - { - DWORD addr = d[i]; - BYTE c = (src[i >> 1] >> ((i & 1) << 2)) & 0x0f; - DWORD shift = (addr & 1) << 2; - dst[addr >> 1] = (dst[addr >> 1] & (0xf0 >> shift)) | (c << shift); - } - } - - #endif } - template static void WriteColumn32(int y, BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteColumn32(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { switch((y >> 1) & 3) { @@ -256,7 +182,7 @@ public: } } - template static void WriteColumn16(int y, BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteColumn16(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { switch((y >> 1) & 3) { @@ -268,7 +194,7 @@ public: } } - template static void WriteColumn8(int y, BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteColumn8(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { switch((y >> 2) & 3) { @@ -280,7 +206,7 @@ public: } } - template static void WriteColumn4(int y, BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteColumn4(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { switch((y >> 2) & 3) { @@ -292,10 +218,8 @@ public: } } - template static void WriteBlock32(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteBlock32(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - WriteColumn32<0, aligned, mask>(dst, src, srcpitch); src += srcpitch * 2; WriteColumn32<1, aligned, mask>(dst, src, srcpitch); @@ -303,33 +227,10 @@ public: WriteColumn32<2, aligned, mask>(dst, src, srcpitch); src += srcpitch * 2; WriteColumn32<3, aligned, mask>(dst, src, srcpitch); - - #else - - const BYTE* d = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, d += 8, src += srcpitch) - { - for(int i = 0; i < 8; i++) - { - if(mask == 0xffffffff) - { - ((DWORD*)dst)[d[i]] = ((DWORD*)src)[i]; - } - else - { - ((DWORD*)dst)[d[i]] = (((DWORD*)dst)[d[i]] & ~mask) | (((DWORD*)src)[i] & mask); - } - } - } - - #endif } - template static void WriteBlock16(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteBlock16(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - WriteColumn16<0, aligned>(dst, src, srcpitch); src += srcpitch * 2; WriteColumn16<1, aligned>(dst, src, srcpitch); @@ -337,26 +238,10 @@ public: WriteColumn16<2, aligned>(dst, src, srcpitch); src += srcpitch * 2; WriteColumn16<3, aligned>(dst, src, srcpitch); - - #else - - const BYTE* d = &columnTable16[0][0]; - - for(int j = 0; j < 8; j++, d += 16, src += srcpitch) - { - for(int i = 0; i < 16; i++) - { - ((WORD*)dst)[d[i]] = ((WORD*)src)[i]; - } - } - - #endif } - template static void WriteBlock8(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteBlock8(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - WriteColumn8<0, aligned>(dst, src, srcpitch); src += srcpitch * 4; WriteColumn8<1, aligned>(dst, src, srcpitch); @@ -364,26 +249,10 @@ public: WriteColumn8<2, aligned>(dst, src, srcpitch); src += srcpitch * 4; WriteColumn8<3, aligned>(dst, src, srcpitch); - - #else - - const BYTE* d = &columnTable8[0][0]; - - for(int j = 0; j < 16; j++, d += 16, src += srcpitch) - { - for(int i = 0; i < 16; i++) - { - dst[d[i]] = src[i]; - } - } - - #endif } - template static void WriteBlock4(BYTE* RESTRICT dst, const BYTE* RESTRICT src, int srcpitch) + template static void WriteBlock4(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) { - #if _M_SSE >= 0x200 - WriteColumn4<0, aligned>(dst, src, srcpitch); src += srcpitch * 4; WriteColumn4<1, aligned>(dst, src, srcpitch); @@ -391,29 +260,10 @@ public: WriteColumn4<2, aligned>(dst, src, srcpitch); src += srcpitch * 4; WriteColumn4<3, aligned>(dst, src, srcpitch); - - #else - - const WORD* d = &columnTable4[0][0]; - - for(int j = 0; j < 16; j++, d += 32, src += srcpitch) - { - for(int i = 0; i < 32; i++) - { - DWORD addr = d[i]; - BYTE c = (src[i >> 1] >> ((i & 1) << 2)) & 0x0f; - DWORD shift = (addr & 1) << 2; - dst[addr >> 1] = (dst[addr >> 1] & (0xf0 >> shift)) | (c << shift); - } - } - - #endif } - template __forceinline static void ReadColumn32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template __forceinline static void ReadColumn32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i v0 = s[i * 4 + 0]; @@ -430,23 +280,9 @@ public: GSVector4i::store(&d0[1], v1); GSVector4i::store(&d1[0], v2); GSVector4i::store(&d1[1], v3); - - #else - - const BYTE* s = &columnTable32[(i & 3) << 1][0]; - - for(int j = 0; j < 2; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[i] = ((DWORD*)src)[s[i]]; - } - } - - #endif } - template __forceinline static void ReadColumn16(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template __forceinline static void ReadColumn16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { #if _M_SSE >= 0x301 @@ -468,7 +304,7 @@ public: GSVector4i::store(&d1[0], v1); GSVector4i::store(&d1[1], v3); - #elif _M_SSE >= 0x200 + #else const GSVector4i* s = (const GSVector4i*)src; @@ -489,22 +325,10 @@ public: GSVector4i::store(&d1[0], v2); GSVector4i::store(&d1[1], v3); - #else - - const BYTE* s = &columnTable16[(i & 3) << 1][0]; - - for(int j = 0; j < 2; j++, s += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - ((WORD*)dst)[i] = ((WORD*)src)[s[i]]; - } - } - #endif } - template __forceinline static void ReadColumn8(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template __forceinline static void ReadColumn8(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { #if _M_SSE >= 0x301 @@ -540,7 +364,7 @@ public: GSVector4i::store(&dst[dstpitch * 2], v1); GSVector4i::store(&dst[dstpitch * 3], v2); - #elif _M_SSE >= 0x200 + #else const GSVector4i* s = (const GSVector4i*)src; @@ -570,22 +394,10 @@ public: GSVector4i::store(&dst[dstpitch * 2], v2); GSVector4i::store(&dst[dstpitch * 3], v3); - #else - - const BYTE* s = &columnTable8[(i & 3) << 2][0]; - - for(int j = 0; j < 4; j++, s += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - dst[i] = src[s[i]]; - } - } - #endif } - template __forceinline static void ReadColumn4(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template __forceinline static void ReadColumn4(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { #if _M_SSE >= 0x301 @@ -619,7 +431,7 @@ public: GSVector4i::store(&dst[dstpitch * 2], v2); GSVector4i::store(&dst[dstpitch * 3], v3); - #elif _M_SSE >= 0x200 + #else const GSVector4i* s = (const GSVector4i*)src; @@ -657,25 +469,10 @@ public: GSVector4i::store(&dst[dstpitch * 2], v2); GSVector4i::store(&dst[dstpitch * 3], v3); - #else - - const WORD* s = &columnTable4[(i & 3) << 2][0]; - - for(int j = 0; j < 4; j++, s += 32, dst += dstpitch) - { - for(int i = 0; i < 32; i++) - { - DWORD addr = s[i]; - BYTE c = (src[addr >> 1] >> ((addr & 1) << 2)) & 0x0f; - int shift = (i & 1) << 2; - dst[i >> 1] = (dst[i >> 1] & (0xf0 >> shift)) | (c << shift); - } - } - #endif } - template static void ReadColumn32(int y, const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadColumn32(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { switch((y >> 1) & 3) { @@ -687,7 +484,7 @@ public: } } - template static void ReadColumn16(int y, const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadColumn16(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { switch((y >> 1) & 3) { @@ -699,7 +496,7 @@ public: } } - template static void ReadColumn8(int y, const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadColumn8(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { switch((y >> 2) & 3) { @@ -711,7 +508,7 @@ public: } } - template static void ReadColumn4(int y, const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadColumn4(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { switch((y >> 2) & 3) { @@ -723,10 +520,8 @@ public: } } - template static void ReadBlock32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadBlock32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - ReadColumn32<0, aligned>(src, dst, dstpitch); dst += dstpitch * 2; ReadColumn32<1, aligned>(src, dst, dstpitch); @@ -734,26 +529,10 @@ public: ReadColumn32<2, aligned>(src, dst, dstpitch); dst += dstpitch * 2; ReadColumn32<3, aligned>(src, dst, dstpitch); - - #else - - const BYTE* s = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[i] = ((DWORD*)src)[s[i]]; - } - } - - #endif } - template static void ReadBlock16(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadBlock16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - ReadColumn16<0, aligned>(src, dst, dstpitch); dst += dstpitch * 2; ReadColumn16<1, aligned>(src, dst, dstpitch); @@ -761,26 +540,10 @@ public: ReadColumn16<2, aligned>(src, dst, dstpitch); dst += dstpitch * 2; ReadColumn16<3, aligned>(src, dst, dstpitch); - - #else - - const BYTE* s = &columnTable16[0][0]; - - for(int j = 0; j < 8; j++, s += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - ((WORD*)dst)[i] = ((WORD*)src)[s[i]]; - } - } - - #endif } - template static void ReadBlock8(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadBlock8(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - ReadColumn8<0, aligned>(src, dst, dstpitch); dst += dstpitch * 4; ReadColumn8<1, aligned>(src, dst, dstpitch); @@ -788,26 +551,10 @@ public: ReadColumn8<2, aligned>(src, dst, dstpitch); dst += dstpitch * 4; ReadColumn8<3, aligned>(src, dst, dstpitch); - - #else - - const BYTE* s = &columnTable8[0][0]; - - for(int j = 0; j < 16; j++, s += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - dst[i] = src[s[i]]; - } - } - - #endif } - template static void ReadBlock4(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + template static void ReadBlock4(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - ReadColumn4<0, aligned>(src, dst, dstpitch); dst += dstpitch * 4; ReadColumn4<1, aligned>(src, dst, dstpitch); @@ -815,29 +562,10 @@ public: ReadColumn4<2, aligned>(src, dst, dstpitch); dst += dstpitch * 4; ReadColumn4<3, aligned>(src, dst, dstpitch); - - #else - - const WORD* s = &columnTable4[0][0]; - - for(int j = 0; j < 16; j++, s += 32, dst += dstpitch) - { - for(int i = 0; i < 32; i++) - { - DWORD addr = s[i]; - BYTE c = (src[addr >> 1] >> ((addr & 1) << 2)) & 0x0f; - int shift = (i & 1) << 2; - dst[i >> 1] = (dst[i >> 1] & (0xf0 >> shift)) | (c << shift); - } - } - - #endif } - __forceinline static void ReadBlock4P(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + __forceinline static void ReadBlock4P(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i v0, v1, v2, v3; @@ -896,18 +624,10 @@ public: dst += dstpitch * 2; } - - #else - - // TODO - - #endif } - __forceinline static void ReadBlock8HP(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + __forceinline static void ReadBlock8HP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i v0, v1, v2, v3; @@ -931,26 +651,10 @@ public: dst += dstpitch; } - - #else - - const BYTE* s = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((BYTE*)dst)[i] = ((DWORD*)src)[s[i]] >> 24; - } - } - - #endif } - __forceinline static void ReadBlock4HLP(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + __forceinline static void ReadBlock4HLP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i v0, v1, v2, v3; @@ -976,26 +680,10 @@ public: dst += dstpitch; } - - #else - - const BYTE* s = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((BYTE*)dst)[i] = (((DWORD*)src)[s[i]] >> 24) & 0xf; - } - } - - #endif } - __forceinline static void ReadBlock4HHP(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch) + __forceinline static void ReadBlock4HHP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i v0, v1, v2, v3; @@ -1019,26 +707,10 @@ public: dst += dstpitch; } - - #else - - const BYTE* s = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((BYTE*)dst)[i] = ((DWORD*)src)[s[i]] >> 28; - } - } - - #endif } - static void UnpackBlock24(const BYTE* RESTRICT src, int srcpitch, DWORD* RESTRICT dst) + static void UnpackBlock24(const uint8* RESTRICT src, int srcpitch, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i mask = GSVector4i::x00ffffff(); for(int i = 0; i < 4; i++, src += srcpitch * 2) @@ -1061,24 +733,10 @@ public: ((GSVector4i*)dst)[i * 4 + 3] = v0.upl32(v0.srl<3>()).upl64(v0.srl<6>().upl32(v0.srl<9>())) & mask; } - - #else - - for(int j = 0, diff = srcpitch - 8 * 3; j < 8; j++, src += diff, dst += 8) - { - for(int i = 0; i < 8; i++, src += 3) - { - dst[i] = (src[2] << 16) | (src[1] << 8) | src[0]; - } - } - - #endif } - static void UnpackBlock8H(const BYTE* RESTRICT src, int srcpitch, DWORD* RESTRICT dst) + static void UnpackBlock8H(const uint8* RESTRICT src, int srcpitch, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i zero = GSVector4i::zero(); for(int i = 0; i < 4; i++, src += srcpitch * 2) @@ -1093,34 +751,20 @@ public: ((GSVector4i*)dst)[i * 4 + 2] = zero.upl16(v1); ((GSVector4i*)dst)[i * 4 + 3] = zero.uph16(v1); } - - #else - - for(int j = 0; j < 8; j++, src += srcpitch, dst += 8) - { - for(int i = 0; i < 8; i++) - { - dst[i] = src[i] << 24; - } - } - - #endif } - static void UnpackBlock4HL(const BYTE* RESTRICT src, int srcpitch, DWORD* RESTRICT dst) + static void UnpackBlock4HL(const uint8* RESTRICT src, int srcpitch, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i zero = GSVector4i::zero(); GSVector4i mask(0x0f0f0f0f); for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = v & mask; GSVector4i hi = (v >> 4) & mask; @@ -1142,35 +786,20 @@ public: ((GSVector4i*)dst)[i * 8 + 6] = zero.upl16(v5); ((GSVector4i*)dst)[i * 8 + 7] = zero.uph16(v5); } - - #else - - for(int j = 0; j < 8; j++, src += srcpitch, dst += 8) - { - for(int i = 0; i < 4; i++) - { - dst[i * 2 + 0] = (src[i] & 0x0f) << 24; - dst[i * 2 + 1] = (src[i] & 0xf0) << 20; - } - } - - #endif } - static void UnpackBlock4HH(const BYTE* RESTRICT src, int srcpitch, DWORD* RESTRICT dst) + static void UnpackBlock4HH(const uint8* RESTRICT src, int srcpitch, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i zero = GSVector4i::zero(); GSVector4i mask(0xf0f0f0f0); for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = (v << 4) & mask; GSVector4i hi = v & mask; @@ -1192,25 +821,10 @@ public: ((GSVector4i*)dst)[i * 8 + 6] = zero.upl16(v5); ((GSVector4i*)dst)[i * 8 + 7] = zero.uph16(v5); } - - #else - - for(int j = 0; j < 8; j++, src += srcpitch, dst += 8) - { - for(int i = 0; i < 4; i++) - { - dst[i * 2 + 0] = (src[i] & 0x0f) << 28; - dst[i * 2 + 1] = (src[i] & 0xf0) << 24; - } - } - - #endif } - template static void ExpandBlock24(const DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) + template static void ExpandBlock24(const uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i TA0(TEXA.TA0 << 24); @@ -1241,35 +855,10 @@ public: d1[1] = v3 | TA0; } } - - #else - - DWORD TA0 = TEXA.TA0 << 24; - - for(int j = 0; j < 8; j++, src += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - DWORD c = src[i] & 0xffffff; - - if(AEM) - { - ((DWORD*)dst)[i] = c | (c ? TA0 : 0); - } - else - { - ((DWORD*)dst)[i] = c | TA0; - } - } - } - - #endif } - static void ExpandBlock16(const WORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) // do not inline, uses too many xmm regs + static void ExpandBlock16(const uint16* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) // do not inline, uses too many xmm regs { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i TA0(TEXA.TA0 << 24); @@ -1321,37 +910,9 @@ public: ((GSVector4i*)dst)[3] = ((h & rm) << 3) | ((h & gm) << 6) | ((h & bm) << 9) | TA0.blend(TA1, h.sra16(15)); } } - - #else - - DWORD TA0 = TEXA.TA0 << 24; - DWORD TA1 = TEXA.TA1 << 24; - - if(TEXA.AEM) - { - for(int j = 0; j < 8; j++, src += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - ((DWORD*)dst)[i] = ((src[i] & 0x8000) ? TA1 : src[i] ? TA0 : 0) | ((src[i] & 0x7c00) << 9) | ((src[i] & 0x03e0) << 6) | ((src[i] & 0x001f) << 3); - } - } - } - else - { - for(int j = 0; j < 8; j++, src += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - ((DWORD*)dst)[i] = ((src[i] & 0x8000) ? TA1 : TA0) | ((src[i] & 0x7c00) << 9) | ((src[i] & 0x03e0) << 6) | ((src[i] & 0x001f) << 3); - } - } - } - - #endif } - __forceinline static void ExpandBlock8_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock8_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -1359,7 +920,7 @@ public: } } - __forceinline static void ExpandBlock8_16(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock8_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -1367,7 +928,7 @@ public: } } - __forceinline static void ExpandBlock4_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -1375,7 +936,7 @@ public: } } - __forceinline static void ExpandBlock4_16(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ExpandBlock4_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -1383,7 +944,7 @@ public: } } - __forceinline static void ExpandBlock8H_32(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock8H_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1394,7 +955,7 @@ public: } } - __forceinline static void ExpandBlock8H_16(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock8H_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1411,14 +972,14 @@ public: for(int i = 0; i < 8; i++) { - ((WORD*)dst)[i] = (WORD)pal[src[j * 8 + i] >> 24]; + ((uint16*)dst)[i] = (uint16)pal[src[j * 8 + i] >> 24]; } #endif } } - __forceinline static void ExpandBlock4HL_32(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock4HL_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1429,7 +990,7 @@ public: } } - __forceinline static void ExpandBlock4HL_16(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock4HL_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1446,14 +1007,14 @@ public: for(int i = 0; i < 8; i++) { - ((WORD*)dst)[i] = (WORD)pal[(src[j * 8 + i] >> 24) & 0xf]; + ((uint16*)dst)[i] = (uint16)pal[(src[j * 8 + i] >> 24) & 0xf]; } #endif } } - __forceinline static void ExpandBlock4HH_32(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock4HH_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1464,7 +1025,7 @@ public: } } - __forceinline static void ExpandBlock4HH_16(DWORD* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ExpandBlock4HH_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { for(int j = 0; j < 8; j++, dst += dstpitch) { @@ -1481,17 +1042,15 @@ public: for(int i = 0; i < 8; i++) { - ((WORD*)dst)[i] = (WORD)pal[src[j * 8 + i] >> 28]; + ((uint16*)dst)[i] = (uint16)pal[src[j * 8 + i] >> 28]; } #endif } } - __forceinline static void UnpackAndWriteBlock24(const BYTE* RESTRICT src, int srcpitch, BYTE* RESTRICT dst) + __forceinline static void UnpackAndWriteBlock24(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i mask(0x00ffffff); for(int i = 0; i < 4; i++, src += srcpitch * 2) @@ -1534,23 +1093,9 @@ public: #endif } - - #else - - const BYTE* d = &columnTable32[0][0]; - - for(int j = 0, diff = srcpitch - 8 * 3; j < 8; j++, src += diff, d += 8) - { - for(int i = 0; i < 8; i++, src += 3) - { - ((DWORD*)dst)[d[i]] = (((DWORD*)dst)[d[i]] & ~0x00ffffff) | (src[2] << 16) | (src[1] << 8) | src[0]; - } - } - - #endif } - __forceinline static void UnpackAndWriteBlock8H(const BYTE* RESTRICT src, int srcpitch, BYTE* RESTRICT dst) + __forceinline static void UnpackAndWriteBlock8H(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) { #if _M_SSE >= 0x301 @@ -1576,7 +1121,7 @@ public: ((GSVector4i*)dst)[i * 4 + 3] = ((GSVector4i*)dst)[i * 4 + 3].blend8(v3, mask); } - #elif _M_SSE >= 0x200 + #else GSVector4i mask(0xff000000); @@ -1600,22 +1145,10 @@ public: ((GSVector4i*)dst)[i * 4 + 3] = ((GSVector4i*)dst)[i * 4 + 3].blend8(v3, mask); } - #else - - const BYTE* d = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, src += srcpitch, dst += 8) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[d[i]] = (((DWORD*)dst)[d[i]] & ~0xff000000) | (src[i] << 24); - } - } - #endif } - __forceinline static void UnpackAndWriteBlock4HL(const BYTE* RESTRICT src, int srcpitch, BYTE* RESTRICT dst) + __forceinline static void UnpackAndWriteBlock4HL(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) { #if _M_SSE >= 0x301 @@ -1629,10 +1162,10 @@ public: for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = v & mask; GSVector4i hi = (v >> 4) & mask; @@ -1666,13 +1199,13 @@ public: } } - #elif _M_SSE >= 0x200 + #else /* - __declspec(align(16)) DWORD block[8 * 8]; + __declspec(align(16)) uint32 block[8 * 8]; UnpackBlock4HL(src, srcpitch, block); - WriteBlock32(dst, (BYTE*)block, sizeof(block) / 8); + WriteBlock32(dst, (uint8*)block, sizeof(block) / 8); */ GSVector4i mask(0x0f0f0f0f); GSVector4i mask2(0x0f000000); @@ -1680,10 +1213,10 @@ public: for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = v & mask; GSVector4i hi = (v >> 4) & mask; @@ -1727,23 +1260,10 @@ public: } } - #else - - const BYTE* d = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, d += 8, src += srcpitch) - { - for(int i = 0; i < 4; i++) - { - ((DWORD*)dst)[d[i * 2 + 0]] = (((DWORD*)dst)[d[i * 2 + 0]] & ~0x0f000000) | ((src[i] & 0x0f) << 24); - ((DWORD*)dst)[d[i * 2 + 1]] = (((DWORD*)dst)[d[i * 2 + 1]] & ~0x0f000000) | ((src[i] & 0xf0) << 20); - } - } - #endif } - __forceinline static void UnpackAndWriteBlock4HH(const BYTE* RESTRICT src, int srcpitch, BYTE* RESTRICT dst) + __forceinline static void UnpackAndWriteBlock4HH(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) { #if _M_SSE >= 0x301 @@ -1757,10 +1277,10 @@ public: for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = (v << 4) & mask; GSVector4i hi = v & mask; @@ -1794,13 +1314,13 @@ public: } } - #elif _M_SSE >= 0x200 + #else /* - __declspec(align(16)) DWORD block[8 * 8]; + __declspec(align(16)) uint32 block[8 * 8]; UnpackBlock4HH(src, srcpitch, block); - WriteBlock32(dst, (BYTE*)block, sizeof(block) / 8); + WriteBlock32(dst, (uint8*)block, sizeof(block) / 8); */ GSVector4i mask(0xf0f0f0f0); GSVector4i mask2(0xf0000000); @@ -1808,10 +1328,10 @@ public: for(int i = 0; i < 2; i++, src += srcpitch * 4) { GSVector4i v( - *(DWORD*)&src[srcpitch * 0], - *(DWORD*)&src[srcpitch * 1], - *(DWORD*)&src[srcpitch * 2], - *(DWORD*)&src[srcpitch * 3]); + *(uint32*)&src[srcpitch * 0], + *(uint32*)&src[srcpitch * 1], + *(uint32*)&src[srcpitch * 2], + *(uint32*)&src[srcpitch * 3]); GSVector4i lo = (v << 4) & mask; GSVector4i hi = v & mask; @@ -1855,26 +1375,11 @@ public: } } - #else - - const BYTE* d = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, d += 8, src += srcpitch) - { - for(int i = 0; i < 4; i++) - { - ((DWORD*)dst)[d[i * 2 + 0]] = (((DWORD*)dst)[d[i * 2 + 0]] & ~0xf0000000) | ((src[i] & 0x0f) << 28); - ((DWORD*)dst)[d[i * 2 + 1]] = (((DWORD*)dst)[d[i * 2 + 1]] & ~0xf0000000) | ((src[i] & 0xf0) << 24); - } - } - #endif } - template __forceinline static void ReadAndExpandBlock24(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) + template __forceinline static void ReadAndExpandBlock24(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) { - #if _M_SSE >= 0x200 - const GSVector4i* s = (const GSVector4i*)src; GSVector4i TA0(TEXA.TA0 << 24); @@ -1912,34 +1417,9 @@ public: d1[1] = v3 | TA0; } } - - #else - - DWORD TA0 = TEXA.TA0 << 24; - - const BYTE* s = &columnTable32[0][0]; - - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - DWORD c = ((DWORD*)src)[s[i]] & 0xffffff; - - if(AEM) - { - ((DWORD*)dst)[i] = c | (c ? TA0 : 0); - } - else - { - ((DWORD*)dst)[i] = c | TA0; - } - } - } - - #endif } - __forceinline static void ReadAndExpandBlock8_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ReadAndExpandBlock8_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { #if _M_SSE >= 0x401 @@ -1985,32 +1465,20 @@ public: dst += dstpitch; } - #elif _M_SSE >= 0x200 - - __declspec(align(16)) BYTE block[16 * 16]; - - ReadBlock8(src, (BYTE*)block, sizeof(block) / 16); - - ExpandBlock8_32(block, dst, dstpitch, pal); - #else - const BYTE* s = &columnTable8[0][0]; + __declspec(align(16)) uint8 block[16 * 16]; - for(int j = 0; j < 16; j++, s += 16, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - ((DWORD*)dst)[i] = pal[src[s[i]]]; - } - } + ReadBlock8(src, (uint8*)block, sizeof(block) / 16); + + ExpandBlock8_32(block, dst, dstpitch, pal); #endif } // TODO: ReadAndExpandBlock8_16 - __forceinline static void ReadAndExpandBlock4_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ReadAndExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) { #if _M_SSE >= 0x401 @@ -2072,38 +1540,20 @@ public: dst += dstpitch; } - #elif _M_SSE >= 0x200 - - __declspec(align(16)) BYTE block[(32 / 2) * 16]; - - ReadBlock4(src, (BYTE*)block, sizeof(block) / 16); - - ExpandBlock4_32(block, dst, dstpitch, pal); - #else - const WORD* s = &columnTable4[0][0]; + __declspec(align(16)) uint8 block[(32 / 2) * 16]; - for(int j = 0; j < 16; j++, s += 32, dst += dstpitch) - { - for(int i = 0; i < 16; i++) - { - BYTE a0 = s[i * 2 + 0]; - BYTE a1 = s[i * 2 + 1]; + ReadBlock4(src, (uint8*)block, sizeof(block) / 16); - BYTE c0 = (src[a0 >> 1] >> ((a0 & 1) << 2)) & 0x0f; - BYTE c1 = (src[a1 >> 1] >> ((a1 & 1) << 2)) & 0x0f; - - ((UINT64*)dst)[i] = pal[(c1 << 4) | c0]; - } - } + ExpandBlock4_32(block, dst, dstpitch, pal); #endif } // TODO: ReadAndExpandBlock4_16 - __forceinline static void ReadAndExpandBlock8H_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ReadAndExpandBlock8H_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { #if _M_SSE >= 0x401 @@ -2131,32 +1581,20 @@ public: dst += dstpitch; } - #elif _M_SSE >= 0x200 - - __declspec(align(16)) DWORD block[8 * 8]; - - ReadBlock32(src, (BYTE*)block, sizeof(block) / 8); - - ExpandBlock8H_32(block, dst, dstpitch, pal); - #else - const BYTE* s = &columnTable32[0][0]; + __declspec(align(16)) uint32 block[8 * 8]; - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[i] = pal[((DWORD*)src)[s[i]] >> 24]; - } - } + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); + + ExpandBlock8H_32(block, dst, dstpitch, pal); #endif } // TODO: ReadAndExpandBlock8H_16 - __forceinline static void ReadAndExpandBlock4HL_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ReadAndExpandBlock4HL_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { #if _M_SSE >= 0x401 @@ -2184,32 +1622,20 @@ public: dst += dstpitch; } - #elif _M_SSE >= 0x200 - - __declspec(align(16)) DWORD block[8 * 8]; - - ReadBlock32(src, (BYTE*)block, sizeof(block) / 8); - - ExpandBlock4HL_32(block, dst, dstpitch, pal); - #else - const BYTE* s = &columnTable32[0][0]; + __declspec(align(16)) uint32 block[8 * 8]; - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[i] = pal[(((DWORD*)src)[s[i]] >> 24) & 0xf]; - } - } + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); + + ExpandBlock4HL_32(block, dst, dstpitch, pal); #endif } // TODO: ReadAndExpandBlock4HL_16 - __forceinline static void ReadAndExpandBlock4HH_32(const BYTE* RESTRICT src, BYTE* RESTRICT dst, int dstpitch, const DWORD* RESTRICT pal) + __forceinline static void ReadAndExpandBlock4HH_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) { #if _M_SSE >= 0x401 @@ -2237,25 +1663,13 @@ public: dst += dstpitch; } - #elif _M_SSE >= 0x200 - - __declspec(align(16)) DWORD block[8 * 8]; - - ReadBlock32(src, (BYTE*)block, sizeof(block) / 8); - - ExpandBlock4HH_32(block, dst, dstpitch, pal); - #else - const BYTE* s = &columnTable32[0][0]; + __declspec(align(16)) uint32 block[8 * 8]; - for(int j = 0; j < 8; j++, s += 8, dst += dstpitch) - { - for(int i = 0; i < 8; i++) - { - ((DWORD*)dst)[i] = pal[((DWORD*)src)[s[i]] >> 28]; - } - } + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); + + ExpandBlock4HH_32(block, dst, dstpitch, pal); #endif } diff --git a/plugins/GSdx/GSCapture.cpp b/plugins/GSdx/GSCapture.cpp index 069e71841b..7fca7cc0b8 100644 --- a/plugins/GSdx/GSCapture.cpp +++ b/plugins/GSdx/GSCapture.cpp @@ -46,24 +46,24 @@ class __declspec(uuid("F8BB6F4F-0965-4ED4-BA74-C6A01E6E6C77")) #endif GSSource : public CBaseFilter, private CCritSec, public IGSSource { - CSize m_size; + GSVector2i m_size; REFERENCE_TIME m_atpf; REFERENCE_TIME m_now; STDMETHODIMP NonDelegatingQueryInterface(REFIID riid, void** ppv) { return - QI(IGSSource) + riid == __uuidof(IGSSource) ? GetInterface((IGSSource*)this, ppv) : __super::NonDelegatingQueryInterface(riid, ppv); } class GSSourceOutputPin : public CBaseOutputPin { - CSize m_size; + GSVector2i m_size; vector m_mts; public: - GSSourceOutputPin(CSize size, REFERENCE_TIME atpf, CBaseFilter* pFilter, CCritSec* pLock, HRESULT& hr) + GSSourceOutputPin(const GSVector2i& size, REFERENCE_TIME atpf, CBaseFilter* pFilter, CCritSec* pLock, HRESULT& hr) : CBaseOutputPin("GSSourceOutputPin", pFilter, pLock, &hr, L"Output") , m_size(size) { @@ -75,36 +75,32 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource memset(&vih, 0, sizeof(vih)); vih.AvgTimePerFrame = atpf; vih.bmiHeader.biSize = sizeof(vih.bmiHeader); - vih.bmiHeader.biWidth = m_size.cx; - vih.bmiHeader.biHeight = m_size.cy; - - #if _M_SSE >= 0x200 + vih.bmiHeader.biWidth = m_size.x; + vih.bmiHeader.biHeight = m_size.y; // YUY2 mt.subtype = MEDIASUBTYPE_YUY2; - mt.lSampleSize = m_size.cx * m_size.cy * 2; + mt.lSampleSize = m_size.x * m_size.y * 2; vih.bmiHeader.biCompression = '2YUY'; vih.bmiHeader.biPlanes = 1; vih.bmiHeader.biBitCount = 16; - vih.bmiHeader.biSizeImage = m_size.cx * m_size.cy * 2; - mt.SetFormat((BYTE*)&vih, sizeof(vih)); + vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 2; + mt.SetFormat((uint8*)&vih, sizeof(vih)); m_mts.push_back(mt); - #endif - // RGB32 mt.subtype = MEDIASUBTYPE_RGB32; - mt.lSampleSize = m_size.cx * m_size.cy * 4; + mt.lSampleSize = m_size.x * m_size.y * 4; vih.bmiHeader.biCompression = BI_RGB; vih.bmiHeader.biPlanes = 1; vih.bmiHeader.biBitCount = 32; - vih.bmiHeader.biSizeImage = m_size.cx * m_size.cy * 4; - mt.SetFormat((BYTE*)&vih, sizeof(vih)); + vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 4; + mt.SetFormat((uint8*)&vih, sizeof(vih)); m_mts.push_back(mt); } @@ -231,17 +227,15 @@ public: const CMediaType& mt = m_output->CurrentMediaType(); - BYTE* src = (BYTE*)bits; + uint8* src = (uint8*)bits; + uint8* dst = NULL; - BYTE* dst = NULL; sample->GetPointer(&dst); - int w = m_size.cx; - int h = m_size.cy; + int w = m_size.x; + int h = m_size.y; int srcpitch = pitch; - #if _M_SSE >= 0x200 - if(mt.subtype == MEDIASUBTYPE_YUY2) { int dstpitch = ((VIDEOINFOHEADER*)mt.Format())->bmiHeader.biWidth * 2; @@ -255,8 +249,8 @@ public: { for(int j = 0; j < h; j++, dst += dstpitch, src += srcpitch) { - DWORD* s = (DWORD*)src; - WORD* d = (WORD*)dst; + uint32* s = (uint32*)src; + uint16* d = (uint16*)dst; for(int i = 0; i < w; i += 2) { @@ -269,7 +263,7 @@ public: GSVector4 c = lo.hadd(hi) + offset; - *((DWORD*)&d[i]) = GSVector4i(c).rgba32(); + *((uint32*)&d[i]) = GSVector4i(c).rgba32(); } } } @@ -277,8 +271,8 @@ public: { for(int j = 0; j < h; j++, dst += dstpitch, src += srcpitch) { - DWORD* s = (DWORD*)src; - WORD* d = (WORD*)dst; + uint32* s = (uint32*)src; + uint16* d = (uint16*)dst; for(int i = 0; i < w; i += 2) { @@ -291,16 +285,12 @@ public: GSVector4 c = lo.hadd(hi) + offset; - *((DWORD*)&d[i]) = GSVector4i(c).rgba32(); + *((uint32*)&d[i]) = GSVector4i(c).rgba32(); } } } } - else - - #endif - - if(mt.subtype == MEDIASUBTYPE_RGB32) + else if(mt.subtype == MEDIASUBTYPE_RGB32) { int dstpitch = ((VIDEOINFOHEADER*)mt.Format())->bmiHeader.biWidth * 4; @@ -323,7 +313,7 @@ public: d[i] = s[i].shuffle8(mask); } - #elif _M_SSE >= 0x200 + #else GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)dst; @@ -333,16 +323,6 @@ public: d[i] = ((s[i] & 0x00ff0000) >> 16) | ((s[i] & 0x000000ff) << 16) | (s[i] & 0x0000ff00); } - #else - - DWORD* s = (DWORD*)src; - DWORD* d = (DWORD*)dst; - - for(int i = 0; i < w; i++) - { - d[i] = ((s[i] & 0x00ff0000) >> 16) | ((s[i] & 0x000000ff) << 16) | (s[i] & 0x0000ff00); - } - #endif } else @@ -429,8 +409,8 @@ bool GSCapture::BeginCapture(int fps) if(IDOK != dlg.DoModal()) return false; - m_size.cx = (dlg.m_width + 7) & ~7; - m_size.cy = (dlg.m_height + 7) & ~7; + m_size.x = (dlg.m_width + 7) & ~7; + m_size.y = (dlg.m_height + 7) & ~7; // @@ -447,7 +427,7 @@ bool GSCapture::BeginCapture(int fps) return false; } - m_src = new GSSource(m_size.cx, m_size.cy, fps, NULL, hr); + m_src = new GSSource(m_size.x, m_size.y, fps, NULL, hr); if(FAILED(hr = m_graph->AddFilter(m_src, L"Source")) || FAILED(hr = m_graph->AddFilter(dlg.m_enc, L"Encoder"))) diff --git a/plugins/GSdx/GSCapture.h b/plugins/GSdx/GSCapture.h index 74a1808160..c6b71a2f12 100644 --- a/plugins/GSdx/GSCapture.h +++ b/plugins/GSdx/GSCapture.h @@ -22,11 +22,12 @@ #pragma once #include "GSCaptureDlg.h" +#include "GSVector.h" class GSCapture : protected CCritSec { bool m_capturing; - CSize m_size; + GSVector2i m_size; CComPtr m_graph; CComPtr m_src; @@ -39,5 +40,5 @@ public: bool EndCapture(); bool IsCapturing() {return m_capturing;} - CSize GetSize() {return m_size;} + GSVector2i GetSize() {return m_size;} }; diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index ed8fad82b2..879764e108 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -26,11 +26,11 @@ GSClut::GSClut(const GSLocalMemory* mem) : m_mem(mem) { - BYTE* p = (BYTE*)VirtualAlloc(NULL, 2 * 4096, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); + uint8* p = (uint8*)VirtualAlloc(NULL, 2 * 4096, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); - m_clut = (WORD*)&p[0]; // 1k + 1k for buffer overruns (sfex: PSM == PSM_PSMT8, CPSM == PSM_PSMCT32, CSA != 0) - m_buff32 = (DWORD*)&p[2048]; // 1k - m_buff64 = (UINT64*)&p[4096]; // 2k + m_clut = (uint16*)&p[0]; // 1k + 1k for buffer overruns (sfex: PSM == PSM_PSMT8, CPSM == PSM_PSMCT32, CSA != 0) + m_buff32 = (uint32*)&p[2048]; // 1k + m_buff64 = (uint64*)&p[4096]; // 2k m_write.dirty = true; m_read.dirty = true; @@ -128,7 +128,7 @@ void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TE { ASSERT(TEX0.CSA == 0); - WriteCLUT_T32_I8_CSM1((DWORD*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T32_I8_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) @@ -137,72 +137,72 @@ void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TE GSVector4i dummy; // this just forces stack alignment and enables inlining the next call - WriteCLUT_T32_I4_CSM1((DWORD*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T32_I4_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { ASSERT(TEX0.CSA < 16); - WriteCLUT_T16_I8_CSM1((WORD*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I8_CSM1((uint16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { ASSERT(TEX0.CSA < 32); - WriteCLUT_T16_I4_CSM1((WORD*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I4_CSM1((uint16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16S_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WriteCLUT_T16_I8_CSM1((WORD*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I8_CSM1((uint16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } void GSClut::WriteCLUT16S_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WriteCLUT_T16_I4_CSM1((WORD*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); + WriteCLUT_T16_I4_CSM1((uint16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } template void GSClut::WriteCLUT32_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WORD* RESTRICT clut = m_clut + (TEX0.CSA << 4); + uint16* RESTRICT clut = m_clut + (TEX0.CSA << 4); - DWORD base = m_mem->PixelAddress32(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); + uint32 base = m_mem->PixelAddress32(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); int* offset = &m_mem->rowOffset32[TEXCLUT.COU << 4]; for(int i = 0; i < n; i++) { - DWORD c = m_mem->ReadPixel32(base + offset[i]); + uint32 c = m_mem->ReadPixel32(base + offset[i]); - clut[i] = (WORD)(c & 0xffff); - clut[i + 256] = (WORD)(c >> 16); + clut[i] = (uint16)(c & 0xffff); + clut[i + 256] = (uint16)(c >> 16); } } template void GSClut::WriteCLUT16_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WORD* RESTRICT clut = m_clut + (TEX0.CSA << 4); + uint16* RESTRICT clut = m_clut + (TEX0.CSA << 4); - DWORD base = m_mem->PixelAddress16(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); + uint32 base = m_mem->PixelAddress16(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); int* offset = &m_mem->rowOffset16[TEXCLUT.COU << 4]; for(int i = 0; i < n; i++) { - clut[i] = (WORD)m_mem->ReadPixel16(base + offset[i]); + clut[i] = (uint16)m_mem->ReadPixel16(base + offset[i]); } } template void GSClut::WriteCLUT16S_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - WORD* RESTRICT clut = m_clut + (TEX0.CSA << 4); + uint16* RESTRICT clut = m_clut + (TEX0.CSA << 4); - DWORD base = m_mem->PixelAddress16S(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); + uint32 base = m_mem->PixelAddress16S(0, TEXCLUT.COV, TEX0.CBP, TEXCLUT.CBW); int* offset = &m_mem->rowOffset16S[TEXCLUT.COU << 4]; for(int i = 0; i < n; i++) { - clut[i] = (WORD)m_mem->ReadPixel16(base + offset[i]); + clut[i] = (uint16)m_mem->ReadPixel16(base + offset[i]); } } @@ -213,7 +213,7 @@ void GSClut::Read(const GIFRegTEX0& TEX0) m_read.TEX0 = TEX0; m_read.dirty = false; - WORD* clut = m_clut + (TEX0.CSA << 4); + uint16* clut = m_clut + (TEX0.CSA << 4); if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -257,7 +257,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) m_read.dirty = false; m_read.adirty = true; - WORD* clut = m_clut + (TEX0.CSA << 4); + uint16* clut = m_clut + (TEX0.CSA << 4); if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -272,7 +272,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) case PSM_PSMT4HH: // TODO: merge these functions ReadCLUT_T32_I4(clut, m_buff32); - ExpandCLUT64_T32_I8(m_buff32, (UINT64*)m_buff64); // sw renderer does not need m_buff64 anymore + ExpandCLUT64_T32_I8(m_buff32, (uint64*)m_buff64); // sw renderer does not need m_buff64 anymore break; } } @@ -289,7 +289,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) case PSM_PSMT4HH: // TODO: merge these functions Expand16(clut, m_buff32, 16, TEXA); - ExpandCLUT64_T32_I8(m_buff32, (UINT64*)m_buff64); // sw renderer does not need m_buff64 anymore + ExpandCLUT64_T32_I8(m_buff32, (uint64*)m_buff64); // sw renderer does not need m_buff64 anymore break; } } @@ -306,9 +306,9 @@ void GSClut::GetAlphaMinMax32(int& amin, int& amax) { m_read.adirty = false; - // DWORD bpp = GSLocalMemory::m_psm[m_read.TEX0.PSM].trbpp; - DWORD cbpp = GSLocalMemory::m_psm[m_read.TEX0.CPSM].trbpp; - DWORD pal = GSLocalMemory::m_psm[m_read.TEX0.PSM].pal; + // uint32 bpp = GSLocalMemory::m_psm[m_read.TEX0.PSM].trbpp; + uint32 cbpp = GSLocalMemory::m_psm[m_read.TEX0.CPSM].trbpp; + uint32 pal = GSLocalMemory::m_psm[m_read.TEX0.PSM].pal; if(cbpp == 24 && m_read.TEXA.AEM == 0) { @@ -352,10 +352,8 @@ void GSClut::GetAlphaMinMax32(int& amin, int& amax) // -void GSClut::WriteCLUT_T32_I8_CSM1(const DWORD* RESTRICT src, WORD* RESTRICT clut) +void GSClut::WriteCLUT_T32_I8_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut) { - #if _M_SSE >= 0x200 - for(int i = 0; i < 64; i += 16) { WriteCLUT_T32_I4_CSM1(&src[i + 0], &clut[i * 2 + 0]); @@ -363,26 +361,10 @@ void GSClut::WriteCLUT_T32_I8_CSM1(const DWORD* RESTRICT src, WORD* RESTRICT clu WriteCLUT_T32_I4_CSM1(&src[i + 128], &clut[i * 2 + 128]); WriteCLUT_T32_I4_CSM1(&src[i + 192], &clut[i * 2 + 144]); } - - #else - - for(int j = 0; j < 2; j++, src += 128, clut += 128) - { - for(int i = 0; i < 128; i++) - { - DWORD c = src[clutTableT32I8[i]]; - clut[i] = (WORD)(c & 0xffff); - clut[i + 256] = (WORD)(c >> 16); - } - } - - #endif } -__forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const DWORD* RESTRICT src, WORD* RESTRICT clut) +__forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)clut; @@ -400,23 +382,10 @@ __forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const DWORD* RESTRICT src, WORD d[1] = v1; d[32] = v2; d[33] = v3; - - #else - - for(int i = 0; i < 16; i++) - { - DWORD c = src[clutTableT32I4[i]]; - clut[i] = (WORD)(c & 0xffff); - clut[i + 256] = (WORD)(c >> 16); - } - - #endif } -void GSClut::WriteCLUT_T16_I8_CSM1(const WORD* RESTRICT src, WORD* RESTRICT clut) +void GSClut::WriteCLUT_T16_I8_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)clut; @@ -436,21 +405,9 @@ void GSClut::WriteCLUT_T16_I8_CSM1(const WORD* RESTRICT src, WORD* RESTRICT clut d[i + 2] = v1; d[i + 3] = v3; } - - #else - - for(int j = 0; j < 8; j++, src += 32, clut += 32) - { - for(int i = 0; i < 32; i++) - { - clut[i] = src[clutTableT16I8[i]]; - } - } - - #endif } -__forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const WORD* RESTRICT src, WORD* RESTRICT clut) +__forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut) { for(int i = 0; i < 16; i++) { @@ -458,29 +415,16 @@ __forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const WORD* RESTRICT src, WORD* } } -void GSClut::ReadCLUT_T32_I8(const WORD* RESTRICT clut, DWORD* RESTRICT dst) +void GSClut::ReadCLUT_T32_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - for(int i = 0; i < 256; i += 16) { ReadCLUT_T32_I4(&clut[i], &dst[i]); } - - #else - - for(int i = 0; i < 256; i++) - { - dst[i] = ((DWORD)clut[i + 256] << 16) | clut[i]; - } - - #endif } -__forceinline void GSClut::ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst) +__forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)clut; GSVector4i* d = (GSVector4i*)dst; @@ -495,21 +439,10 @@ __forceinline void GSClut::ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RES d[1] = v1; d[2] = v2; d[3] = v3; - - #else - - for(int i = 0; i < 16; i++) - { - dst[i] = ((DWORD)clut[i + 256] << 16) | clut[i]; - } - - #endif } -__forceinline void GSClut::ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst32, UINT64* RESTRICT dst64) +__forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)clut; GSVector4i* d32 = (GSVector4i*)dst32; GSVector4i* d64 = (GSVector4i*)dst64; @@ -530,53 +463,18 @@ __forceinline void GSClut::ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RES ExpandCLUT64_T32(s1, s0, s1, s2, s3, &d64[32]); ExpandCLUT64_T32(s2, s0, s1, s2, s3, &d64[64]); ExpandCLUT64_T32(s3, s0, s1, s2, s3, &d64[96]); - - #else - - for(int i = 0; i < 16; i++) - { - dst[i] = ((DWORD)clut[i + 256] << 16) | clut[i]; - } - - DWORD* d = (DWORD*)dst64; - - for(int j = 0; j < 16; j++, d += 32) - { - DWORD hi = dst32[j]; - - for(int i = 0; i < 16; i++) - { - d[i * 2 + 0] = dst32[i]; - d[i * 2 + 1] = hi; - } - } - - #endif } -void GSClut::ReadCLUT_T16_I8(const WORD* RESTRICT clut, DWORD* RESTRICT dst) +void GSClut::ReadCLUT_T16_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - for(int i = 0; i < 256; i += 16) { ReadCLUT_T16_I4(&clut[i], &dst[i]); } - - #else - - for(int i = 0; i < 256; i++) - { - dst[i] = (DWORD)clut[i]; - } - - #endif } -__forceinline void GSClut::ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst) +__forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)clut; GSVector4i* d = (GSVector4i*)dst; @@ -587,21 +485,10 @@ __forceinline void GSClut::ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RES d[1] = v0.uph16(); d[2] = v1.upl16(); d[3] = v1.uph16(); - - #else - - for(int i = 0; i < 16; i++) - { - dst[i] = (DWORD)clut[i]; - } - - #endif } -__forceinline void GSClut::ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst32, UINT64* RESTRICT dst64) +__forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)clut; GSVector4i* d32 = (GSVector4i*)dst32; GSVector4i* d64 = (GSVector4i*)dst64; @@ -623,33 +510,10 @@ __forceinline void GSClut::ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RES ExpandCLUT64_T16(s1, s0, s1, s2, s3, &d64[32]); ExpandCLUT64_T16(s2, s0, s1, s2, s3, &d64[64]); ExpandCLUT64_T16(s3, s0, s1, s2, s3, &d64[96]); - - #else - - for(int i = 0; i < 16; i++) - { - dst32[i] = (DWORD)clut[i]; - } - - DWORD* d = (DWORD*)dst64; - - for(int j = 0; j < 16; j++, d += 32) - { - DWORD hi = dst32[j] << 16; - - for(int i = 0; i < 16; i++) - { - d[i * 2 + 0] = hi | (dst32[i] & 0xffff); - } - } - - #endif } -void GSClut::ExpandCLUT64_T32_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst) +void GSClut::ExpandCLUT64_T32_I8(const uint32* RESTRICT src, uint64* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)dst; @@ -662,23 +526,6 @@ void GSClut::ExpandCLUT64_T32_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst ExpandCLUT64_T32(s1, s0, s1, s2, s3, &d[32]); ExpandCLUT64_T32(s2, s0, s1, s2, s3, &d[64]); ExpandCLUT64_T32(s3, s0, s1, s2, s3, &d[96]); - - #else - - DWORD* d = (DWORD*)dst; - - for(int j = 0; j < 16; j++, d += 32) - { - DWORD hi = src[j]; - - for(int i = 0; i < 16; i++) - { - d[i * 2 + 0] = src[i]; - d[i * 2 + 1] = hi; - } - } - - #endif } __forceinline void GSClut::ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst) @@ -707,10 +554,8 @@ __forceinline void GSClut::ExpandCLUT64_T32(const GSVector4i& hi, const GSVector dst[1] = lo.uph32(hi); } -void GSClut::ExpandCLUT64_T16_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst) +void GSClut::ExpandCLUT64_T16_I8(const uint32* RESTRICT src, uint64* RESTRICT dst) { - #if _M_SSE >= 0x200 - GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)dst; @@ -723,22 +568,6 @@ void GSClut::ExpandCLUT64_T16_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst ExpandCLUT64_T16(s1, s0, s1, s2, s3, &d[32]); ExpandCLUT64_T16(s2, s0, s1, s2, s3, &d[64]); ExpandCLUT64_T16(s3, s0, s1, s2, s3, &d[96]); - - #else - - DWORD* d = (DWORD*)dst; - - for(int j = 0; j < 16; j++, d += 32) - { - DWORD hi = src[j] << 16; - - for(int i = 0; i < 16; i++) - { - d[i * 2 + 0] = hi | (src[i] & 0xffff); - } - } - - #endif } __forceinline void GSClut::ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst) @@ -774,10 +603,8 @@ static const GSVector4i s_bm(0x00007c00); static const GSVector4i s_gm(0x000003e0); static const GSVector4i s_rm(0x0000001f); -void GSClut::Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int w, const GIFRegTEXA& TEXA) +void GSClut::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, const GIFRegTEXA& TEXA) { - #if _M_SSE >= 0x200 - ASSERT((w & 7) == 0); const GSVector4i rm = s_rm; @@ -827,43 +654,21 @@ void GSClut::Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int w, cons d[i * 2 + 1] = ((ch & rm) << 3) | ((ch & gm) << 6) | ((ch & bm) << 9) | TA0.blend8(TA1, ch.sra16(15)).andnot(ch == GSVector4i::zero()); } } - - #else - - DWORD TA0 = (DWORD)TEXA.TA0 << 24; - DWORD TA1 = (DWORD)TEXA.TA1 << 24; - - if(!TEXA.AEM) - { - for(int i = 0; i < w; i++) - { - dst[i] = ((src[i] & 0x8000) ? TA1 : TA0) | ((src[i] & 0x7c00) << 9) | ((src[i] & 0x03e0) << 6) | ((src[i] & 0x001f) << 3); - } - } - else - { - for(int i = 0; i < w; i++) - { - dst[i] = ((src[i] & 0x8000) ? TA1 : src[i] ? TA0 : 0) | ((src[i] & 0x7c00) << 9) | ((src[i] & 0x03e0) << 6) | ((src[i] & 0x001f) << 3); - } - } - - #endif } // bool GSClut::WriteState::IsDirty(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - return dirty || !(GSVector4i::load(this) == GSVector4i::load(&TEX0, &TEXCLUT)).alltrue(); + return dirty || !GSVector4i::load(this).eq(GSVector4i::load(&TEX0, &TEXCLUT)); } bool GSClut::ReadState::IsDirty(const GIFRegTEX0& TEX0) { - return dirty || !(GSVector4i::load(this) == GSVector4i::load(&TEX0, &this->TEXA)).alltrue(); + return dirty || !GSVector4i::load(this).eq(GSVector4i::load(&TEX0, &this->TEXA)); } bool GSClut::ReadState::IsDirty(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) { - return dirty || !(GSVector4i::load(this) == GSVector4i::load(&TEX0, &TEXA)).alltrue(); + return dirty || !GSVector4i::load(this).eq(GSVector4i::load(&TEX0, &TEXA)); } diff --git a/plugins/GSdx/GSClut.h b/plugins/GSdx/GSClut.h index 9c7fa4241a..2d8333cfb4 100644 --- a/plugins/GSdx/GSClut.h +++ b/plugins/GSdx/GSClut.h @@ -32,10 +32,10 @@ __declspec(align(16)) class GSClut : public GSAlignedClass<16> { const GSLocalMemory* m_mem; - DWORD m_CBP[2]; - WORD* m_clut; - DWORD* m_buff32; - UINT64* m_buff64; + uint32 m_CBP[2]; + uint16* m_clut; + uint32* m_buff32; + uint64* m_buff64; __declspec(align(16)) struct WriteState { @@ -73,24 +73,24 @@ __declspec(align(16)) class GSClut : public GSAlignedClass<16> void WriteCLUT_NULL(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) {} - static void WriteCLUT_T32_I8_CSM1(const DWORD* RESTRICT src, WORD* RESTRICT clut); - static void WriteCLUT_T32_I4_CSM1(const DWORD* RESTRICT src, WORD* RESTRICT clut); - static void WriteCLUT_T16_I8_CSM1(const WORD* RESTRICT src, WORD* RESTRICT clut); - static void WriteCLUT_T16_I4_CSM1(const WORD* RESTRICT src, WORD* RESTRICT clut); - static void ReadCLUT_T32_I8(const WORD* RESTRICT clut, DWORD* RESTRICT dst); - static void ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst); - static void ReadCLUT_T32_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst32, UINT64* RESTRICT dst64); - static void ReadCLUT_T16_I8(const WORD* RESTRICT clut, DWORD* RESTRICT dst); - static void ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst); - static void ReadCLUT_T16_I4(const WORD* RESTRICT clut, DWORD* RESTRICT dst32, UINT64* RESTRICT dst64); - static void ExpandCLUT64_T32_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst); + static void WriteCLUT_T32_I8_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut); + static void WriteCLUT_T32_I4_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut); + static void WriteCLUT_T16_I8_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut); + static void WriteCLUT_T16_I4_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut); + static void ReadCLUT_T32_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst); + static void ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst); + static void ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64); + static void ReadCLUT_T16_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst); + static void ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst); + static void ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64); + static void ExpandCLUT64_T32_I8(const uint32* RESTRICT src, uint64* RESTRICT dst); static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst); static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst); - static void ExpandCLUT64_T16_I8(const DWORD* RESTRICT src, UINT64* RESTRICT dst); + static void ExpandCLUT64_T16_I8(const uint32* RESTRICT src, uint64* RESTRICT dst); static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst); static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst); - static void Expand16(const WORD* RESTRICT src, DWORD* RESTRICT dst, int w, const GIFRegTEXA& TEXA); + static void Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, const GIFRegTEXA& TEXA); public: GSClut(const GSLocalMemory* mem); @@ -103,8 +103,8 @@ public: void Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA); void GetAlphaMinMax32(int& amin, int& amax); - DWORD operator [] (size_t i) const {return m_buff32[i];} + uint32 operator [] (size_t i) const {return m_buff32[i];} - operator const DWORD*() const {return m_buff32;} - operator const UINT64*() const {return m_buff64;} + operator const uint32*() const {return m_buff32;} + operator const uint64*() const {return m_buff64;} }; diff --git a/plugins/GSdx/GSCodeBuffer.cpp b/plugins/GSdx/GSCodeBuffer.cpp index 7f90f53569..83b0f45f61 100644 --- a/plugins/GSdx/GSCodeBuffer.cpp +++ b/plugins/GSdx/GSCodeBuffer.cpp @@ -47,14 +47,14 @@ void* GSCodeBuffer::GetBuffer(size_t size) if(m_ptr == NULL || m_pos + size > m_blocksize) { - m_ptr = (BYTE*)VirtualAlloc(NULL, m_blocksize, MEM_COMMIT | MEM_RESERVE, PAGE_EXECUTE_READWRITE); + m_ptr = (uint8*)VirtualAlloc(NULL, m_blocksize, MEM_COMMIT | MEM_RESERVE, PAGE_EXECUTE_READWRITE); m_pos = 0; m_buffers.push_back(m_ptr); } - BYTE* ptr = &m_ptr[m_pos]; + uint8* ptr = &m_ptr[m_pos]; m_reserved = size; diff --git a/plugins/GSdx/GSCodeBuffer.h b/plugins/GSdx/GSCodeBuffer.h index 7011d00950..badaed81a8 100644 --- a/plugins/GSdx/GSCodeBuffer.h +++ b/plugins/GSdx/GSCodeBuffer.h @@ -26,7 +26,7 @@ class GSCodeBuffer list m_buffers; size_t m_blocksize; size_t m_pos, m_reserved; - BYTE* m_ptr; + uint8* m_ptr; public: GSCodeBuffer(size_t blocksize = 4096 * 64); // 256k diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index 85abd05cb5..a4a3fde2e1 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -131,9 +131,9 @@ CRC::Game CRC::m_games[] = {0xFB236A46, SonicUnleashed, US}, }; -hash_map CRC::m_map; +hash_map CRC::m_map; -CRC::Game CRC::Lookup(DWORD crc) +CRC::Game CRC::Lookup(uint32 crc) { if(m_map.empty()) { @@ -143,7 +143,7 @@ CRC::Game CRC::Lookup(DWORD crc) } } - hash_map::iterator i = m_map.find(crc); + hash_map::iterator i = m_map.find(crc); if(i != m_map.end()) { diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index c5187f3b87..731e495a09 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -84,15 +84,15 @@ public: struct Game { - DWORD crc; + uint32 crc; Title title; Region region; }; private: static Game m_games[]; - static hash_map m_map; + static hash_map m_map; public: - static Game Lookup(DWORD crc); + static Game Lookup(uint32 crc); }; diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index 01d06f42cc..bf0595a1b6 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -115,7 +115,7 @@ public: virtual bool IsLost() = 0; - virtual void Present(const CRect& r) = 0; + virtual void Present(const GSVector4i& r) = 0; virtual void BeginScene() = 0; @@ -127,11 +127,11 @@ public: virtual void ClearRenderTarget(Texture& t, const GSVector4& c) = 0; - virtual void ClearRenderTarget(Texture& t, DWORD c) = 0; + virtual void ClearRenderTarget(Texture& t, uint32 c) = 0; virtual void ClearDepth(Texture& t, float c) = 0; - virtual void ClearStencil(Texture& t, BYTE c) = 0; + virtual void ClearStencil(Texture& t, uint8 c) = 0; virtual bool CreateRenderTarget(Texture& t, int w, int h, int format = 0) { @@ -178,11 +178,11 @@ public: t = m_current; } - void Merge(Texture* st, GSVector4* sr, GSVector4* dr, CSize fs, bool slbg, bool mmod, GSVector4& c) + void Merge(Texture* st, GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, GSVector4& c) { - if(!m_merge || m_merge.GetWidth() != fs.cx || m_merge.GetHeight() != fs.cy) + if(!m_merge || m_merge.GetWidth() != fs.x || m_merge.GetHeight() != fs.y) { - CreateRenderTarget(m_merge, fs.cx, fs.cy); + CreateRenderTarget(m_merge, fs.x, fs.y); } // TODO: m_1x1 @@ -192,11 +192,11 @@ public: m_current = m_merge; } - bool Interlace(CSize ds, int field, int mode, float yoffset) + bool Interlace(const GSVector2i& ds, int field, int mode, float yoffset) { - if(!m_weavebob || m_weavebob.GetWidth() != ds.cx || m_weavebob.GetHeight() != ds.cy) + if(!m_weavebob || m_weavebob.GetWidth() != ds.x || m_weavebob.GetHeight() != ds.y) { - CreateRenderTarget(m_weavebob, ds.cx, ds.cy); + CreateRenderTarget(m_weavebob, ds.x, ds.y); } if(mode == 0 || mode == 2) // weave or blend @@ -209,9 +209,9 @@ public: { // blend - if(!m_blend || m_blend.GetWidth() != ds.cx || m_blend.GetHeight() != ds.cy) + if(!m_blend || m_blend.GetWidth() != ds.x || m_blend.GetHeight() != ds.y) { - CreateRenderTarget(m_blend, ds.cx, ds.cy); + CreateRenderTarget(m_blend, ds.x, ds.y); } DoInterlace(m_weavebob, m_blend, 2, false, 0); diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 97465aa013..72d6f65a11 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -80,7 +80,7 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) scd.SampleDesc.Quality = 0; scd.Windowed = TRUE; - UINT flags = 0; + uint32 flags = 0; #ifdef DEBUG flags |= D3D10_CREATE_DEVICE_DEBUG; @@ -256,15 +256,15 @@ bool GSDevice10::Reset(int w, int h, bool fs) return true; } -void GSDevice10::Present(const CRect& r) +void GSDevice10::Present(const GSVector4i& r) { - CRect cr; + GSVector4i cr; - GetClientRect(m_hWnd, &cr); + GetClientRect(m_hWnd, cr); - if(m_backbuffer.GetWidth() != cr.Width() || m_backbuffer.GetHeight() != cr.Height()) + if(m_backbuffer.GetWidth() != cr.width() || m_backbuffer.GetHeight() != cr.height()) { - Reset(cr.Width(), cr.Height(), false); + Reset(cr.width(), cr.height(), false); } float color[4] = {0, 0, 0, 0}; @@ -306,13 +306,13 @@ void GSDevice10::Draw(const string& s) OMSetRenderTargets(m_backbuffer, NULL); - CRect r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); + GSVector4i r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); D3DCOLOR c = D3DCOLOR_ARGB(255, 0, 255, 0); - if(m_font->DrawText(NULL, str, -1, &r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) + if(m_font->DrawText(NULL, str, -1, r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) { - m_font->DrawText(NULL, str, -1, &r, DT_LEFT|DT_WORDBREAK, c); + m_font->DrawText(NULL, str, -1, r, DT_LEFT|DT_WORDBREAK, c); } EndScene(); @@ -360,7 +360,7 @@ void GSDevice10::ClearRenderTarget(Texture& t, const GSVector4& c) m_dev->ClearRenderTargetView(t, c.v); } -void GSDevice10::ClearRenderTarget(Texture& t, DWORD c) +void GSDevice10::ClearRenderTarget(Texture& t, uint32 c) { GSVector4 color = GSVector4(c) * (1.0f / 255); @@ -372,7 +372,7 @@ void GSDevice10::ClearDepth(Texture& t, float c) m_dev->ClearDepthStencilView(t, D3D10_CLEAR_DEPTH, c, 0); } -void GSDevice10::ClearStencil(Texture& t, BYTE c) +void GSDevice10::ClearStencil(Texture& t, uint8 c) { m_dev->ClearDepthStencilView(t, D3D10_CLEAR_STENCIL, 0, c); } @@ -487,11 +487,11 @@ void GSDevice10::DoInterlace(Texture& st, Texture& dt, int shader, bool linear, StretchRect(st, sr, dt, dr, m_interlace.ps[shader], m_interlace.cb, linear); } -void GSDevice10::IASetVertexBuffer(ID3D10Buffer* vb, UINT stride) +void GSDevice10::IASetVertexBuffer(ID3D10Buffer* vb, size_t stride) { if(m_vb != vb || m_vb_stride != stride) { - UINT offset = 0; + uint32 offset = 0; m_dev->IASetVertexBuffers(0, 1, &vb, &stride, &offset); @@ -590,9 +590,9 @@ void GSDevice10::PSSetSamplerState(ID3D10SamplerState* ss0, ID3D10SamplerState* } } -void GSDevice10::RSSet(int width, int height, const RECT* scissor) +void GSDevice10::RSSet(int width, int height, const GSVector4i* scissor) { - if(m_viewport.cx != width || m_viewport.cy != height) + if(m_viewport.x != width || m_viewport.y != height) { D3D10_VIEWPORT vp; @@ -607,20 +607,20 @@ void GSDevice10::RSSet(int width, int height, const RECT* scissor) m_dev->RSSetViewports(1, &vp); - m_viewport = CSize(width, height); + m_viewport = GSVector2i(width, height); } - CRect r = scissor ? *scissor : CRect(0, 0, width, height); + GSVector4i r = scissor ? *scissor : GSVector4i(0, 0, width, height); - if(m_scissor != r) + if(!m_scissor.eq(r)) { - m_dev->RSSetScissorRects(1, &r); + m_dev->RSSetScissorRects(1, r); m_scissor = r; } } -void GSDevice10::OMSetDepthStencilState(ID3D10DepthStencilState* dss, UINT sref) +void GSDevice10::OMSetDepthStencilState(ID3D10DepthStencilState* dss, uint8 sref) { if(m_dss != dss || m_sref != sref) { @@ -655,7 +655,7 @@ void GSDevice10::OMSetRenderTargets(ID3D10RenderTargetView* rtv, ID3D10DepthSten } } -void GSDevice10::DrawPrimitive(UINT count, UINT start) +void GSDevice10::DrawPrimitive(uint32 count, uint32 start) { m_dev->Draw(count, start); } @@ -735,7 +735,7 @@ void GSDevice10::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, cons EndScene(); } -HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) +HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) { HRESULT hr; @@ -753,7 +753,7 @@ HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MAC return hr; } - hr = m_dev->CreateVertexShader((DWORD*)shader->GetBufferPointer(), shader->GetBufferSize(), ps); + hr = m_dev->CreateVertexShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), ps); if(FAILED(hr)) { @@ -770,7 +770,7 @@ HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MAC return hr; } -HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs) +HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs) { HRESULT hr; @@ -788,7 +788,7 @@ HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MAC return hr; } - hr = m_dev->CreateGeometryShader((DWORD*)shader->GetBufferPointer(), shader->GetBufferSize(), gs); + hr = m_dev->CreateGeometryShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), gs); if(FAILED(hr)) { @@ -798,7 +798,7 @@ HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MAC return hr; } -HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps) +HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps) { HRESULT hr; @@ -816,7 +816,7 @@ HRESULT GSDevice10::CompileShader(UINT id, const string& entry, D3D10_SHADER_MAC return hr; } - hr = m_dev->CreatePixelShader((DWORD*)shader->GetBufferPointer(), shader->GetBufferSize(), ps); + hr = m_dev->CreatePixelShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), ps); if(FAILED(hr)) { diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index a3ad0c1a75..81c5b7cecf 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -33,8 +33,8 @@ private: // state cache ID3D10Buffer* m_vb; - UINT m_vb_count; - UINT m_vb_stride; + int m_vb_count; + size_t m_vb_stride; ID3D10InputLayout* m_layout; D3D10_PRIMITIVE_TOPOLOGY m_topology; ID3D10VertexShader* m_vs; @@ -44,10 +44,10 @@ private: ID3D10PixelShader* m_ps; ID3D10Buffer* m_ps_cb; ID3D10SamplerState* m_ps_ss[2]; - CSize m_viewport; - CRect m_scissor; + GSVector2i m_viewport; + GSVector4i m_scissor; ID3D10DepthStencilState* m_dss; - UINT m_sref; + uint8 m_sref; ID3D10BlendState* m_bs; float m_bf; ID3D10RenderTargetView* m_rtv; @@ -100,16 +100,16 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); bool IsLost() {return false;} - void Present(const CRect& r); + void Present(const GSVector4i& r); void BeginScene(); void EndScene(); void Draw(const string& s); bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); void ClearRenderTarget(Texture& t, const GSVector4& c); - void ClearRenderTarget(Texture& t, DWORD c); + void ClearRenderTarget(Texture& t, uint32 c); void ClearDepth(Texture& t, float c); - void ClearStencil(Texture& t, BYTE c); + void ClearStencil(Texture& t, uint8 c); bool CreateRenderTarget(Texture& t, int w, int h, int format = 0); bool CreateDepthStencil(Texture& t, int w, int h, int format = 0); @@ -119,7 +119,7 @@ public: ID3D10Device* operator->() {return m_dev;} operator ID3D10Device*() {return m_dev;} - void IASetVertexBuffer(ID3D10Buffer* vb, UINT stride); + void IASetVertexBuffer(ID3D10Buffer* vb, size_t stride); void IASetInputLayout(ID3D10InputLayout* layout); void IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY topology); void VSSetShader(ID3D10VertexShader* vs, ID3D10Buffer* vs_cb); @@ -127,18 +127,18 @@ public: void PSSetShaderResources(ID3D10ShaderResourceView* srv0, ID3D10ShaderResourceView* srv1); void PSSetShader(ID3D10PixelShader* ps, ID3D10Buffer* ps_cb); void PSSetSamplerState(ID3D10SamplerState* ss0, ID3D10SamplerState* ss1); - void RSSet(int width, int height, const RECT* scissor = NULL); - void OMSetDepthStencilState(ID3D10DepthStencilState* dss, UINT sref); + void RSSet(int width, int height, const GSVector4i* scissor = NULL); + void OMSetDepthStencilState(ID3D10DepthStencilState* dss, uint8 sref); void OMSetBlendState(ID3D10BlendState* bs, float bf); void OMSetRenderTargets(ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv); - void DrawPrimitive(UINT count, UINT start = 0); + void DrawPrimitive(uint32 count, uint32 start = 0); void StretchRect(Texture& st, Texture& dt, const GSVector4& dr, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); - HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il); - HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs); - HRESULT CompileShader(UINT id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps); + HRESULT CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il); + HRESULT CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs); + HRESULT CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10PixelShader** ps); }; diff --git a/plugins/GSdx/GSDevice7.cpp b/plugins/GSdx/GSDevice7.cpp index 8ca5a59f36..477efe6638 100644 --- a/plugins/GSdx/GSDevice7.cpp +++ b/plugins/GSdx/GSDevice7.cpp @@ -112,7 +112,7 @@ bool GSDevice7::Reset(int w, int h, bool fs) HRGN hrgn = CreateRectRgn(0, 0, w, h); - BYTE buff[1024]; + uint8 buff[1024]; GetRegionData(hrgn, sizeof(buff), (RGNDATA*)buff); @@ -129,12 +129,13 @@ bool GSDevice7::Reset(int w, int h, bool fs) return true; } -void GSDevice7::Present(const CRect& r) +void GSDevice7::Present(const GSVector4i& r) { HRESULT hr; - CRect cr; - GetClientRect(m_hWnd, &cr); + GSVector4i cr; + + GetClientRect(m_hWnd, cr); DDSURFACEDESC2 desc; memset(&desc, 0, sizeof(desc)); @@ -142,9 +143,9 @@ void GSDevice7::Present(const CRect& r) hr = m_backbuffer->GetSurfaceDesc(&desc); - if(desc.dwWidth != cr.Width() || desc.dwHeight != cr.Height()) + if(desc.dwWidth != cr.width() || desc.dwHeight != cr.height()) { - Reset(cr.Width(), cr.Height(), false); + Reset(cr.width(), cr.height(), false); } DDBLTFX fx; @@ -156,9 +157,9 @@ void GSDevice7::Present(const CRect& r) hr = m_backbuffer->Blt(NULL, NULL, NULL, DDBLT_WAIT | DDBLT_COLORFILL, &fx); - CRect r2 = r; + GSVector4i r2 = r; - hr = m_backbuffer->Blt(&r2, m_merge, NULL, DDBLT_WAIT, NULL); + hr = m_backbuffer->Blt(r2, m_merge, NULL, DDBLT_WAIT, NULL); r2 = cr; @@ -169,7 +170,7 @@ void GSDevice7::Present(const CRect& r) hr = m_dd->WaitForVerticalBlank(DDWAITVB_BLOCKBEGIN, NULL); } - hr = m_primary->Blt(&r2, m_backbuffer, &cr, DDBLT_WAIT, NULL); + hr = m_primary->Blt(r2, m_backbuffer, cr, DDBLT_WAIT, NULL); if(hr == DDERR_SURFACELOST) { diff --git a/plugins/GSdx/GSDevice7.h b/plugins/GSdx/GSDevice7.h index 42d2853cf7..3029259a92 100644 --- a/plugins/GSdx/GSDevice7.h +++ b/plugins/GSdx/GSDevice7.h @@ -45,14 +45,14 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); bool IsLost() {return false;} - void Present(const CRect& r); + void Present(const GSVector4i& r); void BeginScene() {} void EndScene() {} void Draw(const string& s) {} bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} void ClearRenderTarget(Texture& t, const GSVector4& c) {} - void ClearRenderTarget(Texture& t, DWORD c) {} + void ClearRenderTarget(Texture& t, uint32 c) {} void ClearDepth(Texture& t, float c) {} - void ClearStencil(Texture& t, BYTE c) {} + void ClearStencil(Texture& t, uint8 c) {} }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 73f46d7482..4e38b0af0d 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -237,7 +237,7 @@ bool GSDevice9::Reset(int w, int h, bool fs) m_ps_cb = NULL; m_ps_cb_len = 0; m_ps_ss = NULL; - m_scissor = CRect(0, 0, 0, 0); + m_scissor = GSVector4i::zero(); m_dss = NULL; m_sref = 0; m_bs = NULL; @@ -280,7 +280,7 @@ bool GSDevice9::Reset(int w, int h, bool fs) if(!m_dev) { - UINT flags = D3DCREATE_MULTITHREADED | (m_d3dcaps.VertexProcessingCaps ? D3DCREATE_HARDWARE_VERTEXPROCESSING : D3DCREATE_SOFTWARE_VERTEXPROCESSING); + uint32 flags = D3DCREATE_MULTITHREADED | (m_d3dcaps.VertexProcessingCaps ? D3DCREATE_HARDWARE_VERTEXPROCESSING : D3DCREATE_SOFTWARE_VERTEXPROCESSING); hr = m_d3d->CreateDevice(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, m_hWnd, flags, &m_pp, &m_dev); @@ -347,15 +347,15 @@ bool GSDevice9::IsLost() return hr == D3DERR_DEVICELOST || hr == D3DERR_DEVICENOTRESET; } -void GSDevice9::Present(const CRect& r) +void GSDevice9::Present(const GSVector4i& r) { - CRect cr; + GSVector4i cr; - GetClientRect(m_hWnd, &cr); + GetClientRect(m_hWnd, cr); - if(m_backbuffer.GetWidth() != cr.Width() || m_backbuffer.GetHeight() != cr.Height()) + if(m_backbuffer.GetWidth() != cr.width() || m_backbuffer.GetHeight() != cr.height()) { - Reset(cr.Width(), cr.Height(), false); + Reset(cr.width(), cr.height(), false); } OMSetRenderTargets(m_backbuffer, NULL); @@ -396,13 +396,13 @@ void GSDevice9::Draw(const string& s) OMSetRenderTargets(m_backbuffer, NULL); - CRect r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); + GSVector4i r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); D3DCOLOR c = D3DCOLOR_ARGB(255, 0, 255, 0); - if(m_font->DrawText(NULL, str, -1, &r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) + if(m_font->DrawText(NULL, str, -1, r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) { - m_font->DrawText(NULL, str, -1, &r, DT_LEFT|DT_WORDBREAK, c); + m_font->DrawText(NULL, str, -1, r, DT_LEFT|DT_WORDBREAK, c); } EndScene(); @@ -447,10 +447,10 @@ bool GSDevice9::CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, i void GSDevice9::ClearRenderTarget(Texture& t, const GSVector4& c) { - ClearRenderTarget(t, D3DCOLOR_RGBA((BYTE)(c.r * 255 + 0.5f), (BYTE)(c.g * 255 + 0.5f), (BYTE)(c.b * 255 + 0.5f), (BYTE)(c.a * 255 + 0.5f))); + ClearRenderTarget(t, (c * 255 + 0.5f).zyxw().rgba32()); } -void GSDevice9::ClearRenderTarget(Texture& t, DWORD c) +void GSDevice9::ClearRenderTarget(Texture& t, uint32 c) { CComPtr surface; m_dev->GetRenderTarget(0, &surface); @@ -468,7 +468,7 @@ void GSDevice9::ClearDepth(Texture& t, float c) m_dev->SetDepthStencilSurface(surface); } -void GSDevice9::ClearStencil(Texture& t, BYTE c) +void GSDevice9::ClearStencil(Texture& t, uint8 c) { CComPtr surface; m_dev->GetDepthStencilSurface(&surface); @@ -580,7 +580,7 @@ void GSDevice9::DoInterlace(Texture& st, Texture& dt, int shader, bool linear, f StretchRect(st, sr, dt, dr, m_interlace.ps[shader], (const float*)&cb, 1, linear); } /* -void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, UINT count, const void* vertices, UINT stride) +void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, const void* vertices, size_t stride) { void* data = NULL; @@ -600,7 +600,7 @@ void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, UINT count, const } } */ -void GSDevice9::IASetVertexBuffer(UINT count, const void* vertices, UINT stride) +void GSDevice9::IASetVertexBuffer(int count, const void* vertices, size_t stride) { m_vb_count = count; m_vb_vertices = vertices; @@ -728,19 +728,19 @@ void GSDevice9::PSSetSamplerState(Direct3DSamplerState9* ss) } } -void GSDevice9::RSSet(int width, int height, const RECT* scissor) +void GSDevice9::RSSet(int width, int height, const GSVector4i* scissor) { - CRect r = scissor ? *scissor : CRect(0, 0, width, height); + GSVector4i r = scissor ? *scissor : GSVector4i(0, 0, width, height); - if(m_scissor != r) + if(!m_scissor.eq(r)) { - m_dev->SetScissorRect(&r); + m_dev->SetScissorRect(r); m_scissor = r; } } -void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss, UINT sref) +void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref) { if(m_dss != dss || m_sref != sref) { @@ -770,7 +770,7 @@ void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss, UINT sre } } -void GSDevice9::OMSetBlendState(Direct3DBlendState9* bs, DWORD bf) +void GSDevice9::OMSetBlendState(Direct3DBlendState9* bs, uint32 bf) { if(m_bs != bs || m_bf != bf) { @@ -914,7 +914,7 @@ void GSDevice9::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const // FIXME: D3DXCompileShaderFromResource of d3dx9 v37 (march 2008) calls GetFullPathName on id for some reason and then crashes -static HRESULT LoadShader(UINT id, LPCSTR& data, DWORD& size) +static HRESULT LoadShader(uint32 id, LPCSTR& data, uint32& size) { CComPtr shader, error; @@ -935,7 +935,7 @@ static HRESULT LoadShader(UINT id, LPCSTR& data, DWORD& size) return S_OK; } -HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il) +HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il) { const char* target; @@ -959,7 +959,7 @@ HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); LPCSTR data; - DWORD size; + uint32 size; hr = LoadShader(id, data, size); @@ -995,10 +995,10 @@ HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* return S_OK; } -HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps) +HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps) { const char* target = NULL; - UINT flags = 0; + uint32 flags = 0; if(m_d3dcaps.PixelShaderVersion >= D3DPS_VERSION(3, 0)) { @@ -1021,7 +1021,7 @@ HRESULT GSDevice9::CompileShader(UINT id, const string& entry, const D3DXMACRO* // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, flags, &shader, &error, NULL); LPCSTR data; - DWORD size; + uint32 size; hr = LoadShader(id, data, size); diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 105b3f72f9..c299688715 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -67,9 +67,9 @@ private: // state cache IDirect3DVertexBuffer9* m_vb; - UINT m_vb_count; + int m_vb_count; const void* m_vb_vertices; - UINT m_vb_stride; + size_t m_vb_stride; IDirect3DVertexDeclaration9* m_layout; D3DPRIMITIVETYPE m_topology; IDirect3DVertexShader9* m_vs; @@ -80,11 +80,11 @@ private: float* m_ps_cb; int m_ps_cb_len; Direct3DSamplerState9* m_ps_ss; - CRect m_scissor; + GSVector4i m_scissor; Direct3DDepthStencilState9* m_dss; - UINT m_sref; + uint32 m_sref; Direct3DBlendState9* m_bs; - DWORD m_bf; + uint32 m_bf; IDirect3DSurface9* m_rtv; IDirect3DSurface9* m_dsv; @@ -136,16 +136,16 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); bool IsLost(); - void Present(const CRect& r); + void Present(const GSVector4i& r); void BeginScene(); void EndScene(); void Draw(const string& s); bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); void ClearRenderTarget(Texture& t, const GSVector4& c); - void ClearRenderTarget(Texture& t, DWORD c); + void ClearRenderTarget(Texture& t, uint32 c); void ClearDepth(Texture& t, float c); - void ClearStencil(Texture& t, BYTE c); + void ClearStencil(Texture& t, uint8 c); bool CreateRenderTarget(Texture& t, int w, int h, int format = 0); bool CreateDepthStencil(Texture& t, int w, int h, int format = 0); @@ -155,21 +155,21 @@ public: IDirect3DDevice9* operator->() {return m_dev;} operator IDirect3DDevice9*() {return m_dev;} - // TODO: void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, UINT count, const void* vertices, UINT stride); - void IASetVertexBuffer(UINT count, const void* vertices, UINT stride); + // TODO: void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, const void* vertices, size_t stride); + void IASetVertexBuffer(int count, const void* vertices, size_t stride); void IASetInputLayout(IDirect3DVertexDeclaration9* layout); void IASetPrimitiveTopology(D3DPRIMITIVETYPE topology); void VSSetShader(IDirect3DVertexShader9* vs, const float* vs_cb, int vs_cb_len); void PSSetShaderResources(IDirect3DTexture9* srv0, IDirect3DTexture9* srv1); void PSSetShader(IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len); void PSSetSamplerState(Direct3DSamplerState9* ss); - void RSSet(int width, int height, const RECT* scissor = NULL); - void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, UINT sref); - void OMSetBlendState(Direct3DBlendState9* bs, DWORD bf); + void RSSet(int width, int height, const GSVector4i* scissor = NULL); + void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref); + void OMSetBlendState(Direct3DBlendState9* bs, uint32 bf); void OMSetRenderTargets(IDirect3DSurface9* rtv, IDirect3DSurface9* dsv); void DrawPrimitive(); - template void IASetVertexBuffer(UINT count, T* vertices) + template void IASetVertexBuffer(int count, T* vertices) { IASetVertexBuffer(count, vertices, sizeof(T)); } @@ -179,8 +179,8 @@ public: void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); - HRESULT CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il); - HRESULT CompileShader(UINT id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps); + HRESULT CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il); + HRESULT CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps); virtual bool IsCurrentRGBA() { diff --git a/plugins/GSdx/GSDeviceNull.h b/plugins/GSdx/GSDeviceNull.h index acd471f71d..bc452f7a7e 100644 --- a/plugins/GSdx/GSDeviceNull.h +++ b/plugins/GSdx/GSDeviceNull.h @@ -40,14 +40,14 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); bool IsLost() {return false;} - void Present(const CRect& r) {} + void Present(const GSVector4i& r) {} void BeginScene() {} void EndScene() {} void Draw(const string& s) {} bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} void ClearRenderTarget(Texture& t, const GSVector4& c) {} - void ClearRenderTarget(Texture& t, DWORD c) {} + void ClearRenderTarget(Texture& t, uint32 c) {} void ClearDepth(Texture& t, float c) {} - void ClearStencil(Texture& t, BYTE c) {} + void ClearStencil(Texture& t, uint8 c) {} }; diff --git a/plugins/GSdx/GSDirtyRect.cpp b/plugins/GSdx/GSDirtyRect.cpp index f1d12497bb..9815a4a28c 100644 --- a/plugins/GSdx/GSDirtyRect.cpp +++ b/plugins/GSdx/GSDirtyRect.cpp @@ -23,36 +23,41 @@ #include "GSDirtyRect.h" GSDirtyRect::GSDirtyRect() - : m_psm(PSM_PSMCT32) - , m_rect(0, 0, 0, 0) + : psm(PSM_PSMCT32) { + left = top = right = bottom = 0; } -GSDirtyRect::GSDirtyRect(DWORD psm, CRect rect) +GSDirtyRect::GSDirtyRect(const GSVector4i& r, uint32 psm) + : psm(psm) { - m_psm = psm; - m_rect = rect; + left = r.left; + top = r.top; + right = r.right; + bottom = r.bottom; } -CRect GSDirtyRect::GetDirtyRect(const GIFRegTEX0& TEX0) +GSVector4i GSDirtyRect::GetDirtyRect(const GIFRegTEX0& TEX0) { - CRect r = m_rect; + GSVector4i r; - CSize src = GSLocalMemory::m_psm[m_psm].bs; + GSVector2i src = GSLocalMemory::m_psm[psm].bs; - r.left = (r.left) & ~(src.cx-1); - r.right = (r.right + (src.cx-1) /* + 1 */) & ~(src.cx-1); - r.top = (r.top) & ~(src.cy-1); - r.bottom = (r.bottom + (src.cy-1) /* + 1 */) & ~(src.cy-1); - - if(m_psm != TEX0.PSM) + if(psm != TEX0.PSM) { - CSize dst = GSLocalMemory::m_psm[TEX0.PSM].bs; + GSVector2i dst = GSLocalMemory::m_psm[TEX0.PSM].bs; - r.left = MulDiv(m_rect.left, dst.cx, src.cx); - r.right = MulDiv(m_rect.right, dst.cx, src.cx); - r.top = MulDiv(m_rect.top, dst.cy, src.cy); - r.bottom = MulDiv(m_rect.bottom, dst.cy, src.cy); + r.left = MulDiv(left, dst.x, src.x); + r.right = MulDiv(right, dst.x, src.x); + r.top = MulDiv(top, dst.y, src.y); + r.bottom = MulDiv(bottom, dst.y, src.y); + } + else + { + r.left = left & ~(src.x - 1); + r.right = (right + (src.x - 1) /* + 1 */) & ~(src.x - 1); + r.top = top & ~(src.y - 1); + r.bottom = (bottom + (src.y - 1) /* + 1 */) & ~(src.y - 1); } return r; @@ -60,19 +65,21 @@ CRect GSDirtyRect::GetDirtyRect(const GIFRegTEX0& TEX0) // -CRect GSDirtyRectList::GetDirtyRect(const GIFRegTEX0& TEX0, CSize size) +GSVector4i GSDirtyRectList::GetDirtyRectAndClear(const GIFRegTEX0& TEX0, const GSVector2i& size) { - if(empty()) + if(!empty()) { - return CRect(0, 0, 0, 0); - } - - CRect r(INT_MAX, INT_MAX, 0, 0); + GSVector4i r(INT_MAX, INT_MAX, 0, 0); - for(list::iterator i = begin(); i != end(); i++) - { - r |= i->GetDirtyRect(TEX0); + for(list::iterator i = begin(); i != end(); i++) + { + r = r.runion(i->GetDirtyRect(TEX0)); + } + + clear(); + + return r.rintersect(GSVector4i(0, 0, size.x, size.y)); } - return r & CRect(0, 0, size.cx, size.cy); + return GSVector4i::zero(); } diff --git a/plugins/GSdx/GSDirtyRect.h b/plugins/GSdx/GSDirtyRect.h index a573d79ab2..e9b7699055 100644 --- a/plugins/GSdx/GSDirtyRect.h +++ b/plugins/GSdx/GSDirtyRect.h @@ -25,18 +25,22 @@ class GSDirtyRect { - DWORD m_psm; - CRect m_rect; + int left; + int top; + int right; + int bottom; + + uint32 psm; public: GSDirtyRect(); - GSDirtyRect(DWORD psm, CRect rect); - CRect GetDirtyRect(const GIFRegTEX0& TEX0); + GSDirtyRect(const GSVector4i& r, uint32 psm); + GSVector4i GetDirtyRect(const GIFRegTEX0& TEX0); }; class GSDirtyRectList : public list { public: GSDirtyRectList() {} - CRect GetDirtyRect(const GIFRegTEX0& TEX0, CSize size); + GSVector4i GetDirtyRectAndClear(const GIFRegTEX0& TEX0, const GSVector2i& size); }; \ No newline at end of file diff --git a/plugins/GSdx/GSDrawScanline.cpp b/plugins/GSdx/GSDrawScanline.cpp index 4813bc0440..c3452fe05c 100644 --- a/plugins/GSdx/GSDrawScanline.cpp +++ b/plugins/GSdx/GSDrawScanline.cpp @@ -56,8 +56,8 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) m_env.zm = GSVector4i(p->zm); m_env.aref = GSVector4i((int)context->TEST.AREF); m_env.afix = GSVector4i((int)context->ALPHA.FIX << 7).xxzzlh(); - m_env.frb = GSVector4i((int)env.FOGCOL.ai32[0] & 0x00ff00ff); - m_env.fga = GSVector4i((int)(env.FOGCOL.ai32[0] >> 8) & 0x00ff00ff); + m_env.frb = GSVector4i((int)env.FOGCOL.u32[0] & 0x00ff00ff); + m_env.fga = GSVector4i((int)(env.FOGCOL.u32[0] >> 8) & 0x00ff00ff); m_env.dimx = env.dimx; if(m_sel.fpsm == 1) @@ -212,34 +212,34 @@ void GSDrawScanline::DrawSolidRect(const GSVector4i& r, const GSVertexSW& v) // FIXME: sometimes the frame and z buffer may overlap, the outcome is undefined - DWORD m; + uint32 m; m = m_env.zm.u32[0]; if(m != 0xffffffff) { - DWORD z = (DWORD)(float)v.p.z; + uint32 z = (uint32)v.p.z; if(m_sel.zpsm != 2) { if(m == 0) { - DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); + DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); } else { - DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); + DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); } } else { if(m == 0) { - DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); + DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); } else { - DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); + DrawSolidRectT(m_env.zbr, m_env.zbc[0], r, z, m); } } } @@ -248,7 +248,7 @@ void GSDrawScanline::DrawSolidRect(const GSVector4i& r, const GSVertexSW& v) if(m != 0xffffffff) { - DWORD c = (GSVector4i(v.c) >> 7).rgba32(); + uint32 c = (GSVector4i(v.c) >> 7).rgba32(); if(m_state->m_context->FBA.FBA) { @@ -259,11 +259,11 @@ void GSDrawScanline::DrawSolidRect(const GSVector4i& r, const GSVertexSW& v) { if(m == 0) { - DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); + DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); } else { - DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); + DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); } } else @@ -272,25 +272,25 @@ void GSDrawScanline::DrawSolidRect(const GSVector4i& r, const GSVertexSW& v) if(m == 0) { - DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); + DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); } else { - DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); + DrawSolidRectT(m_env.fbr, m_env.fbc[0], r, c, m); } } } } template -void GSDrawScanline::DrawSolidRectT(const GSVector4i* row, int* col, const GSVector4i& r, DWORD c, DWORD m) +void GSDrawScanline::DrawSolidRectT(const GSVector4i* row, int* col, const GSVector4i& r, uint32 c, uint32 m) { if(m == 0xffffffff) return; GSVector4i color((int)c); GSVector4i mask((int)m); - if(sizeof(T) == sizeof(WORD)) + if(sizeof(T) == sizeof(uint16)) { color = color.xxzzlh(); mask = mask.xxzzlh(); @@ -314,13 +314,13 @@ void GSDrawScanline::DrawSolidRectT(const GSVector4i* row, int* col, const GSVec } template -void GSDrawScanline::FillRect(const GSVector4i* row, int* col, const GSVector4i& r, DWORD c, DWORD m) +void GSDrawScanline::FillRect(const GSVector4i* row, int* col, const GSVector4i& r, uint32 c, uint32 m) { if(r.x >= r.z) return; for(int y = r.y; y < r.w; y++) { - DWORD base = row[y].x; + uint32 base = row[y].x; for(int x = r.x; x < r.z; x++) { @@ -338,7 +338,7 @@ void GSDrawScanline::FillBlock(const GSVector4i* row, int* col, const GSVector4i for(int y = r.y; y < r.w; y += 8) { - DWORD base = row[y].x; + uint32 base = row[y].x; for(int x = r.x; x < r.z; x += 8 * 4 / sizeof(T)) { @@ -363,7 +363,7 @@ GSDrawScanline::GSSetupPrimMap::GSSetupPrimMap(GSScanlineEnvironment& env) { } -GSSetupPrimCodeGenerator* GSDrawScanline::GSSetupPrimMap::Create(UINT64 key, void* ptr, size_t maxsize) +GSSetupPrimCodeGenerator* GSDrawScanline::GSSetupPrimMap::Create(uint64 key, void* ptr, size_t maxsize) { return new GSSetupPrimCodeGenerator(m_env, key, ptr, maxsize); } @@ -376,7 +376,7 @@ GSDrawScanline::GSDrawScanlineMap::GSDrawScanlineMap(GSScanlineEnvironment& env) { } -GSDrawScanlineCodeGenerator* GSDrawScanline::GSDrawScanlineMap::Create(UINT64 key, void* ptr, size_t maxsize) +GSDrawScanlineCodeGenerator* GSDrawScanline::GSDrawScanlineMap::Create(uint64 key, void* ptr, size_t maxsize) { return new GSDrawScanlineCodeGenerator(m_env, key, ptr, maxsize); } diff --git a/plugins/GSdx/GSDrawScanline.h b/plugins/GSdx/GSDrawScanline.h index 2221707dc8..9f9a1ea3d1 100644 --- a/plugins/GSdx/GSDrawScanline.h +++ b/plugins/GSdx/GSDrawScanline.h @@ -35,24 +35,24 @@ class GSDrawScanline : public GSAlignedClass<16>, public IDrawScanline // - class GSSetupPrimMap : public GSCodeGeneratorFunctionMap + class GSSetupPrimMap : public GSCodeGeneratorFunctionMap { GSScanlineEnvironment& m_env; public: GSSetupPrimMap(GSScanlineEnvironment& env); - GSSetupPrimCodeGenerator* Create(UINT64 key, void* ptr, size_t maxsize); + GSSetupPrimCodeGenerator* Create(uint64 key, void* ptr, size_t maxsize); } m_sp; // - class GSDrawScanlineMap : public GSCodeGeneratorFunctionMap + class GSDrawScanlineMap : public GSCodeGeneratorFunctionMap { GSScanlineEnvironment& m_env; public: GSDrawScanlineMap(GSScanlineEnvironment& env); - GSDrawScanlineCodeGenerator* Create(UINT64 key, void* ptr, size_t maxsize); + GSDrawScanlineCodeGenerator* Create(uint64 key, void* ptr, size_t maxsize); } m_ds; // @@ -60,10 +60,10 @@ class GSDrawScanline : public GSAlignedClass<16>, public IDrawScanline void DrawSolidRect(const GSVector4i& r, const GSVertexSW& v); template - void DrawSolidRectT(const GSVector4i* row, int* col, const GSVector4i& r, DWORD c, DWORD m); + void DrawSolidRectT(const GSVector4i* row, int* col, const GSVector4i& r, uint32 c, uint32 m); template - __forceinline void FillRect(const GSVector4i* row, int* col, const GSVector4i& r, DWORD c, DWORD m); + __forceinline void FillRect(const GSVector4i* row, int* col, const GSVector4i& r, uint32 c, uint32 m); template __forceinline void FillBlock(const GSVector4i* row, int* col, const GSVector4i& r, const GSVector4i& c, const GSVector4i& m); diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp index a6eb172b3e..e87c75c4f8 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp @@ -875,10 +875,10 @@ void GSDrawScanlineCodeGenerator::SampleTexture() // xmm1, xmm4, xmm6 = free // xmm7 = used - // c00 = addr00.gather32_32((const DWORD/BYTE*)tex[, clut]); - // c01 = addr01.gather32_32((const DWORD/BYTE*)tex[, clut]); - // c10 = addr10.gather32_32((const DWORD/BYTE*)tex[, clut]); - // c11 = addr11.gather32_32((const DWORD/BYTE*)tex[, clut]); + // c00 = addr00.gather32_32((const uint32/uint8*)tex[, clut]); + // c01 = addr01.gather32_32((const uint32/uint8*)tex[, clut]); + // c10 = addr10.gather32_32((const uint32/uint8*)tex[, clut]); + // c11 = addr11.gather32_32((const uint32/uint8*)tex[, clut]); ReadTexel(xmm6, xmm5, xmm1, xmm4); @@ -994,7 +994,7 @@ void GSDrawScanlineCodeGenerator::SampleTexture() paddd(xmm2, xmm4); - // c00 = addr00.gather32_32((const DWORD/BYTE*)tex[, clut]); + // c00 = addr00.gather32_32((const uint32/uint8*)tex[, clut]); ReadTexel(xmm5, xmm2, xmm0, xmm1); diff --git a/plugins/GSdx/GSDump.cpp b/plugins/GSdx/GSDump.cpp index 7f1095a5b3..4082e7823f 100644 --- a/plugins/GSdx/GSDump.cpp +++ b/plugins/GSdx/GSDump.cpp @@ -36,7 +36,7 @@ GSDump::~GSDump() Close(); } -void GSDump::Open(const string& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs) +void GSDump::Open(const string& fn, uint32 crc, const GSFreezeData& fd, const GSPrivRegSet* regs) { m_gs = fopen((fn + ".gs").c_str(), "wb"); m_obj = fopen((fn + ".obj").c_str(), "wt"); @@ -60,7 +60,7 @@ void GSDump::Close() if(m_obj) {fclose(m_obj); m_obj = NULL;} } -void GSDump::Transfer(int index, BYTE* mem, size_t size) +void GSDump::Transfer(int index, uint8* mem, size_t size) { if(m_gs && size > 0) { @@ -71,7 +71,7 @@ void GSDump::Transfer(int index, BYTE* mem, size_t size) } } -void GSDump::ReadFIFO(UINT32 size) +void GSDump::ReadFIFO(uint32 size) { if(m_gs && size > 0) { diff --git a/plugins/GSdx/GSDump.h b/plugins/GSdx/GSDump.h index 1ffc8167ec..5cc2c88104 100644 --- a/plugins/GSdx/GSDump.h +++ b/plugins/GSdx/GSDump.h @@ -55,10 +55,10 @@ public: GSDump(); virtual ~GSDump(); - void Open(const string& fn, DWORD crc, const GSFreezeData& fd, const GSPrivRegSet* regs); + void Open(const string& fn, uint32 crc, const GSFreezeData& fd, const GSPrivRegSet* regs); void Close(); - void ReadFIFO(UINT32 size); - void Transfer(int index, BYTE* mem, size_t size); + void ReadFIFO(uint32 size); + void Transfer(int index, uint8* mem, size_t size); void VSync(int field, bool last, const GSPrivRegSet* regs); void Object(GSVertexSW* vertices, int count, GS_PRIM_CLASS primclass); operator bool() {return m_gs != NULL;} diff --git a/plugins/GSdx/GSFunctionMap.h b/plugins/GSdx/GSFunctionMap.h index 34074703d0..2d7d052d8d 100644 --- a/plugins/GSdx/GSFunctionMap.h +++ b/plugins/GSdx/GSFunctionMap.h @@ -158,9 +158,9 @@ public: template class GSCodeGeneratorFunctionMap : public GSFunctionMap { - DWORD m_id; + uint32 m_id; string m_name; - hash_map m_cgmap; + hash_map m_cgmap; GSCodeBuffer m_cb; enum {MAX_SIZE = 4096}; @@ -177,7 +177,7 @@ public: virtual ~GSCodeGeneratorFunctionMap() { - for(hash_map::iterator i = m_cgmap.begin(); i != m_cgmap.end(); i++) + for(hash_map::iterator i = m_cgmap.begin(); i != m_cgmap.end(); i++) { delete (*i).second; } @@ -207,7 +207,7 @@ public: // vtune method registration - string name = format("%s<%016I64x>()", m_name.c_str(), (UINT64)key); + string name = format("%s<%016I64x>()", m_name.c_str(), (uint64)key); iJIT_Method_Load ml; diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 4a8a8ebff7..5e1ab20d09 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -31,26 +31,26 @@ ASSERT((r).Width() >= w && (r).Height() >= h && !((r).left&(w-1)) && !((r).top&(h-1)) && !((r).right&(w-1)) && !((r).bottom&(h-1))); \ #define FOREACH_BLOCK_START(w, h, bpp) \ - DWORD bp = TEX0.TBP0; \ - DWORD bw = TEX0.TBW; \ - int offset = dstpitch * h - (r.right - r.left) * bpp / 8; \ - for(int y = r.top; y < r.bottom; y += h, dst += offset) \ + uint32 bp = TEX0.TBP0; \ + uint32 bw = TEX0.TBW; \ + int offset = dstpitch * h - r.width() * bpp / 8; \ + for(int y = r.top, ye = r.bottom; y < ye; y += h, dst += offset) \ { ASSERT_BLOCK(r, w, h); \ - for(int x = r.left; x < r.right; x += w, dst += w * bpp / 8) \ + for(int x = r.left, xe = r.right; x < xe; x += w, dst += w * bpp / 8) \ { \ #define FOREACH_BLOCK_END }} // -DWORD GSLocalMemory::pageOffset32[32][32][64]; -DWORD GSLocalMemory::pageOffset32Z[32][32][64]; -DWORD GSLocalMemory::pageOffset16[32][64][64]; -DWORD GSLocalMemory::pageOffset16S[32][64][64]; -DWORD GSLocalMemory::pageOffset16Z[32][64][64]; -DWORD GSLocalMemory::pageOffset16SZ[32][64][64]; -DWORD GSLocalMemory::pageOffset8[32][64][128]; -DWORD GSLocalMemory::pageOffset4[32][128][128]; +uint32 GSLocalMemory::pageOffset32[32][32][64]; +uint32 GSLocalMemory::pageOffset32Z[32][32][64]; +uint32 GSLocalMemory::pageOffset16[32][64][64]; +uint32 GSLocalMemory::pageOffset16S[32][64][64]; +uint32 GSLocalMemory::pageOffset16Z[32][64][64]; +uint32 GSLocalMemory::pageOffset16SZ[32][64][64]; +uint32 GSLocalMemory::pageOffset8[32][64][128]; +uint32 GSLocalMemory::pageOffset4[32][128][128]; int GSLocalMemory::rowOffset32[2048]; int GSLocalMemory::rowOffset32Z[2048]; @@ -72,8 +72,8 @@ int GSLocalMemory::blockOffset4[256]; // -DWORD GSLocalMemory::m_xtbl[1024]; -DWORD GSLocalMemory::m_ytbl[1024]; +uint32 GSLocalMemory::m_xtbl[1024]; +uint32 GSLocalMemory::m_ytbl[1024]; // @@ -84,7 +84,7 @@ GSLocalMemory::psm_t GSLocalMemory::m_psm[64]; GSLocalMemory::GSLocalMemory() : m_clut(this) { - m_vm8 = (BYTE*)VirtualAlloc(NULL, m_vmsize * 2, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); + m_vm8 = (uint8*)VirtualAlloc(NULL, m_vmsize * 2, MEM_COMMIT | MEM_RESERVE, PAGE_READWRITE); memset(m_vm8, 0, m_vmsize); @@ -218,8 +218,8 @@ GSLocalMemory::GSLocalMemory() m_psm[i].rtxbP = &GSLocalMemory::ReadTextureBlock32; m_psm[i].bpp = m_psm[i].trbpp = 32; m_psm[i].pal = 0; - m_psm[i].bs = CSize(8, 8); - m_psm[i].pgs = CSize(64, 32); + m_psm[i].bs = GSVector2i(8, 8); + m_psm[i].pgs = GSVector2i(64, 32); for(int j = 0; j < 8; j++) m_psm[i].rowOffset[j] = rowOffset32; m_psm[i].blockOffset = blockOffset32; } @@ -432,15 +432,15 @@ GSLocalMemory::GSLocalMemory() m_psm[PSM_PSMZ24].trbpp = 24; m_psm[PSM_PSMZ16].trbpp = m_psm[PSM_PSMZ16S].trbpp = 16; - m_psm[PSM_PSMCT16].bs = m_psm[PSM_PSMCT16S].bs = CSize(16, 8); - m_psm[PSM_PSMT8].bs = CSize(16, 16); - m_psm[PSM_PSMT4].bs = CSize(32, 16); - m_psm[PSM_PSMZ16].bs = m_psm[PSM_PSMZ16S].bs = CSize(16, 8); + m_psm[PSM_PSMCT16].bs = m_psm[PSM_PSMCT16S].bs = GSVector2i(16, 8); + m_psm[PSM_PSMT8].bs = GSVector2i(16, 16); + m_psm[PSM_PSMT4].bs = GSVector2i(32, 16); + m_psm[PSM_PSMZ16].bs = m_psm[PSM_PSMZ16S].bs = GSVector2i(16, 8); - m_psm[PSM_PSMCT16].pgs = m_psm[PSM_PSMCT16S].pgs = CSize(64, 64); - m_psm[PSM_PSMT8].pgs = CSize(128, 64); - m_psm[PSM_PSMT4].pgs = CSize(128, 128); - m_psm[PSM_PSMZ16].pgs = m_psm[PSM_PSMZ16S].pgs = CSize(64, 64); + m_psm[PSM_PSMCT16].pgs = m_psm[PSM_PSMCT16S].pgs = GSVector2i(64, 64); + m_psm[PSM_PSMT8].pgs = GSVector2i(128, 64); + m_psm[PSM_PSMT4].pgs = GSVector2i(128, 128); + m_psm[PSM_PSMZ16].pgs = m_psm[PSM_PSMZ16S].pgs = GSVector2i(64, 64); for(int i = 0; i < 8; i++) m_psm[PSM_PSMCT16].rowOffset[i] = rowOffset16; for(int i = 0; i < 8; i++) m_psm[PSM_PSMCT16S].rowOffset[i] = rowOffset16S; @@ -465,7 +465,7 @@ GSLocalMemory::~GSLocalMemory() { VirtualFree(m_vm8, 0, MEM_RELEASE); - for(hash_map::iterator i = m_omap.begin(); i != m_omap.end(); i++) + for(hash_map::iterator i = m_omap.begin(); i != m_omap.end(); i++) { Offset* o = (*i).second; @@ -477,21 +477,21 @@ GSLocalMemory::~GSLocalMemory() _aligned_free(o); } - for(hash_map::iterator i = m_o4map.begin(); i != m_o4map.end(); i++) + for(hash_map::iterator i = m_o4map.begin(); i != m_o4map.end(); i++) { _aligned_free((*i).second); } } -GSLocalMemory::Offset* GSLocalMemory::GetOffset(DWORD bp, DWORD bw, DWORD psm) +GSLocalMemory::Offset* GSLocalMemory::GetOffset(uint32 bp, uint32 bw, uint32 psm) { if(bw == 0) {ASSERT(0); return NULL;} ASSERT(m_psm[psm].bpp > 8); // only for 16/24/32/8h/4hh/4hl formats where all columns are the same - DWORD hash = bp | (bw << 14) | (psm << 20); + uint32 hash = bp | (bw << 14) | (psm << 20); - hash_map::iterator i = m_omap.find(hash); + hash_map::iterator i = m_omap.find(hash); if(i != m_omap.end()) { @@ -525,22 +525,22 @@ GSLocalMemory::Offset* GSLocalMemory::GetOffset(DWORD bp, DWORD bw, DWORD psm) GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) { - DWORD fbp = FRAME.Block(); - DWORD zbp = ZBUF.Block(); - DWORD fpsm = FRAME.PSM; - DWORD zpsm = ZBUF.PSM; - DWORD bw = FRAME.FBW; + uint32 fbp = FRAME.Block(); + uint32 zbp = ZBUF.Block(); + uint32 fpsm = FRAME.PSM; + uint32 zpsm = ZBUF.PSM; + uint32 bw = FRAME.FBW; ASSERT(m_psm[fpsm].trbpp > 8 || m_psm[zpsm].trbpp > 8); // "(psm & 0x0f) ^ ((psm & 0xf0) >> 2)" creates 4 bit unique identifiers for render target formats (only) - DWORD fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2); - DWORD zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2); + uint32 fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2); + uint32 zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2); - DWORD hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28); + uint32 hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28); - hash_map::iterator i = m_o4map.find(hash); + hash_map::iterator i = m_o4map.find(hash); if(i != m_o4map.end()) { @@ -574,15 +574,15 @@ GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, cons return o; } -bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, DWORD bw) +bool GSLocalMemory::FillRect(const GSVector4i& r, uint32 c, uint32 psm, uint32 bp, uint32 bw) { const psm_t& tbl = m_psm[psm]; writePixel wp = tbl.wp; pixelAddress bn = tbl.bn; - int w = tbl.bs.cx; - int h = tbl.bs.cy; + int w = tbl.bs.x; + int h = tbl.bs.y; int bpp = tbl.bpp; switch(bpp) @@ -593,24 +593,24 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, case 4: c = (c & 0xf) * 0x11111111; break; } - CRect clip; + GSVector4i clip; - clip.left = (r.x + (w - 1)) & ~(w - 1); - clip.top = (r.y + (h - 1)) & ~(h - 1); - clip.right = r.z & ~(w - 1); - clip.bottom = r.w & ~(h - 1); + clip.left = (r.left + (w - 1)) & ~(w - 1); + clip.top = (r.top + (h - 1)) & ~(h - 1); + clip.right = r.right & ~(w - 1); + clip.bottom = r.bottom & ~(h - 1); - for(int y = r.y; y < clip.top; y++) + for(int y = r.top; y < clip.top; y++) { - for(int x = r.x; x < r.z; x++) + for(int x = r.left; x < r.right; x++) { (this->*wp)(x, y, c, bp, bw); } } - for(int y = clip.bottom; y < r.w; y++) + for(int y = clip.bottom; y < r.bottom; y++) { - for(int x = r.x; x < r.z; x++) + for(int x = r.left; x < r.right; x++) { (this->*wp)(x, y, c, bp, bw); } @@ -622,12 +622,12 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, { for(int ys = y, ye = y + h; ys < ye; ys++) { - for(int x = r.x; x < clip.left; x++) + for(int x = r.left; x < clip.left; x++) { (this->*wp)(x, ys, c, bp, bw); } - for(int x = clip.right; x < r.z; x++) + for(int x = clip.right; x < r.right; x++) { (this->*wp)(x, ys, c, bp, bw); } @@ -635,11 +635,10 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, } } + GSVector4i c128(c); + if(psm == PSM_PSMCT24 || psm == PSM_PSMZ24) { - #if _M_SSE >= 0x200 - - GSVector4i c128(c); GSVector4i mask(0x00ffffff); for(int y = clip.top; y < clip.bottom; y += h) @@ -657,35 +656,9 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, } } } - - #else - - c &= 0x00ffffff; - - for(int y = clip.top; y < clip.bottom; y += h) - { - for(int x = clip.left; x < clip.right; x += w) - { - DWORD* p = (DWORD*)&m_vm8[bn(x, y, bp, bw) << 8]; - - for(int i = 0; i < 64; i += 4) - { - p[i + 0] = (p[i + 0] & 0xff000000) | c; - p[i + 1] = (p[i + 1] & 0xff000000) | c; - p[i + 2] = (p[i + 2] & 0xff000000) | c; - p[i + 3] = (p[i + 3] & 0xff000000) | c; - } - } - } - - #endif } else { - #if _M_SSE >= 0x200 - - GSVector4i c128(c); - for(int y = clip.top; y < clip.bottom; y += h) { for(int x = clip.left; x < clip.right; x += w) @@ -701,26 +674,6 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, } } } - - #else - - for(int y = clip.top; y < clip.bottom; y += h) - { - for(int x = clip.left; x < clip.right; x += w) - { - DWORD* p = (DWORD*)&m_vm8[bn(x, y, bp, bw) << 8]; - - for(int i = 0; i < 64; i += 4) - { - p[i + 0] = c; - p[i + 1] = c; - p[i + 2] = c; - p[i + 3] = c; - } - } - } - - #endif } return true; @@ -729,10 +682,10 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, //////////////////// template -void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) +void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) { - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; const int csy = bsy / 4; @@ -758,10 +711,10 @@ void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, BYTE* src, int } template -void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) +void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) { - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; for(int offset = srcpitch * bsy; h >= bsy; h -= bsy, y += bsy, src += offset) { @@ -785,10 +738,10 @@ void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, BYTE* src, int s } template -void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) +void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) { - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; for(; h > 0; y++, h--, src += srcpitch) { @@ -796,14 +749,14 @@ void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, BYTE* src, i { switch(psm) { - case PSM_PSMCT32: WritePixel32(x, y, *(DWORD*)&src[x * 4], bp, bw); break; - case PSM_PSMCT16: WritePixel16(x, y, *(WORD*)&src[x * 2], bp, bw); break; - case PSM_PSMCT16S: WritePixel16S(x, y, *(WORD*)&src[x * 2], bp, bw); break; + case PSM_PSMCT32: WritePixel32(x, y, *(uint32*)&src[x * 4], bp, bw); break; + case PSM_PSMCT16: WritePixel16(x, y, *(uint16*)&src[x * 2], bp, bw); break; + case PSM_PSMCT16S: WritePixel16S(x, y, *(uint16*)&src[x * 2], bp, bw); break; case PSM_PSMT8: WritePixel8(x, y, src[x], bp, bw); break; case PSM_PSMT4: WritePixel4(x, y, src[x >> 1] >> ((x & 1) << 2), bp, bw); break; - case PSM_PSMZ32: WritePixel32Z(x, y, *(DWORD*)&src[x * 4], bp, bw); break; - case PSM_PSMZ16: WritePixel16Z(x, y, *(WORD*)&src[x * 2], bp, bw); break; - case PSM_PSMZ16S: WritePixel16SZ(x, y, *(WORD*)&src[x * 2], bp, bw); break; + case PSM_PSMZ32: WritePixel32Z(x, y, *(uint32*)&src[x * 4], bp, bw); break; + case PSM_PSMZ16: WritePixel16Z(x, y, *(uint16*)&src[x * 2], bp, bw); break; + case PSM_PSMZ16S: WritePixel16SZ(x, y, *(uint16*)&src[x * 2], bp, bw); break; // TODO default: __assume(0); } @@ -812,12 +765,12 @@ void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, BYTE* src, i } template -void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) +void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) { - __declspec(align(16)) BYTE buff[64]; // merge buffer for one column + __declspec(align(16)) uint8 buff[64]; // merge buffer for one column - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; const int csy = bsy / 4; @@ -831,7 +784,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, i for(int x = l; x < r; x += bsx) { - BYTE* dst = NULL; + uint8* dst = NULL; switch(psm) { @@ -912,7 +865,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, i { for(int x = l; x < r; x += bsx) { - BYTE* dst = NULL; + uint8* dst = NULL; switch(psm) { @@ -963,7 +916,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, i } template -void GSLocalMemory::WriteImage(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; @@ -987,7 +940,7 @@ void GSLocalMemory::WriteImage(int& tx, int& ty, BYTE* src, int len, GIFRegBITBL if(ra - la >= bsx && h > 0) // "transfer width" >= "block width" && there is at least one full row { - BYTE* s = &src[-l * trbpp >> 3]; + uint8* s = &src[-l * trbpp >> 3]; src += srcpitch * h; len -= srcpitch * h; @@ -1072,12 +1025,12 @@ void GSLocalMemory::WriteImage(int& tx, int& ty, BYTE* src, int len, GIFRegBITBL #define IsTopLeftAligned(dsax, tx, ty, bw, bh) \ ((((int)dsax) & ((bw)-1)) == 0 && ((tx) & ((bw)-1)) == 0 && ((int)dsax) == (tx) && ((ty) & ((bh)-1)) == 0) -void GSLocalMemory::WriteImage24(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage24(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3; int th = len / srcpitch; @@ -1106,12 +1059,12 @@ void GSLocalMemory::WriteImage24(int& tx, int& ty, BYTE* src, int len, GIFRegBIT } } -void GSLocalMemory::WriteImage8H(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage8H(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW; int th = len / srcpitch; @@ -1140,12 +1093,12 @@ void GSLocalMemory::WriteImage8H(int& tx, int& ty, BYTE* src, int len, GIFRegBIT } } -void GSLocalMemory::WriteImage4HL(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage4HL(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2; int th = len / srcpitch; @@ -1174,12 +1127,12 @@ void GSLocalMemory::WriteImage4HL(int& tx, int& ty, BYTE* src, int len, GIFRegBI } } -void GSLocalMemory::WriteImage4HH(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage4HH(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2; int th = len / srcpitch; @@ -1207,12 +1160,12 @@ void GSLocalMemory::WriteImage4HH(int& tx, int& ty, BYTE* src, int len, GIFRegBI ty = th; } } -void GSLocalMemory::WriteImage24Z(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImage24Z(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(TRXREG.RRW == 0) return; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3; int th = len / srcpitch; @@ -1240,16 +1193,16 @@ void GSLocalMemory::WriteImage24Z(int& tx, int& ty, BYTE* src, int len, GIFRegBI ty = th; } } -void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) +void GSLocalMemory::WriteImageX(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) { if(len <= 0) return; - BYTE* pb = (BYTE*)src; - WORD* pw = (WORD*)src; - DWORD* pd = (DWORD*)src; + uint8* pb = (uint8*)src; + uint16* pw = (uint16*)src; + uint32* pd = (uint32*)src; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; psm_t* psm = &m_psm[BITBLTBUF.DPSM]; int x = tx; @@ -1266,7 +1219,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pd++) @@ -1286,12 +1239,12 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb += 3) { - WritePixel24(addr + offset[x], *(DWORD*)pb); + WritePixel24(addr + offset[x], *(uint32*)pb); } if(x == ex) {x = sx; y++;} @@ -1308,7 +1261,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pw++) @@ -1325,7 +1278,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb++) @@ -1342,7 +1295,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1360,7 +1313,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb++) @@ -1377,7 +1330,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1395,7 +1348,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1416,16 +1369,16 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITB // -void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const +void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const { if(len <= 0) return; - BYTE* pb = (BYTE*)dst; - WORD* pw = (WORD*)dst; - DWORD* pd = (DWORD*)dst; + uint8* pb = (uint8*)dst; + uint16* pw = (uint16*)dst; + uint32* pd = (uint32*)dst; - DWORD bp = BITBLTBUF.SBP; - DWORD bw = BITBLTBUF.SBW; + uint32 bp = BITBLTBUF.SBP; + uint32 bw = BITBLTBUF.SBW; psm_t* psm = &m_psm[BITBLTBUF.SPSM]; int x = tx; @@ -1442,7 +1395,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pd++) @@ -1462,16 +1415,16 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb += 3) { - DWORD dw = ReadPixel32(addr + offset[x]); + uint32 c = ReadPixel32(addr + offset[x]); - pb[0] = ((BYTE*)&dw)[0]; - pb[1] = ((BYTE*)&dw)[1]; - pb[2] = ((BYTE*)&dw)[2]; + pb[0] = ((uint8*)&c)[0]; + pb[1] = ((uint8*)&c)[1]; + pb[2] = ((uint8*)&c)[2]; } if(x == ex) {x = sx; y++;} @@ -1488,7 +1441,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pw++) @@ -1505,7 +1458,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb++) @@ -1522,7 +1475,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1539,7 +1492,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x++, pb++) @@ -1556,7 +1509,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1573,7 +1526,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL while(len > 0) { - DWORD addr = psm->pa(0, y, bp, bw); + uint32 addr = psm->pa(0, y, bp, bw); int* offset = psm->rowOffset[y & 7]; for(; len > 0 && x < ex; len--, x += 2, pb++) @@ -1593,7 +1546,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBL /////////////////// -void GSLocalMemory::ReadTexture32(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture32(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(8, 8, 32) { @@ -1602,7 +1555,7 @@ void GSLocalMemory::ReadTexture32(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture24(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { if(TEXA.AEM) { @@ -1622,35 +1575,35 @@ void GSLocalMemory::ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const } } -void GSLocalMemory::ReadTexture16(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16(BlockPtr16(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture16S(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16S(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16(BlockPtr16S(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16S(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture8(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; FOREACH_BLOCK_START(16, 16, 32) { @@ -1659,9 +1612,9 @@ void GSLocalMemory::ReadTexture8(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const UINT64* pal = m_clut; + const uint64* pal = m_clut; FOREACH_BLOCK_START(32, 16, 32) { @@ -1670,9 +1623,9 @@ void GSLocalMemory::ReadTexture4(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture8H(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8H(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; FOREACH_BLOCK_START(8, 8, 32) { @@ -1681,9 +1634,9 @@ void GSLocalMemory::ReadTexture8H(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4HL(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HL(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; FOREACH_BLOCK_START(8, 8, 32) { @@ -1692,9 +1645,9 @@ void GSLocalMemory::ReadTexture4HL(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4HH(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HH(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; FOREACH_BLOCK_START(8, 8, 32) { @@ -1703,7 +1656,7 @@ void GSLocalMemory::ReadTexture4HH(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture32Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture32Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(8, 8, 32) { @@ -1712,7 +1665,7 @@ void GSLocalMemory::ReadTexture32Z(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture24Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture24Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { if(TEXA.AEM) { @@ -1732,26 +1685,26 @@ void GSLocalMemory::ReadTexture24Z(const CRect& r, BYTE* dst, int dstpitch, cons } } -void GSLocalMemory::ReadTexture16Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16(BlockPtr16Z(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16Z(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture16SZ(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16SZ(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; FOREACH_BLOCK_START(16, 8, 32) { - ReadBlock16(BlockPtr16SZ(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr16SZ(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1760,12 +1713,12 @@ void GSLocalMemory::ReadTexture16SZ(const CRect& r, BYTE* dst, int dstpitch, con /////////////////// -void GSLocalMemory::ReadTextureBlock32(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock32(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock32(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock24(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { if(TEXA.AEM) { @@ -1777,65 +1730,65 @@ void GSLocalMemory::ReadTextureBlock24(DWORD bp, BYTE* dst, int dstpitch, const } } -void GSLocalMemory::ReadTextureBlock16(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock16(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; - ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr(bp), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } -void GSLocalMemory::ReadTextureBlock16S(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock16S(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; - ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr(bp), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } -void GSLocalMemory::ReadTextureBlock8(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock8(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, pal); } -void GSLocalMemory::ReadTextureBlock4(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { const UINT64* pal = m_clut; ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, pal); } -void GSLocalMemory::ReadTextureBlock8H(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock8H(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, pal); } -void GSLocalMemory::ReadTextureBlock4HL(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4HL(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, pal); } -void GSLocalMemory::ReadTextureBlock4HH(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4HH(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; ReadAndExpandBlock4HH_32(BlockPtr(bp), dst, dstpitch, pal); } -void GSLocalMemory::ReadTextureBlock32Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock32Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock32(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock24Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock24Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { if(TEXA.AEM) { @@ -1847,38 +1800,38 @@ void GSLocalMemory::ReadTextureBlock24Z(DWORD bp, BYTE* dst, int dstpitch, const } } -void GSLocalMemory::ReadTextureBlock16Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock16Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; - ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr(bp), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } -void GSLocalMemory::ReadTextureBlock16SZ(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock16SZ(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - __declspec(align(16)) WORD block[16 * 8]; + __declspec(align(16)) uint16 block[16 * 8]; - ReadBlock16(BlockPtr(bp), (BYTE*)block, sizeof(block) / 8); + ReadBlock16(BlockPtr(bp), (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } /////////////////// -void GSLocalMemory::ReadTexture(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) { readTexture rtx = m_psm[TEX0.PSM].rtx; readTexel rt = m_psm[TEX0.PSM].rt; - CSize bs = m_psm[TEX0.PSM].bs; + GSVector2i bs = m_psm[TEX0.PSM].bs; - if(r.Width() < bs.cx || r.Height() < bs.cy - || (r.left & (bs.cx-1)) || (r.top & (bs.cy-1)) - || (r.right & (bs.cx-1)) || (r.bottom & (bs.cy-1)) - || (CLAMP.WMS == 3) || (CLAMP.WMT == 3)) + if(r.width() < bs.x || r.height() < bs.y + || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) + || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1)) + || CLAMP.WMS == 3 || CLAMP.WMT == 3) { - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); } else { @@ -1886,17 +1839,17 @@ void GSLocalMemory::ReadTexture(const CRect& r, BYTE* dst, int dstpitch, const G } } -void GSLocalMemory::ReadTextureNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) { readTexture rtx = m_psm[TEX0.PSM].rtx; readTexel rt = m_psm[TEX0.PSM].rt; - CSize bs = m_psm[TEX0.PSM].bs; + GSVector2i bs = m_psm[TEX0.PSM].bs; - if(r.Width() < bs.cx || r.Height() < bs.cy - || (r.left & (bs.cx-1)) || (r.top & (bs.cy-1)) - || (r.right & (bs.cx-1)) || (r.bottom & (bs.cy-1))) + if(r.width() < bs.x || r.height() < bs.y + || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) + || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1))) { - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); } else { @@ -1905,7 +1858,7 @@ void GSLocalMemory::ReadTextureNC(const CRect& r, BYTE* dst, int dstpitch, const } /////////////////// -void GSLocalMemory::ReadTexture16NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(16, 8, 16) { @@ -1914,7 +1867,7 @@ void GSLocalMemory::ReadTexture16NP(const CRect& r, BYTE* dst, int dstpitch, con FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture16SNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16SNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(16, 8, 16) { @@ -1923,9 +1876,9 @@ void GSLocalMemory::ReadTexture16SNP(const CRect& r, BYTE* dst, int dstpitch, co FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -1939,11 +1892,11 @@ void GSLocalMemory::ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, cons { ASSERT(TEX0.CPSM == PSM_PSMCT16 || TEX0.CPSM == PSM_PSMCT16S); - __declspec(align(16)) BYTE block[16 * 16]; + __declspec(align(16)) uint8 block[16 * 16]; FOREACH_BLOCK_START(16, 16, 16) { - ReadBlock8(BlockPtr8(x, y, bp, bw), (BYTE*)block, sizeof(block) / 16); + ReadBlock8(BlockPtr8(x, y, bp, bw), (uint8*)block, sizeof(block) / 16); ExpandBlock8_16(block, dst, dstpitch, pal); } @@ -1951,7 +1904,7 @@ void GSLocalMemory::ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, cons } } -void GSLocalMemory::ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { const UINT64* pal = m_clut; @@ -1967,11 +1920,11 @@ void GSLocalMemory::ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, cons { ASSERT(TEX0.CPSM == PSM_PSMCT16 || TEX0.CPSM == PSM_PSMCT16S); - __declspec(align(16)) BYTE block[(32 / 2) * 16]; + __declspec(align(16)) uint8 block[(32 / 2) * 16]; FOREACH_BLOCK_START(32, 16, 16) { - ReadBlock4(BlockPtr4(x, y, bp, bw), (BYTE*)block, sizeof(block) / 16); + ReadBlock4(BlockPtr4(x, y, bp, bw), (uint8*)block, sizeof(block) / 16); ExpandBlock4_16(block, dst, dstpitch, pal); } @@ -1979,9 +1932,9 @@ void GSLocalMemory::ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, cons } } -void GSLocalMemory::ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8HNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -1995,11 +1948,11 @@ void GSLocalMemory::ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, con { ASSERT(TEX0.CPSM == PSM_PSMCT16 || TEX0.CPSM == PSM_PSMCT16S); - __declspec(align(16)) DWORD block[8 * 8]; + __declspec(align(16)) uint32 block[8 * 8]; FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock8H_16(block, dst, dstpitch, pal); } @@ -2007,9 +1960,9 @@ void GSLocalMemory::ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, con } } -void GSLocalMemory::ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HLNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -2023,11 +1976,11 @@ void GSLocalMemory::ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, co { ASSERT(TEX0.CPSM == PSM_PSMCT16 || TEX0.CPSM == PSM_PSMCT16S); - __declspec(align(16)) DWORD block[8 * 8]; + __declspec(align(16)) uint32 block[8 * 8]; FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock4HL_16(block, dst, dstpitch, pal); } @@ -2035,9 +1988,9 @@ void GSLocalMemory::ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, co } } -void GSLocalMemory::ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HHNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const DWORD* pal = m_clut; + const uint32* pal = m_clut; if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { @@ -2051,11 +2004,11 @@ void GSLocalMemory::ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, co { ASSERT(TEX0.CPSM == PSM_PSMCT16 || TEX0.CPSM == PSM_PSMCT16S); - __declspec(align(16)) DWORD block[8 * 8]; + __declspec(align(16)) uint32 block[8 * 8]; FOREACH_BLOCK_START(8, 8, 16) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (BYTE*)block, sizeof(block) / 8); + ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); ExpandBlock4HH_16(block, dst, dstpitch, pal); } @@ -2063,7 +2016,7 @@ void GSLocalMemory::ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, co } } -void GSLocalMemory::ReadTexture16ZNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16ZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(16, 8, 16) { @@ -2072,7 +2025,7 @@ void GSLocalMemory::ReadTexture16ZNP(const CRect& r, BYTE* dst, int dstpitch, co FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture16SZNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture16SZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(16, 8, 16) { @@ -2083,18 +2036,18 @@ void GSLocalMemory::ReadTexture16SZNP(const CRect& r, BYTE* dst, int dstpitch, c /////////////////// -void GSLocalMemory::ReadTextureNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) { readTexture rtx = m_psm[TEX0.PSM].rtxNP; readTexel rt = m_psm[TEX0.PSM].rtNP; - CSize bs = m_psm[TEX0.PSM].bs; + GSVector2i bs = m_psm[TEX0.PSM].bs; - if(r.Width() < bs.cx || r.Height() < bs.cy - || (r.left & (bs.cx-1)) || (r.top & (bs.cy-1)) - || (r.right & (bs.cx-1)) || (r.bottom & (bs.cy-1)) - || (CLAMP.WMS == 3) || (CLAMP.WMT == 3)) + if(r.width() < bs.x || r.height() < bs.y + || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) + || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1)) + || CLAMP.WMS == 3 || CLAMP.WMT == 3) { - DWORD psm = TEX0.PSM; + uint32 psm = TEX0.PSM; switch(psm) { @@ -2112,11 +2065,11 @@ void GSLocalMemory::ReadTextureNP(const CRect& r, BYTE* dst, int dstpitch, const default: case PSM_PSMCT32: case PSM_PSMCT24: - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); break; case PSM_PSMCT16: case PSM_PSMCT16S: - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); break; } } @@ -2126,17 +2079,17 @@ void GSLocalMemory::ReadTextureNP(const CRect& r, BYTE* dst, int dstpitch, const } } -void GSLocalMemory::ReadTextureNPNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTextureNPNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) { readTexture rtx = m_psm[TEX0.PSM].rtxNP; readTexel rt = m_psm[TEX0.PSM].rtNP; - CSize bs = m_psm[TEX0.PSM].bs; + GSVector2i bs = m_psm[TEX0.PSM].bs; - if(r.Width() < bs.cx || r.Height() < bs.cy - || (r.left & (bs.cx-1)) || (r.top & (bs.cy-1)) - || (r.right & (bs.cx-1)) || (r.bottom & (bs.cy-1))) + if(r.width() < bs.x || r.height() < bs.y + || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) + || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1))) { - DWORD psm = TEX0.PSM; + uint32 psm = TEX0.PSM; switch(psm) { @@ -2154,11 +2107,11 @@ void GSLocalMemory::ReadTextureNPNC(const CRect& r, BYTE* dst, int dstpitch, con default: case PSM_PSMCT32: case PSM_PSMCT24: - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); break; case PSM_PSMCT16: case PSM_PSMCT16S: - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); break; } } @@ -2170,7 +2123,7 @@ void GSLocalMemory::ReadTextureNPNC(const CRect& r, BYTE* dst, int dstpitch, con // 32/8 -void GSLocalMemory::ReadTexture8P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(16, 16, 8) { @@ -2179,7 +2132,7 @@ void GSLocalMemory::ReadTexture8P(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(32, 16, 8) { @@ -2188,7 +2141,7 @@ void GSLocalMemory::ReadTexture4P(const CRect& r, BYTE* dst, int dstpitch, const FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture8HP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture8HP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(8, 8, 8) { @@ -2197,7 +2150,7 @@ void GSLocalMemory::ReadTexture8HP(const CRect& r, BYTE* dst, int dstpitch, cons FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4HLP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HLP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(8, 8, 8) { @@ -2206,7 +2159,7 @@ void GSLocalMemory::ReadTexture4HLP(const CRect& r, BYTE* dst, int dstpitch, con FOREACH_BLOCK_END } -void GSLocalMemory::ReadTexture4HHP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTexture4HHP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { FOREACH_BLOCK_START(8, 8, 8) { @@ -2217,27 +2170,27 @@ void GSLocalMemory::ReadTexture4HHP(const CRect& r, BYTE* dst, int dstpitch, con // -void GSLocalMemory::ReadTextureBlock8P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock8P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock8(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock4P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock4P(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock8HP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock8HP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock8HP(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock4HLP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4HLP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock4HLP(BlockPtr(bp), dst, dstpitch); } -void GSLocalMemory::ReadTextureBlock4HHP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const +void GSLocalMemory::ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { ReadBlock4HHP(BlockPtr(bp), dst, dstpitch); } @@ -2245,18 +2198,20 @@ void GSLocalMemory::ReadTextureBlock4HHP(DWORD bp, BYTE* dst, int dstpitch, cons // template -void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx) +void GSLocalMemory::ReadTexture(const GSVector4i& r2, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx) { // TODO: this is a mess, make it more simple - DWORD wms = CLAMP.WMS, wmt = CLAMP.WMT; - DWORD minu = CLAMP.MINU, maxu = CLAMP.MAXU; - DWORD minv = CLAMP.MINV, maxv = CLAMP.MAXV; + GSVector4i r = r2; - CSize bs = m_psm[TEX0.PSM].bs; + uint32 wms = CLAMP.WMS, wmt = CLAMP.WMT; + uint32 minu = CLAMP.MINU, maxu = CLAMP.MAXU; + uint32 minv = CLAMP.MINV, maxv = CLAMP.MAXV; - int bsxm = bs.cx - 1; - int bsym = bs.cy - 1; + GSVector2i bs = m_psm[TEX0.PSM].bs; + + int bsxm = bs.x - 1; + int bsym = bs.y - 1; if(wms == 3 || wmt == 3) { @@ -2268,13 +2223,13 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE w = (w + bsxm) & ~bsxm; h = (h + bsym) & ~bsym; - if(w % bs.cx == 0 && maxu % bs.cx == 0 && h % bs.cy == 0 && maxv % bs.cy == 0) + if(w % bs.x == 0 && maxu % bs.x == 0 && h % bs.y == 0 && maxv % bs.y == 0) { // printf("!!! 1 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - (this->*rtx)(CRect(CPoint(maxu, maxv), CSize(w, h)), (BYTE*)buff, w * sizeof(T), TEX0, TEXA); + (this->*rtx)(GSVector4i(maxu, maxv, maxu + w, maxv + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); dst -= r.left * sizeof(T); @@ -2329,7 +2284,7 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE { int w = ((minu + 1) + bsxm) & ~bsxm; - if(w % bs.cx == 0 && maxu % bs.cx == 0) + if(w % bs.x == 0 && maxu % bs.x == 0) { // printf("!!! 2 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); int top = r.top & ~bsym; @@ -2339,7 +2294,7 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - (this->*rtx)(CRect(CPoint(maxu, top), CSize(w, h)), (BYTE*)buff, w * sizeof(T), TEX0, TEXA); + (this->*rtx)(GSVector4i(maxu, top, maxu + w, top + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); dst -= r.left * sizeof(T); @@ -2378,7 +2333,7 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE { int h = (minv + 1 + bsym) & ~bsym; - if(h % bs.cy == 0 && maxv % bs.cy == 0) + if(h % bs.y == 0 && maxv % bs.y == 0) { // printf("!!! 3 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); int left = r.left & ~bsxm; @@ -2388,13 +2343,13 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - (this->*rtx)(CRect(CPoint(left, maxv), CSize(w, h)), (BYTE*)buff, w * sizeof(T), TEX0, TEXA); + (this->*rtx)(GSVector4i(left, maxv, left + w, maxv + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); for(int y = r.top; y < r.bottom; y++, dst += dstpitch) { T* src = &buff[(y & minv) * w + (r.left - left)]; - memcpy(dst, src, sizeof(T) * r.Width()); + memcpy(dst, src, sizeof(T) * r.width()); } _aligned_free(buff); @@ -2425,8 +2380,8 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE { // find a block-aligned rect that fits between r and the region clamped area (if any) - CRect r1 = r; - CRect r2 = r; + GSVector4i r1 = r; + GSVector4i r2 = r; r1.left = (r1.left + bsxm) & ~bsxm; r1.top = (r1.top + bsym) & ~bsym; @@ -2445,11 +2400,11 @@ void GSLocalMemory::ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTE r2.bottom = (maxv + bsym) & ~bsym; } - CRect cr = r1 & r2; + GSVector4i cr = r1.rintersect(r2); bool aligned = ((DWORD_PTR)(dst + (cr.left - r.left) * sizeof(T)) & 0xf) == 0; - if(cr.left >= cr.right && cr.top >= cr.bottom || !aligned) + if(cr.rempty() || !aligned) { // TODO: expand r to block size, read into temp buffer, copy to r (like above) @@ -2469,7 +2424,7 @@ if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); for(int x = r.left, i = 0; x < r.right; x++, i++) ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - if(!cr.IsRectEmpty()) + if(!cr.rempty()) { (this->*rtx)(cr, dst + (cr.left - r.left) * sizeof(T), dstpitch, TEX0, TEXA); } @@ -2490,14 +2445,14 @@ if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); } template -void GSLocalMemory::ReadTextureNC(CRect r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx) +void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx) { - CSize bs = m_psm[TEX0.PSM].bs; + GSVector2i bs = m_psm[TEX0.PSM].bs; - int bsxm = bs.cx - 1; - int bsym = bs.cy - 1; + int bsxm = bs.x - 1; + int bsym = bs.y - 1; - CRect cr; + GSVector4i cr; cr.left = (r.left + bsxm) & ~bsxm; cr.top = (r.top + bsym) & ~bsym; @@ -2506,7 +2461,7 @@ void GSLocalMemory::ReadTextureNC(CRect r, BYTE* dst, int dstpitch, const GIFReg bool aligned = ((DWORD_PTR)(dst + (cr.left - r.left) * sizeof(T)) & 0xf) == 0; - if(cr.left >= cr.right && cr.top >= cr.bottom || !aligned) + if(cr.rempty() || !aligned) { // TODO: expand r to block size, read into temp buffer, copy to r (like above) @@ -2522,7 +2477,7 @@ if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); for(int x = r.left, i = 0; x < r.right; x++, i++) ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - if(!cr.IsRectEmpty()) + if(!cr.rempty()) (this->*rtx)(cr, dst + (cr.left - r.left) * sizeof(T), dstpitch, TEX0, TEXA); for(int y = cr.top; y < cr.bottom; y++, dst += dstpitch) @@ -2539,7 +2494,7 @@ if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); } } -HRESULT GSLocalMemory::SaveBMP(const string& fn, DWORD bp, DWORD bw, DWORD psm, int w, int h) +HRESULT GSLocalMemory::SaveBMP(const string& fn, uint32 bp, uint32 bw, uint32 psm, int w, int h) { int pitch = w * 4; int size = pitch * h; @@ -2557,15 +2512,15 @@ HRESULT GSLocalMemory::SaveBMP(const string& fn, DWORD bp, DWORD bw, DWORD psm, TEXA.TA0 = 0; TEXA.TA1 = 0x80; - // (this->*m_psm[TEX0.PSM].rtx)(CRect(0, 0, w, h), bits, pitch, TEX0, TEXA); + // (this->*m_psm[TEX0.PSM].rtx)(GSVector4i(0, 0, w, h), bits, pitch, TEX0, TEXA); readPixel rp = m_psm[psm].rp; - BYTE* p = (BYTE*)bits; + uint8* p = (uint8*)bits; for(int j = h-1; j >= 0; j--, p += pitch) for(int i = 0; i < w; i++) - ((DWORD*)p)[i] = (this->*rp)(i, j, TEX0.TBP0, TEX0.TBW); + ((uint32*)p)[i] = (this->*rp)(i, j, TEX0.TBP0, TEX0.TBW); if(FILE* fp = fopen(fn.c_str(), "wb")) { diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index fadd7ecee0..4f111289a3 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -32,19 +32,19 @@ class GSLocalMemory : public GSBlock { public: - typedef DWORD (*pixelAddress)(int x, int y, DWORD bp, DWORD bw); - typedef void (GSLocalMemory::*writePixel)(int x, int y, DWORD c, DWORD bp, DWORD bw); - typedef void (GSLocalMemory::*writeFrame)(int x, int y, DWORD c, DWORD bp, DWORD bw); - typedef DWORD (GSLocalMemory::*readPixel)(int x, int y, DWORD bp, DWORD bw) const; - typedef DWORD (GSLocalMemory::*readTexel)(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - typedef void (GSLocalMemory::*writePixelAddr)(DWORD addr, DWORD c); - typedef void (GSLocalMemory::*writeFrameAddr)(DWORD addr, DWORD c); - typedef DWORD (GSLocalMemory::*readPixelAddr)(DWORD addr) const; - typedef DWORD (GSLocalMemory::*readTexelAddr)(DWORD addr, const GIFRegTEXA& TEXA) const; - typedef void (GSLocalMemory::*writeImage)(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - typedef void (GSLocalMemory::*readImage)(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; - typedef void (GSLocalMemory::*readTexture)(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - typedef void (GSLocalMemory::*readTextureBlock)(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + typedef uint32 (*pixelAddress)(int x, int y, uint32 bp, uint32 bw); + typedef void (GSLocalMemory::*writePixel)(int x, int y, uint32 c, uint32 bp, uint32 bw); + typedef void (GSLocalMemory::*writeFrame)(int x, int y, uint32 c, uint32 bp, uint32 bw); + typedef uint32 (GSLocalMemory::*readPixel)(int x, int y, uint32 bp, uint32 bw) const; + typedef uint32 (GSLocalMemory::*readTexel)(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + typedef void (GSLocalMemory::*writePixelAddr)(uint32 addr, uint32 c); + typedef void (GSLocalMemory::*writeFrameAddr)(uint32 addr, uint32 c); + typedef uint32 (GSLocalMemory::*readPixelAddr)(uint32 addr) const; + typedef uint32 (GSLocalMemory::*readTexelAddr)(uint32 addr, const GIFRegTEXA& TEXA) const; + typedef void (GSLocalMemory::*writeImage)(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + typedef void (GSLocalMemory::*readImage)(int& tx, int& ty, uint8* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; + typedef void (GSLocalMemory::*readTexture)(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + typedef void (GSLocalMemory::*readTextureBlock)(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; typedef union { @@ -62,14 +62,14 @@ public: readImage ri; readTexture rtx, rtxNP, rtxP; readTextureBlock rtxb, rtxbP; - WORD bpp, trbpp; - DWORD pal; - CSize bs, pgs; + short bpp, trbpp; + int pal; + GSVector2i bs, pgs; int* rowOffset[8]; int* blockOffset; }; - BYTE dummy[128]; + uint8 dummy[128]; } psm_t; @@ -77,7 +77,7 @@ public: static const int m_vmsize = 1024 * 1024 * 4; - union {BYTE* m_vm8; WORD* m_vm16; DWORD* m_vm32;}; + union {uint8* m_vm8; uint16* m_vm16; uint32* m_vm32;}; GSClut m_clut; @@ -85,7 +85,7 @@ public: { GSVector4i row[2048]; // 0 | 0 | 0 | 0 int* col[4]; // x | x+1 | x+2 | x+3 - DWORD hash; + uint32 hash; }; struct Offset4 @@ -94,18 +94,18 @@ public: GSVector2i row[2048]; // f yn | z yn (n = 0 1 2 ...) GSVector2i col[512]; // f xn | z xn (n = 0 4 8 ...) - DWORD hash; + uint32 hash; }; protected: - static DWORD pageOffset32[32][32][64]; - static DWORD pageOffset32Z[32][32][64]; - static DWORD pageOffset16[32][64][64]; - static DWORD pageOffset16S[32][64][64]; - static DWORD pageOffset16Z[32][64][64]; - static DWORD pageOffset16SZ[32][64][64]; - static DWORD pageOffset8[32][64][128]; - static DWORD pageOffset4[32][128][128]; + static uint32 pageOffset32[32][32][64]; + static uint32 pageOffset32Z[32][32][64]; + static uint32 pageOffset16[32][64][64]; + static uint32 pageOffset16S[32][64][64]; + static uint32 pageOffset16Z[32][64][64]; + static uint32 pageOffset16SZ[32][64][64]; + static uint32 pageOffset8[32][64][128]; + static uint32 pageOffset4[32][128][128]; static int rowOffset32[2048]; static int rowOffset32Z[2048]; @@ -125,12 +125,12 @@ protected: static int blockOffset8[256]; static int blockOffset4[256]; - __forceinline static DWORD Expand24To32(DWORD c, const GIFRegTEXA& TEXA) + __forceinline static uint32 Expand24To32(uint32 c, const GIFRegTEXA& TEXA) { return (((!TEXA.AEM | (c & 0xffffff)) ? TEXA.TA0 : 0) << 24) | (c & 0xffffff); } - __forceinline static DWORD Expand16To32(WORD c, const GIFRegTEXA& TEXA) + __forceinline static uint32 Expand16To32(uint16 c, const GIFRegTEXA& TEXA) { return (((c & 0x8000) ? TEXA.TA1 : (!TEXA.AEM | c) ? TEXA.TA0 : 0) << 24) | ((c & 0x7c00) << 9) | ((c & 0x03e0) << 6) | ((c & 0x001f) << 3); } @@ -141,619 +141,619 @@ protected: // - hash_map m_omap; - hash_map m_o4map; + hash_map m_omap; + hash_map m_o4map; public: GSLocalMemory(); virtual ~GSLocalMemory(); - Offset* GetOffset(DWORD bp, DWORD bw, DWORD psm); + Offset* GetOffset(uint32 bp, uint32 bw, uint32 psm); Offset4* GetOffset4(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF); // address - static DWORD BlockNumber32(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber32(int x, int y, uint32 bp, uint32 bw) { return bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable32[(y >> 3) & 3][(x >> 3) & 7]; } - static DWORD BlockNumber16(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber16(int x, int y, uint32 bp, uint32 bw) { return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD BlockNumber16S(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber16S(int x, int y, uint32 bp, uint32 bw) { return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16S[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD BlockNumber8(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber8(int x, int y, uint32 bp, uint32 bw) { ASSERT((bw & 1) == 0); return bp + ((y >> 1) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f) + blockTable8[(y >> 4) & 3][(x >> 4) & 7]; } - static DWORD BlockNumber4(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber4(int x, int y, uint32 bp, uint32 bw) { ASSERT((bw & 1) == 0); return bp + ((y >> 2) & ~0x1f) * (bw >> 1) + ((x >> 2) & ~0x1f) + blockTable4[(y >> 4) & 7][(x >> 5) & 3]; } - static DWORD BlockNumber32Z(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber32Z(int x, int y, uint32 bp, uint32 bw) { return bp + (y & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable32Z[(y >> 3) & 3][(x >> 3) & 7]; } - static DWORD BlockNumber16Z(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber16Z(int x, int y, uint32 bp, uint32 bw) { return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16Z[(y >> 3) & 7][(x >> 4) & 3]; } - static DWORD BlockNumber16SZ(int x, int y, DWORD bp, DWORD bw) + static uint32 BlockNumber16SZ(int x, int y, uint32 bp, uint32 bw) { return bp + ((y >> 1) & ~0x1f) * bw + ((x >> 1) & ~0x1f) + blockTable16SZ[(y >> 3) & 7][(x >> 4) & 3]; } - BYTE* BlockPtr(DWORD bp) const + uint8* BlockPtr(uint32 bp) const { ASSERT(bp < 16384); return &m_vm8[bp << 8]; } - BYTE* BlockPtr32(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr32(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber32(x, y, bp, bw) << 8]; } - BYTE* BlockPtr16(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr16(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber16(x, y, bp, bw) << 8]; } - BYTE* BlockPtr16S(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr16S(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber16S(x, y, bp, bw) << 8]; } - BYTE* BlockPtr8(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr8(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber8(x, y, bp, bw) << 8]; } - BYTE* BlockPtr4(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr4(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber4(x, y, bp, bw) << 8]; } - BYTE* BlockPtr32Z(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr32Z(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber32Z(x, y, bp, bw) << 8]; } - BYTE* BlockPtr16Z(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr16Z(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber16Z(x, y, bp, bw) << 8]; } - BYTE* BlockPtr16SZ(int x, int y, DWORD bp, DWORD bw) const + uint8* BlockPtr16SZ(int x, int y, uint32 bp, uint32 bw) const { return &m_vm8[BlockNumber16SZ(x, y, bp, bw) << 8]; } - static DWORD PixelAddressOrg32(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg32(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber32(x, y, bp, bw) << 6) + columnTable32[y & 7][x & 7]; } - static DWORD PixelAddressOrg16(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg16(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber16(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } - static DWORD PixelAddressOrg16S(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg16S(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber16S(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } - static DWORD PixelAddressOrg8(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg8(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber8(x, y, bp, bw) << 8) + columnTable8[y & 15][x & 15]; } - static DWORD PixelAddressOrg4(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg4(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber4(x, y, bp, bw) << 9) + columnTable4[y & 15][x & 31]; } - static DWORD PixelAddressOrg32Z(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg32Z(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber32Z(x, y, bp, bw) << 6) + columnTable32[y & 7][x & 7]; } - static DWORD PixelAddressOrg16Z(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg16Z(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber16Z(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } - static DWORD PixelAddressOrg16SZ(int x, int y, DWORD bp, DWORD bw) + static uint32 PixelAddressOrg16SZ(int x, int y, uint32 bp, uint32 bw) { return (BlockNumber16SZ(x, y, bp, bw) << 7) + columnTable16[y & 7][x & 15]; } - static __forceinline DWORD PixelAddress32(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress32(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 5) * bw + (x >> 6); - DWORD word = (page << 11) + pageOffset32[bp & 0x1f][y & 0x1f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 5) * bw + (x >> 6); + uint32 word = (page << 11) + pageOffset32[bp & 0x1f][y & 0x1f][x & 0x3f]; return word; } - static __forceinline DWORD PixelAddress16(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress16(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 6) * bw + (x >> 6); - DWORD word = (page << 12) + pageOffset16[bp & 0x1f][y & 0x3f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 6) * bw + (x >> 6); + uint32 word = (page << 12) + pageOffset16[bp & 0x1f][y & 0x3f][x & 0x3f]; return word; } - static __forceinline DWORD PixelAddress16S(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress16S(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 6) * bw + (x >> 6); - DWORD word = (page << 12) + pageOffset16S[bp & 0x1f][y & 0x3f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 6) * bw + (x >> 6); + uint32 word = (page << 12) + pageOffset16S[bp & 0x1f][y & 0x3f][x & 0x3f]; return word; } - static __forceinline DWORD PixelAddress8(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress8(int x, int y, uint32 bp, uint32 bw) { ASSERT((bw & 1) == 0); - DWORD page = (bp >> 5) + (y >> 6) * (bw >> 1) + (x >> 7); - DWORD word = (page << 13) + pageOffset8[bp & 0x1f][y & 0x3f][x & 0x7f]; + uint32 page = (bp >> 5) + (y >> 6) * (bw >> 1) + (x >> 7); + uint32 word = (page << 13) + pageOffset8[bp & 0x1f][y & 0x3f][x & 0x7f]; return word; } - static __forceinline DWORD PixelAddress4(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress4(int x, int y, uint32 bp, uint32 bw) { ASSERT((bw & 1) == 0); - DWORD page = (bp >> 5) + (y >> 7) * (bw >> 1) + (x >> 7); - DWORD word = (page << 14) + pageOffset4[bp & 0x1f][y & 0x7f][x & 0x7f]; + uint32 page = (bp >> 5) + (y >> 7) * (bw >> 1) + (x >> 7); + uint32 word = (page << 14) + pageOffset4[bp & 0x1f][y & 0x7f][x & 0x7f]; return word; } - static __forceinline DWORD PixelAddress32Z(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress32Z(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 5) * bw + (x >> 6); - DWORD word = (page << 11) + pageOffset32Z[bp & 0x1f][y & 0x1f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 5) * bw + (x >> 6); + uint32 word = (page << 11) + pageOffset32Z[bp & 0x1f][y & 0x1f][x & 0x3f]; return word; } - static __forceinline DWORD PixelAddress16Z(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress16Z(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 6) * bw + (x >> 6); - DWORD word = (page << 12) + pageOffset16Z[bp & 0x1f][y & 0x3f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 6) * bw + (x >> 6); + uint32 word = (page << 12) + pageOffset16Z[bp & 0x1f][y & 0x3f][x & 0x3f]; return word; } - static __forceinline DWORD PixelAddress16SZ(int x, int y, DWORD bp, DWORD bw) + static __forceinline uint32 PixelAddress16SZ(int x, int y, uint32 bp, uint32 bw) { - DWORD page = (bp >> 5) + (y >> 6) * bw + (x >> 6); - DWORD word = (page << 12) + pageOffset16SZ[bp & 0x1f][y & 0x3f][x & 0x3f]; + uint32 page = (bp >> 5) + (y >> 6) * bw + (x >> 6); + uint32 word = (page << 12) + pageOffset16SZ[bp & 0x1f][y & 0x3f][x & 0x3f]; return word; } // pixel R/W - __forceinline DWORD ReadPixel32(DWORD addr) const + __forceinline uint32 ReadPixel32(uint32 addr) const { return m_vm32[addr]; } - __forceinline DWORD ReadPixel24(DWORD addr) const + __forceinline uint32 ReadPixel24(uint32 addr) const { return m_vm32[addr] & 0x00ffffff; } - __forceinline DWORD ReadPixel16(DWORD addr) const + __forceinline uint32 ReadPixel16(uint32 addr) const { - return (DWORD)m_vm16[addr]; + return (uint32)m_vm16[addr]; } - __forceinline DWORD ReadPixel8(DWORD addr) const + __forceinline uint32 ReadPixel8(uint32 addr) const { - return (DWORD)m_vm8[addr]; + return (uint32)m_vm8[addr]; } - __forceinline DWORD ReadPixel4(DWORD addr) const + __forceinline uint32 ReadPixel4(uint32 addr) const { return (m_vm8[addr >> 1] >> ((addr & 1) << 2)) & 0x0f; } - __forceinline DWORD ReadPixel8H(DWORD addr) const + __forceinline uint32 ReadPixel8H(uint32 addr) const { return m_vm32[addr] >> 24; } - __forceinline DWORD ReadPixel4HL(DWORD addr) const + __forceinline uint32 ReadPixel4HL(uint32 addr) const { return (m_vm32[addr] >> 24) & 0x0f; } - __forceinline DWORD ReadPixel4HH(DWORD addr) const + __forceinline uint32 ReadPixel4HH(uint32 addr) const { return (m_vm32[addr] >> 28) & 0x0f; } - __forceinline DWORD ReadFrame24(DWORD addr) const + __forceinline uint32 ReadFrame24(uint32 addr) const { return 0x80000000 | (m_vm32[addr] & 0xffffff); } - __forceinline DWORD ReadFrame16(DWORD addr) const + __forceinline uint32 ReadFrame16(uint32 addr) const { - DWORD c = (DWORD)m_vm16[addr]; + uint32 c = (uint32)m_vm16[addr]; return ((c & 0x8000) << 16) | ((c & 0x7c00) << 9) | ((c & 0x03e0) << 6) | ((c & 0x001f) << 3); } - __forceinline DWORD ReadPixel32(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel32(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel32(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadPixel24(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel24(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel24(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadPixel16(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel16(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel16(PixelAddress16(x, y, bp, bw)); } - __forceinline DWORD ReadPixel16S(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel16S(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel16(PixelAddress16S(x, y, bp, bw)); } - __forceinline DWORD ReadPixel8(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel8(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel8(PixelAddress8(x, y, bp, bw)); } - __forceinline DWORD ReadPixel4(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel4(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel4(PixelAddress4(x, y, bp, bw)); } - __forceinline DWORD ReadPixel8H(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel8H(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel8H(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadPixel4HL(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel4HL(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel4HL(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadPixel4HH(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel4HH(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel4HH(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadPixel32Z(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel32Z(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel32(PixelAddress32Z(x, y, bp, bw)); } - __forceinline DWORD ReadPixel24Z(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel24Z(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel24(PixelAddress32Z(x, y, bp, bw)); } - __forceinline DWORD ReadPixel16Z(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel16Z(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel16(PixelAddress16Z(x, y, bp, bw)); } - __forceinline DWORD ReadPixel16SZ(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadPixel16SZ(int x, int y, uint32 bp, uint32 bw) const { return ReadPixel16(PixelAddress16SZ(x, y, bp, bw)); } - __forceinline DWORD ReadFrame24(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame24(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame24(PixelAddress32(x, y, bp, bw)); } - __forceinline DWORD ReadFrame16(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame16(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame16(PixelAddress16(x, y, bp, bw)); } - __forceinline DWORD ReadFrame16S(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame16S(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame16(PixelAddress16S(x, y, bp, bw)); } - __forceinline DWORD ReadFrame24Z(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame24Z(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame24(PixelAddress32Z(x, y, bp, bw)); } - __forceinline DWORD ReadFrame16Z(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame16Z(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame16(PixelAddress16Z(x, y, bp, bw)); } - __forceinline DWORD ReadFrame16SZ(int x, int y, DWORD bp, DWORD bw) const + __forceinline uint32 ReadFrame16SZ(int x, int y, uint32 bp, uint32 bw) const { return ReadFrame16(PixelAddress16SZ(x, y, bp, bw)); } - __forceinline void WritePixel32(DWORD addr, DWORD c) + __forceinline void WritePixel32(uint32 addr, uint32 c) { m_vm32[addr] = c; } - __forceinline void WritePixel24(DWORD addr, DWORD c) + __forceinline void WritePixel24(uint32 addr, uint32 c) { m_vm32[addr] = (m_vm32[addr] & 0xff000000) | (c & 0x00ffffff); } - __forceinline void WritePixel16(DWORD addr, DWORD c) + __forceinline void WritePixel16(uint32 addr, uint32 c) { - m_vm16[addr] = (WORD)c; + m_vm16[addr] = (uint16)c; } - __forceinline void WritePixel8(DWORD addr, DWORD c) + __forceinline void WritePixel8(uint32 addr, uint32 c) { - m_vm8[addr] = (BYTE)c; + m_vm8[addr] = (uint8)c; } - __forceinline void WritePixel4(DWORD addr, DWORD c) + __forceinline void WritePixel4(uint32 addr, uint32 c) { int shift = (addr & 1) << 2; addr >>= 1; - m_vm8[addr] = (BYTE)((m_vm8[addr] & (0xf0 >> shift)) | ((c & 0x0f) << shift)); + m_vm8[addr] = (uint8)((m_vm8[addr] & (0xf0 >> shift)) | ((c & 0x0f) << shift)); } - __forceinline void WritePixel8H(DWORD addr, DWORD c) + __forceinline void WritePixel8H(uint32 addr, uint32 c) { m_vm32[addr] = (m_vm32[addr] & 0x00ffffff) | (c << 24); } - __forceinline void WritePixel4HL(DWORD addr, DWORD c) + __forceinline void WritePixel4HL(uint32 addr, uint32 c) { m_vm32[addr] = (m_vm32[addr] & 0xf0ffffff) | ((c & 0x0f) << 24); } - __forceinline void WritePixel4HH(DWORD addr, DWORD c) + __forceinline void WritePixel4HH(uint32 addr, uint32 c) { m_vm32[addr] = (m_vm32[addr] & 0x0fffffff) | ((c & 0x0f) << 28); } - __forceinline void WriteFrame16(DWORD addr, DWORD c) + __forceinline void WriteFrame16(uint32 addr, uint32 c) { - DWORD rb = c & 0x00f800f8; - DWORD ga = c & 0x8000f800; + uint32 rb = c & 0x00f800f8; + uint32 ga = c & 0x8000f800; WritePixel16(addr, (ga >> 16) | (rb >> 9) | (ga >> 6) | (rb >> 3)); } - __forceinline void WritePixel32(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel32(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel32(PixelAddress32(x, y, bp, bw), c); } - __forceinline void WritePixel24(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel24(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel24(PixelAddress32(x, y, bp, bw), c); } - __forceinline void WritePixel16(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel16(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel16(PixelAddress16(x, y, bp, bw), c); } - __forceinline void WritePixel16S(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel16S(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel16(PixelAddress16S(x, y, bp, bw), c); } - __forceinline void WritePixel8(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel8(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel8(PixelAddress8(x, y, bp, bw), c); } - __forceinline void WritePixel4(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel4(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel4(PixelAddress4(x, y, bp, bw), c); } - __forceinline void WritePixel8H(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel8H(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel8H(PixelAddress32(x, y, bp, bw), c); } - __forceinline void WritePixel4HL(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel4HL(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel4HL(PixelAddress32(x, y, bp, bw), c); } - __forceinline void WritePixel4HH(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel4HH(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel4HH(PixelAddress32(x, y, bp, bw), c); } - __forceinline void WritePixel32Z(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel32Z(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel32(PixelAddress32Z(x, y, bp, bw), c); } - __forceinline void WritePixel24Z(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel24Z(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel24(PixelAddress32Z(x, y, bp, bw), c); } - __forceinline void WritePixel16Z(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel16Z(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel16(PixelAddress16Z(x, y, bp, bw), c); } - __forceinline void WritePixel16SZ(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WritePixel16SZ(int x, int y, uint32 c, uint32 bp, uint32 bw) { WritePixel16(PixelAddress16SZ(x, y, bp, bw), c); } - __forceinline void WriteFrame16(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WriteFrame16(int x, int y, uint32 c, uint32 bp, uint32 bw) { WriteFrame16(PixelAddress16(x, y, bp, bw), c); } - __forceinline void WriteFrame16S(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WriteFrame16S(int x, int y, uint32 c, uint32 bp, uint32 bw) { WriteFrame16(PixelAddress16S(x, y, bp, bw), c); } - __forceinline void WriteFrame16Z(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WriteFrame16Z(int x, int y, uint32 c, uint32 bp, uint32 bw) { WriteFrame16(PixelAddress16Z(x, y, bp, bw), c); } - __forceinline void WriteFrame16SZ(int x, int y, DWORD c, DWORD bp, DWORD bw) + __forceinline void WriteFrame16SZ(int x, int y, uint32 c, uint32 bp, uint32 bw) { WriteFrame16(PixelAddress16SZ(x, y, bp, bw), c); } - __forceinline DWORD ReadTexel32(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel32(uint32 addr, const GIFRegTEXA& TEXA) const { return m_vm32[addr]; } - __forceinline DWORD ReadTexel24(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel24(uint32 addr, const GIFRegTEXA& TEXA) const { return Expand24To32(m_vm32[addr], TEXA); } - __forceinline DWORD ReadTexel16(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16(uint32 addr, const GIFRegTEXA& TEXA) const { return Expand16To32(m_vm16[addr], TEXA); } - __forceinline DWORD ReadTexel8(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel8(uint32 addr, const GIFRegTEXA& TEXA) const { return m_clut[ReadPixel8(addr)]; } - __forceinline DWORD ReadTexel4(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4(uint32 addr, const GIFRegTEXA& TEXA) const { return m_clut[ReadPixel4(addr)]; } - __forceinline DWORD ReadTexel8H(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel8H(uint32 addr, const GIFRegTEXA& TEXA) const { return m_clut[ReadPixel8H(addr)]; } - __forceinline DWORD ReadTexel4HL(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4HL(uint32 addr, const GIFRegTEXA& TEXA) const { return m_clut[ReadPixel4HL(addr)]; } - __forceinline DWORD ReadTexel4HH(DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4HH(uint32 addr, const GIFRegTEXA& TEXA) const { return m_clut[ReadPixel4HH(addr)]; } - __forceinline DWORD ReadTexel32(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel32(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel32(PixelAddress32(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel24(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel24(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel24(PixelAddress32(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel16(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel16(PixelAddress16(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel16S(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16S(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel16(PixelAddress16S(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel8(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel8(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel8(PixelAddress8(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel4(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel4(PixelAddress4(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel8H(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel8H(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel8H(PixelAddress32(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel4HL(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4HL(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel4HL(PixelAddress32(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel4HH(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel4HH(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel4HH(PixelAddress32(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel32Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel32Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel32(PixelAddress32Z(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel24Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel24Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel24(PixelAddress32Z(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel16Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16Z(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel16(PixelAddress16Z(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel16SZ(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16SZ(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadTexel16(PixelAddress16SZ(x, y, TEX0.TBP0, TEX0.TBW), TEXA); } - __forceinline DWORD ReadTexel16NP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16NP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadPixel16(x, y, TEX0.TBP0, TEX0.TBW); } - __forceinline DWORD ReadTexel16SNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16SNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadPixel16S(x, y, TEX0.TBP0, TEX0.TBW); } - __forceinline DWORD ReadTexel16ZNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16ZNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadPixel16Z(x, y, TEX0.TBP0, TEX0.TBW); } - __forceinline DWORD ReadTexel16SZNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexel16SZNP(int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { return ReadPixel16SZ(x, y, TEX0.TBP0, TEX0.TBW); } // - __forceinline DWORD PixelAddressX(int PSM, int x, int y, DWORD bp, DWORD bw) + __forceinline uint32 PixelAddressX(int PSM, int x, int y, uint32 bp, uint32 bw) { switch(PSM) { @@ -774,7 +774,7 @@ public: } } - __forceinline DWORD ReadPixelX(int PSM, DWORD addr) const + __forceinline uint32 ReadPixelX(int PSM, uint32 addr) const { switch(PSM) { @@ -795,7 +795,7 @@ public: } } - __forceinline DWORD ReadFrameX(int PSM, DWORD addr) const + __forceinline uint32 ReadFrameX(int PSM, uint32 addr) const { switch(PSM) { @@ -811,7 +811,7 @@ public: } } - __forceinline DWORD ReadTexelX(int PSM, DWORD addr, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexelX(int PSM, uint32 addr, const GIFRegTEXA& TEXA) const { switch(PSM) { @@ -832,7 +832,7 @@ public: } } - __forceinline DWORD ReadTexelX(int PSM, int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const + __forceinline uint32 ReadTexelX(int PSM, int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { switch(PSM) { @@ -853,7 +853,7 @@ public: } } - __forceinline void WritePixelX(int PSM, DWORD addr, DWORD c) + __forceinline void WritePixelX(int PSM, uint32 addr, uint32 c) { switch(PSM) { @@ -874,7 +874,7 @@ public: } } - __forceinline void WriteFrameX(int PSM, DWORD addr, DWORD c) + __forceinline void WriteFrameX(int PSM, uint32 addr, uint32 c) { switch(PSM) { @@ -892,108 +892,108 @@ public: // FillRect - bool FillRect(const GSVector4i& r, DWORD c, DWORD psm, DWORD bp, DWORD bw); + bool FillRect(const GSVector4i& r, uint32 c, uint32 psm, uint32 bp, uint32 bw); // template - void WriteImageColumn(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); + void WriteImageColumn(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); template - void WriteImageBlock(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); + void WriteImageBlock(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); template - void WriteImageLeftRight(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); + void WriteImageLeftRight(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); template - void WriteImageTopBottom(int l, int r, int y, int h, BYTE* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); + void WriteImageTopBottom(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); template - void WriteImage(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImage24(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImage8H(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImage4HL(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImage4HH(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImage24Z(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); - void WriteImageX(int& tx, int& ty, BYTE* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage24(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage8H(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage4HL(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage4HH(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImage24Z(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); + void WriteImageX(int& tx, int& ty, uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG); // TODO: ReadImage32/24/... - void ReadImageX(int& tx, int& ty, BYTE* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; + void ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const; // * => 32 - void ReadTexture32(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture24(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16S(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture8(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture8H(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HL(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HH(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture32Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture24Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16Z(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16SZ(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture32(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture24(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16S(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8H(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HL(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HH(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture32Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture24Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16SZ(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - void ReadTextureNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - void ReadTextureBlock32(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock24(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock16(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock16S(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock8(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock8H(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4HL(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4HH(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock32Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock24Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock16Z(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock16SZ(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock32(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16S(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8H(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HL(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HH(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock32Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock24Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock16SZ(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; // * => 32/16 - void ReadTexture16NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16SNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture8NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4NP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture8HNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HLNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HHNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16ZNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture16SZNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16SNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8HNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HLNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HHNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16ZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture16SZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTextureNP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - void ReadTextureNPNC(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTextureNPNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); // pal ? 8 : 32 - void ReadTexture8P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4P(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture8HP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HLP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture4HHP(const CRect& r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture8HP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HLP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; + void ReadTexture4HHP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock8P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4P(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock8HP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4HLP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; - void ReadTextureBlock4HHP(DWORD bp, BYTE* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock8HP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HLP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; + void ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; // - static DWORD m_xtbl[1024], m_ytbl[1024]; + static uint32 m_xtbl[1024], m_ytbl[1024]; - template void ReadTexture(CRect r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx); - template void ReadTextureNC(CRect r, BYTE* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx); + template void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx); + template void ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx); // - HRESULT SaveBMP(const string& fn, DWORD bp, DWORD bw, DWORD psm, int w, int h); + HRESULT SaveBMP(const string& fn, uint32 bp, uint32 bw, uint32 psm, int w, int h); }; #pragma warning(default: 4244) \ No newline at end of file diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index 17795fd9c5..b3d0a640a4 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -94,7 +94,7 @@ void GSRasterizer::DrawPoint(const GSVertexSW* v, const GSVector4i& scissor) GSVector4i p(v->p); - if(scissor.x <= p.x && p.x < scissor.z && scissor.y <= p.y && p.y < scissor.w) + if(scissor.left <= p.x && p.x < scissor.right && scissor.top <= p.y && p.y < scissor.bottom) { if((p.y % m_threads) == m_id) { @@ -153,7 +153,7 @@ void GSRasterizer::DrawLine(const GSVertexSW* v, const GSVector4i& scissor) GSVector4i p(l.p); - if(scissor.y <= p.y && p.y < scissor.w) + if(scissor.top <= p.y && p.y < scissor.bottom) { GSVertexSW dscan = dv / dv.p.xxxx(); @@ -546,29 +546,9 @@ void GSRasterizer::DrawSprite(const GSVertexSW* vertices, const GSVector4i& scis GSVector4i r(v[0].p.xyxy(v[1].p).ceil()); - int& top = r.y; - int& bottom = r.w; - - int& left = r.x; - int& right = r.z; - - #if _M_SSE >= 0x401 - - r = r.sat_i32(scissor); + r = r.rintersect(scissor); - if((r < r.zwzw()).mask() != 0x00ff) return; - - #else - - if(top < scissor.y) top = scissor.y; - if(bottom > scissor.w) bottom = scissor.w; - if(top >= bottom) return; - - if(left < scissor.x) left = scissor.x; - if(right > scissor.z) right = scissor.z; - if(left >= right) return; - - #endif + if(r.rempty()) return; GSVertexSW scan = v[0]; @@ -578,7 +558,7 @@ void GSRasterizer::DrawSprite(const GSVertexSW* vertices, const GSVector4i& scis { (m_ds->*m_dsf.sr)(r, scan); - m_stats.pixels += (r.z - r.x) * (r.w - r.y); + m_stats.pixels += r.width() * r.height(); } return; @@ -599,18 +579,18 @@ void GSRasterizer::DrawSprite(const GSVertexSW* vertices, const GSVector4i& scis dedge.t = (dv.t / dv.p.yyyy()).xyxy(zero).wyww(); dscan.t = (dv.t / dv.p.xxxx()).xyxy(zero).xwww(); - if(scan.p.y < (float)top) scan.t += dedge.t * ((float)top - scan.p.y); - if(scan.p.x < (float)left) scan.t += dscan.t * ((float)left - scan.p.x); + if(scan.p.y < (float)r.top) scan.t += dedge.t * ((float)r.top - scan.p.y); + if(scan.p.x < (float)r.left) scan.t += dscan.t * ((float)r.left - scan.p.x); m_dsf.ssp(v, dscan); - for(; top < bottom; top++, scan.t += dedge.t) + for(; r.top < r.bottom; r.top++, scan.t += dedge.t) { - if((top % m_threads) == m_id) + if((r.top % m_threads) == m_id) { - m_dsf.ssl(right, left, top, scan); + m_dsf.ssl(r.right, r.left, r.top, scan); - m_stats.pixels += right - left; + m_stats.pixels += r.width(); } } } @@ -681,7 +661,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS int xi = x >> 16; int xf = x & 0xffff; - if(scissor.x <= xi && xi < scissor.z && (xi % m_threads) == m_id) + if(scissor.left <= xi && xi < scissor.right && (xi % m_threads) == m_id) { m_stats.pixels++; @@ -709,7 +689,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS int xi = (x >> 16) + 1; int xf = x & 0xffff; - if(scissor.x <= xi && xi < scissor.z && (xi % m_threads) == m_id) + if(scissor.left <= xi && xi < scissor.right && (xi % m_threads) == m_id) { m_stats.pixels++; @@ -779,7 +759,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS int yi = y >> 16; int yf = y & 0xffff; - if(scissor.y <= yi && yi < scissor.w && (yi % m_threads) == m_id) + if(scissor.top <= yi && yi < scissor.bottom && (yi % m_threads) == m_id) { m_stats.pixels++; @@ -807,7 +787,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS int yi = (y >> 16) + 1; int yf = y & 0xffff; - if(scissor.y <= yi && yi < scissor.w && (yi % m_threads) == m_id) + if(scissor.top <= yi && yi < scissor.bottom && (yi % m_threads) == m_id) { m_stats.pixels++; @@ -835,29 +815,17 @@ GSRasterizerMT::GSRasterizerMT(IDrawScanline* ds, int id, int threads, long* syn : GSRasterizer(ds, id, threads) , m_sync(sync) , m_exit(false) - , m_ThreadId(0) - , m_hThread(NULL) , m_data(NULL) { if(id > 0) { - m_hThread = CreateThread(NULL, 0, StaticThreadProc, (LPVOID)this, 0, &m_ThreadId); + CreateThread(); } } GSRasterizerMT::~GSRasterizerMT() { - if(m_hThread != NULL) - { - m_exit = true; - - if(WaitForSingleObject(m_hThread, 5000) != WAIT_OBJECT_0) - { - TerminateThread(m_hThread, 1); - } - - CloseHandle(m_hThread); - } + m_exit = true; } void GSRasterizerMT::Draw(const GSRasterizerData* data) @@ -870,16 +838,11 @@ void GSRasterizerMT::Draw(const GSRasterizerData* data) { m_data = data; - InterlockedBitTestAndSet(m_sync, m_id); + _interlockedbittestandset(m_sync, m_id); } } -DWORD WINAPI GSRasterizerMT::StaticThreadProc(LPVOID lpParam) -{ - return ((GSRasterizerMT*)lpParam)->ThreadProc(); -} - -DWORD GSRasterizerMT::ThreadProc() +void GSRasterizerMT::ThreadProc() { // _mm_setcsr(MXCSR); @@ -889,15 +852,13 @@ DWORD GSRasterizerMT::ThreadProc() { __super::Draw(m_data); - InterlockedBitTestAndReset(m_sync, m_id); + _interlockedbittestandreset(m_sync, m_id); } else { _mm_pause(); } } - - return 0; } // diff --git a/plugins/GSdx/GSRasterizer.h b/plugins/GSdx/GSRasterizer.h index 651899b91a..00a2009baf 100644 --- a/plugins/GSdx/GSRasterizer.h +++ b/plugins/GSdx/GSRasterizer.h @@ -24,6 +24,7 @@ #include "GS.h" #include "GSVertexSW.h" #include "GSFunctionMap.h" +#include "GSThread.h" __declspec(align(16)) class GSRasterizerData { @@ -102,17 +103,13 @@ public: void PrintStats() {m_ds->PrintStats();} }; -class GSRasterizerMT : public GSRasterizer +class GSRasterizerMT : public GSRasterizer, private GSThread { long* m_sync; bool m_exit; - DWORD m_ThreadId; - HANDLE m_hThread; const GSRasterizerData* m_data; - static DWORD WINAPI StaticThreadProc(LPVOID lpParam); - - DWORD ThreadProc(); + void ThreadProc(); public: GSRasterizerMT(IDrawScanline* ds, int id, int threads, long* sync); diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index e25a0fe6f4..923b1e2003 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -95,7 +95,7 @@ public: GSWnd m_wnd; public: - GSRendererBase(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs) + GSRendererBase(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) : GSState(base, mt, irq) , m_osd(true) { @@ -125,8 +125,8 @@ protected: { bool en[2]; - CRect fr[2]; - CRect dr[2]; + GSVector4i fr[2]; + GSVector4i dr[2]; int baseline = INT_MAX; @@ -164,8 +164,8 @@ protected: if(samesrc && m_regs->PMODE.SLBG == 0 && m_regs->PMODE.MMOD == 1 && m_regs->PMODE.ALP == 0x80) { - if(fr[0] == fr[1] + CRect(0, 1, 0, 0) && dr[0] == dr[1] + CRect(0, 0, 0, 1) - || fr[1] == fr[0] + CRect(0, 1, 0, 0) && dr[1] == dr[0] + CRect(0, 0, 0, 1)) + if(fr[0].eq(fr[1] + GSVector4i(0, 1, 0, 0)) && dr[0].eq(dr[1] + GSVector4i(0, 0, 0, 1)) + || fr[1].eq(fr[0] + GSVector4i(0, 1, 0, 0)) && dr[1].eq(dr[0] + GSVector4i(0, 0, 0, 1))) { // persona 4: // @@ -195,7 +195,7 @@ protected: blurdetected = true; } - else if(dr[0] == dr[1] && (fr[0] == fr[1] + CPoint(0, 1) || fr[1] == fr[0] + CPoint(0, 1))) + else if(dr[0].eq(dr[1]) && (fr[0].eq(fr[1] + GSVector4i(0, 1, 0, 1)) || fr[1].eq(fr[0] + GSVector4i(0, 1, 0, 1)))) { // dq5: // @@ -214,8 +214,8 @@ protected: } } - CSize fs(0, 0); - CSize ds(0, 0); + GSVector2i fs(0, 0); + GSVector2i ds(0, 0); Texture tex[2]; @@ -238,13 +238,13 @@ protected: { if(!en[i] || !tex[i]) continue; - CRect r = fr[i]; + GSVector4i r = fr[i]; // overscan hack - if(dr[i].Height() > 512) // hmm + if(dr[i].height() > 512) // hmm { - int y = GetDeviceSize(i).cy; + int y = GetDeviceRect(i).height(); if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) y /= 2; r.bottom = r.top + y; } @@ -280,17 +280,16 @@ protected: dst[i].x = o.x; dst[i].y = o.y; - dst[i].z = o.x + tex[i].m_scale.x * r.Width(); - dst[i].w = o.y + tex[i].m_scale.y * r.Height(); + dst[i].z = o.x + tex[i].m_scale.x * r.width(); + dst[i].w = o.y + tex[i].m_scale.y * r.height(); - fs.cx = max(fs.cx, (int)(dst[i].z + 0.5f)); - fs.cy = max(fs.cy, (int)(dst[i].w + 0.5f)); + fs.x = max(fs.x, (int)(dst[i].z + 0.5f)); + fs.y = max(fs.y, (int)(dst[i].w + 0.5f)); } - ds.cx = fs.cx; - ds.cy = fs.cy; + ds = fs; - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) ds.cy *= 2; + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) ds.y *= 2; bool slbg = m_regs->PMODE.SLBG; bool mmod = m_regs->PMODE.MMOD; @@ -331,7 +330,7 @@ protected: fd.size = 0; fd.data = NULL; Freeze(&fd, true); - fd.data = new BYTE[fd.size]; + fd.data = new uint8[fd.size]; Freeze(&fd, false); m_dump.Open(m_snapshot, m_crc, fd, m_regs); @@ -359,7 +358,7 @@ protected: return; } - CSize size = m_capture.GetSize(); + GSVector2i size = m_capture.GetSize(); Texture current; @@ -367,9 +366,9 @@ protected: Texture offscreen; - if(m_dev.CopyOffscreen(current, GSVector4(0, 0, 1, 1), offscreen, size.cx, size.cy)) + if(m_dev.CopyOffscreen(current, GSVector4(0, 0, 1, 1), offscreen, size.x, size.y)) { - BYTE* bits = NULL; + uint8* bits = NULL; int pitch = 0; if(offscreen.Map(&bits, pitch)) @@ -412,7 +411,7 @@ public: GSCapture m_capture; public: - GSRenderer(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) + GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) : GSRendererBase(base, mt, irq, rs) , m_psrr(psrr) { @@ -466,12 +465,14 @@ public: double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); - string interlace = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; + string s = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; + + GSVector4i r = GetDisplayRect(); s_stats = format( "%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f", - m_perfmon.GetFrame(), GetDisplaySize().cx, GetDisplaySize().cy, fps, (int)(100.0 * fps / GetFPS()), - interlace.c_str(), + m_perfmon.GetFrame(), r.width(), r.height(), fps, (int)(100.0 * fps / GetFPS()), + s.c_str(), GSSettingsDlg::g_interlace[m_interlace].name, GSSettingsDlg::g_aspectratio[m_aspectratio].name, (int)m_perfmon.Get(GSPerfMon::Quad), @@ -516,13 +517,11 @@ public: // - CRect r; + GSVector4i r; - m_wnd.GetClientRect(&r); + m_wnd.GetClientRect(r); - GSUtil::FitRect(r, m_aspectratio); - - m_dev.Present(r); + m_dev.Present(r.fit(m_aspectratio)); // @@ -541,7 +540,7 @@ public: return true; } - virtual void MinMaxUV(int w, int h, CRect& r) {r = CRect(0, 0, w, h);} + virtual void MinMaxUV(int w, int h, GSVector4i& r) {r = GSVector4i(0, 0, w, h);} virtual bool CanUpscale() {return !m_nativeres;} }; @@ -603,7 +602,7 @@ protected: m_maxcount -= 100; } - template __forceinline Vertex* DrawingKick(bool skip, DWORD& count) + template __forceinline Vertex* DrawingKick(bool skip, int& count) { switch(prim) { @@ -683,7 +682,7 @@ protected: virtual void Draw() = 0; public: - GSRendererT(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true) + GSRendererT(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true) : GSRenderer(base, mt, irq, rs, psrr) , m_count(0) , m_maxcount(0) diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 1fb8a3b597..c19c5de96e 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -45,7 +45,7 @@ protected: __super::Reset(); } - void MinMaxUV(int w, int h, CRect& r) + void MinMaxUV(int w, int h, GSVector4i& r) { int wms = m_context->CLAMP.WMS; int wmt = m_context->CLAMP.WMT; @@ -208,18 +208,15 @@ protected: __assume(0); } - r = vr; + r = vr + GSVector4i(-1, -1, 1, 1); // one more pixel because of bilinear filtering - r.InflateRect(1, 1); // one more pixel because of bilinear filtering + GSVector2i bs = GSLocalMemory::m_psm[m_context->TEX0.PSM].bs; + GSVector2i bsm(bs.x - 1, bs.y - 1); - CSize bs = GSLocalMemory::m_psm[m_context->TEX0.PSM].bs; - CSize bsm(bs.cx - 1, bs.cy - 1); - - r.left = max(r.left & ~bsm.cx, 0); - r.right = min((r.right + bsm.cx) & ~bsm.cx, w); - - r.top = max(r.top & ~bsm.cy, 0); - r.bottom = min((r.bottom + bsm.cy) & ~bsm.cy, h); + r.left = max(r.left & ~bsm.x, 0); + r.top = max(r.top & ~bsm.y, 0); + r.right = min((r.right + bsm.x) & ~bsm.x, w); + r.bottom = min((r.bottom + bsm.y) & ~bsm.y, h); } void VSync(int field) @@ -275,14 +272,14 @@ protected: return false; } - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, CRect r) + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { TRACE(_T("[%d] InvalidateVideoMem %d,%d - %d,%d %05x (%d)\n"), (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.DBP, (int)BITBLTBUF.DPSM); m_tc->InvalidateVideoMem(BITBLTBUF, r); } - void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, CRect r) + void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { TRACE(_T("[%d] InvalidateLocalMem %d,%d - %d,%d %05x (%d)\n"), (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.SBP, (int)BITBLTBUF.SPSM); @@ -410,18 +407,18 @@ protected: if(m_game.title == CRC::FFXII && m_game.region == CRC::EU) { - static DWORD* video = NULL; + static uint32* video = NULL; static bool ok = false; - if(prim == GS_POINTLIST && m_count >= 448*448 && m_count <= 448*512) + if(prim == GS_POINTLIST && m_count >= 448 * 448 && m_count <= 448 * 512) { // incoming pixels are stored in columns, one column is 16x512, total res 448x512 or 448x454 - if(!video) video = new DWORD[512*512]; + if(!video) video = new uint32[512 * 512]; for(int x = 0, i = 0, rows = m_count / 448; x < 448; x += 16) { - DWORD* dst = &video[x]; + uint32* dst = &video[x]; for(int y = 0; y < rows; y++, dst += 512) { @@ -436,7 +433,7 @@ protected: return false; } - else if(prim == GS_LINELIST && m_count == 512*2 && ok) + else if(prim == GS_LINELIST && m_count == 512 * 2 && ok) { // normally, this step would copy the video onto screen with 512 texture mapped horizontal lines, // but we use the stored video data to create a new texture, and replace the lines with two triangles @@ -445,7 +442,7 @@ protected: m_dev.CreateTexture(*t, 512, 512); - t->Update(CRect(0, 0, 448, 512), video, 512*4); + t->Update(GSVector4i(0, 0, 448, 512), video, 512 * 4); m_vertices[0] = m_vertices[0]; m_vertices[1] = m_vertices[1]; @@ -467,9 +464,9 @@ protected: if(m_game.title == CRC::FFX) { - DWORD FBP = m_context->FRAME.Block(); - DWORD ZBP = m_context->ZBUF.Block(); - DWORD TBP = m_context->TEX0.TBP0; + uint32 FBP = m_context->FRAME.Block(); + uint32 ZBP = m_context->ZBUF.Block(); + uint32 TBP = m_context->TEX0.TBP0; if((FBP == 0x00d00 || FBP == 0x00000) && ZBP == 0x02100 && PRIM->TME && TBP == 0x01a00 && m_context->TEX0.PSM == PSM_PSMCT16S) { @@ -504,8 +501,8 @@ protected: { if(prim == GS_POINTLIST && !PRIM->TME) { - DWORD bp = m_context->FRAME.Block(); - DWORD bw = m_context->FRAME.FBW; + uint32 bp = m_context->FRAME.Block(); + uint32 bw = m_context->FRAME.FBW; if(bp >= 0x03f40 && (bp & 0x1f) == 0) { @@ -551,9 +548,9 @@ protected: if(m_game.title == CRC::GodOfWar2) { - DWORD FBP = m_context->FRAME.Block(); - DWORD FBW = m_context->FRAME.FBW; - DWORD FPSM = m_context->FRAME.PSM; + uint32 FBP = m_context->FRAME.Block(); + uint32 FBW = m_context->FRAME.FBW; + uint32 FPSM = m_context->FRAME.PSM; if((FBP == 0x00f00 || FBP == 0x00100) && FPSM == PSM_PSMZ24) // ntsc 0xf00, pal 0x100 { @@ -585,8 +582,8 @@ protected: if(m_game.title == CRC::DBZBT2) { - DWORD FBP = m_context->FRAME.Block(); - DWORD TBP0 = m_context->TEX0.TBP0; + uint32 FBP = m_context->FRAME.Block(); + uint32 TBP0 = m_context->TEX0.TBP0; if(PRIM->TME && (FBP == 0x03c00 && TBP0 == 0x03c80 || FBP == 0x03ac0 && TBP0 == 0x03b40)) { @@ -596,7 +593,7 @@ protected: BITBLTBUF.SBW = 1; BITBLTBUF.SPSM = PSM_PSMCT32; - InvalidateLocalMem(BITBLTBUF, CRect(0, 0, 64, 64)); + InvalidateLocalMem(BITBLTBUF, GSVector4i(0, 0, 64, 64)); } } @@ -606,7 +603,7 @@ protected: if(m_game.title == CRC::MajokkoALaMode2) { - DWORD FBP = m_context->FRAME.Block(); + uint32 FBP = m_context->FRAME.Block(); if(!PRIM->TME && FBP == 0x03f40) { @@ -616,7 +613,7 @@ protected: BITBLTBUF.SBW = 1; BITBLTBUF.SPSM = PSM_PSMCT32; - InvalidateLocalMem(BITBLTBUF, CRect(0, 0, 16, 16)); + InvalidateLocalMem(BITBLTBUF, GSVector4i(0, 0, 16, 16)); } } @@ -629,7 +626,7 @@ protected: if(m_game.title == CRC::DBZBT2) { - DWORD FBP = m_context->FRAME.Block(); + uint32 FBP = m_context->FRAME.Block(); if(FBP == 0x03c00 || FBP == 0x03ac0) { @@ -643,7 +640,7 @@ protected: if(m_game.title == CRC::MajokkoALaMode2) { - DWORD FBP = m_context->FRAME.Block(); + uint32 FBP = m_context->FRAME.Block(); if(FBP == 0x03f40) { @@ -657,7 +654,7 @@ protected: if(m_game.title == CRC::TalesOfAbyss) { - DWORD FBP = m_context->FRAME.Block(); + uint32 FBP = m_context->FRAME.Block(); if(FBP == 0x036e0 || FBP == 0x03560 || FBP == 0x038e0) { @@ -671,7 +668,7 @@ protected: } public: - GSRendererHW(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) + GSRendererHW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) : GSRendererT(base, mt, irq, rs, psrr) , m_width(1024) , m_height(1024) @@ -692,7 +689,7 @@ public: delete m_tc; } - void SetGameCRC(DWORD crc, int options) + void SetGameCRC(uint32 crc, int options) { __super::SetGameCRC(crc, options); diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 4069eb711c..f334118a9c 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -24,7 +24,7 @@ #include "GSCrc.h" #include "resource.h" -GSRendererHW10::GSRendererHW10(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs) +GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) : GSRendererHW(base, mt, irq, rs, true) { InitVertexKick(); @@ -70,7 +70,7 @@ bool GSRendererHW10::Create(const string& title) return true; } -template +template void GSRendererHW10::VertexKick(bool skip) { Vertex& dst = m_vl.AddTail(); @@ -83,7 +83,7 @@ void GSRendererHW10::VertexKick(bool skip) GSVector4::storel(&dst.ST, m_v.GetUV()); } - DWORD count = 0; + int count = 0; if(Vertex* v = DrawingKick(skip, count)) { @@ -412,7 +412,7 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachescissor.in) & CRect(0, 0, w, h); + GSVector4i scissor = GSVector4i(GSVector4(rt.m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); // @@ -436,7 +436,7 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCache void VertexKick(bool skip); + template void VertexKick(bool skip); }; \ No newline at end of file diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 48bff43558..036290eea3 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -24,7 +24,7 @@ #include "GSCrc.h" #include "resource.h" -GSRendererHW9::GSRendererHW9(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs) +GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) : GSRendererHW(base, mt, irq, rs, false) { m_fba.enabled = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("fba"), TRUE); @@ -74,7 +74,7 @@ bool GSRendererHW9::Create(const string& title) return true; } -template +template void GSRendererHW9::VertexKick(bool skip) { Vertex& dst = m_vl.AddTail(); @@ -83,8 +83,8 @@ void GSRendererHW9::VertexKick(bool skip) dst.p.y = (float)(int)m_v.XYZ.Y; dst.p.z = (float)m_v.XYZ.Z; - dst.c0 = m_v.RGBAQ.ai32[0]; - dst.c1 = m_v.FOG.ai32[1]; + dst.c0 = m_v.RGBAQ.u32[0]; + dst.c1 = m_v.FOG.u32[1]; if(tme) { @@ -100,7 +100,7 @@ void GSRendererHW9::VertexKick(bool skip) } } - DWORD count = 0; + int count = 0; if(Vertex* v = DrawingKick(skip, count)) { @@ -234,7 +234,7 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheFRAME.FBMSK & 0x00ff0000) != 0x00ff0000; om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; - BYTE bf = context->ALPHA.FIX >= 0x80 ? 0xff : (BYTE)(context->ALPHA.FIX * 2); + uint8 bf = context->ALPHA.FIX >= 0x80 ? 0xff : (uint8)(context->ALPHA.FIX * 2); // vs @@ -386,7 +386,7 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachescissor.in) & CRect(0, 0, w, h); + GSVector4i scissor = GSVector4i(GSVector4(rt.m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); // @@ -410,7 +410,7 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCache void VertexKick(bool skip); + template void VertexKick(bool skip); }; diff --git a/plugins/GSdx/GSRendererNull.h b/plugins/GSdx/GSRendererNull.h index bb9152443e..a12986c203 100644 --- a/plugins/GSdx/GSRendererNull.h +++ b/plugins/GSdx/GSRendererNull.h @@ -37,13 +37,13 @@ protected: } public: - GSRendererNull(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs) + GSRendererNull(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) : GSRendererT(base, mt, irq, rs) { InitVertexKick >(); } - template void VertexKick(bool skip) + template void VertexKick(bool skip) { } }; \ No newline at end of file diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 127e94ecd0..2b70069b02 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -79,16 +79,16 @@ protected: TEX0.TBW = DISPFB.FBW; TEX0.PSM = DISPFB.PSM; - CRect r(0, 0, TEX0.TBW * 64, GetFrameRect(i).bottom); + GSVector4i r(0, 0, TEX0.TBW * 64, GetFrameRect(i).bottom); // TODO: round up bottom - if(m_texture[i].GetWidth() != r.Width() || m_texture[i].GetHeight() != r.Height()) + if(m_texture[i].GetWidth() != r.width() || m_texture[i].GetHeight() != r.height()) { m_texture[i] = Texture(); } - if(!m_texture[i] && !m_dev.CreateTexture(m_texture[i], r.Width(), r.Height())) + if(!m_texture[i] && !m_dev.CreateTexture(m_texture[i], r.width(), r.height())) { return false; } @@ -98,7 +98,7 @@ protected: CLAMP.WMS = CLAMP.WMT = 1; // TODO - static BYTE* buff = (BYTE*)_aligned_malloc(1024 * 1024 * 4, 16); + static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); static int pitch = 1024 * 4; m_mem.ReadTexture(r, buff, pitch, TEX0, m_env.TEXA, CLAMP); @@ -134,9 +134,9 @@ protected: if(PRIM->TME && context->TEX0.TCC) { - DWORD bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; - DWORD cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; - DWORD pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; + uint32 bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; + uint32 cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; + uint32 pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; if(bpp == 32) { @@ -177,6 +177,8 @@ protected: if(a.z > 0xff) a.z = 0xff; break; case TFX_HIGHLIGHT2: + a.x = a.y; + a.z = a.w; break; default: __assume(0); @@ -188,7 +190,7 @@ protected: m_vtrace.m_alpha.valid = true; } - bool TryAlphaTest(DWORD& fm, DWORD& zm) + bool TryAlphaTest(uint32& fm, uint32& zm) { const GSDrawingContext* context = m_context; @@ -440,11 +442,11 @@ protected: } */ - CRect r; - int w = 1 << context->TEX0.TW; int h = 1 << context->TEX0.TH; + GSVector4i r; + MinMaxUV(w, h, r, p.sel.fst); const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, &r); @@ -491,7 +493,7 @@ protected: if(PRIM->ABE && !context->ALPHA.IsOpaque(amin, amax) || PRIM->AA1) { p.sel.abe = PRIM->ABE; - p.sel.ababcd = context->ALPHA.ai32[0]; + p.sel.ababcd = context->ALPHA.u32[0]; if(env.PABE.PABE) { @@ -572,14 +574,14 @@ protected: { s = format("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); - m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);//GetFrameSize(1).cy); + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); } if(s_savez) { s = format("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); - m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512); + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); } s_n++; @@ -604,16 +606,7 @@ protected: m_perfmon.Put(GSPerfMon::Prim, stats.prims); m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); - GSVector4i pos(m_vtrace.m_min.p.xyxy(m_vtrace.m_max.p)); - - GSVector4i scissor = data.scissor; - - CRect r; - - r.left = max(scissor.x, min(scissor.z, pos.x)); - r.top = max(scissor.y, min(scissor.w, pos.y)); - r.right = max(scissor.x, min(scissor.z, pos.z)); - r.bottom = max(scissor.y, min(scissor.w, pos.w)); + GSVector4i r = GSVector4i(m_vtrace.m_min.p.xyxy(m_vtrace.m_max.p)).rintersect(data.scissor); GIFRegBITBLTBUF BITBLTBUF; @@ -645,14 +638,14 @@ protected: { s = format("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); - m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameSize().cx, 512);//GetFrameSize(1).cy); + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); } if(s_savez) { s = format("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); - m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameSize().cx, 512); + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); } s_n++; @@ -668,12 +661,12 @@ protected: } } - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, CRect r) + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { m_tc->InvalidateVideoMem(BITBLTBUF, r); } - void MinMaxUV(int w, int h, CRect& r, DWORD fst) + void MinMaxUV(int w, int h, GSVector4i& r, uint32 fst) { const GSDrawingContext* context = m_context; @@ -778,13 +771,11 @@ protected: } } - r = vr; - - r &= CRect(0, 0, w, h); + r = vr.rintersect(GSVector4i(0, 0, w, h)); } public: - GSRendererSW(BYTE* base, bool mt, void (*irq)(), const GSRendererSettings& rs, int threads) + GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, int threads) : GSRendererT(base, mt, irq, rs) { m_rl.Create(this, threads); @@ -799,12 +790,12 @@ public: delete m_tc; } - template + template void VertexKick(bool skip) { const GSDrawingContext* context = m_context; - GSVector4i xy = GSVector4i::load((int)m_v.XYZ.ai32[0]); + GSVector4i xy = GSVector4i::load((int)m_v.XYZ.u32[0]); xy = xy.insert16<3>(m_v.FOG.F); xy = xy.upl16(); @@ -814,7 +805,7 @@ public: v.p = GSVector4(xy) * g_pos_scale; - v.c = GSVector4(GSVector4i::load((int)m_v.RGBAQ.ai32[0]).u8to32() << 7); + v.c = GSVector4(GSVector4i::load((int)m_v.RGBAQ.u32[0]).u8to32() << 7); if(tme) { @@ -839,9 +830,9 @@ public: dst = v; - dst.p.z = (float)min(m_v.XYZ.Z, 0xffffff00); // max value which can survive the DWORD => float => DWORD conversion + dst.p.z = (float)min(m_v.XYZ.Z, 0xffffff00); // max value which can survive the uint32 => float => uint32 conversion - DWORD count = 0; + int count = 0; if(GSVertexSW* v = DrawingKick(skip, count)) { @@ -951,7 +942,7 @@ if(!m_dump) m_count = 2; - UINT32 tmp = PRIM->PRIM; + uint32 tmp = PRIM->PRIM; PRIM->PRIM = GS_SPRITE; Flush(); diff --git a/plugins/GSdx/GSScanlineEnvironment.h b/plugins/GSdx/GSScanlineEnvironment.h index d959930197..b7ee1cdb9a 100644 --- a/plugins/GSdx/GSScanlineEnvironment.h +++ b/plugins/GSdx/GSScanlineEnvironment.h @@ -28,63 +28,63 @@ union GSScanlineSelector { struct { - DWORD fpsm:2; // 0 - DWORD zpsm:2; // 2 - DWORD ztst:2; // 4 (0: off, 1: write, 2: test (ge), 3: test (g)) - DWORD atst:3; // 6 - DWORD afail:2; // 9 - DWORD iip:1; // 11 - DWORD tfx:3; // 12 - DWORD tcc:1; // 15 - DWORD fst:1; // 16 - DWORD ltf:1; // 17 - DWORD tlu:1; // 18 - DWORD fge:1; // 19 - DWORD date:1; // 20 - DWORD abe:1; // 21 - DWORD aba:2; // 22 - DWORD abb:2; // 24 - DWORD abc:2; // 26 - DWORD abd:2; // 28 - DWORD pabe:1; // 30 - DWORD aa1:1; // 31 + uint32 fpsm:2; // 0 + uint32 zpsm:2; // 2 + uint32 ztst:2; // 4 (0: off, 1: write, 2: test (ge), 3: test (g)) + uint32 atst:3; // 6 + uint32 afail:2; // 9 + uint32 iip:1; // 11 + uint32 tfx:3; // 12 + uint32 tcc:1; // 15 + uint32 fst:1; // 16 + uint32 ltf:1; // 17 + uint32 tlu:1; // 18 + uint32 fge:1; // 19 + uint32 date:1; // 20 + uint32 abe:1; // 21 + uint32 aba:2; // 22 + uint32 abb:2; // 24 + uint32 abc:2; // 26 + uint32 abd:2; // 28 + uint32 pabe:1; // 30 + uint32 aa1:1; // 31 - DWORD fwrite:1; // 32 - DWORD ftest:1; // 33 - DWORD rfb:1; // 34 - DWORD zwrite:1; // 35 - DWORD ztest:1; // 36 - DWORD zoverflow:1; // 37 (z max >= 0x80000000) - DWORD wms:2; // 38 - DWORD wmt:2; // 40 - DWORD datm:1; // 42 - DWORD colclamp:1; // 43 - DWORD fba:1; // 44 - DWORD dthe:1; // 45 - DWORD sprite:1; // 46 - DWORD edge:1; // 47 + uint32 fwrite:1; // 32 + uint32 ftest:1; // 33 + uint32 rfb:1; // 34 + uint32 zwrite:1; // 35 + uint32 ztest:1; // 36 + uint32 zoverflow:1; // 37 (z max >= 0x80000000) + uint32 wms:2; // 38 + uint32 wmt:2; // 40 + uint32 datm:1; // 42 + uint32 colclamp:1; // 43 + uint32 fba:1; // 44 + uint32 dthe:1; // 45 + uint32 sprite:1; // 46 + uint32 edge:1; // 47 }; struct { - DWORD _pad1:22; - DWORD ababcd:8; - DWORD _pad2:2; - DWORD fb:2; - DWORD _pad3:1; - DWORD zb:2; + uint32 _pad1:22; + uint32 ababcd:8; + uint32 _pad2:2; + uint32 fb:2; + uint32 _pad3:1; + uint32 zb:2; }; struct { - DWORD lo; - DWORD hi; + uint32 lo; + uint32 hi; }; - UINT64 key; + uint64 key; - operator DWORD() {return lo;} - operator UINT64() {return key;} + operator uint32() {return lo;} + operator uint64() {return key;} bool IsSolidRect() { @@ -105,22 +105,22 @@ __declspec(align(16)) struct GSScanlineParam void* vm; const void* tex; - const DWORD* clut; - DWORD tw; + const uint32* clut; + uint32 tw; GSLocalMemory::Offset* fbo; GSLocalMemory::Offset* zbo; GSLocalMemory::Offset4* fzbo; - DWORD fm, zm; + uint32 fm, zm; }; __declspec(align(16)) struct GSScanlineEnvironment { void* vm; const void* tex; - const DWORD* clut; - DWORD tw; + const uint32* clut; + uint32 tw; GSVector4i* fbr; GSVector4i* zbr; diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 133cc864d6..0f3d5577fb 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -165,13 +165,13 @@ BOOL GSSettingsDlg::OnInitDialog() if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - UINT ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); - UINT ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); - UINT ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); + uint32 ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); + uint32 ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); + uint32 ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); - UINT nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); + uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); - for(UINT i = 0; i < nModes; i++) + for(uint32 i = 0; i < nModes; i++) { D3DDISPLAYMODE mode; @@ -259,22 +259,22 @@ void GSSettingsDlg::OnOK() if(m_renderer.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("Renderer"), (DWORD)m_renderer.GetItemData(m_renderer.GetCurSel())); + pApp->WriteProfileInt(_T("Settings"), _T("Renderer"), (uint32)m_renderer.GetItemData(m_renderer.GetCurSel())); } if(m_psversion.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("PixelShaderVersion2"), (DWORD)m_psversion.GetItemData(m_psversion.GetCurSel())); + pApp->WriteProfileInt(_T("Settings"), _T("PixelShaderVersion2"), (uint32)m_psversion.GetItemData(m_psversion.GetCurSel())); } if(m_interlace.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("Interlace"), (DWORD)m_interlace.GetItemData(m_interlace.GetCurSel())); + pApp->WriteProfileInt(_T("Settings"), _T("Interlace"), (uint32)m_interlace.GetItemData(m_interlace.GetCurSel())); } if(m_aspectratio.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("AspectRatio"), (DWORD)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); + pApp->WriteProfileInt(_T("Settings"), _T("AspectRatio"), (uint32)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); } pApp->WriteProfileInt(_T("Settings"), _T("filter"), m_filter); diff --git a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp index 4fa97f8ab6..57f204f855 100644 --- a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp +++ b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp @@ -24,7 +24,7 @@ #include "StdAfx.h" #include "GSSetupPrimCodeGenerator.h" -GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, UINT64 key, void* ptr, size_t maxsize) +GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, uint64 key, void* ptr, size_t maxsize) : CodeGenerator(maxsize, ptr) , m_env(env) { diff --git a/plugins/GSdx/GSSetupPrimCodeGenerator.h b/plugins/GSdx/GSSetupPrimCodeGenerator.h index f12913da7b..af7dd116b8 100644 --- a/plugins/GSdx/GSSetupPrimCodeGenerator.h +++ b/plugins/GSdx/GSSetupPrimCodeGenerator.h @@ -38,7 +38,7 @@ class GSSetupPrimCodeGenerator : public CodeGenerator GSScanlineEnvironment& m_env; GSScanlineSelector m_sel; - struct {DWORD z:1, f:1, t:1, c:1;} m_en; + struct {uint32 z:1, f:1, t:1, c:1;} m_en; void Generate(); @@ -47,5 +47,5 @@ class GSSetupPrimCodeGenerator : public CodeGenerator void Color(); public: - GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, UINT64 key, void* ptr, size_t maxsize); + GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, uint64 key, void* ptr, size_t maxsize); }; \ No newline at end of file diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 620f41959b..93721c5236 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -22,7 +22,7 @@ #include "stdafx.h" #include "GSState.h" -GSState::GSState(BYTE* base, bool mt, void (*irq)()) +GSState::GSState(uint8* base, bool mt, void (*irq)()) : m_mt(mt) , m_irq(irq) , m_crc(0) @@ -202,112 +202,65 @@ void GSState::ResetHandlers() m_fpGIFRegHandlers[GIF_A_D_REG_LABEL] = &GSState::GIFRegHandlerLABEL; } -CPoint GSState::GetDisplayPos(int i) +GSVector4i GSState::GetDisplayRect(int i) { - ASSERT(i >= 0 && i < 2); + if(i < 0) i = IsEnabled(1) ? 1 : 0; - CPoint p; + GSVector4i r; - p.x = m_regs->DISP[i].DISPLAY.DX / (m_regs->DISP[i].DISPLAY.MAGH + 1); - p.y = m_regs->DISP[i].DISPLAY.DY / (m_regs->DISP[i].DISPLAY.MAGV + 1); + r.left = m_regs->DISP[i].DISPLAY.DX / (m_regs->DISP[i].DISPLAY.MAGH + 1); + r.top = m_regs->DISP[i].DISPLAY.DY / (m_regs->DISP[i].DISPLAY.MAGV + 1); + r.right = r.left + (m_regs->DISP[i].DISPLAY.DW + 1) / (m_regs->DISP[i].DISPLAY.MAGH + 1); + r.bottom = r.top + (m_regs->DISP[i].DISPLAY.DH + 1) / (m_regs->DISP[i].DISPLAY.MAGV + 1); - return p; + return r; } -CSize GSState::GetDisplaySize(int i) +GSVector4i GSState::GetFrameRect(int i) { - ASSERT(i >= 0 && i < 2); + if(i < 0) i = IsEnabled(1) ? 1 : 0; - CSize s; + GSVector4i r = GetDisplayRect(i); - s.cx = (m_regs->DISP[i].DISPLAY.DW + 1) / (m_regs->DISP[i].DISPLAY.MAGH + 1); - s.cy = (m_regs->DISP[i].DISPLAY.DH + 1) / (m_regs->DISP[i].DISPLAY.MAGV + 1); + int w = r.width(); + int h = r.height(); - return s; + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD && h > 1) h >>= 1; + + r.left = m_regs->DISP[i].DISPFB.DBX; + r.top = m_regs->DISP[i].DISPFB.DBY; + r.right = r.left + w; + r.bottom = r.top + h; + + return r; } -CRect GSState::GetDisplayRect(int i) -{ - return CRect(GetDisplayPos(i), GetDisplaySize(i)); -} - -CSize GSState::GetDisplayPos() -{ - return GetDisplayPos(IsEnabled(1) ? 1 : 0); -} - -CSize GSState::GetDisplaySize() -{ - return GetDisplaySize(IsEnabled(1) ? 1 : 0); -} - -CRect GSState::GetDisplayRect() -{ - return GetDisplayRect(IsEnabled(1) ? 1 : 0); -} - -CPoint GSState::GetFramePos(int i) -{ - ASSERT(i >= 0 && i < 2); - - return CPoint(m_regs->DISP[i].DISPFB.DBX, m_regs->DISP[i].DISPFB.DBY); -} - -CSize GSState::GetFrameSize(int i) -{ - CSize s = GetDisplaySize(i); - - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD && s.cy > 1) s.cy >>= 1; - - return s; -} - -CRect GSState::GetFrameRect(int i) -{ - return CRect(GetFramePos(i), GetFrameSize(i)); -} - -CSize GSState::GetFramePos() -{ - return GetFramePos(IsEnabled(1) ? 1 : 0); -} - -CSize GSState::GetFrameSize() -{ - return GetFrameSize(IsEnabled(1) ? 1 : 0); -} - -CRect GSState::GetFrameRect() -{ - return GetFrameRect(IsEnabled(1) ? 1 : 0); -} - -CSize GSState::GetDeviceSize(int i) +GSVector4i GSState::GetDeviceRect(int i) { // TODO: other params of SMODE1 should affect the true device display size // TODO2: pal games at 60Hz - CSize s = GetDisplaySize(i); + if(i < 0) i = IsEnabled(1) ? 1 : 0; - if(s.cy == 2 * 416 || s.cy == 2 * 448 || s.cy == 2 * 512) + GSVector4i r = GetDisplayRect(i); + + int w = r.width(); + int h = r.height(); + + if(h == 2 * 416 || h == 2 * 448 || h == 2 * 512) { - s.cy /= 2; + h /= 2; } else { - s.cy = (m_regs->SMODE1.CMOD & 1) ? 512 : 448; + h = (m_regs->SMODE1.CMOD & 1) ? 512 : 448; } - return s; + return GSVector4i(0, 0, w, h); } -CSize GSState::GetDeviceSize() -{ - return GetDeviceSize(IsEnabled(1) ? 1 : 0); -} - bool GSState::IsEnabled(int i) { ASSERT(i >= 0 && i < 2); @@ -349,12 +302,12 @@ void GSState::GIFPackedRegHandlerRGBA(GIFPackedReg* r) GSVector4i mask = GSVector4i::load(0x0c080400); GSVector4i v = GSVector4i::load(r).shuffle8(mask); - m_v.RGBAQ.ai32[0] = (UINT32)GSVector4i::store(v); + m_v.RGBAQ.u32[0] = (uint32)GSVector4i::store(v); #elif _M_SSE >= 0x200 GSVector4i v = GSVector4i::load(r) & GSVector4i::x000000ff(); - m_v.RGBAQ.ai32[0] = v.rgba32(); + m_v.RGBAQ.u32[0] = v.rgba32(); #else @@ -372,12 +325,12 @@ void GSState::GIFPackedRegHandlerSTQ(GIFPackedReg* r) { #if defined(_M_AMD64) - m_v.ST.i64 = r->ai64[0]; + m_v.ST.u64 = r->u64[0]; #elif _M_SSE >= 0x200 GSVector4i v = GSVector4i::loadl(r); - GSVector4i::storel(&m_v.ST.i64, v); + GSVector4i::storel(&m_v.ST.u64, v); #else @@ -394,7 +347,7 @@ void GSState::GIFPackedRegHandlerUV(GIFPackedReg* r) #if _M_SSE >= 0x200 GSVector4i v = GSVector4i::loadl(r) & GSVector4i::x00003fff(); - m_v.UV.ai32[0] = (UINT32)GSVector4i::store(v.ps32(v)); + m_v.UV.u32[0] = (uint32)GSVector4i::store(v.ps32(v)); #else @@ -425,12 +378,12 @@ void GSState::GIFPackedRegHandlerXYZ2(GIFPackedReg* r) template void GSState::GIFPackedRegHandlerTEX0(GIFPackedReg* r) { - GIFRegHandlerTEX0((GIFReg*)&r->ai64[0]); + GIFRegHandlerTEX0((GIFReg*)&r->u64[0]); } template void GSState::GIFPackedRegHandlerCLAMP(GIFPackedReg* r) { - GIFRegHandlerCLAMP((GIFReg*)&r->ai64[0]); + GIFRegHandlerCLAMP((GIFReg*)&r->u64[0]); } void GSState::GIFPackedRegHandlerFOG(GIFPackedReg* r) @@ -440,17 +393,17 @@ void GSState::GIFPackedRegHandlerFOG(GIFPackedReg* r) void GSState::GIFPackedRegHandlerXYZF3(GIFPackedReg* r) { - GIFRegHandlerXYZF3((GIFReg*)&r->ai64[0]); + GIFRegHandlerXYZF3((GIFReg*)&r->u64[0]); } void GSState::GIFPackedRegHandlerXYZ3(GIFPackedReg* r) { - GIFRegHandlerXYZ3((GIFReg*)&r->ai64[0]); + GIFRegHandlerXYZ3((GIFReg*)&r->u64[0]); } void GSState::GIFPackedRegHandlerA_D(GIFPackedReg* r) { - (this->*m_fpGIFRegHandlers[(BYTE)r->A_D.ADDR])(&r->r); + (this->*m_fpGIFRegHandlers[r->A_D.ADDR])(&r->r); } void GSState::GIFPackedRegHandlerNOP(GIFPackedReg* r) @@ -470,7 +423,7 @@ void GSState::GIFRegHandlerPRIM(GIFReg* r) if(GSUtil::GetPrimClass(m_env.PRIM.PRIM) == GSUtil::GetPrimClass(r->PRIM.PRIM)) { - if(((m_env.PRIM.i64 ^ r->PRIM.i64) & ~7) != 0) + if(((m_env.PRIM.u64 ^ r->PRIM.u64) & ~7) != 0) { Flush(); } @@ -502,7 +455,7 @@ void GSState::GIFRegHandlerST(GIFReg* r) void GSState::GIFRegHandlerUV(GIFReg* r) { - m_v.UV.ai32[0] = r->UV.ai32[0] & 0x3fff3fff; + m_v.UV.u32[0] = r->UV.u32[0] & 0x3fff3fff; } void GSState::GIFRegHandlerXYZF2(GIFReg* r) @@ -513,9 +466,9 @@ void GSState::GIFRegHandlerXYZF2(GIFReg* r) m_v.XYZ.Z = r->XYZF.Z; m_v.FOG.F = r->XYZF.F; */ - m_v.XYZ.ai32[0] = r->XYZF.ai32[0]; - m_v.XYZ.ai32[1] = r->XYZF.ai32[1] & 0x00ffffff; - m_v.FOG.ai32[1] = r->XYZF.ai32[1] & 0xff000000; + m_v.XYZ.u32[0] = r->XYZF.u32[0]; + m_v.XYZ.u32[1] = r->XYZF.u32[1] & 0x00ffffff; + m_v.FOG.u32[1] = r->XYZF.u32[1] & 0xff000000; VertexKick(false); } @@ -533,7 +486,7 @@ template void GSState::GIFRegHandlerTEX0(GIFReg* r) bool wt = m_mem.m_clut.WriteTest(r->TEX0, m_env.TEXCLUT); - if(wt || PRIM->CTXT == i && !(m_env.CTXT[i].TEX0 == (GSVector4i)r->TEX0).alltrue()) + if(wt || PRIM->CTXT == i && !((GSVector4i)r->TEX0).eq(m_env.CTXT[i].TEX0)) { Flush(); } @@ -558,7 +511,7 @@ template void GSState::GIFRegHandlerTEX0(GIFReg* r) template void GSState::GIFRegHandlerCLAMP(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].CLAMP == (GSVector4i)r->CLAMP).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->CLAMP).eq(m_env.CTXT[i].CLAMP)) { Flush(); } @@ -579,9 +532,9 @@ void GSState::GIFRegHandlerXYZF3(GIFReg* r) m_v.XYZ.Z = r->XYZF.Z; m_v.FOG.F = r->XYZF.F; */ - m_v.XYZ.ai32[0] = r->XYZF.ai32[0]; - m_v.XYZ.ai32[1] = r->XYZF.ai32[1] & 0x00ffffff; - m_v.FOG.ai32[1] = r->XYZF.ai32[1] & 0xff000000; + m_v.XYZ.u32[0] = r->XYZF.u32[0]; + m_v.XYZ.u32[1] = r->XYZF.u32[1] & 0x00ffffff; + m_v.FOG.u32[1] = r->XYZF.u32[1] & 0xff000000; VertexKick(true); } @@ -599,7 +552,7 @@ void GSState::GIFRegHandlerNOP(GIFReg* r) template void GSState::GIFRegHandlerTEX1(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].TEX1 == (GSVector4i)r->TEX1).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->TEX1).eq(m_env.CTXT[i].TEX1)) { Flush(); } @@ -613,7 +566,7 @@ template void GSState::GIFRegHandlerTEX2(GIFReg* r) UINT64 mask = 0xFFFFFFE003F00000ui64; // TEX2 bits - r->i64 = (r->i64 & mask) | (m_env.CTXT[i].TEX0.i64 & ~mask); + r->u64 = (r->u64 & mask) | (m_env.CTXT[i].TEX0.u64 & ~mask); GIFRegHandlerTEX0(r); } @@ -622,7 +575,7 @@ template void GSState::GIFRegHandlerXYOFFSET(GIFReg* r) { GSVector4i o = (GSVector4i)r->XYOFFSET & GSVector4i::x0000ffff(); - if(!(m_env.CTXT[i].XYOFFSET == o).alltrue()) + if(!o.eq(m_env.CTXT[i].XYOFFSET)) { Flush(); } @@ -634,7 +587,7 @@ template void GSState::GIFRegHandlerXYOFFSET(GIFReg* r) void GSState::GIFRegHandlerPRMODECONT(GIFReg* r) { - if(!(m_env.PRMODECONT == (GSVector4i)r->PRMODECONT).alltrue()) + if(!((GSVector4i)r->PRMODECONT).eq(m_env.PRMODECONT)) { Flush(); } @@ -657,7 +610,7 @@ void GSState::GIFRegHandlerPRMODE(GIFReg* r) Flush(); } - UINT32 _PRIM = m_env.PRMODE._PRIM; + uint32 _PRIM = m_env.PRMODE._PRIM; m_env.PRMODE = (GSVector4i)r->PRMODE; m_env.PRMODE._PRIM = _PRIM; @@ -668,7 +621,7 @@ void GSState::GIFRegHandlerPRMODE(GIFReg* r) void GSState::GIFRegHandlerTEXCLUT(GIFReg* r) { - if(!(m_env.TEXCLUT == (GSVector4i)r->TEXCLUT).alltrue()) + if(!((GSVector4i)r->TEXCLUT).eq(m_env.TEXCLUT)) { Flush(); } @@ -678,7 +631,7 @@ void GSState::GIFRegHandlerTEXCLUT(GIFReg* r) void GSState::GIFRegHandlerSCANMSK(GIFReg* r) { - if(!(m_env.SCANMSK == (GSVector4i)r->SCANMSK).alltrue()) + if(!((GSVector4i)r->SCANMSK).eq(m_env.SCANMSK)) { Flush(); } @@ -688,7 +641,7 @@ void GSState::GIFRegHandlerSCANMSK(GIFReg* r) template void GSState::GIFRegHandlerMIPTBP1(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].MIPTBP1 == (GSVector4i)r->MIPTBP1).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->MIPTBP1).eq(m_env.CTXT[i].MIPTBP1)) { Flush(); } @@ -698,7 +651,7 @@ template void GSState::GIFRegHandlerMIPTBP1(GIFReg* r) template void GSState::GIFRegHandlerMIPTBP2(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].MIPTBP2 == (GSVector4i)r->MIPTBP2).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->MIPTBP2).eq(m_env.CTXT[i].MIPTBP2)) { Flush(); } @@ -708,7 +661,7 @@ template void GSState::GIFRegHandlerMIPTBP2(GIFReg* r) void GSState::GIFRegHandlerTEXA(GIFReg* r) { - if(!(m_env.TEXA == (GSVector4i)r->TEXA).alltrue()) + if(!((GSVector4i)r->TEXA).eq(m_env.TEXA)) { Flush(); } @@ -718,7 +671,7 @@ void GSState::GIFRegHandlerTEXA(GIFReg* r) void GSState::GIFRegHandlerFOGCOL(GIFReg* r) { - if(!(m_env.FOGCOL == (GSVector4i)r->FOGCOL).alltrue()) + if(!((GSVector4i)r->FOGCOL).eq(m_env.FOGCOL)) { Flush(); } @@ -735,7 +688,7 @@ void GSState::GIFRegHandlerTEXFLUSH(GIFReg* r) template void GSState::GIFRegHandlerSCISSOR(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].SCISSOR == (GSVector4i)r->SCISSOR).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->SCISSOR).eq(m_env.CTXT[i].SCISSOR)) { Flush(); } @@ -752,7 +705,7 @@ template void GSState::GIFRegHandlerALPHA(GIFReg* r) ASSERT(r->ALPHA.C != 3); ASSERT(r->ALPHA.D != 3); - if(PRIM->CTXT == i && !(m_env.CTXT[i].ALPHA == (GSVector4i)r->ALPHA).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->ALPHA).eq(m_env.CTXT[i].ALPHA)) { Flush(); } @@ -761,14 +714,14 @@ template void GSState::GIFRegHandlerALPHA(GIFReg* r) // A/B/C/D == 3? => 2 - m_env.CTXT[i].ALPHA.ai32[0] = ((~m_env.CTXT[i].ALPHA.ai32[0] >> 1) | 0xAA) & m_env.CTXT[i].ALPHA.ai32[0]; + m_env.CTXT[i].ALPHA.u32[0] = ((~m_env.CTXT[i].ALPHA.u32[0] >> 1) | 0xAA) & m_env.CTXT[i].ALPHA.u32[0]; } void GSState::GIFRegHandlerDIMX(GIFReg* r) { bool update = false; - if(!(m_env.DIMX == (GSVector4i)r->DIMX).alltrue()) + if(!((GSVector4i)r->DIMX).eq(m_env.DIMX)) { Flush(); @@ -785,7 +738,7 @@ void GSState::GIFRegHandlerDIMX(GIFReg* r) void GSState::GIFRegHandlerDTHE(GIFReg* r) { - if(!(m_env.DTHE == (GSVector4i)r->DTHE).alltrue()) + if(!((GSVector4i)r->DTHE).eq(m_env.DTHE)) { Flush(); } @@ -795,7 +748,7 @@ void GSState::GIFRegHandlerDTHE(GIFReg* r) void GSState::GIFRegHandlerCOLCLAMP(GIFReg* r) { - if(!(m_env.COLCLAMP == (GSVector4i)r->COLCLAMP).alltrue()) + if(!((GSVector4i)r->COLCLAMP).eq(m_env.COLCLAMP)) { Flush(); } @@ -805,7 +758,7 @@ void GSState::GIFRegHandlerCOLCLAMP(GIFReg* r) template void GSState::GIFRegHandlerTEST(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].TEST == (GSVector4i)r->TEST).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->TEST).eq(m_env.CTXT[i].TEST)) { Flush(); } @@ -815,7 +768,7 @@ template void GSState::GIFRegHandlerTEST(GIFReg* r) void GSState::GIFRegHandlerPABE(GIFReg* r) { - if(!(m_env.PABE == (GSVector4i)r->PABE).alltrue()) + if(!((GSVector4i)r->PABE).eq(m_env.PABE)) { Flush(); } @@ -825,7 +778,7 @@ void GSState::GIFRegHandlerPABE(GIFReg* r) template void GSState::GIFRegHandlerFBA(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].FBA == (GSVector4i)r->FBA).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->FBA).eq(m_env.CTXT[i].FBA)) { Flush(); } @@ -835,7 +788,7 @@ template void GSState::GIFRegHandlerFBA(GIFReg* r) template void GSState::GIFRegHandlerFRAME(GIFReg* r) { - if(PRIM->CTXT == i && !(m_env.CTXT[i].FRAME == (GSVector4i)r->FRAME).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->FRAME).eq(m_env.CTXT[i].FRAME)) { Flush(); } @@ -845,7 +798,7 @@ template void GSState::GIFRegHandlerFRAME(GIFReg* r) template void GSState::GIFRegHandlerZBUF(GIFReg* r) { - if(r->ZBUF.ai32[0] == 0) + if(r->ZBUF.u32[0] == 0) { // during startup all regs are cleared to 0 (by the bios or something), so we mask z until this register becomes valid @@ -854,7 +807,7 @@ template void GSState::GIFRegHandlerZBUF(GIFReg* r) r->ZBUF.PSM |= 0x30; - if(PRIM->CTXT == i && !(m_env.CTXT[i].ZBUF == (GSVector4i)r->ZBUF).alltrue()) + if(PRIM->CTXT == i && !((GSVector4i)r->ZBUF).eq(m_env.CTXT[i].ZBUF)) { Flush(); } @@ -872,7 +825,7 @@ template void GSState::GIFRegHandlerZBUF(GIFReg* r) void GSState::GIFRegHandlerBITBLTBUF(GIFReg* r) { - if(!(m_env.BITBLTBUF == (GSVector4i)r->BITBLTBUF).alltrue()) + if(!((GSVector4i)r->BITBLTBUF).eq(m_env.BITBLTBUF)) { FlushWrite(); } @@ -892,7 +845,7 @@ void GSState::GIFRegHandlerBITBLTBUF(GIFReg* r) void GSState::GIFRegHandlerTRXPOS(GIFReg* r) { - if(!(m_env.TRXPOS == (GSVector4i)r->TRXPOS).alltrue()) + if(!((GSVector4i)r->TRXPOS).eq(m_env.TRXPOS)) { FlushWrite(); } @@ -902,7 +855,7 @@ void GSState::GIFRegHandlerTRXPOS(GIFReg* r) void GSState::GIFRegHandlerTRXREG(GIFReg* r) { - if(!(m_env.TRXREG == (GSVector4i)r->TRXREG).alltrue()) + if(!((GSVector4i)r->TRXREG).eq(m_env.TRXREG)) { FlushWrite(); } @@ -937,7 +890,7 @@ void GSState::GIFRegHandlerHWREG(GIFReg* r) { ASSERT(m_env.TRXDIR.XDIR == 0); // host => local - Write((BYTE*)r, 8); // hunting ground + Write((uint8*)r, 8); // hunting ground } void GSState::GIFRegHandlerSIGNAL(GIFReg* r) @@ -990,7 +943,7 @@ void GSState::FlushWrite() m_perfmon.Put(GSPerfMon::Swizzle, len); - CRect r; + GSVector4i r; r.left = m_env.TRXPOS.DSAX; r.top = y; @@ -1010,7 +963,7 @@ void GSState::FlushWrite() // -void GSState::Write(BYTE* mem, int len) +void GSState::Write(uint8* mem, int len) { int dx = m_env.TRXPOS.DSAX; int dy = m_env.TRXPOS.DSAY; @@ -1041,7 +994,7 @@ void GSState::Write(BYTE* mem, int len) m_mem.m_clut.Invalidate(); } -void GSState::Read(BYTE* mem, int len) +void GSState::Read(uint8* mem, int len) { if(len <= 0) return; @@ -1059,7 +1012,7 @@ void GSState::Read(BYTE* mem, int len) if(m_tr.x == sx && m_tr.y == sy) { - InvalidateLocalMem(m_env.BITBLTBUF, CRect(CPoint(sx, sy), CSize(w, h))); + InvalidateLocalMem(m_env.BITBLTBUF, GSVector4i(sx, sy, sx + w, sy + h)); } m_mem.ReadImageX(m_tr.x, m_tr.y, mem, len, m_env.BITBLTBUF, m_env.TRXPOS, m_env.TRXREG); @@ -1077,8 +1030,8 @@ void GSState::Move() int w = m_env.TRXREG.RRW; int h = m_env.TRXREG.RRH; - InvalidateLocalMem(m_env.BITBLTBUF, CRect(CPoint(sx, sy), CSize(w, h))); - InvalidateVideoMem(m_env.BITBLTBUF, CRect(CPoint(dx, dy), CSize(w, h))); + InvalidateLocalMem(m_env.BITBLTBUF, GSVector4i(sx, sy, sx + w, sy + h)); + InvalidateVideoMem(m_env.BITBLTBUF, GSVector4i(dx, dy, dx + w, dy + h)); int xinc = 1; int yinc = 1; @@ -1102,10 +1055,10 @@ void GSState::Move() { for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) { - DWORD sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); int* soffset = spsm.rowOffset[sy & 7]; - DWORD dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); int* doffset = dpsm.rowOffset[dy & 7]; for(int x = 0; x < w; x++, sx += xinc, dx += xinc) @@ -1118,10 +1071,10 @@ void GSState::Move() { for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) { - DWORD sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); int* soffset = spsm.rowOffset[sy & 7]; - DWORD dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); int* doffset = dpsm.rowOffset[dy & 7]; for(int x = 0; x < w; x++, sx += xinc, dx += xinc) @@ -1132,7 +1085,7 @@ void GSState::Move() } } -void GSState::SoftReset(BYTE mask) +void GSState::SoftReset(uint32 mask) { if(mask & 1) memset(&m_path[0], 0, sizeof(GIFPath)); if(mask & 2) memset(&m_path[1], 0, sizeof(GIFPath)); @@ -1143,7 +1096,7 @@ void GSState::SoftReset(BYTE mask) m_q = 1; } -void GSState::ReadFIFO(BYTE* mem, int size) +void GSState::ReadFIFO(uint8* mem, int size) { GSPerfMonAutoTimer pmat(m_perfmon); @@ -1159,15 +1112,15 @@ void GSState::ReadFIFO(BYTE* mem, int size) } } -template void GSState::Transfer<0>(BYTE* mem, UINT32 size); -template void GSState::Transfer<1>(BYTE* mem, UINT32 size); -template void GSState::Transfer<2>(BYTE* mem, UINT32 size); +template void GSState::Transfer<0>(uint8* mem, uint32 size); +template void GSState::Transfer<1>(uint8* mem, uint32 size); +template void GSState::Transfer<2>(uint8* mem, uint32 size); -template void GSState::Transfer(BYTE* mem, UINT32 size) +template void GSState::Transfer(uint8* mem, uint32 size) { GSPerfMonAutoTimer pmat(m_perfmon); - BYTE* start = mem; + uint8* start = mem; GIFPath& path = m_path[index]; @@ -1192,7 +1145,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) if(path.tag.PRE && (path.tag.FLG & 2) == 0) { GIFReg r; - r.i64 = path.tag.PRIM; + r.u64 = path.tag.PRIM; (this->*m_fpGIFRegHandlers[GIF_A_D_REG_PRIM])(&r); } } @@ -1211,7 +1164,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) do { - (this->*m_fpGIFRegHandlers[(BYTE)((GIFPackedReg*)mem)->A_D.ADDR])(&((GIFPackedReg*)mem)->r); + (this->*m_fpGIFRegHandlers[((GIFPackedReg*)mem)->A_D.ADDR])(&((GIFPackedReg*)mem)->r); mem += sizeof(GIFPackedReg); } @@ -1221,7 +1174,7 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) { do { - DWORD reg = path.GetReg(); + uint32 reg = path.GetReg(); switch(reg) { @@ -1343,13 +1296,13 @@ template void GSState::Transfer(BYTE* mem, UINT32 size) } } -template static void WriteState(BYTE*& dst, T* src, size_t len = sizeof(T)) +template static void WriteState(uint8*& dst, T* src, size_t len = sizeof(T)) { memcpy(dst, src, len); dst += len; } -template static void ReadState(T* dst, BYTE*& src, size_t len = sizeof(T)) +template static void ReadState(T* dst, uint8*& src, size_t len = sizeof(T)) { memcpy(dst, src, len); src += len; @@ -1370,7 +1323,7 @@ int GSState::Freeze(GSFreezeData* fd, bool sizeonly) Flush(); - BYTE* data = fd->data; + uint8* data = fd->data; WriteState(data, &m_version); WriteState(data, &m_env.PRIM); @@ -1442,7 +1395,7 @@ int GSState::Defrost(const GSFreezeData* fd) return -1; } - BYTE* data = fd->data; + uint8* data = fd->data; int version; @@ -1495,7 +1448,7 @@ int GSState::Defrost(const GSFreezeData* fd) if(version <= 4) { - data += sizeof(DWORD) * 7; // skip + data += sizeof(uint32) * 7; // skip } } @@ -1536,7 +1489,7 @@ m_perfmon.SetFrame(5000); return 0; } -void GSState::SetGameCRC(DWORD crc, int options) +void GSState::SetGameCRC(uint32 crc, int options) { m_crc = crc; m_options = options; @@ -1608,7 +1561,7 @@ GSState::GSTransferBuffer::GSTransferBuffer() { x = y = 0; start = end = total = 0; - buff = (BYTE*)_aligned_malloc(1024 * 1024 * 4, 16); + buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); } GSState::GSTransferBuffer::~GSTransferBuffer() @@ -1653,12 +1606,12 @@ bool GSState::GSTransferBuffer::Update(int tw, int th, int bpp, int& len) struct GSFrameInfo { - DWORD FBP; - DWORD FPSM; - DWORD FBMSK; + uint32 FBP; + uint32 FPSM; + uint32 FBMSK; + uint32 TBP0; + uint32 TPSM; bool TME; - DWORD TBP0; - DWORD TPSM; }; typedef bool (*GetSkipCount)(const GSFrameInfo& fi, int& skip); diff --git a/plugins/GSdx/GSState.h b/plugins/GSdx/GSState.h index d22a814890..595a8b538f 100644 --- a/plugins/GSdx/GSState.h +++ b/plugins/GSdx/GSState.h @@ -117,7 +117,7 @@ class GSState : public GSAlignedClass<16> int x, y; int start, end, total; bool overflow; - BYTE* buff; + uint8* buff; GSTransferBuffer(); virtual ~GSTransferBuffer(); @@ -204,58 +204,49 @@ public: GSDrawingContext* m_context; GSVertex m_v; float m_q; - DWORD m_vprim; + uint32 m_vprim; GSPerfMon m_perfmon; - DWORD m_crc; + uint32 m_crc; int m_options; int m_frameskip; CRC::Game m_game; GSDump m_dump; public: - GSState(BYTE* base, bool mt, void (*irq)()); + GSState(uint8* base, bool mt, void (*irq)()); virtual ~GSState(); void ResetHandlers(); - CPoint GetDisplayPos(int i); - CSize GetDisplaySize(int i); - CRect GetDisplayRect(int i); - CSize GetDisplayPos(); - CSize GetDisplaySize(); - CRect GetDisplayRect(); - CPoint GetFramePos(int i); - CSize GetFrameSize(int i); - CRect GetFrameRect(int i); - CSize GetFramePos(); - CSize GetFrameSize(); - CRect GetFrameRect(); - CSize GetDeviceSize(int i); - CSize GetDeviceSize(); + GSVector4i GetDisplayRect(int i = -1); + GSVector4i GetFrameRect(int i = -1); + GSVector4i GetDeviceRect(int i = -1); + bool IsEnabled(int i); + int GetFPS(); virtual void Reset(); virtual void Flush(); virtual void FlushPrim() = 0; virtual void ResetPrim() = 0; - virtual void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, CRect r) {} - virtual void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, CRect r) {} + virtual void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) {} + virtual void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) {} virtual void InvalidateTextureCache() {} void Move(); - void Write(BYTE* mem, int len); - void Read(BYTE* mem, int len); + void Write(uint8* mem, int len); + void Read(uint8* mem, int len); - void SoftReset(BYTE mask); - void WriteCSR(UINT32 csr) {m_regs->CSR.ai32[1] = csr;} - void ReadFIFO(BYTE* mem, int size); - template void Transfer(BYTE* mem, UINT32 size); + void SoftReset(uint32 mask); + void WriteCSR(uint32 csr) {m_regs->CSR.u32[1] = csr;} + void ReadFIFO(uint8* mem, int size); + template void Transfer(uint8* mem, uint32 size); int Freeze(GSFreezeData* fd, bool sizeonly); int Defrost(const GSFreezeData* fd); - void GetLastTag(UINT32* tag) {*tag = m_path3hack; m_path3hack = 0;} - virtual void SetGameCRC(DWORD crc, int options); + void GetLastTag(uint32* tag) {*tag = m_path3hack; m_path3hack = 0;} + virtual void SetGameCRC(uint32 crc, int options); void SetFrameSkip(int frameskip); }; diff --git a/plugins/GSdx/GSTables.cpp b/plugins/GSdx/GSTables.cpp index ec0aa149e7..f24b9ec179 100644 --- a/plugins/GSdx/GSTables.cpp +++ b/plugins/GSdx/GSTables.cpp @@ -22,21 +22,21 @@ #include "StdAfx.h" #include "GSTables.h" -const BYTE blockTable32[4][8] = { +const uint8 blockTable32[4][8] = { { 0, 1, 4, 5, 16, 17, 20, 21}, { 2, 3, 6, 7, 18, 19, 22, 23}, { 8, 9, 12, 13, 24, 25, 28, 29}, { 10, 11, 14, 15, 26, 27, 30, 31} }; -const BYTE blockTable32Z[4][8] = { +const uint8 blockTable32Z[4][8] = { { 24, 25, 28, 29, 8, 9, 12, 13}, { 26, 27, 30, 31, 10, 11, 14, 15}, { 16, 17, 20, 21, 0, 1, 4, 5}, { 18, 19, 22, 23, 2, 3, 6, 7} }; -const BYTE blockTable16[8][4] = { +const uint8 blockTable16[8][4] = { { 0, 2, 8, 10 }, { 1, 3, 9, 11 }, { 4, 6, 12, 14 }, @@ -47,7 +47,7 @@ const BYTE blockTable16[8][4] = { { 21, 23, 29, 31 } }; -const BYTE blockTable16S[8][4] = { +const uint8 blockTable16S[8][4] = { { 0, 2, 16, 18 }, { 1, 3, 17, 19 }, { 8, 10, 24, 26 }, @@ -58,7 +58,7 @@ const BYTE blockTable16S[8][4] = { { 13, 15, 29, 31 } }; -const BYTE blockTable16Z[8][4] = { +const uint8 blockTable16Z[8][4] = { { 24, 26, 16, 18 }, { 25, 27, 17, 19 }, { 28, 30, 20, 22 }, @@ -69,7 +69,7 @@ const BYTE blockTable16Z[8][4] = { { 13, 15, 5, 7 } }; -const BYTE blockTable16SZ[8][4] = { +const uint8 blockTable16SZ[8][4] = { { 24, 26, 8, 10 }, { 25, 27, 9, 11 }, { 16, 18, 0, 2 }, @@ -80,14 +80,14 @@ const BYTE blockTable16SZ[8][4] = { { 21, 23, 5, 7 } }; -const BYTE blockTable8[4][8] = { +const uint8 blockTable8[4][8] = { { 0, 1, 4, 5, 16, 17, 20, 21}, { 2, 3, 6, 7, 18, 19, 22, 23}, { 8, 9, 12, 13, 24, 25, 28, 29}, { 10, 11, 14, 15, 26, 27, 30, 31} }; -const BYTE blockTable4[8][4] = { +const uint8 blockTable4[8][4] = { { 0, 2, 8, 10 }, { 1, 3, 9, 11 }, { 4, 6, 12, 14 }, @@ -98,7 +98,7 @@ const BYTE blockTable4[8][4] = { { 21, 23, 29, 31 } }; -const BYTE columnTable32[8][8] = { +const uint8 columnTable32[8][8] = { { 0, 1, 4, 5, 8, 9, 12, 13 }, { 2, 3, 6, 7, 10, 11, 14, 15 }, { 16, 17, 20, 21, 24, 25, 28, 29 }, @@ -109,7 +109,7 @@ const BYTE columnTable32[8][8] = { { 50, 51, 54, 55, 58, 59, 62, 63 }, }; -const BYTE columnTable16[8][16] = { +const uint8 columnTable16[8][16] = { { 0, 2, 8, 10, 16, 18, 24, 26, 1, 3, 9, 11, 17, 19, 25, 27 }, { 4, 6, 12, 14, 20, 22, 28, 30, @@ -128,7 +128,7 @@ const BYTE columnTable16[8][16] = { 101, 103, 109, 111, 117, 119, 125, 127 }, }; -const BYTE columnTable8[16][16] = { +const uint8 columnTable8[16][16] = { { 0, 4, 16, 20, 32, 36, 48, 52, // column 0 2, 6, 18, 22, 34, 38, 50, 54 }, { 8, 12, 24, 28, 40, 44, 56, 60, @@ -163,7 +163,7 @@ const BYTE columnTable8[16][16] = { 203, 207, 219, 223, 235, 239, 251, 255 }, }; -const WORD columnTable4[16][32] = { +const uint16 columnTable4[16][32] = { { 0, 8, 32, 40, 64, 72, 96, 104, // column 0 2, 10, 34, 42, 66, 74, 98, 106, 4, 12, 36, 44, 68, 76, 100, 108, @@ -230,7 +230,7 @@ const WORD columnTable4[16][32] = { 407, 415, 439, 447, 471, 479, 503, 511 }, }; -const BYTE clutTableT32I8[128] = +const uint8 clutTableT32I8[128] = { 0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15, 64, 65, 68, 69, 72, 73, 76, 77, 66, 67, 70, 71, 74, 75, 78, 79, @@ -242,13 +242,13 @@ const BYTE clutTableT32I8[128] = 112, 113, 116, 117, 120, 121, 124, 125, 114, 115, 118, 119, 122, 123, 126, 127 }; -const BYTE clutTableT32I4[16] = +const uint8 clutTableT32I4[16] = { 0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15 }; -const BYTE clutTableT16I8[32] = +const uint8 clutTableT16I8[32] = { 0, 2, 8, 10, 16, 18, 24, 26, 4, 6, 12, 14, 20, 22, 28, 30, @@ -256,7 +256,7 @@ const BYTE clutTableT16I8[32] = 5, 7, 13, 15, 21, 23, 29, 31 }; -const BYTE clutTableT16I4[16] = +const uint8 clutTableT16I4[16] = { 0, 2, 8, 10, 16, 18, 24, 26, 4, 6, 12, 14, 20, 22, 28, 30 diff --git a/plugins/GSdx/GSTables.h b/plugins/GSdx/GSTables.h index 1ed40beea8..258a22d0e8 100644 --- a/plugins/GSdx/GSTables.h +++ b/plugins/GSdx/GSTables.h @@ -21,19 +21,19 @@ #pragma once -extern const BYTE blockTable32[4][8]; -extern const BYTE blockTable32Z[4][8]; -extern const BYTE blockTable16[8][4]; -extern const BYTE blockTable16S[8][4]; -extern const BYTE blockTable16Z[8][4]; -extern const BYTE blockTable16SZ[8][4]; -extern const BYTE blockTable8[4][8]; -extern const BYTE blockTable4[8][4]; -extern const BYTE columnTable32[8][8]; -extern const BYTE columnTable16[8][16]; -extern const BYTE columnTable8[16][16]; -extern const WORD columnTable4[16][32]; -extern const BYTE clutTableT32I8[128]; -extern const BYTE clutTableT32I4[16]; -extern const BYTE clutTableT16I8[32]; -extern const BYTE clutTableT16I4[16]; +extern const uint8 blockTable32[4][8]; +extern const uint8 blockTable32Z[4][8]; +extern const uint8 blockTable16[8][4]; +extern const uint8 blockTable16S[8][4]; +extern const uint8 blockTable16Z[8][4]; +extern const uint8 blockTable16SZ[8][4]; +extern const uint8 blockTable8[4][8]; +extern const uint8 blockTable4[8][4]; +extern const uint8 columnTable32[8][8]; +extern const uint8 columnTable16[8][16]; +extern const uint8 columnTable8[16][16]; +extern const uint16 columnTable4[16][32]; +extern const uint8 clutTableT32I8[128]; +extern const uint8 clutTableT32I4[16]; +extern const uint8 clutTableT16I8[32]; +extern const uint8 clutTableT16I4[16]; diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index 39d8af3a27..6f485905bd 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -39,10 +39,10 @@ public: virtual int GetWidth() const = 0; virtual int GetHeight() const = 0; virtual int GetFormat() const = 0; - virtual bool Update(const CRect& r, const void* data, int pitch) = 0; - virtual bool Map(BYTE** bits, int& pitch, const RECT* r = NULL) = 0; + virtual bool Update(const GSVector4i& r, const void* data, int pitch) = 0; + virtual bool Map(uint8** bits, int& pitch) = 0; virtual void Unmap() = 0; virtual bool Save(const string& fn, bool dds = false) = 0; - CSize GetSize() {return CSize(GetWidth(), GetHeight());} + GSVector2i GetSize() {return GSVector2i(GetWidth(), GetHeight());} }; diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 4b14de0539..051cbb07fd 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -69,7 +69,7 @@ int GSTexture10::GetFormat() const return m_desc.Format; } -bool GSTexture10::Update(const CRect& r, const void* data, int pitch) +bool GSTexture10::Update(const GSVector4i& r, const void* data, int pitch) { if(m_dev && m_texture) { @@ -83,7 +83,7 @@ bool GSTexture10::Update(const CRect& r, const void* data, int pitch) return false; } -bool GSTexture10::Map(BYTE** bits, int& pitch, const RECT* r) +bool GSTexture10::Map(uint8** bits, int& pitch) { if(m_texture) { @@ -91,7 +91,7 @@ bool GSTexture10::Map(BYTE** bits, int& pitch, const RECT* r) if(SUCCEEDED(m_texture->Map(0, D3D10_MAP_READ_WRITE, 0, &map))) { - *bits = (BYTE*)map.pData; + *bits = (uint8*)map.pData; pitch = (int)map.RowPitch; return true; @@ -143,14 +143,14 @@ bool GSTexture10::Save(const string& fn, bool dds) hr = src->Map(0, D3D10_MAP_READ, 0, &sm); hr = dst->Map(0, D3D10_MAP_WRITE, 0, &dm); - BYTE* s = (BYTE*)sm.pData; - BYTE* d = (BYTE*)dm.pData; + uint8* s = (uint8*)sm.pData; + uint8* d = (uint8*)dm.pData; - for(UINT y = 0; y < desc.Height; y++, s += sm.RowPitch, d += dm.RowPitch) + for(uint32 y = 0; y < desc.Height; y++, s += sm.RowPitch, d += dm.RowPitch) { - for(UINT x = 0; x < desc.Width; x++) + for(uint32 x = 0; x < desc.Width; x++) { - ((UINT*)d)[x] = (UINT)(((float*)s)[x*2] * UINT_MAX); + ((uint32*)d)[x] = (uint32)(((float*)s)[x*2] * UINT_MAX); } } diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index 3609224696..e01ae272f2 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -43,8 +43,8 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; - bool Update(const CRect& r, const void* data, int pitch); - bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); + bool Update(const GSVector4i& r, const void* data, int pitch); + bool Map(uint8** bits, int& pitch); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture7.cpp b/plugins/GSdx/GSTexture7.cpp index c817e2130d..fd2d2dcbef 100644 --- a/plugins/GSdx/GSTexture7.cpp +++ b/plugins/GSdx/GSTexture7.cpp @@ -80,11 +80,11 @@ int GSTexture7::GetFormat() const return (int)m_desc.ddpfPixelFormat.dwFourCC; } -bool GSTexture7::Update(const CRect& r, const void* data, int pitch) +bool GSTexture7::Update(const GSVector4i& r, const void* data, int pitch) { HRESULT hr; - CRect r2 = r; + GSVector4i r2 = r; DDSURFACEDESC2 desc; @@ -92,14 +92,14 @@ bool GSTexture7::Update(const CRect& r, const void* data, int pitch) desc.dwSize = sizeof(desc); - if(SUCCEEDED(hr = m_system->Lock(&r2, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR | DDLOCK_WRITEONLY, NULL))) + if(SUCCEEDED(hr = m_system->Lock(r2, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR | DDLOCK_WRITEONLY, NULL))) { - BYTE* src = (BYTE*)data; - BYTE* dst = (BYTE*)desc.lpSurface; + uint8* src = (uint8*)data; + uint8* dst = (uint8*)desc.lpSurface; int bytes = min(pitch, desc.lPitch); - for(int i = 0, j = r.Height(); i < j; i++, src += pitch, dst += desc.lPitch) + for(int i = 0, j = r.height(); i < j; i++, src += pitch, dst += desc.lPitch) { // memcpy(dst, src, bytes); @@ -120,11 +120,11 @@ bool GSTexture7::Update(const CRect& r, const void* data, int pitch) } } - hr = m_system->Unlock(&r2); + hr = m_system->Unlock(r2); if(m_video) { - hr = m_video->Blt(&r2, m_system, &r2, DDBLT_WAIT, NULL); + hr = m_video->Blt(r2, m_system, r2, DDBLT_WAIT, NULL); } return true; @@ -133,21 +133,17 @@ bool GSTexture7::Update(const CRect& r, const void* data, int pitch) return false; } -bool GSTexture7::Map(BYTE** bits, int& pitch, const RECT* r) +bool GSTexture7::Map(uint8** bits, int& pitch) { HRESULT hr; - CRect r2 = r; - DDSURFACEDESC2 desc; - if(SUCCEEDED(hr = m_system->Lock(&r2, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR, NULL))) + if(SUCCEEDED(hr = m_system->Lock(NULL, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR, NULL))) { - *bits = (BYTE*)desc.lpSurface; + *bits = (uint8*)desc.lpSurface; pitch = (int)desc.lPitch; - m_lr = r; - return true; } @@ -162,7 +158,7 @@ void GSTexture7::Unmap() if(m_video) { - hr = m_video->Blt(&m_lr, m_system, &m_lr, DDBLT_WAIT, NULL); + hr = m_video->Blt(NULL, m_system, NULL, DDBLT_WAIT, NULL); } } diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index b2bdc0656a..fafccbd863 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -30,7 +30,6 @@ class GSTexture7 : public GSTexture CComPtr m_system; CComPtr m_video; DDSURFACEDESC2 m_desc; - CRect m_lr; public: GSTexture7(); @@ -44,8 +43,8 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; - bool Update(const CRect& r, const void* data, int pitch); - bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); + bool Update(const GSVector4i& r, const void* data, int pitch); + bool Map(uint8** bits, int& pitch); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index b12fc2f684..bbb42e0420 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -83,7 +83,7 @@ int GSTexture9::GetFormat() const return m_desc.Format; } -bool GSTexture9::Update(const CRect& r, const void* data, int pitch) +bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) { if(CComPtr surface = *this) { @@ -91,12 +91,12 @@ bool GSTexture9::Update(const CRect& r, const void* data, int pitch) if(SUCCEEDED(surface->LockRect(&lr, r, 0))) { - BYTE* src = (BYTE*)data; - BYTE* dst = (BYTE*)lr.pBits; + uint8* src = (uint8*)data; + uint8* dst = (uint8*)lr.pBits; int bytes = min(pitch, lr.Pitch); - for(int i = 0, j = r.Height(); i < j; i++, src += pitch, dst += lr.Pitch) + for(int i = 0, j = r.height(); i < j; i++, src += pitch, dst += lr.Pitch) { memcpy(dst, src, bytes); } @@ -110,7 +110,7 @@ bool GSTexture9::Update(const CRect& r, const void* data, int pitch) return false; } -bool GSTexture9::Map(BYTE** bits, int& pitch, const RECT* r) +bool GSTexture9::Map(uint8** bits, int& pitch) { HRESULT hr; @@ -118,9 +118,9 @@ bool GSTexture9::Map(BYTE** bits, int& pitch, const RECT* r) { D3DLOCKED_RECT lr; - if(SUCCEEDED(hr = surface->LockRect(&lr, r, 0))) + if(SUCCEEDED(hr = surface->LockRect(&lr, NULL, 0))) { - *bits = (BYTE*)lr.pBits; + *bits = (uint8*)lr.pBits; pitch = (int)lr.Pitch; return true; @@ -160,12 +160,12 @@ bool GSTexture9::Save(const string& fn, bool dds) hr = m_surface->LockRect(&slr, NULL, 0); hr = surface->LockRect(&dlr, NULL, 0); - BYTE* s = (BYTE*)slr.pBits; - BYTE* d = (BYTE*)dlr.pBits; + uint8* s = (uint8*)slr.pBits; + uint8* d = (uint8*)dlr.pBits; - for(UINT y = 0; y < desc.Height; y++, s += slr.Pitch, d += dlr.Pitch) + for(uint32 y = 0; y < desc.Height; y++, s += slr.Pitch, d += dlr.Pitch) { - for(UINT x = 0; x < desc.Width; x++) + for(uint32 x = 0; x < desc.Width; x++) { ((float*)d)[x] = ((float*)s)[x]; } diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index fe46918e34..dff127d4ac 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -42,8 +42,8 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; - bool Update(const CRect& r, const void* data, int pitch); - bool Map(BYTE** bits, int& pitch, const RECT* r = NULL); + bool Update(const GSVector4i& r, const void* data, int pitch); + bool Map(uint8** bits, int& pitch); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 6168bc88a4..d409d54f30 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -28,7 +28,7 @@ template class GSTextureCache typedef typename Device::Texture Texture; public: - class GSSurface + class GSSurface : public GSAlignedClass<16> { protected: GSRenderer* m_renderer; @@ -48,7 +48,7 @@ public: , m_age(0) , m_initpalette(false) { - m_TEX0.TBP0 = (UINT32)~0; + m_TEX0.TBP0 = (uint32)~0; } virtual ~GSSurface() @@ -81,7 +81,7 @@ public: return m_renderer->m_dev.CreateRenderTarget(m_texture, w, h); } - virtual void Read(CRect r) = 0; + virtual void Read(const GSVector4i& r) = 0; }; class GSDepthStencil : public GSSurface @@ -107,30 +107,30 @@ public: class GSTexture : public GSSurface { protected: - bool GetDirtyRect(CRect& rr) + bool GetDirtyRect(GSVector4i& rr) { int w = 1 << m_TEX0.TW; int h = 1 << m_TEX0.TH; - CRect r(0, 0, w, h); + GSVector4i r(0, 0, w, h); for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) { - const CRect& dirty = i->GetDirtyRect(m_TEX0) & r; + const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(r); - if(!(m_valid & dirty).IsRectEmpty()) + if(!m_valid.rintersect(dirty).rempty()) { // find the rect having the largest area, outside dirty, inside m_valid - CRect left(m_valid.left, m_valid.top, min(m_valid.right, dirty.left), m_valid.bottom); - CRect top(m_valid.left, m_valid.top, m_valid.right, min(m_valid.bottom, dirty.top)); - CRect right(max(m_valid.left, dirty.right), m_valid.top, m_valid.right, m_valid.bottom); - CRect bottom(m_valid.left, max(m_valid.top, dirty.bottom), m_valid.right, m_valid.bottom); + GSVector4i left(m_valid.left, m_valid.top, min(m_valid.right, dirty.left), m_valid.bottom); + GSVector4i top(m_valid.left, m_valid.top, m_valid.right, min(m_valid.bottom, dirty.top)); + GSVector4i right(max(m_valid.left, dirty.right), m_valid.top, m_valid.right, m_valid.bottom); + GSVector4i bottom(m_valid.left, max(m_valid.top, dirty.bottom), m_valid.right, m_valid.bottom); - int leftsize = !left.IsRectEmpty() ? left.Width() * left.Height() : 0; - int topsize = !top.IsRectEmpty() ? top.Width() * top.Height() : 0; - int rightsize = !right.IsRectEmpty() ? right.Width() * right.Height() : 0; - int bottomsize = !bottom.IsRectEmpty() ? bottom.Width() * bottom.Height() : 0; + int leftsize = !left.rempty() ? left.width() * left.height() : 0; + int topsize = !top.rempty() ? top.width() * top.height() : 0; + int rightsize = !right.rempty() ? right.width() * right.height() : 0; + int bottomsize = !bottom.rempty() ? bottom.width() * bottom.height() : 0; // TODO: sort @@ -139,7 +139,7 @@ public: topsize > 0 ? top : rightsize > 0 ? right : bottomsize > 0 ? bottom : - CRect(0, 0, 0, 0); + GSVector4i::zero(); } } @@ -167,10 +167,10 @@ public: } else { - r |= m_valid; + r = r.runion(m_valid); } - if(r.IsRectEmpty()) + if(r.rempty()) { return false; } @@ -182,8 +182,8 @@ public: public: GIFRegCLAMP m_CLAMP; - DWORD* m_clut; // * - CRect m_valid; + uint32* m_clut; // * + GSVector4i m_valid; int m_bpp; int m_bpp2; bool m_rendered; @@ -195,7 +195,7 @@ public: , m_bpp2(0) , m_rendered(false) { - m_clut = (DWORD*)_aligned_malloc(256 * sizeof(DWORD), 16); + m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); memset(m_clut, 0, sizeof(m_clut)); } @@ -205,6 +205,41 @@ public: _aligned_free(m_clut); } + void Update() + { + __super::Update(); + + if(m_rendered) + { + return; + } + + GSVector4i r; + + if(!GetDirtyRect(r)) + { + return; + } + + m_valid = m_valid.runion(r); + + static uint8* bits = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); + static int pitch = 1024 * 4; + + if(m_renderer->m_psrr) + { + m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + else + { + m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + + m_texture.Update(r, bits, pitch); + + m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * m_bpp >> 3); + } + virtual bool Create() = 0; virtual bool Create(GSRenderTarget* rt) = 0; virtual bool Create(GSDepthStencil* ds) = 0; @@ -335,10 +370,12 @@ public: if(m_renderer->CanUpscale()) { - int ww = (int)(m_renderer->GetFramePos().cx + rt->m_TEX0.TBW * 64); - int hh = (int)(m_renderer->GetFramePos().cy + m_renderer->GetDisplaySize().cy); + GSVector4i fr = m_renderer->GetFrameRect(); - if(hh <= m_renderer->GetDeviceSize().cy / 2) + int ww = (int)(fr.left + rt->m_TEX0.TBW * 64); + int hh = (int)(fr.top + m_renderer->GetDisplayRect().height()); + + if(hh <= m_renderer->GetDeviceRect().height() / 2) { hh *= 2; } @@ -412,7 +449,7 @@ public: const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; - const DWORD* clut = m_renderer->m_mem.m_clut; + const uint32* clut = m_renderer->m_mem.m_clut; const int pal = GSLocalMemory::m_psm[TEX0.PSM].pal; if(pal > 0) @@ -480,7 +517,7 @@ public: { if(TEX0.PSM == t->m_TEX0.PSM && TEX0.TBW == t->m_TEX0.TBW && TEX0.TW == t->m_TEX0.TW && TEX0.TH == t->m_TEX0.TH - && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.i64 == t->m_CLAMP.i64)) + && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.u64 == t->m_CLAMP.u64)) && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) { m_tex.splice(m_tex.begin(), m_tex, i); @@ -563,14 +600,14 @@ public: if(t->m_initpalette) { memcpy(t->m_clut, clut, size); - t->m_palette.Update(CRect(0, 0, pal, 1), t->m_clut, size); + t->m_palette.Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); t->m_initpalette = false; } else { if(GSVector4i::update(t->m_clut, clut, size)) { - t->m_palette.Update(CRect(0, 0, pal, 1), t->m_clut, size); + t->m_palette.Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); } } } @@ -603,7 +640,7 @@ public: } } - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& r) + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { bool found = false; @@ -617,7 +654,7 @@ public: { if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) { - t->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); + t->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); found = true; } @@ -636,16 +673,16 @@ public: if(rowsize > 0 && offset % rowsize == 0) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.cy * offset / rowsize; + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - CRect r2(r.left, r.top + y, r.right, r.bottom + y); + GSVector4i r2(r.left, r.top + y, r.right, r.bottom + y); int w = 1 << t->m_TEX0.TW; int h = 1 << t->m_TEX0.TH; if(r2.bottom > 0 && r2.top < h && r2.right > 0 && r2.left < w) { - t->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r2)); + t->m_dirty.push_back(GSDirtyRect(r2, BITBLTBUF.DPSM)); } } } @@ -662,7 +699,7 @@ public: { if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM)) { - rt->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); + rt->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); rt->m_TEX0.TBW = BITBLTBUF.DBW; } else @@ -675,17 +712,17 @@ public: if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM) && BITBLTBUF.DBP < rt->m_TEX0.TBP0) { - DWORD rowsize = BITBLTBUF.DBW * 8192; - DWORD offset = (DWORD)((rt->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); + uint32 rowsize = BITBLTBUF.DBW * 8192; + uint32 offset = (uint32)((rt->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); if(rowsize > 0 && offset % rowsize == 0) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.cy * offset / rowsize; + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; if(r.bottom > y) { // TODO: do not add this rect above too - rt->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); + rt->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); rt->m_TEX0.TBW = BITBLTBUF.DBW; continue; } @@ -705,7 +742,7 @@ public: { if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM)) { - ds->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, r)); + ds->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); ds->m_TEX0.TBW = BITBLTBUF.DBW; } else @@ -718,17 +755,17 @@ public: if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM) && BITBLTBUF.DBP < ds->m_TEX0.TBP0) { - DWORD rowsize = BITBLTBUF.DBW * 8192; - DWORD offset = (DWORD)((ds->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); + uint32 rowsize = BITBLTBUF.DBW * 8192; + uint32 offset = (uint32)((ds->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); if(rowsize > 0 && offset % rowsize == 0) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.cy * offset / rowsize; + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; if(r.bottom > y) { // TODO: do not add this rect above too - ds->m_dirty.push_back(GSDirtyRect(BITBLTBUF.DPSM, CRect(r.left, r.top - y, r.right, r.bottom - y))); + ds->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); ds->m_TEX0.TBW = BITBLTBUF.DBW; continue; } @@ -738,7 +775,7 @@ public: } } - void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& r) + void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) { @@ -757,7 +794,7 @@ public: { // ffx-2 riku changing to her default (shoots some reflecting glass at the end), 16-bit rt read as 32-bit - rt->Read(CRect(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); + rt->Read(GSVector4i(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); return; } else @@ -784,12 +821,12 @@ public: { // ffx2 pause screen background - DWORD rowsize = BITBLTBUF.SBW * 8192; - DWORD offset = (DWORD)((BITBLTBUF.SBP - rt->m_TEX0.TBP0) * 256); + uint32 rowsize = BITBLTBUF.SBW * 8192; + uint32 offset = (uint32)((BITBLTBUF.SBP - rt->m_TEX0.TBP0) * 256); if(rowsize > 0 && offset % rowsize == 0) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.SPSM].pgs.cy * offset / rowsize; + int y = m_renderer->m_mem.m_psm[BITBLTBUF.SPSM].pgs.y * offset / rowsize; if(y < ymin && y < 512) { @@ -802,7 +839,7 @@ public: if(rt2) { - rt2->Read(CRect(r.left, r.top + ymin, r.right, r.bottom + ymin)); + rt2->Read(GSVector4i(r.left, r.top + ymin, r.right, r.bottom + ymin)); } // TODO: ds diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 0ab8210bdc..6a59e815dd 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -37,16 +37,14 @@ void GSTextureCache10::GSRenderTargetHW10::Update() // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - CRect r = m_dirty.GetDirtyRect(m_TEX0, m_texture.GetSize()); + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture.GetSize()); - m_dirty.clear(); + if(r.rempty()) return; - if(r.IsRectEmpty()) return; + int w = r.width(); + int h = r.height(); - int w = r.Width(); - int h = r.Height(); - - static BYTE* buff = (BYTE*)_aligned_malloc(1024 * 1024 * 4, 16); + static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); static int pitch = 1024 * 4; GIFRegTEXA TEXA; @@ -69,7 +67,7 @@ void GSTextureCache10::GSRenderTargetHW10::Update() if(!m_renderer->m_dev.CreateTexture(texture, w, h)) return; - texture.Update(CRect(0, 0, w, h), buff, pitch); + texture.Update(GSVector4i(0, 0, w, h), buff, pitch); GSVector4 dr( m_texture.m_scale.x * r.left, @@ -82,7 +80,7 @@ void GSTextureCache10::GSRenderTargetHW10::Update() m_renderer->m_dev.Recycle(texture); } -void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) +void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) { if(m_TEX0.PSM != PSM_PSMCT32 && m_TEX0.PSM != PSM_PSMCT24 @@ -102,8 +100,8 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) // m_renderer->m_perfmon.Put(GSPerfMon::ReadRT, 1); - int w = r.Width(); - int h = r.Height(); + int w = r.width(); + int h = r.height(); GSVector4 src; @@ -119,15 +117,15 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) if(!m_renderer->m_dev.CopyOffscreen(m_texture, src, offscreen, w, h, format)) return; - BYTE* bits; + uint8* bits; int pitch; if(offscreen.Map(&bits, pitch)) { // TODO: block level write - DWORD bp = m_TEX0.TBP0; - DWORD bw = m_TEX0.TBW; + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; @@ -135,12 +133,12 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((DWORD*)bits)[i]); + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); } } } @@ -148,12 +146,12 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((DWORD*)bits)[i]); + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); } } } @@ -161,12 +159,12 @@ void GSTextureCache10::GSRenderTargetHW10::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel16(addr + offset[x], ((WORD*)bits)[i]); + m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)bits)[i]); } } } @@ -199,7 +197,7 @@ bool GSTextureCache10::GSTextureHW10::Create() m_TEX0 = m_renderer->m_context->TEX0; m_CLAMP = m_renderer->m_context->CLAMP; - DWORD psm = m_TEX0.PSM; + uint32 psm = m_TEX0.PSM; switch(psm) { @@ -413,40 +411,3 @@ bool GSTextureCache10::GSTextureHW10::Create(GSDepthStencil* ds) return false; } - -void GSTextureCache10::GSTextureHW10::Update() -{ - __super::Update(); - - if(m_rendered) - { - return; - } - - CRect r; - - if(!GetDirtyRect(r)) - { - return; - } - - m_valid |= r; - - //TRACE(_T("GSTexture::Update %d,%d - %d,%d (%08x)\n"), r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); - - static BYTE* bits = (BYTE*)::_aligned_malloc(1024 * 1024 * 4, 16); - int pitch = 1024 * 4; - - if(m_renderer->m_psrr) - { - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - else - { - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - - m_texture.Update(r, bits, pitch); - - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * m_bpp >> 3); -} diff --git a/plugins/GSdx/GSTextureCache10.h b/plugins/GSdx/GSTextureCache10.h index 48b304325a..7076c62594 100644 --- a/plugins/GSdx/GSTextureCache10.h +++ b/plugins/GSdx/GSTextureCache10.h @@ -34,7 +34,7 @@ class GSTextureCache10 : public GSTextureCache explicit GSRenderTargetHW10(GSRenderer* renderer) : GSRenderTarget(renderer) {} void Update(); - void Read(CRect r); + void Read(const GSVector4i& r); }; class GSDepthStencilHW10 : public GSDepthStencil @@ -53,7 +53,6 @@ class GSTextureCache10 : public GSTextureCache bool Create(); bool Create(GSRenderTarget* rt); bool Create(GSDepthStencil* ds); - void Update(); }; protected: diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index 52f733b4de..8ae4d90c42 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -37,21 +37,19 @@ void GSTextureCache9::GSRenderTarget9::Update() // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - CRect r = m_dirty.GetDirtyRect(m_TEX0, m_texture.GetSize()); + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture.GetSize()); - m_dirty.clear(); + if(r.rempty()) return; - if(r.IsRectEmpty()) return; - - int w = r.Width(); - int h = r.Height(); + int w = r.width(); + int h = r.height(); Texture texture; if(!m_renderer->m_dev.CreateTexture(texture, w, h)) return; - BYTE* bits; + uint8* bits; int pitch; if(texture.Map(&bits, pitch)) @@ -85,7 +83,7 @@ void GSTextureCache9::GSRenderTarget9::Update() m_renderer->m_dev.Recycle(texture); } -void GSTextureCache9::GSRenderTarget9::Read(CRect r) +void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) { if(m_TEX0.PSM != PSM_PSMCT32 && m_TEX0.PSM != PSM_PSMCT24 @@ -105,8 +103,8 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) // m_renderer->m_perfmon.Put(GSPerfMon::ReadRT, 1); - int w = r.Width(); - int h = r.Height(); + int w = r.width(); + int h = r.height(); GSVector4 src; @@ -120,15 +118,15 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) if(!m_renderer->m_dev.CopyOffscreen(m_texture, src, offscreen, w, h)) return; - BYTE* bits; + uint8* bits; int pitch; if(offscreen.Map(&bits, pitch)) { // TODO: block level write - DWORD bp = m_TEX0.TBP0; - DWORD bw = m_TEX0.TBW; + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; @@ -136,12 +134,12 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((DWORD*)bits)[i]); + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); } } } @@ -149,12 +147,12 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((DWORD*)bits)[i]); + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); } } } @@ -162,12 +160,12 @@ void GSTextureCache9::GSRenderTarget9::Read(CRect r) { for(int y = r.top; y < r.bottom; y++, bits += pitch) { - DWORD addr = pa(0, y, bp, bw); + uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WriteFrame16(addr + offset[x], ((DWORD*)bits)[i]); + m_renderer->m_mem.WriteFrame16(addr + offset[x], ((uint32*)bits)[i]); } } } @@ -200,7 +198,7 @@ bool GSTextureCache9::GSTexture9::Create() m_TEX0 = m_renderer->m_context->TEX0; m_CLAMP = m_renderer->m_context->CLAMP; - DWORD psm = m_TEX0.PSM; + uint32 psm = m_TEX0.PSM; switch(psm) { @@ -358,7 +356,7 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) { - CRect r(0, 0, w, h); + GSVector4i r(0, 0, w, h); m_renderer->m_dev->StretchRect(*st, r, *dt, r, D3DTEXF_POINT); } @@ -413,41 +411,3 @@ bool GSTextureCache9::GSTexture9::Create(GSDepthStencil* ds) return false; } - -void GSTextureCache9::GSTexture9::Update() -{ - __super::Update(); - - if(m_rendered) - { - return; - } - - CRect r; - - if(!GetDirtyRect(r)) - { - return; - } - - m_valid |= r; - - BYTE* bits; - int pitch; - - if(m_texture.Map(&bits, pitch, &r)) - { - if(m_renderer->m_psrr) - { - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - else - { - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - - m_texture.Unmap(); - } - - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * m_bpp >> 3); -} diff --git a/plugins/GSdx/GSTextureCache9.h b/plugins/GSdx/GSTextureCache9.h index e0f73ec01a..e7ebef5ea1 100644 --- a/plugins/GSdx/GSTextureCache9.h +++ b/plugins/GSdx/GSTextureCache9.h @@ -34,7 +34,7 @@ class GSTextureCache9 : public GSTextureCache explicit GSRenderTarget9(GSRenderer* renderer) : GSRenderTarget(renderer) {} void Update(); - void Read(CRect r); + void Read(const GSVector4i& r); }; class GSDepthStencil9 : public GSDepthStencil @@ -53,7 +53,6 @@ class GSTextureCache9 : public GSTextureCache bool Create(); bool Create(GSRenderTarget* rt); bool Create(GSDepthStencil* ds); - void Update(); }; protected: diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index 2aed3b6b64..a9433b9ccb 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -32,7 +32,7 @@ GSTextureCacheSW::~GSTextureCacheSW() RemoveAll(); } -const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* r) +const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r) { GSLocalMemory& mem = m_state->m_mem; @@ -47,12 +47,12 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE GSTexture* t2 = (*i).first; // if(t2->m_TEX0.TBP0 != TEX0.TBP0 || t2->m_TEX0.TBW != TEX0.TBW || t2->m_TEX0.PSM != TEX0.PSM || t2->m_TEX0.TW != TEX0.TW || t2->m_TEX0.TH != TEX0.TH) - if(((t2->m_TEX0.ai32[0] ^ TEX0.ai32[0]) | ((t2->m_TEX0.ai32[1] ^ TEX0.ai32[1]) & 3)) != 0) + if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) { continue; } - if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && !(t2->m_TEXA == (GSVector4i)TEXA).alltrue())) + if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && !((GSVector4i)TEXA).eq(t2->m_TEXA))) { continue; } @@ -73,18 +73,18 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE int tw = 1 << TEX0.TW; int th = 1 << TEX0.TH; - DWORD bp = TEX0.TBP0; - DWORD bw = TEX0.TBW; + uint32 bp = TEX0.TBP0; + uint32 bw = TEX0.TBW; - CSize s = (bp & 31) == 0 ? psm.pgs : psm.bs; + GSVector2i s = (bp & 31) == 0 ? psm.pgs : psm.bs; - for(int y = 0; y < th; y += s.cy) + for(int y = 0; y < th; y += s.y) { - DWORD base = psm.bn(0, y, bp, bw); + uint32 base = psm.bn(0, y, bp, bw); - for(int x = 0; x < tw; x += s.cx) + for(int x = 0; x < tw; x += s.x) { - DWORD page = (base + psm.blockOffset[x >> 3]) >> 5; + uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; if(page >= MAX_PAGES) { @@ -152,29 +152,29 @@ void GSTextureCacheSW::IncAge() } } -void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& rect) +void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& rect) { const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[BITBLTBUF.DPSM]; - DWORD bp = BITBLTBUF.DBP; - DWORD bw = BITBLTBUF.DBW; + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; - CSize s = (bp & 31) == 0 ? psm.pgs : psm.bs; + GSVector2i s = (bp & 31) == 0 ? psm.pgs : psm.bs; - CRect r; + GSVector4i r; - r.left = rect.left & ~(s.cx - 1); - r.top = rect.top & ~(s.cy - 1); - r.right = (rect.right + (s.cx - 1)) & ~(s.cx - 1); - r.bottom = (rect.bottom + (s.cy - 1)) & ~(s.cy - 1); + r.left = rect.left & ~(s.x - 1); + r.top = rect.top & ~(s.y - 1); + r.right = (rect.right + (s.x - 1)) & ~(s.x - 1); + r.bottom = (rect.bottom + (s.y - 1)) & ~(s.y - 1); - for(int y = r.top; y < r.bottom; y += s.cy) + for(int y = r.top; y < r.bottom; y += s.y) { - DWORD base = psm.bn(0, y, bp, bw); + uint32 base = psm.bn(0, y, bp, bw); - for(int x = r.left; x < r.right; x += s.cx) + for(int x = r.left; x < r.right; x += s.x) { - DWORD page = (base + psm.blockOffset[x >> 3]) >> 5; + uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; if(page >= MAX_PAGES) { @@ -215,7 +215,7 @@ GSTextureCacheSW::GSTexture::~GSTexture() } } -bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* rect) +bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* rect) { if(m_complete) { @@ -229,17 +229,17 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - DWORD bp = TEX0.TBP0; - DWORD bw = TEX0.TBW; + uint32 bp = TEX0.TBP0; + uint32 bw = TEX0.TBW; - CSize s = psm.bs; + GSVector2i s = psm.bs; - int tw = max(1 << TEX0.TW, s.cx); - int th = max(1 << TEX0.TH, s.cy); + int tw = max(1 << TEX0.TW, s.x); + int th = max(1 << TEX0.TH, s.y); if(m_buff == NULL) { - m_buff = _aligned_malloc(tw * th * sizeof(DWORD), 16); + m_buff = _aligned_malloc(tw * th * sizeof(uint32), 16); if(m_buff == NULL) { @@ -249,14 +249,14 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX m_tw = max(psm.pal > 0 ? 5 : 3, TEX0.TW); // makes one row 32 bytes at least, matches the smallest block size that is allocated above for m_buff } - CRect r(0, 0, tw, th); + GSVector4i r(0, 0, tw, th); if(rect) { - r.left = rect->left & ~(s.cx - 1); - r.top = rect->top & ~(s.cy - 1); - r.right = (rect->right + (s.cx - 1)) & ~(s.cx - 1); - r.bottom = (rect->bottom + (s.cy - 1)) & ~(s.cy - 1); + r.left = rect->left & ~(s.x - 1); + r.top = rect->top & ~(s.y - 1); + r.right = (rect->right + (s.x - 1)) & ~(s.x - 1); + r.bottom = (rect->bottom + (s.y - 1)) & ~(s.y - 1); } if(r.left == 0 && r.top == 0 && r.right == tw && r.bottom == th) @@ -268,26 +268,26 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX int bytes = psm.pal > 0 ? 1 : 4; - DWORD pitch = (1 << m_tw) * bytes; + uint32 pitch = (1 << m_tw) * bytes; - BYTE* dst = (BYTE*)m_buff + pitch * r.top; + uint8* dst = (uint8*)m_buff + pitch * r.top; - DWORD blocks = 0; + uint32 blocks = 0; if(tw <= (bw << 6)) { - for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) + for(int y = r.top, o = pitch * s.y; y < r.bottom; y += s.y, dst += o) { - DWORD base = psm.bn(0, y, bp, bw); + uint32 base = psm.bn(0, y, bp, bw); - for(int x = r.left; x < r.right; x += s.cx) + for(int x = r.left; x < r.right; x += s.x) { - DWORD block = base + psm.blockOffset[x >> 3]; + uint32 block = base + psm.blockOffset[x >> 3]; if(block < MAX_BLOCKS) { - DWORD row = block >> 5; - DWORD col = 1 << (block & 31); + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); if((m_valid[row] & col) == 0) { @@ -308,18 +308,18 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX // TODO: still bogus if those repeated parts aren't fetched together - for(int y = r.top, o = pitch * s.cy; y < r.bottom; y += s.cy, dst += o) + for(int y = r.top, o = pitch * s.y; y < r.bottom; y += s.y, dst += o) { - DWORD base = psm.bn(0, y, bp, bw); + uint32 base = psm.bn(0, y, bp, bw); - for(int x = r.left; x < r.right; x += s.cx) + for(int x = r.left; x < r.right; x += s.x) { - DWORD block = base + psm.blockOffset[x >> 3]; + uint32 block = base + psm.blockOffset[x >> 3]; if(block < MAX_BLOCKS) { - DWORD row = block >> 5; - DWORD col = 1 << (block & 31); + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); if((m_valid[row] & col) == 0) { @@ -331,18 +331,18 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX } } - for(int y = r.top; y < r.bottom; y += s.cy) + for(int y = r.top; y < r.bottom; y += s.y) { - DWORD base = psm.bn(0, y, bp, bw); + uint32 base = psm.bn(0, y, bp, bw); - for(int x = r.left; x < r.right; x += s.cx) + for(int x = r.left; x < r.right; x += s.x) { - DWORD block = base + psm.blockOffset[x >> 3]; + uint32 block = base + psm.blockOffset[x >> 3]; if(block < MAX_BLOCKS) { - DWORD row = block >> 5; - DWORD col = 1 << (block & 31); + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); m_valid[row] |= col; } @@ -350,7 +350,7 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX } } - m_state->m_perfmon.Put(GSPerfMon::Unswizzle, s.cx * s.cy * bytes * blocks); + m_state->m_perfmon.Put(GSPerfMon::Unswizzle, s.x * s.y * bytes * blocks); return true; } diff --git a/plugins/GSdx/GSTextureCacheSW.h b/plugins/GSdx/GSTextureCacheSW.h index 540cdd6e1b..ea5128c566 100644 --- a/plugins/GSdx/GSTextureCacheSW.h +++ b/plugins/GSdx/GSTextureCacheSW.h @@ -36,15 +36,15 @@ public: GIFRegTEX0 m_TEX0; GIFRegTEXA m_TEXA; void* m_buff; - DWORD m_tw; - DWORD m_valid[MAX_PAGES]; // each DWORD bits map to the 32 blocks of that page - DWORD m_age; + uint32 m_tw; + uint32 m_valid[MAX_PAGES]; // each uint32 bits map to the 32 blocks of that page + uint32 m_age; bool m_complete; explicit GSTexture(GSState* state); virtual ~GSTexture(); - bool Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* r = NULL); + bool Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r = NULL); }; protected: @@ -56,9 +56,9 @@ public: GSTextureCacheSW(GSState* state); virtual ~GSTextureCacheSW(); - const GSTexture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const CRect* r = NULL); + const GSTexture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r = NULL); void RemoveAll(); void IncAge(); - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const CRect& r); + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); }; diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 80167bcc3a..e81b5f5d77 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -163,7 +163,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { CComPtr vs; - hash_map >::iterator i = m_vs.find(sel); + hash_map >::iterator i = m_vs.find(sel); if(i != m_vs.end()) { @@ -229,7 +229,7 @@ bool GSTextureFX10::SetupGS(GSSelector sel) if(sel.prim > 0 && (sel.iip == 0 || sel.prim == 3)) // geometry shader works in every case, but not needed { - hash_map >::iterator i = m_gs.find(sel); + hash_map >::iterator i = m_gs.find(sel); if(i != m_gs.end()) { @@ -275,7 +275,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl CComPtr ps; - hash_map >::iterator i = m_ps.find(sel); + hash_map >::iterator i = m_ps.find(sel); if(i != m_ps.end()) { @@ -338,7 +338,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl ssel.min = ssel.mag = 0; } - hash_map >::iterator i = m_ps_ss.find(ssel); + hash_map >::iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) { @@ -378,7 +378,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl m_dev->PSSetSamplerState(ss0, ss1); } -void GSTextureFX10::SetupRS(UINT w, UINT h, const RECT& scissor) +void GSTextureFX10::SetupRS(int w, int h, const GSVector4i& scissor) { m_dev->RSSet(w, h, &scissor); } @@ -396,7 +396,7 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr dss; - hash_map >::iterator i = m_om_dss.find(dssel); + hash_map >::iterator i = m_om_dss.find(dssel); if(i != m_om_dss.end()) { @@ -447,7 +447,7 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr bs; - hash_map >::iterator j = m_om_bs.find(bsel); + hash_map >::iterator j = m_om_bs.find(bsel); if(j != m_om_bs.end()) { diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 7c554e66c1..2774b721ba 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -63,16 +63,16 @@ public: { struct { - DWORD bpp:3; - DWORD bppz:2; - DWORD tme:1; - DWORD fst:1; - DWORD prim:3; + uint32 bpp:3; + uint32 bppz:2; + uint32 tme:1; + uint32 fst:1; + uint32 prim:3; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x3ff;} + operator uint32() {return key & 0x3ff;} }; __declspec(align(16)) struct PSConstantBuffer @@ -82,10 +82,10 @@ public: float MAXU; float MINV; float MAXV; - DWORD UMSK; - DWORD UFIX; - DWORD VMSK; - DWORD VFIX; + uint32 UMSK; + uint32 UFIX; + uint32 VMSK; + uint32 VFIX; float TA0; float TA1; float AREF; @@ -125,87 +125,87 @@ public: { struct { - DWORD fst:1; - DWORD wms:2; - DWORD wmt:2; - DWORD bpp:3; - DWORD aem:1; - DWORD tfx:3; - DWORD tcc:1; - DWORD ate:1; - DWORD atst:3; - DWORD fog:1; - DWORD clr1:1; - DWORD fba:1; - DWORD aout:1; + uint32 fst:1; + uint32 wms:2; + uint32 wmt:2; + uint32 bpp:3; + uint32 aem:1; + uint32 tfx:3; + uint32 tcc:1; + uint32 ate:1; + uint32 atst:3; + uint32 fog:1; + uint32 clr1:1; + uint32 fba:1; + uint32 aout:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x1fffff;} + operator uint32() {return key & 0x1fffff;} }; union GSSelector { struct { - DWORD iip:1; - DWORD prim:2; + uint32 iip:1; + uint32 prim:2; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x7;} + operator uint32() {return key & 0x7;} }; union PSSamplerSelector { struct { - DWORD tau:1; - DWORD tav:1; - DWORD min:1; - DWORD mag:1; + uint32 tau:1; + uint32 tav:1; + uint32 min:1; + uint32 mag:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0xf;} + operator uint32() {return key & 0xf;} }; union OMDepthStencilSelector { struct { - DWORD zte:1; - DWORD ztst:2; - DWORD zwe:1; - DWORD date:1; + uint32 zte:1; + uint32 ztst:2; + uint32 zwe:1; + uint32 date:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x1f;} + operator uint32() {return key & 0x1f;} }; union OMBlendSelector { struct { - DWORD abe:1; - DWORD a:2; - DWORD b:2; - DWORD c:2; - DWORD d:2; - DWORD wr:1; - DWORD wg:1; - DWORD wb:1; - DWORD wa:1; + uint32 abe:1; + uint32 a:2; + uint32 b:2; + uint32 c:2; + uint32 d:2; + uint32 wr:1; + uint32 wg:1; + uint32 wb:1; + uint32 wa:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x1fff;} + operator uint32() {return key & 0x1fff;} }; #pragma pack(pop) @@ -213,15 +213,15 @@ public: private: GSDevice10* m_dev; CComPtr m_il; - hash_map > m_vs; + hash_map > m_vs; CComPtr m_vs_cb; - hash_map > m_gs; - hash_map > m_ps; + hash_map > m_gs; + hash_map > m_ps; CComPtr m_ps_cb; - hash_map > m_ps_ss; + hash_map > m_ps_ss; CComPtr m_palette_ss; - hash_map > m_om_dss; - hash_map > m_om_bs; + hash_map > m_om_dss; + hash_map > m_om_bs; CComPtr m_vb, m_vb_old; int m_vb_max; @@ -241,7 +241,7 @@ public: bool SetupGS(GSSelector sel); bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, ID3D10ShaderResourceView* tex, ID3D10ShaderResourceView* pal); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); - void SetupRS(UINT w, UINT h, const RECT& scissor); + void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); void Draw(); diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index c6709d9e1f..e108c6bd6c 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -46,11 +46,11 @@ bool GSTextureFX9::Create(GSDevice9* dev) return true; } -bool GSTextureFX9::CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix) +bool GSTextureFX9::CreateMskFix(GSTexture9& t, uint32 size, uint32 msk, uint32 fix) { - DWORD hash = (size << 20) | (msk << 10) | fix; + uint32 hash = (size << 20) | (msk << 10) | fix; - hash_map::iterator i = m_mskfix.find(hash); + hash_map::iterator i = m_mskfix.find(hash); if(i != m_mskfix.end()) { @@ -63,12 +63,12 @@ bool GSTextureFX9::CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix) return false; } - BYTE* bits; + uint8* bits; int pitch; if(t.Map(&bits, pitch)) { - for(DWORD i = 0; i < size; i++) + for(uint32 i = 0; i < size; i++) { ((float*)bits)[i] = (float)((i & msk) | fix) / size; } @@ -82,7 +82,7 @@ bool GSTextureFX9::CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix) return true; } -bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, UINT count, D3DPRIMITIVETYPE prim) +bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim) { m_dev->IASetVertexBuffer(count, vertices); m_dev->IASetInputLayout(m_il); @@ -95,7 +95,7 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { CComPtr vs; - hash_map >::iterator i = m_vs.find(sel); + hash_map >::iterator i = m_vs.find(sel); if(i != m_vs.end()) { @@ -191,7 +191,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample CComPtr ps; - hash_map >::iterator i = m_ps.find(sel); + hash_map >::iterator i = m_ps.find(sel); if(i != m_ps.end()) { @@ -247,7 +247,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample ssel.min = ssel.mag = 0; } - hash_map::iterator i = m_ps_ss.find(ssel); + hash_map::iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) { @@ -274,23 +274,23 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample m_dev->PSSetSamplerState(ss); } -void GSTextureFX9::SetupRS(int w, int h, const RECT& scissor) +void GSTextureFX9::SetupRS(int w, int h, const GSVector4i& scissor) { m_dev->RSSet(w, h, &scissor); } -void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, BYTE bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds) +void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds) { UpdateOM(dssel, bsel, bf); m_dev->OMSetRenderTargets(rt, ds); } -void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, BYTE bf) +void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf) { Direct3DDepthStencilState9* dss = NULL; - hash_map::iterator i = m_om_dss.find(dssel); + hash_map::iterator i = m_om_dss.find(dssel); if(i != m_om_dss.end()) { @@ -335,7 +335,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, Direct3DBlendState9* bs = NULL; - hash_map::iterator j = m_om_bs.find(bsel); + hash_map::iterator j = m_om_bs.find(bsel); if(j != m_om_bs.end()) { diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index a8c1ab9a3f..6cca7aad7b 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -40,15 +40,15 @@ public: { struct { - DWORD bppz:2; - DWORD tme:1; - DWORD fst:1; - DWORD logz:1; + uint32 bppz:2; + uint32 tme:1; + uint32 fst:1; + uint32 logz:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x1f;} + operator uint32() {return key & 0x1f;} }; struct PSConstantBuffer @@ -58,10 +58,10 @@ public: float MAXU; float MINV; float MAXV; - DWORD UMSK; - DWORD UFIX; - DWORD VMSK; - DWORD VFIX; + uint32 UMSK; + uint32 UFIX; + uint32 VMSK; + uint32 VFIX; float TA0; float TA1; float AREF; @@ -74,74 +74,74 @@ public: { struct { - DWORD fst:1; - DWORD wms:2; - DWORD wmt:2; - DWORD bpp:3; - DWORD aem:1; - DWORD tfx:3; - DWORD tcc:1; - DWORD ate:1; - DWORD atst:3; - DWORD fog:1; - DWORD clr1:1; - DWORD rt:1; + uint32 fst:1; + uint32 wms:2; + uint32 wmt:2; + uint32 bpp:3; + uint32 aem:1; + uint32 tfx:3; + uint32 tcc:1; + uint32 ate:1; + uint32 atst:3; + uint32 fog:1; + uint32 clr1:1; + uint32 rt:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0xfffff;} + operator uint32() {return key & 0xfffff;} }; union PSSamplerSelector { struct { - DWORD tau:1; - DWORD tav:1; - DWORD min:1; - DWORD mag:1; + uint32 tau:1; + uint32 tav:1; + uint32 min:1; + uint32 mag:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0xf;} + operator uint32() {return key & 0xf;} }; union OMDepthStencilSelector { struct { - DWORD zte:1; - DWORD ztst:2; - DWORD zwe:1; - DWORD date:1; - DWORD fba:1; + uint32 zte:1; + uint32 ztst:2; + uint32 zwe:1; + uint32 date:1; + uint32 fba:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x3f;} + operator uint32() {return key & 0x3f;} }; union OMBlendSelector { struct { - DWORD abe:1; - DWORD a:2; - DWORD b:2; - DWORD c:2; - DWORD d:2; - DWORD wr:1; - DWORD wg:1; - DWORD wb:1; - DWORD wa:1; + uint32 abe:1; + uint32 a:2; + uint32 b:2; + uint32 c:2; + uint32 d:2; + uint32 wr:1; + uint32 wg:1; + uint32 wb:1; + uint32 wa:1; }; - DWORD dw; + uint32 key; - operator DWORD() {return dw & 0x1fff;} + operator uint32() {return key & 0x1fff;} }; #pragma pack(pop) @@ -149,25 +149,25 @@ public: private: GSDevice9* m_dev; CComPtr m_il; - hash_map > m_vs; + hash_map > m_vs; D3DXHANDLE m_vs_params; - hash_map > m_ps; - hash_map m_ps_ss; - hash_map m_om_dss; - hash_map m_om_bs; - hash_map m_mskfix; + hash_map > m_ps; + hash_map m_ps_ss; + hash_map m_om_dss; + hash_map m_om_bs; + hash_map m_mskfix; public: GSTextureFX9(); bool Create(GSDevice9* dev); - bool CreateMskFix(GSTexture9& t, DWORD size, DWORD msk, DWORD fix); + bool CreateMskFix(GSTexture9& t, uint32 size, uint32 msk, uint32 fix); - bool SetupIA(const GSVertexHW9* vertices, UINT count, D3DPRIMITIVETYPE prim); + bool SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim); bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, IDirect3DTexture9* tex, IDirect3DTexture9* pal, bool psrr); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, bool psrr); - void SetupRS(int w, int h, const RECT& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, BYTE bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds); - void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, BYTE bf); + void SetupRS(int w, int h, const GSVector4i& scissor); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds); + void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); }; diff --git a/plugins/GSdx/GSTextureNull.h b/plugins/GSdx/GSTextureNull.h index 1d6e07868b..cc094240a8 100644 --- a/plugins/GSdx/GSTextureNull.h +++ b/plugins/GSdx/GSTextureNull.h @@ -37,8 +37,8 @@ public: int GetWidth() const {return m_desc.w;} int GetHeight() const {return m_desc.h;} int GetFormat() const {return m_desc.format;} - bool Update(const CRect& r, const void* data, int pitch) {return true;} - bool Map(BYTE** bits, int& pitch, const RECT* r = NULL) {return true;} + bool Update(const GSVector4i& r, const void* data, int pitch) {return true;} + bool Map(uint8** bits, int& pitch) {return true;} void Unmap() {} bool Save(const string& fn, bool dds = false) {return false;} }; diff --git a/plugins/GSdx/GSThread.cpp b/plugins/GSdx/GSThread.cpp new file mode 100644 index 0000000000..8b64fdce55 --- /dev/null +++ b/plugins/GSdx/GSThread.cpp @@ -0,0 +1,63 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "StdAfx.h" +#include "GSThread.h" + +GSThread::GSThread() + : m_ThreadId(0) + , m_hThread(NULL) +{ +} + +GSThread::~GSThread() +{ + CloseThread(); +} + +DWORD WINAPI GSThread::StaticThreadProc(LPVOID lpParam) +{ + ((GSThread*)lpParam)->ThreadProc(); + + return 0; +} + +void GSThread::CreateThread() +{ + m_hThread = ::CreateThread(NULL, 0, StaticThreadProc, (LPVOID)this, 0, &m_ThreadId); +} + +void GSThread::CloseThread() +{ + if(m_hThread != NULL) + { + if(WaitForSingleObject(m_hThread, 5000) != WAIT_OBJECT_0) + { + TerminateThread(m_hThread, 1); + } + + CloseHandle(m_hThread); + + m_hThread = NULL; + m_ThreadId = 0; + } +} + diff --git a/plugins/GSdx/GSThread.h b/plugins/GSdx/GSThread.h new file mode 100644 index 0000000000..79dca94f9e --- /dev/null +++ b/plugins/GSdx/GSThread.h @@ -0,0 +1,42 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +class GSThread +{ + // TODO: linux + + DWORD m_ThreadId; + HANDLE m_hThread; + + static DWORD WINAPI StaticThreadProc(LPVOID lpParam); + +protected: + virtual void ThreadProc() = 0; + + void CreateThread(); + void CloseThread(); + +public: + GSThread(); + virtual ~GSThread(); +}; diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index b6f391ddf3..57b75cf6f2 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -23,10 +23,11 @@ #include "GS.h" #include "GSUtil.h" #include "svnrev.h" +#include "xbyak/xbyak_util.h" static struct GSUtilMaps { - BYTE PrimClassField[8]; + uint8 PrimClassField[8]; bool CompatibleBitsField[64][64]; bool SharedBitsField[64][64]; @@ -72,80 +73,45 @@ static struct GSUtilMaps } s_maps; -GS_PRIM_CLASS GSUtil::GetPrimClass(DWORD prim) +GS_PRIM_CLASS GSUtil::GetPrimClass(uint32 prim) { return (GS_PRIM_CLASS)s_maps.PrimClassField[prim]; } -bool GSUtil::HasSharedBits(DWORD spsm, DWORD dpsm) +bool GSUtil::HasSharedBits(uint32 spsm, uint32 dpsm) { return s_maps.SharedBitsField[spsm][dpsm]; } -bool GSUtil::HasSharedBits(DWORD sbp, DWORD spsm, DWORD dbp, DWORD dpsm) +bool GSUtil::HasSharedBits(uint32 sbp, uint32 spsm, uint32 dbp, uint32 dpsm) { if(sbp != dbp) return false; return HasSharedBits(spsm, dpsm); } -bool GSUtil::HasCompatibleBits(DWORD spsm, DWORD dpsm) +bool GSUtil::HasCompatibleBits(uint32 spsm, uint32 dpsm) { if(spsm == dpsm) return true; return s_maps.CompatibleBitsField[spsm][dpsm]; } -bool GSUtil::IsRectInRect(const CRect& inner, const CRect& outer) +bool GSUtil::IsRectInRect(const GSVector4i& inner, const GSVector4i& outer) { return outer.left <= inner.left && inner.right <= outer.right && outer.top <= inner.top && inner.bottom <= outer.bottom; } -bool GSUtil::IsRectInRectH(const CRect& inner, const CRect& outer) +bool GSUtil::IsRectInRectH(const GSVector4i& inner, const GSVector4i& outer) { return outer.top <= inner.top && inner.bottom <= outer.bottom; } -bool GSUtil::IsRectInRectV(const CRect& inner, const CRect& outer) +bool GSUtil::IsRectInRectV(const GSVector4i& inner, const GSVector4i& outer) { return outer.left <= inner.left && inner.right <= outer.right; } -void GSUtil::FitRect(CRect& r, int aspectratio) -{ - static const int ar[][2] = {{0, 0}, {4, 3}, {16, 9}}; - - if(aspectratio <= 0 || aspectratio >= countof(ar)) - { - return; - } - - int arx = ar[aspectratio][0]; - int ary = ar[aspectratio][1]; - - CRect r2 = r; - - if(arx > 0 && ary > 0) - { - if(r.Width() * ary > r.Height() * arx) - { - int w = r.Height() * arx / ary; - r.left = r.CenterPoint().x - w / 2; - if(r.left & 1) r.left++; - r.right = r.left + w; - } - else - { - int h = r.Width() * ary / arx; - r.top = r.CenterPoint().y - h / 2; - if(r.top & 1) r.top++; - r.bottom = r.top + h; - } - } - - r &= r2; -} - bool GSUtil::CheckDirectX() { if(HINSTANCE hDll = LoadLibrary(format("d3dx9_%d.dll", D3DX_SDK_VERSION).c_str())) @@ -167,33 +133,22 @@ bool GSUtil::CheckDirectX() return true; } -static bool _CheckSSE() -{ - __try - { - static __m128i m; - - #if _M_SSE >= 0x402 - m.m128i_i32[0] = _mm_popcnt_u32(1234); - #elif _M_SSE >= 0x401 - m = _mm_packus_epi32(m, m); - #elif _M_SSE >= 0x301 - m = _mm_alignr_epi8(m, m, 1); - #elif _M_SSE >= 0x200 - m = _mm_packs_epi32(m, m); - #endif - } - __except(EXCEPTION_EXECUTE_HANDLER) - { - return false; - } - - return true; -} - bool GSUtil::CheckSSE() { - if(!_CheckSSE()) + Xbyak::util::Cpu cpu; + Xbyak::util::Cpu::Type type; + + #if _M_SSE >= 0x402 + type = Xbyak::util::Cpu::tSSE42; + #elif _M_SSE >= 0x401 + type = Xbyak::util::Cpu::tSSE41; + #elif _M_SSE >= 0x301 + type = Xbyak::util::Cpu::tSSSE3; + #elif _M_SSE >= 0x200 + type = Xbyak::util::Cpu::tSSE2; + #endif + + if(!cpu.has(type)) { string s = format("This CPU does not support SSE %d.%02d", _M_SSE >> 8, _M_SSE & 0xff); @@ -223,29 +178,31 @@ char* GSUtil::GetLibName() if(SVN_MODS) str += "m"; -#if _M_AMD64 + #if _M_AMD64 str += " 64-bit"; -#endif + #endif list sl; -#ifdef __INTEL_COMPILER - sl.push_back(format("Intel C++ %d.%02d", __INTEL_COMPILER / 100, __INTEL_COMPILER % 100)); -#elif _MSC_VER - sl.push_back(format("MSVC %d.%02d", _MSC_VER / 100, _MSC_VER % 100)); -#endif + // TODO: gcc -#if _M_SSE >= 0x402 + #ifdef __INTEL_COMPILER + sl.push_back(format("Intel C++ %d.%02d", __INTEL_COMPILER / 100, __INTEL_COMPILER % 100)); + #elif _MSC_VER + sl.push_back(format("MSVC %d.%02d", _MSC_VER / 100, _MSC_VER % 100)); + #endif + + #if _M_SSE >= 0x402 sl.push_back("SSE42"); -#elif _M_SSE >= 0x401 + #elif _M_SSE >= 0x401 sl.push_back("SSE41"); -#elif _M_SSE >= 0x301 + #elif _M_SSE >= 0x301 sl.push_back("SSSE3"); -#elif _M_SSE >= 0x200 + #elif _M_SSE >= 0x200 sl.push_back("SSE2"); -#elif _M_SSE >= 0x100 + #elif _M_SSE >= 0x100 sl.push_back("SSE"); -#endif + #endif for(list::iterator i = sl.begin(); i != sl.end(); ) { diff --git a/plugins/GSdx/GSUtil.h b/plugins/GSdx/GSUtil.h index 2d0eacb569..bd3f580c6b 100644 --- a/plugins/GSdx/GSUtil.h +++ b/plugins/GSdx/GSUtil.h @@ -26,19 +26,17 @@ class GSUtil { public: - static GS_PRIM_CLASS GetPrimClass(DWORD prim); + static GS_PRIM_CLASS GetPrimClass(uint32 prim); - static bool HasSharedBits(DWORD spsm, DWORD dpsm); - static bool HasSharedBits(DWORD sbp, DWORD spsm, DWORD dbp, DWORD dpsm); - static bool HasCompatibleBits(DWORD spsm, DWORD dpsm); + static bool HasSharedBits(uint32 spsm, uint32 dpsm); + static bool HasSharedBits(uint32 sbp, uint32 spsm, uint32 dbp, uint32 dpsm); + static bool HasCompatibleBits(uint32 spsm, uint32 dpsm); - static bool IsRectInRect(const CRect& inner, const CRect& outer); - static bool IsRectInRectH(const CRect& inner, const CRect& outer); - static bool IsRectInRectV(const CRect& inner, const CRect& outer); - - static void FitRect(CRect& r, int aspectratio); + static bool IsRectInRect(const GSVector4i& inner, const GSVector4i& outer); + static bool IsRectInRectH(const GSVector4i& inner, const GSVector4i& outer); + static bool IsRectInRectV(const GSVector4i& inner, const GSVector4i& outer); - static int EncodePSM(int psm) + static uint32 EncodePSM(uint32 psm) { switch(psm) { diff --git a/plugins/GSdx/GSVector.cpp b/plugins/GSdx/GSVector.cpp index 18a2e9fcc1..206fa163b9 100644 --- a/plugins/GSdx/GSVector.cpp +++ b/plugins/GSdx/GSVector.cpp @@ -46,3 +46,55 @@ GSVector4 GSVector4::cast(const GSVector4i& v) { return GSVector4(_mm_castsi128_ps(v.m)); } + +GSVector4i GSVector4i::fit(int arx, int ary) const +{ + GSVector4i r = *this; + + if(arx > 0 && ary > 0) + { + int w = width(); + int h = height(); + + if(w * ary > h * arx) + { + int w2 = w * arx / ary; + r.left = (r.left + r.right - w2) >> 1; + if(r.left & 1) r.left++; + r.right = r.left + w2; + } + else + { + int h2 = w * ary / arx; + r.top = (r.top + r.bottom - h2) >> 1; + if(r.top & 1) r.top++; + r.bottom = r.top + h2; + } + + r = r.rintersect(*this); + } + else + { + r = *this; + } + + return r; +} + +GSVector4i GSVector4i::fit(int preset) const +{ + GSVector4i r; + + static const int ar[][2] = {{0, 0}, {4, 3}, {16, 9}}; + + if(preset > 0 && preset < countof(ar)) + { + r = fit(ar[preset][0], ar[preset][1]); + } + else + { + r = *this; + } + + return r; +} diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 2f413e3a0a..90dce5b963 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -29,9 +29,11 @@ public: *this = v; } - void operator = (const GSVector2T& v) + const GSVector2T& operator = (const GSVector2T& v) { _mm_storel_epi64((__m128i*)this, _mm_loadl_epi64((__m128i*)&v)); + + return *this; } }; @@ -47,16 +49,17 @@ public: { struct {int x, y, z, w;}; struct {int r, g, b, a;}; + struct {int left, top, right, bottom;}; int v[4]; float f32[4]; - unsigned __int64 u64[2]; - __int8 i8[16]; - __int16 i16[8]; - __int32 i32[4]; - __int64 i64[2]; - unsigned __int8 u8[16]; - unsigned __int16 u16[8]; - unsigned __int32 u32[4]; + int8 i8[16]; + int16 i16[8]; + int32 i32[4]; + int64 i64[2]; + uint8 u8[16]; + uint16 u16[8]; + uint32 u32[4]; + uint64 u64[2]; __m128i m; }; @@ -113,11 +116,6 @@ public: this->m = m; } - explicit GSVector4i(const CRect& r) - { - *this = GSVector4i(r.left, r.top, r.right, r.bottom); - } - explicit GSVector4i(const GSVector4& v) { *this = v; @@ -140,29 +138,70 @@ public: this->m = m; } - void operator = (const CRect& r) - { - m = GSVector4i(r); - } - operator __m128i() const { return m; } - operator CRect() const + // rect + + int width() const { - return *(CRect*)&m; + return right - left; } - UINT32 rgba32() const + int height() const + { + return bottom - top; + } + + bool rempty() const + { + return (*this < zwzw()).mask() != 0x00ff; + } + + GSVector4i runion(const GSVector4i& a) const + { + #if _M_SSE >= 0x401 + + return min_i32(a).upl64(max_i32(a).srl<8>()); + + #else + + return GSVector4i(min(x, a.x), min(y, a.y), max(z, a.z), max(x, a.w)); + + #endif + } + + GSVector4i rintersect(const GSVector4i& a) const + { + return sat_i32(a); + } + + GSVector4i fit(int arx, int ary) const; + + GSVector4i fit(int preset) const; + + operator LPCRECT() const + { + return (LPCRECT)this; + } + + operator LPRECT() + { + return (LPRECT)this; + } + + // + + uint32 rgba32() const { GSVector4i v = *this; v = v.ps32(v); v = v.pu16(v); - return (UINT32)store(v); + return (uint32)store(v); } static GSVector4i cast(const GSVector4& v); @@ -203,6 +242,32 @@ public: return max_i32(a.xyxy()).min_i32(a.zwzw()); } + #else + + GSVector4i sat_i32(const GSVector4i& a, const GSVector4i& b) const + { + GSVector4i v; + + v.x = min(max(x, a.x), b.x); + v.y = min(max(y, a.y), b.y); + v.z = min(max(z, a.z), b.z); + v.w = min(max(x, a.w), b.w); + + return v; + } + + GSVector4i sat_i32(const GSVector4i& a) const + { + GSVector4i v; + + v.x = min(max(x, a.x), b.z); + v.y = min(max(y, a.y), b.w); + v.z = min(max(z, a.x), b.z); + v.w = min(max(x, a.y), b.w); + + return v; + } + #endif GSVector4i sat_u8(const GSVector4i& a, const GSVector4i& b) const @@ -841,6 +906,18 @@ public: return sll16(shift + 1).mul16hs(f); } + bool eq(const GSVector4i& v) const + { + #if _M_SSE >= 0x401 + // pxor, ptest, je + GSVector4i t = *this ^ v; + return _mm_testz_si128(t, t) != 0; + #else + // pcmpeqd, pmovmskb, cmp, je + return eq32(v).alltrue(); + #endif + } + GSVector4i eq8(const GSVector4i& v) const { return GSVector4i(_mm_cmpeq_epi8(m, v.m)); @@ -1357,7 +1434,7 @@ public: dst[1] = gather8_4<8>(ptr); } - __forceinline void gather8_8(const BYTE* RESTRICT ptr, GSVector4i* RESTRICT dst) const + __forceinline void gather8_8(const uint8* RESTRICT ptr, GSVector4i* RESTRICT dst) const { dst[0] = gather8_8<>(ptr); } @@ -2077,16 +2154,17 @@ public: { struct {float x, y, z, w;}; struct {float r, g, b, a;}; + struct {int left, top, right, bottom;}; float v[4]; float f32[4]; - unsigned __int64 u64[2]; - __int8 i8[16]; - __int16 i16[8]; - __int32 i32[4]; - __int64 i64[2]; - unsigned __int8 u8[16]; - unsigned __int16 u16[8]; - unsigned __int32 u32[4]; + int8 i8[16]; + int16 i16[8]; + int32 i32[4]; + int64 i64[2]; + uint8 u8[16]; + uint16 u16[8]; + uint32 u32[4]; + uint64 u64[2]; __m128 m; }; @@ -2140,14 +2218,9 @@ public: this->m = m; } - explicit GSVector4(CRect r) + explicit GSVector4(uint32 u32) { - m = _mm_set_ps((float)r.bottom, (float)r.right, (float)r.top, (float)r.left); - } - - explicit GSVector4(DWORD dw) - { - *this = GSVector4(GSVector4i::load((int)dw).u8to32()); + *this = GSVector4(GSVector4i::load((int)u32).u8to32()); } explicit GSVector4(const GSVector4i& v) @@ -2172,14 +2245,9 @@ public: this->m = m; } - void operator = (DWORD dw) + void operator = (uint32 u32) { - *this = GSVector4(GSVector4i::load((int)dw).u8to32()); - } - - void operator = (CRect r) - { - *this = GSVector4(GSVector4i(r.left, r.top, r.right, r.bottom)); + *this = GSVector4(GSVector4i::load((int)u32).u8to32()); } operator __m128() const @@ -2187,7 +2255,7 @@ public: return m; } - UINT32 rgba32() const + uint32 rgba32() const { return GSVector4i(*this).rgba32(); } diff --git a/plugins/GSdx/GSVertex.h b/plugins/GSdx/GSVertex.h index bfe77ee796..9caff2d673 100644 --- a/plugins/GSdx/GSVertex.h +++ b/plugins/GSdx/GSVertex.h @@ -48,7 +48,7 @@ __declspec(align(16)) struct GSVertex GSVertex() {memset(this, 0, sizeof(*this));} - GSVector4 GetUV() const {return GSVector4(GSVector4i::load(UV.ai32[0]).upl16());} + GSVector4 GetUV() const {return GSVector4(GSVector4i::load(UV.u32[0]).upl16());} }; struct GSVertexOld diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index add5e9eeeb..7afaf65ac4 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -30,17 +30,19 @@ __declspec(align(16)) union GSVertexHW9 struct { GSVector2 t; - union {struct {BYTE r, g, b, a;}; DWORD c0;}; - union {struct {BYTE ta0, ta1, res, f;}; DWORD c1;}; + union {struct {uint8 r, g, b, a;}; uint32 c0;}; + union {struct {uint8 ta0, ta1, res, f;}; uint32 c1;}; GSVector4 p; }; + + #if _M_SSE >= 0x200 struct {__m128i m128i[2];}; struct {__m128 m128[2];}; -#if _M_SSE >= 0x200 GSVertexHW9& operator = (GSVertexHW9& v) {m128i[0] = v.m128i[0]; m128i[1] = v.m128i[1]; return *this;} -#endif + + #endif float GetQ() {return p.w;} }; @@ -57,29 +59,31 @@ __declspec(align(16)) union GSVertexHW10 union { - struct {union {struct {WORD x, y;}; DWORD xy;}; DWORD z;} p; + struct {union {struct {uint16 x, y;}; uint32 xy;}; uint32 z;} p; GIFRegXYZ XYZ; }; union { - union {struct {BYTE r, g, b, a; float q;}; DWORD c0;}; + union {struct {uint8 r, g, b, a; float q;}; uint32 c0;}; GIFRegRGBAQ RGBAQ; }; union { - struct {DWORD _pad[1]; union {struct {BYTE ta0, ta1, res, f;}; DWORD c1;};}; + struct {uint32 _pad; union {struct {uint8 ta0, ta1, res, f;}; uint32 c1;};}; GIFRegFOG FOG; }; }; + + #if _M_SSE >= 0x200 struct {__m128i m128i[2];}; struct {__m128 m128[2];}; -#if _M_SSE >= 0x200 GSVertexHW10& operator = (GSVertexHW10& v) {m128i[0] = v.m128i[0]; m128i[1] = v.m128i[1]; return *this;} -#endif + + #endif float GetQ() {return q;} }; diff --git a/plugins/GSdx/GSVertexList.h b/plugins/GSdx/GSVertexList.h index c7152a8ff0..c6edc3dc61 100644 --- a/plugins/GSdx/GSVertexList.h +++ b/plugins/GSdx/GSVertexList.h @@ -25,7 +25,7 @@ template class GSVertexList { void* m_base; Vertex* m_v[3]; - DWORD m_count; + int m_count; public: GSVertexList() @@ -80,7 +80,7 @@ public: v = *m_v[i]; } - DWORD GetCount() + int GetCount() { return m_count; } diff --git a/plugins/GSdx/GSVertexSW.cpp b/plugins/GSdx/GSVertexSW.cpp index d037c1611f..43702ed972 100644 --- a/plugins/GSdx/GSVertexSW.cpp +++ b/plugins/GSdx/GSVertexSW.cpp @@ -26,7 +26,7 @@ using namespace Xbyak; -GSVertexTrace::GSVertexTraceCodeGenerator::GSVertexTraceCodeGenerator(DWORD key, void* ptr, size_t maxsize) +GSVertexTrace::GSVertexTraceCodeGenerator::GSVertexTraceCodeGenerator(uint32 key, void* ptr, size_t maxsize) : CodeGenerator(maxsize, ptr) { #if _M_AMD64 @@ -35,10 +35,10 @@ GSVertexTrace::GSVertexTraceCodeGenerator::GSVertexTraceCodeGenerator(DWORD key, const int params = 0; - DWORD primclass = (key >> 0) & 3; - DWORD iip = (key >> 2) & 1; - DWORD tme = (key >> 3) & 1; - DWORD color = (key >> 4) & 1; + uint32 primclass = (key >> 0) & 3; + uint32 iip = (key >> 2) & 1; + uint32 tme = (key >> 3) & 1; + uint32 color = (key >> 4) & 1; int n = 1; diff --git a/plugins/GSdx/GSVertexSW.h b/plugins/GSdx/GSVertexSW.h index 146aa1531b..8766fc9249 100644 --- a/plugins/GSdx/GSVertexSW.h +++ b/plugins/GSdx/GSVertexSW.h @@ -222,16 +222,16 @@ __declspec(align(16)) class GSVertexTrace class GSVertexTraceCodeGenerator : public Xbyak::CodeGenerator { public: - GSVertexTraceCodeGenerator(DWORD key, void* ptr, size_t maxsize); + GSVertexTraceCodeGenerator(uint32 key, void* ptr, size_t maxsize); }; typedef void (*VertexTracePtr)(const GSVertexSW* v, int count, GSVertexSW& min, GSVertexSW& max); - class GSVertexTraceMap : public GSCodeGeneratorFunctionMap + class GSVertexTraceMap : public GSCodeGeneratorFunctionMap { public: GSVertexTraceMap() : GSCodeGeneratorFunctionMap("VertexTrace") {} - GSVertexTraceCodeGenerator* Create(DWORD key, void* ptr, size_t maxsize) {return new GSVertexTraceCodeGenerator(key, ptr, maxsize);} + GSVertexTraceCodeGenerator* Create(uint32 key, void* ptr, size_t maxsize) {return new GSVertexTraceCodeGenerator(key, ptr, maxsize);} } m_map; public: @@ -241,14 +241,14 @@ public: union { - DWORD value; - struct {DWORD x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1, r:1, g:1, b:1, a:1;}; - struct {DWORD xyzf:4, stq:4, rgba:4;}; + uint32 value; + struct {uint32 x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1, r:1, g:1, b:1, a:1;}; + struct {uint32 xyzf:4, stq:4, rgba:4;}; } m_eq; - void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, DWORD iip, DWORD tme, DWORD tfx, DWORD tcc) + void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, uint32 iip, uint32 tme, uint32 tfx, uint32 tcc) { - DWORD key = primclass | (iip << 2) | (tme << 3); + uint32 key = primclass | (iip << 2) | (tme << 3); if(!(tme && tfx == TFX_DECAL && tcc)) { diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 8fa577240e..b012278704 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -82,452 +82,6 @@ Name="VCPostBuildEventTool" /> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + @@ -1191,14 +1199,6 @@ AssemblerOutput="4" /> - - - + + @@ -1420,6 +1424,10 @@ RelativePath=".\GSTextureNull.cpp" > + + @@ -1459,62 +1467,6 @@ UsePrecompiledHeader="1" /> - - - - - - - - - - - - - - - - - - - - - @@ -1523,6 +1475,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1531,6 +1491,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1539,6 +1507,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1547,6 +1523,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1555,6 +1539,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1563,6 +1555,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1571,6 +1571,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1730,6 +1738,10 @@ RelativePath=".\GSRasterizer.h" > + + @@ -1822,6 +1834,10 @@ RelativePath=".\GSTextureNull.h" > + + @@ -1850,10 +1866,6 @@ RelativePath=".\GSWnd.h" > - - @@ -1947,62 +1959,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2011,6 +1967,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2019,6 +1983,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2027,6 +1999,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2035,6 +2015,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2043,6 +2031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2051,6 +2047,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2059,6 +2063,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2083,62 +2095,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2147,6 +2103,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2155,6 +2119,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2163,6 +2135,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2171,6 +2151,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2179,6 +2167,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2187,6 +2183,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2195,6 +2199,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2219,62 +2231,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2283,6 +2239,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2291,6 +2255,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2299,6 +2271,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2307,6 +2287,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2315,6 +2303,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2323,6 +2319,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2331,6 +2335,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2355,62 +2367,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2419,6 +2375,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2427,6 +2391,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2435,6 +2407,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2443,6 +2423,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2451,6 +2439,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2459,6 +2455,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2467,6 +2471,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2491,62 +2503,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2555,6 +2511,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2563,6 +2527,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2571,6 +2543,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2579,6 +2559,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2587,6 +2575,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2595,6 +2591,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2603,6 +2607,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2627,62 +2639,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2691,6 +2647,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2699,6 +2663,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2707,6 +2679,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2715,6 +2695,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2723,6 +2711,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2731,6 +2727,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2739,6 +2743,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2779,62 +2791,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2843,6 +2799,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2851,6 +2815,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2859,6 +2831,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2867,6 +2847,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2875,6 +2863,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2883,6 +2879,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2891,6 +2895,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2915,62 +2927,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -2979,6 +2935,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2987,6 +2951,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2995,6 +2967,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3003,6 +2983,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3011,6 +2999,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3019,6 +3015,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3027,6 +3031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3051,62 +3063,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3115,6 +3071,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3123,6 +3087,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3131,6 +3103,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3139,6 +3119,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3147,6 +3135,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3155,6 +3151,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3163,6 +3167,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3187,62 +3199,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3251,6 +3207,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3259,6 +3223,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3267,6 +3239,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3275,6 +3255,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3283,6 +3271,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3291,6 +3287,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3299,6 +3303,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3323,62 +3335,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3387,6 +3343,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3395,6 +3359,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3403,6 +3375,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3411,6 +3391,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3419,6 +3407,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3427,6 +3423,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3435,6 +3439,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3463,62 +3475,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3527,6 +3483,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3535,6 +3499,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3543,6 +3515,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3551,6 +3531,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3559,6 +3547,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3567,6 +3563,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3575,6 +3579,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3599,62 +3611,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3663,6 +3619,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3671,6 +3635,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3679,6 +3651,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3687,6 +3667,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3695,6 +3683,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3703,6 +3699,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3711,6 +3715,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3735,62 +3747,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3799,6 +3755,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3807,6 +3771,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3815,6 +3787,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3823,6 +3803,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3831,6 +3819,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3839,6 +3835,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3847,6 +3851,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3871,62 +3883,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -3935,6 +3891,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3943,6 +3907,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3951,6 +3923,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3959,6 +3939,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3967,6 +3955,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3975,6 +3971,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3983,6 +3987,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4011,62 +4023,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4075,6 +4031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4083,6 +4047,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4091,6 +4063,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4099,6 +4079,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4107,6 +4095,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4115,6 +4111,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4123,6 +4127,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4147,62 +4159,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4211,6 +4167,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4219,6 +4183,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4227,6 +4199,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4235,6 +4215,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4243,6 +4231,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4251,6 +4247,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4259,6 +4263,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4283,62 +4295,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4347,6 +4303,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4355,6 +4319,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4363,6 +4335,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4371,6 +4351,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4379,6 +4367,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4387,6 +4383,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4395,6 +4399,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4419,62 +4431,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4483,6 +4439,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4491,6 +4455,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4499,6 +4471,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4507,6 +4487,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4515,6 +4503,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4523,6 +4519,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4531,6 +4535,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4555,62 +4567,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4619,6 +4575,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4627,6 +4591,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4635,6 +4607,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4643,6 +4623,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4651,6 +4639,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4659,6 +4655,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4667,6 +4671,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4691,62 +4703,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4755,6 +4711,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4763,6 +4727,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4771,6 +4743,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4779,6 +4759,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4787,6 +4775,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4795,6 +4791,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4803,6 +4807,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4827,62 +4839,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -4891,6 +4847,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4899,6 +4863,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4907,6 +4879,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4915,6 +4895,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4923,6 +4911,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4931,6 +4927,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4939,6 +4943,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4963,62 +4975,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - - - - @@ -5027,6 +4983,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5035,6 +4999,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5043,6 +5015,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5051,6 +5031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5059,6 +5047,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5067,6 +5063,14 @@ UsePrecompiledHeader="0" /> + + + @@ -5075,6 +5079,14 @@ UsePrecompiledHeader="0" /> + + + diff --git a/plugins/GSdx/sse.h b/plugins/GSdx/sse.h deleted file mode 100644 index d1b35dd2d7..0000000000 --- a/plugins/GSdx/sse.h +++ /dev/null @@ -1,80 +0,0 @@ -/* - * Copyright (C) 2007-2009 Gabest - * http://www.gabest.org - * - * This Program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2, or (at your option) - * any later version. - * - * This Program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with GNU Make; see the file COPYING. If not, write to - * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. - * http://www.gnu.org/copyleft/gpl.html - * - */ - -#pragma once - -// sse2 - -#if _M_SSE >= 0x200 - - #include - #include - - #ifndef _MM_DENORMALS_ARE_ZERO - #define _MM_DENORMALS_ARE_ZERO 0x0040 - #endif - - #define MXCSR (_MM_DENORMALS_ARE_ZERO | _MM_MASK_MASK | _MM_ROUND_NEAREST | _MM_FLUSH_ZERO_ON) - - #if _MSC_VER < 1500 - - __forceinline __m128i _mm_castps_si128(__m128 a) {return *(__m128i*)&a;} - __forceinline __m128 _mm_castsi128_ps(__m128i a) {return *(__m128*)&a;} - __forceinline __m128i _mm_castpd_si128(__m128d a) {return *(__m128i*)&a;} - __forceinline __m128d _mm_castsi128_pd(__m128i a) {return *(__m128d*)&a;} - __forceinline __m128d _mm_castps_pd(__m128 a) {return *(__m128d*)&a;} - __forceinline __m128 _mm_castpd_ps(__m128d a) {return *(__m128*)&a;} - - #endif - - #define _MM_TRANSPOSE4_SI128(row0, row1, row2, row3) \ - { \ - __m128 tmp0 = _mm_shuffle_ps(_mm_castsi128_ps(row0), _mm_castsi128_ps(row1), 0x44); \ - __m128 tmp2 = _mm_shuffle_ps(_mm_castsi128_ps(row0), _mm_castsi128_ps(row1), 0xEE); \ - __m128 tmp1 = _mm_shuffle_ps(_mm_castsi128_ps(row2), _mm_castsi128_ps(row3), 0x44); \ - __m128 tmp3 = _mm_shuffle_ps(_mm_castsi128_ps(row2), _mm_castsi128_ps(row3), 0xEE); \ - (row0) = _mm_castps_si128(_mm_shuffle_ps(tmp0, tmp1, 0x88)); \ - (row1) = _mm_castps_si128(_mm_shuffle_ps(tmp0, tmp1, 0xDD)); \ - (row2) = _mm_castps_si128(_mm_shuffle_ps(tmp2, tmp3, 0x88)); \ - (row3) = _mm_castps_si128(_mm_shuffle_ps(tmp2, tmp3, 0xDD)); \ - } - -#else - -#error TODO: GSVector4 and GSRasterizer needs SSE2 - -#endif - -// sse3 - -#if _M_SSE >= 0x301 - - #include - -#endif - -// sse4 - -#if _M_SSE >= 0x401 - - #include - -#endif diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index 8b3acecd95..5d2b8eae29 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -4,7 +4,7 @@ #pragma once -#pragma warning(disable: 4996) +#pragma warning(disable: 4996 4995 4324 4100) #ifndef VC_EXTRALEAN #define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers @@ -36,11 +36,9 @@ #include // MFC support for Windows Common Controls #endif // _AFX_NO_AFXCMN_SUPPORT //#include -#include -#include -#include -#include -#include + +// stdc + #include #include @@ -54,13 +52,20 @@ using namespace stdext; extern string format(const char* fmt, ...); -#if !defined(_M_SSE) && (defined(_M_AMD64) || defined(_M_IX86_FP) && _M_IX86_FP >= 2) -#define _M_SSE 0x200 -#endif +// syntactic sugar -#include "sse.h" +// put these into vc9/common7/ide/usertype.dat to have them highlighted -#define countof(a) (sizeof(a)/sizeof(a[0])) +typedef unsigned char uint8; +typedef signed char int8; +typedef unsigned short uint16; +typedef signed short int16; +typedef unsigned int uint32; +typedef signed int int32; +typedef unsigned long long uint64; +typedef signed long long int64; + +#define countof(a) (sizeof(a) / sizeof(a[0])) #define EXPORT_C extern "C" __declspec(dllexport) void __stdcall #define EXPORT_C_(type) extern "C" __declspec(dllexport) type __stdcall @@ -68,16 +73,80 @@ extern string format(const char* fmt, ...); #ifndef RESTRICT #ifdef __INTEL_COMPILER #define RESTRICT restrict - #elif _MSC_VER >= 1400 + #elif _MSC_VER >= 1400 // TODO: gcc #define RESTRICT __restrict #else #define RESTRICT #endif #endif -#pragma warning(disable : 4995 4324 4100) +// directx -#define D3DCOLORWRITEENABLE_RGB (D3DCOLORWRITEENABLE_RED|D3DCOLORWRITEENABLE_GREEN|D3DCOLORWRITEENABLE_BLUE) -#define D3DCOLORWRITEENABLE_RGBA (D3DCOLORWRITEENABLE_RGB|D3DCOLORWRITEENABLE_ALPHA) +#include +#include +#include +#include +#include + +#define D3DCOLORWRITEENABLE_RGBA (D3DCOLORWRITEENABLE_RED | D3DCOLORWRITEENABLE_GREEN | D3DCOLORWRITEENABLE_BLUE | D3DCOLORWRITEENABLE_ALPHA) + +// sse + +#if !defined(_M_SSE) && (defined(_M_AMD64) || defined(_M_IX86_FP) && _M_IX86_FP >= 2) + + #define _M_SSE 0x200 + +#endif + +#if _M_SSE >= 0x200 + + #include + #include + + #ifndef _MM_DENORMALS_ARE_ZERO + #define _MM_DENORMALS_ARE_ZERO 0x0040 + #endif + + #define MXCSR (_MM_DENORMALS_ARE_ZERO | _MM_MASK_MASK | _MM_ROUND_NEAREST | _MM_FLUSH_ZERO_ON) + + #if _MSC_VER < 1500 + + __forceinline __m128i _mm_castps_si128(__m128 a) {return *(__m128i*)&a;} + __forceinline __m128 _mm_castsi128_ps(__m128i a) {return *(__m128*)&a;} + __forceinline __m128i _mm_castpd_si128(__m128d a) {return *(__m128i*)&a;} + __forceinline __m128d _mm_castsi128_pd(__m128i a) {return *(__m128d*)&a;} + __forceinline __m128d _mm_castps_pd(__m128 a) {return *(__m128d*)&a;} + __forceinline __m128 _mm_castpd_ps(__m128d a) {return *(__m128*)&a;} + + #endif + + #define _MM_TRANSPOSE4_SI128(row0, row1, row2, row3) \ + { \ + __m128 tmp0 = _mm_shuffle_ps(_mm_castsi128_ps(row0), _mm_castsi128_ps(row1), 0x44); \ + __m128 tmp2 = _mm_shuffle_ps(_mm_castsi128_ps(row0), _mm_castsi128_ps(row1), 0xEE); \ + __m128 tmp1 = _mm_shuffle_ps(_mm_castsi128_ps(row2), _mm_castsi128_ps(row3), 0x44); \ + __m128 tmp3 = _mm_shuffle_ps(_mm_castsi128_ps(row2), _mm_castsi128_ps(row3), 0xEE); \ + (row0) = _mm_castps_si128(_mm_shuffle_ps(tmp0, tmp1, 0x88)); \ + (row1) = _mm_castps_si128(_mm_shuffle_ps(tmp0, tmp1, 0xDD)); \ + (row2) = _mm_castps_si128(_mm_shuffle_ps(tmp2, tmp3, 0x88)); \ + (row3) = _mm_castps_si128(_mm_shuffle_ps(tmp2, tmp3, 0xDD)); \ + } + +#else + +#error TODO: GSVector4 and GSRasterizer needs SSE2 + +#endif + +#if _M_SSE >= 0x301 + + #include + +#endif + +#if _M_SSE >= 0x401 + + #include + +#endif -#define QI(i) (riid == __uuidof(i)) ? GetInterface((i*)this, ppv) : diff --git a/plugins/GSdx/xbyak/xbyak.h b/plugins/GSdx/xbyak/xbyak.h index 6282d9b4bd..564cca6262 100644 --- a/plugins/GSdx/xbyak/xbyak.h +++ b/plugins/GSdx/xbyak/xbyak.h @@ -62,7 +62,7 @@ enum { DEFAULT_MAX_CODE_SIZE = 2048, VERSION = 0x2070, /* 0xABCD = A.BC(D) */ }; - +/* #ifndef MIE_DEFINED_UINT32 #define MIE_DEFINED_UINT32 #ifdef _MSC_VER @@ -81,7 +81,7 @@ enum { #endif #endif #endif - +*/ enum Error { ERR_NONE = 0, ERR_BAD_ADDRESSING, From 729d90c1cea98eeb800329e01aef35e258214198 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 14 May 2009 19:34:17 +0000 Subject: [PATCH 057/296] microVU: * Smartened up the microprogram cache. Programs are (quite accurately) selected based on how often and how recently they have been used. Solves constant recompilation issues in games like FFXII and Tekken 5 [but not the SPS, sorry!]. * Changed memory allocation so that microprogram blocks are allocated *on demand* -- this saves a lot of memory and makes it possible to test 64-program caches (which would previously fail due to exceeding 2gb ram allocations!). Also fixed a few small memory bugs in Init/Alloc, and made it so mVU only allocates memory once instead of on every reset. :) * Tweaked uses of progSize in microVU.h so that it's consistent [all instances of progSize are (vumem_size / 4) now] git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1180 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 4 +- pcsx2/Counters.cpp | 5 ++ pcsx2/x86/microVU.cpp | 104 +++++++++++++++++++++++++++------- pcsx2/x86/microVU.h | 20 +++++-- pcsx2/x86/microVU_Alloc.h | 4 +- pcsx2/x86/microVU_Compile.inl | 13 +++++ 6 files changed, 118 insertions(+), 32 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index d26587a84b..df3c3eff2b 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -44,8 +44,8 @@ extern SessionOverrideFlags g_Session; ////////////////////////////////////////////////////////////////////////// // Pcsx2 User Configuration Options! -//#define PCSX2_MICROVU // Use Micro VU recs instead of Zero VU Recs -//#define PCSX2_MICROVU_ // Fully enable Micro VU recs (temporary option for now) +#define PCSX2_MICROVU // Use Micro VU recs instead of Zero VU Recs +#define PCSX2_MICROVU_ // Fully enable Micro VU recs (temporary option for now) #define PCSX2_GSMULTITHREAD 1 // uses multi-threaded gs #define PCSX2_EEREC 0x10 #define PCSX2_VU0REC 0x20 diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index 5dc192f0ec..777d954cbb 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -447,6 +447,11 @@ __forceinline bool rcntUpdate_vSync() // Accumulate hsync rounding errors: hsyncCounter.sCycle += vSyncInfo.hSyncError; +#ifdef PCSX2_MICROVU + extern void mVUvsyncUpdate(); + mVUvsyncUpdate(); +#endif + # ifdef VSYNC_DEBUG vblankinc++; if( vblankinc > 1 ) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index bcf478c048..cd8d56207a 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -47,6 +47,9 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr) { memset(&mVU->prog, 0, sizeof(mVU->prog)); mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); + mVU->cache = SysMmapEx((vuIndex ? 0x0f240000 : 0x0e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); + if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", (u32)mVU->cache)); + mVUreset(); } @@ -54,14 +57,17 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr) { microVUt(void) mVUreset() { microVU* mVU = mVUx; - mVUclose(); // Close - mVUprint((vuIndex) ? "microVU1: reset" : "microVU0: reset"); + // Delete Block Managers + for (int i = 0; i <= mVU->prog.max; i++) { + for (u32 j = 0; j < (mVU->progSize / 2); j++) { + safe_delete( mVU->prog.prog[i].block[j] ); + } + } + // Dynarec Cache - mVU->cache = SysMmapEx((vuIndex ? 0x1e840000 : 0x0e840000), mVU->cacheSize, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); - if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", params (u32)mVU->cache)); - memset(mVU->cache, 0xcc, mVU->cacheSize); + memset(mVU->cache, 0xcc, mVU->cacheSize + 0x1000); // Setup Entrance/Exit Points x86SetPtr(mVU->cache); @@ -72,11 +78,12 @@ microVUt(void) mVUreset() { memset(&mVU->prog, 0, sizeof(mVU->prog)); // Create Block Managers - for (int i = 0; i <= mVU->prog.max; i++) { + // Block managers are now allocated "on-demand" by the recompiler -- air + /*for (int i = 0; i <= mVU->prog.max; i++) { for (u32 j = 0; j < (mVU->progSize / 2); j++) { mVU->prog.prog[i].block[j] = new microBlockManager(); } - } + }*/ // Program Variables mVU->prog.finished = 1; @@ -87,7 +94,7 @@ microVUt(void) mVUreset() { //mVU->prog.lpState = &mVU->prog.prog[15].allocInfo.block.pState; // Blank Pipeline State (ToDo: finish implementation) // Setup Dynarec Cache Limits for Each Program - u8* z = (mVU->cache + 512); // Dispatcher Code is in first 512 bytes + u8* z = (mVU->cache + 0x1000); // Dispatcher Code is in first page of cache for (int i = 0; i <= mVU->prog.max; i++) { mVU->prog.prog[i].x86start = z; mVU->prog.prog[i].x86ptr = z; @@ -107,7 +114,9 @@ microVUt(void) mVUclose() { // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { for (u32 j = 0; j < (mVU->progSize / 2); j++) { - if (mVU->prog.prog[i].block[j]) delete mVU->prog.prog[i].block[j]; + if (mVU->prog.prog[i].block[j]) { + safe_delete( mVU->prog.prog[i].block[j] ); + } } } } @@ -131,7 +140,8 @@ microVUt(void) mVUclearProg(int progIndex) { mVU->prog.prog[progIndex].sFlagHack = 0; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; for (u32 i = 0; i < (mVU->progSize / 2); i++) { - mVU->prog.prog[progIndex].block[i]->reset(); + if( mVU->prog.prog[progIndex].block[i] ) + mVU->prog.prog[progIndex].block[i]->reset(); } } @@ -153,30 +163,80 @@ microVUt(int) mVUfindLeastUsedProg() { return mVU->prog.total; } else { - int j = (mVU->prog.cur + 1) & mVU->prog.max; - /*u32 smallest = mVU->prog.prog[j].used; - for (int i = ((j+1)&mVU->prog.max), z = 0; z < mVU->prog.max; i = (i+1)&mVU->prog.max, z++) { - if (smallest > mVU->prog.prog[i].used) { - smallest = mVU->prog.prog[i].used; - j = i; + + int startidx = (mVU->prog.cur + 1) & mVU->prog.max; + int endidx = mVU->prog.cur; + int smallidx = startidx; + u32 smallval = mVU->prog.prog[startidx].used; + + for (int i = startidx; i != endidx; i = (i+1)&mVU->prog.max) + { + u32 used = mVU->prog.prog[i].used; + if (smallval > used) + { + smallval = used; + smallidx = i; } - }*/ - mVUclearProg(j); // Clear old data if overwriting old program - mVUcacheProg(j); // Cache Micro Program - //Console::Notice("microVU%d: MicroProgram Cache Full!", params vuIndex); - return j; + } + + mVUclearProg(smallidx); // Clear old data if overwriting old program + mVUcacheProg(smallidx); // Cache Micro Program + Console::Notice("microVU%d: Overwriting existing program in slot %d [%d times used]", params vuIndex, smallidx, smallval ); + return smallidx; } } +// mVUvsyncUpdate --> +// This should be run at 30fps intervals from Counters.cpp (or 60fps works too, but 30fps is +// probably all we need for accurate results) +// +// To fix the program cache to more efficiently dispose of "obsolete" programs, we need to use a +// frame-based decrementing system in combination with a program-execution-based incrementing +// system. In english: if last_used >= 2 it means the program has been used for the current +// or prev frame. if it's 0, the program hasn't been used for a while. +// +microVUt(void) __mVUvsyncUpdate() { + + microVU* mVU = mVUx; + + if (mVU->prog.total < mVU->prog.max) return; + + for (int i = 0; i <= mVU->prog.total; i++) { + if( mVU->prog.prog[i].last_used != 0 ) + { + if( mVU->prog.prog[i].last_used >= 3 ) + { + // program has been used recently. Give it's program execution counter a + // 'weighted' bonus signifying it's importance: + if( mVU->prog.prog[i].used < 0x4fffffff ) + mVU->prog.prog[i].used += 0x200; + } + mVU->prog.prog[i].last_used--; + } + else + mVU->prog.prog[i].used /= 2; // penalize unused programs. + } +} + +void mVUvsyncUpdate() +{ + __mVUvsyncUpdate<0>(); + __mVUvsyncUpdate<1>(); +} + // Searches for Cached Micro Program and sets prog.cur to it (returns 1 if program found, else returns 0) microVUt(int) mVUsearchProg() { microVU* mVU = mVUx; + if (mVU->prog.cleared) { // If cleared, we need to search for new program for (int i = 0; i <= mVU->prog.total; i++) { if (!memcmp_mmx(mVU->prog.prog[i].data, mVU->regs->Micro, mVU->microSize)) { mVU->prog.cur = i; mVU->prog.cleared = 0; - mVU->prog.prog[i].used++; + if( mVU->prog.prog[i].used < 0x7fffffff ) // avoid overflows on well-used programs + mVU->prog.prog[i].used++; + + mVU->prog.prog[i].last_used = 3; // add me to the mVU structs return 1; } } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 55c773c780..7dc1f4d804 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -19,6 +19,7 @@ #pragma once //#define mVUdebug // Prints Extra Info to Console //#define mVUlogProg // Dumps MicroPrograms into microVU0.txt/microVU1.txt + #include "Common.h" #include "VU.h" #include "GS.h" @@ -64,20 +65,21 @@ public: } }; -template +template // progSize = VU program memory size / 4 struct microProgram { - u32 data[progSize/4]; + u32 data[progSize]; u32 used; // Number of times its been used + u32 last_used; // counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vsync) u32 sFlagHack; // Optimize out Status Flag Updates if Program doesn't use Status Flags u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache - microBlockManager* block[progSize/8]; + microBlockManager* block[progSize/2]; microAllocInfo allocInfo; }; -#define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) -template +#define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) +template // pSize = VU program memory size / 4 struct microProgManager { microProgram prog[mMaxProg]; // Store MicroPrograms in memory static const int max = mMaxProg - 1; @@ -95,7 +97,7 @@ struct microVU { u32 progSize; // VU Micro Program Size (microSize/4) u32 cacheSize; // VU Cache Size - microProgManager<0x4000> prog; // Micro Program Data + microProgManager<0x4000/4> prog; // Micro Program Data FILE* logFile; // Log File Pointer VURegs* regs; // VU Regs Struct @@ -112,6 +114,12 @@ struct microVU { u32 totalCycles; // Total Cycles that mVU is expected to run for u32 cycles; // Cycles Counter + // WARNING! MSVC does not reliably guarantee alignment on structure or class member variables, + // failing around 10-20% of the time to align (random depending on various circumstances). + // GCC fails to align the members at all, failing about 50-80% of the time (barring occasional + // random luck). If you want these to be guaranteed aligned, move them to the top of the + // struct, and ensure the struct itself is aligned. :) -- air + PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_Alloc.h index dc4e81d867..c9fd35b4a3 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_Alloc.h @@ -76,6 +76,6 @@ struct microAllocInfo { u32 curPC; // Current PC u32 startPC; // Start PC for Cur Block u32 sFlagHack; // Optimize out all Status flag updates if microProgram doesn't use Status flags - u32 info[pSize/8]; // Info for Instructions in current block - u8 stall[pSize/8]; // Info on how much each instruction stalled (stores the max amount of cycles to stall for the current opcodes) + u32 info[pSize/2]; // Info for Instructions in current block + u8 stall[pSize/2]; // Info on how much each instruction stalled (stores the max amount of cycles to stall for the current opcodes) }; diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 4d30137cf5..f87a3e2080 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -27,6 +27,8 @@ mVUsetupBranch(xStatus, xMac, xClip, xCycles); \ CMP16ItoM((uptr)&mVU->branch, 0); \ incPC2(1); \ + if( mVUblocks[iPC/2] == NULL ) \ + mVUblocks[iPC/2] = new microBlockManager(); \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ @@ -160,6 +162,9 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (startPC > ((vuIndex) ? 0x3fff : 0xfff)) { mVUprint("microVU: invalid startPC"); } startPC &= (vuIndex ? 0x3ff8 : 0xff8); + if( mVUblocks[startPC/8] == NULL ) + mVUblocks[startPC/8] = new microBlockManager(); + // Searches for Existing Compiled Block (if found, then returns; else, compile) microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); if (pBlock) { return pBlock->x86ptrStart; } @@ -225,6 +230,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { else { microBlock* bBlock = NULL; u32* ajmp = 0; + switch (mVUbranch) { case 3: branchCase(JE32, JNE32); // IBEQ case 4: branchCase(JGE32, JNGE32); // IBGEZ @@ -238,6 +244,9 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { incPC(-3); // Go back to branch opcode (to get branch imm addr) mVUsetupBranch(xStatus, xMac, xClip, xCycles); + if( mVUblocks[branchAddr/8] == NULL ) + mVUblocks[branchAddr/8] = new microBlockManager(); + // Check if branch-block has already been compiled pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); if (pBlock) { JMP32((uptr)pBlock->x86ptrStart - ((uptr)x86Ptr + 5)); } @@ -265,6 +274,10 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile conditional branch"); if (bBlock) { // Branch non-taken has already been compiled incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if( mVUblocks[branchAddr/8] == NULL ) + mVUblocks[branchAddr/8] = new microBlockManager(); + // Check if branch-block has already been compiled pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); if (pBlock) { JMP32((uptr)pBlock->x86ptrStart - ((uptr)x86Ptr + 5)); } From 1f67f63ed5bf3d1365a008013a77e10cf4413f48 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 14 May 2009 19:40:54 +0000 Subject: [PATCH 058/296] Removed the microVU enabler define (oops >_<), and removed some SPU2 counters message since the code in question is behaving as intended. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1181 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 4 ++-- pcsx2/IopDma.cpp | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index df3c3eff2b..d26587a84b 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -44,8 +44,8 @@ extern SessionOverrideFlags g_Session; ////////////////////////////////////////////////////////////////////////// // Pcsx2 User Configuration Options! -#define PCSX2_MICROVU // Use Micro VU recs instead of Zero VU Recs -#define PCSX2_MICROVU_ // Fully enable Micro VU recs (temporary option for now) +//#define PCSX2_MICROVU // Use Micro VU recs instead of Zero VU Recs +//#define PCSX2_MICROVU_ // Fully enable Micro VU recs (temporary option for now) #define PCSX2_GSMULTITHREAD 1 // uses multi-threaded gs #define PCSX2_EEREC 0x10 #define PCSX2_VU0REC 0x20 diff --git a/pcsx2/IopDma.cpp b/pcsx2/IopDma.cpp index 7138ca8ccf..1e9231a99d 100644 --- a/pcsx2/IopDma.cpp +++ b/pcsx2/IopDma.cpp @@ -57,7 +57,7 @@ static void __fastcall psxDmaGeneric(u32 madr, u32 bcr, u32 chcr, u32 spuCore, _ if((g_psxNextBranchCycle - psxNextsCounter) > (u32)psxNextCounter) { - DevCon::Notice("SPU2async Setting new counter branch, old %x new %x ((%x - %x = %x) > %x delta)", params g_psxNextBranchCycle, psxNextsCounter + psxNextCounter, g_psxNextBranchCycle, psxNextsCounter, (g_psxNextBranchCycle - psxNextsCounter), psxNextCounter); + //DevCon::Notice("SPU2async Setting new counter branch, old %x new %x ((%x - %x = %x) > %x delta)", params g_psxNextBranchCycle, psxNextsCounter + psxNextCounter, g_psxNextBranchCycle, psxNextsCounter, (g_psxNextBranchCycle - psxNextsCounter), psxNextCounter); g_psxNextBranchCycle = psxNextsCounter + psxNextCounter; } } From bdedc325a8882397461cb17bd5db7e8040b69b98 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 14 May 2009 20:44:17 +0000 Subject: [PATCH 059/296] GSdx: just fixing the typos... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1182 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSVector.h | 10 +++++----- plugins/GSdx/GSdx_vs2008.vcproj | 8 -------- 2 files changed, 5 insertions(+), 13 deletions(-) diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 90dce5b963..46a6194ff7 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -260,10 +260,10 @@ public: { GSVector4i v; - v.x = min(max(x, a.x), b.z); - v.y = min(max(y, a.y), b.w); - v.z = min(max(z, a.x), b.z); - v.w = min(max(x, a.y), b.w); + v.x = min(max(x, a.x), a.z); + v.y = min(max(y, a.y), a.w); + v.z = min(max(z, a.x), a.z); + v.w = min(max(x, a.y), a.w); return v; } @@ -2154,7 +2154,7 @@ public: { struct {float x, y, z, w;}; struct {float r, g, b, a;}; - struct {int left, top, right, bottom;}; + struct {float left, top, right, bottom;}; float v[4]; float f32[4]; int8 i8[16]; diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index b012278704..2523f25124 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1268,10 +1268,6 @@ /> - - @@ -1738,10 +1734,6 @@ RelativePath=".\GSRasterizer.h" > - - From bddfd67bae804b69653d8c4445f54b3f62280fcd Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 14 May 2009 21:50:03 +0000 Subject: [PATCH 060/296] Minor tweaks here and there. Removes a couple compiler warnings, adds in a -m32 flag I missed, and updates the Makefile a bit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1183 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/HwWrite.cpp | 2 +- pcsx2/Linux/ConfigDlg.cpp | 8 ++++---- pcsx2/Linux/ConfigDlg.h | 4 ++-- pcsx2/x86/Makefile.am | 15 ++++++++------- pcsx2/x86/iR3000A.cpp | 13 +++++++------ pcsx2/x86/ix86-32/iR5900LoadStore.cpp | 7 ++++--- pcsx2/x86/ix86-32/recVTLB.cpp | 2 +- pcsx2/x86/microVU.h | 5 +---- plugins/PeopsSPU2/Makefile | 4 ++-- 9 files changed, 30 insertions(+), 30 deletions(-) diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index fc978a6f67..b9a346bd95 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -169,7 +169,7 @@ void hwWrite8(u32 mem, u8 value) { case 0x1000f180: { - bool flush = false; + //bool flush = false; // Terminate lines on CR or full buffers, and ignore \n's if the string contents // are empty (otherwise terminate on \n too!) diff --git a/pcsx2/Linux/ConfigDlg.cpp b/pcsx2/Linux/ConfigDlg.cpp index 8ba14b1a08..f37a141cc3 100644 --- a/pcsx2/Linux/ConfigDlg.cpp +++ b/pcsx2/Linux/ConfigDlg.cpp @@ -88,7 +88,7 @@ static void ConfPlugin(plugin_types type, plugin_callback call, bool pullcombo = void OnConf_Menu(GtkMenuItem *menuitem, gpointer user_data) { - char *name = gtk_widget_get_name(GTK_WIDGET(menuitem)); + char *name = (char*)gtk_widget_get_name(GTK_WIDGET(menuitem)); plugin_types type = strToPluginType(name); gtk_widget_set_sensitive(MainWindow, FALSE); @@ -138,7 +138,7 @@ void OnConfConf_Ok(GtkButton *button, gpointer user_data) { PluginConf *confs = ConfS(type); - if (!GetComboText(confs->Combo, confs->plist, PluginName(type))) + if (!GetComboText(confs->Combo, confs->plist, (char*)PluginName(type))) applychanges = FALSE; } @@ -153,7 +153,7 @@ void OnConfConf_Ok(GtkButton *button, gpointer user_data) void OnConfButton(GtkButton *button, gpointer user_data) { - char *name = gtk_widget_get_name(GTK_WIDGET(button)); + char *name = (char*)gtk_widget_get_name(GTK_WIDGET(button)); plugin_types type = strToPluginType(name); plugin_callback call = strToPluginCall(name); @@ -196,7 +196,7 @@ void UpdateConfDlg() sprintf(tmp, "GtkCombo_%s", PluginTypeToStr(type)); confs->Combo = lookup_widget(ConfDlg, tmp); SetComboToGList(GTK_COMBO_BOX(confs->Combo), confs->PluginNameList); - FindComboText(confs->Combo, confs->plist, confs->PluginNameList, PluginName(type)); + FindComboText(confs->Combo, confs->plist, confs->PluginNameList, (char*)PluginName(type)); } } diff --git a/pcsx2/Linux/ConfigDlg.h b/pcsx2/Linux/ConfigDlg.h index 5a8f971e11..081e2c0a3f 100644 --- a/pcsx2/Linux/ConfigDlg.h +++ b/pcsx2/Linux/ConfigDlg.h @@ -171,7 +171,7 @@ __forceinline const char *PluginCallbackName(int type, int call) return NULL; } -__forceinline int strToPluginCall(char *s) +__forceinline plugin_callback strToPluginCall(char *s) { char *sub = NULL; @@ -187,7 +187,7 @@ __forceinline int strToPluginCall(char *s) return PLUGIN_NULL; } -__forceinline int strToPluginType(char *s) +__forceinline plugin_types strToPluginType(char *s) { char *sub = NULL; diff --git a/pcsx2/x86/Makefile.am b/pcsx2/x86/Makefile.am index 77f3d29c0e..2e77cc0168 100644 --- a/pcsx2/x86/Makefile.am +++ b/pcsx2/x86/Makefile.am @@ -10,18 +10,19 @@ ix86-32/iR5900MultDiv.cpp ix86-32/iCore-32.cpp ix86-32/aR5900-32.S ix86-32/iR590 libx86recomp_a_SOURCES = \ BaseblockEx.cpp iCOP0.cpp iCOP2.cpp iCore.cpp iFPU.cpp iFPUd.cpp iMMI.cpp iPsxMem.cpp iR3000A.cpp iR3000Atables.cpp \ iR5900Misc.cpp iVU0micro.cpp iVU1micro.cpp iVUmicro.cpp iVUmicroLower.cpp iVUmicroUpper.cpp iVUzerorec.cpp iVif.cpp \ -ir5900tables.cpp fast_routines.S aR3000A.S aVUzerorec.S aVif.S $(archfiles) +ir5900tables.cpp fast_routines.S aR3000A.S aVUzerorec.S aVif.S $(archfiles) + +#ifdef PCSX2_MICROVU +#libx86recomp_a_SOURCES += \ +#microVU.cpp microVU_Misc.inl microVU_Log.inl microVU_Analyze.inl microVU_Alloc.inl microVU_Upper.inl microVU_Lower.inl \ +#microVU_Tables.inl microVU_Flags.inl microVU_Compile.inl microVU_Execute.inl \ +#microVU.h microVU_Alloc.h microVU_Misc.h +#endif libx86recomp_a_SOURCES += \ BaseblockEx.h iCOP0.h iCore.h iFPU.h iMMI.h iR3000A.h iR5900.h iR5900Arit.h iR5900AritImm.h iR5900Branch.h iR5900Jump.h \ iR5900LoadStore.h iR5900Move.h iR5900MultDiv.h iR5900Shift.h iVUmicro.h iVUops.h iVUzerorec.h -#ifdef PCSX2_MICROVU -#libx86recomp_a_SOURCES += \ -#microVU.cpp microVU_Alloc.cpp microVU_Compile.cpp microVU_Lower.cpp microVU_Misc.cpp microVU_Upper.cpp aMicroVU.S \ -#microVU.h microVU_Alloc.h microVU_Misc.h microVU_Tables.h -#endif - libx86recomp_a_DEPENDENCIES = ix86/libix86.a SUBDIRS = ix86 \ No newline at end of file diff --git a/pcsx2/x86/iR3000A.cpp b/pcsx2/x86/iR3000A.cpp index 074222ea11..d2a20cdaff 100644 --- a/pcsx2/x86/iR3000A.cpp +++ b/pcsx2/x86/iR3000A.cpp @@ -212,15 +212,15 @@ static void iIopDumpBlock( int startpc, u8 * ptr ) fclose( f ); #ifdef __LINUX__ - // dump the asm - f = fopen( "mydump1", "wb" ); + // dump the asm + f = fopen( "mydump1", "wb" ); fwrite( ptr, 1, (uptr)x86Ptr - (uptr)ptr, f ); fclose( f ); sprintf( command, "objdump -D --target=binary --architecture=i386 -M intel mydump1 | cat %s - > tempdump", filename ); system( command ); - sprintf(command, "mv tempdump %s", filename); - system(command); - f = fopen( filename.c_str(), "a+" ); + sprintf(command, "mv tempdump %s", filename); + system(command); + f = fopen( filename.c_str(), "a+" ); #endif } #endif @@ -889,6 +889,7 @@ void psxRecompileNextInstruction(int delayslot) { static u8 s_bFlushReg = 1; + // pblock isn't used elsewhere in this function. BASEBLOCK* pblock = PSX_GETBLOCK(psxpc); #ifdef _DEBUG @@ -982,7 +983,7 @@ void iopRecRecompile(u32 startpc) if(!s_pCurBlockEx || s_pCurBlockEx->startpc != HWADDR(startpc)) s_pCurBlockEx = recBlocks.New(HWADDR(startpc), (uptr)recPtr); - psxbranch = 0; + psxbranch = 0; s_pCurBlock->SetFnptr( (uptr)x86Ptr ); s_psxBlockCycles = 0; diff --git a/pcsx2/x86/ix86-32/iR5900LoadStore.cpp b/pcsx2/x86/ix86-32/iR5900LoadStore.cpp index aaf93d1d47..ba09e7f748 100644 --- a/pcsx2/x86/ix86-32/iR5900LoadStore.cpp +++ b/pcsx2/x86/ix86-32/iR5900LoadStore.cpp @@ -67,9 +67,6 @@ REC_FUNC(SQC2); #else PCSX2_ALIGNED16(u64 retValues[2]); -static u32 s_bCachingMem = 0; -static u32 s_nAddMemOffset = 0; -static u32 s_tempaddr = 0; void _eeOnLoadWrite(int reg) { @@ -94,6 +91,10 @@ void _eeOnLoadWrite(int reg) #ifdef PCSX2_VIRTUAL_MEM +static u32 s_bCachingMem = 0; +static u32 s_nAddMemOffset = 0; +static u32 s_tempaddr = 0; + //////////////////////////////////////////////////// //#define REC_SLOWREAD //#define REC_SLOWWRITE diff --git a/pcsx2/x86/ix86-32/recVTLB.cpp b/pcsx2/x86/ix86-32/recVTLB.cpp index 5ce11a9d58..16f76b193d 100644 --- a/pcsx2/x86/ix86-32/recVTLB.cpp +++ b/pcsx2/x86/ix86-32/recVTLB.cpp @@ -293,7 +293,7 @@ void vtlb_dynarec_init() // clear the buffer to 0xcc (easier debugging). memset_8<0xcc,0x1000>( m_IndirectDispatchers ); - u8* baseptr = m_IndirectDispatchers; + //u8* baseptr = m_IndirectDispatchers; for( int mode=0; mode<2; ++mode ) { diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7dc1f4d804..88be602ebb 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -147,11 +147,8 @@ microVUt(void) mVUcacheProg(microVU* mVU, int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); -#ifndef __LINUX__ typedef void (__fastcall *mVUrecCall)(u32, u32); -#else -typedef void (*mVUrecCall)(u32, u32) __attribute__((__fastcall)); // Not sure if this is correct syntax (should be close xD) -#endif + // Include all the *.inl files (Needed because C++ sucks with templates and *.cpp files) #include "microVU_Misc.inl" diff --git a/plugins/PeopsSPU2/Makefile b/plugins/PeopsSPU2/Makefile index c620587c0c..e53dcd2c0e 100644 --- a/plugins/PeopsSPU2/Makefile +++ b/plugins/PeopsSPU2/Makefile @@ -28,11 +28,11 @@ LIB = -lc -lm ifeq ($(USEALSA), TRUE) OBJ+= alsa.o LIB+= -lasound - LINKFLAGS = -shared -Wl,-soname,libspu2PeopsALSA.so -o libspu2PeopsALSA.so.1.0.3 + LINKFLAGS = -m32 -shared -Wl,-soname,libspu2PeopsALSA.so -o libspu2PeopsALSA.so.1.0.3 CCFLAGS3+= -DUSEALSA else OBJ+= oss.o - LINKFLAGS = -shared -Wl,-soname,libspu2PeopsOSS.so.1.6 -fPIC -fomit-frame-pointer -o libspu2PeopsOSS.so.1.6 -m32 + LINKFLAGS = -m32 -shared -Wl,-soname,libspu2PeopsOSS.so.1.6 -fPIC -fomit-frame-pointer -o libspu2PeopsOSS.so.1.6 endif ifeq ($(NOTHREADLIB), TRUE) From 5b4935dd86911db3065e75bab98d343580689fa0 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 00:42:05 +0000 Subject: [PATCH 061/296] Zeropad: Archiving some hat work. Still disabled and buggy, but I want to be able to revert back here if neccessary. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1184 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/gui.cpp | 5 ++ plugins/zeropad/Linux/linux.cpp | 93 ++++++++++++++++----------------- plugins/zeropad/analog.cpp | 31 +++++++++++ plugins/zeropad/analog.h | 1 + plugins/zeropad/zeropad.h | 1 + 5 files changed, 83 insertions(+), 48 deletions(-) diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index 3c075e778a..0111b4a6d4 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -136,6 +136,7 @@ void on_joydevicescombo_changed(GtkComboBox *combobox, gpointer user_data) void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) conf.options |= mask; else @@ -145,6 +146,7 @@ void on_checkbutton_reverselx_toggled(GtkToggleButton *togglebutton, gpointer us void on_checkbutton_reversely_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTLY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) conf.options |= mask; else @@ -163,6 +165,7 @@ void on_checkbutton_reverserx_toggled(GtkToggleButton *togglebutton, gpointer us void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTRY << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) conf.options |= mask; else @@ -172,11 +175,13 @@ void on_checkbutton_reversery_toggled(GtkToggleButton *togglebutton, gpointer us void on_forcefeedback_toggled(GtkToggleButton *togglebutton, gpointer user_data) { int mask = PADOPTION_REVERTLX << (16 * s_selectedpad); + if (gtk_toggle_button_get_active(togglebutton)) { conf.options |= mask; int joyid = gtk_combo_box_get_active(GTK_COMBO_BOX(s_devicecombo)); + if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) s_vjoysticks[joyid]->TestForce(); } else diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 4bbaaf62c2..48a5756b53 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -126,28 +126,16 @@ void CALLBACK PADupdate(int pad) switch (i) { case PAD_R_LEFT: - Analog::ConfigurePad(PAD_RX, pad, DEF_VALUE); - break; case PAD_R_UP: - Analog::ConfigurePad(PAD_RY, pad, DEF_VALUE); - break; case PAD_L_LEFT: - Analog::ConfigurePad(PAD_LX, pad, DEF_VALUE); - break; case PAD_L_UP: - Analog::ConfigurePad(PAD_LY, pad, DEF_VALUE); - break; - case PAD_R_DOWN: - Analog::ConfigurePad(PAD_RY, pad, -DEF_VALUE); + Analog::ConfigurePad(Analog::AnalogToPad(i), pad, DEF_VALUE); break; case PAD_R_RIGHT: - Analog::ConfigurePad(PAD_RX, pad, -DEF_VALUE); - break; - case PAD_L_DOWN: - Analog::ConfigurePad(PAD_LY, pad, -DEF_VALUE); - break; + case PAD_R_DOWN: case PAD_L_RIGHT: - Analog::ConfigurePad(PAD_LX, pad, -DEF_VALUE); + case PAD_L_DOWN: + Analog::ConfigurePad(Analog::AnalogToPad(i), pad, -DEF_VALUE); break; } i += 0xff00; @@ -172,34 +160,8 @@ void CALLBACK PADupdate(int pad) if ((i > PAD_RY) && (i <= PAD_R_LEFT)) { - switch (i) - { - case PAD_R_LEFT: - Analog::ResetPad(PAD_RX, pad); - break; - case PAD_R_UP: - Analog::ResetPad(PAD_RY, pad); - break; - case PAD_L_LEFT: - Analog::ResetPad(PAD_LX, pad); - break; - case PAD_L_UP: - Analog::ResetPad(PAD_LY, pad); - break; - case PAD_R_DOWN: - Analog::ResetPad(PAD_RY, pad); - break; - case PAD_R_RIGHT: - Analog::ResetPad(PAD_RX, pad); - break; - case PAD_L_DOWN: - Analog::ResetPad(PAD_LY, pad); - break; - case PAD_L_RIGHT: - Analog::ResetPad(PAD_LX, pad); - break; - } - i += 0xff00; + Analog::ResetPad(Analog::AnalogToPad(i), pad); + i += 0xff00; } #endif if (i != -1) @@ -255,6 +217,7 @@ void CALLBACK PADupdate(int pad) pjoy = s_vjoysticks[joyid]; int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); + switch (i) { case PAD_LX: @@ -279,12 +242,46 @@ void CALLBACK PADupdate(int pad) int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); + int temp; - //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVSIGN(key), key); - if PAD_GETPOVSIGN(key) + //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); + if (value != 0) + { + if (PAD_GETPOVDIR(key) == value) + set_bit(status[pad], i); + /*else + clear_bit(status[pad], i);*/ + + switch (value) + { + case SDL_HAT_UP: + PAD_LOG("Up!\n"); + break; + + case SDL_HAT_RIGHT: + PAD_LOG("Right!\n"); + break; + + case SDL_HAT_DOWN: + PAD_LOG("Down!\n"); + break; + + case SDL_HAT_LEFT: + PAD_LOG("Left!\n"); + break; + case SDL_HAT_CENTER: + clear_bit(status[pad], i); + break; + } + } + else + { + //clear_bit(status[pad], i); + } + /*if PAD_GETPOVSIGN(key) set_bit(status[pad], i); else - clear_bit(status[pad], i); + clear_bit(status[pad], i);*/ } } #else @@ -357,7 +354,7 @@ void UpdateConf(int pad) #ifdef EXPERAMENTAL_POV_CODE { tmp.resize(28); - switch(PAD_GETPOVSIGN(conf.keys[pad][i])) + switch(PAD_GETPOVDIR(conf.keys[pad][i])) { case SDL_HAT_UP: sprintf(&tmp[0], "JPOVU-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); diff --git a/plugins/zeropad/analog.cpp b/plugins/zeropad/analog.cpp index 67e63ba1e9..e0fd490eb9 100644 --- a/plugins/zeropad/analog.cpp +++ b/plugins/zeropad/analog.cpp @@ -128,4 +128,35 @@ namespace Analog PAD_LOG("Setting pad[%d]@%d to %d from %d\n", padvalue, i, value, temp); } + int AnalogToPad(int padvalue) + { + switch (padvalue) + { + case PAD_R_LEFT: + return PAD_RX; + break; + case PAD_R_UP: + return PAD_RY; + break; + case PAD_L_LEFT: + return PAD_LX; + break; + case PAD_L_UP: + return PAD_LY; + break; + case PAD_R_DOWN: + return PAD_RY; + break; + case PAD_R_RIGHT: + return PAD_RX; + break; + case PAD_L_DOWN: + return PAD_LY; + break; + case PAD_L_RIGHT: + return PAD_LX; + break; + } + return 0; + } } diff --git a/plugins/zeropad/analog.h b/plugins/zeropad/analog.h index 2de476df81..ee4055bbf2 100644 --- a/plugins/zeropad/analog.h +++ b/plugins/zeropad/analog.h @@ -29,4 +29,5 @@ extern void ResetPad(int padvalue, u8 i); extern void ConfigurePad(int padvalue, u8 i, int value); extern int KeypadToPad(u8 keypress); + extern int AnalogToPad(int padvalue); } \ No newline at end of file diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index c9f4cae778..92609b15ac 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -65,6 +65,7 @@ extern char libraryName[256]; #define PAD_JOYSTICK(joyid, axisid) (0x20000 | ((joyid) << 12) | (axisid)) #define PAD_POV(joyid, sign, axisid) (0x30000 | ((joyid) << 12) | ((sign) << 8) | (axisid)) #define PAD_GETPOVSIGN(key) (((key) & 0x100) >> 8) +#define PAD_GETPOVDIR(key) (((key) & ~ 0x30000) >> 8) #ifdef __LINUX__ #define PADKEYS 28 From c671bba45a6244d838526bd4798a1e544deba898 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 01:06:59 +0000 Subject: [PATCH 062/296] GSnull: Add some Gif Transfer code from pcsx2 (not currently hooked in). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1185 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GifTransfer.cpp | 239 +++++++++++++++++++++++++++++++++ plugins/GSnull/GifTransfer.h | 88 ++++++++++++ plugins/GSnull/Makefile.am | 3 +- 3 files changed, 329 insertions(+), 1 deletion(-) create mode 100644 plugins/GSnull/GifTransfer.cpp create mode 100644 plugins/GSnull/GifTransfer.h diff --git a/plugins/GSnull/GifTransfer.cpp b/plugins/GSnull/GifTransfer.cpp new file mode 100644 index 0000000000..a692018424 --- /dev/null +++ b/plugins/GSnull/GifTransfer.cpp @@ -0,0 +1,239 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + // Processes a GIFtag & packet, and throws out some gsIRQs as needed. +// Used to keep interrupts in sync with the EE, while the GS itself +// runs potentially several frames behind. +// size - size of the packet in simd128's + +#include "GS.h" +#include "GifTransfer.h" + +using namespace std; + +static void RegHandlerSIGNAL(const u32* data) +{ + //MTGS_LOG("MTGS SIGNAL data %x_%x CSRw %x\n",data[0], data[1], CSRw); + + //GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); + + //if ((CSRw & 0x1)) GSCSRr |= 1; // signal + + //if (!(GSIMR&0x100) ) gsIrq(); +} + +static void RegHandlerFINISH(const u32* data) +{ + //MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); + + //if ((CSRw & 0x2)) GSCSRr |= 2; // finish + + //if (!(GSIMR&0x200) ) gsIrq(); + +} + +static void RegHandlerLABEL(const u32* data) +{ + //GSSIGLBLID->LBLID = (GSSIGLBLID->LBLID&~data[1])|(data[0]&data[1]); +} + +typedef void (*GIFRegHandler)(const u32* data); +static GIFRegHandler s_GSHandlers[3] = +{ + RegHandlerSIGNAL, RegHandlerFINISH, RegHandlerLABEL +}; + +__forceinline void GIFPath::PrepRegs() +{ + if( tag.nreg == 0 ) + { + u32 tempreg = tag.regs[0]; + for(u32 i=0; i<16; ++i, tempreg >>= 4) + { + if( i == 8 ) tempreg = tag.regs[1]; + assert( (tempreg&0xf) < 0x64 ); + regs[i] = tempreg & 0xf; + } + } + else + { + u32 tempreg = tag.regs[0]; + for(u32 i=0; i>= 4) + { + assert( (tempreg&0xf) < 0x64 ); + regs[i] = tempreg & 0xf; + } + } +} + +void GIFPath::SetTag(const void* mem) +{ + tag = *((GIFTAG*)mem); + curreg = 0; + + PrepRegs(); +} + +u32 GIFPath::GetReg() +{ + return regs[curreg]; +} + +__forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) +{ + GIFPath& path = m_path[pathidx]; + + while(size > 0) + { + bool eop = false; + + if(path.tag.nloop == 0) + { + path.SetTag( pMem ); + + pMem += sizeof(GIFTAG); + --size; + + if(pathidx == 2 && path.tag.eop) + Path3transfer = FALSE; + + if( pathidx == 0 ) + { + // hack: if too much data for VU1, just ignore. + + // The GIF is evil : if nreg is 0, it's really 16. Otherwise it's the value in nreg. + const int numregs = ((path.tag.nreg - 1) & 15) + 1; + + if((path.tag.nloop * numregs) > (size * ((path.tag.flg == 1) ? 2 : 1))) + { + path.tag.nloop = 0; + return ++size; + } + } + + if(path.tag.eop) + { + eop = true; + } + else if(path.tag.nloop == 0) + { + if(pathidx == 0) + continue; + + eop = true; + } + } + + if(path.tag.nloop > 0) + { + switch(path.tag.flg) + { + case GIF_FLG_PACKED: + + while(size > 0) + { + if( path.GetReg() == 0xe ) + { + const int handler = pMem[8]; + if(handler >= 0x60 && handler < 0x63) + s_GSHandlers[handler & 0x3]((const u32*)pMem); + } + size--; + pMem += 16; // 128 bits! //sizeof(GIFPackedReg); + + if((++path.curreg & 0xf) == path.tag.nreg) + { + path.curreg = 0; + path.tag.nloop--; + + if(path.tag.nloop == 0) + break; + } + } + break; + + case GIF_FLG_REGLIST: + + size *= 2; + + while(size > 0) + { + const int handler = path.GetReg(); + if (handler >= 0x60 && handler < 0x63) + s_GSHandlers[handler&0x3]((const u32*)pMem); + + size--; + pMem += 8; //sizeof(GIFReg); -- 64 bits! + + if((++path.curreg & 0xf) == path.tag.nreg) + { + path.curreg = 0; + path.tag.nloop--; + + if(path.tag.nloop == 0) + { + break; + } + } + } + + if (size & 1) pMem += 8; //sizeof(GIFReg); + size /= 2; + + break; + + case GIF_FLG_IMAGE2: // hmmm + assert(0); + path.tag.nloop = 0; + + break; + + case GIF_FLG_IMAGE: + { + int len = (int)min(size, path.tag.nloop); + + pMem += len * 16; + path.tag.nloop -= len; + size -= len; + } + break; + + jNO_DEFAULT; + + } + } + + if (eop && ((int)size <= 0 || pathidx == 0)) + { + break; + } + } + + if(pathidx == 0) + { + if (!path.tag.eop && path.tag.nloop > 0) + { + path.tag.nloop = 0; + SysPrintf( "path1 hack! " ); + + // This means that the giftag data got screwly somewhere + // along the way (often means curreg was in a bad state or something) + } + } + return size; +} diff --git a/plugins/GSnull/GifTransfer.h b/plugins/GSnull/GifTransfer.h new file mode 100644 index 0000000000..019e99f521 --- /dev/null +++ b/plugins/GSnull/GifTransfer.h @@ -0,0 +1,88 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include + + struct GSRegSIGBLID +{ + u32 SIGID; + u32 LBLID; +}; + +enum GIF_FLG +{ + GIF_FLG_PACKED = 0, + GIF_FLG_REGLIST = 1, + GIF_FLG_IMAGE = 2, + GIF_FLG_IMAGE2 = 3 +}; + +enum GIF_PATH +{ + GIF_PATH_1 = 0, + GIF_PATH_2, + GIF_PATH_3, +}; + +enum GIF_REG +{ + GIF_REG_PRIM = 0x00, + GIF_REG_RGBA = 0x01, + GIF_REG_STQ = 0x02, + GIF_REG_UV = 0x03, + GIF_REG_XYZF2 = 0x04, + GIF_REG_XYZ2 = 0x05, + GIF_REG_TEX0_1 = 0x06, + GIF_REG_TEX0_2 = 0x07, + GIF_REG_CLAMP_1 = 0x08, + GIF_REG_CLAMP_2 = 0x09, + GIF_REG_FOG = 0x0a, + GIF_REG_XYZF3 = 0x0c, + GIF_REG_XYZ3 = 0x0d, + GIF_REG_A_D = 0x0e, + GIF_REG_NOP = 0x0f, +}; + +struct GIFTAG +{ + u32 nloop : 15; + u32 eop : 1; + u32 dummy0 : 16; + u32 dummy1 : 14; + u32 pre : 1; + u32 prim : 11; + u32 flg : 2; + u32 nreg : 4; + u32 regs[2]; +}; + +struct GIFPath +{ + GIFTAG tag; + u32 curreg; + u32 _pad[3]; + u8 regs[16]; + + __forceinline void PrepRegs(); + void SetTag(const void* mem); + u32 GetReg(); +}; + + GIFPath m_path[3]; + bool Path3transfer; diff --git a/plugins/GSnull/Makefile.am b/plugins/GSnull/Makefile.am index d976e49677..872aa0e2a0 100644 --- a/plugins/GSnull/Makefile.am +++ b/plugins/GSnull/Makefile.am @@ -28,7 +28,8 @@ FW.h Linux/Config.h Linux/Linux.h libGSnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ -Linux/callbacks.h +Linux/callbacks.h \ +GifTransfer.cpp GifTransfer.h #Linux/callbacks.c #SUBDIRS = Linux \ No newline at end of file From fde4685f77c293278f7dbd68064b6939a612947a Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 01:57:00 +0000 Subject: [PATCH 063/296] GSnull: A bit more work on the gif transfers. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1186 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 4 ++ plugins/GSnull/GS.h | 2 + plugins/GSnull/GifTransfer.cpp | 48 ++++++++++++++++++--- plugins/GSnull/GifTransfer.h | 13 +++++- plugins/GSnull/Windows/GSnull_vc2008.vcproj | 8 ++++ 5 files changed, 67 insertions(+), 8 deletions(-) diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index d3cc3bbe55..7dfe2bdaf4 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -27,6 +27,7 @@ using namespace std; #include "GS.h" + #ifdef __LINUX__ Display *display; int screen; @@ -174,14 +175,17 @@ EXPORT_C_(void) GSvsync(int field) EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) { + _GSgifTransfer1(pMem, addr); } EXPORT_C_(void) GSgifTransfer2(u32 *pMem, u32 size) { + _GSgifTransfer2(pMem, size); } EXPORT_C_(void) GSgifTransfer3(u32 *pMem, u32 size) { + _GSgifTransfer3(pMem, size); } // returns the last tag processed (64 bits) diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h index 9cd5c0287f..a322279618 100644 --- a/plugins/GSnull/GS.h +++ b/plugins/GSnull/GS.h @@ -44,6 +44,8 @@ extern "C" } #endif +#include "GifTransfer.h" + /*#ifdef _MSC_VER #define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK #else diff --git a/plugins/GSnull/GifTransfer.cpp b/plugins/GSnull/GifTransfer.cpp index a692018424..37a22819bb 100644 --- a/plugins/GSnull/GifTransfer.cpp +++ b/plugins/GSnull/GifTransfer.cpp @@ -26,30 +26,39 @@ using namespace std; +u32 CSRw; +GIFPath m_path[3]; +bool Path3transfer; + +PCSX2_ALIGNED16( u8 g_RealGSMem[0x2000] ); +#define GSCSRr *((u64*)(g_RealGSMem+0x1000)) +#define GSIMR *((u32*)(g_RealGSMem+0x1010)) +#define GSSIGLBLID ((GSRegSIGBLID*)(g_RealGSMem+0x1080)) + static void RegHandlerSIGNAL(const u32* data) { //MTGS_LOG("MTGS SIGNAL data %x_%x CSRw %x\n",data[0], data[1], CSRw); - //GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); + GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); - //if ((CSRw & 0x1)) GSCSRr |= 1; // signal + if ((CSRw & 0x1)) GSCSRr |= 1; // signal - //if (!(GSIMR&0x100) ) gsIrq(); + if (!(GSIMR & 0x100) ) GSirq(); } static void RegHandlerFINISH(const u32* data) { //MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); - //if ((CSRw & 0x2)) GSCSRr |= 2; // finish + if ((CSRw & 0x2)) GSCSRr |= 2; // finish - //if (!(GSIMR&0x200) ) gsIrq(); + if (!(GSIMR & 0x200) ) GSirq(); } static void RegHandlerLABEL(const u32* data) { - //GSSIGLBLID->LBLID = (GSSIGLBLID->LBLID&~data[1])|(data[0]&data[1]); + GSSIGLBLID->LBLID = (GSSIGLBLID->LBLID&~data[1])|(data[0]&data[1]); } typedef void (*GIFRegHandler)(const u32* data); @@ -237,3 +246,30 @@ __forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) } return size; } + +// This currently segfaults in the beginning of KH1 if defined. +//#define DO_GIF_TRANSFERS + +void _GSgifTransfer1(u32 *pMem, u32 addr) +{ +#ifdef DO_GIF_TRANSFERS + /* This needs looking at, since I quickly grabbed it from ZeroGS. */ + addr &= 0x3fff; + _gifTransfer( GIF_PATH_1, ((u8*)pMem+(u8)addr), (0x4000-addr)/16); +#endif +} + +void _GSgifTransfer2(u32 *pMem, u32 size) +{ +#ifdef DO_GIF_TRANSFERS + _gifTransfer( GIF_PATH_2, (u8*)pMem, size); +#endif +} + +void _GSgifTransfer3(u32 *pMem, u32 size) +{ +#ifdef DO_GIF_TRANSFERS + _gifTransfer( GIF_PATH_3, (u8*)pMem, size); +#endif +} + diff --git a/plugins/GSnull/GifTransfer.h b/plugins/GSnull/GifTransfer.h index 019e99f521..bf30e19756 100644 --- a/plugins/GSnull/GifTransfer.h +++ b/plugins/GSnull/GifTransfer.h @@ -16,6 +16,9 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + #ifndef __GIF_TRANSFER_H__ + #define __GIF_TRANSFER_H__ + #include #include @@ -84,5 +87,11 @@ struct GIFPath u32 GetReg(); }; - GIFPath m_path[3]; - bool Path3transfer; +extern GIFPath m_path[3]; +extern bool Path3transfer; + +extern void _GSgifTransfer1(u32 *pMem, u32 addr); +extern void _GSgifTransfer2(u32 *pMem, u32 size); +extern void _GSgifTransfer3(u32 *pMem, u32 size); + +#endif diff --git a/plugins/GSnull/Windows/GSnull_vc2008.vcproj b/plugins/GSnull/Windows/GSnull_vc2008.vcproj index e08a6c645c..65a952655a 100644 --- a/plugins/GSnull/Windows/GSnull_vc2008.vcproj +++ b/plugins/GSnull/Windows/GSnull_vc2008.vcproj @@ -164,6 +164,10 @@ RelativePath="..\GS.cpp" > + + @@ -177,6 +181,10 @@ RelativePath="..\GS.h" > + + From c67bfbe6486b8f5610d0a65b05b044d698d89383 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 02:46:35 +0000 Subject: [PATCH 064/296] microVU: fixed up microVU so that it compiles in Linux if the appropriate defines are uncommented, and runs surrealistically. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1187 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/Makefile.am | 14 +++++++------- pcsx2/x86/microVU.h | 10 +++++++++- 2 files changed, 16 insertions(+), 8 deletions(-) diff --git a/pcsx2/x86/Makefile.am b/pcsx2/x86/Makefile.am index 2e77cc0168..bf6ed92239 100644 --- a/pcsx2/x86/Makefile.am +++ b/pcsx2/x86/Makefile.am @@ -12,17 +12,17 @@ BaseblockEx.cpp iCOP0.cpp iCOP2.cpp iCore.cpp iFPU.cpp iFPUd.cpp iMMI.cpp iPsxMe iR5900Misc.cpp iVU0micro.cpp iVU1micro.cpp iVUmicro.cpp iVUmicroLower.cpp iVUmicroUpper.cpp iVUzerorec.cpp iVif.cpp \ ir5900tables.cpp fast_routines.S aR3000A.S aVUzerorec.S aVif.S $(archfiles) -#ifdef PCSX2_MICROVU -#libx86recomp_a_SOURCES += \ -#microVU.cpp microVU_Misc.inl microVU_Log.inl microVU_Analyze.inl microVU_Alloc.inl microVU_Upper.inl microVU_Lower.inl \ -#microVU_Tables.inl microVU_Flags.inl microVU_Compile.inl microVU_Execute.inl \ -#microVU.h microVU_Alloc.h microVU_Misc.h -#endif - libx86recomp_a_SOURCES += \ BaseblockEx.h iCOP0.h iCore.h iFPU.h iMMI.h iR3000A.h iR5900.h iR5900Arit.h iR5900AritImm.h iR5900Branch.h iR5900Jump.h \ iR5900LoadStore.h iR5900Move.h iR5900MultDiv.h iR5900Shift.h iVUmicro.h iVUops.h iVUzerorec.h +# Only active if PCSX2_MICROVU is defined. +libx86recomp_a_SOURCES += \ +microVU.cpp microVU_Misc.inl microVU_Log.inl microVU_Analyze.inl microVU_Alloc.inl microVU_Upper.inl microVU_Lower.inl \ +microVU_Tables.inl microVU_Flags.inl microVU_Compile.inl microVU_Execute.inl \ +microVU.h microVU_Alloc.h microVU_Misc.h + + libx86recomp_a_DEPENDENCIES = ix86/libix86.a SUBDIRS = ix86 \ No newline at end of file diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 88be602ebb..6bf6ff3906 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -139,10 +139,18 @@ microVUt(void) mVUreset(); microVUt(void) mVUclose(); microVUt(void) mVUclear(u32, u32); +// Prototypes for Linux. +void __fastcall mVUcleanUpVU0(); +void __fastcall mVUcleanUpVU1(); +void* __fastcall mVUcompileVU0(u32 startPC, uptr pState); +void* __fastcall mVUcompileVU1(u32 startPC, uptr pState); +microVUf(void) mVUopU(); +microVUf(void) mVUopL(); + // Private Functions microVUt(void) mVUclearProg(microVU* mVU, int progIndex); microVUt(int) mVUfindLeastUsedProg(microVU* mVU); -microVUt(int) mVUsearchProg(microVU* mVU); +microVUt(int) mVUsearchProg(/*microVU* mVU*/); microVUt(void) mVUcacheProg(microVU* mVU, int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); From 9390d5583acb189ccc8f2c1550fa2686f737216c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 15 May 2009 04:51:53 +0000 Subject: [PATCH 065/296] microVU: -implemented Jake's suggestion for proper alignment in the microVU struct. -minor optimizations/changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1188 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 2 +- pcsx2/x86/microVU.h | 21 ++++++---------- pcsx2/x86/microVU_Compile.inl | 10 ++++---- pcsx2/x86/microVU_Misc.inl | 47 +++++++++++++++++------------------ 4 files changed, 37 insertions(+), 43 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index cd8d56207a..e69f1de53c 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -181,7 +181,7 @@ microVUt(int) mVUfindLeastUsedProg() { mVUclearProg(smallidx); // Clear old data if overwriting old program mVUcacheProg(smallidx); // Cache Micro Program - Console::Notice("microVU%d: Overwriting existing program in slot %d [%d times used]", params vuIndex, smallidx, smallval ); + //Console::Notice("microVU%d: Overwriting existing program in slot %d [%d times used]", params vuIndex, smallidx, smallval ); return smallidx; } } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 6bf6ff3906..774469f7c1 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -65,7 +65,7 @@ public: } }; -template // progSize = VU program memory size / 4 +template // progSize = VU program memory size / 4 struct microProgram { u32 data[progSize]; u32 used; // Number of times its been used @@ -92,6 +92,11 @@ struct microProgManager { #define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) struct microVU { + + PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) + PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) + PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ + u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size u32 progSize; // VU Micro Program Size (microSize/4) @@ -113,16 +118,6 @@ struct microVU { u32 q; // Holds current Q instance index u32 totalCycles; // Total Cycles that mVU is expected to run for u32 cycles; // Cycles Counter - - // WARNING! MSVC does not reliably guarantee alignment on structure or class member variables, - // failing around 10-20% of the time to align (random depending on various circumstances). - // GCC fails to align the members at all, failing about 50-80% of the time (barring occasional - // random luck). If you want these to be guaranteed aligned, move them to the top of the - // struct, and ensure the struct itself is aligned. :) -- air - - PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) - PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) - PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ }; // microVU rec structs @@ -139,7 +134,7 @@ microVUt(void) mVUreset(); microVUt(void) mVUclose(); microVUt(void) mVUclear(u32, u32); -// Prototypes for Linux. +// Prototypes for Linux void __fastcall mVUcleanUpVU0(); void __fastcall mVUcleanUpVU1(); void* __fastcall mVUcompileVU0(u32 startPC, uptr pState); @@ -150,7 +145,7 @@ microVUf(void) mVUopL(); // Private Functions microVUt(void) mVUclearProg(microVU* mVU, int progIndex); microVUt(int) mVUfindLeastUsedProg(microVU* mVU); -microVUt(int) mVUsearchProg(/*microVU* mVU*/); +microVUt(int) mVUsearchProg(); microVUt(void) mVUcacheProg(microVU* mVU, int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index f87a3e2080..661caf6e12 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -27,8 +27,7 @@ mVUsetupBranch(xStatus, xMac, xClip, xCycles); \ CMP16ItoM((uptr)&mVU->branch, 0); \ incPC2(1); \ - if( mVUblocks[iPC/2] == NULL ) \ - mVUblocks[iPC/2] = new microBlockManager(); \ + if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = new microBlockManager(); } \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ @@ -159,11 +158,12 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { microVU* mVU = mVUx; u8* thisPtr = x86Ptr; - if (startPC > ((vuIndex) ? 0x3fff : 0xfff)) { mVUprint("microVU: invalid startPC"); } + if (startPC > ((vuIndex) ? 0x3fff : 0xfff)) { Console::Error("microVU%d: invalid startPC", params vuIndex); } startPC &= (vuIndex ? 0x3ff8 : 0xff8); - if( mVUblocks[startPC/8] == NULL ) + if (mVUblocks[startPC/8] == NULL) { mVUblocks[startPC/8] = new microBlockManager(); + } // Searches for Existing Compiled Block (if found, then returns; else, compile) microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); @@ -275,7 +275,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (bBlock) { // Branch non-taken has already been compiled incPC(-3); // Go back to branch opcode (to get branch imm addr) - if( mVUblocks[branchAddr/8] == NULL ) + if (mVUblocks[branchAddr/8] == NULL) mVUblocks[branchAddr/8] = new microBlockManager(); // Check if branch-block has already been compiled diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index a88093e03b..05647a0614 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -105,20 +105,20 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { return;*/ switch ( xyzw ) { - case 5: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xe1); //WZXY + case 5: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xe1); //WZXY SSE_MOVSS_XMM_to_M32(offset+4, reg); - SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // YW - case 6: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc9); + case 6: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xc9); SSE_MOVLPS_XMM_to_M64(offset+4, reg); break; // YZ - case 7: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x93); //ZYXW + case 7: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x93); //ZYXW SSE_MOVHPS_XMM_to_M64(offset+4, reg); SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // YZW case 9: SSE_MOVSS_XMM_to_M32(offset, reg); - SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // XW case 10: SSE_MOVSS_XMM_to_M32(offset, reg); @@ -128,7 +128,7 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { case 11: SSE_MOVSS_XMM_to_M32(offset, reg); SSE_MOVHPS_XMM_to_M64(offset+8, reg); break; //XZW - case 13: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x4b); //YXZW + case 13: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x4b); //YXZW SSE_MOVHPS_XMM_to_M64(offset, reg); SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // XYW @@ -163,20 +163,20 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { return;*/ switch ( xyzw ) { - case 5: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xe1); //WZXY + case 5: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xe1); //WZXY SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+4); - SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // YW - case 6: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc9); + case 6: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xc9); SSE_MOVLPS_XMM_to_Rm(gprReg, reg, offset+4); break; // YZ - case 7: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x93); //ZYXW + case 7: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x93); //ZYXW SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset+4); SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // YZW case 9: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); - SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xff); //WWWW + SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0xff); //WWWW SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // XW case 10: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); @@ -186,7 +186,7 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { case 11: SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset); SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset+8); break; //XZW - case 13: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x4b); //YXZW + case 13: SSE2_PSHUFD_XMM_to_XMM(reg, reg, 0x4b); //YXZW SSE_MOVHPS_XMM_to_Rm(gprReg, reg, offset); SSE_MOVSS_XMM_to_Rm(gprReg, reg, offset+12); break; // XYW @@ -316,7 +316,7 @@ microVUt(void) mVUcheckSflag(int progIndex) { } } -static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; +static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0x7fffffff, 0x80000000, 0x7fffffff, 0x80000000}; static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK2[4]) = {0x00000000, 0x40000000, 0x00000000, 0x40000000}; // Warning: Modifies xmmT1 and xmmT2 @@ -330,28 +330,27 @@ void MIN_MAX_(x86SSERegType to, x86SSERegType from, bool min) { SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); if (min) SSE2_MINPD_XMM_to_XMM(xmmT1, xmmT2); - else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); + else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x88); - mVUmergeRegs(to, xmmT1, 0xc); // ZW - SSE2_PSHUFD_XMM_to_XMM(xmmT1, to, 0xfa); - SSE2_PAND_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK1); - SSE2_POR_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(to, to, 0xfa); + SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0xfa); SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); - if (min) SSE2_MINPD_XMM_to_XMM(xmmT1, xmmT2); - else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); - SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x88); - mVUmergeRegs(to, xmmT1, 0x3); + if (min) SSE2_MINPD_XMM_to_XMM(to, xmmT2); + else SSE2_MAXPD_XMM_to_XMM(to, xmmT2); + SSE2_PSHUFD_XMM_to_XMM(to, to, 0x88); + SSE2_MOVSD_XMM_to_XMM (to, xmmT1); } // Warning: Modifies from and to's upper 3 vectors void MIN_MAX_SS(x86SSERegType to, x86SSERegType from, bool min) { SSE2_PSHUFD_XMM_to_XMM(to, to, 0x50); - SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); - SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); + SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); SSE2_PSHUFD_XMM_to_XMM(from, from, 0x50); SSE2_PAND_M128_to_XMM (from, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (from, (uptr)MIN_MAX_MASK2); From 5615929ce3531cb50e9bc7aa39b03ab1ee0acec9 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 06:08:35 +0000 Subject: [PATCH 066/296] Linux: Work on dialog boxes for a bunch of plugins. mostly null plugins. Not finished, by any means, but this seemed like a good commit point. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1189 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/interface.c | 16 +- pcsx2/Linux/pcsx2.glade | 2 +- plugins/CDVDnull/CDVD.cpp | 1 - plugins/CDVDnull/Linux/Config.cpp | 2 +- .../Linux/Linux.h => CDVDnull/Linux/Config.h} | 4 +- plugins/CDVDnull/Linux/callbacks.c | 34 ++ plugins/CDVDnull/Linux/callbacks.h | 14 + plugins/CDVDnull/Linux/interface.c | 173 +++++++ plugins/CDVDnull/Linux/interface.h | 6 + plugins/CDVDnull/Linux/support.c | 144 ++++++ plugins/CDVDnull/Linux/support.h | 69 +++ plugins/FWnull/FW.cpp | 4 +- plugins/FWnull/Linux/Config.cpp | 41 +- plugins/FWnull/Linux/Config.h | 7 +- plugins/FWnull/Linux/firewire.glade | 252 +--------- plugins/FWnull/Linux/interface.c | 105 +---- plugins/FWnull/Makefile.am | 4 +- plugins/SPU2null/Linux/Config.cpp | 90 +++- .../Linux.cpp => SPU2null/Linux/Config.h} | 47 +- plugins/SPU2null/Linux/callbacks.c | 34 ++ plugins/SPU2null/Linux/callbacks.h | 14 + plugins/SPU2null/Linux/interface.c | 173 +++++++ plugins/SPU2null/Linux/interface.h | 6 + plugins/SPU2null/Linux/spu2.glade | 241 ++++++++++ plugins/SPU2null/Linux/support.c | 144 ++++++ plugins/SPU2null/Linux/support.h | 69 +++ plugins/SPU2null/Makefile.am | 7 +- plugins/SPU2null/SPU2.cpp | 5 +- plugins/SPU2null/SPU2.h | 2 +- plugins/USBnull/Linux/Config.cpp | 29 +- plugins/USBnull/Linux/Config.h | 8 +- plugins/USBnull/Linux/Linux.cpp | 8 +- plugins/USBnull/Linux/interface.c | 150 ++---- plugins/USBnull/Linux/usbnull.glade | 256 ++-------- plugins/USBnull/USB.cpp | 2 +- plugins/dev9null/Linux/callbacks.c | 34 ++ plugins/dev9null/Linux/callbacks.h | 14 + plugins/dev9null/Linux/dev9.glade | 444 ++++++++++++++++++ plugins/dev9null/Linux/interface.c | 240 ++++++++++ plugins/dev9null/Linux/interface.h | 6 + plugins/dev9null/Linux/support.c | 144 ++++++ plugins/dev9null/Linux/support.h | 69 +++ plugins/zeropad/Linux/callbacks.c | 9 + plugins/zeropad/Linux/callbacks.h | 5 + plugins/zeropad/Linux/gui.cpp | 5 +- plugins/zeropad/Linux/interface.c | 5 + plugins/zeropad/Linux/zeropad.glade | 3 +- 47 files changed, 2370 insertions(+), 771 deletions(-) rename plugins/{FWnull/Linux/Linux.h => CDVDnull/Linux/Config.h} (93%) create mode 100644 plugins/CDVDnull/Linux/callbacks.c create mode 100644 plugins/CDVDnull/Linux/callbacks.h create mode 100644 plugins/CDVDnull/Linux/interface.c create mode 100644 plugins/CDVDnull/Linux/interface.h create mode 100644 plugins/CDVDnull/Linux/support.c create mode 100644 plugins/CDVDnull/Linux/support.h rename plugins/{FWnull/Linux/Linux.cpp => SPU2null/Linux/Config.h} (62%) create mode 100644 plugins/SPU2null/Linux/callbacks.c create mode 100644 plugins/SPU2null/Linux/callbacks.h create mode 100644 plugins/SPU2null/Linux/interface.c create mode 100644 plugins/SPU2null/Linux/interface.h create mode 100644 plugins/SPU2null/Linux/spu2.glade create mode 100644 plugins/SPU2null/Linux/support.c create mode 100644 plugins/SPU2null/Linux/support.h create mode 100644 plugins/dev9null/Linux/callbacks.c create mode 100644 plugins/dev9null/Linux/callbacks.h create mode 100644 plugins/dev9null/Linux/dev9.glade create mode 100644 plugins/dev9null/Linux/interface.c create mode 100644 plugins/dev9null/Linux/interface.h create mode 100644 plugins/dev9null/Linux/support.c create mode 100644 plugins/dev9null/Linux/support.h diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index 87e6a49abf..df94254c7e 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -2036,7 +2036,7 @@ create_ConfDlg (void) GtkWidget *hbuttonbox24; GtkWidget *GtkButton_FWconfigure; GtkWidget *GtkButton_FWtest; - GtkWidget *GtkButton_FireWireabout; + GtkWidget *GtkButton_FWabout; GtkWidget *label30; GtkWidget *GtkLabel_Bios; GtkWidget *hbox5; @@ -2385,11 +2385,11 @@ create_ConfDlg (void) gtk_container_add (GTK_CONTAINER (hbuttonbox24), GtkButton_FWtest); GTK_WIDGET_SET_FLAGS (GtkButton_FWtest, GTK_CAN_DEFAULT); - GtkButton_FireWireabout = gtk_button_new_with_mnemonic (_("About")); - gtk_widget_set_name (GtkButton_FireWireabout, "GtkButton_FireWireabout"); - gtk_widget_show (GtkButton_FireWireabout); - gtk_container_add (GTK_CONTAINER (hbuttonbox24), GtkButton_FireWireabout); - GTK_WIDGET_SET_FLAGS (GtkButton_FireWireabout, GTK_CAN_DEFAULT); + GtkButton_FWabout = gtk_button_new_with_mnemonic (_("About")); + gtk_widget_set_name (GtkButton_FWabout, "GtkButton_FWabout"); + gtk_widget_show (GtkButton_FWabout); + gtk_container_add (GTK_CONTAINER (hbuttonbox24), GtkButton_FWabout); + GTK_WIDGET_SET_FLAGS (GtkButton_FWabout, GTK_CAN_DEFAULT); label30 = gtk_label_new (_("FireWire")); gtk_widget_set_name (label30, "label30"); @@ -2518,7 +2518,7 @@ create_ConfDlg (void) g_signal_connect ((gpointer) GtkButton_FWtest, "clicked", G_CALLBACK (OnConfButton), NULL); - g_signal_connect ((gpointer) GtkButton_FireWireabout, "clicked", + g_signal_connect ((gpointer) GtkButton_FWabout, "clicked", G_CALLBACK (OnConfButton), NULL); g_signal_connect ((gpointer) GtkButton_SelectPluginsDir, "clicked", @@ -2585,7 +2585,7 @@ create_ConfDlg (void) GLADE_HOOKUP_OBJECT (ConfDlg, hbuttonbox24, "hbuttonbox24"); GLADE_HOOKUP_OBJECT (ConfDlg, GtkButton_FWconfigure, "GtkButton_FWconfigure"); GLADE_HOOKUP_OBJECT (ConfDlg, GtkButton_FWtest, "GtkButton_FWtest"); - GLADE_HOOKUP_OBJECT (ConfDlg, GtkButton_FireWireabout, "GtkButton_FireWireabout"); + GLADE_HOOKUP_OBJECT (ConfDlg, GtkButton_FWabout, "GtkButton_FWabout"); GLADE_HOOKUP_OBJECT (ConfDlg, label30, "label30"); GLADE_HOOKUP_OBJECT (ConfDlg, GtkLabel_Bios, "GtkLabel_Bios"); GLADE_HOOKUP_OBJECT (ConfDlg, hbox5, "hbox5"); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index bb90d59e3d..a53cc57b60 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -4413,7 +4413,7 @@ Configure them here.</b> - + True True True diff --git a/plugins/CDVDnull/CDVD.cpp b/plugins/CDVDnull/CDVD.cpp index 7d922872a7..29ed1642cb 100644 --- a/plugins/CDVDnull/CDVD.cpp +++ b/plugins/CDVDnull/CDVD.cpp @@ -17,7 +17,6 @@ */ #include - #include "CDVD.h" const char *LibName = "CDVDnull Driver"; diff --git a/plugins/CDVDnull/Linux/Config.cpp b/plugins/CDVDnull/Linux/Config.cpp index fead9422be..1c641a0ffd 100644 --- a/plugins/CDVDnull/Linux/Config.cpp +++ b/plugins/CDVDnull/Linux/Config.cpp @@ -46,7 +46,7 @@ void SysMessage(const char *fmt, ...) MsgDlg = gtk_window_new(GTK_WINDOW_POPUP); gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); - gtk_window_set_title(GTK_WINDOW(MsgDlg), "SPU2null Msg"); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "CDVDnull Msg"); gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); Box = gtk_vbox_new(5, 0); diff --git a/plugins/FWnull/Linux/Linux.h b/plugins/CDVDnull/Linux/Config.h similarity index 93% rename from plugins/FWnull/Linux/Linux.h rename to plugins/CDVDnull/Linux/Config.h index ed6cd2d1cb..ac1b6a60b5 100644 --- a/plugins/FWnull/Linux/Linux.h +++ b/plugins/CDVDnull/Linux/Config.h @@ -1,5 +1,5 @@ -/* FWnull - * Copyright (C) 2004-2009 PCSX2 Team +/* CDVDnull + * Copyright (C) 2002-2009 Pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by diff --git a/plugins/CDVDnull/Linux/callbacks.c b/plugins/CDVDnull/Linux/callbacks.c new file mode 100644 index 0000000000..e87ed8be2d --- /dev/null +++ b/plugins/CDVDnull/Linux/callbacks.c @@ -0,0 +1,34 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/CDVDnull/Linux/callbacks.h b/plugins/CDVDnull/Linux/callbacks.h new file mode 100644 index 0000000000..e606c242f0 --- /dev/null +++ b/plugins/CDVDnull/Linux/callbacks.h @@ -0,0 +1,14 @@ +#include + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/CDVDnull/Linux/interface.c b/plugins/CDVDnull/Linux/interface.c new file mode 100644 index 0000000000..f290071448 --- /dev/null +++ b/plugins/CDVDnull/Linux/interface.c @@ -0,0 +1,173 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + +#define GLADE_HOOKUP_OBJECT(component,widget,name) \ + g_object_set_data_full (G_OBJECT (component), name, \ + gtk_widget_ref (widget), (GDestroyNotify) gtk_widget_unref) + +#define GLADE_HOOKUP_OBJECT_NO_REF(component,widget,name) \ + g_object_set_data (G_OBJECT (component), name, widget) + +GtkWidget* +create_Config (void) +{ + GtkWidget *Config; + GtkWidget *vbox1; + GtkWidget *frame3; + GtkWidget *alignment1; + GtkWidget *check_logging; + GtkWidget *label15; + GtkWidget *hbuttonbox1; + GtkWidget *button1; + GtkWidget *button2; + + Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); + gtk_container_set_border_width (GTK_CONTAINER (Config), 5); + gtk_window_set_title (GTK_WINDOW (Config), _("CDVDconfig")); + + vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); + gtk_widget_show (vbox1); + gtk_container_add (GTK_CONTAINER (Config), vbox1); + gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); + + frame3 = gtk_frame_new (NULL); + gtk_widget_set_name (frame3, "frame3"); + gtk_widget_show (frame3); + gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + gtk_frame_set_shadow_type (GTK_FRAME (frame3), GTK_SHADOW_NONE); + + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame3), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); + + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); + + label15 = gtk_label_new (_("Logging")); + gtk_widget_set_name (label15, "label15"); + gtk_widget_show (label15); + gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + gtk_label_set_use_markup (GTK_LABEL (label15), TRUE); + + hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); + gtk_widget_show (hbuttonbox1); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + + button1 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button1, "button1"); + gtk_widget_show (button1); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); + GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + + button2 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button2, "button2"); + gtk_widget_show (button2); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); + GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button1, "clicked", + G_CALLBACK (OnConf_Ok), + NULL); + g_signal_connect ((gpointer) button2, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); + GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); + GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); + GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); + GLADE_HOOKUP_OBJECT (Config, button1, "button1"); + GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + + return Config; +} + +GtkWidget* +create_About (void) +{ + GtkWidget *About; + GtkWidget *vbox2; + GtkWidget *label2; + GtkWidget *label3; + GtkWidget *hbuttonbox2; + GtkWidget *button3; + + About = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_title (GTK_WINDOW (About), _("CDVDabout")); + + vbox2 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox2, "vbox2"); + gtk_widget_show (vbox2); + gtk_container_add (GTK_CONTAINER (About), vbox2); + gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); + + label2 = gtk_label_new (_("CDVD Null Driver")); + gtk_widget_set_name (label2, "label2"); + gtk_widget_show (label2); + gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); + + label3 = gtk_label_new (_("Original Author: (Not sure. linuzappz?)\nRevised by arcum42@gmail.com")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); + + hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); + gtk_widget_show (hbuttonbox2); + gtk_box_pack_start (GTK_BOX (vbox2), hbuttonbox2, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox2), 30); + + button3 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button3, "button3"); + gtk_widget_show (button3); + gtk_container_add (GTK_CONTAINER (hbuttonbox2), button3); + GTK_WIDGET_SET_FLAGS (button3, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button3, "clicked", + G_CALLBACK (OnAbout_Ok), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); + GLADE_HOOKUP_OBJECT (About, vbox2, "vbox2"); + GLADE_HOOKUP_OBJECT (About, label2, "label2"); + GLADE_HOOKUP_OBJECT (About, label3, "label3"); + GLADE_HOOKUP_OBJECT (About, hbuttonbox2, "hbuttonbox2"); + GLADE_HOOKUP_OBJECT (About, button3, "button3"); + + return About; +} + diff --git a/plugins/CDVDnull/Linux/interface.h b/plugins/CDVDnull/Linux/interface.h new file mode 100644 index 0000000000..69c303ba3b --- /dev/null +++ b/plugins/CDVDnull/Linux/interface.h @@ -0,0 +1,6 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +GtkWidget* create_Config (void); +GtkWidget* create_About (void); diff --git a/plugins/CDVDnull/Linux/support.c b/plugins/CDVDnull/Linux/support.c new file mode 100644 index 0000000000..00aff29822 --- /dev/null +++ b/plugins/CDVDnull/Linux/support.c @@ -0,0 +1,144 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include + +#include "support.h" + +GtkWidget* +lookup_widget (GtkWidget *widget, + const gchar *widget_name) +{ + GtkWidget *parent, *found_widget; + + for (;;) + { + if (GTK_IS_MENU (widget)) + parent = gtk_menu_get_attach_widget (GTK_MENU (widget)); + else + parent = widget->parent; + if (!parent) + parent = (GtkWidget*) g_object_get_data (G_OBJECT (widget), "GladeParentKey"); + if (parent == NULL) + break; + widget = parent; + } + + found_widget = (GtkWidget*) g_object_get_data (G_OBJECT (widget), + widget_name); + if (!found_widget) + g_warning ("Widget not found: %s", widget_name); + return found_widget; +} + +static GList *pixmaps_directories = NULL; + +/* Use this function to set the directory containing installed pixmaps. */ +void +add_pixmap_directory (const gchar *directory) +{ + pixmaps_directories = g_list_prepend (pixmaps_directories, + g_strdup (directory)); +} + +/* This is an internally used function to find pixmap files. */ +static gchar* +find_pixmap_file (const gchar *filename) +{ + GList *elem; + + /* We step through each of the pixmaps directory to find it. */ + elem = pixmaps_directories; + while (elem) + { + gchar *pathname = g_strdup_printf ("%s%s%s", (gchar*)elem->data, + G_DIR_SEPARATOR_S, filename); + if (g_file_test (pathname, G_FILE_TEST_EXISTS)) + return pathname; + g_free (pathname); + elem = elem->next; + } + return NULL; +} + +/* This is an internally used function to create pixmaps. */ +GtkWidget* +create_pixmap (GtkWidget *widget, + const gchar *filename) +{ + gchar *pathname = NULL; + GtkWidget *pixmap; + + if (!filename || !filename[0]) + return gtk_image_new (); + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return gtk_image_new (); + } + + pixmap = gtk_image_new_from_file (pathname); + g_free (pathname); + return pixmap; +} + +/* This is an internally used function to create pixmaps. */ +GdkPixbuf* +create_pixbuf (const gchar *filename) +{ + gchar *pathname = NULL; + GdkPixbuf *pixbuf; + GError *error = NULL; + + if (!filename || !filename[0]) + return NULL; + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return NULL; + } + + pixbuf = gdk_pixbuf_new_from_file (pathname, &error); + if (!pixbuf) + { + fprintf (stderr, "Failed to load pixbuf file: %s: %s\n", + pathname, error->message); + g_error_free (error); + } + g_free (pathname); + return pixbuf; +} + +/* This is used to set ATK action descriptions. */ +void +glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description) +{ + gint n_actions, i; + + n_actions = atk_action_get_n_actions (action); + for (i = 0; i < n_actions; i++) + { + if (!strcmp (atk_action_get_name (action, i), action_name)) + atk_action_set_description (action, i, description); + } +} + diff --git a/plugins/CDVDnull/Linux/support.h b/plugins/CDVDnull/Linux/support.h new file mode 100644 index 0000000000..a32649e53c --- /dev/null +++ b/plugins/CDVDnull/Linux/support.h @@ -0,0 +1,69 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* + * Standard gettext macros. + */ +#ifdef ENABLE_NLS +# include +# undef _ +# define _(String) dgettext (PACKAGE, String) +# define Q_(String) g_strip_context ((String), gettext (String)) +# ifdef gettext_noop +# define N_(String) gettext_noop (String) +# else +# define N_(String) (String) +# endif +#else +# define textdomain(String) (String) +# define gettext(String) (String) +# define dgettext(Domain,Message) (Message) +# define dcgettext(Domain,Message,Type) (Message) +# define bindtextdomain(Domain,Directory) (Domain) +# define _(String) (String) +# define Q_(String) g_strip_context ((String), (String)) +# define N_(String) (String) +#endif + + +/* + * Public Functions. + */ + +/* + * This function returns a widget in a component created by Glade. + * Call it with the toplevel widget in the component (i.e. a window/dialog), + * or alternatively any widget in the component, and the name of the widget + * you want returned. + */ +GtkWidget* lookup_widget (GtkWidget *widget, + const gchar *widget_name); + + +/* Use this function to set the directory containing installed pixmaps. */ +void add_pixmap_directory (const gchar *directory); + + +/* + * Private Functions. + */ + +/* This is used to create the pixmaps used in the interface. */ +GtkWidget* create_pixmap (GtkWidget *widget, + const gchar *filename); + +/* This is used to create the pixbufs used in the interface. */ +GdkPixbuf* create_pixbuf (const gchar *filename); + +/* This is used to set ATK action descriptions. */ +void glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description); + diff --git a/plugins/FWnull/FW.cpp b/plugins/FWnull/FW.cpp index 93a9a7a1d4..c4503cab30 100644 --- a/plugins/FWnull/FW.cpp +++ b/plugins/FWnull/FW.cpp @@ -29,10 +29,12 @@ const unsigned char revision = 0; const unsigned char build = 5; // increase that with each version static char *libraryName = "FWnull Driver"; +string s_strIniPath="inis/FWnull.ini"; + s8 *fwregs; FILE *fwLog; Config conf; -string s_strIniPath="inis/FWnull.ini"; + void (*FWirq)(); EXPORT_C_(u32) PS2EgetLibType() diff --git a/plugins/FWnull/Linux/Config.cpp b/plugins/FWnull/Linux/Config.cpp index 5cd7583574..c40b4055e8 100644 --- a/plugins/FWnull/Linux/Config.cpp +++ b/plugins/FWnull/Linux/Config.cpp @@ -51,7 +51,7 @@ void OnMsg_Ok() gtk_main_quit(); } -void cfgSysMessage(char *fmt, ...) +void SysMessage(char *fmt, ...) { GtkWidget *Ok,*Txt; GtkWidget *Box,*Box1; @@ -92,21 +92,23 @@ void cfgSysMessage(char *fmt, ...) gtk_main(); } - + void OnAbout_Ok(GtkButton *button, gpointer user_data) { gtk_widget_destroy(About); gtk_main_quit(); } -void CFGabout() +EXPORT_C_(void) FWabout() { About = create_About(); gtk_widget_show_all(About); gtk_main(); } -void OnConf_Ok(GtkButton *button, gpointer user_data) { +void OnConf_Ok(GtkButton *button, gpointer user_data) +{ + conf.Log = is_checked(Conf, "check_logging"); SaveConfig(); gtk_widget_destroy(Conf); @@ -119,36 +121,16 @@ void OnConf_Cancel(GtkButton *button, gpointer user_data) gtk_main_quit(); } -void CFGconfigure() +EXPORT_C_(void) FWconfigure() { + LoadConfig(); Conf = create_Config(); - LoadConfig(); - + set_checked(Conf, "check_logging", conf.Log); gtk_widget_show_all(Conf); gtk_main(); } -long CFGmessage(char *msg) { - cfgSysMessage(msg); - - return 0; -} - -/*int main(int argc, char *argv[]) { - gtk_init(NULL, NULL); - - if (!strcmp(argv[1], "configure")) { - CFGconfigure(); - } else if (!strcmp(argv[1], "about")) { - CFGabout(); - } else if (!strcmp(argv[1], "message")) { - CFGmessage(argv[2]); - } - - return 0; -}*/ - void LoadConfig() { FILE *f; @@ -162,6 +144,7 @@ void LoadConfig() SaveConfig();//save and return return; } + fscanf(f, "logging = %hhx\n", &conf.Log); //fscanf(f, "options = %hhx\n", &confOptions); fclose(f); } @@ -175,10 +158,10 @@ void SaveConfig() f = fopen(cfg,"w"); if (f == NULL) { - printf("failed to open %s\n", s_strIniPath.c_str()); + printf("failed to open '%s'\n", s_strIniPath.c_str()); return; } - + fprintf(f, "logging = %hhx\n", conf.Log); //fprintf(f, "options = %hhx\n", confOptions); fclose(f); } diff --git a/plugins/FWnull/Linux/Config.h b/plugins/FWnull/Linux/Config.h index 824a7dafe7..7251ed01eb 100644 --- a/plugins/FWnull/Linux/Config.h +++ b/plugins/FWnull/Linux/Config.h @@ -18,8 +18,7 @@ void SaveConf(); void LoadConf(); +void SysMessage(char *fmt, ...); -extern long CFGmessage(char *msg); -extern void CFGconfigure(); -extern void cfgSysMessage(char *fmt, ...); -extern void CFGabout(); +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) diff --git a/plugins/FWnull/Linux/firewire.glade b/plugins/FWnull/Linux/firewire.glade index cef4eab65e..17c3f83051 100644 --- a/plugins/FWnull/Linux/firewire.glade +++ b/plugins/FWnull/Linux/firewire.glade @@ -6,7 +6,7 @@ 5 True - DEV9config + FWconfig GTK_WINDOW_TOPLEVEL GTK_WIN_POS_NONE False @@ -27,240 +27,37 @@ False 5 - - - True - 0 - 0.5 - GTK_SHADOW_ETCHED_IN - - - - 5 - True - True - 5 - - - - True - Device: - False - False - GTK_JUSTIFY_CENTER - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - 0 - False - False - - - - - - True - False - True - False - True - False - - - - True - True - True - True - 0 - - True - * - False - - - - - - True - GTK_SELECTION_BROWSE - - - - True - - - - True - - False - False - GTK_JUSTIFY_LEFT - False - False - 0 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - - - - - - - 0 - False - False - - - - - - - - True - Ethernet - False - False - GTK_JUSTIFY_LEFT - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - label_item - - - - - 0 - True - True - - - True 0 0.5 - GTK_SHADOW_ETCHED_IN + GTK_SHADOW_NONE - - 5 + True - True - 5 + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 - + True - Device: - False - False - GTK_JUSTIFY_CENTER - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 + True + Enable Logging + True + GTK_RELIEF_NORMAL + True + False + False + True - - 0 - False - False - - - - - - True - False - True - False - True - False - - - - True - True - True - True - 0 - - True - * - False - - - - - - True - GTK_SELECTION_BROWSE - - - - True - - - - True - - False - False - GTK_JUSTIFY_LEFT - False - False - 0 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - - - - - - - 0 - False - False - @@ -268,9 +65,9 @@ True - Hdd + <b>Logging</b> False - False + True GTK_JUSTIFY_LEFT False False @@ -340,7 +137,7 @@ 5 True - DEV9about + FWabout GTK_WINDOW_TOPLEVEL GTK_WIN_POS_NONE False @@ -389,7 +186,8 @@ True - Author: linuzappz <linuzappz@hotmail.com> + Original Author: linuzappz <linuzappz@hotmail.com> +Revised by arcum42@gmail.com False False GTK_JUSTIFY_LEFT diff --git a/plugins/FWnull/Linux/interface.c b/plugins/FWnull/Linux/interface.c index 5c23358a21..13f79cea74 100644 --- a/plugins/FWnull/Linux/interface.c +++ b/plugins/FWnull/Linux/interface.c @@ -31,19 +31,9 @@ create_Config (void) { GtkWidget *Config; GtkWidget *vbox1; - GtkWidget *frame2; - GtkWidget *hbox1; - GtkWidget *label4; - GtkWidget *GtkCombo_Eth; - GList *GtkCombo_Eth_items = NULL; - GtkWidget *combo_entry1; - GtkWidget *label1; GtkWidget *frame3; - GtkWidget *hbox2; - GtkWidget *label5; - GtkWidget *GtkCombo_Hdd; - GList *GtkCombo_Hdd_items = NULL; - GtkWidget *entry1; + GtkWidget *alignment1; + GtkWidget *check_logging; GtkWidget *label15; GtkWidget *hbuttonbox1; GtkWidget *button1; @@ -52,7 +42,7 @@ create_Config (void) Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); gtk_widget_set_name (Config, "Config"); gtk_container_set_border_width (GTK_CONTAINER (Config), 5); - gtk_window_set_title (GTK_WINDOW (Config), _("DEV9config")); + gtk_window_set_title (GTK_WINDOW (Config), _("FWconfig")); vbox1 = gtk_vbox_new (FALSE, 5); gtk_widget_set_name (vbox1, "vbox1"); @@ -60,77 +50,28 @@ create_Config (void) gtk_container_add (GTK_CONTAINER (Config), vbox1); gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); - frame2 = gtk_frame_new (NULL); - gtk_widget_set_name (frame2, "frame2"); - gtk_widget_show (frame2); - gtk_box_pack_start (GTK_BOX (vbox1), frame2, TRUE, TRUE, 0); - - hbox1 = gtk_hbox_new (TRUE, 5); - gtk_widget_set_name (hbox1, "hbox1"); - gtk_widget_show (hbox1); - gtk_container_add (GTK_CONTAINER (frame2), hbox1); - gtk_container_set_border_width (GTK_CONTAINER (hbox1), 5); - - label4 = gtk_label_new (_("Device:")); - gtk_widget_set_name (label4, "label4"); - gtk_widget_show (label4); - gtk_box_pack_start (GTK_BOX (hbox1), label4, FALSE, FALSE, 0); - gtk_label_set_justify (GTK_LABEL (label4), GTK_JUSTIFY_CENTER); - - GtkCombo_Eth = gtk_combo_new (); - g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Eth)->popwin), - "GladeParentKey", GtkCombo_Eth); - gtk_widget_set_name (GtkCombo_Eth, "GtkCombo_Eth"); - gtk_widget_show (GtkCombo_Eth); - gtk_box_pack_start (GTK_BOX (hbox1), GtkCombo_Eth, FALSE, FALSE, 0); - GtkCombo_Eth_items = g_list_append (GtkCombo_Eth_items, (gpointer) ""); - gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Eth), GtkCombo_Eth_items); - g_list_free (GtkCombo_Eth_items); - - combo_entry1 = GTK_COMBO (GtkCombo_Eth)->entry; - gtk_widget_set_name (combo_entry1, "combo_entry1"); - gtk_widget_show (combo_entry1); - - label1 = gtk_label_new (_("Ethernet")); - gtk_widget_set_name (label1, "label1"); - gtk_widget_show (label1); - gtk_frame_set_label_widget (GTK_FRAME (frame2), label1); - frame3 = gtk_frame_new (NULL); gtk_widget_set_name (frame3, "frame3"); gtk_widget_show (frame3); gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + gtk_frame_set_shadow_type (GTK_FRAME (frame3), GTK_SHADOW_NONE); - hbox2 = gtk_hbox_new (TRUE, 5); - gtk_widget_set_name (hbox2, "hbox2"); - gtk_widget_show (hbox2); - gtk_container_add (GTK_CONTAINER (frame3), hbox2); - gtk_container_set_border_width (GTK_CONTAINER (hbox2), 5); + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame3), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); - label5 = gtk_label_new (_("Device:")); - gtk_widget_set_name (label5, "label5"); - gtk_widget_show (label5); - gtk_box_pack_start (GTK_BOX (hbox2), label5, FALSE, FALSE, 0); - gtk_label_set_justify (GTK_LABEL (label5), GTK_JUSTIFY_CENTER); + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); - GtkCombo_Hdd = gtk_combo_new (); - g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Hdd)->popwin), - "GladeParentKey", GtkCombo_Hdd); - gtk_widget_set_name (GtkCombo_Hdd, "GtkCombo_Hdd"); - gtk_widget_show (GtkCombo_Hdd); - gtk_box_pack_start (GTK_BOX (hbox2), GtkCombo_Hdd, FALSE, FALSE, 0); - GtkCombo_Hdd_items = g_list_append (GtkCombo_Hdd_items, (gpointer) ""); - gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Hdd), GtkCombo_Hdd_items); - g_list_free (GtkCombo_Hdd_items); - - entry1 = GTK_COMBO (GtkCombo_Hdd)->entry; - gtk_widget_set_name (entry1, "entry1"); - gtk_widget_show (entry1); - - label15 = gtk_label_new (_("Hdd")); + label15 = gtk_label_new (_("Logging")); gtk_widget_set_name (label15, "label15"); gtk_widget_show (label15); gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + gtk_label_set_use_markup (GTK_LABEL (label15), TRUE); hbuttonbox1 = gtk_hbutton_box_new (); gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); @@ -160,17 +101,9 @@ create_Config (void) /* Store pointers to all widgets, for use by lookup_widget(). */ GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); - GLADE_HOOKUP_OBJECT (Config, frame2, "frame2"); - GLADE_HOOKUP_OBJECT (Config, hbox1, "hbox1"); - GLADE_HOOKUP_OBJECT (Config, label4, "label4"); - GLADE_HOOKUP_OBJECT (Config, GtkCombo_Eth, "GtkCombo_Eth"); - GLADE_HOOKUP_OBJECT (Config, combo_entry1, "combo_entry1"); - GLADE_HOOKUP_OBJECT (Config, label1, "label1"); GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); - GLADE_HOOKUP_OBJECT (Config, hbox2, "hbox2"); - GLADE_HOOKUP_OBJECT (Config, label5, "label5"); - GLADE_HOOKUP_OBJECT (Config, GtkCombo_Hdd, "GtkCombo_Hdd"); - GLADE_HOOKUP_OBJECT (Config, entry1, "entry1"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); GLADE_HOOKUP_OBJECT (Config, label15, "label15"); GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); GLADE_HOOKUP_OBJECT (Config, button1, "button1"); @@ -192,7 +125,7 @@ create_About (void) About = gtk_window_new (GTK_WINDOW_TOPLEVEL); gtk_widget_set_name (About, "About"); gtk_container_set_border_width (GTK_CONTAINER (About), 5); - gtk_window_set_title (GTK_WINDOW (About), _("DEV9about")); + gtk_window_set_title (GTK_WINDOW (About), _("FWabout")); vbox2 = gtk_vbox_new (FALSE, 5); gtk_widget_set_name (vbox2, "vbox2"); @@ -206,7 +139,7 @@ create_About (void) gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); - label3 = gtk_label_new (_("Author: linuzappz ")); + label3 = gtk_label_new (_("Original Author: linuzappz \nRevised by arcum42@gmail.com")); gtk_widget_set_name (label3, "label3"); gtk_widget_show (label3); gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); diff --git a/plugins/FWnull/Makefile.am b/plugins/FWnull/Makefile.am index 60060ae259..bde89b4aef 100644 --- a/plugins/FWnull/Makefile.am +++ b/plugins/FWnull/Makefile.am @@ -22,13 +22,11 @@ libFWnull_LDFLAGS= @SHARED_LDFLAGS@ libFWnull_LDFLAGS+=-Wl,-soname,@libFWnull_SONAME@ libFWnull_LDADD=$(libFWnull_a_OBJECTS) -libFWnull_a_SOURCES = FW.cpp Linux/Config.cpp Linux/Linux.cpp \ -FW.h Linux/Config.h Linux/Linux.h +libFWnull_a_SOURCES = FW.cpp FW.h Linux/Config.cpp Linux/Config.h libFWnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ Linux/callbacks.h -#Linux/callbacks.c #SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/SPU2null/Linux/Config.cpp b/plugins/SPU2null/Linux/Config.cpp index a42fc933a1..2e6e4aed6d 100644 --- a/plugins/SPU2null/Linux/Config.cpp +++ b/plugins/SPU2null/Linux/Config.cpp @@ -16,12 +16,25 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include "Config.h" #include "SPU2.h" -#include -#include + #ifdef __cplusplus +extern "C" +{ +#endif -GtkWidget *MsgDlg; +#include "support.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus +} +#endif + +using namespace std; + +GtkWidget *MsgDlg, *About, *Conf; void OnMsg_Ok() { @@ -71,16 +84,81 @@ void SysMessage(char *fmt, ...) gtk_main(); } +void OnAbout_Ok(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(About); + gtk_main_quit(); +} + +void OnConf_Ok(GtkButton *button, gpointer user_data) +{ + conf.Log = is_checked(Conf, "check_logging"); + SaveConfig(); + + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void OnConf_Cancel(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + EXPORT_C_(void) SPU2configure() { - SysMessage("Nothing to Configure"); + //SysMessage("Nothing to Configure"); + + Conf = create_Config(); + + LoadConfig(); + + set_checked(Conf, "check_logging", conf.Log); + gtk_widget_show_all(Conf); + gtk_main(); } EXPORT_C_(void) SPU2about() { - SysMessage("%s %d.%d", libraryName, version, build); + //SysMessage("%s %d.%d", libraryName, version, build); + + About = create_About(); + gtk_widget_show_all(About); + gtk_main(); } -void LoadConfig() +void LoadConfig() { + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "r"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + SaveConfig();//save and return + return; + } + fscanf(f, "logging = %hhx\n", &conf.Log); + //fscanf(f, "options = %hhx\n", &confOptions); + fclose(f); } + +void SaveConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg,"w"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + return; + } + + fprintf(f, "logging = %hhx\n", conf.Log); + //fprintf(f, "options = %hhx\n", confOptions); + fclose(f); +} \ No newline at end of file diff --git a/plugins/FWnull/Linux/Linux.cpp b/plugins/SPU2null/Linux/Config.h similarity index 62% rename from plugins/FWnull/Linux/Linux.cpp rename to plugins/SPU2null/Linux/Config.h index 626584ce52..ea196262a9 100644 --- a/plugins/FWnull/Linux/Linux.cpp +++ b/plugins/SPU2null/Linux/Config.h @@ -1,5 +1,5 @@ -/* FWnull - * Copyright (C) 2004-2009 PCSX2 Team +/* SPU2null + * Copyright (C) 2002-2005 SPU2null Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -14,40 +14,17 @@ * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - + */ + #include -#include #include -#include -#include -#include -#include -#include +#include +#include +using namespace std; -#include "FW.h" -#include "Config.h" - -void SysMessage(char *fmt, ...) -{ - va_list list; - char msg[512]; - char cmd[512]; - - va_start(list, fmt); - vsprintf(msg, fmt, list); - va_end(list); - - cfgSysMessage(msg); -} - -void FWconfigure() -{ - CFGconfigure(); -} - -void FWabout() -{ - CFGabout(); -} +void SaveConf(); +void LoadConf(); +extern string s_strIniPath; +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) diff --git a/plugins/SPU2null/Linux/callbacks.c b/plugins/SPU2null/Linux/callbacks.c new file mode 100644 index 0000000000..e87ed8be2d --- /dev/null +++ b/plugins/SPU2null/Linux/callbacks.c @@ -0,0 +1,34 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/SPU2null/Linux/callbacks.h b/plugins/SPU2null/Linux/callbacks.h new file mode 100644 index 0000000000..e606c242f0 --- /dev/null +++ b/plugins/SPU2null/Linux/callbacks.h @@ -0,0 +1,14 @@ +#include + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/SPU2null/Linux/interface.c b/plugins/SPU2null/Linux/interface.c new file mode 100644 index 0000000000..1efa3adbbd --- /dev/null +++ b/plugins/SPU2null/Linux/interface.c @@ -0,0 +1,173 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + +#define GLADE_HOOKUP_OBJECT(component,widget,name) \ + g_object_set_data_full (G_OBJECT (component), name, \ + gtk_widget_ref (widget), (GDestroyNotify) gtk_widget_unref) + +#define GLADE_HOOKUP_OBJECT_NO_REF(component,widget,name) \ + g_object_set_data (G_OBJECT (component), name, widget) + +GtkWidget* +create_Config (void) +{ + GtkWidget *Config; + GtkWidget *vbox1; + GtkWidget *frame3; + GtkWidget *alignment1; + GtkWidget *check_logging; + GtkWidget *label15; + GtkWidget *hbuttonbox1; + GtkWidget *button1; + GtkWidget *button2; + + Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); + gtk_container_set_border_width (GTK_CONTAINER (Config), 5); + gtk_window_set_title (GTK_WINDOW (Config), _("SPU2config")); + + vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); + gtk_widget_show (vbox1); + gtk_container_add (GTK_CONTAINER (Config), vbox1); + gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); + + frame3 = gtk_frame_new (NULL); + gtk_widget_set_name (frame3, "frame3"); + gtk_widget_show (frame3); + gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + gtk_frame_set_shadow_type (GTK_FRAME (frame3), GTK_SHADOW_NONE); + + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame3), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); + + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); + + label15 = gtk_label_new (_("Logging")); + gtk_widget_set_name (label15, "label15"); + gtk_widget_show (label15); + gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + gtk_label_set_use_markup (GTK_LABEL (label15), TRUE); + + hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); + gtk_widget_show (hbuttonbox1); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + + button1 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button1, "button1"); + gtk_widget_show (button1); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); + GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + + button2 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button2, "button2"); + gtk_widget_show (button2); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); + GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button1, "clicked", + G_CALLBACK (OnConf_Ok), + NULL); + g_signal_connect ((gpointer) button2, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); + GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); + GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); + GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); + GLADE_HOOKUP_OBJECT (Config, button1, "button1"); + GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + + return Config; +} + +GtkWidget* +create_About (void) +{ + GtkWidget *About; + GtkWidget *vbox2; + GtkWidget *label2; + GtkWidget *label3; + GtkWidget *hbuttonbox2; + GtkWidget *button3; + + About = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_title (GTK_WINDOW (About), _("SPU2about")); + + vbox2 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox2, "vbox2"); + gtk_widget_show (vbox2); + gtk_container_add (GTK_CONTAINER (About), vbox2); + gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); + + label2 = gtk_label_new (_("SPU2 Null Driver")); + gtk_widget_set_name (label2, "label2"); + gtk_widget_show (label2); + gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); + + label3 = gtk_label_new (_("Original Author: ? (zerofrog? linuz?)\nRevised by arcum42@gmail.com")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); + + hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); + gtk_widget_show (hbuttonbox2); + gtk_box_pack_start (GTK_BOX (vbox2), hbuttonbox2, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox2), 30); + + button3 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button3, "button3"); + gtk_widget_show (button3); + gtk_container_add (GTK_CONTAINER (hbuttonbox2), button3); + GTK_WIDGET_SET_FLAGS (button3, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button3, "clicked", + G_CALLBACK (OnAbout_Ok), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); + GLADE_HOOKUP_OBJECT (About, vbox2, "vbox2"); + GLADE_HOOKUP_OBJECT (About, label2, "label2"); + GLADE_HOOKUP_OBJECT (About, label3, "label3"); + GLADE_HOOKUP_OBJECT (About, hbuttonbox2, "hbuttonbox2"); + GLADE_HOOKUP_OBJECT (About, button3, "button3"); + + return About; +} + diff --git a/plugins/SPU2null/Linux/interface.h b/plugins/SPU2null/Linux/interface.h new file mode 100644 index 0000000000..69c303ba3b --- /dev/null +++ b/plugins/SPU2null/Linux/interface.h @@ -0,0 +1,6 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +GtkWidget* create_Config (void); +GtkWidget* create_About (void); diff --git a/plugins/SPU2null/Linux/spu2.glade b/plugins/SPU2null/Linux/spu2.glade new file mode 100644 index 0000000000..75ce9f6b74 --- /dev/null +++ b/plugins/SPU2null/Linux/spu2.glade @@ -0,0 +1,241 @@ + + + + + + + 5 + True + SPU2config + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + 0 + 0.5 + GTK_SHADOW_NONE + + + + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + True + Enable Logging + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + + + + + True + <b>Logging</b> + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + + True + True + True + Cancel + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + + 5 + True + SPU2about + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + SPU2 Null Driver + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + Original Author: ? (zerofrog? linuz?) +Revised by arcum42@gmail.com + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + diff --git a/plugins/SPU2null/Linux/support.c b/plugins/SPU2null/Linux/support.c new file mode 100644 index 0000000000..00aff29822 --- /dev/null +++ b/plugins/SPU2null/Linux/support.c @@ -0,0 +1,144 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include + +#include "support.h" + +GtkWidget* +lookup_widget (GtkWidget *widget, + const gchar *widget_name) +{ + GtkWidget *parent, *found_widget; + + for (;;) + { + if (GTK_IS_MENU (widget)) + parent = gtk_menu_get_attach_widget (GTK_MENU (widget)); + else + parent = widget->parent; + if (!parent) + parent = (GtkWidget*) g_object_get_data (G_OBJECT (widget), "GladeParentKey"); + if (parent == NULL) + break; + widget = parent; + } + + found_widget = (GtkWidget*) g_object_get_data (G_OBJECT (widget), + widget_name); + if (!found_widget) + g_warning ("Widget not found: %s", widget_name); + return found_widget; +} + +static GList *pixmaps_directories = NULL; + +/* Use this function to set the directory containing installed pixmaps. */ +void +add_pixmap_directory (const gchar *directory) +{ + pixmaps_directories = g_list_prepend (pixmaps_directories, + g_strdup (directory)); +} + +/* This is an internally used function to find pixmap files. */ +static gchar* +find_pixmap_file (const gchar *filename) +{ + GList *elem; + + /* We step through each of the pixmaps directory to find it. */ + elem = pixmaps_directories; + while (elem) + { + gchar *pathname = g_strdup_printf ("%s%s%s", (gchar*)elem->data, + G_DIR_SEPARATOR_S, filename); + if (g_file_test (pathname, G_FILE_TEST_EXISTS)) + return pathname; + g_free (pathname); + elem = elem->next; + } + return NULL; +} + +/* This is an internally used function to create pixmaps. */ +GtkWidget* +create_pixmap (GtkWidget *widget, + const gchar *filename) +{ + gchar *pathname = NULL; + GtkWidget *pixmap; + + if (!filename || !filename[0]) + return gtk_image_new (); + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return gtk_image_new (); + } + + pixmap = gtk_image_new_from_file (pathname); + g_free (pathname); + return pixmap; +} + +/* This is an internally used function to create pixmaps. */ +GdkPixbuf* +create_pixbuf (const gchar *filename) +{ + gchar *pathname = NULL; + GdkPixbuf *pixbuf; + GError *error = NULL; + + if (!filename || !filename[0]) + return NULL; + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return NULL; + } + + pixbuf = gdk_pixbuf_new_from_file (pathname, &error); + if (!pixbuf) + { + fprintf (stderr, "Failed to load pixbuf file: %s: %s\n", + pathname, error->message); + g_error_free (error); + } + g_free (pathname); + return pixbuf; +} + +/* This is used to set ATK action descriptions. */ +void +glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description) +{ + gint n_actions, i; + + n_actions = atk_action_get_n_actions (action); + for (i = 0; i < n_actions; i++) + { + if (!strcmp (atk_action_get_name (action, i), action_name)) + atk_action_set_description (action, i, description); + } +} + diff --git a/plugins/SPU2null/Linux/support.h b/plugins/SPU2null/Linux/support.h new file mode 100644 index 0000000000..a32649e53c --- /dev/null +++ b/plugins/SPU2null/Linux/support.h @@ -0,0 +1,69 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* + * Standard gettext macros. + */ +#ifdef ENABLE_NLS +# include +# undef _ +# define _(String) dgettext (PACKAGE, String) +# define Q_(String) g_strip_context ((String), gettext (String)) +# ifdef gettext_noop +# define N_(String) gettext_noop (String) +# else +# define N_(String) (String) +# endif +#else +# define textdomain(String) (String) +# define gettext(String) (String) +# define dgettext(Domain,Message) (Message) +# define dcgettext(Domain,Message,Type) (Message) +# define bindtextdomain(Domain,Directory) (Domain) +# define _(String) (String) +# define Q_(String) g_strip_context ((String), (String)) +# define N_(String) (String) +#endif + + +/* + * Public Functions. + */ + +/* + * This function returns a widget in a component created by Glade. + * Call it with the toplevel widget in the component (i.e. a window/dialog), + * or alternatively any widget in the component, and the name of the widget + * you want returned. + */ +GtkWidget* lookup_widget (GtkWidget *widget, + const gchar *widget_name); + + +/* Use this function to set the directory containing installed pixmaps. */ +void add_pixmap_directory (const gchar *directory); + + +/* + * Private Functions. + */ + +/* This is used to create the pixmaps used in the interface. */ +GtkWidget* create_pixmap (GtkWidget *widget, + const gchar *filename); + +/* This is used to create the pixbufs used in the interface. */ +GdkPixbuf* create_pixbuf (const gchar *filename); + +/* This is used to set ATK action descriptions. */ +void glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description); + diff --git a/plugins/SPU2null/Makefile.am b/plugins/SPU2null/Makefile.am index 394dba22a3..768683e86f 100644 --- a/plugins/SPU2null/Makefile.am +++ b/plugins/SPU2null/Makefile.am @@ -22,6 +22,11 @@ libSPU2null_LDFLAGS= @SHARED_LDFLAGS@ libSPU2null_LDFLAGS+=-Wl,-soname,@libSPU2null_SONAME@ libSPU2null_LDADD=$(libSPU2null_a_OBJECTS) -libSPU2null_a_SOURCES = SPU2.cpp SPU2.h Linux/Config.cpp +libSPU2null_a_SOURCES = SPU2.cpp SPU2.h Linux/Config.cpp + +libSPU2null_a_SOURCES += \ +Linux/interface.h Linux/support.c \ +Linux/interface.c Linux/support.h \ +Linux/callbacks.h #SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/SPU2null/SPU2.cpp b/plugins/SPU2null/SPU2.cpp index e8a8e23697..7040b31178 100644 --- a/plugins/SPU2null/SPU2.cpp +++ b/plugins/SPU2null/SPU2.cpp @@ -20,7 +20,9 @@ #include #include - +#include +using namespace std; + const unsigned char version = PS2E_SPU2_VERSION; const unsigned char revision = 0; const unsigned char build = 7; // increase that with each version @@ -38,6 +40,7 @@ char *libraryName = "SPU2null (Debug)"; #else char *libraryName = "SPU2null "; #endif +string s_strIniPath="inis/SPU2null.ini"; FILE *spu2Log; Config conf; diff --git a/plugins/SPU2null/SPU2.h b/plugins/SPU2null/SPU2.h index 842ff27b85..4b6b1bf7b5 100644 --- a/plugins/SPU2null/SPU2.h +++ b/plugins/SPU2null/SPU2.h @@ -15,7 +15,7 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - + #ifndef __SPU2_H__ #define __SPU2_H__ diff --git a/plugins/USBnull/Linux/Config.cpp b/plugins/USBnull/Linux/Config.cpp index f7b43f4852..a931f6cfde 100644 --- a/plugins/USBnull/Linux/Config.cpp +++ b/plugins/USBnull/Linux/Config.cpp @@ -26,6 +26,9 @@ #include using namespace std; +#include "USB.h" +#include "Config.h" + #ifdef __cplusplus extern "C" { @@ -39,8 +42,6 @@ extern "C" } #endif -#include "USB.h" - GtkWidget *MsgDlg, *About, *Conf; extern string s_strIniPath; @@ -50,7 +51,7 @@ void OnMsg_Ok() gtk_main_quit(); } -void cfgSysMessage(char *fmt, ...) +void SysMessage(char *fmt, ...) { GtkWidget *Ok, *Txt; GtkWidget *Box, *Box1; @@ -98,7 +99,7 @@ void OnAbout_Ok(GtkButton *button, gpointer user_data) gtk_main_quit(); } -void CFGabout() +EXPORT_C_(void) USBabout() { About = create_About(); gtk_widget_show_all(About); @@ -107,6 +108,9 @@ void CFGabout() void OnConf_Ok(GtkButton *button, gpointer user_data) { + conf.Log = is_checked(Conf, "check_logging"); + SaveConfig(); + gtk_widget_destroy(Conf); gtk_main_quit(); } @@ -117,21 +121,16 @@ void OnConf_Cancel(GtkButton *button, gpointer user_data) gtk_main_quit(); } -void CFGconfigure() +EXPORT_C_(void) USBconfigure() { + LoadConfig(); Conf = create_Config(); + set_checked(Conf, "check_logging", conf.Log); gtk_widget_show_all(Conf); gtk_main(); } -long CFGmessage(char *msg) -{ - cfgSysMessage(msg); - - return 0; -} - void LoadConfig() { FILE *f; @@ -141,10 +140,11 @@ void LoadConfig() f = fopen(cfg, "r"); if (f == NULL) { - printf("failed to open %s\n", s_strIniPath.c_str()); + printf("failed to open '%s'\n", s_strIniPath.c_str()); SaveConfig();//save and return return; } + fscanf(f, "logging = %hhx\n", &conf.Log); //fscanf(f, "options = %hhx\n", &confOptions); fclose(f); } @@ -158,10 +158,11 @@ void SaveConfig() f = fopen(cfg,"w"); if (f == NULL) { - printf("failed to open %s\n", s_strIniPath.c_str()); + printf("failed to open '%s'\n", s_strIniPath.c_str()); return; } + fprintf(f, "logging = %hhx\n", conf.Log); //fprintf(f, "options = %hhx\n", confOptions); fclose(f); } diff --git a/plugins/USBnull/Linux/Config.h b/plugins/USBnull/Linux/Config.h index c4036d30ed..54fddc414f 100644 --- a/plugins/USBnull/Linux/Config.h +++ b/plugins/USBnull/Linux/Config.h @@ -18,7 +18,7 @@ void SaveConfig(); void LoadConfig(); -void cfgSysMessage(char *fmt, ...); -void CFGabout(); -void CFGconfigure(); -long CFGmessage(char *msg); +void SysMessage(char *fmt, ...); + +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) diff --git a/plugins/USBnull/Linux/Linux.cpp b/plugins/USBnull/Linux/Linux.cpp index 739a7a8162..8ab4944954 100644 --- a/plugins/USBnull/Linux/Linux.cpp +++ b/plugins/USBnull/Linux/Linux.cpp @@ -28,7 +28,7 @@ #include "USB.h" #include "Config.h" -void SysMessage(char *fmt, ...) +/*void SysMessage(char *fmt, ...) { va_list list; char msg[512]; @@ -39,9 +39,9 @@ void SysMessage(char *fmt, ...) va_end(list); cfgSysMessage(msg); -} +}*/ -void USBconfigure() +/*void USBconfigure() { CFGconfigure(); } @@ -49,5 +49,5 @@ void USBconfigure() void USBabout() { CFGabout(); -} +}*/ diff --git a/plugins/USBnull/Linux/interface.c b/plugins/USBnull/Linux/interface.c index 5c23358a21..0e304b63fe 100644 --- a/plugins/USBnull/Linux/interface.c +++ b/plugins/USBnull/Linux/interface.c @@ -31,28 +31,18 @@ create_Config (void) { GtkWidget *Config; GtkWidget *vbox1; - GtkWidget *frame2; - GtkWidget *hbox1; - GtkWidget *label4; - GtkWidget *GtkCombo_Eth; - GList *GtkCombo_Eth_items = NULL; - GtkWidget *combo_entry1; - GtkWidget *label1; GtkWidget *frame3; - GtkWidget *hbox2; - GtkWidget *label5; - GtkWidget *GtkCombo_Hdd; - GList *GtkCombo_Hdd_items = NULL; - GtkWidget *entry1; + GtkWidget *alignment1; + GtkWidget *check_logging; GtkWidget *label15; - GtkWidget *hbuttonbox1; - GtkWidget *button1; - GtkWidget *button2; + GtkWidget *hbuttonbox3; + GtkWidget *button4; + GtkWidget *button5; Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); gtk_widget_set_name (Config, "Config"); gtk_container_set_border_width (GTK_CONTAINER (Config), 5); - gtk_window_set_title (GTK_WINDOW (Config), _("DEV9config")); + gtk_window_set_title (GTK_WINDOW (Config), _("USBconfig")); vbox1 = gtk_vbox_new (FALSE, 5); gtk_widget_set_name (vbox1, "vbox1"); @@ -60,121 +50,63 @@ create_Config (void) gtk_container_add (GTK_CONTAINER (Config), vbox1); gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); - frame2 = gtk_frame_new (NULL); - gtk_widget_set_name (frame2, "frame2"); - gtk_widget_show (frame2); - gtk_box_pack_start (GTK_BOX (vbox1), frame2, TRUE, TRUE, 0); - - hbox1 = gtk_hbox_new (TRUE, 5); - gtk_widget_set_name (hbox1, "hbox1"); - gtk_widget_show (hbox1); - gtk_container_add (GTK_CONTAINER (frame2), hbox1); - gtk_container_set_border_width (GTK_CONTAINER (hbox1), 5); - - label4 = gtk_label_new (_("Device:")); - gtk_widget_set_name (label4, "label4"); - gtk_widget_show (label4); - gtk_box_pack_start (GTK_BOX (hbox1), label4, FALSE, FALSE, 0); - gtk_label_set_justify (GTK_LABEL (label4), GTK_JUSTIFY_CENTER); - - GtkCombo_Eth = gtk_combo_new (); - g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Eth)->popwin), - "GladeParentKey", GtkCombo_Eth); - gtk_widget_set_name (GtkCombo_Eth, "GtkCombo_Eth"); - gtk_widget_show (GtkCombo_Eth); - gtk_box_pack_start (GTK_BOX (hbox1), GtkCombo_Eth, FALSE, FALSE, 0); - GtkCombo_Eth_items = g_list_append (GtkCombo_Eth_items, (gpointer) ""); - gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Eth), GtkCombo_Eth_items); - g_list_free (GtkCombo_Eth_items); - - combo_entry1 = GTK_COMBO (GtkCombo_Eth)->entry; - gtk_widget_set_name (combo_entry1, "combo_entry1"); - gtk_widget_show (combo_entry1); - - label1 = gtk_label_new (_("Ethernet")); - gtk_widget_set_name (label1, "label1"); - gtk_widget_show (label1); - gtk_frame_set_label_widget (GTK_FRAME (frame2), label1); - frame3 = gtk_frame_new (NULL); gtk_widget_set_name (frame3, "frame3"); gtk_widget_show (frame3); gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); - hbox2 = gtk_hbox_new (TRUE, 5); - gtk_widget_set_name (hbox2, "hbox2"); - gtk_widget_show (hbox2); - gtk_container_add (GTK_CONTAINER (frame3), hbox2); - gtk_container_set_border_width (GTK_CONTAINER (hbox2), 5); + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame3), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); - label5 = gtk_label_new (_("Device:")); - gtk_widget_set_name (label5, "label5"); - gtk_widget_show (label5); - gtk_box_pack_start (GTK_BOX (hbox2), label5, FALSE, FALSE, 0); - gtk_label_set_justify (GTK_LABEL (label5), GTK_JUSTIFY_CENTER); + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); - GtkCombo_Hdd = gtk_combo_new (); - g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Hdd)->popwin), - "GladeParentKey", GtkCombo_Hdd); - gtk_widget_set_name (GtkCombo_Hdd, "GtkCombo_Hdd"); - gtk_widget_show (GtkCombo_Hdd); - gtk_box_pack_start (GTK_BOX (hbox2), GtkCombo_Hdd, FALSE, FALSE, 0); - GtkCombo_Hdd_items = g_list_append (GtkCombo_Hdd_items, (gpointer) ""); - gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Hdd), GtkCombo_Hdd_items); - g_list_free (GtkCombo_Hdd_items); - - entry1 = GTK_COMBO (GtkCombo_Hdd)->entry; - gtk_widget_set_name (entry1, "entry1"); - gtk_widget_show (entry1); - - label15 = gtk_label_new (_("Hdd")); + label15 = gtk_label_new (_("Logging")); gtk_widget_set_name (label15, "label15"); gtk_widget_show (label15); gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + gtk_label_set_use_markup (GTK_LABEL (label15), TRUE); - hbuttonbox1 = gtk_hbutton_box_new (); - gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); - gtk_widget_show (hbuttonbox1); - gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); - gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + hbuttonbox3 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox3, "hbuttonbox3"); + gtk_widget_show (hbuttonbox3); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox3, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox3), 30); - button1 = gtk_button_new_with_mnemonic (_("Ok")); - gtk_widget_set_name (button1, "button1"); - gtk_widget_show (button1); - gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); - GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + button4 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button4, "button4"); + gtk_widget_show (button4); + gtk_container_add (GTK_CONTAINER (hbuttonbox3), button4); + GTK_WIDGET_SET_FLAGS (button4, GTK_CAN_DEFAULT); - button2 = gtk_button_new_with_mnemonic (_("Cancel")); - gtk_widget_set_name (button2, "button2"); - gtk_widget_show (button2); - gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); - GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + button5 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button5, "button5"); + gtk_widget_show (button5); + gtk_container_add (GTK_CONTAINER (hbuttonbox3), button5); + GTK_WIDGET_SET_FLAGS (button5, GTK_CAN_DEFAULT); - g_signal_connect ((gpointer) button1, "clicked", + g_signal_connect ((gpointer) button4, "clicked", G_CALLBACK (OnConf_Ok), NULL); - g_signal_connect ((gpointer) button2, "clicked", + g_signal_connect ((gpointer) button5, "clicked", G_CALLBACK (OnConf_Cancel), NULL); /* Store pointers to all widgets, for use by lookup_widget(). */ GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); - GLADE_HOOKUP_OBJECT (Config, frame2, "frame2"); - GLADE_HOOKUP_OBJECT (Config, hbox1, "hbox1"); - GLADE_HOOKUP_OBJECT (Config, label4, "label4"); - GLADE_HOOKUP_OBJECT (Config, GtkCombo_Eth, "GtkCombo_Eth"); - GLADE_HOOKUP_OBJECT (Config, combo_entry1, "combo_entry1"); - GLADE_HOOKUP_OBJECT (Config, label1, "label1"); GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); - GLADE_HOOKUP_OBJECT (Config, hbox2, "hbox2"); - GLADE_HOOKUP_OBJECT (Config, label5, "label5"); - GLADE_HOOKUP_OBJECT (Config, GtkCombo_Hdd, "GtkCombo_Hdd"); - GLADE_HOOKUP_OBJECT (Config, entry1, "entry1"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); GLADE_HOOKUP_OBJECT (Config, label15, "label15"); - GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); - GLADE_HOOKUP_OBJECT (Config, button1, "button1"); - GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox3, "hbuttonbox3"); + GLADE_HOOKUP_OBJECT (Config, button4, "button4"); + GLADE_HOOKUP_OBJECT (Config, button5, "button5"); return Config; } @@ -200,13 +132,13 @@ create_About (void) gtk_container_add (GTK_CONTAINER (About), vbox2); gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); - label2 = gtk_label_new (_("FireWire Driver")); + label2 = gtk_label_new (_("USB Null Driver")); gtk_widget_set_name (label2, "label2"); gtk_widget_show (label2); gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); - label3 = gtk_label_new (_("Author: linuzappz ")); + label3 = gtk_label_new (_("Original Author: linuzappz \nRevised by arcum42@gmail.com.")); gtk_widget_set_name (label3, "label3"); gtk_widget_show (label3); gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); diff --git a/plugins/USBnull/Linux/usbnull.glade b/plugins/USBnull/Linux/usbnull.glade index cef4eab65e..03afe746ac 100644 --- a/plugins/USBnull/Linux/usbnull.glade +++ b/plugins/USBnull/Linux/usbnull.glade @@ -6,7 +6,7 @@ 5 True - DEV9config + USBconfig GTK_WINDOW_TOPLEVEL GTK_WIN_POS_NONE False @@ -27,140 +27,6 @@ False 5 - - - True - 0 - 0.5 - GTK_SHADOW_ETCHED_IN - - - - 5 - True - True - 5 - - - - True - Device: - False - False - GTK_JUSTIFY_CENTER - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - 0 - False - False - - - - - - True - False - True - False - True - False - - - - True - True - True - True - 0 - - True - * - False - - - - - - True - GTK_SELECTION_BROWSE - - - - True - - - - True - - False - False - GTK_JUSTIFY_LEFT - False - False - 0 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - - - - - - - 0 - False - False - - - - - - - - True - Ethernet - False - False - GTK_JUSTIFY_LEFT - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - label_item - - - - - 0 - True - True - - - True @@ -169,98 +35,29 @@ GTK_SHADOW_ETCHED_IN - - 5 + True - True - 5 + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 - + True - Device: - False - False - GTK_JUSTIFY_CENTER - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 + True + Enable Logging + True + GTK_RELIEF_NORMAL + True + False + False + True - - 0 - False - False - - - - - - True - False - True - False - True - False - - - - True - True - True - True - 0 - - True - * - False - - - - - - True - GTK_SELECTION_BROWSE - - - - True - - - - True - - False - False - GTK_JUSTIFY_LEFT - False - False - 0 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - - - - - - - 0 - False - False - @@ -268,9 +65,9 @@ True - Hdd + <b>Logging</b> False - False + True GTK_JUSTIFY_LEFT False False @@ -296,13 +93,13 @@ - + True GTK_BUTTONBOX_DEFAULT_STYLE 30 - + True True True @@ -315,7 +112,7 @@ - + True True True @@ -364,7 +161,7 @@ True - FireWire Driver + USB Null Driver False False GTK_JUSTIFY_CENTER @@ -389,7 +186,8 @@ True - Author: linuzappz <linuzappz@hotmail.com> + Original Author: linuzappz <linuzappz@hotmail.com> +Revised by arcum42@gmail.com. False False GTK_JUSTIFY_LEFT diff --git a/plugins/USBnull/USB.cpp b/plugins/USBnull/USB.cpp index 6aa6bab4bc..ba540e73a1 100644 --- a/plugins/USBnull/USB.cpp +++ b/plugins/USBnull/USB.cpp @@ -23,13 +23,13 @@ using namespace std; #include "USB.h" -string s_strIniPath; const unsigned char version = PS2E_USB_VERSION; const unsigned char revision = 0; const unsigned char build = 6; // increase that with each version static char *libraryName = "USBnull Driver"; +string s_strIniPath="inis/USBnull.ini"; //void (*USBirq)(); USBcallback USBirq; Config conf; diff --git a/plugins/dev9null/Linux/callbacks.c b/plugins/dev9null/Linux/callbacks.c new file mode 100644 index 0000000000..e87ed8be2d --- /dev/null +++ b/plugins/dev9null/Linux/callbacks.c @@ -0,0 +1,34 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/dev9null/Linux/callbacks.h b/plugins/dev9null/Linux/callbacks.h new file mode 100644 index 0000000000..e606c242f0 --- /dev/null +++ b/plugins/dev9null/Linux/callbacks.h @@ -0,0 +1,14 @@ +#include + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/dev9null/Linux/dev9.glade b/plugins/dev9null/Linux/dev9.glade new file mode 100644 index 0000000000..10b8a7f4d1 --- /dev/null +++ b/plugins/dev9null/Linux/dev9.glade @@ -0,0 +1,444 @@ + + + + + + + 5 + True + DEV9config + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + 5 + True + True + 5 + + + + True + Device: + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + False + True + False + True + False + + + + True + True + True + True + 0 + + True + * + False + + + + + + True + GTK_SELECTION_BROWSE + + + + True + + + + True + + False + False + GTK_JUSTIFY_LEFT + False + False + 0 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + + + + + + + 0 + False + False + + + + + + + + True + Ethernet + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + 5 + True + True + 5 + + + + True + Device: + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + False + True + False + True + False + + + + True + True + True + True + 0 + + True + * + False + + + + + + True + GTK_SELECTION_BROWSE + + + + True + + + + True + + False + False + GTK_JUSTIFY_LEFT + False + False + 0 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + + + + + + + 0 + False + False + + + + + + + + True + Hdd + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + + True + True + True + Cancel + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + + 5 + True + DEV9about + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + Dev9 null Driver + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + Original Author: linuzappz <linuzappz@hotmail.com> +Revised by arcum42@gmail.com + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + diff --git a/plugins/dev9null/Linux/interface.c b/plugins/dev9null/Linux/interface.c new file mode 100644 index 0000000000..eef82eaf94 --- /dev/null +++ b/plugins/dev9null/Linux/interface.c @@ -0,0 +1,240 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + +#define GLADE_HOOKUP_OBJECT(component,widget,name) \ + g_object_set_data_full (G_OBJECT (component), name, \ + gtk_widget_ref (widget), (GDestroyNotify) gtk_widget_unref) + +#define GLADE_HOOKUP_OBJECT_NO_REF(component,widget,name) \ + g_object_set_data (G_OBJECT (component), name, widget) + +GtkWidget* +create_Config (void) +{ + GtkWidget *Config; + GtkWidget *vbox1; + GtkWidget *frame2; + GtkWidget *hbox1; + GtkWidget *label4; + GtkWidget *GtkCombo_Eth; + GList *GtkCombo_Eth_items = NULL; + GtkWidget *combo_entry1; + GtkWidget *label1; + GtkWidget *frame3; + GtkWidget *hbox2; + GtkWidget *label5; + GtkWidget *GtkCombo_Hdd; + GList *GtkCombo_Hdd_items = NULL; + GtkWidget *entry1; + GtkWidget *label15; + GtkWidget *hbuttonbox1; + GtkWidget *button1; + GtkWidget *button2; + + Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); + gtk_container_set_border_width (GTK_CONTAINER (Config), 5); + gtk_window_set_title (GTK_WINDOW (Config), _("DEV9config")); + + vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); + gtk_widget_show (vbox1); + gtk_container_add (GTK_CONTAINER (Config), vbox1); + gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); + + frame2 = gtk_frame_new (NULL); + gtk_widget_set_name (frame2, "frame2"); + gtk_widget_show (frame2); + gtk_box_pack_start (GTK_BOX (vbox1), frame2, TRUE, TRUE, 0); + + hbox1 = gtk_hbox_new (TRUE, 5); + gtk_widget_set_name (hbox1, "hbox1"); + gtk_widget_show (hbox1); + gtk_container_add (GTK_CONTAINER (frame2), hbox1); + gtk_container_set_border_width (GTK_CONTAINER (hbox1), 5); + + label4 = gtk_label_new (_("Device:")); + gtk_widget_set_name (label4, "label4"); + gtk_widget_show (label4); + gtk_box_pack_start (GTK_BOX (hbox1), label4, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label4), GTK_JUSTIFY_CENTER); + + GtkCombo_Eth = gtk_combo_new (); + g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Eth)->popwin), + "GladeParentKey", GtkCombo_Eth); + gtk_widget_set_name (GtkCombo_Eth, "GtkCombo_Eth"); + gtk_widget_show (GtkCombo_Eth); + gtk_box_pack_start (GTK_BOX (hbox1), GtkCombo_Eth, FALSE, FALSE, 0); + GtkCombo_Eth_items = g_list_append (GtkCombo_Eth_items, (gpointer) ""); + gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Eth), GtkCombo_Eth_items); + g_list_free (GtkCombo_Eth_items); + + combo_entry1 = GTK_COMBO (GtkCombo_Eth)->entry; + gtk_widget_set_name (combo_entry1, "combo_entry1"); + gtk_widget_show (combo_entry1); + + label1 = gtk_label_new (_("Ethernet")); + gtk_widget_set_name (label1, "label1"); + gtk_widget_show (label1); + gtk_frame_set_label_widget (GTK_FRAME (frame2), label1); + + frame3 = gtk_frame_new (NULL); + gtk_widget_set_name (frame3, "frame3"); + gtk_widget_show (frame3); + gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + + hbox2 = gtk_hbox_new (TRUE, 5); + gtk_widget_set_name (hbox2, "hbox2"); + gtk_widget_show (hbox2); + gtk_container_add (GTK_CONTAINER (frame3), hbox2); + gtk_container_set_border_width (GTK_CONTAINER (hbox2), 5); + + label5 = gtk_label_new (_("Device:")); + gtk_widget_set_name (label5, "label5"); + gtk_widget_show (label5); + gtk_box_pack_start (GTK_BOX (hbox2), label5, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label5), GTK_JUSTIFY_CENTER); + + GtkCombo_Hdd = gtk_combo_new (); + g_object_set_data (G_OBJECT (GTK_COMBO (GtkCombo_Hdd)->popwin), + "GladeParentKey", GtkCombo_Hdd); + gtk_widget_set_name (GtkCombo_Hdd, "GtkCombo_Hdd"); + gtk_widget_show (GtkCombo_Hdd); + gtk_box_pack_start (GTK_BOX (hbox2), GtkCombo_Hdd, FALSE, FALSE, 0); + GtkCombo_Hdd_items = g_list_append (GtkCombo_Hdd_items, (gpointer) ""); + gtk_combo_set_popdown_strings (GTK_COMBO (GtkCombo_Hdd), GtkCombo_Hdd_items); + g_list_free (GtkCombo_Hdd_items); + + entry1 = GTK_COMBO (GtkCombo_Hdd)->entry; + gtk_widget_set_name (entry1, "entry1"); + gtk_widget_show (entry1); + + label15 = gtk_label_new (_("Hdd")); + gtk_widget_set_name (label15, "label15"); + gtk_widget_show (label15); + gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + + hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); + gtk_widget_show (hbuttonbox1); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + + button1 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button1, "button1"); + gtk_widget_show (button1); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); + GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + + button2 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button2, "button2"); + gtk_widget_show (button2); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); + GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button1, "clicked", + G_CALLBACK (OnConf_Ok), + NULL); + g_signal_connect ((gpointer) button2, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); + GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); + GLADE_HOOKUP_OBJECT (Config, frame2, "frame2"); + GLADE_HOOKUP_OBJECT (Config, hbox1, "hbox1"); + GLADE_HOOKUP_OBJECT (Config, label4, "label4"); + GLADE_HOOKUP_OBJECT (Config, GtkCombo_Eth, "GtkCombo_Eth"); + GLADE_HOOKUP_OBJECT (Config, combo_entry1, "combo_entry1"); + GLADE_HOOKUP_OBJECT (Config, label1, "label1"); + GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); + GLADE_HOOKUP_OBJECT (Config, hbox2, "hbox2"); + GLADE_HOOKUP_OBJECT (Config, label5, "label5"); + GLADE_HOOKUP_OBJECT (Config, GtkCombo_Hdd, "GtkCombo_Hdd"); + GLADE_HOOKUP_OBJECT (Config, entry1, "entry1"); + GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); + GLADE_HOOKUP_OBJECT (Config, button1, "button1"); + GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + + return Config; +} + +GtkWidget* +create_About (void) +{ + GtkWidget *About; + GtkWidget *vbox2; + GtkWidget *label2; + GtkWidget *label3; + GtkWidget *hbuttonbox2; + GtkWidget *button3; + + About = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_title (GTK_WINDOW (About), _("DEV9about")); + + vbox2 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox2, "vbox2"); + gtk_widget_show (vbox2); + gtk_container_add (GTK_CONTAINER (About), vbox2); + gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); + + label2 = gtk_label_new (_("Dev9 null Driver")); + gtk_widget_set_name (label2, "label2"); + gtk_widget_show (label2); + gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); + + label3 = gtk_label_new (_("Original Author: linuzappz \nRevised by arcum42@gmail.com")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); + + hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); + gtk_widget_show (hbuttonbox2); + gtk_box_pack_start (GTK_BOX (vbox2), hbuttonbox2, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox2), 30); + + button3 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button3, "button3"); + gtk_widget_show (button3); + gtk_container_add (GTK_CONTAINER (hbuttonbox2), button3); + GTK_WIDGET_SET_FLAGS (button3, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button3, "clicked", + G_CALLBACK (OnAbout_Ok), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); + GLADE_HOOKUP_OBJECT (About, vbox2, "vbox2"); + GLADE_HOOKUP_OBJECT (About, label2, "label2"); + GLADE_HOOKUP_OBJECT (About, label3, "label3"); + GLADE_HOOKUP_OBJECT (About, hbuttonbox2, "hbuttonbox2"); + GLADE_HOOKUP_OBJECT (About, button3, "button3"); + + return About; +} + diff --git a/plugins/dev9null/Linux/interface.h b/plugins/dev9null/Linux/interface.h new file mode 100644 index 0000000000..69c303ba3b --- /dev/null +++ b/plugins/dev9null/Linux/interface.h @@ -0,0 +1,6 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +GtkWidget* create_Config (void); +GtkWidget* create_About (void); diff --git a/plugins/dev9null/Linux/support.c b/plugins/dev9null/Linux/support.c new file mode 100644 index 0000000000..00aff29822 --- /dev/null +++ b/plugins/dev9null/Linux/support.c @@ -0,0 +1,144 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include + +#include "support.h" + +GtkWidget* +lookup_widget (GtkWidget *widget, + const gchar *widget_name) +{ + GtkWidget *parent, *found_widget; + + for (;;) + { + if (GTK_IS_MENU (widget)) + parent = gtk_menu_get_attach_widget (GTK_MENU (widget)); + else + parent = widget->parent; + if (!parent) + parent = (GtkWidget*) g_object_get_data (G_OBJECT (widget), "GladeParentKey"); + if (parent == NULL) + break; + widget = parent; + } + + found_widget = (GtkWidget*) g_object_get_data (G_OBJECT (widget), + widget_name); + if (!found_widget) + g_warning ("Widget not found: %s", widget_name); + return found_widget; +} + +static GList *pixmaps_directories = NULL; + +/* Use this function to set the directory containing installed pixmaps. */ +void +add_pixmap_directory (const gchar *directory) +{ + pixmaps_directories = g_list_prepend (pixmaps_directories, + g_strdup (directory)); +} + +/* This is an internally used function to find pixmap files. */ +static gchar* +find_pixmap_file (const gchar *filename) +{ + GList *elem; + + /* We step through each of the pixmaps directory to find it. */ + elem = pixmaps_directories; + while (elem) + { + gchar *pathname = g_strdup_printf ("%s%s%s", (gchar*)elem->data, + G_DIR_SEPARATOR_S, filename); + if (g_file_test (pathname, G_FILE_TEST_EXISTS)) + return pathname; + g_free (pathname); + elem = elem->next; + } + return NULL; +} + +/* This is an internally used function to create pixmaps. */ +GtkWidget* +create_pixmap (GtkWidget *widget, + const gchar *filename) +{ + gchar *pathname = NULL; + GtkWidget *pixmap; + + if (!filename || !filename[0]) + return gtk_image_new (); + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return gtk_image_new (); + } + + pixmap = gtk_image_new_from_file (pathname); + g_free (pathname); + return pixmap; +} + +/* This is an internally used function to create pixmaps. */ +GdkPixbuf* +create_pixbuf (const gchar *filename) +{ + gchar *pathname = NULL; + GdkPixbuf *pixbuf; + GError *error = NULL; + + if (!filename || !filename[0]) + return NULL; + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return NULL; + } + + pixbuf = gdk_pixbuf_new_from_file (pathname, &error); + if (!pixbuf) + { + fprintf (stderr, "Failed to load pixbuf file: %s: %s\n", + pathname, error->message); + g_error_free (error); + } + g_free (pathname); + return pixbuf; +} + +/* This is used to set ATK action descriptions. */ +void +glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description) +{ + gint n_actions, i; + + n_actions = atk_action_get_n_actions (action); + for (i = 0; i < n_actions; i++) + { + if (!strcmp (atk_action_get_name (action, i), action_name)) + atk_action_set_description (action, i, description); + } +} + diff --git a/plugins/dev9null/Linux/support.h b/plugins/dev9null/Linux/support.h new file mode 100644 index 0000000000..a32649e53c --- /dev/null +++ b/plugins/dev9null/Linux/support.h @@ -0,0 +1,69 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* + * Standard gettext macros. + */ +#ifdef ENABLE_NLS +# include +# undef _ +# define _(String) dgettext (PACKAGE, String) +# define Q_(String) g_strip_context ((String), gettext (String)) +# ifdef gettext_noop +# define N_(String) gettext_noop (String) +# else +# define N_(String) (String) +# endif +#else +# define textdomain(String) (String) +# define gettext(String) (String) +# define dgettext(Domain,Message) (Message) +# define dcgettext(Domain,Message,Type) (Message) +# define bindtextdomain(Domain,Directory) (Domain) +# define _(String) (String) +# define Q_(String) g_strip_context ((String), (String)) +# define N_(String) (String) +#endif + + +/* + * Public Functions. + */ + +/* + * This function returns a widget in a component created by Glade. + * Call it with the toplevel widget in the component (i.e. a window/dialog), + * or alternatively any widget in the component, and the name of the widget + * you want returned. + */ +GtkWidget* lookup_widget (GtkWidget *widget, + const gchar *widget_name); + + +/* Use this function to set the directory containing installed pixmaps. */ +void add_pixmap_directory (const gchar *directory); + + +/* + * Private Functions. + */ + +/* This is used to create the pixmaps used in the interface. */ +GtkWidget* create_pixmap (GtkWidget *widget, + const gchar *filename); + +/* This is used to create the pixbufs used in the interface. */ +GdkPixbuf* create_pixbuf (const gchar *filename); + +/* This is used to set ATK action descriptions. */ +void glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description); + diff --git a/plugins/zeropad/Linux/callbacks.c b/plugins/zeropad/Linux/callbacks.c index e3a92ed51e..cb65ceeb21 100644 --- a/plugins/zeropad/Linux/callbacks.c +++ b/plugins/zeropad/Linux/callbacks.c @@ -112,3 +112,12 @@ OnConf_Ok (GtkButton *button, } + +void +OnAbout_Ok (GtkDialog *dialog, + gint response_id, + gpointer user_data) +{ + +} + diff --git a/plugins/zeropad/Linux/callbacks.h b/plugins/zeropad/Linux/callbacks.h index d9c6d64ae1..3e57b0e5d4 100644 --- a/plugins/zeropad/Linux/callbacks.h +++ b/plugins/zeropad/Linux/callbacks.h @@ -52,3 +52,8 @@ OnConf_Cancel (GtkButton *button, void OnConf_Ok (GtkButton *button, gpointer user_data); + +void +OnAbout_Ok (GtkDialog *dialog, + gint response_id, + gpointer user_data); diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index 0111b4a6d4..199da357a3 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -270,9 +270,10 @@ void OnConf_Cancel(GtkButton *button, gpointer user_data) GtkWidget *About = NULL; -void OnAbout_Ok(GtkButton *button, gpointer user_data) +//void OnAbout_Ok(GtkButton *button, gpointer user_data) +void OnAbout_Ok(GtkDialog *About, gint response_id, gpointer user_data) { - gtk_widget_destroy(About); + gtk_widget_destroy(GTK_WIDGET(About)); gtk_main_quit(); } diff --git a/plugins/zeropad/Linux/interface.c b/plugins/zeropad/Linux/interface.c index 808afbc3e6..d8e669e802 100644 --- a/plugins/zeropad/Linux/interface.c +++ b/plugins/zeropad/Linux/interface.c @@ -44,11 +44,16 @@ create_About (void) About = gtk_about_dialog_new (); gtk_widget_set_name (About, "About"); gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_destroy_with_parent (GTK_WINDOW (About), TRUE); gtk_about_dialog_set_version (GTK_ABOUT_DIALOG (About), VERSION); gtk_about_dialog_set_name (GTK_ABOUT_DIALOG (About), _("ZeroPAD")); gtk_about_dialog_set_authors (GTK_ABOUT_DIALOG (About), authors); gtk_about_dialog_set_translator_credits (GTK_ABOUT_DIALOG (About), translators); + g_signal_connect ((gpointer) About, "response", + G_CALLBACK (OnAbout_Ok), + NULL); + /* Store pointers to all widgets, for use by lookup_widget(). */ GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); diff --git a/plugins/zeropad/Linux/zeropad.glade b/plugins/zeropad/Linux/zeropad.glade index c8fc069ca7..ef754e23a4 100644 --- a/plugins/zeropad/Linux/zeropad.glade +++ b/plugins/zeropad/Linux/zeropad.glade @@ -6,7 +6,7 @@ 5 True - False + True ZeroPAD False Author: zerofrog(@gmail.com) @@ -15,6 +15,7 @@ Thanks to: florin sasu <florinsasu@hotmail.com> and SSSPSXPad, TwinPad authors translator-credits + From 667581e4be100715cd9009204b9181ec4784a020 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 07:02:51 +0000 Subject: [PATCH 067/296] Linux: Fix focus problems with various dialogs when hitting escape, and hook in dev9's dialog box. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1190 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/LnxMain.cpp | 5 ++ pcsx2/Linux/callbacks.h | 4 ++ pcsx2/Linux/interface.c | 18 +++++ pcsx2/Linux/pcsx2.glade | 6 ++ plugins/USBnull/Makefile.am | 1 - plugins/USBnull/USB.cpp | 13 ++-- plugins/USBnull/USB.h | 5 +- plugins/dev9null/DEV9.cpp | 42 +++++++++--- plugins/dev9null/DEV9.h | 30 ++++++++- plugins/dev9null/Linux/Config.cpp | 102 +++++++++++++++++++++++++++-- plugins/dev9null/Linux/dev9.glade | 65 ++++++++++++++++++ plugins/dev9null/Linux/interface.c | 31 +++++++++ plugins/dev9null/Makefile.am | 6 +- 13 files changed, 301 insertions(+), 27 deletions(-) diff --git a/pcsx2/Linux/LnxMain.cpp b/pcsx2/Linux/LnxMain.cpp index 4f764cfac7..251d9f13d1 100644 --- a/pcsx2/Linux/LnxMain.cpp +++ b/pcsx2/Linux/LnxMain.cpp @@ -216,6 +216,11 @@ void On_Dialog_Cancelled(GtkButton* button, gpointer user_data) gtk_main_quit(); } +void On_Dialog_Close(GtkDialog* dialog, gpointer user_data) +{ + gtk_widget_set_sensitive(MainWindow, TRUE); +} + void RefreshMenuSlots() { GtkWidget *Item; diff --git a/pcsx2/Linux/callbacks.h b/pcsx2/Linux/callbacks.h index 771418a3ca..7e20155d17 100644 --- a/pcsx2/Linux/callbacks.h +++ b/pcsx2/Linux/callbacks.h @@ -257,3 +257,7 @@ OnLogging_Ok (GtkButton *button, void OnMemcards_Ok (GtkButton *button, gpointer user_data); + +void +On_Dialog_Close (GtkDialog *dialog, + gpointer user_data); diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index df94254c7e..5c1fcf9ed7 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -523,6 +523,9 @@ create_AdvDlg (void) gtk_dialog_add_action_widget (GTK_DIALOG (AdvDlg), button80, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (button80, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) AdvDlg, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) AdvDefaultBtn, "clicked", G_CALLBACK (on_Advanced_Defaults), NULL); @@ -831,6 +834,9 @@ create_SpeedHacksDlg (void) gtk_dialog_add_action_widget (GTK_DIALOG (SpeedHacksDlg), button98, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (button98, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) SpeedHacksDlg, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) EECycleHackScale, "value_changed", G_CALLBACK (on_ee_slider_changed), NULL); @@ -969,6 +975,9 @@ create_GameFixDlg (void) gtk_dialog_add_action_widget (GTK_DIALOG (GameFixDlg), button83, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (button83, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) GameFixDlg, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) cancelbutton1, "clicked", G_CALLBACK (on_Game_Fix_OK), NULL); @@ -4156,6 +4165,9 @@ create_CpuDlg (void) gtk_dialog_add_action_widget (GTK_DIALOG (CpuDlg), button97, 0); GTK_WIDGET_SET_FLAGS (button97, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) CpuDlg, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) button96, "clicked", G_CALLBACK (OnCpu_Ok), NULL); @@ -4521,6 +4533,9 @@ create_Logging (void) gtk_dialog_add_action_widget (GTK_DIALOG (Logging), Logging2Cancel, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (Logging2Cancel, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) Logging, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) Logging_Ok, "clicked", G_CALLBACK (OnLogging_Ok), NULL); @@ -4747,6 +4762,9 @@ create_MemDlg (void) gtk_dialog_add_action_widget (GTK_DIALOG (MemDlg), okbutton1, GTK_RESPONSE_CANCEL); GTK_WIDGET_SET_FLAGS (okbutton1, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) MemDlg, "close", + G_CALLBACK (On_Dialog_Close), + NULL); g_signal_connect ((gpointer) memcardcancelbutton, "clicked", G_CALLBACK (OnMemcards_Ok), NULL); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index a53cc57b60..397d21bb1b 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -19,6 +19,7 @@ True False True + @@ -1188,6 +1189,7 @@ Denormals are Zero - Your CPU makes Floating Point Denormals become Zero, so it True False True + @@ -1777,6 +1779,7 @@ If you have problems, Disable all of these and try again. True False True + @@ -6418,6 +6421,7 @@ Version x.x True False True + @@ -7295,6 +7299,7 @@ Version x.x True False True + @@ -8102,6 +8107,7 @@ Version x.x True False True + diff --git a/plugins/USBnull/Makefile.am b/plugins/USBnull/Makefile.am index 838ab78919..b341b28e91 100644 --- a/plugins/USBnull/Makefile.am +++ b/plugins/USBnull/Makefile.am @@ -29,6 +29,5 @@ libUSBnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ Linux/callbacks.h -#Linux/callbacks.c #SUBDIRS = Linux \ No newline at end of file diff --git a/plugins/USBnull/USB.cpp b/plugins/USBnull/USB.cpp index ba540e73a1..3afa17d50f 100644 --- a/plugins/USBnull/USB.cpp +++ b/plugins/USBnull/USB.cpp @@ -95,35 +95,35 @@ EXPORT_C_(void) USBclose() EXPORT_C_(u8 ) USBread8(u32 addr) { - USB_LOG("*UnKnown 8bit read at address %lx ", addr); + USB_LOG("*Unknown 8bit read at address %lx ", addr); return 0; } EXPORT_C_(u16) USBread16(u32 addr) { - USB_LOG("*UnKnown 16bit read at address %lx", addr); + USB_LOG("*Unknown 16bit read at address %lx", addr); return 0; } EXPORT_C_(u32) USBread32(u32 addr) { - USB_LOG("*UnKnown 32bit read at address %lx", addr); + USB_LOG("*Unknown 32bit read at address %lx", addr); return 0; } EXPORT_C_(void) USBwrite8(u32 addr, u8 value) { - USB_LOG("*UnKnown 8bit write at address %lx value %x\n", addr, value); + USB_LOG("*Unknown 8bit write at address %lx value %x\n", addr, value); } EXPORT_C_(void) USBwrite16(u32 addr, u16 value) { - USB_LOG("*UnKnown 16bit write at address %lx value %x\n", addr, value); + USB_LOG("*Unknown 16bit write at address %lx value %x\n", addr, value); } EXPORT_C_(void) USBwrite32(u32 addr, u32 value) { - USB_LOG("*UnKnown 32bit write at address %lx value %lx\n", addr, value); + USB_LOG("*Unknown 32bit write at address %lx value %lx\n", addr, value); } EXPORT_C_(void) USBirqCallback(USBcallback callback) @@ -143,6 +143,7 @@ EXPORT_C_(USBhandler) USBirqHandler(void) EXPORT_C_(void) USBsetRAM(void *mem) { + USB_LOG("*Setting ram.\n"); } // extended funcs diff --git a/plugins/USBnull/USB.h b/plugins/USBnull/USB.h index b10424e5d1..699ba1a432 100644 --- a/plugins/USBnull/USB.h +++ b/plugins/USBnull/USB.h @@ -1,5 +1,5 @@ -/* USBlinuz - * Copyright (C) 2002-2004 USBlinuz Team +/* USBnull + * Copyright (C) 2002-2009 pcsx2 Team * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -16,7 +16,6 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - #ifndef __USB_H__ #define __USB_H__ diff --git a/plugins/dev9null/DEV9.cpp b/plugins/dev9null/DEV9.cpp index 7f050a4723..7e8ce7ecb9 100644 --- a/plugins/dev9null/DEV9.cpp +++ b/plugins/dev9null/DEV9.cpp @@ -16,11 +16,12 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include -#include -#include +#include #include -#include "PS2Etypes.h" +#include +#include +using namespace std; + #include "DEV9.h" const unsigned char version = PS2E_DEV9_VERSION; @@ -28,9 +29,11 @@ const unsigned char revision = 0; const unsigned char build = 4; // increase that with each version const char *libraryName = "DEV9null Driver"; +string s_strIniPath="inis/DEV9null.ini"; -void (*DEV9irq)(); +void (*DEV9irq)(int); FILE *dev9Log; +Config conf; EXPORT_C_(u32) PS2EgetLibType() { @@ -47,10 +50,12 @@ EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) return (version<<16) | (revision<<8) | build; } -void __Log(char *fmt, ...) +void __Log(char *fmt, ...) { va_list list; + if (!conf.Log || dev9Log == NULL) return; + va_start(list, fmt); vfprintf(dev9Log, fmt, list); va_end(list); @@ -58,11 +63,23 @@ void __Log(char *fmt, ...) EXPORT_C_(s32) DEV9init() { + LoadConfig(); + +#ifdef DEV9_LOG + dev9Log = fopen("logs/dev9Log.txt", "w"); + if (dev9Log) setvbuf(dev9Log, NULL, _IONBF, 0); + DEV9_LOG("dev9null plugin version %d,%d\n", revision, build); + DEV9_LOG("DEV9init\n"); +#endif + return 0; } EXPORT_C_(void) DEV9shutdown() { +#ifdef DEV9_LOG + if (dev9Log) fclose(dev9Log); +#endif } EXPORT_C_(s32) DEV9open(void *pDsp) @@ -77,42 +94,50 @@ EXPORT_C_(void) DEV9close() EXPORT_C_(u8) DEV9read8(u32 addr) { + DEV9_LOG("*Unknown 8bit read at address %lx ", addr); return 0; } EXPORT_C_(u16 ) DEV9read16(u32 addr) { + DEV9_LOG("*Unknown 16bit read at address %lx ", addr); return 0; } EXPORT_C_(u32 ) DEV9read32(u32 addr) { + DEV9_LOG("*Unknown 32bit read at address %lx ", addr); return 0; } EXPORT_C_(void) DEV9write8(u32 addr, u8 value) { + DEV9_LOG("*Unknown 8bit write at address %lx value %x\n", addr, value); } EXPORT_C_(void) DEV9write16(u32 addr, u16 value) { + DEV9_LOG("*Unknown 16bit write at address %lx value %x\n", addr, value); } EXPORT_C_(void) DEV9write32(u32 addr, u32 value) { + DEV9_LOG("*Unknown 32bit write at address %lx value %x\n", addr, value); } EXPORT_C_(void) DEV9readDMA8Mem(u32 *pMem, int size) { + DEV9_LOG("Reading DMA8 Mem."); } EXPORT_C_(void) DEV9writeDMA8Mem(u32* pMem, int size) { + DEV9_LOG("Writing DMA8 Mem."); } EXPORT_C_(void) DEV9irqCallback(DEV9callback callback) { - + DEV9irq = callback; } EXPORT_C_(DEV9handler) DEV9irqHandler(void) @@ -127,6 +152,7 @@ EXPORT_C_(s32) DEV9test() return 0; } +#ifdef _WIN32 EXPORT_C_(void) DEV9configure() { SysMessage("Nothing to Configure"); @@ -136,8 +162,6 @@ EXPORT_C_(void) DEV9about() { } -#ifdef _WIN32 - HINSTANCE hInst; void SysMessage(const char *fmt, ...) diff --git a/plugins/dev9null/DEV9.h b/plugins/dev9null/DEV9.h index a827a0e019..a20dd450db 100644 --- a/plugins/dev9null/DEV9.h +++ b/plugins/dev9null/DEV9.h @@ -1,3 +1,21 @@ +/* dev9null + * Copyright (C) 2002-2009 pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + #ifndef __DEV9_H__ #define __DEV9_H__ @@ -17,6 +35,13 @@ extern "C" #include "PS2Edefs.h" } +typedef struct { + int Log; +} Config; + +extern Config conf; +#define DEV9_LOG __Log + /*#ifdef _MSC_VER #define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK #else @@ -35,7 +60,10 @@ extern const unsigned char build; extern const unsigned int minor; extern const char *libraryName; -extern void (*DEV9irq)(); +void SaveConfig(); +void LoadConfig(); + +extern void (*DEV9irq)(int); extern void SysMessage(const char *fmt, ...); extern FILE *dev9Log; diff --git a/plugins/dev9null/Linux/Config.cpp b/plugins/dev9null/Linux/Config.cpp index 89b686d779..fb99c4381d 100644 --- a/plugins/dev9null/Linux/Config.cpp +++ b/plugins/dev9null/Linux/Config.cpp @@ -16,15 +16,35 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + #include +#include +#include +#include +#include +#include +#include +#include +using namespace std; + +#include "Config.h" #include "DEV9.h" + #ifdef __cplusplus + +extern "C" +{ +#endif -#include -#include -#include -#include +#include "support.h" +#include "callbacks.h" +#include "interface.h" -GtkWidget *MsgDlg; +#ifdef __cplusplus +} +#endif + +GtkWidget *MsgDlg, *About, *Conf; +extern string s_strIniPath; void OnMsg_Ok() { @@ -74,6 +94,76 @@ void SysMessage(const char *fmt, ...) gtk_main(); } -void LoadConfig() +void OnAbout_Ok(GtkButton *button, gpointer user_data) { + gtk_widget_destroy(About); + gtk_main_quit(); } + +EXPORT_C_(void) DEV9about() +{ + About = create_About(); + gtk_widget_show_all(About); + gtk_main(); +} + +void OnConf_Ok(GtkButton *button, gpointer user_data) +{ + conf.Log = is_checked(Conf, "check_logging"); + SaveConfig(); + + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void OnConf_Cancel(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +EXPORT_C_(void) DEV9configure() +{ + LoadConfig(); + Conf = create_Config(); + + set_checked(Conf, "check_logging", conf.Log); + gtk_widget_show_all(Conf); + gtk_main(); +} + +void LoadConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "r"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + SaveConfig();//save and return + return; + } + fscanf(f, "logging = %hhx\n", &conf.Log); + //fscanf(f, "options = %hhx\n", &confOptions); + fclose(f); +} + +void SaveConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg,"w"); + if (f == NULL) + { + printf("failed to open '%s'\n", s_strIniPath.c_str()); + return; + } + fprintf(f, "logging = %hhx\n", conf.Log); + //fprintf(f, "options = %hhx\n", confOptions); + fclose(f); +} + diff --git a/plugins/dev9null/Linux/dev9.glade b/plugins/dev9null/Linux/dev9.glade index 10b8a7f4d1..204ecc3db2 100644 --- a/plugins/dev9null/Linux/dev9.glade +++ b/plugins/dev9null/Linux/dev9.glade @@ -295,6 +295,71 @@ + + + True + 0 + 0.5 + GTK_SHADOW_NONE + + + + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + True + Enable Logging + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + + + + + True + <b>Logging</b> + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + True diff --git a/plugins/dev9null/Linux/interface.c b/plugins/dev9null/Linux/interface.c index eef82eaf94..9625f99b79 100644 --- a/plugins/dev9null/Linux/interface.c +++ b/plugins/dev9null/Linux/interface.c @@ -45,6 +45,10 @@ create_Config (void) GList *GtkCombo_Hdd_items = NULL; GtkWidget *entry1; GtkWidget *label15; + GtkWidget *frame4; + GtkWidget *alignment1; + GtkWidget *check_logging; + GtkWidget *label16; GtkWidget *hbuttonbox1; GtkWidget *button1; GtkWidget *button2; @@ -132,6 +136,29 @@ create_Config (void) gtk_widget_show (label15); gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + frame4 = gtk_frame_new (NULL); + gtk_widget_set_name (frame4, "frame4"); + gtk_widget_show (frame4); + gtk_box_pack_start (GTK_BOX (vbox1), frame4, TRUE, TRUE, 0); + gtk_frame_set_shadow_type (GTK_FRAME (frame4), GTK_SHADOW_NONE); + + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame4), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); + + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); + + label16 = gtk_label_new (_("Logging")); + gtk_widget_set_name (label16, "label16"); + gtk_widget_show (label16); + gtk_frame_set_label_widget (GTK_FRAME (frame4), label16); + gtk_label_set_use_markup (GTK_LABEL (label16), TRUE); + hbuttonbox1 = gtk_hbutton_box_new (); gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); gtk_widget_show (hbuttonbox1); @@ -172,6 +199,10 @@ create_Config (void) GLADE_HOOKUP_OBJECT (Config, GtkCombo_Hdd, "GtkCombo_Hdd"); GLADE_HOOKUP_OBJECT (Config, entry1, "entry1"); GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, frame4, "frame4"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); + GLADE_HOOKUP_OBJECT (Config, label16, "label16"); GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); GLADE_HOOKUP_OBJECT (Config, button1, "button1"); GLADE_HOOKUP_OBJECT (Config, button2, "button2"); diff --git a/plugins/dev9null/Makefile.am b/plugins/dev9null/Makefile.am index 708b882df6..d8137623c7 100644 --- a/plugins/dev9null/Makefile.am +++ b/plugins/dev9null/Makefile.am @@ -22,6 +22,10 @@ libDEV9null_LDFLAGS= @SHARED_LDFLAGS@ libDEV9null_LDFLAGS+=-Wl,-soname,@libDEV9null_SONAME@ libDEV9null_LDADD=$(libDEV9null_a_OBJECTS) -libDEV9null_a_SOURCES = DEV9.cpp DEV9.h Linux/Config.cpp +libDEV9null_a_SOURCES = DEV9.cpp DEV9.h Linux/Config.cpp Linux/Config.h +libDEV9null_a_SOURCES += \ +Linux/interface.h Linux/support.c \ +Linux/interface.c Linux/support.h \ +Linux/callbacks.h #SUBDIRS = Linux \ No newline at end of file From a1507e60447f6d8fa57c2633275f050ab0f0c9a2 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 15 May 2009 07:48:51 +0000 Subject: [PATCH 068/296] GSdx: this should fix d3d9 mode and the aspect ratio bug git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1191 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSLocalMemory.cpp | 2 +- plugins/GSdx/GSTexture9.cpp | 5 ++++- plugins/GSdx/GSTextureCache.h | 3 ++- plugins/GSdx/GSVector.cpp | 2 +- 4 files changed, 8 insertions(+), 4 deletions(-) diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 5e1ab20d09..0d4605b566 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -28,7 +28,7 @@ #include "GSLocalMemory.h" #define ASSERT_BLOCK(r, w, h) \ - ASSERT((r).Width() >= w && (r).Height() >= h && !((r).left&(w-1)) && !((r).top&(h-1)) && !((r).right&(w-1)) && !((r).bottom&(h-1))); \ + ASSERT((r).width() >= w && (r).height() >= h && !((r).left&(w-1)) && !((r).top&(h-1)) && !((r).right&(w-1)) && !((r).bottom&(h-1))); \ #define FOREACH_BLOCK_START(w, h, bpp) \ uint32 bp = TEX0.TBP0; \ diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index bbb42e0420..e0eaa212b6 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -94,7 +94,10 @@ bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) uint8* src = (uint8*)data; uint8* dst = (uint8*)lr.pBits; - int bytes = min(pitch, lr.Pitch); + int bytes = r.width() << (m_desc.Format == D3DFMT_A1R5G5B5 ? 1 : 2); + + bytes = min(bytes, pitch); + bytes = min(bytes, lr.Pitch); for(int i = 0, j = r.height(); i < j; i++, src += pitch, dst += lr.Pitch) { diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index d409d54f30..282e0b18ca 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -224,7 +224,8 @@ public: m_valid = m_valid.runion(r); static uint8* bits = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - static int pitch = 1024 * 4; + + int pitch = ((r.width() + 3) & ~3) * 4; if(m_renderer->m_psrr) { diff --git a/plugins/GSdx/GSVector.cpp b/plugins/GSdx/GSVector.cpp index 206fa163b9..1256e96a6f 100644 --- a/plugins/GSdx/GSVector.cpp +++ b/plugins/GSdx/GSVector.cpp @@ -58,7 +58,7 @@ GSVector4i GSVector4i::fit(int arx, int ary) const if(w * ary > h * arx) { - int w2 = w * arx / ary; + int w2 = h * arx / ary; r.left = (r.left + r.right - w2) >> 1; if(r.left & 1) r.left++; r.right = r.left + w2; From 441aee367e2c0872050a9667d3f84351a8a33b52 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 15 May 2009 11:11:26 +0000 Subject: [PATCH 069/296] GSdx: fixed a few more typos in GSVector4i, the black screen bug should be gone git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1192 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPURenderer.h | 2 +- plugins/GSdx/GSBlock.h | 6 ++-- plugins/GSdx/GSDrawScanlineCodeGenerator.cpp | 2 +- plugins/GSdx/GSDrawScanlineCodeGenerator.h | 2 +- plugins/GSdx/GSFunctionMap.h | 22 ++++++------- plugins/GSdx/GSLocalMemory.cpp | 4 +-- plugins/GSdx/GSPerfMon.cpp | 2 -- plugins/GSdx/GSPerfMon.h | 6 ++-- plugins/GSdx/GSRasterizer.cpp | 4 +-- plugins/GSdx/GSRenderer.h | 2 +- plugins/GSdx/GSRendererHW.h | 4 +-- plugins/GSdx/GSRendererSW.h | 4 +-- plugins/GSdx/GSState.cpp | 2 +- plugins/GSdx/GSVector.cpp | 12 +++---- plugins/GSdx/GSVector.h | 34 ++++++++++---------- plugins/GSdx/stdafx.h | 7 +++- 16 files changed, 59 insertions(+), 56 deletions(-) diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index bd683a67c5..d6a091fec1 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -331,7 +331,7 @@ public: // osd - static UINT64 s_frame = 0; + static uint64 s_frame = 0; static string s_stats; if(m_perfmon.GetFrame() - s_frame >= 30) diff --git a/plugins/GSdx/GSBlock.h b/plugins/GSdx/GSBlock.h index 2ac27a0f12..7b548bf3bf 100644 --- a/plugins/GSdx/GSBlock.h +++ b/plugins/GSdx/GSBlock.h @@ -928,7 +928,7 @@ public: } } - __forceinline static void ExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -936,7 +936,7 @@ public: } } - __forceinline static void ExpandBlock4_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ExpandBlock4_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) { for(int j = 0; j < 16; j++, dst += dstpitch) { @@ -1478,7 +1478,7 @@ public: // TODO: ReadAndExpandBlock8_16 - __forceinline static void ReadAndExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const UINT64* RESTRICT pal) + __forceinline static void ReadAndExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) { #if _M_SSE >= 0x401 diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp index e87c75c4f8..4ba2fb75b0 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp @@ -25,7 +25,7 @@ #include "StdAfx.h" #include "GSDrawScanlineCodeGenerator.h" -GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(GSScanlineEnvironment& env, UINT64 key, void* ptr, size_t maxsize) +GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(GSScanlineEnvironment& env, uint64 key, void* ptr, size_t maxsize) : CodeGenerator(maxsize, ptr) , m_env(env) { diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.h b/plugins/GSdx/GSDrawScanlineCodeGenerator.h index 2803037200..a131f8c678 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.h +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.h @@ -73,5 +73,5 @@ class GSDrawScanlineCodeGenerator : public CodeGenerator void blendr(const Xmm& b, const Xmm& a, const Xmm& mask); public: - GSDrawScanlineCodeGenerator(GSScanlineEnvironment& env, UINT64 key, void* ptr, size_t maxsize); + GSDrawScanlineCodeGenerator(GSScanlineEnvironment& env, uint64 key, void* ptr, size_t maxsize); }; \ No newline at end of file diff --git a/plugins/GSdx/GSFunctionMap.h b/plugins/GSdx/GSFunctionMap.h index 2d7d052d8d..b0acb7daec 100644 --- a/plugins/GSdx/GSFunctionMap.h +++ b/plugins/GSdx/GSFunctionMap.h @@ -26,7 +26,7 @@ struct GSRasterizerStats { - __int64 ticks; + int64 ticks; int prims, pixels; GSRasterizerStats() @@ -46,8 +46,8 @@ template class GSFunctionMap protected: struct ActivePtr { - UINT64 frame, frames; - __int64 ticks, pixels; + uint64 frame, frames; + int64 ticks, pixels; VALUE f; }; @@ -90,7 +90,7 @@ public: memset(p, 0, sizeof(*p)); - p->frame = (UINT64)-1; + p->frame = (uint64)-1; p->f = i != m_map.end() ? (*i).second : GetDefaultFunction(key); @@ -102,7 +102,7 @@ public: return m_active->f; } - void UpdateStats(const GSRasterizerStats& stats, UINT64 frame) + void UpdateStats(const GSRasterizerStats& stats, uint64 frame) { if(m_active) { @@ -119,7 +119,7 @@ public: virtual void PrintStats() { - __int64 ttpf = 0; + int64 ttpf = 0; for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { @@ -138,12 +138,12 @@ public: if(p->frames > 0) { - __int64 tpp = p->pixels > 0 ? p->ticks / p->pixels : 0; - __int64 tpf = p->frames > 0 ? p->ticks / p->frames : 0; - __int64 ppf = p->frames > 0 ? p->pixels / p->frames : 0; + int64 tpp = p->pixels > 0 ? p->ticks / p->pixels : 0; + int64 tpf = p->frames > 0 ? p->ticks / p->frames : 0; + int64 ppf = p->frames > 0 ? p->pixels / p->frames : 0; printf("[%012I64x]%c %6.2f%% | %5.2f%% | f %4I64d | p %10I64d | tpp %4I64d | tpf %9I64d | ppf %7I64d\n", - (UINT64)key, m_map.find(key) == m_map.end() ? '*' : ' ', + (uint64)key, m_map.find(key) == m_map.end() ? '*' : ' ', (float)(tpf * 10000 / 50000000) / 100, (float)(tpf * 10000 / ttpf) / 100, p->frames, p->pixels, @@ -187,7 +187,7 @@ public: { CG* cg = NULL; - hash_map::iterator i = m_cgmap.find(key); + hash_map::iterator i = m_cgmap.find(key); if(i != m_cgmap.end()) { diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 0d4605b566..b9c499518d 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -1757,7 +1757,7 @@ void GSLocalMemory::ReadTextureBlock8(uint32 bp, uint8* dst, int dstpitch, const void GSLocalMemory::ReadTextureBlock4(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const UINT64* pal = m_clut; + const uint64* pal = m_clut; ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, pal); } @@ -1906,7 +1906,7 @@ void GSLocalMemory::ReadTexture8NP(const GSVector4i& r, uint8* dst, int dstpitch void GSLocalMemory::ReadTexture4NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - const UINT64* pal = m_clut; + const uint64* pal = m_clut; if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { diff --git a/plugins/GSdx/GSPerfMon.cpp b/plugins/GSdx/GSPerfMon.cpp index 3a5f2d297e..82c044c361 100644 --- a/plugins/GSdx/GSPerfMon.cpp +++ b/plugins/GSdx/GSPerfMon.cpp @@ -22,8 +22,6 @@ #include "stdafx.h" #include "GSPerfMon.h" -extern "C" unsigned __int64 __rdtsc(); - GSPerfMon::GSPerfMon() : m_total(0) , m_begin(0) diff --git a/plugins/GSdx/GSPerfMon.h b/plugins/GSdx/GSPerfMon.h index 42463551c5..e6d97e338c 100644 --- a/plugins/GSdx/GSPerfMon.h +++ b/plugins/GSdx/GSPerfMon.h @@ -29,7 +29,7 @@ public: protected: double m_counters[CounterLast]; double m_stats[CounterLast]; - UINT64 m_begin, m_total, m_start, m_frame; + uint64 m_begin, m_total, m_start, m_frame; clock_t m_lastframe; int m_count; @@ -41,8 +41,8 @@ protected: public: GSPerfMon(); - void SetFrame(UINT64 frame) {m_frame = frame;} - UINT64 GetFrame() {return m_frame;} + void SetFrame(uint64 frame) {m_frame = frame;} + uint64 GetFrame() {return m_frame;} void Put(counter_t c, double val = 0); double Get(counter_t c) {return m_stats[c];} void Update(); diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index b3d0a640a4..741c2dbeca 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -51,7 +51,7 @@ void GSRasterizer::Draw(const GSRasterizerData* data) m_stats.Reset(); - __int64 start = __rdtsc(); + int64 start = __rdtsc(); switch(data->primclass) { @@ -897,7 +897,7 @@ void GSRasterizerList::Draw(const GSRasterizerData* data) m_stats.Reset(); - __int64 start = __rdtsc(); + int64 start = __rdtsc(); for(list::reverse_iterator i = rbegin(); i != rend(); i++) { diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 923b1e2003..6ba6ce086b 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -454,7 +454,7 @@ public: // osd - static UINT64 s_frame = 0; + static uint64 s_frame = 0; static string s_stats; if(m_perfmon.GetFrame() - s_frame >= 30) diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index c19c5de96e..79cc37c9ad 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -321,7 +321,7 @@ protected: if(s_dump) { - UINT64 frame = m_perfmon.GetFrame(); + uint64 frame = m_perfmon.GetFrame(); string s; @@ -375,7 +375,7 @@ protected: if(s_dump) { - UINT64 frame = m_perfmon.GetFrame(); + uint64 frame = m_perfmon.GetFrame(); string s; diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 2b70069b02..8c3d50bc42 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -557,7 +557,7 @@ protected: if(s_dump) { - UINT64 frame = m_perfmon.GetFrame(); + uint64 frame = m_perfmon.GetFrame(); string s; @@ -630,7 +630,7 @@ protected: if(s_dump) { - UINT64 frame = m_perfmon.GetFrame(); + uint64 frame = m_perfmon.GetFrame(); string s; diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 93721c5236..20fd4124e1 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -564,7 +564,7 @@ template void GSState::GIFRegHandlerTEX2(GIFReg* r) { // m_env.CTXT[i].TEX2 = r->TEX2; // not used - UINT64 mask = 0xFFFFFFE003F00000ui64; // TEX2 bits + uint64 mask = 0xFFFFFFE003F00000ui64; // TEX2 bits r->u64 = (r->u64 & mask) | (m_env.CTXT[i].TEX0.u64 & ~mask); diff --git a/plugins/GSdx/GSVector.cpp b/plugins/GSdx/GSVector.cpp index 1256e96a6f..a1741d5e0a 100644 --- a/plugins/GSdx/GSVector.cpp +++ b/plugins/GSdx/GSVector.cpp @@ -58,17 +58,17 @@ GSVector4i GSVector4i::fit(int arx, int ary) const if(w * ary > h * arx) { - int w2 = h * arx / ary; - r.left = (r.left + r.right - w2) >> 1; + w = h * arx / ary; + r.left = (r.left + r.right - w) >> 1; if(r.left & 1) r.left++; - r.right = r.left + w2; + r.right = r.left + w; } else { - int h2 = w * ary / arx; - r.top = (r.top + r.bottom - h2) >> 1; + h = w * ary / arx; + r.top = (r.top + r.bottom - h) >> 1; if(r.top & 1) r.top++; - r.bottom = r.top + h2; + r.bottom = r.top + h; } r = r.rintersect(*this); diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 46a6194ff7..d3dd190e0f 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -168,7 +168,7 @@ public: #else - return GSVector4i(min(x, a.x), min(y, a.y), max(z, a.z), max(x, a.w)); + return GSVector4i(min(x, a.x), min(y, a.y), max(z, a.z), max(w, a.w)); #endif } @@ -251,7 +251,7 @@ public: v.x = min(max(x, a.x), b.x); v.y = min(max(y, a.y), b.y); v.z = min(max(z, a.z), b.z); - v.w = min(max(x, a.w), b.w); + v.w = min(max(w, a.w), b.w); return v; } @@ -263,7 +263,7 @@ public: v.x = min(max(x, a.x), a.z); v.y = min(max(y, a.y), a.w); v.z = min(max(z, a.x), a.z); - v.w = min(max(x, a.y), a.w); + v.w = min(max(w, a.y), a.w); return v; } @@ -1049,14 +1049,14 @@ public: #if _M_SSE >= 0x401 - template GSVector4i insert64(__int64 a) const + template GSVector4i insert64(int64 a) const { return GSVector4i(_mm_insert_epi64(m, a, i)); } #endif - template __int64 extract64() const + template int64 extract64() const { if(i == 0) return GSVector4i::storeq(*this); #if _M_SSE >= 0x401 @@ -1340,8 +1340,8 @@ public: { GSVector4i v; - v = loadq((__int64)ptr[extract8() & 0xf]); - v = v.insert64<1>((__int64)ptr[extract8() >> 4]); + v = loadq((int64)ptr[extract8() & 0xf]); + v = v.insert64<1>((int64)ptr[extract8() >> 4]); return v; } @@ -1350,8 +1350,8 @@ public: { GSVector4i v; - v = loadq((__int64)ptr[extract8()]); - v = v.insert64<1>((__int64)ptr[extract8()]); + v = loadq((int64)ptr[extract8()]); + v = v.insert64<1>((int64)ptr[extract8()]); return v; } @@ -1360,8 +1360,8 @@ public: { GSVector4i v; - v = loadq((__int64)ptr[extract16()]); - v = v.insert64<1>((__int64)ptr[extract16()]); + v = loadq((int64)ptr[extract16()]); + v = v.insert64<1>((int64)ptr[extract16()]); return v; } @@ -1370,8 +1370,8 @@ public: { GSVector4i v; - v = loadq((__int64)ptr[extract32()]); - v = v.insert64<1>((__int64)ptr[extract32()]); + v = loadq((int64)ptr[extract32()]); + v = v.insert64<1>((int64)ptr[extract32()]); return v; } @@ -1380,8 +1380,8 @@ public: { GSVector4i v; - v = loadq((__int64)ptr[extract64<0>()]); - v = v.insert64<1>((__int64)ptr[extract64<1>()]); + v = loadq((int64)ptr[extract64<0>()]); + v = v.insert64<1>((int64)ptr[extract64<1>()]); return v; } @@ -1595,7 +1595,7 @@ public: #ifdef _M_AMD64 - static GSVector4i loadq(__int64 i) + static GSVector4i loadq(int64 i) { return GSVector4i(_mm_cvtsi64_si128(i)); } @@ -1636,7 +1636,7 @@ public: #ifdef _M_AMD64 - static __int64 storeq(const GSVector4i& v) + static int64 storeq(const GSVector4i& v) { return _mm_cvtsi128_si64(v.m); } diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index 5d2b8eae29..adf3463f2d 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -80,6 +80,12 @@ typedef signed long long int64; #endif #endif +#ifdef __x86_64__ + #define _M_AMD64 +#endif + +extern "C" uint64 __rdtsc(); // TODO: gcc + // directx #include @@ -149,4 +155,3 @@ typedef signed long long int64; #include #endif - From fc7bc70cfc002477e62e118cc65118a1707883ff Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 15 May 2009 19:45:34 +0000 Subject: [PATCH 070/296] Linux/Plugins: I left this out at some point... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1193 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/dev9null/Linux/Config.h | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100644 plugins/dev9null/Linux/Config.h diff --git a/plugins/dev9null/Linux/Config.h b/plugins/dev9null/Linux/Config.h new file mode 100644 index 0000000000..a5de746902 --- /dev/null +++ b/plugins/dev9null/Linux/Config.h @@ -0,0 +1,24 @@ +/* dev9null + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +void SaveConfig(); +void LoadConfig(); +void SysMessage(char *fmt, ...); + +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) From b535039ee77b128f488ea132be53a45c7262f4f2 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 15 May 2009 21:07:58 +0000 Subject: [PATCH 071/296] GSdx: blur removal works again git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1194 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRenderer.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 6ba6ce086b..bb3fb7c34e 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -164,8 +164,8 @@ protected: if(samesrc && m_regs->PMODE.SLBG == 0 && m_regs->PMODE.MMOD == 1 && m_regs->PMODE.ALP == 0x80) { - if(fr[0].eq(fr[1] + GSVector4i(0, 1, 0, 0)) && dr[0].eq(dr[1] + GSVector4i(0, 0, 0, 1)) - || fr[1].eq(fr[0] + GSVector4i(0, 1, 0, 0)) && dr[1].eq(dr[0] + GSVector4i(0, 0, 0, 1))) + if(fr[0].eq(fr[1] + GSVector4i(0, -1, 0, 0)) && dr[0].eq(dr[1] + GSVector4i(0, 0, 0, 1)) + || fr[1].eq(fr[0] + GSVector4i(0, -1, 0, 0)) && dr[1].eq(dr[0] + GSVector4i(0, 0, 0, 1))) { // persona 4: // From 728eb75e7ae940ae29bc57ea2e14c0cc19931646 Mon Sep 17 00:00:00 2001 From: refraction Date: Fri, 15 May 2009 23:19:22 +0000 Subject: [PATCH 072/296] Half fix for Gran Turismo 3-4 menus. Some text is still garbled but better than it was. Oh this was done by implementing "Wait for GS transfers" a bit better than the previous attempt git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1197 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 60 +++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 52 insertions(+), 8 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 37c25ed4e3..f80de93ab0 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -813,7 +813,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i if((u32)(((size / ft->gsize) / vifRegs->cycle.cl) * vifRegs->cycle.wl) < vifRegs->num) DevCon::Notice("Filling write warning! %x < %x and CL = %x WL = %x", params (size / ft->gsize), vifRegs->num, vifRegs->cycle.cl, vifRegs->cycle.wl); - DevCon::Notice("filling write %d cl %d, wl %d mask %x mode %x unpacktype %x addr %x", params vifRegs->num, vifRegs->cycle.cl, vifRegs->cycle.wl, vifRegs->mask, vifRegs->mode, unpackType, vif->tag.addr); + //DevCon::Notice("filling write %d cl %d, wl %d mask %x mode %x unpacktype %x addr %x", params vifRegs->num, vifRegs->cycle.cl, vifRegs->cycle.wl, vifRegs->mask, vifRegs->mode, unpackType, vif->tag.addr); while (vifRegs->num > 0) { if (vif->cl == vifRegs->cycle.wl) @@ -2009,6 +2009,7 @@ static void Vif1CMDSTMod() // STMOD vif1.cmd &= ~0x7f; } + static void Vif1CMDMskPath3() // MSKPATH3 { @@ -2018,6 +2019,8 @@ static void Vif1CMDMskPath3() // MSKPATH3 #ifdef GSPATH3FIX if ((vif1Regs->code >> 15) & 0x1) { + psHu32(GIF_STAT) |= 0x2; + psHu32(GIF_MODE) |= 0x1; while ((gif->chcr & 0x100)) //Can be done 2 different ways, depends on the game/company { if (!path3hack && !Path3transfer && (gif->qwc == 0)) break; @@ -2025,14 +2028,16 @@ static void Vif1CMDMskPath3() // MSKPATH3 gsInterrupt(); if (path3hack && (gif->qwc == 0)) break; //add games not working with it to elfheader.c to enable this instead + } - psHu32(GIF_STAT) |= 0x2; + } else { // fixme: This is the *only* reason 'transferred' is global. Otherwise it'd be local to Vif1Transfer. if (gif->chcr & 0x100) CPU_INT(2, (transferred >> 2) * BIAS); // Restart Path3 on its own, time it right! psHu32(GIF_STAT) &= ~0x2; + psHu32(GIF_MODE) &= ~0x1; } #else if (vif1Regs->mskpath3) @@ -2055,17 +2060,22 @@ static void Vif1CMDMark() // MARK vif1Regs->stat |= VIF1_STAT_MRK; vif1.cmd &= ~0x7f; } + + + static void Vif1CMDFlush() // FLUSH/E/A { vif1FLUSH(); - if ((vif1.cmd & 0x7f) == 0x13) + if((vif1.cmd & 0x7f) == 0x13) { while ((gif->chcr & 0x100)) { if (!Path3transfer && gif->qwc == 0) break; - gsInterrupt(); + vif1Regs->stat |= 0x4; + break; + } } @@ -2073,6 +2083,7 @@ static void Vif1CMDFlush() // FLUSH/E/A } static void Vif1CMDMSCALF() //MSCAL/F { + vif1FLUSH(); vuExecMicro((u16)(vif1Regs->code) << 3, VIF1dmanum); vif1.cmd &= ~0x7f; } @@ -2102,6 +2113,9 @@ static void Vif1CMDMPGTransfer() // MPG vif1.tag.addr = (u16)((vif1Regs->code) << 3) & 0x3fff; vif1.tag.size = vifNum * 2; } + + + static void Vif1CMDDirectHL() // DIRECT/HL { int vifImm; @@ -2112,10 +2126,15 @@ static void Vif1CMDDirectHL() // DIRECT/HL else vif1.tag.size = vifImm << 2; - while ((gif->chcr & 0x100) && (vif1.cmd & 0x7f) == 0x51) + + if((gif->chcr & 0x100) && (vif1.cmd & 0x7f) == 0x51) { - gsInterrupt(); //DirectHL flushes the lot + //if(vif1Regs->mskpath3)CPU_INT(2, vif1ch->qwc - (vif1.vifpacketsize >> 2) * BIAS); + //DirectHL flushes the lot + vif1Regs->stat |= 0x4; + } + } static void Vif1CMDNull() // invalid opcode { @@ -2200,6 +2219,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size != 0) DevCon::Error("no vif1 cmd but tag size is left last cmd read %x", params vif1Regs->code); if (vif1.irq) break; + if(vif1Regs->stat & 0x4) break; vif1.cmd = (data[0] >> 24); vif1Regs->code = data[0]; @@ -2244,7 +2264,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size == 0) break; } } - + if(vif1Regs->stat & 0x4) break; } // End of Transfer loop transferred += size - vif1.vifpacketsize; @@ -2282,6 +2302,13 @@ int VIF1transfer(u32 *data, int size, int istag) vif1ch->qwc -= transferred; } + if(vif1Regs->stat & 0x4) + { + vif1.vifstalled = true; + return -1; + } + + if (vif1ch->qwc == 0 && (vif1.irqoffset == 0 || istag == 1)) vif1.inprogress = 0; @@ -2462,6 +2489,23 @@ __forceinline void vif1Interrupt() VIF_LOG("vif1Interrupt: %8.8x", cpuRegs.cycle); g_vifCycles = 0; + if((vif1Regs->stat & 0x4) && + ((psHu32(GIF_MODE) & 0x1) ? (Path3transfer == true || gif->qwc > 0) : (gif->chcr & 0x100))) + { + int delay = 0; + + if ((psHu32(GIF_MODE) & 0x1)) + delay = min( 8, (int)gif->qwc ); + else + delay = gif->qwc * BIAS; + + if((psHu32(GIF_MODE) & 0x1)) gsInterrupt(); + //else CPU_INT(2, min( 64, (int)gif->qwc ) * BIAS); + CPU_INT(1, delay); + return; + } + vif1Regs->stat &= ~0x4; + if ((vif1ch->chcr & 0x100) == 0) Console::WriteLn("Vif1 running when CHCR == %x", params vif1ch->chcr); @@ -2500,7 +2544,7 @@ __forceinline void vif1Interrupt() if ((vif1.inprogress & 0x1) == 0) vif1SetupTransfer(); - CPU_INT(1, vif1ch->qwc * BIAS); + CPU_INT(1, (vif1ch->qwc * BIAS) - (vif1.vifpacketsize >> 1)); return; } From e93b101f49ff75198a18ba069a2ab95d05195643 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 16 May 2009 00:43:46 +0000 Subject: [PATCH 073/296] Cleanup: Moved some rather intrusive instances of global namespace pollution in Sio.h into a new sio_internal.h, so that the rest of the world doesn't have to ensure the pain of having BREAK and RTS macro'd to integers. ;) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1198 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IopSio2.cpp | 2 +- pcsx2/Sio.cpp | 1 + pcsx2/Sio.h | 61 +-------------------- pcsx2/sio_internal.h | 62 ++++++++++++++++++++++ pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 ++ plugins/dev9null/DEV9.cpp | 2 + 6 files changed, 71 insertions(+), 61 deletions(-) create mode 100644 pcsx2/sio_internal.h diff --git a/pcsx2/IopSio2.cpp b/pcsx2/IopSio2.cpp index d3ffd8271e..369591f265 100644 --- a/pcsx2/IopSio2.cpp +++ b/pcsx2/IopSio2.cpp @@ -17,8 +17,8 @@ */ #include "PrecompiledHeader.h" - #include "IopCommon.h" +#include "sio_internal.h" sio2Struct sio2; diff --git a/pcsx2/Sio.cpp b/pcsx2/Sio.cpp index 52687d7698..d944ebd299 100644 --- a/pcsx2/Sio.cpp +++ b/pcsx2/Sio.cpp @@ -20,6 +20,7 @@ #include "IopCommon.h" #include "MemoryCard.h" +#include "sio_internal.h" _sio sio; diff --git a/pcsx2/Sio.h b/pcsx2/Sio.h index c3ec7c74bb..541f583143 100644 --- a/pcsx2/Sio.h +++ b/pcsx2/Sio.h @@ -16,27 +16,7 @@ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ - -#ifndef _SIO_H_ -#define _SIO_H_ - -// SIO IRQ Timings... -// Scheduling ints into the future is a purist approach to emulation, and -// is mostly cosmetic since the emulator itself performs all actions instantly -// (as far as the emulated CPU is concerned). In some cases this can actually -// cause more sync problems than it supposedly solves, due to accumulated -// delays incurred by the recompiler's low cycle update rate and also Pcsx2 -// failing to properly handle pre-emptive DMA/IRQs or cpu exceptions. - -// The SIO is one of these cases, where-by many games seem to be a lot happier -// if the SIO handles its IRQs instantly instead of scheduling them. -// Uncomment the line below for SIO instant-IRQ mode. It improves responsiveness -// considerably, fixes PAD latency problems in some games, and may even reduce the -// chance of saves getting corrupted (untested). But it lacks the purist touch, -// so it's not enabled by default. - -//#define SIO_INLINE_IRQS - +#pragma once struct _sio { u16 StatReg; @@ -79,27 +59,6 @@ struct _sio { extern _sio sio; -// Status Flags -#define TX_RDY 0x0001 -#define RX_RDY 0x0002 -#define TX_EMPTY 0x0004 -#define PARITY_ERR 0x0008 -#define RX_OVERRUN 0x0010 -#define FRAMING_ERR 0x0020 -#define SYNC_DETECT 0x0040 -#define DSR 0x0080 -#define CTS 0x0100 -#define IRQ 0x0200 - -// Control Flags -#define TX_PERM 0x0001 -#define DTR 0x0002 -#define RX_PERM 0x0004 -#define BREAK 0x0008 -#define RESET_ERR 0x0010 -#define RTS 0x0020 -#define SIO_RESET 0x0040 - extern void sioInit(); extern void sioShutdown(); extern void psxSIOShutdown(); @@ -110,21 +69,3 @@ extern void sioInterrupt(); extern void InitializeSIO(u8 value); extern void sioEjectCard( uint mcdId ); -#ifdef _MSC_VER -#pragma pack(1) -#endif -struct mc_command_0x26_tag{ - u8 field_151; //+02 flags - u16 sectorSize; //+03 divide to it - u16 field_2C; //+05 divide to it - u32 mc_size; //+07 - u8 mc_xor; //+0b don't forget to recalculate it!!! - u8 Z; //+0c -#ifdef _MSC_VER -}; -#pragma pack() -#else -} __attribute__((packed)); -#endif - -#endif diff --git a/pcsx2/sio_internal.h b/pcsx2/sio_internal.h new file mode 100644 index 0000000000..3a10d2c232 --- /dev/null +++ b/pcsx2/sio_internal.h @@ -0,0 +1,62 @@ +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +#pragma once + +// sio_internal.h -- contains defines and structs used by sio and sio2, which +// are of little or no use to the rest of the world. + +// Status Flags +static const int + TX_RDY = 0x0001, + RX_RDY = 0x0002, + TX_EMPTY = 0x0004, + PARITY_ERR = 0x0008, + RX_OVERRUN = 0x0010, + FRAMING_ERR = 0x0020, + SYNC_DETECT = 0x0040, + DSR = 0x0080, + CTS = 0x0100, + IRQ = 0x0200; + +// Control Flags +static const int + TX_PERM = 0x0001, + DTR = 0x0002, + RX_PERM = 0x0004, + BREAK = 0x0008, + RESET_ERR = 0x0010, + RTS = 0x0020, + SIO_RESET = 0x0040; + +#ifdef _MSC_VER +#pragma pack(1) +#endif +struct mc_command_0x26_tag{ + u8 field_151; //+02 flags + u16 sectorSize; //+03 divide to it + u16 field_2C; //+05 divide to it + u32 mc_size; //+07 + u8 mc_xor; //+0b don't forget to recalculate it!!! + u8 Z; //+0c +#ifdef _MSC_VER +}; +#pragma pack() +#else +} __attribute__((packed)); +#endif diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 589cc0a396..fd4f9ce1d4 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -1524,6 +1524,10 @@ RelativePath="..\..\Sio.h" > + + diff --git a/plugins/dev9null/DEV9.cpp b/plugins/dev9null/DEV9.cpp index 7e8ce7ecb9..03971aae0d 100644 --- a/plugins/dev9null/DEV9.cpp +++ b/plugins/dev9null/DEV9.cpp @@ -63,7 +63,9 @@ void __Log(char *fmt, ...) EXPORT_C_(s32) DEV9init() { +#ifdef __LINUX__ // for until we get a win32 version sorted out / implemented... LoadConfig(); +#endif #ifdef DEV9_LOG dev9Log = fopen("logs/dev9Log.txt", "w"); From d676d2ee44e38dbf817da1de8b4dbba85a1ad941 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 16 May 2009 01:32:49 +0000 Subject: [PATCH 074/296] Break out a bunch of clutter from CDVD.h into its own file. Meant to do this a while back, but forgot about it. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1199 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/CDVD.cpp | 1 + pcsx2/CDVD.h | 234 ++------------------- pcsx2/CDVD_internal.h | 233 ++++++++++++++++++++ pcsx2/Makefile.am | 2 +- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 + 5 files changed, 251 insertions(+), 223 deletions(-) create mode 100644 pcsx2/CDVD_internal.h diff --git a/pcsx2/CDVD.cpp b/pcsx2/CDVD.cpp index 79e6a4b30f..e93117da87 100644 --- a/pcsx2/CDVD.cpp +++ b/pcsx2/CDVD.cpp @@ -23,6 +23,7 @@ #include "IopCommon.h" #include "CDVDiso.h" +#include "CDVD_internal.h" static cdvdStruct cdvd; diff --git a/pcsx2/CDVD.h b/pcsx2/CDVD.h index 3aaa5647a1..ca2917e7ee 100644 --- a/pcsx2/CDVD.h +++ b/pcsx2/CDVD.h @@ -21,6 +21,9 @@ #include "IopCommon.h" +#define btoi(b) ((b)/16*10 + (b)%16) /* BCD to u_char */ +#define itob(i) ((i)/10*16 + (i)%10) /* u_char to BCD */ + struct cdvdRTC { u8 status; u8 second; @@ -89,230 +92,17 @@ struct cdvdStruct { bool Spinning; // indicates if the Cdvd is spinning or needs a spinup delay }; -/* -Interrupts - values are flag bits. +extern void cdvdReset(); +extern void cdvdVsync(); +extern void cdvdActionInterrupt(); +extern void cdvdReadInterrupt(); -0x00 No interrupt -0x01 Data Ready -0x02 Command Complete -0x03 Acknowledge (reserved) -0x04 End of Data Detected -0x05 Error Detected -0x06 Drive Not Ready - -In limited experimentation I found that PS2 apps respond actively to use of the -'Data Ready' flag -- in that they'll almost immediately initiate a DMA transfer -after receiving an Irq with that as the cause. But the question is, of course, -*when* to use it. Adding it into some locations of CDVD reading only slowed -games down and broke things. - -Using Drive Not Ready also invokes basic error handling from the Iop Bios, but -without proper emulation of the cdvd status flag it also tends to break things. - -*/ - -enum CdvdIrqId -{ - Irq_None = 0 -, Irq_DataReady = 0 -, Irq_CommandComplete -, Irq_Acknowledge -, Irq_EndOfData -, Irq_Error -, Irq_NotReady - -}; - -/* is cdvd.Status only for NCMDS? (linuzappz) */ -enum cdvdStatus -{ - CDVD_STATUS_NONE = 0x00, // not sure ;) - CDVD_STATUS_SEEK_COMPLETE = 0x0A, -}; - -enum cdvdready -{ - CDVD_NOTREADY = 0x00, - CDVD_READY1 = 0x40, - CDVD_READY2 = 0x4e // This is used in a few places for some reason. - //It would be worth checking if this was just a typo made at some point. -}; - -// Cdvd actions tell the emulator how and when to respond to certain requests. -// Actions are handled by the cdvdInterrupt() -enum cdvdActions -{ - cdvdAction_None = 0 -, cdvdAction_Seek -, cdvdAction_Standby -, cdvdAction_Stop -, cdvdAction_Break -, cdvdAction_Read // note: not used yet. -}; - -////////////////////////////////////////////////////////////////////////////////////////// -// Cdvd Block Read Cycle Timings -// -// The PS2 CDVD effectively has two seek modes -- the normal/slow one (est. avg seeks being -// around 120-160ms), and a faster seek which has an estimated seek time of about 35-40ms. -// Fast seeks happen when the destination sector is within a certain range of the starting -// point, such that abs(start-dest) is less than the value in the tbl_FastSeekDelta. -// -// CDVDs also have a secondary seeking method used when the destination is close enough -// that a contiguous sector read can reach the sector faster than initiating a full seek. -// Typically this value is very low. - -enum CDVD_MODE_TYPE -{ - MODE_CDROM = 0, - MODE_DVDROM, -}; - -static const uint tbl_FastSeekDelta[3] = -{ - 4371, // CD-ROM - 14764, // Single-layer DVD-ROM - 13360 // dual-layer DVD-ROM [currently unused] -}; - -// if a seek is within this many blocks, read instead of seek. -// These values are arbitrary assumptions. Not sure what the real PS2 uses. -static const uint tbl_ContigiousSeekDelta[3] = -{ - 8, // CD-ROM - 16, // single-layer DVD-ROM - 16, // dual-layer DVD-ROM [currently unused] -}; - -// Note: DVD read times are modified to be faster, because games seem to be a lot more -// concerned with accurate(ish) seek delays and less concerned with actual block read speeds. -// Translation: it's a minor speedhack :D - -static const uint PSX_CD_READSPEED = 153600; // 1 Byte Time @ x1 (150KB = cd x 1) -static const uint PSX_DVD_READSPEED = 1382400 + 256000; // normal is 1 Byte Time @ x1 (1350KB = dvd x 1). - - -// Legacy Note: FullSeek timing causes many games to load very slow, but it likely not the real problem. -// Games breaking with it set to PSXCLK*40 : "wrath unleashed" and "Shijou Saikyou no Deshi Kenichi". - -static const uint Cdvd_FullSeek_Cycles = (PSXCLK*100) / 1000; // average number of cycles per fullseek (100ms) -static const uint Cdvd_FastSeek_Cycles = (PSXCLK*30) / 1000; // average number of cycles per fastseek (37ms) - -static const __unused char *mg_zones[8] = {"Japan", "USA", "Europe", "Oceania", "Asia", "Russia", "China", "Mexico"}; - -static const __unused char *nCmdName[0x100]= { - "CdSync", - "CdNop", - "CdStandby", - "CdStop", - "CdPause", - "CdSeek", - "CdRead", - "CdReadCDDA", - "CdReadDVDV", - "CdGetToc", - "", - "NCMD_B", - "CdReadKey", - "", - "sceCdReadXCDDA", - "sceCdChgSpdlCtrl", -}; - -enum nCmds -{ - N_CD_SYNC = 0x00, // CdSync - N_CD_NOP = 0x01, // CdNop - N_CD_STANDBY = 0x02, // CdStandby - N_CD_STOP = 0x03, // CdStop - N_CD_PAUSE = 0x04, // CdPause - N_CD_SEEK = 0x05, // CdSeek - N_CD_READ = 0x06, // CdRead - N_CD_READ_CDDA = 0x07, // CdReadCDDA - N_DVD_READ = 0x08, // DvdRead - N_CD_GET_TOC = 0x09, // CdGetToc & cdvdman_call19 - N_CMD_B = 0x0B, // CdReadKey - N_CD_READ_KEY = 0x0C, // CdReadKey - N_CD_READ_XCDDA = 0x0E, // CdReadXCDDA - N_CD_CHG_SPDL_CTRL = 0x0F, // CdChgSpdlCtrl -}; - -static const __unused char *sCmdName[0x100]= { - "", "sceCdGetDiscType", "sceCdReadSubQ", "subcommands",//sceCdGetMecaconVersion, read/write console id, read renewal date - "", "sceCdTrayState", "sceCdTrayCtrl", "", - "sceCdReadClock", "sceCdWriteClock", "sceCdReadNVM", "sceCdWriteNVM", - "sceCdSetHDMode", "", "", "sceCdPowerOff", - "", "", "sceCdReadILinkID", "sceCdWriteILinkID", /*10*/ - "sceAudioDigitalOut", "sceForbidDVDP", "sceAutoAdjustCtrl", "sceCdReadModelNumber", - "sceWriteModelNumber", "sceCdForbidCD", "sceCdBootCertify", "sceCdCancelPOffRdy", - "sceCdBlueLEDCtl", "", "sceRm2Read", "sceRemote2_7",//Rm2PortGetConnection? - "sceRemote2_6", "sceCdWriteWakeUpTime", "sceCdReadWakeUpTime", "", /*20*/ - "sceCdRcBypassCtl", "", "", "", - "", "sceCdNoticeGameStart", "", "", - "sceCdXBSPowerCtl", "sceCdXLEDCtl", "sceCdBuzzerCtl", "", - "", "sceCdSetMediumRemoval", "sceCdGetMediumRemoval", "sceCdXDVRPReset", /*30*/ - "", "", "__sceCdReadRegionParams", "__sceCdReadMAC", - "__sceCdWriteMAC", "", "", "", - "", "", "__sceCdWriteRegionParams", "", - "sceCdOpenConfig", "sceCdReadConfig", "sceCdWriteConfig", "sceCdCloseConfig", /*40*/ - "", "", "", "", - "", "", "", "", - "", "", "", "", - "", "", "", "", /*50*/ - "", "", "", "", - "", "", "", "", - "", "", "", "", - "", "", "", "", /*60*/ - "", "", "", "", - "", "", "", "", - "", "", "", "", - "", "", "", "", /*70*/ - "", "", "", "", - "", "", "", "", - "", "", "", "", - "mechacon_auth_0x80", "mechacon_auth_0x81", "mechacon_auth_0x82", "mechacon_auth_0x83", /*80*/ - "mechacon_auth_0x84", "mechacon_auth_0x85", "mechacon_auth_0x86", "mechacon_auth_0x87", - "mechacon_auth_0x88", "", "", "", - "", "sceMgWriteData", "sceMgReadData", "mechacon_auth_0x8F", - "sceMgWriteHeaderStart", "sceMgReadBITLength", "sceMgWriteDatainLength", "sceMgWriteDataoutLength", /*90*/ - "sceMgReadKbit", "sceMgReadKbit2", "sceMgReadKcon", "sceMgReadKcon2", - "sceMgReadIcvPs2", "", "", "", - "", "", "", "", - /*A0, no sCmds above?*/ -}; - -// NVM (eeprom) layout info -struct NVMLayout { - u32 biosVer; // bios version that this eeprom layout is for - s32 config0; // offset of 1st config block - s32 config1; // offset of 2nd config block - s32 config2; // offset of 3rd config block - s32 consoleId; // offset of console id (?) - s32 ilinkId; // offset of ilink id (ilink mac address) - s32 modelNum; // offset of ps2 model number (eg "SCPH-70002") - s32 regparams; // offset of RegionParams for PStwo - s32 mac; // offset of the value written to 0xFFFE0188 and 0xFFFE018C on PStwo -}; - -#define NVM_FORMAT_MAX 2 -static NVMLayout nvmlayouts[NVM_FORMAT_MAX] = -{ - {0x000, 0x280, 0x300, 0x200, 0x1C8, 0x1C0, 0x1A0, 0x180, 0x198}, // eeproms from bios v0.00 and up - {0x146, 0x270, 0x2B0, 0x200, 0x1C8, 0x1E0, 0x1B0, 0x180, 0x198}, // eeproms from bios v1.70 and up -}; - -#define btoi(b) ((b)/16*10 + (b)%16) /* BCD to u_char */ -#define itob(i) ((i)/10*16 + (i)%10) /* u_char to BCD */ - -void cdvdReset(); -void cdvdVsync(); -extern void cdvdActionInterrupt(); -extern void cdvdReadInterrupt(); -void cdvdNewDiskCB(); -u8 cdvdRead(u8 key); -void cdvdWrite(u8 key, u8 rt); +// We really should not have a function with the exact same name as a callback except for case! +extern void cdvdNewDiskCB(); +extern u8 cdvdRead(u8 key); +extern void cdvdWrite(u8 key, u8 rt); // Platform dependent system time assignment (see WinMisc / LnxMisc) -extern void cdvdSetSystemTime( cdvdStruct& setme ); +extern void cdvdSetSystemTime(cdvdStruct& setme); #endif /* __CDVD_H__ */ diff --git a/pcsx2/CDVD_internal.h b/pcsx2/CDVD_internal.h new file mode 100644 index 0000000000..c51c46e309 --- /dev/null +++ b/pcsx2/CDVD_internal.h @@ -0,0 +1,233 @@ +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +#ifndef __CDVD_INTERNAL_H__ +#define __CDVD_INTERNAL_H__ + +/* +Interrupts - values are flag bits. + +0x00 No interrupt +0x01 Data Ready +0x02 Command Complete +0x03 Acknowledge (reserved) +0x04 End of Data Detected +0x05 Error Detected +0x06 Drive Not Ready + +In limited experimentation I found that PS2 apps respond actively to use of the +'Data Ready' flag -- in that they'll almost immediately initiate a DMA transfer +after receiving an Irq with that as the cause. But the question is, of course, +*when* to use it. Adding it into some locations of CDVD reading only slowed +games down and broke things. + +Using Drive Not Ready also invokes basic error handling from the Iop Bios, but +without proper emulation of the cdvd status flag it also tends to break things. + +*/ + +enum CdvdIrqId +{ + Irq_None = 0 +, Irq_DataReady = 0 +, Irq_CommandComplete +, Irq_Acknowledge +, Irq_EndOfData +, Irq_Error +, Irq_NotReady + +}; + +/* is cdvd.Status only for NCMDS? (linuzappz) */ +enum cdvdStatus +{ + CDVD_STATUS_NONE = 0x00, // not sure ;) + CDVD_STATUS_SEEK_COMPLETE = 0x0A, +}; + +enum cdvdready +{ + CDVD_NOTREADY = 0x00, + CDVD_READY1 = 0x40, + CDVD_READY2 = 0x4e // This is used in a few places for some reason. + //It would be worth checking if this was just a typo made at some point. +}; + +// Cdvd actions tell the emulator how and when to respond to certain requests. +// Actions are handled by the cdvdInterrupt() +enum cdvdActions +{ + cdvdAction_None = 0 +, cdvdAction_Seek +, cdvdAction_Standby +, cdvdAction_Stop +, cdvdAction_Break +, cdvdAction_Read // note: not used yet. +}; + +////////////////////////////////////////////////////////////////////////////////////////// +// Cdvd Block Read Cycle Timings +// +// The PS2 CDVD effectively has two seek modes -- the normal/slow one (est. avg seeks being +// around 120-160ms), and a faster seek which has an estimated seek time of about 35-40ms. +// Fast seeks happen when the destination sector is within a certain range of the starting +// point, such that abs(start-dest) is less than the value in the tbl_FastSeekDelta. +// +// CDVDs also have a secondary seeking method used when the destination is close enough +// that a contiguous sector read can reach the sector faster than initiating a full seek. +// Typically this value is very low. + +enum CDVD_MODE_TYPE +{ + MODE_CDROM = 0, + MODE_DVDROM, +}; + +static const uint tbl_FastSeekDelta[3] = +{ + 4371, // CD-ROM + 14764, // Single-layer DVD-ROM + 13360 // dual-layer DVD-ROM [currently unused] +}; + +// if a seek is within this many blocks, read instead of seek. +// These values are arbitrary assumptions. Not sure what the real PS2 uses. +static const uint tbl_ContigiousSeekDelta[3] = +{ + 8, // CD-ROM + 16, // single-layer DVD-ROM + 16, // dual-layer DVD-ROM [currently unused] +}; + +// Note: DVD read times are modified to be faster, because games seem to be a lot more +// concerned with accurate(ish) seek delays and less concerned with actual block read speeds. +// Translation: it's a minor speedhack :D + +static const uint PSX_CD_READSPEED = 153600; // 1 Byte Time @ x1 (150KB = cd x 1) +static const uint PSX_DVD_READSPEED = 1382400 + 256000; // normal is 1 Byte Time @ x1 (1350KB = dvd x 1). + +// Legacy Note: FullSeek timing causes many games to load very slow, but it likely not the real problem. +// Games breaking with it set to PSXCLK*40 : "wrath unleashed" and "Shijou Saikyou no Deshi Kenichi". + +static const uint Cdvd_FullSeek_Cycles = (PSXCLK*100) / 1000; // average number of cycles per fullseek (100ms) +static const uint Cdvd_FastSeek_Cycles = (PSXCLK*30) / 1000; // average number of cycles per fastseek (37ms) + +static const __unused char *mg_zones[8] = {"Japan", "USA", "Europe", "Oceania", "Asia", "Russia", "China", "Mexico"}; + +static const __unused char *nCmdName[0x100]= { + "CdSync", + "CdNop", + "CdStandby", + "CdStop", + "CdPause", + "CdSeek", + "CdRead", + "CdReadCDDA", + "CdReadDVDV", + "CdGetToc", + "", + "NCMD_B", + "CdReadKey", + "", + "sceCdReadXCDDA", + "sceCdChgSpdlCtrl", +}; + +enum nCmds +{ + N_CD_SYNC = 0x00, // CdSync + N_CD_NOP = 0x01, // CdNop + N_CD_STANDBY = 0x02, // CdStandby + N_CD_STOP = 0x03, // CdStop + N_CD_PAUSE = 0x04, // CdPause + N_CD_SEEK = 0x05, // CdSeek + N_CD_READ = 0x06, // CdRead + N_CD_READ_CDDA = 0x07, // CdReadCDDA + N_DVD_READ = 0x08, // DvdRead + N_CD_GET_TOC = 0x09, // CdGetToc & cdvdman_call19 + N_CMD_B = 0x0B, // CdReadKey + N_CD_READ_KEY = 0x0C, // CdReadKey + N_CD_READ_XCDDA = 0x0E, // CdReadXCDDA + N_CD_CHG_SPDL_CTRL = 0x0F, // CdChgSpdlCtrl +}; + +static const __unused char *sCmdName[0x100]= { + "", "sceCdGetDiscType", "sceCdReadSubQ", "subcommands",//sceCdGetMecaconVersion, read/write console id, read renewal date + "", "sceCdTrayState", "sceCdTrayCtrl", "", + "sceCdReadClock", "sceCdWriteClock", "sceCdReadNVM", "sceCdWriteNVM", + "sceCdSetHDMode", "", "", "sceCdPowerOff", + "", "", "sceCdReadILinkID", "sceCdWriteILinkID", /*10*/ + "sceAudioDigitalOut", "sceForbidDVDP", "sceAutoAdjustCtrl", "sceCdReadModelNumber", + "sceWriteModelNumber", "sceCdForbidCD", "sceCdBootCertify", "sceCdCancelPOffRdy", + "sceCdBlueLEDCtl", "", "sceRm2Read", "sceRemote2_7",//Rm2PortGetConnection? + "sceRemote2_6", "sceCdWriteWakeUpTime", "sceCdReadWakeUpTime", "", /*20*/ + "sceCdRcBypassCtl", "", "", "", + "", "sceCdNoticeGameStart", "", "", + "sceCdXBSPowerCtl", "sceCdXLEDCtl", "sceCdBuzzerCtl", "", + "", "sceCdSetMediumRemoval", "sceCdGetMediumRemoval", "sceCdXDVRPReset", /*30*/ + "", "", "__sceCdReadRegionParams", "__sceCdReadMAC", + "__sceCdWriteMAC", "", "", "", + "", "", "__sceCdWriteRegionParams", "", + "sceCdOpenConfig", "sceCdReadConfig", "sceCdWriteConfig", "sceCdCloseConfig", /*40*/ + "", "", "", "", + "", "", "", "", + "", "", "", "", + "", "", "", "", /*50*/ + "", "", "", "", + "", "", "", "", + "", "", "", "", + "", "", "", "", /*60*/ + "", "", "", "", + "", "", "", "", + "", "", "", "", + "", "", "", "", /*70*/ + "", "", "", "", + "", "", "", "", + "", "", "", "", + "mechacon_auth_0x80", "mechacon_auth_0x81", "mechacon_auth_0x82", "mechacon_auth_0x83", /*80*/ + "mechacon_auth_0x84", "mechacon_auth_0x85", "mechacon_auth_0x86", "mechacon_auth_0x87", + "mechacon_auth_0x88", "", "", "", + "", "sceMgWriteData", "sceMgReadData", "mechacon_auth_0x8F", + "sceMgWriteHeaderStart", "sceMgReadBITLength", "sceMgWriteDatainLength", "sceMgWriteDataoutLength", /*90*/ + "sceMgReadKbit", "sceMgReadKbit2", "sceMgReadKcon", "sceMgReadKcon2", + "sceMgReadIcvPs2", "", "", "", + "", "", "", "", + /*A0, no sCmds above?*/ +}; + +// NVM (eeprom) layout info +struct NVMLayout { + u32 biosVer; // bios version that this eeprom layout is for + s32 config0; // offset of 1st config block + s32 config1; // offset of 2nd config block + s32 config2; // offset of 3rd config block + s32 consoleId; // offset of console id (?) + s32 ilinkId; // offset of ilink id (ilink mac address) + s32 modelNum; // offset of ps2 model number (eg "SCPH-70002") + s32 regparams; // offset of RegionParams for PStwo + s32 mac; // offset of the value written to 0xFFFE0188 and 0xFFFE018C on PStwo +}; + +#define NVM_FORMAT_MAX 2 +static NVMLayout nvmlayouts[NVM_FORMAT_MAX] = +{ + {0x000, 0x280, 0x300, 0x200, 0x1C8, 0x1C0, 0x1A0, 0x180, 0x198}, // eeproms from bios v0.00 and up + {0x146, 0x270, 0x2B0, 0x200, 0x1C8, 0x1E0, 0x1B0, 0x180, 0x198}, // eeproms from bios v1.70 and up +}; + +#endif \ No newline at end of file diff --git a/pcsx2/Makefile.am b/pcsx2/Makefile.am index 4a8ef1cf76..8de7b044d6 100644 --- a/pcsx2/Makefile.am +++ b/pcsx2/Makefile.am @@ -25,7 +25,7 @@ CDVDiso.h Counters.h HostGui.h IopHw.h MemoryTypes.h R3000A.h CDVDisodrv.h Decode_XA.h Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ CDVDlib.h Dump.h IopBios.h IopSio2.h NakedAsm.h R5900Exceptions.h Sif.h VU.h vtlb.h \ COP0.h Elfheader.h IopBios2.h Mdec.h Patch.h R5900OpcodeTables.h Sifcmd.h VUflags.h \ -Cache.h Exceptions.h IopCommon.h MemcpyFast.h Paths.h Sio.h VUmicro.h \ +Cache.h Exceptions.h IopCommon.h MemcpyFast.h Paths.h Sio.h Sio_internal.h VUmicro.h \ CdRom.h GS.h IopCounters.h Memory.h Plugins.h SPR.h Stats.h VUops.h #RedtapeWindows.h diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index fd4f9ce1d4..3ccacd31e2 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -1464,6 +1464,10 @@ RelativePath="..\..\CDVD.h" > + + From 15821b465a24faaae9f8e0d8271059b42e2889cc Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 16 May 2009 04:26:10 +0000 Subject: [PATCH 075/296] Work on pcsx2's Makefiles a bit. Move CDVD*.* and CdRom.* to a separate subdirectory, to make things a bit less cluttered. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1201 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/{ => CDVD}/CDVD.cpp | 0 pcsx2/{ => CDVD}/CDVD.h | 0 pcsx2/{ => CDVD}/CDVD_internal.h | 0 pcsx2/{ => CDVD}/CDVDiso.cpp | 0 pcsx2/{ => CDVD}/CDVDiso.h | 0 pcsx2/{ => CDVD}/CDVDisodrv.cpp | 0 pcsx2/{ => CDVD}/CDVDisodrv.h | 0 pcsx2/{ => CDVD}/CDVDlib.h | 0 pcsx2/{ => CDVD}/CdRom.cpp | 0 pcsx2/{ => CDVD}/CdRom.h | 0 pcsx2/CDVD/Makefile.am | 18 ++++++++++++ pcsx2/DebugTools/Makefile.am | 12 +++++++- pcsx2/IPU/Makefile.am | 12 +++++++- pcsx2/IPU/mpeg2lib/Makefile.am | 12 +++++++- pcsx2/Linux/Makefile.am | 18 ++++++++++-- pcsx2/Makefile.am | 34 +++++++++++++++------- pcsx2/RDebug/Makefile.am | 12 +++++++- pcsx2/configure.ac | 3 ++ pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 18 ++++++------ pcsx2/x86/Makefile.am | 12 +++++++- pcsx2/x86/ix86/Makefile.am | 12 +++++++- 21 files changed, 135 insertions(+), 28 deletions(-) rename pcsx2/{ => CDVD}/CDVD.cpp (100%) rename pcsx2/{ => CDVD}/CDVD.h (100%) rename pcsx2/{ => CDVD}/CDVD_internal.h (100%) rename pcsx2/{ => CDVD}/CDVDiso.cpp (100%) rename pcsx2/{ => CDVD}/CDVDiso.h (100%) rename pcsx2/{ => CDVD}/CDVDisodrv.cpp (100%) rename pcsx2/{ => CDVD}/CDVDisodrv.h (100%) rename pcsx2/{ => CDVD}/CDVDlib.h (100%) rename pcsx2/{ => CDVD}/CdRom.cpp (100%) rename pcsx2/{ => CDVD}/CdRom.h (100%) create mode 100644 pcsx2/CDVD/Makefile.am diff --git a/pcsx2/CDVD.cpp b/pcsx2/CDVD/CDVD.cpp similarity index 100% rename from pcsx2/CDVD.cpp rename to pcsx2/CDVD/CDVD.cpp diff --git a/pcsx2/CDVD.h b/pcsx2/CDVD/CDVD.h similarity index 100% rename from pcsx2/CDVD.h rename to pcsx2/CDVD/CDVD.h diff --git a/pcsx2/CDVD_internal.h b/pcsx2/CDVD/CDVD_internal.h similarity index 100% rename from pcsx2/CDVD_internal.h rename to pcsx2/CDVD/CDVD_internal.h diff --git a/pcsx2/CDVDiso.cpp b/pcsx2/CDVD/CDVDiso.cpp similarity index 100% rename from pcsx2/CDVDiso.cpp rename to pcsx2/CDVD/CDVDiso.cpp diff --git a/pcsx2/CDVDiso.h b/pcsx2/CDVD/CDVDiso.h similarity index 100% rename from pcsx2/CDVDiso.h rename to pcsx2/CDVD/CDVDiso.h diff --git a/pcsx2/CDVDisodrv.cpp b/pcsx2/CDVD/CDVDisodrv.cpp similarity index 100% rename from pcsx2/CDVDisodrv.cpp rename to pcsx2/CDVD/CDVDisodrv.cpp diff --git a/pcsx2/CDVDisodrv.h b/pcsx2/CDVD/CDVDisodrv.h similarity index 100% rename from pcsx2/CDVDisodrv.h rename to pcsx2/CDVD/CDVDisodrv.h diff --git a/pcsx2/CDVDlib.h b/pcsx2/CDVD/CDVDlib.h similarity index 100% rename from pcsx2/CDVDlib.h rename to pcsx2/CDVD/CDVDlib.h diff --git a/pcsx2/CdRom.cpp b/pcsx2/CDVD/CdRom.cpp similarity index 100% rename from pcsx2/CdRom.cpp rename to pcsx2/CDVD/CdRom.cpp diff --git a/pcsx2/CdRom.h b/pcsx2/CDVD/CdRom.h similarity index 100% rename from pcsx2/CdRom.h rename to pcsx2/CDVD/CdRom.h diff --git a/pcsx2/CDVD/Makefile.am b/pcsx2/CDVD/Makefile.am new file mode 100644 index 0000000000..be2de4296b --- /dev/null +++ b/pcsx2/CDVD/Makefile.am @@ -0,0 +1,18 @@ +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(top_srcdir) -I$(x86_dir) -I$(common_dir) -I$(third_dir) +noinst_LIBRARIES = libps2_cdvd.a + +libps2_cdvd_a_SOURCES = \ +CDVD.cpp CDVDiso.cpp CDVDisodrv.cpp CdRom.cpp \ +CDVD.h CDVD_internal.h CDVDiso.h CDVDisodrv.h CDVDlib.h CdRom.h + +#SUBDIRS = \ No newline at end of file diff --git a/pcsx2/DebugTools/Makefile.am b/pcsx2/DebugTools/Makefile.am index 24e7e0bf18..fe0dc8ab59 100644 --- a/pcsx2/DebugTools/Makefile.am +++ b/pcsx2/DebugTools/Makefile.am @@ -1,4 +1,14 @@ -INCLUDES = -I@srcdir@/../ -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(top_srcdir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libDebugTools.a libDebugTools_a_SOURCES = \ diff --git a/pcsx2/IPU/Makefile.am b/pcsx2/IPU/Makefile.am index c6ba0fd177..3a6a835956 100644 --- a/pcsx2/IPU/Makefile.am +++ b/pcsx2/IPU/Makefile.am @@ -1,4 +1,14 @@ -INCLUDES = -I@srcdir@/../ -I@srcdir@/../x86 -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(top_srcdir) -I$(x86_dir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libIPU.a libIPU_a_SOURCES = IPU.cpp yuv2rgb.cpp coroutine.cpp \ diff --git a/pcsx2/IPU/mpeg2lib/Makefile.am b/pcsx2/IPU/mpeg2lib/Makefile.am index 84af02e631..ca369bdc59 100644 --- a/pcsx2/IPU/mpeg2lib/Makefile.am +++ b/pcsx2/IPU/mpeg2lib/Makefile.am @@ -1,4 +1,14 @@ -INCLUDES = -I@srcdir@/../ -I@srcdir@/../../ -I@srcdir@/../../../common/include -I@srcdir@/../../../3rdparty +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(ipu_dir) -I$(top_srcdir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libmpeg2IPU.a libmpeg2IPU_a_SOURCES = Idct.cpp Mpeg.cpp Mpeg.h Vlc.h \ No newline at end of file diff --git a/pcsx2/Linux/Makefile.am b/pcsx2/Linux/Makefile.am index 28de7eb964..c9c9dae65a 100644 --- a/pcsx2/Linux/Makefile.am +++ b/pcsx2/Linux/Makefile.am @@ -1,5 +1,17 @@ AUTOMAKE_OPTIONS = foreign -INCLUDES = $(shell pkg-config --cflags gtk+-2.0) -I@srcdir@/../ -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty + +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux +gtk_includes= $(shell pkg-config --cflags gtk+-2.0) + +INCLUDES = $(gtk_includes) -I$(top_srcdir) -I$(cdvd_dir) -I$(common_dir) -I$(third_dir) bin_PROGRAMS = pcsx2 @@ -14,7 +26,9 @@ pcsx2_LDFLAGS = pcsx2_DEPENDENCIES = ../libpcsx2.a ../IPU/libIPU.a ../IPU/mpeg2lib/libmpeg2IPU.a ../RDebug/libRDebug.a ../tinyxml/libtinyxml.a pcsx2_DEPENDENCIES += ../x86/libx86recomp.a ../x86/ix86/libix86.a pcsx2_DEPENDENCIES += ../DebugTools/libDebugTools.a +pcsx2_DEPENDENCIES += ../CDVD/libps2_cdvd.a pcsx2_LDADD = ../libpcsx2.a ../IPU/libIPU.a ../IPU/mpeg2lib/libmpeg2IPU.a ../RDebug/libRDebug.a ../tinyxml/libtinyxml.a pcsx2_LDADD += ../x86/libx86recomp.a ../x86/ix86/libix86.a -pcsx2_LDADD += ../DebugTools/libDebugTools.a \ No newline at end of file +pcsx2_LDADD += ../DebugTools/libDebugTools.a +pcsx2_LDADD += ../CDVD/libps2_cdvd.a \ No newline at end of file diff --git a/pcsx2/Makefile.am b/pcsx2/Makefile.am index 8de7b044d6..8b666ac2bd 100644 --- a/pcsx2/Makefile.am +++ b/pcsx2/Makefile.am @@ -1,16 +1,28 @@ AUTOMAKE_OPTIONS = foreign -INCLUDES = -I@srcdir@/x86/ -I@srcdir@/../common/include -I@srcdir@/../3rdparty +ACLOCAL_AMFLAGS = -I m4 --install + +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(x86_dir) -I$(cdvd_dir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libpcsx2.a libpcsx2_a_SOURCES = \ AlignedMalloc.cpp Decode_XA.cpp HwRead.cpp MTGS.cpp R3000AInterpreter.cpp Sio.cpp VUflags.cpp \ -CDVD.cpp Dump.cpp HwWrite.cpp R3000AOpcodeTables.cpp SourceLog.cpp VUmicroMem.cpp \ -CDVDiso.cpp Elfheader.cpp Interpreter.cpp Memory.cpp R5900.cpp Stats.cpp VUops.cpp \ -CDVDisodrv.cpp FPU.cpp IopBios.cpp MemoryCard.cpp R5900OpcodeImpl.cpp System.cpp Vif.cpp \ +Dump.cpp HwWrite.cpp R3000AOpcodeTables.cpp SourceLog.cpp VUmicroMem.cpp \ +Elfheader.cpp Interpreter.cpp Memory.cpp R5900.cpp Stats.cpp VUops.cpp \ +FPU.cpp IopBios.cpp MemoryCard.cpp R5900OpcodeImpl.cpp System.cpp Vif.cpp \ COP0.cpp FPU2.cpp IopCounters.cpp Misc.cpp R5900OpcodeTables.cpp ThreadTools.cpp VifDma.cpp \ COP2.cpp FiFo.cpp IopDma.cpp Patch.cpp RecoverySystem.cpp VU0.cpp vssprintf.cpp \ Cache.cpp GS.cpp IopHw.cpp PathUtils.cpp SPR.cpp VU0micro.cpp vtlb.cpp \ -CdRom.cpp Gif.cpp IopMem.cpp Plugins.cpp SaveState.cpp VU0microInterp.cpp xmlpatchloader.cpp \ +Gif.cpp IopMem.cpp Plugins.cpp SaveState.cpp VU0microInterp.cpp xmlpatchloader.cpp \ Console.cpp IopSio2.cpp PrecompiledHeader.cpp Saveslots.cpp VU1micro.cpp \ Counters.cpp Hw.cpp MMI.cpp R3000A.cpp Sif.cpp VU1microInterp.cpp #Mdec.cpp HashTools.cpp @@ -20,13 +32,13 @@ libpcsx2_a_SOURCES += \ ps2/Iop/IopHwRead.cpp ps2/Iop/IopHwWrite.cpp ps2/Iop/IopHw_Internal.h libpcsx2_a_SOURCES += \ -CDVD.h Common.h HashMap.h IopDma.h MemoryCard.h PrecompiledHeader.h SafeArray.h StringUtils.h Vif.h \ -CDVDiso.h Counters.h HostGui.h IopHw.h MemoryTypes.h R3000A.h SamplProf.h System.h VifDma.h \ -CDVDisodrv.h Decode_XA.h Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ -CDVDlib.h Dump.h IopBios.h IopSio2.h NakedAsm.h R5900Exceptions.h Sif.h VU.h vtlb.h \ +Common.h HashMap.h IopDma.h MemoryCard.h PrecompiledHeader.h SafeArray.h StringUtils.h Vif.h \ +Counters.h HostGui.h IopHw.h MemoryTypes.h R3000A.h SamplProf.h System.h VifDma.h \ +Decode_XA.h Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ +Dump.h IopBios.h IopSio2.h NakedAsm.h R5900Exceptions.h Sif.h VU.h vtlb.h \ COP0.h Elfheader.h IopBios2.h Mdec.h Patch.h R5900OpcodeTables.h Sifcmd.h VUflags.h \ Cache.h Exceptions.h IopCommon.h MemcpyFast.h Paths.h Sio.h Sio_internal.h VUmicro.h \ -CdRom.h GS.h IopCounters.h Memory.h Plugins.h SPR.h Stats.h VUops.h +GS.h IopCounters.h Memory.h Plugins.h SPR.h Stats.h VUops.h #RedtapeWindows.h -SUBDIRS = x86 . DebugTools IPU RDebug tinyxml Linux \ No newline at end of file +SUBDIRS = x86 . CDVD DebugTools IPU RDebug tinyxml Linux \ No newline at end of file diff --git a/pcsx2/RDebug/Makefile.am b/pcsx2/RDebug/Makefile.am index 80878c894d..8b67f3ed4b 100644 --- a/pcsx2/RDebug/Makefile.am +++ b/pcsx2/RDebug/Makefile.am @@ -1,4 +1,14 @@ -INCLUDES = -I@srcdir@/../ -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty +pcsx2_dir = $(top_srcdir)/.. +common_dir = $(pcsx2_dir)/common/include +third_dir = $(pcsx2_dir)/3rdparty +cdvd_dir = $(top_srcdir)/CDVD +x86_dir = $(top_srcdir)/x86 +rdebug_dir = $(top_srcdir)/RDebug +debugtools_dir = $(top_srcdir)/DebugTools +ipu_dir = $(top_srcdir)/IPU +linux_dir = $(top_srcdir)/Linux + +INCLUDES = -I$(top_srcdir) -I$(cdvd_dir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libRDebug.a libRDebug_a_SOURCES = \ diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index 33ef288569..417bf5e948 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -9,6 +9,7 @@ AC_PROG_RANLIB dnl necessary for compiling assembly AM_PROG_AS +AC_CONFIG_MACRO_DIR([m4]) svnrev="`svn info | grep Revision:`" @@ -110,6 +111,7 @@ fi AC_OUTPUT([ Makefile + CDVD/Makefile DebugTools/Makefile Linux/Makefile IPU/Makefile @@ -132,3 +134,4 @@ echo " local plugin inis? $localinis" echo " custom cflags? $customcflags" echo " memcpy_fast? $memcpyfast" #echo " microVU? $microVU" +echo " pcsx2_dir = '$pcsx2_dir'" diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 3ccacd31e2..fa733283cd 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -1449,23 +1449,23 @@ Name="Iop" > Date: Sat, 16 May 2009 07:24:31 +0000 Subject: [PATCH 076/296] Let's try that again... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1202 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Elfheader.cpp | 2 +- pcsx2/IopCommon.h | 4 +- pcsx2/Misc.cpp | 2 +- pcsx2/SaveState.cpp | 2 +- pcsx2/configure.ac | 1 - pcsx2/windows/WinMisc.cpp | 2 +- pcsx2/x86/iVif.cpp | 82 +++++++++++++++++++++++++++++++-------- 7 files changed, 71 insertions(+), 24 deletions(-) diff --git a/pcsx2/Elfheader.cpp b/pcsx2/Elfheader.cpp index 208acd3ae6..c264f53a38 100644 --- a/pcsx2/Elfheader.cpp +++ b/pcsx2/Elfheader.cpp @@ -19,7 +19,7 @@ #include "PrecompiledHeader.h" #include "Common.h" -#include "CDVDisodrv.h" +#include "CDVD/CDVDisodrv.h" using namespace std; diff --git a/pcsx2/IopCommon.h b/pcsx2/IopCommon.h index 18b2161e32..c153504a55 100644 --- a/pcsx2/IopCommon.h +++ b/pcsx2/IopCommon.h @@ -23,8 +23,8 @@ #include "R3000A.h" #include "Common.h" -#include "CdRom.h" -#include "CDVD.h" +#include "CDVD/CdRom.h" +#include "CDVD/CDVD.h" #include "Sio.h" #include "Sif.h" diff --git a/pcsx2/Misc.cpp b/pcsx2/Misc.cpp index 7899499157..77992cdc70 100644 --- a/pcsx2/Misc.cpp +++ b/pcsx2/Misc.cpp @@ -29,7 +29,7 @@ #include "IopCommon.h" #include "HostGui.h" -#include "CDVDisodrv.h" +#include "CDVD/CDVDisodrv.h" #include "VUmicro.h" #include "VU.h" #include "iCore.h" diff --git a/pcsx2/SaveState.cpp b/pcsx2/SaveState.cpp index f0a437bd1d..569e4e045a 100644 --- a/pcsx2/SaveState.cpp +++ b/pcsx2/SaveState.cpp @@ -21,7 +21,7 @@ #include "IopCommon.h" #include "SaveState.h" -#include "CDVDisodrv.h" +#include "CDVD/CDVDisodrv.h" #include "VUmicro.h" #include "VU.h" #include "iCore.h" diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index 417bf5e948..baff113474 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -134,4 +134,3 @@ echo " local plugin inis? $localinis" echo " custom cflags? $customcflags" echo " memcpy_fast? $memcpyfast" #echo " microVU? $microVU" -echo " pcsx2_dir = '$pcsx2_dir'" diff --git a/pcsx2/windows/WinMisc.cpp b/pcsx2/windows/WinMisc.cpp index 23c812e0cf..2e050fe316 100644 --- a/pcsx2/windows/WinMisc.cpp +++ b/pcsx2/windows/WinMisc.cpp @@ -17,7 +17,7 @@ */ #include "Win32.h" -#include "cdvd.h" +#include "CDVD/CDVD.h" static LARGE_INTEGER lfreq; diff --git a/pcsx2/x86/iVif.cpp b/pcsx2/x86/iVif.cpp index 64d63a99f9..c068ecd8e1 100644 --- a/pcsx2/x86/iVif.cpp +++ b/pcsx2/x86/iVif.cpp @@ -25,37 +25,40 @@ #include #include +//#define USE_OLD_IVIF_CODE + // sse2 highly optimized vif (~200 separate functions are built) zerofrog(@gmail.com) extern u32 g_vif1Masks[48], g_vif0Masks[48]; extern u32 g_vif1HasMask3[4], g_vif0HasMask3[4]; // arranged in writearr, rowarr, colarr, updatearr static PCSX2_ALIGNED16(u32 s_maskarr[16][4]) = { - 0xffffffff, 0x00000000, 0x00000000, 0xffffffff, - 0xffff0000, 0x0000ffff, 0x00000000, 0xffffffff, - 0xffff0000, 0x00000000, 0x0000ffff, 0xffffffff, - 0xffff0000, 0x00000000, 0x00000000, 0xffff0000, - 0x0000ffff, 0xffff0000, 0x00000000, 0xffffffff, - 0x00000000, 0xffffffff, 0x00000000, 0xffffffff, - 0x00000000, 0xffff0000, 0x0000ffff, 0xffffffff, - 0x00000000, 0xffff0000, 0x00000000, 0xffff0000, - 0x0000ffff, 0x00000000, 0xffff0000, 0xffffffff, - 0x00000000, 0x0000ffff, 0xffff0000, 0xffffffff, - 0x00000000, 0x00000000, 0xffffffff, 0xffffffff, - 0x00000000, 0x00000000, 0xffff0000, 0xffff0000, - 0x0000ffff, 0x00000000, 0x00000000, 0x0000ffff, - 0x00000000, 0x0000ffff, 0x00000000, 0x0000ffff, - 0x00000000, 0x00000000, 0x0000ffff, 0x0000ffff, - 0x00000000, 0x00000000, 0x00000000, 0x00000000 + {0xffffffff, 0x00000000, 0x00000000, 0xffffffff}, + {0xffff0000, 0x0000ffff, 0x00000000, 0xffffffff}, + {0xffff0000, 0x00000000, 0x0000ffff, 0xffffffff}, + {0xffff0000, 0x00000000, 0x00000000, 0xffff0000}, + {0x0000ffff, 0xffff0000, 0x00000000, 0xffffffff}, + {0x00000000, 0xffffffff, 0x00000000, 0xffffffff}, + {0x00000000, 0xffff0000, 0x0000ffff, 0xffffffff}, + {0x00000000, 0xffff0000, 0x00000000, 0xffff0000}, + {0x0000ffff, 0x00000000, 0xffff0000, 0xffffffff}, + {0x00000000, 0x0000ffff, 0xffff0000, 0xffffffff}, + {0x00000000, 0x00000000, 0xffffffff, 0xffffffff}, + {0x00000000, 0x00000000, 0xffff0000, 0xffff0000}, + {0x0000ffff, 0x00000000, 0x00000000, 0x0000ffff}, + {0x00000000, 0x0000ffff, 0x00000000, 0x0000ffff}, + {0x00000000, 0x00000000, 0x0000ffff, 0x0000ffff}, + {0x00000000, 0x00000000, 0x00000000, 0x00000000} }; extern u8 s_maskwrite[256]; extern "C" PCSX2_ALIGNED16(u32 s_TempDecompress[4]) = {0}; +#if defined(_MSC_VER) || !defined(USE_OLD_IVIF_CODE) void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) { - u32 i; + u32 i; u32 prev = 0; FreezeXMMRegs(1); for(i = 0; i < 4; ++i, mask >>= 8, oldmask >>= 8, vif1masks += 16) { @@ -86,3 +89,48 @@ void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) } FreezeXMMRegs(0); } + +#else // gcc +// After some experimentation, I'm putting the old code back in for now for testing purposes, as the +// other version reliably SegFaults when loading YuGiOh: Duelist of the Roses on Linux (when setting +// r0 to _mm_load_si128, when vif1masks=0x846f670, hasmask=0x846f454, mask=0, and oldmask=5). +// Seems to work everywhere else. It'll stay disabled for now, but it's easier for me to fiddle with if its in here. -arcum42 +void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) +{ + u32 i; + u32 prev = 0; + FreezeXMMRegs(1); + + for(i = 0; i < 4; ++i, mask >>= 8, oldmask >>= 8, vif1masks += 16) { + + prev |= s_maskwrite[mask&0xff];//((mask&3)==3)||((mask&0xc)==0xc)||((mask&0x30)==0x30)||((mask&0xc0)==0xc0); + hasmask[i] = prev; + + if( (mask&0xff) != (oldmask&0xff) ) { + u8* p0 = (u8*)&s_maskarr[mask&15][0]; + u8* p1 = (u8*)&s_maskarr[(mask>>4)&15][0]; + + __asm__(".intel_syntax noprefix\n" + "movaps xmm0, [%0]\n" + "movaps xmm1, [%1]\n" + "movaps xmm2, xmm0\n" + "punpcklwd xmm0, xmm0\n" + "punpckhwd xmm2, xmm2\n" + "movaps xmm3, xmm1\n" + "punpcklwd xmm1, xmm1\n" + "punpckhwd xmm3, xmm3\n" + "movq [%2], xmm0\n" + "movq [%2+8], xmm1\n" + "movhps [%2+16], xmm0\n" + "movhps [%2+24], xmm1\n" + "movq [%2+32], xmm2\n" + "movq [%2+40], xmm3\n" + "movhps [%2+48], xmm2\n" + "movhps [%2+56], xmm3\n" + ".att_syntax\n" : : "r"(p0), "r"(p1), "r"(vif1masks) ); + } + } + FreezeXMMRegs(0); +} + +#endif From 2c3043862d20dd823acb35461718981bf577367c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 16 May 2009 08:43:56 +0000 Subject: [PATCH 077/296] microVU: - Cached Program search is faster/smarter thanks to an idea Jake had to search recently used blocks first. - Fixed a problem with block searching. - Minor changes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1203 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 65 ++++++++++++++++++++--------------- pcsx2/x86/microVU.h | 15 ++++---- pcsx2/x86/microVU_Execute.inl | 2 +- 3 files changed, 47 insertions(+), 35 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index e69f1de53c..de7a4f7ccb 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -53,7 +53,7 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr) { mVUreset(); } -// Will Optimize later +// Resets Rec Data microVUt(void) mVUreset() { microVU* mVU = mVUx; @@ -86,7 +86,6 @@ microVUt(void) mVUreset() { }*/ // Program Variables - mVU->prog.finished = 1; mVU->prog.cleared = 1; mVU->prog.cur = -1; mVU->prog.total = -1; @@ -125,8 +124,10 @@ microVUt(void) mVUclose() { microVUt(void) mVUclear(u32 addr, u32 size) { microVU* mVU = mVUx; - memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); - mVU->prog.cleared = 1; // Next execution searches/creates a new microprogram + if (!mVU->prog.cleared) { + memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); + mVU->prog.cleared = 1; // Next execution searches/creates a new microprogram + } } //------------------------------------------------------------------ @@ -198,30 +199,36 @@ microVUt(int) mVUfindLeastUsedProg() { microVUt(void) __mVUvsyncUpdate() { microVU* mVU = mVUx; - if (mVU->prog.total < mVU->prog.max) return; for (int i = 0; i <= mVU->prog.total; i++) { - if( mVU->prog.prog[i].last_used != 0 ) - { - if( mVU->prog.prog[i].last_used >= 3 ) - { - // program has been used recently. Give it's program execution counter a - // 'weighted' bonus signifying it's importance: - if( mVU->prog.prog[i].used < 0x4fffffff ) - mVU->prog.prog[i].used += 0x200; + if (mVU->prog.prog[i].last_used != 0) { + if (mVU->prog.prog[i].last_used >= 3) { + + if (mVU->prog.prog[i].used < 0x4fffffff) // program has been used recently. Give it a + mVU->prog.prog[i].used += 0x200; // 'weighted' bonus signifying it's importance } mVU->prog.prog[i].last_used--; } - else - mVU->prog.prog[i].used /= 2; // penalize unused programs. + else mVU->prog.prog[i].used /= 2; // penalize unused programs. } } -void mVUvsyncUpdate() -{ - __mVUvsyncUpdate<0>(); - __mVUvsyncUpdate<1>(); +microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck) { + microVU* mVU = mVUx; + + if (progUsed) { + if (!memcmp_mmx(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize)) { + mVU->prog.cur = progIndex; + mVU->prog.cleared = 0; + mVU->prog.prog[progIndex].last_used = 3; + if (!needOverflowCheck || mVU->prog.prog[progIndex].used < 0x7fffffff) { + mVU->prog.prog[progIndex].used++; // increment 'used' (avoiding overflows if necessary) + } + return 1; + } + } + return 0; } // Searches for Cached Micro Program and sets prog.cur to it (returns 1 if program found, else returns 0) @@ -230,15 +237,12 @@ microVUt(int) mVUsearchProg() { if (mVU->prog.cleared) { // If cleared, we need to search for new program for (int i = 0; i <= mVU->prog.total; i++) { - if (!memcmp_mmx(mVU->prog.prog[i].data, mVU->regs->Micro, mVU->microSize)) { - mVU->prog.cur = i; - mVU->prog.cleared = 0; - if( mVU->prog.prog[i].used < 0x7fffffff ) // avoid overflows on well-used programs - mVU->prog.prog[i].used++; - - mVU->prog.prog[i].last_used = 3; // add me to the mVU structs - return 1; - } + if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 1)) + return 1; // Check Recently Used Programs + } + for (int i = 0; i <= mVU->prog.total; i++) { + if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0)) + return 1; // Check Older Programs } mVU->prog.cur = mVUfindLeastUsedProg(); // If cleared and program not found, make a new program instance mVU->prog.cleared = 0; @@ -297,4 +301,9 @@ void runVUrec(u32 startPC, u32 cycles, const int vuIndex) { else startVU1(startPC, cycles); } +void mVUvsyncUpdate() { + __mVUvsyncUpdate<0>(); + __mVUvsyncUpdate<1>(); +} + #endif // PCSX2_MICROVU diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 774469f7c1..e55571221e 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -27,7 +27,7 @@ #include "microVU_Alloc.h" #include "microVU_Misc.h" -#define mMaxBlocks 32 // Max Blocks With Different Pipeline States (For n = 1, 2, 4, 8, 16, etc...) +#define mMaxBlocks 32 // Max Blocks With Different Pipeline States class microBlockManager { private: static const int MaxBlocks = mMaxBlocks - 1; @@ -42,7 +42,7 @@ public: microBlock* thisBlock = search(&pBlock->pState); if (!thisBlock) { listSize++; - if (listSize > MaxBlocks) { Console::Error("microVU Warning: Block List Overflow"); listSize &= MaxBlocks; } + if (listSize > MaxBlocks) { Console::Error("microVU Warning: Block List Overflow"); listSize = 0; } memcpy_fast(&blockList[listSize], pBlock, sizeof(microBlock)); thisBlock = &blockList[listSize]; } @@ -57,8 +57,10 @@ public: } else { // Can do Simple Search (Only Matches the Important Pipeline Stuff) for (int i = 0; i <= listSize; i++) { - if ((blockList[i].pState.q == pState->q) && (blockList[i].pState.p == pState->p) - && (blockList[i].pState.flags == pState->flags)) { return &blockList[i]; } + if ((blockList[i].pState.q == pState->q) + && (blockList[i].pState.p == pState->p) + && (blockList[i].pState.flags == pState->flags) + && !(blockList[i].pState.needExactMatch & 0xf0f)) { return &blockList[i]; } } } return NULL; @@ -69,8 +71,9 @@ template // progSize = VU program memory size / 4 struct microProgram { u32 data[progSize]; u32 used; // Number of times its been used - u32 last_used; // counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vsync) + u32 last_used; // Counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vSync) u32 sFlagHack; // Optimize out Status Flag Updates if Program doesn't use Status Flags + s32 range[2]; // The range of microMemory that has already been recompiled for the current program u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache @@ -85,8 +88,8 @@ struct microProgManager { static const int max = mMaxProg - 1; int cur; // Index to Current MicroProgram thats running (-1 = uncached) int total; // Total Number of valid MicroPrograms minus 1 + int isSame; // Current cached microProgram is Exact Same program as mVU->regs->Micro int cleared; // Micro Program is Indeterminate so must be searched for (and if no matches are found then recompile a new one) - int finished; // Completed MicroProgram by E-bit Termination microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 85285c3c6b..37dd5dbea2 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -108,7 +108,7 @@ microVUt(void) mVUdispatcherB() { if (isMMX(1)) EMMS(); RET(); - mVUcacheCheck(x86Ptr, mVU->cache, 512); + mVUcacheCheck(x86Ptr, mVU->cache, 0x1000); } //------------------------------------------------------------------ From 757098e66a135ac791efc6ddef638af84714db50 Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 16 May 2009 16:10:16 +0000 Subject: [PATCH 078/296] GT menus look much better now (GT4 at least) - Bugfix from r1197 Note to users: Does this fix ToA or Persona? git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1204 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index f80de93ab0..a71ca56c26 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2073,7 +2073,7 @@ static void Vif1CMDFlush() // FLUSH/E/A while ((gif->chcr & 0x100)) { if (!Path3transfer && gif->qwc == 0) break; - vif1Regs->stat |= 0x4; + vif1Regs->stat |= VIF1_STAT_VGW; break; } @@ -2131,7 +2131,7 @@ static void Vif1CMDDirectHL() // DIRECT/HL { //if(vif1Regs->mskpath3)CPU_INT(2, vif1ch->qwc - (vif1.vifpacketsize >> 2) * BIAS); //DirectHL flushes the lot - vif1Regs->stat |= 0x4; + vif1Regs->stat |= VIF1_STAT_VGW; } @@ -2219,7 +2219,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size != 0) DevCon::Error("no vif1 cmd but tag size is left last cmd read %x", params vif1Regs->code); if (vif1.irq) break; - if(vif1Regs->stat & 0x4) break; + if(vif1Regs->stat & VIF1_STAT_VGW) break; vif1.cmd = (data[0] >> 24); vif1Regs->code = data[0]; @@ -2264,7 +2264,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size == 0) break; } } - if(vif1Regs->stat & 0x4) break; + if(vif1Regs->stat & VIF1_STAT_VGW) break; } // End of Transfer loop transferred += size - vif1.vifpacketsize; @@ -2302,7 +2302,7 @@ int VIF1transfer(u32 *data, int size, int istag) vif1ch->qwc -= transferred; } - if(vif1Regs->stat & 0x4) + if(vif1Regs->stat & VIF1_STAT_VGW) { vif1.vifstalled = true; return -1; @@ -2489,7 +2489,7 @@ __forceinline void vif1Interrupt() VIF_LOG("vif1Interrupt: %8.8x", cpuRegs.cycle); g_vifCycles = 0; - if((vif1Regs->stat & 0x4) && + if((vif1Regs->stat & VIF1_STAT_VGW) && ((psHu32(GIF_MODE) & 0x1) ? (Path3transfer == true || gif->qwc > 0) : (gif->chcr & 0x100))) { int delay = 0; @@ -2504,7 +2504,7 @@ __forceinline void vif1Interrupt() CPU_INT(1, delay); return; } - vif1Regs->stat &= ~0x4; + vif1Regs->stat &= ~VIF1_STAT_VGW; if ((vif1ch->chcr & 0x100) == 0) Console::WriteLn("Vif1 running when CHCR == %x", params vif1ch->chcr); From c07a79741edb461ad2c250c8b3dfa8a39d696b58 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 16 May 2009 19:45:30 +0000 Subject: [PATCH 079/296] Linux: Added a console log regarding page protection, based on a hunch I have for what's causing Issue 220. [need a linux person to check it and see if the console logs the new error or not] git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1205 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/LnxSysExec.cpp | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/pcsx2/Linux/LnxSysExec.cpp b/pcsx2/Linux/LnxSysExec.cpp index 5ce58e3b1c..b0536465d8 100644 --- a/pcsx2/Linux/LnxSysExec.cpp +++ b/pcsx2/Linux/LnxSysExec.cpp @@ -382,10 +382,18 @@ namespace HostSys void MemProtect( void* baseaddr, size_t size, PageProtectionMode mode, bool allowExecution ) { + // Breakpoint this to trap potentially inappropriate use of page protection, which would + // be caused by failed aligned directives on global vars (dunno if it's happening or not yet) + Console::Error( + "*PCSX2/Linux Warning* Inappropriate use of page protection detected.\n" + "\tbaseaddr not page aligned: 0x%08X", params (uptr)baseaddr + ); + int lnxmode = 0; - // make sure size is aligned to the system page size: + // make sure base and size are aligned to the system page size: size = (size + m_pagemask) & ~m_pagemask; + baseaddr = (void*)( ((uptr)baseaddr) & ~mpagemask ); switch( mode ) { From e6627c5e78b367c4e5b75899d36d102fb96c8436 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 16 May 2009 21:07:30 +0000 Subject: [PATCH 080/296] Linux: Typo fix from prev commit. >_< git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1206 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/LnxSysExec.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/Linux/LnxSysExec.cpp b/pcsx2/Linux/LnxSysExec.cpp index b0536465d8..4fb0bbbf77 100644 --- a/pcsx2/Linux/LnxSysExec.cpp +++ b/pcsx2/Linux/LnxSysExec.cpp @@ -393,7 +393,7 @@ namespace HostSys // make sure base and size are aligned to the system page size: size = (size + m_pagemask) & ~m_pagemask; - baseaddr = (void*)( ((uptr)baseaddr) & ~mpagemask ); + baseaddr = (void*)( ((uptr)baseaddr) & ~m_pagemask ); switch( mode ) { From e1b38852b651dae0ef770e5b556f1d4c667af737 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 17 May 2009 02:21:39 +0000 Subject: [PATCH 081/296] Linux: Take back out the logging message from 1204-1205, since Jake seems to be done with it. Remove the messages printed in Exception handing, and set up a test to abort if it's endlessly looping on one offset. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1207 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/LnxSysExec.cpp | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-) diff --git a/pcsx2/Linux/LnxSysExec.cpp b/pcsx2/Linux/LnxSysExec.cpp index 4fb0bbbf77..37de74572b 100644 --- a/pcsx2/Linux/LnxSysExec.cpp +++ b/pcsx2/Linux/LnxSysExec.cpp @@ -22,7 +22,8 @@ static bool sinit = false; GtkWidget *FileSel; - +static uptr current_offset = 0; +static uptr offset_counter = 0; bool Slots[5] = { false, false, false, false, false }; void InstallLinuxExceptionHandler() @@ -47,13 +48,27 @@ void SysPageFaultExceptionFilter( int signal, siginfo_t *info, void * ) { // get bad virtual address uptr offset = (u8*)info->si_addr - psM; - - DevCon::Status( "Protected memory cleanup. Offset 0x%x", params offset ); + + if (offset != current_offset) + { + current_offset = offset; + offset_counter = 0; + } + else + { + offset_counter++; + if (offset_counter > 500) + { + DevCon::Status( "Offset 0x%x endlessly repeating. Aborting.", params offset ); + assert( false ); + } + } if (offset>=Ps2MemSize::Base) { // Bad mojo! Completely invalid address. // Instigate a crash or abort emulation or something. + DevCon::Status( "Offset 0x%x invalid. Legit SIGSEGV. Aborting.", params offset ); assert( false ); } @@ -384,10 +399,10 @@ namespace HostSys { // Breakpoint this to trap potentially inappropriate use of page protection, which would // be caused by failed aligned directives on global vars (dunno if it's happening or not yet) - Console::Error( + /*Console::Error( "*PCSX2/Linux Warning* Inappropriate use of page protection detected.\n" "\tbaseaddr not page aligned: 0x%08X", params (uptr)baseaddr - ); + );*/ int lnxmode = 0; From f6ce237b8e8a4546f1b20cfc12d59022d3825faf Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sun, 17 May 2009 03:57:18 +0000 Subject: [PATCH 082/296] Did some general cleanups to vtlb's memory protection and recompiled block integrity checking systems. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1208 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/COP0.cpp | 8 +- pcsx2/Linux/LnxSysExec.cpp | 13 +-- pcsx2/Memory.cpp | 144 +++++++++++++++++++++++--------- pcsx2/Memory.h | 4 +- pcsx2/Sif.cpp | 3 +- pcsx2/x86/ix86-32/iR5900-32.cpp | 20 ++--- 6 files changed, 132 insertions(+), 60 deletions(-) diff --git a/pcsx2/COP0.cpp b/pcsx2/COP0.cpp index 6042250b18..500f2eeaf9 100644 --- a/pcsx2/COP0.cpp +++ b/pcsx2/COP0.cpp @@ -69,7 +69,7 @@ void MapTLB(int i) for (addr=saddr; addr> 12) & mask)) { //match memSetPageAddr(addr << 12, tlb[i].PFN0 + ((addr - saddr) << 12)); - Cpu->Clear(addr << 12, 1); + Cpu->Clear(addr << 12, 0x400); } } } @@ -82,7 +82,7 @@ void MapTLB(int i) for (addr=saddr; addr> 12) & mask)) { //match memSetPageAddr(addr << 12, tlb[i].PFN1 + ((addr - saddr) << 12)); - Cpu->Clear(addr << 12, 1); + Cpu->Clear(addr << 12, 0x400); } } } @@ -109,7 +109,7 @@ void UnmapTLB(int i) for (addr=saddr; addr> 12) & mask)) { //match memClearPageAddr(addr << 12); - Cpu->Clear(addr << 12, 1); + Cpu->Clear(addr << 12, 0x400); } } } @@ -122,7 +122,7 @@ void UnmapTLB(int i) for (addr=saddr; addr> 12) & mask)) { //match memClearPageAddr(addr << 12); - Cpu->Clear(addr << 12, 1); + Cpu->Clear(addr << 12, 0x400); } } } diff --git a/pcsx2/Linux/LnxSysExec.cpp b/pcsx2/Linux/LnxSysExec.cpp index 37de74572b..8ec471d00e 100644 --- a/pcsx2/Linux/LnxSysExec.cpp +++ b/pcsx2/Linux/LnxSysExec.cpp @@ -398,11 +398,14 @@ namespace HostSys void MemProtect( void* baseaddr, size_t size, PageProtectionMode mode, bool allowExecution ) { // Breakpoint this to trap potentially inappropriate use of page protection, which would - // be caused by failed aligned directives on global vars (dunno if it's happening or not yet) - /*Console::Error( - "*PCSX2/Linux Warning* Inappropriate use of page protection detected.\n" - "\tbaseaddr not page aligned: 0x%08X", params (uptr)baseaddr - );*/ + // be caused by failed aligned directives on global vars. + /*if( (baseaddr & m_pagemask) != 0 ) + { + Console::Error( + "*PCSX2/Linux Warning* Inappropriate use of page protection detected.\n" + "\tbaseaddr not page aligned: 0x%08X", params (uptr)baseaddr + ); + }*/ int lnxmode = 0; diff --git a/pcsx2/Memory.cpp b/pcsx2/Memory.cpp index 217e72bc0b..050a5cc44d 100644 --- a/pcsx2/Memory.cpp +++ b/pcsx2/Memory.cpp @@ -117,8 +117,6 @@ void loadBiosRom( const char *ext, u8 *dest, long maxSize ) fclose(fp); } -static u32 psMPWC[(Ps2MemSize::Base/32)>>12]; -static std::vector psMPWVA[Ps2MemSize::Base>>12]; u8 *psM = NULL; //32mb Main Ram u8 *psR = NULL; //4mb rom area @@ -828,52 +826,122 @@ void memReset() loadBiosRom("erom", PS2MEM_EROM, Ps2MemSize::ERom); } -int mmap_GetRamPageInfo(void* ptr) +////////////////////////////////////////////////////////////////////////////////////////// +// Memory Protection and Block Checking, vtlb Style! +// +// For the first time code is recompiled (executed), the PS2 ram page for that code is +// protected using Virtual Memory (mprotect). If the game modifies its own code then this +// protection causes an *exception* to be raised (signal in Linux), which is handled by +// unprotecting the page and switching the recompiled block to "manual" protection. +// +// Manual protection uses a simple brute-force memcmp of the recompiled code to the code +// currently in RAM for *each time* the block is executed. Fool-proof, but slow, which +// is why we default to using the exception-based protection scheme described above. +// +// Why manual blocks? Because many games contain code and data in the same 4k page, so +// we *cannot* automatically recompile and reprotect pages, lest we end up recompiling and +// reprotecting them constantly (Which would be very slow). As a counter, the R5900 side +// of the block checking code does try to periodically re-protect blocks [going from manual +// back to protected], so that blocks which underwent a single invalidation don't need to +// incur a permanent performance penalty. +// +// Page Granularity: +// Fortunately for us MIPS and x86 use the same page granularity for TLB and memory +// protection, so we can use a 1:1 correspondence when protecting pages. Page granularity +// is 4096 (4k), which is why you'll see a lot of 0xfff's, >><< 12's, and 0x1000's in the +// code below. +// + +enum vtlb_ProtectionMode { - u32 offset=((u8*)ptr-psM); - if (offset>=Ps2MemSize::Base) + ProtMode_None = 0, // page is 'unaccounted' -- neither protected nor unprotected + ProtMode_Write, // page is under write protection (exception handler) + ProtMode_Manual // page is under manual protection (self-checked at execution) +}; + +struct vtlb_PageProtectionInfo +{ + // Ram De-mapping -- used to convert fully translated/mapped offsets into psM back + // into their originating ps2 physical ram address. Values are assigned when pages + // are marked for protection. + u32 ReverseRamMap; + + vtlb_ProtectionMode Mode; +}; + +PCSX2_ALIGNED16( static vtlb_PageProtectionInfo m_PageProtectInfo[Ps2MemSize::Base >> 12] ); + + +// returns: +// -1 - unchecked block (resides in ROM, thus is integrity is constant) +// 0 - page is using Write protection +// 1 - page is using manual protection (recompiler must include execution-time +// self-checking of block integrity) +// +int mmap_GetRamPageInfo( u32 paddr ) +{ + paddr &= ~0xfff; + + uptr ptr = (uptr)PSM( paddr ); + uptr rampage = ptr - (uptr)psM; + + if (rampage >= Ps2MemSize::Base) return -1; //not in ram, no tracking done ... - offset>>=12; - return (psMPWC[(offset/32)]&(1<<(offset&31)))?1:0; + + rampage >>= 12; + return ( m_PageProtectInfo[rampage].Mode == ProtMode_Manual ) ? 1 : 0; } -void mmap_MarkCountedRamPage(void* ptr,u32 vaddr) +// paddr - physically mapped address +void mmap_MarkCountedRamPage( u32 paddr ) { - HostSys::MemProtect( ptr, 1, Protect_ReadOnly ); + paddr &= ~0xfff; - u32 offset=((u8*)ptr-psM); - offset>>=12; - psMPWC[(offset/32)] &= ~(1<<(offset&31)); + uptr ptr = (uptr)PSM( paddr ); + int rampage = (ptr - (uptr)psM) >> 12; - for (u32 i=0;ipsM mapping + // (and clear blocks accordingly), but don't necessarily clear the protection status. + m_PageProtectInfo[rampage].ReverseRamMap = paddr; -void mmap_ResetBlockTracking() -{ - DevCon::WriteLn("vtlb/mmap: Block Tracking reset..."); - memzero_obj(psMPWC); - for(u32 i=0;i<(Ps2MemSize::Base>>12);i++) - { - psMPWVA[i].clear(); - } - HostSys::MemProtect( psM, Ps2MemSize::Base, Protect_ReadWrite ); + if( m_PageProtectInfo[rampage].Mode == ProtMode_Write ) + return; // skip town if we're already protected. + + if( m_PageProtectInfo[rampage].Mode == ProtMode_Manual ) + DbgCon::WriteLn( "dyna_page_reset @ 0x%05x", params paddr>>12 ); + else + DbgCon::WriteLn( "Write-protected page @ 0x%05x", params paddr>>12 ); + + m_PageProtectInfo[rampage].Mode = ProtMode_Write; + HostSys::MemProtect( &psM[rampage<<12], 1, Protect_ReadOnly ); } +// offset - offset of address relative to psM. The exception handler for the platform/host +// OS should ensure that only addresses within psM address space are passed. Anything else +// will produce undefined results (ie, crashes). void mmap_ClearCpuBlock( uint offset ) { - HostSys::MemProtect( &psM[offset], 1, Protect_ReadWrite ); - - offset>>=12; - psMPWC[(offset/32)]|=(1<<(offset&31)); + int rampage = offset >> 12; - for (u32 i=0;iClear(psMPWVA[offset][i],0x400); - } - psMPWVA[offset].clear(); -} \ No newline at end of file + // Assertion: This function should never be run on a block that's already under + // manual protection. Indicates a logic error in the recompiler or protection code. + jASSUME( m_PageProtectInfo[rampage].Mode != ProtMode_Manual ); + + #ifndef __LINUX__ // this function is called from the signal handler + DbgCon::WriteLn( "Manual page @ 0x%05x", params m_PageProtectInfo[rampage].ReverseRamMap>>12 ); + #endif + + HostSys::MemProtect( &psM[rampage<<12], 1, Protect_ReadWrite ); + m_PageProtectInfo[rampage].Mode = ProtMode_Manual; + Cpu->Clear( m_PageProtectInfo[rampage].ReverseRamMap, 0x400 ); +} + +// Clears all block tracking statuses, manual protection flags, and write protection. +// This does not clear any recompiler blocks. IT is assumed (and necessary) for the caller +// to ensure the EErec is also reset in conjunction with calling this function. +void mmap_ResetBlockTracking() +{ + DevCon::WriteLn( "vtlb/mmap: Block Tracking reset..." ); + memzero_obj( m_PageProtectInfo ); + HostSys::MemProtect( psM, Ps2MemSize::Base, Protect_ReadWrite ); +} diff --git a/pcsx2/Memory.h b/pcsx2/Memory.h index 3051b73bbb..45cb2098b3 100644 --- a/pcsx2/Memory.h +++ b/pcsx2/Memory.h @@ -141,8 +141,8 @@ extern void memClearPageAddr(u32 vaddr); extern void memMapVUmicro(); -extern int mmap_GetRamPageInfo(void* ptr); -extern void mmap_MarkCountedRamPage(void* ptr,u32 vaddr); +extern int mmap_GetRamPageInfo( u32 paddr ); +extern void mmap_MarkCountedRamPage( u32 paddr ); extern void mmap_ResetBlockTracking(); extern void mmap_ClearCpuBlock( uint offset ); diff --git a/pcsx2/Sif.cpp b/pcsx2/Sif.cpp index 23576f767f..5fdaa6ca05 100644 --- a/pcsx2/Sif.cpp +++ b/pcsx2/Sif.cpp @@ -209,7 +209,8 @@ __forceinline void SIF0Dma() SIF0read((u32*)ptag, readSize << 2); - Cpu->Clear(sif0dma->madr, readSize*4); + // Clearing handled by vtlb memory protection and manual blocks. + //Cpu->Clear(sif0dma->madr, readSize*4); cycles += readSize * BIAS; // fixme : BIAS is factored in below sif0dma->qwc -= readSize; diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 17f820ff57..982540b379 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -1215,19 +1215,20 @@ void badespfn() { assert(0); } +// Called when a block under manual protection fails it's pre-execution integrity check. void __fastcall dyna_block_discard(u32 start,u32 sz) { - DevCon::WriteLn("dyna_block_discard .. start: %08X count=%d", params start,sz); + DevCon::WriteLn("dyna_block_discard .. start=0x%08X size=%d", params start, sz*4); recClear(start, sz); } - +// called when a block under manual protection has been run enough times to be a +// candidate for being reset under the faster vtlb write protection. void __fastcall dyna_page_reset(u32 start,u32 sz) { - DevCon::WriteLn("dyna_page_reset .. start=%08X size=%d", params start,sz*4); recClear(start & ~0xfffUL, 0x400); manual_counter[start >> 12]++; - mmap_MarkCountedRamPage(PSM(start), start & ~0xfffUL); + mmap_MarkCountedRamPage( start ); } void recRecompile( const u32 startpc ) @@ -1492,15 +1493,14 @@ StartRecomp: // note: blocks are guaranteed to reside within the confines of a single page. - const int PageType = mmap_GetRamPageInfo((u32*)PSM(inpage_ptr)); - const u32 inpage_offs = inpage_ptr & 0xFFF; + const int PageType = mmap_GetRamPageInfo( inpage_ptr ); //const u32 pgsz = std::min(0x1000 - inpage_offs, inpage_sz); const u32 pgsz = inpage_sz; if(PageType!=-1) { if (PageType==0) { - mmap_MarkCountedRamPage(PSM(inpage_ptr),inpage_ptr&~0xFFF); + mmap_MarkCountedRamPage( inpage_ptr ); manual_page[inpage_ptr >> 12] = 0; } else @@ -1550,13 +1550,13 @@ StartRecomp: xJC( dyna_page_reset ); // note: clearcnt is measured per-page, not per-block! - DbgCon::WriteLn( "Manual block @ %08X : size=%3d page/offs=%05X/%03X inpgsz=%d clearcnt=%d", - params startpc, sz, inpage_ptr>>12, inpage_offs, inpage_sz, manual_counter[inpage_ptr >> 12] ); + //DbgCon::WriteLn( "Manual block @ %08X : size=%3d page/offs=%05X/%03X inpgsz=%d clearcnt=%d", + // params startpc, sz, inpage_ptr>>12, inpage_ptr&0xfff, inpage_sz, manual_counter[inpage_ptr >> 12] ); } else { DbgCon::Notice( "Uncounted Manual block @ %08X : size=%3d page/offs=%05X/%03X inpgsz=%d", - params startpc, sz, inpage_ptr>>12, inpage_offs, pgsz, inpage_sz ); + params startpc, sz, inpage_ptr>>12, inpage_ptr&0xfff, pgsz, inpage_sz ); } } From ac00db5dd5645a8c6d28bd5454f9648e8680b2a2 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 17 May 2009 08:22:00 +0000 Subject: [PATCH 083/296] microVU: - Added partial program searching support (helps to recompile less in games that send a lot of small programs like Tekken 5) - Optimized logical min/max code some more (knocked off 2 instructions) - Fixed some cases where 'last_used' should have been set. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1209 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 37 ++++++++++++++++++++++------------- pcsx2/x86/microVU.h | 2 +- pcsx2/x86/microVU_Compile.inl | 33 +++++++++++++++++++++++++++++++ pcsx2/x86/microVU_Misc.h | 19 +++++++++--------- pcsx2/x86/microVU_Misc.inl | 6 ++---- 5 files changed, 68 insertions(+), 29 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index de7a4f7ccb..bc2bfb096c 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -77,15 +77,8 @@ microVUt(void) mVUreset() { // Clear All Program Data memset(&mVU->prog, 0, sizeof(mVU->prog)); - // Create Block Managers - // Block managers are now allocated "on-demand" by the recompiler -- air - /*for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progSize / 2); j++) { - mVU->prog.prog[i].block[j] = new microBlockManager(); - } - }*/ - // Program Variables + mVU->prog.isSame = -1; mVU->prog.cleared = 1; mVU->prog.cur = -1; mVU->prog.total = -1; @@ -99,6 +92,8 @@ microVUt(void) mVUreset() { mVU->prog.prog[i].x86ptr = z; z += (mVU->cacheSize / (mVU->prog.max + 1)); mVU->prog.prog[i].x86end = z; + mVU->prog.prog[i].range[0] = -1; // Set range to + mVU->prog.prog[i].range[1] = -1; // indeterminable status } } @@ -138,7 +133,10 @@ microVUt(void) mVUclear(u32 addr, u32 size) { microVUt(void) mVUclearProg(int progIndex) { microVU* mVU = mVUx; mVU->prog.prog[progIndex].used = 1; + mVU->prog.prog[progIndex].last_used = 3; mVU->prog.prog[progIndex].sFlagHack = 0; + mVU->prog.prog[progIndex].range[0] = -1; + mVU->prog.prog[progIndex].range[1] = -1; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; for (u32 i = 0; i < (mVU->progSize / 2); i++) { if( mVU->prog.prog[progIndex].block[i] ) @@ -160,6 +158,8 @@ microVUt(int) mVUfindLeastUsedProg() { if (mVU->prog.total < mVU->prog.max) { mVU->prog.total++; mVUcacheProg(mVU->prog.total); // Cache Micro Program + mVU->prog.prog[mVU->prog.total].used = 1; + mVU->prog.prog[mVU->prog.total].last_used = 3; Console::Notice("microVU%d: Cached MicroPrograms = %d", params vuIndex, mVU->prog.total+1); return mVU->prog.total; } @@ -214,13 +214,16 @@ microVUt(void) __mVUvsyncUpdate() { } } -microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck) { +// Compare Cached microProgram to mVU->regs->Micro +microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, bool cmpWholeProg) { microVU* mVU = mVUx; - + if (progUsed) { - if (!memcmp_mmx(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize)) { + if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microSize)) || + (!cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data + mVUprogI.range[0], (u8*)mVU->regs->Micro + mVUprogI.range[0], ((mVUprogI.range[1] + 8) - mVUprogI.range[0]))))) { mVU->prog.cur = progIndex; mVU->prog.cleared = 0; + mVU->prog.isSame = cmpWholeProg ? 1 : -1; mVU->prog.prog[progIndex].last_used = 3; if (!needOverflowCheck || mVU->prog.prog[progIndex].used < 0x7fffffff) { mVU->prog.prog[progIndex].used++; // increment 'used' (avoiding overflows if necessary) @@ -237,18 +240,24 @@ microVUt(int) mVUsearchProg() { if (mVU->prog.cleared) { // If cleared, we need to search for new program for (int i = 0; i <= mVU->prog.total; i++) { - if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 1)) + if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 1, 0)) return 1; // Check Recently Used Programs } for (int i = 0; i <= mVU->prog.total; i++) { - if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0)) + if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0, 0)) return 1; // Check Older Programs - } + } + /*for (int i = 0; i <= mVU->prog.total; i++) { + if (mVUcmpProg(i, 1, 1, 0)) + return 1; // Check Partial Program + }*/ mVU->prog.cur = mVUfindLeastUsedProg(); // If cleared and program not found, make a new program instance mVU->prog.cleared = 0; + mVU->prog.isSame = 1; return 0; } mVU->prog.prog[mVU->prog.cur].used++; + mVU->prog.prog[mVU->prog.cur].last_used = 3; return 1; // If !cleared, then we're still on the same program as last-time ;) } /* diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index e55571221e..b6d4106c3c 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -88,7 +88,7 @@ struct microProgManager { static const int max = mMaxProg - 1; int cur; // Index to Current MicroProgram thats running (-1 = uncached) int total; // Total Number of valid MicroPrograms minus 1 - int isSame; // Current cached microProgram is Exact Same program as mVU->regs->Micro + int isSame; // Current cached microProgram is Exact Same program as mVU->regs->Micro (-1 = unknown, 0 = No, 1 = Yes) int cleared; // Micro Program is Indeterminate so must be searched for (and if no matches are found then recompile a new one) microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 661caf6e12..5e21be02e4 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -150,6 +150,34 @@ microVUt(void) mVUtestCycles() { SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } +microVUt(void) mVUcheckIsSame() { + microVU* mVU = mVUx; + + if (mVU->prog.isSame == -1) { + mVU->prog.isSame = !!memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); + } + if (mVU->prog.isSame == 0) { + mVUcacheProg(mVU->prog.cur); + } +} + +microVUt(void) mVUsetupRange(u32 pc) { + microVU* mVU = mVUx; + + if (mVUcurProg.range[0] == -1) { + mVUcurProg.range[0] = (s32)pc; + mVUcurProg.range[1] = (s32)pc; + } + else if (mVUcurProg.range[0] > (s32)pc) { + mVUcurProg.range[0] = (s32)pc; + mVUcheckIsSame(); + } + else if (mVUcurProg.range[1] < (s32)pc) { + mVUcurProg.range[1] = (s32)pc; + mVUcheckIsSame(); + } +} + //------------------------------------------------------------------ // Recompiler //------------------------------------------------------------------ @@ -169,6 +197,9 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); if (pBlock) { return pBlock->x86ptrStart; } + // Setup Program Bounds/Range + mVUsetupRange(startPC); + // First Pass iPC = startPC / 4; setCode(); @@ -230,6 +261,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { else { microBlock* bBlock = NULL; u32* ajmp = 0; + mVUsetupRange(xPC); switch (mVUbranch) { case 3: branchCase(JE32, JNE32); // IBEQ @@ -308,6 +340,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } // Do E-bit end stuff here + mVUsetupRange(xPC - 8); mVUendProgram(findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); return thisPtr; //ToDo: Save pipeline state? diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index b995fc9c0e..e8c34d0886 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -150,6 +150,7 @@ declareAllVariables #define pass4 if (recPass == 3) // Misc Macros... +#define mVUprogI mVU->prog.prog[progIndex] #define mVUcurProg mVU->prog.prog[mVU->prog.cur] #define mVUblocks mVU->prog.prog[mVU->prog.cur].block #define mVUallocInfo mVU->prog.prog[mVU->prog.cur].allocInfo @@ -161,13 +162,14 @@ declareAllVariables #define mVUregs mVUallocInfo.block.pState #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC -#define mVUflagInfo mVUregs.needExactMatch #define mVUsFlagHack mVUallocInfo.sFlagHack #define mVUinfo mVUallocInfo.info[iPC / 2] #define mVUstall mVUallocInfo.stall[iPC / 2] #define mVUstartPC mVUallocInfo.startPC +#define mVUflagInfo mVUregs.needExactMatch +#define mVUflagHack (mVUcurProg.sFlagHack) #define xPC ((iPC / 2) * 8) -#define curI mVUcurProg.data[iPC] +#define curI ((u32*)mVU->regs->Micro)[iPC]//mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } #define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } #define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } @@ -175,7 +177,11 @@ declareAllVariables #define bSaveAddr (((xPC + (2 * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) -#define mVUflagHack (mVUcurProg.sFlagHack) + +// Flag Info +#define __Status (mVUflagInfo & (0xf<<0)) +#define __Mac (mVUflagInfo & (0xf<<4)) +#define __Clip (mVUflagInfo & (0xf<<8)) // Pass 1 uses these to set mVUinfo #define _isNOP (1<<0) // Skip Lower Instruction @@ -254,13 +260,6 @@ declareAllVariables #define mVUlogQ() { mVUlog(", Q"); } #define mVUlogCLIP() { mVUlog("w.xyz vf%02d, vf%02dw", _Fs_, _Ft_); } -// Flag Info -//#define __NeedExact (1<<12) -//#define __ExactMatch (mVUregs.needExactMatch & (1<<12)) -#define __Status (mVUflagInfo & (0xf<<0)) -#define __Mac (mVUflagInfo & (0xf<<4)) -#define __Clip (mVUflagInfo & (0xf<<8)) - // Store VI regs in mmx regs? #define isMMX(_VIreg_) 0 //(_VIreg_ >= 1 && _VIreg_ <=8) #define mmVI(_VIreg_) (_VIreg_ - 1) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 05647a0614..9810b00336 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -348,12 +348,10 @@ void MIN_MAX_(x86SSERegType to, x86SSERegType from, bool min) { // Warning: Modifies from and to's upper 3 vectors void MIN_MAX_SS(x86SSERegType to, x86SSERegType from, bool min) { - SSE2_PSHUFD_XMM_to_XMM(to, to, 0x50); + SSE_SHUFPS_XMM_to_XMM (to, from, 0); SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); - SSE2_PSHUFD_XMM_to_XMM(from, from, 0x50); - SSE2_PAND_M128_to_XMM (from, (uptr)MIN_MAX_MASK1); - SSE2_POR_M128_to_XMM (from, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(from, to, 0xee); if (min) SSE2_MINPD_XMM_to_XMM(to, from); else SSE2_MAXPD_XMM_to_XMM(to, from); } From 0391b76015ea6ed791b2332a2af57497b3d17748 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 17 May 2009 09:23:32 +0000 Subject: [PATCH 084/296] Linux: Fix the exeption check up a bit, and mess with the configure.ac file. It should now be possible to compile with versions of gcc lower then 4.3, though I haven't tested that. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1210 96395faa-99c1-11dd-bbfe-3dabce05a288 --- build.sh | 12 ++- pcsx2/Linux/LnxSysExec.cpp | 27 +++-- pcsx2/build.sh | 2 +- pcsx2/configure.ac | 203 +++++++++++++++++++++++-------------- 4 files changed, 157 insertions(+), 87 deletions(-) diff --git a/build.sh b/build.sh index 94bc4c8acd..fbf599a8d6 100644 --- a/build.sh +++ b/build.sh @@ -11,8 +11,14 @@ #Optimized, but a devbuild export PCSX2OPTIONS="--enable-devbuild --prefix `pwd`" -#Debug / Devbuild version -#export PCSX2OPTIONS="--enable-debug --enable-devbuild --prefix `pwd`" +#Debug version (which now implies a devbuild) +#export PCSX2OPTIONS="--enable-debug --prefix `pwd`" + +#Normal, but unoptimized +#export PCSX2OPTIONS="--disable-optimization --prefix `pwd`" + +#Normal, with warnings +#export PCSX2OPTIONS="--enable-warnings --prefix `pwd`" #ZeroGS Normal mode export ZEROGSOPTIONS="--enable-sse2" @@ -27,6 +33,7 @@ option=$@ export PCSX2PLUGINS="`pwd`/bin/plugins" curdir=`pwd` +echo echo "Building the Pcsx2 Suite." echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies." cd ${curdir}/plugins @@ -38,6 +45,7 @@ echo Error with building plugins exit 1 fi +echo echo "Building Pcsx2." echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies." cd ${curdir}/pcsx2 diff --git a/pcsx2/Linux/LnxSysExec.cpp b/pcsx2/Linux/LnxSysExec.cpp index 8ec471d00e..e5da6434f0 100644 --- a/pcsx2/Linux/LnxSysExec.cpp +++ b/pcsx2/Linux/LnxSysExec.cpp @@ -26,7 +26,7 @@ static uptr current_offset = 0; static uptr offset_counter = 0; bool Slots[5] = { false, false, false, false, false }; -void InstallLinuxExceptionHandler() +__noinline void InstallLinuxExceptionHandler() { struct sigaction sa; @@ -36,15 +36,25 @@ void InstallLinuxExceptionHandler() sigaction(SIGSEGV, &sa, NULL); } -void ReleaseLinuxExceptionHandler() +__noinline void ReleaseLinuxExceptionHandler() { - // Code this later. + // This may be called too early or something, since implementing it causes all games to segfault. + // I'll look in to it. --arcum42 } +__noinline void KillLinuxExceptionHandler() +{ + struct sigaction sa; + + sigemptyset(&sa.sa_mask); + sa.sa_flags = SA_RESETHAND; + //sa.sa_sigaction = &SysPageFaultExceptionFilter; + sigaction(SIGSEGV, &sa, NULL); +} static const uptr m_pagemask = getpagesize()-1; // Linux implementation of SIGSEGV handler. Bind it using sigaction(). -void SysPageFaultExceptionFilter( int signal, siginfo_t *info, void * ) +__noinline void SysPageFaultExceptionFilter( int signal, siginfo_t *info, void * ) { // get bad virtual address uptr offset = (u8*)info->si_addr - psM; @@ -60,6 +70,7 @@ void SysPageFaultExceptionFilter( int signal, siginfo_t *info, void * ) if (offset_counter > 500) { DevCon::Status( "Offset 0x%x endlessly repeating. Aborting.", params offset ); + KillLinuxExceptionHandler(); assert( false ); } } @@ -69,6 +80,7 @@ void SysPageFaultExceptionFilter( int signal, siginfo_t *info, void * ) // Bad mojo! Completely invalid address. // Instigate a crash or abort emulation or something. DevCon::Status( "Offset 0x%x invalid. Legit SIGSEGV. Aborting.", params offset ); + KillLinuxExceptionHandler(); assert( false ); } @@ -205,7 +217,6 @@ void OnStates_Load(GtkMenuItem *menuitem, gpointer user_data) } sscanf(name, "Slot %d", &i); - //if (States_Load(i)) ExecuteCpu(); States_Load(i); RefreshMenuSlots(); } @@ -218,8 +229,6 @@ void OnLoadOther_Ok(GtkButton* button, gpointer user_data) File = (gchar*)gtk_file_selection_get_filename(GTK_FILE_SELECTION(FileSel)); strcpy(str, File); gtk_widget_destroy(FileSel); - - //if (States_Load(str)) ExecuteCpu(); States_Load(str); RefreshMenuSlots(); } @@ -399,13 +408,13 @@ namespace HostSys { // Breakpoint this to trap potentially inappropriate use of page protection, which would // be caused by failed aligned directives on global vars. - /*if( (baseaddr & m_pagemask) != 0 ) + if( ((uptr)baseaddr & m_pagemask) != 0 ) { Console::Error( "*PCSX2/Linux Warning* Inappropriate use of page protection detected.\n" "\tbaseaddr not page aligned: 0x%08X", params (uptr)baseaddr ); - }*/ + } int lnxmode = 0; diff --git a/pcsx2/build.sh b/pcsx2/build.sh index 9b88972756..e7f54a886c 100644 --- a/pcsx2/build.sh +++ b/pcsx2/build.sh @@ -18,7 +18,7 @@ # #Normal -#export PCSX2OPTIONS="--enable-debug --enable-devbuild --enable-sse3 --enable-sse4 --prefix `pwd`" +#export PCSX2OPTIONS="--prefix `pwd`" echo --------------- echo Building Pcsx2 diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index baff113474..cb8bdeebe6 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -1,4 +1,4 @@ -AC_INIT(pcsx2,0.9.6,zerofrog@gmail.com) +AC_INIT(pcsx2,0.9.6,arcum42@gmail.com) AM_INIT_AUTOMAKE(pcsx2,0.9.6) AC_PROG_CC([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) @@ -11,6 +11,81 @@ dnl necessary for compiling assembly AM_PROG_AS AC_CONFIG_MACRO_DIR([m4]) +if test "x$GCC" = "xyes"; then + dnl get gcc version + AC_MSG_CHECKING([gcc version]) + gccver=$($CC -dumpversion) + gccvermajor=$(echo $gccver | cut -d . -f1) + gccverminor=$(echo $gccver | cut -d . -f2) + gccverrevision=$(echo $gccver | cut -d . -f3) + gccvernum=$(expr $gccvermajor "*" 100 + $gccverminor "*" 10 + $gccverrevision) + AC_MSG_RESULT($gccver) +fi + +AC_ARG_ENABLE(warnings, AC_HELP_STRING([--enable-warnings], [Add warning flags when compiling]), +warnings=$enableval,warnings=no) + +AC_ARG_ENABLE(customcflags, AC_HELP_STRING([--enable-customcflags], [Use custom CFLAGS]), +customcflags=$enableval,customcflags=no) + +AC_ARG_ENABLE(optimization, AC_HELP_STRING([--disable-optimization], [No optimizations when compiling]), +optimization=$enableval,optimization=yes) + +AC_ARG_ENABLE(unsupportedspeedfreak, AC_HELP_STRING([--enable-unsupportedspeedfreak], [Set the compiler to do unreasonable optimizations that will probably make all your games crash.]), +unsupportedspeedfreak=$enableval,unsupportedspeedfreak=no) + +dnl Check for debug build +AC_MSG_CHECKING(debug build) +AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), + debug=$enableval,debug=no) + +if test "x$debug" == "xyes" +then + AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEVBUILD,1,[PCSX2_DEVBUILD]) + MYOBJDIR="Debug" + optimization="no" + warnings="yes" +else + AC_DEFINE(NDEBUG,1,[NDEBUG]) + MYOBJDIR="Release" +fi +AC_MSG_RESULT($debug) + +dnl Check for dev build +AC_MSG_CHECKING(for development build) +AC_ARG_ENABLE(devbuild, AC_HELP_STRING([--enable-devbuild], [Special Build for developers that simplifies testing and adds extra checks]), + devbuild=$enableval,devbuild=no) + +if test "x$debug" == "xyes" +then + dnl Force devbuild on, because Debug mode never seems to stay working properly without Devbuild on. + devbuild=yes +fi +if test "x$devbuild" == "xyes" +then + AC_DEFINE(PCSX2_DEVBUILD,1,[PCSX2_DEVBUILD]) + MYOBJDIR="Dev" + nowarnings="no" +fi +AC_MSG_RESULT($devbuild) + +AC_ARG_ENABLE(nls, AC_HELP_STRING([--disable-nls], [Disable nls support]), +nls=$enableval,nls=yes) + +if test "x$nls" == xyes +then + AC_CHECK_HEADER([libintl.h], [AC_DEFINE(ENABLE_NLS,1,[__x86_64__])]) +fi + +AC_ARG_ENABLE(local-inis, AC_HELP_STRING([--enable-local-inis], [Enable local plugin inis]), +localinis=$enableval, localinis=no) + +if test "x$localinis" == xyes +then + AC_DEFINE(LOCAL_PLUGIN_INIS) +fi + svnrev="`svn info | grep Revision:`" if test -n "${svnrev}" ; then @@ -23,69 +98,60 @@ else [Define to be the subversion revision number]) fi -AC_ARG_ENABLE(customcflags, AC_HELP_STRING([--enable-customcflags], [Use custom CFLAGS]), -customcflags=$enableval,customcflags=no) - -if test "x$customcflags" == xno +if test "x$customcflags" == "xno" then -CFLAGS= -CPPFLAGS= -CXXFLAGS= -CCASFLAGS= -else -DEBUG_FLAGS=" -O0 -g " + CFLAGS= + CPPFLAGS= + CXXFLAGS= + CCASFLAGS= fi -WARNING_FLAGS="-Wno-format -Wno-unused-parameter -Wno-unused-value -Wunused-variable " -EXTRA_WARNING_FLAGS="-Wall -Wextra" -NORMAL_FLAGS=" -pipe -msse -msse2 -O2 ${WARNING_FLAGS}" -# These optimizations seem to cause issues with GCC 4.3.3, so we'll turn them off. Comment if not on 4.3+ -NORMAL_FLAGS+=" -fno-guess-branch-probability -fno-dse -fno-tree-dse " +WARNING_FLAGS= +NORMAL_FLAGS= -DEBUG_FLAGS+=" -g -msse -msse2 ${EXTRA_WARNING_FLAGS} ${WARNING_FLAGS} " - -dnl Check for debug build -AC_MSG_CHECKING(debug build) -AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), - debug=$enableval,debug=no) -if test "x$debug" == xyes +if test "x$warnings" == "xyes" then - AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+=" ${DEBUG_FLAGS} -Xlinker -zmuldefs -m32 " - CXXFLAGS+=" ${DEBUG_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " - CCASFLAGS+=" -D_DEBUG ${DEBUG_FLAGS} -m32 " - MYOBJDIR="Debug" + WARNING_FLAGS+=" -Wall -Wextra -Wunused-variable " +fi + +WARNING_FLAGS+=" -Wno-format -Wno-unused-parameter -Wno-unused-value " +NORMAL_FLAGS+=" -msse -msse2 ${WARNING_FLAGS} " + +if test "x$unsupportedspeedfreak" == "xyes" +then + dnl Don't even try to get support for this. + NORMAL_FLAGS+=" -pipe -O3 -march=native -funroll-loops -ffast-math " else - AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+=" ${NORMAL_FLAGS} -m32 " + if test "x$optimization" == "xyes" + then + NORMAL_FLAGS+=" -pipe -O2 " + else + NORMAL_FLAGS+=" -O0 " + fi +fi + +if test "x$unsupportedspeedfreak" == "xno" +then +dnl These optimizations seem to cause issues with GCC 4.3.3, so we'll turn them off. +dnl I really need to run some tests and see if they are still needed. + if test "$gccvernum" -ge "433"; then + NORMAL_FLAGS+=" -fno-guess-branch-probability -fno-dse -fno-tree-dse " + fi +fi + + CFLAGS+=" ${NORMAL_FLAGS} -Xlinker -zmuldefs -m32 " CXXFLAGS+=" ${NORMAL_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " - CCASFLAGS+=" ${NORMAL_FLAGS} -m32 " - MYOBJDIR="Release" + CCASFLAGS+=" ${NORMAL_FLAGS} -m32 " + +if test "x$debug" == "xyes" +then + CFLAGS+=" -g " + CXXFLAGS+=" -g " + CCASFLAGS+=" -D_DEBUG " fi -AC_MSG_RESULT($debug) AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) -#AC_MSG_CHECKING(turn on microVU) -#AC_ARG_ENABLE(microVU, AC_HELP_STRING([--enable-microVU], [Turns on the currently incomplete microVU files - Not a good idea]), -# microVU=$enableval,microVU=no) -#if test "x$microVU" == xyes -#then -# AC_DEFINE(PCSX2_MICROVU,1,[PCSX2_MICROVU]) -#fi -#AC_MSG_RESULT($microVU) - -dnl Check for dev build -AC_MSG_CHECKING(for development build) -AC_ARG_ENABLE(devbuild, AC_HELP_STRING([--enable-devbuild], [Special Build for developers that simplifies testing and adds extra checks]), - devbuild=$enableval,devbuild=no) -if test "x$devbuild" == xyes -then - AC_DEFINE(PCSX2_DEVBUILD,1,[PCSX2_DEVBUILD]) - MYOBJDIR=$MYOBJDIR"Dev" -fi -AC_MSG_RESULT($devbuild) - dnl gtk AC_MSG_CHECKING(gtk+) AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) @@ -95,20 +161,6 @@ dnl AC_CHECK_LIB(pthread,main,[LIBS="$LIBS -lpthread"]) AC_CHECK_LIB(stdc++,main,[LIBS="$LIBS -lstdc++"]) AC_CHECK_LIB(z,main,[LIBS="$LIBS -lz"]) -AC_ARG_ENABLE(nls, AC_HELP_STRING([--disable-nls], [Disable nls support]), -nls=$enableval,nls=yes) -if test "x$nls" == xyes -then -AC_CHECK_HEADER([libintl.h], [AC_DEFINE(ENABLE_NLS,1,[__x86_64__])]) -fi - -AC_ARG_ENABLE(local-inis, AC_HELP_STRING([--enable-local-inis], [Enable local plugin inis]), -localinis=$enableval, localinis=no) -if test "x$localinis" == xyes -then -AC_DEFINE(LOCAL_PLUGIN_INIS) -fi - AC_OUTPUT([ Makefile CDVD/Makefile @@ -126,11 +178,12 @@ AC_OUTPUT([ dnl bindir = pcsx2exe echo "Configuration:" -echo " Target system type: $target" -echo " Debug build? $debug" -echo " Dev build? $devbuild" -echo " nls support? $nls" -echo " local plugin inis? $localinis" -echo " custom cflags? $customcflags" -echo " memcpy_fast? $memcpyfast" -#echo " microVU? $microVU" +echo " Target system type: $target" +echo " Debug build? $debug" +echo " Dev build? $devbuild" +echo " NLS support? $nls" +echo " Local plugin inis? $localinis" +echo " Custom cflags? $customcflags" +echo " Compiler warnings? $warnings" +echo " Compiler optimizations? $optimization" + From 942e7e92318cd98bae458619662b6e822431766a Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 17 May 2009 11:34:00 +0000 Subject: [PATCH 085/296] microVU: - Crashes when "microVU Error: Program went over its cache limit" message occurs should be fixed. - Miscellaneous code changes and cleanups. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1211 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 37 +++------------------ pcsx2/x86/microVU.h | 18 ++++++----- pcsx2/x86/microVU_Analyze.inl | 35 +++++++++++++++++--- pcsx2/x86/microVU_Compile.inl | 61 ++++++++++++++++++----------------- pcsx2/x86/microVU_Flags.inl | 5 ++- pcsx2/x86/microVU_Misc.h | 9 ++++-- pcsx2/x86/microVU_Misc.inl | 3 -- 7 files changed, 84 insertions(+), 84 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index bc2bfb096c..d4564b78c9 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -30,6 +30,7 @@ PCSX2_ALIGNED16(microVU microVU0); PCSX2_ALIGNED16(microVU microVU1); declareAllVariables // Declares All Global Variables :D + //------------------------------------------------------------------ // Micro VU - Main Functions //------------------------------------------------------------------ @@ -117,7 +118,6 @@ microVUt(void) mVUclose() { // Clears Block Data in specified range microVUt(void) mVUclear(u32 addr, u32 size) { - microVU* mVU = mVUx; if (!mVU->prog.cleared) { memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); @@ -139,7 +139,7 @@ microVUt(void) mVUclearProg(int progIndex) { mVU->prog.prog[progIndex].range[1] = -1; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; for (u32 i = 0; i < (mVU->progSize / 2); i++) { - if( mVU->prog.prog[progIndex].block[i] ) + if (mVU->prog.prog[progIndex].block[i]) mVU->prog.prog[progIndex].block[i]->reset(); } } @@ -170,11 +170,9 @@ microVUt(int) mVUfindLeastUsedProg() { int smallidx = startidx; u32 smallval = mVU->prog.prog[startidx].used; - for (int i = startidx; i != endidx; i = (i+1)&mVU->prog.max) - { + for (int i = startidx; i != endidx; i = (i+1)&mVU->prog.max) { u32 used = mVU->prog.prog[i].used; - if (smallval > used) - { + if (smallval > used) { smallval = used; smallidx = i; } @@ -195,7 +193,6 @@ microVUt(int) mVUfindLeastUsedProg() { // frame-based decrementing system in combination with a program-execution-based incrementing // system. In english: if last_used >= 2 it means the program has been used for the current // or prev frame. if it's 0, the program hasn't been used for a while. -// microVUt(void) __mVUvsyncUpdate() { microVU* mVU = mVUx; @@ -246,11 +243,7 @@ microVUt(int) mVUsearchProg() { for (int i = 0; i <= mVU->prog.total; i++) { if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0, 0)) return 1; // Check Older Programs - } - /*for (int i = 0; i <= mVU->prog.total; i++) { - if (mVUcmpProg(i, 1, 1, 0)) - return 1; // Check Partial Program - }*/ + } mVU->prog.cur = mVUfindLeastUsedProg(); // If cleared and program not found, make a new program instance mVU->prog.cleared = 0; mVU->prog.isSame = 1; @@ -260,27 +253,7 @@ microVUt(int) mVUsearchProg() { mVU->prog.prog[mVU->prog.cur].last_used = 3; return 1; // If !cleared, then we're still on the same program as last-time ;) } -/* -// Block Invalidation -__forceinline void mVUinvalidateBlock(microVU* mVU, u32 addr, u32 size) { - int i = addr/8; - int end = i+((size+(8-(size&7)))/8); // ToDo: Can be simplified to addr+size if Size is always a multiple of 8 - - if (!mVU->prog.cleared) { - for ( ; i < end; i++) { - if ( mVU->prog.prog[mVU->prog.cur].block[i]->clear() ) { - mVU->prog.cleared = 1; - i++; - break; - } - } - } - for ( ; i < end; i++) { - mVU->prog.prog[mVU->prog.cur].block[i]->clearFast(); - } -} -*/ //------------------------------------------------------------------ // Wrapper Functions - Called by other parts of the Emu //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index b6d4106c3c..827102865d 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -17,8 +17,8 @@ */ #pragma once -//#define mVUdebug // Prints Extra Info to Console -//#define mVUlogProg // Dumps MicroPrograms into microVU0.txt/microVU1.txt +//#define mVUdebug // Prints Extra Info to Console +//#define mVUlogProg // Dumps MicroPrograms to \logs\*.html #include "Common.h" #include "VU.h" @@ -27,24 +27,26 @@ #include "microVU_Alloc.h" #include "microVU_Misc.h" -#define mMaxBlocks 32 // Max Blocks With Different Pipeline States +#define mMaxBlocks 32 // Max Blocks With Different Pipeline States (For n = 1, 2, 4, 8, 16, etc...) class microBlockManager { private: static const int MaxBlocks = mMaxBlocks - 1; int listSize; // Total Items - 1 + int listI; // Index to Add new block microBlock blockList[mMaxBlocks]; public: microBlockManager() { reset(); } ~microBlockManager() {} - void reset() { listSize = -1; }; + void reset() { listSize = -1; listI = -1; }; microBlock* add(microBlock* pBlock) { microBlock* thisBlock = search(&pBlock->pState); if (!thisBlock) { - listSize++; - if (listSize > MaxBlocks) { Console::Error("microVU Warning: Block List Overflow"); listSize = 0; } - memcpy_fast(&blockList[listSize], pBlock, sizeof(microBlock)); - thisBlock = &blockList[listSize]; + listI++; + if (listSize < MaxBlocks) { listSize++; } + if (listI > MaxBlocks) { Console::Error("microVU Warning: Block List Overflow"); listI = 0; } + memcpy_fast(&blockList[listI], pBlock, sizeof(microBlock)); + thisBlock = &blockList[listI]; } return thisBlock; } diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index e10803938d..c61872be38 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -49,6 +49,19 @@ } \ } +#define analyzeReg1b(reg) { \ + if (reg) { \ + analyzeReg1(reg); \ + if (mVUregsTemp.VFreg[0] == reg) { \ + if ((mVUregsTemp.VF[0].x && _X) \ + || (mVUregsTemp.VF[0].y && _Y) \ + || (mVUregsTemp.VF[0].z && _Z) \ + || (mVUregsTemp.VF[0].w && _W)) \ + { mVUinfo |= _swapOps; } \ + } \ + } \ +} + microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { microVU* mVU = mVUx; mVUinfo |= _doStatus; @@ -136,6 +149,13 @@ microVUt(void) mVUanalyzeIALU2(int Is, int It) { if (_Y) { mVUstall = aMax(mVUstall, aReg(reg).z); } \ if (_Z) { mVUstall = aMax(mVUstall, aReg(reg).w); } \ if (_W) { mVUstall = aMax(mVUstall, aReg(reg).x); } \ + if (mVUregsTemp.VFreg[0] == reg) { \ + if ((mVUregsTemp.VF[0].y && _X) \ + || (mVUregsTemp.VF[0].z && _Y) \ + || (mVUregsTemp.VF[0].w && _Z) \ + || (mVUregsTemp.VF[0].x && _W)) \ + { mVUinfo |= _swapOps; } \ + } \ } \ } @@ -158,6 +178,13 @@ microVUt(void) mVUanalyzeMR32(int Fs, int Ft) { case 2: mVUstall = aMax(mVUstall, aReg(reg).z); break; \ case 3: mVUstall = aMax(mVUstall, aReg(reg).w); break; \ } \ + if (mVUregsTemp.VFreg[0] == reg) { \ + if ((mVUregsTemp.VF[0].x && (fxf == 0)) \ + || (mVUregsTemp.VF[0].y && (fxf == 1)) \ + || (mVUregsTemp.VF[0].z && (fxf == 2)) \ + || (mVUregsTemp.VF[0].w && (fxf == 3))) \ + { mVUinfo |= _swapOps; } \ + } \ } \ } @@ -186,7 +213,7 @@ microVUt(void) mVUanalyzeEFU1(int Fs, int Fsf, u8 xCycles) { microVUt(void) mVUanalyzeEFU2(int Fs, u8 xCycles) { microVU* mVU = mVUx; mVUprint("microVU: EFU Opcode"); - analyzeReg1(Fs); + analyzeReg1b(Fs); analyzePreg(xCycles); } @@ -207,8 +234,7 @@ microVUt(void) mVUanalyzeMFP(int Ft) { microVUt(void) mVUanalyzeMOVE(int Fs, int Ft) { microVU* mVU = mVUx; if (!Ft || (Ft == Fs)) { mVUinfo |= _isNOP; } - if (mVUregsTemp.VFreg[0] == Fs) { mVUinfo |= _swapOps; } - analyzeReg1(Fs); + analyzeReg1b(Fs); analyzeReg2(Ft, 1); } @@ -231,10 +257,9 @@ microVUt(void) mVUanalyzeLQ(int Ft, int Is, bool writeIs) { microVUt(void) mVUanalyzeSQ(int Fs, int It, bool writeIt) { microVU* mVU = mVUx; - analyzeReg1(Fs); + analyzeReg1b(Fs); analyzeVIreg1(It); if (writeIt) { analyzeVIreg2(It, 1); } - if (mVUregsTemp.VFreg[0] == Fs) { mVUinfo |= _swapOps; } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 5e21be02e4..3af44862f8 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -45,6 +45,36 @@ // Helper Functions //------------------------------------------------------------------ +// Used by mVUsetupRange +microVUt(void) mVUcheckIsSame() { + microVU* mVU = mVUx; + + if (mVU->prog.isSame == -1) { + mVU->prog.isSame = !!memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); + } + if (mVU->prog.isSame == 0) { + mVUcacheProg(mVU->prog.cur); + } +} + +// Sets up microProgram PC ranges based on whats been recompiled +microVUt(void) mVUsetupRange(u32 pc) { + microVU* mVU = mVUx; + + if (mVUcurProg.range[0] == -1) { + mVUcurProg.range[0] = (s32)pc; + mVUcurProg.range[1] = (s32)pc; + } + else if (mVUcurProg.range[0] > (s32)pc) { + mVUcurProg.range[0] = (s32)pc; + mVUcheckIsSame(); + } + else if (mVUcurProg.range[1] < (s32)pc) { + mVUcurProg.range[1] = (s32)pc; + mVUcheckIsSame(); + } +} + // Recompiles Code for Proper Flags and Q/P regs on Block Linkings microVUt(void) mVUsetupBranch(int* xStatus, int* xMac, int* xClip, int xCycles) { microVU* mVU = mVUx; @@ -150,34 +180,6 @@ microVUt(void) mVUtestCycles() { SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } -microVUt(void) mVUcheckIsSame() { - microVU* mVU = mVUx; - - if (mVU->prog.isSame == -1) { - mVU->prog.isSame = !!memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); - } - if (mVU->prog.isSame == 0) { - mVUcacheProg(mVU->prog.cur); - } -} - -microVUt(void) mVUsetupRange(u32 pc) { - microVU* mVU = mVUx; - - if (mVUcurProg.range[0] == -1) { - mVUcurProg.range[0] = (s32)pc; - mVUcurProg.range[1] = (s32)pc; - } - else if (mVUcurProg.range[0] > (s32)pc) { - mVUcurProg.range[0] = (s32)pc; - mVUcheckIsSame(); - } - else if (mVUcurProg.range[1] < (s32)pc) { - mVUcurProg.range[1] = (s32)pc; - mVUcheckIsSame(); - } -} - //------------------------------------------------------------------ // Recompiler //------------------------------------------------------------------ @@ -276,7 +278,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { incPC(-3); // Go back to branch opcode (to get branch imm addr) mVUsetupBranch(xStatus, xMac, xClip, xCycles); - if( mVUblocks[branchAddr/8] == NULL ) + if (mVUblocks[branchAddr/8] == NULL) mVUblocks[branchAddr/8] = new microBlockManager(); // Check if branch-block has already been compiled @@ -293,7 +295,6 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbackupRegs(); MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) - //AND32ItoR(gprT2, (vuIndex)?0x3ff8:0xff8); // Ensure valid jump address MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) if (!vuIndex) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 7091fc711a..4dc2c8b226 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -26,7 +26,7 @@ microVUt(void) mVUdivSet() { if (doDivFlag) { getFlagReg(flagReg1, fsInstance); if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } - AND16ItoR(flagReg1, 0x0fcf); + AND32ItoR(flagReg1, 0x0fcf); OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } } @@ -223,8 +223,7 @@ microVUt(void) mVUsetFlagInfo() { mVUflagInfo = 0; incPC(4); // Branch Not Taken mVUpass4(xPC); - incPC(-3); - //if (mVUflagInfo != backupFlagInfo) { mVUflagInfo |= __NeedExact; } + incPC(-3); mVUflagInfo |= backupFlagInfo; } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index e8c34d0886..9e1ad49a4d 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -293,7 +293,10 @@ declareAllVariables #define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack // Cache Limit Check -#define mVUcacheCheck(ptr, start, limit) { \ - uptr diff = ptr - start; \ - if (diff >= limit) { Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); } \ +#define mVUcacheCheck(ptr, start, limit) { \ + uptr diff = ptr - start; \ + if (diff >= limit) { \ + Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); \ + mVUreset(); \ + } \ } diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 9810b00336..1d4f16ffba 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -302,12 +302,9 @@ microVUt(void) mVUrestoreRegs() { // Reads entire microProgram and finds out if Status Flag is Used microVUt(void) mVUcheckSflag(int progIndex) { if (CHECK_VU_FLAGHACK) { - microVU* mVU = mVUx; mVUsFlagHack = 1; for (u32 i = 0; i < mVU->progSize; i+=2) { - mVU->code = mVU->prog.prog[progIndex].data[i+1]; - mVUopU(); mVU->code = mVU->prog.prog[progIndex].data[i]; mVUopL(); } From 1b012e9dd920ccd71934f69df67c8090a7ac459e Mon Sep 17 00:00:00 2001 From: refraction Date: Sun, 17 May 2009 12:50:16 +0000 Subject: [PATCH 086/296] Half revert for earlier commit, didnt really make any difference to GT games anyway, just hoped itd effect path3 masking :P Btw, if it doesnt work, Blame Krakatos, he said it does! git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1212 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 17 ++++++----------- 1 file changed, 6 insertions(+), 11 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index a71ca56c26..c1de89843c 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2019,8 +2019,7 @@ static void Vif1CMDMskPath3() // MSKPATH3 #ifdef GSPATH3FIX if ((vif1Regs->code >> 15) & 0x1) { - psHu32(GIF_STAT) |= 0x2; - psHu32(GIF_MODE) |= 0x1; + while ((gif->chcr & 0x100)) //Can be done 2 different ways, depends on the game/company { if (!path3hack && !Path3transfer && (gif->qwc == 0)) break; @@ -2030,14 +2029,13 @@ static void Vif1CMDMskPath3() // MSKPATH3 if (path3hack && (gif->qwc == 0)) break; //add games not working with it to elfheader.c to enable this instead } - + psHu32(GIF_STAT) |= 0x2; } else { // fixme: This is the *only* reason 'transferred' is global. Otherwise it'd be local to Vif1Transfer. if (gif->chcr & 0x100) CPU_INT(2, (transferred >> 2) * BIAS); // Restart Path3 on its own, time it right! psHu32(GIF_STAT) &= ~0x2; - psHu32(GIF_MODE) &= ~0x1; } #else if (vif1Regs->mskpath3) @@ -2073,10 +2071,8 @@ static void Vif1CMDFlush() // FLUSH/E/A while ((gif->chcr & 0x100)) { if (!Path3transfer && gif->qwc == 0) break; - vif1Regs->stat |= VIF1_STAT_VGW; - break; - - } + gsInterrupt(); + } } vif1.cmd &= ~0x7f; @@ -2490,16 +2486,15 @@ __forceinline void vif1Interrupt() g_vifCycles = 0; if((vif1Regs->stat & VIF1_STAT_VGW) && - ((psHu32(GIF_MODE) & 0x1) ? (Path3transfer == true || gif->qwc > 0) : (gif->chcr & 0x100))) + (gif->chcr & 0x100)) { int delay = 0; - + if ((psHu32(GIF_MODE) & 0x1)) delay = min( 8, (int)gif->qwc ); else delay = gif->qwc * BIAS; - if((psHu32(GIF_MODE) & 0x1)) gsInterrupt(); //else CPU_INT(2, min( 64, (int)gif->qwc ) * BIAS); CPU_INT(1, delay); return; From 1a53f45b2c1a2676135a3a9fc6832bd0939e9a70 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 17 May 2009 21:57:24 +0000 Subject: [PATCH 087/296] Some work on variable type consistancy, and getting rid of 64 bit processor checks. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1214 96395faa-99c1-11dd-bbfe-3dabce05a288 --- 3rdparty/SoundTouch/Makefile.am | 5 - 3rdparty/SoundTouch/configure.ac | 28 ----- pcsx2/Linux/LnxThreads.cpp | 24 ---- plugins/SPU2null/SPU2.cpp | 102 +++++++++------- plugins/SPU2null/SPU2.h | 58 ++++----- plugins/spu2-x/src/Makefile.am | 5 - plugins/spu2-x/src/configure.ac | 26 ---- plugins/zerogs/opengl/Linux/Makefile.am | 6 - plugins/zerogs/opengl/Makefile.am | 11 -- plugins/zerogs/opengl/configure | 95 --------------- plugins/zerogs/opengl/configure.ac | 27 ---- plugins/zeropad/Makefile.am | 5 - plugins/zeropad/configure.ac | 26 ---- plugins/zerospu2/Makefile.am | 5 - plugins/zerospu2/configure.ac | 27 ---- plugins/zerospu2/misc.h | 4 +- plugins/zerospu2/voices.cpp | 23 ++-- plugins/zerospu2/zerodma.cpp | 10 +- plugins/zerospu2/zerospu2.cpp | 156 ++++++++++++------------ plugins/zerospu2/zerospu2.h | 100 +++++++-------- 20 files changed, 230 insertions(+), 513 deletions(-) diff --git a/3rdparty/SoundTouch/Makefile.am b/3rdparty/SoundTouch/Makefile.am index baf08c838f..eb39bcc96f 100644 --- a/3rdparty/SoundTouch/Makefile.am +++ b/3rdparty/SoundTouch/Makefile.am @@ -24,13 +24,8 @@ AUTOMAKE_OPTIONS = foreign noinst_HEADERS=AAFilter.h cpu_detect.h FIRFilter.h RateTransposer.h TDStretch.h cpu_detect_x86_gcc.cpp noinst_LIBRARIES = libSoundTouch.a -if X86_64 -libSoundTouch_a_CXXFLAGS = -fPIC -libSoundTouch_a_CFLAGS = -fPIC -else libSoundTouch_a_CXXFLAGS = -msse -mmmx libSoundTouch_a_CFLAGS = -msse -mmmx -endif #lib_LTLIBRARIES=libSoundTouch.la # the mmx_gcc.cpp and cpu_detect_x86_gcc.cpp may need to be conditionally included here from things discovered in configure.ac diff --git a/3rdparty/SoundTouch/configure.ac b/3rdparty/SoundTouch/configure.ac index 73f71cf3d0..458551d5e8 100644 --- a/3rdparty/SoundTouch/configure.ac +++ b/3rdparty/SoundTouch/configure.ac @@ -20,34 +20,6 @@ CFLAGS+=" -m32 " CPPFLAGS+=" -m32 " CXXFLAGS+=" -m32 " CCASFLAGS+=" -m32 " - -dnl Check for 64bit CPU -AC_MSG_CHECKING(for a x86-64 CPU) -dnl if test "$build_os" == "target_os" -dnl then -AC_TRY_RUN([ -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} -],cpu64=yes,cpu64=no,) -dnl else -dnl cpu64=no -dnl fi -if test "x$cpu64" == xyes -then -AC_DEFINE(__x86_64__,1,[__x86_64__]) -fi -AC_MSG_RESULT($cpu64) -AM_CONDITIONAL(X86_64, test x$cpu64 = xyes) - # Checks for header files. AC_CHECK_HEADERS([limits.h memory.h stdlib.h string.h]) diff --git a/pcsx2/Linux/LnxThreads.cpp b/pcsx2/Linux/LnxThreads.cpp index 42d27dcc32..29551fe3ce 100644 --- a/pcsx2/Linux/LnxThreads.cpp +++ b/pcsx2/Linux/LnxThreads.cpp @@ -169,28 +169,4 @@ namespace Threading return result; } - - #ifdef __x86_64__ - __forceinline void pcsx2_InterlockedExchange64(volatile s64* Target, s64 Value) - { - __asm__ __volatile__( - ".intel_syntax noprefix\n" - "lock xchg [%0], rax\n" - ".att_syntax\n" : : "r"(Target), "a"(Value) : "memory" - ); - return 0; - } - - __forceinline s64 pcsx2_InterlockedCompareExchange64(volatile s64* dest, s64 exch, s64 comp) - { - s64 old; - __asm__ __volatile__( - "lock; cmpxchgq %q2, %q1" - : "=a" (old) - : "r" (exch), "m" (*dest), "a" (comp) - ); - return old; - } -#endif - } diff --git a/plugins/SPU2null/SPU2.cpp b/plugins/SPU2null/SPU2.cpp index 7040b31178..720be35249 100644 --- a/plugins/SPU2null/SPU2.cpp +++ b/plugins/SPU2null/SPU2.cpp @@ -23,10 +23,10 @@ #include using namespace std; -const unsigned char version = PS2E_SPU2_VERSION; -const unsigned char revision = 0; -const unsigned char build = 7; // increase that with each version -const unsigned int minor = 1; // increase that with each version +const u8 version = PS2E_SPU2_VERSION; +const u8 revision = 0; +const u8 build = 8; // increase that with each version +const u32 minor = 0; // increase that with each version // ADSR constants #define ATTACK_MS 494L @@ -50,16 +50,16 @@ ADMA Adma7; u32 MemAddr[2]; u32 g_nSpuInit = 0; -unsigned short interrupt = 0; +u16 interrupt = 0; s8 *spu2regs = NULL; u16* spu2mem = NULL; u16* pSpuIrq[2] = {NULL}; u32 dwEndChannel2[2] = {0}; // keeps track of what channels have ended -unsigned long dwNoiseVal = 1; // global noise generator +u32 dwNoiseVal = 1; // global noise generator -int SPUCycles = 0, SPUWorkerCycles = 0; -int SPUStartCycle[2]; -int SPUTargetCycle[2]; +s32 SPUCycles = 0, SPUWorkerCycles = 0; +s32 SPUStartCycle[2]; +s32 SPUTargetCycle[2]; int ADMAS4Write(); int ADMAS7Write(); @@ -70,7 +70,7 @@ void (*irqCallbackSPU2)(); // func of main emu, called on spu i void (*irqCallbackDMA4)() = 0; // func of main emu, called on spu irq void (*irqCallbackDMA7)() = 0; // func of main emu, called on spu irq -const int f[5][2] = { +const s32 f[5][2] = { { 0, 0 }, { 60, 0 }, { 115, -52 }, @@ -221,9 +221,9 @@ EXPORT_C_(void) SPU2async(u32 cycle) void InitADSR() // INIT ADSR { - unsigned long r, rs, rd; - int i; - memset(RateTable, 0, sizeof(unsigned long)*160); // build the rate table according to Neill's rules (see at bottom of file) + u32 r, rs, rd; + s32 i; + memset(RateTable, 0, sizeof(u32)*160); // build the rate table according to Neill's rules (see at bottom of file) r = 3; rs = 1; @@ -678,9 +678,9 @@ int ADMAS4Write() spuaddr = C0_SPUADDR; // SPU2 Deinterleaves the Left and Right Channels - memcpy((short*)(spu2mem + spuaddr + 0x2000), (short*)Adma4.MemAddr, 512); + memcpy((s16*)(spu2mem + spuaddr + 0x2000), (s16*)Adma4.MemAddr, 512); Adma4.MemAddr += 256; - memcpy((short*)(spu2mem + spuaddr + 0x2200), (short*)Adma4.MemAddr, 512); + memcpy((s16*)(spu2mem + spuaddr + 0x2200), (s16*)Adma4.MemAddr, 512); Adma4.MemAddr += 256; spuaddr = (spuaddr + 256) & 511; C0_SPUADDR_SET(spuaddr); @@ -704,9 +704,9 @@ int ADMAS7Write() spuaddr = C1_SPUADDR; // SPU2 Deinterleaves the Left and Right Channels - memcpy((short*)(spu2mem + spuaddr + 0x2400), (short*)Adma7.MemAddr, 512); + memcpy((s16*)(spu2mem + spuaddr + 0x2400), (s16*)Adma7.MemAddr, 512); Adma7.MemAddr += 256; - memcpy((short*)(spu2mem + spuaddr + 0x2600), (short*)Adma7.MemAddr, 512); + memcpy((s16*)(spu2mem + spuaddr + 0x2600), (s16*)Adma7.MemAddr, 512); Adma7.MemAddr += 256; spuaddr = (spuaddr + 256) & 511; C1_SPUADDR_SET(spuaddr); @@ -740,7 +740,7 @@ EXPORT_C_(void) SPU2writeDMA4Mem(u16* pMem, int size) } spuaddr = C0_SPUADDR; - memcpy((unsigned char*)(spu2mem + spuaddr), (unsigned char*)pMem, size << 1); + memcpy((u8*)(spu2mem + spuaddr), (u8*)pMem, size << 1); spuaddr += size; C0_SPUADDR_SET(spuaddr); @@ -779,7 +779,7 @@ EXPORT_C_(void) SPU2writeDMA7Mem(u16* pMem, int size) } spuaddr = C1_SPUADDR; - memcpy((unsigned char*)(spu2mem + spuaddr), (unsigned char*)pMem, size << 1); + memcpy((u8*)(spu2mem + spuaddr), (u8*)pMem, size << 1); spuaddr += size; C1_SPUADDR_SET(spuaddr); @@ -820,9 +820,9 @@ EXPORT_C_(void) SPU2interruptDMA7() } // turn channels on -void SoundOn(int start, int end, unsigned short val) // SOUND ON PSX COMAND +void SoundOn(s32 start, s32 end, u16 val) // SOUND ON PSX COMAND { - for (int ch = start;ch < end;ch++, val >>= 1) // loop channels + for (s32 ch = start;ch < end;ch++, val >>= 1) // loop channels { if ((val&1) && voices[ch].pStart) // mmm... start has to be set before key on !?! { @@ -833,16 +833,16 @@ void SoundOn(int start, int end, unsigned short val) // SOUND ON PSX COMAND } // turn channels off -void SoundOff(int start, int end, unsigned short val) // SOUND OFF PSX COMMAND +void SoundOff(s32 start, s32 end, u16 val) // SOUND OFF PSX COMMAND { - for (int ch = start;ch < end;ch++, val >>= 1) // loop channels + for (s32 ch = start;ch < end;ch++, val >>= 1) // loop channels { if (val&1) // && s_chan[i].bOn) mmm... voices[ch].bStop = true; } } -void FModOn(int start, int end, unsigned short val) // FMOD ON PSX COMMAND +void FModOn(s32 start, s32 end, u16 val) // FMOD ON PSX COMMAND { int ch; @@ -891,8 +891,10 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) case 4: { int NP; - if (value > 0x3fff) NP = 0x3fff; // get pitch val - else NP = value; + if (value > 0x3fff) + NP = 0x3fff; // get pitch val + else + NP = value; pvoice->pvoice->pitch = NP; @@ -924,8 +926,8 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) // more channel info if ((r >= 0x01c0 && r <= 0x02E0) || (r >= 0x05c0 && r <= 0x06E0)) { - int ch = 0; - unsigned long rx = r; + s32 ch = 0; + u32 rx = r; if (rx >= 0x400) { ch = 24; @@ -939,7 +941,7 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) switch (rx) { case 0x1C0: - pvoice->iStartAddr = (((unsigned long)value & 0x3f) << 16) | (pvoice->iStartAddr & 0xFFFF); + pvoice->iStartAddr = (((u32)value & 0x3f) << 16) | (pvoice->iStartAddr & 0xFFFF); pvoice->pStart = (u8*)(spu2mem + pvoice->iStartAddr); break; case 0x1C2: @@ -947,7 +949,7 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) pvoice->pStart = (u8*)(spu2mem + pvoice->iStartAddr); break; case 0x1C4: - pvoice->iLoopAddr = (((unsigned long)value & 0x3f) << 16) | (pvoice->iLoopAddr & 0xFFFF); + pvoice->iLoopAddr = (((u32)value & 0x3f) << 16) | (pvoice->iLoopAddr & 0xFFFF); pvoice->pLoop = (u8*)(spu2mem + pvoice->iLoopAddr); pvoice->bIgnoreLoop = pvoice->iLoopAddr > 0; break; @@ -958,7 +960,7 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) break; case 0x1C8: // unused... check if it gets written as well - pvoice->iNextAddr = (((unsigned long)value & 0x3f) << 16) | (pvoice->iNextAddr & 0xFFFF); + pvoice->iNextAddr = (((u32)value & 0x3f) << 16) | (pvoice->iNextAddr & 0xFFFF); break; case 0x1CA: // unused... check if it gets written as well @@ -982,8 +984,8 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) IRQINFO |= 4; irqCallbackSPU2(); } - if (spuaddr > 0xFFFFE) - spuaddr = 0x2800; + if (spuaddr > 0xFFFFE) spuaddr = 0x2800; + C0_SPUADDR_SET(spuaddr); spu2Ru16(REG_C0_SPUSTAT) &= ~0x80; spu2Ru16(REG_C0_CTRL) &= ~0x30; @@ -998,8 +1000,8 @@ EXPORT_C_(void) SPU2write(u32 mem, u16 value) IRQINFO |= 8; irqCallbackSPU2(); } - if (spuaddr > 0xFFFFE) - spuaddr = 0x2800; + if (spuaddr > 0xFFFFE) spuaddr = 0x2800; + C1_SPUADDR_SET(spuaddr); spu2Ru16(REG_C1_SPUSTAT) &= ~0x80; spu2Ru16(REG_C1_CTRL) &= ~0x30; @@ -1082,23 +1084,27 @@ EXPORT_C_(u16) SPU2read(u32 mem) if ((r >= 0x0000 && r <= 0x0180) || (r >= 0x0400 && r <= 0x0580)) // some channel info? { - int ch = 0; - if (r >= 0x400) ch = ((r - 0x400) >> 4) + 24; - else ch = (r >> 4); + s32 ch = 0; + + if (r >= 0x400) + ch = ((r - 0x400) >> 4) + 24; + else + ch = (r >> 4); VOICE_PROCESSED* pvoice = &voices[ch]; switch (r&0x0f) { case 10: - return (unsigned short)(pvoice->ADSRX.EnvelopeVol >> 16); + return (u16)(pvoice->ADSRX.EnvelopeVol >> 16); } } if ((r > 0x01c0 && r <= 0x02E0) || (r > 0x05c0 && r <= 0x06E0)) // some channel info? { - int ch = 0; - unsigned long rx = r; + s32 ch = 0; + u32 rx = r; + if (rx >= 0x400) { ch = 24; @@ -1132,16 +1138,15 @@ EXPORT_C_(u16) SPU2read(u32 mem) spuaddr = C0_SPUADDR; ret = spu2mem[spuaddr]; spuaddr++; - if (spuaddr > 0xfffff) - spuaddr = 0; + if (spuaddr > 0xfffff) spuaddr = 0; C0_SPUADDR_SET(spuaddr); break; + case REG_C1_SPUDATA: spuaddr = C1_SPUADDR; ret = spu2mem[spuaddr]; spuaddr++; - if (spuaddr > 0xfffff) - spuaddr = 0; + if (spuaddr > 0xfffff) spuaddr = 0; C1_SPUADDR_SET(spuaddr); break; @@ -1196,7 +1201,7 @@ void VOICE_PROCESSED::SetVolume(int iProcessRight) if (vol&0x8000) // sweep not working { - short sInc = 1; // -> sweep up? + s16 sInc = 1; // -> sweep up? if (vol&0x2000) sInc = -1; // -> or down? if (vol&0x1000) vol ^= 0xffff; // -> mmm... phase inverted? have to investigate this vol = ((vol & 0x7f) + 1) / 2; // -> sweep: 0..127 -> 0..64 @@ -1270,7 +1275,10 @@ EXPORT_C_(s32) SPU2freeze(int mode, freezeData *data) { memcpy(spu2regs, spud->spu2regs, 0x10000); } - else printf("SPU2null wrong format\n"); + else + { + printf("SPU2null wrong format\n"); + } } else if (mode == FREEZE_SAVE) diff --git a/plugins/SPU2null/SPU2.h b/plugins/SPU2null/SPU2.h index 4b6b1bf7b5..eb65db59ef 100644 --- a/plugins/SPU2null/SPU2.h +++ b/plugins/SPU2null/SPU2.h @@ -46,15 +46,15 @@ extern "C" extern FILE *spu2Log; #define SPU2_LOG __Log //debug mode -extern const unsigned char version; -extern const unsigned char revision; -extern const unsigned char build; -extern const unsigned int minor; +extern const u8 version; +extern const u8 revision; +extern const u8 build; +extern const u32 minor; extern char *libraryName; typedef struct { - int Log; + s32 Log; } Config; extern Config conf; @@ -218,18 +218,18 @@ struct _SPU_VOICE // ADSR INFOS PER CHANNEL struct ADSRInfoEx { - int State; - int AttackModeExp; - int AttackRate; - int DecayRate; - int SustainLevel; - int SustainModeExp; - int SustainIncrease; - int SustainRate; - int ReleaseModeExp; - int ReleaseRate; - int EnvelopeVol; - long lVolume; + s32 State; + s32 AttackModeExp; + s32 AttackRate; + s32 DecayRate; + s32 SustainLevel; + s32 SustainModeExp; + s32 SustainIncrease; + s32 SustainRate; + s32 ReleaseModeExp; + s32 ReleaseRate; + s32 EnvelopeVol; + s32 lVolume; }; #define NSSIZE 48 // ~ 1 ms of data @@ -257,18 +257,18 @@ struct VOICE_PROCESSED // start save state - int iSBPos; // mixing stuff - int spos; - int sinc; + s32 iSBPos; // mixing stuff + s32 spos; + s32 sinc; - int iActFreq; // current psx pitch - int iUsedFreq; // current pc pitch + s32 iActFreq; // current psx pitch + s32 iUsedFreq; // current pc pitch - int iStartAddr, iLoopAddr, iNextAddr; + s32 iStartAddr, iLoopAddr, iNextAddr; ADSRInfoEx ADSRX; // next ADSR settings (will be moved to active on sample start) bool bIgnoreLoop, bNew, bNoise, bReverb, bOn, bStop, bVolChanged; - int memoffset; // if first core, 0, if second, 0x400 + s32 memoffset; // if first core, 0, if second, 0x400 // end save state @@ -283,11 +283,11 @@ struct VOICE_PROCESSED struct ADMA { - unsigned short * MemAddr; - long IntPointer; - int Index; - int AmountLeft; - int Enabled; + u16 * MemAddr; + s32 IntPointer; + s32 Index; + s32 AmountLeft; + s32 Enabled; }; #endif /* __SPU2_H__ */ diff --git a/plugins/spu2-x/src/Makefile.am b/plugins/spu2-x/src/Makefile.am index 41449973a1..ffd70c5832 100644 --- a/plugins/spu2-x/src/Makefile.am +++ b/plugins/spu2-x/src/Makefile.am @@ -6,11 +6,6 @@ INCLUDES = -I@srcdir@/../../../common/include -I@srcdir@/3rdparty -I@srcdir@/.. libSPU2X_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) -liconv libSPU2X_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) -liconv -if X86_64 -libSPU2X_a_CXXFLAGS += -fPIC -libSPU2X_a_CFLAGS += -fPIC -endif - # Create a shared object by faking an exe (thanks to ODE makefiles) traplibdir=$(prefix) diff --git a/plugins/spu2-x/src/configure.ac b/plugins/spu2-x/src/configure.ac index 37d1dd2f14..feacc310e1 100644 --- a/plugins/spu2-x/src/configure.ac +++ b/plugins/spu2-x/src/configure.ac @@ -63,32 +63,6 @@ dnl gtk AC_MSG_CHECKING(gtk2+) AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) LIBS+=$(pkg-config --libs gtk+-2.0) -dnl Check for 64bit CPU -AC_MSG_CHECKING(for a x86-64 CPU) -dnl if test "$build_os" == "target_os" -dnl then -AC_TRY_RUN([ -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} -],cpu64=yes,cpu64=no,) -dnl else -dnl cpu64=no -dnl fi -if test "x$cpu64" == xyes -then -AC_DEFINE(__x86_64__,1,[__x86_64__]) -fi -AC_MSG_RESULT($cpu64) -AM_CONDITIONAL(X86_64, test x$cpu64 = xyes) dnl bindir = pcsx2exe diff --git a/plugins/zerogs/opengl/Linux/Makefile.am b/plugins/zerogs/opengl/Linux/Makefile.am index aa98f152c5..e1631ec3ac 100644 --- a/plugins/zerogs/opengl/Linux/Makefile.am +++ b/plugins/zerogs/opengl/Linux/Makefile.am @@ -1,10 +1,4 @@ noinst_LIBRARIES = libZeroGSLinux.a -if X86_64 -libZeroGSLinux_a_CPPFLAGS = -fPIC -libZeroGSLinux_a_CXXFLAGS = -fPIC -libZeroGSLinux_a_CFLAGS = -fPIC -endif - INCLUDES = $(shell pkg-config --cflags gtk+-2.0) -I@srcdir@/../ -I@srcdir@/../common libZeroGSLinux_a_SOURCES = callbacks.c Conf.cpp interface.c Linux.cpp support.c diff --git a/plugins/zerogs/opengl/Makefile.am b/plugins/zerogs/opengl/Makefile.am index c66a43a85b..d84ca8397b 100644 --- a/plugins/zerogs/opengl/Makefile.am +++ b/plugins/zerogs/opengl/Makefile.am @@ -6,16 +6,9 @@ INCLUDES = -I@srcdir@/../../../common/include -I@srcdir@/../../../3rdparty libZeroGSogl_a_CPPFLAGS = $(shell pkg-config --cflags gtk+-2.0) libZeroGSogl_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) -if X86_64 -libZeroGSogl_a_CPPFLAGS += -fPIC -libZeroGSogl_a_CXXFLAGS += -fPIC -CCASFLAGS += -fPIC -endif - # Create a shared object by faking an exe (thanks to ODE makefiles) traplibdir=$(prefix) - if RELEASE_TO_PUBLIC preext=r endif @@ -33,11 +26,7 @@ libZeroGSogl_a_SOURCES = \ GSmain.cpp memcpy_amd.cpp Regs.cpp x86.cpp zpipe.cpp \ Mem.cpp rasterfont.cpp targets.cpp zerogs.cpp -if X86_64 -libZeroGSogl_a_SOURCES += x86-64.S -else libZeroGSogl_a_SOURCES += x86-32.S -endif if SSE2 CCASFLAGS+= -DZEROGS_SSE2 diff --git a/plugins/zerogs/opengl/configure b/plugins/zerogs/opengl/configure index c0c7bb22fe..e81eb7f7f7 100644 --- a/plugins/zerogs/opengl/configure +++ b/plugins/zerogs/opengl/configure @@ -643,8 +643,6 @@ GREP SHARED_LDFLAGS so_ext GTK_CONFIG -X86_64_FALSE -X86_64_TRUE SSE2_FALSE SSE2_TRUE RELEASE_TO_PUBLIC_FALSE @@ -4627,92 +4625,6 @@ else fi -{ $as_echo "$as_me:$LINENO: checking for a x86-64 CPU" >&5 -$as_echo_n "checking for a x86-64 CPU... " >&6; } -if test "$cross_compiling" = yes; then - { { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: cannot run test program while cross compiling -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: cannot run test program while cross compiling -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} - -_ACEOF -rm -f conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { ac_try='./conftest$ac_exeext' - { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; }; then - cpu64=yes -else - $as_echo "$as_me: program exited with status $ac_status" >&5 -$as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - -( exit $ac_status ) -cpu64=no -fi -rm -rf conftest.dSYM -rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext conftest.$ac_objext conftest.$ac_ext -fi - - -if test "x$cpu64" == xyes -then - -cat >>confdefs.h <<\_ACEOF -#define __x86_64__ 1 -_ACEOF - -fi -{ $as_echo "$as_me:$LINENO: result: $cpu64" >&5 -$as_echo "$cpu64" >&6; } - if test x$cpu64 = xyes; then - X86_64_TRUE= - X86_64_FALSE='#' -else - X86_64_TRUE='#' - X86_64_FALSE= -fi - - { $as_echo "$as_me:$LINENO: checking gtk2+" >&5 $as_echo_n "checking gtk2+... " >&6; } # Extract the first word of "pkg-config", so it can be a program name with args. @@ -6127,13 +6039,6 @@ $as_echo "$as_me: error: conditional \"SSE2\" was never defined. Usually this means the macro was only invoked conditionally." >&2;} { (exit 1); exit 1; }; } fi -if test -z "${X86_64_TRUE}" && test -z "${X86_64_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"X86_64\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"X86_64\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi : ${CONFIG_STATUS=./config.status} ac_write_fail=0 diff --git a/plugins/zerogs/opengl/configure.ac b/plugins/zerogs/opengl/configure.ac index 040f6f7de5..c097cedbec 100644 --- a/plugins/zerogs/opengl/configure.ac +++ b/plugins/zerogs/opengl/configure.ac @@ -68,33 +68,6 @@ fi AC_MSG_RESULT($sse2) AM_CONDITIONAL(SSE2, test x$sse2 = xyes) -dnl Check for 64bit CPU -AC_MSG_CHECKING(for a x86-64 CPU) -dnl if test "$build_os" == "target_os" -dnl then -AC_TRY_RUN([ -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} -],cpu64=yes,cpu64=no,) -dnl else -dnl cpu64=no -dnl fi -if test "x$cpu64" == xyes -then -AC_DEFINE(__x86_64__,1,[__x86_64__]) -fi -AC_MSG_RESULT($cpu64) -AM_CONDITIONAL(X86_64, test x$cpu64 = xyes) - dnl gtk AC_MSG_CHECKING(gtk2+) AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) diff --git a/plugins/zeropad/Makefile.am b/plugins/zeropad/Makefile.am index 9376619c69..987d3f3dd2 100644 --- a/plugins/zeropad/Makefile.am +++ b/plugins/zeropad/Makefile.am @@ -6,11 +6,6 @@ INCLUDES = -I@srcdir@/../../common/include libZeroPAD_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) libZeroPAD_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) -if X86_64 -libZeroPAD_a_CXXFLAGS += -fPIC -libZeroPAD_a_CFLAGS += -fPIC -endif - # Create a shared object by faking an exe (thanks to ODE makefiles) traplibdir=$(prefix) diff --git a/plugins/zeropad/configure.ac b/plugins/zeropad/configure.ac index 95b38496c2..db2f1a4f10 100644 --- a/plugins/zeropad/configure.ac +++ b/plugins/zeropad/configure.ac @@ -41,32 +41,6 @@ fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) -dnl Check for 64bit CPU -AC_MSG_CHECKING(for a x86-64 CPU) -dnl if test "$build_os" == "target_os" -dnl then -AC_TRY_RUN([ -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} -],cpu64=yes,cpu64=no,) -dnl else -dnl cpu64=no -dnl fi -if test "x$cpu64" == xyes -then -AC_DEFINE(__x86_64__,1,[__x86_64__]) -fi -AC_MSG_RESULT($cpu64) -AM_CONDITIONAL(X86_64, test x$cpu64 = xyes) AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) diff --git a/plugins/zerospu2/Makefile.am b/plugins/zerospu2/Makefile.am index 54d3daf5c5..3d6a00768a 100644 --- a/plugins/zerospu2/Makefile.am +++ b/plugins/zerospu2/Makefile.am @@ -6,11 +6,6 @@ INCLUDES = -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty libZeroSPU2_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) libZeroSPU2_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) -if X86_64 -libZeroSPU2_a_CXXFLAGS += -fPIC -libZeroSPU2_a_CFLAGS += -fPIC -endif - # Create a shared object by faking an exe (thanks to ODE makefiles) traplibdir=$(prefix) diff --git a/plugins/zerospu2/configure.ac b/plugins/zerospu2/configure.ac index b0b2ce69cb..613f9be657 100644 --- a/plugins/zerospu2/configure.ac +++ b/plugins/zerospu2/configure.ac @@ -55,33 +55,6 @@ fi AC_MSG_RESULT($devbuild) AM_CONDITIONAL(RELEASE_TO_PUBLIC, test x$devbuild = xno) -dnl Check for 64bit CPU -AC_MSG_CHECKING(for a x86-64 CPU) -dnl if test "$build_os" == "target_os" -dnl then -AC_TRY_RUN([ -int main() -{ -int a = 0; -int*pa = &a; -asm(".intel_syntax\n" - "mov %%rax, %0\n" - "mov %%eax, [%%rax]\n" - ".att_syntax\n" - : : "r"(pa) : "%rax"); -return 0; -} -],cpu64=yes,cpu64=no,) -dnl else -dnl cpu64=no -dnl fi -if test "x$cpu64" == xyes -then -AC_DEFINE(__x86_64__,1,[__x86_64__]) -fi -AC_MSG_RESULT($cpu64) -AM_CONDITIONAL(X86_64, test x$cpu64 = xyes) - AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) dnl gtk diff --git a/plugins/zerospu2/misc.h b/plugins/zerospu2/misc.h index 17c3426b45..b1bf3b8315 100644 --- a/plugins/zerospu2/misc.h +++ b/plugins/zerospu2/misc.h @@ -23,7 +23,7 @@ #define Sleep(ms) usleep(1000*ms) -inline unsigned long timeGetTime() +inline u32 timeGetTime() { #ifdef _WIN32 _timeb t; @@ -33,7 +33,7 @@ inline unsigned long timeGetTime() ftime(&t); #endif - return (unsigned long)(t.time*1000+t.millitm); + return (u32)(t.time*1000+t.millitm); } #include diff --git a/plugins/zerospu2/voices.cpp b/plugins/zerospu2/voices.cpp index d46ee1592f..bdf8e5e2f2 100644 --- a/plugins/zerospu2/voices.cpp +++ b/plugins/zerospu2/voices.cpp @@ -31,7 +31,7 @@ SPU_CONTROL_* VOICE_PROCESSED::GetCtrl() return ((SPU_CONTROL_*)(spu2regs+memoffset+REG_C0_CTRL)); } -void VOICE_PROCESSED::SetVolume(int iProcessRight) +void VOICE_PROCESSED::SetVolume(s32 iProcessRight) { u16 vol = iProcessRight ? pvoice->right.word : pvoice->left.word; @@ -101,8 +101,8 @@ void VOICE_PROCESSED::InterpolateUp() { if (SB[32]==1) // flag == 1? calc step and set flag... and don't change the value in this pass { - const int id1=SB[30]-SB[29]; // curr delta to next val - const int id2=SB[31]-SB[30]; // and next delta to next-next val :) + const s32 id1=SB[30]-SB[29]; // curr delta to next val + const s32 id2=SB[31]-SB[30]; // and next delta to next-next val :) SB[32]=0; @@ -159,9 +159,9 @@ void VOICE_PROCESSED::InterpolateDown() } } -void VOICE_PROCESSED::FModChangeFrequency(int ns) +void VOICE_PROCESSED::FModChangeFrequency(s32 ns) { - int NP=pvoice->pitch; + s32 NP=pvoice->pitch; NP=((32768L+iFMod[ns])*NP)/32768L; @@ -173,8 +173,7 @@ void VOICE_PROCESSED::FModChangeFrequency(int ns) iActFreq=NP; iUsedFreq=NP; sinc=(((NP/10)<<16)/4800); - if (!sinc) - sinc=1; + if (!sinc) sinc=1; // freq change in simple interpolation mode SB[32]=1; @@ -188,7 +187,7 @@ static void __forceinline GetNoiseValues(s32& VD) if(Seed&0x100) VD = (s32)((Seed&0xff)<<8); - else if(!(Seed&0xffff)) + else if (!(Seed&0xffff)) VD = (s32)0x8000; else VD = (s32)0x7fff; @@ -226,7 +225,7 @@ static void __forceinline GetNoiseValues(s32& VD) // and sometimes the noise will be used as fmod modulation... pfff int VOICE_PROCESSED::iGetNoiseVal() { - int fa; + s32 fa; /*if ((dwNoiseVal<<=1)&0x80000000L) { @@ -248,7 +247,7 @@ int VOICE_PROCESSED::iGetNoiseVal() return fa; } -void VOICE_PROCESSED::StoreInterpolationVal(int fa) +void VOICE_PROCESSED::StoreInterpolationVal(s32 fa) { if (bFMod==2) // fmod freq channel SB[29]=fa; @@ -269,9 +268,9 @@ void VOICE_PROCESSED::StoreInterpolationVal(int fa) } } -int VOICE_PROCESSED::iGetInterpolationVal() +s32 VOICE_PROCESSED::iGetInterpolationVal() { - int fa; + s32 fa; if (bFMod==2) return SB[29]; diff --git a/plugins/zerospu2/zerodma.cpp b/plugins/zerospu2/zerodma.cpp index 85cc6a4a2c..e25c6027c0 100644 --- a/plugins/zerospu2/zerodma.cpp +++ b/plugins/zerospu2/zerodma.cpp @@ -27,7 +27,7 @@ void CALLBACK SPU2readDMAMem(u16 *pMem, int size, int core) { u32 spuaddr; - int i, dma, offset; + s32 i, dma, offset; if ( core == 0) { @@ -44,7 +44,7 @@ void CALLBACK SPU2readDMAMem(u16 *pMem, int size, int core) SPU2_LOG("SPU2 readDMA%dMem size %x, addr: %x\n", dma, size, pMem); - for (i=0; i < size; i++) + for (i = 0; i < size; i++) { *pMem++ = *(u16*)(spu2mem + spuaddr); if ((spu2Rs16(REG_C0_CTRL + offset) & 0x40) && (C_IRQA(core) == spuaddr)) @@ -94,7 +94,7 @@ int ADMASWrite(int core) { u32 spuaddr; ADMA *Adma; - int dma, offset; + s32 dma, offset; if (core == 0) { @@ -159,7 +159,7 @@ void CALLBACK SPU2writeDMAMem(u16* pMem, int size, int core) { u32 spuaddr; ADMA *Adma; - int dma, offset; + s32 dma, offset; if (core == 0) { @@ -240,7 +240,7 @@ void CALLBACK SPU2writeDMA7Mem(u16* pMem, int size) void CALLBACK SPU2interruptDMA(int core) { - int dma, offset; + s32 dma, offset; if (core == 0) { diff --git a/plugins/zerospu2/zerospu2.cpp b/plugins/zerospu2/zerospu2.cpp index 337b7828b8..a0b5351572 100644 --- a/plugins/zerospu2/zerospu2.cpp +++ b/plugins/zerospu2/zerospu2.cpp @@ -44,14 +44,14 @@ u16* spu2mem = NULL; u16* pSpuIrq[2] = {NULL}; u32 dwNewChannel2[2] = {0}; // keeps track of what channels that have been turned on u32 dwEndChannel2[2] = {0}; // keeps track of what channels have ended -unsigned long dwNoiseVal=1; // global noise generator +u32 dwNoiseVal=1; // global noise generator bool g_bPlaySound = true; // if true, will output sound, otherwise no -int iFMod[NSSIZE]; -int s_buffers[NSSIZE][2]; // left and right buffers +s32 iFMod[NSSIZE]; +s32 s_buffers[NSSIZE][2]; // left and right buffers // mixer thread variables static bool s_bThreadExit = true; -static int s_nDropPacket = 0; +static s32 s_nDropPacket = 0; string s_strIniPath="inis/zerospu2.ini"; #ifdef _WIN32 @@ -66,7 +66,7 @@ void* SPU2ThreadProc(void*); #endif static AUDIOBUFFER s_pAudioBuffers[NSPACKETS]; -static int s_nCurBuffer = 0, s_nQueuedBuffers = 0; +static s32 s_nCurBuffer = 0, s_nQueuedBuffers = 0; static s16* s_pCurOutput = NULL; static u32 g_startcount=0xffffffff; static u32 g_packetcount=0; @@ -76,15 +76,15 @@ soundtouch::SoundTouch* pSoundTouch=NULL; WavOutFile* g_pWavRecord=NULL; // used for recording static u64 s_GlobalTimeStamp = 0; -static int s_nDurations[64]={0}; -static int s_nCurDuration=0; -static int s_nTotalDuration=0; +static s32 s_nDurations[64]={0}; +static s32 s_nCurDuration=0; +static s32 s_nTotalDuration=0; -int SPUCycles = 0, SPUWorkerCycles = 0; -int SPUStartCycle[2]; -int SPUTargetCycle[2]; +s32 SPUCycles = 0, SPUWorkerCycles = 0; +s32 SPUStartCycle[2]; +s32 SPUTargetCycle[2]; -int g_logsound=0; +s32 g_logsound=0; int ADMASWrite(int c); @@ -97,7 +97,7 @@ void (*irqCallbackDMA7)()=0; uptr g_pDMABaseAddr=0; -const int f[5][2] = { +const s32 f[5][2] = { { 0, 0 }, { 60, 0 }, { 115, -52 }, @@ -239,7 +239,7 @@ s32 CALLBACK SPU2init() memset(voices, 0, sizeof(voices)); // last 24 channels have higher mem offset - for (int i = 0; i < 24; ++i) + for (s32 i = 0; i < 24; ++i) voices[i+24].memoffset = 0x400; // init each channel @@ -305,7 +305,7 @@ s32 CALLBACK SPU2open(void *pDsp) s_pCurOutput = (s16*)s_pAudioBuffers[0].pbuf; assert( s_pCurOutput != NULL); - for (int i = 0; i < ArraySize(s_nDurations); ++i) + for (s32 i = 0; i < ArraySize(s_nDurations); ++i) { s_nDurations[i] = NSFRAMES*1000; } @@ -410,7 +410,7 @@ void CALLBACK SPU2async(u32 cycle) void InitADSR() // INIT ADSR { u32 r,rs,rd; - int i; + s32 i; memset(RateTable,0,sizeof(u32)*160); // build the rate table according to Neill's rules (see at bottom of file) r=3;rs=1;rd=0; @@ -434,7 +434,7 @@ void InitADSR() // INIT ADSR } } -int MixADSR(VOICE_PROCESSED* pvoice) // MIX ADSR +s32 MixADSR(VOICE_PROCESSED* pvoice) // MIX ADSR { u32 rateadd[8] = { 0, 4, 6, 8, 9, 10, 11, 12 }; @@ -536,11 +536,11 @@ int MixADSR(VOICE_PROCESSED* pvoice) // MIX ADSR return 0; } -void MixChannels(int core) +void MixChannels(s32 core) { // mix all channels - int c_offset = 0x0400 * core; - int dma, left_vol, right_vol; + s32 c_offset = 0x0400 * core; + s32 dma, left_vol, right_vol; ADMA *Adma; if (core == 0) @@ -560,12 +560,12 @@ void MixChannels(int core) if ((spu2Ru16(REG_C0_MMIX + c_offset) & 0xF0) && (spu2Ru16(REG_C0_ADMAS + c_offset) & (0x1 + core))) { - for (int ns=0;nsIndex]*(int)spu2Ru16(left_vol))>>16; + s_buffers[ns][0] += (((s16*)spu2mem)[0x2000 + c_offset +Adma->Index]*(s32)spu2Ru16(left_vol))>>16; if ((spu2Ru16(REG_C0_MMIX + c_offset) & 0x40)) - s_buffers[ns][1] += (((short*)spu2mem)[0x2200 + c_offset +Adma->Index]*(int)spu2Ru16(right_vol))>>16; + s_buffers[ns][1] += (((s16*)spu2mem)[0x2200 + c_offset +Adma->Index]*(s32)spu2Ru16(right_vol))>>16; Adma->Index +=1; MemAddr[core] += 4; @@ -602,7 +602,7 @@ void SPU2Worker() { u8* start; u32 nSample; - int ch, predict_nr, shift_factor, flags; + s32 ch, predict_nr, shift_factor, flags; // assume s_buffers are zeroed out if ( dwNewChannel2[0] || dwNewChannel2[1] ) @@ -624,11 +624,11 @@ void SPU2Worker() pChannel->VoiceChangeFrequency(); // loop until 1 ms of data is reached - int ns = 0; + s32 ns = 0; while(nsbFMod==1 && iFMod[ns]) pChannel->FModChangeFrequency(ns); @@ -662,8 +662,8 @@ void SPU2Worker() for (nSample=0; nSample<28; ++start) { - int s; - int d = (int)*start; + s32 s; + s32 d = (s32)*start; s = ((d & 0xf)<<12); if (s & 0x8000) s |= 0xffff0000; @@ -685,7 +685,7 @@ void SPU2Worker() } // irq occurs no matter what core access the address - for (int core = 0; core < 2; ++core) + for (s32 core = 0; core < 2; ++core) { if (((SPU_CONTROL_*)(spu2regs + (0x400 * core) + REG_C0_CTRL))->irq) // some callback and irq active? { @@ -736,7 +736,7 @@ void SPU2Worker() else fa=pChannel->iGetInterpolationVal(); // get sample val - int sval = (MixADSR(pChannel) * fa) / 1023; // mix adsr + s32 sval = (MixADSR(pChannel) * fa) / 1023; // mix adsr if (pChannel->bFMod == 2) // fmod freq channel { @@ -767,7 +767,7 @@ ENDX: { assert( s_pCurOutput != NULL); - for (int ns=0; ns 5000) { WARN_LOG("ZeroSPU2: recording\n"); @@ -840,17 +840,17 @@ ENDX: } // resamples pStereoSamples -void ResampleLinear(s16* pStereoSamples, int oldsamples, s16* pNewSamples, int newsamples) +void ResampleLinear(s16* pStereoSamples, s32 oldsamples, s16* pNewSamples, s32 newsamples) { - for (int i = 0; i < newsamples; ++i) + for (s32 i = 0; i < newsamples; ++i) { - int io = i * oldsamples; - int old = io / newsamples; - int rem = io - old * newsamples; + s32 io = i * oldsamples; + s32 old = io / newsamples; + s32 rem = io - old * newsamples; old *= 2; - int newsampL = pStereoSamples[old] * (newsamples - rem) + pStereoSamples[old+2] * rem; - int newsampR = pStereoSamples[old+1] * (newsamples - rem) + pStereoSamples[old+3] * rem; + s32 newsampL = pStereoSamples[old] * (newsamples - rem) + pStereoSamples[old+2] * rem; + s32 newsampR = pStereoSamples[old+1] * (newsamples - rem) + pStereoSamples[old+3] * rem; pNewSamples[2 * i] = newsampL / newsamples; pNewSamples[2 * i + 1] = newsampR / newsamples; } @@ -868,7 +868,7 @@ DWORD WINAPI SPU2ThreadProc(LPVOID) void* SPU2ThreadProc(void* lpParam) #endif { - int nReadBuf = 0; + s32 nReadBuf = 0; while (!s_bThreadExit) { @@ -901,13 +901,13 @@ void* SPU2ThreadProc(void* lpParam) } - //int ps2delay = timeGetTime() - s_pAudioBuffers[nReadBuf].timestamp; - int NewSamples = s_pAudioBuffers[nReadBuf].avgtime; + //s32 ps2delay = timeGetTime() - s_pAudioBuffers[nReadBuf].timestamp; + s32 NewSamples = s_pAudioBuffers[nReadBuf].avgtime; if ( (conf.options & OPTION_TIMESTRETCH) ) { - int bytesbuf = SoundGetBytesBuffered(); + s32 bytesbuf = SoundGetBytesBuffered(); if ( bytesbuf < 8000 ) NewSamples += 1000; // check the current timestamp, if too far apart, speed up audio @@ -928,25 +928,25 @@ void* SPU2ThreadProc(void* lpParam) NewSamples = min(NewSamples, NSFRAMES * NSSIZE * 3); - int oldsamples = s_pAudioBuffers[nReadBuf].len / 4; + s32 oldsamples = s_pAudioBuffers[nReadBuf].len / 4; if ((nReadBuf & 3) == 0) // wow, this if statement makes the whole difference pSoundTouch->setTempoChange(100.0f*(float)oldsamples/(float)NewSamples - 100.0f); - for( int sx=0; sxputSamples(s_floatBuffer, oldsamples); // extract 2*NSFRAMES ms at a time - int nOutSamples; + s32 nOutSamples; do { nOutSamples = pSoundTouch->receiveSamples(s_floatBuffer, NSSIZE * NSFRAMES * 5); if ( nOutSamples > 0 ) { - for( int sx=0; sx>=1) // loop channels + for (s32 ch=start;ch>=1) // loop channels { if ((val&1) && voices[ch].pStart) // mmm... start has to be set before key on !?! { @@ -983,17 +983,17 @@ void SoundOn(int start,int end,unsigned short val) // SOUND ON PSX COMAND } // turn channels off -void SoundOff(int start,int end,unsigned short val) // SOUND OFF PSX COMMAND +void SoundOff(s32 start,s32 end,u16 val) // SOUND OFF PSX COMMAND { - for (int ch=start;ch>=1) // loop channels + for (s32 ch=start;ch>=1) // loop channels { if (val&1) voices[ch].bStop=true; // && s_chan[i].bOn) mmm... } } -void FModOn(int start,int end,unsigned short val) // FMOD ON PSX COMMAND +void FModOn(s32 start,s32 end,u16 val) // FMOD ON PSX COMMAND { - int ch; + s32 ch; for (ch=start;ch>=1) // loop channels { @@ -1010,9 +1010,9 @@ void FModOn(int start,int end,unsigned short val) // FMOD ON PSX COMMAND } } -void VolumeOn(int start,int end,unsigned short val,int iRight) // VOLUME ON PSX COMMAND +void VolumeOn(s32 start,s32 end,u16 val,s32 iRight) // VOLUME ON PSX COMMAND { - int ch; + s32 ch; for (ch=start;ch>=1) // loop channels { @@ -1048,7 +1048,7 @@ void CALLBACK SPU2write(u32 mem, u16 value) // channel info if ((r<0x0180) || (r>=0x0400 && r<0x0580)) // u32s are always >= 0. { - int ch=0; + s32 ch=0; if (r >= 0x400) ch = ((r - 0x400) >> 4) + 24; else @@ -1064,7 +1064,7 @@ void CALLBACK SPU2write(u32 mem, u16 value) break; case 4: { - int NP; + s32 NP; if (value> 0x3fff) NP=0x3fff; // get pitch val else @@ -1100,8 +1100,8 @@ void CALLBACK SPU2write(u32 mem, u16 value) // more channel info if ((r>=0x01c0 && r<0x02E0)||(r>=0x05c0 && r<0x06E0)) { - int ch=0; - unsigned long rx=r; + s32 ch=0; + u32 rx=r; if (rx>=0x400) { ch=24; @@ -1257,7 +1257,7 @@ u16 CALLBACK SPU2read(u32 mem) // if the register is any of the regs before core 0, or is somewhere between core 0 and 1... if ((r < 0x0180) || (r >= 0x0400 && r < 0x0580)) // u32s are always >= 0. { - int ch = 0; + s32 ch = 0; if (r >= 0x400) ch=((r - 0x400) >> 4) + 24; @@ -1272,8 +1272,8 @@ u16 CALLBACK SPU2read(u32 mem) if ((r>=REG_VA_SSA && r=0x05c0 && r<0x06E0)) // some channel info? { - int ch=0; - unsigned long rx = r; + s32 ch=0; + u32 rx = r; if (rx >=0x400) { @@ -1390,24 +1390,24 @@ s32 CALLBACK SPU2test() } // size is in bytes -void LogPacketSound(void* packet, int memsize) +void LogPacketSound(void* packet, s32 memsize) { u16 buf[28]; u8* pstart = (u8*)packet; - int s_1 = 0, s_2=0; + s32 s_1 = 0, s_2=0; - for (int i = 0; i < memsize; i += 16) + for (s32 i = 0; i < memsize; i += 16) { - int predict_nr=(int)pstart[0]; - int shift_factor=predict_nr&0xf; + s32 predict_nr=(s32)pstart[0]; + s32 shift_factor=predict_nr&0xf; predict_nr >>= 4; pstart += 2; - for (int nSample=0;nSample<28; ++pstart) + for (s32 nSample=0;nSample<28; ++pstart) { - int d=(int)*pstart; - int s, fa; + s32 d=(s32)*pstart; + s32 s, fa; s =((d & 0xf) << 12); SetPacket(s); @@ -1420,7 +1420,7 @@ void LogPacketSound(void* packet, int memsize) } } -void LogRawSound(void* pleft, int leftstride, void* pright, int rightstride, int numsamples) +void LogRawSound(void* pleft, s32 leftstride, void* pright, s32 rightstride, s32 numsamples) { if (g_pWavRecord == NULL ) g_pWavRecord = new WavOutFile(RECORD_FILENAME, SAMPLE_RATE, 16, 2); @@ -1431,7 +1431,7 @@ void LogRawSound(void* pleft, int leftstride, void* pright, int rightstride, int tempbuf.resize(2 * numsamples); - for (int i = 0; i < numsamples; ++i) + for (s32 i = 0; i < numsamples; ++i) { tempbuf[2*i+0] = *(s16*)left; tempbuf[2*i+1] = *(s16*)right; @@ -1462,7 +1462,7 @@ int CALLBACK SPU2setupRecording(int start, void* pData) void save_data(freezeData *data) { SPU2freezeData *spud; - int i; + s32 i; spud = (SPU2freezeData*)data->data; spud->version = 0x70000001; @@ -1470,8 +1470,8 @@ void save_data(freezeData *data) memcpy(spud->spu2regs, spu2regs, 0x10000); memcpy(spud->spu2mem, spu2mem, 0x200000); - spud->nSpuIrq[0] = (int)(pSpuIrq[0] - spu2mem); - spud->nSpuIrq[1] = (int)(pSpuIrq[1] - spu2mem); + spud->nSpuIrq[0] = (s32)(pSpuIrq[0] - spu2mem); + spud->nSpuIrq[1] = (s32)(pSpuIrq[1] - spu2mem); memcpy(spud->dwNewChannel2, dwNewChannel2, 4*2); memcpy(spud->dwEndChannel2, dwEndChannel2, 4*2); @@ -1518,7 +1518,7 @@ void save_data(freezeData *data) void load_data(freezeData *data) { SPU2freezeData *spud; - int i; + s32 i; spud = (SPU2freezeData*)data->data; @@ -1556,7 +1556,7 @@ void load_data(freezeData *data) for (i = 0; i < ArraySize(voices); ++i) { - memcpy(&voices[i], &spud->voices[i], min((int)SPU_VOICE_STATE_SIZE, spud->voicesize)); + memcpy(&voices[i], &spud->voices[i], min((s32)SPU_VOICE_STATE_SIZE, spud->voicesize)); voices[i].pStart = (u8*)((uptr)spud->voices[i].pStart+(uptr)spu2mem); voices[i].pLoop = (u8*)((uptr)spud->voices[i].pLoop+(uptr)spu2mem); voices[i].pCurr = (u8*)((uptr)spud->voices[i].pCurr+(uptr)spu2mem); @@ -1565,7 +1565,7 @@ void load_data(freezeData *data) s_GlobalTimeStamp = 0; g_startcount = 0xffffffff; - for (int i = 0; i < ArraySize(s_nDurations); ++i) + for (s32 i = 0; i < ArraySize(s_nDurations); ++i) { s_nDurations[i] = NSFRAMES*1000; } diff --git a/plugins/zerospu2/zerospu2.h b/plugins/zerospu2/zerospu2.h index d97b9e112c..37698ddf1c 100644 --- a/plugins/zerospu2/zerospu2.h +++ b/plugins/zerospu2/zerospu2.h @@ -88,24 +88,24 @@ extern FILE *spu2Log; extern s8 *spu2regs; extern u16* spu2mem; -extern int iFMod[NSSIZE]; +extern s32 iFMod[NSSIZE]; extern u32 MemAddr[2]; -extern unsigned long dwNoiseVal; // global noise generator +extern u32 dwNoiseVal; // global noise generator // functions of main emu, called on spu irq extern void (*irqCallbackSPU2)(); extern void (*irqCallbackDMA4)(); extern void (*irqCallbackDMA7)(); -extern int SPUCycles, SPUWorkerCycles; -extern int SPUStartCycle[2]; -extern int SPUTargetCycle[2]; +extern s32 SPUCycles, SPUWorkerCycles; +extern s32 SPUStartCycle[2]; +extern s32 SPUTargetCycle[2]; extern u16 interrupt; typedef struct { - int Log; - int options; + s32 Log; + s32 options; } Config; extern Config conf; @@ -173,7 +173,7 @@ static __forceinline u32 C1_IRQA() return SPU2_GET32BIT(REG_C1_IRQA_LO, REG_C1_IRQA_HI); } -static __forceinline u32 C_IRQA(int c) +static __forceinline u32 C_IRQA(s32 c) { if (c == 0) return C0_IRQA(); @@ -191,7 +191,7 @@ static __forceinline u32 C1_SPUADDR() return SPU2_GET32BIT(REG_C1_SPUADDR_LO, REG_C1_SPUADDR_HI); } -static __forceinline u32 C_SPUADDR(int c) +static __forceinline u32 C_SPUADDR(s32 c) { if (c == 0) return C0_SPUADDR(); @@ -209,7 +209,7 @@ static __forceinline void C1_SPUADDR_SET(u32 value) SPU2_SET32BIT(value, REG_C1_SPUADDR_LO, REG_C1_SPUADDR_HI); } -static __forceinline void C_SPUADDR_SET(u32 value, int c) +static __forceinline void C_SPUADDR_SET(u32 value, s32 c) { if (c == 0) C0_SPUADDR_SET(value); @@ -282,18 +282,18 @@ struct _SPU_VOICE // ADSR INFOS PER CHANNEL struct ADSRInfoEx { - int State; - int AttackModeExp; - int AttackRate; - int DecayRate; - int SustainLevel; - int SustainModeExp; - int SustainIncrease; - int SustainRate; - int ReleaseModeExp; - int ReleaseRate; - int EnvelopeVol; - long lVolume; + s32 State; + s32 AttackModeExp; + s32 AttackRate; + s32 DecayRate; + s32 SustainLevel; + s32 SustainModeExp; + s32 SustainIncrease; + s32 SustainRate; + s32 ReleaseModeExp; + s32 ReleaseRate; + s32 EnvelopeVol; + s32 lVolume; }; #define SPU_VOICE_STATE_SIZE (sizeof(VOICE_PROCESSED)-4*sizeof(void*)) @@ -311,34 +311,34 @@ struct VOICE_PROCESSED void InterpolateUp(); void InterpolateDown(); void FModChangeFrequency(int ns); - int iGetNoiseVal(); + s32 iGetNoiseVal(); void StoreInterpolationVal(int fa); - int iGetInterpolationVal(); + s32 iGetInterpolationVal(); void Stop(); SPU_CONTROL_* GetCtrl(); // start save state - int leftvol, rightvol; // left right volumes + s32 leftvol, rightvol; // left right volumes - int iSBPos; // mixing stuff - int SB[32+32]; - int spos; - int sinc; + s32 iSBPos; // mixing stuff + s32 SB[32+32]; + s32 spos; + s32 sinc; - int iIrqDone; // debug irq done flag - int s_1; // last decoding infos - int s_2; - int iOldNoise; // old noise val for this channel - int iActFreq; // current psx pitch - int iUsedFreq; // current pc pitch + s32 iIrqDone; // debug irq done flag + s32 s_1; // last decoding infos + s32 s_2; + s32 iOldNoise; // old noise val for this channel + s32 iActFreq; // current psx pitch + s32 iUsedFreq; // current pc pitch - int iStartAddr, iLoopAddr, iNextAddr; - int bFMod; + s32 iStartAddr, iLoopAddr, iNextAddr; + s32 bFMod; ADSRInfoEx ADSRX; // next ADSR settings (will be moved to active on sample start) - int memoffset; // if first core, 0, if second, 0x400 - int chanid; // channel id + s32 memoffset; // if first core, 0, if second, 0x400 + s32 chanid; // channel id bool bIgnoreLoop, bNew, bNoise, bReverb, bOn, bStop, bVolChanged; bool bVolumeR, bVolumeL; @@ -363,15 +363,15 @@ struct AUDIOBUFFER // Variable used to smooth out sound by concentrating on new voices u32 timestamp; // in microseconds, only used for time stretching u32 avgtime; - int newchannels; + s32 newchannels; }; struct ADMA { - unsigned short * MemAddr; - int Index; - int AmountLeft; - int Enabled; + u16* MemAddr; + s32 Index; + s32 AmountLeft; + s32 Enabled; // used to make sure that ADMA doesn't get interrupted with a writeDMA call }; @@ -384,19 +384,19 @@ struct SPU2freezeData u8 spu2regs[0x10000]; u8 spu2mem[0x200000]; u16 interrupt; - int nSpuIrq[2]; + s32 nSpuIrq[2]; u32 dwNewChannel2[2], dwEndChannel2[2]; u32 dwNoiseVal; - int iFMod[NSSIZE]; + s32 iFMod[NSSIZE]; u32 MemAddr[2]; ADMA adma[2]; u32 Adma4MemAddr, Adma7MemAddr; - int SPUCycles, SPUWorkerCycles; - int SPUStartCycle[2]; - int SPUTargetCycle[2]; + s32 SPUCycles, SPUWorkerCycles; + s32 SPUStartCycle[2]; + s32 SPUTargetCycle[2]; - int voicesize; + s32 voicesize; VOICE_PROCESSED voices[SPU_NUMBER_VOICES+1]; }; From 4b5117c078e6b2a3278307736404089c631e9273 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 18 May 2009 03:07:23 +0000 Subject: [PATCH 088/296] microVU: - Fixed a bug with P-reg's instances; this fixes problems/SPS in a lot of games such as (FFXII, NaN3, Digimon Rumble Arena 2, etc...) Note: All the games mentioned still have other problems (ffxii for example has missing geometry on characters, but looks a lot better than before :p). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1215 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Lower.inl | 4 ++-- pcsx2/x86/microVU_Misc.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index c874cdebe5..1837c7ec0a 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -261,14 +261,14 @@ microVUf(void) mVU_EEXP() { } microVUt(void) mVU_sumXYZ() { - // regd.x = x ^ 2 + y ^ 2 + z ^ 2 + // xmmPQ.x = x ^ 2 + y ^ 2 + z ^ 2 if( cpucaps.hasStreamingSIMD4Extensions ) { SSE4_DPPS_XMM_to_XMM(xmmFs, xmmFs, 0x71); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); } else { SSE_MULPS_XMM_to_XMM(xmmFs, xmmFs); // wzyx ^ 2 - SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); + SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); // x ^ 2 SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFs, 0xe1); // wzyx -> wzxy SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFs); // x ^ 2 + y ^ 2 SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFs, 0xD2); // wzxy -> wxyz diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 9e1ad49a4d..e0c0530713 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -222,7 +222,7 @@ declareAllVariables #define isSflag (mVUinfo & (1<<4)) #define writeQ ((mVUinfo >> 5) & 1) #define readQ ((mVUinfo >> 6) & 1) -#define writeP ((mVUinfo >> 7) & 1) +#define writeP (((mVUinfo >> 7) + 1) & 1) #define readP ((mVUinfo >> 7) & 1) // same as writeP #define doFlags (mVUinfo & (3<<8)) #define doMac (mVUinfo & (1<<8)) From a974cff9bd0dd5cd55770f4890a639c0fa0067dc Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 18 May 2009 06:24:12 +0000 Subject: [PATCH 089/296] microVU: - Raised the run-cycles amount. This fixes FFXII's missing geometry (the game appears perfect now :p) - mVU prints a message to console now if games are exiting early due to infinite loops (I'm interested in games that do this, so post any game you find that does this. Dark Cloud does this for example...) - Added some debug stuff to iVU1micro.cpp. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1216 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU0micro.cpp | 6 ++-- pcsx2/x86/iVU1micro.cpp | 60 +++++++++++++++++++++++++++++++++-- pcsx2/x86/microVU_Compile.inl | 9 ++++-- pcsx2/x86/microVU_Misc.h | 2 +- 4 files changed, 68 insertions(+), 9 deletions(-) diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index 8500f60665..ccdfa0580d 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -80,13 +80,13 @@ namespace VU0micro static void recStep() {} static void recExecuteBlock() { - if((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; + if ((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); //FreezeMMXRegs(1); - runVUrec(VU0.VI[REG_TPC].UL, 5000, 0); - FreezeXMMRegs(0); + runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); //FreezeMMXRegs(0); + FreezeXMMRegs(0); } } diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 4393cf9100..5eba1f8782 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -29,7 +29,49 @@ #ifdef _DEBUG extern u32 vudump; #endif + +//#define DEBUG_COMPARE +#ifdef DEBUG_COMPARE + +#include +static int runAmount = 0; + +void VUtestPause() { + + runAmount++; + if (runAmount < 100) return; + #ifndef PCSX2_MICROVU_ + SysPrintf("Super VU - Pass %d\n", runAmount); +#else + SysPrintf("Micro VU - Pass %d\n", runAmount); +#endif + + for (int i = 0; i < 32; i++) { + SysPrintf("VF%02d = {%f, %f, %f, %f}\n", i, VU1.VF[i].F[0], VU1.VF[i].F[1], VU1.VF[i].F[2], VU1.VF[i].F[3]); + } + + for (int i = 0; i < 16; i++) { + SysPrintf("VI%02d = % 8d ($%08x)\n", i, (s16)VU1.VI[i].UL, (s16)VU1.VI[i].UL); + } + + u32 j = 0; + for (int i = 0; i < (0x4000 / 4); i++) { + j ^= ((u32*)(VU1.Mem))[i]; + } + SysPrintf("VU Mem CRC = 0x%08x\n", j); + SysPrintf("EndPC = 0x%04x\n", VU1.VI[REG_TPC].UL); + + for (int i = 0; i < 10000000; i++) { + Sleep(1000); + } +} +#else +void VUtestPause() {} +#endif + +#ifndef PCSX2_MICROVU_ + namespace VU1micro { void recAlloc() @@ -113,12 +155,17 @@ namespace VU1micro } assert( (VU1.VI[ REG_TPC ].UL&7) == 0 ); +#ifdef DEBUG_COMPARE + SysPrintf("StartPC = 0x%04x\n", VU1.VI[REG_TPC].UL); +#endif FreezeXMMRegs(1); do { // while loop needed since not always will return finished SuperVUExecuteProgram(VU1.VI[ REG_TPC ].UL & 0x3fff, 1); } while( VU0.VI[ REG_VPU_STAT ].UL&0x100 ); FreezeXMMRegs(0); + + VUtestPause(); } } #else @@ -129,6 +176,7 @@ extern void resetVUrec(const int vuIndex); extern void clearVUrec(u32 addr, u32 size, const int vuIndex); extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); + namespace VU1micro { void recAlloc() { initVUrec(&VU1, 1); } @@ -138,14 +186,20 @@ namespace VU1micro static void recStep() {} static void recExecuteBlock() { - if((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; + if ((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; assert( (VU1.VI[REG_TPC].UL&7) == 0 ); +#ifdef DEBUG_COMPARE + SysPrintf("StartPC = 0x%04x\n", VU1.VI[REG_TPC].UL); +#endif + FreezeXMMRegs(1); //FreezeMMXRegs(1); - runVUrec(VU1.VI[REG_TPC].UL, 5000, 1); - FreezeXMMRegs(0); + runVUrec(VU1.VI[REG_TPC].UL, 300000 /*0x7fffffff*/, 1); //FreezeMMXRegs(0); + FreezeXMMRegs(0); + + VUtestPause(); } } #endif diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 3af44862f8..8639ace77e 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -160,8 +160,8 @@ microVUt(void) mVUendProgram(int fStatus, int fMac, int fClip) { //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) - //AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag - AND32ItoM((uptr)µVU0.regs->VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag + //AND32ItoM((uptr)µVU0.regs->VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag + AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); @@ -170,11 +170,16 @@ microVUt(void) mVUendProgram(int fStatus, int fMac, int fClip) { #define sI ((mVUpBlock->pState.needExactMatch & 0x000f) ? 0 : ((mVUpBlock->pState.flags >> 0) & 3)) #define cI ((mVUpBlock->pState.needExactMatch & 0x0f00) ? 0 : ((mVUpBlock->pState.flags >> 2) & 3)) +void mVUwarning() { Console::Error("microVU Warning: Exiting Execution Early (Possible Infinite Loop)"); } + microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; iPC = mVUstartPC; CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); + PUSH32R(gprR); + CALLFunc((uptr)mVUwarning); + POP32R(gprR); mVUendProgram(sI, 0, cI); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index e0c0530713..66d996594a 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -169,7 +169,7 @@ declareAllVariables #define mVUflagInfo mVUregs.needExactMatch #define mVUflagHack (mVUcurProg.sFlagHack) #define xPC ((iPC / 2) * 8) -#define curI ((u32*)mVU->regs->Micro)[iPC]//mVUcurProg.data[iPC] +#define curI ((u32*)mVU->regs->Micro)[iPC] //mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } #define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } #define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } From 3f9a45c8aa563ebebc88ea1175783a1003688b5c Mon Sep 17 00:00:00 2001 From: gabest11 Date: Mon, 18 May 2009 10:50:43 +0000 Subject: [PATCH 090/296] Cleaned up the dummy MTGS gif transfer function a bit, and let the F5-F7 keys to be passed in GSkeyEvent (allows me to remove windows specific keyboard handling from the gs). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1218 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/GS.h | 15 ++++++++ pcsx2/MTGS.cpp | 74 +++++++++--------------------------- pcsx2/windows/WinSysExec.cpp | 2 +- 3 files changed, 34 insertions(+), 57 deletions(-) diff --git a/pcsx2/GS.h b/pcsx2/GS.h index 983d13a9b1..9f6f8f70f2 100644 --- a/pcsx2/GS.h +++ b/pcsx2/GS.h @@ -99,6 +99,21 @@ struct GIFPath __forceinline void PrepRegs(); void SetTag(const void* mem); u32 GetReg(); + + __forceinline bool StepReg() + { + if((++curreg & 0xf) == tag.nreg) + { + curreg = 0; + + if(--tag.nloop == 0) + { + return false; + } + } + + return true; + } }; diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index dacea2eeea..253345695f 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -280,8 +280,6 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u while(size > 0) { - bool eop = false; - if(path.tag.nloop == 0) { path.SetTag( pMem ); @@ -290,7 +288,9 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u --size; if(pathidx == 2 && path.tag.eop) + { Path3transfer = FALSE; + } if( pathidx == 0 ) { @@ -305,42 +305,16 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u return ++size; } } - - - /*f(path.tag.pre) - { - assert(path.tag.flg != GIF_FLG_IMAGE); // kingdom hearts, ffxii, tales of abyss - - if((path.tag.flg & 2) == 0) - { - // Primitive handler... Nothing for the Dummy to do here. - - //GIFReg r; - //r.i64 = path.tag.PRIM; - //(this->*m_fpGIFRegHandlers[GIF_A_D_REG_PRIM])(&r); - } - }*/ - - if(path.tag.eop) - { - eop = true; - } - else if(path.tag.nloop == 0) - { - if(pathidx == 0) - continue; - - eop = true; - } } - - if(path.tag.nloop > 0) + else { + // NOTE: size > 0 => do {} while(size > 0); should be faster than while(size > 0) {} + switch(path.tag.flg) { case GIF_FLG_PACKED: - while(size > 0) + do { if( path.GetReg() == 0xe ) { @@ -348,25 +322,19 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u if(handler >= 0x60 && handler < 0x63) s_GSHandlers[handler&0x3]((const u32*)pMem); } + size--; pMem += 16; // 128 bits! //sizeof(GIFPackedReg); - - if((++path.curreg & 0xf) == path.tag.nreg) - { - path.curreg = 0; - path.tag.nloop--; - - if(path.tag.nloop == 0) - break; - } } + while(path.StepReg() && size > 0); + break; case GIF_FLG_REGLIST: size *= 2; - while(size > 0) + do { const int handler = path.GetReg(); if(handler >= 0x60 && handler < 0x63) @@ -374,20 +342,11 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u size--; pMem += 8; //sizeof(GIFReg); -- 64 bits! - - if((++path.curreg & 0xf) == path.tag.nreg) - { - path.curreg = 0; - path.tag.nloop--; - - if(path.tag.nloop == 0) - { - break; - } - } } + while(path.StepReg() && size > 0); if(size & 1) pMem += 8; //sizeof(GIFReg); + size /= 2; break; @@ -413,15 +372,18 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u } } - if(eop && ((int)size <= 0 || pathidx == 0)) + if(pathidx == 0) { - break; + if(path.tag.eop && path.tag.nloop == 0) + { + break; + } } } if(pathidx == 0) { - if(!path.tag.eop && path.tag.nloop > 0) + if(size == 0 && path.tag.nloop > 0) { path.tag.nloop = 0; DevCon::Write( "path1 hack! " ); diff --git a/pcsx2/windows/WinSysExec.cpp b/pcsx2/windows/WinSysExec.cpp index fb8e48e106..cde8acee50 100644 --- a/pcsx2/windows/WinSysExec.cpp +++ b/pcsx2/windows/WinSysExec.cpp @@ -413,7 +413,7 @@ namespace HostGui case VK_CAPITAL: keymod->capslock = TRUE; break; case VK_F1: case VK_F2: case VK_F3: case VK_F4: - case VK_F5: case VK_F6: case VK_F7: case VK_F8: + case VK_F8: // note: VK_F5-VK_F7 are reserved for GS case VK_F9: case VK_F10: case VK_F11: case VK_F12: try { From 5f392aba63e51816d146d0735782e86801e8bfbc Mon Sep 17 00:00:00 2001 From: gabest11 Date: Mon, 18 May 2009 11:08:04 +0000 Subject: [PATCH 091/296] GSdx: now I think only the window classes of mfc are left to be replaced (main window, config dialog), trying to get familiar with wx... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1219 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 67 ++++++++------- plugins/GSdx/GPURenderer.h | 11 ++- plugins/GSdx/GPUSettingsDlg.cpp | 48 +++++------ plugins/GSdx/GS.cpp | 91 +++++++++++++-------- plugins/GSdx/GS.h | 2 +- plugins/GSdx/GSCaptureDlg.cpp | 17 ++-- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSDevice9.cpp | 11 +-- plugins/GSdx/GSRenderer.h | 139 +++++++++++++------------------- plugins/GSdx/GSRendererHW.h | 4 +- plugins/GSdx/GSRendererHW9.cpp | 4 +- plugins/GSdx/GSSettingsDlg.cpp | 72 ++++++++--------- plugins/GSdx/GSVector.h | 4 + plugins/GSdx/GSWnd.cpp | 3 +- plugins/GSdx/GSdx.cpp | 27 +++++-- plugins/GSdx/GSdx.h | 17 ++-- plugins/GSdx/stdafx.h | 22 ++--- 17 files changed, 291 insertions(+), 249 deletions(-) diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index 4583b48b90..a1abe6c9ec 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -20,6 +20,7 @@ */ #include "stdafx.h" +#include "GSdx.h" #include "GSUtil.h" #include "GPURendererSW.h" #include "GSDevice7.h" @@ -52,30 +53,22 @@ EXPORT_C_(uint32) PSEgetLibVersion() EXPORT_C_(int32) GPUinit() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - // TODO - return 0; } EXPORT_C_(int32) GPUshutdown() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - // TODO - return 0; } EXPORT_C_(int32) GPUclose() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - delete s_gpu; s_gpu = NULL; +#ifdef _WINDOWS + if(SUCCEEDED(s_hr)) { ::CoUninitialize(); @@ -83,32 +76,47 @@ EXPORT_C_(int32) GPUclose() s_hr = E_FAIL; } +#endif + return 0; } EXPORT_C_(int32) GPUopen(HWND hWnd) { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - if(!GSUtil::CheckDirectX() || !GSUtil::CheckSSE()) - { - return -1; - } - GPUclose(); GPURendererSettings rs; - rs.m_filter = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("filter"), 0); - rs.m_dither = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("dithering"), 1); - rs.m_aspectratio = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("AspectRatio"), 1); - rs.m_vsync = !!AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("vsync"), FALSE); - rs.m_scale.x = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_x"), 0); - rs.m_scale.y = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("scale_y"), 0); + int threads = 1; + int renderer = 1; - int threads = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("swthreads"), 1); +#ifdef _WINDOWS - int renderer = AfxGetApp()->GetProfileInt(_T("GPUSettings"), _T("Renderer"), 1); + AFX_MANAGE_STATE(AfxGetStaticModuleState()); + + s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); + + if(!GSUtil::CheckDirectX()) + { + return -1; + } + +#endif + + if(!GSUtil::CheckSSE()) + { + return -1; + } + + rs.m_filter = theApp.GetConfig("filter", 0); + rs.m_dither = theApp.GetConfig("dithering", 1); + rs.m_aspectratio = theApp.GetConfig("AspectRatio", 1); + rs.m_vsync = !!theApp.GetConfig("vsync", 0); + rs.m_scale.x = theApp.GetConfig("scale_x", 0); + rs.m_scale.y = theApp.GetConfig("scale_y", 0); + + threads = theApp.GetConfig("swthreads", 1); + renderer = theApp.GetConfig("Renderer", 1); switch(renderer) { @@ -119,8 +127,6 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) // TODO: case 3: s_gpu = new GPURendererNull(rs, threads); break; } - s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); - if(!s_gpu->Create(hWnd)) { GPUclose(); @@ -133,6 +139,9 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) EXPORT_C_(int32) GPUconfigure() { + +#ifdef _WINDOWS + AFX_MANAGE_STATE(AfxGetStaticModuleState()); GPUSettingsDlg dlg; @@ -143,13 +152,13 @@ EXPORT_C_(int32) GPUconfigure() GPUinit(); } +#endif + return 0; } EXPORT_C_(int32) GPUtest() { - // TODO - return 0; } diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index d6a091fec1..71125b9227 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -381,6 +381,15 @@ public: virtual bool MakeSnapshot(const string& path) { - return m_dev.SaveCurrent(format("%s_%s", path.c_str(), CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))) + ".bmp"); + time_t t = time(NULL); + + char buff[16]; + + if(!strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) + { + return false; + } + + return m_dev.SaveCurrent(format("%s_%s.bmp", path.c_str(), buff)); } }; diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index b1bbc9aa17..96b9e5c28b 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -130,8 +130,6 @@ BOOL GPUSettingsDlg::OnInitDialog() { __super::OnInitDialog(); - CWinApp* pApp = AfxGetApp(); - D3DCAPS9 caps; memset(&caps, 0, sizeof(caps)); caps.PixelShaderVersion = D3DPS_VERSION(0, 0); @@ -154,9 +152,9 @@ BOOL GPUSettingsDlg::OnInitDialog() if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - uint32 ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); - uint32 ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); - uint32 ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); + uint32 ModeWidth = theApp.GetConfig("ModeWidth", 0); + uint32 ModeHeight = theApp.GetConfig("ModeHeight", 0); + uint32 ModeRefreshRate = theApp.GetConfig("ModeRefreshRate", 0); uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); @@ -194,12 +192,12 @@ BOOL GPUSettingsDlg::OnInitDialog() renderers.push_back(g_renderers[i]); } - GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, pApp->GetProfileInt(_T("GPUSettings"), _T("Renderer"), 1)); - GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, pApp->GetProfileInt(_T("Settings"), _T("PixelShaderVersion2"), D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); - GSSetting::InitComboBox(g_filter, countof(g_filter), m_filter, pApp->GetProfileInt(_T("GPUSettings"), _T("filter"), 0)); - GSSetting::InitComboBox(g_dithering, countof(g_dithering), m_dithering, pApp->GetProfileInt(_T("GPUSettings"), _T("dithering"), 1)); - GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, pApp->GetProfileInt(_T("GPUSettings"), _T("AspectRatio"), 1)); - GSSetting::InitComboBox(g_internalresolution, countof(g_internalresolution), m_internalresolution, pApp->GetProfileInt(_T("GPUSettings"), _T("scale_x"), 0) | (pApp->GetProfileInt(_T("GPUSettings"), _T("scale_y"), 0) << 2)); + GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, theApp.GetConfig("Renderer", 1)); + GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); + GSSetting::InitComboBox(g_filter, countof(g_filter), m_filter, theApp.GetConfig("filter", 0)); + GSSetting::InitComboBox(g_dithering, countof(g_dithering), m_dithering, theApp.GetConfig("dithering", 1)); + GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, theApp.GetConfig("AspectRatio", 1)); + GSSetting::InitComboBox(g_internalresolution, countof(g_internalresolution), m_internalresolution, theApp.GetConfig("scale_x", 0) | (theApp.GetConfig("scale_y", 0) << 2)); OnCbnSelchangeCombo1(); @@ -207,7 +205,7 @@ BOOL GPUSettingsDlg::OnInitDialog() // m_swthreads.SetRange(1, 16); - m_swthreads.SetPos(pApp->GetProfileInt(_T("GPUSettings"), _T("swthreads"), 1)); + m_swthreads.SetPos(theApp.GetConfig("swthreads", 1)); // @@ -219,53 +217,51 @@ BOOL GPUSettingsDlg::OnInitDialog() void GPUSettingsDlg::OnOK() { - CWinApp* pApp = AfxGetApp(); - UpdateData(); if(m_resolution.GetCurSel() >= 0) { D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); - pApp->WriteProfileInt(_T("Settings"), _T("ModeWidth"), mode.Width); - pApp->WriteProfileInt(_T("Settings"), _T("ModeHeight"), mode.Height); - pApp->WriteProfileInt(_T("Settings"), _T("ModeRefreshRate"), mode.RefreshRate); + theApp.SetConfig("ModeWidth", (int)mode.Width); + theApp.SetConfig("ModeHeight", (int)mode.Height); + theApp.SetConfig("ModeRefreshRate", (int)mode.RefreshRate); } if(m_renderer.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("GPUSettings"), _T("Renderer"), (DWORD)m_renderer.GetItemData(m_renderer.GetCurSel())); + theApp.SetConfig("Renderer", (int)m_renderer.GetItemData(m_renderer.GetCurSel())); } if(m_psversion.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("PixelShaderVersion2"), (DWORD)m_psversion.GetItemData(m_psversion.GetCurSel())); + theApp.SetConfig("PixelShaderVersion2", (int)m_psversion.GetItemData(m_psversion.GetCurSel())); } if(m_filter.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("GPUSettings"), _T("filter"), (DWORD)m_filter.GetItemData(m_filter.GetCurSel())); + theApp.SetConfig("filter", (int)m_filter.GetItemData(m_filter.GetCurSel())); } if(m_dithering.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("GPUSettings"), _T("dithering"), (DWORD)m_dithering.GetItemData(m_dithering.GetCurSel())); + theApp.SetConfig("dithering", (int)m_dithering.GetItemData(m_dithering.GetCurSel())); } if(m_aspectratio.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("GPUSettings"), _T("AspectRatio"), (DWORD)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); + theApp.SetConfig("AspectRatio", (int)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); } if(m_internalresolution.GetCurSel() >= 0) { - DWORD value = (DWORD)m_internalresolution.GetItemData(m_internalresolution.GetCurSel()); + int value = (int)m_internalresolution.GetItemData(m_internalresolution.GetCurSel()); - pApp->WriteProfileInt(_T("GPUSettings"), _T("scale_x"), value & 3); - pApp->WriteProfileInt(_T("GPUSettings"), _T("scale_y"), (value >> 2) & 3); + theApp.SetConfig("scale_x", value & 3); + theApp.SetConfig("scale_y", (value >> 2) & 3); } - pApp->WriteProfileInt(_T("GPUSettings"), _T("swthreads"), m_swthreads.GetPos()); + theApp.SetConfig("swthreads", m_swthreads.GetPos()); __super::OnOK(); } diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index a06decfe00..ff0133ac0e 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -71,14 +71,11 @@ EXPORT_C GSsetBaseMem(uint8* mem) EXPORT_C_(INT32) GSinit() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - return 0; } EXPORT_C GSshutdown() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); } EXPORT_C GSclose() @@ -87,36 +84,51 @@ EXPORT_C GSclose() s_gs = NULL; +#ifdef _WINDOWS + if(SUCCEEDED(s_hr)) { ::CoUninitialize(); s_hr = E_FAIL; } + +#endif } static INT32 GSopen(void* dsp, char* title, int mt, int renderer) { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - if(!GSUtil::CheckDirectX() || !GSUtil::CheckSSE()) - { - return -1; - } - GSclose(); GSRendererSettings rs; - rs.m_interlace = AfxGetApp()->GetProfileInt(_T("Settings"), _T("interlace"), 0); - rs.m_aspectratio = AfxGetApp()->GetProfileInt(_T("Settings"), _T("aspectratio"), 1); - rs.m_filter = AfxGetApp()->GetProfileInt(_T("Settings"), _T("filter"), 1); - rs.m_vsync = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("vsync"), FALSE); - rs.m_nativeres = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("nativeres"), FALSE); - rs.m_aa1 = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("aa1"), FALSE); - rs.m_blur = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("blur"), FALSE); + int threads = 1; - int threads = AfxGetApp()->GetProfileInt(_T("Settings"), _T("swthreads"), 1); +#ifdef _WINDOWS + + s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); + + if(!GSUtil::CheckDirectX()) + { + return -1; + } + +#endif + + if(!GSUtil::CheckSSE()) + { + return -1; + } + + rs.m_interlace = theApp.GetConfig("interlace", 0); + rs.m_aspectratio = theApp.GetConfig("aspectratio", 1); + rs.m_filter = theApp.GetConfig("filter", 1); + rs.m_vsync = !!theApp.GetConfig("vsync", 0); + rs.m_nativeres = !!theApp.GetConfig("nativeres", 0); + rs.m_aa1 = !!theApp.GetConfig("aa1", 0); + rs.m_blur = !!theApp.GetConfig("blur", 0); + + threads = theApp.GetConfig("swthreads", 1); switch(renderer) { @@ -131,8 +143,6 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs); break; } - s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); - if(!s_gs->Create(title)) { GSclose(); @@ -151,8 +161,12 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) EXPORT_C_(INT32) GSopen(void* dsp, char* title, int mt) { +#ifdef _WINDOWS + AFX_MANAGE_STATE(AfxGetStaticModuleState()); +#endif + int renderer; if(mt == 2) @@ -164,7 +178,7 @@ EXPORT_C_(INT32) GSopen(void* dsp, char* title, int mt) else { // normal init - renderer = AfxGetApp()->GetProfileInt(_T("Settings"), _T("renderer"), 0); + renderer = theApp.GetConfig("renderer", 0); } return GSopen(dsp, title, mt, renderer); @@ -212,6 +226,20 @@ EXPORT_C GSgifTransfer3(uint8* mem, uint32 size) EXPORT_C GSvsync(int field) { +#ifdef _WINDOWS + + MSG msg; + + memset(&msg, 0, sizeof(msg)); + + while(msg.message != WM_QUIT && PeekMessage(&msg, NULL, 0, 0, PM_REMOVE)) + { + TranslateMessage(&msg); + DispatchMessage(&msg); + } + +#endif + s_gs->VSync(field); } @@ -220,8 +248,9 @@ EXPORT_C_(uint32) GSmakeSnapshot(char* path) return s_gs->MakeSnapshot(string(path) + "gsdx"); } -EXPORT_C GSkeyEvent(keyEvent* ev) +EXPORT_C GSkeyEvent(GSKeyEventData* e) { + s_gs->KeyEvent(e); } EXPORT_C_(int) GSfreeze(int mode, GSFreezeData* data) @@ -244,6 +273,8 @@ EXPORT_C_(int) GSfreeze(int mode, GSFreezeData* data) EXPORT_C GSconfigure() { +#ifdef _WINDOWS + AFX_MANAGE_STATE(AfxGetStaticModuleState()); GSSettingsDlg dlg; @@ -253,21 +284,14 @@ EXPORT_C GSconfigure() GSshutdown(); GSinit(); } + +#endif + } EXPORT_C_(INT32) GStest() { return 0; - - // TODO - - /* - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - CComPtr dev; - - return SUCCEEDED(D3D10CreateDevice(NULL, D3D10_DRIVER_TYPE_HARDWARE, NULL, 0, D3D10_SDK_VERSION, &dev)) ? 0 : -1; - */ } EXPORT_C GSabout() @@ -294,6 +318,8 @@ EXPORT_C GSsetFrameSkip(int frameskip) s_gs->SetFrameSkip(frameskip); } +#ifdef _WINDOWS + EXPORT_C GSReplay(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow) { int renderer = -1; @@ -597,3 +623,4 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow fclose(file); } +#endif \ No newline at end of file diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index 90ddc3f126..39a799a0f1 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -1163,7 +1163,7 @@ struct GSPrivRegSet #pragma pack(pop) enum {KEYPRESS=1, KEYRELEASE=2}; -struct keyEvent {uint32 key, event;}; +struct GSKeyEventData {uint32 key, type;}; enum {FREEZE_LOAD=0, FREEZE_SAVE=1, FREEZE_SIZE=2}; struct GSFreezeData {int size; uint8* data;}; diff --git a/plugins/GSdx/GSCaptureDlg.cpp b/plugins/GSdx/GSCaptureDlg.cpp index 32432a5df4..9e3ae7c2e3 100644 --- a/plugins/GSdx/GSCaptureDlg.cpp +++ b/plugins/GSdx/GSCaptureDlg.cpp @@ -21,6 +21,7 @@ #include "stdafx.h" #include +#include "GSdx.h" #include "GSCaptureDlg.h" // GSCaptureDlg dialog @@ -29,9 +30,9 @@ IMPLEMENT_DYNAMIC(GSCaptureDlg, CDialog) GSCaptureDlg::GSCaptureDlg(CWnd* pParent /*=NULL*/) : CDialog(GSCaptureDlg::IDD, pParent) { - m_width = AfxGetApp()->GetProfileInt(_T("Capture"), _T("Width"), 640); - m_height = AfxGetApp()->GetProfileInt(_T("Capture"), _T("Height"), 480); - m_filename = AfxGetApp()->GetProfileString(_T("Capture"), _T("FileName")); + m_width = theApp.GetConfig("CaptureWidth", 640); + m_height = theApp.GetConfig("CaptureHeight", 480); + m_filename = theApp.GetConfig("CaptureFileName", "").c_str(); } GSCaptureDlg::~GSCaptureDlg() @@ -123,7 +124,7 @@ BOOL GSCaptureDlg::OnInitDialog() // - CString DisplayNameToFind = AfxGetApp()->GetProfileString(_T("Capture"), _T("VideoCodecDisplayName")); + CString DisplayNameToFind = theApp.GetConfig("CaptureVideoCodecDisplayName", "").c_str(); for(int i = 0; i < m_codeclist.GetCount(); i++) { @@ -230,10 +231,10 @@ void GSCaptureDlg::OnBnClickedOk() m_enc = c.filter; - AfxGetApp()->WriteProfileInt(_T("Capture"), _T("Width"), m_width); - AfxGetApp()->WriteProfileInt(_T("Capture"), _T("Height"), m_height); - AfxGetApp()->WriteProfileString(_T("Capture"), _T("FileName"), m_filename); - AfxGetApp()->WriteProfileString(_T("Capture"), _T("VideoCodecDisplayName"), CString(c.DisplayName)); + theApp.SetConfig("CaptureWidth", m_width); + theApp.SetConfig("CaptureHeight", m_height); + theApp.SetConfig("CaptureFileName", m_filename); + theApp.SetConfig("CaptureVideoCodecDisplayName", CString(c.DisplayName)); OnOK(); } diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index a4a3fde2e1..b6b98060dc 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -82,6 +82,7 @@ CRC::Game CRC::m_games[] = {0xC19A374E, SoTC, US}, {0x7D8F539A, SoTC, EU}, {0x3122B508, OnePieceGrandAdventure, US}, + {0x8DF14A24, OnePieceGrandAdventure, Unknown}, {0x6F8545DB, ICO, US}, {0xB01A4C95, ICO, JP}, {0x5C991F4E, ICO, Unknown}, diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 4e38b0af0d..461cb30842 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -20,6 +20,7 @@ */ #include "stdafx.h" +#include "GSdx.h" #include "GSDevice9.h" #include "resource.h" @@ -102,7 +103,7 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) m_d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &m_d3dcaps); - bool fs = AfxGetApp()->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0) > 0; + bool fs = theApp.GetConfig("ModeWidth", 0) > 0; if(!Reset(1, 1, fs)) return false; @@ -110,7 +111,7 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) // shaders - DWORD psver = AfxGetApp()->GetProfileInt(_T("Settings"), _T("PixelShaderVersion2"), D3DPS_VERSION(2, 0)); + uint32 psver = (uint32)theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)); if(psver > m_d3dcaps.PixelShaderVersion) { @@ -262,9 +263,9 @@ bool GSDevice9::Reset(int w, int h, bool fs) // m_pp.Flags |= D3DPRESENTFLAG_VIDEO; // enables tv-out (but I don't think anyone would still use a regular tv...) - int mw = AfxGetApp()->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); - int mh = AfxGetApp()->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); - int mrr = AfxGetApp()->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); + int mw = theApp.GetConfig("ModeWidth", 0); + int mh = theApp.GetConfig("ModeHeight", 0); + int mrr = theApp.GetConfig("ModeRefreshRate", 0); if(fs && mw > 0 && mh > 0 && mrr >= 0) { diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index bb3fb7c34e..a5fc7f3577 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -21,6 +21,7 @@ #pragma once +#include "GSdx.h" #include "GSWnd.h" #include "GSState.h" #include "GSVertexList.h" @@ -43,56 +44,10 @@ class GSRendererBase : public GSState, protected GSRendererSettings protected: bool m_osd; - void ProcessWindowMessages() - { - MSG msg; - - memset(&msg, 0, sizeof(msg)); - - while(msg.message != WM_QUIT && PeekMessage(&msg, NULL, 0, 0, PM_REMOVE)) - { - if(OnMessage(msg)) - { - continue; - } - - TranslateMessage(&msg); - DispatchMessage(&msg); - } - } - - virtual bool OnMessage(const MSG& msg) - { - if(msg.message == WM_KEYDOWN) - { - int step = (::GetAsyncKeyState(VK_SHIFT) & 0x8000) ? -1 : 1; - - switch(msg.wParam) - { - case VK_F5: - m_interlace = (m_interlace + 7 + step) % 7; - return true; - case VK_F6: - m_aspectratio = (m_aspectratio + 3 + step) % 3; - return true; - case VK_F7: - m_wnd.SetWindowText(_T("PCSX2")); - m_osd = !m_osd; - return true; - case VK_DELETE: - m_aa1 = !m_aa1; - return true; - case VK_END: - m_blur = !m_blur; - return true; - } - } - - return false; - } - public: GSWnd m_wnd; + GSCapture m_capture; + string m_snapshot; public: GSRendererBase(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) @@ -108,9 +63,59 @@ public: m_blur = rs.m_blur; }; + void KeyEvent(GSKeyEventData* e) + { + if(e->type == KEYPRESS) + { + // TODO: linux + + int step = (::GetAsyncKeyState(VK_SHIFT) & 0x8000) ? -1 : 1; + + switch(e->key) + { + case VK_F5: + m_interlace = (m_interlace + 7 + step) % 7; + return; + case VK_F6: + m_aspectratio = (m_aspectratio + 3 + step) % 3; + return; + case VK_F7: + m_wnd.SetWindowText(_T("PCSX2")); + m_osd = !m_osd; + return; + case VK_F12: + if(m_capture.IsCapturing()) m_capture.EndCapture(); + else m_capture.BeginCapture(GetFPS()); + return; + case VK_DELETE: + m_aa1 = !m_aa1; + return; + case VK_END: + m_blur = !m_blur; + return; + } + } + } + + bool MakeSnapshot(const string& path) + { + if(m_snapshot.empty()) + { + time_t t = time(NULL); + + char buff[16]; + + if(strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) + { + m_snapshot = format("%s_%s", path.c_str(), buff); + } + } + + return true; + } + virtual bool Create(const string& title) = 0; virtual void VSync(int field) = 0; - virtual bool MakeSnapshot(const string& path) = 0; }; template class GSRenderer : public GSRendererBase @@ -382,21 +387,6 @@ protected: } } - virtual bool OnMessage(const MSG& msg) - { - if(msg.message == WM_KEYDOWN) - { - switch(msg.wParam) - { - case VK_F12: - if(m_capture.IsCapturing()) m_capture.EndCapture(); - else m_capture.BeginCapture(GetFPS()); - return true; - } - } - - return __super::OnMessage(msg); - } public: Device m_dev; @@ -407,18 +397,15 @@ public: bool s_save; bool s_savez; - string m_snapshot; - GSCapture m_capture; - public: GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) : GSRendererBase(base, mt, irq, rs) , m_psrr(psrr) { s_n = 0; - s_dump = !!AfxGetApp()->GetProfileInt(_T("Debug"), _T("dump"), 0); - s_save = !!AfxGetApp()->GetProfileInt(_T("Debug"), _T("save"), 0); - s_savez = !!AfxGetApp()->GetProfileInt(_T("Debug"), _T("savez"), 0); + s_dump = !!theApp.GetConfig("dump", 0); + s_save = !!theApp.GetConfig("save", 0); + s_savez = !!theApp.GetConfig("savez", 0); } bool Create(const string& title) @@ -446,8 +433,6 @@ public: m_perfmon.Put(GSPerfMon::Frame); - ProcessWindowMessages(); - field = field ? 1 : 0; if(!Merge(field)) return; @@ -530,16 +515,6 @@ public: DoCapture(); } - bool MakeSnapshot(const string& path) - { - if(m_snapshot.empty()) - { - m_snapshot = format("%s_%s", path.c_str(), CTime::GetCurrentTime().Format(_T("%Y%m%d%H%M%S"))); - } - - return true; - } - virtual void MinMaxUV(int w, int h, GSVector4i& r) {r = GSVector4i(0, 0, w, h);} virtual bool CanUpscale() {return !m_nativeres;} }; diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 79cc37c9ad..34df556fbb 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -677,8 +677,8 @@ public: { if(!m_nativeres) { - m_width = AfxGetApp()->GetProfileInt(_T("Settings"), _T("resx"), m_width); - m_height = AfxGetApp()->GetProfileInt(_T("Settings"), _T("resy"), m_height); + m_width = theApp.GetConfig("resx", m_width); + m_height = theApp.GetConfig("resy", m_height); } m_tc = new TextureCache(this); diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 036290eea3..a5ee0df4fd 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -27,8 +27,8 @@ GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) : GSRendererHW(base, mt, irq, rs, false) { - m_fba.enabled = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("fba"), TRUE); - m_logz = !!AfxGetApp()->GetProfileInt(_T("Settings"), _T("logz"), FALSE); + m_fba.enabled = !!theApp.GetConfig("fba", 1); + m_logz = !!theApp.GetConfig("logz", 0); InitVertexKick(); } diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 0f3d5577fb..928fa7ecb2 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -141,8 +141,6 @@ BOOL GSSettingsDlg::OnInitDialog() { __super::OnInitDialog(); - CWinApp* pApp = AfxGetApp(); - D3DCAPS9 caps; memset(&caps, 0, sizeof(caps)); caps.PixelShaderVersion = D3DPS_VERSION(0, 0); @@ -165,9 +163,9 @@ BOOL GSSettingsDlg::OnInitDialog() if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - uint32 ModeWidth = pApp->GetProfileInt(_T("Settings"), _T("ModeWidth"), 0); - uint32 ModeHeight = pApp->GetProfileInt(_T("Settings"), _T("ModeHeight"), 0); - uint32 ModeRefreshRate = pApp->GetProfileInt(_T("Settings"), _T("ModeRefreshRate"), 0); + uint32 ModeWidth = theApp.GetConfig("ModeWidth", 0); + uint32 ModeHeight = theApp.GetConfig("ModeHeight", 0); + uint32 ModeRefreshRate = theApp.GetConfig("ModeRefreshRate", 0); uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); @@ -204,27 +202,27 @@ BOOL GSSettingsDlg::OnInitDialog() renderers.push_back(g_renderers[i]); } - GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, pApp->GetProfileInt(_T("Settings"), _T("Renderer"), 0)); - GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, pApp->GetProfileInt(_T("Settings"), _T("PixelShaderVersion2"), D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); - GSSetting::InitComboBox(g_interlace, countof(g_interlace), m_interlace, pApp->GetProfileInt(_T("Settings"), _T("Interlace"), 0)); - GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, pApp->GetProfileInt(_T("Settings"), _T("AspectRatio"), 1)); + GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, theApp.GetConfig("Renderer", 0)); + GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); + GSSetting::InitComboBox(g_interlace, countof(g_interlace), m_interlace, theApp.GetConfig("Interlace", 0)); + GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, theApp.GetConfig("AspectRatio", 1)); OnCbnSelchangeCombo1(); // - m_filter = pApp->GetProfileInt(_T("Settings"), _T("filter"), 1); - m_vsync = !!pApp->GetProfileInt(_T("Settings"), _T("vsync"), FALSE); - m_logz = !!pApp->GetProfileInt(_T("Settings"), _T("logz"), FALSE); - m_fba = !!pApp->GetProfileInt(_T("Settings"), _T("fba"), TRUE); - m_aa1 = !!pApp->GetProfileInt(_T("Settings"), _T("aa1"), FALSE); - m_blur = !!pApp->GetProfileInt(_T("Settings"), _T("blur"), FALSE); + m_filter = theApp.GetConfig("filter", 1); + m_vsync = !!theApp.GetConfig("vsync", 0); + m_logz = !!theApp.GetConfig("logz", 0); + m_fba = !!theApp.GetConfig("fba", 1); + m_aa1 = !!theApp.GetConfig("aa1", 0); + m_blur = !!theApp.GetConfig("blur", 0); m_resx.SetRange(512, 4096); m_resy.SetRange(512, 4096); - m_resx.SetPos(pApp->GetProfileInt(_T("Settings"), _T("resx"), 1024)); - m_resy.SetPos(pApp->GetProfileInt(_T("Settings"), _T("resy"), 1024)); - m_nativeres = !!pApp->GetProfileInt(_T("Settings"), _T("nativeres"), FALSE); + m_resx.SetPos(theApp.GetConfig("resx", 1024)); + m_resy.SetPos(theApp.GetConfig("resy", 1024)); + m_nativeres = !!theApp.GetConfig("nativeres", 0); m_resx.EnableWindow(!m_nativeres); m_resy.EnableWindow(!m_nativeres); @@ -232,7 +230,7 @@ BOOL GSSettingsDlg::OnInitDialog() m_resyedit.EnableWindow(!m_nativeres); m_swthreads.SetRange(1, 16); - m_swthreads.SetPos(pApp->GetProfileInt(_T("Settings"), _T("swthreads"), 1)); + m_swthreads.SetPos(theApp.GetConfig("swthreads", 1)); // @@ -244,50 +242,48 @@ BOOL GSSettingsDlg::OnInitDialog() void GSSettingsDlg::OnOK() { - CWinApp* pApp = AfxGetApp(); - UpdateData(); if(m_resolution.GetCurSel() >= 0) { D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); - pApp->WriteProfileInt(_T("Settings"), _T("ModeWidth"), mode.Width); - pApp->WriteProfileInt(_T("Settings"), _T("ModeHeight"), mode.Height); - pApp->WriteProfileInt(_T("Settings"), _T("ModeRefreshRate"), mode.RefreshRate); + theApp.SetConfig("ModeWidth", (int)mode.Width); + theApp.SetConfig("ModeHeight", (int)mode.Height); + theApp.SetConfig("ModeRefreshRate", (int)mode.RefreshRate); } if(m_renderer.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("Renderer"), (uint32)m_renderer.GetItemData(m_renderer.GetCurSel())); + theApp.SetConfig("Renderer", (uint32)m_renderer.GetItemData(m_renderer.GetCurSel())); } if(m_psversion.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("PixelShaderVersion2"), (uint32)m_psversion.GetItemData(m_psversion.GetCurSel())); + theApp.SetConfig("PixelShaderVersion2", (uint32)m_psversion.GetItemData(m_psversion.GetCurSel())); } if(m_interlace.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("Interlace"), (uint32)m_interlace.GetItemData(m_interlace.GetCurSel())); + theApp.SetConfig("Interlace", (uint32)m_interlace.GetItemData(m_interlace.GetCurSel())); } if(m_aspectratio.GetCurSel() >= 0) { - pApp->WriteProfileInt(_T("Settings"), _T("AspectRatio"), (uint32)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); + theApp.SetConfig("AspectRatio", (uint32)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); } - pApp->WriteProfileInt(_T("Settings"), _T("filter"), m_filter); - pApp->WriteProfileInt(_T("Settings"), _T("vsync"), m_vsync); - pApp->WriteProfileInt(_T("Settings"), _T("logz"), m_logz); - pApp->WriteProfileInt(_T("Settings"), _T("fba"), m_fba); - pApp->WriteProfileInt(_T("Settings"), _T("aa1"), m_aa1); - pApp->WriteProfileInt(_T("Settings"), _T("blur"), m_blur); + theApp.SetConfig("filter", m_filter); + theApp.SetConfig("vsync", m_vsync); + theApp.SetConfig("logz", m_logz); + theApp.SetConfig("fba", m_fba); + theApp.SetConfig("aa1", m_aa1); + theApp.SetConfig("blur", m_blur); - pApp->WriteProfileInt(_T("Settings"), _T("resx"), m_resx.GetPos()); - pApp->WriteProfileInt(_T("Settings"), _T("resy"), m_resy.GetPos()); - pApp->WriteProfileInt(_T("Settings"), _T("swthreads"), m_swthreads.GetPos()); - pApp->WriteProfileInt(_T("Settings"), _T("nativeres"), m_nativeres); + theApp.SetConfig("resx", m_resx.GetPos()); + theApp.SetConfig("resy", m_resy.GetPos()); + theApp.SetConfig("swthreads", m_swthreads.GetPos()); + theApp.SetConfig("nativeres", m_nativeres); __super::OnOK(); } diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index d3dd190e0f..80a102c146 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -182,6 +182,8 @@ public: GSVector4i fit(int preset) const; + #ifdef _WINDOWS + operator LPCRECT() const { return (LPCRECT)this; @@ -192,6 +194,8 @@ public: return (LPRECT)this; } + #endif + // uint32 rgba32() const diff --git a/plugins/GSdx/GSWnd.cpp b/plugins/GSdx/GSWnd.cpp index c01f3390bb..bcb8ff9caa 100644 --- a/plugins/GSdx/GSWnd.cpp +++ b/plugins/GSdx/GSWnd.cpp @@ -20,6 +20,7 @@ */ #include "StdAfx.h" +#include "GSdx.h" #include "GSWnd.h" BEGIN_MESSAGE_MAP(GSWnd, CWnd) @@ -51,7 +52,7 @@ bool GSWnd::Create(const string& title) r = CRect(r.CenterPoint() - CSize(s.cx / 2, s.cy / 2), s); - LPCTSTR wc = AfxRegisterWndClass(CS_VREDRAW|CS_HREDRAW|CS_DBLCLKS, AfxGetApp()->LoadStandardCursor(IDC_ARROW), 0, 0); + LPCTSTR wc = AfxRegisterWndClass(CS_VREDRAW|CS_HREDRAW|CS_DBLCLKS, theApp.LoadStandardCursor(IDC_ARROW), 0, 0); return !!CreateEx(0, wc, title.c_str(), WS_OVERLAPPEDWINDOW, r, NULL, 0); } diff --git a/plugins/GSdx/GSdx.cpp b/plugins/GSdx/GSdx.cpp index 48a30d5e92..339263ba7f 100644 --- a/plugins/GSdx/GSdx.cpp +++ b/plugins/GSdx/GSdx.cpp @@ -51,15 +51,12 @@ // details. // -BEGIN_MESSAGE_MAP(GSdxApp, CWinApp) -END_MESSAGE_MAP() +GSdxApp theApp; GSdxApp::GSdxApp() { } -GSdxApp theApp; - BOOL GSdxApp::InitInstance() { __super::InitInstance(); @@ -96,4 +93,24 @@ BOOL GSdxApp::InitInstance() m_pszProfileName = _tcsdup((LPCTSTR)path); return TRUE; -} \ No newline at end of file +} + +string GSdxApp::GetConfig(const char* entry, const char* value) +{ + return string(GetProfileString("Settings", entry, value)); +} + +void GSdxApp::SetConfig(const char* entry, const char* value) +{ + WriteProfileString("Settings", entry, value); +} + +int GSdxApp::GetConfig(const char* entry, int value) +{ + return GetProfileInt("Settings", entry, value); +} + +void GSdxApp::SetConfig(const char* entry, int value) +{ + WriteProfileInt("Settings", entry, value); +} diff --git a/plugins/GSdx/GSdx.h b/plugins/GSdx/GSdx.h index 5d3be89e27..e1abf14876 100644 --- a/plugins/GSdx/GSdx.h +++ b/plugins/GSdx/GSdx.h @@ -21,17 +21,18 @@ #pragma once -#ifndef __AFXWIN_H__ - #error include 'stdafx.h' before including this file for PCH -#endif - class GSdxApp : public CWinApp { +protected: + virtual BOOL InitInstance(); + public: GSdxApp(); -public: - virtual BOOL InitInstance(); - - DECLARE_MESSAGE_MAP() + string GetConfig(const char* entry, const char* value); + void SetConfig(const char* entry, const char* value); + int GetConfig(const char* entry, int value); + void SetConfig(const char* entry, int value); }; + +extern GSdxApp theApp; \ No newline at end of file diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index adf3463f2d..49d9a216c5 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -6,6 +6,8 @@ #pragma warning(disable: 4996 4995 4324 4100) +#ifdef _WINDOWS + #ifndef VC_EXTRALEAN #define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers #endif @@ -35,11 +37,13 @@ #ifndef _AFX_NO_AFXCMN_SUPPORT #include // MFC support for Windows Common Controls #endif // _AFX_NO_AFXCMN_SUPPORT -//#include + +#endif // stdc #include +#include #include #include @@ -71,17 +75,17 @@ typedef signed long long int64; #define EXPORT_C_(type) extern "C" __declspec(dllexport) type __stdcall #ifndef RESTRICT - #ifdef __INTEL_COMPILER - #define RESTRICT restrict - #elif _MSC_VER >= 1400 // TODO: gcc - #define RESTRICT __restrict - #else - #define RESTRICT - #endif +#ifdef __INTEL_COMPILER + #define RESTRICT restrict +#elif _MSC_VER >= 1400 // TODO: gcc + #define RESTRICT __restrict +#else + #define RESTRICT +#endif #endif #ifdef __x86_64__ - #define _M_AMD64 +#define _M_AMD64 #endif extern "C" uint64 __rdtsc(); // TODO: gcc From 3019f7d02c62f6d5da8a18023f7e4aa06109f958 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 18 May 2009 11:13:30 +0000 Subject: [PATCH 092/296] microVU: - Fixed the bugs with the logical min/max code (DaZ should work fine again) - Optimized some lower instructions a bit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1220 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Lower.inl | 15 +++++++++------ pcsx2/x86/microVU_Misc.inl | 24 +++++++++++------------- 2 files changed, 20 insertions(+), 19 deletions(-) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 1837c7ec0a..b0f4b5dd1e 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -73,8 +73,9 @@ microVUf(void) mVU_DIV() { mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); - mVUunpack_xyzw(xmmFs, xmmFs, 0); - mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("DIV Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } @@ -91,8 +92,9 @@ microVUf(void) mVU_SQRT() { if (CHECK_VU_OVERFLOW) SSE_MINSS_XMM_to_XMM(xmmFt, xmmMax); // Clamp infinities (only need to do positive clamp since xmmFt is positive) SSE_SQRTSS_XMM_to_XMM(xmmFt, xmmFt); - mVUunpack_xyzw(xmmFt, xmmFt, 0); - mVUmergeRegs(xmmPQ, xmmFt, writeQ ? 4 : 8); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFt); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("SQRT Q, vf%02d%s", _Ft_, _Ftf_String); } } @@ -129,8 +131,9 @@ microVUf(void) mVU_RSQRT() { mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); - mVUunpack_xyzw(xmmFs, xmmFs, 0); - mVUmergeRegs(xmmPQ, xmmFs, writeQ ? 4 : 8); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); + if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("RSQRT Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 1d4f16ffba..adf3ac37bc 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -313,34 +313,33 @@ microVUt(void) mVUcheckSflag(int progIndex) { } } -static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0x7fffffff, 0x80000000, 0x7fffffff, 0x80000000}; +static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK2[4]) = {0x00000000, 0x40000000, 0x00000000, 0x40000000}; // Warning: Modifies xmmT1 and xmmT2 void MIN_MAX_(x86SSERegType to, x86SSERegType from, bool min) { - // XY - SSE2_PSHUFD_XMM_to_XMM(xmmT1, to, 0x50); + // ZW + SSE2_PSHUFD_XMM_to_XMM(xmmT1, to, 0xfa); SSE2_PAND_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (xmmT1, (uptr)MIN_MAX_MASK2); - SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0x50); + SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0xfa); SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); if (min) SSE2_MINPD_XMM_to_XMM(xmmT1, xmmT2); else SSE2_MAXPD_XMM_to_XMM(xmmT1, xmmT2); - SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x88); - // ZW - SSE2_PSHUFD_XMM_to_XMM(to, to, 0xfa); - SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); - SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); - SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0xfa); + // XY + SSE2_PSHUFD_XMM_to_XMM(xmmT2, from, 0x50); SSE2_PAND_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK1); SSE2_POR_M128_to_XMM (xmmT2, (uptr)MIN_MAX_MASK2); + SSE2_PSHUFD_XMM_to_XMM(to, to, 0x50); + SSE2_PAND_M128_to_XMM (to, (uptr)MIN_MAX_MASK1); + SSE2_POR_M128_to_XMM (to, (uptr)MIN_MAX_MASK2); if (min) SSE2_MINPD_XMM_to_XMM(to, xmmT2); else SSE2_MAXPD_XMM_to_XMM(to, xmmT2); - SSE2_PSHUFD_XMM_to_XMM(to, to, 0x88); - SSE2_MOVSD_XMM_to_XMM (to, xmmT1); + + SSE_SHUFPS_XMM_to_XMM(to, xmmT1, 0x88); } // Warning: Modifies from and to's upper 3 vectors @@ -370,5 +369,4 @@ void SSE_MIN2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { else { MIN_MAX_SS(to, from, 1); } } - #endif //PCSX2_MICROVU From 134f02babbb8b7f7afcda9feb3523d05a0afd538 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 18 May 2009 13:08:35 +0000 Subject: [PATCH 093/296] Linux/microVU: Uninline a function that isn't inlinable according to gcc, so microVU is compilable on Linux. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1221 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 2 +- pcsx2/x86/microVU.h | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index d4564b78c9..971dcc01be 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -55,7 +55,7 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr) { } // Resets Rec Data -microVUt(void) mVUreset() { +microVUx(void) mVUreset() { microVU* mVU = mVUx; mVUprint((vuIndex) ? "microVU1: reset" : "microVU0: reset"); diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 827102865d..079b086633 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -135,7 +135,7 @@ extern void (*mVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); // Main Functions microVUt(void) mVUinit(VURegs*); -microVUt(void) mVUreset(); +microVUx(void) mVUreset(); microVUt(void) mVUclose(); microVUt(void) mVUclear(u32, u32); @@ -151,7 +151,7 @@ microVUf(void) mVUopL(); microVUt(void) mVUclearProg(microVU* mVU, int progIndex); microVUt(int) mVUfindLeastUsedProg(microVU* mVU); microVUt(int) mVUsearchProg(); -microVUt(void) mVUcacheProg(microVU* mVU, int progIndex); +microVUt(void) mVUcacheProg(int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); From 9eda5dc951a722baf33f5452d4cb591f71319bea Mon Sep 17 00:00:00 2001 From: arcum42 Date: Tue, 19 May 2009 13:25:09 +0000 Subject: [PATCH 094/296] Linux: Fear and Loathing in GCC 4.4.0... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1222 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Defs.h | 27 +++++++++++++++ pcsx2/configure.ac | 71 +++++++++++++++++++++++++++----------- 2 files changed, 78 insertions(+), 20 deletions(-) diff --git a/common/include/Pcsx2Defs.h b/common/include/Pcsx2Defs.h index 85c83d4a66..cb2aa7937c 100644 --- a/common/include/Pcsx2Defs.h +++ b/common/include/Pcsx2Defs.h @@ -98,6 +98,29 @@ default: \ #else +// GCC 4.4.0 is a bit nutty, as compilers go. it gets a define to itself. +#define GCC_VERSION (__GNUC__ * 10000 \ + + __GNUC_MINOR__ * 100 \ + + __GNUC_PATCHLEVEL__) + +/* Test for GCC > 4.4.0; Should be adjusted when new versions come out */ +#if GCC_VERSION >= 40400 +#define THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 +#define __nooptimization __attribute__((optimize("O0")) +#endif + +/* +This theoretically unoptimizes. Not having much luck so far. +#ifdef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 +#pragma GCC optimize ("O0") +#endif + +#ifdef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 +#pragma GCC reset_options +#endif + +*/ + // fixme - is this needed for recent versions of GCC? Or can we just use the first two macros // instead for both definitions (implementations) and declarations (includes)? -- air #define PCSX2_ALIGNED(alig,x) x __attribute((aligned(alig))) @@ -116,6 +139,10 @@ default: \ #define __noinline __attribute__((noinline)) #endif +#ifndef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 +#define __nooptimization +#endif + typedef struct { int size; s8 *data; diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index cb8bdeebe6..d6e41960b1 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -117,28 +117,47 @@ fi WARNING_FLAGS+=" -Wno-format -Wno-unused-parameter -Wno-unused-value " NORMAL_FLAGS+=" -msse -msse2 ${WARNING_FLAGS} " -if test "x$unsupportedspeedfreak" == "xyes" -then - dnl Don't even try to get support for this. - NORMAL_FLAGS+=" -pipe -O3 -march=native -funroll-loops -ffast-math " -else - if test "x$optimization" == "xyes" - then - NORMAL_FLAGS+=" -pipe -O2 " - else - NORMAL_FLAGS+=" -O0 " - fi -fi - -if test "x$unsupportedspeedfreak" == "xno" -then dnl These optimizations seem to cause issues with GCC 4.3.3, so we'll turn them off. dnl I really need to run some tests and see if they are still needed. - if test "$gccvernum" -ge "433"; then - NORMAL_FLAGS+=" -fno-guess-branch-probability -fno-dse -fno-tree-dse " - fi -fi - + if test "$gccvernum" -ge "433"; then + NORMAL_FLAGS+=" -fno-guess-branch-probability -fno-dse -fno-tree-dse " + fi + +dnl Anything higher then -O0 seem to cause issues with GCC 4.4.0. + if test "x$optimization" == "xyes" + then + dnl Oddly, turning on all the optimizations up to -O2 one by one doesn't cause issues in GCC 4.4.0. + if test "$gccvernum" -ge "440"; then + NORMAL_FLAGS+=" -O0 " + dnl -O1, flag by flag, minus the 3 taken out on gcc 4.3.3 + NORMAL_FLAGS+=" -ftree-builtin-call-dce -ftree-ccp -ftree-ch -ftree-copyrename " + NORMAL_FLAGS+=" -ftree-dce -ftree-dominator-opts " + NORMAL_FLAGS+=" -ftree-fre " + NORMAL_FLAGS+=" -ftree-sra " + NORMAL_FLAGS+=" -ftree-ter " + NORMAL_FLAGS+=" -fauto-inc-dec -fcprop-registers -fdce -fdefer-pop " + NORMAL_FLAGS+=" -fif-conversion2 -fif-conversion " + NORMAL_FLAGS+=" -finline-small-functions -fipa-pure-const -fipa-reference -fmerge-constants " + NORMAL_FLAGS+=" -fsplit-wide-types -funit-at-a-time -fno-omit-frame-pointer" + dnl And now for -O2. Comment out one that breaks compilation when turned on individually. + NORMAL_FLAGS+=" -fthread-jumps -falign-functions -falign-jumps -falign-loops -falign-labels " + NORMAL_FLAGS+=" -fcaller-saves -fcrossjumping -fcse-follow-jumps -fcse-skip-blocks " + NORMAL_FLAGS+=" -fdelete-null-pointer-checks -fexpensive-optimizations -fgcse -fgcse-lm " + NORMAL_FLAGS+=" -findirect-inlining -foptimize-sibling-calls -fpeephole2 -fregmove " + NORMAL_FLAGS+=" -freorder-blocks -freorder-functions -frerun-cse-after-loop " + NORMAL_FLAGS+=" -fsched-interblock " + NORMAL_FLAGS+=" -fsched-spec " + dnl NORMAL_FLAGS+=" -fschedule-insns " + NORMAL_FLAGS+=" -fschedule-insns2 " + NORMAL_FLAGS+=" -fstrict-aliasing -fstrict-overflow " + NORMAL_FLAGS+=" -ftree-switch-conversion -ftree-pre -ftree-vrp " + else + NORMAL_FLAGS+=" -pipe -O2 " + fi + else + NORMAL_FLAGS+=" -O0 " + fi + CFLAGS+=" ${NORMAL_FLAGS} -Xlinker -zmuldefs -m32 " CXXFLAGS+=" ${NORMAL_FLAGS} -fpermissive -Xlinker -zmuldefs -m32 " CCASFLAGS+=" ${NORMAL_FLAGS} -m32 " @@ -187,3 +206,15 @@ echo " Custom cflags? $customcflags" echo " Compiler warnings? $warnings" echo " Compiler optimizations? $optimization" + if test "$gccvernum" -ge "440"; then + echo + echo "*Congratulations*! You are running GCC 4.4.0! Thanks to this wonderful fact, you get" + echo "millions of command line switches, probably seriously slowing down your compilation time." + echo + echo "GCC 4.4.0 causes pcsx2 to crash at *any* optimization above 0, yet mysteriously" + echo "works properly if you add in the flags for all the optimizations yourself, which I have done." + echo "Enjoy!" + echo + fi + + From faf535a4611049bc4a674f90edef57cf0a3eae58 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Tue, 19 May 2009 23:37:20 +0000 Subject: [PATCH 095/296] Added COMI.SS/COMI.SD to the emitter (and moved UCOMI from moremovs to comparisons, where it belongs -- oops) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1224 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/ix86/implement/xmm/comparisons.h | 16 ++++++ pcsx2/x86/ix86/implement/xmm/moremovs.h | 12 ----- pcsx2/x86/ix86/ix86_instructions.h | 63 +++++++++++----------- pcsx2/x86/ix86/ix86_legacy_sse.cpp | 1 + pcsx2/x86/ix86/ix86_simd.cpp | 30 +++++------ 5 files changed, 64 insertions(+), 58 deletions(-) diff --git a/pcsx2/x86/ix86/implement/xmm/comparisons.h b/pcsx2/x86/ix86/implement/xmm/comparisons.h index e74d25b1f9..ce6e12702e 100644 --- a/pcsx2/x86/ix86/implement/xmm/comparisons.h +++ b/pcsx2/x86/ix86/implement/xmm/comparisons.h @@ -55,6 +55,22 @@ public: SimdImpl_Compare() {} //GCWhat? }; +////////////////////////////////////////////////////////////////////////////////////////// +// Compare scalar floating point values and set EFLAGS (Ordered or Unordered) +// +template< bool Ordered > +class SimdImpl_COMI +{ +protected: + static const u16 OpcodeSSE = Ordered ? 0x2f : 0x2e; + +public: + const SimdImpl_DestRegSSE<0x00,OpcodeSSE> SS; + const SimdImpl_DestRegSSE<0x66,OpcodeSSE> SD; + + SimdImpl_COMI() {} +}; + ////////////////////////////////////////////////////////////////////////////////////////// // diff --git a/pcsx2/x86/ix86/implement/xmm/moremovs.h b/pcsx2/x86/ix86/implement/xmm/moremovs.h index fa30d4f441..4ea3a5a01d 100644 --- a/pcsx2/x86/ix86/implement/xmm/moremovs.h +++ b/pcsx2/x86/ix86/implement/xmm/moremovs.h @@ -165,18 +165,6 @@ public: } }; - -////////////////////////////////////////////////////////////////////////////////////////// -// -template< u8 AltPrefix, u16 OpcodeSSE > -class SimdImpl_UcomI -{ -public: - const SimdImpl_DestRegSSE<0x00,OpcodeSSE> SS; - const SimdImpl_DestRegSSE SD; - SimdImpl_UcomI() {} -}; - ////////////////////////////////////////////////////////////////////////////////////////// // Blend - Conditional copying of values in src into dest. // diff --git a/pcsx2/x86/ix86/ix86_instructions.h b/pcsx2/x86/ix86/ix86_instructions.h index 889848d4f8..fd12beaa2d 100644 --- a/pcsx2/x86/ix86/ix86_instructions.h +++ b/pcsx2/x86/ix86/ix86_instructions.h @@ -451,16 +451,17 @@ namespace x86Emitter extern const Internal::SimdImpl_DestRegEither<0x66,0xeb> xPOR; extern const Internal::SimdImpl_DestRegEither<0x66,0xef> xPXOR; - extern const Internal::SimdImpl_AndNot xANDN; + extern const Internal::SimdImpl_AndNot xANDN; - extern const Internal::SimdImpl_UcomI<0x66,0x2e> xUCOMI; - extern const Internal::SimdImpl_rSqrt<0x53> xRCP; - extern const Internal::SimdImpl_rSqrt<0x52> xRSQRT; - extern const Internal::SimdImpl_Sqrt<0x51> xSQRT; + extern const Internal::SimdImpl_COMI xCOMI; + extern const Internal::SimdImpl_COMI xUCOMI; + extern const Internal::SimdImpl_rSqrt<0x53> xRCP; + extern const Internal::SimdImpl_rSqrt<0x52> xRSQRT; + extern const Internal::SimdImpl_Sqrt<0x51> xSQRT; - extern const Internal::SimdImpl_MinMax<0x5f> xMAX; - extern const Internal::SimdImpl_MinMax<0x5d> xMIN; - extern const Internal::SimdImpl_Shuffle<0xc6> xSHUF; + extern const Internal::SimdImpl_MinMax<0x5f> xMAX; + extern const Internal::SimdImpl_MinMax<0x5d> xMIN; + extern const Internal::SimdImpl_Shuffle<0xc6> xSHUF; // ------------------------------------------------------------------------ @@ -510,35 +511,35 @@ namespace x86Emitter // ------------------------------------------------------------------------ - extern const Internal::SimdImpl_Shift<0xf0, 6> xPSLL; - extern const Internal::SimdImpl_Shift<0xd0, 2> xPSRL; + extern const Internal::SimdImpl_Shift<0xf0, 6> xPSLL; + extern const Internal::SimdImpl_Shift<0xd0, 2> xPSRL; extern const Internal::SimdImpl_ShiftWithoutQ<0xe0, 4> xPSRA; - extern const Internal::SimdImpl_AddSub<0xdc, 0xd4> xPADD; - extern const Internal::SimdImpl_AddSub<0xd8, 0xfb> xPSUB; - extern const Internal::SimdImpl_PMinMax<0xde,0x3c> xPMAX; - extern const Internal::SimdImpl_PMinMax<0xda,0x38> xPMIN; + extern const Internal::SimdImpl_AddSub<0xdc, 0xd4> xPADD; + extern const Internal::SimdImpl_AddSub<0xd8, 0xfb> xPSUB; + extern const Internal::SimdImpl_PMinMax<0xde,0x3c> xPMAX; + extern const Internal::SimdImpl_PMinMax<0xda,0x38> xPMIN; - extern const Internal::SimdImpl_PMul xPMUL; - extern const Internal::SimdImpl_PCompare xPCMP; - extern const Internal::SimdImpl_PShuffle xPSHUF; - extern const Internal::SimdImpl_PUnpack xPUNPCK; - extern const Internal::SimdImpl_Unpack xUNPCK; - extern const Internal::SimdImpl_Pack xPACK; + extern const Internal::SimdImpl_PMul xPMUL; + extern const Internal::SimdImpl_PCompare xPCMP; + extern const Internal::SimdImpl_PShuffle xPSHUF; + extern const Internal::SimdImpl_PUnpack xPUNPCK; + extern const Internal::SimdImpl_Unpack xUNPCK; + extern const Internal::SimdImpl_Pack xPACK; - extern const Internal::SimdImpl_PAbsolute xPABS; - extern const Internal::SimdImpl_PSign xPSIGN; - extern const Internal::SimdImpl_PInsert xPINSR; - extern const Internal::SimdImpl_PExtract xPEXTR; - extern const Internal::SimdImpl_PMultAdd xPMADD; - extern const Internal::SimdImpl_HorizAdd xHADD; + extern const Internal::SimdImpl_PAbsolute xPABS; + extern const Internal::SimdImpl_PSign xPSIGN; + extern const Internal::SimdImpl_PInsert xPINSR; + extern const Internal::SimdImpl_PExtract xPEXTR; + extern const Internal::SimdImpl_PMultAdd xPMADD; + extern const Internal::SimdImpl_HorizAdd xHADD; - extern const Internal::SimdImpl_Blend xBLEND; - extern const Internal::SimdImpl_DotProduct xDP; - extern const Internal::SimdImpl_Round xROUND; + extern const Internal::SimdImpl_Blend xBLEND; + extern const Internal::SimdImpl_DotProduct xDP; + extern const Internal::SimdImpl_Round xROUND; - extern const Internal::SimdImpl_PMove xPMOVSX; - extern const Internal::SimdImpl_PMove xPMOVZX; + extern const Internal::SimdImpl_PMove xPMOVSX; + extern const Internal::SimdImpl_PMove xPMOVZX; } diff --git a/pcsx2/x86/ix86/ix86_legacy_sse.cpp b/pcsx2/x86/ix86/ix86_legacy_sse.cpp index 169f6ac100..ed1bddcbce 100644 --- a/pcsx2/x86/ix86/ix86_legacy_sse.cpp +++ b/pcsx2/x86/ix86/ix86_legacy_sse.cpp @@ -209,6 +209,7 @@ DEFINE_LEGACY_CMP_OPCODE( NLT ) DEFINE_LEGACY_CMP_OPCODE( NLE ) DEFINE_LEGACY_CMP_OPCODE( ORD ) +DEFINE_LEGACY_SSSD_OPCODE( COMI ) DEFINE_LEGACY_SSSD_OPCODE( UCOMI ) DEFINE_LEGACY_RSQRT_OPCODE( RCP ) DEFINE_LEGACY_RSQRT_OPCODE( RSQRT ) diff --git a/pcsx2/x86/ix86/ix86_simd.cpp b/pcsx2/x86/ix86/ix86_simd.cpp index 860e8ffc0a..2c1c18f469 100644 --- a/pcsx2/x86/ix86/ix86_simd.cpp +++ b/pcsx2/x86/ix86/ix86_simd.cpp @@ -90,27 +90,27 @@ const SimdImpl_MoveSSE<0x66,true> xMOVAPD; const SimdImpl_MoveSSE<0x66,false> xMOVUPD; #endif -const MovhlImplAll<0x16> xMOVH; -const MovhlImplAll<0x12> xMOVL; -const MovhlImpl_RtoR<0x16> xMOVLH; -const MovhlImpl_RtoR<0x12> xMOVHL; +const MovhlImplAll<0x16> xMOVH; +const MovhlImplAll<0x12> xMOVL; +const MovhlImpl_RtoR<0x16> xMOVLH; +const MovhlImpl_RtoR<0x12> xMOVHL; + +const SimdImpl_AndNot xANDN; +const SimdImpl_COMI xCOMI; +const SimdImpl_COMI xUCOMI; +const SimdImpl_rSqrt<0x53> xRCP; +const SimdImpl_rSqrt<0x52> xRSQRT; +const SimdImpl_Sqrt<0x51> xSQRT; + +const SimdImpl_MinMax<0x5f> xMAX; +const SimdImpl_MinMax<0x5d> xMIN; +const SimdImpl_Shuffle<0xc6> xSHUF; const SimdImpl_DestRegEither<0x66,0xdb> xPAND; const SimdImpl_DestRegEither<0x66,0xdf> xPANDN; const SimdImpl_DestRegEither<0x66,0xeb> xPOR; const SimdImpl_DestRegEither<0x66,0xef> xPXOR; -const SimdImpl_AndNot xANDN; - -const SimdImpl_UcomI<0x66,0x2e> xUCOMI; -const SimdImpl_rSqrt<0x53> xRCP; -const SimdImpl_rSqrt<0x52> xRSQRT; -const SimdImpl_Sqrt<0x51> xSQRT; - -const SimdImpl_MinMax<0x5f> xMAX; -const SimdImpl_MinMax<0x5d> xMIN; -const SimdImpl_Shuffle<0xc6> xSHUF; - // ------------------------------------------------------------------------ // [SSE-4.1] Performs a bitwise AND of dest against src, and sets the ZF flag From 23c833e922c406094b3942b5b7a5745e09850461 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 20 May 2009 03:05:36 +0000 Subject: [PATCH 096/296] microVU: - Fixed a Q/P instance bug. This fixed some graphical corruption in Haunting Ground, might fix some other games too. - Minor Changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1225 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 1 - pcsx2/x86/microVU_Alloc.inl | 14 +++++++------- pcsx2/x86/microVU_Compile.inl | 16 +++++++++------- pcsx2/x86/microVU_Lower.inl | 4 ++-- pcsx2/x86/microVU_Misc.inl | 4 ++-- 5 files changed, 20 insertions(+), 19 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 079b086633..de834dd70a 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -115,7 +115,6 @@ struct microVU { u8* startFunct; // Ptr Function to the Start code for recompiled programs u8* exitFunct; // Ptr Function to the Exit code for recompiled programs u32 code; // Contains the current Instruction - u32 iReg; // iReg (only used in recompilation, not execution) u32 divFlag; // 1 instance of I/D flags u32 VIbackup[2]; // Holds a backup of a VI reg if modified before a branch u32 branch; // Holds branch compare result (IBxx) OR Holds address to Jump to (JALR/JR) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index f9e7111a07..34a2732195 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -29,7 +29,7 @@ #define getReg(reg, _reg_) { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _X_Y_Z_W); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, _X_Y_Z_W); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _X_Y_Z_W); \ } #define getZero(reg) { \ @@ -83,12 +83,12 @@ microVUt(void) mVUallocFMAC2b(int& Ft) { #define getReg3SS(reg, _reg_) { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, (1 << (3 - _bc_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ } #define getReg3(reg, _reg_) { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, (1 << (3 - _bc_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ mVUunpack_xyzw(reg, reg, 0); \ } @@ -135,7 +135,7 @@ microVUt(void) mVUallocFMAC3b(int& Fd) { #define getReg4(reg, _reg_) { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _xyzw_ACC); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, _xyzw_ACC); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _xyzw_ACC); \ } #define getZero4(reg) { \ @@ -204,7 +204,7 @@ microVUt(void) mVUallocFMAC5b(int& ACC, int& Fs) { #define getIreg(reg, modXYZW) { \ MOV32MtoR(gprT1, (uptr)&mVU->regs->VI[REG_I].UL); \ SSE2_MOVD_R_to_XMM(reg, gprT1); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 8); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 8); \ if (!((_XYZW_SS && modXYZW) || (_X_Y_Z_W == 8))) { mVUunpack_xyzw(reg, reg, 0); } \ } @@ -496,7 +496,7 @@ microVUt(void) mVUallocFMAC16b(int& ACCw, int& ACCr) { #define getReg9(reg, _reg_) { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], 1); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 1); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 1); \ mVUunpack_xyzw(reg, reg, 0); \ } @@ -754,7 +754,7 @@ microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { } \ else { \ mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _fxf_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, (1 << (3 - _fxf_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _fxf_))); \ } \ } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 8639ace77e..e7c6730bed 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -135,16 +135,14 @@ microVUt(void) mVUsetCycles() { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUendProgram(int fStatus, int fMac, int fClip) { +microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fClip) { microVU* mVU = mVUx; - incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) - mVUcycles -= 100; // Save P/Q Regs - if (mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } + if (qInst) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_Q].UL, xmmPQ); if (vuIndex) { - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVU->p ? 3 : 2); + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, pInst ? 3 : 2); SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_P].UL, xmmPQ); } @@ -180,7 +178,7 @@ microVUt(void) mVUtestCycles() { PUSH32R(gprR); CALLFunc((uptr)mVUwarning); POP32R(gprR); - mVUendProgram(sI, 0, cI); + mVUendProgram(0, 0, sI, 0, cI); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } @@ -345,9 +343,13 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile ebit"); if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } + + incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) + mVUcycles -= 100; + // Do E-bit end stuff here mVUsetupRange(xPC - 8); - mVUendProgram(findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); + mVUendProgram(mVU->q, mVU->p, findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); return thisPtr; //ToDo: Save pipeline state? } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index b0f4b5dd1e..d85db8f2ff 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -70,7 +70,7 @@ microVUf(void) mVU_DIV() { x86SetJ8(cjmp); MOV32ItoM((uptr)&mVU->divFlag, 0); // Clear I/D flags SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - mVUclamp1(xmmFs, xmmFt, 8); + if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); @@ -128,7 +128,7 @@ microVUf(void) mVU_RSQRT() { djmp = JMP8(0); x86SetJ8(ajmp); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - mVUclamp1(xmmFs, xmmFt, 8); + if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index adf3ac37bc..0011b852cb 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -99,7 +99,7 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { if (modXYZW && (xyzw == 8 || xyzw == 4 || xyzw == 2 || xyzw == 1)) { mVUunpack_xyzw(reg, reg, 0); } - mVUmergeRegs(xmmT2, reg, xyzw); + mVUmergeRegs(xmmT2, reg, xyzw); SSE_MOVAPS_XMM_to_M128(offset, xmmT2); return;*/ @@ -158,7 +158,7 @@ microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { if (xyzw == 8 || xyzw == 4 || xyzw == 2 || xyzw == 1) { mVUunpack_xyzw(reg, reg, 0); } - mVUmergeRegs(xmmT2, reg, xyzw); + mVUmergeRegs(xmmT2, reg, xyzw); SSE_MOVAPSRtoRm(gprReg, xmmT2, offset); return;*/ From 09ee76d9eddd4c3c09e7cf62bd1a19e5bb33647b Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 20 May 2009 15:35:31 +0000 Subject: [PATCH 097/296] GSdx: optimized local-local transfer a bit, it boosts fps in games like ffxii or ico, which do a lot of moves in local memory. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1226 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GS.h | 5 +- plugins/GSdx/GSClut.cpp | 14 ---- plugins/GSdx/GSDirtyRect.cpp | 7 +- plugins/GSdx/GSLocalMemory.cpp | 7 +- plugins/GSdx/GSRenderer.h | 38 +++------ plugins/GSdx/GSRendererHW.h | 8 +- plugins/GSdx/GSRendererSW.h | 2 +- plugins/GSdx/GSState.cpp | 133 ++++++++++++++++++++++-------- plugins/GSdx/GSState.h | 2 +- plugins/GSdx/GSTextureCache.h | 2 +- plugins/GSdx/GSTextureCache10.cpp | 33 +++----- plugins/GSdx/GSTextureCache9.cpp | 32 +++---- plugins/GSdx/GSTextureCacheSW.cpp | 23 ++---- plugins/GSdx/GSTextureCacheSW.h | 4 +- plugins/GSdx/GSVector.h | 36 +++++++- plugins/GSdx/GSdx_vs2008.vcproj | 8 ++ 16 files changed, 194 insertions(+), 160 deletions(-) diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index 39a799a0f1..e62751a8eb 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -221,6 +221,8 @@ union name \ uint64 u64; \ uint32 u32[2]; \ void operator = (const GSVector4i& v) {GSVector4i::storel(this, v);} \ + bool operator == (const union name& r) const {return ((GSVector4i)r).eq(*this);} \ + bool operator != (const union name& r) const {return !((GSVector4i)r).eq(*this);} \ operator GSVector4i() const {return GSVector4i::loadl(this);} \ struct { \ @@ -864,7 +866,8 @@ REG64_(GIFReg, TRXPOS) uint32 DSAX:11; uint32 _PAD3:5; uint32 DSAY:11; - uint32 DIR:2; + uint32 DIRY:1; + uint32 DIRX:1; uint32 _PAD4:3; REG_END diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index 879764e108..c6e3385fd4 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -598,7 +598,6 @@ __forceinline void GSClut::ExpandCLUT64_T16(const GSVector4i& hi, const GSVector // TODO -static const GSVector4i s_am(0x00008000); static const GSVector4i s_bm(0x00007c00); static const GSVector4i s_gm(0x000003e0); static const GSVector4i s_rm(0x0000001f); @@ -610,7 +609,6 @@ void GSClut::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, c const GSVector4i rm = s_rm; const GSVector4i gm = s_gm; const GSVector4i bm = s_bm; - // const GSVector4i am = s_am; GSVector4i TA0(TEXA.TA0 << 24); GSVector4i TA1(TEXA.TA1 << 24); @@ -625,12 +623,6 @@ void GSClut::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, c for(int i = 0, j = w >> 3; i < j; i++) { c = s[i]; - /* - cl = c.upl16(); - ch = c.uph16(); - d[i * 2 + 0] = ((cl & rm) << 3) | ((cl & gm) << 6) | ((cl & bm) << 9) | TA1.blend(TA0, cl < am); - d[i * 2 + 1] = ((ch & rm) << 3) | ((ch & gm) << 6) | ((ch & bm) << 9) | TA1.blend(TA0, ch < am); - */ cl = c.upl16(c); ch = c.uph16(c); d[i * 2 + 0] = ((cl & rm) << 3) | ((cl & gm) << 6) | ((cl & bm) << 9) | TA0.blend8(TA1, cl.sra16(15)); @@ -642,12 +634,6 @@ void GSClut::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, c for(int i = 0, j = w >> 3; i < j; i++) { c = s[i]; - /* - cl = c.upl16(); - ch = c.uph16(); - d[i * 2 + 0] = ((cl & rm) << 3) | ((cl & gm) << 6) | ((cl & bm) << 9) | TA1.blend(TA0, cl < am).andnot(cl == GSVector4i::zero()); - d[i * 2 + 1] = ((ch & rm) << 3) | ((ch & gm) << 6) | ((ch & bm) << 9) | TA1.blend(TA0, ch < am).andnot(ch == GSVector4i::zero()); - */ cl = c.upl16(c); ch = c.uph16(c); d[i * 2 + 0] = ((cl & rm) << 3) | ((cl & gm) << 6) | ((cl & bm) << 9) | TA0.blend8(TA1, cl.sra16(15)).andnot(cl == GSVector4i::zero()); diff --git a/plugins/GSdx/GSDirtyRect.cpp b/plugins/GSdx/GSDirtyRect.cpp index 9815a4a28c..4276d0d47e 100644 --- a/plugins/GSdx/GSDirtyRect.cpp +++ b/plugins/GSdx/GSDirtyRect.cpp @@ -48,16 +48,13 @@ GSVector4i GSDirtyRect::GetDirtyRect(const GIFRegTEX0& TEX0) GSVector2i dst = GSLocalMemory::m_psm[TEX0.PSM].bs; r.left = MulDiv(left, dst.x, src.x); - r.right = MulDiv(right, dst.x, src.x); r.top = MulDiv(top, dst.y, src.y); + r.right = MulDiv(right, dst.x, src.x); r.bottom = MulDiv(bottom, dst.y, src.y); } else { - r.left = left & ~(src.x - 1); - r.right = (right + (src.x - 1) /* + 1 */) & ~(src.x - 1); - r.top = top & ~(src.y - 1); - r.bottom = (bottom + (src.y - 1) /* + 1 */) & ~(src.y - 1); + r = GSVector4i(left, top, right, bottom).ralign(src); } return r; diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index b9c499518d..b5f24fbb56 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -593,13 +593,8 @@ bool GSLocalMemory::FillRect(const GSVector4i& r, uint32 c, uint32 psm, uint32 b case 4: c = (c & 0xf) * 0x11111111; break; } - GSVector4i clip; + GSVector4i clip = r.ralign(tbl.bs); - clip.left = (r.left + (w - 1)) & ~(w - 1); - clip.top = (r.top + (h - 1)) & ~(h - 1); - clip.right = r.right & ~(w - 1); - clip.bottom = r.bottom & ~(h - 1); - for(int y = r.top; y < clip.top; y++) { for(int x = r.left; x < r.right; x++) diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index a5fc7f3577..8722cf12cd 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -249,7 +249,7 @@ protected: if(dr[i].height() > 512) // hmm { - int y = GetDeviceRect(i).height(); + int y = GetDeviceSize(i).y; if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) y /= 2; r.bottom = r.top + y; } @@ -258,35 +258,26 @@ protected: if(m_blur && blurdetected && i == 1) { - src[i].x = tex[i].m_scale.x * r.left / tex[i].GetWidth(); - src[i].y = (tex[i].m_scale.y * r.top + 1) / tex[i].GetHeight(); - src[i].z = tex[i].m_scale.x * r.right / tex[i].GetWidth(); - src[i].w = (tex[i].m_scale.y * r.bottom + 1) / tex[i].GetHeight(); - } - else - { - src[i].x = tex[i].m_scale.x * r.left / tex[i].GetWidth(); - src[i].y = tex[i].m_scale.y * r.top / tex[i].GetHeight(); - src[i].z = tex[i].m_scale.x * r.right / tex[i].GetWidth(); - src[i].w = tex[i].m_scale.y * r.bottom / tex[i].GetHeight(); + r += GSVector4i(0, 1).xyxy(); } - GSVector2 o; + GSVector4 scale = GSVector4(tex[i].m_scale).xyxy(); - o.x = 0; - o.y = 0; + src[i] = GSVector4(r) * scale / GSVector4(tex[i].GetSize()).xyxy(); + + GSVector2 o(0, 0); if(dr[i].top - baseline >= 4) // 2? { o.y = tex[i].m_scale.y * (dr[i].top - baseline); } - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) o.y /= 2; + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) + { + o.y /= 2; + } - dst[i].x = o.x; - dst[i].y = o.y; - dst[i].z = o.x + tex[i].m_scale.x * r.width(); - dst[i].w = o.y + tex[i].m_scale.y * r.height(); + dst[i] = GSVector4(o).xyxy() + scale * GSVector4(r.rsize()); fs.x = max(fs.x, (int)(dst[i].z + 0.5f)); fs.y = max(fs.y, (int)(dst[i].w + 0.5f)); @@ -301,12 +292,7 @@ protected: if(tex[0] || tex[1]) { - GSVector4 c; - - c.r = (float)m_regs->BGCOLOR.R / 255; - c.g = (float)m_regs->BGCOLOR.G / 255; - c.b = (float)m_regs->BGCOLOR.B / 255; - c.a = (float)m_regs->PMODE.ALP / 255; + GSVector4 c = GSVector4((int)m_regs->BGCOLOR.R, (int)m_regs->BGCOLOR.G, (int)m_regs->BGCOLOR.B, (int)m_regs->PMODE.ALP) / 255; m_dev.Merge(tex, src, dst, fs, slbg, mmod, c); diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 34df556fbb..609790327e 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -93,7 +93,7 @@ protected: maxv = maxv.maxv(v0); } - mm = minv.xyxy(maxv) * GSVector4(16 << m_context->TEX0.TW, 16 << m_context->TEX0.TH, 16 << m_context->TEX0.TW, 16 << m_context->TEX0.TH).rcpnr(); + mm = minv.xyxy(maxv) * GSVector4(16 << m_context->TEX0.TW, 16 << m_context->TEX0.TH).xyxy().rcpnr(); } else { @@ -211,12 +211,8 @@ protected: r = vr + GSVector4i(-1, -1, 1, 1); // one more pixel because of bilinear filtering GSVector2i bs = GSLocalMemory::m_psm[m_context->TEX0.PSM].bs; - GSVector2i bsm(bs.x - 1, bs.y - 1); - r.left = max(r.left & ~bsm.x, 0); - r.top = max(r.top & ~bsm.y, 0); - r.right = min((r.right + bsm.x) & ~bsm.x, w); - r.bottom = min((r.bottom + bsm.y) & ~bsm.y, h); + r = r.ralign(bs).rintersect(GSVector4i(0, 0, w, h)); } void VSync(int field) diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 8c3d50bc42..aa70b60454 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -449,7 +449,7 @@ protected: MinMaxUV(w, h, r, p.sel.fst); - const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, &r); + const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, r); if(!t) {ASSERT(0); return;} diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 20fd4124e1..31300861bf 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -235,7 +235,7 @@ GSVector4i GSState::GetFrameRect(int i) return r; } -GSVector4i GSState::GetDeviceRect(int i) +GSVector2i GSState::GetDeviceSize(int i) { // TODO: other params of SMODE1 should affect the true device display size @@ -257,7 +257,7 @@ GSVector4i GSState::GetDeviceRect(int i) h = (m_regs->SMODE1.CMOD & 1) ? 512 : 448; } - return GSVector4i(0, 0, w, h); + return GSVector2i(w, h); } @@ -279,7 +279,7 @@ bool GSState::IsEnabled(int i) int GSState::GetFPS() { - return ((m_regs->SMODE1.CMOD & 1) ? 50 : 60) / (m_regs->SMODE2.INT ? 1 : 2); + return ((m_regs->SMODE1.CMOD & 1) ? 50 : 60) >> (1 - m_regs->SMODE2.INT); } // GIFPackedRegHandler* @@ -486,7 +486,7 @@ template void GSState::GIFRegHandlerTEX0(GIFReg* r) bool wt = m_mem.m_clut.WriteTest(r->TEX0, m_env.TEXCLUT); - if(wt || PRIM->CTXT == i && !((GSVector4i)r->TEX0).eq(m_env.CTXT[i].TEX0)) + if(wt || PRIM->CTXT == i && r->TEX0 != m_env.CTXT[i].TEX0) { Flush(); } @@ -511,7 +511,7 @@ template void GSState::GIFRegHandlerTEX0(GIFReg* r) template void GSState::GIFRegHandlerCLAMP(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->CLAMP).eq(m_env.CTXT[i].CLAMP)) + if(PRIM->CTXT == i && r->CLAMP != m_env.CTXT[i].CLAMP) { Flush(); } @@ -552,7 +552,7 @@ void GSState::GIFRegHandlerNOP(GIFReg* r) template void GSState::GIFRegHandlerTEX1(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->TEX1).eq(m_env.CTXT[i].TEX1)) + if(PRIM->CTXT == i && r->TEX1 != m_env.CTXT[i].TEX1) { Flush(); } @@ -587,7 +587,7 @@ template void GSState::GIFRegHandlerXYOFFSET(GIFReg* r) void GSState::GIFRegHandlerPRMODECONT(GIFReg* r) { - if(!((GSVector4i)r->PRMODECONT).eq(m_env.PRMODECONT)) + if(r->PRMODECONT != m_env.PRMODECONT) { Flush(); } @@ -621,7 +621,7 @@ void GSState::GIFRegHandlerPRMODE(GIFReg* r) void GSState::GIFRegHandlerTEXCLUT(GIFReg* r) { - if(!((GSVector4i)r->TEXCLUT).eq(m_env.TEXCLUT)) + if(r->TEXCLUT != m_env.TEXCLUT) { Flush(); } @@ -631,7 +631,7 @@ void GSState::GIFRegHandlerTEXCLUT(GIFReg* r) void GSState::GIFRegHandlerSCANMSK(GIFReg* r) { - if(!((GSVector4i)r->SCANMSK).eq(m_env.SCANMSK)) + if(r->SCANMSK != m_env.SCANMSK) { Flush(); } @@ -641,7 +641,7 @@ void GSState::GIFRegHandlerSCANMSK(GIFReg* r) template void GSState::GIFRegHandlerMIPTBP1(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->MIPTBP1).eq(m_env.CTXT[i].MIPTBP1)) + if(PRIM->CTXT == i && r->MIPTBP1 != m_env.CTXT[i].MIPTBP1) { Flush(); } @@ -651,7 +651,7 @@ template void GSState::GIFRegHandlerMIPTBP1(GIFReg* r) template void GSState::GIFRegHandlerMIPTBP2(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->MIPTBP2).eq(m_env.CTXT[i].MIPTBP2)) + if(PRIM->CTXT == i && r->MIPTBP2 != m_env.CTXT[i].MIPTBP2) { Flush(); } @@ -661,7 +661,7 @@ template void GSState::GIFRegHandlerMIPTBP2(GIFReg* r) void GSState::GIFRegHandlerTEXA(GIFReg* r) { - if(!((GSVector4i)r->TEXA).eq(m_env.TEXA)) + if(r->TEXA != m_env.TEXA) { Flush(); } @@ -671,7 +671,7 @@ void GSState::GIFRegHandlerTEXA(GIFReg* r) void GSState::GIFRegHandlerFOGCOL(GIFReg* r) { - if(!((GSVector4i)r->FOGCOL).eq(m_env.FOGCOL)) + if(r->FOGCOL != m_env.FOGCOL) { Flush(); } @@ -688,7 +688,7 @@ void GSState::GIFRegHandlerTEXFLUSH(GIFReg* r) template void GSState::GIFRegHandlerSCISSOR(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->SCISSOR).eq(m_env.CTXT[i].SCISSOR)) + if(PRIM->CTXT == i && r->SCISSOR != m_env.CTXT[i].SCISSOR) { Flush(); } @@ -705,7 +705,7 @@ template void GSState::GIFRegHandlerALPHA(GIFReg* r) ASSERT(r->ALPHA.C != 3); ASSERT(r->ALPHA.D != 3); - if(PRIM->CTXT == i && !((GSVector4i)r->ALPHA).eq(m_env.CTXT[i].ALPHA)) + if(PRIM->CTXT == i && r->ALPHA != m_env.CTXT[i].ALPHA) { Flush(); } @@ -721,7 +721,7 @@ void GSState::GIFRegHandlerDIMX(GIFReg* r) { bool update = false; - if(!((GSVector4i)r->DIMX).eq(m_env.DIMX)) + if(r->DIMX != m_env.DIMX) { Flush(); @@ -738,7 +738,7 @@ void GSState::GIFRegHandlerDIMX(GIFReg* r) void GSState::GIFRegHandlerDTHE(GIFReg* r) { - if(!((GSVector4i)r->DTHE).eq(m_env.DTHE)) + if(r->DTHE != m_env.DTHE) { Flush(); } @@ -748,7 +748,7 @@ void GSState::GIFRegHandlerDTHE(GIFReg* r) void GSState::GIFRegHandlerCOLCLAMP(GIFReg* r) { - if(!((GSVector4i)r->COLCLAMP).eq(m_env.COLCLAMP)) + if(r->COLCLAMP != m_env.COLCLAMP) { Flush(); } @@ -758,7 +758,7 @@ void GSState::GIFRegHandlerCOLCLAMP(GIFReg* r) template void GSState::GIFRegHandlerTEST(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->TEST).eq(m_env.CTXT[i].TEST)) + if(PRIM->CTXT == i && r->TEST != m_env.CTXT[i].TEST) { Flush(); } @@ -768,7 +768,7 @@ template void GSState::GIFRegHandlerTEST(GIFReg* r) void GSState::GIFRegHandlerPABE(GIFReg* r) { - if(!((GSVector4i)r->PABE).eq(m_env.PABE)) + if(r->PABE != m_env.PABE) { Flush(); } @@ -778,7 +778,7 @@ void GSState::GIFRegHandlerPABE(GIFReg* r) template void GSState::GIFRegHandlerFBA(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->FBA).eq(m_env.CTXT[i].FBA)) + if(PRIM->CTXT == i && r->FBA != m_env.CTXT[i].FBA) { Flush(); } @@ -788,7 +788,7 @@ template void GSState::GIFRegHandlerFBA(GIFReg* r) template void GSState::GIFRegHandlerFRAME(GIFReg* r) { - if(PRIM->CTXT == i && !((GSVector4i)r->FRAME).eq(m_env.CTXT[i].FRAME)) + if(PRIM->CTXT == i && r->FRAME != m_env.CTXT[i].FRAME) { Flush(); } @@ -807,7 +807,7 @@ template void GSState::GIFRegHandlerZBUF(GIFReg* r) r->ZBUF.PSM |= 0x30; - if(PRIM->CTXT == i && !((GSVector4i)r->ZBUF).eq(m_env.CTXT[i].ZBUF)) + if(PRIM->CTXT == i && r->ZBUF != m_env.CTXT[i].ZBUF) { Flush(); } @@ -825,7 +825,7 @@ template void GSState::GIFRegHandlerZBUF(GIFReg* r) void GSState::GIFRegHandlerBITBLTBUF(GIFReg* r) { - if(!((GSVector4i)r->BITBLTBUF).eq(m_env.BITBLTBUF)) + if(r->BITBLTBUF != m_env.BITBLTBUF) { FlushWrite(); } @@ -845,7 +845,7 @@ void GSState::GIFRegHandlerBITBLTBUF(GIFReg* r) void GSState::GIFRegHandlerTRXPOS(GIFReg* r) { - if(!((GSVector4i)r->TRXPOS).eq(m_env.TRXPOS)) + if(r->TRXPOS != m_env.TRXPOS) { FlushWrite(); } @@ -855,7 +855,7 @@ void GSState::GIFRegHandlerTRXPOS(GIFReg* r) void GSState::GIFRegHandlerTRXREG(GIFReg* r) { - if(!((GSVector4i)r->TRXREG).eq(m_env.TRXREG)) + if(r->TRXREG != m_env.TRXREG) { FlushWrite(); } @@ -1036,9 +1036,15 @@ void GSState::Move() int xinc = 1; int yinc = 1; - if(sx < dx) {sx += w - 1; dx += w - 1; xinc = -1;} - if(sy < dy) {sy += h - 1; dy += h - 1; yinc = -1;} - + if(m_env.TRXPOS.DIRX) {sx += w - 1; dx += w - 1; xinc = -1;} + if(m_env.TRXPOS.DIRY) {sy += h - 1; dy += h - 1; yinc = -1;} +/* + printf("%05x %d %d => %05x %d %d (%d%d), %d %d %d %d %d %d\n", + m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW, m_env.BITBLTBUF.SPSM, + m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, + m_env.TRXPOS.DIRX, m_env.TRXPOS.DIRY, + sx, sy, dx, dy, w, h); +*/ /* GSLocalMemory::readPixel rp = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].rp; GSLocalMemory::writePixel wp = GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM].wp; @@ -1051,19 +1057,80 @@ void GSState::Move() const GSLocalMemory::psm_t& spsm = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM]; const GSLocalMemory::psm_t& dpsm = GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM]; - if(m_env.BITBLTBUF.SPSM == PSM_PSMCT32 && m_env.BITBLTBUF.DPSM == PSM_PSMCT32) + if(spsm.trbpp == dpsm.trbpp && spsm.trbpp >= 16) + { + int* soffset = spsm.rowOffset[0]; + int* doffset = dpsm.rowOffset[0]; + + if(spsm.trbpp == 32) + { + for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) + { + uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + + for(int x = 0; x < w; x++, sx += xinc, dx += xinc) + { + m_mem.WritePixel32(dbase + doffset[dx], m_mem.ReadPixel32(sbase + soffset[sx])); + } + } + } + else if(spsm.trbpp == 24) + { + for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) + { + uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + + for(int x = 0; x < w; x++, sx += xinc, dx += xinc) + { + m_mem.WritePixel24(dbase + doffset[dx], m_mem.ReadPixel24(sbase + soffset[sx])); + } + } + } + else // if(spsm.trbpp == 16) + { + for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) + { + uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + + for(int x = 0; x < w; x++, sx += xinc, dx += xinc) + { + m_mem.WritePixel16(dbase + doffset[dx], m_mem.ReadPixel16(sbase + soffset[sx])); + } + } + } + } + else if(m_env.BITBLTBUF.SPSM == PSM_PSMT8 && m_env.BITBLTBUF.DPSM == PSM_PSMT8) { for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) { - uint32 sbase = spsm.pa(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + uint32 sbase = GSLocalMemory::PixelAddress8(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); int* soffset = spsm.rowOffset[sy & 7]; - uint32 dbase = dpsm.pa(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + uint32 dbase = GSLocalMemory::PixelAddress8(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); int* doffset = dpsm.rowOffset[dy & 7]; for(int x = 0; x < w; x++, sx += xinc, dx += xinc) { - m_mem.WritePixel32(dbase + doffset[dx], m_mem.ReadPixel32(sbase + soffset[sx])); + m_mem.WritePixel8(dbase + doffset[dx], m_mem.ReadPixel8(sbase + soffset[sx])); + } + } + } + else if(m_env.BITBLTBUF.SPSM == PSM_PSMT4 && m_env.BITBLTBUF.DPSM == PSM_PSMT4) + { + for(int y = 0; y < h; y++, sy += yinc, dy += yinc, sx -= xinc * w, dx -= xinc * w) + { + uint32 sbase = GSLocalMemory::PixelAddress4(0, sy, m_env.BITBLTBUF.SBP, m_env.BITBLTBUF.SBW); + int* soffset = spsm.rowOffset[sy & 7]; + + uint32 dbase = GSLocalMemory::PixelAddress4(0, dy, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW); + int* doffset = dpsm.rowOffset[dy & 7]; + + for(int x = 0; x < w; x++, sx += xinc, dx += xinc) + { + m_mem.WritePixel4(dbase + doffset[dx], m_mem.ReadPixel4(sbase + soffset[sx])); } } } diff --git a/plugins/GSdx/GSState.h b/plugins/GSdx/GSState.h index 595a8b538f..0f7163aa0b 100644 --- a/plugins/GSdx/GSState.h +++ b/plugins/GSdx/GSState.h @@ -221,7 +221,7 @@ public: GSVector4i GetDisplayRect(int i = -1); GSVector4i GetFrameRect(int i = -1); - GSVector4i GetDeviceRect(int i = -1); + GSVector2i GetDeviceSize(int i = -1); bool IsEnabled(int i); diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 282e0b18ca..6dd495118c 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -376,7 +376,7 @@ public: int ww = (int)(fr.left + rt->m_TEX0.TBW * 64); int hh = (int)(fr.top + m_renderer->GetDisplayRect().height()); - if(hh <= m_renderer->GetDeviceRect().height() / 2) + if(hh <= m_renderer->GetDeviceSize().y / 2) { hh *= 2; } diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 6a59e815dd..4670658e84 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -69,11 +69,7 @@ void GSTextureCache10::GSRenderTargetHW10::Update() texture.Update(GSVector4i(0, 0, w, h), buff, pitch); - GSVector4 dr( - m_texture.m_scale.x * r.left, - m_texture.m_scale.y * r.top, - m_texture.m_scale.x * r.right, - m_texture.m_scale.y * r.bottom); + GSVector4 dr = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy(); m_renderer->m_dev.StretchRect(texture, m_texture, dr); @@ -103,12 +99,7 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) int w = r.width(); int h = r.height(); - GSVector4 src; - - src.x = m_texture.m_scale.x * r.left / m_texture.GetWidth(); - src.y = m_texture.m_scale.y * r.top / m_texture.GetHeight(); - src.z = m_texture.m_scale.x * r.right / m_texture.GetWidth(); - src.w = m_texture.m_scale.y * r.bottom / m_texture.GetHeight(); + GSVector4 src = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy() / GSVector4(m_texture.GetSize()).xyxy(); DXGI_FORMAT format = m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S ? DXGI_FORMAT_R16_UINT : DXGI_FORMAT_R8G8B8A8_UNORM; @@ -269,9 +260,14 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) m_renderer->m_dev.CreateRenderTarget(m_texture, rt->m_texture.GetWidth(), rt->m_texture.GetHeight()); + GSVector4 size = GSVector4(rt->m_texture.GetSize()).xyxy(); + GSVector4 scale = GSVector4(rt->m_texture.m_scale).xyxy(); + int bw = 64; int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + GSVector4i br(0, 0, bw, bh); + int sw = (int)rt->m_TEX0.TBW << 6; int dw = (int)m_TEX0.TBW << 6; @@ -287,18 +283,9 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) int sx = o % sw; int sy = o / sw; - GSVector4 src, dst; - - src.x = rt->m_texture.m_scale.x * sx / rt->m_texture.GetWidth(); - src.y = rt->m_texture.m_scale.y * sy / rt->m_texture.GetHeight(); - src.z = rt->m_texture.m_scale.x * (sx + bw) / rt->m_texture.GetWidth(); - src.w = rt->m_texture.m_scale.y * (sy + bh) / rt->m_texture.GetHeight(); - - dst.x = rt->m_texture.m_scale.x * dx; - dst.y = rt->m_texture.m_scale.y * dy; - dst.z = rt->m_texture.m_scale.x * (dx + bw); - dst.w = rt->m_texture.m_scale.y * (dy + bh); - + GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; + m_renderer->m_dev.StretchRect(rt->m_texture, src, m_texture, dst); // TODO: this is quite a lot of StretchRect, do it with one Draw diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index 8ae4d90c42..fa8e028e4f 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -71,11 +71,7 @@ void GSTextureCache9::GSRenderTarget9::Update() // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * 4); - GSVector4 dr( - m_texture.m_scale.x * r.left, - m_texture.m_scale.y * r.top, - m_texture.m_scale.x * r.right, - m_texture.m_scale.y * r.bottom); + GSVector4 dr = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy(); m_renderer->m_dev.StretchRect(texture, m_texture, dr); } @@ -106,12 +102,7 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) int w = r.width(); int h = r.height(); - GSVector4 src; - - src.x = m_texture.m_scale.x * r.left / m_texture.GetWidth(); - src.y = m_texture.m_scale.y * r.top / m_texture.GetHeight(); - src.z = m_texture.m_scale.x * r.right / m_texture.GetWidth(); - src.w = m_texture.m_scale.y * r.bottom / m_texture.GetHeight(); + GSVector4 src = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy() / GSVector4(m_texture.GetSize()).xyxy(); Texture offscreen; @@ -270,14 +261,20 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) m_renderer->m_dev.CreateRenderTarget(m_texture, rt->m_texture.GetWidth(), rt->m_texture.GetHeight()); + GSVector4 size = GSVector4(rt->m_texture.GetSize()).xyxy(); + GSVector4 scale = GSVector4(rt->m_texture.m_scale).xyxy(); + int bw = 64; int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + GSVector4i br(0, 0, bw, bh); + int sw = (int)rt->m_TEX0.TBW << 6; int dw = (int)m_TEX0.TBW << 6; int dh = 1 << m_TEX0.TH; + if(sw != 0) for(int dy = 0; dy < dh; dy += bh) { for(int dx = 0; dx < dw; dx += bw) @@ -287,17 +284,8 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) int sx = o % sw; int sy = o / sw; - GSVector4 src, dst; - - src.x = rt->m_texture.m_scale.x * sx / rt->m_texture.GetWidth(); - src.y = rt->m_texture.m_scale.y * sy / rt->m_texture.GetHeight(); - src.z = rt->m_texture.m_scale.x * (sx + bw) / rt->m_texture.GetWidth(); - src.w = rt->m_texture.m_scale.y * (sy + bh) / rt->m_texture.GetHeight(); - - dst.x = rt->m_texture.m_scale.x * dx; - dst.y = rt->m_texture.m_scale.y * dy; - dst.z = rt->m_texture.m_scale.x * (dx + bw); - dst.w = rt->m_texture.m_scale.y * (dy + bh); + GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; m_renderer->m_dev.StretchRect(rt->m_texture, src, m_texture, dst); diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index a9433b9ccb..524521e5ec 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -32,7 +32,7 @@ GSTextureCacheSW::~GSTextureCacheSW() RemoveAll(); } -const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r) +const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r) { GSLocalMemory& mem = m_state->m_mem; @@ -52,7 +52,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE continue; } - if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && !((GSVector4i)TEXA).eq(t2->m_TEXA))) + if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && TEXA != t2->m_TEXA)) { continue; } @@ -161,12 +161,7 @@ void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, cons GSVector2i s = (bp & 31) == 0 ? psm.pgs : psm.bs; - GSVector4i r; - - r.left = rect.left & ~(s.x - 1); - r.top = rect.top & ~(s.y - 1); - r.right = (rect.right + (s.x - 1)) & ~(s.x - 1); - r.bottom = (rect.bottom + (s.y - 1)) & ~(s.y - 1); + GSVector4i r = rect.ralign(s); for(int y = r.top; y < r.bottom; y += s.y) { @@ -215,7 +210,7 @@ GSTextureCacheSW::GSTexture::~GSTexture() } } -bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* rect) +bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect) { if(m_complete) { @@ -249,15 +244,7 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX m_tw = max(psm.pal > 0 ? 5 : 3, TEX0.TW); // makes one row 32 bytes at least, matches the smallest block size that is allocated above for m_buff } - GSVector4i r(0, 0, tw, th); - - if(rect) - { - r.left = rect->left & ~(s.x - 1); - r.top = rect->top & ~(s.y - 1); - r.right = (rect->right + (s.x - 1)) & ~(s.x - 1); - r.bottom = (rect->bottom + (s.y - 1)) & ~(s.y - 1); - } + GSVector4i r = rect.ralign(s); if(r.left == 0 && r.top == 0 && r.right == tw && r.bottom == th) { diff --git a/plugins/GSdx/GSTextureCacheSW.h b/plugins/GSdx/GSTextureCacheSW.h index ea5128c566..4e99396ca9 100644 --- a/plugins/GSdx/GSTextureCacheSW.h +++ b/plugins/GSdx/GSTextureCacheSW.h @@ -44,7 +44,7 @@ public: explicit GSTexture(GSState* state); virtual ~GSTexture(); - bool Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r = NULL); + bool Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r); }; protected: @@ -56,7 +56,7 @@ public: GSTextureCacheSW(GSState* state); virtual ~GSTextureCacheSW(); - const GSTexture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i* r = NULL); + const GSTexture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r); void RemoveAll(); void IncAge(); diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 80a102c146..fdeeac7286 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -155,6 +155,11 @@ public: return bottom - top; } + GSVector4i rsize() const + { + return *this - xyxy(); // same as GSVector4i(0, 0, width(), height()); + } + bool rempty() const { return (*this < zwzw()).mask() != 0x00ff; @@ -178,6 +183,28 @@ public: return sat_i32(a); } + enum RoundMode {Outside, Inside, NegInf, PosInf}; + + template GSVector4i ralign(const GSVector2i& a) const + { + // a must be 1 << n + + GSVector4i mask = GSVector4i(a) - GSVector4i(1, 1); + + GSVector4i v; + + switch(mode) + { + case Inside: v = *this + mask; break; + case Outside: v = *this + mask.zwxy(); break; + case NegInf: v = *this; break; + case PosInf: v = *this + mask.zwzw(); break; + default: ASSERT(0); break; + } + + return v.andnot(mask.xyxy()); + } + GSVector4i fit(int arx, int ary) const; GSVector4i fit(int preset) const; @@ -2194,7 +2221,9 @@ public: GSVector4(int x, int y, int z, int w) { - m = _mm_cvtepi32_ps(_mm_set_epi32(w, z, y, x)); + GSVector4i v(x, y, z, w); + + m = _mm_cvtepi32_ps(v.m); } GSVector4(int x, int y) @@ -2212,6 +2241,11 @@ public: m = _mm_castsi128_ps(_mm_loadl_epi64((__m128i*)&v)); } + explicit GSVector4(const GSVector2i& v) + { + m = _mm_cvtepi32_ps(_mm_loadl_epi64((__m128i*)&v)); + } + explicit GSVector4(float f) { m = _mm_set1_ps(f); diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 2523f25124..7cd1211c08 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1399,6 +1399,14 @@ + + + Date: Wed, 20 May 2009 20:15:34 +0000 Subject: [PATCH 098/296] mVU: minor fix for branches, improving compat a wee bit.. like lots :p git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1227 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 2 -- 1 file changed, 2 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index c61872be38..a4b2a7ac5e 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -296,7 +296,6 @@ microVUt(void) mVUanalyzeSflag(int It) { // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } - analyzeVIreg2(It, 1); } microVUt(void) mVUanalyzeFSSET() { @@ -327,7 +326,6 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { iPC = curPC; } analyzeVIreg1(Is); - analyzeVIreg2(It, 1); } //------------------------------------------------------------------ From 2aec2318cce0ba4eb5c26ba975e13728c9ec1731 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 20 May 2009 21:22:37 +0000 Subject: [PATCH 099/296] GSdx: found a compiler def for stl that removes a lot of iterator checks in the release build. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1228 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/vsprops/release.vsprops | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/GSdx/vsprops/release.vsprops b/plugins/GSdx/vsprops/release.vsprops index 5cc40308ee..618c9e8888 100644 --- a/plugins/GSdx/vsprops/release.vsprops +++ b/plugins/GSdx/vsprops/release.vsprops @@ -12,7 +12,7 @@ FavorSizeOrSpeed="1" OmitFramePointers="true" WholeProgramOptimization="true" - PreprocessorDefinitions="NDEBUG" + PreprocessorDefinitions="NDEBUG;_SECURE_SCL=0" StringPooling="true" BufferSecurityCheck="false" /> From 31c5c234ef316ec955c786e0570de2d2f8564a31 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 20 May 2009 23:18:19 +0000 Subject: [PATCH 100/296] microVU: minor changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1229 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 6 +++++- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 8 ++++---- pcsx2/x86/microVU_Misc.h | 2 +- 4 files changed, 11 insertions(+), 7 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index a4b2a7ac5e..0301d2a476 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -122,6 +122,7 @@ microVUt(void) mVUanalyzeFMAC4(int Fs, int Ft) { #define analyzeVIreg1(reg) { if (reg) { mVUstall = aMax(mVUstall, mVUregs.VI[reg]); } } #define analyzeVIreg2(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; mVUinfo |= _writesVI; mVU->VIbackup[0] = reg; } } +#define analyzeVIreg3(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; } } microVUt(void) mVUanalyzeIALU1(int Id, int Is, int It) { microVU* mVU = mVUx; @@ -296,6 +297,7 @@ microVUt(void) mVUanalyzeSflag(int It) { // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } + analyzeVIreg3(It, 1); } microVUt(void) mVUanalyzeFSSET() { @@ -326,16 +328,18 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { iPC = curPC; } analyzeVIreg1(Is); + analyzeVIreg3(It, 1); } //------------------------------------------------------------------ // Cflag - Clip Flag Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeCflag() { +microVUt(void) mVUanalyzeCflag(int It) { microVU* mVU = mVUx; mVUinfo |= _swapOps; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } + analyzeVIreg3(It, 1); } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 4dc2c8b226..aac54d574f 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -215,7 +215,7 @@ microVUt(void) mVUpass4(int startPC) { microVUt(void) mVUsetFlagInfo() { microVU* mVU = mVUx; branchType1 { incPC(-1); mVUpass4(branchAddr); incPC(1); } - branchType2 { mVUflagInfo |= 0xffffffff; } + branchType2 { mVUflagInfo |= 0xfff; } branchType3 { incPC(-1); mVUpass4(branchAddr); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index d85db8f2ff..3880f57766 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -434,7 +434,7 @@ microVUf(void) mVU_ESUM() { microVUf(void) mVU_FCAND() { microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(); } + pass1 { mVUanalyzeCflag(1); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); AND32ItoR(gprT1, _Imm24_); @@ -448,7 +448,7 @@ microVUf(void) mVU_FCAND() { microVUf(void) mVU_FCEQ() { microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(); } + pass1 { mVUanalyzeCflag(1); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); XOR32ItoR(gprT1, _Imm24_); @@ -462,7 +462,7 @@ microVUf(void) mVU_FCEQ() { microVUf(void) mVU_FCGET() { microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(); } + pass1 { mVUanalyzeCflag(_It_); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); AND32ItoR(gprT1, 0xfff); @@ -474,7 +474,7 @@ microVUf(void) mVU_FCGET() { microVUf(void) mVU_FCOR() { microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(); } + pass1 { mVUanalyzeCflag(1); } pass2 { mVUallocCFLAGa(gprT1, fvcInstance); OR32ItoR(gprT1, _Imm24_); diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 66d996594a..f5a2e60174 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -207,7 +207,7 @@ declareAllVariables #define _backupVI (1<<22) // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) #define _memReadIs (1<<23) // Read Is (VI reg) from memory (used by branches) #define _memReadIt (1<<24) // Read If (VI reg) from memory (used by branches) -#define _writesVI (1<<25) // Current Instruction writes to VI +#define _writesVI (1<<25) // Current Instruction writes to VI (used by branches; note that flag-modifying opcodes shouldn't set this) #define _swapOps (1<<26) // Runs Lower Instruction Before Upper Instruction #define _isFSSET (1<<27) // Cur Instruction is FSSET #define _doDivFlag (1<<28) // Transfer Div flag to Status Flag From 996440bab05643188d7ff299e6702dd8408adcd7 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 21 May 2009 07:09:39 +0000 Subject: [PATCH 101/296] microVU: - Fixed a problem that was causing microPrograms to be cached when they shouldn't be (introduced in r1209). This also fixes broken logging in some games, and is a small-speedup. - Minor Changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1230 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 2 +- pcsx2/x86/microVU_Compile.inl | 13 ++++++++++--- pcsx2/x86/microVU_Flags.inl | 1 + 3 files changed, 12 insertions(+), 4 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 971dcc01be..8956d4b266 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -222,7 +222,7 @@ microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, b mVU->prog.cleared = 0; mVU->prog.isSame = cmpWholeProg ? 1 : -1; mVU->prog.prog[progIndex].last_used = 3; - if (!needOverflowCheck || mVU->prog.prog[progIndex].used < 0x7fffffff) { + if (needOverflowCheck && (mVU->prog.prog[progIndex].used < 0x7fffffff)) { mVU->prog.prog[progIndex].used++; // increment 'used' (avoiding overflows if necessary) } return 1; diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index e7c6730bed..016dff7217 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -34,6 +34,13 @@ else { ajmp = JMPcc((uptr)0); } \ break +#define branchWarning() { \ + if (mVUbranch) { \ + Console::Error("microVU%d Warning: Branch in E-bit/Branch delay slot! [%04x]", params vuIndex, xPC); \ + mVUinfo |= _isNOP; \ + } \ +} + #define startLoop() { mVUdebug1(); mVUstall = 0; memset(&mVUregsTemp, 0, sizeof(mVUregsTemp)); } #define calcCycles(reg, x) { reg = ((reg > x) ? (reg - x) : 0); } #define tCycles(dest, src) { dest = aMax(dest, src); } @@ -50,7 +57,7 @@ microVUt(void) mVUcheckIsSame() { microVU* mVU = mVUx; if (mVU->prog.isSame == -1) { - mVU->prog.isSame = !!memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); + mVU->prog.isSame = !memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); } if (mVU->prog.isSame == 0) { mVUcacheProg(mVU->prog.cur); @@ -235,7 +242,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (mVU->p) { mVUinfo |= _readP; } if (mVU->q) { mVUinfo |= _readQ; } else { mVUinfo |= _writeQ; } - if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; if (mVUbranch) { Console::Error("microVU Warning: Branch in E-bit/Branch delay slot!"); mVUinfo |= _isNOP; } break; } + if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } if (mVUbranch) { mVUsetFlagInfo(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } incPC(1); @@ -308,7 +315,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { } // Conditional Branches mVUprint("mVUcompile conditional branch"); - if (bBlock) { // Branch non-taken has already been compiled + if (bBlock) { // Branch non-taken has already been compiled incPC(-3); // Go back to branch opcode (to get branch imm addr) if (mVUblocks[branchAddr/8] == NULL) diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index aac54d574f..dd68d718b7 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -202,6 +202,7 @@ microVUt(void) mVUpass4(int startPC) { if (mVUbranch) { branch = 3; mVUbranch = 0; } incPC(1); } + //if (mVUcount == 4) { mVUflagInfo |= 0xfff; } // Is this Too Slow? 99% of games probably don't need this. iPC = oldPC; mVUcount = oldCount; mVUbranch = oldBranch; From 7922f33f4fc4c815121a895c5bfff85956303da7 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 21 May 2009 21:12:01 +0000 Subject: [PATCH 102/296] Minor variable type changes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1233 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/CDVD/CdRom.cpp | 14 ++-- pcsx2/CDVD/CdRom.h | 6 +- pcsx2/DebugTools/DisR3000asm.cpp | 8 +- pcsx2/Decode_XA.cpp | 128 ++++++++++++++--------------- pcsx2/Decode_XA.h | 20 +++-- pcsx2/IPU/coroutine.cpp | 6 +- pcsx2/Linux/DebugDlg.cpp | 4 +- pcsx2/MTGS.cpp | 13 +-- pcsx2/Mdec.h | 10 +-- pcsx2/RDebug/deci2.h | 8 +- pcsx2/x86/ix86-32/iCore-32.cpp | 8 +- plugins/CDVDiso/src/CDVDiso.h | 6 +- plugins/CDVDiso/src/CDVDisop.cpp | 10 +-- plugins/CDVDiso/src/Linux/Config.h | 4 +- plugins/FWnull/FW.cpp | 6 +- plugins/FWnull/FW.h | 2 +- 16 files changed, 128 insertions(+), 125 deletions(-) diff --git a/pcsx2/CDVD/CdRom.cpp b/pcsx2/CDVD/CdRom.cpp index 07bcf3a6ac..947961a5ac 100644 --- a/pcsx2/CDVD/CdRom.cpp +++ b/pcsx2/CDVD/CdRom.cpp @@ -71,7 +71,7 @@ const char *CmdName[0x100]= { }; cdrStruct cdr; -long LoadCdBios; +s32 LoadCdBios; u8 Test04[] = { 0 }; u8 Test05[] = { 0 }; @@ -83,12 +83,12 @@ u8 Test23[] = { 0x43, 0x58, 0x44, 0x32, 0x39 ,0x34, 0x30, 0x51 }; // PSXCLK = 1 sec in the ps // so (PSXCLK / 75) / BIAS = cdr read time (linuzappz) //#define cdReadTime ((PSXCLK / 75) / BIAS) -unsigned long cdReadTime;// = ((PSXCLK / 75) / BIAS); +u32 cdReadTime;// = ((PSXCLK / 75) / BIAS); #define CDR_INT(eCycle) PSX_INT(IopEvt_Cdrom, eCycle) #define CDREAD_INT(eCycle) PSX_INT(IopEvt_CdromRead, eCycle) -static __forceinline void StartReading(unsigned long type) { +static __forceinline void StartReading(u32 type) { cdr.Reading = type; cdr.FirstSector = 1; cdr.Readed = 0xff; @@ -149,7 +149,7 @@ static void ReadTrack() { #define DataEnd 4 #define DiskError 5 -static void AddIrqQueue(u8 irq, unsigned long ecycle) { +static void AddIrqQueue(u8 irq, u32 ecycle) { cdr.Irq = irq; if (cdr.Stat) { cdr.eCycle = ecycle; @@ -649,7 +649,7 @@ void cdrWrite1(u8 rt) { for (i=0; i<3; i++) cdr.SetSector[i] = btoi(cdr.Param[i]); cdr.SetSector[3] = 0; if ((cdr.SetSector[0] | cdr.SetSector[1] | cdr.SetSector[2]) == 0) { - *(unsigned long *)cdr.SetSector = *(unsigned long *)cdr.SetSectorSeek; + *(u32 *)cdr.SetSector = *(u32 *)cdr.SetSectorSeek; } cdr.Ctrl|= 0x80; cdr.Stat = NoIntr; @@ -780,14 +780,14 @@ void cdrWrite1(u8 rt) { break; case CdlSeekL: - ((unsigned long *)cdr.SetSectorSeek)[0] = ((unsigned long *)cdr.SetSector)[0]; + ((u32 *)cdr.SetSectorSeek)[0] = ((u32 *)cdr.SetSector)[0]; cdr.Ctrl|= 0x80; cdr.Stat = NoIntr; AddIrqQueue(cdr.Cmd, 0x800); break; case CdlSeekP: - ((unsigned long *)cdr.SetSectorSeek)[0] = ((unsigned long *)cdr.SetSector)[0]; + ((u32 *)cdr.SetSectorSeek)[0] = ((u32 *)cdr.SetSector)[0]; cdr.Ctrl|= 0x80; cdr.Stat = NoIntr; AddIrqQueue(cdr.Cmd, 0x800); diff --git a/pcsx2/CDVD/CdRom.h b/pcsx2/CDVD/CdRom.h index 36ceba98a2..41b3d9b280 100644 --- a/pcsx2/CDVD/CdRom.h +++ b/pcsx2/CDVD/CdRom.h @@ -48,7 +48,7 @@ struct cdrStruct u8 ResultReady; u8 Cmd; u8 Readed; - unsigned long Reading; + u32 Reading; cdvdTN ResultTN; u8 ResultTD[4]; @@ -67,12 +67,12 @@ struct cdrStruct int Init; u8 Irq; - unsigned long eCycle; + u32 eCycle; char Unused[4087]; }; -void AddIrqQueue(u8 irq, unsigned long ecycle); +void AddIrqQueue(u8 irq, u32 ecycle); extern cdrStruct cdr; diff --git a/pcsx2/DebugTools/DisR3000asm.cpp b/pcsx2/DebugTools/DisR3000asm.cpp index 840adaaa03..b3c2dc9333 100644 --- a/pcsx2/DebugTools/DisR3000asm.cpp +++ b/pcsx2/DebugTools/DisR3000asm.cpp @@ -24,7 +24,7 @@ namespace R3000A { -unsigned long IOP_opcode_addr; +u32 IOP_opcode_addr; const char *GPR_IOP_REG[32] = { "zero", "at", "v0", "v1", "a0", "a1", "a2", "a3", @@ -223,7 +223,7 @@ char *disR3000Fasm(u32 code, u32 pc) { char *IOP_jump_decode(void) { static char buf[256]; - unsigned long addr; + u32 addr; addr = (IOP_opcode_addr & 0xf0000000)|((psxRegs.code&0x3ffffff)<<2); sprintf(buf, "0x%08lX", addr); return buf; @@ -231,8 +231,8 @@ char *IOP_jump_decode(void) char *IOP_offset_decode(void) { static char buf[256]; - unsigned long addr; - addr = ((((short)( psxRegs.code & 0xFFFF) * 4) + IOP_opcode_addr + 4)); + u32 addr; + addr = ((((s16)( psxRegs.code & 0xFFFF) * 4) + IOP_opcode_addr + 4)); sprintf(buf, "0x%08lX", addr); return buf; } diff --git a/pcsx2/Decode_XA.cpp b/pcsx2/Decode_XA.cpp index 51de78a58e..e9029de18b 100644 --- a/pcsx2/Decode_XA.cpp +++ b/pcsx2/Decode_XA.cpp @@ -30,10 +30,6 @@ #pragma warning(disable:4244) #endif -typedef unsigned char U8; -typedef unsigned short U16; -typedef unsigned long U32; - #define NOT(_X_) (!(_X_)) #define CLAMP(_X_,_MI_,_MA_) {if(_X_<_MI_)_X_=_MI_;if(_X_>_MA_)_X_=_MA_;} @@ -71,13 +67,13 @@ void ADPCM_InitDecode( ADPCM_Decode_t *decp ) #define IK0(fid) ((int)((-K0[fid]) * (1<> 4) & 0x0f; range = (filter_range >> 0) & 0x0f; @@ -85,14 +81,14 @@ void ADPCM_DecodeBlock16( ADPCM_Decode_t *decp, U8 filter_range, const void *vbl fy1 = decp->y1; for (i = BLKSIZ/4; i; --i) { - long y; - long x0, x1, x2, x3; + s32 y; + s32 x0, x1, x2, x3; y = *blockp++; - x3 = (short)( y & 0xf000) >> range; x3 <<= SH; - x2 = (short)((y << 4) & 0xf000) >> range; x2 <<= SH; - x1 = (short)((y << 8) & 0xf000) >> range; x1 <<= SH; - x0 = (short)((y << 12) & 0xf000) >> range; x0 <<= SH; + x3 = (s16)( y & 0xf000) >> range; x3 <<= SH; + x2 = (s16)((y << 4) & 0xf000) >> range; x2 <<= SH; + x1 = (s16)((y << 8) & 0xf000) >> range; x1 <<= SH; + x0 = (s16)((y << 12) & 0xf000) >> range; x0 <<= SH; x0 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x0; x1 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x1; @@ -111,12 +107,12 @@ void ADPCM_DecodeBlock16( ADPCM_Decode_t *decp, U8 filter_range, const void *vbl static int headtable[4] = {0,2,8,10}; //=========================================== -static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { - const U8 *sound_groupsp; - const U8 *sound_datap, *sound_datap2; - int i, j, k, nbits; - U16 data[4096], *datap; - short *destp; +static void xa_decode_data( xa_decode_t *xdp, u8 *srcp ) { + const u8 *sound_groupsp; + const u8 *sound_datap, *sound_datap2; + s32 i, j, k, nbits; + u16 data[4096], *datap; + s16 *destp; destp = xdp->pcm; nbits = xdp->nbits == 4 ? 4 : 2; @@ -131,15 +127,16 @@ static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { sound_datap2 = sound_datap + i; if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (U16)sound_datap2[0] | - (U16)(sound_datap2[4] << 8); + *(datap++) = (u16)sound_datap2[0] | + (u16)(sound_datap2[4] << 8); } - } else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (U16)(sound_datap2[ 0] & 0x0f) | - ((U16)(sound_datap2[ 4] & 0x0f) << 4) | - ((U16)(sound_datap2[ 8] & 0x0f) << 8) | - ((U16)(sound_datap2[12] & 0x0f) << 12); + } + else { // level B/C + for (k=0; k < 7; k++, sound_datap2 += 16) { + *(datap++) = (u16)(sound_datap2[ 0] & 0x0f) | + ((u16)(sound_datap2[ 4] & 0x0f) << 4) | + ((u16)(sound_datap2[ 8] & 0x0f) << 8) | + ((u16)(sound_datap2[12] & 0x0f) << 12); } } ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+0], data, @@ -149,15 +146,16 @@ static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { sound_datap2 = sound_datap + i; if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (U16)sound_datap2[0] | - (U16)(sound_datap2[4] << 8); + *(datap++) = (u16)sound_datap2[0] | + (u16)(sound_datap2[4] << 8); } - } else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (U16)(sound_datap2[ 0] >> 4) | - ((U16)(sound_datap2[ 4] >> 4) << 4) | - ((U16)(sound_datap2[ 8] >> 4) << 8) | - ((U16)(sound_datap2[12] >> 4) << 12); + } + else { // level B/C + for (k=0; k < 7; k++, sound_datap2 += 16) { + *(datap++) = (u16)(sound_datap2[ 0] >> 4) | + ((u16)(sound_datap2[ 4] >> 4) << 4) | + ((u16)(sound_datap2[ 8] >> 4) << 8) | + ((u16)(sound_datap2[12] >> 4) << 12); } } ADPCM_DecodeBlock16( &xdp->right, sound_groupsp[headtable[i]+1], data, @@ -177,15 +175,16 @@ static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { sound_datap2 = sound_datap + i; if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (U16)sound_datap2[0] | - (U16)(sound_datap2[4] << 8); + *(datap++) = (u16)sound_datap2[0] | + (u16)(sound_datap2[4] << 8); } - } else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (U16)(sound_datap2[ 0] & 0x0f) | - ((U16)(sound_datap2[ 4] & 0x0f) << 4) | - ((U16)(sound_datap2[ 8] & 0x0f) << 8) | - ((U16)(sound_datap2[12] & 0x0f) << 12); + } + else { // level B/C + for (k=0; k < 7; k++, sound_datap2 += 16) { + *(datap++) = (u16)(sound_datap2[ 0] & 0x0f) | + ((u16)(sound_datap2[ 4] & 0x0f) << 4) | + ((u16)(sound_datap2[ 8] & 0x0f) << 8) | + ((u16)(sound_datap2[12] & 0x0f) << 12); } } ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+0], data, @@ -197,15 +196,16 @@ static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { sound_datap2 = sound_datap + i; if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (U16)sound_datap2[0] | - (U16)(sound_datap2[4] << 8); + *(datap++) = (u16)sound_datap2[0] | + (u16)(sound_datap2[4] << 8); } - } else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (U16)(sound_datap2[ 0] >> 4) | - ((U16)(sound_datap2[ 4] >> 4) << 4) | - ((U16)(sound_datap2[ 8] >> 4) << 8) | - ((U16)(sound_datap2[12] >> 4) << 12); + } + else { // level B/C + for (k=0; k < 7; k++, sound_datap2 += 16) { + *(datap++) = (u16)(sound_datap2[ 0] >> 4) | + ((u16)(sound_datap2[ 4] >> 4) << 4) | + ((u16)(sound_datap2[ 8] >> 4) << 8) | + ((u16)(sound_datap2[12] >> 4) << 12); } } ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+1], data, @@ -221,15 +221,15 @@ static void xa_decode_data( xa_decode_t *xdp, unsigned char *srcp ) { //=== XA SPECIFIC ROUTINES //============================================ typedef struct { -U8 filenum; -U8 channum; -U8 submode; -U8 coding; +u8 filenum; +u8 channum; +u8 submode; +u8 coding; -U8 filenum2; -U8 channum2; -U8 submode2; -U8 coding2; +u8 filenum2; +u8 channum2; +u8 submode2; +u8 coding2; } xa_subheader_t; #define SUB_SUB_EOF (1<<7) // end of file @@ -251,10 +251,10 @@ U8 coding2; #define SUB_AUDIO 2 //============================================ -static int parse_xa_audio_sector( xa_decode_t *xdp, +static s32 parse_xa_audio_sector( xa_decode_t *xdp, xa_subheader_t *subheadp, - unsigned char *sectorp, - int is_first_sector ) { + u8*sectorp, + s32 is_first_sector ) { if ( is_first_sector ) { switch ( AUDIO_CODING_GET_FREQ(subheadp->coding) ) { case 0: xdp->freq = 37800; break; @@ -295,8 +295,8 @@ static int parse_xa_audio_sector( xa_decode_t *xdp, //=== - 0 for any other successive sector //=== return -1 if error //================================================================ -long xa_decode_sector( xa_decode_t *xdp, - unsigned char *sectorp, int is_first_sector ) { +s32 xa_decode_sector( xa_decode_t *xdp, + u8 *sectorp, s32 is_first_sector ) { if (parse_xa_audio_sector(xdp, (xa_subheader_t *)sectorp, sectorp + sizeof(xa_subheader_t), is_first_sector)) return -1; diff --git a/pcsx2/Decode_XA.h b/pcsx2/Decode_XA.h index 0aeab19749..85736bf835 100644 --- a/pcsx2/Decode_XA.h +++ b/pcsx2/Decode_XA.h @@ -23,21 +23,23 @@ #ifndef DECODEXA_H #define DECODEXA_H +#include "PrecompiledHeader.h" + typedef struct { - long y0, y1; + s32 y0, y1; } ADPCM_Decode_t; typedef struct { - int freq; - int nbits; - int stereo; - int nsamples; + s32 freq; + s32 nbits; + s32 stereo; + s32 nsamples; ADPCM_Decode_t left, right; - short pcm[16384]; + s16 pcm[16384]; } xa_decode_t; -long xa_decode_sector( xa_decode_t *xdp, - unsigned char *sectorp, - int is_first_sector ); +s32 xa_decode_sector( xa_decode_t *xdp, + u8*sectorp, + s32 is_first_sector ); #endif diff --git a/pcsx2/IPU/coroutine.cpp b/pcsx2/IPU/coroutine.cpp index ac89043d8d..3ccfc84e1b 100644 --- a/pcsx2/IPU/coroutine.cpp +++ b/pcsx2/IPU/coroutine.cpp @@ -26,8 +26,8 @@ struct coroutine { uptr storeebx, storeesi, storeedi, storeebp; - int restore; // if nonzero, restore the registers - int alloc; + s32 restore; // if nonzero, restore the registers + s32 alloc; //struct s_coroutine *caller; //struct s_coroutine *restarget; @@ -45,7 +45,7 @@ coroutine_t so_create(void (*func)(void *), void *data, void *stack, int size) int alloc = 0; // r = CO_STK_COROSIZE; coroutine *co; - if ((size &= ~(sizeof(long) - 1)) < CO_MIN_SIZE) return NULL; + if ((size &= ~(sizeof(s32) - 1)) < CO_MIN_SIZE) return NULL; if (!stack) { size = (size + sizeof(coroutine) + CO_STK_ALIGN - 1) & ~(CO_STK_ALIGN - 1); stack = malloc(size); diff --git a/pcsx2/Linux/DebugDlg.cpp b/pcsx2/Linux/DebugDlg.cpp index 7e4b11e50e..4f0b19210b 100644 --- a/pcsx2/Linux/DebugDlg.cpp +++ b/pcsx2/Linux/DebugDlg.cpp @@ -18,7 +18,7 @@ #include "DebugDlg.h" using namespace R5900; -unsigned long DebuggerPC = 0; +u32 DebuggerPC = 0; /*void UpdateDebugger() { @@ -70,7 +70,7 @@ unsigned long DebuggerPC = 0; void UpdateDebugger(void) { - /* unsigned long t; + /* u32 t; int cnt; if (DebuggerPC == 0) DebuggerPC = cpuRegs.pc; //- 0x00000038; diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 253345695f..d2e22bdd2f 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -769,15 +769,16 @@ static u32 GSRingBufCopySz = 0; int mtgsThreadObject::PrepDataPacket( GIF_PATH pathidx, const u8* srcdata, u32 size ) { #ifdef PCSX2_GSRING_TX_STATS - ringtx_s+=size; - ringtx_s_ulg+=size&0x7F; - ringtx_s_min=min(ringtx_s_min,size); - ringtx_s_max=max(ringtx_s_max,size); + ringtx_s += size; + ringtx_s_ulg += size&0x7F; + ringtx_s_min = min(ringtx_s_min,size); + ringtx_s_max = max(ringtx_s_max,size); ringtx_c++; - unsigned long tx_sz; + u32 tx_sz; + if (_BitScanReverse(&tx_sz,size)) { - unsigned long tx_algn; + u32 tx_algn; _BitScanForward(&tx_algn,size); ringtx_inf[tx_sz][tx_algn]++; ringtx_inf_s[tx_sz]+=size; diff --git a/pcsx2/Mdec.h b/pcsx2/Mdec.h index e54dae3a9e..5e77ec2937 100644 --- a/pcsx2/Mdec.h +++ b/pcsx2/Mdec.h @@ -94,16 +94,16 @@ void psxDma0(u32 madr, u32 bcr, u32 chcr); void psxDma1(u32 madr, u32 bcr, u32 chcr); int mdecFreeze(gzFile f, int Mode); -unsigned short* rl2blk(int *blk,unsigned short *mdec_rl); +u16* rl2blk(int *blk,u16 *mdec_rl); void iqtab_init(int *iqtab,unsigned char *iq_y); void round_init(void); void yuv2rgb24(int *blk,unsigned char *image); -void yuv2rgb15(int *blk,unsigned short *image); +void yuv2rgb15(int *blk,u16 *image); struct { - unsigned long command; - unsigned long status; - unsigned short *rl; + u32 command; + u32 status; + u16 *rl; int rlsize; } mdec; diff --git a/pcsx2/RDebug/deci2.h b/pcsx2/RDebug/deci2.h index b35bfb9ba0..a9d7a2aa88 100644 --- a/pcsx2/RDebug/deci2.h +++ b/pcsx2/RDebug/deci2.h @@ -54,13 +54,13 @@ struct DECI2_DBGP_BRK{ #define RUN 1 extern DECI2_DBGP_BRK ebrk[32], ibrk[32]; -extern int ebrk_count, ibrk_count; -extern volatile long runStatus; -extern int runCode, runCount; +extern s32 ebrk_count, ibrk_count; +extern volatile s32 runStatus; +extern s32 runCode, runCount; extern Threading::Semaphore* runEvent; -extern int connected; +extern s32 connected; //when add linux code this might change int writeData(const u8 *result); diff --git a/pcsx2/x86/ix86-32/iCore-32.cpp b/pcsx2/x86/ix86-32/iCore-32.cpp index e90e051a22..70cffa7fcc 100644 --- a/pcsx2/x86/ix86-32/iCore-32.cpp +++ b/pcsx2/x86/ix86-32/iCore-32.cpp @@ -139,10 +139,10 @@ void _flushConstReg(int reg) void _flushConstRegs() { - int i, j; - int zero_cnt = 0, minusone_cnt = 0; - int eaxval = 1; // 0, -1 - unsigned long done[4] = {0, 0, 0, 0}; + s32 i, j; + s32 zero_cnt = 0, minusone_cnt = 0; + s32 eaxval = 1; // 0, -1 + u32 done[4] = {0, 0, 0, 0}; u8* rewindPtr; // flush constants diff --git a/plugins/CDVDiso/src/CDVDiso.h b/plugins/CDVDiso/src/CDVDiso.h index faddb1b616..8226a6e912 100644 --- a/plugins/CDVDiso/src/CDVDiso.h +++ b/plugins/CDVDiso/src/CDVDiso.h @@ -106,9 +106,9 @@ extern _cdIso cdIso[8]; #define MSF2SECT(m,s,f) (((m)*60+(s)-2)*75+(f)) -extern const unsigned char version; -extern const unsigned char revision; -extern const unsigned char build; +extern const u8 version; +extern const u8 revision; +extern const u8 build; extern char IsoFile[256]; extern char IsoCWD[256]; diff --git a/plugins/CDVDiso/src/CDVDisop.cpp b/plugins/CDVDiso/src/CDVDisop.cpp index bae8f753c9..b2260b9798 100755 --- a/plugins/CDVDiso/src/CDVDisop.cpp +++ b/plugins/CDVDiso/src/CDVDisop.cpp @@ -53,11 +53,11 @@ char *LibName = "Linuz Iso CDVD (Debug) "; char *LibName = "Linuz Iso CDVD "; #endif -const unsigned char version = PS2E_CDVD_VERSION; -const unsigned char revision = 0; -const unsigned char build = 9; +const u8 version = PS2E_CDVD_VERSION; +const u8 revision = 0; +const u8 build = 9; -unsigned char cdbuffer[CD_FRAMESIZE_RAW * 10] = {0}; +u8 cdbuffer[CD_FRAMESIZE_RAW * 10] = {0}; s32 msf_to_lba(u8 m, u8 s, u8 f) { @@ -279,7 +279,7 @@ EXPORT_C_(s32) CDVDgetTD(u8 Track, cdvdTD *Buffer) return 0; } -static int layer1start = -1; +static s32 layer1start = -1; EXPORT_C_(s32) CDVDgetTOC(void* toc) { u8 type = CDVDgetDiskType(); diff --git a/plugins/CDVDiso/src/Linux/Config.h b/plugins/CDVDiso/src/Linux/Config.h index 5009dd0f3d..c8cb20e594 100644 --- a/plugins/CDVDiso/src/Linux/Config.h +++ b/plugins/CDVDiso/src/Linux/Config.h @@ -52,8 +52,8 @@ extern void SysMessageLoc(char *fmt, ...); extern char *LibName; -extern const unsigned char revision; -extern const unsigned char build; +extern const u8 revision; +extern const u8 build; extern GtkWidget *AboutDlg, *ConfDlg, *MsgDlg, *FileSel; extern GtkWidget *Edit, *CdEdit; diff --git a/plugins/FWnull/FW.cpp b/plugins/FWnull/FW.cpp index c4503cab30..174d1a356a 100644 --- a/plugins/FWnull/FW.cpp +++ b/plugins/FWnull/FW.cpp @@ -24,9 +24,9 @@ using namespace std; #include "FW.h" -const unsigned char version = PS2E_FW_VERSION; -const unsigned char revision = 0; -const unsigned char build = 5; // increase that with each version +const u8 version = PS2E_FW_VERSION; +const u8 revision = 0; +const u8 build = 5; // increase that with each version static char *libraryName = "FWnull Driver"; string s_strIniPath="inis/FWnull.ini"; diff --git a/plugins/FWnull/FW.h b/plugins/FWnull/FW.h index da97459e62..4613f22088 100644 --- a/plugins/FWnull/FW.h +++ b/plugins/FWnull/FW.h @@ -62,7 +62,7 @@ extern "C" typedef struct { - int Log; + s32 Log; } Config; extern Config conf; From b1fd826954e156ae401ca268c6190010b71f4e04 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 21 May 2009 22:19:54 +0000 Subject: [PATCH 103/296] GSnull: Update the GifTransfer code. Start to add keyboard code. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1234 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 68 +++++++++++++++- plugins/GSnull/GifTransfer.cpp | 86 +++++++++------------ plugins/GSnull/GifTransfer.h | 1 + plugins/GSnull/Makefile.am | 6 +- plugins/GSnull/Windows/GSnull_vc2008.vcproj | 8 ++ plugins/GSnull/null/GSnull.cpp | 39 ++++++++++ plugins/GSnull/null/GSnull.h | 22 ++++++ 7 files changed, 177 insertions(+), 53 deletions(-) create mode 100644 plugins/GSnull/null/GSnull.cpp create mode 100644 plugins/GSnull/null/GSnull.h diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index 7dfe2bdaf4..92385b393c 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -27,6 +27,7 @@ using namespace std; #include "GS.h" +#include "null/GSnull.h" #ifdef __LINUX__ Display *display; @@ -43,6 +44,9 @@ const unsigned char build = 1; // increase that with each version static char *libraryName = "GSnull Driver"; FILE *gsLog; Config conf; +u32 GSKeyEvent = 0; +bool GSShift = false, GSAlt = false; + string s_strIniPath="inis/GSnull.ini"; void (*GSirq)(); @@ -168,9 +172,37 @@ EXPORT_C_(s32) GStest() return 0; } +void ProcessMessages() +{ +#ifdef __LINUX__ + if ( GSKeyEvent ) + { + int myKeyEvent = GSKeyEvent; + bool myShift = GSShift; + GSKeyEvent = 0; + + switch ( myKeyEvent ) + { + case XK_F5: + OnKeyboardF5(myShift); + break; + case XK_F6: + OnKeyboardF6(myShift); + break; + case XK_F7: + OnKeyboardF7(myShift); + break; + case XK_F9: + OnKeyboardF9(myShift); + break; + } + } +#endif +} EXPORT_C_(void) GSvsync(int field) { + ProcessMessages(); } EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) @@ -211,7 +243,41 @@ EXPORT_C_(void) GSreadFIFO2(u64 *mem, int qwc) // GSkeyEvent gets called when there is a keyEvent from the PAD plugin EXPORT_C_(void) GSkeyEvent(keyEvent *ev) { - //SysPrintf("There is a key event.\n"); + switch(ev->evt) { + case KEYPRESS: + switch(ev->key) { + case XK_F5: + case XK_F6: + case XK_F7: + case XK_F9: + GSKeyEvent = ev->key ; + break; + case XK_Escape: + break; + case XK_Shift_L: + case XK_Shift_R: + //bShift = true; + GSShift = true; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = true; + break; + } + break; + case KEYRELEASE: + switch(ev->key) { + case XK_Shift_L: + case XK_Shift_R: + //bShift = false; + GSShift = false; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = false; + break; + } + } } EXPORT_C_(void) GSchangeSaveState(int, const char* filename) diff --git a/plugins/GSnull/GifTransfer.cpp b/plugins/GSnull/GifTransfer.cpp index 37a22819bb..34e69fa3bb 100644 --- a/plugins/GSnull/GifTransfer.cpp +++ b/plugins/GSnull/GifTransfer.cpp @@ -103,14 +103,24 @@ u32 GIFPath::GetReg() return regs[curreg]; } -__forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) +__forceinline bool GIFPath::StepReg() { - GIFPath& path = m_path[pathidx]; + if ((++curreg & 0xf) == tag.nreg) + { + curreg = 0; + + if(--tag.nloop == 0) + { + return false; + } + } +} + +__forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) +{ GIFPath& path = m_path[pathidx]; while(size > 0) { - bool eop = false; - if(path.tag.nloop == 0) { path.SetTag( pMem ); @@ -119,14 +129,16 @@ __forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) --size; if(pathidx == 2 && path.tag.eop) + { Path3transfer = FALSE; + } if( pathidx == 0 ) { // hack: if too much data for VU1, just ignore. // The GIF is evil : if nreg is 0, it's really 16. Otherwise it's the value in nreg. - const int numregs = ((path.tag.nreg - 1) & 15) + 1; + const int numregs = ((path.tag.nreg-1)&15)+1; if((path.tag.nloop * numregs) > (size * ((path.tag.flg == 1) ? 2 : 1))) { @@ -134,74 +146,48 @@ __forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) return ++size; } } - - if(path.tag.eop) - { - eop = true; - } - else if(path.tag.nloop == 0) - { - if(pathidx == 0) - continue; - - eop = true; - } } - - if(path.tag.nloop > 0) + else { + // NOTE: size > 0 => do {} while(size > 0); should be faster than while(size > 0) {} + switch(path.tag.flg) { case GIF_FLG_PACKED: - while(size > 0) + do { if( path.GetReg() == 0xe ) { const int handler = pMem[8]; if(handler >= 0x60 && handler < 0x63) - s_GSHandlers[handler & 0x3]((const u32*)pMem); + s_GSHandlers[handler&0x3]((const u32*)pMem); } + size--; pMem += 16; // 128 bits! //sizeof(GIFPackedReg); - - if((++path.curreg & 0xf) == path.tag.nreg) - { - path.curreg = 0; - path.tag.nloop--; - - if(path.tag.nloop == 0) - break; - } } + while(path.StepReg() && size > 0); + break; case GIF_FLG_REGLIST: size *= 2; - while(size > 0) + do { const int handler = path.GetReg(); - if (handler >= 0x60 && handler < 0x63) + if(handler >= 0x60 && handler < 0x63) s_GSHandlers[handler&0x3]((const u32*)pMem); size--; pMem += 8; //sizeof(GIFReg); -- 64 bits! - - if((++path.curreg & 0xf) == path.tag.nreg) - { - path.curreg = 0; - path.tag.nloop--; - - if(path.tag.nloop == 0) - { - break; - } - } } + while(path.StepReg() && size > 0); - if (size & 1) pMem += 8; //sizeof(GIFReg); + if(size & 1) pMem += 8; //sizeof(GIFReg); + size /= 2; break; @@ -227,23 +213,27 @@ __forceinline u32 _gifTransfer( GIF_PATH pathidx, const u8* pMem, u32 size ) } } - if (eop && ((int)size <= 0 || pathidx == 0)) + if(pathidx == 0) { - break; + if(path.tag.eop && path.tag.nloop == 0) + { + break; + } } } if(pathidx == 0) { - if (!path.tag.eop && path.tag.nloop > 0) + if(size == 0 && path.tag.nloop > 0) { path.tag.nloop = 0; - SysPrintf( "path1 hack! " ); + SysPrintf( "path1 hack! \n" ); // This means that the giftag data got screwly somewhere // along the way (often means curreg was in a bad state or something) } } + return size; } diff --git a/plugins/GSnull/GifTransfer.h b/plugins/GSnull/GifTransfer.h index bf30e19756..0e16a99d62 100644 --- a/plugins/GSnull/GifTransfer.h +++ b/plugins/GSnull/GifTransfer.h @@ -85,6 +85,7 @@ struct GIFPath __forceinline void PrepRegs(); void SetTag(const void* mem); u32 GetReg(); + __forceinline bool StepReg(); }; extern GIFPath m_path[3]; diff --git a/plugins/GSnull/Makefile.am b/plugins/GSnull/Makefile.am index 872aa0e2a0..21f39566fa 100644 --- a/plugins/GSnull/Makefile.am +++ b/plugins/GSnull/Makefile.am @@ -29,7 +29,5 @@ libGSnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ Linux/callbacks.h \ -GifTransfer.cpp GifTransfer.h -#Linux/callbacks.c - -#SUBDIRS = Linux \ No newline at end of file +GifTransfer.cpp GifTransfer.h \ +null/GSnull.cpp null/GSnull.h \ No newline at end of file diff --git a/plugins/GSnull/Windows/GSnull_vc2008.vcproj b/plugins/GSnull/Windows/GSnull_vc2008.vcproj index 65a952655a..5772e553be 100644 --- a/plugins/GSnull/Windows/GSnull_vc2008.vcproj +++ b/plugins/GSnull/Windows/GSnull_vc2008.vcproj @@ -164,6 +164,10 @@ RelativePath="..\GS.cpp" > + + @@ -185,6 +189,10 @@ RelativePath="..\GifTransfer.h" > + + diff --git a/plugins/GSnull/null/GSnull.cpp b/plugins/GSnull/null/GSnull.cpp new file mode 100644 index 0000000000..78d848f517 --- /dev/null +++ b/plugins/GSnull/null/GSnull.cpp @@ -0,0 +1,39 @@ + /* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include "GS.h" + +void OnKeyboardF5(int myShift) +{ + SysPrintf("F5 pressed!\n"); +} + +void OnKeyboardF6(int myShift) +{ + SysPrintf("F6 pressed!\n"); +} + +void OnKeyboardF7(int myShift) +{ + SysPrintf("F7 pressed!\n"); +} + +void OnKeyboardF9(int myShift) +{ + SysPrintf("F9 pressed!\n"); +} \ No newline at end of file diff --git a/plugins/GSnull/null/GSnull.h b/plugins/GSnull/null/GSnull.h new file mode 100644 index 0000000000..53c25d1398 --- /dev/null +++ b/plugins/GSnull/null/GSnull.h @@ -0,0 +1,22 @@ + /* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +extern void OnKeyboardF5(int myShift); +extern void OnKeyboardF6(int myShift); +extern void OnKeyboardF7(int myShift); +extern void OnKeyboardF9(int myShift); \ No newline at end of file From 8e3d3080c02027495603e4d34dd704507f612dc4 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 22 May 2009 00:35:09 +0000 Subject: [PATCH 104/296] Get rid of Decode_XA. A few minor changes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1235 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/CDVD/CdRom.cpp | 3 + pcsx2/CDVD/CdRom.h | 18 +- pcsx2/Decode_XA.cpp | 322 -------------------------------- pcsx2/Decode_XA.h | 45 ----- pcsx2/MMI.cpp | 2 +- pcsx2/Makefile.am | 4 +- pcsx2/RDebug/deci2.h | 2 +- pcsx2/x86/iR5900.h | 1 - pcsx2/x86/ix86-32/iR5900-32.cpp | 1 + pcsx2/x86/ix86/ix86_jmp.cpp | 6 +- 10 files changed, 28 insertions(+), 376 deletions(-) delete mode 100644 pcsx2/Decode_XA.cpp delete mode 100644 pcsx2/Decode_XA.h diff --git a/pcsx2/CDVD/CdRom.cpp b/pcsx2/CDVD/CdRom.cpp index 947961a5ac..69286aa481 100644 --- a/pcsx2/CDVD/CdRom.cpp +++ b/pcsx2/CDVD/CdRom.cpp @@ -88,6 +88,9 @@ u32 cdReadTime;// = ((PSXCLK / 75) / BIAS); #define CDR_INT(eCycle) PSX_INT(IopEvt_Cdrom, eCycle) #define CDREAD_INT(eCycle) PSX_INT(IopEvt_CdromRead, eCycle) + +static void AddIrqQueue(u8 irq, u32 ecycle); + static __forceinline void StartReading(u32 type) { cdr.Reading = type; cdr.FirstSector = 1; diff --git a/pcsx2/CDVD/CdRom.h b/pcsx2/CDVD/CdRom.h index 41b3d9b280..cb46373957 100644 --- a/pcsx2/CDVD/CdRom.h +++ b/pcsx2/CDVD/CdRom.h @@ -20,9 +20,23 @@ #define __CDROM_H__ #include "IopCommon.h" -#include "Decode_XA.h" #include "PS2Edefs.h" +// Not used. +typedef struct { + s32 y0, y1; +} ADPCM_Decode_t; + +// Not used. +typedef struct { + s32 freq; + s32 nbits; + s32 stereo; + s32 nsamples; + ADPCM_Decode_t left, right; + s16 pcm[16384]; +} xa_decode_t; + struct cdrStruct { u8 OCUP; @@ -72,8 +86,6 @@ struct cdrStruct char Unused[4087]; }; -void AddIrqQueue(u8 irq, u32 ecycle); - extern cdrStruct cdr; void cdrReset(); diff --git a/pcsx2/Decode_XA.cpp b/pcsx2/Decode_XA.cpp deleted file mode 100644 index e9029de18b..0000000000 --- a/pcsx2/Decode_XA.cpp +++ /dev/null @@ -1,322 +0,0 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2008 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA - */ -//============================================ -//=== Audio XA decoding -//=== Kazzuya -//============================================ -//=== Modified by linuzappz -//============================================ - -#include - -#include "Decode_XA.h" - -#ifdef __WIN32__ -#pragma warning(disable:4244) -#endif - -#define NOT(_X_) (!(_X_)) -#define CLAMP(_X_,_MI_,_MA_) {if(_X_<_MI_)_X_=_MI_;if(_X_>_MA_)_X_=_MA_;} - -//============================================ -//=== ADPCM DECODING ROUTINES -//============================================ - -static double K0[4] = { - 0.0, - 0.9375, - 1.796875, - 1.53125 -}; - -static double K1[4] = { - 0.0, - 0.0, - -0.8125, - -0.859375 -}; - -#define BLKSIZ 28 /* block size (32 - 4 nibbles) */ - -//=========================================== -void ADPCM_InitDecode( ADPCM_Decode_t *decp ) -{ - decp->y0 = 0; - decp->y1 = 0; -} - -//=========================================== -#define SH 4 -#define SHC 10 - -#define IK0(fid) ((int)((-K0[fid]) * (1<> 4) & 0x0f; - range = (filter_range >> 0) & 0x0f; - - fy0 = decp->y0; - fy1 = decp->y1; - - for (i = BLKSIZ/4; i; --i) { - s32 y; - s32 x0, x1, x2, x3; - - y = *blockp++; - x3 = (s16)( y & 0xf000) >> range; x3 <<= SH; - x2 = (s16)((y << 4) & 0xf000) >> range; x2 <<= SH; - x1 = (s16)((y << 8) & 0xf000) >> range; x1 <<= SH; - x0 = (s16)((y << 12) & 0xf000) >> range; x0 <<= SH; - - x0 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x0; - x1 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x1; - x2 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x2; - x3 -= (IK0(filterid) * fy0 + (IK1(filterid) * fy1)) >> SHC; fy1 = fy0; fy0 = x3; - - CLAMP( x0, -32768<> SH; destp += inc; - CLAMP( x1, -32768<> SH; destp += inc; - CLAMP( x2, -32768<> SH; destp += inc; - CLAMP( x3, -32768<> SH; destp += inc; - } - decp->y0 = fy0; - decp->y1 = fy1; -} - -static int headtable[4] = {0,2,8,10}; - -//=========================================== -static void xa_decode_data( xa_decode_t *xdp, u8 *srcp ) { - const u8 *sound_groupsp; - const u8 *sound_datap, *sound_datap2; - s32 i, j, k, nbits; - u16 data[4096], *datap; - s16 *destp; - - destp = xdp->pcm; - nbits = xdp->nbits == 4 ? 4 : 2; - - if (xdp->stereo) { // stereo - for (j=0; j < 18; j++) { - sound_groupsp = srcp + j * 128; // sound groups header - sound_datap = sound_groupsp + 16; // sound data just after the header - - for (i=0; i < nbits; i++) { - datap = data; - sound_datap2 = sound_datap + i; - if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A - for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (u16)sound_datap2[0] | - (u16)(sound_datap2[4] << 8); - } - } - else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (u16)(sound_datap2[ 0] & 0x0f) | - ((u16)(sound_datap2[ 4] & 0x0f) << 4) | - ((u16)(sound_datap2[ 8] & 0x0f) << 8) | - ((u16)(sound_datap2[12] & 0x0f) << 12); - } - } - ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+0], data, - destp+0, 2 ); - - datap = data; - sound_datap2 = sound_datap + i; - if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A - for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (u16)sound_datap2[0] | - (u16)(sound_datap2[4] << 8); - } - } - else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (u16)(sound_datap2[ 0] >> 4) | - ((u16)(sound_datap2[ 4] >> 4) << 4) | - ((u16)(sound_datap2[ 8] >> 4) << 8) | - ((u16)(sound_datap2[12] >> 4) << 12); - } - } - ADPCM_DecodeBlock16( &xdp->right, sound_groupsp[headtable[i]+1], data, - destp+1, 2 ); - - destp += 28*2; - } - } - } - else { // mono - for (j=0; j < 18; j++) { - sound_groupsp = srcp + j * 128; // sound groups header - sound_datap = sound_groupsp + 16; // sound data just after the header - - for (i=0; i < nbits; i++) { - datap = data; - sound_datap2 = sound_datap + i; - if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A - for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (u16)sound_datap2[0] | - (u16)(sound_datap2[4] << 8); - } - } - else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (u16)(sound_datap2[ 0] & 0x0f) | - ((u16)(sound_datap2[ 4] & 0x0f) << 4) | - ((u16)(sound_datap2[ 8] & 0x0f) << 8) | - ((u16)(sound_datap2[12] & 0x0f) << 12); - } - } - ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+0], data, - destp, 1 ); - - destp += 28; - - datap = data; - sound_datap2 = sound_datap + i; - if ((xdp->nbits == 8) && (xdp->freq == 37800)) { // level A - for (k=0; k < 14; k++, sound_datap2 += 8) { - *(datap++) = (u16)sound_datap2[0] | - (u16)(sound_datap2[4] << 8); - } - } - else { // level B/C - for (k=0; k < 7; k++, sound_datap2 += 16) { - *(datap++) = (u16)(sound_datap2[ 0] >> 4) | - ((u16)(sound_datap2[ 4] >> 4) << 4) | - ((u16)(sound_datap2[ 8] >> 4) << 8) | - ((u16)(sound_datap2[12] >> 4) << 12); - } - } - ADPCM_DecodeBlock16( &xdp->left, sound_groupsp[headtable[i]+1], data, - destp, 1 ); - - destp += 28; - } - } - } -} - -//============================================ -//=== XA SPECIFIC ROUTINES -//============================================ -typedef struct { -u8 filenum; -u8 channum; -u8 submode; -u8 coding; - -u8 filenum2; -u8 channum2; -u8 submode2; -u8 coding2; -} xa_subheader_t; - -#define SUB_SUB_EOF (1<<7) // end of file -#define SUB_SUB_RT (1<<6) // real-time sector -#define SUB_SUB_FORM (1<<5) // 0 form1 1 form2 -#define SUB_SUB_TRIGGER (1<<4) // used for interrupt -#define SUB_SUB_DATA (1<<3) // contains data -#define SUB_SUB_AUDIO (1<<2) // contains audio -#define SUB_SUB_VIDEO (1<<1) // contains video -#define SUB_SUB_EOR (1<<0) // end of record - -#define AUDIO_CODING_GET_STEREO(_X_) ( (_X_) & 3) -#define AUDIO_CODING_GET_FREQ(_X_) (((_X_) >> 2) & 3) -#define AUDIO_CODING_GET_BPS(_X_) (((_X_) >> 4) & 3) -#define AUDIO_CODING_GET_EMPHASIS(_X_) (((_X_) >> 6) & 1) - -#define SUB_UNKNOWN 0 -#define SUB_VIDEO 1 -#define SUB_AUDIO 2 - -//============================================ -static s32 parse_xa_audio_sector( xa_decode_t *xdp, - xa_subheader_t *subheadp, - u8*sectorp, - s32 is_first_sector ) { - if ( is_first_sector ) { - switch ( AUDIO_CODING_GET_FREQ(subheadp->coding) ) { - case 0: xdp->freq = 37800; break; - case 1: xdp->freq = 18900; break; - default: xdp->freq = 0; break; - } - switch ( AUDIO_CODING_GET_BPS(subheadp->coding) ) { - case 0: xdp->nbits = 4; break; - case 1: xdp->nbits = 8; break; - default: xdp->nbits = 0; break; - } - switch ( AUDIO_CODING_GET_STEREO(subheadp->coding) ) { - case 0: xdp->stereo = 0; break; - case 1: xdp->stereo = 1; break; - default: xdp->stereo = 0; break; - } - - if ( xdp->freq == 0 ) - return -1; - - ADPCM_InitDecode( &xdp->left ); - ADPCM_InitDecode( &xdp->right ); - - xdp->nsamples = 18 * 28 * 8; - if (xdp->stereo == 1) xdp->nsamples /= 2; - } - xa_decode_data( xdp, sectorp ); - - return 0; -} - -//================================================================ -//=== THIS IS WHAT YOU HAVE TO CALL -//=== xdp - structure were all important data are returned -//=== sectorp - data in input -//=== pcmp - data in output -//=== is_first_sector - 1 if it's the 1st sector of the stream -//=== - 0 for any other successive sector -//=== return -1 if error -//================================================================ -s32 xa_decode_sector( xa_decode_t *xdp, - u8 *sectorp, s32 is_first_sector ) { - if (parse_xa_audio_sector(xdp, (xa_subheader_t *)sectorp, sectorp + sizeof(xa_subheader_t), is_first_sector)) - return -1; - - return 0; -} - -/* EXAMPLE: -"nsamples" is the number of 16 bit samples -every sample is 2 bytes in mono and 4 bytes in stereo - -xa_decode_t xa; - - sectorp = read_first_sector(); - xa_decode_sector( &xa, sectorp, 1 ); - play_wave( xa.pcm, xa.freq, xa.nsamples ); - - while ( --n_sectors ) - { - sectorp = read_next_sector(); - xa_decode_sector( &xa, sectorp, 0 ); - play_wave( xa.pcm, xa.freq, xa.nsamples ); - } -*/ diff --git a/pcsx2/Decode_XA.h b/pcsx2/Decode_XA.h deleted file mode 100644 index 85736bf835..0000000000 --- a/pcsx2/Decode_XA.h +++ /dev/null @@ -1,45 +0,0 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2008 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA - */ -//============================================ -//=== Audio XA decoding -//=== Kazzuya -//============================================ - -#ifndef DECODEXA_H -#define DECODEXA_H - -#include "PrecompiledHeader.h" - -typedef struct { - s32 y0, y1; -} ADPCM_Decode_t; - -typedef struct { - s32 freq; - s32 nbits; - s32 stereo; - s32 nsamples; - ADPCM_Decode_t left, right; - s16 pcm[16384]; -} xa_decode_t; - -s32 xa_decode_sector( xa_decode_t *xdp, - u8*sectorp, - s32 is_first_sector ); - -#endif diff --git a/pcsx2/MMI.cpp b/pcsx2/MMI.cpp index 448c1bacff..1dfbf605d3 100644 --- a/pcsx2/MMI.cpp +++ b/pcsx2/MMI.cpp @@ -181,7 +181,7 @@ __forceinline void PMFHL_CLAMP(u16 dst, u16 src) { if ((int)src > (int)0x00007fff) dst = 0x7fff; - else if ((int)src < (int)0xffff8000) + else if ((int)src < (int)0xffff8000) // Ints only go up to 0x7FFFFFFF. Something's not right here. --arcum42 dst = 0x8000; else dst = (u16)src; diff --git a/pcsx2/Makefile.am b/pcsx2/Makefile.am index 8b666ac2bd..04a0bf84a9 100644 --- a/pcsx2/Makefile.am +++ b/pcsx2/Makefile.am @@ -15,7 +15,7 @@ INCLUDES = -I$(x86_dir) -I$(cdvd_dir) -I$(common_dir) -I$(third_dir) noinst_LIBRARIES = libpcsx2.a libpcsx2_a_SOURCES = \ -AlignedMalloc.cpp Decode_XA.cpp HwRead.cpp MTGS.cpp R3000AInterpreter.cpp Sio.cpp VUflags.cpp \ +AlignedMalloc.cpp HwRead.cpp MTGS.cpp R3000AInterpreter.cpp Sio.cpp VUflags.cpp \ Dump.cpp HwWrite.cpp R3000AOpcodeTables.cpp SourceLog.cpp VUmicroMem.cpp \ Elfheader.cpp Interpreter.cpp Memory.cpp R5900.cpp Stats.cpp VUops.cpp \ FPU.cpp IopBios.cpp MemoryCard.cpp R5900OpcodeImpl.cpp System.cpp Vif.cpp \ @@ -34,7 +34,7 @@ ps2/Iop/IopHwRead.cpp ps2/Iop/IopHwWrite.cpp ps2/Iop/IopHw_Internal.h libpcsx2_a_SOURCES += \ Common.h HashMap.h IopDma.h MemoryCard.h PrecompiledHeader.h SafeArray.h StringUtils.h Vif.h \ Counters.h HostGui.h IopHw.h MemoryTypes.h R3000A.h SamplProf.h System.h VifDma.h \ -Decode_XA.h Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ +Hw.h IopMem.h Misc.h R5900.h SaveState.h Threading.h cheatscpp.h \ Dump.h IopBios.h IopSio2.h NakedAsm.h R5900Exceptions.h Sif.h VU.h vtlb.h \ COP0.h Elfheader.h IopBios2.h Mdec.h Patch.h R5900OpcodeTables.h Sifcmd.h VUflags.h \ Cache.h Exceptions.h IopCommon.h MemcpyFast.h Paths.h Sio.h Sio_internal.h VUmicro.h \ diff --git a/pcsx2/RDebug/deci2.h b/pcsx2/RDebug/deci2.h index a9d7a2aa88..a01f84f334 100644 --- a/pcsx2/RDebug/deci2.h +++ b/pcsx2/RDebug/deci2.h @@ -55,7 +55,7 @@ struct DECI2_DBGP_BRK{ extern DECI2_DBGP_BRK ebrk[32], ibrk[32]; extern s32 ebrk_count, ibrk_count; -extern volatile s32 runStatus; +extern volatile long runStatus; extern s32 runCode, runCount; extern Threading::Semaphore* runEvent; diff --git a/pcsx2/x86/iR5900.h b/pcsx2/x86/iR5900.h index af168fc238..edd4cf446d 100644 --- a/pcsx2/x86/iR5900.h +++ b/pcsx2/x86/iR5900.h @@ -90,7 +90,6 @@ void LoadBranchState(); void recompileNextInstruction(int delayslot); void SetBranchReg( u32 reg ); void SetBranchImm( u32 imm ); -u32 eeScaleBlockCycles(); void iFlushCall(int flushtype); void recBranchCall( void (*func)() ); diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 982540b379..50626dfe4e 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -111,6 +111,7 @@ static u32 dumplog = 0; static void iBranchTest(u32 newpc = 0xffffffff, bool noDispatch=false); static void ClearRecLUT(BASEBLOCK* base, int count); +static u32 eeScaleBlockCycles(); #ifdef PCSX2_VM_COISSUE static u8 _eeLoadWritesRs(u32 tempcode) diff --git a/pcsx2/x86/ix86/ix86_jmp.cpp b/pcsx2/x86/ix86/ix86_jmp.cpp index fb304af18b..4574af60b9 100644 --- a/pcsx2/x86/ix86/ix86_jmp.cpp +++ b/pcsx2/x86/ix86/ix86_jmp.cpp @@ -119,7 +119,11 @@ __emitinline void Internal::xJccKnownTarget( JccComparisonType comparison, const // if the following assert fails it means we accidentally used slideForard on a backward // jump (which is an invalid operation since there's nothing to slide forward). - if( slideForward ) jASSUME( displacement8 >= 0 ); + if( slideForward ) + { + // jASSUME has an else statement in it that would be abiguous without the brackets. + jASSUME( displacement8 >= 0 ); + } if( is_s8( displacement8 ) ) xJcc8( comparison, displacement8 ); From 3b22adc4b350d0451f0fca99b75732650ff5afc1 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 22 May 2009 01:22:52 +0000 Subject: [PATCH 105/296] GSdx: converted a few template classes to pointers and virtual functions, the dll got more than 100k smaller, since there were many changes, backup your old version first :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1236 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 10 +- plugins/GSdx/GPURenderer.cpp | 79 +- plugins/GSdx/GPURenderer.h | 154 +--- plugins/GSdx/GPURendererSW.cpp | 207 +++++ plugins/GSdx/GPURendererSW.h | 202 +---- plugins/GSdx/GPUSetupPrimCodeGenerator.cpp | 2 - plugins/GSdx/GS.cpp | 14 +- plugins/GSdx/GS.h | 4 + plugins/GSdx/GSClut.cpp | 6 +- plugins/GSdx/GSDevice.cpp | 192 +++++ plugins/GSdx/GSDevice.h | 218 +----- plugins/GSdx/GSDevice10.cpp | 337 ++++---- plugins/GSdx/GSDevice10.h | 52 +- plugins/GSdx/GSDevice7.cpp | 46 +- plugins/GSdx/GSDevice7.h | 24 +- plugins/GSdx/GSDevice9.cpp | 335 ++++---- plugins/GSdx/GSDevice9.h | 59 +- plugins/GSdx/GSDeviceNull.cpp | 6 +- plugins/GSdx/GSDeviceNull.h | 23 +- plugins/GSdx/GSDrawScanlineCodeGenerator.cpp | 5 - plugins/GSdx/GSLocalMemory.cpp | 36 +- plugins/GSdx/GSRasterizer.cpp | 20 +- plugins/GSdx/GSRenderer.cpp | 418 ++++++++++ plugins/GSdx/GSRenderer.h | 470 +---------- plugins/GSdx/GSRendererHW.h | 56 +- plugins/GSdx/GSRendererHW10.cpp | 225 +++--- plugins/GSdx/GSRendererHW10.h | 10 +- plugins/GSdx/GSRendererHW9.cpp | 237 +++--- plugins/GSdx/GSRendererHW9.h | 12 +- plugins/GSdx/GSRendererNull.h | 14 +- plugins/GSdx/GSRendererSW.cpp | 772 ++++++++++++++++++ plugins/GSdx/GSRendererSW.h | 769 +----------------- plugins/GSdx/GSState.cpp | 5 +- plugins/GSdx/GSTexture.h | 2 +- plugins/GSdx/GSTextureCache.cpp | 767 ++++++++++++++++++ plugins/GSdx/GSTextureCache.h | 781 +------------------ plugins/GSdx/GSTextureCache10.cpp | 217 +++--- plugins/GSdx/GSTextureCache10.h | 16 +- plugins/GSdx/GSTextureCache9.cpp | 221 +++--- plugins/GSdx/GSTextureCache9.h | 16 +- plugins/GSdx/GSTextureFX10.cpp | 6 +- plugins/GSdx/GSTextureFX10.h | 4 +- plugins/GSdx/GSTextureFX9.cpp | 62 +- plugins/GSdx/GSTextureFX9.h | 9 +- plugins/GSdx/GSVector.h | 5 + plugins/GSdx/GSdx_vs2008.vcproj | 25 +- plugins/GSdx/stdafx.h | 4 +- plugins/GSdx/vsprops/common.vsprops | 7 +- 48 files changed, 3620 insertions(+), 3541 deletions(-) diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index a1abe6c9ec..8df8d3a4bc 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -31,7 +31,7 @@ #define PSE_LT_GPU 2 static HRESULT s_hr = E_FAIL; -static GPURendererBase* s_gpu = NULL; +static GPURenderer* s_gpu = NULL; EXPORT_C_(uint32) PSEgetLibType() { @@ -121,10 +121,10 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) switch(renderer) { default: - case 0: s_gpu = new GPURendererSW(rs, threads); break; - case 1: s_gpu = new GPURendererSW(rs, threads); break; - case 2: s_gpu = new GPURendererSW(rs, threads); break; - // TODO: case 3: s_gpu = new GPURendererNull(rs, threads); break; + case 0: s_gpu = new GPURendererSW(rs, new GSDevice7(), threads); break; + case 1: s_gpu = new GPURendererSW(rs, new GSDevice9(), threads); break; + case 2: s_gpu = new GPURendererSW(rs, new GSDevice10(), threads); break; + // TODO: case 3: s_gpu = new GPURendererNull(rs, new GSDeviceNull(), threads); break; } if(!s_gpu->Create(hWnd)) diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index e6cd2ade78..b3c2670771 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -22,4 +22,81 @@ #include "StdAfx.h" #include "GPURenderer.h" -map GPURendererBase::m_wnd2gpu; \ No newline at end of file +map GPURenderer::m_wnd2gpu; + +GPURenderer::GPURenderer(const GPURendererSettings& rs) + : GPUState(rs.m_scale) + , m_hWnd(NULL) + , m_wndproc(NULL) +{ + m_filter = rs.m_filter; + m_dither = rs.m_dither; + m_aspectratio = rs.m_aspectratio; + m_vsync = rs.m_vsync; + m_scale = m_mem.GetScale(); +} + +GPURenderer::~GPURenderer() +{ + if(m_wndproc) + { + SetWindowLongPtr(m_hWnd, GWLP_WNDPROC, (LONG_PTR)m_wndproc); + + m_wnd2gpu.erase(m_hWnd); + } +} + +bool GPURenderer::Create(HWND hWnd) +{ + m_hWnd = hWnd; + + m_wndproc = (WNDPROC)GetWindowLongPtr(hWnd, GWLP_WNDPROC); + SetWindowLongPtr(hWnd, GWLP_WNDPROC, (LONG_PTR)WndProc); + + m_wnd2gpu[hWnd] = this; + + DWORD style = GetWindowLong(hWnd, GWL_STYLE); + style |= WS_OVERLAPPEDWINDOW; + SetWindowLong(hWnd, GWL_STYLE, style); + UpdateWindow(hWnd); + + ShowWindow(hWnd, SW_SHOWNORMAL); + + return true; +} + +LRESULT CALLBACK GPURenderer::WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) +{ + map::iterator i = m_wnd2gpu.find(hWnd); + + if(i != m_wnd2gpu.end()) + { + return (*i).second->OnMessage(message, wParam, lParam); + } + + ASSERT(0); + + return 0; +} + +LRESULT GPURenderer::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) +{ + if(message == WM_KEYUP) + { + switch(wParam) + { + case VK_DELETE: + m_filter = (m_filter + 1) % 3; + return 0; + case VK_END: + m_dither = m_dither ? 0 : 1; + return 0; + case VK_NEXT: + m_aspectratio = (m_aspectratio + 1) % 3; + return 0; + } + } + + return m_wndproc(m_hWnd, message, wParam, lParam); +} + diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index 71125b9227..d7e21136a6 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -23,6 +23,7 @@ #include "GPUState.h" #include "GSVertexList.h" +#include "GSDevice.h" struct GPURendererSettings { @@ -33,100 +34,30 @@ struct GPURendererSettings GSVector2i m_scale; }; -class GPURendererBase : public GPUState, protected GPURendererSettings +class GPURenderer : public GPUState, protected GPURendererSettings { protected: HWND m_hWnd; WNDPROC m_wndproc; - static map m_wnd2gpu; + static map m_wnd2gpu; - static LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) - { - map::iterator i = m_wnd2gpu.find(hWnd); + static LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam); - if(i != m_wnd2gpu.end()) - { - return (*i).second->OnMessage(message, wParam, lParam); - } - - ASSERT(0); - - return 0; - } - - LRESULT OnMessage(UINT message, WPARAM wParam, LPARAM lParam) - { - if(message == WM_KEYUP) - { - switch(wParam) - { - case VK_DELETE: - m_filter = (m_filter + 1) % 3; - return 0; - case VK_END: - m_dither = m_dither ? 0 : 1; - return 0; - case VK_NEXT: - m_aspectratio = (m_aspectratio + 1) % 3; - return 0; - } - } - - return m_wndproc(m_hWnd, message, wParam, lParam); - } + LRESULT OnMessage(UINT message, WPARAM wParam, LPARAM lParam); public: - GPURendererBase(const GPURendererSettings& rs) - : GPUState(rs.m_scale) - , m_hWnd(NULL) - , m_wndproc(NULL) - { - m_filter = rs.m_filter; - m_dither = rs.m_dither; - m_aspectratio = rs.m_aspectratio; - m_vsync = rs.m_vsync; - m_scale = m_mem.GetScale(); - } - - virtual ~GPURendererBase() - { - if(m_wndproc) - { - SetWindowLongPtr(m_hWnd, GWLP_WNDPROC, (LONG_PTR)m_wndproc); - - m_wnd2gpu.erase(m_hWnd); - } - } - - virtual bool Create(HWND hWnd) - { - m_hWnd = hWnd; - - m_wndproc = (WNDPROC)GetWindowLongPtr(hWnd, GWLP_WNDPROC); - SetWindowLongPtr(hWnd, GWLP_WNDPROC, (LONG_PTR)WndProc); - - m_wnd2gpu[hWnd] = this; - - DWORD style = GetWindowLong(hWnd, GWL_STYLE); - style |= WS_OVERLAPPEDWINDOW; - SetWindowLong(hWnd, GWL_STYLE, style); - UpdateWindow(hWnd); - - ShowWindow(hWnd, SW_SHOWNORMAL); - - return true; - } + GPURenderer(const GPURendererSettings& rs); + virtual ~GPURenderer(); + virtual bool Create(HWND hWnd); virtual void VSync() = 0; virtual bool MakeSnapshot(const string& path) = 0; }; -template -class GPURenderer : public GPURendererBase +template +class GPURendererT : public GPURenderer { protected: - typedef typename Device::Texture Texture; - Vertex* m_vertices; int m_count; int m_maxcount; @@ -191,7 +122,7 @@ protected: m_count += count; } - typedef void (GPURenderer::*DrawingKickHandler)(Vertex* v, int& count); + typedef void (GPURendererT::*DrawingKickHandler)(Vertex* v, int& count); DrawingKickHandler m_fpDrawingKickHandlers[4]; @@ -235,46 +166,36 @@ protected: virtual void ResetDevice() {} virtual void Draw() = 0; - virtual bool GetOutput(Texture& t) = 0; + virtual GSTexture* GetOutput() = 0; bool Merge() { - Texture st[2]; + GSTexture* st[2] = {GetOutput(), NULL}; - if(!GetOutput(st[0])) + if(!st[0]) { return false; } - GSVector2i s = st[0].GetSize(); - + GSVector2i s = st[0]->GetSize(); + GSVector4 sr[2]; - - sr[0].x = 0; - sr[0].y = 0; - sr[0].z = 1.0f; - sr[0].w = 1.0f; - GSVector4 dr[2]; - dr[0].x = 0; - dr[0].y = 0; - dr[0].z = (float)s.x; - dr[0].w = (float)s.y; + sr[0] = GSVector4(0, 0, 1, 1); + dr[0] = GSVector4(0, 0, s.x, s.y); - GSVector4 c(0, 0, 0, 1); - - m_dev.Merge(st, sr, dr, s, 1, 1, c); + m_dev->Merge(st, sr, dr, s, 1, 1, GSVector4(0, 0, 0, 1)); return true; } public: - Device m_dev; + GSDevice* m_dev; public: - GPURenderer(const GPURendererSettings& rs) - : GPURendererBase(rs) + GPURendererT(const GPURendererSettings& rs) + : GPURenderer(rs) , m_count(0) , m_maxcount(10000) { @@ -283,11 +204,11 @@ public: for(int i = 0; i < countof(m_fpDrawingKickHandlers); i++) { - m_fpDrawingKickHandlers[i] = &GPURenderer::DrawingKickNull; + m_fpDrawingKickHandlers[i] = &GPURendererT::DrawingKickNull; } } - virtual ~GPURenderer() + virtual ~GPURendererT() { if(m_vertices) _aligned_free(m_vertices); } @@ -299,7 +220,7 @@ public: return false; } - if(!m_dev.Create(hWnd, m_vsync)) + if(!m_dev->Create(hWnd, m_vsync)) { return false; } @@ -313,21 +234,15 @@ public: { GSPerfMonAutoTimer pmat(m_perfmon); + m_perfmon.Put(GSPerfMon::Frame); + // m_env.STATUS.LCF = ~m_env.STATUS.LCF; // ? - if(!IsWindow(m_hWnd)) - { - return; - } + if(!IsWindow(m_hWnd)) return; Flush(); - m_perfmon.Put(GSPerfMon::Frame); - - if(!Merge()) - { - return; - } + if(!Merge()) return; // osd @@ -367,7 +282,7 @@ public: SetWindowText(m_hWnd, s_stats.c_str()); } - if(m_dev.IsLost()) + if(m_dev->IsLost()) { ResetDevice(); } @@ -376,7 +291,7 @@ public: GetClientRect(m_hWnd, r); - m_dev.Present(r.fit(m_aspectratio)); + m_dev->Present(r.fit(m_aspectratio)); } virtual bool MakeSnapshot(const string& path) @@ -390,6 +305,11 @@ public: return false; } - return m_dev.SaveCurrent(format("%s_%s.bmp", path.c_str(), buff)); + if(GSTexture* t = m_dev->GetCurrent()) + { + return t->Save(format("%s_%s.bmp", path.c_str(), buff)); + } + + return false; } }; diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index 71e1d083f6..eef66a2962 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -21,3 +21,210 @@ #include "StdAfx.h" #include "GPURendererSW.h" + +GPURendererSW::GPURendererSW(const GPURendererSettings& rs, GSDevice* dev, int threads) + : GPURendererT(rs) + , m_texture(NULL) +{ + m_dev = dev; + + m_rl.Create(this, threads); + + m_fpDrawingKickHandlers[GPU_POLYGON] = (DrawingKickHandler)&GPURendererSW::DrawingKickTriangle; + m_fpDrawingKickHandlers[GPU_LINE] = (DrawingKickHandler)&GPURendererSW::DrawingKickLine; + m_fpDrawingKickHandlers[GPU_SPRITE] = (DrawingKickHandler)&GPURendererSW::DrawingKickSprite; +} + +GPURendererSW::~GPURendererSW() +{ +} + +void GPURendererSW::ResetDevice() +{ + delete m_texture; + + m_texture = NULL; +} + +GSVector4i GPURendererSW::GetScissor() +{ + GSVector4i r; + + r.left = (int)m_env.DRAREATL.X << m_scale.x; + r.top = (int)m_env.DRAREATL.Y << m_scale.y; + r.right = min((int)(m_env.DRAREABR.X + 1) << m_scale.x, m_mem.GetWidth()); + r.bottom = min((int)(m_env.DRAREABR.Y + 1) << m_scale.y, m_mem.GetHeight()); + + return r; +} + +void GPURendererSW::VertexKick() +{ + GSVertexSW& v = m_vl.AddTail(); + + // TODO: x/y + off.x/y should wrap around at +/-1024 + + int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.x; + int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.y; + + int s = m_v.UV.X; + int t = m_v.UV.Y; + + GSVector4 pt(x, y, s, t); + + v.p = pt.xyxy(GSVector4::zero()); + v.t = (pt.zwzw(GSVector4::zero()) + GSVector4(0.125f)) * 256.0f; + v.c = GSVector4(m_v.RGB.u32) * 128.0f; + + __super::VertexKick(); +} + +void GPURendererSW::DrawingKickTriangle(GSVertexSW* v, int& count) +{ + // TODO +} + +void GPURendererSW::DrawingKickLine(GSVertexSW* v, int& count) +{ + // TODO +} + +void GPURendererSW::DrawingKickSprite(GSVertexSW* v, int& count) +{ + // TODO +} + +void GPURendererSW::Draw() +{ + const GPUDrawingEnvironment& env = m_env; + + // + + GPUScanlineParam p; + + p.sel.key = 0; + p.sel.iip = env.PRIM.IIP; + p.sel.me = env.STATUS.ME; + + if(env.PRIM.ABE) + { + p.sel.abe = env.PRIM.ABE; + p.sel.abr = env.STATUS.ABR; + } + + p.sel.tge = env.PRIM.TGE; + + if(env.PRIM.TME) + { + p.sel.tme = env.PRIM.TME; + p.sel.tlu = env.STATUS.TP < 2; + p.sel.twin = (env.TWIN.u32 & 0xfffff) != 0; + p.sel.ltf = m_filter == 1 && env.PRIM.TYPE == GPU_POLYGON || m_filter == 2 ? 1 : 0; + + const void* t = m_mem.GetTexture(env.STATUS.TP, env.STATUS.TX, env.STATUS.TY); + + if(!t) {ASSERT(0); return;} + + p.tex = t; + p.clut = m_mem.GetCLUT(env.STATUS.TP, env.CLUT.X, env.CLUT.Y); + } + + p.sel.dtd = m_dither ? env.STATUS.DTD : 0; + p.sel.md = env.STATUS.MD; + p.sel.sprite = env.PRIM.TYPE == GPU_SPRITE; + p.sel.scalex = m_mem.GetScale().x; + + // + + GSRasterizerData data; + + data.scissor = GetScissor(); + data.vertices = m_vertices; + data.count = m_count; + data.param = &p; + + switch(env.PRIM.TYPE) + { + case GPU_POLYGON: data.primclass = GS_TRIANGLE_CLASS; break; + case GPU_LINE: data.primclass = GS_LINE_CLASS; break; + case GPU_SPRITE: data.primclass = GS_SPRITE_CLASS; break; + default: __assume(0); + } + + m_rl.Draw(&data); + + GSRasterizerStats stats; + + m_rl.GetStats(stats); + + m_perfmon.Put(GSPerfMon::Draw, 1); + m_perfmon.Put(GSPerfMon::Prim, stats.prims); + m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); + + // TODO + + { + GSVector4 tl(+1e10f); + GSVector4 br(-1e10f); + + for(int i = 0, j = m_count; i < j; i++) + { + GSVector4 p = m_vertices[i].p; + + tl = tl.minv(p); + br = br.maxv(p); + } + + GSVector4i r = GSVector4i(tl.xyxy(br)).rintersect(data.scissor); + + r.left >>= m_scale.x; + r.top >>= m_scale.y; + r.right >>= m_scale.x; + r.bottom >>= m_scale.y; + + Invalidate(r); + } +} + +GSTexture* GPURendererSW::GetOutput() +{ + GSVector4i r = m_env.GetDisplayRect(); + + r.left <<= m_scale.x; + r.top <<= m_scale.y; + r.right <<= m_scale.x; + r.bottom <<= m_scale.y; + + // TODO + static uint32* buff = (uint32*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(uint32), 16); + + m_mem.ReadFrame32(r, buff, !!m_env.STATUS.ISRGB24); + + int w = r.width(); + int h = r.height(); + + if(m_texture) + { + if(m_texture->GetWidth() != w || m_texture->GetHeight() != h) + { + delete m_texture; + + m_texture = NULL; + } + } + + if(!m_texture) + { + m_texture = m_dev->CreateTexture(w, h); + + if(!m_texture) + { + return NULL; + } + } + + m_texture->Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); + + return m_texture; +} + diff --git a/plugins/GSdx/GPURendererSW.h b/plugins/GSdx/GPURendererSW.h index 4ea9d29082..58b5910bf9 100644 --- a/plugins/GSdx/GPURendererSW.h +++ b/plugins/GSdx/GPURendererSW.h @@ -24,204 +24,26 @@ #include "GPURenderer.h" #include "GPUDrawScanline.h" -template -class GPURendererSW : public GPURenderer +class GPURendererSW : public GPURendererT { protected: GSRasterizerList m_rl; - Texture m_texture; + GSTexture* m_texture; - void ResetDevice() - { - m_texture = Texture(); - } + void ResetDevice(); - bool GetOutput(Texture& t) - { - GSVector4i r = m_env.GetDisplayRect(); + GSVector4i GetScissor(); - r.left <<= m_scale.x; - r.top <<= m_scale.y; - r.right <<= m_scale.x; - r.bottom <<= m_scale.y; + void VertexKick(); + void DrawingKickTriangle(GSVertexSW* v, int& count); + void DrawingKickLine(GSVertexSW* v, int& count); + void DrawingKickSprite(GSVertexSW* v, int& count); - // TODO - static uint32* buff = (uint32*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(uint32), 16); + void Draw(); - m_mem.ReadFrame32(r, buff, !!m_env.STATUS.ISRGB24); - - int w = r.width(); - int h = r.height(); - - if(m_texture.GetWidth() != w || m_texture.GetHeight() != h) - { - m_texture = Texture(); - } - - if(!m_texture && !m_dev.CreateTexture(m_texture, w, h)) - { - return false; - } - - m_texture.Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); - - t = m_texture; - - return true; - } - - void VertexKick() - { - GSVertexSW& v = m_vl.AddTail(); - - // TODO: x/y + off.x/y should wrap around at +/-1024 - - int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.x; - int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.y; - - int s = m_v.UV.X; - int t = m_v.UV.Y; - - GSVector4 pt(x, y, s, t); - - v.p = pt.xyxy(GSVector4::zero()); - v.t = (pt.zwzw(GSVector4::zero()) + GSVector4(0.125f)) * 256.0f; - v.c = GSVector4(m_v.RGB.u32) * 128.0f; - - __super::VertexKick(); - } - - void DrawingKickTriangle(GSVertexSW* v, int& count) - { - // TODO - } - - void DrawingKickLine(GSVertexSW* v, int& count) - { - // TODO - } - - void DrawingKickSprite(GSVertexSW* v, int& count) - { - // TODO - } - - GSVector4i GetScissor() - { - GSVector4i r; - - r.left = (int)m_env.DRAREATL.X << m_scale.x; - r.top = (int)m_env.DRAREATL.Y << m_scale.y; - r.right = min((int)(m_env.DRAREABR.X + 1) << m_scale.x, m_mem.GetWidth()); - r.bottom = min((int)(m_env.DRAREABR.Y + 1) << m_scale.y, m_mem.GetHeight()); - - return r; - } - - void Draw() - { - const GPUDrawingEnvironment& env = m_env; - - // - - GPUScanlineParam p; - - p.sel.key = 0; - p.sel.iip = env.PRIM.IIP; - p.sel.me = env.STATUS.ME; - - if(env.PRIM.ABE) - { - p.sel.abe = env.PRIM.ABE; - p.sel.abr = env.STATUS.ABR; - } - - p.sel.tge = env.PRIM.TGE; - - if(env.PRIM.TME) - { - p.sel.tme = env.PRIM.TME; - p.sel.tlu = env.STATUS.TP < 2; - p.sel.twin = (env.TWIN.u32 & 0xfffff) != 0; - p.sel.ltf = m_filter == 1 && env.PRIM.TYPE == GPU_POLYGON || m_filter == 2 ? 1 : 0; - - const void* t = m_mem.GetTexture(env.STATUS.TP, env.STATUS.TX, env.STATUS.TY); - - if(!t) {ASSERT(0); return;} - - p.tex = t; - p.clut = m_mem.GetCLUT(env.STATUS.TP, env.CLUT.X, env.CLUT.Y); - } - - p.sel.dtd = m_dither ? env.STATUS.DTD : 0; - p.sel.md = env.STATUS.MD; - p.sel.sprite = env.PRIM.TYPE == GPU_SPRITE; - p.sel.scalex = m_mem.GetScale().x; - - // - - GSRasterizerData data; - - data.scissor = GetScissor(); - data.vertices = m_vertices; - data.count = m_count; - data.param = &p; - - switch(env.PRIM.TYPE) - { - case GPU_POLYGON: data.primclass = GS_TRIANGLE_CLASS; break; - case GPU_LINE: data.primclass = GS_LINE_CLASS; break; - case GPU_SPRITE: data.primclass = GS_SPRITE_CLASS; break; - default: __assume(0); - } - - m_rl.Draw(&data); - - GSRasterizerStats stats; - - m_rl.GetStats(stats); - - m_perfmon.Put(GSPerfMon::Draw, 1); - m_perfmon.Put(GSPerfMon::Prim, stats.prims); - m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); - - // TODO - - { - GSVector4 tl(+1e10f); - GSVector4 br(-1e10f); - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 p = m_vertices[i].p; - - tl = tl.minv(p); - br = br.maxv(p); - } - - GSVector4i r = GSVector4i(tl.xyxy(br)).rintersect(data.scissor); - - r.left >>= m_scale.x; - r.top >>= m_scale.y; - r.right >>= m_scale.x; - r.bottom >>= m_scale.y; - - Invalidate(r); - } - } + GSTexture* GetOutput(); public: - GPURendererSW(const GPURendererSettings& rs, int threads) - : GPURenderer(rs) - { - m_rl.Create(this, threads); - - m_fpDrawingKickHandlers[GPU_POLYGON] = (DrawingKickHandler)&GPURendererSW::DrawingKickTriangle; - m_fpDrawingKickHandlers[GPU_LINE] = (DrawingKickHandler)&GPURendererSW::DrawingKickLine; - m_fpDrawingKickHandlers[GPU_SPRITE] = (DrawingKickHandler)&GPURendererSW::DrawingKickSprite; - } - - virtual ~GPURendererSW() - { - } + GPURendererSW(const GPURendererSettings& rs, GSDevice* dev, int threads); + virtual ~GPURendererSW(); }; diff --git a/plugins/GSdx/GPUSetupPrimCodeGenerator.cpp b/plugins/GSdx/GPUSetupPrimCodeGenerator.cpp index d70903dbbb..43ebb94d93 100644 --- a/plugins/GSdx/GPUSetupPrimCodeGenerator.cpp +++ b/plugins/GSdx/GPUSetupPrimCodeGenerator.cpp @@ -38,8 +38,6 @@ GPUSetupPrimCodeGenerator::GPUSetupPrimCodeGenerator(GPUScanlineEnvironment& env void GPUSetupPrimCodeGenerator::Generate() { - const int params = 0; - if(m_env.sel.tme && !m_env.sel.twin) { pcmpeqd(xmm0, xmm0); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index ff0133ac0e..b6612429b8 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -33,7 +33,7 @@ #define PS2E_X86_64 0x02 // 64 bit static HRESULT s_hr = E_FAIL; -static GSRendererBase* s_gs = NULL; +static GSRenderer* s_gs = NULL; static void (*s_irq)() = NULL; static uint8* s_basemem = NULL; @@ -134,13 +134,13 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) { default: case 0: s_gs = new GSRendererHW9(s_basemem, !!mt, s_irq, rs); break; - case 1: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, threads); break; - case 2: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs); break; + case 1: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDevice9(), threads); break; + case 2: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDevice9()); break; case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq, rs); break; - case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, threads); break; - case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs); break; - case 6: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, threads); break; - case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs); break; + case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDevice10(), threads); break; + case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDevice10()); break; + case 6: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDeviceNull(), threads); break; + case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDeviceNull()); break; } if(!s_gs->Create(title)) diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index e62751a8eb..2e523d9003 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -440,18 +440,22 @@ REG64_(GSReg, SMODE2) REG_END REG64_(GSReg, SRFSH) + uint32 _DUMMY; // TODO REG_END REG64_(GSReg, SYNCH1) + uint32 _DUMMY; // TODO REG_END REG64_(GSReg, SYNCH2) + uint32 _DUMMY; // TODO REG_END REG64_(GSReg, SYNCV) + uint64 _DUMMY; // TODO REG_END diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index c6e3385fd4..27084e87b1 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -126,6 +126,8 @@ void GSClut::Write(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { + ALIGN_STACK(16); + ASSERT(TEX0.CSA == 0); WriteCLUT_T32_I8_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); @@ -133,9 +135,9 @@ void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TE void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) { - ASSERT(TEX0.CSA < 16); + ALIGN_STACK(16); - GSVector4i dummy; // this just forces stack alignment and enables inlining the next call + ASSERT(TEX0.CSA < 16); WriteCLUT_T32_I4_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); } diff --git a/plugins/GSdx/GSDevice.cpp b/plugins/GSdx/GSDevice.cpp index fe40a0da83..c76a18dc13 100644 --- a/plugins/GSdx/GSDevice.cpp +++ b/plugins/GSdx/GSDevice.cpp @@ -21,3 +21,195 @@ #include "StdAfx.h" #include "GSDevice.h" + +GSDevice::GSDevice() + : m_hWnd(NULL) + , m_backbuffer(NULL) + , m_merge(NULL) + , m_weavebob(NULL) + , m_blend(NULL) + , m_1x1(NULL) +{ +} + +GSDevice::~GSDevice() +{ +} + +bool GSDevice::Create(HWND hWnd, bool vsync) +{ + m_hWnd = hWnd; + m_vsync = vsync; + + return true; +} + +bool GSDevice::Reset(int w, int h, bool fs) +{ + for(list::iterator i = m_pool.begin(); i != m_pool.end(); i++) + { + delete *i; + } + + m_pool.clear(); + + delete m_backbuffer; + delete m_merge; + delete m_weavebob; + delete m_blend; + delete m_1x1; + + m_backbuffer = NULL; + m_merge = NULL; + m_weavebob = NULL; + m_blend = NULL; + m_1x1 = NULL; + + m_current = NULL; // current is special, points to other textures, no need to delete + + return true; +} + +void GSDevice::Present(const GSVector4i& r) +{ + GSVector4i cr; + + GetClientRect(m_hWnd, cr); + + if(m_backbuffer->GetWidth() != cr.width() || m_backbuffer->GetHeight() != cr.height()) + { + Reset(cr.width(), cr.height(), false); + } + + ClearRenderTarget(m_backbuffer, 0); + + if(m_current) + { + StretchRect(m_current, m_backbuffer, GSVector4(r)); + } + + Flip(); +} + +GSTexture* GSDevice::Fetch(int type, int w, int h, int format) +{ + GSVector2i size(w, h); + + for(list::iterator i = m_pool.begin(); i != m_pool.end(); i++) + { + GSTexture* t = *i; + + if(t->GetType() == type && t->GetFormat() == format && t->GetSize() == size) + { + m_pool.erase(i); + + return t; + } + } + + return Create(type, w, h, format); +} + +void GSDevice::Recycle(GSTexture* t) +{ + if(t) + { + m_pool.push_front(t); + + while(m_pool.size() > 200) + { + delete m_pool.back(); + + m_pool.pop_back(); + } + } +} + +GSTexture* GSDevice::CreateRenderTarget(int w, int h, int format) +{ + return Fetch(GSTexture::RenderTarget, w, h, format); +} + +GSTexture* GSDevice::CreateDepthStencil(int w, int h, int format) +{ + return Fetch(GSTexture::DepthStencil, w, h, format); +} + +GSTexture* GSDevice::CreateTexture(int w, int h, int format) +{ + return Fetch(GSTexture::Texture, w, h, format); +} + +GSTexture* GSDevice::CreateOffscreen(int w, int h, int format) +{ + return Fetch(GSTexture::Offscreen, w, h, format); +} + +void GSDevice::StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, bool linear) +{ + StretchRect(st, GSVector4(0, 0, 1, 1), dt, dr, linear); +} + +GSTexture* GSDevice::GetCurrent() +{ + return m_current; +} + +void GSDevice::Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, const GSVector4& c) +{ + if(!m_merge || !(m_merge->GetSize() == fs)) + { + m_merge = CreateRenderTarget(fs.x, fs.y); + } + + // TODO: m_1x1 + + DoMerge(st, sr, dr, m_merge, slbg, mmod, c); + + m_current = m_merge; +} + +bool GSDevice::Interlace(const GSVector2i& ds, int field, int mode, float yoffset) +{ + if(!m_weavebob || !(m_weavebob->GetSize() == ds)) + { + m_weavebob = CreateRenderTarget(ds.x, ds.y); + } + + if(mode == 0 || mode == 2) // weave or blend + { + // weave first + + DoInterlace(m_merge, m_weavebob, field, false, 0); + + if(mode == 2) + { + // blend + + if(!m_blend || !(m_blend->GetSize() == ds)) + { + m_blend = CreateRenderTarget(ds.x, ds.y); + } + + DoInterlace(m_weavebob, m_blend, 2, false, 0); + + m_current = m_blend; + } + else + { + m_current = m_weavebob; + } + } + else if(mode == 1) // bob + { + DoInterlace(m_merge, m_weavebob, 3, true, yoffset * field); + + m_current = m_weavebob; + } + else + { + m_current = m_merge; + } + + return true; +} \ No newline at end of file diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index bf0595a1b6..ca41beac55 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -23,6 +23,7 @@ #include "GSTexture.h" #include "GSVertex.h" +#include "GSAlignedClass.h" #pragma pack(push, 1) @@ -44,201 +45,64 @@ struct InterlaceConstantBuffer #pragma pack(pop) -template class GSDevice +class GSDevice : public GSAlignedClass<16> { - list m_pool; + list m_pool; + + GSTexture* Fetch(int type, int w, int h, int format); protected: HWND m_hWnd; bool m_vsync; - Texture m_backbuffer; - Texture m_merge; - Texture m_weavebob; - Texture m_blend; - Texture m_1x1; - Texture m_current; + GSTexture* m_backbuffer; + GSTexture* m_merge; + GSTexture* m_weavebob; + GSTexture* m_blend; + GSTexture* m_1x1; + GSTexture* m_current; - bool Fetch(int type, Texture& t, int w, int h, int format) - { - Recycle(t); + virtual GSTexture* Create(int type, int w, int h, int format) = 0; - for(list::iterator i = m_pool.begin(); i != m_pool.end(); i++) - { - const Texture& t2 = *i; - - if(t2.GetType() == type && t2.GetWidth() == w && t2.GetHeight() == h && t2.GetFormat() == format) - { - t = t2; - - m_pool.erase(i); - - return true; - } - } - - return Create(type, t, w, h, format); - } - - virtual bool Create(int type, Texture& t, int w, int h, int format) = 0; - virtual void DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c) = 0; - virtual void DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset) = 0; + virtual void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) = 0; + virtual void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) = 0; public: - GSDevice() : m_hWnd(NULL) - { - } + GSDevice(); + virtual ~GSDevice(); - virtual ~GSDevice() - { - } + void Recycle(GSTexture* t); - virtual bool Create(HWND hWnd, bool vsync) - { - m_hWnd = hWnd; - m_vsync = vsync; + virtual bool Create(HWND hWnd, bool vsync); + virtual bool Reset(int w, int h, bool fs); - return true; - } - - virtual bool Reset(int w, int h, bool fs) - { - m_pool.clear(); - m_backbuffer = Texture(); - m_merge = Texture(); - m_weavebob = Texture(); - m_blend = Texture(); - m_1x1 = Texture(); - m_current = Texture(); + virtual bool IsLost() {return false;} + virtual void Present(const GSVector4i& r); + virtual void Flip() {}; - return true; - } + virtual void BeginScene() {}; + virtual void EndScene() {}; - virtual bool IsLost() = 0; + virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {}; + virtual void ClearRenderTarget(GSTexture* t, uint32 c) {}; + virtual void ClearDepth(GSTexture* t, float c) {}; + virtual void ClearStencil(GSTexture* t, uint8 c) {}; - virtual void Present(const GSVector4i& r) = 0; + virtual GSTexture* CreateRenderTarget(int w, int h, int format = 0); + virtual GSTexture* CreateDepthStencil(int w, int h, int format = 0); + virtual GSTexture* CreateTexture(int w, int h, int format = 0); + virtual GSTexture* CreateOffscreen(int w, int h, int format = 0); - virtual void BeginScene() = 0; + virtual GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0) {return NULL;} - virtual void EndScene() = 0; + virtual void StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, bool linear = true); + virtual void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true) {} - virtual void Draw(const string& s) = 0; + GSTexture* GetCurrent(); + virtual bool IsCurrentRGBA() {return true;} - virtual bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) = 0; + void Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, const GSVector4& c); + bool Interlace(const GSVector2i& ds, int field, int mode, float yoffset); - virtual void ClearRenderTarget(Texture& t, const GSVector4& c) = 0; - - virtual void ClearRenderTarget(Texture& t, uint32 c) = 0; - - virtual void ClearDepth(Texture& t, float c) = 0; - - virtual void ClearStencil(Texture& t, uint8 c) = 0; - - virtual bool CreateRenderTarget(Texture& t, int w, int h, int format = 0) - { - return Fetch(GSTexture::RenderTarget, t, w, h, format); - } - - virtual bool CreateDepthStencil(Texture& t, int w, int h, int format = 0) - { - return Fetch(GSTexture::DepthStencil, t, w, h, format); - } - - virtual bool CreateTexture(Texture& t, int w, int h, int format = 0) - { - return Fetch(GSTexture::Texture, t, w, h, format); - } - - virtual bool CreateOffscreen(Texture& t, int w, int h, int format = 0) - { - return Fetch(GSTexture::Offscreen, t, w, h, format); - } - - void Recycle(Texture& t) - { - if(t) - { - m_pool.push_front(t); - - while(m_pool.size() > 200) - { - m_pool.pop_back(); - } - - t = Texture(); - } - } - - bool SaveCurrent(const string& fn) - { - return m_current.Save(fn); - } - - void GetCurrent(Texture& t) - { - t = m_current; - } - - void Merge(Texture* st, GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, GSVector4& c) - { - if(!m_merge || m_merge.GetWidth() != fs.x || m_merge.GetHeight() != fs.y) - { - CreateRenderTarget(m_merge, fs.x, fs.y); - } - - // TODO: m_1x1 - - DoMerge(st, sr, dr, m_merge, slbg, mmod, c); - - m_current = m_merge; - } - - bool Interlace(const GSVector2i& ds, int field, int mode, float yoffset) - { - if(!m_weavebob || m_weavebob.GetWidth() != ds.x || m_weavebob.GetHeight() != ds.y) - { - CreateRenderTarget(m_weavebob, ds.x, ds.y); - } - - if(mode == 0 || mode == 2) // weave or blend - { - // weave first - - DoInterlace(m_merge, m_weavebob, field, false, 0); - - if(mode == 2) - { - // blend - - if(!m_blend || m_blend.GetWidth() != ds.x || m_blend.GetHeight() != ds.y) - { - CreateRenderTarget(m_blend, ds.x, ds.y); - } - - DoInterlace(m_weavebob, m_blend, 2, false, 0); - - m_current = m_blend; - } - else - { - m_current = m_weavebob; - } - } - else if(mode == 1) // bob - { - DoInterlace(m_merge, m_weavebob, 3, true, yoffset * field); - - m_current = m_weavebob; - } - else - { - m_current = m_merge; - } - - return true; - } - - virtual bool IsCurrentRGBA() - { - return true; - } + virtual void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) {} + virtual void OMSetRenderTargets(GSTexture* rt, GSTexture* ds) {}; }; diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 72d6f65a11..8eadbb5751 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -251,31 +251,13 @@ bool GSDevice10::Reset(int w, int h, bool fs) CComPtr backbuffer; m_swapchain->GetBuffer(0, __uuidof(ID3D10Texture2D), (void**)&backbuffer); - m_backbuffer = Texture(backbuffer); + m_backbuffer = new GSTexture10(backbuffer); return true; } -void GSDevice10::Present(const GSVector4i& r) +void GSDevice10::Flip() { - GSVector4i cr; - - GetClientRect(m_hWnd, cr); - - if(m_backbuffer.GetWidth() != cr.width() || m_backbuffer.GetHeight() != cr.height()) - { - Reset(cr.width(), cr.height(), false); - } - - float color[4] = {0, 0, 0, 0}; - - m_dev->ClearRenderTargetView(m_backbuffer, color); - - if(m_current) - { - StretchRect(m_current, m_backbuffer, GSVector4(r)); - } - m_swapchain->Present(m_vsync ? 1 : 0, 0); } @@ -292,92 +274,29 @@ void GSDevice10::EndScene() // OMSetRenderTargets(NULL, NULL); } -void GSDevice10::Draw(const string& s) +void GSDevice10::ClearRenderTarget(GSTexture* t, const GSVector4& c) { - /* - BOOL fs; - CComPtr target; - - m_swapchain->GetFullscreenState(&fs, &target); - - if(fs) - { - BeginScene(); - - OMSetRenderTargets(m_backbuffer, NULL); - - GSVector4i r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); - - D3DCOLOR c = D3DCOLOR_ARGB(255, 0, 255, 0); - - if(m_font->DrawText(NULL, str, -1, r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) - { - m_font->DrawText(NULL, str, -1, r, DT_LEFT|DT_WORDBREAK, c); - } - - EndScene(); - } - */ + m_dev->ClearRenderTargetView(*(GSTexture10*)t, c.v); } -bool GSDevice10::CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format) -{ - dst = Texture(); - - if(format == 0) - { - format = DXGI_FORMAT_R8G8B8A8_UNORM; - } - - if(format != DXGI_FORMAT_R8G8B8A8_UNORM && format != DXGI_FORMAT_R16_UINT) - { - ASSERT(0); - - return false; - } - - Texture rt; - - if(CreateRenderTarget(rt, w, h, format)) - { - GSVector4 dr(0, 0, w, h); - - StretchRect(src, sr, rt, dr, m_convert.ps[format == DXGI_FORMAT_R16_UINT ? 1 : 0], NULL); - - if(CreateOffscreen(dst, w, h, format)) - { - m_dev->CopyResource(dst, rt); - } - } - - Recycle(rt); - - return !!dst; -} - -void GSDevice10::ClearRenderTarget(Texture& t, const GSVector4& c) -{ - m_dev->ClearRenderTargetView(t, c.v); -} - -void GSDevice10::ClearRenderTarget(Texture& t, uint32 c) +void GSDevice10::ClearRenderTarget(GSTexture* t, uint32 c) { GSVector4 color = GSVector4(c) * (1.0f / 255); - m_dev->ClearRenderTargetView(t, color.v); + m_dev->ClearRenderTargetView(*(GSTexture10*)t, color.v); } -void GSDevice10::ClearDepth(Texture& t, float c) +void GSDevice10::ClearDepth(GSTexture* t, float c) { - m_dev->ClearDepthStencilView(t, D3D10_CLEAR_DEPTH, c, 0); + m_dev->ClearDepthStencilView(*(GSTexture10*)t, D3D10_CLEAR_DEPTH, c, 0); } -void GSDevice10::ClearStencil(Texture& t, uint8 c) +void GSDevice10::ClearStencil(GSTexture* t, uint8 c) { - m_dev->ClearDepthStencilView(t, D3D10_CLEAR_STENCIL, 0, c); + m_dev->ClearDepthStencilView(*(GSTexture10*)t, D3D10_CLEAR_STENCIL, 0, c); } -bool GSDevice10::Create(int type, Texture& t, int w, int h, int format) +GSTexture* GSDevice10::Create(int type, int w, int h, int format) { HRESULT hr; @@ -411,13 +330,15 @@ bool GSDevice10::Create(int type, Texture& t, int w, int h, int format) break; } + GSTexture10* t = NULL; + CComPtr texture; hr = m_dev->CreateTexture2D(&desc, NULL, &texture); if(SUCCEEDED(hr)) { - t = Texture(texture); + t = new GSTexture10(texture); switch(type) { @@ -428,34 +349,139 @@ bool GSDevice10::Create(int type, Texture& t, int w, int h, int format) ClearDepth(t, 0); break; } - - return true; } - return false; + return t; } -bool GSDevice10::CreateRenderTarget(Texture& t, int w, int h, int format) +GSTexture* GSDevice10::CreateRenderTarget(int w, int h, int format) { - return __super::CreateRenderTarget(t, w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); + return __super::CreateRenderTarget(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); } -bool GSDevice10::CreateDepthStencil(Texture& t, int w, int h, int format) +GSTexture* GSDevice10::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(t, w, h, format ? format : DXGI_FORMAT_D32_FLOAT_S8X24_UINT); + return __super::CreateDepthStencil(w, h, format ? format : DXGI_FORMAT_D32_FLOAT_S8X24_UINT); } -bool GSDevice10::CreateTexture(Texture& t, int w, int h, int format) +GSTexture* GSDevice10::CreateTexture(int w, int h, int format) { - return __super::CreateTexture(t, w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); + return __super::CreateTexture(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); } -bool GSDevice10::CreateOffscreen(Texture& t, int w, int h, int format) +GSTexture* GSDevice10::CreateOffscreen(int w, int h, int format) { - return __super::CreateOffscreen(t, w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); + return __super::CreateOffscreen(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); } -void GSDevice10::DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c) +GSTexture* GSDevice10::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format) +{ + GSTexture* dst = NULL; + + if(format == 0) + { + format = DXGI_FORMAT_R8G8B8A8_UNORM; + } + + if(format != DXGI_FORMAT_R8G8B8A8_UNORM && format != DXGI_FORMAT_R16_UINT) + { + ASSERT(0); + + return false; + } + + if(GSTexture* rt = CreateRenderTarget(w, h, format)) + { + GSVector4 dr(0, 0, w, h); + + StretchRect(src, sr, rt, dr, m_convert.ps[format == DXGI_FORMAT_R16_UINT ? 1 : 0], NULL); + + dst = CreateOffscreen(w, h, format); + + if(dst) + { + m_dev->CopyResource(*(GSTexture10*)dst, *(GSTexture10*)rt); + } + + Recycle(rt); + } + + return dst; +} + +void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear) +{ + StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, linear); +} + +void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear) +{ + StretchRect(st, sr, dt, dr, ps, ps_cb, m_convert.bs, linear); +} + +void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear) +{ + BeginScene(); + + GSVector2i ds = dt->GetSize(); + + // om + + OMSetDepthStencilState(m_convert.dss, 0); + OMSetBlendState(bs, 0); + OMSetRenderTargets(dt, NULL); + + // ia + + float left = dr.x * 2 / ds.x - 1.0f; + float top = 1.0f - dr.y * 2 / ds.y; + float right = dr.z * 2 / ds.x - 1.0f; + float bottom = 1.0f - dr.w * 2 / ds.y; + + GSVertexPT1 vertices[] = + { + {GSVector4(left, top, 0.5f, 1.0f), GSVector2(sr.x, sr.y)}, + {GSVector4(right, top, 0.5f, 1.0f), GSVector2(sr.z, sr.y)}, + {GSVector4(left, bottom, 0.5f, 1.0f), GSVector2(sr.x, sr.w)}, + {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, + }; + + D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; + + m_dev->UpdateSubresource(m_convert.vb, 0, &box, vertices, 0, 0); + + IASetVertexBuffer(m_convert.vb, sizeof(vertices[0])); + IASetInputLayout(m_convert.il); + IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + VSSetShader(m_convert.vs, NULL); + + // gs + + GSSetShader(NULL); + + // ps + + PSSetShader(ps, ps_cb); + PSSetSamplerState(linear ? m_convert.ln : m_convert.pt, NULL); + PSSetShaderResources(st, NULL); + + // rs + + RSSet(ds.x, ds.y); + + // + + DrawPrimitive(countof(vertices)); + + // + + EndScene(); +} + +void GSDevice10::DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) { ClearRenderTarget(dt, c); @@ -472,15 +498,17 @@ void GSDevice10::DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, } } -void GSDevice10::DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset) +void GSDevice10::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) { + GSVector4 s = GSVector4(dt->GetSize()); + GSVector4 sr(0, 0, 1, 1); - GSVector4 dr(0.0f, yoffset, (float)dt.GetWidth(), (float)dt.GetHeight() + yoffset); + GSVector4 dr(0.0f, yoffset, s.x, s.y + yoffset); InterlaceConstantBuffer cb; - cb.ZrH = GSVector2(0, 1.0f / dt.GetHeight()); - cb.hH = (float)dt.GetHeight() / 2; + cb.ZrH = GSVector2(0, 1.0f / s.y); + cb.hH = s.y / 2; m_dev->UpdateSubresource(m_interlace.cb, 0, NULL, &cb, 0, 0); @@ -547,8 +575,14 @@ void GSDevice10::GSSetShader(ID3D10GeometryShader* gs) } } -void GSDevice10::PSSetShaderResources(ID3D10ShaderResourceView* srv0, ID3D10ShaderResourceView* srv1) +void GSDevice10::PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) { + ID3D10ShaderResourceView* srv0 = NULL; + ID3D10ShaderResourceView* srv1 = NULL; + + if(sr0) srv0 = *(GSTexture10*)sr0; + if(sr1) srv1 = *(GSTexture10*)sr1; + if(m_ps_srv[0] != srv0 || m_ps_srv[1] != srv1) { ID3D10ShaderResourceView* srvs[] = {srv0, srv1}; @@ -644,8 +678,14 @@ void GSDevice10::OMSetBlendState(ID3D10BlendState* bs, float bf) } } -void GSDevice10::OMSetRenderTargets(ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv) +void GSDevice10::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) { + ID3D10RenderTargetView* rtv = NULL; + ID3D10DepthStencilView* dsv = NULL; + + if(rt) rtv = *(GSTexture10*)rt; + if(ds) dsv = *(GSTexture10*)ds; + if(m_rtv != rtv || m_dsv != dsv) { m_dev->OMSetRenderTargets(1, &rtv, dsv); @@ -660,81 +700,6 @@ void GSDevice10::DrawPrimitive(uint32 count, uint32 start) m_dev->Draw(count, start); } -void GSDevice10::StretchRect(Texture& st, Texture& dt, const GSVector4& dr, bool linear) -{ - StretchRect(st, GSVector4(0, 0, 1, 1), dt, dr, linear); -} - -void GSDevice10::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, bool linear) -{ - StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, linear); -} - -void GSDevice10::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear) -{ - StretchRect(st, sr, dt, dr, ps, ps_cb, m_convert.bs, linear); -} - -void GSDevice10::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear) -{ - BeginScene(); - - // om - - OMSetDepthStencilState(m_convert.dss, 0); - OMSetBlendState(bs, 0); - OMSetRenderTargets(dt, NULL); - - // ia - - float left = dr.x * 2 / dt.GetWidth() - 1.0f; - float top = 1.0f - dr.y * 2 / dt.GetHeight(); - float right = dr.z * 2 / dt.GetWidth() - 1.0f; - float bottom = 1.0f - dr.w * 2 / dt.GetHeight(); - - GSVertexPT1 vertices[] = - { - {GSVector4(left, top, 0.5f, 1.0f), GSVector2(sr.x, sr.y)}, - {GSVector4(right, top, 0.5f, 1.0f), GSVector2(sr.z, sr.y)}, - {GSVector4(left, bottom, 0.5f, 1.0f), GSVector2(sr.x, sr.w)}, - {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, - }; - - D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; - - m_dev->UpdateSubresource(m_convert.vb, 0, &box, vertices, 0, 0); - - IASetVertexBuffer(m_convert.vb, sizeof(vertices[0])); - IASetInputLayout(m_convert.il); - IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); - - // vs - - VSSetShader(m_convert.vs, NULL); - - // gs - - GSSetShader(NULL); - - // ps - - PSSetShader(ps, ps_cb); - PSSetSamplerState(linear ? m_convert.ln : m_convert.pt, NULL); - PSSetShaderResources(st, NULL); - - // rs - - RSSet(dt.GetWidth(), dt.GetHeight()); - - // - - DrawPrimitive(countof(vertices)); - - // - - EndScene(); -} - HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) { HRESULT hr; diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index 81c5b7cecf..3179aea568 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -24,11 +24,8 @@ #include "GSDevice.h" #include "GSTexture10.h" -class GSDevice10 : public GSDevice +class GSDevice10 : public GSDevice { -public: - typedef GSTexture10 Texture; - private: // state cache @@ -55,9 +52,10 @@ private: // - bool Create(int type, Texture& t, int w, int h, int format); - void DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c); - void DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset = 0); + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0); // @@ -99,44 +97,44 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); - bool IsLost() {return false;} - void Present(const GSVector4i& r); + + void Flip(); + void BeginScene(); void EndScene(); - void Draw(const string& s); - bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); - void ClearRenderTarget(Texture& t, const GSVector4& c); - void ClearRenderTarget(Texture& t, uint32 c); - void ClearDepth(Texture& t, float c); - void ClearStencil(Texture& t, uint8 c); + void ClearRenderTarget(GSTexture* t, const GSVector4& c); + void ClearRenderTarget(GSTexture* t, uint32 c); + void ClearDepth(GSTexture* t, float c); + void ClearStencil(GSTexture* t, uint8 c); - bool CreateRenderTarget(Texture& t, int w, int h, int format = 0); - bool CreateDepthStencil(Texture& t, int w, int h, int format = 0); - bool CreateTexture(Texture& t, int w, int h, int format = 0); - bool CreateOffscreen(Texture& t, int w, int h, int format = 0); + GSTexture* CreateRenderTarget(int w, int h, int format = 0); + GSTexture* CreateDepthStencil(int w, int h, int format = 0); + GSTexture* CreateTexture(int w, int h, int format = 0); + GSTexture* CreateOffscreen(int w, int h, int format = 0); - ID3D10Device* operator->() {return m_dev;} - operator ID3D10Device*() {return m_dev;} + GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); + + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); void IASetVertexBuffer(ID3D10Buffer* vb, size_t stride); void IASetInputLayout(ID3D10InputLayout* layout); void IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY topology); void VSSetShader(ID3D10VertexShader* vs, ID3D10Buffer* vs_cb); void GSSetShader(ID3D10GeometryShader* gs); - void PSSetShaderResources(ID3D10ShaderResourceView* srv0, ID3D10ShaderResourceView* srv1); + void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1); void PSSetShader(ID3D10PixelShader* ps, ID3D10Buffer* ps_cb); void PSSetSamplerState(ID3D10SamplerState* ss0, ID3D10SamplerState* ss1); void RSSet(int width, int height, const GSVector4i* scissor = NULL); void OMSetDepthStencilState(ID3D10DepthStencilState* dss, uint8 sref); void OMSetBlendState(ID3D10BlendState* bs, float bf); - void OMSetRenderTargets(ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv); + void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); void DrawPrimitive(uint32 count, uint32 start = 0); - void StretchRect(Texture& st, Texture& dt, const GSVector4& dr, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); + ID3D10Device* operator->() {return m_dev;} + operator ID3D10Device*() {return m_dev;} HRESULT CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il); HRESULT CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10GeometryShader** gs); diff --git a/plugins/GSdx/GSDevice7.cpp b/plugins/GSdx/GSDevice7.cpp index 477efe6638..32b09d41e7 100644 --- a/plugins/GSdx/GSDevice7.cpp +++ b/plugins/GSdx/GSDevice7.cpp @@ -83,8 +83,6 @@ bool GSDevice7::Reset(int w, int h, bool fs) if(!__super::Reset(w, h, fs)) return false; - m_backbuffer = NULL; - DDSURFACEDESC2 desc; memset(&desc, 0, sizeof(desc)); @@ -95,11 +93,15 @@ bool GSDevice7::Reset(int w, int h, bool fs) desc.dwWidth = w; desc.dwHeight = h; - if(FAILED(m_dd->CreateSurface(&desc, &m_backbuffer, NULL))) + CComPtr backbuffer; + + if(FAILED(m_dd->CreateSurface(&desc, &backbuffer, NULL))) { return false; } + m_backbuffer = new GSTexture7(GSTexture::RenderTarget, backbuffer); + CComPtr clipper; if(FAILED(m_dd->CreateClipper(0, &clipper, NULL))) @@ -120,7 +122,7 @@ bool GSDevice7::Reset(int w, int h, bool fs) clipper->SetClipList((RGNDATA*)buff, 0); - if(FAILED(m_backbuffer->SetClipper(clipper))) + if(FAILED(backbuffer->SetClipper(clipper))) { return false; } @@ -137,17 +139,13 @@ void GSDevice7::Present(const GSVector4i& r) GetClientRect(m_hWnd, cr); - DDSURFACEDESC2 desc; - memset(&desc, 0, sizeof(desc)); - desc.dwSize = sizeof(desc); - - hr = m_backbuffer->GetSurfaceDesc(&desc); - - if(desc.dwWidth != cr.width() || desc.dwHeight != cr.height()) + if(m_backbuffer->GetWidth() != cr.width() || m_backbuffer->GetHeight() != cr.height()) { Reset(cr.width(), cr.height(), false); } + CComPtr backbuffer = *(GSTexture7*)m_backbuffer; + DDBLTFX fx; memset(&fx, 0, sizeof(fx)); @@ -155,11 +153,13 @@ void GSDevice7::Present(const GSVector4i& r) fx.dwSize = sizeof(fx); fx.dwFillColor = 0; - hr = m_backbuffer->Blt(NULL, NULL, NULL, DDBLT_WAIT | DDBLT_COLORFILL, &fx); + hr = backbuffer->Blt(NULL, NULL, NULL, DDBLT_WAIT | DDBLT_COLORFILL, &fx); GSVector4i r2 = r; - hr = m_backbuffer->Blt(r2, m_merge, NULL, DDBLT_WAIT, NULL); + hr = backbuffer->Blt(r2, *(GSTexture7*)m_merge, NULL, DDBLT_WAIT, NULL); + + // if ClearRenderTarget was implemented the parent class could handle these tasks until this point r2 = cr; @@ -170,7 +170,7 @@ void GSDevice7::Present(const GSVector4i& r) hr = m_dd->WaitForVerticalBlank(DDWAITVB_BLOCKBEGIN, NULL); } - hr = m_primary->Blt(r2, m_backbuffer, cr, DDBLT_WAIT, NULL); + hr = m_primary->Blt(r2, backbuffer, cr, DDBLT_WAIT, NULL); if(hr == DDERR_SURFACELOST) { @@ -184,12 +184,10 @@ void GSDevice7::Present(const GSVector4i& r) } } -bool GSDevice7::Create(int type, Texture& t, int w, int h, int format) +GSTexture* GSDevice7::Create(int type, int w, int h, int format) { HRESULT hr; - t = Texture(); - DDSURFACEDESC2 desc; memset(&desc, 0, sizeof(desc)); @@ -207,6 +205,8 @@ bool GSDevice7::Create(int type, Texture& t, int w, int h, int format) desc.ddpfPixelFormat.dwGBitMask = 0x0000ff00; desc.ddpfPixelFormat.dwBBitMask = 0x000000ff; + GSTexture7* t = NULL; + CComPtr system, video; switch(type) @@ -218,25 +218,25 @@ bool GSDevice7::Create(int type, Texture& t, int w, int h, int format) if(FAILED(hr = m_dd->CreateSurface(&desc, &system, NULL))) return false; desc.ddsCaps.dwCaps = DDSCAPS_TEXTURE | DDSCAPS_VIDEOMEMORY; if(FAILED(hr = m_dd->CreateSurface(&desc, &video, NULL))) return false; - t = Texture(type, system, video); + t = new GSTexture7(type, system, video); break; case GSTexture::Offscreen: desc.ddsCaps.dwCaps = DDSCAPS_OFFSCREENPLAIN | DDSCAPS_VIDEOMEMORY; if(FAILED(hr = m_dd->CreateSurface(&desc, &system, NULL))) return false; - t = Texture(type, system); + t = new GSTexture7(type, system); break; } - return !!t; + return t; } -void GSDevice7::DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c) +void GSDevice7::DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) { HRESULT hr; - hr = dt->Blt(NULL, st[0], NULL, DDBLT_WAIT, NULL); + hr = (*(GSTexture7*)dt)->Blt(NULL, *(GSTexture7*)st[0], NULL, DDBLT_WAIT, NULL); } -void GSDevice7::DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset) +void GSDevice7::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) { } diff --git a/plugins/GSdx/GSDevice7.h b/plugins/GSdx/GSDevice7.h index 3029259a92..e0327f003b 100644 --- a/plugins/GSdx/GSDevice7.h +++ b/plugins/GSdx/GSDevice7.h @@ -24,19 +24,16 @@ #include "GSDevice.h" #include "GSTexture7.h" -class GSDevice7 : public GSDevice +class GSDevice7 : public GSDevice { -public: - typedef GSTexture7 Texture; - private: CComPtr m_dd; CComPtr m_primary; - CComPtr m_backbuffer; - bool Create(int type, Texture& t, int w, int h, int format); - void DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c); - void DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset = 0); + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0); public: GSDevice7(); @@ -44,15 +41,6 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); - bool IsLost() {return false;} - void Present(const GSVector4i& r); - void BeginScene() {} - void EndScene() {} - void Draw(const string& s) {} - bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} - void ClearRenderTarget(Texture& t, const GSVector4& c) {} - void ClearRenderTarget(Texture& t, uint32 c) {} - void ClearDepth(Texture& t, float c) {} - void ClearStencil(Texture& t, uint8 c) {} + void Present(const GSVector4i& r); }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 461cb30842..bb13ecb389 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -215,15 +215,14 @@ bool GSDevice9::Reset(int w, int h, bool fs) CComPtr backbuffer; hr = m_swapchain->GetBackBuffer(0, D3DBACKBUFFER_TYPE_MONO, &backbuffer); - m_backbuffer = Texture(backbuffer); + m_backbuffer = new GSTexture9(backbuffer); return true; } m_swapchain = NULL; - m_backbuffer = Texture(); - if(m_font) m_font->OnLostDevice(); - m_font = NULL; + + if(m_font) {m_font->OnLostDevice(); m_font = NULL;} if(m_vs_cb) _aligned_free(m_vs_cb); if(m_ps_cb) _aligned_free(m_ps_cb); @@ -325,7 +324,7 @@ bool GSDevice9::Reset(int w, int h, bool fs) hr = m_dev->GetBackBuffer(0, 0, D3DBACKBUFFER_TYPE_MONO, &backbuffer); } - m_backbuffer = Texture(backbuffer); + m_backbuffer = new GSTexture9(backbuffer); D3DXFONT_DESC fd; memset(&fd, 0, sizeof(fd)); @@ -348,26 +347,8 @@ bool GSDevice9::IsLost() return hr == D3DERR_DEVICELOST || hr == D3DERR_DEVICENOTRESET; } -void GSDevice9::Present(const GSVector4i& r) +void GSDevice9::Flip() { - GSVector4i cr; - - GetClientRect(m_hWnd, cr); - - if(m_backbuffer.GetWidth() != cr.width() || m_backbuffer.GetHeight() != cr.height()) - { - Reset(cr.width(), cr.height(), false); - } - - OMSetRenderTargets(m_backbuffer, NULL); - - m_dev->Clear(0, NULL, D3DCLEAR_TARGET, 0, 1.0f, 0); - - if(m_current) - { - StretchRect(m_current, m_backbuffer, GSVector4(r)); - } - if(m_swapchain) { m_swapchain->Present(NULL, NULL, NULL, NULL, 0); @@ -388,97 +369,39 @@ void GSDevice9::EndScene() m_dev->EndScene(); } -void GSDevice9::Draw(const string& s) -{ - /* - if(!m_pp.Windowed) - { - BeginScene(); - - OMSetRenderTargets(m_backbuffer, NULL); - - GSVector4i r(0, 0, m_backbuffer.GetWidth(), m_backbuffer.GetHeight()); - - D3DCOLOR c = D3DCOLOR_ARGB(255, 0, 255, 0); - - if(m_font->DrawText(NULL, str, -1, r, DT_CALCRECT|DT_LEFT|DT_WORDBREAK, c)) - { - m_font->DrawText(NULL, str, -1, r, DT_LEFT|DT_WORDBREAK, c); - } - - EndScene(); - } - */ -} - -bool GSDevice9::CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format) -{ - dst = Texture(); - - if(format == 0) - { - format = D3DFMT_A8R8G8B8; - } - - if(format != D3DFMT_A8R8G8B8) - { - ASSERT(0); - - return false; - } - - Texture rt; - - if(CreateRenderTarget(rt, w, h, format)) - { - GSVector4 dr(0, 0, w, h); - - StretchRect(src, sr, rt, dr, m_convert.ps[1], NULL, 0); - - if(CreateOffscreen(dst, w, h, format)) - { - m_dev->GetRenderTargetData(rt, dst); - } - } - - Recycle(rt); - - return !!dst; -} - -void GSDevice9::ClearRenderTarget(Texture& t, const GSVector4& c) +void GSDevice9::ClearRenderTarget(GSTexture* t, const GSVector4& c) { ClearRenderTarget(t, (c * 255 + 0.5f).zyxw().rgba32()); } -void GSDevice9::ClearRenderTarget(Texture& t, uint32 c) +void GSDevice9::ClearRenderTarget(GSTexture* t, uint32 c) { CComPtr surface; m_dev->GetRenderTarget(0, &surface); - m_dev->SetRenderTarget(0, t); + m_dev->SetRenderTarget(0, *(GSTexture9*)t); m_dev->Clear(0, NULL, D3DCLEAR_TARGET, c, 0, 0); m_dev->SetRenderTarget(0, surface); } -void GSDevice9::ClearDepth(Texture& t, float c) +void GSDevice9::ClearDepth(GSTexture* t, float c) { CComPtr surface; m_dev->GetDepthStencilSurface(&surface); - m_dev->SetDepthStencilSurface(t); + m_dev->SetDepthStencilSurface(*(GSTexture9*)t); m_dev->Clear(0, NULL, D3DCLEAR_ZBUFFER, 0, c, 0); m_dev->SetDepthStencilSurface(surface); } -void GSDevice9::ClearStencil(Texture& t, uint8 c) +void GSDevice9::ClearStencil(GSTexture* t, uint8 c) { CComPtr surface; m_dev->GetDepthStencilSurface(&surface); - m_dev->SetDepthStencilSurface(t); + m_dev->SetDepthStencilSurface(*(GSTexture9*)t); m_dev->Clear(0, NULL, D3DCLEAR_STENCIL, 0, 0, c); m_dev->SetDepthStencilSurface(surface); } -bool GSDevice9::Create(int type, Texture& t, int w, int h, int format) +GSTexture* GSDevice9::Create(int type, int w, int h, int format) { HRESULT hr; @@ -501,14 +424,16 @@ bool GSDevice9::Create(int type, Texture& t, int w, int h, int format) break; } + GSTexture9* t = NULL; + if(surface) { - t = Texture(surface); + t = new GSTexture9(surface); } if(texture) { - t = Texture(texture); + t = new GSTexture9(texture); } if(t) @@ -522,34 +447,137 @@ bool GSDevice9::Create(int type, Texture& t, int w, int h, int format) ClearDepth(t, 0); break; } - - return t; } - return false; + return t; } -bool GSDevice9::CreateRenderTarget(Texture& t, int w, int h, int format) +GSTexture* GSDevice9::CreateRenderTarget(int w, int h, int format) { - return __super::CreateRenderTarget(t, w, h, format ? format : D3DFMT_A8R8G8B8); + return __super::CreateRenderTarget(w, h, format ? format : D3DFMT_A8R8G8B8); } -bool GSDevice9::CreateDepthStencil(Texture& t, int w, int h, int format) +GSTexture* GSDevice9::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(t, w, h, format ? format : D3DFMT_D24S8); + return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D24S8); } -bool GSDevice9::CreateTexture(Texture& t, int w, int h, int format) +GSTexture* GSDevice9::CreateTexture(int w, int h, int format) { - return __super::CreateTexture(t, w, h, format ? format : D3DFMT_A8R8G8B8); + return __super::CreateTexture(w, h, format ? format : D3DFMT_A8R8G8B8); } -bool GSDevice9::CreateOffscreen(Texture& t, int w, int h, int format) +GSTexture* GSDevice9::CreateOffscreen(int w, int h, int format) { - return __super::CreateOffscreen(t, w, h, format ? format : D3DFMT_A8R8G8B8); + return __super::CreateOffscreen(w, h, format ? format : D3DFMT_A8R8G8B8); } -void GSDevice9::DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c) +GSTexture* GSDevice9::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format) +{ + GSTexture* dst = NULL; + + if(format == 0) + { + format = D3DFMT_A8R8G8B8; + } + + if(format != D3DFMT_A8R8G8B8) + { + ASSERT(0); + + return false; + } + + if(GSTexture* rt = CreateRenderTarget(w, h, format)) + { + GSVector4 dr(0, 0, w, h); + + StretchRect(src, sr, rt, dr, m_convert.ps[1], NULL, 0); + + dst = CreateOffscreen(w, h, format); + + if(dst) + { + m_dev->GetRenderTargetData(*(GSTexture9*)rt, *(GSTexture9*)dst); + } + + Recycle(rt); + } + + return dst; +} + +void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear) +{ + StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, 0, linear); +} + +void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear) +{ + StretchRect(st, sr, dt, dr, ps, ps_cb, ps_cb_len, &m_convert.bs, linear); +} + +void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear) +{ + BeginScene(); + + GSVector2i ds = dt->GetSize(); + + // om + + OMSetDepthStencilState(&m_convert.dss, 0); + OMSetBlendState(bs, 0); + OMSetRenderTargets(dt, NULL); + + // ia + + float left = dr.x * 2 / ds.x - 1.0f; + float top = 1.0f - dr.y * 2 / ds.y; + float right = dr.z * 2 / ds.x - 1.0f; + float bottom = 1.0f - dr.w * 2 / ds.y; + + GSVertexPT1 vertices[] = + { + {GSVector4(left, top, 0.5f, 1.0f), GSVector2(sr.x, sr.y)}, + {GSVector4(right, top, 0.5f, 1.0f), GSVector2(sr.z, sr.y)}, + {GSVector4(left, bottom, 0.5f, 1.0f), GSVector2(sr.x, sr.w)}, + {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, + }; + + for(int i = 0; i < countof(vertices); i++) + { + vertices[i].p.x -= 1.0f / ds.x; + vertices[i].p.y += 1.0f / ds.y; + } + + IASetVertexBuffer(4, vertices); + IASetInputLayout(m_convert.il); + IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); + + // vs + + VSSetShader(m_convert.vs, NULL, 0); + + // ps + + PSSetShader(ps, ps_cb, ps_cb_len); + PSSetSamplerState(linear ? &m_convert.ln : &m_convert.pt); + PSSetShaderResources(st, NULL); + + // rs + + RSSet(ds.x, ds.y); + + // + + DrawPrimitive(); + + // + + EndScene(); +} + +void GSDevice9::DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) { ClearRenderTarget(dt, c); @@ -568,15 +596,17 @@ void GSDevice9::DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, } } -void GSDevice9::DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset) +void GSDevice9::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) { + GSVector4 s = GSVector4(dt->GetSize()); + GSVector4 sr(0, 0, 1, 1); - GSVector4 dr(0.0f, yoffset, (float)dt.GetWidth(), (float)dt.GetHeight() + yoffset); + GSVector4 dr(0.0f, yoffset, s.x, s.y + yoffset); InterlaceConstantBuffer cb; - cb.ZrH = GSVector2(0, 1.0f / dt.GetHeight()); - cb.hH = (float)dt.GetHeight() / 2; + cb.ZrH = GSVector2(0, 1.0f / s.y); + cb.hH = (float)s.y / 2; StretchRect(st, sr, dt, dr, m_interlace.ps[shader], (const float*)&cb, 1, linear); } @@ -656,8 +686,14 @@ void GSDevice9::VSSetShader(IDirect3DVertexShader9* vs, const float* vs_cb, int } } -void GSDevice9::PSSetShaderResources(IDirect3DTexture9* srv0, IDirect3DTexture9* srv1) +void GSDevice9::PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) { + IDirect3DTexture9* srv0 = NULL; + IDirect3DTexture9* srv1 = NULL; + + if(sr0) srv0 = *(GSTexture9*)sr0; + if(sr1) srv1 = *(GSTexture9*)sr1; + if(m_ps_srvs[0] != srv0) { m_dev->SetTexture(0, srv0); @@ -796,8 +832,14 @@ void GSDevice9::OMSetBlendState(Direct3DBlendState9* bs, uint32 bf) } } -void GSDevice9::OMSetRenderTargets(IDirect3DSurface9* rtv, IDirect3DSurface9* dsv) +void GSDevice9::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) { + IDirect3DSurface9* rtv = NULL; + IDirect3DSurface9* dsv = NULL; + + if(rt) rtv = *(GSTexture9*)rt; + if(ds) dsv = *(GSTexture9*)ds; + if(m_rtv != rtv) { m_dev->SetRenderTarget(0, rtv); @@ -840,79 +882,6 @@ void GSDevice9::DrawPrimitive() m_dev->DrawPrimitiveUP(m_topology, prims, m_vb_vertices, m_vb_stride); } -void GSDevice9::StretchRect(Texture& st, Texture& dt, const GSVector4& dr, bool linear) -{ - StretchRect(st, GSVector4(0, 0, 1, 1), dt, dr, linear); -} - -void GSDevice9::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, bool linear) -{ - StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, 0, linear); -} - -void GSDevice9::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear) -{ - StretchRect(st, sr, dt, dr, ps, ps_cb, ps_cb_len, &m_convert.bs, linear); -} - -void GSDevice9::StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear) -{ - BeginScene(); - - // om - - OMSetDepthStencilState(&m_convert.dss, 0); - OMSetBlendState(bs, 0); - OMSetRenderTargets(dt, NULL); - - // ia - - float left = dr.x * 2 / dt.GetWidth() - 1.0f; - float top = 1.0f - dr.y * 2 / dt.GetHeight(); - float right = dr.z * 2 / dt.GetWidth() - 1.0f; - float bottom = 1.0f - dr.w * 2 / dt.GetHeight(); - - GSVertexPT1 vertices[] = - { - {GSVector4(left, top, 0.5f, 1.0f), GSVector2(sr.x, sr.y)}, - {GSVector4(right, top, 0.5f, 1.0f), GSVector2(sr.z, sr.y)}, - {GSVector4(left, bottom, 0.5f, 1.0f), GSVector2(sr.x, sr.w)}, - {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, - }; - - for(int i = 0; i < countof(vertices); i++) - { - vertices[i].p.x -= 1.0f / dt.GetWidth(); - vertices[i].p.y += 1.0f / dt.GetHeight(); - } - - IASetVertexBuffer(4, vertices); - IASetInputLayout(m_convert.il); - IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); - - // vs - - VSSetShader(m_convert.vs, NULL, 0); - - // ps - - PSSetShader(ps, ps_cb, ps_cb_len); - PSSetSamplerState(linear ? &m_convert.ln : &m_convert.pt); - PSSetShaderResources(st, NULL); - - // rs - - RSSet(dt.GetWidth(), dt.GetHeight()); - - // - - DrawPrimitive(); - - // - - EndScene(); -} - // FIXME: D3DXCompileShaderFromResource of d3dx9 v37 (march 2008) calls GetFullPathName on id for some reason and then crashes static HRESULT LoadShader(uint32 id, LPCSTR& data, uint32& size) diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index c299688715..23870ef97d 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -58,11 +58,8 @@ struct Direct3DBlendState9 UINT8 RenderTargetWriteMask; }; -class GSDevice9 : public GSDevice +class GSDevice9 : public GSDevice { -public: - typedef GSTexture9 Texture; - private: // state cache @@ -90,9 +87,10 @@ private: // - bool Create(int type, Texture& t, int w, int h, int format); - void DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c); - void DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset = 0); + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0); // @@ -101,7 +99,6 @@ private: CComPtr m_d3d; CComPtr m_dev; CComPtr m_swapchain; - Texture m_backbuffer; public: // TODO D3DPRESENT_PARAMETERS m_pp; @@ -135,38 +132,43 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); + bool IsLost(); - void Present(const GSVector4i& r); + void Flip(); + void BeginScene(); void EndScene(); - void Draw(const string& s); - bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0); - void ClearRenderTarget(Texture& t, const GSVector4& c); - void ClearRenderTarget(Texture& t, uint32 c); - void ClearDepth(Texture& t, float c); - void ClearStencil(Texture& t, uint8 c); + void ClearRenderTarget(GSTexture* t, const GSVector4& c); + void ClearRenderTarget(GSTexture* t, uint32 c); + void ClearDepth(GSTexture* t, float c); + void ClearStencil(GSTexture* t, uint8 c); - bool CreateRenderTarget(Texture& t, int w, int h, int format = 0); - bool CreateDepthStencil(Texture& t, int w, int h, int format = 0); - bool CreateTexture(Texture& t, int w, int h, int format = 0); - bool CreateOffscreen(Texture& t, int w, int h, int format = 0); + GSTexture* CreateRenderTarget(int w, int h, int format = 0); + GSTexture* CreateDepthStencil(int w, int h, int format = 0); + GSTexture* CreateTexture(int w, int h, int format = 0); + GSTexture* CreateOffscreen(int w, int h, int format = 0); - IDirect3DDevice9* operator->() {return m_dev;} - operator IDirect3DDevice9*() {return m_dev;} + GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); + + virtual bool IsCurrentRGBA() {return false;} + + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); // TODO: void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, const void* vertices, size_t stride); void IASetVertexBuffer(int count, const void* vertices, size_t stride); void IASetInputLayout(IDirect3DVertexDeclaration9* layout); void IASetPrimitiveTopology(D3DPRIMITIVETYPE topology); void VSSetShader(IDirect3DVertexShader9* vs, const float* vs_cb, int vs_cb_len); - void PSSetShaderResources(IDirect3DTexture9* srv0, IDirect3DTexture9* srv1); + void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1); void PSSetShader(IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len); void PSSetSamplerState(Direct3DSamplerState9* ss); void RSSet(int width, int height, const GSVector4i* scissor = NULL); void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref); void OMSetBlendState(Direct3DBlendState9* bs, uint32 bf); - void OMSetRenderTargets(IDirect3DSurface9* rtv, IDirect3DSurface9* dsv); + void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); void DrawPrimitive(); template void IASetVertexBuffer(int count, T* vertices) @@ -174,16 +176,9 @@ public: IASetVertexBuffer(count, vertices, sizeof(T)); } - void StretchRect(Texture& st, Texture& dt, const GSVector4& dr, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); - void StretchRect(Texture& st, const GSVector4& sr, Texture& dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); + IDirect3DDevice9* operator->() {return m_dev;} + operator IDirect3DDevice9*() {return m_dev;} HRESULT CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il); HRESULT CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps); - - virtual bool IsCurrentRGBA() - { - return false; - } }; diff --git a/plugins/GSdx/GSDeviceNull.cpp b/plugins/GSdx/GSDeviceNull.cpp index 4d8aa3b8b3..b9ec152c6c 100644 --- a/plugins/GSdx/GSDeviceNull.cpp +++ b/plugins/GSdx/GSDeviceNull.cpp @@ -42,9 +42,7 @@ bool GSDeviceNull::Reset(int w, int h, bool fs) return true; } -bool GSDeviceNull::Create(int type, Texture& t, int w, int h, int format) +GSTexture* GSDeviceNull::Create(int type, int w, int h, int format) { - t = Texture(type, w, h, format); - - return true; + return new GSTextureNull(type, w, h, format); } diff --git a/plugins/GSdx/GSDeviceNull.h b/plugins/GSdx/GSDeviceNull.h index bc452f7a7e..5b417529e2 100644 --- a/plugins/GSdx/GSDeviceNull.h +++ b/plugins/GSdx/GSDeviceNull.h @@ -24,30 +24,17 @@ #include "GSDevice.h" #include "GSTextureNull.h" -class GSDeviceNull : public GSDevice +class GSDeviceNull : public GSDevice { -public: - typedef GSTextureNull Texture; - private: - bool Create(int type, Texture& t, int w, int h, int format); - void DoMerge(Texture* st, GSVector4* sr, GSVector4* dr, Texture& dt, bool slbg, bool mmod, GSVector4& c) {} - void DoInterlace(Texture& st, Texture& dt, int shader, bool linear, float yoffset = 0) {} + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) {} + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0) {} public: GSDeviceNull() {} bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); - bool IsLost() {return false;} - void Present(const GSVector4i& r) {} - void BeginScene() {} - void EndScene() {} - void Draw(const string& s) {} - bool CopyOffscreen(Texture& src, const GSVector4& sr, Texture& dst, int w, int h, int format = 0) {return false;} - - void ClearRenderTarget(Texture& t, const GSVector4& c) {} - void ClearRenderTarget(Texture& t, uint32 c) {} - void ClearDepth(Texture& t, float c) {} - void ClearStencil(Texture& t, uint8 c) {} }; diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp index 4ba2fb75b0..6d61fb2b5b 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp @@ -35,11 +35,6 @@ GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(GSScanlineEnvironment& m_sel.key = key; - if(m_sel.tfx == TFX_DECAL && m_sel.tcc == 0) - { - printf("*** TFX_DECAL && !TCC *** \n"); - } - Generate(); } diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index b5f24fbb56..54346dabbe 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -1710,11 +1710,15 @@ void GSLocalMemory::ReadTexture16SZ(const GSVector4i& r, uint8* dst, int dstpitc void GSLocalMemory::ReadTextureBlock32(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock32(BlockPtr(bp), dst, dstpitch); } void GSLocalMemory::ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + if(TEXA.AEM) { ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); @@ -1745,46 +1749,50 @@ void GSLocalMemory::ReadTextureBlock16S(uint32 bp, uint8* dst, int dstpitch, con void GSLocalMemory::ReadTextureBlock8(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const uint32* pal = m_clut; + ALIGN_STACK(16); - ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, pal); + ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, m_clut); } void GSLocalMemory::ReadTextureBlock4(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const uint64* pal = m_clut; + ALIGN_STACK(16); - ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, pal); + ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, m_clut); } void GSLocalMemory::ReadTextureBlock8H(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const uint32* pal = m_clut; + ALIGN_STACK(16); - ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, pal); + ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, m_clut); } void GSLocalMemory::ReadTextureBlock4HL(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const uint32* pal = m_clut; + ALIGN_STACK(16); - ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, pal); + ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, m_clut); } void GSLocalMemory::ReadTextureBlock4HH(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { - const uint32* pal = m_clut; + ALIGN_STACK(16); - ReadAndExpandBlock4HH_32(BlockPtr(bp), dst, dstpitch, pal); + ReadAndExpandBlock4HH_32(BlockPtr(bp), dst, dstpitch, m_clut); } void GSLocalMemory::ReadTextureBlock32Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock32(BlockPtr(bp), dst, dstpitch); } void GSLocalMemory::ReadTextureBlock24Z(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + if(TEXA.AEM) { ReadAndExpandBlock24(BlockPtr(bp), dst, dstpitch, TEXA); @@ -2172,21 +2180,29 @@ void GSLocalMemory::ReadTextureBlock8P(uint32 bp, uint8* dst, int dstpitch, cons void GSLocalMemory::ReadTextureBlock4P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock4P(BlockPtr(bp), dst, dstpitch); } void GSLocalMemory::ReadTextureBlock8HP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock8HP(BlockPtr(bp), dst, dstpitch); } void GSLocalMemory::ReadTextureBlock4HLP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock4HLP(BlockPtr(bp), dst, dstpitch); } void GSLocalMemory::ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const { + ALIGN_STACK(16); + ReadBlock4HHP(BlockPtr(bp), dst, dstpitch); } diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index 741c2dbeca..7249849e03 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -277,7 +277,7 @@ void GSRasterizer::DrawTriangleTop(GSVertexSW* v, const GSVector4i& scissor) longest.p = v[2].p - v[1].p; - int i = (longest.p > GSVector4::zero()).upl(longest.p == GSVector4::zero()).mask(); + int i = longest.p.upl(longest.p == GSVector4::zero()).mask(); if(i & 2) return; @@ -305,8 +305,8 @@ void GSRasterizer::DrawTriangleTop(GSVertexSW* v, const GSVector4i& scissor) GSVertexSW dscan = longest * longest.p.xxxx().rcp(); - GSVertexSW vl = v[2 - i] - l; - GSVector4 vr = v[1 + i].p - r; + GSVertexSW vl = v[1 + i] - l; + GSVector4 vr = v[2 - i].p - r; GSVertexSW dl = vl / vl.p.yyyy(); GSVector4 dr = vr / vr.yyyy(); @@ -329,14 +329,14 @@ void GSRasterizer::DrawTriangleBottom(GSVertexSW* v, const GSVector4i& scissor) longest.p = v[1].p - v[0].p; - int i = (longest.p > GSVector4::zero()).upl(longest.p == GSVector4::zero()).mask(); + int i = longest.p.upl(longest.p == GSVector4::zero()).mask(); if(i & 2) return; i &= 1; - GSVertexSW& l = v[1 - i]; - GSVector4& r = v[i].p; + GSVertexSW& l = v[i]; + GSVector4& r = v[1 - i].p; GSVector4 fscissor(scissor); @@ -384,7 +384,7 @@ void GSRasterizer::DrawTriangleTopBottom(GSVertexSW* v, const GSVector4i& scisso GSVertexSW longest = dv[1] * (dv[0].p / dv[1].p).yyyy() - dv[0]; - int i = (longest.p > GSVector4::zero()).upl(longest.p == GSVector4::zero()).mask(); + int i = longest.p.upl(longest.p == GSVector4::zero()).mask(); if(i & 2) return; @@ -409,8 +409,8 @@ void GSRasterizer::DrawTriangleTopBottom(GSVertexSW* v, const GSVector4i& scisso GSVertexSW& l = v[0]; GSVector4 r = v[0].p; - GSVertexSW dl = dv[1 - i] / dv[1 - i].p.yyyy(); - GSVector4 dr = dv[i].p / dv[i].p.yyyy(); + GSVertexSW dl = dv[i] / dv[i].p.yyyy(); + GSVector4 dr = dv[1 - i].p / dv[1 - i].p.yyyy(); GSVector4 dy = tbmax.xxxx() - l.p.yyyy(); @@ -427,7 +427,7 @@ void GSRasterizer::DrawTriangleTopBottom(GSVertexSW* v, const GSVector4i& scisso if(top < bottom) { - if(i) + if(i == 0) { l = v[1]; dv[2] = v[2] - v[1]; diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 8d4460dc03..1fe11ac825 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -21,3 +21,421 @@ #include "StdAfx.h" #include "GSRenderer.h" + +GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) + : GSState(base, mt, irq) + , m_dev(NULL) + , m_osd(true) + , m_psrr(psrr) +{ + m_interlace = rs.m_interlace; + m_aspectratio = rs.m_aspectratio; + m_filter = rs.m_filter; + m_vsync = rs.m_vsync; + m_nativeres = rs.m_nativeres; + m_aa1 = rs.m_aa1; + m_blur = rs.m_blur; + + s_n = 0; + s_dump = !!theApp.GetConfig("dump", 0); + s_save = !!theApp.GetConfig("save", 0); + s_savez = !!theApp.GetConfig("savez", 0); +} + +GSRenderer::~GSRenderer() +{ + delete m_dev; +} + +bool GSRenderer::Create(const string& title) +{ + if(!m_wnd.Create(title.c_str())) + { + return false; + } + + ASSERT(m_dev); + + if(!m_dev->Create(m_wnd, m_vsync)) + { + return false; + } + + Reset(); + + return true; +} + +bool GSRenderer::Merge(int field) +{ + bool en[2]; + + GSVector4i fr[2]; + GSVector4i dr[2]; + + int baseline = INT_MAX; + + for(int i = 0; i < 2; i++) + { + en[i] = IsEnabled(i); + + if(en[i]) + { + fr[i] = GetFrameRect(i); + dr[i] = GetDisplayRect(i); + + baseline = min(dr[i].top, baseline); + + // printf("[%d]: %d %d %d %d, %d %d %d %d\n", i, fr[i], dr[i]); + } + } + + if(!en[0] && !en[1]) + { + return false; + } + + // try to avoid fullscreen blur, could be nice on tv but on a monitor it's like double vision, hurts my eyes (persona 4, guitar hero) + // + // NOTE: probably the technique explained in graphtip.pdf (Antialiasing by Supersampling / 4. Reading Odd/Even Scan Lines Separately with the PCRTC then Blending) + + bool samesrc = + en[0] && en[1] && + m_regs->DISP[0].DISPFB.FBP == m_regs->DISP[1].DISPFB.FBP && + m_regs->DISP[0].DISPFB.FBW == m_regs->DISP[1].DISPFB.FBW && + m_regs->DISP[0].DISPFB.PSM == m_regs->DISP[1].DISPFB.PSM; + + bool blurdetected = false; + + if(samesrc && m_regs->PMODE.SLBG == 0 && m_regs->PMODE.MMOD == 1 && m_regs->PMODE.ALP == 0x80) + { + if(fr[0].eq(fr[1] + GSVector4i(0, -1, 0, 0)) && dr[0].eq(dr[1] + GSVector4i(0, 0, 0, 1)) + || fr[1].eq(fr[0] + GSVector4i(0, -1, 0, 0)) && dr[1].eq(dr[0] + GSVector4i(0, 0, 0, 1))) + { + // persona 4: + // + // fr[0] = 0 0 640 448 + // fr[1] = 0 1 640 448 + // dr[0] = 159 50 779 498 + // dr[1] = 159 50 779 497 + // + // second image shifted up by 1 pixel and blended over itself + // + // god of war: + // + // fr[0] = 0 1 512 448 + // fr[1] = 0 0 512 448 + // dr[0] = 127 50 639 497 + // dr[1] = 127 50 639 498 + // + // same just the first image shifted + + int top = min(fr[0].top, fr[1].top); + int bottom = max(dr[0].bottom, dr[1].bottom); + + fr[0].top = top; + fr[1].top = top; + dr[0].bottom = bottom; + dr[1].bottom = bottom; + + blurdetected = true; + } + else if(dr[0].eq(dr[1]) && (fr[0].eq(fr[1] + GSVector4i(0, 1, 0, 1)) || fr[1].eq(fr[0] + GSVector4i(0, 1, 0, 1)))) + { + // dq5: + // + // fr[0] = 0 1 512 445 + // fr[1] = 0 0 512 444 + // dr[0] = 127 50 639 494 + // dr[1] = 127 50 639 494 + + int top = min(fr[0].top, fr[1].top); + int bottom = min(fr[0].bottom, fr[1].bottom); + + fr[0].top = fr[1].top = top; + fr[0].bottom = fr[1].bottom = bottom; + + blurdetected = true; + } + } + + GSVector2i fs(0, 0); + GSVector2i ds(0, 0); + + GSTexture* tex[2] = {NULL, NULL}; + + if(samesrc && fr[0].bottom == fr[1].bottom) + { + tex[0] = GetOutput(0); + tex[1] = tex[0]; // saves one texture fetch + } + else + { + if(en[0]) tex[0] = GetOutput(0); + if(en[1]) tex[1] = GetOutput(1); + } + + GSVector4 src[2]; + GSVector4 dst[2]; + + for(int i = 0; i < 2; i++) + { + if(!en[i] || !tex[i]) continue; + + GSVector4i r = fr[i]; + + // overscan hack + + if(dr[i].height() > 512) // hmm + { + int y = GetDeviceSize(i).y; + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) y /= 2; + r.bottom = r.top + y; + } + + // + + if(m_blur && blurdetected && i == 1) + { + r += GSVector4i(0, 1).xyxy(); + } + + GSVector4 scale = GSVector4(tex[i]->m_scale).xyxy(); + + src[i] = GSVector4(r) * scale / GSVector4(tex[i]->GetSize()).xyxy(); + + GSVector2 o(0, 0); + + if(dr[i].top - baseline >= 4) // 2? + { + o.y = tex[i]->m_scale.y * (dr[i].top - baseline); + + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) + { + o.y /= 2; + } + } + + dst[i] = GSVector4(o).xyxy() + scale * GSVector4(r.rsize()); + + fs.x = max(fs.x, (int)(dst[i].z + 0.5f)); + fs.y = max(fs.y, (int)(dst[i].w + 0.5f)); + } + + ds = fs; + + if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) + { + ds.y *= 2; + } + + bool slbg = m_regs->PMODE.SLBG; + bool mmod = m_regs->PMODE.MMOD; + + if(tex[0] || tex[1]) + { + GSVector4 c = GSVector4((int)m_regs->BGCOLOR.R, (int)m_regs->BGCOLOR.G, (int)m_regs->BGCOLOR.B, (int)m_regs->PMODE.ALP) / 255; + + m_dev->Merge(tex, src, dst, fs, slbg, mmod, c); + + if(m_regs->SMODE2.INT && m_interlace > 0) + { + int field2 = 1 - ((m_interlace - 1) & 1); + int mode = (m_interlace - 1) >> 1; + + if(!m_dev->Interlace(ds, field ^ field2, mode, tex[1]->m_scale.y)) + { + return false; + } + } + } + + return true; +} + + +void GSRenderer::VSync(int field) +{ + GSPerfMonAutoTimer pmat(m_perfmon); + + m_perfmon.Put(GSPerfMon::Frame); + + Flush(); + + field = field ? 1 : 0; + + if(!Merge(field)) return; + + // osd + + static uint64 s_frame = 0; + static string s_stats; + + if(m_perfmon.GetFrame() - s_frame >= 30) + { + m_perfmon.Update(); + + s_frame = m_perfmon.GetFrame(); + + double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); + + string s = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; + + GSVector4i r = GetDisplayRect(); + + s_stats = format( + "%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f", + m_perfmon.GetFrame(), r.width(), r.height(), fps, (int)(100.0 * fps / GetFPS()), + s.c_str(), + GSSettingsDlg::g_interlace[m_interlace].name, + GSSettingsDlg::g_aspectratio[m_aspectratio].name, + (int)m_perfmon.Get(GSPerfMon::Quad), + (int)m_perfmon.Get(GSPerfMon::Prim), + (int)m_perfmon.Get(GSPerfMon::Draw), + m_perfmon.CPU(), + m_perfmon.Get(GSPerfMon::Swizzle) / 1024, + m_perfmon.Get(GSPerfMon::Unswizzle) / 1024 + ); + + double fillrate = m_perfmon.Get(GSPerfMon::Fillrate); + + if(fillrate > 0) + { + s_stats += format(" | %.2f mpps", fps * fillrate / (1024 * 1024)); + } + + if(m_capture.IsCapturing()) + { + s_stats += " | Recording..."; + } + + m_wnd.SetWindowText(s_stats.c_str()); + } + + if(m_frameskip) + { + return; + } + + // present + + if(m_dev->IsLost()) + { + ResetDevice(); + } + + GSVector4i r; + + m_wnd.GetClientRect(r); + + m_dev->Present(r.fit(m_aspectratio)); + + // snapshot + + if(!m_snapshot.empty()) + { + if(!m_dump && (::GetAsyncKeyState(VK_SHIFT) & 0x8000)) + { + GSFreezeData fd; + fd.size = 0; + fd.data = NULL; + Freeze(&fd, true); + fd.data = new uint8[fd.size]; + Freeze(&fd, false); + + m_dump.Open(m_snapshot, m_crc, fd, m_regs); + + delete [] fd.data; + } + + if(GSTexture* t = m_dev->GetCurrent()) + { + t->Save(m_snapshot + ".bmp"); + } + + m_snapshot.clear(); + } + else + { + if(m_dump) + { + m_dump.VSync(field, !(::GetAsyncKeyState(VK_CONTROL) & 0x8000), m_regs); + } + } + + // capture + + if(m_capture.IsCapturing()) + { + if(GSTexture* current = m_dev->GetCurrent()) + { + GSVector2i size = m_capture.GetSize(); + + if(GSTexture* offscreen = m_dev->CopyOffscreen(current, GSVector4(0, 0, 1, 1), size.x, size.y)) + { + uint8* bits = NULL; + int pitch = 0; + + if(offscreen->Map(&bits, pitch)) + { + m_capture.DeliverFrame(bits, pitch, m_dev->IsCurrentRGBA()); + + offscreen->Unmap(); + } + + m_dev->Recycle(offscreen); + } + } + } +} + +void GSRenderer::KeyEvent(GSKeyEventData* e) +{ + if(e->type == KEYPRESS) + { + // TODO: linux + + int step = (::GetAsyncKeyState(VK_SHIFT) & 0x8000) ? -1 : 1; + + switch(e->key) + { + case VK_F5: + m_interlace = (m_interlace + 7 + step) % 7; + return; + case VK_F6: + m_aspectratio = (m_aspectratio + 3 + step) % 3; + return; + case VK_F7: + m_wnd.SetWindowText(_T("PCSX2")); + m_osd = !m_osd; + return; + case VK_F12: + if(m_capture.IsCapturing()) m_capture.EndCapture(); + else m_capture.BeginCapture(GetFPS()); + return; + case VK_DELETE: + m_aa1 = !m_aa1; + return; + case VK_END: + m_blur = !m_blur; + return; + } + } +} + +bool GSRenderer::MakeSnapshot(const string& path) +{ + if(m_snapshot.empty()) + { + time_t t = time(NULL); + + char buff[16]; + + if(strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) + { + m_snapshot = format("%s_%s", path.c_str(), buff); + } + } + + return true; +} diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 8722cf12cd..fe7947aeb5 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -39,343 +39,20 @@ struct GSRendererSettings bool m_blur; }; -class GSRendererBase : public GSState, protected GSRendererSettings +class GSRenderer : public GSState, protected GSRendererSettings { + bool Merge(int field); + protected: - bool m_osd; + virtual void ResetDevice() {} + virtual GSTexture* GetOutput(int i) = 0; public: GSWnd m_wnd; + GSDevice* m_dev; GSCapture m_capture; string m_snapshot; - -public: - GSRendererBase(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSState(base, mt, irq) - , m_osd(true) - { - m_interlace = rs.m_interlace; - m_aspectratio = rs.m_aspectratio; - m_filter = rs.m_filter; - m_vsync = rs.m_vsync; - m_nativeres = rs.m_nativeres; - m_aa1 = rs.m_aa1; - m_blur = rs.m_blur; - }; - - void KeyEvent(GSKeyEventData* e) - { - if(e->type == KEYPRESS) - { - // TODO: linux - - int step = (::GetAsyncKeyState(VK_SHIFT) & 0x8000) ? -1 : 1; - - switch(e->key) - { - case VK_F5: - m_interlace = (m_interlace + 7 + step) % 7; - return; - case VK_F6: - m_aspectratio = (m_aspectratio + 3 + step) % 3; - return; - case VK_F7: - m_wnd.SetWindowText(_T("PCSX2")); - m_osd = !m_osd; - return; - case VK_F12: - if(m_capture.IsCapturing()) m_capture.EndCapture(); - else m_capture.BeginCapture(GetFPS()); - return; - case VK_DELETE: - m_aa1 = !m_aa1; - return; - case VK_END: - m_blur = !m_blur; - return; - } - } - } - - bool MakeSnapshot(const string& path) - { - if(m_snapshot.empty()) - { - time_t t = time(NULL); - - char buff[16]; - - if(strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) - { - m_snapshot = format("%s_%s", path.c_str(), buff); - } - } - - return true; - } - - virtual bool Create(const string& title) = 0; - virtual void VSync(int field) = 0; -}; - -template class GSRenderer : public GSRendererBase -{ -protected: - typedef typename Device::Texture Texture; - - virtual void ResetDevice() {} - virtual bool GetOutput(int i, Texture& t) = 0; - - bool Merge(int field) - { - bool en[2]; - - GSVector4i fr[2]; - GSVector4i dr[2]; - - int baseline = INT_MAX; - - for(int i = 0; i < 2; i++) - { - en[i] = IsEnabled(i); - - if(en[i]) - { - fr[i] = GetFrameRect(i); - dr[i] = GetDisplayRect(i); - - baseline = min(dr[i].top, baseline); - - // printf("[%d]: %d %d %d %d, %d %d %d %d\n", i, fr[i], dr[i]); - } - } - - if(!en[0] && !en[1]) - { - return false; - } - - // try to avoid fullscreen blur, could be nice on tv but on a monitor it's like double vision, hurts my eyes (persona 4, guitar hero) - // - // NOTE: probably the technique explained in graphtip.pdf (Antialiasing by Supersampling / 4. Reading Odd/Even Scan Lines Separately with the PCRTC then Blending) - - bool samesrc = - en[0] && en[1] && - m_regs->DISP[0].DISPFB.FBP == m_regs->DISP[1].DISPFB.FBP && - m_regs->DISP[0].DISPFB.FBW == m_regs->DISP[1].DISPFB.FBW && - m_regs->DISP[0].DISPFB.PSM == m_regs->DISP[1].DISPFB.PSM; - - bool blurdetected = false; - - if(samesrc && m_regs->PMODE.SLBG == 0 && m_regs->PMODE.MMOD == 1 && m_regs->PMODE.ALP == 0x80) - { - if(fr[0].eq(fr[1] + GSVector4i(0, -1, 0, 0)) && dr[0].eq(dr[1] + GSVector4i(0, 0, 0, 1)) - || fr[1].eq(fr[0] + GSVector4i(0, -1, 0, 0)) && dr[1].eq(dr[0] + GSVector4i(0, 0, 0, 1))) - { - // persona 4: - // - // fr[0] = 0 0 640 448 - // fr[1] = 0 1 640 448 - // dr[0] = 159 50 779 498 - // dr[1] = 159 50 779 497 - // - // second image shifted up by 1 pixel and blended over itself - // - // god of war: - // - // fr[0] = 0 1 512 448 - // fr[1] = 0 0 512 448 - // dr[0] = 127 50 639 497 - // dr[1] = 127 50 639 498 - // - // same just the first image shifted - - int top = min(fr[0].top, fr[1].top); - int bottom = max(dr[0].bottom, dr[1].bottom); - - fr[0].top = top; - fr[1].top = top; - dr[0].bottom = bottom; - dr[1].bottom = bottom; - - blurdetected = true; - } - else if(dr[0].eq(dr[1]) && (fr[0].eq(fr[1] + GSVector4i(0, 1, 0, 1)) || fr[1].eq(fr[0] + GSVector4i(0, 1, 0, 1)))) - { - // dq5: - // - // fr[0] = 0 1 512 445 - // fr[1] = 0 0 512 444 - // dr[0] = 127 50 639 494 - // dr[1] = 127 50 639 494 - - int top = min(fr[0].top, fr[1].top); - int bottom = min(fr[0].bottom, fr[1].bottom); - - fr[0].top = fr[1].top = top; - fr[0].bottom = fr[1].bottom = bottom; - - blurdetected = true; - } - } - - GSVector2i fs(0, 0); - GSVector2i ds(0, 0); - - Texture tex[2]; - - if(samesrc && fr[0].bottom == fr[1].bottom) - { - GetOutput(0, tex[0]); - - tex[1] = tex[0]; // saves one texture fetch - } - else - { - if(en[0]) GetOutput(0, tex[0]); - if(en[1]) GetOutput(1, tex[1]); - } - - GSVector4 src[2]; - GSVector4 dst[2]; - - for(int i = 0; i < 2; i++) - { - if(!en[i] || !tex[i]) continue; - - GSVector4i r = fr[i]; - - // overscan hack - - if(dr[i].height() > 512) // hmm - { - int y = GetDeviceSize(i).y; - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) y /= 2; - r.bottom = r.top + y; - } - - // - - if(m_blur && blurdetected && i == 1) - { - r += GSVector4i(0, 1).xyxy(); - } - - GSVector4 scale = GSVector4(tex[i].m_scale).xyxy(); - - src[i] = GSVector4(r) * scale / GSVector4(tex[i].GetSize()).xyxy(); - - GSVector2 o(0, 0); - - if(dr[i].top - baseline >= 4) // 2? - { - o.y = tex[i].m_scale.y * (dr[i].top - baseline); - } - - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) - { - o.y /= 2; - } - - dst[i] = GSVector4(o).xyxy() + scale * GSVector4(r.rsize()); - - fs.x = max(fs.x, (int)(dst[i].z + 0.5f)); - fs.y = max(fs.y, (int)(dst[i].w + 0.5f)); - } - - ds = fs; - - if(m_regs->SMODE2.INT && m_regs->SMODE2.FFMD) ds.y *= 2; - - bool slbg = m_regs->PMODE.SLBG; - bool mmod = m_regs->PMODE.MMOD; - - if(tex[0] || tex[1]) - { - GSVector4 c = GSVector4((int)m_regs->BGCOLOR.R, (int)m_regs->BGCOLOR.G, (int)m_regs->BGCOLOR.B, (int)m_regs->PMODE.ALP) / 255; - - m_dev.Merge(tex, src, dst, fs, slbg, mmod, c); - - if(m_regs->SMODE2.INT && m_interlace > 0) - { - int field2 = 1 - ((m_interlace - 1) & 1); - int mode = (m_interlace - 1) >> 1; - - if(!m_dev.Interlace(ds, field ^ field2, mode, tex[1].m_scale.y)) - { - return false; - } - } - } - - return true; - } - - void DoSnapshot(int field) - { - if(!m_snapshot.empty()) - { - if(!m_dump && (::GetAsyncKeyState(VK_SHIFT) & 0x8000)) - { - GSFreezeData fd; - fd.size = 0; - fd.data = NULL; - Freeze(&fd, true); - fd.data = new uint8[fd.size]; - Freeze(&fd, false); - - m_dump.Open(m_snapshot, m_crc, fd, m_regs); - - delete [] fd.data; - } - - m_dev.SaveCurrent(m_snapshot + ".bmp"); - - m_snapshot.clear(); - } - else - { - if(m_dump) - { - m_dump.VSync(field, !(::GetAsyncKeyState(VK_CONTROL) & 0x8000), m_regs); - } - } - } - - void DoCapture() - { - if(!m_capture.IsCapturing()) - { - return; - } - - GSVector2i size = m_capture.GetSize(); - - Texture current; - - m_dev.GetCurrent(current); - - Texture offscreen; - - if(m_dev.CopyOffscreen(current, GSVector4(0, 0, 1, 1), offscreen, size.x, size.y)) - { - uint8* bits = NULL; - int pitch = 0; - - if(offscreen.Map(&bits, pitch)) - { - m_capture.DeliverFrame(bits, pitch, m_dev.IsCurrentRGBA()); - - offscreen.Unmap(); - } - - m_dev.Recycle(offscreen); - } - } - - -public: - Device m_dev; + bool m_osd; bool m_psrr; int s_n; @@ -384,128 +61,29 @@ public: bool s_savez; public: - GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) - : GSRendererBase(base, mt, irq, rs) - , m_psrr(psrr) + GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true); + virtual ~GSRenderer(); + + virtual bool Create(const string& title); + + virtual void VSync(int field); + + virtual void KeyEvent(GSKeyEventData* e); + + virtual bool MakeSnapshot(const string& path); + + virtual void MinMaxUV(int w, int h, GSVector4i& r) { - s_n = 0; - s_dump = !!theApp.GetConfig("dump", 0); - s_save = !!theApp.GetConfig("save", 0); - s_savez = !!theApp.GetConfig("savez", 0); + r = GSVector4i(0, 0, w, h); } - bool Create(const string& title) + virtual bool CanUpscale() { - if(!m_wnd.Create(title.c_str())) - { - return false; - } - - if(!m_dev.Create(m_wnd, m_vsync)) - { - return false; - } - - Reset(); - - return true; + return !m_nativeres; } - - void VSync(int field) - { - GSPerfMonAutoTimer pmat(m_perfmon); - - Flush(); - - m_perfmon.Put(GSPerfMon::Frame); - - field = field ? 1 : 0; - - if(!Merge(field)) return; - - // osd - - static uint64 s_frame = 0; - static string s_stats; - - if(m_perfmon.GetFrame() - s_frame >= 30) - { - m_perfmon.Update(); - - s_frame = m_perfmon.GetFrame(); - - double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); - - string s = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; - - GSVector4i r = GetDisplayRect(); - - s_stats = format( - "%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f", - m_perfmon.GetFrame(), r.width(), r.height(), fps, (int)(100.0 * fps / GetFPS()), - s.c_str(), - GSSettingsDlg::g_interlace[m_interlace].name, - GSSettingsDlg::g_aspectratio[m_aspectratio].name, - (int)m_perfmon.Get(GSPerfMon::Quad), - (int)m_perfmon.Get(GSPerfMon::Prim), - (int)m_perfmon.Get(GSPerfMon::Draw), - m_perfmon.CPU(), - m_perfmon.Get(GSPerfMon::Swizzle) / 1024, - m_perfmon.Get(GSPerfMon::Unswizzle) / 1024 - ); - - double fillrate = m_perfmon.Get(GSPerfMon::Fillrate); - - if(fillrate > 0) - { - s_stats += format(" | %.2f mpps", fps * fillrate / (1024 * 1024)); - } - - if(m_capture.IsCapturing()) - { - s_stats += " | Recording..."; - } - - m_wnd.SetWindowText(s_stats.c_str()); - } - - if(m_osd) - { - m_dev.Draw(s_stats + "\n\nF5: interlace mode\nF6: aspect ratio\nF7: OSD"); - } - - if(m_frameskip) - { - return; - } - - // - - if(m_dev.IsLost()) - { - ResetDevice(); - } - - // - - GSVector4i r; - - m_wnd.GetClientRect(r); - - m_dev.Present(r.fit(m_aspectratio)); - - // - - DoSnapshot(field); - - DoCapture(); - } - - virtual void MinMaxUV(int w, int h, GSVector4i& r) {r = GSVector4i(0, 0, w, h);} - virtual bool CanUpscale() {return !m_nativeres;} }; -template class GSRendererT : public GSRenderer +template class GSRendererT : public GSRenderer { protected: Vertex* m_vertices; @@ -644,7 +222,7 @@ protected: public: GSRendererT(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true) - : GSRenderer(base, mt, irq, rs, psrr) + : GSRenderer(base, mt, irq, rs, psrr) , m_count(0) , m_maxcount(0) , m_vertices(NULL) diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 609790327e..e5ca38975a 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -25,16 +25,17 @@ #include "GSTextureCache.h" #include "GSCrc.h" -template -class GSRendererHW : public GSRendererT +template +class GSRendererHW : public GSRendererT { - TextureCache* m_tc; int m_width; int m_height; int m_skip; bool m_reset; protected: + GSTextureCache* m_tc; + void Reset() { // TODO: GSreset can come from the main thread too => crash @@ -236,7 +237,7 @@ protected: m_tc->RemoveAll(); } - bool GetOutput(int i, Texture& t) + GSTexture* GetOutput(int i) { const GSRegDISPFB& DISPFB = m_regs->DISP[i].DISPFB; @@ -248,7 +249,9 @@ protected: TRACE(_T("[%d] GetOutput %d %05x (%d)\n"), (int)m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM); - if(GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height, true)) + GSTexture* t = NULL; + + if(GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height, true)) { t = rt->m_texture; @@ -256,16 +259,14 @@ protected: { if(s_save) { - rt->m_texture.Save(format("c:\\temp2\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); + t->Save(format("c:\\temp2\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); } s_n++; } - - return true; } - return false; + return t; } void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) @@ -298,15 +299,15 @@ protected: TEX0.TBW = context->FRAME.FBW; TEX0.PSM = context->FRAME.PSM; - GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height); + GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height); TEX0.TBP0 = context->ZBUF.Block(); TEX0.TBW = context->FRAME.FBW; TEX0.PSM = context->ZBUF.PSM; - GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height); + GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height); - GSTextureCache::GSTexture* tex = NULL; + GSTextureCache::GSCachedTexture* tex = NULL; if(PRIM->TME) { @@ -329,13 +330,13 @@ protected: (int)context->CLAMP.MINU, (int)context->CLAMP.MAXU, (int)context->CLAMP.MINV, (int)context->CLAMP.MAXV); - tex->m_texture.Save(s, true); + tex->m_texture->Save(s, true); if(tex->m_palette) { s = format("c:\\temp2\\_%05d_f%I64d_tpx_%05x_%d.dds", s_n, frame, context->TEX0.CBP, context->TEX0.CPSM); - tex->m_palette.Save(s, true); + tex->m_palette->Save(s, true); } } @@ -345,14 +346,14 @@ protected: { s = format("c:\\temp2\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, context->FRAME.Block(), context->FRAME.PSM); - rt->m_texture.Save(s); + rt->m_texture->Save(s); } if(s_savez) { s = format("c:\\temp2\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, context->ZBUF.Block(), context->ZBUF.PSM); - ds->m_texture.Save(s); + ds->m_texture->Save(s); } s_n++; @@ -360,7 +361,7 @@ protected: int prim = PRIM->PRIM; - if(!OverrideInput(prim, rt->m_texture, ds->m_texture, tex ? &tex->m_texture : NULL)) + if(!OverrideInput(prim, rt->m_texture, ds->m_texture, tex ? tex->m_texture : NULL)) { return; } @@ -379,14 +380,14 @@ protected: { s = format("c:\\temp2\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, context->FRAME.Block(), context->FRAME.PSM); - rt->m_texture.Save(s); + rt->m_texture->Save(s); } if(s_savez) { s = format("c:\\temp2\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, context->ZBUF.Block(), context->ZBUF.PSM); - ds->m_texture.Save(s); + ds->m_texture->Save(s); } s_n++; @@ -395,9 +396,9 @@ protected: m_tc->InvalidateTextures(context->FRAME, context->ZBUF); } - virtual void Draw(int prim, Texture& rt, Texture& ds, typename GSTextureCache::GSTexture* tex) = 0; + virtual void Draw(int prim, GSTexture* rt, GSTexture* ds, typename GSTextureCache::GSCachedTexture* tex) = 0; - virtual bool OverrideInput(int& prim, Texture& rt, Texture& ds, Texture* t) + virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTexture* t) { #pragma region ffxii pal video conversion @@ -436,7 +437,7 @@ protected: ok = false; - m_dev.CreateTexture(*t, 512, 512); + t = m_dev->CreateTexture(512, 512); t->Update(GSVector4i(0, 0, 448, 512), video, 512 * 4); @@ -466,7 +467,7 @@ protected: if((FBP == 0x00d00 || FBP == 0x00000) && ZBP == 0x02100 && PRIM->TME && TBP == 0x01a00 && m_context->TEX0.PSM == PSM_PSMCT16S) { - m_dev.ClearDepth(ds, 0); + m_dev->ClearDepth(ds, 0); } return true; @@ -556,9 +557,9 @@ protected: TEX0.TBW = FBW; TEX0.PSM = FPSM; - if(GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height)) + if(GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height)) { - m_dev.ClearDepth(ds->m_texture, 0); + m_dev->ClearDepth(ds->m_texture, 0); } return false; @@ -665,7 +666,8 @@ protected: public: GSRendererHW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) - : GSRendererT(base, mt, irq, rs, psrr) + : GSRendererT(base, mt, irq, rs, psrr) + , m_tc(NULL) , m_width(1024) , m_height(1024) , m_skip(0) @@ -676,8 +678,6 @@ public: m_width = theApp.GetConfig("resx", m_width); m_height = theApp.GetConfig("resy", m_height); } - - m_tc = new TextureCache(this); } virtual ~GSRendererHW() diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index f334118a9c..29a0f0c1c3 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -25,8 +25,11 @@ #include "resource.h" GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSRendererHW(base, mt, irq, rs, true) + : GSRendererHW(base, mt, irq, rs, true) { + m_dev = new GSDevice10(); + m_tc = new GSTextureCache10(this); + InitVertexKick(); } @@ -35,7 +38,7 @@ bool GSRendererHW10::Create(const string& title) if(!__super::Create(title)) return false; - if(!m_tfx.Create(&m_dev)) + if(!m_tfx.Create((GSDevice10*)m_dev)) return false; // @@ -57,13 +60,13 @@ bool GSRendererHW10::Create(const string& title) dsd.BackFace.StencilFailOp = D3D10_STENCIL_OP_KEEP; dsd.BackFace.StencilDepthFailOp = D3D10_STENCIL_OP_KEEP; - m_dev->CreateDepthStencilState(&dsd, &m_date.dss); + (*(GSDevice10*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); D3D10_BLEND_DESC bd; memset(&bd, 0, sizeof(bd)); - m_dev->CreateBlendState(&bd, &m_date.bs); + (*(GSDevice10*)m_dev)->CreateBlendState(&bd, &m_date.bs); // @@ -73,7 +76,7 @@ bool GSRendererHW10::Create(const string& title) template void GSRendererHW10::VertexKick(bool skip) { - Vertex& dst = m_vl.AddTail(); + GSVertexHW10& dst = m_vl.AddTail(); dst.m128i[0] = m_v.m128i[0]; dst.m128i[1] = m_v.m128i[1]; @@ -85,7 +88,7 @@ void GSRendererHW10::VertexKick(bool skip) int count = 0; - if(Vertex* v = DrawingKick(skip, count)) + if(GSVertexHW10* v = DrawingKick(skip, count)) { GSVector4i scissor = m_context->scissor.dx10; @@ -169,7 +172,7 @@ void GSRendererHW10::VertexKick(bool skip) } } -void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCache::GSTexture* tex) +void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; @@ -223,7 +226,7 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheBeginScene(); // om @@ -285,8 +288,8 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachem_scale.x / (rt->GetWidth() * 16); + float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); float ox = (float)(int)context->XYOFFSET.OFX; float oy = (float)(int)context->XYOFFSET.OFY; @@ -396,8 +399,8 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachem_texture.GetWidth(); - float h = (float)tex->m_texture.GetHeight(); + float w = (float)tex->m_texture->GetWidth(); + float h = (float)tex->m_texture->GetHeight(); ps_cb.WH = GSVector2(w, h); ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); @@ -409,10 +412,10 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheGetWidth(); + int h = rt->GetHeight(); - GSVector4i scissor = GSVector4i(GSVector4(rt.m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); + GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); // @@ -420,9 +423,7 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachem_texture : NULL, - tex ? (ID3D10ShaderResourceView*)tex->m_palette : NULL); + m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); m_tfx.SetupRS(w, h, scissor); // draw @@ -471,7 +472,7 @@ void GSRendererHW10::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheEndScene(); } bool GSRendererHW10::WrapZ(uint32 maxz) @@ -489,112 +490,114 @@ bool GSRendererHW10::WrapZ(uint32 maxz) return true; } -void GSRendererHW10::SetupDATE(Texture& rt, Texture& ds) +void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) { if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) - // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + GSDevice10* dev = (GSDevice10*)m_dev; - GSVector4 mm; + int w = rt->GetWidth(); + int h = rt->GetHeight(); - // TODO - - mm = GSVector4(-1, -1, 1, 1); - - // if(m_count < 100) + if(GSTexture* t = dev->CreateRenderTarget(w, h)) { - GSVector4 pmin(65535, 65535, 0, 0); - GSVector4 pmax = GSVector4::zero(); + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - for(int i = 0, j = m_count; i < j; i++) + GSVector4 mm; + + // TODO + + mm = GSVector4(-1, -1, 1, 1); + + // if(m_count < 100) { - GSVector4 p(GSVector4i(m_vertices[i].m128i[0]).uph16()); + GSVector4 pmin(65535, 65535, 0, 0); + GSVector4 pmax = GSVector4::zero(); - pmin = p.minv(pmin); - pmax = p.maxv(pmax); + for(int i = 0, j = m_count; i < j; i++) + { + GSVector4 p(GSVector4i(m_vertices[i].m128i[0]).uph16()); + + pmin = p.minv(pmin); + pmax = p.maxv(pmax); + } + + mm += pmin.xyxy(pmax); + + float sx = 2.0f * rt->m_scale.x / (w * 16); + float sy = 2.0f * rt->m_scale.y / (h * 16); + float ox = (float)(int)m_context->XYOFFSET.OFX; + float oy = (float)(int)m_context->XYOFFSET.OFY; + + mm.x = (mm.x - ox) * sx - 1; + mm.y = (mm.y - oy) * sy - 1; + mm.z = (mm.z - ox) * sx - 1; + mm.w = (mm.w - oy) * sy - 1; + + if(mm.x < -1) mm.x = -1; + if(mm.y < -1) mm.y = -1; + if(mm.z > +1) mm.z = +1; + if(mm.w > +1) mm.w = +1; } - mm += pmin.xyxy(pmax); + GSVector4 uv = (mm + 1.0f) / 2.0f; - int w = rt.GetWidth(); - int h = rt.GetHeight(); + // - float sx = 2.0f * rt.m_scale.x / (w * 16); - float sy = 2.0f * rt.m_scale.y / (h * 16); - float ox = (float)(int)m_context->XYOFFSET.OFX; - float oy = (float)(int)m_context->XYOFFSET.OFY; + dev->BeginScene(); - mm.x = (mm.x - ox) * sx - 1; - mm.y = (mm.y - oy) * sy - 1; - mm.z = (mm.z - ox) * sx - 1; - mm.w = (mm.w - oy) * sy - 1; + dev->ClearStencil(ds, 0); - if(mm.x < -1) mm.x = -1; - if(mm.y < -1) mm.y = -1; - if(mm.z > +1) mm.z = +1; - if(mm.w > +1) mm.w = +1; + // om + + dev->OMSetDepthStencilState(m_date.dss, 1); + dev->OMSetBlendState(m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVertexPT1 vertices[] = + { + {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, + {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, + {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, + {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + }; + + D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; + + (*dev)->UpdateSubresource(dev->m_convert.vb, 0, &box, vertices, 0, 0); + + dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL); + + // gs + + dev->GSSetShader(NULL); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); + dev->PSSetSamplerState(dev->m_convert.pt, NULL); + + // rs + + dev->RSSet(w, h); + + // set + + dev->DrawPrimitive(countof(vertices)); + + // + + dev->EndScene(); + + dev->Recycle(t); } - - GSVector4 uv = (mm + 1.0f) / 2.0f; - - // - - m_dev.BeginScene(); - - // om - - GSTexture10 tmp; - - m_dev.CreateRenderTarget(tmp, rt.GetWidth(), rt.GetHeight()); - - m_dev.OMSetRenderTargets(tmp, ds); - m_dev.OMSetDepthStencilState(m_date.dss, 1); - m_dev.OMSetBlendState(m_date.bs, 0); - - m_dev->ClearDepthStencilView(ds, D3D10_CLEAR_STENCIL, 0, 0); - - // ia - - GSVertexPT1 vertices[] = - { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, - }; - - D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; - m_dev->UpdateSubresource(m_dev.m_convert.vb, 0, &box, vertices, 0, 0); - - m_dev.IASetVertexBuffer(m_dev.m_convert.vb, sizeof(vertices[0])); - m_dev.IASetInputLayout(m_dev.m_convert.il); - m_dev.IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); - - // vs - - m_dev.VSSetShader(m_dev.m_convert.vs, NULL); - - // gs - - m_dev.GSSetShader(NULL); - - // ps - - m_dev.PSSetShaderResources(rt, NULL); - m_dev.PSSetShader(m_dev.m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); - m_dev.PSSetSamplerState(m_dev.m_convert.pt, NULL); - - // rs - - m_dev.RSSet(tmp.GetWidth(), tmp.GetHeight()); - - // set - - m_dev.DrawPrimitive(countof(vertices)); - - // - - m_dev.EndScene(); - - m_dev.Recycle(tmp); } diff --git a/plugins/GSdx/GSRendererHW10.h b/plugins/GSdx/GSRendererHW10.h index 46a7c30a15..ef6c3548bd 100644 --- a/plugins/GSdx/GSRendererHW10.h +++ b/plugins/GSdx/GSRendererHW10.h @@ -26,18 +26,14 @@ #include "GSTextureCache10.h" #include "GSTextureFX10.h" -class GSRendererHW10 : public GSRendererHW +class GSRendererHW10 : public GSRendererHW { - typedef GSDevice10 Device; - typedef GSVertexHW10 Vertex; - typedef GSTextureCache10 TextureCache; - bool WrapZ(uint32 maxz); protected: GSTextureFX10 m_tfx; - void Draw(int prim, Texture& rt, Texture& ds, GSTextureCache::GSTexture* tex); + void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); struct { @@ -45,7 +41,7 @@ protected: CComPtr bs; } m_date; - void SetupDATE(Texture& rt, Texture& ds); + void SetupDATE(GSTexture* rt, GSTexture* ds); public: GSRendererHW10(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs); diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index a5ee0df4fd..1f3dc149d4 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -25,11 +25,14 @@ #include "resource.h" GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSRendererHW(base, mt, irq, rs, false) + : GSRendererHW(base, mt, irq, rs, false) { m_fba.enabled = !!theApp.GetConfig("fba", 1); m_logz = !!theApp.GetConfig("logz", 0); + m_dev = new GSDevice9(); + m_tc = new GSTextureCache9(this); + InitVertexKick(); } @@ -38,7 +41,7 @@ bool GSRendererHW9::Create(const string& title) if(!__super::Create(title)) return false; - if(!m_tfx.Create(&m_dev)) + if(!m_tfx.Create((GSDevice9*)m_dev)) return false; // @@ -77,7 +80,7 @@ bool GSRendererHW9::Create(const string& title) template void GSRendererHW9::VertexKick(bool skip) { - Vertex& dst = m_vl.AddTail(); + GSVertexHW9& dst = m_vl.AddTail(); dst.p.x = (float)(int)m_v.XYZ.X; dst.p.y = (float)(int)m_v.XYZ.Y; @@ -102,7 +105,7 @@ void GSRendererHW9::VertexKick(bool skip) int count = 0; - if(Vertex* v = DrawingKick(skip, count)) + if(GSVertexHW9* v = DrawingKick(skip, count)) { GSVector4 scissor = m_context->scissor.dx9; @@ -169,7 +172,7 @@ void GSRendererHW9::VertexKick(bool skip) } } -void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCache::GSTexture* tex) +void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; @@ -208,9 +211,9 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheBeginScene(); - m_dev->SetRenderState(D3DRS_SHADEMODE, PRIM->IIP ? D3DSHADE_GOURAUD : D3DSHADE_FLAT); // TODO + (*(GSDevice9*)m_dev)->SetRenderState(D3DRS_SHADEMODE, PRIM->IIP ? D3DSHADE_GOURAUD : D3DSHADE_FLAT); // TODO // om @@ -267,8 +270,8 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachem_scale.x / (rt->GetWidth() * 16); + float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); float ox = (float)(int)context->XYOFFSET.OFX; float oy = (float)(int)context->XYOFFSET.OFY; @@ -370,8 +373,8 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCachem_texture.GetWidth(); - float h = (float)tex->m_texture.GetHeight(); + float w = (float)tex->m_texture->GetWidth(); + float h = (float)tex->m_texture->GetHeight(); ps_cb.WH = GSVector2(w, h); ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); @@ -383,27 +386,24 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheGetWidth(); + int h = rt->GetHeight(); - GSVector4i scissor = GSVector4i(GSVector4(rt.m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); + GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); // m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); m_tfx.SetupIA(m_vertices, m_count, topology); m_tfx.SetupVS(vs_sel, &vs_cb); - m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, - tex ? (IDirect3DTexture9*)tex->m_texture : NULL, - tex ? (IDirect3DTexture9*)tex->m_palette : NULL, - m_psrr); + m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL, m_psrr); m_tfx.SetupRS(w, h, scissor); // draw if(context->TEST.DoFirstPass()) { - m_dev.DrawPrimitive(); + ((GSDevice9*)m_dev)->DrawPrimitive(); } if(context->TEST.DoSecondPass()) @@ -441,11 +441,11 @@ void GSRendererHW9::Draw(int prim, Texture& rt, Texture& ds, GSTextureCacheDrawPrimitive(); } } - m_dev.EndScene(); + m_dev->EndScene(); if(om_dssel.fba) UpdateFBA(rt); } @@ -465,130 +465,133 @@ bool GSRendererHW9::WrapZ(float maxz) return true; } -void GSRendererHW9::SetupDATE(Texture& rt, Texture& ds) +void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) { if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) - // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + GSDevice9* dev = (GSDevice9*)m_dev; - GSVector4 mm; + int w = rt->GetWidth(); + int h = rt->GetHeight(); - // TODO - - mm = GSVector4(-1, -1, 1, 1); - - // if(m_count < 100) + if(GSTexture* t = dev->CreateRenderTarget(w, h)) { - GSVector4 pmin(65535, 65535, 0, 0); - GSVector4 pmax = GSVector4::zero(); + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - for(int i = 0, j = m_count; i < j; i++) + GSVector4 mm; + + // TODO + + mm = GSVector4(-1, -1, 1, 1); + + // if(m_count < 100) { - GSVector4 p(m_vertices[i].p); + GSVector4 pmin(65535, 65535, 0, 0); + GSVector4 pmax = GSVector4::zero(); - pmin = p.minv(pmin); - pmax = p.maxv(pmax); + for(int i = 0, j = m_count; i < j; i++) + { + GSVector4 p(m_vertices[i].p); + + pmin = p.minv(pmin); + pmax = p.maxv(pmax); + } + + mm += pmin.xyxy(pmax); + + float sx = 2.0f * rt->m_scale.x / (w * 16); + float sy = 2.0f * rt->m_scale.y / (h * 16); + float ox = (float)(int)m_context->XYOFFSET.OFX; + float oy = (float)(int)m_context->XYOFFSET.OFY; + + mm.x = (mm.x - ox) * sx - 1; + mm.y = (mm.y - oy) * sy - 1; + mm.z = (mm.z - ox) * sx - 1; + mm.w = (mm.w - oy) * sy - 1; + + if(mm.x < -1) mm.x = -1; + if(mm.y < -1) mm.y = -1; + if(mm.z > +1) mm.z = +1; + if(mm.w > +1) mm.w = +1; } - mm += pmin.xyxy(pmax); + GSVector4 uv = (mm + 1.0f) / 2.0f; - int w = rt.GetWidth(); - int h = rt.GetHeight(); + // - float sx = 2.0f * rt.m_scale.x / (w * 16); - float sy = 2.0f * rt.m_scale.y / (h * 16); - float ox = (float)(int)m_context->XYOFFSET.OFX; - float oy = (float)(int)m_context->XYOFFSET.OFY; + dev->BeginScene(); - mm.x = (mm.x - ox) * sx - 1; - mm.y = (mm.y - oy) * sy - 1; - mm.z = (mm.z - ox) * sx - 1; - mm.w = (mm.w - oy) * sy - 1; + dev->ClearStencil(ds, 0); - if(mm.x < -1) mm.x = -1; - if(mm.y < -1) mm.y = -1; - if(mm.z > +1) mm.z = +1; - if(mm.w > +1) mm.w = +1; + // om + + dev->OMSetDepthStencilState(&m_date.dss, 1); + dev->OMSetBlendState(&m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVertexPT1 vertices[] = + { + {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, + {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, + {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, + {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + }; + + dev->IASetVertexBuffer(4, vertices); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL, 0); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL, 0); + dev->PSSetSamplerState(&dev->m_convert.pt); + + // rs + + dev->RSSet(w, h); + + // + + dev->DrawPrimitive(); + + // + + dev->EndScene(); + + dev->Recycle(t); } - - GSVector4 uv = (mm + 1.0f) / 2.0f; - - // - - m_dev.BeginScene(); - - // om - - GSTexture9 tmp; - - m_dev.CreateRenderTarget(tmp, rt.GetWidth(), rt.GetHeight()); - - m_dev.OMSetRenderTargets(tmp, ds); - m_dev.OMSetDepthStencilState(&m_date.dss, 1); - m_dev.OMSetBlendState(&m_date.bs, 0); - - m_dev->Clear(0, NULL, D3DCLEAR_STENCIL, 0, 0, 0); - - // ia - - GSVertexPT1 vertices[] = - { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, - }; - - m_dev.IASetVertexBuffer(4, vertices); - m_dev.IASetInputLayout(m_dev.m_convert.il); - m_dev.IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); - - // vs - - m_dev.VSSetShader(m_dev.m_convert.vs, NULL, 0); - - // ps - - m_dev.PSSetShaderResources(rt, NULL); - m_dev.PSSetShader(m_dev.m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL, 0); - m_dev.PSSetSamplerState(&m_dev.m_convert.pt); - - // rs - - m_dev.RSSet(tmp.GetWidth(), tmp.GetHeight()); - - // - - m_dev.DrawPrimitive(); - - // - - m_dev.EndScene(); - - m_dev.Recycle(tmp); } -void GSRendererHW9::UpdateFBA(Texture& rt) +void GSRendererHW9::UpdateFBA(GSTexture* rt) { - m_dev.BeginScene(); + GSDevice9* dev = (GSDevice9*)m_dev; + + dev->BeginScene(); // om - m_dev.OMSetDepthStencilState(&m_fba.dss, 2); - m_dev.OMSetBlendState(&m_fba.bs, 0); + dev->OMSetDepthStencilState(&m_fba.dss, 2); + dev->OMSetBlendState(&m_fba.bs, 0); // vs - m_dev.VSSetShader(NULL, NULL, 0); + dev->VSSetShader(NULL, NULL, 0); // ps - m_dev.PSSetShader(m_dev.m_convert.ps[4], NULL, 0); + dev->PSSetShader(dev->m_convert.ps[4], NULL, 0); // - int w = rt.GetWidth(); - int h = rt.GetHeight(); + int w = rt->GetWidth(); + int h = rt->GetHeight(); GSVertexP vertices[] = { @@ -598,11 +601,11 @@ void GSRendererHW9::UpdateFBA(Texture& rt) {GSVector4(w, h, 0, 0)}, }; - m_dev->SetFVF(D3DFVF_XYZRHW); + (*dev)->SetFVF(D3DFVF_XYZRHW); - m_dev->DrawPrimitiveUP(D3DPT_TRIANGLESTRIP, 2, vertices, sizeof(vertices[0])); + (*dev)->DrawPrimitiveUP(D3DPT_TRIANGLESTRIP, 2, vertices, sizeof(vertices[0])); // - m_dev.EndScene(); + dev->EndScene(); } diff --git a/plugins/GSdx/GSRendererHW9.h b/plugins/GSdx/GSRendererHW9.h index dd08dd467b..e6235b45f4 100644 --- a/plugins/GSdx/GSRendererHW9.h +++ b/plugins/GSdx/GSRendererHW9.h @@ -26,19 +26,15 @@ #include "GSTextureCache9.h" #include "GSTextureFX9.h" -class GSRendererHW9 : public GSRendererHW +class GSRendererHW9 : public GSRendererHW { - typedef GSDevice9 Device; - typedef GSVertexHW9 Vertex; - typedef GSTextureCache9 TextureCache; - bool WrapZ(float maxz); protected: GSTextureFX9 m_tfx; bool m_logz; - void Draw(int prim, Texture& rt, Texture& ds, GSTextureCache::GSTexture* tex); + void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); struct { @@ -53,8 +49,8 @@ protected: Direct3DBlendState9 bs; } m_fba; - void SetupDATE(Texture& rt, Texture& ds); - void UpdateFBA(Texture& rt); + void SetupDATE(GSTexture* rt, GSTexture* ds); + void UpdateFBA(GSTexture* rt); public: GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs); diff --git a/plugins/GSdx/GSRendererNull.h b/plugins/GSdx/GSRendererNull.h index a12986c203..d23738ee17 100644 --- a/plugins/GSdx/GSRendererNull.h +++ b/plugins/GSdx/GSRendererNull.h @@ -24,23 +24,25 @@ #include "GSRenderer.h" #include "GSDeviceNull.h" -template class GSRendererNull : public GSRendererT +class GSRendererNull : public GSRendererT { protected: void Draw() { } - bool GetOutput(int i, Texture& t) + GSTexture* GetOutput(int i) { - return false; + return NULL; } public: - GSRendererNull(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSRendererT(base, mt, irq, rs) + GSRendererNull(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev) + : GSRendererT(base, mt, irq, rs) { - InitVertexKick >(); + m_dev = dev; + + InitVertexKick(); } template void VertexKick(bool skip) diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index b422422e62..8433888d90 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -23,3 +23,775 @@ #include "GSRendererSW.h" const GSVector4 g_pos_scale(1.0f / 16, 1.0f / 16, 1.0f, 128.0f); + +GSRendererSW::GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev, int threads) + : GSRendererT(base, mt, irq, rs) +{ + m_dev = dev; + + m_tc = new GSTextureCacheSW(this); + + memset(m_texture, 0, sizeof(m_texture)); + + m_rl.Create(this, threads); + + InitVertexKick(); +} + +GSRendererSW::~GSRendererSW() +{ + delete m_tc; +} + +void GSRendererSW::Reset() +{ + // TODO: GSreset can come from the main thread too => crash + // m_tc->RemoveAll(); + + m_reset = true; + + __super::Reset(); +} + +void GSRendererSW::VSync(int field) +{ + __super::VSync(field); + + m_tc->IncAge(); + + if(m_reset) + { + m_tc->RemoveAll(); + + m_reset = false; + } + + // if((m_perfmon.GetFrame() & 255) == 0) m_rl.PrintStats(); +} + +void GSRendererSW::ResetDevice() +{ + for(int i = 0; i < countof(m_texture); i++) + { + delete m_texture[i]; + + m_texture[i] = NULL; + } +} + +GSTexture* GSRendererSW::GetOutput(int i) +{ + const GSRegDISPFB& DISPFB = m_regs->DISP[i].DISPFB; + + GIFRegTEX0 TEX0; + + TEX0.TBP0 = DISPFB.Block(); + TEX0.TBW = DISPFB.FBW; + TEX0.PSM = DISPFB.PSM; + + GSVector4i r(0, 0, TEX0.TBW * 64, GetFrameRect(i).bottom); + + // TODO: round up bottom + + int w = r.width(); + int h = r.height(); + + if(m_texture[i]) + { + if(m_texture[i]->GetWidth() != w || m_texture[i]->GetHeight() != h) + { + delete m_texture[i]; + + m_texture[i] = NULL; + } + } + + if(!m_texture[i]) + { + m_texture[i] = m_dev->CreateTexture(w, h); + + if(!m_texture[i]) + { + return NULL; + } + } + + GIFRegCLAMP CLAMP; + + CLAMP.WMS = CLAMP.WMT = 1; + + // TODO + static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); + static int pitch = 1024 * 4; + + m_mem.ReadTexture(r, buff, pitch, TEX0, m_env.TEXA, CLAMP); + + m_texture[i]->Update(r, buff, pitch); + + if(s_dump) + { + if(s_save) + { + m_texture[i]->Save(format("c:\\temp1\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); + } + + s_n++; + } + + return m_texture[i]; +} + +void GSRendererSW::Draw() +{ + GS_PRIM_CLASS primclass = GSUtil::GetPrimClass(PRIM->PRIM); + + m_vtrace.Update(m_vertices, m_count, primclass, PRIM->IIP, PRIM->TME, m_context->TEX0.TFX, m_context->TEX0.TCC); + + if(m_dump) + { + m_dump.Object(m_vertices, m_count, primclass); + } + + GSScanlineParam p; + + GetScanlineParam(p, primclass); + + if((p.fm & p.zm) == 0xffffffff) + { + return; + } + + if(s_dump) + { + uint64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save && PRIM->TME) + { + s = format("c:\\temp1\\_%05d_f%I64d_tex_%05x_%d.bmp", s_n, frame, (int)m_context->TEX0.TBP0, (int)m_context->TEX0.PSM); + + m_mem.SaveBMP(s, m_context->TEX0.TBP0, m_context->TEX0.TBW, m_context->TEX0.PSM, 1 << m_context->TEX0.TW, 1 << m_context->TEX0.TH); + } + + s_n++; + + if(s_save) + { + s = format("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); + + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); + } + + if(s_savez) + { + s = format("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); + + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); + } + + s_n++; + } + + GSRasterizerData data; + + data.scissor = GSVector4i(m_context->scissor.in); + data.scissor.z = min(data.scissor.z, (int)m_context->FRAME.FBW * 64); // TODO: find a game that overflows and check which one is the right behaviour + data.primclass = primclass; + data.vertices = m_vertices; + data.count = m_count; + data.param = &p; + + m_rl.Draw(&data); + + GSRasterizerStats stats; + + m_rl.GetStats(stats); + + m_perfmon.Put(GSPerfMon::Draw, 1); + m_perfmon.Put(GSPerfMon::Prim, stats.prims); + m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); + + GSVector4i r = GSVector4i(m_vtrace.m_min.p.xyxy(m_vtrace.m_max.p)).rintersect(data.scissor); + + GIFRegBITBLTBUF BITBLTBUF; + + BITBLTBUF.DBW = m_context->FRAME.FBW; + + if(p.fm != 0xffffffff) + { + BITBLTBUF.DBP = m_context->FRAME.Block(); + BITBLTBUF.DPSM = m_context->FRAME.PSM; + + m_tc->InvalidateVideoMem(BITBLTBUF, r); + } + + if(p.zm != 0xffffffff) + { + BITBLTBUF.DBP = m_context->ZBUF.Block(); + BITBLTBUF.DPSM = m_context->ZBUF.PSM; + + m_tc->InvalidateVideoMem(BITBLTBUF, r); + } + + if(s_dump) + { + uint64 frame = m_perfmon.GetFrame(); + + string s; + + if(s_save) + { + s = format("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); + + m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); + } + + if(s_savez) + { + s = format("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); + + m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); + } + + s_n++; + } + + if(0)//stats.ticks > 5000000) + { + printf("* [%I64d | %012I64x] ticks %I64d prims %d (%d) pixels %d (%d)\n", + m_perfmon.GetFrame(), p.sel.key, + stats.ticks, + stats.prims, stats.prims > 0 ? (int)(stats.ticks / stats.prims) : -1, + stats.pixels, stats.pixels > 0 ? (int)(stats.ticks / stats.pixels) : -1); + } +} + +void GSRendererSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) +{ + m_tc->InvalidateVideoMem(BITBLTBUF, r); +} + +void GSRendererSW::GetTextureMinMax(int w, int h, GSVector4i& r, uint32 fst) +{ + const GSDrawingContext* context = m_context; + + int wms = context->CLAMP.WMS; + int wmt = context->CLAMP.WMT; + + int minu = (int)context->CLAMP.MINU; + int minv = (int)context->CLAMP.MINV; + int maxu = (int)context->CLAMP.MAXU; + int maxv = (int)context->CLAMP.MAXV; + + GSVector4i vr(0, 0, w, h); + + switch(wms) + { + case CLAMP_REPEAT: + break; + case CLAMP_CLAMP: + break; + case CLAMP_REGION_CLAMP: + if(vr.x < minu) vr.x = minu; + if(vr.z > maxu + 1) vr.z = maxu + 1; + break; + case CLAMP_REGION_REPEAT: + vr.x = maxu; + vr.z = vr.x + (minu + 1); + break; + default: + __assume(0); + } + + switch(wmt) + { + case CLAMP_REPEAT: + break; + case CLAMP_CLAMP: + break; + case CLAMP_REGION_CLAMP: + if(vr.y < minv) vr.y = minv; + if(vr.w > maxv + 1) vr.w = maxv + 1; + break; + case CLAMP_REGION_REPEAT: + vr.y = maxv; + vr.w = vr.y + (minv + 1); + break; + default: + __assume(0); + } + + if(fst) + { + GSVector4i uv = GSVector4i(m_vtrace.m_min.t.xyxy(m_vtrace.m_max.t)).sra32(16); + + GSVector4i u, v; + + int mask = 0; + + if(wms == CLAMP_REPEAT || wmt == CLAMP_REPEAT) + { + int tw = context->TEX0.TW; + int th = context->TEX0.TH; + + u = uv & GSVector4i::xffffffff().srl32(32 - tw); + v = uv & GSVector4i::xffffffff().srl32(32 - th); + + GSVector4i uu = uv.sra32(tw); + GSVector4i vv = uv.sra32(th); + + mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); + } + + switch(wms) + { + case CLAMP_REPEAT: + if(mask & 0x000f) {if(vr.x < u.x) vr.x = u.x; if(vr.z > u.z + 1) vr.z = u.z + 1;} + break; + case CLAMP_CLAMP: + case CLAMP_REGION_CLAMP: + if(vr.x < uv.x) vr.x = uv.x; + if(vr.z > uv.z + 1) vr.z = uv.z + 1; + break; + case CLAMP_REGION_REPEAT: // TODO + break; + default: + __assume(0); + } + + switch(wmt) + { + case CLAMP_REPEAT: + if(mask & 0xf000) {if(vr.y < v.y) vr.y = v.y; if(vr.w > v.w + 1) vr.w = v.w + 1;} + break; + case CLAMP_CLAMP: + case CLAMP_REGION_CLAMP: + if(vr.y < uv.y) vr.y = uv.y; + if(vr.w > uv.w + 1) vr.w = uv.w + 1; + break; + case CLAMP_REGION_REPEAT: // TODO + break; + default: + __assume(0); + } + } + + r = vr.rintersect(GSVector4i(0, 0, w, h)); +} + +void GSRendererSW::GetAlphaMinMax() +{ + if(m_vtrace.m_alpha.valid) + { + return; + } + + const GSDrawingEnvironment& env = m_env; + const GSDrawingContext* context = m_context; + + GSVector4i a = GSVector4i(m_vtrace.m_min.c.wwww(m_vtrace.m_max.c)) >> 7; + + if(PRIM->TME && context->TEX0.TCC) + { + uint32 bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; + uint32 cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; + uint32 pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; + + if(bpp == 32) + { + a.y = 0; + a.w = 0xff; + } + else if(bpp == 24) + { + a.y = env.TEXA.AEM ? 0 : env.TEXA.TA0; + a.w = env.TEXA.TA0; + } + else if(bpp == 16) + { + a.y = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); + a.w = max(env.TEXA.TA0, env.TEXA.TA1); + } + else + { + m_mem.m_clut.GetAlphaMinMax32(a.y, a.w); + } + + switch(context->TEX0.TFX) + { + case TFX_MODULATE: + a.x = (a.x * a.y) >> 7; + a.z = (a.z * a.w) >> 7; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_DECAL: + a.x = a.y; + a.z = a.w; + break; + case TFX_HIGHLIGHT: + a.x = a.x + a.y; + a.z = a.z + a.w; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_HIGHLIGHT2: + a.x = a.y; + a.z = a.w; + break; + default: + __assume(0); + } + } + + m_vtrace.m_alpha.min = a.x; + m_vtrace.m_alpha.max = a.z; + m_vtrace.m_alpha.valid = true; +} + +bool GSRendererSW::TryAlphaTest(uint32& fm, uint32& zm) +{ + const GSDrawingContext* context = m_context; + + bool pass = true; + + if(context->TEST.ATST == ATST_NEVER) + { + pass = false; + } + else if(context->TEST.ATST != ATST_ALWAYS) + { + GetAlphaMinMax(); + + int amin = m_vtrace.m_alpha.min; + int amax = m_vtrace.m_alpha.max; + + int aref = context->TEST.AREF; + + switch(context->TEST.ATST) + { + case ATST_NEVER: + pass = false; + break; + case ATST_ALWAYS: + pass = true; + break; + case ATST_LESS: + if(amax < aref) pass = true; + else if(amin >= aref) pass = false; + else return false; + break; + case ATST_LEQUAL: + if(amax <= aref) pass = true; + else if(amin > aref) pass = false; + else return false; + break; + case ATST_EQUAL: + if(amin == aref && amax == aref) pass = true; + else if(amin > aref || amax < aref) pass = false; + else return false; + break; + case ATST_GEQUAL: + if(amin >= aref) pass = true; + else if(amax < aref) pass = false; + else return false; + break; + case ATST_GREATER: + if(amin > aref) pass = true; + else if(amax <= aref) pass = false; + else return false; + break; + case ATST_NOTEQUAL: + if(amin == aref && amax == aref) pass = false; + else if(amin > aref || amax < aref) pass = true; + else return false; + break; + default: + __assume(0); + } + } + + if(!pass) + { + switch(context->TEST.AFAIL) + { + case AFAIL_KEEP: fm = zm = 0xffffffff; break; + case AFAIL_FB_ONLY: zm = 0xffffffff; break; + case AFAIL_ZB_ONLY: fm = 0xffffffff; break; + case AFAIL_RGB_ONLY: fm |= 0xff000000; zm = 0xffffffff; break; + default: __assume(0); + } + } + + return true; +} + +void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) +{ + const GSDrawingEnvironment& env = m_env; + const GSDrawingContext* context = m_context; + + p.vm = m_mem.m_vm8; + + p.fbo = m_mem.GetOffset(context->FRAME.Block(), context->FRAME.FBW, context->FRAME.PSM); + p.zbo = m_mem.GetOffset(context->ZBUF.Block(), context->FRAME.FBW, context->ZBUF.PSM); + p.fzbo = m_mem.GetOffset4(context->FRAME, context->ZBUF); + + p.sel.key = 0; + + p.sel.fpsm = 3; + p.sel.zpsm = 3; + p.sel.atst = ATST_ALWAYS; + p.sel.tfx = TFX_NONE; + p.sel.ababcd = 255; + p.sel.sprite = primclass == GS_SPRITE_CLASS ? 1 : 0; + + p.fm = context->FRAME.FBMSK; + p.zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0; + + if(context->TEST.ZTE && context->TEST.ZTST == ZTST_NEVER) + { + p.fm = 0xffffffff; + p.zm = 0xffffffff; + } + + if(PRIM->TME) + { + m_mem.m_clut.Read32(context->TEX0, env.TEXA); + } + + if(context->TEST.ATE) + { + if(!TryAlphaTest(p.fm, p.zm)) + { + p.sel.atst = context->TEST.ATST; + p.sel.afail = context->TEST.AFAIL; + } + } + + bool fwrite = p.fm != 0xffffffff; + bool ftest = p.sel.atst != ATST_ALWAYS || context->TEST.DATE && context->FRAME.PSM != PSM_PSMCT24; + + p.sel.fwrite = fwrite; + p.sel.ftest = ftest; + + if(fwrite || ftest) + { + p.sel.fpsm = GSUtil::EncodePSM(context->FRAME.PSM); + + if((primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS) && m_vtrace.m_eq.rgba != 15) + { + p.sel.iip = PRIM->IIP; + } + + if(PRIM->TME) + { + p.sel.tfx = context->TEX0.TFX; + p.sel.tcc = context->TEX0.TCC; + p.sel.fst = PRIM->FST; + p.sel.ltf = context->TEX1.IsLinear(); + p.sel.tlu = GSLocalMemory::m_psm[context->TEX0.PSM].pal > 0; + p.sel.wms = context->CLAMP.WMS; + p.sel.wmt = context->CLAMP.WMT; + + if(p.sel.iip == 0 && p.sel.tfx == TFX_MODULATE && p.sel.tcc) + { + if(m_vtrace.m_eq.rgba == 15 && (m_vtrace.m_min.c == GSVector4(128.0f * 128.0f)).alltrue()) + { + // modulate does not do anything when vertex color is 0x80 + + p.sel.tfx = TFX_DECAL; + } + } + + if(p.sel.fst == 0) + { + // skip per pixel division if q is constant + + GSVertexSW* v = m_vertices; + + if(m_vtrace.m_eq.q) + { + p.sel.fst = 1; + + if(v[0].t.z != 1.0f) + { + GSVector4 w = v[0].t.zzzz().rcpnr(); + + for(int i = 0, j = m_count; i < j; i++) + { + v[i].t *= w; + } + + m_vtrace.m_min.t *= w; + m_vtrace.m_max.t *= w; + } + } + else if(primclass == GS_SPRITE_CLASS) + { + p.sel.fst = 1; + + GSVector4 tmin = GSVector4(FLT_MAX); + GSVector4 tmax = GSVector4(-FLT_MAX); + + for(int i = 0, j = m_count; i < j; i += 2) + { + GSVector4 w = v[i + 1].t.zzzz().rcpnr(); + + GSVector4 v0 = v[i + 0].t * w; + GSVector4 v1 = v[i + 1].t * w; + + v[i + 0].t = v0; + v[i + 1].t = v1; + + tmin = tmin.minv(v0).minv(v1); + tmax = tmax.maxv(v0).maxv(v1); + } + + m_vtrace.m_max.t = tmax; + m_vtrace.m_min.t = tmin; + } + } + + if(p.sel.fst) + { + // if q is constant we can do the half pel shift for bilinear sampling on the vertices + + if(p.sel.ltf) + { + GSVector4 half(0x8000, 0x8000); + + GSVertexSW* v = m_vertices; + + for(int i = 0, j = m_count; i < j; i++) + { + v[i].t -= half; + } + + m_vtrace.m_min.t -= half; + m_vtrace.m_max.t += half; + } + } + /* + else + { + GSVector4 tmin = GSVector4(FLT_MAX); + GSVector4 tmax = GSVector4(-FLT_MAX); + + GSVertexSW* v = m_vertices; + + for(int i = 0, j = m_count; i < j; i++) + { + GSVector4 v0 = v[i].t * v[i].t.zzzz().rcpnr(); + + tmin = tmin.minv(v0); + tmax = tmax.maxv(v0); + } + + if(p.sel.ltf) + { + GSVector4 half(0x8000, 0x8000); + + tmin -= half; + tmax += half; + } + + m_vtrace.min.t = tmin; + m_vtrace.max.t = tmax; + } + */ + + int w = 1 << context->TEX0.TW; + int h = 1 << context->TEX0.TH; + + GSVector4i r; + + GetTextureMinMax(w, h, r, p.sel.fst); + + const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, r); + + if(!t) {ASSERT(0); return;} + + p.tex = t->m_buff; + p.clut = m_mem.m_clut; + p.tw = t->m_tw; + } + + p.sel.fge = PRIM->FGE; + + if(context->FRAME.PSM != PSM_PSMCT24) + { + p.sel.date = context->TEST.DATE; + p.sel.datm = context->TEST.DATM; + } + + int amin = 0, amax = 0xff; + + if(PRIM->ABE && context->ALPHA.A != context->ALPHA.B && !PRIM->AA1) + { + if(context->ALPHA.C == 0) + { + GetAlphaMinMax(); + + amin = m_vtrace.m_alpha.min; + amax = m_vtrace.m_alpha.max; + } + else if(context->ALPHA.C == 1) + { + if(p.sel.fpsm == 1) + { + amin = amax = 0x80; + } + } + else if(context->ALPHA.C == 1) + { + amin = amax = context->ALPHA.FIX; + } + } + + if(PRIM->ABE && !context->ALPHA.IsOpaque(amin, amax) || PRIM->AA1) + { + p.sel.abe = PRIM->ABE; + p.sel.ababcd = context->ALPHA.u32[0]; + + if(env.PABE.PABE) + { + p.sel.pabe = 1; + } + + if(PRIM->AA1 && (primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS)) + { + p.sel.aa1 = m_aa1 ? 1 : 0; + } + } + + if(p.sel.date + || p.sel.aba == 1 || p.sel.abb == 1 || p.sel.abc == 1 || p.sel.abd == 1 + || p.sel.atst != ATST_ALWAYS && p.sel.afail == AFAIL_RGB_ONLY + || p.sel.fpsm == 0 && p.fm != 0 && p.fm != 0xffffffff + || p.sel.fpsm == 1 && (p.fm & 0x00ffffff) != 0 && (p.fm & 0x00ffffff) != 0x00ffffff + || p.sel.fpsm == 2 && (p.fm & 0x80f8f8f8) != 0 && (p.fm & 0x80f8f8f8) != 0x80f8f8f8) + { + p.sel.rfb = 1; + } + + p.sel.colclamp = env.COLCLAMP.CLAMP; + p.sel.fba = context->FBA.FBA; + p.sel.dthe = env.DTHE.DTHE; + } + + bool zwrite = p.zm != 0xffffffff; + bool ztest = context->TEST.ZTE && context->TEST.ZTST > 1; + + p.sel.zwrite = zwrite; + p.sel.ztest = ztest; + + if(zwrite || ztest) + { + p.sel.zpsm = GSUtil::EncodePSM(context->ZBUF.PSM); + p.sel.ztst = ztest ? context->TEST.ZTST : 1; + p.sel.zoverflow = GSVector4i(m_vtrace.m_max.p).z == 0x80000000; + } +} + diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index aa70b60454..7ccdd9b136 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -27,768 +27,31 @@ extern const GSVector4 g_pos_scale; -template -class GSRendererSW : public GSRendererT +class GSRendererSW : public GSRendererT { protected: GSRasterizerList m_rl; GSTextureCacheSW* m_tc; GSVertexTrace m_vtrace; - Texture m_texture[2]; + GSTexture* m_texture[2]; bool m_reset; - void Reset() - { - // TODO: GSreset can come from the main thread too => crash - // m_tc->RemoveAll(); - - m_reset = true; - - __super::Reset(); - } - - void VSync(int field) - { - __super::VSync(field); - - m_tc->IncAge(); - - if(m_reset) - { - m_tc->RemoveAll(); - - m_reset = false; - } - - // if((m_perfmon.GetFrame() & 255) == 0) m_rl.PrintStats(); - } - - void ResetDevice() - { - m_texture[0] = Texture(); - m_texture[1] = Texture(); - } - - bool GetOutput(int i, Texture& t) - { - const GSRegDISPFB& DISPFB = m_regs->DISP[i].DISPFB; - - GIFRegTEX0 TEX0; - - TEX0.TBP0 = DISPFB.Block(); - TEX0.TBW = DISPFB.FBW; - TEX0.PSM = DISPFB.PSM; - - GSVector4i r(0, 0, TEX0.TBW * 64, GetFrameRect(i).bottom); - - // TODO: round up bottom - - if(m_texture[i].GetWidth() != r.width() || m_texture[i].GetHeight() != r.height()) - { - m_texture[i] = Texture(); - } - - if(!m_texture[i] && !m_dev.CreateTexture(m_texture[i], r.width(), r.height())) - { - return false; - } - - GIFRegCLAMP CLAMP; - - CLAMP.WMS = CLAMP.WMT = 1; - - // TODO - static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); - static int pitch = 1024 * 4; - - m_mem.ReadTexture(r, buff, pitch, TEX0, m_env.TEXA, CLAMP); - - m_texture[i].Update(r, buff, pitch); - - t = m_texture[i]; - - if(s_dump) - { - if(s_save) - { - t.Save(format("c:\\temp1\\_%05d_f%I64d_fr%d_%05x_%d.bmp", s_n, m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM)); - } - - s_n++; - } - - return true; - } - - void GetAlphaMinMax() - { - if(m_vtrace.m_alpha.valid) - { - return; - } - - const GSDrawingEnvironment& env = m_env; - const GSDrawingContext* context = m_context; - - GSVector4i a = GSVector4i(m_vtrace.m_min.c.wwww(m_vtrace.m_max.c)) >> 7; - - if(PRIM->TME && context->TEX0.TCC) - { - uint32 bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; - uint32 cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; - uint32 pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; - - if(bpp == 32) - { - a.y = 0; - a.w = 0xff; - } - else if(bpp == 24) - { - a.y = env.TEXA.AEM ? 0 : env.TEXA.TA0; - a.w = env.TEXA.TA0; - } - else if(bpp == 16) - { - a.y = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); - a.w = max(env.TEXA.TA0, env.TEXA.TA1); - } - else - { - m_mem.m_clut.GetAlphaMinMax32(a.y, a.w); - } - - switch(context->TEX0.TFX) - { - case TFX_MODULATE: - a.x = (a.x * a.y) >> 7; - a.z = (a.z * a.w) >> 7; - if(a.x > 0xff) a.x = 0xff; - if(a.z > 0xff) a.z = 0xff; - break; - case TFX_DECAL: - a.x = a.y; - a.z = a.w; - break; - case TFX_HIGHLIGHT: - a.x = a.x + a.y; - a.z = a.z + a.w; - if(a.x > 0xff) a.x = 0xff; - if(a.z > 0xff) a.z = 0xff; - break; - case TFX_HIGHLIGHT2: - a.x = a.y; - a.z = a.w; - break; - default: - __assume(0); - } - } - - m_vtrace.m_alpha.min = a.x; - m_vtrace.m_alpha.max = a.z; - m_vtrace.m_alpha.valid = true; - } - - bool TryAlphaTest(uint32& fm, uint32& zm) - { - const GSDrawingContext* context = m_context; - - bool pass = true; - - if(context->TEST.ATST == ATST_NEVER) - { - pass = false; - } - else if(context->TEST.ATST != ATST_ALWAYS) - { - GetAlphaMinMax(); - - int amin = m_vtrace.m_alpha.min; - int amax = m_vtrace.m_alpha.max; - - int aref = context->TEST.AREF; - - switch(context->TEST.ATST) - { - case ATST_NEVER: - pass = false; - break; - case ATST_ALWAYS: - pass = true; - break; - case ATST_LESS: - if(amax < aref) pass = true; - else if(amin >= aref) pass = false; - else return false; - break; - case ATST_LEQUAL: - if(amax <= aref) pass = true; - else if(amin > aref) pass = false; - else return false; - break; - case ATST_EQUAL: - if(amin == aref && amax == aref) pass = true; - else if(amin > aref || amax < aref) pass = false; - else return false; - break; - case ATST_GEQUAL: - if(amin >= aref) pass = true; - else if(amax < aref) pass = false; - else return false; - break; - case ATST_GREATER: - if(amin > aref) pass = true; - else if(amax <= aref) pass = false; - else return false; - break; - case ATST_NOTEQUAL: - if(amin == aref && amax == aref) pass = false; - else if(amin > aref || amax < aref) pass = true; - else return false; - break; - default: - __assume(0); - } - } - - if(!pass) - { - switch(context->TEST.AFAIL) - { - case AFAIL_KEEP: fm = zm = 0xffffffff; break; - case AFAIL_FB_ONLY: zm = 0xffffffff; break; - case AFAIL_ZB_ONLY: fm = 0xffffffff; break; - case AFAIL_RGB_ONLY: fm |= 0xff000000; zm = 0xffffffff; break; - default: __assume(0); - } - } - - return true; - } - - void GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) - { - const GSDrawingEnvironment& env = m_env; - const GSDrawingContext* context = m_context; - - p.vm = m_mem.m_vm8; - - p.fbo = m_mem.GetOffset(context->FRAME.Block(), context->FRAME.FBW, context->FRAME.PSM); - p.zbo = m_mem.GetOffset(context->ZBUF.Block(), context->FRAME.FBW, context->ZBUF.PSM); - p.fzbo = m_mem.GetOffset4(context->FRAME, context->ZBUF); - - p.sel.key = 0; - - p.sel.fpsm = 3; - p.sel.zpsm = 3; - p.sel.atst = ATST_ALWAYS; - p.sel.tfx = TFX_NONE; - p.sel.ababcd = 255; - p.sel.sprite = primclass == GS_SPRITE_CLASS ? 1 : 0; - - p.fm = context->FRAME.FBMSK; - p.zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0; - - if(context->TEST.ZTE && context->TEST.ZTST == ZTST_NEVER) - { - p.fm = 0xffffffff; - p.zm = 0xffffffff; - } - - if(PRIM->TME) - { - m_mem.m_clut.Read32(context->TEX0, env.TEXA); - } - - if(context->TEST.ATE) - { - if(!TryAlphaTest(p.fm, p.zm)) - { - p.sel.atst = context->TEST.ATST; - p.sel.afail = context->TEST.AFAIL; - } - } - - bool fwrite = p.fm != 0xffffffff; - bool ftest = p.sel.atst != ATST_ALWAYS || context->TEST.DATE && context->FRAME.PSM != PSM_PSMCT24; - - p.sel.fwrite = fwrite; - p.sel.ftest = ftest; - - if(fwrite || ftest) - { - p.sel.fpsm = GSUtil::EncodePSM(context->FRAME.PSM); - - if((primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS) && m_vtrace.m_eq.rgba != 15) - { - p.sel.iip = PRIM->IIP; - } - - if(PRIM->TME) - { - p.sel.tfx = context->TEX0.TFX; - p.sel.tcc = context->TEX0.TCC; - p.sel.fst = PRIM->FST; - p.sel.ltf = context->TEX1.IsLinear(); - p.sel.tlu = GSLocalMemory::m_psm[context->TEX0.PSM].pal > 0; - p.sel.wms = context->CLAMP.WMS; - p.sel.wmt = context->CLAMP.WMT; - - if(p.sel.iip == 0 && p.sel.tfx == TFX_MODULATE && p.sel.tcc) - { - if(m_vtrace.m_eq.rgba == 15 && (m_vtrace.m_min.c == GSVector4(128.0f * 128.0f)).alltrue()) - { - // modulate does not do anything when vertex color is 0x80 - - p.sel.tfx = TFX_DECAL; - } - } - - if(p.sel.fst == 0) - { - // skip per pixel division if q is constant - - GSVertexSW* v = m_vertices; - - if(m_vtrace.m_eq.q) - { - p.sel.fst = 1; - - if(v[0].t.z != 1.0f) - { - GSVector4 w = v[0].t.zzzz().rcpnr(); - - for(int i = 0, j = m_count; i < j; i++) - { - v[i].t *= w; - } - - m_vtrace.m_min.t *= w; - m_vtrace.m_max.t *= w; - } - } - else if(primclass == GS_SPRITE_CLASS) - { - p.sel.fst = 1; - - GSVector4 tmin = GSVector4(FLT_MAX); - GSVector4 tmax = GSVector4(-FLT_MAX); - - for(int i = 0, j = m_count; i < j; i += 2) - { - GSVector4 w = v[i + 1].t.zzzz().rcpnr(); - - GSVector4 v0 = v[i + 0].t * w; - GSVector4 v1 = v[i + 1].t * w; - - v[i + 0].t = v0; - v[i + 1].t = v1; - - tmin = tmin.minv(v0).minv(v1); - tmax = tmax.maxv(v0).maxv(v1); - } - - m_vtrace.m_max.t = tmax; - m_vtrace.m_min.t = tmin; - } - } - - if(p.sel.fst) - { - // if q is constant we can do the half pel shift for bilinear sampling on the vertices - - if(p.sel.ltf) - { - GSVector4 half(0x8000, 0x8000); - - GSVertexSW* v = m_vertices; - - for(int i = 0, j = m_count; i < j; i++) - { - v[i].t -= half; - } - - m_vtrace.m_min.t -= half; - m_vtrace.m_max.t += half; - } - } - /* - else - { - GSVector4 tmin = GSVector4(FLT_MAX); - GSVector4 tmax = GSVector4(-FLT_MAX); - - GSVertexSW* v = m_vertices; - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 v0 = v[i].t * v[i].t.zzzz().rcpnr(); - - tmin = tmin.minv(v0); - tmax = tmax.maxv(v0); - } - - if(p.sel.ltf) - { - GSVector4 half(0x8000, 0x8000); - - tmin -= half; - tmax += half; - } - - m_vtrace.min.t = tmin; - m_vtrace.max.t = tmax; - } - */ - - int w = 1 << context->TEX0.TW; - int h = 1 << context->TEX0.TH; - - GSVector4i r; - - MinMaxUV(w, h, r, p.sel.fst); - - const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, r); - - if(!t) {ASSERT(0); return;} - - p.tex = t->m_buff; - p.clut = m_mem.m_clut; - p.tw = t->m_tw; - } - - p.sel.fge = PRIM->FGE; - - if(context->FRAME.PSM != PSM_PSMCT24) - { - p.sel.date = context->TEST.DATE; - p.sel.datm = context->TEST.DATM; - } - - int amin = 0, amax = 0xff; - - if(PRIM->ABE && context->ALPHA.A != context->ALPHA.B && !PRIM->AA1) - { - if(context->ALPHA.C == 0) - { - GetAlphaMinMax(); - - amin = m_vtrace.m_alpha.min; - amax = m_vtrace.m_alpha.max; - } - else if(context->ALPHA.C == 1) - { - if(p.sel.fpsm == 1) - { - amin = amax = 0x80; - } - } - else if(context->ALPHA.C == 1) - { - amin = amax = context->ALPHA.FIX; - } - } - - if(PRIM->ABE && !context->ALPHA.IsOpaque(amin, amax) || PRIM->AA1) - { - p.sel.abe = PRIM->ABE; - p.sel.ababcd = context->ALPHA.u32[0]; - - if(env.PABE.PABE) - { - p.sel.pabe = 1; - } - - if(PRIM->AA1 && (primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS)) - { - p.sel.aa1 = m_aa1 ? 1 : 0; - } - } - - if(p.sel.date - || p.sel.aba == 1 || p.sel.abb == 1 || p.sel.abc == 1 || p.sel.abd == 1 - || p.sel.atst != ATST_ALWAYS && p.sel.afail == AFAIL_RGB_ONLY - || p.sel.fpsm == 0 && p.fm != 0 && p.fm != 0xffffffff - || p.sel.fpsm == 1 && (p.fm & 0x00ffffff) != 0 && (p.fm & 0x00ffffff) != 0x00ffffff - || p.sel.fpsm == 2 && (p.fm & 0x80f8f8f8) != 0 && (p.fm & 0x80f8f8f8) != 0x80f8f8f8) - { - p.sel.rfb = 1; - } - - p.sel.colclamp = env.COLCLAMP.CLAMP; - p.sel.fba = context->FBA.FBA; - p.sel.dthe = env.DTHE.DTHE; - } - - bool zwrite = p.zm != 0xffffffff; - bool ztest = context->TEST.ZTE && context->TEST.ZTST > 1; - - p.sel.zwrite = zwrite; - p.sel.ztest = ztest; - - if(zwrite || ztest) - { - p.sel.zpsm = GSUtil::EncodePSM(context->ZBUF.PSM); - p.sel.ztst = ztest ? context->TEST.ZTST : 1; - p.sel.zoverflow = GSVector4i(m_vtrace.m_max.p).z == 0x80000000; - } - } - - void Draw() - { - GS_PRIM_CLASS primclass = GSUtil::GetPrimClass(PRIM->PRIM); - - m_vtrace.Update(m_vertices, m_count, primclass, PRIM->IIP, PRIM->TME, m_context->TEX0.TFX, m_context->TEX0.TCC); - - if(m_dump) - { - m_dump.Object(m_vertices, m_count, primclass); - } - - GSScanlineParam p; - - GetScanlineParam(p, primclass); - - if((p.fm & p.zm) == 0xffffffff) - { - return; - } - - if(s_dump) - { - uint64 frame = m_perfmon.GetFrame(); - - string s; - - if(s_save && PRIM->TME) - { - s = format("c:\\temp1\\_%05d_f%I64d_tex_%05x_%d.bmp", s_n, frame, (int)m_context->TEX0.TBP0, (int)m_context->TEX0.PSM); - - m_mem.SaveBMP(s, m_context->TEX0.TBP0, m_context->TEX0.TBW, m_context->TEX0.PSM, 1 << m_context->TEX0.TW, 1 << m_context->TEX0.TH); - } - - s_n++; - - if(s_save) - { - s = format("c:\\temp1\\_%05d_f%I64d_rt0_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); - - m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); - } - - if(s_savez) - { - s = format("c:\\temp1\\_%05d_f%I64d_rz0_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); - - m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); - } - - s_n++; - } - - GSRasterizerData data; - - data.scissor = GSVector4i(m_context->scissor.in); - data.scissor.z = min(data.scissor.z, (int)m_context->FRAME.FBW * 64); // TODO: find a game that overflows and check which one is the right behaviour - data.primclass = primclass; - data.vertices = m_vertices; - data.count = m_count; - data.param = &p; - - m_rl.Draw(&data); - - GSRasterizerStats stats; - - m_rl.GetStats(stats); - - m_perfmon.Put(GSPerfMon::Draw, 1); - m_perfmon.Put(GSPerfMon::Prim, stats.prims); - m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); - - GSVector4i r = GSVector4i(m_vtrace.m_min.p.xyxy(m_vtrace.m_max.p)).rintersect(data.scissor); - - GIFRegBITBLTBUF BITBLTBUF; - - BITBLTBUF.DBW = m_context->FRAME.FBW; - - if(p.fm != 0xffffffff) - { - BITBLTBUF.DBP = m_context->FRAME.Block(); - BITBLTBUF.DPSM = m_context->FRAME.PSM; - - m_tc->InvalidateVideoMem(BITBLTBUF, r); - } - - if(p.zm != 0xffffffff) - { - BITBLTBUF.DBP = m_context->ZBUF.Block(); - BITBLTBUF.DPSM = m_context->ZBUF.PSM; - - m_tc->InvalidateVideoMem(BITBLTBUF, r); - } - - if(s_dump) - { - uint64 frame = m_perfmon.GetFrame(); - - string s; - - if(s_save) - { - s = format("c:\\temp1\\_%05d_f%I64d_rt1_%05x_%d.bmp", s_n, frame, m_context->FRAME.Block(), m_context->FRAME.PSM); - - m_mem.SaveBMP(s, m_context->FRAME.Block(), m_context->FRAME.FBW, m_context->FRAME.PSM, GetFrameRect().width(), 512);//GetFrameSize(1).cy); - } - - if(s_savez) - { - s = format("c:\\temp1\\_%05d_f%I64d_rz1_%05x_%d.bmp", s_n, frame, m_context->ZBUF.Block(), m_context->ZBUF.PSM); - - m_mem.SaveBMP(s, m_context->ZBUF.Block(), m_context->FRAME.FBW, m_context->ZBUF.PSM, GetFrameRect().width(), 512); - } - - s_n++; - } - - if(0)//stats.ticks > 5000000) - { - printf("* [%I64d | %012I64x] ticks %I64d prims %d (%d) pixels %d (%d)\n", - m_perfmon.GetFrame(), p.sel.key, - stats.ticks, - stats.prims, stats.prims > 0 ? (int)(stats.ticks / stats.prims) : -1, - stats.pixels, stats.pixels > 0 ? (int)(stats.ticks / stats.pixels) : -1); - } - } - - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) - { - m_tc->InvalidateVideoMem(BITBLTBUF, r); - } - - void MinMaxUV(int w, int h, GSVector4i& r, uint32 fst) - { - const GSDrawingContext* context = m_context; - - int wms = context->CLAMP.WMS; - int wmt = context->CLAMP.WMT; - - int minu = (int)context->CLAMP.MINU; - int minv = (int)context->CLAMP.MINV; - int maxu = (int)context->CLAMP.MAXU; - int maxv = (int)context->CLAMP.MAXV; - - GSVector4i vr(0, 0, w, h); - - switch(wms) - { - case CLAMP_REPEAT: - break; - case CLAMP_CLAMP: - break; - case CLAMP_REGION_CLAMP: - if(vr.x < minu) vr.x = minu; - if(vr.z > maxu + 1) vr.z = maxu + 1; - break; - case CLAMP_REGION_REPEAT: - vr.x = maxu; - vr.z = vr.x + (minu + 1); - break; - default: - __assume(0); - } - - switch(wmt) - { - case CLAMP_REPEAT: - break; - case CLAMP_CLAMP: - break; - case CLAMP_REGION_CLAMP: - if(vr.y < minv) vr.y = minv; - if(vr.w > maxv + 1) vr.w = maxv + 1; - break; - case CLAMP_REGION_REPEAT: - vr.y = maxv; - vr.w = vr.y + (minv + 1); - break; - default: - __assume(0); - } - - if(fst) - { - GSVector4i uv = GSVector4i(m_vtrace.m_min.t.xyxy(m_vtrace.m_max.t)).sra32(16); - - GSVector4i u, v; - - int mask; - - if(wms == CLAMP_REPEAT || wmt == CLAMP_REPEAT) - { - int tw = context->TEX0.TW; - int th = context->TEX0.TH; - - u = uv & GSVector4i::xffffffff().srl32(32 - tw); - v = uv & GSVector4i::xffffffff().srl32(32 - th); - - GSVector4i uu = uv.sra32(tw); - GSVector4i vv = uv.sra32(th); - - mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); - } - - switch(wms) - { - case CLAMP_REPEAT: - if(mask & 0x000f) {if(vr.x < u.x) vr.x = u.x; if(vr.z > u.z + 1) vr.z = u.z + 1;} - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - if(vr.x < uv.x) vr.x = uv.x; - if(vr.z > uv.z + 1) vr.z = uv.z + 1; - break; - case CLAMP_REGION_REPEAT: // TODO - break; - default: - __assume(0); - } - - switch(wmt) - { - case CLAMP_REPEAT: - if(mask & 0xf000) {if(vr.y < v.y) vr.y = v.y; if(vr.w > v.w + 1) vr.w = v.w + 1;} - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - if(vr.y < uv.y) vr.y = uv.y; - if(vr.w > uv.w + 1) vr.w = uv.w + 1; - break; - case CLAMP_REGION_REPEAT: // TODO - break; - default: - __assume(0); - } - } - - r = vr.rintersect(GSVector4i(0, 0, w, h)); - } + void Reset(); + void VSync(int field); + void ResetDevice(); + GSTexture* GetOutput(int i); + + void Draw(); + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); + + void GetTextureMinMax(int w, int h, GSVector4i& r, uint32 fst); + void GetAlphaMinMax(); + bool TryAlphaTest(uint32& fm, uint32& zm); + void GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass); public: - GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, int threads) - : GSRendererT(base, mt, irq, rs) - { - m_rl.Create(this, threads); - - m_tc = new GSTextureCacheSW(this); - - InitVertexKick >(); - } - - virtual ~GSRendererSW() - { - delete m_tc; - } + GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev, int threads); + virtual ~GSRendererSW(); template void VertexKick(bool skip) diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 31300861bf..4295e9adb3 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -965,8 +965,6 @@ void GSState::FlushWrite() void GSState::Write(uint8* mem, int len) { - int dx = m_env.TRXPOS.DSAX; - int dy = m_env.TRXPOS.DSAY; int w = m_env.TRXREG.RRW; int h = m_env.TRXREG.RRH; @@ -1057,6 +1055,8 @@ void GSState::Move() const GSLocalMemory::psm_t& spsm = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM]; const GSLocalMemory::psm_t& dpsm = GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM]; + // TODO: unroll inner loops (width has special size requirement, must be multiples of 1 << n, depending on the format) + if(spsm.trbpp == dpsm.trbpp && spsm.trbpp >= 16) { int* soffset = spsm.rowOffset[0]; @@ -2164,6 +2164,7 @@ bool GSC_SonicUnleashed(const GSFrameInfo& fi, int& skip) return true; } + bool GSState::IsBadFrame(int& skip) { GSFrameInfo fi; diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index 6f485905bd..d21d778425 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -44,5 +44,5 @@ public: virtual void Unmap() = 0; virtual bool Save(const string& fn, bool dds = false) = 0; - GSVector2i GetSize() {return GSVector2i(GetWidth(), GetHeight());} + GSVector2i GetSize() const {return GSVector2i(GetWidth(), GetHeight());} }; diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 9174325f7a..90f717fd8a 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -21,3 +21,770 @@ #include "StdAfx.h" #include "GSTextureCache.h" + +GSTextureCache::GSTextureCache(GSRenderer* r) + : m_renderer(r) +{ +} + +GSTextureCache::~GSTextureCache() +{ + RemoveAll(); +} + +void GSTextureCache::RemoveAll() +{ + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + { + delete *i; + } + + m_rt.clear(); + + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) + { + delete *i; + } + + m_ds.clear(); + + for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) + { + delete *i; + } + + m_tex.clear(); +} + +GSTextureCache::GSRenderTarget* GSTextureCache::GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb) +{ + GSRenderTarget* rt = NULL; + + if(rt == NULL) + { + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + { + GSRenderTarget* rt2 = *i; + + if(rt2->m_TEX0.TBP0 == TEX0.TBP0) + { + m_rt.splice(m_rt.begin(), m_rt, i); + + rt = rt2; + + if(!fb) rt->m_TEX0 = TEX0; + + rt->Update(); + + break; + } + } + } + + if(rt == NULL && fb) + { + // HACK: try to find something close to the base pointer + + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + { + GSRenderTarget* rt2 = *i; + + if(rt2->m_TEX0.TBP0 <= TEX0.TBP0 && TEX0.TBP0 < rt2->m_TEX0.TBP0 + 0x700 && (!rt || rt2->m_TEX0.TBP0 >= rt->m_TEX0.TBP0)) + { + rt = rt2; + } + } + + if(rt) + { + rt->Update(); + } + } + + if(rt == NULL) + { + rt = CreateRenderTarget(); + + rt->m_TEX0 = TEX0; + + if(!rt->Create(w, h)) + { + delete rt; + + return NULL; + } + + m_rt.push_front(rt); + } + + if(m_renderer->CanUpscale()) + { + GSVector4i fr = m_renderer->GetFrameRect(); + + int ww = (int)(fr.left + rt->m_TEX0.TBW * 64); + int hh = (int)(fr.top + m_renderer->GetDisplayRect().height()); + + if(hh <= m_renderer->GetDeviceSize().y / 2) + { + hh *= 2; + } + + if(ww > 0 && hh > 0) + { + rt->m_texture->m_scale.x = (float)w / ww; + rt->m_texture->m_scale.y = (float)h / hh; + } + } + + if(!fb) + { + rt->m_used = true; + } + + return rt; +} + +GSTextureCache::GSDepthStencil* GSTextureCache::GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h) +{ + GSDepthStencil* ds = NULL; + + if(ds == NULL) + { + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) + { + GSDepthStencil* ds2 = *i; + + if(ds2->m_TEX0.TBP0 == TEX0.TBP0) + { + m_ds.splice(m_ds.begin(), m_ds, i); + + ds = ds2; + + ds->m_TEX0 = TEX0; + + ds->Update(); + + break; + } + } + } + + if(ds == NULL) + { + ds = CreateDepthStencil(); + + ds->m_TEX0 = TEX0; + + if(!ds->Create(w, h)) + { + delete ds; + + return NULL; + } + + m_ds.push_front(ds); + } + + if(m_renderer->m_context->DepthWrite()) + { + ds->m_used = true; + } + + return ds; +} + +GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() +{ + const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; + const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; + + const uint32* clut = m_renderer->m_mem.m_clut; + const int pal = GSLocalMemory::m_psm[TEX0.PSM].pal; + + if(pal > 0) + { + m_renderer->m_mem.m_clut.Read(TEX0); + + /* + POSITION pos = m_tex.GetHeadPosition(); + + while(pos) + { + POSITION cur = pos; + + GSSurface* s = m_tex.GetNext(pos); + + if(s->m_TEX0.TBP0 == TEX0.CBP) + { + m_tex.RemoveAt(cur); + + delete s; + } + } + + pos = m_rt.GetHeadPosition(); + + while(pos) + { + POSITION cur = pos; + + GSSurface* s = m_rt.GetNext(pos); + + if(s->m_TEX0.TBP0 == TEX0.CBP) + { + m_rt.RemoveAt(cur); + + delete s; + } + } + + pos = m_ds.GetHeadPosition(); + + while(pos) + { + POSITION cur = pos; + + GSSurface* s = m_ds.GetNext(pos); + + if(s->m_TEX0.TBP0 == TEX0.CBP) + { + m_ds.RemoveAt(cur); + + delete s; + } + } + */ + } + + GSCachedTexture* t = NULL; + + for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) + { + t = *i; + + if(GSUtil::HasSharedBits(t->m_TEX0.TBP0, t->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + { + if(TEX0.PSM == t->m_TEX0.PSM && TEX0.TBW == t->m_TEX0.TBW + && TEX0.TW == t->m_TEX0.TW && TEX0.TH == t->m_TEX0.TH + && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.u64 == t->m_CLAMP.u64)) + && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) + { + m_tex.splice(m_tex.begin(), m_tex, i); + + break; + } + } + + t = NULL; + } + + if(t == NULL) + { + for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + { + GSRenderTarget* rt = *i; + + if(rt->m_dirty.empty() && GSUtil::HasSharedBits(rt->m_TEX0.TBP0, rt->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + { + t = CreateTexture(); + + if(!t->Create(rt)) + { + delete t; + + return NULL; + } + + m_tex.push_front(t); + + break; + } + } + } + + if(t == NULL) + { + for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) + { + GSDepthStencil* ds = *i; + + if(ds->m_dirty.empty() && ds->m_used && GSUtil::HasSharedBits(ds->m_TEX0.TBP0, ds->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + { + t = CreateTexture(); + + if(!t->Create(ds)) + { + delete t; + + return NULL; + } + + m_tex.push_front(t); + + break; + } + } + } + + if(t == NULL) + { + t = CreateTexture(); + + if(!t->Create()) + { + delete t; + + return NULL; + } + + m_tex.push_front(t); + } + + if(pal > 0) + { + int size = pal * sizeof(clut[0]); + + if(t->m_palette) + { + if(t->m_initpalette) + { + memcpy(t->m_clut, clut, size); + t->m_palette->Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); + t->m_initpalette = false; + } + else + { + if(GSVector4i::update(t->m_clut, clut, size)) + { + t->m_palette->Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); + } + } + } + else + { + memcpy(t->m_clut, clut, size); + } + } + + t->Update(); + + return t; +} + +void GSTextureCache::InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) +{ + for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) + { + list::iterator j = i++; + + GSCachedTexture* t = *j; + + if(GSUtil::HasSharedBits(FRAME.Block(), FRAME.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM) + || GSUtil::HasSharedBits(ZBUF.Block(), ZBUF.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) + { + m_tex.erase(j); + + delete t; + } + } +} + +void GSTextureCache::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) +{ + bool found = false; + + for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) + { + list::iterator j = i++; + + GSCachedTexture* t = *j; + + if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) + { + if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) + { + t->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); + + found = true; + } + else + { + m_tex.erase(j); + delete t; + } + } + else if(GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, t->m_TEX0.PSM)) + { + if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) + { + int rowsize = (int)BITBLTBUF.DBW * 8192; + int offset = ((int)BITBLTBUF.DBP - (int)t->m_TEX0.TBP0) * 256; + + if(rowsize > 0 && offset % rowsize == 0) + { + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; + + GSVector4i r2(r.left, r.top + y, r.right, r.bottom + y); + + int w = 1 << t->m_TEX0.TW; + int h = 1 << t->m_TEX0.TH; + + if(r2.bottom > 0 && r2.top < h && r2.right > 0 && r2.left < w) + { + t->m_dirty.push_back(GSDirtyRect(r2, BITBLTBUF.DPSM)); + } + } + } + } + } + + for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) + { + list::iterator j = i++; + + GSRenderTarget* rt = *j; + + if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) + { + if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM)) + { + rt->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); + rt->m_TEX0.TBW = BITBLTBUF.DBW; + } + else + { + m_rt.erase(j); + delete rt; + continue; + } + } + + if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM) && BITBLTBUF.DBP < rt->m_TEX0.TBP0) + { + uint32 rowsize = BITBLTBUF.DBW * 8192; + uint32 offset = (uint32)((rt->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); + + if(rowsize > 0 && offset % rowsize == 0) + { + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; + + if(r.bottom > y) + { + // TODO: do not add this rect above too + rt->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); + rt->m_TEX0.TBW = BITBLTBUF.DBW; + continue; + } + } + } + } + + // copypaste for ds + + for(list::iterator i = m_ds.begin(); i != m_ds.end(); ) + { + list::iterator j = i++; + + GSDepthStencil* ds = *j; + + if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, ds->m_TEX0.TBP0, ds->m_TEX0.PSM)) + { + if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM)) + { + ds->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); + ds->m_TEX0.TBW = BITBLTBUF.DBW; + } + else + { + m_ds.erase(j); + delete ds; + continue; + } + } + + if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM) && BITBLTBUF.DBP < ds->m_TEX0.TBP0) + { + uint32 rowsize = BITBLTBUF.DBW * 8192; + uint32 offset = (uint32)((ds->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); + + if(rowsize > 0 && offset % rowsize == 0) + { + int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; + + if(r.bottom > y) + { + // TODO: do not add this rect above too + ds->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); + ds->m_TEX0.TBW = BITBLTBUF.DBW; + continue; + } + } + } + } +} + +void GSTextureCache::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) +{ + for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) + { + list::iterator j = i++; + + GSRenderTarget* rt = *j; + + if(GSUtil::HasSharedBits(BITBLTBUF.SBP, BITBLTBUF.SPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) + { + if(GSUtil::HasCompatibleBits(BITBLTBUF.SPSM, rt->m_TEX0.PSM)) + { + rt->Read(r); + return; + } + else if(BITBLTBUF.SPSM == PSM_PSMCT32 && (rt->m_TEX0.PSM == PSM_PSMCT16 || rt->m_TEX0.PSM == PSM_PSMCT16S)) + { + // ffx-2 riku changing to her default (shoots some reflecting glass at the end), 16-bit rt read as 32-bit + + rt->Read(GSVector4i(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); + return; + } + else + { + m_rt.erase(j); + delete rt; + continue; + } + } + } +/* + // no good, ffx does a lot of readback after exiting menu, at 0x02f00 this wrongly finds rt 0x02100 (0,448 - 512,480) + + GSRenderTarget* rt2 = NULL; + int ymin = INT_MAX; + + pos = m_rt.GetHeadPosition(); + + while(pos) + { + GSRenderTarget* rt = m_rt.GetNext(pos); + + if(HasSharedBits(BITBLTBUF.SPSM, rt->m_TEX0.PSM) && BITBLTBUF.SBP > rt->m_TEX0.TBP0) + { + // ffx2 pause screen background + + uint32 rowsize = BITBLTBUF.SBW * 8192; + uint32 offset = (uint32)((BITBLTBUF.SBP - rt->m_TEX0.TBP0) * 256); + + if(rowsize > 0 && offset % rowsize == 0) + { + int y = m_renderer->m_mem.m_psm[BITBLTBUF.SPSM].pgs.y * offset / rowsize; + + if(y < ymin && y < 512) + { + rt2 = rt; + ymin = y; + } + } + } + } + + if(rt2) + { + rt2->Read(GSVector4i(r.left, r.top + ymin, r.right, r.bottom + ymin)); + } + + // TODO: ds +*/ +} + +void GSTextureCache::IncAge() +{ + RecycleByAge(m_tex, 2); + RecycleByAge(m_rt); + RecycleByAge(m_ds); +} + +// GSTextureCache::GSSurface + +GSTextureCache::GSSurface::GSSurface(GSRenderer* r) + : m_renderer(r) + , m_texture(NULL) + , m_palette(NULL) + , m_initpalette(false) + , m_age(0) +{ + m_TEX0.TBP0 = (uint32)~0; +} + +GSTextureCache::GSSurface::~GSSurface() +{ + m_renderer->m_dev->Recycle(m_texture); + m_renderer->m_dev->Recycle(m_palette); + + m_texture = NULL; + m_palette = NULL; +} + +void GSTextureCache::GSSurface::Update() +{ + m_age = 0; +} + +// GSTextureCache::GSRenderTarget + +GSTextureCache::GSRenderTarget::GSRenderTarget(GSRenderer* r) + : GSSurface(r) + , m_used(true) +{ +} + +bool GSTextureCache::GSRenderTarget::Create(int w, int h) +{ + // FIXME: initial data should be unswizzled from local mem in Update() if dirty + + m_texture = m_renderer->m_dev->CreateRenderTarget(w, h); + + return m_texture != NULL; +} + +// GSTextureCache::GSDepthStencil + +GSTextureCache::GSDepthStencil::GSDepthStencil(GSRenderer* r) + : GSSurface(r) + , m_used(false) +{ +} + +bool GSTextureCache::GSDepthStencil::Create(int w, int h) +{ + // FIXME: initial data should be unswizzled from local mem in Update() if dirty + + m_texture = m_renderer->m_dev->CreateDepthStencil(w, h); + + return m_texture != NULL; +} + +// GSTextureCache::GSCachedTexture + +GSTextureCache::GSCachedTexture::GSCachedTexture(GSRenderer* r) + : GSSurface(r) + , m_valid(0, 0, 0, 0) + , m_bpp(0) + , m_bpp2(0) + , m_rendered(false) +{ + m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); + + memset(m_clut, 0, sizeof(m_clut)); +} + +GSTextureCache::GSCachedTexture::~GSCachedTexture() +{ + _aligned_free(m_clut); +} + +void GSTextureCache::GSCachedTexture::Update() +{ + __super::Update(); + + if(m_rendered) + { + return; + } + + GSVector4i r; + + if(!GetDirtyRect(r)) + { + return; + } + + m_valid = m_valid.runion(r); + + static uint8* bits = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); + + int pitch = ((r.width() + 3) & ~3) * 4; + + if(m_renderer->m_psrr) + { + m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + else + { + m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + + m_texture->Update(r, bits, pitch); + + m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * m_bpp >> 3); +} + +bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& rr) +{ + int w = 1 << m_TEX0.TW; + int h = 1 << m_TEX0.TH; + + GSVector4i r(0, 0, w, h); + + for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) + { + const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(r); + + if(!m_valid.rintersect(dirty).rempty()) + { + // find the rect having the largest area, outside dirty, inside m_valid + + GSVector4i left(m_valid.left, m_valid.top, min(m_valid.right, dirty.left), m_valid.bottom); + GSVector4i top(m_valid.left, m_valid.top, m_valid.right, min(m_valid.bottom, dirty.top)); + GSVector4i right(max(m_valid.left, dirty.right), m_valid.top, m_valid.right, m_valid.bottom); + GSVector4i bottom(m_valid.left, max(m_valid.top, dirty.bottom), m_valid.right, m_valid.bottom); + + int leftsize = !left.rempty() ? left.width() * left.height() : 0; + int topsize = !top.rempty() ? top.width() * top.height() : 0; + int rightsize = !right.rempty() ? right.width() * right.height() : 0; + int bottomsize = !bottom.rempty() ? bottom.width() * bottom.height() : 0; + + // TODO: sort + + m_valid = + leftsize > 0 ? left : + topsize > 0 ? top : + rightsize > 0 ? right : + bottomsize > 0 ? bottom : + GSVector4i::zero(); + } + } + + m_dirty.clear(); + + m_renderer->MinMaxUV(w, h, r); + + if(GSUtil::IsRectInRect(r, m_valid)) + { + return false; + } + else if(GSUtil::IsRectInRectH(r, m_valid) && (r.left >= m_valid.left || r.right <= m_valid.right)) + { + r.top = m_valid.top; + r.bottom = m_valid.bottom; + if(r.left < m_valid.left) r.right = m_valid.left; + else r.left = m_valid.right; // if(r.right > m_valid.right) + } + else if(GSUtil::IsRectInRectV(r, m_valid) && (r.top >= m_valid.top || r.bottom <= m_valid.bottom)) + { + r.left = m_valid.left; + r.right = m_valid.right; + if(r.top < m_valid.top) r.bottom = m_valid.top; + else r.top = m_valid.bottom; // if(r.bottom > m_valid.bottom) + } + else + { + r = r.runion(m_valid); + } + + if(r.rempty()) + { + return false; + } + + rr = r; + + return true; +} \ No newline at end of file diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 6dd495118c..66daec1834 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -23,44 +23,26 @@ #include "GSRenderer.h" -template class GSTextureCache +class GSTextureCache { - typedef typename Device::Texture Texture; - public: class GSSurface : public GSAlignedClass<16> { protected: - GSRenderer* m_renderer; - - typedef typename Device::Texture Texture; + GSRenderer* m_renderer; public: - Texture m_texture; - Texture m_palette; + GSTexture* m_texture; + GSTexture* m_palette; bool m_initpalette; int m_age; GSDirtyRectList m_dirty; GIFRegTEX0 m_TEX0; - explicit GSSurface(GSRenderer* renderer) - : m_renderer(renderer) - , m_age(0) - , m_initpalette(false) - { - m_TEX0.TBP0 = (uint32)~0; - } + explicit GSSurface(GSRenderer* r); + virtual ~GSSurface(); - virtual ~GSSurface() - { - m_renderer->m_dev.Recycle(m_texture); - m_renderer->m_dev.Recycle(m_palette); - } - - virtual void Update() - { - m_age = 0; - } + virtual void Update(); }; class GSRenderTarget : public GSSurface @@ -68,19 +50,9 @@ public: public: bool m_used; - explicit GSRenderTarget(GSRenderer* renderer) - : GSSurface(renderer) - , m_used(true) - { - } - - virtual bool Create(int w, int h) - { - // FIXME: initial data should be unswizzled from local mem in Update() if dirty - - return m_renderer->m_dev.CreateRenderTarget(m_texture, w, h); - } + explicit GSRenderTarget(GSRenderer* r); + virtual bool Create(int w, int h); virtual void Read(const GSVector4i& r) = 0; }; @@ -89,96 +61,15 @@ public: public: bool m_used; - explicit GSDepthStencil(GSRenderer* renderer) - : GSSurface(renderer) - , m_used(false) - { - } - - virtual bool Create(int w, int h) - { - // FIXME: initial data should be unswizzled from local mem in Update() if dirty - - return m_renderer->m_dev.CreateDepthStencil(m_texture, w, h); - } + explicit GSDepthStencil(GSRenderer* renderer); + virtual bool Create(int w, int h); }; - class GSTexture : public GSSurface + class GSCachedTexture : public GSSurface { protected: - bool GetDirtyRect(GSVector4i& rr) - { - int w = 1 << m_TEX0.TW; - int h = 1 << m_TEX0.TH; - - GSVector4i r(0, 0, w, h); - - for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) - { - const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(r); - - if(!m_valid.rintersect(dirty).rempty()) - { - // find the rect having the largest area, outside dirty, inside m_valid - - GSVector4i left(m_valid.left, m_valid.top, min(m_valid.right, dirty.left), m_valid.bottom); - GSVector4i top(m_valid.left, m_valid.top, m_valid.right, min(m_valid.bottom, dirty.top)); - GSVector4i right(max(m_valid.left, dirty.right), m_valid.top, m_valid.right, m_valid.bottom); - GSVector4i bottom(m_valid.left, max(m_valid.top, dirty.bottom), m_valid.right, m_valid.bottom); - - int leftsize = !left.rempty() ? left.width() * left.height() : 0; - int topsize = !top.rempty() ? top.width() * top.height() : 0; - int rightsize = !right.rempty() ? right.width() * right.height() : 0; - int bottomsize = !bottom.rempty() ? bottom.width() * bottom.height() : 0; - - // TODO: sort - - m_valid = - leftsize > 0 ? left : - topsize > 0 ? top : - rightsize > 0 ? right : - bottomsize > 0 ? bottom : - GSVector4i::zero(); - } - } - - m_dirty.clear(); - - m_renderer->MinMaxUV(w, h, r); - - if(GSUtil::IsRectInRect(r, m_valid)) - { - return false; - } - else if(GSUtil::IsRectInRectH(r, m_valid) && (r.left >= m_valid.left || r.right <= m_valid.right)) - { - r.top = m_valid.top; - r.bottom = m_valid.bottom; - if(r.left < m_valid.left) r.right = m_valid.left; - else r.left = m_valid.right; // if(r.right > m_valid.right) - } - else if(GSUtil::IsRectInRectV(r, m_valid) && (r.top >= m_valid.top || r.bottom <= m_valid.bottom)) - { - r.left = m_valid.left; - r.right = m_valid.right; - if(r.top < m_valid.top) r.bottom = m_valid.top; - else r.top = m_valid.bottom; // if(r.bottom > m_valid.bottom) - } - else - { - r = r.runion(m_valid); - } - - if(r.rempty()) - { - return false; - } - - rr = r; - - return true; - } + bool GetDirtyRect(GSVector4i& r); public: GIFRegCLAMP m_CLAMP; @@ -188,58 +79,10 @@ public: int m_bpp2; bool m_rendered; - explicit GSTexture(GSRenderer* renderer) - : GSSurface(renderer) - , m_valid(0, 0, 0, 0) - , m_bpp(0) - , m_bpp2(0) - , m_rendered(false) - { - m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); + explicit GSCachedTexture(GSRenderer* renderer); + virtual ~GSCachedTexture(); - memset(m_clut, 0, sizeof(m_clut)); - } - - ~GSTexture() - { - _aligned_free(m_clut); - } - - void Update() - { - __super::Update(); - - if(m_rendered) - { - return; - } - - GSVector4i r; - - if(!GetDirtyRect(r)) - { - return; - } - - m_valid = m_valid.runion(r); - - static uint8* bits = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - - int pitch = ((r.width() + 3) & ~3) * 4; - - if(m_renderer->m_psrr) - { - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - else - { - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - - m_texture.Update(r, bits, pitch); - - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * m_bpp >> 3); - } + void Update(); virtual bool Create() = 0; virtual bool Create(GSRenderTarget* rt) = 0; @@ -247,10 +90,11 @@ public: }; protected: - GSRenderer* m_renderer; + GSRenderer* m_renderer; + list m_rt; list m_ds; - list m_tex; + list m_tex; template void RecycleByAge(list& l, int maxage = 60) { @@ -271,586 +115,21 @@ protected: virtual GSRenderTarget* CreateRenderTarget() = 0; virtual GSDepthStencil* CreateDepthStencil() = 0; - virtual GSTexture* CreateTexture() = 0; + virtual GSCachedTexture* CreateTexture() = 0; public: - GSTextureCache(GSRenderer* renderer) - : m_renderer(renderer) - { - } + GSTextureCache(GSRenderer* r); + virtual ~GSTextureCache(); - virtual ~GSTextureCache() - { - RemoveAll(); - } + void RemoveAll(); - void RemoveAll() - { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) - { - delete *i; - } + GSRenderTarget* GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb = false); + GSDepthStencil* GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h); + GSCachedTexture* GetTexture(); - m_rt.clear(); + void InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF); + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); + void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - delete *i; - } - - m_ds.clear(); - - for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) - { - delete *i; - } - - m_tex.clear(); - } - - GSRenderTarget* GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb = false) - { - GSRenderTarget* rt = NULL; - - if(rt == NULL) - { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) - { - GSRenderTarget* rt2 = *i; - - if(rt2->m_TEX0.TBP0 == TEX0.TBP0) - { - m_rt.splice(m_rt.begin(), m_rt, i); - - rt = rt2; - - if(!fb) rt->m_TEX0 = TEX0; - - rt->Update(); - - break; - } - } - } - - if(rt == NULL && fb) - { - // HACK: try to find something close to the base pointer - - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) - { - GSRenderTarget* rt2 = *i; - - if(rt2->m_TEX0.TBP0 <= TEX0.TBP0 && TEX0.TBP0 < rt2->m_TEX0.TBP0 + 0x700 && (!rt || rt2->m_TEX0.TBP0 >= rt->m_TEX0.TBP0)) - { - rt = rt2; - } - } - - if(rt) - { - rt->Update(); - } - } - - if(rt == NULL) - { - rt = CreateRenderTarget(); - - rt->m_TEX0 = TEX0; - - if(!rt->Create(w, h)) - { - delete rt; - - return NULL; - } - - m_rt.push_front(rt); - } - - if(m_renderer->CanUpscale()) - { - GSVector4i fr = m_renderer->GetFrameRect(); - - int ww = (int)(fr.left + rt->m_TEX0.TBW * 64); - int hh = (int)(fr.top + m_renderer->GetDisplayRect().height()); - - if(hh <= m_renderer->GetDeviceSize().y / 2) - { - hh *= 2; - } - - if(ww > 0 && hh > 0) - { - rt->m_texture.m_scale.x = (float)w / ww; - rt->m_texture.m_scale.y = (float)h / hh; - } - } - - if(!fb) - { - rt->m_used = true; - } - - return rt; - } - - GSDepthStencil* GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h) - { - GSDepthStencil* ds = NULL; - - if(ds == NULL) - { - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - GSDepthStencil* ds2 = *i; - - if(ds2->m_TEX0.TBP0 == TEX0.TBP0) - { - m_ds.splice(m_ds.begin(), m_ds, i); - - ds = ds2; - - ds->m_TEX0 = TEX0; - - ds->Update(); - - break; - } - } - } - - if(ds == NULL) - { - ds = CreateDepthStencil(); - - ds->m_TEX0 = TEX0; - - if(!ds->Create(w, h)) - { - delete ds; - - return NULL; - } - - m_ds.push_front(ds); - } - - if(m_renderer->m_context->DepthWrite()) - { - ds->m_used = true; - } - - return ds; - } - - GSTexture* GetTexture() - { - const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; - const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; - - const uint32* clut = m_renderer->m_mem.m_clut; - const int pal = GSLocalMemory::m_psm[TEX0.PSM].pal; - - if(pal > 0) - { - m_renderer->m_mem.m_clut.Read(TEX0); - - /* - POSITION pos = m_tex.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_tex.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_tex.RemoveAt(cur); - - delete s; - } - } - - pos = m_rt.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_rt.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_rt.RemoveAt(cur); - - delete s; - } - } - - pos = m_ds.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_ds.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_ds.RemoveAt(cur); - - delete s; - } - } - */ - } - - GSTexture* t = NULL; - - for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) - { - t = *i; - - if(GSUtil::HasSharedBits(t->m_TEX0.TBP0, t->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) - { - if(TEX0.PSM == t->m_TEX0.PSM && TEX0.TBW == t->m_TEX0.TBW - && TEX0.TW == t->m_TEX0.TW && TEX0.TH == t->m_TEX0.TH - && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.u64 == t->m_CLAMP.u64)) - && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) - { - m_tex.splice(m_tex.begin(), m_tex, i); - - break; - } - } - - t = NULL; - } - - if(t == NULL) - { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) - { - GSRenderTarget* rt = *i; - - if(rt->m_dirty.empty() && GSUtil::HasSharedBits(rt->m_TEX0.TBP0, rt->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) - { - t = CreateTexture(); - - if(!t->Create(rt)) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - - break; - } - } - } - - if(t == NULL) - { - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - GSDepthStencil* ds = *i; - - if(ds->m_dirty.empty() && ds->m_used && GSUtil::HasSharedBits(ds->m_TEX0.TBP0, ds->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) - { - t = CreateTexture(); - - if(!t->Create(ds)) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - - break; - } - } - } - - if(t == NULL) - { - t = CreateTexture(); - - if(!t->Create()) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - } - - if(pal > 0) - { - int size = pal * sizeof(clut[0]); - - if(t->m_palette) - { - if(t->m_initpalette) - { - memcpy(t->m_clut, clut, size); - t->m_palette.Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); - t->m_initpalette = false; - } - else - { - if(GSVector4i::update(t->m_clut, clut, size)) - { - t->m_palette.Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); - } - } - } - else - { - memcpy(t->m_clut, clut, size); - } - } - - t->Update(); - - return t; - } - - void InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) - { - for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) - { - list::iterator j = i++; - - GSTexture* t = *j; - - if(GSUtil::HasSharedBits(FRAME.Block(), FRAME.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM) - || GSUtil::HasSharedBits(ZBUF.Block(), ZBUF.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) - { - m_tex.erase(j); - - delete t; - } - } - } - - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) - { - bool found = false; - - for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) - { - list::iterator j = i++; - - GSTexture* t = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) - { - if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) - { - t->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); - - found = true; - } - else - { - m_tex.erase(j); - delete t; - } - } - else if(GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, t->m_TEX0.PSM)) - { - if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) - { - int rowsize = (int)BITBLTBUF.DBW * 8192; - int offset = ((int)BITBLTBUF.DBP - (int)t->m_TEX0.TBP0) * 256; - - if(rowsize > 0 && offset % rowsize == 0) - { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - - GSVector4i r2(r.left, r.top + y, r.right, r.bottom + y); - - int w = 1 << t->m_TEX0.TW; - int h = 1 << t->m_TEX0.TH; - - if(r2.bottom > 0 && r2.top < h && r2.right > 0 && r2.left < w) - { - t->m_dirty.push_back(GSDirtyRect(r2, BITBLTBUF.DPSM)); - } - } - } - } - } - - for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) - { - list::iterator j = i++; - - GSRenderTarget* rt = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) - { - if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM)) - { - rt->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); - rt->m_TEX0.TBW = BITBLTBUF.DBW; - } - else - { - m_rt.erase(j); - delete rt; - continue; - } - } - - if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM) && BITBLTBUF.DBP < rt->m_TEX0.TBP0) - { - uint32 rowsize = BITBLTBUF.DBW * 8192; - uint32 offset = (uint32)((rt->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); - - if(rowsize > 0 && offset % rowsize == 0) - { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - - if(r.bottom > y) - { - // TODO: do not add this rect above too - rt->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); - rt->m_TEX0.TBW = BITBLTBUF.DBW; - continue; - } - } - } - } - - // copypaste for ds - - for(list::iterator i = m_ds.begin(); i != m_ds.end(); ) - { - list::iterator j = i++; - - GSDepthStencil* ds = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, ds->m_TEX0.TBP0, ds->m_TEX0.PSM)) - { - if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM)) - { - ds->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); - ds->m_TEX0.TBW = BITBLTBUF.DBW; - } - else - { - m_ds.erase(j); - delete ds; - continue; - } - } - - if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM) && BITBLTBUF.DBP < ds->m_TEX0.TBP0) - { - uint32 rowsize = BITBLTBUF.DBW * 8192; - uint32 offset = (uint32)((ds->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); - - if(rowsize > 0 && offset % rowsize == 0) - { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - - if(r.bottom > y) - { - // TODO: do not add this rect above too - ds->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); - ds->m_TEX0.TBW = BITBLTBUF.DBW; - continue; - } - } - } - - } - } - - void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) - { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) - { - list::iterator j = i++; - - GSRenderTarget* rt = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.SBP, BITBLTBUF.SPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) - { - if(GSUtil::HasCompatibleBits(BITBLTBUF.SPSM, rt->m_TEX0.PSM)) - { - rt->Read(r); - return; - } - else if(BITBLTBUF.SPSM == PSM_PSMCT32 && (rt->m_TEX0.PSM == PSM_PSMCT16 || rt->m_TEX0.PSM == PSM_PSMCT16S)) - { - // ffx-2 riku changing to her default (shoots some reflecting glass at the end), 16-bit rt read as 32-bit - - rt->Read(GSVector4i(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); - return; - } - else - { - m_rt.erase(j); - delete rt; - continue; - } - } - } -/* - // no good, ffx does a lot of readback after exiting menu, at 0x02f00 this wrongly finds rt 0x02100 (0,448 - 512,480) - - GSRenderTarget* rt2 = NULL; - int ymin = INT_MAX; - - pos = m_rt.GetHeadPosition(); - - while(pos) - { - GSRenderTarget* rt = m_rt.GetNext(pos); - - if(HasSharedBits(BITBLTBUF.SPSM, rt->m_TEX0.PSM) && BITBLTBUF.SBP > rt->m_TEX0.TBP0) - { - // ffx2 pause screen background - - uint32 rowsize = BITBLTBUF.SBW * 8192; - uint32 offset = (uint32)((BITBLTBUF.SBP - rt->m_TEX0.TBP0) * 256); - - if(rowsize > 0 && offset % rowsize == 0) - { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.SPSM].pgs.y * offset / rowsize; - - if(y < ymin && y < 512) - { - rt2 = rt; - ymin = y; - } - } - } - } - - if(rt2) - { - rt2->Read(GSVector4i(r.left, r.top + ymin, r.right, r.bottom + ymin)); - } - - // TODO: ds -*/ - } - - void IncAge() - { - RecycleByAge(m_tex, 2); - RecycleByAge(m_rt); - RecycleByAge(m_ds); - } + void IncAge(); }; diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 4670658e84..7b27dea300 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -24,8 +24,8 @@ // GSTextureCache10 -GSTextureCache10::GSTextureCache10(GSRenderer* renderer) - : GSTextureCache(renderer) +GSTextureCache10::GSTextureCache10(GSRenderer* r) + : GSTextureCache(r) { } @@ -37,43 +37,41 @@ void GSTextureCache10::GSRenderTargetHW10::Update() // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture.GetSize()); + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); if(r.rempty()) return; int w = r.width(); int h = r.height(); - static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); - static int pitch = 1024 * 4; + if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) + { + static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); + static int pitch = 1024 * 4; - GIFRegTEXA TEXA; + GIFRegTEXA TEXA; - TEXA.AEM = 1; - TEXA.TA0 = 0; - TEXA.TA1 = 0x80; + TEXA.AEM = 1; + TEXA.TA0 = 0; + TEXA.TA1 = 0x80; - GIFRegCLAMP CLAMP; + GIFRegCLAMP CLAMP; - CLAMP.WMS = 0; - CLAMP.WMT = 0; + CLAMP.WMS = 0; + CLAMP.WMT = 0; - m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA, CLAMP); - - // s->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); + m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA, CLAMP); + + // s->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); - Texture texture; + t->Update(GSVector4i(0, 0, w, h), buff, pitch); - if(!m_renderer->m_dev.CreateTexture(texture, w, h)) - return; + GSVector4 dr = GSVector4(r) * GSVector4(m_texture->m_scale).xyxy(); - texture.Update(GSVector4i(0, 0, w, h), buff, pitch); + m_renderer->m_dev->StretchRect(t, m_texture, dr); - GSVector4 dr = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy(); - - m_renderer->m_dev.StretchRect(texture, m_texture, dr); - - m_renderer->m_dev.Recycle(texture); + m_renderer->m_dev->Recycle(t); + } } void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) @@ -99,75 +97,73 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) int w = r.width(); int h = r.height(); - GSVector4 src = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy() / GSVector4(m_texture.GetSize()).xyxy(); + GSVector4 src = GSVector4(r) * GSVector4(m_texture->m_scale).xyxy() / GSVector4(m_texture->GetSize()).xyxy(); DXGI_FORMAT format = m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S ? DXGI_FORMAT_R16_UINT : DXGI_FORMAT_R8G8B8A8_UNORM; - Texture offscreen; - - if(!m_renderer->m_dev.CopyOffscreen(m_texture, src, offscreen, w, h, format)) - return; - - uint8* bits; - int pitch; - - if(offscreen.Map(&bits, pitch)) + if(GSTexture* offscreen = m_renderer->m_dev->CopyOffscreen(m_texture, src, w, h, format)) { - // TODO: block level write + uint8* bits; + int pitch; - uint32 bp = m_TEX0.TBP0; - uint32 bw = m_TEX0.TBW; - - GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; - - if(m_TEX0.PSM == PSM_PSMCT32) + if(offscreen->Map(&bits, pitch)) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) - { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + // TODO: block level write - for(int x = r.left, i = 0; x < r.right; x++, i++) + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; + + GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; + + if(m_TEX0.PSM == PSM_PSMCT32) + { + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + } } } - } - else if(m_TEX0.PSM == PSM_PSMCT24) - { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + else if(m_TEX0.PSM == PSM_PSMCT24) { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; - - for(int x = r.left, i = 0; x < r.right; x++, i++) + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + } } } - } - else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) - { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; - - for(int x = r.left, i = 0; x < r.right; x++, i++) + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)bits)[i]); + } } } - } - else - { - ASSERT(0); + else + { + ASSERT(0); + } + + offscreen->Unmap(); } - offscreen.Unmap(); + m_renderer->m_dev->Recycle(offscreen); } - - m_renderer->m_dev.Recycle(offscreen); } // GSDepthStencilHW10 @@ -181,7 +177,7 @@ void GSTextureCache10::GSDepthStencilHW10::Update() // GSTextureHW10 -bool GSTextureCache10::GSTextureHW10::Create() +bool GSTextureCache10::GSCachedTextureHW10::Create() { // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); @@ -228,11 +224,17 @@ bool GSTextureCache10::GSTextureHW10::Create() int w = 1 << m_TEX0.TW; int h = 1 << m_TEX0.TH; - return m_renderer->m_dev.CreateTexture(m_texture, w, h, format); + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(w, h, format); + + return m_texture != NULL; } -bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) +bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) { + // TODO: clean up this mess + rt->Update(); // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); @@ -247,8 +249,10 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - int w = (int)(rt->m_texture.m_scale.x * tw); - int h = (int)(rt->m_texture.m_scale.y * th); + int w = (int)(rt->m_texture->m_scale.x * tw); + int h = (int)(rt->m_texture->m_scale.y * th); + + GSVector2i rtsize = rt->m_texture->GetSize(); // pitch conversion @@ -258,10 +262,12 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) - m_renderer->m_dev.CreateRenderTarget(m_texture, rt->m_texture.GetWidth(), rt->m_texture.GetHeight()); + ASSERT(m_texture == NULL); - GSVector4 size = GSVector4(rt->m_texture.GetSize()).xyxy(); - GSVector4 scale = GSVector4(rt->m_texture.m_scale).xyxy(); + m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); + + GSVector4 size = GSVector4(rtsize).xyxy(); + GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); int bw = 64; int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; @@ -286,7 +292,7 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; - m_renderer->m_dev.StretchRect(rt->m_texture, src, m_texture, dst); + m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); // TODO: this is quite a lot of StretchRect, do it with one Draw } @@ -304,65 +310,56 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) // width/height conversion - GSVector2 scale = rt->m_texture.m_scale; + GSVector2 scale = rt->m_texture->m_scale; GSVector4 dst(0, 0, w, h); - if(w > rt->m_texture.GetWidth()) + if(w > rtsize.x) { - scale.x = (float)rt->m_texture.GetWidth() / tw; - dst.z = (float)rt->m_texture.GetWidth() * scale.x / rt->m_texture.m_scale.x; - w = rt->m_texture.GetWidth(); + scale.x = (float)rtsize.x / tw; + dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; + w = rtsize.x; } - if(h > rt->m_texture.GetHeight()) + if(h > rtsize.y) { - scale.y = (float)rt->m_texture.GetHeight() / th; - dst.w = (float)rt->m_texture.GetHeight() * scale.y / rt->m_texture.m_scale.y; - h = rt->m_texture.GetHeight(); + scale.y = (float)rtsize.y / th; + dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; + h = rtsize.y; } GSVector4 src(0, 0, w, h); - Texture* st; - Texture* dt; - Texture tmp; + GSTexture* st = m_texture ? m_texture : rt->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); if(!m_texture) { - st = &rt->m_texture; - dt = &m_texture; + m_texture = dt; } - else - { - st = &m_texture; - dt = &tmp; - } - - m_renderer->m_dev.CreateRenderTarget(*dt, w, h); if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) { D3D10_BOX box = {0, 0, 0, w, h, 1}; - m_renderer->m_dev->CopySubresourceRegion(*dt, 0, 0, 0, 0, *st, 0, &box); + (*(GSDevice10*)m_renderer->m_dev)->CopySubresourceRegion(*(GSTexture10*)dt, 0, 0, 0, 0, *(GSTexture10*)st, 0, &box); } else { src.z /= st->GetWidth(); src.w /= st->GetHeight(); - m_renderer->m_dev.StretchRect(*st, src, *dt, dst); + m_renderer->m_dev->StretchRect(st, src, dt, dst); } - if(tmp) + if(dt != m_texture) { - m_renderer->m_dev.Recycle(m_texture); + m_renderer->m_dev->Recycle(m_texture); - m_texture = tmp; + m_texture = dt; } - m_texture.m_scale = scale; + m_texture->m_scale = scale; switch(m_TEX0.PSM) { @@ -378,7 +375,7 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) break; case PSM_PSMT8H: m_bpp2 = 3; - m_renderer->m_dev.CreateTexture(m_palette, 256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? DXGI_FORMAT_R8G8B8A8_UNORM : DXGI_FORMAT_R16_UNORM); // + m_palette = m_renderer->m_dev->CreateTexture(256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? DXGI_FORMAT_R8G8B8A8_UNORM : DXGI_FORMAT_R16_UNORM); // m_initpalette = true; break; case PSM_PSMT4HL: @@ -390,7 +387,7 @@ bool GSTextureCache10::GSTextureHW10::Create(GSRenderTarget* rt) return true; } -bool GSTextureCache10::GSTextureHW10::Create(GSDepthStencil* ds) +bool GSTextureCache10::GSCachedTextureHW10::Create(GSDepthStencil* ds) { m_rendered = true; diff --git a/plugins/GSdx/GSTextureCache10.h b/plugins/GSdx/GSTextureCache10.h index 7076c62594..c2c221cf32 100644 --- a/plugins/GSdx/GSTextureCache10.h +++ b/plugins/GSdx/GSTextureCache10.h @@ -24,14 +24,12 @@ #include "GSTextureCache.h" #include "GSDevice10.h" -class GSTextureCache10 : public GSTextureCache +class GSTextureCache10 : public GSTextureCache { - typedef GSDevice10::Texture Texture; - class GSRenderTargetHW10 : public GSRenderTarget { public: - explicit GSRenderTargetHW10(GSRenderer* renderer) : GSRenderTarget(renderer) {} + explicit GSRenderTargetHW10(GSRenderer* r) : GSRenderTarget(r) {} void Update(); void Read(const GSVector4i& r); @@ -40,15 +38,15 @@ class GSTextureCache10 : public GSTextureCache class GSDepthStencilHW10 : public GSDepthStencil { public: - explicit GSDepthStencilHW10(GSRenderer* renderer) : GSDepthStencil(renderer) {} + explicit GSDepthStencilHW10(GSRenderer* r) : GSDepthStencil(r) {} void Update(); }; - class GSTextureHW10 : public GSTexture + class GSCachedTextureHW10 : public GSCachedTexture { public: - explicit GSTextureHW10(GSRenderer* renderer) : GSTexture(renderer) {} + explicit GSCachedTextureHW10(GSRenderer* r) : GSCachedTexture(r) {} bool Create(); bool Create(GSRenderTarget* rt); @@ -58,8 +56,8 @@ class GSTextureCache10 : public GSTextureCache protected: GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetHW10(m_renderer);} GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilHW10(m_renderer);} - GSTexture* CreateTexture() {return new GSTextureHW10(m_renderer);} + GSCachedTexture* CreateTexture() {return new GSCachedTextureHW10(m_renderer);} public: - GSTextureCache10(GSRenderer* renderer); + GSTextureCache10(GSRenderer* r); }; diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index fa8e028e4f..d83f118d27 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -24,8 +24,8 @@ // GSTextureCache9 -GSTextureCache9::GSTextureCache9(GSRenderer* renderer) - : GSTextureCache(renderer) +GSTextureCache9::GSTextureCache9(GSRenderer* r) + : GSTextureCache(r) { } @@ -37,46 +37,42 @@ void GSTextureCache9::GSRenderTarget9::Update() // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture.GetSize()); + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); if(r.rempty()) return; int w = r.width(); int h = r.height(); - Texture texture; - - if(!m_renderer->m_dev.CreateTexture(texture, w, h)) - return; - - uint8* bits; - int pitch; - - if(texture.Map(&bits, pitch)) + if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) { - GIFRegTEXA TEXA; + uint8* bits; + int pitch; - TEXA.AEM = 1; - TEXA.TA0 = 0; - TEXA.TA1 = 0x80; + if(t->Map(&bits, pitch)) + { + GIFRegTEXA TEXA; - GIFRegCLAMP CLAMP; + TEXA.AEM = 1; + TEXA.TA0 = 0; + TEXA.TA1 = 0x80; - CLAMP.WMS = 0; - CLAMP.WMT = 0; + GIFRegCLAMP CLAMP; - m_renderer->m_mem.ReadTexture(r, bits, pitch, m_TEX0, TEXA, CLAMP); + CLAMP.WMS = 0; + CLAMP.WMT = 0; - texture.Unmap(); - - // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * 4); + m_renderer->m_mem.ReadTexture(r, bits, pitch, m_TEX0, TEXA, CLAMP); - GSVector4 dr = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy(); + t->Unmap(); + + // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * 4); - m_renderer->m_dev.StretchRect(texture, m_texture, dr); + m_renderer->m_dev->StretchRect(t, m_texture, GSVector4(r) * GSVector4(m_texture->m_scale).xyxy()); + } + + m_renderer->m_dev->Recycle(t); } - - m_renderer->m_dev.Recycle(texture); } void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) @@ -102,73 +98,71 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) int w = r.width(); int h = r.height(); - GSVector4 src = GSVector4(r) * GSVector4(m_texture.m_scale).xyxy() / GSVector4(m_texture.GetSize()).xyxy(); + GSVector4 src = GSVector4(r) * GSVector4(m_texture->m_scale).xyxy() / GSVector4(m_texture->GetSize()).xyxy(); - Texture offscreen; - - if(!m_renderer->m_dev.CopyOffscreen(m_texture, src, offscreen, w, h)) - return; - - uint8* bits; - int pitch; - - if(offscreen.Map(&bits, pitch)) + if(GSTexture* offscreen = m_renderer->m_dev->CopyOffscreen(m_texture, src, w, h)) { - // TODO: block level write + uint8* bits; + int pitch; - uint32 bp = m_TEX0.TBP0; - uint32 bw = m_TEX0.TBW; - - GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; - - if(m_TEX0.PSM == PSM_PSMCT32) + if(offscreen->Map(&bits, pitch)) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) - { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + // TODO: block level write - for(int x = r.left, i = 0; x < r.right; x++, i++) + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; + + GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; + + if(m_TEX0.PSM == PSM_PSMCT32) + { + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + } } } - } - else if(m_TEX0.PSM == PSM_PSMCT24) - { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + else if(m_TEX0.PSM == PSM_PSMCT24) { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; - - for(int x = r.left, i = 0; x < r.right; x++, i++) + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + } } } - } - else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) - { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) { - uint32 addr = pa(0, y, bp, bw); - int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; - - for(int x = r.left, i = 0; x < r.right; x++, i++) + for(int y = r.top; y < r.bottom; y++, bits += pitch) { - m_renderer->m_mem.WriteFrame16(addr + offset[x], ((uint32*)bits)[i]); + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WriteFrame16(addr + offset[x], ((uint32*)bits)[i]); + } } } - } - else - { - ASSERT(0); + else + { + ASSERT(0); + } + + offscreen->Unmap(); } - offscreen.Unmap(); + m_renderer->m_dev->Recycle(offscreen); } - - m_renderer->m_dev.Recycle(offscreen); } // GSDepthStencil9 @@ -182,7 +176,7 @@ void GSTextureCache9::GSDepthStencil9::Update() // GSTexture9 -bool GSTextureCache9::GSTexture9::Create() +bool GSTextureCache9::GSCachedTexture9::Create() { // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); @@ -229,11 +223,17 @@ bool GSTextureCache9::GSTexture9::Create() int w = 1 << m_TEX0.TW; int h = 1 << m_TEX0.TH; - return m_renderer->m_dev.CreateTexture(m_texture, w, h, format); + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(w, h, format); + + return m_texture != NULL; } -bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) +bool GSTextureCache9::GSCachedTexture9::Create(GSRenderTarget* rt) { + // TODO: clean up this mess + rt->Update(); // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); @@ -248,8 +248,10 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - int w = (int)(rt->m_texture.m_scale.x * tw); - int h = (int)(rt->m_texture.m_scale.y * th); + int w = (int)(rt->m_texture->m_scale.x * tw); + int h = (int)(rt->m_texture->m_scale.y * th); + + GSVector2i rtsize = rt->m_texture->GetSize(); // pitch conversion @@ -259,10 +261,12 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) - m_renderer->m_dev.CreateRenderTarget(m_texture, rt->m_texture.GetWidth(), rt->m_texture.GetHeight()); + ASSERT(m_texture == NULL); - GSVector4 size = GSVector4(rt->m_texture.GetSize()).xyxy(); - GSVector4 scale = GSVector4(rt->m_texture.m_scale).xyxy(); + m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); + + GSVector4 size = GSVector4(rtsize).xyxy(); + GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); int bw = 64; int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; @@ -287,7 +291,7 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; - m_renderer->m_dev.StretchRect(rt->m_texture, src, m_texture, dst); + m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); // TODO: this is quite a lot of StretchRect, do it with one Draw } @@ -305,65 +309,56 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) // width/height conversion - GSVector2 scale = rt->m_texture.m_scale; + GSVector2 scale = rt->m_texture->m_scale; GSVector4 dst(0, 0, w, h); - if(w > rt->m_texture.GetWidth()) + if(w > rtsize.x) { - scale.x = (float)rt->m_texture.GetWidth() / tw; - dst.z = (float)rt->m_texture.GetWidth() * scale.x / rt->m_texture.m_scale.x; - w = rt->m_texture.GetWidth(); + scale.x = (float)rtsize.x / tw; + dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; + w = rtsize.x; } - if(h > rt->m_texture.GetHeight()) + if(h > rtsize.y) { - scale.y = (float)rt->m_texture.GetHeight() / th; - dst.w = (float)rt->m_texture.GetHeight() * scale.y / rt->m_texture.m_scale.y; - h = rt->m_texture.GetHeight(); + scale.y = (float)rtsize.y / th; + dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; + h = rtsize.y; } GSVector4 src(0, 0, w, h); - Texture* st; - Texture* dt; - Texture tmp; + GSTexture* st = m_texture ? m_texture : rt->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); if(!m_texture) { - st = &rt->m_texture; - dt = &m_texture; + m_texture = dt; } - else - { - st = &m_texture; - dt = &tmp; - } - - m_renderer->m_dev.CreateRenderTarget(*dt, w, h); if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) { GSVector4i r(0, 0, w, h); - m_renderer->m_dev->StretchRect(*st, r, *dt, r, D3DTEXF_POINT); + (*(GSDevice9*)m_renderer->m_dev)->StretchRect(*(GSTexture9*)st, r, *(GSTexture9*)dt, r, D3DTEXF_POINT); } else { src.z /= st->GetWidth(); src.w /= st->GetHeight(); - m_renderer->m_dev.StretchRect(*st, src, *dt, dst); + m_renderer->m_dev->StretchRect(st, src, dt, dst); } - if(tmp) + if(dt != m_texture) { - m_renderer->m_dev.Recycle(m_texture); + m_renderer->m_dev->Recycle(m_texture); - m_texture = tmp; + m_texture = dt; } - m_texture.m_scale = scale; + m_texture->m_scale = scale; switch(m_TEX0.PSM) { @@ -379,7 +374,7 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) break; case PSM_PSMT8H: m_bpp2 = 3; - m_renderer->m_dev.CreateTexture(m_palette, 256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? D3DFMT_A8R8G8B8 : D3DFMT_A1R5G5B5); + m_palette = m_renderer->m_dev->CreateTexture(256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? D3DFMT_A8R8G8B8 : D3DFMT_A1R5G5B5); m_initpalette = true; break; case PSM_PSMT4HL: @@ -391,7 +386,7 @@ bool GSTextureCache9::GSTexture9::Create(GSRenderTarget* rt) return true; } -bool GSTextureCache9::GSTexture9::Create(GSDepthStencil* ds) +bool GSTextureCache9::GSCachedTexture9::Create(GSDepthStencil* ds) { m_rendered = true; diff --git a/plugins/GSdx/GSTextureCache9.h b/plugins/GSdx/GSTextureCache9.h index e7ebef5ea1..1b794db260 100644 --- a/plugins/GSdx/GSTextureCache9.h +++ b/plugins/GSdx/GSTextureCache9.h @@ -24,14 +24,12 @@ #include "GSTextureCache.h" #include "GSDevice9.h" -class GSTextureCache9 : public GSTextureCache +class GSTextureCache9 : public GSTextureCache { - typedef GSDevice9::Texture Texture; - class GSRenderTarget9 : public GSRenderTarget { public: - explicit GSRenderTarget9(GSRenderer* renderer) : GSRenderTarget(renderer) {} + explicit GSRenderTarget9(GSRenderer* r) : GSRenderTarget(r) {} void Update(); void Read(const GSVector4i& r); @@ -40,15 +38,15 @@ class GSTextureCache9 : public GSTextureCache class GSDepthStencil9 : public GSDepthStencil { public: - explicit GSDepthStencil9(GSRenderer* renderer) : GSDepthStencil(renderer) {} + explicit GSDepthStencil9(GSRenderer* r) : GSDepthStencil(r) {} void Update(); }; - class GSTexture9 : public GSTexture + class GSCachedTexture9 : public GSCachedTexture { public: - explicit GSTexture9(GSRenderer* renderer) : GSTexture(renderer) {} + explicit GSCachedTexture9(GSRenderer* r) : GSCachedTexture(r) {} bool Create(); bool Create(GSRenderTarget* rt); @@ -58,8 +56,8 @@ class GSTextureCache9 : public GSTextureCache protected: GSRenderTarget* CreateRenderTarget() {return new GSRenderTarget9(m_renderer);} GSDepthStencil* CreateDepthStencil() {return new GSDepthStencil9(m_renderer);} - GSTexture* CreateTexture() {return new GSTexture9(m_renderer);} + GSCachedTexture* CreateTexture() {return new GSCachedTexture9(m_renderer);} public: - GSTextureCache9(GSRenderer* renderer); + GSTextureCache9(GSRenderer* r); }; \ No newline at end of file diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index e81b5f5d77..9045bf9970 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -260,7 +260,7 @@ bool GSTextureFX10::SetupGS(GSSelector sel) return true; } -bool GSTextureFX10::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, ID3D10ShaderResourceView* tex, ID3D10ShaderResourceView* pal) +bool GSTextureFX10::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) { m_dev->PSSetShaderResources(tex, pal); @@ -383,11 +383,11 @@ void GSTextureFX10::SetupRS(int w, int h, const GSVector4i& scissor) m_dev->RSSet(w, h, &scissor); } -void GSTextureFX10::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv) +void GSTextureFX10::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds) { UpdateOM(dssel, bsel, bf); - m_dev->OMSetRenderTargets(rtv, dsv); + m_dev->OMSetRenderTargets(rt, ds); } void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf) diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 2774b721ba..f87306972d 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -239,10 +239,10 @@ public: bool SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMITIVE_TOPOLOGY prim); bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); bool SetupGS(GSSelector sel); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, ID3D10ShaderResourceView* tex, ID3D10ShaderResourceView* pal); + bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupRS(int w, int h, const GSVector4i& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, ID3D10RenderTargetView* rtv, ID3D10DepthStencilView* dsv); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); void Draw(); }; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index e108c6bd6c..81f356cace 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -46,11 +46,13 @@ bool GSTextureFX9::Create(GSDevice9* dev) return true; } -bool GSTextureFX9::CreateMskFix(GSTexture9& t, uint32 size, uint32 msk, uint32 fix) +GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) { + GSTexture* t = NULL; + uint32 hash = (size << 20) | (msk << 10) | fix; - hash_map::iterator i = m_mskfix.find(hash); + hash_map::iterator i = m_mskfix.find(hash); if(i != m_mskfix.end()) { @@ -58,28 +60,28 @@ bool GSTextureFX9::CreateMskFix(GSTexture9& t, uint32 size, uint32 msk, uint32 f } else { - if(!m_dev->CreateTexture(t, size, 1, D3DFMT_R32F)) - { - return false; - } + t = m_dev->CreateTexture(size, 1, D3DFMT_R32F); - uint8* bits; - int pitch; - - if(t.Map(&bits, pitch)) + if(t) { - for(uint32 i = 0; i < size; i++) + uint8* bits; + int pitch; + + if(t->Map(&bits, pitch)) { - ((float*)bits)[i] = (float)((i & msk) | fix) / size; + for(uint32 i = 0; i < size; i++) + { + ((float*)bits)[i] = (float)((i & msk) | fix) / size; + } + + t->Unmap(); } - t.Unmap(); + m_mskfix[hash] = t; } - - m_mskfix[hash] = t; } - return true; + return t; } bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim) @@ -145,32 +147,26 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) return true; } -bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, IDirect3DTexture9* tex, IDirect3DTexture9* pal, bool psrr) +bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal, bool psrr) { m_dev->PSSetShaderResources(tex, pal); - if(tex && psrr) + if(tex && psrr && (sel.wms == 3 || sel.wmt == 3)) { if(sel.wms == 3) { - D3DSURFACE_DESC desc; - tex->GetLevelDesc(0, &desc); - - GSTexture9 t; - CreateMskFix(t, desc.Width, cb->UMSK, cb->UFIX); - - (*m_dev)->SetTexture(2, t); + if(GSTexture* t = CreateMskFix(tex->GetWidth(), cb->UMSK, cb->UFIX)) + { + (*m_dev)->SetTexture(2, *(GSTexture9*)t); + } } if(sel.wmt == 3) { - D3DSURFACE_DESC desc; - tex->GetLevelDesc(0, &desc); - - GSTexture9 t; - CreateMskFix(t, desc.Height, cb->VMSK, cb->VFIX); - - (*m_dev)->SetTexture(3, t); + if(GSTexture* t = CreateMskFix(tex->GetHeight(), cb->VMSK, cb->VFIX)) + { + (*m_dev)->SetTexture(3, *(GSTexture9*)t); + } } } @@ -279,7 +275,7 @@ void GSTextureFX9::SetupRS(int w, int h, const GSVector4i& scissor) m_dev->RSSet(w, h, &scissor); } -void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds) +void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds) { UpdateOM(dssel, bsel, bf); diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 6cca7aad7b..8a94a24b83 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -155,19 +155,20 @@ private: hash_map m_ps_ss; hash_map m_om_dss; hash_map m_om_bs; - hash_map m_mskfix; + hash_map m_mskfix; + + GSTexture* CreateMskFix(uint32 size, uint32 msk, uint32 fix); public: GSTextureFX9(); bool Create(GSDevice9* dev); - bool CreateMskFix(GSTexture9& t, uint32 size, uint32 msk, uint32 fix); bool SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim); bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, IDirect3DTexture9* tex, IDirect3DTexture9* pal, bool psrr); + bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal, bool psrr); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, bool psrr); void SetupRS(int w, int h, const GSVector4i& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, IDirect3DSurface9* rt, IDirect3DSurface9* ds); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); }; diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index fdeeac7286..75f798cf7b 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -35,6 +35,11 @@ public: return *this; } + + bool operator == (const GSVector2T& v) const + { + return x == v.x && y == v.y; + } }; typedef GSVector2T GSVector2; diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 7cd1211c08..bca134a194 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -22,7 +22,7 @@ @@ -149,7 +149,7 @@ @@ -405,7 +404,7 @@ @@ -784,7 +783,7 @@ @@ -910,7 +909,7 @@ + + + diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index 49d9a216c5..f982943160 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -4,7 +4,7 @@ #pragma once -#pragma warning(disable: 4996 4995 4324 4100) +#pragma warning(disable: 4996 4995 4324 4100 4101) #ifdef _WINDOWS @@ -74,6 +74,8 @@ typedef signed long long int64; #define EXPORT_C extern "C" __declspec(dllexport) void __stdcall #define EXPORT_C_(type) extern "C" __declspec(dllexport) type __stdcall +#define ALIGN_STACK(n) __declspec(align(n)) int __dummy; + #ifndef RESTRICT #ifdef __INTEL_COMPILER #define RESTRICT restrict diff --git a/plugins/GSdx/vsprops/common.vsprops b/plugins/GSdx/vsprops/common.vsprops index 002a838f84..782976e7bf 100644 --- a/plugins/GSdx/vsprops/common.vsprops +++ b/plugins/GSdx/vsprops/common.vsprops @@ -3,6 +3,7 @@ ProjectType="Visual C++" Version="8.00" Name="common" + OutputDirectory="$(SolutionDir)\bin\$(PcsxSubsection)" IntermediateDirectory="$(PlatformName)\$(ConfigurationName)" > + From f9cf9051550a15e5b5689aefb5d2a5b655fab2cf Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 22 May 2009 01:49:39 +0000 Subject: [PATCH 106/296] GSnull: Make it work better with Windows and LilyPad. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1237 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 721 +++++++++++++++++---------------- plugins/GSnull/null/GSnull.cpp | 2 +- 2 files changed, 379 insertions(+), 344 deletions(-) diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index 92385b393c..d23c07cdf1 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -1,344 +1,379 @@ -/* GSnull - * Copyright (C) 2004-2009 PCSX2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include -#include -#include -#include - -#include -#include - -using namespace std; - -#include "GS.h" -#include "null/GSnull.h" - -#ifdef __LINUX__ -Display *display; -int screen; -#endif -#ifdef _WIN32 -HINSTANCE HInst; -#endif - -const unsigned char version = PS2E_GS_VERSION; -const unsigned char revision = 0; -const unsigned char build = 1; // increase that with each version - -static char *libraryName = "GSnull Driver"; -FILE *gsLog; -Config conf; -u32 GSKeyEvent = 0; -bool GSShift = false, GSAlt = false; - -string s_strIniPath="inis/GSnull.ini"; -void (*GSirq)(); - -EXPORT_C_(u32) PS2EgetLibType() -{ - return PS2E_LT_GS; -} - -EXPORT_C_(char*) PS2EgetLibName() -{ - return libraryName; -} - -EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) -{ - return (version<<16) | (revision<<8) | build; -} - -void __Log(char *fmt, ...) -{ - va_list list; - - if (!conf.Log || gsLog == NULL) return; - - va_start(list, fmt); - vfprintf(gsLog, fmt, list); - va_end(list); -} - -EXPORT_C_(void) GSprintf(int timeout, char *fmt, ...) -{ - va_list list; - char msg[512]; - - va_start(list, fmt); - vsprintf(msg, fmt, list); - va_end(list); - - GS_LOG("GSprintf:%s", msg); - printf("GSprintf:%s", msg); -} - -void SysPrintf(const char *fmt, ...) -{ - va_list list; - char msg[512]; - - va_start(list, fmt); - vsprintf(msg, fmt, list); - va_end(list); - - GS_LOG(msg); - printf("GSnull:%s", msg); -} - -// basic funcs - -EXPORT_C_(s32) GSinit() -{ - LoadConfig(); - -#ifdef GS_LOG - gsLog = fopen("logs/gsLog.txt", "w"); - if (gsLog) setvbuf(gsLog, NULL, _IONBF, 0); - GS_LOG("GSnull plugin version %d,%d\n",revision,build); - GS_LOG("GS init\n"); -#endif - - SysPrintf("Initializing GSnull\n"); - return 0; -} - -EXPORT_C_(void) GSshutdown() -{ -#ifdef GS_LOG - if (gsLog) fclose(gsLog); -#endif - - SysPrintf("Shutting down GSnull\n"); -} - -EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) -{ -#ifdef GS_LOG - GS_LOG("GS open\n"); -#endif - //assert( GSirq != NULL ); - -#ifdef __LINUX__ - display = XOpenDisplay(0); - screen = DefaultScreen(display); - - if( pDsp != NULL ) *(Display**)pDsp = display; -#else - if( pDsp != NULL ) *(int*)pDsp = (int)GetNextWindow(GetForegroundWindow(),GW_HWNDNEXT); -(NULL); -#endif - SysPrintf("Opening GSnull\n"); - return 0; -} - -EXPORT_C_(void) GSclose() -{ - SysPrintf("Closing GSnull\n"); -#ifdef __LINUX__ - XCloseDisplay(display); -#endif -} - -EXPORT_C_(void) GSirqCallback(void (*callback)()) -{ - GSirq = callback; -} - -EXPORT_C_(s32) GSfreeze(int mode, freezeData *data) -{ - return 0; -} - -EXPORT_C_(s32) GStest() -{ - SysPrintf("Testing GSnull\n"); - return 0; -} - -void ProcessMessages() -{ -#ifdef __LINUX__ - if ( GSKeyEvent ) - { - int myKeyEvent = GSKeyEvent; - bool myShift = GSShift; - GSKeyEvent = 0; - - switch ( myKeyEvent ) - { - case XK_F5: - OnKeyboardF5(myShift); - break; - case XK_F6: - OnKeyboardF6(myShift); - break; - case XK_F7: - OnKeyboardF7(myShift); - break; - case XK_F9: - OnKeyboardF9(myShift); - break; - } - } -#endif -} - -EXPORT_C_(void) GSvsync(int field) -{ - ProcessMessages(); -} - -EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) -{ - _GSgifTransfer1(pMem, addr); -} - -EXPORT_C_(void) GSgifTransfer2(u32 *pMem, u32 size) -{ - _GSgifTransfer2(pMem, size); -} - -EXPORT_C_(void) GSgifTransfer3(u32 *pMem, u32 size) -{ - _GSgifTransfer3(pMem, size); -} - - // returns the last tag processed (64 bits) -EXPORT_C_(void) GSgetLastTag(u64* ptag) -{ -} - -EXPORT_C_(void) GSgifSoftReset(u32 mask) -{ - SysPrintf("Doing a soft reset of the GS plugin.\n"); -} - -EXPORT_C_(void) GSreadFIFO(u64 *mem) -{ -} - -EXPORT_C_(void) GSreadFIFO2(u64 *mem, int qwc) -{ -} - -// extended funcs - -// GSkeyEvent gets called when there is a keyEvent from the PAD plugin -EXPORT_C_(void) GSkeyEvent(keyEvent *ev) -{ - switch(ev->evt) { - case KEYPRESS: - switch(ev->key) { - case XK_F5: - case XK_F6: - case XK_F7: - case XK_F9: - GSKeyEvent = ev->key ; - break; - case XK_Escape: - break; - case XK_Shift_L: - case XK_Shift_R: - //bShift = true; - GSShift = true; - break; - case XK_Alt_L: - case XK_Alt_R: - GSAlt = true; - break; - } - break; - case KEYRELEASE: - switch(ev->key) { - case XK_Shift_L: - case XK_Shift_R: - //bShift = false; - GSShift = false; - break; - case XK_Alt_L: - case XK_Alt_R: - GSAlt = false; - break; - } - } -} - -EXPORT_C_(void) GSchangeSaveState(int, const char* filename) -{ -} - -EXPORT_C_(void) GSmakeSnapshot(char *path) -{ - - SysPrintf("Taking a snapshot.\n"); -} - -EXPORT_C_(void) GSmakeSnapshot2(char *pathname, int* snapdone, int savejpg) -{ - SysPrintf("Taking a snapshot to %s.\n", pathname); -} - -EXPORT_C_(void) GSsetBaseMem(void*) -{ -} - -EXPORT_C_(void) GSsetGameCRC(int crc, int gameoptions) -{ - SysPrintf("Setting the crc to '%x' with 0x%x for options.\n", crc, gameoptions); -} - -// controls frame skipping in the GS, if this routine isn't present, frame skipping won't be done -EXPORT_C_(void) GSsetFrameSkip(int frameskip) -{ - SysPrintf("Frameskip set to %d.\n", frameskip); -} - -// if start is 1, starts recording spu2 data, else stops -// returns a non zero value if successful -// for now, pData is not used -EXPORT_C_(int) GSsetupRecording(int start, void* pData) -{ - if (start) - SysPrintf("Pretending to record.\n"); - else - SysPrintf("Pretending to stop recording.\n"); - - return 1; -} - -EXPORT_C_(void) GSreset() -{ - SysPrintf("Doing a reset of the GS plugin."); -} - -EXPORT_C_(void) GSwriteCSR(u32 value) -{ -} - -EXPORT_C_(void) GSgetDriverInfo(GSdriverInfo *info) -{ -} - -#ifdef _WIN32 -EXPORT_C_(s32) GSsetWindowInfo(winInfo *info) -{ - return 0; -} +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include + +#include +#include + +using namespace std; + +#include "GS.h" +#include "null/GSnull.h" + +#ifdef __LINUX__ +Display *display; +int screen; +#endif +#ifdef _WIN32 +HINSTANCE HInst; +HWND GShwnd; +#endif + +const unsigned char version = PS2E_GS_VERSION; +const unsigned char revision = 0; +const unsigned char build = 1; // increase that with each version + +static char *libraryName = "GSnull Driver"; +FILE *gsLog; +Config conf; +u32 GSKeyEvent = 0; +bool GSShift = false, GSAlt = false; + +string s_strIniPath="inis/GSnull.ini"; +void (*GSirq)(); + +EXPORT_C_(u32) PS2EgetLibType() +{ + return PS2E_LT_GS; +} + +EXPORT_C_(char*) PS2EgetLibName() +{ + return libraryName; +} + +EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) +{ + return (version<<16) | (revision<<8) | build; +} + +void __Log(char *fmt, ...) +{ + va_list list; + + if (!conf.Log || gsLog == NULL) return; + + va_start(list, fmt); + vfprintf(gsLog, fmt, list); + va_end(list); +} + +EXPORT_C_(void) GSprintf(int timeout, char *fmt, ...) +{ + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + GS_LOG("GSprintf:%s", msg); + printf("GSprintf:%s", msg); +} + +void SysPrintf(const char *fmt, ...) +{ + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + GS_LOG(msg); + printf("GSnull:%s", msg); +} + +// basic funcs + +EXPORT_C_(s32) GSinit() +{ + LoadConfig(); + +#ifdef GS_LOG + gsLog = fopen("logs/gsLog.txt", "w"); + if (gsLog) setvbuf(gsLog, NULL, _IONBF, 0); + GS_LOG("GSnull plugin version %d,%d\n",revision,build); + GS_LOG("GS init\n"); +#endif + + SysPrintf("Initializing GSnull\n"); + return 0; +} + +EXPORT_C_(void) GSshutdown() +{ +#ifdef GS_LOG + if (gsLog) fclose(gsLog); +#endif + + SysPrintf("Shutting down GSnull\n"); +} + +#ifndef __LINUX__ +LRESULT CALLBACK MsgProc(HWND hwnd, UINT msg, WPARAM wParam, LPARAM lParam) +{ + switch(msg) + { + case WM_CLOSE: + DestroyWindow(hwnd); + break; + case WM_DESTROY: + PostQuitMessage(0); + break; + default: + return DefWindowProc(hwnd, msg, wParam, lParam); + } + return 0; +} + +#endif + +EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) +{ +#ifdef GS_LOG + GS_LOG("GS open\n"); +#endif + //assert( GSirq != NULL ); + +#ifdef __LINUX__ + display = XOpenDisplay(0); + screen = DefaultScreen(display); + + if( pDsp != NULL ) *(Display**)pDsp = display; +#else + WNDCLASSEX wc = { sizeof(WNDCLASSEX), CS_CLASSDC, MsgProc, 0L, 0L, + GetModuleHandle(NULL), NULL, NULL, NULL, NULL, + "PS2EMU_GSNULL", NULL }; + RegisterClassEx( &wc ); + + GShwnd = CreateWindowEx( WS_EX_CLIENTEDGE, "PS2EMU_GSNULL", "The title of my window", + WS_OVERLAPPEDWINDOW, CW_USEDEFAULT, CW_USEDEFAULT, 240, 120, NULL, NULL, wc.hInstance, NULL); + + if(GShwnd == NULL) + { + GS_LOG("Failed to create window. Exiting..."); + return -1; + } + + if( pDsp != NULL ) *(int*)pDsp = (int)GShwnd; +#endif + SysPrintf("Opening GSnull\n"); + return 0; +} + +EXPORT_C_(void) GSclose() +{ + SysPrintf("Closing GSnull\n"); +#ifdef __LINUX__ + XCloseDisplay(display); +#endif +} + +EXPORT_C_(void) GSirqCallback(void (*callback)()) +{ + GSirq = callback; +} + +EXPORT_C_(s32) GSfreeze(int mode, freezeData *data) +{ + return 0; +} + +EXPORT_C_(s32) GStest() +{ + SysPrintf("Testing GSnull\n"); + return 0; +} + +void ProcessMessages() +{ +#ifdef __LINUX__ + if ( GSKeyEvent ) + { + int myKeyEvent = GSKeyEvent; + bool myShift = GSShift; + GSKeyEvent = 0; + + switch ( myKeyEvent ) + { + case XK_F5: + OnKeyboardF5(myShift); + break; + case XK_F6: + OnKeyboardF6(myShift); + break; + case XK_F7: + OnKeyboardF7(myShift); + break; + case XK_F9: + OnKeyboardF9(myShift); + break; + } + } +#endif +} + +EXPORT_C_(void) GSvsync(int field) +{ + ProcessMessages(); +} + +EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) +{ + _GSgifTransfer1(pMem, addr); +} + +EXPORT_C_(void) GSgifTransfer2(u32 *pMem, u32 size) +{ + _GSgifTransfer2(pMem, size); +} + +EXPORT_C_(void) GSgifTransfer3(u32 *pMem, u32 size) +{ + _GSgifTransfer3(pMem, size); +} + + // returns the last tag processed (64 bits) +EXPORT_C_(void) GSgetLastTag(u64* ptag) +{ +} + +EXPORT_C_(void) GSgifSoftReset(u32 mask) +{ + SysPrintf("Doing a soft reset of the GS plugin.\n"); +} + +EXPORT_C_(void) GSreadFIFO(u64 *mem) +{ +} + +EXPORT_C_(void) GSreadFIFO2(u64 *mem, int qwc) +{ +} + +// extended funcs + +// GSkeyEvent gets called when there is a keyEvent from the PAD plugin +EXPORT_C_(void) GSkeyEvent(keyEvent *ev) +{ +#ifdef __LINUX__ + switch(ev->evt) { + case KEYPRESS: + switch(ev->key) { + case XK_F5: + case XK_F6: + case XK_F7: + case XK_F9: + GSKeyEvent = ev->key ; + break; + case XK_Escape: + break; + case XK_Shift_L: + case XK_Shift_R: + //bShift = true; + GSShift = true; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = true; + break; + } + break; + case KEYRELEASE: + switch(ev->key) { + case XK_Shift_L: + case XK_Shift_R: + //bShift = false; + GSShift = false; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = false; + break; + } + } +#endif +} + +EXPORT_C_(void) GSchangeSaveState(int, const char* filename) +{ +} + +EXPORT_C_(void) GSmakeSnapshot(char *path) +{ + + SysPrintf("Taking a snapshot.\n"); +} + +EXPORT_C_(void) GSmakeSnapshot2(char *pathname, int* snapdone, int savejpg) +{ + SysPrintf("Taking a snapshot to %s.\n", pathname); +} + +EXPORT_C_(void) GSsetBaseMem(void*) +{ +} + +EXPORT_C_(void) GSsetGameCRC(int crc, int gameoptions) +{ + SysPrintf("Setting the crc to '%x' with 0x%x for options.\n", crc, gameoptions); +} + +// controls frame skipping in the GS, if this routine isn't present, frame skipping won't be done +EXPORT_C_(void) GSsetFrameSkip(int frameskip) +{ + SysPrintf("Frameskip set to %d.\n", frameskip); +} + +// if start is 1, starts recording spu2 data, else stops +// returns a non zero value if successful +// for now, pData is not used +EXPORT_C_(int) GSsetupRecording(int start, void* pData) +{ + if (start) + SysPrintf("Pretending to record.\n"); + else + SysPrintf("Pretending to stop recording.\n"); + + return 1; +} + +EXPORT_C_(void) GSreset() +{ + SysPrintf("Doing a reset of the GS plugin."); +} + +EXPORT_C_(void) GSwriteCSR(u32 value) +{ +} + +EXPORT_C_(void) GSgetDriverInfo(GSdriverInfo *info) +{ +} + +#ifdef _WIN32 +EXPORT_C_(s32) GSsetWindowInfo(winInfo *info) +{ + return 0; +} #endif \ No newline at end of file diff --git a/plugins/GSnull/null/GSnull.cpp b/plugins/GSnull/null/GSnull.cpp index 78d848f517..c7a22b64bf 100644 --- a/plugins/GSnull/null/GSnull.cpp +++ b/plugins/GSnull/null/GSnull.cpp @@ -16,7 +16,7 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - #include "GS.h" + #include "../GS.h" void OnKeyboardF5(int myShift) { From 82f0143d23f5b87655b923cb7dfb3ba8ef07e5de Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 22 May 2009 02:29:01 +0000 Subject: [PATCH 107/296] GSnull: Modularise things a bit more, so we don't have a bunch of platform-specific code in the main code. (Note: the Linux side hasn't been tested yet. I'll be doing that a bit later.) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1238 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.cpp | 132 +------------------- plugins/GSnull/GS.h | 11 +- plugins/GSnull/Linux/GSLinux.cpp | 104 +++++++++++++++ plugins/GSnull/Linux/GSLinux.h | 27 ++++ plugins/GSnull/Makefile.am | 12 +- plugins/GSnull/Windows/GSnull_vc2008.vcproj | 58 ++++++++- plugins/GSnull/Windows/GSwin.cpp | 77 ++++++++++++ plugins/GSnull/Windows/GSwin.h | 36 ++++++ 8 files changed, 312 insertions(+), 145 deletions(-) create mode 100644 plugins/GSnull/Linux/GSLinux.cpp create mode 100644 plugins/GSnull/Linux/GSLinux.h create mode 100644 plugins/GSnull/Windows/GSwin.cpp create mode 100644 plugins/GSnull/Windows/GSwin.h diff --git a/plugins/GSnull/GS.cpp b/plugins/GSnull/GS.cpp index d23c07cdf1..b524ce8349 100644 --- a/plugins/GSnull/GS.cpp +++ b/plugins/GSnull/GS.cpp @@ -29,15 +29,6 @@ using namespace std; #include "GS.h" #include "null/GSnull.h" -#ifdef __LINUX__ -Display *display; -int screen; -#endif -#ifdef _WIN32 -HINSTANCE HInst; -HWND GShwnd; -#endif - const unsigned char version = PS2E_GS_VERSION; const unsigned char revision = 0; const unsigned char build = 1; // increase that with each version @@ -129,64 +120,24 @@ EXPORT_C_(void) GSshutdown() SysPrintf("Shutting down GSnull\n"); } -#ifndef __LINUX__ -LRESULT CALLBACK MsgProc(HWND hwnd, UINT msg, WPARAM wParam, LPARAM lParam) -{ - switch(msg) - { - case WM_CLOSE: - DestroyWindow(hwnd); - break; - case WM_DESTROY: - PostQuitMessage(0); - break; - default: - return DefWindowProc(hwnd, msg, wParam, lParam); - } - return 0; -} - -#endif - EXPORT_C_(s32) GSopen(void *pDsp, char *Title, int multithread) { + int err = 0; #ifdef GS_LOG GS_LOG("GS open\n"); #endif //assert( GSirq != NULL ); -#ifdef __LINUX__ - display = XOpenDisplay(0); - screen = DefaultScreen(display); + err = GSOpenWindow(pDsp, Title); - if( pDsp != NULL ) *(Display**)pDsp = display; -#else - WNDCLASSEX wc = { sizeof(WNDCLASSEX), CS_CLASSDC, MsgProc, 0L, 0L, - GetModuleHandle(NULL), NULL, NULL, NULL, NULL, - "PS2EMU_GSNULL", NULL }; - RegisterClassEx( &wc ); - - GShwnd = CreateWindowEx( WS_EX_CLIENTEDGE, "PS2EMU_GSNULL", "The title of my window", - WS_OVERLAPPEDWINDOW, CW_USEDEFAULT, CW_USEDEFAULT, 240, 120, NULL, NULL, wc.hInstance, NULL); - - if(GShwnd == NULL) - { - GS_LOG("Failed to create window. Exiting..."); - return -1; - } - - if( pDsp != NULL ) *(int*)pDsp = (int)GShwnd; -#endif SysPrintf("Opening GSnull\n"); - return 0; + return err; } EXPORT_C_(void) GSclose() { SysPrintf("Closing GSnull\n"); -#ifdef __LINUX__ - XCloseDisplay(display); -#endif + GSCloseWindow(); } EXPORT_C_(void) GSirqCallback(void (*callback)()) @@ -205,37 +156,9 @@ EXPORT_C_(s32) GStest() return 0; } -void ProcessMessages() -{ -#ifdef __LINUX__ - if ( GSKeyEvent ) - { - int myKeyEvent = GSKeyEvent; - bool myShift = GSShift; - GSKeyEvent = 0; - - switch ( myKeyEvent ) - { - case XK_F5: - OnKeyboardF5(myShift); - break; - case XK_F6: - OnKeyboardF6(myShift); - break; - case XK_F7: - OnKeyboardF7(myShift); - break; - case XK_F9: - OnKeyboardF9(myShift); - break; - } - } -#endif -} - EXPORT_C_(void) GSvsync(int field) { - ProcessMessages(); + GSProcessMessages(); } EXPORT_C_(void) GSgifTransfer1(u32 *pMem, u32 addr) @@ -276,43 +199,7 @@ EXPORT_C_(void) GSreadFIFO2(u64 *mem, int qwc) // GSkeyEvent gets called when there is a keyEvent from the PAD plugin EXPORT_C_(void) GSkeyEvent(keyEvent *ev) { -#ifdef __LINUX__ - switch(ev->evt) { - case KEYPRESS: - switch(ev->key) { - case XK_F5: - case XK_F6: - case XK_F7: - case XK_F9: - GSKeyEvent = ev->key ; - break; - case XK_Escape: - break; - case XK_Shift_L: - case XK_Shift_R: - //bShift = true; - GSShift = true; - break; - case XK_Alt_L: - case XK_Alt_R: - GSAlt = true; - break; - } - break; - case KEYRELEASE: - switch(ev->key) { - case XK_Shift_L: - case XK_Shift_R: - //bShift = false; - GSShift = false; - break; - case XK_Alt_L: - case XK_Alt_R: - GSAlt = false; - break; - } - } -#endif + HandleKeyEvent(ev); } EXPORT_C_(void) GSchangeSaveState(int, const char* filename) @@ -370,10 +257,3 @@ EXPORT_C_(void) GSwriteCSR(u32 value) EXPORT_C_(void) GSgetDriverInfo(GSdriverInfo *info) { } - -#ifdef _WIN32 -EXPORT_C_(s32) GSsetWindowInfo(winInfo *info) -{ - return 0; -} -#endif \ No newline at end of file diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h index a322279618..58877bac70 100644 --- a/plugins/GSnull/GS.h +++ b/plugins/GSnull/GS.h @@ -22,16 +22,9 @@ #include #ifdef _WIN32 - -#include -#include - +#include "Windows/GSwin.h" #else - -#include -#include -#include - +#include "Linux/GSLinux.h" #endif #ifdef __cplusplus diff --git a/plugins/GSnull/Linux/GSLinux.cpp b/plugins/GSnull/Linux/GSLinux.cpp new file mode 100644 index 0000000000..34c7d5149c --- /dev/null +++ b/plugins/GSnull/Linux/GSLinux.cpp @@ -0,0 +1,104 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "GSLinux.h" + +Display *display; +int screen; + +int GSOpenWindow(void *pDsp, char *Title) +{ + display = XOpenDisplay(0); + screen = DefaultScreen(display); + + if( pDsp != NULL ) + *(Display**)pDsp = display; + else + return -1; + + return 0; +} + +void GSCloseWindow() +{ + XCloseDisplay(display); +} + +void GSProcessMessages() +{ + if ( GSKeyEvent ) + { + int myKeyEvent = GSKeyEvent; + bool myShift = GSShift; + GSKeyEvent = 0; + + switch ( myKeyEvent ) + { + case XK_F5: + OnKeyboardF5(myShift); + break; + case XK_F6: + OnKeyboardF6(myShift); + break; + case XK_F7: + OnKeyboardF7(myShift); + break; + case XK_F9: + OnKeyboardF9(myShift); + break; + } + } +} + + +void HandleKeyEvent(keyEvent *ev) +{ + switch(ev->evt) { + case KEYPRESS: + switch(ev->key) { + case XK_F5: + case XK_F6: + case XK_F7: + case XK_F9: + GSKeyEvent = ev->key ; + break; + case XK_Escape: + break; + case XK_Shift_L: + case XK_Shift_R: + GSShift = true; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = true; + break; + } + break; + case KEYRELEASE: + switch(ev->key) { + case XK_Shift_L: + case XK_Shift_R: + GSShift = false; + break; + case XK_Alt_L: + case XK_Alt_R: + GSAlt = false; + break; + } + } +} diff --git a/plugins/GSnull/Linux/GSLinux.h b/plugins/GSnull/Linux/GSLinux.h new file mode 100644 index 0000000000..99a686efd6 --- /dev/null +++ b/plugins/GSnull/Linux/GSLinux.h @@ -0,0 +1,27 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include "../GS.h" + +int GSOpenWindow(void *pDsp, char *Title); +void GSCloseWindow(); +void GSProcessMessages(); +void HandleKeyEvent(keyEvent *ev); \ No newline at end of file diff --git a/plugins/GSnull/Makefile.am b/plugins/GSnull/Makefile.am index 21f39566fa..da1ecce6a1 100644 --- a/plugins/GSnull/Makefile.am +++ b/plugins/GSnull/Makefile.am @@ -22,12 +22,14 @@ libGSnull_LDFLAGS= @SHARED_LDFLAGS@ libGSnull_LDFLAGS+=-Wl,-soname,@libGSnull_SONAME@ libGSnull_LDADD=$(libGSnull_a_OBJECTS) -libGSnull_a_SOURCES = GS.cpp Linux/Config.cpp Linux/Linux.cpp \ -FW.h Linux/Config.h Linux/Linux.h +libGSnull_a_SOURCES = \ +GS.cpp \ +Linux/Config.cpp Linux/Linux.cpp Linux/GSLinux.cpp +Linux/GSLinux.h Linux/Config.h Linux/Linux.h \ +GifTransfer.cpp GifTransfer.h \ +null/GSnull.cpp null/GSnull.h libGSnull_a_SOURCES += \ Linux/interface.h Linux/support.c \ Linux/interface.c Linux/support.h \ -Linux/callbacks.h \ -GifTransfer.cpp GifTransfer.h \ -null/GSnull.cpp null/GSnull.h \ No newline at end of file +Linux/callbacks.h \ No newline at end of file diff --git a/plugins/GSnull/Windows/GSnull_vc2008.vcproj b/plugins/GSnull/Windows/GSnull_vc2008.vcproj index 5772e553be..95b476419f 100644 --- a/plugins/GSnull/Windows/GSnull_vc2008.vcproj +++ b/plugins/GSnull/Windows/GSnull_vc2008.vcproj @@ -160,16 +160,40 @@ RelativePath=".\Config.cpp" > + + + + + + + + + + - - + + + + + + + + + + + + diff --git a/plugins/GSnull/Windows/GSwin.cpp b/plugins/GSnull/Windows/GSwin.cpp new file mode 100644 index 0000000000..fa3c66a335 --- /dev/null +++ b/plugins/GSnull/Windows/GSwin.cpp @@ -0,0 +1,77 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "GSwin.h" + +HINSTANCE HInst; +HWND GShwnd; + +LRESULT CALLBACK MsgProc(HWND hwnd, UINT msg, WPARAM wParam, LPARAM lParam) +{ + switch(msg) + { + case WM_CLOSE: + DestroyWindow(hwnd); + break; + case WM_DESTROY: + PostQuitMessage(0); + break; + default: + return DefWindowProc(hwnd, msg, wParam, lParam); + } + return 0; +} + +int GSOpenWindow(void *pDsp, char *Title) +{ + WNDCLASSEX wc = { sizeof(WNDCLASSEX), CS_CLASSDC, MsgProc, 0L, 0L, + GetModuleHandle(NULL), NULL, NULL, NULL, NULL, + "PS2EMU_GSNULL", NULL }; + RegisterClassEx( &wc ); + + GShwnd = CreateWindowEx( WS_EX_CLIENTEDGE, "PS2EMU_GSNULL", Title, + WS_OVERLAPPEDWINDOW, CW_USEDEFAULT, CW_USEDEFAULT, 240, 120, NULL, NULL, wc.hInstance, NULL); + + if(GShwnd == NULL) + { + GS_LOG("Failed to create window. Exiting..."); + return -1; + } + + if( pDsp != NULL ) *(int*)pDsp = (int)GShwnd; + + return 0; +} + +void GSCloseWindow() +{ +} + +void GSProcessMessages() +{ +} + +// GSkeyEvent gets called when there is a keyEvent from the PAD plugin +void HandleKeyEvent(keyEvent *ev) +{ +} + +EXPORT_C_(s32) GSsetWindowInfo(winInfo *info) +{ + return 0; +} \ No newline at end of file diff --git a/plugins/GSnull/Windows/GSwin.h b/plugins/GSnull/Windows/GSwin.h new file mode 100644 index 0000000000..8e7777fe87 --- /dev/null +++ b/plugins/GSnull/Windows/GSwin.h @@ -0,0 +1,36 @@ +/* GSnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include "../GS.h" + +#ifdef __cplusplus +extern "C" +{ +#endif +#define GSdefs +#include "PS2Edefs.h" +#ifdef __cplusplus +} +#endif + +int GSOpenWindow(void *pDsp, char *Title); +void GSCloseWindow(); +void GSProcessMessages(); +void HandleKeyEvent(keyEvent *ev); \ No newline at end of file From 8d119f3928688106d65df9d90126277d421b054a Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 22 May 2009 23:23:38 +0000 Subject: [PATCH 108/296] GSdx: more source cleanup and something interesting on F7 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1239 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 23 +++++------------------ plugins/GSdx/GPULocalMemory.cpp | 7 ++++--- plugins/GSdx/GPULocalMemory.h | 2 +- plugins/GSdx/GPURenderer.cpp | 13 +++++++------ plugins/GSdx/GPURenderer.h | 21 +++++++++------------ plugins/GSdx/GPURendererSW.cpp | 9 ++++----- plugins/GSdx/GPURendererSW.h | 2 +- plugins/GSdx/GPUState.cpp | 5 ++--- plugins/GSdx/GPUState.h | 2 +- plugins/GSdx/GS.cpp | 30 ++++++++---------------------- plugins/GSdx/GSDevice.cpp | 10 ++++++---- plugins/GSdx/GSDevice.h | 6 +++--- plugins/GSdx/GSDevice10.cpp | 11 ++--------- plugins/GSdx/GSDevice10.h | 4 ++-- plugins/GSdx/GSDevice7.cpp | 2 +- plugins/GSdx/GSDevice7.h | 2 +- plugins/GSdx/GSDevice9.cpp | 4 ++-- plugins/GSdx/GSDevice9.h | 4 ++-- plugins/GSdx/GSRenderer.cpp | 25 ++++++++++++------------- plugins/GSdx/GSRenderer.h | 26 +++++++++++--------------- plugins/GSdx/GSRendererHW.h | 8 ++++---- plugins/GSdx/GSRendererHW10.cpp | 7 ++----- plugins/GSdx/GSRendererHW10.h | 2 +- plugins/GSdx/GSRendererHW9.cpp | 7 ++----- plugins/GSdx/GSRendererHW9.h | 2 +- plugins/GSdx/GSRendererNull.h | 6 ++---- plugins/GSdx/GSRendererSW.cpp | 8 +++----- plugins/GSdx/GSRendererSW.h | 2 +- plugins/GSdx/GSUtil.cpp | 2 +- plugins/GSdx/res/convert10.fx | 28 ++++++++++++++++++++++++++++ plugins/GSdx/res/convert9.fx | 27 +++++++++++++++++++++++++++ 31 files changed, 156 insertions(+), 151 deletions(-) diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index 8df8d3a4bc..d4f1501aa7 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -85,11 +85,6 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) { GPUclose(); - GPURendererSettings rs; - - int threads = 1; - int renderer = 1; - #ifdef _WINDOWS AFX_MANAGE_STATE(AfxGetStaticModuleState()); @@ -108,23 +103,15 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) return -1; } - rs.m_filter = theApp.GetConfig("filter", 0); - rs.m_dither = theApp.GetConfig("dithering", 1); - rs.m_aspectratio = theApp.GetConfig("AspectRatio", 1); - rs.m_vsync = !!theApp.GetConfig("vsync", 0); - rs.m_scale.x = theApp.GetConfig("scale_x", 0); - rs.m_scale.y = theApp.GetConfig("scale_y", 0); - - threads = theApp.GetConfig("swthreads", 1); - renderer = theApp.GetConfig("Renderer", 1); + int renderer = theApp.GetConfig("Renderer", 1); switch(renderer) { default: - case 0: s_gpu = new GPURendererSW(rs, new GSDevice7(), threads); break; - case 1: s_gpu = new GPURendererSW(rs, new GSDevice9(), threads); break; - case 2: s_gpu = new GPURendererSW(rs, new GSDevice10(), threads); break; - // TODO: case 3: s_gpu = new GPURendererNull(rs, new GSDeviceNull(), threads); break; + case 0: s_gpu = new GPURendererSW(new GSDevice7()); break; + case 1: s_gpu = new GPURendererSW(new GSDevice9()); break; + case 2: s_gpu = new GPURendererSW(new GSDevice10()); break; + // TODO: case 3: s_gpu = new GPURendererNull(new GSDeviceNull()); break; } if(!s_gpu->Create(hWnd)) diff --git a/plugins/GSdx/GPULocalMemory.cpp b/plugins/GSdx/GPULocalMemory.cpp index c0bfefb076..65a82716ba 100644 --- a/plugins/GSdx/GPULocalMemory.cpp +++ b/plugins/GSdx/GPULocalMemory.cpp @@ -21,16 +21,17 @@ #include "StdAfx.h" #include "GPULocalMemory.h" +#include "GSdx.h" const GSVector4i GPULocalMemory::m_xxxa(0x00008000); const GSVector4i GPULocalMemory::m_xxbx(0x00007c00); const GSVector4i GPULocalMemory::m_xgxx(0x000003e0); const GSVector4i GPULocalMemory::m_rxxx(0x0000001f); -GPULocalMemory::GPULocalMemory(const GSVector2i& scale) +GPULocalMemory::GPULocalMemory() { - m_scale.x = min(max(scale.x, 0), 2); - m_scale.y = min(max(scale.y, 0), 2); + m_scale.x = min(max(theApp.GetConfig("scale_x", 0), 0), 2); + m_scale.y = min(max(theApp.GetConfig("scale_y", 0), 0), 2); // diff --git a/plugins/GSdx/GPULocalMemory.h b/plugins/GSdx/GPULocalMemory.h index 5759779a7e..e693bad81a 100644 --- a/plugins/GSdx/GPULocalMemory.h +++ b/plugins/GSdx/GPULocalMemory.h @@ -50,7 +50,7 @@ class GPULocalMemory GSVector2i m_scale; public: - GPULocalMemory(const GSVector2i& scale); + GPULocalMemory(); virtual ~GPULocalMemory(); GSVector2i GetScale() {return m_scale;} diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index b3c2670771..5d04451f46 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -21,18 +21,19 @@ #include "StdAfx.h" #include "GPURenderer.h" +#include "GSdx.h" map GPURenderer::m_wnd2gpu; -GPURenderer::GPURenderer(const GPURendererSettings& rs) - : GPUState(rs.m_scale) +GPURenderer::GPURenderer(GSDevice* dev) + : m_dev(dev) , m_hWnd(NULL) , m_wndproc(NULL) { - m_filter = rs.m_filter; - m_dither = rs.m_dither; - m_aspectratio = rs.m_aspectratio; - m_vsync = rs.m_vsync; + m_filter = theApp.GetConfig("filter", 0); + m_dither = theApp.GetConfig("dithering", 1); + m_aspectratio = theApp.GetConfig("AspectRatio", 1); + m_vsync = !!theApp.GetConfig("vsync", 0); m_scale = m_mem.GetScale(); } diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index d7e21136a6..852e3be0f6 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -25,18 +25,15 @@ #include "GSVertexList.h" #include "GSDevice.h" -struct GPURendererSettings +class GPURenderer : public GPUState { +protected: int m_filter; int m_dither; int m_aspectratio; bool m_vsync; GSVector2i m_scale; -}; -class GPURenderer : public GPUState, protected GPURendererSettings -{ -protected: HWND m_hWnd; WNDPROC m_wndproc; static map m_wnd2gpu; @@ -46,7 +43,10 @@ protected: LRESULT OnMessage(UINT message, WPARAM wParam, LPARAM lParam); public: - GPURenderer(const GPURendererSettings& rs); + GSDevice* m_dev; + +public: + GPURenderer(GSDevice* dev); virtual ~GPURenderer(); virtual bool Create(HWND hWnd); @@ -191,11 +191,8 @@ protected: } public: - GSDevice* m_dev; - -public: - GPURendererT(const GPURendererSettings& rs) - : GPURenderer(rs) + GPURendererT(GSDevice* dev) + : GPURenderer(dev) , m_count(0) , m_maxcount(10000) { @@ -291,7 +288,7 @@ public: GetClientRect(m_hWnd, r); - m_dev->Present(r.fit(m_aspectratio)); + m_dev->Present(r.fit(m_aspectratio), 0); } virtual bool MakeSnapshot(const string& path) diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index eef66a2962..4fbec76f77 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -21,14 +21,13 @@ #include "StdAfx.h" #include "GPURendererSW.h" +#include "GSdx.h" -GPURendererSW::GPURendererSW(const GPURendererSettings& rs, GSDevice* dev, int threads) - : GPURendererT(rs) +GPURendererSW::GPURendererSW(GSDevice* dev) + : GPURendererT(dev) , m_texture(NULL) { - m_dev = dev; - - m_rl.Create(this, threads); + m_rl.Create(this, theApp.GetConfig("swthreads", 1)); m_fpDrawingKickHandlers[GPU_POLYGON] = (DrawingKickHandler)&GPURendererSW::DrawingKickTriangle; m_fpDrawingKickHandlers[GPU_LINE] = (DrawingKickHandler)&GPURendererSW::DrawingKickLine; diff --git a/plugins/GSdx/GPURendererSW.h b/plugins/GSdx/GPURendererSW.h index 58b5910bf9..f797bc3080 100644 --- a/plugins/GSdx/GPURendererSW.h +++ b/plugins/GSdx/GPURendererSW.h @@ -44,6 +44,6 @@ protected: GSTexture* GetOutput(); public: - GPURendererSW(const GPURendererSettings& rs, GSDevice* dev, int threads); + GPURendererSW(GSDevice* dev); virtual ~GPURendererSW(); }; diff --git a/plugins/GSdx/GPUState.cpp b/plugins/GSdx/GPUState.cpp index dae0e4cc1a..11e213c373 100644 --- a/plugins/GSdx/GPUState.cpp +++ b/plugins/GSdx/GPUState.cpp @@ -22,9 +22,8 @@ #include "stdafx.h" #include "GPUState.h" -GPUState::GPUState(const GSVector2i& scale) - : m_mem(scale) - , s_n(0) +GPUState::GPUState() + : s_n(0) { memset(m_status, 0, sizeof(m_status)); diff --git a/plugins/GSdx/GPUState.h b/plugins/GSdx/GPUState.h index db55189d3f..15d4d9c767 100644 --- a/plugins/GSdx/GPUState.h +++ b/plugins/GSdx/GPUState.h @@ -119,7 +119,7 @@ public: uint32 m_status[256]; public: - GPUState(const GSVector2i& scale); + GPUState(); virtual ~GPUState(); virtual void Reset(); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index b6612429b8..7b0aefabcc 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -100,10 +100,6 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) { GSclose(); - GSRendererSettings rs; - - int threads = 1; - #ifdef _WINDOWS s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); @@ -120,27 +116,17 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) return -1; } - rs.m_interlace = theApp.GetConfig("interlace", 0); - rs.m_aspectratio = theApp.GetConfig("aspectratio", 1); - rs.m_filter = theApp.GetConfig("filter", 1); - rs.m_vsync = !!theApp.GetConfig("vsync", 0); - rs.m_nativeres = !!theApp.GetConfig("nativeres", 0); - rs.m_aa1 = !!theApp.GetConfig("aa1", 0); - rs.m_blur = !!theApp.GetConfig("blur", 0); - - threads = theApp.GetConfig("swthreads", 1); - switch(renderer) { default: - case 0: s_gs = new GSRendererHW9(s_basemem, !!mt, s_irq, rs); break; - case 1: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDevice9(), threads); break; - case 2: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDevice9()); break; - case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq, rs); break; - case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDevice10(), threads); break; - case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDevice10()); break; - case 6: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, rs, new GSDeviceNull(), threads); break; - case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, rs, new GSDeviceNull()); break; + case 0: s_gs = new GSRendererHW9(s_basemem, !!mt, s_irq); break; + case 1: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice9()); break; + case 2: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice9()); break; + case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; + case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; + case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; + case 6: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; + case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; } if(!s_gs->Create(title)) diff --git a/plugins/GSdx/GSDevice.cpp b/plugins/GSdx/GSDevice.cpp index c76a18dc13..7a36bc98d5 100644 --- a/plugins/GSdx/GSDevice.cpp +++ b/plugins/GSdx/GSDevice.cpp @@ -70,7 +70,7 @@ bool GSDevice::Reset(int w, int h, bool fs) return true; } -void GSDevice::Present(const GSVector4i& r) +void GSDevice::Present(const GSVector4i& r, int shader) { GSVector4i cr; @@ -85,7 +85,9 @@ void GSDevice::Present(const GSVector4i& r) if(m_current) { - StretchRect(m_current, m_backbuffer, GSVector4(r)); + static int s_shader[3] = {0, 5, 6}; // FIXME + + StretchRect(m_current, m_backbuffer, GSVector4(r), s_shader[shader]); } Flip(); @@ -145,9 +147,9 @@ GSTexture* GSDevice::CreateOffscreen(int w, int h, int format) return Fetch(GSTexture::Offscreen, w, h, format); } -void GSDevice::StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, bool linear) +void GSDevice::StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, int shader, bool linear) { - StretchRect(st, GSVector4(0, 0, 1, 1), dt, dr, linear); + StretchRect(st, GSVector4(0, 0, 1, 1), dt, dr, shader, linear); } GSTexture* GSDevice::GetCurrent() diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index ca41beac55..3a45f8631c 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -76,7 +76,7 @@ public: virtual bool Reset(int w, int h, bool fs); virtual bool IsLost() {return false;} - virtual void Present(const GSVector4i& r); + virtual void Present(const GSVector4i& r, int shader); virtual void Flip() {}; virtual void BeginScene() {}; @@ -94,8 +94,8 @@ public: virtual GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0) {return NULL;} - virtual void StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, bool linear = true); - virtual void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true) {} + virtual void StretchRect(GSTexture* st, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); + virtual void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true) {} GSTexture* GetCurrent(); virtual bool IsCurrentRGBA() {return true;} diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 8eadbb5751..f492eef496 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -228,13 +228,6 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) Reset(1, 1, true); // -/* - if(!m_mergefx.Create(this)) - { - return false; - } -*/ - // return true; } @@ -409,9 +402,9 @@ GSTexture* GSDevice10::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, return dst; } -void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear) +void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader, bool linear) { - StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, linear); + StretchRect(st, sr, dt, dr, m_convert.ps[shader], NULL, linear); } void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear) diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index 3179aea568..6d94f998a8 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -71,7 +71,7 @@ public: // TODO CComPtr vb; CComPtr il; CComPtr vs; - CComPtr ps[5]; + CComPtr ps[7]; CComPtr ln; CComPtr pt; CComPtr dss; @@ -115,7 +115,7 @@ public: GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); - void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); diff --git a/plugins/GSdx/GSDevice7.cpp b/plugins/GSdx/GSDevice7.cpp index 32b09d41e7..f09d75b6bc 100644 --- a/plugins/GSdx/GSDevice7.cpp +++ b/plugins/GSdx/GSDevice7.cpp @@ -131,7 +131,7 @@ bool GSDevice7::Reset(int w, int h, bool fs) return true; } -void GSDevice7::Present(const GSVector4i& r) +void GSDevice7::Present(const GSVector4i& r, int shader) { HRESULT hr; diff --git a/plugins/GSdx/GSDevice7.h b/plugins/GSdx/GSDevice7.h index e0327f003b..ac47094df2 100644 --- a/plugins/GSdx/GSDevice7.h +++ b/plugins/GSdx/GSDevice7.h @@ -42,5 +42,5 @@ public: bool Create(HWND hWnd, bool vsync); bool Reset(int w, int h, bool fs); - void Present(const GSVector4i& r); + void Present(const GSVector4i& r, int shader); }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index bb13ecb389..ac0c436262 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -507,9 +507,9 @@ GSTexture* GSDevice9::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, return dst; } -void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear) +void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader, bool linear) { - StretchRect(st, sr, dt, dr, m_convert.ps[0], NULL, 0, linear); + StretchRect(st, sr, dt, dr, m_convert.ps[shader], NULL, 0, linear); } void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear) diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 23870ef97d..3d6b69086c 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -108,7 +108,7 @@ public: // TODO { CComPtr vs; CComPtr il; - CComPtr ps[5]; + CComPtr ps[7]; Direct3DSamplerState9 ln; Direct3DSamplerState9 pt; Direct3DDepthStencilState9 dss; @@ -153,7 +153,7 @@ public: virtual bool IsCurrentRGBA() {return false;} - void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 1fe11ac825..53da363389 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -22,19 +22,19 @@ #include "StdAfx.h" #include "GSRenderer.h" -GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) +GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr) : GSState(base, mt, irq) - , m_dev(NULL) - , m_osd(true) + , m_dev(dev) + , m_shader(0) , m_psrr(psrr) { - m_interlace = rs.m_interlace; - m_aspectratio = rs.m_aspectratio; - m_filter = rs.m_filter; - m_vsync = rs.m_vsync; - m_nativeres = rs.m_nativeres; - m_aa1 = rs.m_aa1; - m_blur = rs.m_blur; + m_interlace = theApp.GetConfig("interlace", 0); + m_aspectratio = theApp.GetConfig("aspectratio", 1); + m_filter = theApp.GetConfig("filter", 1); + m_vsync = !!theApp.GetConfig("vsync", 0); + m_nativeres = !!theApp.GetConfig("nativeres", 0); + m_aa1 = !!theApp.GetConfig("aa1", 0); + m_blur = !!theApp.GetConfig("blur", 0); s_n = 0; s_dump = !!theApp.GetConfig("dump", 0); @@ -328,7 +328,7 @@ void GSRenderer::VSync(int field) m_wnd.GetClientRect(r); - m_dev->Present(r.fit(m_aspectratio)); + m_dev->Present(r.fit(m_aspectratio), m_shader); // snapshot @@ -406,8 +406,7 @@ void GSRenderer::KeyEvent(GSKeyEventData* e) m_aspectratio = (m_aspectratio + 3 + step) % 3; return; case VK_F7: - m_wnd.SetWindowText(_T("PCSX2")); - m_osd = !m_osd; + m_shader = (m_shader + 3 + step) % 3; return; case VK_F12: if(m_capture.IsCapturing()) m_capture.EndCapture(); diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index fe7947aeb5..b4d1f4c3d7 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -28,8 +28,15 @@ #include "GSSettingsDlg.h" #include "GSCapture.h" -struct GSRendererSettings +class GSRenderer : public GSState { + GSCapture m_capture; + string m_snapshot; + int m_shader; + + bool Merge(int field); + +protected: int m_interlace; int m_aspectratio; int m_filter; @@ -37,22 +44,13 @@ struct GSRendererSettings bool m_nativeres; bool m_aa1; bool m_blur; -}; -class GSRenderer : public GSState, protected GSRendererSettings -{ - bool Merge(int field); - -protected: virtual void ResetDevice() {} virtual GSTexture* GetOutput(int i) = 0; public: GSWnd m_wnd; GSDevice* m_dev; - GSCapture m_capture; - string m_snapshot; - bool m_osd; bool m_psrr; int s_n; @@ -61,15 +59,13 @@ public: bool s_savez; public: - GSRenderer(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true); + GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr = true); virtual ~GSRenderer(); virtual bool Create(const string& title); virtual void VSync(int field); - virtual void KeyEvent(GSKeyEventData* e); - virtual bool MakeSnapshot(const string& path); virtual void MinMaxUV(int w, int h, GSVector4i& r) @@ -221,8 +217,8 @@ protected: virtual void Draw() = 0; public: - GSRendererT(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr = true) - : GSRenderer(base, mt, irq, rs, psrr) + GSRendererT(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr = true) + : GSRenderer(base, mt, irq, dev, psrr) , m_count(0) , m_maxcount(0) , m_vertices(NULL) diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index e5ca38975a..1fcde06e94 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -247,7 +247,7 @@ protected: TEX0.TBW = DISPFB.FBW; TEX0.PSM = DISPFB.PSM; - TRACE(_T("[%d] GetOutput %d %05x (%d)\n"), (int)m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM); + // TRACE(_T("[%d] GetOutput %d %05x (%d)\n"), (int)m_perfmon.GetFrame(), i, (int)TEX0.TBP0, (int)TEX0.PSM); GSTexture* t = NULL; @@ -665,9 +665,9 @@ protected: } public: - GSRendererHW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, bool psrr) - : GSRendererT(base, mt, irq, rs, psrr) - , m_tc(NULL) + GSRendererHW(uint8* base, bool mt, void (*irq)(), GSDevice* dev, GSTextureCache* tc, bool psrr) + : GSRendererT(base, mt, irq, dev, psrr) + , m_tc(tc) , m_width(1024) , m_height(1024) , m_skip(0) diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 29a0f0c1c3..f26c67cfd1 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -24,12 +24,9 @@ #include "GSCrc.h" #include "resource.h" -GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSRendererHW(base, mt, irq, rs, true) +GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)()) + : GSRendererHW(base, mt, irq, new GSDevice10(), new GSTextureCache10(this), true) { - m_dev = new GSDevice10(); - m_tc = new GSTextureCache10(this); - InitVertexKick(); } diff --git a/plugins/GSdx/GSRendererHW10.h b/plugins/GSdx/GSRendererHW10.h index ef6c3548bd..51cd9648b4 100644 --- a/plugins/GSdx/GSRendererHW10.h +++ b/plugins/GSdx/GSRendererHW10.h @@ -44,7 +44,7 @@ protected: void SetupDATE(GSTexture* rt, GSTexture* ds); public: - GSRendererHW10(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs); + GSRendererHW10(uint8* base, bool mt, void (*irq)()); bool Create(const string& title); diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 1f3dc149d4..4f4e0cf3e9 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -24,15 +24,12 @@ #include "GSCrc.h" #include "resource.h" -GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs) - : GSRendererHW(base, mt, irq, rs, false) +GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)()) + : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this), false) { m_fba.enabled = !!theApp.GetConfig("fba", 1); m_logz = !!theApp.GetConfig("logz", 0); - m_dev = new GSDevice9(); - m_tc = new GSTextureCache9(this); - InitVertexKick(); } diff --git a/plugins/GSdx/GSRendererHW9.h b/plugins/GSdx/GSRendererHW9.h index e6235b45f4..9b8ff4221a 100644 --- a/plugins/GSdx/GSRendererHW9.h +++ b/plugins/GSdx/GSRendererHW9.h @@ -53,7 +53,7 @@ protected: void UpdateFBA(GSTexture* rt); public: - GSRendererHW9(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs); + GSRendererHW9(uint8* base, bool mt, void (*irq)()); bool Create(const string& title); diff --git a/plugins/GSdx/GSRendererNull.h b/plugins/GSdx/GSRendererNull.h index d23738ee17..285cdc4d2f 100644 --- a/plugins/GSdx/GSRendererNull.h +++ b/plugins/GSdx/GSRendererNull.h @@ -37,11 +37,9 @@ protected: } public: - GSRendererNull(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev) - : GSRendererT(base, mt, irq, rs) + GSRendererNull(uint8* base, bool mt, void (*irq)(), GSDevice* dev) + : GSRendererT(base, mt, irq, dev) { - m_dev = dev; - InitVertexKick(); } diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index 8433888d90..778cb3e4c9 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -24,16 +24,14 @@ const GSVector4 g_pos_scale(1.0f / 16, 1.0f / 16, 1.0f, 128.0f); -GSRendererSW::GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev, int threads) - : GSRendererT(base, mt, irq, rs) +GSRendererSW::GSRendererSW(uint8* base, bool mt, void (*irq)(), GSDevice* dev) + : GSRendererT(base, mt, irq, dev) { - m_dev = dev; - m_tc = new GSTextureCacheSW(this); memset(m_texture, 0, sizeof(m_texture)); - m_rl.Create(this, threads); + m_rl.Create(this, theApp.GetConfig("swthreads", 1)); InitVertexKick(); } diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 7ccdd9b136..367f999d79 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -50,7 +50,7 @@ protected: void GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass); public: - GSRendererSW(uint8* base, bool mt, void (*irq)(), const GSRendererSettings& rs, GSDevice* dev, int threads); + GSRendererSW(uint8* base, bool mt, void (*irq)(), GSDevice* dev); virtual ~GSRendererSW(); template diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index 57b75cf6f2..5656f48c45 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -120,7 +120,7 @@ bool GSUtil::CheckDirectX() } else { - int res = AfxMessageBox(_T("Please update DirectX!\n\nWould you like to open the download page in your browser?"), MB_YESNO); + int res = AfxMessageBox(_T("You need to update directx, would you like to do it now?"), MB_YESNO); if(res == IDYES) { diff --git a/plugins/GSdx/res/convert10.fx b/plugins/GSdx/res/convert10.fx index 05044323cc..73742f9d3f 100644 --- a/plugins/GSdx/res/convert10.fx +++ b/plugins/GSdx/res/convert10.fx @@ -65,3 +65,31 @@ float4 ps_main4(PS_INPUT input) : SV_Target0 return fmod(c * 255 + 0.5f, 256) / 255; } + +float4 ps_crt(PS_INPUT input, uint i) +{ + float4 mask[4] = + { + float4(1, 0, 0, 0), + float4(0, 1, 0, 0), + float4(0, 0, 1, 0), + float4(1, 1, 1, 0) + }; + + return Texture.Sample(Sampler, input.t) * saturate(mask[i] + 0.25f); +} + +float4 ps_main5(PS_INPUT input) : SV_Target0 // triangular +{ + uint4 p = (uint4)input.p; + + // return ps_crt(input, ((p.x + (p.y & 1) * 3) >> 1) % 3); + return ps_crt(input, ((p.x + ((p.y >> 1) & 1) * 3) >> 1) % 3); +} + +float4 ps_main6(PS_INPUT input) : SV_Target0 // diagonal +{ + uint4 p = (uint4)input.p; + + return ps_crt(input, (p.x + (p.y % 3)) % 3); +} \ No newline at end of file diff --git a/plugins/GSdx/res/convert9.fx b/plugins/GSdx/res/convert9.fx index 4f1f8be5da..1341091681 100644 --- a/plugins/GSdx/res/convert9.fx +++ b/plugins/GSdx/res/convert9.fx @@ -53,3 +53,30 @@ float4 ps_main4() : COLOR return 1; } +float4 ps_crt(float2 t, int i) +{ + float4 mask[4] = + { + float4(1, 0, 0, 0), + float4(0, 1, 0, 0), + float4(0, 0, 1, 0), + float4(1, 1, 1, 0) + }; + + return tex2D(Texture, t) * saturate(mask[i] + 0.5f); +} + +float4 ps_main5(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // triangular +{ + int4 p = (int4)vPos; + + // return ps_crt(t, ((p.x + (p.y % 2) * 3) / 2) % 3); + return ps_crt(t, ((p.x + ((p.y / 2) % 2) * 3) / 2) % 3); +} + +float4 ps_main6(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // diagonal +{ + int4 p = (int4)vPos; + + return ps_crt(t, (p.x + (p.y % 3)) % 3); +} From 3ff7f2f191779ca9ebb14c8989d869830317a167 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 23 May 2009 01:11:28 +0000 Subject: [PATCH 109/296] Meaningless Commits Inc presents: Made jASSUME to be more scope-proof, and fixed a minor signed/unsigned warning in cdvd. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1240 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Defs.h | 4 ++-- pcsx2/CDVD/CDVD.cpp | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/common/include/Pcsx2Defs.h b/common/include/Pcsx2Defs.h index cb2aa7937c..646feabe6f 100644 --- a/common/include/Pcsx2Defs.h +++ b/common/include/Pcsx2Defs.h @@ -50,12 +50,12 @@ # define jASSUME(exp) ((void) sizeof(exp)) # endif # else -# if defined(_MSC_VER) +# ifdef _MSC_VER # define jBREAKPOINT() do { __asm int 3 } while(0) # else # define jBREAKPOINT() ((void) *(volatile char *) 0) # endif -# define jASSUME(exp) if(exp) ; else jBREAKPOINT() +# define jASSUME(exp) do { if(exp) ; else jBREAKPOINT() } while(0) # endif #endif diff --git a/pcsx2/CDVD/CDVD.cpp b/pcsx2/CDVD/CDVD.cpp index e93117da87..675a070ec7 100644 --- a/pcsx2/CDVD/CDVD.cpp +++ b/pcsx2/CDVD/CDVD.cpp @@ -812,7 +812,7 @@ static uint cdvdStartSeek( uint newsector, CDVD_MODE_TYPE mode ) { cdvd.SeekToSector = newsector; - uint delta = abs(cdvd.SeekToSector - cdvd.Sector); + uint delta = abs( (s32)(cdvd.SeekToSector - cdvd.Sector) ); uint seektime; cdvd.Ready = CDVD_NOTREADY; From 995b9d1eaa43be5e6dc2efbb028aa004525cecfc Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 23 May 2009 03:27:53 +0000 Subject: [PATCH 110/296] microVU: - Implemented the case where an Unconditional branch has the E-bit Set. This solves a few games from getting stuck in infinite loops. (I still need to handle conditional branches and jumps). - Max/Mini Opcodes shouldn't set status/mac flags. This should be a minor speedup. - Minor Changes/Cleanups... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1241 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 42 ++++++++++++++++++++++------------- pcsx2/x86/microVU_Execute.inl | 7 ------ pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 4 ++-- pcsx2/x86/microVU_Upper.inl | 31 +++++++++++++++----------- 5 files changed, 47 insertions(+), 39 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 016dff7217..0857fd144a 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -41,6 +41,15 @@ } \ } +#define branchEbit() { \ + if (branch == 2) { \ + if (!((mVUbranch == 1) || (mVUbranch == 2))) { \ + Console::Error("microVU%d Warning: Jump with E-bit not Implemented [%04x]", params vuIndex, xPC); \ + } \ + else { eBitBranch = 1; } \ + } \ +} + #define startLoop() { mVUdebug1(); mVUstall = 0; memset(&mVUregsTemp, 0, sizeof(mVUregsTemp)); } #define calcCycles(reg, x) { reg = ((reg > x) ? (reg - x) : 0); } #define tCycles(dest, src) { dest = aMax(dest, src); } @@ -163,9 +172,7 @@ microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fC MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); - //memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - //MOV32ItoM((uptr)&mVU->prog.lpState, (int)&mVUblock.pState); // Save pipeline state (clipflag instance) - //AND32ItoM((uptr)µVU0.regs->VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag + // Clear 'is busy' Flags, Save PC, and Jump to Exit Point AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); @@ -175,7 +182,8 @@ microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fC #define sI ((mVUpBlock->pState.needExactMatch & 0x000f) ? 0 : ((mVUpBlock->pState.flags >> 0) & 3)) #define cI ((mVUpBlock->pState.needExactMatch & 0x0f00) ? 0 : ((mVUpBlock->pState.flags >> 2) & 3)) -void mVUwarning() { Console::Error("microVU Warning: Exiting Execution Early (Possible Infinite Loop)"); } +void __fastcall mVUwarning0(u32 PC) { Console::Error("microVU0 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } +void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; @@ -183,7 +191,9 @@ microVUt(void) mVUtestCycles() { CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); PUSH32R(gprR); - CALLFunc((uptr)mVUwarning); + MOV32ItoR(gprT2, xPC); + if (!vuIndex) CALLFunc((uptr)mVUwarning0); + else CALLFunc((uptr)mVUwarning1); POP32R(gprR); mVUendProgram(0, 0, sI, 0, cI); x86SetJ8(jmp8); @@ -224,9 +234,10 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info mVUblock.x86ptrStart = thisPtr; pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager - mVUpBlock = pBlock; - mVUregs.flags = 0; - mVUflagInfo = 0; + mVUpBlock = pBlock; + mVUregs.flags = 0; + mVUflagInfo = 0; + bool eBitBranch = 0; // E-bit Set on Branch for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); @@ -235,7 +246,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { startLoop(); mVUopU(); if (curI & _Ebit_) { branch = 1; } - if (curI & _MDTbit_) { branch = 2; } + if (curI & _MDTbit_) { branch = 4; } if (curI & _Ibit_) { mVUinfo |= _isNOP; } else { incPC(-1); mVUopL(); incPC(1); } mVUsetCycles(); @@ -244,7 +255,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { else { mVUinfo |= _writeQ; } if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } + if (mVUbranch) { mVUsetFlagInfo(); branchEbit(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } incPC(1); mVUcount++; } @@ -253,10 +264,6 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { int xStatus[4], xMac[4], xClip[4]; int xCycles = mVUsetFlags(xStatus, xMac, xClip); mVUtestCycles(); - //SysPrintf("bS[0] = %08x, bS[1] = %08x, bS[2] = %08x, bS[3] = %08x\n", bStatus[0], bStatus[1], bStatus[2], bStatus[3]); - //SysPrintf("bM[0] = %08x, bM[1] = %08x, bM[2] = %08x, bM[3] = %08x\n", bMac[0], bMac[1], bMac[2], bMac[3]); - //SysPrintf("mVUcount = %d\n", mVUcount); - //SysPrintf("mVUflagInfo = %d\n", mVUflagInfo); // Second Pass iPC = mVUstartPC; @@ -286,6 +293,8 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile B/BAL"); incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if (eBitBranch) { iPC = branchAddr/4; goto eBitTemination; } // E-bit Was Set on Branch mVUsetupBranch(xStatus, xMac, xClip, xCycles); if (mVUblocks[branchAddr/8] == NULL) @@ -347,10 +356,11 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { return thisPtr; } } - mVUprint("mVUcompile ebit"); if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } +eBitTemination: + mVUprint("mVUcompile ebit"); incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; @@ -358,7 +368,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUsetupRange(xPC - 8); mVUendProgram(mVU->q, mVU->p, findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); - return thisPtr; //ToDo: Save pipeline state? + return thisPtr; } void* __fastcall mVUcompileVU0(u32 startPC, uptr pState) { return mVUcompile<0>(startPC, pState); } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 37dd5dbea2..148e0fd0c5 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -23,8 +23,6 @@ // Dispatcher Functions //------------------------------------------------------------------ -void testFunction() { mVUprint("microVU: Entered Execution Mode"); } - // Generates the code for entering recompiled blocks microVUt(void) mVUdispatcherA() { microVU* mVU = mVUx; @@ -70,11 +68,6 @@ microVUt(void) mVUdispatcherA() { if (isMMX(i)) { MOVQMtoR(mmVI(i), (uptr)&mVU->regs->VI[i].UL); } } - //PUSH32R(EAX); - //CALLFunc((uptr)testFunction); - //write8(0xcc); - //POP32R(EAX); - // Jump to Recompiled Code Block JMPR(EAX); } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index dd68d718b7..b513f8f89c 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -202,7 +202,7 @@ microVUt(void) mVUpass4(int startPC) { if (mVUbranch) { branch = 3; mVUbranch = 0; } incPC(1); } - //if (mVUcount == 4) { mVUflagInfo |= 0xfff; } // Is this Too Slow? 99% of games probably don't need this. + //if (mVUcount < 4) { mVUflagInfo |= 0xfff; } // Is this Too Slow? 99% of games probably don't need this. iPC = oldPC; mVUcount = oldCount; mVUbranch = oldBranch; diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 3880f57766..9a192c958f 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -590,8 +590,8 @@ microVUf(void) mVU_FSSET() { int flagReg1, flagReg2; getFlagReg(flagReg1, fsInstance); if (!(doStatus||doDivFlag)) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction - AND16ItoR(flagReg1, 0x03f); // Remember not to modify upper 16 bits because of mac flag - OR16ItoR (flagReg1, (_Imm12_ & 0xfc0)); + AND32ItoR(flagReg1, 0x03f); + OR32ItoR (flagReg1, (_Imm12_ & 0xfc0)); } pass3 { mVUlog("FSSET $%x", _Imm12_); } pass4 { mVUsFlagHack = 0; } diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index eb51cc2b42..37a92f8442 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -461,6 +461,11 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX pass3 { mVUlog(OPname); mVUlogACC(); mVUlogQ(); } \ } +// FMAC27~29 - MAX/MINI FMAC Opcodes +#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } +#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } +#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } + //------------------------------------------------------------------ // Micro VU Micromode Upper instructions //------------------------------------------------------------------ @@ -546,21 +551,21 @@ microVUf(void) mVU_MSUBAx() { mVU_FMAC15(SUB, "MSUBAx"); } microVUf(void) mVU_MSUBAy() { mVU_FMAC15(SUB, "MSUBAy"); } microVUf(void) mVU_MSUBAz() { mVU_FMAC15(SUB, "MSUBAz"); } microVUf(void) mVU_MSUBAw() { mVU_FMAC15(SUB, "MSUBAw"); } -microVUf(void) mVU_MAX() { mVU_FMAC1 (MAX2, "MAX"); } -microVUf(void) mVU_MAXi() { mVU_FMAC6 (MAX2, "MAXi"); } -microVUf(void) mVU_MAXx() { mVU_FMAC3 (MAX2, "MAXx"); } -microVUf(void) mVU_MAXy() { mVU_FMAC3 (MAX2, "MAXy"); } -microVUf(void) mVU_MAXz() { mVU_FMAC3 (MAX2, "MAXz"); } -microVUf(void) mVU_MAXw() { mVU_FMAC3 (MAX2, "MAXw"); } -microVUf(void) mVU_MINI() { mVU_FMAC1 (MIN2, "MINI"); } -microVUf(void) mVU_MINIi() { mVU_FMAC6 (MIN2, "MINIi"); } -microVUf(void) mVU_MINIx() { mVU_FMAC3 (MIN2, "MINIx"); } -microVUf(void) mVU_MINIy() { mVU_FMAC3 (MIN2, "MINIy"); } -microVUf(void) mVU_MINIz() { mVU_FMAC3 (MIN2, "MINIz"); } -microVUf(void) mVU_MINIw() { mVU_FMAC3 (MIN2, "MINIw"); } +microVUf(void) mVU_MAX() { mVU_FMAC27(MAX2, "MAX"); } +microVUf(void) mVU_MAXi() { mVU_FMAC28(MAX2, "MAXi"); } +microVUf(void) mVU_MAXx() { mVU_FMAC29(MAX2, "MAXx"); } +microVUf(void) mVU_MAXy() { mVU_FMAC29(MAX2, "MAXy"); } +microVUf(void) mVU_MAXz() { mVU_FMAC29(MAX2, "MAXz"); } +microVUf(void) mVU_MAXw() { mVU_FMAC29(MAX2, "MAXw"); } +microVUf(void) mVU_MINI() { mVU_FMAC27(MIN2, "MINI"); } +microVUf(void) mVU_MINIi() { mVU_FMAC28(MIN2, "MINIi"); } +microVUf(void) mVU_MINIx() { mVU_FMAC29(MIN2, "MINIx"); } +microVUf(void) mVU_MINIy() { mVU_FMAC29(MIN2, "MINIy"); } +microVUf(void) mVU_MINIz() { mVU_FMAC29(MIN2, "MINIz"); } +microVUf(void) mVU_MINIw() { mVU_FMAC29(MIN2, "MINIw"); } microVUf(void) mVU_OPMULA() { mVU_FMAC18(MUL, "OPMULA"); } microVUf(void) mVU_OPMSUB() { mVU_FMAC19(SUB, "OPMSUB"); } -microVUf(void) mVU_NOP() { pass3 { mVUlog("NOP"); } } +microVUf(void) mVU_NOP() { pass3 { mVUlog("NOP"); } } microVUq(void) mVU_FTOIx(uptr addr) { microVU* mVU = mVUx; pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } From 47bc9722d4cc340fe967b031908de19e8943c9e6 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 05:15:03 +0000 Subject: [PATCH 111/296] GSnull/PadNull: Add a null Pad plugin (Currently Linux only). Fix up GSnull on the Linux side. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1242 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSnull/GS.h | 9 +- plugins/GSnull/Linux/GSLinux.cpp | 14 +- plugins/GSnull/Linux/GSLinux.h | 16 +- plugins/GSnull/Makefile.am | 4 +- plugins/PadNull/License.txt | 342 ++++++++++++++++++ plugins/PadNull/Linux/Config.cpp | 168 +++++++++ plugins/PadNull/Linux/Config.h | 24 ++ plugins/PadNull/Linux/PadLinux.cpp | 83 +++++ plugins/PadNull/Linux/PadLinux.h | 30 ++ plugins/PadNull/Linux/callbacks.c | 34 ++ plugins/PadNull/Linux/callbacks.h | 14 + plugins/PadNull/Linux/interface.c | 173 +++++++++ plugins/PadNull/Linux/interface.h | 6 + plugins/PadNull/Linux/pad.glade | 240 ++++++++++++ plugins/PadNull/Linux/support.c | 144 ++++++++ plugins/PadNull/Linux/support.h | 69 ++++ plugins/PadNull/Makefile.am | 31 ++ plugins/PadNull/Pad.cpp | 170 +++++++++ plugins/PadNull/Pad.h | 71 ++++ plugins/PadNull/ReadMe.txt | 23 ++ plugins/PadNull/Windows/Config.cpp | 51 +++ plugins/PadNull/Windows/PadNull.def | 19 + plugins/PadNull/Windows/PadNull.rc | 116 ++++++ plugins/PadNull/Windows/PadNull_vc2008.cbp | 44 +++ plugins/PadNull/Windows/PadNull_vc2008.layout | 4 + plugins/PadNull/Windows/PadNull_vc2008.vcproj | 201 ++++++++++ .../PadNull/Windows/ProjectRootDir.vsprops | 11 + plugins/PadNull/Windows/Win32.cpp | 81 +++++ plugins/PadNull/Windows/resource.h | 21 ++ plugins/PadNull/build.sh | 32 ++ plugins/PadNull/configure.ac | 83 +++++ 31 files changed, 2314 insertions(+), 14 deletions(-) create mode 100644 plugins/PadNull/License.txt create mode 100644 plugins/PadNull/Linux/Config.cpp create mode 100644 plugins/PadNull/Linux/Config.h create mode 100644 plugins/PadNull/Linux/PadLinux.cpp create mode 100644 plugins/PadNull/Linux/PadLinux.h create mode 100644 plugins/PadNull/Linux/callbacks.c create mode 100644 plugins/PadNull/Linux/callbacks.h create mode 100644 plugins/PadNull/Linux/interface.c create mode 100644 plugins/PadNull/Linux/interface.h create mode 100644 plugins/PadNull/Linux/pad.glade create mode 100644 plugins/PadNull/Linux/support.c create mode 100644 plugins/PadNull/Linux/support.h create mode 100644 plugins/PadNull/Makefile.am create mode 100644 plugins/PadNull/Pad.cpp create mode 100644 plugins/PadNull/Pad.h create mode 100644 plugins/PadNull/ReadMe.txt create mode 100644 plugins/PadNull/Windows/Config.cpp create mode 100644 plugins/PadNull/Windows/PadNull.def create mode 100644 plugins/PadNull/Windows/PadNull.rc create mode 100644 plugins/PadNull/Windows/PadNull_vc2008.cbp create mode 100644 plugins/PadNull/Windows/PadNull_vc2008.layout create mode 100644 plugins/PadNull/Windows/PadNull_vc2008.vcproj create mode 100644 plugins/PadNull/Windows/ProjectRootDir.vsprops create mode 100644 plugins/PadNull/Windows/Win32.cpp create mode 100644 plugins/PadNull/Windows/resource.h create mode 100644 plugins/PadNull/build.sh create mode 100644 plugins/PadNull/configure.ac diff --git a/plugins/GSnull/GS.h b/plugins/GSnull/GS.h index 58877bac70..cd2dd0c071 100644 --- a/plugins/GSnull/GS.h +++ b/plugins/GSnull/GS.h @@ -23,8 +23,6 @@ #ifdef _WIN32 #include "Windows/GSwin.h" -#else -#include "Linux/GSLinux.h" #endif #ifdef __cplusplus @@ -37,7 +35,12 @@ extern "C" } #endif +#ifdef __LINUX__ +#include "Linux/GSLinux.h" +#endif + #include "GifTransfer.h" +#include "null/GSnull.h" /*#ifdef _MSC_VER #define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK @@ -60,6 +63,8 @@ typedef struct extern Config conf; extern FILE *gsLog; +extern u32 GSKeyEvent; +extern bool GSShift, GSAlt; extern void (*GSirq)(); diff --git a/plugins/GSnull/Linux/GSLinux.cpp b/plugins/GSnull/Linux/GSLinux.cpp index 34c7d5149c..67385d2bc3 100644 --- a/plugins/GSnull/Linux/GSLinux.cpp +++ b/plugins/GSnull/Linux/GSLinux.cpp @@ -15,7 +15,8 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - + +#include "GS.h" #include "GSLinux.h" Display *display; @@ -68,9 +69,11 @@ void GSProcessMessages() void HandleKeyEvent(keyEvent *ev) { - switch(ev->evt) { + switch(ev->evt) + { case KEYPRESS: - switch(ev->key) { + switch(ev->key) + { case XK_F5: case XK_F6: case XK_F7: @@ -87,10 +90,11 @@ void HandleKeyEvent(keyEvent *ev) case XK_Alt_R: GSAlt = true; break; - } + } break; case KEYRELEASE: - switch(ev->key) { + switch(ev->key) + { case XK_Shift_L: case XK_Shift_R: GSShift = false; diff --git a/plugins/GSnull/Linux/GSLinux.h b/plugins/GSnull/Linux/GSLinux.h index 99a686efd6..d4f1303a9b 100644 --- a/plugins/GSnull/Linux/GSLinux.h +++ b/plugins/GSnull/Linux/GSLinux.h @@ -16,12 +16,18 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#ifndef __GSLINUX_H__ +#define __GSLINUX_H__ + #include #include #include -#include "../GS.h" + +#include "GS.h" -int GSOpenWindow(void *pDsp, char *Title); -void GSCloseWindow(); -void GSProcessMessages(); -void HandleKeyEvent(keyEvent *ev); \ No newline at end of file +extern int GSOpenWindow(void *pDsp, char *Title); +extern void GSCloseWindow(); +extern void GSProcessMessages(); +extern void HandleKeyEvent(keyEvent *ev); + +#endif diff --git a/plugins/GSnull/Makefile.am b/plugins/GSnull/Makefile.am index da1ecce6a1..1e86125335 100644 --- a/plugins/GSnull/Makefile.am +++ b/plugins/GSnull/Makefile.am @@ -1,7 +1,7 @@ # Create a shared library libGSnull AUTOMAKE_OPTIONS = foreign noinst_LIBRARIES = libGSnull.a -INCLUDES = -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty -I@srcdir@/Linux +INCLUDES = -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty -I@srcdir@/Linux -I@srcdir@/null libGSnull_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) libGSnull_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) @@ -24,7 +24,7 @@ libGSnull_LDADD=$(libGSnull_a_OBJECTS) libGSnull_a_SOURCES = \ GS.cpp \ -Linux/Config.cpp Linux/Linux.cpp Linux/GSLinux.cpp +Linux/Config.cpp Linux/Linux.cpp Linux/GSLinux.cpp \ Linux/GSLinux.h Linux/Config.h Linux/Linux.h \ GifTransfer.cpp GifTransfer.h \ null/GSnull.cpp null/GSnull.h diff --git a/plugins/PadNull/License.txt b/plugins/PadNull/License.txt new file mode 100644 index 0000000000..7d1f8605f8 --- /dev/null +++ b/plugins/PadNull/License.txt @@ -0,0 +1,342 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc. + 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +License is intended to guarantee your freedom to share and change free +software--to make sure the software is free for all its users. This +General Public License applies to most of the Free Software +Foundation's software and to any other program whose authors commit to +using it. (Some other Free Software Foundation software is covered by +the GNU Library General Public License instead.) You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +this service if you wish), that you receive source code or can get it +if you want it, that you can change the software or use pieces of it +in new free programs; and that you know you can do these things. + + To protect your rights, we need to make restrictions that forbid +anyone to deny you these rights or to ask you to surrender the rights. +These restrictions translate to certain responsibilities for you if you +distribute copies of the software, or if you modify it. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must give the recipients all the rights that +you have. You must make sure that they, too, receive or can get the +source code. And you must show them these terms so they know their +rights. + + We protect your rights with two steps: (1) copyright the software, and +(2) offer you this license which gives you legal permission to copy, +distribute and/or modify the software. + + Also, for each author's protection and ours, we want to make certain +that everyone understands that there is no warranty for this free +software. If the software is modified by someone else and passed on, we +want its recipients to know that what they have is not the original, so +that any problems introduced by others will not reflect on the original +authors' reputations. + + Finally, any free program is threatened constantly by software +patents. We wish to avoid the danger that redistributors of a free +program will individually obtain patent licenses, in effect making the +program proprietary. To prevent this, we have made it clear that any +patent must be licensed for everyone's free use or not licensed at all. + + The precise terms and conditions for copying, distribution and +modification follow. + + GNU GENERAL PUBLIC LICENSE + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License applies to any program or other work which contains +a notice placed by the copyright holder saying it may be distributed +under the terms of this General Public License. The "Program", below, +refers to any such program or work, and a "work based on the Program" +means either the Program or any derivative work under copyright law: +that is to say, a work containing the Program or a portion of it, +either verbatim or with modifications and/or translated into another +language. (Hereinafter, translation is included without limitation in +the term "modification".) Each licensee is addressed as "you". + +Activities other than copying, distribution and modification are not +covered by this License; they are outside its scope. The act of +running the Program is not restricted, and the output from the Program +is covered only if its contents constitute a work based on the +Program (independent of having been made by running the Program). +Whether that is true depends on what the Program does. + + 1. You may copy and distribute verbatim copies of the Program's +source code as you receive it, in any medium, provided that you +conspicuously and appropriately publish on each copy an appropriate +copyright notice and disclaimer of warranty; keep intact all the +notices that refer to this License and to the absence of any warranty; +and give any other recipients of the Program a copy of this License +along with the Program. + +You may charge a fee for the physical act of transferring a copy, and +you may at your option offer warranty protection in exchange for a fee. + + 2. You may modify your copy or copies of the Program or any portion +of it, thus forming a work based on the Program, and copy and +distribute such modifications or work under the terms of Section 1 +above, provided that you also meet all of these conditions: + + a) You must cause the modified files to carry prominent notices + stating that you changed the files and the date of any change. + + b) You must cause any work that you distribute or publish, that in + whole or in part contains or is derived from the Program or any + part thereof, to be licensed as a whole at no charge to all third + parties under the terms of this License. + + c) If the modified program normally reads commands interactively + when run, you must cause it, when started running for such + interactive use in the most ordinary way, to print or display an + announcement including an appropriate copyright notice and a + notice that there is no warranty (or else, saying that you provide + a warranty) and that users may redistribute the program under + these conditions, and telling the user how to view a copy of this + License. (Exception: if the Program itself is interactive but + does not normally print such an announcement, your work based on + the Program is not required to print an announcement.) + +These requirements apply to the modified work as a whole. If +identifiable sections of that work are not derived from the Program, +and can be reasonably considered independent and separate works in +themselves, then this License, and its terms, do not apply to those +sections when you distribute them as separate works. But when you +distribute the same sections as part of a whole which is a work based +on the Program, the distribution of the whole must be on the terms of +this License, whose permissions for other licensees extend to the +entire whole, and thus to each and every part regardless of who wrote it. + +Thus, it is not the intent of this section to claim rights or contest +your rights to work written entirely by you; rather, the intent is to +exercise the right to control the distribution of derivative or +collective works based on the Program. + +In addition, mere aggregation of another work not based on the Program +with the Program (or with a work based on the Program) on a volume of +a storage or distribution medium does not bring the other work under +the scope of this License. + + 3. You may copy and distribute the Program (or a work based on it, +under Section 2) in object code or executable form under the terms of +Sections 1 and 2 above provided that you also do one of the following: + + a) Accompany it with the complete corresponding machine-readable + source code, which must be distributed under the terms of Sections + 1 and 2 above on a medium customarily used for software interchange; or, + + b) Accompany it with a written offer, valid for at least three + years, to give any third party, for a charge no more than your + cost of physically performing source distribution, a complete + machine-readable copy of the corresponding source code, to be + distributed under the terms of Sections 1 and 2 above on a medium + customarily used for software interchange; or, + + c) Accompany it with the information you received as to the offer + to distribute corresponding source code. (This alternative is + allowed only for noncommercial distribution and only if you + received the program in object code or executable form with such + an offer, in accord with Subsection b above.) + +The source code for a work means the preferred form of the work for +making modifications to it. For an executable work, complete source +code means all the source code for all modules it contains, plus any +associated interface definition files, plus the scripts used to +control compilation and installation of the executable. However, as a +special exception, the source code distributed need not include +anything that is normally distributed (in either source or binary +form) with the major components (compiler, kernel, and so on) of the +operating system on which the executable runs, unless that component +itself accompanies the executable. + +If distribution of executable or object code is made by offering +access to copy from a designated place, then offering equivalent +access to copy the source code from the same place counts as +distribution of the source code, even though third parties are not +compelled to copy the source along with the object code. + + 4. You may not copy, modify, sublicense, or distribute the Program +except as expressly provided under this License. Any attempt +otherwise to copy, modify, sublicense or distribute the Program is +void, and will automatically terminate your rights under this License. +However, parties who have received copies, or rights, from you under +this License will not have their licenses terminated so long as such +parties remain in full compliance. + + 5. You are not required to accept this License, since you have not +signed it. However, nothing else grants you permission to modify or +distribute the Program or its derivative works. These actions are +prohibited by law if you do not accept this License. Therefore, by +modifying or distributing the Program (or any work based on the +Program), you indicate your acceptance of this License to do so, and +all its terms and conditions for copying, distributing or modifying +the Program or works based on it. + + 6. Each time you redistribute the Program (or any work based on the +Program), the recipient automatically receives a license from the +original licensor to copy, distribute or modify the Program subject to +these terms and conditions. You may not impose any further +restrictions on the recipients' exercise of the rights granted herein. +You are not responsible for enforcing compliance by third parties to +this License. + + 7. If, as a consequence of a court judgment or allegation of patent +infringement or for any other reason (not limited to patent issues), +conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot +distribute so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you +may not distribute the Program at all. For example, if a patent +license would not permit royalty-free redistribution of the Program by +all those who receive copies directly or indirectly through you, then +the only way you could satisfy both it and this License would be to +refrain entirely from distribution of the Program. + +If any portion of this section is held invalid or unenforceable under +any particular circumstance, the balance of the section is intended to +apply and the section as a whole is intended to apply in other +circumstances. + +It is not the purpose of this section to induce you to infringe any +patents or other property right claims or to contest validity of any +such claims; this section has the sole purpose of protecting the +integrity of the free software distribution system, which is +implemented by public license practices. Many people have made +generous contributions to the wide range of software distributed +through that system in reliance on consistent application of that +system; it is up to the author/donor to decide if he or she is willing +to distribute software through any other system and a licensee cannot +impose that choice. + +This section is intended to make thoroughly clear what is believed to +be a consequence of the rest of this License. + + 8. If the distribution and/or use of the Program is restricted in +certain countries either by patents or by copyrighted interfaces, the +original copyright holder who places the Program under this License +may add an explicit geographical distribution limitation excluding +those countries, so that distribution is permitted only in or among +countries not thus excluded. In such case, this License incorporates +the limitation as if written in the body of this License. + + 9. The Free Software Foundation may publish revised and/or new versions +of the General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + +Each version is given a distinguishing version number. If the Program +specifies a version number of this License which applies to it and "any +later version", you have the option of following the terms and conditions +either of that version or of any later version published by the Free +Software Foundation. If the Program does not specify a version number of +this License, you may choose any version ever published by the Free Software +Foundation. + + 10. If you wish to incorporate parts of the Program into other free +programs whose distribution conditions are different, write to the author +to ask for permission. For software which is copyrighted by the Free +Software Foundation, write to the Free Software Foundation; we sometimes +make exceptions for this. Our decision will be guided by the two goals +of preserving the free status of all derivatives of our free software and +of promoting the sharing and reuse of software generally. + + NO WARRANTY + + 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY +FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN +OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES +PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED +OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS +TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE +PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, +REPAIR OR CORRECTION. + + 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR +REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, +INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING +OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED +TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY +YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER +PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE +POSSIBILITY OF SUCH DAMAGES. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +convey the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + + +Also add information on how to contact you by electronic and paper mail. + +If the program is interactive, make it output a short notice like this +when it starts in an interactive mode: + + Gnomovision version 69, Copyright (C) year name of author + Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, the commands you use may +be called something other than `show w' and `show c'; they could even be +mouse-clicks or menu items--whatever suits your program. + +You should also get your employer (if you work as a programmer) or your +school, if any, to sign a "copyright disclaimer" for the program, if +necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the program + `Gnomovision' (which makes passes at compilers) written by James Hacker. + + , 1 April 1989 + Ty Coon, President of Vice + +This General Public License does not permit incorporating your program into +proprietary programs. If your program is a subroutine library, you may +consider it more useful to permit linking proprietary applications with the +library. If this is what you want to do, use the GNU Library General +Public License instead of this License. + + diff --git a/plugins/PadNull/Linux/Config.cpp b/plugins/PadNull/Linux/Config.cpp new file mode 100644 index 0000000000..05f91b2528 --- /dev/null +++ b/plugins/PadNull/Linux/Config.cpp @@ -0,0 +1,168 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include +#include +#include +#include +#include +using namespace std; + +#include "Pad.h" +#include "Config.h" + + #ifdef __cplusplus +extern "C" +{ +#endif + +#include "support.h" +#include "callbacks.h" +#include "interface.h" + +#ifdef __cplusplus +} +#endif + +GtkWidget *MsgDlg, *About, *Conf; +extern string s_strIniPath; + +void OnMsg_Ok() +{ + gtk_widget_destroy(MsgDlg); + gtk_main_quit(); +} + +void SysMessage(char *fmt, ...) +{ + GtkWidget *Ok,*Txt; + GtkWidget *Box,*Box1; + va_list list; + char msg[512]; + + va_start(list, fmt); + vsprintf(msg, fmt, list); + va_end(list); + + if (msg[strlen(msg) - 1] == '\n') msg[strlen(msg)-1] = 0; + + MsgDlg = gtk_window_new (GTK_WINDOW_POPUP); + gtk_window_set_position(GTK_WINDOW(MsgDlg), GTK_WIN_POS_CENTER); + gtk_window_set_title(GTK_WINDOW(MsgDlg), "FireWire Msg"); + gtk_container_set_border_width(GTK_CONTAINER(MsgDlg), 5); + + Box = gtk_vbox_new(5, 0); + gtk_container_add(GTK_CONTAINER(MsgDlg), Box); + gtk_widget_show(Box); + + Txt = gtk_label_new(msg); + + gtk_box_pack_start(GTK_BOX(Box), Txt, FALSE, FALSE, 5); + gtk_widget_show(Txt); + + Box1 = gtk_hbutton_box_new(); + gtk_box_pack_start(GTK_BOX(Box), Box1, FALSE, FALSE, 0); + gtk_widget_show(Box1); + + Ok = gtk_button_new_with_label("Ok"); + gtk_signal_connect (GTK_OBJECT(Ok), "clicked", GTK_SIGNAL_FUNC(OnMsg_Ok), NULL); + gtk_container_add(GTK_CONTAINER(Box1), Ok); + GTK_WIDGET_SET_FLAGS(Ok, GTK_CAN_DEFAULT); + gtk_widget_show(Ok); + + gtk_widget_show(MsgDlg); + + gtk_main(); +} + +void OnAbout_Ok(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(About); + gtk_main_quit(); +} + +EXPORT_C_(void) PADabout() +{ + About = create_About(); + gtk_widget_show_all(About); + gtk_main(); +} + +void OnConf_Ok(GtkButton *button, gpointer user_data) +{ + conf.Log = is_checked(Conf, "check_logging"); + SaveConfig(); + + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +void OnConf_Cancel(GtkButton *button, gpointer user_data) +{ + gtk_widget_destroy(Conf); + gtk_main_quit(); +} + +EXPORT_C_(void) PADconfigure() +{ + LoadConfig(); + Conf = create_Config(); + + set_checked(Conf, "check_logging", conf.Log); + gtk_widget_show_all(Conf); + gtk_main(); +} + +void LoadConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg, "r"); + if (f == NULL) + { + printf("failed to open %s\n", s_strIniPath.c_str()); + SaveConfig();//save and return + return; + } + fscanf(f, "logging = %hhx\n", &conf.Log); + //fscanf(f, "options = %hhx\n", &confOptions); + fclose(f); +} + +void SaveConfig() +{ + FILE *f; + char cfg[255]; + + strcpy(cfg, s_strIniPath.c_str()); + f = fopen(cfg,"w"); + if (f == NULL) + { + printf("failed to open '%s'\n", s_strIniPath.c_str()); + return; + } + fprintf(f, "logging = %hhx\n", conf.Log); + //fprintf(f, "options = %hhx\n", confOptions); + fclose(f); +} + diff --git a/plugins/PadNull/Linux/Config.h b/plugins/PadNull/Linux/Config.h new file mode 100644 index 0000000000..7251ed01eb --- /dev/null +++ b/plugins/PadNull/Linux/Config.h @@ -0,0 +1,24 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +void SaveConf(); +void LoadConf(); +void SysMessage(char *fmt, ...); + +#define is_checked(main_widget, widget_name) (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)))) +#define set_checked(main_widget,widget_name, state) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(main_widget, widget_name)), state) diff --git a/plugins/PadNull/Linux/PadLinux.cpp b/plugins/PadNull/Linux/PadLinux.cpp new file mode 100644 index 0000000000..87003a772b --- /dev/null +++ b/plugins/PadNull/Linux/PadLinux.cpp @@ -0,0 +1,83 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "PadLinux.h" + +Display *GSdsp; +int autoRepeatMode; + +void _PadUpdate(int pad) +{ + XEvent evt; + KeySym key; + + // keyboard input + while (XPending(GSdsp) > 0) + { + XNextEvent(GSdsp, &evt); + switch (evt.type) + { + case KeyPress: + key = XLookupKeysym((XKeyEvent *) &evt, 0); + + // Add code to check if it's one of the keys we configured here on a real pda plugin.. + + event.evt = KEYPRESS; + event.key = key; + break; + + case KeyRelease: + key = XLookupKeysym((XKeyEvent *) &evt, 0); + + // Add code to check if it's one of the keys we configured here on a real pda plugin.. + + event.evt = KEYRELEASE; + event.key = key; + break; + + case FocusIn: + XAutoRepeatOff(GSdsp); + break; + + case FocusOut: + XAutoRepeatOn(GSdsp); + break; + } + } +} + +s32 _PADOpen(void *pDsp) +{ + //XKeyboardState *settings; + + GSdsp = *(Display**)pDsp; + //XGetKeyboardControl(GSdsp, settings); + //autoRepeatMode = settings->global_auto_repeat; + XAutoRepeatOff(GSdsp); + + return 0; +} + +void _PADClose() +{ + XAutoRepeatOn(GSdsp); + //XKeyboardControl *settings; + //settings->auto_repeat_mode = autoRepeatMode; + + //XChangeKeyboardControl(GSdsp, KBAutoRepeatMode, settings); +} \ No newline at end of file diff --git a/plugins/PadNull/Linux/PadLinux.h b/plugins/PadNull/Linux/PadLinux.h new file mode 100644 index 0000000000..daa4c72b9f --- /dev/null +++ b/plugins/PadNull/Linux/PadLinux.h @@ -0,0 +1,30 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __PADLINUX_H__ +#define __PADLINUX_H__ + +#include "Pad.h" +#include +#include + +void _PadUpdate(int pad); +s32 _PADOpen(void *pDsp); +void _PADClose(); + +#endif \ No newline at end of file diff --git a/plugins/PadNull/Linux/callbacks.c b/plugins/PadNull/Linux/callbacks.c new file mode 100644 index 0000000000..e87ed8be2d --- /dev/null +++ b/plugins/PadNull/Linux/callbacks.c @@ -0,0 +1,34 @@ +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data) +{ + +} + + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data) +{ + +} + diff --git a/plugins/PadNull/Linux/callbacks.h b/plugins/PadNull/Linux/callbacks.h new file mode 100644 index 0000000000..e606c242f0 --- /dev/null +++ b/plugins/PadNull/Linux/callbacks.h @@ -0,0 +1,14 @@ +#include + + +void +OnConf_Ok (GtkButton *button, + gpointer user_data); + +void +OnConf_Cancel (GtkButton *button, + gpointer user_data); + +void +OnAbout_Ok (GtkButton *button, + gpointer user_data); diff --git a/plugins/PadNull/Linux/interface.c b/plugins/PadNull/Linux/interface.c new file mode 100644 index 0000000000..18efe01d59 --- /dev/null +++ b/plugins/PadNull/Linux/interface.c @@ -0,0 +1,173 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include +#include + +#include "callbacks.h" +#include "interface.h" +#include "support.h" + +#define GLADE_HOOKUP_OBJECT(component,widget,name) \ + g_object_set_data_full (G_OBJECT (component), name, \ + gtk_widget_ref (widget), (GDestroyNotify) gtk_widget_unref) + +#define GLADE_HOOKUP_OBJECT_NO_REF(component,widget,name) \ + g_object_set_data (G_OBJECT (component), name, widget) + +GtkWidget* +create_Config (void) +{ + GtkWidget *Config; + GtkWidget *vbox1; + GtkWidget *frame3; + GtkWidget *alignment1; + GtkWidget *check_logging; + GtkWidget *label15; + GtkWidget *hbuttonbox1; + GtkWidget *button1; + GtkWidget *button2; + + Config = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (Config, "Config"); + gtk_container_set_border_width (GTK_CONTAINER (Config), 5); + gtk_window_set_title (GTK_WINDOW (Config), _("Padconfig")); + + vbox1 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox1, "vbox1"); + gtk_widget_show (vbox1); + gtk_container_add (GTK_CONTAINER (Config), vbox1); + gtk_container_set_border_width (GTK_CONTAINER (vbox1), 5); + + frame3 = gtk_frame_new (NULL); + gtk_widget_set_name (frame3, "frame3"); + gtk_widget_show (frame3); + gtk_box_pack_start (GTK_BOX (vbox1), frame3, TRUE, TRUE, 0); + gtk_frame_set_shadow_type (GTK_FRAME (frame3), GTK_SHADOW_NONE); + + alignment1 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment1, "alignment1"); + gtk_widget_show (alignment1); + gtk_container_add (GTK_CONTAINER (frame3), alignment1); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment1), 0, 0, 12, 0); + + check_logging = gtk_check_button_new_with_mnemonic (_("Enable Logging")); + gtk_widget_set_name (check_logging, "check_logging"); + gtk_widget_show (check_logging); + gtk_container_add (GTK_CONTAINER (alignment1), check_logging); + + label15 = gtk_label_new (_("Logging")); + gtk_widget_set_name (label15, "label15"); + gtk_widget_show (label15); + gtk_frame_set_label_widget (GTK_FRAME (frame3), label15); + gtk_label_set_use_markup (GTK_LABEL (label15), TRUE); + + hbuttonbox1 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox1, "hbuttonbox1"); + gtk_widget_show (hbuttonbox1); + gtk_box_pack_start (GTK_BOX (vbox1), hbuttonbox1, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox1), 30); + + button1 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button1, "button1"); + gtk_widget_show (button1); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button1); + GTK_WIDGET_SET_FLAGS (button1, GTK_CAN_DEFAULT); + + button2 = gtk_button_new_with_mnemonic (_("Cancel")); + gtk_widget_set_name (button2, "button2"); + gtk_widget_show (button2); + gtk_container_add (GTK_CONTAINER (hbuttonbox1), button2); + GTK_WIDGET_SET_FLAGS (button2, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button1, "clicked", + G_CALLBACK (OnConf_Ok), + NULL); + g_signal_connect ((gpointer) button2, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (Config, Config, "Config"); + GLADE_HOOKUP_OBJECT (Config, vbox1, "vbox1"); + GLADE_HOOKUP_OBJECT (Config, frame3, "frame3"); + GLADE_HOOKUP_OBJECT (Config, alignment1, "alignment1"); + GLADE_HOOKUP_OBJECT (Config, check_logging, "check_logging"); + GLADE_HOOKUP_OBJECT (Config, label15, "label15"); + GLADE_HOOKUP_OBJECT (Config, hbuttonbox1, "hbuttonbox1"); + GLADE_HOOKUP_OBJECT (Config, button1, "button1"); + GLADE_HOOKUP_OBJECT (Config, button2, "button2"); + + return Config; +} + +GtkWidget* +create_About (void) +{ + GtkWidget *About; + GtkWidget *vbox2; + GtkWidget *label2; + GtkWidget *label3; + GtkWidget *hbuttonbox2; + GtkWidget *button3; + + About = gtk_window_new (GTK_WINDOW_TOPLEVEL); + gtk_widget_set_name (About, "About"); + gtk_container_set_border_width (GTK_CONTAINER (About), 5); + gtk_window_set_title (GTK_WINDOW (About), _("FWabout")); + + vbox2 = gtk_vbox_new (FALSE, 5); + gtk_widget_set_name (vbox2, "vbox2"); + gtk_widget_show (vbox2); + gtk_container_add (GTK_CONTAINER (About), vbox2); + gtk_container_set_border_width (GTK_CONTAINER (vbox2), 5); + + label2 = gtk_label_new (_("PadNull Driver")); + gtk_widget_set_name (label2, "label2"); + gtk_widget_show (label2); + gtk_box_pack_start (GTK_BOX (vbox2), label2, FALSE, FALSE, 0); + gtk_label_set_justify (GTK_LABEL (label2), GTK_JUSTIFY_CENTER); + + label3 = gtk_label_new (_("Original Author: arcum42@gmail.com")); + gtk_widget_set_name (label3, "label3"); + gtk_widget_show (label3); + gtk_box_pack_start (GTK_BOX (vbox2), label3, FALSE, FALSE, 0); + + hbuttonbox2 = gtk_hbutton_box_new (); + gtk_widget_set_name (hbuttonbox2, "hbuttonbox2"); + gtk_widget_show (hbuttonbox2); + gtk_box_pack_start (GTK_BOX (vbox2), hbuttonbox2, TRUE, TRUE, 0); + gtk_box_set_spacing (GTK_BOX (hbuttonbox2), 30); + + button3 = gtk_button_new_with_mnemonic (_("Ok")); + gtk_widget_set_name (button3, "button3"); + gtk_widget_show (button3); + gtk_container_add (GTK_CONTAINER (hbuttonbox2), button3); + GTK_WIDGET_SET_FLAGS (button3, GTK_CAN_DEFAULT); + + g_signal_connect ((gpointer) button3, "clicked", + G_CALLBACK (OnAbout_Ok), + NULL); + + /* Store pointers to all widgets, for use by lookup_widget(). */ + GLADE_HOOKUP_OBJECT_NO_REF (About, About, "About"); + GLADE_HOOKUP_OBJECT (About, vbox2, "vbox2"); + GLADE_HOOKUP_OBJECT (About, label2, "label2"); + GLADE_HOOKUP_OBJECT (About, label3, "label3"); + GLADE_HOOKUP_OBJECT (About, hbuttonbox2, "hbuttonbox2"); + GLADE_HOOKUP_OBJECT (About, button3, "button3"); + + return About; +} + diff --git a/plugins/PadNull/Linux/interface.h b/plugins/PadNull/Linux/interface.h new file mode 100644 index 0000000000..69c303ba3b --- /dev/null +++ b/plugins/PadNull/Linux/interface.h @@ -0,0 +1,6 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +GtkWidget* create_Config (void); +GtkWidget* create_About (void); diff --git a/plugins/PadNull/Linux/pad.glade b/plugins/PadNull/Linux/pad.glade new file mode 100644 index 0000000000..8a407e0fb7 --- /dev/null +++ b/plugins/PadNull/Linux/pad.glade @@ -0,0 +1,240 @@ + + + + + + + 5 + True + Padconfig + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + 0 + 0.5 + GTK_SHADOW_NONE + + + + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + True + Enable Logging + True + GTK_RELIEF_NORMAL + True + False + False + True + + + + + + + + True + <b>Logging</b> + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + + True + True + True + Cancel + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + + 5 + True + FWabout + GTK_WINDOW_TOPLEVEL + GTK_WIN_POS_NONE + False + True + False + True + False + False + GDK_WINDOW_TYPE_HINT_NORMAL + GDK_GRAVITY_NORTH_WEST + True + False + + + + 5 + True + False + 5 + + + + True + PadNull Driver + False + False + GTK_JUSTIFY_CENTER + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + Original Author: arcum42@gmail.com + False + False + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + 0 + False + False + + + + + + True + GTK_BUTTONBOX_DEFAULT_STYLE + 30 + + + + True + True + True + Ok + True + GTK_RELIEF_NORMAL + True + + + + + + 0 + True + True + + + + + + + diff --git a/plugins/PadNull/Linux/support.c b/plugins/PadNull/Linux/support.c new file mode 100644 index 0000000000..00aff29822 --- /dev/null +++ b/plugins/PadNull/Linux/support.c @@ -0,0 +1,144 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include +#include +#include +#include + +#include + +#include "support.h" + +GtkWidget* +lookup_widget (GtkWidget *widget, + const gchar *widget_name) +{ + GtkWidget *parent, *found_widget; + + for (;;) + { + if (GTK_IS_MENU (widget)) + parent = gtk_menu_get_attach_widget (GTK_MENU (widget)); + else + parent = widget->parent; + if (!parent) + parent = (GtkWidget*) g_object_get_data (G_OBJECT (widget), "GladeParentKey"); + if (parent == NULL) + break; + widget = parent; + } + + found_widget = (GtkWidget*) g_object_get_data (G_OBJECT (widget), + widget_name); + if (!found_widget) + g_warning ("Widget not found: %s", widget_name); + return found_widget; +} + +static GList *pixmaps_directories = NULL; + +/* Use this function to set the directory containing installed pixmaps. */ +void +add_pixmap_directory (const gchar *directory) +{ + pixmaps_directories = g_list_prepend (pixmaps_directories, + g_strdup (directory)); +} + +/* This is an internally used function to find pixmap files. */ +static gchar* +find_pixmap_file (const gchar *filename) +{ + GList *elem; + + /* We step through each of the pixmaps directory to find it. */ + elem = pixmaps_directories; + while (elem) + { + gchar *pathname = g_strdup_printf ("%s%s%s", (gchar*)elem->data, + G_DIR_SEPARATOR_S, filename); + if (g_file_test (pathname, G_FILE_TEST_EXISTS)) + return pathname; + g_free (pathname); + elem = elem->next; + } + return NULL; +} + +/* This is an internally used function to create pixmaps. */ +GtkWidget* +create_pixmap (GtkWidget *widget, + const gchar *filename) +{ + gchar *pathname = NULL; + GtkWidget *pixmap; + + if (!filename || !filename[0]) + return gtk_image_new (); + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return gtk_image_new (); + } + + pixmap = gtk_image_new_from_file (pathname); + g_free (pathname); + return pixmap; +} + +/* This is an internally used function to create pixmaps. */ +GdkPixbuf* +create_pixbuf (const gchar *filename) +{ + gchar *pathname = NULL; + GdkPixbuf *pixbuf; + GError *error = NULL; + + if (!filename || !filename[0]) + return NULL; + + pathname = find_pixmap_file (filename); + + if (!pathname) + { + g_warning (_("Couldn't find pixmap file: %s"), filename); + return NULL; + } + + pixbuf = gdk_pixbuf_new_from_file (pathname, &error); + if (!pixbuf) + { + fprintf (stderr, "Failed to load pixbuf file: %s: %s\n", + pathname, error->message); + g_error_free (error); + } + g_free (pathname); + return pixbuf; +} + +/* This is used to set ATK action descriptions. */ +void +glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description) +{ + gint n_actions, i; + + n_actions = atk_action_get_n_actions (action); + for (i = 0; i < n_actions; i++) + { + if (!strcmp (atk_action_get_name (action, i), action_name)) + atk_action_set_description (action, i, description); + } +} + diff --git a/plugins/PadNull/Linux/support.h b/plugins/PadNull/Linux/support.h new file mode 100644 index 0000000000..a32649e53c --- /dev/null +++ b/plugins/PadNull/Linux/support.h @@ -0,0 +1,69 @@ +/* + * DO NOT EDIT THIS FILE - it is generated by Glade. + */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* + * Standard gettext macros. + */ +#ifdef ENABLE_NLS +# include +# undef _ +# define _(String) dgettext (PACKAGE, String) +# define Q_(String) g_strip_context ((String), gettext (String)) +# ifdef gettext_noop +# define N_(String) gettext_noop (String) +# else +# define N_(String) (String) +# endif +#else +# define textdomain(String) (String) +# define gettext(String) (String) +# define dgettext(Domain,Message) (Message) +# define dcgettext(Domain,Message,Type) (Message) +# define bindtextdomain(Domain,Directory) (Domain) +# define _(String) (String) +# define Q_(String) g_strip_context ((String), (String)) +# define N_(String) (String) +#endif + + +/* + * Public Functions. + */ + +/* + * This function returns a widget in a component created by Glade. + * Call it with the toplevel widget in the component (i.e. a window/dialog), + * or alternatively any widget in the component, and the name of the widget + * you want returned. + */ +GtkWidget* lookup_widget (GtkWidget *widget, + const gchar *widget_name); + + +/* Use this function to set the directory containing installed pixmaps. */ +void add_pixmap_directory (const gchar *directory); + + +/* + * Private Functions. + */ + +/* This is used to create the pixmaps used in the interface. */ +GtkWidget* create_pixmap (GtkWidget *widget, + const gchar *filename); + +/* This is used to create the pixbufs used in the interface. */ +GdkPixbuf* create_pixbuf (const gchar *filename); + +/* This is used to set ATK action descriptions. */ +void glade_set_atk_action_description (AtkAction *action, + const gchar *action_name, + const gchar *description); + diff --git a/plugins/PadNull/Makefile.am b/plugins/PadNull/Makefile.am new file mode 100644 index 0000000000..6263b08c0a --- /dev/null +++ b/plugins/PadNull/Makefile.am @@ -0,0 +1,31 @@ +# Create a shared library libPadnull +AUTOMAKE_OPTIONS = foreign +noinst_LIBRARIES = libPadnull.a +INCLUDES = -I@srcdir@/../../common/include -I@srcdir@/../../3rdparty -I@srcdir@/Linux + +libPadnull_a_CXXFLAGS = $(shell pkg-config --cflags gtk+-2.0) +libPadnull_a_CFLAGS = $(shell pkg-config --cflags gtk+-2.0) + +# Create a shared object by faking an exe (thanks to ODE makefiles) +traplibdir=$(prefix) + +if DEBUGBUILD +preext=d +endif + +EXEEXT=$(preext)@so_ext@ + +traplib_PROGRAMS=libPadnull +libPadnull_SOURCES= +libPadnull_DEPENDENCIES = libPadnull.a +libPadnull_LDFLAGS= @SHARED_LDFLAGS@ +libPadnull_LDFLAGS+=-Wl,-soname,@libPadnull_SONAME@ +libPadnull_LDADD=$(libPadnull_a_OBJECTS) + +libPadnull_a_SOURCES = Pad.cpp Pad.h Linux/Config.cpp Linux/Config.h \ +Linux/PadLinux.cpp Linux/PadLinux.h + +libPadnull_a_SOURCES += \ +Linux/interface.h Linux/support.c \ +Linux/interface.c Linux/support.h \ +Linux/callbacks.h \ No newline at end of file diff --git a/plugins/PadNull/Pad.cpp b/plugins/PadNull/Pad.cpp new file mode 100644 index 0000000000..9c4d52afb6 --- /dev/null +++ b/plugins/PadNull/Pad.cpp @@ -0,0 +1,170 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include +#include +#include +#include +using namespace std; + +#include "Pad.h" + +const u8 version = PS2E_PAD_VERSION; +const u8 revision = 0; +const u8 build = 1; // increase that with each version + +static char *libraryName = "Padnull Driver"; +string s_strIniPath="inis/Padnull.ini"; + +FILE *padLog; +Config conf; +keyEvent event; +static keyEvent s_event; + +EXPORT_C_(u32) PS2EgetLibType() +{ + return PS2E_LT_PAD; +} + +EXPORT_C_(char*) PS2EgetLibName() +{ + return libraryName; +} + +EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) +{ + return (version<<16) | (revision<<8) | build; +} + +void __Log(const char *fmt, ...) +{ + va_list list; + + if (padLog == NULL) return; + va_start(list, fmt); + vfprintf(padLog, fmt, list); + va_end(list); +} + +void __LogToConsole(const char *fmt, ...) +{ + va_list list; + + va_start(list, fmt); + + if (padLog != NULL) vfprintf(padLog, fmt, list); + + printf("PadNull: "); + vprintf(fmt, list); + va_end(list); +} + +EXPORT_C_(s32) PADinit(u32 flags) +{ + LoadConfig(); + +#ifdef PAD_LOG + if (padLog == NULL) + { + padLog = fopen("logs/padLog.txt", "w"); + if (padLog) setvbuf(padLog, NULL, _IONBF, 0); + } + PAD_LOG("PADinit\n"); +#endif + + return 0; +} + +EXPORT_C_(void) PADshutdown() +{ +#ifdef PAD_LOG + if (padLog != NULL) + { + fclose(padLog); + padLog = NULL; + } +#endif +} + +EXPORT_C_(s32) PADopen(void *pDsp) +{ + memset(&event, 0, sizeof(event)); + + return _PADOpen(pDsp); + + return 0; +} + +EXPORT_C_(void) PADclose() +{ + _PADClose(); +} + +// PADkeyEvent is called every vsync (return NULL if no event) +EXPORT_C_(keyEvent*) PADkeyEvent() +{ + + s_event = event; + event.evt = 0; + + return &s_event; +} + +EXPORT_C_(u8) PADstartPoll(int pad) +{ + return 0; +} + +EXPORT_C_(u8) PADpoll(u8 value) +{ + return 0; +} + +// call to give a hint to the PAD plugin to query for the keyboard state. A +// good plugin will query the OS for keyboard state ONLY in this function. +// This function is necessary when multithreading because otherwise +// the PAD plugin can get into deadlocks with the thread that really owns +// the window (and input). Note that PADupdate can be called from a different +// thread than the other functions, so mutex or other multithreading primitives +// have to be added to maintain data integrity. +EXPORT_C_(u32) PADquery() +// returns: 1 if supported pad1 +// 2 if supported pad2 +// 3 if both are supported +{ + return 3; +} + +EXPORT_C_(void) PADupdate(int pad) +{ + _PadUpdate(pad); +} + +EXPORT_C_(void) PADgsDriverInfo(GSdriverInfo *info) +{ +} + +EXPORT_C_(s32) PADfreeze(int mode, freezeData *data) +{ + return 0; +} + +EXPORT_C_(s32) PADtest() +{ + return 0; +} diff --git a/plugins/PadNull/Pad.h b/plugins/PadNull/Pad.h new file mode 100644 index 0000000000..4d503611db --- /dev/null +++ b/plugins/PadNull/Pad.h @@ -0,0 +1,71 @@ +/* FWnull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __PAD_H__ +#define __PAD_H__ + +#include + +#ifdef __cplusplus +extern "C" +{ +#endif +#define PADdefs +#include "PS2Edefs.h" +#ifdef __cplusplus +} +#endif + +#ifdef _WIN32 + +#include +#include + +#else +#include "PadLinux.h" +#endif + +/*#ifdef _MSC_VER +#define EXPORT_C_(type) extern "C" __declspec(dllexport) type CALLBACK +#else +#define EXPORT_C_(type) extern "C" type +#endif*/ + +#ifdef _MSC_VER +#define EXPORT_C_(type) extern "C" type CALLBACK +#else +#define EXPORT_C_(type) extern "C" type +#endif + +#define PAD_LOG __Log + +typedef struct +{ + s32 Log; +} Config; + +extern Config conf; +extern FILE *padLog; +extern keyEvent event; + +extern void __Log(char *fmt, ...); +extern void SysMessage(char *fmt, ...); +extern void SaveConfig(); +extern void LoadConfig(); + +#endif diff --git a/plugins/PadNull/ReadMe.txt b/plugins/PadNull/ReadMe.txt new file mode 100644 index 0000000000..75019a2573 --- /dev/null +++ b/plugins/PadNull/ReadMe.txt @@ -0,0 +1,23 @@ +PadNull v0.1 +------------- + + This is an extension to use with play station2 emulators + as PCSX2 (only one right now). + The plugin is free open source code. + +Usage: +----- + Place the file "libPadNull.so.0.1.0" (linux) or "PadNull.dll" (win32) + at the Plugin directory of the Emulator to use it. + +Changes: +------- + v0.1: + * First Release + * Tested with Pcsx2 + +Authors: +------- +arcum42@gmail.com + + diff --git a/plugins/PadNull/Windows/Config.cpp b/plugins/PadNull/Windows/Config.cpp new file mode 100644 index 0000000000..62db91dd61 --- /dev/null +++ b/plugins/PadNull/Windows/Config.cpp @@ -0,0 +1,51 @@ +#include + +#include "../FW.h" + +extern HINSTANCE hInst; +void SaveConfig() +{ + + Config *Conf1 = &conf; + char *szTemp; + char szIniFile[256], szValue[256]; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if(!szTemp) return; + strcpy(szTemp, "\\inis\\fwnull.ini"); + sprintf(szValue,"%u",Conf1->Log); + WritePrivateProfileString("Interface", "Logging",szValue,szIniFile); + +} + +void LoadConfig() { + FILE *fp; + + + Config *Conf1 = &conf; + char *szTemp; + char szIniFile[256], szValue[256]; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if(!szTemp) return ; + strcpy(szTemp, "\\inis\\fwnull.ini"); + fp=fopen("inis\\fwnull.ini","rt");//check if firewirenull.ini really exists + if (!fp) + { + CreateDirectory("inis",NULL); + memset(&conf, 0, sizeof(conf)); + conf.Log = 0;//default value + SaveConfig();//save and return + return ; + } + fclose(fp); + GetPrivateProfileString("Interface", "Logging", NULL, szValue, 20, szIniFile); + Conf1->Log = strtoul(szValue, NULL, 10); + return ; + +} + diff --git a/plugins/PadNull/Windows/PadNull.def b/plugins/PadNull/Windows/PadNull.def new file mode 100644 index 0000000000..51e786a1fd --- /dev/null +++ b/plugins/PadNull/Windows/PadNull.def @@ -0,0 +1,19 @@ +; FireWire.def : Declares the module parameters for the DLL. + +LIBRARY "FWnull" + +EXPORTS + ; Explicit exports can go here + PS2EgetLibType @2 + PS2EgetLibName @3 + PS2EgetLibVersion2 @4 + FWinit @5 + FWshutdown @6 + FWopen @7 + FWclose @8 + FWread32 @11 + FWwrite32 @14 + FWirqCallback @15 + FWconfigure @17 + FWtest @18 + FWabout @19 diff --git a/plugins/PadNull/Windows/PadNull.rc b/plugins/PadNull/Windows/PadNull.rc new file mode 100644 index 0000000000..a505f6108d --- /dev/null +++ b/plugins/PadNull/Windows/PadNull.rc @@ -0,0 +1,116 @@ +// Microsoft Visual C++ generated resource script. +// +#include "resource.h" + +#define APSTUDIO_READONLY_SYMBOLS +///////////////////////////////////////////////////////////////////////////// +// +// Generated from the TEXTINCLUDE 2 resource. +// +#include "afxresmw.h" + +///////////////////////////////////////////////////////////////////////////// +#undef APSTUDIO_READONLY_SYMBOLS + +///////////////////////////////////////////////////////////////////////////// +// Spanish (Argentina) resources + +#if !defined(AFX_RESOURCE_DLL) || defined(AFX_TARG_ESS) +#ifdef _WIN32 +LANGUAGE LANG_SPANISH, SUBLANG_SPANISH_ARGENTINA +#pragma code_page(1252) +#endif //_WIN32 + +#ifdef APSTUDIO_INVOKED +///////////////////////////////////////////////////////////////////////////// +// +// TEXTINCLUDE +// + +1 TEXTINCLUDE +BEGIN + "resource.h\0" +END + +2 TEXTINCLUDE +BEGIN + "#include ""afxresmw.h""\r\0" +END + +3 TEXTINCLUDE +BEGIN + "\r\0" +END + +#endif // APSTUDIO_INVOKED + + +///////////////////////////////////////////////////////////////////////////// +// +// Dialog +// + +IDD_CONFIG DIALOGEX 0, 0, 212, 121 +STYLE DS_SETFONT | DS_MODALFRAME | WS_POPUP | WS_CAPTION | WS_SYSMENU +CAPTION "Firewireconfigure" +FONT 8, "MS Sans Serif", 0, 0, 0x0 +BEGIN + DEFPUSHBUTTON "OK",IDOK,48,100,50,14 + PUSHBUTTON "Cancel",IDCANCEL,113,100,50,14 + CONTROL "Enable Logging (for develop use only)",IDC_LOGGING, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,7,187,13 +END + +IDD_ABOUT DIALOGEX 0, 0, 177, 106 +STYLE DS_SETFONT | DS_MODALFRAME | DS_CENTER | WS_POPUP | WS_CAPTION | WS_SYSMENU +CAPTION "FireWire About" +FONT 8, "MS Sans Serif", 0, 0, 0x0 +BEGIN + DEFPUSHBUTTON "OK",IDOK,65,85,50,14 + LTEXT "FireWire Driver",IDC_NAME,70,10,48,8 + GROUPBOX "",IDC_STATIC,5,35,170,40 + LTEXT "Author: Shadow and linuzappz",IDC_STATIC,29,19,141,10 +END + + +///////////////////////////////////////////////////////////////////////////// +// +// DESIGNINFO +// + +#ifdef APSTUDIO_INVOKED +GUIDELINES DESIGNINFO +BEGIN + IDD_CONFIG, DIALOG + BEGIN + LEFTMARGIN, 7 + RIGHTMARGIN, 205 + TOPMARGIN, 7 + BOTTOMMARGIN, 114 + END + + IDD_ABOUT, DIALOG + BEGIN + LEFTMARGIN, 7 + RIGHTMARGIN, 170 + TOPMARGIN, 7 + BOTTOMMARGIN, 99 + END +END +#endif // APSTUDIO_INVOKED + +#endif // Spanish (Argentina) resources +///////////////////////////////////////////////////////////////////////////// + + + +#ifndef APSTUDIO_INVOKED +///////////////////////////////////////////////////////////////////////////// +// +// Generated from the TEXTINCLUDE 3 resource. +// + + +///////////////////////////////////////////////////////////////////////////// +#endif // not APSTUDIO_INVOKED + diff --git a/plugins/PadNull/Windows/PadNull_vc2008.cbp b/plugins/PadNull/Windows/PadNull_vc2008.cbp new file mode 100644 index 0000000000..a0815f31d6 --- /dev/null +++ b/plugins/PadNull/Windows/PadNull_vc2008.cbp @@ -0,0 +1,44 @@ + + + + + + diff --git a/plugins/PadNull/Windows/PadNull_vc2008.layout b/plugins/PadNull/Windows/PadNull_vc2008.layout new file mode 100644 index 0000000000..d55539c58b --- /dev/null +++ b/plugins/PadNull/Windows/PadNull_vc2008.layout @@ -0,0 +1,4 @@ + + + + diff --git a/plugins/PadNull/Windows/PadNull_vc2008.vcproj b/plugins/PadNull/Windows/PadNull_vc2008.vcproj new file mode 100644 index 0000000000..7ca94e36c7 --- /dev/null +++ b/plugins/PadNull/Windows/PadNull_vc2008.vcproj @@ -0,0 +1,201 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/plugins/PadNull/Windows/ProjectRootDir.vsprops b/plugins/PadNull/Windows/ProjectRootDir.vsprops new file mode 100644 index 0000000000..b8400f4c9c --- /dev/null +++ b/plugins/PadNull/Windows/ProjectRootDir.vsprops @@ -0,0 +1,11 @@ + + + + diff --git a/plugins/PadNull/Windows/Win32.cpp b/plugins/PadNull/Windows/Win32.cpp new file mode 100644 index 0000000000..c09edf89f0 --- /dev/null +++ b/plugins/PadNull/Windows/Win32.cpp @@ -0,0 +1,81 @@ +#include +#include +#include + + +#include "resource.h" +#include "../FW.h" + +HINSTANCE hInst; + +void SysMessage(char *fmt, ...) { + va_list list; + char tmp[512]; + + va_start(list,fmt); + vsprintf(tmp,fmt,list); + va_end(list); + MessageBox(0, tmp, "FW Plugin Msg", 0); +} + +BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { + + switch(uMsg) { + case WM_INITDIALOG: + LoadConfig(); + if (conf.Log) CheckDlgButton(hW, IDC_LOGGING, TRUE); + return TRUE; + + case WM_COMMAND: + switch(LOWORD(wParam)) { + case IDCANCEL: + EndDialog(hW, TRUE); + return TRUE; + case IDOK: + if (IsDlgButtonChecked(hW, IDC_LOGGING)) + conf.Log = 1; + else conf.Log = 0; + SaveConfig(); + EndDialog(hW, FALSE); + return TRUE; + } + } + return FALSE; +} + +BOOL CALLBACK AboutDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { + switch(uMsg) { + case WM_INITDIALOG: + return TRUE; + + case WM_COMMAND: + switch(LOWORD(wParam)) { + case IDOK: + EndDialog(hW, FALSE); + return TRUE; + } + } + return FALSE; +} + +void CALLBACK FWconfigure() { + DialogBox(hInst, + MAKEINTRESOURCE(IDD_CONFIG), + GetActiveWindow(), + (DLGPROC)ConfigureDlgProc); +} + +void CALLBACK FWabout() { + DialogBox(hInst, + MAKEINTRESOURCE(IDD_ABOUT), + GetActiveWindow(), + (DLGPROC)AboutDlgProc); +} + +BOOL APIENTRY DllMain(HANDLE hModule, // DLL INIT + DWORD dwReason, + LPVOID lpReserved) { + hInst = (HINSTANCE)hModule; + return TRUE; // very quick :) +} + diff --git a/plugins/PadNull/Windows/resource.h b/plugins/PadNull/Windows/resource.h new file mode 100644 index 0000000000..6ea8fead11 --- /dev/null +++ b/plugins/PadNull/Windows/resource.h @@ -0,0 +1,21 @@ +//{{NO_DEPENDENCIES}} +// Microsoft Visual C++ generated include file. +// Used by FireWireNull.rc +// +#define IDD_CONFDLG 101 +#define IDD_CONFIG 101 +#define IDD_ABOUT 103 +#define IDC_NAME 1000 +#define IDC_CHECK1 1007 +#define IDC_LOGGING 1007 + +// Next default values for new objects +// +#ifdef APSTUDIO_INVOKED +#ifndef APSTUDIO_READONLY_SYMBOLS +#define _APS_NEXT_RESOURCE_VALUE 105 +#define _APS_NEXT_COMMAND_VALUE 40001 +#define _APS_NEXT_CONTROL_VALUE 1008 +#define _APS_NEXT_SYMED_VALUE 101 +#endif +#endif diff --git a/plugins/PadNull/build.sh b/plugins/PadNull/build.sh new file mode 100644 index 0000000000..86f311f53f --- /dev/null +++ b/plugins/PadNull/build.sh @@ -0,0 +1,32 @@ +#!/bin/sh + +echo --------------- +echo Building Padnull +echo --------------- + +curdir=`pwd` + + +if test "${PadnullOPTIONS+set}" != set ; then +export PadnullOPTIONS="" +fi + +if [ $# -gt 0 ] && [ $1 = "all" ] +then + +aclocal +automake -a +autoconf + +./configure ${PadnullOPTIONS} --prefix=${PCSX2PLUGINS} +make clean +make install + +else +make $@ +fi + +if [ $? -ne 0 ] +then +exit 1 +fi \ No newline at end of file diff --git a/plugins/PadNull/configure.ac b/plugins/PadNull/configure.ac new file mode 100644 index 0000000000..e8617e98dd --- /dev/null +++ b/plugins/PadNull/configure.ac @@ -0,0 +1,83 @@ +AC_INIT(Padnull, 0.1,arcum42@gmail.com) + +AM_INIT_AUTOMAKE(Padnull,0.1) + +AC_PROG_CC([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CXX([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) +AC_PROG_CPP([gcc g++ cl KCC CC cxx cc++ xlC aCC c++]) + +AC_PROG_INSTALL +AC_PROG_RANLIB + +dnl necessary for compiling assembly +AM_PROG_AS + +AC_SUBST(Padnull_CURRENT, 0) +AC_SUBST(Padnull_REVISION, 1) +AC_SUBST(Padnull_AGE, 0) +AC_SUBST(Padnull_RELEASE,[$Padnull_CURRENT].[$Padnull_REVISION].[$Padnull_AGE]) +AC_SUBST(Padnull_SONAME,libPadnull.so.[$Padnull_CURRENT].[$Padnull_REVISION].[$Padnull_AGE]) + +CFLAGS= +CPPFLAGS= +CXXFLAGS= +CCASFLAGS= + +dnl Check for debug build +AC_MSG_CHECKING(debug build) +AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), +debug=$enableval,debug=no) +if test "x$debug" == xyes +then + AC_DEFINE(_DEBUG,1,[_DEBUG]) + CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " +else + AC_DEFINE(NDEBUG,1,[NDEBUG]) + CFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CPPFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CXXFLAGS+="-O3 -fomit-frame-pointer -fPIC -Wall -Wno-unused-value -m32 " + CCASFLAGS+=" -m32 " +fi +AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) +AC_MSG_RESULT($debug) + +AC_DEFINE(__LINUX__,1,[__LINUX__]) + +dnl Check for dev build +AC_MSG_CHECKING(for development build...) +AC_ARG_ENABLE(devbuild, AC_HELP_STRING([--enable-devbuild], [Special Build for developers that simplifies testing and adds extra checks]), + devbuild=$enableval,devbuild=no) +if test "x$devbuild" == xyes +then + AC_DEFINE(Padnull_DEVBUILD,1,[Padnull_DEVBUILD]) +fi +AC_MSG_RESULT($devbuild) +AM_CONDITIONAL(RELEASE_TO_PUBLIC, test x$devbuild = xno) + +AC_CHECK_FUNCS([ _aligned_malloc _aligned_free ], AC_DEFINE(HAVE_ALIGNED_MALLOC)) + +dnl gtk +AC_MSG_CHECKING(gtk2+) +AC_CHECK_PROG(GTK_CONFIG, pkg-config, pkg-config) +LIBS+=$(pkg-config --libs gtk+-2.0) + +dnl bindir = pcsx2exe + +dnl assuming linux environment +so_ext=".so.$Padnull_RELEASE" +SHARED_LDFLAGS="-shared" +AC_SUBST(so_ext) +AC_SUBST(SHARED_LDFLAGS) + +AC_CHECK_LIB(stdc++,main,[LIBS="$LIBS -lstdc++"]) + +AC_OUTPUT([ + Makefile + ]) + +echo "Configuration:" +echo " Debug build? $debug" +echo " Dev build? $devbuild" \ No newline at end of file From a180b105d30ae4aa2923eab6746a2e4d1e1ef5fb Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 06:11:48 +0000 Subject: [PATCH 112/296] PadNull: It compiles in Windows. (It does not pass key events, such as escape, though, so should not be used yet.) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1243 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/PadNull/Linux/PadLinux.cpp | 4 +-- plugins/PadNull/Linux/PadLinux.h | 2 +- plugins/PadNull/Pad.h | 5 +-- plugins/PadNull/Windows/Config.cpp | 20 ++++++++++- plugins/PadNull/Windows/PadNull.def | 26 ++++++++------- plugins/PadNull/Windows/PadNull_vc2008.vcproj | 22 +++++++++---- plugins/PadNull/Windows/PadWin.cpp | 33 +++++++++++++++++++ plugins/PadNull/Windows/PadWin.h | 30 +++++++++++++++++ plugins/PadNull/Windows/Win32.cpp | 29 +++++++++++++--- 9 files changed, 140 insertions(+), 31 deletions(-) create mode 100644 plugins/PadNull/Windows/PadWin.cpp create mode 100644 plugins/PadNull/Windows/PadWin.h diff --git a/plugins/PadNull/Linux/PadLinux.cpp b/plugins/PadNull/Linux/PadLinux.cpp index 87003a772b..4a3d23f76f 100644 --- a/plugins/PadNull/Linux/PadLinux.cpp +++ b/plugins/PadNull/Linux/PadLinux.cpp @@ -1,4 +1,4 @@ -/* FWnull +/* PadNull * Copyright (C) 2004-2009 PCSX2 Team * * This program is free software; you can redistribute it and/or modify @@ -80,4 +80,4 @@ void _PADClose() //settings->auto_repeat_mode = autoRepeatMode; //XChangeKeyboardControl(GSdsp, KBAutoRepeatMode, settings); -} \ No newline at end of file +} diff --git a/plugins/PadNull/Linux/PadLinux.h b/plugins/PadNull/Linux/PadLinux.h index daa4c72b9f..6b61c90aee 100644 --- a/plugins/PadNull/Linux/PadLinux.h +++ b/plugins/PadNull/Linux/PadLinux.h @@ -1,4 +1,4 @@ -/* FWnull +/* PadNull * Copyright (C) 2004-2009 PCSX2 Team * * This program is free software; you can redistribute it and/or modify diff --git a/plugins/PadNull/Pad.h b/plugins/PadNull/Pad.h index 4d503611db..4f391f5c25 100644 --- a/plugins/PadNull/Pad.h +++ b/plugins/PadNull/Pad.h @@ -32,10 +32,7 @@ extern "C" #endif #ifdef _WIN32 - -#include -#include - +#include "PadWin.h" #else #include "PadLinux.h" #endif diff --git a/plugins/PadNull/Windows/Config.cpp b/plugins/PadNull/Windows/Config.cpp index 62db91dd61..43621bedd7 100644 --- a/plugins/PadNull/Windows/Config.cpp +++ b/plugins/PadNull/Windows/Config.cpp @@ -1,6 +1,24 @@ +/* PadNull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + #include -#include "../FW.h" +#include "../Pad.h" extern HINSTANCE hInst; void SaveConfig() diff --git a/plugins/PadNull/Windows/PadNull.def b/plugins/PadNull/Windows/PadNull.def index 51e786a1fd..b9d8488af1 100644 --- a/plugins/PadNull/Windows/PadNull.def +++ b/plugins/PadNull/Windows/PadNull.def @@ -1,19 +1,23 @@ ; FireWire.def : Declares the module parameters for the DLL. -LIBRARY "FWnull" +LIBRARY "PadNull" EXPORTS ; Explicit exports can go here PS2EgetLibType @2 PS2EgetLibName @3 PS2EgetLibVersion2 @4 - FWinit @5 - FWshutdown @6 - FWopen @7 - FWclose @8 - FWread32 @11 - FWwrite32 @14 - FWirqCallback @15 - FWconfigure @17 - FWtest @18 - FWabout @19 + PADinit @5 + PADshutdown @6 + PADopen @7 + PADclose @8 + PADkeyEvent @9 + PADstartPoll @10 + PADpoll @11 + PADquery @12 + PADupdate @13 + PADgsDriverInfo @14 + PADfreeze @15 + PADconfigure @16 + PADtest @17 + PADabout @18 \ No newline at end of file diff --git a/plugins/PadNull/Windows/PadNull_vc2008.vcproj b/plugins/PadNull/Windows/PadNull_vc2008.vcproj index 7ca94e36c7..8d8eade5da 100644 --- a/plugins/PadNull/Windows/PadNull_vc2008.vcproj +++ b/plugins/PadNull/Windows/PadNull_vc2008.vcproj @@ -2,9 +2,9 @@ @@ -122,7 +122,7 @@ /> @@ -161,7 +161,11 @@ > + + + + diff --git a/plugins/PadNull/Windows/PadWin.cpp b/plugins/PadNull/Windows/PadWin.cpp new file mode 100644 index 0000000000..1cd4bddb87 --- /dev/null +++ b/plugins/PadNull/Windows/PadWin.cpp @@ -0,0 +1,33 @@ +/* PadNull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "PadWin.h" + +void _PadUpdate(int pad) +{ +} + +s32 _PADOpen(void *pDsp) +{ + + return 0; +} + +void _PADClose() +{ +} \ No newline at end of file diff --git a/plugins/PadNull/Windows/PadWin.h b/plugins/PadNull/Windows/PadWin.h new file mode 100644 index 0000000000..5bd2bd73e1 --- /dev/null +++ b/plugins/PadNull/Windows/PadWin.h @@ -0,0 +1,30 @@ +/* PadNull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __PADLINUX_H__ +#define __PADLINUX_H__ + +#include "../Pad.h" +#include +#include + +void _PadUpdate(int pad); +s32 _PADOpen(void *pDsp); +void _PADClose(); + +#endif \ No newline at end of file diff --git a/plugins/PadNull/Windows/Win32.cpp b/plugins/PadNull/Windows/Win32.cpp index c09edf89f0..3720d897f7 100644 --- a/plugins/PadNull/Windows/Win32.cpp +++ b/plugins/PadNull/Windows/Win32.cpp @@ -1,10 +1,28 @@ +/* PadNull + * Copyright (C) 2004-2009 PCSX2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + #include #include #include #include "resource.h" -#include "../FW.h" +#include "../Pad.h" HINSTANCE hInst; @@ -15,7 +33,7 @@ void SysMessage(char *fmt, ...) { va_start(list,fmt); vsprintf(tmp,fmt,list); va_end(list); - MessageBox(0, tmp, "FW Plugin Msg", 0); + MessageBox(0, tmp, "Pad Plugin Msg", 0); } BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { @@ -34,7 +52,8 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) case IDOK: if (IsDlgButtonChecked(hW, IDC_LOGGING)) conf.Log = 1; - else conf.Log = 0; + else + conf.Log = 0; SaveConfig(); EndDialog(hW, FALSE); return TRUE; @@ -58,14 +77,14 @@ BOOL CALLBACK AboutDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { return FALSE; } -void CALLBACK FWconfigure() { +EXPORT_C_(void) PADconfigure() { DialogBox(hInst, MAKEINTRESOURCE(IDD_CONFIG), GetActiveWindow(), (DLGPROC)ConfigureDlgProc); } -void CALLBACK FWabout() { +EXPORT_C_(void) PADabout() { DialogBox(hInst, MAKEINTRESOURCE(IDD_ABOUT), GetActiveWindow(), From fea158f392cd6ff84e7acefd69479d5ccb7597c6 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 06:50:10 +0000 Subject: [PATCH 113/296] PadNull: A bit more work on the Windows side. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1244 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/PadNull/Linux/PadLinux.cpp | 8 ----- plugins/PadNull/Pad.cpp | 2 -- plugins/PadNull/Windows/PadWin.cpp | 56 +++++++++++++++++++++++++++--- 3 files changed, 52 insertions(+), 14 deletions(-) diff --git a/plugins/PadNull/Linux/PadLinux.cpp b/plugins/PadNull/Linux/PadLinux.cpp index 4a3d23f76f..2337ea5d79 100644 --- a/plugins/PadNull/Linux/PadLinux.cpp +++ b/plugins/PadNull/Linux/PadLinux.cpp @@ -63,11 +63,7 @@ void _PadUpdate(int pad) s32 _PADOpen(void *pDsp) { - //XKeyboardState *settings; - GSdsp = *(Display**)pDsp; - //XGetKeyboardControl(GSdsp, settings); - //autoRepeatMode = settings->global_auto_repeat; XAutoRepeatOff(GSdsp); return 0; @@ -76,8 +72,4 @@ s32 _PADOpen(void *pDsp) void _PADClose() { XAutoRepeatOn(GSdsp); - //XKeyboardControl *settings; - //settings->auto_repeat_mode = autoRepeatMode; - - //XChangeKeyboardControl(GSdsp, KBAutoRepeatMode, settings); } diff --git a/plugins/PadNull/Pad.cpp b/plugins/PadNull/Pad.cpp index 9c4d52afb6..ca4985d9f1 100644 --- a/plugins/PadNull/Pad.cpp +++ b/plugins/PadNull/Pad.cpp @@ -106,8 +106,6 @@ EXPORT_C_(s32) PADopen(void *pDsp) memset(&event, 0, sizeof(event)); return _PADOpen(pDsp); - - return 0; } EXPORT_C_(void) PADclose() diff --git a/plugins/PadNull/Windows/PadWin.cpp b/plugins/PadNull/Windows/PadWin.cpp index 1cd4bddb87..48e3e580a2 100644 --- a/plugins/PadNull/Windows/PadWin.cpp +++ b/plugins/PadNull/Windows/PadWin.cpp @@ -16,18 +16,66 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include "PadWin.h" +#include "PadWin.h" + +LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); +WNDPROC GSwndProc = NULL; +HWND GShwnd = NULL; + +LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) +{ + switch (msg) + { + case WM_KEYDOWN: + if (lParam & 0x40000000)return TRUE; + + event.evt = KEYPRESS; + event.key = wParam; + break; + + case WM_KEYUP: + event.evt = KEYRELEASE; + event.key = wParam; + break; + + case WM_DESTROY: + case WM_QUIT: + event.evt = KEYPRESS; + event.key = VK_ESCAPE; + return GSwndProc(hWnd, msg, wParam, lParam); + + default: + return GSwndProc(hWnd, msg, wParam, lParam); + }; + + return TRUE; +} void _PadUpdate(int pad) { } s32 _PADOpen(void *pDsp) -{ - +{ + GShwnd = (HWND)*(long*)pDsp; + + if (GShwnd != NULL && GSwndProc != NULL) + { + // revert + SetWindowLongPtr(GShwnd, GWLP_WNDPROC, (LPARAM)(WNDPROC)(GSwndProc)); + } + + GSwndProc = (WNDPROC)GetWindowLongPtr(GShwnd, GWLP_WNDPROC); + GSwndProc = ((WNDPROC)SetWindowLongPtr(GShwnd, GWLP_WNDPROC, (LPARAM)(WNDPROC)(PADwndProc))); return 0; } void _PADClose() -{ +{ + if (GShwnd != NULL && GSwndProc != NULL) + { + SetWindowLongPtr(GShwnd, GWLP_WNDPROC, (LPARAM)(WNDPROC)(GSwndProc)); + GSwndProc = NULL; + GShwnd = NULL; + } } \ No newline at end of file From ecfed766cd40d7371eb4f406d8f1e16e5484967d Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 11:57:28 +0000 Subject: [PATCH 114/296] ZeroPad: Refactor a bunch of code. Escape now sets the key value to 'Unknown' when configuring. Work some more on the hat code (still disabled). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1245 96395faa-99c1-11dd-bbfe-3dabce05a288 --- build.sh | 3 + plugins/zeropad/Linux/gui.cpp | 6 +- plugins/zeropad/Linux/linux.cpp | 324 +++++++++++++++++++------------- plugins/zeropad/Linux/linux.h | 1 + plugins/zeropad/zeropad.h | 5 +- 5 files changed, 206 insertions(+), 133 deletions(-) diff --git a/build.sh b/build.sh index fbf599a8d6..4da94673a5 100644 --- a/build.sh +++ b/build.sh @@ -29,6 +29,9 @@ export ZEROGSOPTIONS="--enable-sse2" #ZeroSPU2 Debug mode (Don't enable right now) #export ZEROSPU2OPTIONS="--enable-debug --enable-devbuild" +#GSnull debug options. +#export GSnullOPTIONS="--enable-debug" + option=$@ export PCSX2PLUGINS="`pwd`/bin/plugins" curdir=`pwd` diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index 199da357a3..d2dd512a7d 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -36,7 +36,8 @@ extern "C" } extern string s_strIniPath; -GtkWidget *Conf = NULL, *s_devicecombo = NULL; +GtkWidget *Conf = NULL, *About = NULL; +GtkWidget *s_devicecombo = NULL; extern void UpdateConf(int pad); void SaveConfig() @@ -268,9 +269,6 @@ void OnConf_Cancel(GtkButton *button, gpointer user_data) LoadConfig(); // load previous config } -GtkWidget *About = NULL; - -//void OnAbout_Ok(GtkButton *button, gpointer user_data) void OnAbout_Ok(GtkDialog *About, gint response_id, gpointer user_data) { gtk_widget_destroy(GTK_WIDGET(About)); diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 48a5756b53..8134d044fe 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -143,8 +143,8 @@ void CALLBACK PADupdate(int pad) #endif if (i != -1) { - keyPress |= (1 << i); - keyRelease &= ~(1 << i); + clear_bit(keyRelease, i); + set_bit(keyPress, i); } //PAD_LOG("Key pressed:%d\n", i); @@ -232,7 +232,7 @@ void CALLBACK PADupdate(int pad) } } } -#ifdef EXPERAMENTAL_POV_CODE +#ifdef EXPERIMENTAL_POV_CODE else if (IS_POV(key)) { int joyid = PAD_GETJOYID(key); @@ -245,12 +245,12 @@ void CALLBACK PADupdate(int pad) int temp; //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); - if (value != 0) + if (value != SDL_HAT_CENTERED) { if (PAD_GETPOVDIR(key) == value) set_bit(status[pad], i); - /*else - clear_bit(status[pad], i);*/ + else + clear_bit(status[pad], i); switch (value) { @@ -269,19 +269,12 @@ void CALLBACK PADupdate(int pad) case SDL_HAT_LEFT: PAD_LOG("Left!\n"); break; - case SDL_HAT_CENTER: - clear_bit(status[pad], i); - break; } } else { - //clear_bit(status[pad], i); + clear_bit(status[pad], i); } - /*if PAD_GETPOVSIGN(key) - set_bit(status[pad], i); - else - clear_bit(status[pad], i);*/ } } #else @@ -351,7 +344,7 @@ void UpdateConf(int pad) sprintf(&tmp[0], "JAxis %d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); } else if (IS_POV(conf.keys[pad][i])) -#ifdef EXPERAMENTAL_POV_CODE +#ifdef EXPERIMENTAL_POV_CODE { tmp.resize(28); switch(PAD_GETPOVDIR(conf.keys[pad][i])) @@ -416,27 +409,27 @@ void UpdateConf(int pad) set_checked(Conf, "forcefeedback", padopts & PADOPTION_FORCEFEEDBACK); } -void OnConf_Key(GtkButton *button, gpointer user_data) +GtkWidget* GetLabelWidget(GtkButton *button) { - GdkEvent *ev; - const char* buttonname = gtk_widget_get_name(GTK_WIDGET(button));//gtk_button_get_label(button); + const char* buttonname = gtk_widget_get_name(GTK_WIDGET(button)); const char* labelname = GetLabelFromButton(buttonname).c_str(); - GtkWidget* label = lookup_widget(Conf, labelname); + return lookup_widget(Conf, labelname); +} + +int GetLabelId(GtkWidget *label) +{ if (label == NULL) { PAD_LOG("couldn't find correct label\n"); - return; + return -1; } - int id = (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); - int pad = id / PADKEYS; - int key = id % PADKEYS; - PAD_LOG("Button = '%s', Label = '%s', id = %d, pad = %d, key = %d\n", buttonname, labelname, id, pad, key); - u32 *pkey = &conf.keys[pad][key]; + return (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); +} - // save the states -#ifdef JOYSTICK_SUPPORT - vector::iterator itjoy = s_vjoysticks.begin(); +void UpdateJoysticks(vector::iterator itjoy) +{ + itjoy = s_vjoysticks.begin(); SDL_JoystickUpdate(); @@ -446,26 +439,155 @@ void OnConf_Key(GtkButton *button, gpointer user_data) (*itjoy)->SaveState(); itjoy++; } +} + +bool PollKeyboard(char* &temp, u32* &pkey) +{ + GdkEvent *ev = gdk_event_get(); + + if (ev != NULL) + { + if (ev->type == GDK_KEY_PRESS) + { + *pkey = ev->key.keyval; + + if (ev->key.keyval == GDK_Escape) + temp = "Unknown"; + else + temp = XKeysymToString(ev->key.keyval); + + return true; + } + } + + return false; +} + +bool PollButtons(vector::iterator itjoy, int &jbutton, u32* &pkey) +{ + // MAKE sure to look for changes in the state!! + for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) + { + int but = SDL_JoystickGetButton((*itjoy)->GetJoy(), i); + + if (but != (*itjoy)->GetButtonState(i)) + { + if (!but) // released, we don't really want this + { + (*itjoy)->SetButtonState(i, 0); + break; + } + + *pkey = PAD_JOYBUTTON((*itjoy)->GetId(), i); + jbutton = i; + return true; + } + } + + return false; +} + +bool PollAxes(vector::iterator itjoy, bool pov, int &jbutton, bool &negative, u32* &pkey) +{ + for (int i = 0; i < (*itjoy)->GetNumAxes(); ++i) + { + int value = SDL_JoystickGetAxis((*itjoy)->GetJoy(), i); + + if (value != (*itjoy)->GetAxisState(i)) + { + PAD_LOG("Change in joystick %d: %d.\n", i, value); + + if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this + { + // released, we don't really want this + (*itjoy)->SetAxisState(i, value); + break; + } + + if (abs(value) > 0x3fff) + { + jbutton = i; + + if (pov) + { + *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); + negative = (value < 0); + } + else // axis + { + *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); + } + return true; + } + } + } + return false; +} + +/*SDL_HAT_CENTERED, +SDL_HAT_UP, SDL_HAT_RIGHT, +SDL_HAT_DOWN, SDL_HAT_LEFT, +SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, +SDL_HAT_LEFTUP, SDL_HAT_LEFTDOWN*/ + +bool PollPOV(vector::iterator itjoy, int &jbutton, int &dir, u32* &pkey) +{ +#ifdef EXPERIMENTAL_POV_CODE + for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) + { + int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); + + if (value != SDL_HAT_CENTERED) + { + switch (value) + { + case SDL_HAT_UP: + case SDL_HAT_RIGHT: + case SDL_HAT_DOWN: + case SDL_HAT_LEFT: + *pkey = PAD_POV((*itjoy)->GetId(), value, i); + jbutton = i; + dir = value; + PAD_LOG("Hat Pressed!"); + return true; + default: + break; + } + } + } +#endif + return false; +} + +void OnConf_Key(GtkButton *button, gpointer user_data) +{ + u32 *pkey; + GtkWidget* label = GetLabelWidget(button); + int id = GetLabelId(label); + + if (id == -1) return; + + int pad = id / PADKEYS; + int key = id % PADKEYS; + pkey = &conf.keys[pad][key]; + + // save the states +#ifdef JOYSTICK_SUPPORT + vector::iterator itjoy; + + UpdateJoysticks(itjoy); #endif for (;;) { - ev = gdk_event_get(); - if (ev != NULL) + char *tmp; + + if (PollKeyboard(tmp, pkey)) { - if (ev->type == GDK_KEY_PRESS) - { - *pkey = ev->key.keyval; - - char* tmp = XKeysymToString(*pkey); - if (tmp != NULL) - gtk_entry_set_text(GTK_ENTRY(label), tmp); - else - gtk_entry_set_text(GTK_ENTRY(label), "Unknown"); - return; - } + gtk_entry_set_text(GTK_ENTRY(label), tmp); + return; } - + #ifdef JOYSTICK_SUPPORT itjoy = s_vjoysticks.begin(); @@ -473,101 +595,51 @@ void OnConf_Key(GtkButton *button, gpointer user_data) while (itjoy != s_vjoysticks.end()) { - // MAKE sure to look for changes in the state!! - for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) + int jbutton, direction; + + if (PollButtons(itjoy, jbutton, pkey)) { - int but = SDL_JoystickGetButton((*itjoy)->GetJoy(), i); - if (but != (*itjoy)->GetButtonState(i)) - { - if (!but) // released, we don't really want this - { - (*itjoy)->SetButtonState(i, 0); - break; - } - - *pkey = PAD_JOYBUTTON((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JBut %d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } + char str[32]; + + sprintf(str, "JBut %d", jbutton); + gtk_entry_set_text(GTK_ENTRY(label), str); + return; } - for (int i = 0; i < (*itjoy)->GetNumAxes(); ++i) + bool negative = false, pov = (key < 16); + + if (PollAxes(itjoy, pov, jbutton, negative, pkey)) { - int value = SDL_JoystickGetAxis((*itjoy)->GetJoy(), i); - - if (value != (*itjoy)->GetAxisState(i)) - { - PAD_LOG("Change in joystick %d: %d.\n", i, value); - - if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this - { - // released, we don't really want this - (*itjoy)->SetAxisState(i, value); - break; - } - - if (abs(value) > 0x3fff) - { - if (key < 16) // POV - { - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); - char str[32]; - sprintf(str, "JPOV %d%s", i, value < 0 ? "-" : "+"); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } - else // axis - { - *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); - char str[32]; - sprintf(str, "JAxis %d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - } - } - } + char str[32]; + + if (pov) + sprintf(str, "JPOV %d%s", jbutton, (negative) ? "-" : "+"); + else + sprintf(str, "JAxis %d", jbutton); + + gtk_entry_set_text(GTK_ENTRY(label), str); + + return; } -#ifdef EXPERAMENTAL_POV_CODE - for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) +#ifdef EXPERIMENTAL_POV_CODE + if (PollPOV(itjoy, jbutton, direction, pkey)) { - int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); - - if (value != (*itjoy)->GetPOVState(i)) + char str[32]; + + switch (direction) { - switch (value) - { - char str[32]; - - case SDL_HAT_UP: - *pkey = PAD_POV((*itjoy)->GetId(), value, i); - sprintf(str, "JPOVU-%d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - case SDL_HAT_RIGHT: - *pkey = PAD_POV((*itjoy)->GetId(), value, i); - sprintf(str, "JPOVR-%d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - case SDL_HAT_DOWN: - *pkey = PAD_POV((*itjoy)->GetId(), value, i); - sprintf(str, "JPOVD-%d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - case SDL_HAT_LEFT: - *pkey = PAD_POV((*itjoy)->GetId(), value, i); - sprintf(str, "JPOVL-%d", i); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; - // Not handling SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, - // SDL_HAT_LEFTUP, or SDL_HAT_LEFTDOWN here. They should be - // handled in the PADUpdate code, though. - } + case SDL_HAT_UP: sprintf(str, "JPOVU-%d", jbutton); break; + case SDL_HAT_RIGHT: sprintf(str, "JPOVR-%d", jbutton); break; + case SDL_HAT_DOWN: sprintf(str, "JPOVD-%d", jbutton); break; + case SDL_HAT_LEFT: sprintf(str, "JPOVL-%d", jbutton); break; } + + gtk_entry_set_text(GTK_ENTRY(label), str); + return; } #endif + itjoy++; } #endif diff --git a/plugins/zeropad/Linux/linux.h b/plugins/zeropad/Linux/linux.h index e121503a1d..0dedcde2f1 100644 --- a/plugins/zeropad/Linux/linux.h +++ b/plugins/zeropad/Linux/linux.h @@ -18,6 +18,7 @@ #include #include +#include #include #define JOYSTICK_SUPPORT diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index 92609b15ac..aa866ca7c2 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -85,7 +85,7 @@ extern char libraryName[256]; #define PADSUBKEYS 2 #endif -//#define EXPERAMENTAL_POV_CODE +//#define EXPERIMENTAL_POV_CODE extern int PadEnum[2][2]; typedef struct @@ -154,8 +154,7 @@ enum gamePadValues PAD_L2 = 0 }; -// Put in the code for bolche's analog contols hack, ifdeffed out, so I don't forget to -// add a gui some day and activate it. +// Activate bolche's analog contols hack #define ANALOG_CONTROLS_HACK // DEF_VALUE is the strength you press the control. // Code taken from http://forums.pcsx2.net/thread-4699.html From 342da0e2a850ed5b3a40e53aba294f685504c1eb Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 23 May 2009 12:40:01 +0000 Subject: [PATCH 115/296] microVU: - Added a new status-flag speedhack after learning how evil sVU is with status flags (mVU with both flag-speedhacks ON is less evil than sVU's default setting which does something even hackier) - Minor Changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1246 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 1 + pcsx2/x86/microVU_Compile.inl | 2 +- pcsx2/x86/microVU_Flags.inl | 18 ++++++++++-------- pcsx2/x86/microVU_Misc.h | 15 ++++++++------- pcsx2/x86/microVU_Upper.inl | 2 +- 5 files changed, 21 insertions(+), 17 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 0301d2a476..214f3fe440 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -291,6 +291,7 @@ microVUt(void) mVUanalyzeSflag(int It) { if (!It) { mVUinfo |= _isNOP; } else { mVUinfo |= _swapOps; + mVUsFlagHack = 0; // Don't Optimize Out Status Flags for this block if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf /*<< mVUcount*/; } if (mVUcount >= 1) { incPC2(-2); mVUinfo |= _isSflag; incPC2(2); } // Note: _isSflag is used for status flag optimizations. diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 0857fd144a..db6a44131e 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -237,6 +237,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUpBlock = pBlock; mVUregs.flags = 0; mVUflagInfo = 0; + mVUsFlagHack = CHECK_VU_FLAGHACK2 | mVUflagHack; bool eBitBranch = 0; // E-bit Set on Branch for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { @@ -252,7 +253,6 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUsetCycles(); if (mVU->p) { mVUinfo |= _readP; } if (mVU->q) { mVUinfo |= _readQ; } - else { mVUinfo |= _writeQ; } if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } if (mVUbranch) { mVUsetFlagInfo(); branchEbit(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index b513f8f89c..95f4527efb 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -72,6 +72,8 @@ void sortFlag(int* fFlag, int* bFlag, int cycles) { } } +#define sFlagCond ((doStatus && !mVUsFlagHack) || isFSSET || doDivFlag) + // Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { microVU* mVU = mVUx; @@ -132,9 +134,9 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { mVUinfo |= xM << 10; // _fmInstance mVUinfo |= xC << 14; // _fcInstance - if (doStatus || isFSSET || doDivFlag) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } - if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } - if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } + if (sFlagCond) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } + if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } + if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } cycles++; incPC2(2); @@ -183,7 +185,7 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { } } -microVUt(void) mVUpass4(int startPC) { +microVUt(void) mVUflagPass(int startPC) { microVU* mVU = mVUx; int oldPC = iPC; @@ -195,7 +197,7 @@ microVUt(void) mVUpass4(int startPC) { for (int branch = 0; mVUcount < 4; mVUcount++) { incPC(1); if ( curI & _Ebit_ ) { branch = 1; } - if ( curI & _MDTbit_ ) { branch = 2; } + if ( curI & _MDTbit_ ) { branch = 4; } if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(); incPC(1); } if (branch >= 2) { break; } else if (branch == 1) { branch = 2; } @@ -215,15 +217,15 @@ microVUt(void) mVUpass4(int startPC) { microVUt(void) mVUsetFlagInfo() { microVU* mVU = mVUx; - branchType1 { incPC(-1); mVUpass4(branchAddr); incPC(1); } + branchType1 { incPC(-1); mVUflagPass(branchAddr); incPC(1); } branchType2 { mVUflagInfo |= 0xfff; } branchType3 { incPC(-1); - mVUpass4(branchAddr); + mVUflagPass(branchAddr); int backupFlagInfo = mVUflagInfo; mVUflagInfo = 0; incPC(4); // Branch Not Taken - mVUpass4(xPC); + mVUflagPass(xPC); incPC(-3); mVUflagInfo |= backupFlagInfo; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index f5a2e60174..306e9d5686 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -189,8 +189,8 @@ declareAllVariables #define _isEOB (1<<2) // End of Block #define _isBdelay (1<<3) // Cur Instruction in Branch Delay slot #define _isSflag (1<<4) // Cur Instruction uses status flag -#define _writeQ (1<<5) -#define _readQ (1<<6) +#define _writeQ (1<<6) +#define _readQ (1<<6) // same as writeQ #define _writeP (1<<7) #define _readP (1<<7) // same as writeP #define _doFlags (3<<8) @@ -220,10 +220,10 @@ declareAllVariables #define isEOB (mVUinfo & (1<<2)) #define isBdelay (mVUinfo & (1<<3)) #define isSflag (mVUinfo & (1<<4)) -#define writeQ ((mVUinfo >> 5) & 1) -#define readQ ((mVUinfo >> 6) & 1) -#define writeP (((mVUinfo >> 7) + 1) & 1) +#define readQ ((mVUinfo >> 6) & 1) // same as writeQ #define readP ((mVUinfo >> 7) & 1) // same as writeP +#define writeQ (((mVUinfo >> 6) + 1) & 1) +#define writeP (((mVUinfo >> 7) + 1) & 1) #define doFlags (mVUinfo & (3<<8)) #define doMac (mVUinfo & (1<<8)) #define doStatus (mVUinfo & (1<<9)) @@ -289,8 +289,9 @@ declareAllVariables #endif // Speed Hacks (Set to 1 to turn On) -#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack -#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack +#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack (Very-Compatible) +#define CHECK_VU_FLAGHACK2 0 // Status Flag Speed Hack (More-Hacky) +#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack (Semi-Compatible) // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 37a92f8442..3ecdb57fb7 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -35,7 +35,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; //SysPrintf("doStatus = %d; doMac = %d\n", doStatus>>9, doMac>>8); - if (mVUflagHack) { mVUinfo &= ~_doStatus; } + if (mVUsFlagHack) { mVUinfo &= ~_doStatus; } if (!doFlags) return; if (!doMac || (_XYZW_SS && modXYZW)) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw From f1be4728baa9415f268957d8d3808329f8a9540f Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 22:20:16 +0000 Subject: [PATCH 116/296] Zeropad: Fix a glitch in the new esc code, and the hat code should not save in the same spots as the axis code. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1248 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 57 ++++++++++++++++++--------------- plugins/zeropad/zeropad.h | 10 +++--- 2 files changed, 37 insertions(+), 30 deletions(-) diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 8134d044fe..32d3583137 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -232,8 +232,26 @@ void CALLBACK PADupdate(int pad) } } } -#ifdef EXPERIMENTAL_POV_CODE else if (IS_POV(key)) + { + int joyid = PAD_GETJOYID(key); + if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) + { + pjoy = s_vjoysticks[joyid]; + + int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); + int pad = (pjoy)->GetPAD(); + + if (PAD_GETPOVSIGN(key) && (value < -2048)) + clear_bit(status[pad], i); + else if (!PAD_GETPOVSIGN(key) && (value > 2048)) + clear_bit(status[pad], i); + else + set_bit(status[pad], i); + } + } +#ifdef EXPERIMENTAL_POV_CODE + else if (IS_HAT(key)) { int joyid = PAD_GETJOYID(key); if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) @@ -247,7 +265,7 @@ void CALLBACK PADupdate(int pad) //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); if (value != SDL_HAT_CENTERED) { - if (PAD_GETPOVDIR(key) == value) + if (PAD_GETHATDIR(key) == value) set_bit(status[pad], i); else clear_bit(status[pad], i); @@ -271,29 +289,10 @@ void CALLBACK PADupdate(int pad) break; } } - else + /*else { clear_bit(status[pad], i); - } - } - } -#else - else if (IS_POV(key)) - { - int joyid = PAD_GETJOYID(key); - if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) - { - pjoy = s_vjoysticks[joyid]; - - int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - int pad = (pjoy)->GetPAD(); - - if (PAD_GETPOVSIGN(key) && (value < -2048)) - clear_bit(status[pad], i); - else if (!PAD_GETPOVSIGN(key) && (value > 2048)) - clear_bit(status[pad], i); - else - set_bit(status[pad], i); + }*/ } } #endif @@ -347,7 +346,7 @@ void UpdateConf(int pad) #ifdef EXPERIMENTAL_POV_CODE { tmp.resize(28); - switch(PAD_GETPOVDIR(conf.keys[pad][i])) + switch(PAD_GETHATDIR(conf.keys[pad][i])) { case SDL_HAT_UP: sprintf(&tmp[0], "JPOVU-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); @@ -449,12 +448,18 @@ bool PollKeyboard(char* &temp, u32* &pkey) { if (ev->type == GDK_KEY_PRESS) { - *pkey = ev->key.keyval; if (ev->key.keyval == GDK_Escape) + { temp = "Unknown"; + *pkey = NULL; + } else + { temp = XKeysymToString(ev->key.keyval); + *pkey = ev->key.keyval; + } + return true; } @@ -545,7 +550,7 @@ bool PollPOV(vector::iterator itjoy, int &jbutton, int &dir, u32* case SDL_HAT_RIGHT: case SDL_HAT_DOWN: case SDL_HAT_LEFT: - *pkey = PAD_POV((*itjoy)->GetId(), value, i); + *pkey = PAD_HAT((*itjoy)->GetId(), value, i); jbutton = i; dir = value; PAD_LOG("Hat Pressed!"); diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index aa866ca7c2..d6c5d2ba98 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -49,13 +49,13 @@ extern "C" extern char libraryName[256]; -#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); (it)++) - #define IS_KEYBOARD(key) (key < 0x10000) #define IS_JOYBUTTONS(key) (key >= 0x10000 && key < 0x20000) // buttons #define IS_JOYSTICK(key) (key >= 0x20000 && key < 0x30000) // analog #define IS_POV(key) (key >= 0x30000 && key < 0x40000) // uses analog as buttons (cares about sign) -#define IS_MOUSE(key) (key >= 0x40000 && key < 0x50000) // mouse +#define IS_HAT(key) (key >= 0x40000 && key < 0x50000) // uses hat as buttons (cares about sign) + +//#define IS_MOUSE(key) (key >= 0x40000 && key < 0x50000) // mouse #define PAD_GETKEY(key) ((key) & 0xffff) #define PAD_GETJOYID(key) (((key) & 0xf000) >> 12) @@ -64,8 +64,10 @@ extern char libraryName[256]; #define PAD_JOYBUTTON(joyid, buttonid) (0x10000 | ((joyid) << 12) | (buttonid)) #define PAD_JOYSTICK(joyid, axisid) (0x20000 | ((joyid) << 12) | (axisid)) #define PAD_POV(joyid, sign, axisid) (0x30000 | ((joyid) << 12) | ((sign) << 8) | (axisid)) +#define PAD_HAT(joyid, dir, axisid) (0x40000 | ((joyid) << 12) | ((dir) << 8) | (axisid)) + #define PAD_GETPOVSIGN(key) (((key) & 0x100) >> 8) -#define PAD_GETPOVDIR(key) (((key) & ~ 0x30000) >> 8) +#define PAD_GETHATDIR(key) (((key) & ~ 0x40000) >> 8) #ifdef __LINUX__ #define PADKEYS 28 From 20f0e52a1a9af448eb8adf78662daf74156cfa07 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 23 May 2009 23:07:53 +0000 Subject: [PATCH 117/296] Zeropad: All around my hat... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1249 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/callbacks.c | 68 ++++++++++++++--------------- plugins/zeropad/Linux/callbacks.h | 36 +++++++-------- plugins/zeropad/Linux/interface.c | 22 +++++----- plugins/zeropad/Linux/linux.cpp | 54 +++++++++++++++-------- plugins/zeropad/Linux/zeropad.glade | 30 ++++++------- plugins/zeropad/zeropad.cpp | 6 +-- plugins/zeropad/zeropad.h | 5 +-- 7 files changed, 119 insertions(+), 102 deletions(-) diff --git a/plugins/zeropad/Linux/callbacks.c b/plugins/zeropad/Linux/callbacks.c index cb65ceeb21..81945c91cb 100644 --- a/plugins/zeropad/Linux/callbacks.c +++ b/plugins/zeropad/Linux/callbacks.c @@ -9,6 +9,15 @@ #include "support.h" +void +OnAbout_Ok (GtkDialog *dialog, + gint response_id, + gpointer user_data) +{ + +} + + void OnConf_Pad1 (GtkButton *button, gpointer user_data) @@ -57,6 +66,30 @@ OnConf_Key (GtkButton *button, } +void +on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_checkbutton_reverserx_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + +void +on_forcefeedback_toggled (GtkToggleButton *togglebutton, + gpointer user_data) +{ + +} + + void on_checkbutton_reverselx_toggled (GtkToggleButton *togglebutton, gpointer user_data) @@ -74,23 +107,7 @@ on_checkbutton_reversely_toggled (GtkToggleButton *togglebutton, void -on_checkbutton_reverserx_toggled (GtkToggleButton *togglebutton, - gpointer user_data) -{ - -} - - -void -on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, - gpointer user_data) -{ - -} - - -void -on_forcefeedback_toggled (GtkToggleButton *togglebutton, +OnConf_Ok (GtkButton *button, gpointer user_data) { @@ -104,20 +121,3 @@ OnConf_Cancel (GtkButton *button, } - -void -OnConf_Ok (GtkButton *button, - gpointer user_data) -{ - -} - - -void -OnAbout_Ok (GtkDialog *dialog, - gint response_id, - gpointer user_data) -{ - -} - diff --git a/plugins/zeropad/Linux/callbacks.h b/plugins/zeropad/Linux/callbacks.h index 3e57b0e5d4..ff61cd234e 100644 --- a/plugins/zeropad/Linux/callbacks.h +++ b/plugins/zeropad/Linux/callbacks.h @@ -1,6 +1,11 @@ #include +void +OnAbout_Ok (GtkDialog *dialog, + gint response_id, + gpointer user_data); + void OnConf_Pad1 (GtkButton *button, gpointer user_data); @@ -25,6 +30,18 @@ void OnConf_Key (GtkButton *button, gpointer user_data); +void +on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, + gpointer user_data); + +void +on_checkbutton_reverserx_toggled (GtkToggleButton *togglebutton, + gpointer user_data); + +void +on_forcefeedback_toggled (GtkToggleButton *togglebutton, + gpointer user_data); + void on_checkbutton_reverselx_toggled (GtkToggleButton *togglebutton, gpointer user_data); @@ -34,26 +51,9 @@ on_checkbutton_reversely_toggled (GtkToggleButton *togglebutton, gpointer user_data); void -on_checkbutton_reverserx_toggled (GtkToggleButton *togglebutton, - gpointer user_data); - -void -on_checkbutton_reversery_toggled (GtkToggleButton *togglebutton, - gpointer user_data); - -void -on_forcefeedback_toggled (GtkToggleButton *togglebutton, +OnConf_Ok (GtkButton *button, gpointer user_data); void OnConf_Cancel (GtkButton *button, gpointer user_data); - -void -OnConf_Ok (GtkButton *button, - gpointer user_data); - -void -OnAbout_Ok (GtkDialog *dialog, - gint response_id, - gpointer user_data); diff --git a/plugins/zeropad/Linux/interface.c b/plugins/zeropad/Linux/interface.c index d8e669e802..dd5158b82c 100644 --- a/plugins/zeropad/Linux/interface.c +++ b/plugins/zeropad/Linux/interface.c @@ -145,8 +145,8 @@ create_Conf (void) GtkWidget *Analog; GtkWidget *label6; GtkWidget *dialog_action_area1; - GtkWidget *cancelbutton1; GtkWidget *okbutton1; + GtkWidget *cancelbutton1; Conf = gtk_dialog_new (); gtk_widget_set_name (Conf, "Conf"); @@ -672,18 +672,18 @@ create_Conf (void) gtk_widget_show (dialog_action_area1); gtk_button_box_set_layout (GTK_BUTTON_BOX (dialog_action_area1), GTK_BUTTONBOX_END); - cancelbutton1 = gtk_button_new_from_stock ("gtk-cancel"); - gtk_widget_set_name (cancelbutton1, "cancelbutton1"); - gtk_widget_show (cancelbutton1); - gtk_dialog_add_action_widget (GTK_DIALOG (Conf), cancelbutton1, GTK_RESPONSE_CANCEL); - GTK_WIDGET_SET_FLAGS (cancelbutton1, GTK_CAN_DEFAULT); - okbutton1 = gtk_button_new_from_stock ("gtk-ok"); gtk_widget_set_name (okbutton1, "okbutton1"); gtk_widget_show (okbutton1); gtk_dialog_add_action_widget (GTK_DIALOG (Conf), okbutton1, GTK_RESPONSE_OK); GTK_WIDGET_SET_FLAGS (okbutton1, GTK_CAN_DEFAULT); + cancelbutton1 = gtk_button_new_from_stock ("gtk-cancel"); + gtk_widget_set_name (cancelbutton1, "cancelbutton1"); + gtk_widget_show (cancelbutton1); + gtk_dialog_add_action_widget (GTK_DIALOG (Conf), cancelbutton1, GTK_RESPONSE_CANCEL); + GTK_WIDGET_SET_FLAGS (cancelbutton1, GTK_CAN_DEFAULT); + g_signal_connect ((gpointer) PAD1, "clicked", G_CALLBACK (OnConf_Pad1), NULL); @@ -801,12 +801,12 @@ create_Conf (void) g_signal_connect ((gpointer) Analog, "clicked", G_CALLBACK (OnConf_Key), NULL); - g_signal_connect ((gpointer) cancelbutton1, "clicked", - G_CALLBACK (OnConf_Cancel), - NULL); g_signal_connect ((gpointer) okbutton1, "clicked", G_CALLBACK (OnConf_Ok), NULL); + g_signal_connect ((gpointer) cancelbutton1, "clicked", + G_CALLBACK (OnConf_Cancel), + NULL); /* Store pointers to all widgets, for use by lookup_widget(). */ GLADE_HOOKUP_OBJECT_NO_REF (Conf, Conf, "Conf"); @@ -890,8 +890,8 @@ create_Conf (void) GLADE_HOOKUP_OBJECT (Conf, Analog, "Analog"); GLADE_HOOKUP_OBJECT (Conf, label6, "label6"); GLADE_HOOKUP_OBJECT_NO_REF (Conf, dialog_action_area1, "dialog_action_area1"); - GLADE_HOOKUP_OBJECT (Conf, cancelbutton1, "cancelbutton1"); GLADE_HOOKUP_OBJECT (Conf, okbutton1, "okbutton1"); + GLADE_HOOKUP_OBJECT (Conf, cancelbutton1, "cancelbutton1"); return Conf; } diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 32d3583137..8378fe5295 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -120,7 +120,8 @@ void CALLBACK PADupdate(int pad) key = XLookupKeysym((XKeyEvent *) & E, 0); i = FindKey(key, pad); -#ifdef ANALOG_CONTROLS_HACK + + // Analog controls. if ((i > PAD_RY) && (i <= PAD_R_LEFT)) { switch (i) @@ -140,7 +141,7 @@ void CALLBACK PADupdate(int pad) } i += 0xff00; } -#endif + if (i != -1) { clear_bit(keyRelease, i); @@ -156,14 +157,14 @@ void CALLBACK PADupdate(int pad) key = XLookupKeysym((XKeyEvent *) & E, 0); i = FindKey(key, pad); -#ifdef ANALOG_CONTROLS_HACK + // Analog Controls. if ((i > PAD_RY) && (i <= PAD_R_LEFT)) { Analog::ResetPad(Analog::AnalogToPad(i), pad); i += 0xff00; } -#endif + if (i != -1) { clear_bit(keyPress, i); @@ -260,39 +261,57 @@ void CALLBACK PADupdate(int pad) int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); - int temp; //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); - if (value != SDL_HAT_CENTERED) + if ((value != SDL_HAT_CENTERED) && (PAD_GETHATDIR(key) == value)) { - if (PAD_GETHATDIR(key) == value) + if ((value == SDL_HAT_UP) || (value == SDL_HAT_RIGHT) || (value == SDL_HAT_DOWN) ||(value == SDL_HAT_LEFT)) + { set_bit(status[pad], i); + } else + { clear_bit(status[pad], i); + } + switch (i) + { + case PAD_LEFT: + PAD_LOG("Hat Left!\n"); + break; + case PAD_DOWN: + PAD_LOG("Hat Down!\n"); + break; + case PAD_RIGHT: + PAD_LOG("Hat Right!\n"); + break; + case PAD_UP: + PAD_LOG("Hat Up!\n"); + break; + } switch (value) { case SDL_HAT_UP: - PAD_LOG("Up!\n"); + PAD_LOG("D-pad Up!\n"); break; case SDL_HAT_RIGHT: - PAD_LOG("Right!\n"); + PAD_LOG("D-pad Right!\n"); break; case SDL_HAT_DOWN: - PAD_LOG("Down!\n"); + PAD_LOG("D-pad Down!\n"); break; case SDL_HAT_LEFT: - PAD_LOG("Left!\n"); + PAD_LOG("D-pad Left!\n"); break; } } - /*else + else { clear_bit(status[pad], i); - }*/ + } } } #endif @@ -342,8 +361,8 @@ void UpdateConf(int pad) tmp.resize(28); sprintf(&tmp[0], "JAxis %d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); } - else if (IS_POV(conf.keys[pad][i])) #ifdef EXPERIMENTAL_POV_CODE + else if (IS_HAT(conf.keys[pad][i])) { tmp.resize(28); switch(PAD_GETHATDIR(conf.keys[pad][i])) @@ -366,6 +385,7 @@ void UpdateConf(int pad) } } #else + else if (IS_POV(conf.keys[pad][i])) { tmp.resize(28); sprintf(&tmp[0], "JPOV %d%s", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i]), PAD_GETPOVSIGN(conf.keys[pad][i]) ? "-" : "+"); @@ -515,8 +535,8 @@ bool PollAxes(vector::iterator itjoy, bool pov, int &jbutton, boo if (pov) { - *pkey = PAD_POV((*itjoy)->GetId(), value < 0, i); negative = (value < 0); + *pkey = PAD_POV((*itjoy)->GetId(), negative, i); } else // axis { @@ -535,7 +555,7 @@ SDL_HAT_DOWN, SDL_HAT_LEFT, SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, SDL_HAT_LEFTUP, SDL_HAT_LEFTDOWN*/ -bool PollPOV(vector::iterator itjoy, int &jbutton, int &dir, u32* &pkey) +bool PollHAT(vector::iterator itjoy, int &jbutton, int &dir, u32* &pkey) { #ifdef EXPERIMENTAL_POV_CODE for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) @@ -628,7 +648,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) } #ifdef EXPERIMENTAL_POV_CODE - if (PollPOV(itjoy, jbutton, direction, pkey)) + if (PollHAT(itjoy, jbutton, direction, pkey)) { char str[32]; diff --git a/plugins/zeropad/Linux/zeropad.glade b/plugins/zeropad/Linux/zeropad.glade index ef754e23a4..7c33fea000 100644 --- a/plugins/zeropad/Linux/zeropad.glade +++ b/plugins/zeropad/Linux/zeropad.glade @@ -46,20 +46,6 @@ and SSSPSXPad, TwinPad authors True GTK_BUTTONBOX_END - - - True - True - True - gtk-cancel - True - GTK_RELIEF_NORMAL - True - -6 - - - - True @@ -70,7 +56,21 @@ and SSSPSXPad, TwinPad authors GTK_RELIEF_NORMAL True -5 - + + + + + + + True + True + True + gtk-cancel + True + GTK_RELIEF_NORMAL + True + -6 + diff --git a/plugins/zeropad/zeropad.cpp b/plugins/zeropad/zeropad.cpp index b67549fc54..13582dfe27 100644 --- a/plugins/zeropad/zeropad.cpp +++ b/plugins/zeropad/zeropad.cpp @@ -309,7 +309,7 @@ void PADsetMode(int pad, int mode) u8 CALLBACK PADstartPoll(int pad) { - PAD_LOG("PADstartPoll: %d\n", pad); + //PAD_LOG("PADstartPoll: %d\n", pad); curPad = pad - 1; curByte = 0; @@ -325,7 +325,7 @@ u8 _PADpoll(u8 value) { curByte++; - PAD_LOG("PADpoll: cmd: %x\n", value); + //PAD_LOG("PADpoll: cmd: %x\n", value); curCmd = value; switch (value) @@ -541,7 +541,7 @@ u8 CALLBACK PADpoll(u8 value) u8 ret; ret = _PADpoll(value); - PAD_LOG("PADpoll: %x (%d: %x)\n", value, curByte, ret); + //PAD_LOG("PADpoll: %x (%d: %x)\n", value, curByte, ret); return ret; } diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index d6c5d2ba98..8e3127b1f0 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -156,14 +156,11 @@ enum gamePadValues PAD_L2 = 0 }; -// Activate bolche's analog contols hack -#define ANALOG_CONTROLS_HACK +// Activate bolche's analog controls hack // DEF_VALUE is the strength you press the control. // Code taken from http://forums.pcsx2.net/thread-4699.html -#ifdef ANALOG_CONTROLS_HACK #define DEF_VALUE 32766 -#endif /* end of pad.h */ From 081006bcbe15d9e6eab8f6538f306a5beda0acc0 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 23 May 2009 23:52:53 +0000 Subject: [PATCH 118/296] GSdx: dx9 hw mode got a vertex buffer too (as dx10), it was an old TODO... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1250 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSClut.cpp | 2 +- plugins/GSdx/GSDevice10.h | 3 -- plugins/GSdx/GSDevice9.cpp | 57 ++++++++++++---------------- plugins/GSdx/GSDevice9.h | 17 ++------- plugins/GSdx/GSRendererHW9.cpp | 68 ++++++++++++++++++++++++---------- plugins/GSdx/GSState.cpp | 4 +- plugins/GSdx/GSTextureFX10.cpp | 28 +++++++------- plugins/GSdx/GSTextureFX9.cpp | 60 +++++++++++++++++++++++++++++- plugins/GSdx/GSTextureFX9.h | 7 ++++ plugins/GSdx/GSVector.h | 27 +++++++++++--- plugins/GSdx/res/convert10.fx | 2 +- 11 files changed, 181 insertions(+), 94 deletions(-) diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index 27084e87b1..f428b29db4 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -659,4 +659,4 @@ bool GSClut::ReadState::IsDirty(const GIFRegTEX0& TEX0) bool GSClut::ReadState::IsDirty(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) { return dirty || !GSVector4i::load(this).eq(GSVector4i::load(&TEX0, &TEXA)); -} +} \ No newline at end of file diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index 6d94f998a8..cd5266acc7 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -27,10 +27,7 @@ class GSDevice10 : public GSDevice { private: - // state cache - ID3D10Buffer* m_vb; - int m_vb_count; size_t m_vb_stride; ID3D10InputLayout* m_layout; D3D10_PRIMITIVE_TOPOLOGY m_topology; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index ac0c436262..0da5457c6d 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -26,8 +26,6 @@ GSDevice9::GSDevice9() : m_vb(NULL) - , m_vb_count(0) - , m_vb_vertices(NULL) , m_vb_stride(0) , m_layout(NULL) , m_topology((D3DPRIMITIVETYPE)0) @@ -145,6 +143,8 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) CompileShader(IDR_CONVERT9_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } + m_dev->CreateVertexBuffer(4 * sizeof(GSVertexPT1), D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_convert.vb, NULL); + m_convert.dss.DepthEnable = false; m_convert.dss.StencilEnable = false; @@ -550,7 +550,16 @@ void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, c vertices[i].p.y += 1.0f / ds.y; } - IASetVertexBuffer(4, vertices); + void* buff = NULL; + + if(SUCCEEDED(m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) + { + memcpy(buff, vertices, sizeof(vertices)); + + m_convert.vb->Unlock(); + } + + IASetVertexBuffer(m_convert.vb, sizeof(vertices[0])); IASetInputLayout(m_convert.il); IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); @@ -570,7 +579,7 @@ void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, c // - DrawPrimitive(); + DrawPrimitive(countof(vertices)); // @@ -610,18 +619,9 @@ void GSDevice9::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linea StretchRect(st, sr, dt, dr, m_interlace.ps[shader], (const float*)&cb, 1, linear); } -/* -void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, const void* vertices, size_t stride) + +void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, size_t stride) { - void* data = NULL; - - if(SUCCEEDED(vb->Lock(0, count * stride, &data, D3DLOCK_DISCARD))) - { - memcpy(data, vertices, count * stride); - - vb->Unlock(); - } - if(m_vb != vb || m_vb_stride != stride) { m_dev->SetStreamSource(0, vb, 0, stride); @@ -630,23 +630,14 @@ void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, cons m_vb_stride = stride; } } -*/ -void GSDevice9::IASetVertexBuffer(int count, const void* vertices, size_t stride) -{ - m_vb_count = count; - m_vb_vertices = vertices; - m_vb_stride = stride; -} void GSDevice9::IASetInputLayout(IDirect3DVertexDeclaration9* layout) { - // TODO: get rid of all SetFVF before enabling this - - // if(m_layout != layout) + if(m_layout != layout) { m_dev->SetVertexDeclaration(layout); - // m_layout = layout; + m_layout = layout; } } @@ -855,31 +846,31 @@ void GSDevice9::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) } } -void GSDevice9::DrawPrimitive() +void GSDevice9::DrawPrimitive(uint32 count, uint32 start) { int prims = 0; switch(m_topology) { case D3DPT_TRIANGLELIST: - prims = m_vb_count / 3; + prims = count / 3; break; case D3DPT_LINELIST: - prims = m_vb_count / 2; + prims = count / 2; break; case D3DPT_POINTLIST: - prims = m_vb_count; + prims = count; break; case D3DPT_TRIANGLESTRIP: case D3DPT_TRIANGLEFAN: - prims = m_vb_count - 2; + prims = count - 2; break; case D3DPT_LINESTRIP: - prims = m_vb_count - 1; + prims = count - 1; break; } - m_dev->DrawPrimitiveUP(m_topology, prims, m_vb_vertices, m_vb_stride); + m_dev->DrawPrimitive(m_topology, start, prims); } // FIXME: D3DXCompileShaderFromResource of d3dx9 v37 (march 2008) calls GetFullPathName on id for some reason and then crashes diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 3d6b69086c..dc13b3d894 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -61,11 +61,7 @@ struct Direct3DBlendState9 class GSDevice9 : public GSDevice { private: - // state cache - IDirect3DVertexBuffer9* m_vb; - int m_vb_count; - const void* m_vb_vertices; size_t m_vb_stride; IDirect3DVertexDeclaration9* m_layout; D3DPRIMITIVETYPE m_topology; @@ -106,8 +102,9 @@ public: // TODO struct { - CComPtr vs; + CComPtr vb; CComPtr il; + CComPtr vs; CComPtr ps[7]; Direct3DSamplerState9 ln; Direct3DSamplerState9 pt; @@ -157,8 +154,7 @@ public: void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); - // TODO: void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, uint32 count, const void* vertices, size_t stride); - void IASetVertexBuffer(int count, const void* vertices, size_t stride); + void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, size_t stride); void IASetInputLayout(IDirect3DVertexDeclaration9* layout); void IASetPrimitiveTopology(D3DPRIMITIVETYPE topology); void VSSetShader(IDirect3DVertexShader9* vs, const float* vs_cb, int vs_cb_len); @@ -169,12 +165,7 @@ public: void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref); void OMSetBlendState(Direct3DBlendState9* bs, uint32 bf); void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); - void DrawPrimitive(); - - template void IASetVertexBuffer(int count, T* vertices) - { - IASetVertexBuffer(count, vertices, sizeof(T)); - } + void DrawPrimitive(uint32 count, uint32 start = 0); IDirect3DDevice9* operator->() {return m_dev;} operator IDirect3DDevice9*() {return m_dev;} diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 4f4e0cf3e9..4d366ea251 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -400,7 +400,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: if(context->TEST.DoFirstPass()) { - ((GSDevice9*)m_dev)->DrawPrimitive(); + m_tfx.Draw(); } if(context->TEST.DoSecondPass()) @@ -438,7 +438,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: m_tfx.UpdateOM(om_dssel, om_bsel, bf); - ((GSDevice9*)m_dev)->DrawPrimitive(); + m_tfx.Draw(); } } @@ -536,7 +536,16 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, }; - dev->IASetVertexBuffer(4, vertices); + void* buff = NULL; + + if(SUCCEEDED(dev->m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) + { + memcpy(buff, vertices, sizeof(vertices)); + + dev->m_convert.vb->Unlock(); + } + + dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); dev->IASetInputLayout(dev->m_convert.il); dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); @@ -556,7 +565,7 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) // - dev->DrawPrimitive(); + dev->DrawPrimitive(countof(vertices)); // @@ -570,6 +579,11 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) { GSDevice9* dev = (GSDevice9*)m_dev; + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + GSVector4 mm = GSVector4(-1, -1, 1, 1); + dev->BeginScene(); // om @@ -577,32 +591,46 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) dev->OMSetDepthStencilState(&m_fba.dss, 2); dev->OMSetBlendState(&m_fba.bs, 0); + // ia + + GSVertexPT1 vertices[] = + { + {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(0, 0)}, + }; + + void* buff = NULL; + + if(SUCCEEDED(dev->m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) + { + memcpy(buff, vertices, sizeof(vertices)); + + dev->m_convert.vb->Unlock(); + } + + dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); + // vs - dev->VSSetShader(NULL, NULL, 0); + dev->VSSetShader(dev->m_convert.vs, NULL, 0); // ps dev->PSSetShader(dev->m_convert.ps[4], NULL, 0); + // rs + + dev->RSSet(w, h); + // - int w = rt->GetWidth(); - int h = rt->GetHeight(); + dev->DrawPrimitive(countof(vertices)); - GSVertexP vertices[] = - { - {GSVector4(0, 0, 0, 0)}, - {GSVector4(w, 0, 0, 0)}, - {GSVector4(0, h, 0, 0)}, - {GSVector4(w, h, 0, 0)}, - }; - - (*dev)->SetFVF(D3DFVF_XYZRHW); - - (*dev)->DrawPrimitiveUP(D3DPT_TRIANGLESTRIP, 2, vertices, sizeof(vertices[0])); - - // + // dev->EndScene(); } diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 4295e9adb3..ab2f4ea660 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -1765,9 +1765,9 @@ bool GSC_SFEX3(const GSFrameInfo& fi, int& skip) { if(skip == 0) { - if(fi.TME && fi.FBP == 0x00f00 && fi.FPSM == PSM_PSMCT16 && (fi.TBP0 == 0x00500 || fi.TBP0 == 0x00000) && fi.TPSM == PSM_PSMCT32) + if(fi.TME && fi.FBP == 0x00500 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x00f00 && fi.TPSM == PSM_PSMCT16) { - skip = 4; + skip = 2; // blur } } diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 9045bf9970..925ddd0b9b 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -37,8 +37,6 @@ bool GSTextureFX10::Create(GSDevice10* dev) { m_dev = dev; - // - VSSelector sel; sel.bppz = 0; @@ -127,19 +125,7 @@ bool GSTextureFX10::SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMI int next = m_vb_start + m_vb_count; - if(next + count > m_vb_max) - { - if(SUCCEEDED(m_vb->Map(D3D10_MAP_WRITE_DISCARD, 0, (void**)&v))) - { - memcpy(v, vertices, count * sizeof(vertices[0])); - - m_vb->Unmap(); - } - - m_vb_start = 0; - m_vb_count = count; - } - else + if(next + count <= m_vb_max) { if(SUCCEEDED(m_vb->Map(D3D10_MAP_WRITE_NO_OVERWRITE, 0, (void**)&v))) { @@ -151,6 +137,18 @@ bool GSTextureFX10::SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMI m_vb_start = next; m_vb_count = count; } + else + { + if(SUCCEEDED(m_vb->Map(D3D10_MAP_WRITE_DISCARD, 0, (void**)&v))) + { + memcpy(v, vertices, count * sizeof(vertices[0])); + + m_vb->Unmap(); + } + + m_vb_start = 0; + m_vb_count = count; + } m_dev->IASetVertexBuffer(m_vb, sizeof(vertices[0])); m_dev->IASetInputLayout(m_il); diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 81f356cace..433b8565a7 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -25,6 +25,9 @@ GSTextureFX9::GSTextureFX9() : m_dev(NULL) + , m_vb_max(0) + , m_vb_start(0) + , m_vb_count(0) { } @@ -86,7 +89,57 @@ GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim) { - m_dev->IASetVertexBuffer(count, vertices); + HRESULT hr; + + if(max(count * 3 / 2, 10000) > m_vb_max) + { + m_vb_old = m_vb; + m_vb = NULL; + m_vb_max = max(count * 2, 10000); + m_vb_start = 0; + m_vb_count = 0; + } + + if(!m_vb) + { + hr = (*m_dev)->CreateVertexBuffer(m_vb_max * sizeof(vertices[0]), D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_vb, NULL); + + if(FAILED(hr)) return false; + } + + GSVertexHW9* v = NULL; + + int next = m_vb_start + m_vb_count; + int size = count * sizeof(vertices[0]); + + if(next + count <= m_vb_max) + { + int offset = next * sizeof(vertices[0]); + + if(SUCCEEDED(m_vb->Lock(offset, size, (void**)&v, D3DLOCK_NOOVERWRITE))) + { + memcpy(v, vertices, size); + + m_vb->Unlock(); + } + + m_vb_start = next; + m_vb_count = count; + } + else + { + if(SUCCEEDED(m_vb->Lock(0, size, (void**)&v, D3DLOCK_DISCARD))) + { + memcpy(v, vertices, size); + + m_vb->Unlock(); + } + + m_vb_start = 0; + m_vb_count = count; + } + + m_dev->IASetVertexBuffer(m_vb, sizeof(vertices[0])); m_dev->IASetInputLayout(m_il); m_dev->IASetPrimitiveTopology(prim); @@ -470,3 +523,8 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, m_dev->OMSetBlendState(bs, 0x010101 * bf); } + +void GSTextureFX9::Draw() +{ + m_dev->DrawPrimitive(m_vb_count, m_vb_start); +} diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 8a94a24b83..831f72e1f6 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -155,6 +155,12 @@ private: hash_map m_ps_ss; hash_map m_om_dss; hash_map m_om_bs; + + CComPtr m_vb, m_vb_old; + int m_vb_max; + int m_vb_start; + int m_vb_count; + hash_map m_mskfix; GSTexture* CreateMskFix(uint32 size, uint32 msk, uint32 fix); @@ -171,4 +177,5 @@ public: void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); + void Draw(); }; diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index 75f798cf7b..cd28cebad4 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -172,15 +172,32 @@ public: GSVector4i runion(const GSVector4i& a) const { - #if _M_SSE >= 0x401 + int i = (upl64(a) < uph64(a)).mask(); - return min_i32(a).upl64(max_i32(a).srl<8>()); + if(i == 0xffff) + { + #if _M_SSE >= 0x401 - #else + return min_i32(a).upl64(max_i32(a).srl<8>()); - return GSVector4i(min(x, a.x), min(y, a.y), max(z, a.z), max(w, a.w)); + #else - #endif + return GSVector4i(min(x, a.x), min(y, a.y), max(z, a.z), max(w, a.w)); + + #endif + } + + if((i & 0x00ff) == 0x00ff) + { + return *this; + } + + if((i & 0xff00) == 0xff00) + { + return a; + } + + return GSVector4i::zero(); } GSVector4i rintersect(const GSVector4i& a) const diff --git a/plugins/GSdx/res/convert10.fx b/plugins/GSdx/res/convert10.fx index 73742f9d3f..bcdfb22d7a 100644 --- a/plugins/GSdx/res/convert10.fx +++ b/plugins/GSdx/res/convert10.fx @@ -76,7 +76,7 @@ float4 ps_crt(PS_INPUT input, uint i) float4(1, 1, 1, 0) }; - return Texture.Sample(Sampler, input.t) * saturate(mask[i] + 0.25f); + return Texture.Sample(Sampler, input.t) * saturate(mask[i] + 0.5f); } float4 ps_main5(PS_INPUT input) : SV_Target0 // triangular From 08d3360376a2e36b3be2a4c06b1294735ce3aa49 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 24 May 2009 00:31:46 +0000 Subject: [PATCH 119/296] ZeroGS: Strictly in the interests of maintainability, I'm moving these functions to a separate file. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1251 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zerogs/opengl/GSmain.cpp | 296 +----------------------- plugins/zerogs/opengl/GifTransfer.cpp | 312 ++++++++++++++++++++++++++ plugins/zerogs/opengl/GifTransfer.h | 31 +++ plugins/zerogs/opengl/Makefile.am | 2 +- 4 files changed, 345 insertions(+), 296 deletions(-) create mode 100644 plugins/zerogs/opengl/GifTransfer.cpp create mode 100644 plugins/zerogs/opengl/GifTransfer.h diff --git a/plugins/zerogs/opengl/GSmain.cpp b/plugins/zerogs/opengl/GSmain.cpp index abdcf05c0a..be3a6c309e 100644 --- a/plugins/zerogs/opengl/GSmain.cpp +++ b/plugins/zerogs/opengl/GSmain.cpp @@ -470,8 +470,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) { return -1; } - if( pDsp != NULL ) - *(HWND*)pDsp = GShwnd; + if( pDsp != NULL ) *(HWND*)pDsp = GShwnd; ERROR_LOG("creating zerogs\n"); //if (conf.record) recOpen(); @@ -827,299 +826,6 @@ void CALLBACK GSvsync(int interlace) GL_REPORT_ERRORD(); } -void GIFtag(pathInfo *path, u32 *data) { - - path->tag.nloop = data[0] & 0x7fff; - path->tag.eop = (data[0] >> 15) & 0x1; - u32 tagpre = (data[1] >> 14) & 0x1; - u32 tagprim = (data[1] >> 15) & 0x7ff; - u32 tagflg = (data[1] >> 26) & 0x3; - path->tag.nreg = (data[1] >> 28)<<2; - - if (path->tag.nreg == 0) path->tag.nreg = 64; - - gs.q = 1; - -// GS_LOG("GIFtag: %8.8lx_%8.8lx_%8.8lx_%8.8lx: EOP=%d, NLOOP=%x, FLG=%x, NREG=%d, PRE=%d\n", -// data[3], data[2], data[1], data[0], -// path->tag.eop, path->tag.nloop, tagflg, path->tag.nreg, tagpre); - - path->mode = tagflg+1; - - switch (tagflg) { - case 0x0: - path->regs = *(u64 *)(data+2); - path->regn = 0; - if (tagpre) - GIFRegHandlerPRIM((u32*)&tagprim); - - break; - - case 0x1: - path->regs = *(u64 *)(data+2); - path->regn = 0; - break; - } -} - -void _GSgifPacket(pathInfo *path, u32 *pMem) { // 128bit - - int reg = (int)((path->regs >> path->regn) & 0xf); - g_GIFPackedRegHandlers[reg](pMem); - - path->regn += 4; - if (path->tag.nreg == path->regn) { - path->regn = 0; - path->tag.nloop--; - } -} - -void _GSgifRegList(pathInfo *path, u32 *pMem) { // 64bit - int reg; - - reg = (int)((path->regs >> path->regn) & 0xf); - - g_GIFRegHandlers[reg](pMem); - path->regn += 4; - if (path->tag.nreg == path->regn) { - path->regn = 0; - path->tag.nloop--; - } -} - -static int nPath3Hack = 0; - -void CALLBACK GSgetLastTag(u64* ptag) -{ - *(u32*)ptag = nPath3Hack; - nPath3Hack = 0; -} - -void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size) -{ -#ifdef _WIN32 - assert( g_hCurrentThread == GetCurrentThread() ); -#endif - -#ifdef _DEBUG - if( conf.log & 0x20 ) { - static int nSaveIndex=0; - GS_LOG("%d: p:%d %x\n", nSaveIndex++, (path==&gs.path3)?3:(path==&gs.path2?2:1), size); - int vals[4] = {0}; - for(int i = 0; i < size; i++) { - for(int j = 0; j < 4; ++j ) - vals[j] ^= pMem[4*i+j]; - } - GS_LOG("%x %x %x %x\n", vals[0], vals[1], vals[2], vals[3]); - } -#endif - - while(size > 0) - { - //LOG(_T("Transfer(%08x, %d) START\n"), pMem, size); - if (path->tag.nloop == 0) - { - GIFtag(path, pMem); - pMem+= 4; - size--; - - if ((g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop) - nPath3Hack = 1; - - if (path == &gs.path1) - { - // if too much data, just ignore - if (path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1)) - { - static int lasttime = 0; - if( timeGetTime() - lasttime > 5000 ) - { - ERROR_LOG("VU1 too much data, ignore if gfx are fine\n"); - lasttime = timeGetTime(); - } - path->tag.nloop = 0; - return; - } - - if (path->mode == 1) - { - // check if 0xb is in any reg, if yes, exit (kh2) - for(int i = 0; i < path->tag.nreg; i += 4) - { - if (((path->regs >> i)&0xf) == 11) - { - static int lasttime = 0; - if( timeGetTime() - lasttime > 5000 ) - { - ERROR_LOG("Invalid unpack type\n"); - lasttime = timeGetTime(); - } - path->tag.nloop = 0; - return; - } - } - } - } - - if(path->tag.nloop == 0 ) - { - if( path == &gs.path1 ) - { - // ffx hack - /*if( g_GameSettings & GAME_FFXHACK ) - {*/ - if( path->tag.eop ) - return; - continue; - /*}*/ - - return; - } - - if( !path->tag.eop ) - { - //DEBUG_LOG("continuing from eop\n"); - continue; - } - - continue; - } - } - - switch(path->mode) { - case 1: // PACKED - { - assert( path->tag.nloop > 0 ); - for(; size > 0; size--, pMem += 4) - { - int reg = (int)((path->regs >> path->regn) & 0xf); - - g_GIFPackedRegHandlers[reg](pMem); - - path->regn += 4; - if (path->tag.nreg == path->regn) - { - path->regn = 0; - if( path->tag.nloop-- <= 1 ) - { - size--; - pMem += 4; - break; - } - } - } - break; - } - case 2: // REGLIST - { - //GS_LOG("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4); - assert( path->tag.nloop > 0 ); - size *= 2; - for(; size > 0; pMem+= 2, size--) - { - int reg = (int)((path->regs >> path->regn) & 0xf); - g_GIFRegHandlers[reg](pMem); - path->regn += 4; - if (path->tag.nreg == path->regn) - { - path->regn = 0; - if( path->tag.nloop-- <= 1 ) - { - size--; - pMem += 2; - break; - } - } - } - - if( size & 1 ) pMem += 2; - size /= 2; - break; - } - case 3: // GIF_IMAGE (FROM_VFRAM) - case 4: // Used in the DirectX version, so we'll use it here too. - { - if(gs.imageTransfer >= 0 && gs.imageTransfer <= 1) - { - int process = min((int)size, path->tag.nloop); - - if( process > 0 ) - { - if ( gs.imageTransfer ) - ZeroGS::TransferLocalHost(pMem, process); - else - ZeroGS::TransferHostLocal(pMem, process*4); - - path->tag.nloop -= process; - pMem += process*4; size -= process; - - assert( size == 0 || path->tag.nloop == 0 ); - } - break; - } - else - { - // simulate - int process = min((int)size, path->tag.nloop); - path->tag.nloop -= process; - pMem += process*4; size -= process; - } - - break; - } - default: // GIF_IMAGE - GS_LOG("*** WARNING **** Unexpected GIFTag flag\n"); - assert(0); - path->tag.nloop = 0; - break; - } - - if( path == &gs.path1 && path->tag.eop ) - return; - } -} - -void CALLBACK GSgifTransfer2(u32 *pMem, u32 size) -{ - //GS_LOG("GSgifTransfer2 size = %lx (mode %d, gs.path2.tag.nloop = %d)\n", size, gs.path2.mode, gs.path2.tag.nloop); - - _GSgifTransfer(&gs.path2, pMem, size); -} - -void CALLBACK GSgifTransfer3(u32 *pMem, u32 size) -{ - //GS_LOG("GSgifTransfer3 size = %lx (mode %d, gs.path3.tag.nloop = %d)\n", size, gs.path3.mode, gs.path3.tag.nloop); - - nPath3Hack = 0; - _GSgifTransfer(&gs.path3, pMem, size); -} - -static int count = 0; -void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr) -{ - pathInfo *path = &gs.path1; - - //GS_LOG("GSgifTransfer1 0x%x (mode %d)\n", addr, path->mode); - - addr &= 0x3fff; - -#ifdef _DEBUG - PRIM_LOG("count: %d\n", count); - count++; -#endif - - gs.path1.tag.nloop = 0; - gs.path1.tag.eop = 0; - _GSgifTransfer(&gs.path1, (u32*)((u8*)pMem+addr), (0x4000-addr)/16); - - if( !gs.path1.tag.eop && gs.path1.tag.nloop > 0 ) { - assert( (addr&0xf) == 0 ); //BUG - gs.path1.tag.nloop = 0; - ERROR_LOG("Transfer1 - 2\n"); - return; - } -} - void CALLBACK GSreadFIFO(u64 *pMem) { //GS_LOG("GSreadFIFO\n"); diff --git a/plugins/zerogs/opengl/GifTransfer.cpp b/plugins/zerogs/opengl/GifTransfer.cpp new file mode 100644 index 0000000000..bfed0dac67 --- /dev/null +++ b/plugins/zerogs/opengl/GifTransfer.cpp @@ -0,0 +1,312 @@ +/* ZeroGS KOSMOS + * Copyright (C) 2005-2006 zerofrog@gmail.com + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include "GifTransfer.h" + + void GIFtag(pathInfo *path, u32 *data) { + + path->tag.nloop = data[0] & 0x7fff; + path->tag.eop = (data[0] >> 15) & 0x1; + u32 tagpre = (data[1] >> 14) & 0x1; + u32 tagprim = (data[1] >> 15) & 0x7ff; + u32 tagflg = (data[1] >> 26) & 0x3; + path->tag.nreg = (data[1] >> 28)<<2; + + if (path->tag.nreg == 0) path->tag.nreg = 64; + + gs.q = 1; + +// GS_LOG("GIFtag: %8.8lx_%8.8lx_%8.8lx_%8.8lx: EOP=%d, NLOOP=%x, FLG=%x, NREG=%d, PRE=%d\n", +// data[3], data[2], data[1], data[0], +// path->tag.eop, path->tag.nloop, tagflg, path->tag.nreg, tagpre); + + path->mode = tagflg+1; + + switch (tagflg) { + case 0x0: + path->regs = *(u64 *)(data+2); + path->regn = 0; + if (tagpre) + GIFRegHandlerPRIM((u32*)&tagprim); + + break; + + case 0x1: + path->regs = *(u64 *)(data+2); + path->regn = 0; + break; + } +} + +void _GSgifPacket(pathInfo *path, u32 *pMem) { // 128bit + + int reg = (int)((path->regs >> path->regn) & 0xf); + g_GIFPackedRegHandlers[reg](pMem); + + path->regn += 4; + if (path->tag.nreg == path->regn) { + path->regn = 0; + path->tag.nloop--; + } +} + +void _GSgifRegList(pathInfo *path, u32 *pMem) { // 64bit + int reg; + + reg = (int)((path->regs >> path->regn) & 0xf); + + g_GIFRegHandlers[reg](pMem); + path->regn += 4; + if (path->tag.nreg == path->regn) { + path->regn = 0; + path->tag.nloop--; + } +} + +static int nPath3Hack = 0; + +void CALLBACK GSgetLastTag(u64* ptag) +{ + *(u32*)ptag = nPath3Hack; + nPath3Hack = 0; +} + +void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size) +{ +#ifdef _WIN32 + assert( g_hCurrentThread == GetCurrentThread() ); +#endif + +#ifdef _DEBUG + if( conf.log & 0x20 ) { + static int nSaveIndex=0; + GS_LOG("%d: p:%d %x\n", nSaveIndex++, (path==&gs.path3)?3:(path==&gs.path2?2:1), size); + int vals[4] = {0}; + for(int i = 0; i < size; i++) { + for(int j = 0; j < 4; ++j ) + vals[j] ^= pMem[4*i+j]; + } + GS_LOG("%x %x %x %x\n", vals[0], vals[1], vals[2], vals[3]); + } +#endif + + while(size > 0) + { + //LOG(_T("Transfer(%08x, %d) START\n"), pMem, size); + if (path->tag.nloop == 0) + { + GIFtag(path, pMem); + pMem+= 4; + size--; + + if ((g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop) + nPath3Hack = 1; + + if (path == &gs.path1) + { + // if too much data, just ignore + if (path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1)) + { + static int lasttime = 0; + if( timeGetTime() - lasttime > 5000 ) + { + ERROR_LOG("VU1 too much data, ignore if gfx are fine\n"); + lasttime = timeGetTime(); + } + path->tag.nloop = 0; + return; + } + + if (path->mode == 1) + { + // check if 0xb is in any reg, if yes, exit (kh2) + for(int i = 0; i < path->tag.nreg; i += 4) + { + if (((path->regs >> i)&0xf) == 11) + { + static int lasttime = 0; + if( timeGetTime() - lasttime > 5000 ) + { + ERROR_LOG("Invalid unpack type\n"); + lasttime = timeGetTime(); + } + path->tag.nloop = 0; + return; + } + } + } + } + + if(path->tag.nloop == 0 ) + { + if( path == &gs.path1 ) + { + // ffx hack + /*if( g_GameSettings & GAME_FFXHACK ) + {*/ + if( path->tag.eop ) + return; + continue; + /*}*/ + + return; + } + + if( !path->tag.eop ) + { + //DEBUG_LOG("continuing from eop\n"); + continue; + } + + continue; + } + } + + switch(path->mode) { + case 1: // PACKED + { + assert( path->tag.nloop > 0 ); + for(; size > 0; size--, pMem += 4) + { + int reg = (int)((path->regs >> path->regn) & 0xf); + + g_GIFPackedRegHandlers[reg](pMem); + + path->regn += 4; + if (path->tag.nreg == path->regn) + { + path->regn = 0; + if( path->tag.nloop-- <= 1 ) + { + size--; + pMem += 4; + break; + } + } + } + break; + } + case 2: // REGLIST + { + //GS_LOG("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4); + assert( path->tag.nloop > 0 ); + size *= 2; + for(; size > 0; pMem+= 2, size--) + { + int reg = (int)((path->regs >> path->regn) & 0xf); + g_GIFRegHandlers[reg](pMem); + path->regn += 4; + if (path->tag.nreg == path->regn) + { + path->regn = 0; + if( path->tag.nloop-- <= 1 ) + { + size--; + pMem += 2; + break; + } + } + } + + if( size & 1 ) pMem += 2; + size /= 2; + break; + } + case 3: // GIF_IMAGE (FROM_VFRAM) + case 4: // Used in the DirectX version, so we'll use it here too. + { + if(gs.imageTransfer >= 0 && gs.imageTransfer <= 1) + { + int process = min((int)size, path->tag.nloop); + + if( process > 0 ) + { + if ( gs.imageTransfer ) + ZeroGS::TransferLocalHost(pMem, process); + else + ZeroGS::TransferHostLocal(pMem, process*4); + + path->tag.nloop -= process; + pMem += process*4; size -= process; + + assert( size == 0 || path->tag.nloop == 0 ); + } + break; + } + else + { + // simulate + int process = min((int)size, path->tag.nloop); + path->tag.nloop -= process; + pMem += process*4; size -= process; + } + + break; + } + default: // GIF_IMAGE + GS_LOG("*** WARNING **** Unexpected GIFTag flag\n"); + assert(0); + path->tag.nloop = 0; + break; + } + + if( path == &gs.path1 && path->tag.eop ) + return; + } +} + +void CALLBACK GSgifTransfer2(u32 *pMem, u32 size) +{ + //GS_LOG("GSgifTransfer2 size = %lx (mode %d, gs.path2.tag.nloop = %d)\n", size, gs.path2.mode, gs.path2.tag.nloop); + + _GSgifTransfer(&gs.path2, pMem, size); +} + +void CALLBACK GSgifTransfer3(u32 *pMem, u32 size) +{ + //GS_LOG("GSgifTransfer3 size = %lx (mode %d, gs.path3.tag.nloop = %d)\n", size, gs.path3.mode, gs.path3.tag.nloop); + + nPath3Hack = 0; + _GSgifTransfer(&gs.path3, pMem, size); +} + +static int count = 0; +void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr) +{ + pathInfo *path = &gs.path1; + + //GS_LOG("GSgifTransfer1 0x%x (mode %d)\n", addr, path->mode); + + addr &= 0x3fff; + +#ifdef _DEBUG + PRIM_LOG("count: %d\n", count); + count++; +#endif + + gs.path1.tag.nloop = 0; + gs.path1.tag.eop = 0; + _GSgifTransfer(&gs.path1, (u32*)((u8*)pMem+addr), (0x4000-addr)/16); + + if( !gs.path1.tag.eop && gs.path1.tag.nloop > 0 ) { + assert( (addr&0xf) == 0 ); //BUG + gs.path1.tag.nloop = 0; + ERROR_LOG("Transfer1 - 2\n"); + return; + } +} diff --git a/plugins/zerogs/opengl/GifTransfer.h b/plugins/zerogs/opengl/GifTransfer.h new file mode 100644 index 0000000000..28b43a79d0 --- /dev/null +++ b/plugins/zerogs/opengl/GifTransfer.h @@ -0,0 +1,31 @@ +/* ZeroGS KOSMOS + * Copyright (C) 2005-2006 zerofrog@gmail.com + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + +#include "GS.h" +#include "Mem.h" +#include "Regs.h" +#include "zerogs.h" + +void GIFtag(pathInfo *path, u32 *data); +void _GSgifPacket(pathInfo *path, u32 *pMem); +void _GSgifRegList(pathInfo *path, u32 *pMem); +void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size); + +extern GIFRegHandler g_GIFPackedRegHandlers[]; +extern GIFRegHandler g_GIFRegHandlers[]; \ No newline at end of file diff --git a/plugins/zerogs/opengl/Makefile.am b/plugins/zerogs/opengl/Makefile.am index d84ca8397b..3523948ac0 100644 --- a/plugins/zerogs/opengl/Makefile.am +++ b/plugins/zerogs/opengl/Makefile.am @@ -24,7 +24,7 @@ libZeroGSogl_LDADD=$(libZeroGSogl_a_OBJECTS) libZeroGSogl_a_SOURCES = \ GSmain.cpp memcpy_amd.cpp Regs.cpp x86.cpp zpipe.cpp \ -Mem.cpp rasterfont.cpp targets.cpp zerogs.cpp +Mem.cpp rasterfont.cpp targets.cpp zerogs.cpp GifTransfer.cpp libZeroGSogl_a_SOURCES += x86-32.S From 7519fb8113fab11644e0eb2d23a3d8128b053c0f Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 24 May 2009 04:34:23 +0000 Subject: [PATCH 120/296] Some quick fixes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1252 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Defs.h | 8 +++---- .../zerogs/opengl/Win32/zerogsogl_2008.vcproj | 23 ++++++++++++++++--- 2 files changed, 24 insertions(+), 7 deletions(-) diff --git a/common/include/Pcsx2Defs.h b/common/include/Pcsx2Defs.h index 646feabe6f..9060c0423c 100644 --- a/common/include/Pcsx2Defs.h +++ b/common/include/Pcsx2Defs.h @@ -51,11 +51,11 @@ # endif # else # ifdef _MSC_VER -# define jBREAKPOINT() do { __asm int 3 } while(0) +# define jBREAKPOINT() do { __asm int 3 } while(0); # else # define jBREAKPOINT() ((void) *(volatile char *) 0) # endif -# define jASSUME(exp) do { if(exp) ; else jBREAKPOINT() } while(0) +# define jASSUME(exp) do { if(exp) ; else jBREAKPOINT() } while(0); # endif #endif @@ -71,9 +71,9 @@ // #ifndef jNO_DEFAULT #define jNO_DEFAULT \ -{ \ default: \ - jASSUME(0); \ +{ \ + jASSUME(0); \ break; \ } #endif diff --git a/plugins/zerogs/opengl/Win32/zerogsogl_2008.vcproj b/plugins/zerogs/opengl/Win32/zerogsogl_2008.vcproj index 19a3b591a8..09d5998433 100644 --- a/plugins/zerogs/opengl/Win32/zerogsogl_2008.vcproj +++ b/plugins/zerogs/opengl/Win32/zerogsogl_2008.vcproj @@ -17,8 +17,8 @@ + + + @@ -278,6 +287,14 @@ RelativePath=".\Conf.cpp" > + + + + From a8094d578626b69ce5b49167e694717ea16c1608 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 24 May 2009 12:20:33 +0000 Subject: [PATCH 121/296] microVU: - Implemented an idea by Jake to use a GPR as a reg-offset index to save a few bytes from reg-moves (only did it for VI regs, didn't have enough free GPR's to do it for VF regs) - Fixed a minor problem with micro-program logs git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1254 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Alloc.inl | 6 +++--- pcsx2/x86/microVU_Compile.inl | 3 +-- pcsx2/x86/microVU_Execute.inl | 3 +-- pcsx2/x86/microVU_Flags.inl | 3 +-- pcsx2/x86/microVU_Lower.inl | 31 +++++++++++++++++-------------- pcsx2/x86/microVU_Misc.h | 4 +++- pcsx2/x86/microVU_Misc.inl | 3 +-- 7 files changed, 27 insertions(+), 26 deletions(-) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index 34a2732195..e54a0b2046 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -716,9 +716,9 @@ microVUt(void) mVUallocCFLAGb(int reg, int fInstance) { microVUt(void) mVUallocVIa(int GPRreg, int _reg_) { microVU* mVU = mVUx; - if (!_reg_ || _reg_>15) { XOR32RtoR(GPRreg, GPRreg); } + if (!_reg_) { XOR32RtoR(GPRreg, GPRreg); } else if (isMMX(_reg_)) { MOVD32MMXtoR(GPRreg, mmVI(_reg_)); } - else { MOVZX32M16toR(GPRreg, (uptr)&mVU->regs->VI[_reg_].UL); } + else { MOVZX32Rm16toR(GPRreg, gprR, (_reg_ - 9) * 16); } } microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { @@ -731,7 +731,7 @@ microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { } if (_reg_ == 0) { return; } else if (isMMX(_reg_)) { MOVD32RtoMMX(mmVI(_reg_), GPRreg); } - else if (_reg_ < 16) { MOV16RtoM((uptr)&mVU->regs->VI[_reg_].UL, GPRreg); } + else if (_reg_ < 16) { MOV32RtoRm(gprR, GPRreg, (_reg_ - 9) * 16); } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index db6a44131e..af92930a10 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -190,11 +190,10 @@ microVUt(void) mVUtestCycles() { iPC = mVUstartPC; CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); - PUSH32R(gprR); MOV32ItoR(gprT2, xPC); if (!vuIndex) CALLFunc((uptr)mVUwarning0); else CALLFunc((uptr)mVUwarning1); - POP32R(gprR); + MOV32ItoR(gprR, Roffset); // Restore gprR mVUendProgram(0, 0, sI, 0, cI); x86SetJ8(jmp8); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 148e0fd0c5..5c61f33ca6 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -42,7 +42,7 @@ microVUt(void) mVUdispatcherA() { SSE_LDMXCSR((uptr)&g_sseVUMXCSR); // Load Regs - MOV32MtoR(gprR, (uptr)&mVU->regs->VI[REG_R].UL); + MOV32ItoR(gprR, Roffset); // Load VI Reg Offset MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); AND32ItoR(gprF0, 0xffff); MOV32RtoR(gprF1, gprF0); @@ -81,7 +81,6 @@ microVUt(void) mVUdispatcherB() { SSE_LDMXCSR((uptr)&g_sseMXCSR); // Save Regs (Other Regs Saved in mVUcompile) - MOV32RtoM((uptr)&mVU->regs->VI[REG_R].UL, gprR); SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); for (int i = 1; i < 16; i++) { diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 95f4527efb..2e3e13e6ea 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -157,7 +157,6 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { if (__Status && !mVUflagHack) { int bStatus[4]; sortFlag(xStatus, bStatus, cycles); - PUSH32R(gprR); // Backup gprR MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); MOV32RtoR(gprR, getFlagReg1(bStatus[2])); @@ -165,7 +164,7 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { MOV32RtoR(gprF0, gprT1); MOV32RtoR(gprF1, gprT2); MOV32RtoR(gprF2, gprR); - POP32R(gprR); // Restore gprR + MOV32ItoR(gprR, Roffset); // Restore gprR } if (__Mac) { diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 9a192c958f..3362827ce5 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -990,29 +990,30 @@ microVUf(void) mVU_RINIT() { pass1 { mVUanalyzeR1(_Fs_, _Fsf_); } pass2 { if (_Fs_ || (_Fsf_ == 3)) { - getReg8(gprR, _Fs_, _Fsf_); - AND32ItoR(gprR, 0x007fffff); - OR32ItoR (gprR, 0x3f800000); + getReg8(gprT1, _Fs_, _Fsf_); + AND32ItoR(gprT1, 0x007fffff); + OR32ItoR (gprT1, 0x3f800000); + MOV32RtoM(Rmem, gprT1); } - else MOV32ItoR(gprR, 0x3f800000); + else MOV32ItoM(Rmem, 0x3f800000); } pass3 { mVUlog("RINIT R, vf%02d%s", _Fs_, _Fsf_String); } } -microVUt(void) mVU_RGET_() { +microVUt(void) mVU_RGET_(int Rreg) { microVU* mVU = mVUx; if (!noWriteVF) { - if (_X) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[0], gprR); - if (_Y) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[1], gprR); - if (_Z) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[2], gprR); - if (_W) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[3], gprR); + if (_X) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[0], Rreg); + if (_Y) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[1], Rreg); + if (_Z) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[2], Rreg); + if (_W) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[3], Rreg); } } microVUf(void) mVU_RGET() { microVU* mVU = mVUx; pass1 { mVUanalyzeR2(_Ft_, 1); } - pass2 { mVU_RGET_(); } + pass2 { MOV32MtoR(gprT1, Rmem); mVU_RGET_(gprT1); } pass3 { mVUlog("RGET.%s vf%02d, R", _XYZW_String, _Ft_); } } @@ -1021,6 +1022,7 @@ microVUf(void) mVU_RNEXT() { pass1 { mVUanalyzeR2(_Ft_, 0); } pass2 { // algorithm from www.project-fao.org + MOV32MtoR(gprR, Rmem); MOV32RtoR(gprT1, gprR); SHR32ItoR(gprT1, 4); AND32ItoR(gprT1, 1); @@ -1034,7 +1036,8 @@ microVUf(void) mVU_RNEXT() { XOR32RtoR(gprR, gprT1); AND32ItoR(gprR, 0x007fffff); OR32ItoR (gprR, 0x3f800000); - mVU_RGET_(); + mVU_RGET_(gprR); + MOV32ItoR(gprR, Roffset); // Restore gprR } pass3 { mVUlog("RNEXT.%s vf%02d, R", _XYZW_String, _Ft_); } } @@ -1046,7 +1049,7 @@ microVUf(void) mVU_RXOR() { if (_Fs_ || (_Fsf_ == 3)) { getReg8(gprT1, _Fs_, _Fsf_); AND32ItoR(gprT1, 0x7fffff); - XOR32RtoR(gprR, gprT1); + XOR32RtoM(Rmem, gprT1); } } pass3 { mVUlog("RXOR R, vf%02d%s", _Fs_, _Fsf_String); } @@ -1128,8 +1131,8 @@ microVUf(void) mVU_XGKICK() { //------------------------------------------------------------------ #define setBranchA(x, _x_) { \ - pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; return; } mVUbranch = x; } \ - pass2 { mVUbranch = x; } \ + mVUbranch = x; \ + pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; mVUbranch = 0; return; } } \ } microVUf(void) mVU_B() { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 306e9d5686..bcd5ae2afa 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -129,7 +129,7 @@ declareAllVariables #define gprT1 0 // Temp Reg #define gprT2 1 // Temp Reg -#define gprR 2 // R Reg +#define gprR 2 // VI Reg Offset #define gprF0 3 // Status Flag 0 #define gprESP 4 // Don't use? #define gprF1 5 // Status Flag 1 @@ -177,6 +177,8 @@ declareAllVariables #define bSaveAddr (((xPC + (2 * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) +#define Rmem (uptr)&mVU->regs->VI[REG_R].UL +#define Roffset (uptr)&mVU->regs->VI[9].UL // Flag Info #define __Status (mVUflagInfo & (0xf<<0)) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 0011b852cb..e0927813ca 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -286,7 +286,6 @@ microVUt(void) mVUbackupRegs() { microVU* mVU = mVUx; SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); SSE_MOVAPS_XMM_to_M128((uptr)&mVU->xmmPQb[0], xmmPQ); - PUSH32R(gprR); // Backup EDX } // Restore Volatile Regs @@ -296,7 +295,7 @@ microVUt(void) mVUrestoreRegs() { SSE_MOVAPS_M128_to_XMM(xmmPQ, (uptr)&mVU->xmmPQb[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); SSE_MOVAPS_M128_to_XMM(xmmMin, (uptr)mVU_minvals); - POP32R(gprR); // Restore EDX + MOV32ItoR(gprR, Roffset); // Restore gprR } // Reads entire microProgram and finds out if Status Flag is Used From bd9886b33214b7f3e13cf2251a2e529b485415bd Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 25 May 2009 01:59:17 +0000 Subject: [PATCH 122/296] Zeropad: Yes it's a bunch *more* cleanup work on Zeropad. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1256 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/gui.cpp | 3 - plugins/zeropad/Linux/linux.cpp | 324 ++++++----------------- plugins/zeropad/Linux/linux.h | 3 - plugins/zeropad/Makefile.am | 4 +- plugins/zeropad/analog.cpp | 2 +- plugins/zeropad/configure.ac | 12 +- plugins/zeropad/{Linux => }/joystick.cpp | 145 +++++++++- plugins/zeropad/{Linux => }/joystick.h | 28 +- 8 files changed, 252 insertions(+), 269 deletions(-) rename plugins/zeropad/{Linux => }/joystick.cpp (54%) rename plugins/zeropad/{Linux => }/joystick.h (80%) diff --git a/plugins/zeropad/Linux/gui.cpp b/plugins/zeropad/Linux/gui.cpp index d2dd512a7d..b7b6bccfb2 100644 --- a/plugins/zeropad/Linux/gui.cpp +++ b/plugins/zeropad/Linux/gui.cpp @@ -20,10 +20,7 @@ #include #include -#define JOYSTICK_SUPPORT -#ifdef JOYSTICK_SUPPORT #include "joystick.h" -#endif #include "zeropad.h" #include "linux.h" diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 8378fe5295..250e956fd1 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -48,9 +48,7 @@ s32 _PADopen(void *pDsp) s_keyRelease[0] = s_keyRelease[1] = 0; XAutoRepeatOff(GSdsp); -#ifdef JOYSTICK_SUPPORT JoystickInfo::EnumerateJoysticks(s_vjoysticks); -#endif return 0; } @@ -100,6 +98,16 @@ int _GetJoystickIdFromPAD(int pad) } } } + + if ((joyid < 0) || (joyid >= (int)s_vjoysticks.size())) + { + // get first unused joystick + for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) + { + if (s_vjoysticks[joyid]->GetPAD() < 0) break; + } + } + return joyid; } @@ -186,8 +194,6 @@ void CALLBACK PADupdate(int pad) } // joystick info -#ifdef JOYSTICK_SUPPORT - SDL_JoystickUpdate(); for (int i = 0; i < PADKEYS; i++) @@ -202,22 +208,24 @@ void CALLBACK PADupdate(int pad) if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) { pjoy = s_vjoysticks[joyid]; + int pad = (pjoy)->GetPAD(); + int value = SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key)); - if (SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key))) - clear_bit(status[(pjoy)->GetPAD()], i); // pressed + if (value) + clear_bit(status[pad], i); // pressed else - set_bit(status[(pjoy)->GetPAD()], i); // pressed + set_bit(status[pad], i); // pressed } } else if (IS_JOYSTICK(key)) { int joyid = PAD_GETJOYID(key); + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) { - pjoy = s_vjoysticks[joyid]; - int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); + int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); switch (i) { @@ -233,34 +241,16 @@ void CALLBACK PADupdate(int pad) } } } - else if (IS_POV(key)) - { - int joyid = PAD_GETJOYID(key); - if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) - { - pjoy = s_vjoysticks[joyid]; - - int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - int pad = (pjoy)->GetPAD(); - - if (PAD_GETPOVSIGN(key) && (value < -2048)) - clear_bit(status[pad], i); - else if (!PAD_GETPOVSIGN(key) && (value > 2048)) - clear_bit(status[pad], i); - else - set_bit(status[pad], i); - } - } #ifdef EXPERIMENTAL_POV_CODE else if (IS_HAT(key)) { int joyid = PAD_GETJOYID(key); + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) { pjoy = s_vjoysticks[joyid]; - - int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); int pad = (pjoy)->GetPAD(); + int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); if ((value != SDL_HAT_CENTERED) && (PAD_GETHATDIR(key) == value)) @@ -268,45 +258,12 @@ void CALLBACK PADupdate(int pad) if ((value == SDL_HAT_UP) || (value == SDL_HAT_RIGHT) || (value == SDL_HAT_DOWN) ||(value == SDL_HAT_LEFT)) { set_bit(status[pad], i); + PAD_LOG("Registered %s. Set (%d)\n", HatName(value), i); } else { clear_bit(status[pad], i); } - switch (i) - { - case PAD_LEFT: - PAD_LOG("Hat Left!\n"); - break; - case PAD_DOWN: - PAD_LOG("Hat Down!\n"); - break; - case PAD_RIGHT: - PAD_LOG("Hat Right!\n"); - break; - case PAD_UP: - PAD_LOG("Hat Up!\n"); - break; - } - - switch (value) - { - case SDL_HAT_UP: - PAD_LOG("D-pad Up!\n"); - break; - - case SDL_HAT_RIGHT: - PAD_LOG("D-pad Right!\n"); - break; - - case SDL_HAT_DOWN: - PAD_LOG("D-pad Down!\n"); - break; - - case SDL_HAT_LEFT: - PAD_LOG("D-pad Left!\n"); - break; - } } else { @@ -315,8 +272,27 @@ void CALLBACK PADupdate(int pad) } } #endif + else if (IS_POV(key)) + { + int joyid = PAD_GETJOYID(key); + + if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) + { + pjoy = s_vjoysticks[joyid]; + int pad = (pjoy)->GetPAD(); + int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); + + if (PAD_GETPOVSIGN(key) && (value < -2048)) + clear_bit(status[pad], i); + else if (!PAD_GETPOVSIGN(key) && (value > 2048)) + clear_bit(status[pad], i); + else + set_bit(status[pad], i); + + + } + } } -#endif pthread_spin_lock(&s_mutexStatus); s_keyPress[pad] |= keyPress; @@ -335,6 +311,7 @@ void UpdateConf(int pad) GtkWidget *Btn; for (i = 0; i < ArraySize(s_pGuiKeyMap); i++) { + string tmp; if (s_pGuiKeyMap[i] == NULL) continue; @@ -344,8 +321,7 @@ void UpdateConf(int pad) PAD_LOG("ZeroPAD: cannot find key %s\n", s_pGuiKeyMap[i]); continue; } - - string tmp; + if (IS_KEYBOARD(conf.keys[pad][i])) { char* pstr = XKeysymToString(PAD_GETKEY(conf.keys[pad][i])); @@ -353,44 +329,49 @@ void UpdateConf(int pad) } else if (IS_JOYBUTTONS(conf.keys[pad][i])) { + int button = PAD_GETJOYBUTTON(conf.keys[pad][i]); tmp.resize(28); - sprintf(&tmp[0], "JBut %d", PAD_GETJOYBUTTON(conf.keys[pad][i])); + + sprintf(&tmp[0], "JBut %d", button); } else if (IS_JOYSTICK(conf.keys[pad][i])) { + int axis = PAD_GETJOYSTICK_AXIS(conf.keys[pad][i]); tmp.resize(28); - sprintf(&tmp[0], "JAxis %d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + + sprintf(&tmp[0], "JAxis %d", axis); } #ifdef EXPERIMENTAL_POV_CODE else if (IS_HAT(conf.keys[pad][i])) { + int axis = PAD_GETJOYSTICK_AXIS(conf.keys[pad][i]); tmp.resize(28); + switch(PAD_GETHATDIR(conf.keys[pad][i])) { case SDL_HAT_UP: - sprintf(&tmp[0], "JPOVU-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + sprintf(&tmp[0], "JPOVU-%d", axis); break; case SDL_HAT_RIGHT: - sprintf(&tmp[0], "JPOVR-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + sprintf(&tmp[0], "JPOVR-%d", axis); break; case SDL_HAT_DOWN: - sprintf(&tmp[0], "JPOVD-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + sprintf(&tmp[0], "JPOVD-%d", axis); break; case SDL_HAT_LEFT: - sprintf(&tmp[0], "JPOVL-%d", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i])); + sprintf(&tmp[0], "JPOVL-%d", axis); break; } } -#else +#endif else if (IS_POV(conf.keys[pad][i])) { tmp.resize(28); sprintf(&tmp[0], "JPOV %d%s", PAD_GETJOYSTICK_AXIS(conf.keys[pad][i]), PAD_GETPOVSIGN(conf.keys[pad][i]) ? "-" : "+"); } -#endif if (tmp.size() > 0) { @@ -404,17 +385,8 @@ void UpdateConf(int pad) // check bounds int joyid = _GetJoystickIdFromPAD(pad); - - if ((joyid < 0) || (joyid >= (int)s_vjoysticks.size())) - { - // get first unused joystick - for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) - { - if (s_vjoysticks[joyid]->GetPAD() < 0) break; - } - } - if (joyid >= 0 && joyid < (int)s_vjoysticks.size()) + if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), joyid); // select the combo else gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), s_vjoysticks.size()); // no gamepad @@ -446,21 +418,7 @@ int GetLabelId(GtkWidget *label) return (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); } -void UpdateJoysticks(vector::iterator itjoy) -{ - itjoy = s_vjoysticks.begin(); - - SDL_JoystickUpdate(); - - // Save everything in the vector s_vjoysticks. - while (itjoy != s_vjoysticks.end()) - { - (*itjoy)->SaveState(); - itjoy++; - } -} - -bool PollKeyboard(char* &temp, u32* &pkey) +bool PollKeyboard(char* &temp, u32 &pkey) { GdkEvent *ev = gdk_event_get(); @@ -472,15 +430,14 @@ bool PollKeyboard(char* &temp, u32* &pkey) if (ev->key.keyval == GDK_Escape) { temp = "Unknown"; - *pkey = NULL; + pkey = NULL; } else { temp = XKeysymToString(ev->key.keyval); - *pkey = ev->key.keyval; + pkey = ev->key.keyval; } - return true; } } @@ -488,170 +445,64 @@ bool PollKeyboard(char* &temp, u32* &pkey) return false; } -bool PollButtons(vector::iterator itjoy, int &jbutton, u32* &pkey) -{ - // MAKE sure to look for changes in the state!! - for (int i = 0; i < (*itjoy)->GetNumButtons(); ++i) - { - int but = SDL_JoystickGetButton((*itjoy)->GetJoy(), i); - - if (but != (*itjoy)->GetButtonState(i)) - { - if (!but) // released, we don't really want this - { - (*itjoy)->SetButtonState(i, 0); - break; - } - - *pkey = PAD_JOYBUTTON((*itjoy)->GetId(), i); - jbutton = i; - return true; - } - } - - return false; -} - -bool PollAxes(vector::iterator itjoy, bool pov, int &jbutton, bool &negative, u32* &pkey) -{ - for (int i = 0; i < (*itjoy)->GetNumAxes(); ++i) - { - int value = SDL_JoystickGetAxis((*itjoy)->GetJoy(), i); - - if (value != (*itjoy)->GetAxisState(i)) - { - PAD_LOG("Change in joystick %d: %d.\n", i, value); - - if (abs(value) <= (*itjoy)->GetAxisState(i)) // we don't want this - { - // released, we don't really want this - (*itjoy)->SetAxisState(i, value); - break; - } - - if (abs(value) > 0x3fff) - { - jbutton = i; - - if (pov) - { - negative = (value < 0); - *pkey = PAD_POV((*itjoy)->GetId(), negative, i); - } - else // axis - { - *pkey = PAD_JOYSTICK((*itjoy)->GetId(), i); - } - return true; - } - } - } - return false; -} - -/*SDL_HAT_CENTERED, -SDL_HAT_UP, SDL_HAT_RIGHT, -SDL_HAT_DOWN, SDL_HAT_LEFT, -SDL_HAT_RIGHTUP, SDL_HAT_RIGHTDOWN, -SDL_HAT_LEFTUP, SDL_HAT_LEFTDOWN*/ - -bool PollHAT(vector::iterator itjoy, int &jbutton, int &dir, u32* &pkey) -{ -#ifdef EXPERIMENTAL_POV_CODE - for (int i = 0; i < (*itjoy)->GetNumPOV(); ++i) - { - int value = SDL_JoystickGetHat((*itjoy)->GetJoy(), i); - - if (value != SDL_HAT_CENTERED) - { - switch (value) - { - case SDL_HAT_UP: - case SDL_HAT_RIGHT: - case SDL_HAT_DOWN: - case SDL_HAT_LEFT: - *pkey = PAD_HAT((*itjoy)->GetId(), value, i); - jbutton = i; - dir = value; - PAD_LOG("Hat Pressed!"); - return true; - default: - break; - } - } - } -#endif - return false; -} - void OnConf_Key(GtkButton *button, gpointer user_data) { - u32 *pkey; GtkWidget* label = GetLabelWidget(button); + bool captured = false; + char str[32]; int id = GetLabelId(label); if (id == -1) return; int pad = id / PADKEYS; int key = id % PADKEYS; - pkey = &conf.keys[pad][key]; - // save the states -#ifdef JOYSTICK_SUPPORT - vector::iterator itjoy; + // save the joystick states + UpdateJoysticks(); - UpdateJoysticks(itjoy); -#endif - - for (;;) + while (!captured) { + vector::iterator itjoy; char *tmp; - if (PollKeyboard(tmp, pkey)) + if (PollKeyboard(tmp, conf.keys[pad][key])) { - gtk_entry_set_text(GTK_ENTRY(label), tmp); - return; + strcpy(str, tmp); + captured = true; + break; } -#ifdef JOYSTICK_SUPPORT - itjoy = s_vjoysticks.begin(); - SDL_JoystickUpdate(); - while (itjoy != s_vjoysticks.end()) + itjoy = s_vjoysticks.begin(); + while ((itjoy != s_vjoysticks.end()) && (!captured)) { int jbutton, direction; - if (PollButtons(itjoy, jbutton, pkey)) + if ((*itjoy)->PollButtons(jbutton, conf.keys[pad][key])) { - char str[32]; sprintf(str, "JBut %d", jbutton); - gtk_entry_set_text(GTK_ENTRY(label), str); - return; + captured = true; + break; } - bool negative = false, pov = (key < 16); + bool negative = false; + bool pov = (!((key == PAD_RY) || (key == PAD_LY) || (key == PAD_RX) || (key == PAD_LX))); - if (PollAxes(itjoy, pov, jbutton, negative, pkey)) + if ((*itjoy)->PollAxes(pov, jbutton, negative, conf.keys[pad][key])) { - char str[32]; - if (pov) sprintf(str, "JPOV %d%s", jbutton, (negative) ? "-" : "+"); else sprintf(str, "JAxis %d", jbutton); - - gtk_entry_set_text(GTK_ENTRY(label), str); - - return; + captured = true; + break; } #ifdef EXPERIMENTAL_POV_CODE - if (PollHAT(itjoy, jbutton, direction, pkey)) + if ((*itjoy)->PollHats(jbutton, direction, conf.keys[pad][key])) { - char str[32]; - switch (direction) { case SDL_HAT_UP: sprintf(str, "JPOVU-%d", jbutton); break; @@ -659,16 +510,15 @@ void OnConf_Key(GtkButton *button, gpointer user_data) case SDL_HAT_DOWN: sprintf(str, "JPOVD-%d", jbutton); break; case SDL_HAT_LEFT: sprintf(str, "JPOVL-%d", jbutton); break; } - - gtk_entry_set_text(GTK_ENTRY(label), str); - return; + captured = true; + break; } #endif - itjoy++; } -#endif } + + gtk_entry_set_text(GTK_ENTRY(label), str); } void CALLBACK PADconfigure() @@ -683,9 +533,7 @@ void CALLBACK PADconfigure() Conf = create_Conf(); // recreate -#ifdef JOYSTICK_SUPPORT JoystickInfo::EnumerateJoysticks(s_vjoysticks); -#endif s_devicecombo = lookup_widget(Conf, "joydevicescombo"); @@ -696,8 +544,8 @@ void CALLBACK PADconfigure() // Delete everything in the vector vjoysticks. while (it != s_vjoysticks.end()) { - sprintf(str, "%d: %s - but: %d, axes: %d, pov: %d", (*it)->GetId(), (*it)->GetName().c_str(), - (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumPOV()); + sprintf(str, "%d: %s - but: %d, axes: %d, hats: %d", (*it)->GetId(), (*it)->GetName().c_str(), + (*it)->GetNumButtons(), (*it)->GetNumAxes(), (*it)->GetNumHats()); gtk_combo_box_append_text(GTK_COMBO_BOX(s_devicecombo), str); it++; } diff --git a/plugins/zeropad/Linux/linux.h b/plugins/zeropad/Linux/linux.h index 0dedcde2f1..5a12f8c145 100644 --- a/plugins/zeropad/Linux/linux.h +++ b/plugins/zeropad/Linux/linux.h @@ -21,10 +21,7 @@ #include #include -#define JOYSTICK_SUPPORT -#ifdef JOYSTICK_SUPPORT #include "joystick.h" -#endif #include "zeropad.h" diff --git a/plugins/zeropad/Makefile.am b/plugins/zeropad/Makefile.am index 987d3f3dd2..3937ec3c90 100644 --- a/plugins/zeropad/Makefile.am +++ b/plugins/zeropad/Makefile.am @@ -22,5 +22,5 @@ libZeroPAD_LDFLAGS= @SHARED_LDFLAGS@ libZeroPAD_LDFLAGS+=-Wl,-soname,@ZEROPAD_SONAME@ libZeroPAD_LDADD=$(libZeroPAD_a_OBJECTS) -libZeroPAD_a_SOURCES = analog.cpp analog.h zeropad.cpp zeropad.h \ -Linux/gui.cpp Linux/linux.cpp Linux/joystick.cpp Linux/support.c Linux/interface.c +libZeroPAD_a_SOURCES = joystick.cpp analog.cpp analog.h zeropad.cpp zeropad.h \ +Linux/gui.cpp Linux/linux.cpp Linux/support.c Linux/interface.c diff --git a/plugins/zeropad/analog.cpp b/plugins/zeropad/analog.cpp index e0fd490eb9..009f45c8db 100644 --- a/plugins/zeropad/analog.cpp +++ b/plugins/zeropad/analog.cpp @@ -126,7 +126,7 @@ namespace Analog if (RevertPad(padvalue)) InvertPad(padvalue, i); SetPad(padvalue, i, Pad(padvalue, i) + 0x80); - PAD_LOG("Setting pad[%d]@%d to %d from %d\n", padvalue, i, value, temp); + //PAD_LOG("Setting pad[%d]@%d to %d from %d\n", padvalue, i, value, temp); } int AnalogToPad(int padvalue) { diff --git a/plugins/zeropad/configure.ac b/plugins/zeropad/configure.ac index db2f1a4f10..89e5c48c0a 100644 --- a/plugins/zeropad/configure.ac +++ b/plugins/zeropad/configure.ac @@ -29,14 +29,14 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), if test "x$debug" == xyes then AC_DEFINE(_DEBUG,1,[_DEBUG]) - CFLAGS+="-g -m32 " - CXXFLAGS+="-g -m32 " - CCASFLAGS+=" -m32 " + CFLAGS+="-g -m32 -fpic " + CXXFLAGS+="-g -m32 -fpic " + CCASFLAGS+=" -m32 -fpic " else AC_DEFINE(NDEBUG,1,[NDEBUG]) - CFLAGS+="-O2 -fomit-frame-pointer -m32 " - CXXFLAGS+="-O2 -fomit-frame-pointer -m32 " - CCASFLAGS+=" -m32 " + CFLAGS+="-O2 -fomit-frame-pointer -m32 -fpic " + CXXFLAGS+="-O2 -fomit-frame-pointer -m32 -fpic " + CCASFLAGS+=" -m32 -fpic " fi AM_CONDITIONAL(DEBUGBUILD, test x$debug = xyes) AC_MSG_RESULT($debug) diff --git a/plugins/zeropad/Linux/joystick.cpp b/plugins/zeropad/joystick.cpp similarity index 54% rename from plugins/zeropad/Linux/joystick.cpp rename to plugins/zeropad/joystick.cpp index f1d05204e7..579490b694 100644 --- a/plugins/zeropad/Linux/joystick.cpp +++ b/plugins/zeropad/joystick.cpp @@ -26,6 +26,49 @@ int s_selectedpad = 0; vector s_vjoysticks; static u32 s_bSDLInit = false; +void UpdateJoysticks() +{ + vector::iterator itjoy = s_vjoysticks.begin(); + + SDL_JoystickUpdate(); + + // Save everything in the vector s_vjoysticks. + while (itjoy != s_vjoysticks.end()) + { + (*itjoy)->SaveState(); + itjoy++; + } +} + +const char *HatName(int value) +{ + switch(value) + { + case SDL_HAT_CENTERED: + return "SDL_HAT_CENTERED"; + case SDL_HAT_UP: + return "SDL_HAT_UP"; + case SDL_HAT_RIGHT: + return "SDL_HAT_RIGHT"; + case SDL_HAT_DOWN: + return "SDL_HAT_DOWN"; + case SDL_HAT_LEFT: + return "SDL_HAT_LEFT"; + case SDL_HAT_RIGHTUP: + return "SDL_HAT_RIGHTUP"; + case SDL_HAT_RIGHTDOWN: + return "SDL_HAT_RIGHTDOWN"; + case SDL_HAT_LEFTUP: + return "SDL_HAT_LEFTUP"; + case SDL_HAT_LEFTDOWN: + return "SDL_HAT_LEFTDOWN"; + default: + return "Unknown"; + } + + return "Unknown"; +} + // opens handles to all possible joysticks void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) { @@ -107,14 +150,14 @@ bool JoystickInfo::Init(int id, bool bStartThread) numaxes = SDL_JoystickNumAxes(joy); numbuttons = SDL_JoystickNumButtons(joy); - numpov = SDL_JoystickNumHats(joy); + numhats = SDL_JoystickNumHats(joy); devname = SDL_JoystickName(id); vaxisstate.resize(numaxes); - vbutstate.resize(numbuttons); - vpovstate.resize(numpov); + vbuttonstate.resize(numbuttons); + vhatstate.resize(numhats); - //PAD_LOG("There are %d buttons, %d axises, and %d povs.\n", numbuttons, numaxes, numpov); + //PAD_LOG("There are %d buttons, %d axises, and %d hats.\n", numbuttons, numaxes, numhats); return true; } @@ -146,10 +189,100 @@ void JoystickInfo::SaveState() SetButtonState(i, SDL_JoystickGetButton(joy, i)); for (int i = 0; i < numaxes; ++i) SetAxisState(i, SDL_JoystickGetAxis(joy, i)); - for (int i = 0; i < numpov; ++i) - SetPOVState(i, SDL_JoystickGetHat(joy, i)); + for (int i = 0; i < numhats; ++i) + SetHatState(i, SDL_JoystickGetHat(joy, i)); } void JoystickInfo::TestForce() { } + +bool JoystickInfo::PollButtons(int &jbutton, u32 &pkey) +{ + // MAKE sure to look for changes in the state!! + for (int i = 0; i < GetNumButtons(); ++i) + { + int but = SDL_JoystickGetButton(GetJoy(), i); + + if (but != GetButtonState(i)) + { + if (!but) // released, we don't really want this + { + SetButtonState(i, 0); + break; + } + + pkey = PAD_JOYBUTTON(GetId(), i); + jbutton = i; + return true; + } + } + + return false; +} + +bool JoystickInfo::PollAxes(bool pov, int &jbutton, bool &negative, u32 &pkey) +{ + for (int i = 0; i < GetNumAxes(); ++i) + { + int value = SDL_JoystickGetAxis(GetJoy(), i); + + if (value != GetAxisState(i)) + { + PAD_LOG("Change in joystick %d: %d.\n", i, value); + + if (abs(value) <= GetAxisState(i)) // we don't want this + { + // released, we don't really want this + SetAxisState(i, value); + break; + } + + if (abs(value) > 0x3fff) + { + jbutton = i; + + if (pov) + { + negative = (value < 0); + pkey = PAD_POV(GetId(), negative, i); + } + else // axis + { + pkey = PAD_JOYSTICK(GetId(), i); + } + return true; + } + } + } + return false; +} + +bool JoystickInfo::PollHats(int &jbutton, int &dir, u32 &pkey) +{ +#ifdef EXPERIMENTAL_POV_CODE + for (int i = 0; i < GetNumHats(); ++i) + { + int value = SDL_JoystickGetHat(GetJoy(), i); + + if (value != SDL_HAT_CENTERED) + { + switch (value) + { + case SDL_HAT_UP: + case SDL_HAT_RIGHT: + case SDL_HAT_DOWN: + case SDL_HAT_LEFT: + pkey = PAD_HAT(GetId(), value, i); + jbutton = i; + dir = value; + PAD_LOG("Hat Pressed!"); + return true; + default: + break; + } + } + } +#endif + return false; +} diff --git a/plugins/zeropad/Linux/joystick.h b/plugins/zeropad/joystick.h similarity index 80% rename from plugins/zeropad/Linux/joystick.h rename to plugins/zeropad/joystick.h index 43cbce9f90..4f9d5b0052 100644 --- a/plugins/zeropad/Linux/joystick.h +++ b/plugins/zeropad/joystick.h @@ -41,6 +41,12 @@ class JoystickInfo void TestForce(); + bool PollButtons(int &jbutton, u32 &pkey); + + bool PollAxes(bool pov, int &jbutton, bool &negative, u32 &pkey); + + bool PollHats(int &jbutton, int &dir, u32 &pkey); + const string& GetName() { return devname; @@ -56,9 +62,9 @@ class JoystickInfo return numaxes; } - int GetNumPOV() + int GetNumHats() { - return numpov; + return numhats; } int GetId() @@ -80,7 +86,7 @@ class JoystickInfo int GetButtonState(int i) { - return vbutstate[i]; + return vbuttonstate[i]; } int GetAxisState(int i) @@ -88,15 +94,15 @@ class JoystickInfo return vaxisstate[i]; } - int GetPOVState(int i) + int GetHatState(int i) { //PAD_LOG("Getting POV State of %d.\n", i); - return vpovstate[i]; + return vhatstate[i]; } void SetButtonState(int i, int state) { - vbutstate[i] = state; + vbuttonstate[i] = state; } void SetAxisState(int i, int value) @@ -104,10 +110,10 @@ class JoystickInfo vaxisstate[i] = value; } - void SetPOVState(int i, int value) + void SetHatState(int i, int value) { //PAD_LOG("We should set %d to %d.\n", i, value); - vpovstate[i] = value; + vhatstate[i] = value; } SDL_Joystick* GetJoy() @@ -119,11 +125,11 @@ class JoystickInfo string devname; // pretty device name int _id; - int numbuttons, numaxes, numpov; + int numbuttons, numaxes, numhats; int axisrange, deadzone; int pad; - vector vbutstate, vaxisstate, vpovstate; + vector vbuttonstate, vaxisstate, vhatstate; SDL_Joystick* joy; }; @@ -131,4 +137,6 @@ class JoystickInfo extern int s_selectedpad; extern vector s_vjoysticks; +extern void UpdateJoysticks(); +extern const char *HatName(int value); #endif From 44378b31f7e41d7fae38b121c698f354bcc31be0 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 07:07:59 +0000 Subject: [PATCH 123/296] Path3 Masking mystery solved, Simpsons (Issue 24) GT4 (Issue 211), tenchu and many many other path3 masking games should be fixed. This will only work on MTGS. Also semi-removed my gifsplit thing to resolve the tekken 5 menus Issue 209 Note: GT4 is still missing its logo, this is down to a really bizzare timing issue i cant pinpoint. Removed redundant path3hack git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1257 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Elfheader.cpp | 5 -- pcsx2/GS.cpp | 2 - pcsx2/Gif.cpp | 135 ++++++++++++++++++++------------------- pcsx2/MTGS.cpp | 41 +++++++++--- pcsx2/Patch.cpp | 6 -- pcsx2/Patch.h | 2 - pcsx2/VifDma.cpp | 68 ++++++++------------ pcsx2/VifDma.h | 2 +- pcsx2/x86/iVUzerorec.cpp | 4 +- 9 files changed, 133 insertions(+), 132 deletions(-) diff --git a/pcsx2/Elfheader.cpp b/pcsx2/Elfheader.cpp index c264f53a38..5bf94204b3 100644 --- a/pcsx2/Elfheader.cpp +++ b/pcsx2/Elfheader.cpp @@ -571,7 +571,6 @@ int loadElfFile(const char *filename) } #include "VU.h" -extern bool path3hack; int g_VUGameFixes = 0; // fixme - this should be moved to patches or eliminated @@ -584,9 +583,5 @@ void LoadGameSpecificSettings() case 0xb99379b7: // erementar gerad (discolored chars) g_VUGameFixes |= VUFIX_XGKICKDELAY2; // Tested - still needed - arcum42 break; - //case 0xa08c4057: //Sprint Cars (SLUS) - //case 0x8b0725d5: //Flinstones Bedrock Racing (SLES) - //path3hack = TRUE; // We can move this to patch files right now - //break; } } diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index 24b440538e..8b8fe7e958 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -243,8 +243,6 @@ void gsReset() memzero_obj(g_RealGSMem); - Path3transfer = FALSE; - GSCSRr = 0x551B400F; // Set the FINISH bit to 1 for now GSIMR = 0x7f00; psHu32(GIF_STAT) = 0; diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index edb8ee7069..9e189856ae 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -28,7 +28,7 @@ using std::min; -#define gifsplit 64 +#define gifsplit 0x10000 enum gifstate_t { GIF_STATE_READY = 0, @@ -57,6 +57,9 @@ __forceinline void gsInterrupt() { //Console::WriteLn("Eh? why are you still interrupting! chcr %x, qwc %x, done = %x", params gif->chcr, gif->qwc, done); return; } + + if(Path3progress == 2) vif1Regs->stat &= ~VIF1_STAT_VGW; + if (gif->qwc > 0 || gspath3done == 0) { if (!(psHu32(DMAC_CTRL) & 0x1)) { Console::Notice("gs dma masked, re-scheduling..."); @@ -66,19 +69,11 @@ __forceinline void gsInterrupt() { } GIFdma(); -#ifdef GSPATH3FIX - // re-raise the IRQ as part of the mysterious Path3fix. - // fixme - this hack *should* have the gs_irq raised from the VIF, I think. It would be - // more efficient and more correct. (air) - /*if (!(vif1Regs->mskpath3 && (vif1ch->chcr & 0x100)) || (psHu32(GIF_MODE) & 0x1)) - CPU_INT( 2, 64 );*/ -#endif - if(gspath3done == 0 || gif->qwc > 0) return; + return; } gspath3done = 0; gscycles = 0; - Path3transfer = FALSE; gif->chcr &= ~0x100; GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x4000; //FIFO empty @@ -89,12 +84,10 @@ __forceinline void gsInterrupt() { } -static void WRITERING_DMA(u32 *pMem, u32 qwc) +static u32 WRITERING_DMA(u32 *pMem, u32 qwc) { psHu32(GIF_STAT) |= 0xE00; - // Path3 transfer will be set to false by the GIFtag handler. - Path3transfer = TRUE; if( mtgsThread != NULL ) { @@ -114,15 +107,14 @@ static void WRITERING_DMA(u32 *pMem, u32 qwc) memcpy_aligned(pgsmem, pMem, sizetoread<<4); mtgsThread->SendDataPacket(); + return sizetoread; } else { GSGIFTRANSFER3(pMem, qwc); - if (GSgetLastTag != NULL) - { - GSgetLastTag(&s_gstag); - if (s_gstag == 1) Path3transfer = FALSE; /* fixes SRS and others */ - } + gif->madr+= qwc*16; + gif->qwc -= qwc; + return qwc; } } @@ -145,11 +137,11 @@ int _GIFchain() { Console::Notice( "Hackfix - NULL GIFchain" ); return -1; } - WRITERING_DMA(pMem, qwc); + qwc = WRITERING_DMA(pMem, qwc); - gif->madr+= qwc*16; - gif->qwc -= qwc; - return (qwc)*2; + /*gif->madr+= qwc*16; + gif->qwc -= qwc;*/ + return (qwc); } static __forceinline void GIFchain() @@ -161,10 +153,7 @@ static __forceinline void GIFchain() static __forceinline void dmaGIFend() { - if ((psHu32(GIF_MODE) & 0x4) && gif->qwc != 0) - CPU_INT(2, min( 8, (int)gif->qwc ) /** BIAS*/); - else - CPU_INT(2, min( gifsplit, (int)gif->qwc ) /** BIAS*/); + CPU_INT(2, 2); } // These could probably be consolidated into one function, @@ -172,10 +161,10 @@ static __forceinline void dmaGIFend() // not to do the gif->qwc != 0 check. --arcum42 static __forceinline void GIFdmaEnd() { - if (psHu32(GIF_MODE) & 0x4) - CPU_INT(2, min( 8, (int)gif->qwc ) /** BIAS*/); - else - CPU_INT(2, min( gifsplit, (int)gif->qwc ) /** BIAS*/); + //if (psHu32(GIF_MODE) & 0x4) + CPU_INT(2, gscycles * BIAS); + /*else + CPU_INT(2, min( gifsplit, (int)gif->qwc ) * BIAS);*/ } void GIFdma() @@ -183,7 +172,7 @@ void GIFdma() u32 *ptag; u32 id; - gscycles= prevcycles ? prevcycles: gscycles; + gscycles= prevcycles ? prevcycles: 0; if ((psHu32(GIF_CTRL) & 8)) { // temporarily stop Console::WriteLn("Gif dma temp paused?"); @@ -219,6 +208,13 @@ void GIFdma() GSCSRr |= 0x8000; //FIFO full psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [ used to be 0xE00; // OPH=1 | APATH=3] + if(((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == 0) //Path2 gets priority in intermittent mode + { + //GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); + dmaGIFend(); + return; + } + #ifdef GSPATH3FIX if (vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1)) { if(gif->qwc == 0) { @@ -243,28 +239,26 @@ void GIFdma() gspath3done |= 1; } } + } + + if(Path3progress == 2/* && gif->qwc != 0*/) + { + dmaGIFend(); + return; } - - GIFchain(); - if((gif->qwc == 0) && ((gspath3done == 1) || (gif->chcr & 0xc) == 0)){ - gspath3done = 0; - gif->chcr &= ~0x100; - GSCSRr &= ~0xC000; - GSCSRr |= 0x4000; - Path3transfer = FALSE; - psHu32(GIF_STAT)&= ~0x1F000E00; // OPH=0 | APATH=0 | QFC=0 - hwDmacIrq(DMAC_GIF); - } - //Dont unfreeze xmm regs here, Masked PATH3 can only be called by VIF, which is already handling it. + GIFchain(); + GIFdmaEnd(); return; } #endif // Transfer Dn_QWC from Dn_MADR to GIF if ((gif->chcr & 0xc) == 0 || gif->qwc > 0) { // Normal Mode + if ((((psHu32(DMAC_CTRL) & 0xC0) == 0x80) && ((gif->chcr & 0xc) == 0))) { Console::WriteLn("DMA Stall Control on GIF normal"); } + GIFchain(); //Transfers the data set by the switch if (((gif->qwc == 0) && (gif->chcr & 0xc) == 0)) @@ -272,9 +266,8 @@ void GIFdma() else if(gif->qwc > 0) { GIFdmaEnd(); - return; - - } + return; + } } if ((gif->chcr & 0xc) == 0x4 && gspath3done == 0) { @@ -322,23 +315,23 @@ void GIFdma() prevcycles = 0; if (!(vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1))) { - if (gspath3done == 0 || gif->qwc > 0) + if (gspath3done == 0 && gif->qwc == 0) { - if (gif->qwc != 0) - { - GIFdmaEnd(); - } - else - { - ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR - gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag - gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 - - GIFdmaEnd(); - gif->qwc = 0; - return; + + ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR + gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag + gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 + gif->madr = ptag[1]; + + gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); + if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag + GIF_LOG("dmaIrq Set"); + gspath3done = 1; } - } + GIFdmaEnd(); + return; + + } else GIFdmaEnd(); gscycles = 0; } } @@ -358,23 +351,33 @@ void dmaGIF() { GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x8000; //FIFO full psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [used to be 0xE00; // OPH=1 | APATH=3] + Path3progress = 2; if ((gif->qwc == 0) && ((gif->chcr & 0xc) != 0)){ u32 *ptag; ptag = (u32*)dmaGetAddr(gif->tadr); gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 - - //gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); - GIFdmaEnd(); - gif->qwc = 0; + gif->madr = ptag[1]; + + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28), gif->madr); + + + gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); + if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag + GIF_LOG("dmaIrq Set"); + gspath3done = 1; + } + + GIFdma(); + //gif->qwc = 0; return; } //Halflife sets a QWC amount in chain mode, no tadr set. if((gif->qwc > 0) && ((gif->chcr & 0x4) == 0x4)) gspath3done = 1; - dmaGIFend(); + GIFdma(); } // called from only one location, so forceinline it: diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index d2e22bdd2f..25560e85b0 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -287,9 +287,11 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u pMem += sizeof(GIFTAG); --size; - if(pathidx == 2 && path.tag.eop) - { - Path3transfer = FALSE; + if(pathidx == 2) + { + if(path.tag.flg != GIF_FLG_IMAGE)Path3progress = 1; //Other mode + else Path3progress = 0; //IMAGE mode + //if(pathidx == 2) GIF_LOG("Set Giftag NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); } if( pathidx == 0 ) @@ -305,11 +307,11 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u return ++size; } } - } - else + }else { // NOTE: size > 0 => do {} while(size > 0); should be faster than while(size > 0) {} - + + //if(pathidx == 2) GIF_LOG("PATH3 NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); switch(path.tag.flg) { case GIF_FLG_PACKED: @@ -371,11 +373,12 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u } } - - if(pathidx == 0) + + if(pathidx == 0 || pathidx == 2) { if(path.tag.eop && path.tag.nloop == 0) { + //if(pathidx == 2) GIF_LOG("BREAK PATH3 NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); break; } } @@ -392,6 +395,23 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u // along the way (often means curreg was in a bad state or something) } } + + + if(pathidx == 2) + { + if(path.tag.nloop == 0 ) + { + //DevCon::Notice("Finishing Giftag NLoop %d EOP %x Mode %d nregs %d Path3progress %d Vifstat VGW %x", + //params path.tag.nloop, path.tag.eop, path.tag.flg, path.tag.nreg, Path3progress, vif1Regs->stat & VIF1_STAT_VGW); + if(path.tag.eop) + { + Path3progress = 2; + //GIF_LOG("Set progress NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); + } + + } + + } #ifdef PCSX2_GSRING_SAMPLING_STATS __asm { @@ -838,6 +858,11 @@ int mtgsThreadObject::PrepDataPacket( GIF_PATH pathidx, const u8* srcdata, u32 s // enough room for size - retval: int retval = _gifTransferDummy( pathidx, srcdata, size ); + if(pathidx == 2) + { + gif->madr += (size - retval) * 16; + gif->qwc -= size - retval; + } size = size - retval; m_packet_size = size; size++; // takes into account our command qword. diff --git a/pcsx2/Patch.cpp b/pcsx2/Patch.cpp index 9262dcf3b0..618e6f5041 100644 --- a/pcsx2/Patch.cpp +++ b/pcsx2/Patch.cpp @@ -57,7 +57,6 @@ PatchTextTable commands[] = // X - EE rounding mode (default is NEAR) // Y - VU rounding mode (default is CHOP) { "zerogs", 6, patchFunc_zerogs }, // zerogs=hex - { "path3hack", 7, patchFunc_path3hack }, { "vunanmode",8, patchFunc_vunanmode }, { "ffxhack",9, patchFunc_ffxhack}, { "xkickdelay",10, patchFunc_xkickdelay}, @@ -674,11 +673,6 @@ void patchFunc_vunanmode( char * cmd, char * param ) // Doesn't do anything anymore } -void patchFunc_path3hack( char * cmd, char * param ) -{ - path3hack = TRUE; -} - void patchFunc_roundmode( char * cmd, char * param ) { int index; diff --git a/pcsx2/Patch.h b/pcsx2/Patch.h index b077ae67fd..cd3e100e44 100644 --- a/pcsx2/Patch.h +++ b/pcsx2/Patch.h @@ -82,7 +82,6 @@ void patchFunc_comment( char * text1, char * text2 ); void patchFunc_gametitle( char * text1, char * text2 ); void patchFunc_patch( char * text1, char * text2 ); void patchFunc_fastmemory( char * text1, char * text2 ); -void patchFunc_path3hack( char * text1, char * text2 ); void patchFunc_roundmode( char * text1, char * text2 ); void patchFunc_zerogs( char * text1, char * text2 ); void patchFunc_vunanmode( char * text1, char * text2 ); @@ -109,7 +108,6 @@ int AddPatch(int Mode, int Place, int Address, int Size, u64 data); extern void SetFastMemory(int); // iR5900LoadStore.c -extern bool path3hack; //extern int g_VUGameFixes; extern int g_ZeroGSOptions; extern u32 g_sseMXCSR; diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index c1de89843c..8adc7586bd 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -58,8 +58,7 @@ static const unsigned int VIF0dmanum = 0; static const unsigned int VIF1dmanum = 1; int g_vifCycles = 0; -bool path3hack = false; -bool Path3transfer = false; +int Path3progress = 2; //0 = Image Mode (DirectHL), 1 = transferring, 2 = Stopped at End of Packet u32 splittransfer[4]; u32 splitptr = 0; @@ -1891,6 +1890,7 @@ static int __fastcall Vif1TransDirectHL(u32 *data) } else { + psHu32(GIF_STAT) &= ~0x80; ret = vif1.tag.size; vif1.tag.size = 0; vif1.cmd = 0; @@ -2019,22 +2019,11 @@ static void Vif1CMDMskPath3() // MSKPATH3 #ifdef GSPATH3FIX if ((vif1Regs->code >> 15) & 0x1) { - - while ((gif->chcr & 0x100)) //Can be done 2 different ways, depends on the game/company - { - if (!path3hack && !Path3transfer && (gif->qwc == 0)) break; - - gsInterrupt(); - - if (path3hack && (gif->qwc == 0)) break; //add games not working with it to elfheader.c to enable this instead - - } psHu32(GIF_STAT) |= 0x2; } else { - // fixme: This is the *only* reason 'transferred' is global. Otherwise it'd be local to Vif1Transfer. - if (gif->chcr & 0x100) CPU_INT(2, (transferred >> 2) * BIAS); // Restart Path3 on its own, time it right! + Path3progress = 1; //Let the Gif know it can transfer again (making sure any vif stall isnt unset prematurely) psHu32(GIF_STAT) &= ~0x2; } #else @@ -2068,11 +2057,10 @@ static void Vif1CMDFlush() // FLUSH/E/A if((vif1.cmd & 0x7f) == 0x13) { - while ((gif->chcr & 0x100)) - { - if (!Path3transfer && gif->qwc == 0) break; - gsInterrupt(); - } + if(Path3progress != 2 && gif->chcr & 0x100) // Gif is already transferring so wait for it. + { + vif1Regs->stat |= VIF1_STAT_VGW; + } } vif1.cmd &= ~0x7f; @@ -2123,13 +2111,16 @@ static void Vif1CMDDirectHL() // DIRECT/HL vif1.tag.size = vifImm << 2; - if((gif->chcr & 0x100) && (vif1.cmd & 0x7f) == 0x51) + if((vif1.cmd & 0x7f) == 0x51) { - //if(vif1Regs->mskpath3)CPU_INT(2, vif1ch->qwc - (vif1.vifpacketsize >> 2) * BIAS); - //DirectHL flushes the lot - vif1Regs->stat |= VIF1_STAT_VGW; + if(gif->chcr & 0x100 && Path3progress == 0) //PATH3 is in image mode, so wait for end of transfer + { + //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); + vif1Regs->stat |= VIF1_STAT_VGW; + } } + psHu32(GIF_STAT) |= 0x80; } static void Vif1CMDNull() // invalid opcode @@ -2215,7 +2206,6 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size != 0) DevCon::Error("no vif1 cmd but tag size is left last cmd read %x", params vif1Regs->code); if (vif1.irq) break; - if(vif1Regs->stat & VIF1_STAT_VGW) break; vif1.cmd = (data[0] >> 24); vif1Regs->code = data[0]; @@ -2459,7 +2449,7 @@ __forceinline void vif1SetupTransfer() else ret = VIF1transfer(vif1ptag + 2, 2, 1); //Transfer Tag - if (ret < 0) + if (ret < 0 && vif1.irqoffset < 2) { vif1.inprogress = 0; //Better clear this so it has to do it again (Jak 1) return; //There has been an error or an interrupt @@ -2485,22 +2475,18 @@ __forceinline void vif1Interrupt() VIF_LOG("vif1Interrupt: %8.8x", cpuRegs.cycle); g_vifCycles = 0; - if((vif1Regs->stat & VIF1_STAT_VGW) && - (gif->chcr & 0x100)) + + if((vif1Regs->stat & VIF1_STAT_VGW)) { - int delay = 0; - - if ((psHu32(GIF_MODE) & 0x1)) - delay = min( 8, (int)gif->qwc ); - else - delay = gif->qwc * BIAS; - - //else CPU_INT(2, min( 64, (int)gif->qwc ) * BIAS); - CPU_INT(1, delay); - return; - } - vif1Regs->stat &= ~VIF1_STAT_VGW; + if(gif->chcr & 0x100) + { + CPU_INT(1, 2); + return; + } + else vif1Regs->stat &= ~VIF1_STAT_VGW; + } + if ((vif1ch->chcr & 0x100) == 0) Console::WriteLn("Vif1 running when CHCR == %x", params vif1ch->chcr); @@ -2539,7 +2525,7 @@ __forceinline void vif1Interrupt() if ((vif1.inprogress & 0x1) == 0) vif1SetupTransfer(); - CPU_INT(1, (vif1ch->qwc * BIAS) - (vif1.vifpacketsize >> 1)); + CPU_INT(1, g_vifCycles); return; } @@ -2606,7 +2592,7 @@ void dmaVIF1() // Chain Mode vif1.done = false; - CPU_INT(1, 0); + vif1Interrupt(); } diff --git a/pcsx2/VifDma.h b/pcsx2/VifDma.h index 21d45e4a4e..f210d49e2c 100644 --- a/pcsx2/VifDma.h +++ b/pcsx2/VifDma.h @@ -47,7 +47,7 @@ struct vifStruct { }; extern vifStruct vif0, vif1; -extern bool Path3transfer; +extern int Path3progress; void __fastcall UNPACK_S_32( u32 *dest, u32 *data, int size ); diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 39b0930833..4dbb9c6e3e 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -4409,7 +4409,7 @@ void recVUMI_XGKICK_(VURegs *VU) CALLFunc((uptr)GSgifTransfer1); #endif } - + psHu32(GIF_STAT) &= ~0x100; s_ScheduleXGKICK = 0; } @@ -4430,6 +4430,8 @@ void recVUMI_XGKICK(VURegs *VU, int info) SHL32ItoR(isreg, 4); AND32ItoR(isreg, 0x3fff); s_XGKICKReg = isreg; + psHu32(GIF_STAT) |= 0x100; + if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) { From f8505c9eaf33c5b5ec9e42d7bfe93ee331c3e711 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 10:42:26 +0000 Subject: [PATCH 124/296] Altered time between GIF resume checks, 2 cycles was a bit ambitious, 16 cycles (8 qw's transferred) is much more realistic and doesnt slow games down as much. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1259 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 9e189856ae..886f3f4531 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -153,7 +153,7 @@ static __forceinline void GIFchain() static __forceinline void dmaGIFend() { - CPU_INT(2, 2); + CPU_INT(2, 16); } // These could probably be consolidated into one function, From 1d4ad79349b72b52fa23ef3cd42bc3f7823e34b9 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 12:16:02 +0000 Subject: [PATCH 125/296] Fixed GIF MFIFO for Front Mission 4 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1260 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 28 ++++++++++++++++------------ 1 file changed, 16 insertions(+), 12 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 886f3f4531..f2e01e47be 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -328,6 +328,7 @@ void GIFdma() GIF_LOG("dmaIrq Set"); gspath3done = 1; } + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, id, gif->madr); GIFdmaEnd(); return; @@ -397,12 +398,15 @@ static __forceinline int mfifoGIFrbTransfer() { /* it does, so first copy 's1' bytes from 'addr' to 'data' */ src = (u32*)PSM(gif->madr); if (src == NULL) return -1; - WRITERING_DMA(src, s1); + s1 = WRITERING_DMA(src, s1); - /* and second copy 's2' bytes from 'maddr' to '&data[s1]' */ - src = (u32*)PSM(psHu32(DMAC_RBOR)); - if (src == NULL) return -1; - WRITERING_DMA(src, (mfifoqwc - s1)); + if(s1 == (((psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR)+16) - gif->madr) >> 4)) + { + /* and second copy 's2' bytes from 'maddr' to '&data[s1]' */ + src = (u32*)PSM(psHu32(DMAC_RBOR)); + if (src == NULL) return -1; + mfifoqwc = WRITERING_DMA(src, (mfifoqwc - s1)) + s1; + } } else @@ -412,13 +416,13 @@ static __forceinline int mfifoGIFrbTransfer() { src = (u32*)PSM(gif->madr); if (src == NULL) return -1; - WRITERING_DMA(src, mfifoqwc); + mfifoqwc = WRITERING_DMA(src, mfifoqwc); gif->madr = psHu32(DMAC_RBOR) + (gif->madr & psHu32(DMAC_RBSR)); } gifqwc -= mfifoqwc; - gif->qwc -= mfifoqwc; - gif->madr += mfifoqwc*16; + //gif->qwc -= mfifoqwc; + //gif->madr += mfifoqwc*16; //mfifocycles += (mfifoqwc) * 2; /* guessing */ return 0; @@ -442,8 +446,8 @@ static __forceinline int mfifoGIFchain() { if (pMem == NULL) return -1; WRITERING_DMA(pMem, mfifoqwc); - gif->madr += mfifoqwc*16; - gif->qwc -= mfifoqwc; + //gif->madr += mfifoqwc*16; + //gif->qwc -= mfifoqwc; mfifocycles += (mfifoqwc) * 2; /* guessing */ } @@ -464,7 +468,7 @@ void mfifoGIFtransfer(int qwc) { gifstate &= ~GIF_STATE_EMPTY; } - SPR_LOG("mfifoGIFtransfer %x madr %x, tadr %x", gif->chcr, gif->madr, gif->tadr); + GIF_LOG("mfifoGIFtransfer %x madr %x, tadr %x", gif->chcr, gif->madr, gif->tadr); if (gif->qwc == 0) { if (gif->tadr == spr0->madr) { @@ -482,7 +486,7 @@ void mfifoGIFtransfer(int qwc) { mfifocycles += 2; gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); - SPR_LOG("dmaChain %8.8x_%8.8x size=%d, id=%d, madr=%lx, tadr=%lx mfifo qwc = %x spr0 madr = %x", + GIF_LOG("dmaChain %8.8x_%8.8x size=%d, id=%d, madr=%lx, tadr=%lx mfifo qwc = %x spr0 madr = %x", ptag[1], ptag[0], gif->qwc, id, gif->madr, gif->tadr, gifqwc, spr0->madr); gifqwc--; From ed963ae5ea18f51ad0a60bbe561d4cd019bec118 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 13:29:48 +0000 Subject: [PATCH 126/296] Temp fix for Soul Calibur 3 so i can close Issue 233. Other small fixes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1261 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 2 +- pcsx2/VifDma.cpp | 11 ++++++----- 2 files changed, 7 insertions(+), 6 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index f2e01e47be..892441607d 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -328,7 +328,7 @@ void GIFdma() GIF_LOG("dmaIrq Set"); gspath3done = 1; } - GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, id, gif->madr); + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28) & 0x7, gif->madr); GIFdmaEnd(); return; diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 8adc7586bd..d1604902ae 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2110,13 +2110,14 @@ static void Vif1CMDDirectHL() // DIRECT/HL else vif1.tag.size = vifImm << 2; - + //FIXME: This should have timing in both cases, see note below. if((vif1.cmd & 0x7f) == 0x51) { - if(gif->chcr & 0x100 && Path3progress == 0) //PATH3 is in image mode, so wait for end of transfer + if(gif->chcr & 0x100 /*&& Path3progress == 0*/) //PATH3 is in image mode, so wait for end of transfer { //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); - vif1Regs->stat |= VIF1_STAT_VGW; + if(vif1Regs->mskpath3)vif1Regs->stat |= VIF1_STAT_VGW; + else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( } } @@ -2480,7 +2481,7 @@ __forceinline void vif1Interrupt() { if(gif->chcr & 0x100) { - CPU_INT(1, 2); + CPU_INT(1, 16); return; } else vif1Regs->stat &= ~VIF1_STAT_VGW; @@ -2542,7 +2543,7 @@ __forceinline void vif1Interrupt() vif1ch->chcr &= ~0x100; g_vifCycles = 0; hwDmacIrq(DMAC_VIF1); - if (vif1Regs->mskpath3 == 0 || (vif1ch->chcr & 0x1) == 0x1)vif1Regs->stat &= ~0x1F000000; // FQC=0 + vif1Regs->stat &= ~0x1F000000; // FQC=0 } void dmaVIF1() From 0c5c45061b62b82702c98c45d120261e64464485 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 14:13:57 +0000 Subject: [PATCH 127/296] Guess im just a jackass, do do do, oh *ahem*... This fixes Jackass :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1262 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IPU/IPU.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/IPU/IPU.cpp b/pcsx2/IPU/IPU.cpp index e8f9fa72a6..9967133f71 100644 --- a/pcsx2/IPU/IPU.cpp +++ b/pcsx2/IPU/IPU.cpp @@ -1386,7 +1386,7 @@ int IPU1dma() assert(!(g_nDMATransfer & IPU_DMA_TIE1)); //We need to make sure GIF has flushed before sending IPU data, it seems to REALLY screw FFX videos - while(gif->chcr & 0x100) + while(gif->chcr & 0x100 && vif1Regs->mskpath3 == 0) { GIF_LOG("Flushing gif chcr %x tadr %x madr %x qwc %x", gif->chcr, gif->tadr, gif->madr, gif->qwc); gsInterrupt(); From a46f1538f0df19a7d34941fdec8dd0c3aa4c4bdb Mon Sep 17 00:00:00 2001 From: sudonim1 Date: Mon, 25 May 2009 14:43:15 +0000 Subject: [PATCH 128/296] =?UTF-8?q?Another=20=C2=B5VU=20typo,=20fixes=20si?= =?UTF-8?q?lent=20hill=202=20but=20probably=20not=20nothing=20else.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1263 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 214f3fe440..f1fc7e589b 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -28,7 +28,7 @@ //------------------------------------------------------------------ #define aReg(x) mVUregs.VF[x] -#define bReg(x, y) mVUregsTemp.VFreg[y] = x; mVUregsTemp.VF[0] +#define bReg(x, y) mVUregsTemp.VFreg[y] = x; mVUregsTemp.VF[y] #define aMax(x, y) ((x > y) ? x : y) #define analyzeReg1(reg) { \ From 0900ef5420ddabbc054440915187bf75b31e6bb6 Mon Sep 17 00:00:00 2001 From: refraction Date: Mon, 25 May 2009 22:09:18 +0000 Subject: [PATCH 129/296] Added VIF wait to MFIFO, fixes Crash of the Titans graphics git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1264 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Vif.cpp | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/pcsx2/Vif.cpp b/pcsx2/Vif.cpp index 0da6e70175..73e874c9f7 100644 --- a/pcsx2/Vif.cpp +++ b/pcsx2/Vif.cpp @@ -533,6 +533,17 @@ void vifMFIFOInterrupt() { g_vifCycles = 0; + if((vif1Regs->stat & VIF1_STAT_VGW)) + { + if(gif->chcr & 0x100) + { + CPU_INT(10, 16); + return; + } + else vif1Regs->stat &= ~VIF1_STAT_VGW; + + } + if((spr0->chcr & 0x100) && spr0->qwc == 0) { spr0->chcr &= ~0x100; From 22535ad886fe1d545aade7af611f497bd6ff297c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 26 May 2009 01:00:38 +0000 Subject: [PATCH 130/296] microVU: - Properly fixed problems from r1254, a lot of games should be fixed. pcsx2/microVU: - Added some debug code that runs sVU and mVU, and compares their results. If they differ, then the game is halted, and debug info is given. (this can be enabled in iVU1micro.cpp with the DEBUG_COMPARE2 macro)... the code is a bit messy, but it gets the job done for now. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1265 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU1micro.cpp | 316 +++++++++++++++++++++++----------- pcsx2/x86/microVU.h | 3 + pcsx2/x86/microVU_Alloc.inl | 2 +- pcsx2/x86/microVU_Compile.inl | 6 +- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 6 +- pcsx2/x86/microVU_Misc.h | 9 + 7 files changed, 237 insertions(+), 107 deletions(-) diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 5eba1f8782..a54bdd6bed 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -30,7 +30,11 @@ extern u32 vudump; #endif -//#define DEBUG_COMPARE +int mVUdebugNow = 0; + +//#define DEBUG_COMPARE // Run sVU or mVU and print results +//#define DEBUG_COMPARE2 // Runs both VU recs and breaks when results differ + #ifdef DEBUG_COMPARE #include @@ -39,7 +43,7 @@ static int runAmount = 0; void VUtestPause() { runAmount++; - if (runAmount < 100) return; + if (runAmount < 654) return; #ifndef PCSX2_MICROVU_ SysPrintf("Super VU - Pass %d\n", runAmount); @@ -48,13 +52,27 @@ void VUtestPause() { #endif for (int i = 0; i < 32; i++) { - SysPrintf("VF%02d = {%f, %f, %f, %f}\n", i, VU1.VF[i].F[0], VU1.VF[i].F[1], VU1.VF[i].F[2], VU1.VF[i].F[3]); + SysPrintf("VF%02d = {%f, %f, %f, %f}\n", i, VU1.VF[i].F[0], VU1.VF[i].F[1], VU1.VF[i].F[2], VU1.VF[i].F[3]); } + SysPrintf("ACC = {%f, %f, %f, %f}\n", VU1.ACC.F[0], VU1.ACC.F[1], VU1.ACC.F[2], VU1.ACC.F[3]); + for (int i = 0; i < 16; i++) { - SysPrintf("VI%02d = % 8d ($%08x)\n", i, (s16)VU1.VI[i].UL, (s16)VU1.VI[i].UL); + SysPrintf("VI%02d = % 8d ($%08x)\n", i, (s16)VU1.VI[i].UL, (s16)VU1.VI[i].UL); } + SysPrintf("Stat = % 8d ($%08x)\n", (s16)VU1.VI[REG_STATUS_FLAG].UL, (s16)VU1.VI[REG_STATUS_FLAG].UL); + SysPrintf("MAC = % 8d ($%08x)\n", (s16)VU1.VI[REG_MAC_FLAG].UL, (s16)VU1.VI[REG_MAC_FLAG].UL); + SysPrintf("CLIP = % 8d ($%08x)\n", (s16)VU1.VI[REG_CLIP_FLAG].UL, (s16)VU1.VI[REG_CLIP_FLAG].UL); + + SysPrintf("Q-reg = %f ($%08x)\n", VU1.VI[REG_Q].F, (s32)VU1.VI[REG_Q].UL); + SysPrintf("P-reg = %f ($%08x)\n", VU1.VI[REG_P].F, (s32)VU1.VI[REG_P].UL); + SysPrintf("I-reg = %f ($%08x)\n", VU1.VI[REG_I].F, (s32)VU1.VI[REG_I].UL); + + SysPrintf("_Stat = % 8d ($%08x)\n", (s16)VU1.statusflag, (s16)VU1.statusflag); + SysPrintf("_MAC = % 8d ($%08x)\n", (s16)VU1.macflag, (s16)VU1.macflag); + SysPrintf("_CLIP = % 8d ($%08x)\n", (s16)VU1.clipflag, (s16)VU1.clipflag); + u32 j = 0; for (int i = 0; i < (0x4000 / 4); i++) { j ^= ((u32*)(VU1.Mem))[i]; @@ -70,112 +88,178 @@ void VUtestPause() { void VUtestPause() {} #endif -#ifndef PCSX2_MICROVU_ -namespace VU1micro -{ - void recAlloc() - { - SuperVUAlloc(1); - } +#ifdef DEBUG_COMPARE2 - void __fastcall recClear( u32 Addr, u32 Size ) - { - assert( (Addr&7) == 0 ); - SuperVUClear(Addr, Size, 1); // Size should be a multiple of 8 bytes! - } - - void recShutdown() - { - SuperVUDestroy( 1 ); - } - - // commented out because I'm not sure it actually works anymore with SuperVU (air) - /*static void iVU1DumpBlock() - { - FILE *f; - char filename[ g_MaxPath ]; - u32 *mem; - u32 i; - - #ifdef _WIN32 - CreateDirectory("dumps", NULL); - sprintf_s( filename, g_MaxPath, "dumps\\vu%.4X.txt", VU1.VI[ REG_TPC ].UL ); - #else - mkdir("dumps", 0755); - sprintf( filename, "dumps/vu%.4X.txt", VU1.VI[ REG_TPC ].UL ); - #endif - Console::WriteLn( "dump1 %x => %x (%s)", params VU1.VI[ REG_TPC ].UL, pc, filename ); - - f = fopen( filename, "wb" ); - for ( i = VU1.VI[REG_TPC].UL; i < pc; i += 8 ) { - char* pstr; - mem = (u32*)&VU1.Micro[i]; - - pstr = disVU1MicroUF( mem[1], i+4 ); - fprintf(f, "%x: %-40s ", i, pstr); - - pstr = disVU1MicroLF( mem[0], i ); - fprintf(f, "%s\n", pstr); - } - fclose( f ); - }*/ - - static void recReset() - { - SuperVUReset(1); - - // these shouldn't be needed, but shouldn't hurt anything either. - x86FpuState = FPU_STATE; - } - - static void recStep() - { - } - - static void recExecuteBlock(void) - { - #ifdef _DEBUG - static u32 vuprogcount = 0; - vuprogcount++; - if( vudump & 8 ) __Log("start vu1: %x %x", VU1.VI[ REG_TPC ].UL, vuprogcount); - #endif - - if((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0){ - //Console::WriteLn("Execute block VU1, VU1 not busy"); - return; - } - - if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) - { - Console::Error("VU1 memory overflow!!: %x", params VU1.VI[REG_TPC].UL); - /*VU0.VI[REG_VPU_STAT].UL&= ~0x100; - VU1.cycle++; - return;*/ - } - - assert( (VU1.VI[ REG_TPC ].UL&7) == 0 ); -#ifdef DEBUG_COMPARE - SysPrintf("StartPC = 0x%04x\n", VU1.VI[REG_TPC].UL); +#ifndef DEBUG_COMPARE +#include #endif - - FreezeXMMRegs(1); - do { // while loop needed since not always will return finished - SuperVUExecuteProgram(VU1.VI[ REG_TPC ].UL & 0x3fff, 1); - } while( VU0.VI[ REG_VPU_STAT ].UL&0x100 ); - FreezeXMMRegs(0); - - VUtestPause(); - } -} -#else - extern void initVUrec(VURegs* vuRegs, const int vuIndex); extern void closeVUrec(const int vuIndex); extern void resetVUrec(const int vuIndex); extern void clearVUrec(u32 addr, u32 size, const int vuIndex); extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); +PCSX2_ALIGNED16(u8 backVUregs[sizeof(VURegs)]); +PCSX2_ALIGNED16(u8 cmpVUregs [sizeof(VURegs)]); +PCSX2_ALIGNED16(u8 backVUmem [0x4000]); +PCSX2_ALIGNED16(u8 cmpVUmem [0x4000]); +static u32 runCount = 0; +#define VU3 ((VURegs)*((VURegs*)cmpVUregs)) +#define cmpA Console::Error +#define cmpB Console::WriteLn +#define cmpPrint(cond) { \ + if (cond) { \ + cmpA("%s", params str1); \ + cmpA("%s", params str2); \ + } \ + else { \ + cmpB("%s", params str1); \ + cmpB("%s", params str2); \ + } \ +} + +namespace VU1micro +{ + void recAlloc() { SuperVUAlloc(1); initVUrec(&VU1, 1); } + void __fastcall recClear(u32 Addr, u32 Size) { SuperVUClear(Addr, Size, 1); clearVUrec(Addr, Size, 1); } + void recShutdown() { SuperVUDestroy(1); closeVUrec(1); } + static void recReset() { SuperVUReset(1); resetVUrec(1); x86FpuState = FPU_STATE; } + static void recStep() {} + + static void recExecuteBlock(void) + { + if((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; + assert((VU1.VI[ REG_TPC ].UL&7) == 0); + if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { Console::Error("VU1 memory overflow!!: %x", params VU1.VI[REG_TPC].UL); } + +#ifdef DEBUG_COMPARE + SysPrintf("(%08d) StartPC = 0x%04x\n", runAmount, VU1.VI[REG_TPC].UL); +#endif + + FreezeXMMRegs(1); + + runCount++; + memcpy_fast((u8*)backVUregs, (u8*)&VU1, sizeof(VURegs)); + memcpy_fast((u8*)backVUmem, (u8*)VU1.Mem, 0x4000); + + do { // while loop needed since not always will return finished + SuperVUExecuteProgram(VU1.VI[ REG_TPC ].UL & 0x3fff, 1); + } while( VU0.VI[ REG_VPU_STAT ].UL&0x100 ); + + memcpy_fast((u8*)cmpVUregs, (u8*)&VU1, sizeof(VURegs)); + memcpy_fast((u8*)cmpVUmem, (u8*)VU1.Mem, 0x4000); + memcpy_fast((u8*)&VU1, (u8*)backVUregs, sizeof(VURegs)); + memcpy_fast((u8*)VU1.Mem, (u8*)backVUmem, 0x4000); + + runVUrec(VU1.VI[REG_TPC].UL, 300000 /*0x7fffffff*/, 1); + + if ((memcmp((u8*)cmpVUregs, (u8*)&VU1, (16*32) + (16*16))) || (memcmp((u8*)cmpVUmem, (u8*)VU1.Mem, 0x4000))) { + char str1[150]; + char str2[150]; + SysPrintf("\n\n"); + SysPrintf("-----------------------------------------------\n"); + Console::Notice("Problem Occurred!"); + SysPrintf("-----------------------------------------------\n"); + SysPrintf("runCount = %d\n", runCount); + SysPrintf("StartPC [%04x]\n", ((VURegs*)backVUregs)->VI[REG_TPC].UL); + SysPrintf("-----------------------------------------------\n\n"); + + SysPrintf("-----------------------------------------------\n"); + Console::Notice("Super VU / Micro VU"); + SysPrintf("-----------------------------------------------\n"); + + for (int i = 0; i < 32; i++) { + sprintf(str1, "VF%02d = {%f, %f, %f, %f}", i, VU3.VF[i].F[0], VU3.VF[i].F[1], VU3.VF[i].F[2], VU3.VF[i].F[3]); + sprintf(str2, "VF%02d = {%f, %f, %f, %f}", i, VU1.VF[i].F[0], VU1.VF[i].F[1], VU1.VF[i].F[2], VU1.VF[i].F[3]); + cmpPrint(((VU1.VF[i].UL[0] != VU3.VF[i].UL[0]) || (VU1.VF[i].UL[1] != VU3.VF[i].UL[1]) || (VU1.VF[i].UL[2] != VU3.VF[i].UL[2]) || (VU1.VF[i].UL[3] != VU3.VF[i].UL[3]))); + } + + sprintf(str1, "ACC = {%f, %f, %f, %f}", VU3.ACC.F[0], VU3.ACC.F[1], VU3.ACC.F[2], VU3.ACC.F[3]); + sprintf(str2, "ACC = {%f, %f, %f, %f}", VU1.ACC.F[0], VU1.ACC.F[1], VU1.ACC.F[2], VU1.ACC.F[3]); + cmpPrint(((VU1.ACC.F[0] != VU3.ACC.F[0]) || (VU1.ACC.F[1] != VU3.ACC.F[1]) || (VU1.ACC.F[2] != VU3.ACC.F[2]) || (VU1.ACC.F[3] != VU3.ACC.F[3]))); + + for (int i = 0; i < 16; i++) { + sprintf(str1, "VI%02d = % 8d ($%08x)", i, (s16)VU3.VI[i].UL, VU3.VI[i].UL); + sprintf(str2, "VI%02d = % 8d ($%08x)", i, (s16)VU1.VI[i].UL, VU1.VI[i].UL); + cmpPrint((VU1.VI[i].UL != VU3.VI[i].UL)); + } + + sprintf(str1, "Stat = % 8d ($%08x)", (s16)VU3.VI[REG_STATUS_FLAG].UL, VU3.VI[REG_STATUS_FLAG].UL); + sprintf(str2, "Stat = % 8d ($%08x)", (s16)VU1.VI[REG_STATUS_FLAG].UL, VU1.VI[REG_STATUS_FLAG].UL); + cmpPrint((VU1.VI[REG_STATUS_FLAG].UL != VU3.VI[REG_STATUS_FLAG].UL)); + + sprintf(str1, "MAC = % 8d ($%08x)", (s16)VU3.VI[REG_MAC_FLAG].UL, VU3.VI[REG_MAC_FLAG].UL); + sprintf(str2, "MAC = % 8d ($%08x)", (s16)VU1.VI[REG_MAC_FLAG].UL, VU1.VI[REG_MAC_FLAG].UL); + cmpPrint((VU1.VI[REG_MAC_FLAG].UL != VU3.VI[REG_MAC_FLAG].UL)); + + sprintf(str1, "CLIP = % 8d ($%08x)", (s16)VU3.VI[REG_CLIP_FLAG].UL, VU3.VI[REG_CLIP_FLAG].UL); + sprintf(str2, "CLIP = % 8d ($%08x)", (s16)VU1.VI[REG_CLIP_FLAG].UL, VU1.VI[REG_CLIP_FLAG].UL); + cmpPrint((VU1.VI[REG_CLIP_FLAG].UL != VU3.VI[REG_CLIP_FLAG].UL)); + + sprintf(str1, "Q-reg = %f ($%08x)", VU3.VI[REG_Q].F, VU3.VI[REG_Q].UL); + sprintf(str2, "Q-reg = %f ($%08x)", VU1.VI[REG_Q].F, VU1.VI[REG_Q].UL); + cmpPrint((VU1.VI[REG_Q].UL != VU3.VI[REG_Q].UL)); + + sprintf(str1, "P-reg = %f ($%08x)", VU3.VI[REG_P].F, VU3.VI[REG_P].UL); + sprintf(str2, "P-reg = %f ($%08x)", VU1.VI[REG_P].F, VU1.VI[REG_P].UL); + cmpPrint((VU1.VI[REG_P].UL != VU3.VI[REG_P].UL)); + + sprintf(str1, "I-reg = %f ($%08x)", VU3.VI[REG_I].F, VU3.VI[REG_I].UL); + sprintf(str2, "I-reg = %f ($%08x)", VU1.VI[REG_I].F, VU1.VI[REG_I].UL); + cmpPrint((VU1.VI[REG_I].UL != VU3.VI[REG_I].UL)); + + sprintf(str1, "_Stat = % 8d ($%08x)", (s16)VU3.statusflag, VU3.statusflag); + sprintf(str2, "_Stat = % 8d ($%08x)", (s16)VU1.statusflag, VU1.statusflag); + cmpPrint((VU1.statusflag != VU3.statusflag)); + + sprintf(str1, "_MAC = % 8d ($%08x)", (s16)VU3.macflag, VU3.macflag); + sprintf(str2, "_MAC = % 8d ($%08x)", (s16)VU1.macflag, VU1.macflag); + cmpPrint((VU1.macflag != VU3.macflag)); + + sprintf(str1, "_CLIP = % 8d ($%08x)", (s16)VU3.clipflag, VU3.clipflag); + sprintf(str2, "_CLIP = % 8d ($%08x)", (s16)VU1.clipflag, VU1.clipflag); + cmpPrint((VU1.clipflag != VU3.clipflag)); + + u32 j = 0; + u32 z = 0; + for (int i = 0; i < (0x4000 / 4); i++) { + j ^= ((u32*)(cmpVUmem))[i]; + z ^= ((u32*)(VU1.Mem)) [i]; + } + sprintf(str1, "VU Mem CRC = 0x%08x", j); + sprintf(str2, "VU Mem CRC = 0x%08x", z); + cmpPrint((j != z)); + + sprintf(str1, "EndPC = 0x%04x", VU3.VI[REG_TPC].UL); + sprintf(str2, "EndPC = 0x%04x", VU1.VI[REG_TPC].UL); + cmpPrint((VU1.VI[REG_TPC].UL != VU3.VI[REG_TPC].UL)); + + SysPrintf("-----------------------------------------------\n\n"); + + mVUdebugNow = 1; + resetVUrec(1); + memcpy_fast((u8*)&VU1, (u8*)backVUregs, sizeof(VURegs)); + memcpy_fast((u8*)VU1.Mem, (u8*)backVUmem, 0x4000); + + runVUrec(VU1.VI[REG_TPC].UL, 300000 /*0x7fffffff*/, 1); + + for (int i = 0; i < 10000000; i++) { + Sleep(1000); + } + } + VUtestPause(); + FreezeXMMRegs(0); + } +} +#else +#ifdef PCSX2_MICROVU_ + +extern void initVUrec(VURegs* vuRegs, const int vuIndex); +extern void closeVUrec(const int vuIndex); +extern void resetVUrec(const int vuIndex); +extern void clearVUrec(u32 addr, u32 size, const int vuIndex); +extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); namespace VU1micro { @@ -190,7 +274,7 @@ namespace VU1micro assert( (VU1.VI[REG_TPC].UL&7) == 0 ); #ifdef DEBUG_COMPARE - SysPrintf("StartPC = 0x%04x\n", VU1.VI[REG_TPC].UL); + SysPrintf("(%08d) StartPC = 0x%04x\n", runAmount, VU1.VI[REG_TPC].UL); #endif FreezeXMMRegs(1); @@ -202,6 +286,34 @@ namespace VU1micro VUtestPause(); } } +#else + +namespace VU1micro +{ + void recAlloc() { SuperVUAlloc(1); } + void __fastcall recClear(u32 Addr, u32 Size) { SuperVUClear(Addr, Size, 1); } + void recShutdown() { SuperVUDestroy(1); } + static void recReset() { SuperVUReset(1); x86FpuState = FPU_STATE; } + static void recStep() {} + + static void recExecuteBlock(void) + { + if((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; + if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { Console::Error("VU1 memory overflow!!: %x", params VU1.VI[REG_TPC].UL); } + assert((VU1.VI[ REG_TPC ].UL&7) == 0); + +#ifdef DEBUG_COMPARE + SysPrintf("(%08d) StartPC = 0x%04x\n", runAmount, VU1.VI[REG_TPC].UL); +#endif + + FreezeXMMRegs(1); + do { // while loop needed since not always will return finished + SuperVUExecuteProgram(VU1.VI[ REG_TPC ].UL & 0x3fff, 1); + } while( VU0.VI[ REG_VPU_STAT ].UL&0x100 ); + FreezeXMMRegs(0); + } +} +#endif #endif using namespace VU1micro; diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index de834dd70a..c4c0dcdda6 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -132,6 +132,9 @@ extern PCSX2_ALIGNED16(microVU microVU1); extern void (*mVU_UPPER_OPCODE[64])( VURegs* VU, s32 info ); extern void (*mVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); +// Debug Helper +extern int mVUdebugNow; + // Main Functions microVUt(void) mVUinit(VURegs*); microVUx(void) mVUreset(); diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index e54a0b2046..c0d9216f53 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -731,7 +731,7 @@ microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { } if (_reg_ == 0) { return; } else if (isMMX(_reg_)) { MOVD32RtoMMX(mmVI(_reg_), GPRreg); } - else if (_reg_ < 16) { MOV32RtoRm(gprR, GPRreg, (_reg_ - 9) * 16); } + else if (_reg_ < 16) { MOV16RtoRm(gprR, GPRreg, (_reg_ - 9) * 16); } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index af92930a10..e8f1aaf0f8 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -184,10 +184,13 @@ microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fC void __fastcall mVUwarning0(u32 PC) { Console::Error("microVU0 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } +void __fastcall mVUprintPC1(u32 PC) { Console::WriteLn("Block startPC [%04x]", params PC); } +void __fastcall mVUprintPC2(u32 PC) { Console::WriteLn("Block endPC [%04x]\n", params PC); } microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; iPC = mVUstartPC; + mVUdebugNOW(0); CMP32ItoM((uptr)&mVU->cycles, 0); u8* jmp8 = JG8(0); MOV32ItoR(gprT2, xPC); @@ -209,7 +212,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (startPC > ((vuIndex) ? 0x3fff : 0xfff)) { Console::Error("microVU%d: invalid startPC", params vuIndex); } startPC &= (vuIndex ? 0x3ff8 : 0xff8); - + if (mVUblocks[startPC/8] == NULL) { mVUblocks[startPC/8] = new microBlockManager(); } @@ -280,6 +283,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { microBlock* bBlock = NULL; u32* ajmp = 0; mVUsetupRange(xPC); + mVUdebugNOW(1); switch (mVUbranch) { case 3: branchCase(JE32, JNE32); // IBEQ diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 2e3e13e6ea..5f29d43c00 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -203,7 +203,7 @@ microVUt(void) mVUflagPass(int startPC) { if (mVUbranch) { branch = 3; mVUbranch = 0; } incPC(1); } - //if (mVUcount < 4) { mVUflagInfo |= 0xfff; } // Is this Too Slow? 99% of games probably don't need this. + if (mVUcount < 4) { mVUflagInfo |= 0xfff; } iPC = oldPC; mVUcount = oldCount; mVUbranch = oldBranch; diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 3362827ce5..ad619cfb87 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1131,8 +1131,10 @@ microVUf(void) mVU_XGKICK() { //------------------------------------------------------------------ #define setBranchA(x, _x_) { \ - mVUbranch = x; \ - pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; mVUbranch = 0; return; } } \ + pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; return; } mVUbranch = x; } \ + pass2 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ + pass3 { mVUbranch = x; } \ + pass4 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ } microVUf(void) mVU_B() { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index bcd5ae2afa..f5314d5910 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -303,3 +303,12 @@ declareAllVariables mVUreset(); \ } \ } + +#define mVUdebugNOW(isEndPC) { \ + if (mVUdebugNow) { \ + MOV32ItoR(gprT2, xPC); \ + if (isEndPC) { CALLFunc((uptr)mVUprintPC2); } \ + else { CALLFunc((uptr)mVUprintPC1); } \ + MOV32ItoR(gprR, Roffset); \ + } \ +} From 6c58f3039a7b93c99cfc77f169fa12fed0352e9c Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 26 May 2009 03:40:31 +0000 Subject: [PATCH 131/296] GSdx: bit more work on the vertex buffer, and broken ffxii fmv fixed again. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1266 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDevice10.cpp | 97 +++++++++---- plugins/GSdx/GSDevice10.h | 11 +- plugins/GSdx/GSDevice9.cpp | 133 ++++++++++------- plugins/GSdx/GSDevice9.h | 10 +- plugins/GSdx/GSRendererHW.h | 12 +- plugins/GSdx/GSRendererHW10.cpp | 8 +- plugins/GSdx/GSRendererHW9.cpp | 35 +---- plugins/GSdx/GSRendererSW.cpp | 168 ++++++++++++++++++++++ plugins/GSdx/GSRendererSW.h | 168 +--------------------- plugins/GSdx/GSSetupPrimCodeGenerator.cpp | 2 - plugins/GSdx/GSState.cpp | 2 +- plugins/GSdx/GSTextureFX10.cpp | 63 +------- plugins/GSdx/GSTextureFX9.cpp | 57 +------- plugins/GSdx/GSTextureFX9.h | 6 - plugins/GSdx/GSVector.h | 39 ++++- 15 files changed, 398 insertions(+), 413 deletions(-) diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index f492eef496..bf22deb23d 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -44,6 +44,11 @@ GSDevice10::GSDevice10() { memset(m_ps_srv, 0, sizeof(m_ps_srv)); memset(m_ps_ss, 0, sizeof(m_ps_ss)); + + m_vertices.stride = 0; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = 0; } GSDevice10::~GSDevice10() @@ -115,16 +120,6 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) hr = CompileShader(IDR_CONVERT10_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } - memset(&bd, 0, sizeof(bd)); - - bd.Usage = D3D10_USAGE_DEFAULT; - bd.BindFlags = D3D10_BIND_VERTEX_BUFFER; - bd.CPUAccessFlags = 0; - bd.MiscFlags = 0; - bd.ByteWidth = 4 * sizeof(GSVertexPT1); - - hr = m_dev->CreateBuffer(&bd, NULL, &m_convert.vb); - memset(&dsd, 0, sizeof(dsd)); dsd.DepthEnable = false; @@ -258,6 +253,11 @@ void GSDevice10::BeginScene() { } +void GSDevice10::DrawPrimitive() +{ + m_dev->Draw(m_vertices.count, m_vertices.start); +} + void GSDevice10::EndScene() { PSSetShaderResources(NULL, NULL); @@ -265,6 +265,9 @@ void GSDevice10::EndScene() // not clearing the rt/ds gives a little fps boost in complex games (5-10%) // OMSetRenderTargets(NULL, NULL); + + m_vertices.start += m_vertices.count; + m_vertices.count = 0; } void GSDevice10::ClearRenderTarget(GSTexture* t, const GSVector4& c) @@ -439,11 +442,7 @@ void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, }; - D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; - - m_dev->UpdateSubresource(m_convert.vb, 0, &box, vertices, 0, 0); - - IASetVertexBuffer(m_convert.vb, sizeof(vertices[0])); + IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); IASetInputLayout(m_convert.il); IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); @@ -467,7 +466,7 @@ void GSDevice10::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, // - DrawPrimitive(countof(vertices)); + DrawPrimitive(); // @@ -508,6 +507,61 @@ void GSDevice10::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool line StretchRect(st, sr, dt, dr, m_interlace.ps[shader], m_interlace.cb, linear); } +void GSDevice10::IASetVertexBuffer(const void* vertices, size_t stride, size_t count) +{ + ASSERT(m_vertices.count == 0); + + if(count > m_vertices.limit) + { + m_vertices.vb_old = m_vertices.vb; + m_vertices.vb = NULL; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = max(count * 3 / 2, 10000); + } + + if(m_vertices.vb == NULL) + { + D3D10_BUFFER_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + bd.Usage = D3D10_USAGE_DYNAMIC; + bd.ByteWidth = m_vertices.limit * stride; + bd.BindFlags = D3D10_BIND_VERTEX_BUFFER; + bd.CPUAccessFlags = D3D10_CPU_ACCESS_WRITE; + + HRESULT hr; + + hr = m_dev->CreateBuffer(&bd, NULL, &m_vertices.vb); + + if(FAILED(hr)) return; + } + + D3D10_MAP type = D3D10_MAP_WRITE_NO_OVERWRITE; + + if(m_vertices.start + count > m_vertices.limit || stride != m_vertices.stride) + { + m_vertices.start = 0; + + type = D3D10_MAP_WRITE_DISCARD; + } + + void* v = NULL; + + if(SUCCEEDED(m_vertices.vb->Map(type, 0, &v))) + { + GSVector4i::storent((uint8*)v + m_vertices.start * stride, vertices, count * stride); + + m_vertices.vb->Unmap(); + } + + m_vertices.count = count; + m_vertices.stride = stride; + + IASetVertexBuffer(m_vertices.vb, stride); +} + void GSDevice10::IASetVertexBuffer(ID3D10Buffer* vb, size_t stride) { if(m_vb != vb || m_vb_stride != stride) @@ -688,11 +742,6 @@ void GSDevice10::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) } } -void GSDevice10::DrawPrimitive(uint32 count, uint32 start) -{ - m_dev->Draw(count, start); -} - HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) { HRESULT hr; @@ -703,7 +752,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M if(error) { - TRACE(_T("%s\n"), CString((LPCSTR)error->GetBufferPointer())); + printf("%s\n", (const char*)error->GetBufferPointer()); } if(FAILED(hr)) @@ -738,7 +787,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M if(error) { - TRACE(_T("%s\n"), CString((LPCSTR)error->GetBufferPointer())); + printf("%s\n", (const char*)error->GetBufferPointer()); } if(FAILED(hr)) @@ -766,7 +815,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M if(error) { - TRACE(_T("%s\n"), CString((LPCSTR)error->GetBufferPointer())); + printf("%s\n", (const char*)error->GetBufferPointer()); } if(FAILED(hr)) diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index cd5266acc7..38793b42b6 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -58,14 +58,18 @@ private: CComPtr m_dev; CComPtr m_swapchain; - CComPtr m_font; + + struct + { + CComPtr vb, vb_old; + size_t stride, start, count, limit; + } m_vertices; public: // TODO CComPtr m_rs; struct { - CComPtr vb; CComPtr il; CComPtr vs; CComPtr ps[7]; @@ -98,6 +102,7 @@ public: void Flip(); void BeginScene(); + void DrawPrimitive(); void EndScene(); void ClearRenderTarget(GSTexture* t, const GSVector4& c); @@ -116,6 +121,7 @@ public: void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D10PixelShader* ps, ID3D10Buffer* ps_cb, ID3D10BlendState* bs, bool linear = true); + void IASetVertexBuffer(const void* vertices, size_t stride, size_t count); void IASetVertexBuffer(ID3D10Buffer* vb, size_t stride); void IASetInputLayout(ID3D10InputLayout* layout); void IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY topology); @@ -128,7 +134,6 @@ public: void OMSetDepthStencilState(ID3D10DepthStencilState* dss, uint8 sref); void OMSetBlendState(ID3D10BlendState* bs, float bf); void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); - void DrawPrimitive(uint32 count, uint32 start = 0); ID3D10Device* operator->() {return m_dev;} operator ID3D10Device*() {return m_dev;} diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 0da5457c6d..e84029c303 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -48,6 +48,11 @@ GSDevice9::GSDevice9() memset(&m_ddcaps, 0, sizeof(m_ddcaps)); memset(&m_d3dcaps, 0, sizeof(m_d3dcaps)); memset(m_ps_srvs, 0, sizeof(m_ps_srvs)); + + m_vertices.stride = 0; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = 0; } GSDevice9::~GSDevice9() @@ -143,8 +148,6 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) CompileShader(IDR_CONVERT9_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } - m_dev->CreateVertexBuffer(4 * sizeof(GSVertexPT1), D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_convert.vb, NULL); - m_convert.dss.DepthEnable = false; m_convert.dss.StencilEnable = false; @@ -364,9 +367,39 @@ void GSDevice9::BeginScene() m_dev->BeginScene(); } +void GSDevice9::DrawPrimitive() +{ + int prims = 0; + + switch(m_topology) + { + case D3DPT_TRIANGLELIST: + prims = m_vertices.count / 3; + break; + case D3DPT_LINELIST: + prims = m_vertices.count / 2; + break; + case D3DPT_POINTLIST: + prims = m_vertices.count; + break; + case D3DPT_TRIANGLESTRIP: + case D3DPT_TRIANGLEFAN: + prims = m_vertices.count - 2; + break; + case D3DPT_LINESTRIP: + prims = m_vertices.count - 1; + break; + } + + m_dev->DrawPrimitive(m_topology, m_vertices.start, prims); +} + void GSDevice9::EndScene() { m_dev->EndScene(); + + m_vertices.start += m_vertices.count; + m_vertices.count = 0; } void GSDevice9::ClearRenderTarget(GSTexture* t, const GSVector4& c) @@ -550,16 +583,7 @@ void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, c vertices[i].p.y += 1.0f / ds.y; } - void* buff = NULL; - - if(SUCCEEDED(m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) - { - memcpy(buff, vertices, sizeof(vertices)); - - m_convert.vb->Unlock(); - } - - IASetVertexBuffer(m_convert.vb, sizeof(vertices[0])); + IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); IASetInputLayout(m_convert.il); IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); @@ -579,7 +603,7 @@ void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, c // - DrawPrimitive(countof(vertices)); + DrawPrimitive(); // @@ -620,6 +644,52 @@ void GSDevice9::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linea StretchRect(st, sr, dt, dr, m_interlace.ps[shader], (const float*)&cb, 1, linear); } +void GSDevice9::IASetVertexBuffer(const void* vertices, size_t stride, size_t count) +{ + ASSERT(m_vertices.count == 0); + + if(count > m_vertices.limit) + { + m_vertices.vb_old = m_vertices.vb; + m_vertices.vb = NULL; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = max(count * 3 / 2, 10000); + } + + if(m_vertices.vb == NULL) + { + HRESULT hr; + + hr = m_dev->CreateVertexBuffer(m_vertices.limit * stride, D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_vertices.vb, NULL); + + if(FAILED(hr)) return; + } + + uint32 flags = D3DLOCK_NOOVERWRITE; + + if(m_vertices.start + count > m_vertices.limit || stride != m_vertices.stride) + { + m_vertices.start = 0; + + flags = D3DLOCK_DISCARD; + } + + void* v = NULL; + + if(SUCCEEDED(m_vertices.vb->Lock(m_vertices.start * stride, count * stride, &v, flags))) + { + GSVector4i::storent(v, vertices, count * stride); + + m_vertices.vb->Unlock(); + } + + m_vertices.count = count; + m_vertices.stride = stride; + + IASetVertexBuffer(m_vertices.vb, stride); +} + void GSDevice9::IASetVertexBuffer(IDirect3DVertexBuffer9* vb, size_t stride) { if(m_vb != vb || m_vb_stride != stride) @@ -846,33 +916,6 @@ void GSDevice9::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) } } -void GSDevice9::DrawPrimitive(uint32 count, uint32 start) -{ - int prims = 0; - - switch(m_topology) - { - case D3DPT_TRIANGLELIST: - prims = count / 3; - break; - case D3DPT_LINELIST: - prims = count / 2; - break; - case D3DPT_POINTLIST: - prims = count; - break; - case D3DPT_TRIANGLESTRIP: - case D3DPT_TRIANGLEFAN: - prims = count - 2; - break; - case D3DPT_LINESTRIP: - prims = count - 1; - break; - } - - m_dev->DrawPrimitive(m_topology, start, prims); -} - // FIXME: D3DXCompileShaderFromResource of d3dx9 v37 (march 2008) calls GetFullPathName on id for some reason and then crashes static HRESULT LoadShader(uint32 id, LPCSTR& data, uint32& size) @@ -934,9 +977,7 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO } else if(error) { - LPCSTR msg = (LPCSTR)error->GetBufferPointer(); - - TRACE(_T("%s\n"), CString(msg)); + printf("%s\n", (const char*)error->GetBufferPointer()); } ASSERT(SUCCEEDED(hr)); @@ -993,14 +1034,10 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO if(SUCCEEDED(hr)) { hr = m_dev->CreatePixelShader((DWORD*)shader->GetBufferPointer(), ps); - - ASSERT(SUCCEEDED(hr)); } else if(error) { - LPCSTR msg = (LPCSTR)error->GetBufferPointer(); - - TRACE(_T("%s\n"), CString(msg)); + printf("%s\n", (const char*)error->GetBufferPointer()); } ASSERT(SUCCEEDED(hr)); diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index dc13b3d894..a9a1123341 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -96,13 +96,18 @@ private: CComPtr m_dev; CComPtr m_swapchain; + struct + { + CComPtr vb, vb_old; + size_t stride, start, count, limit; + } m_vertices; + public: // TODO D3DPRESENT_PARAMETERS m_pp; CComPtr m_font; struct { - CComPtr vb; CComPtr il; CComPtr vs; CComPtr ps[7]; @@ -134,6 +139,7 @@ public: void Flip(); void BeginScene(); + void DrawPrimitive(); void EndScene(); void ClearRenderTarget(GSTexture* t, const GSVector4& c); @@ -154,6 +160,7 @@ public: void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, bool linear = true); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len, Direct3DBlendState9* bs, bool linear = true); + void IASetVertexBuffer(const void* vertices, size_t stride, size_t count); void IASetVertexBuffer(IDirect3DVertexBuffer9* vb, size_t stride); void IASetInputLayout(IDirect3DVertexDeclaration9* layout); void IASetPrimitiveTopology(D3DPRIMITIVETYPE topology); @@ -165,7 +172,6 @@ public: void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref); void OMSetBlendState(Direct3DBlendState9* bs, uint32 bf); void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); - void DrawPrimitive(uint32 count, uint32 start = 0); IDirect3DDevice9* operator->() {return m_dev;} operator IDirect3DDevice9*() {return m_dev;} diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 1fcde06e94..98bcb59d65 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -361,7 +361,7 @@ protected: int prim = PRIM->PRIM; - if(!OverrideInput(prim, rt->m_texture, ds->m_texture, tex ? tex->m_texture : NULL)) + if(!OverrideInput(prim, rt->m_texture, ds->m_texture, tex)) { return; } @@ -396,9 +396,9 @@ protected: m_tc->InvalidateTextures(context->FRAME, context->ZBUF); } - virtual void Draw(int prim, GSTexture* rt, GSTexture* ds, typename GSTextureCache::GSCachedTexture* tex) = 0; + virtual void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) = 0; - virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTexture* t) + virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* t) { #pragma region ffxii pal video conversion @@ -437,9 +437,11 @@ protected: ok = false; - t = m_dev->CreateTexture(512, 512); + m_dev->Recycle(t->m_texture); - t->Update(GSVector4i(0, 0, 448, 512), video, 512 * 4); + t->m_texture = m_dev->CreateTexture(512, 512); + + t->m_texture->Update(GSVector4i(0, 0, 448, 512), video, 512 * 4); m_vertices[0] = m_vertices[0]; m_vertices[1] = m_vertices[1]; diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index f26c67cfd1..5f085798a1 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -561,11 +561,7 @@ void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, }; - D3D10_BOX box = {0, 0, 0, sizeof(vertices), 1, 1}; - - (*dev)->UpdateSubresource(dev->m_convert.vb, 0, &box, vertices, 0, 0); - - dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); dev->IASetInputLayout(dev->m_convert.il); dev->IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); @@ -589,7 +585,7 @@ void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) // set - dev->DrawPrimitive(countof(vertices)); + dev->DrawPrimitive(); // diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 4d366ea251..851a02b50d 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -536,16 +536,7 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, }; - void* buff = NULL; - - if(SUCCEEDED(dev->m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) - { - memcpy(buff, vertices, sizeof(vertices)); - - dev->m_convert.vb->Unlock(); - } - - dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); dev->IASetInputLayout(dev->m_convert.il); dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); @@ -565,7 +556,7 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) // - dev->DrawPrimitive(countof(vertices)); + dev->DrawPrimitive(); // @@ -579,11 +570,6 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) { GSDevice9* dev = (GSDevice9*)m_dev; - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - GSVector4 mm = GSVector4(-1, -1, 1, 1); - dev->BeginScene(); // om @@ -593,6 +579,8 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) // ia + GSVector4 mm = GSVector4(-1, -1, 1, 1); + GSVertexPT1 vertices[] = { {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(0, 0)}, @@ -601,16 +589,7 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(0, 0)}, }; - void* buff = NULL; - - if(SUCCEEDED(dev->m_convert.vb->Lock(0, 0, &buff, D3DLOCK_DISCARD))) - { - memcpy(buff, vertices, sizeof(vertices)); - - dev->m_convert.vb->Unlock(); - } - - dev->IASetVertexBuffer(dev->m_convert.vb, sizeof(vertices[0])); + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); dev->IASetInputLayout(dev->m_convert.il); dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); @@ -624,11 +603,11 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) // rs - dev->RSSet(w, h); + dev->RSSet(rt->GetWidth(), rt->GetHeight()); // - dev->DrawPrimitive(countof(vertices)); + dev->DrawPrimitive(); // diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index 778cb3e4c9..2b78701155 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -793,3 +793,171 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) } } +template +void GSRendererSW::VertexKick(bool skip) +{ + const GSDrawingContext* context = m_context; + + GSVector4i xy = GSVector4i::load((int)m_v.XYZ.u32[0]); + + xy = xy.insert16<3>(m_v.FOG.F); + xy = xy.upl16(); + xy -= context->XYOFFSET; + + GSVertexSW v; + + v.p = GSVector4(xy) * g_pos_scale; + + v.c = GSVector4(GSVector4i::load((int)m_v.RGBAQ.u32[0]).u8to32() << 7); + + if(tme) + { + float q; + + if(fst) + { + v.t = GSVector4(((GSVector4i)m_v.UV).upl16() << (16 - 4)); + q = 1.0f; + } + else + { + v.t = GSVector4(m_v.ST.S, m_v.ST.T); + v.t *= GSVector4(0x10000 << context->TEX0.TW, 0x10000 << context->TEX0.TH); + q = m_v.RGBAQ.Q; + } + + v.t = v.t.xyxy(GSVector4::load(q)); + } + + GSVertexSW& dst = m_vl.AddTail(); + + dst = v; + + dst.p.z = (float)min(m_v.XYZ.Z, 0xffffff00); // max value which can survive the uint32 => float => uint32 conversion + + int count = 0; + + if(GSVertexSW* v = DrawingKick(skip, count)) + { +if(!m_dump) +{ + GSVector4 pmin, pmax; + + switch(prim) + { + case GS_POINTLIST: + pmin = v[0].p; + pmax = v[0].p; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + pmin = v[0].p.minv(v[1].p); + pmax = v[0].p.maxv(v[1].p); + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + pmin = v[0].p.minv(v[1].p).minv(v[2].p); + pmax = v[0].p.maxv(v[1].p).maxv(v[2].p); + break; + } + + GSVector4 scissor = context->scissor.ex; + + GSVector4 test = (pmax < scissor) | (pmin > scissor.zwxy()); + + switch(prim) + { + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + case GS_SPRITE: + test |= pmin.ceil() == pmax.ceil(); + break; + } + + switch(prim) + { + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + // are in line or just two of them are the same (cross product == 0) + GSVector4 tmp = (v[1].p - v[0].p) * (v[2].p - v[0].p).yxwz(); + test |= tmp == tmp.yxwz(); + break; + } + + if(test.mask() & 3) + { + return; + } +} + switch(prim) + { + case GS_POINTLIST: + break; + case GS_LINELIST: + case GS_LINESTRIP: + if(PRIM->IIP == 0) {v[0].c = v[1].c;} + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + if(PRIM->IIP == 0) {v[0].c = v[2].c; v[1].c = v[2].c;} + break; + case GS_SPRITE: + break; + } + + if(m_count < 30 && m_count >= 3) + { + GSVertexSW* v = &m_vertices[m_count - 3]; + + int tl = 0; + int br = 0; + + bool isquad = false; + + switch(prim) + { + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + case GS_TRIANGLELIST: + isquad = GSVertexSW::IsQuad(v, tl, br); + break; + } + + if(isquad) + { + m_count -= 3; + + if(m_count > 0) + { + tl += m_count; + br += m_count; + + Flush(); + } + + if(tl != 0) m_vertices[0] = m_vertices[tl]; + if(br != 1) m_vertices[1] = m_vertices[br]; + + m_count = 2; + + uint32 tmp = PRIM->PRIM; + PRIM->PRIM = GS_SPRITE; + + Flush(); + + PRIM->PRIM = tmp; + + m_perfmon.Put(GSPerfMon::Quad, 1); + + return; + } + } + + m_count += count; + } +} diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 367f999d79..a3e083d799 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -54,171 +54,5 @@ public: virtual ~GSRendererSW(); template - void VertexKick(bool skip) - { - const GSDrawingContext* context = m_context; - - GSVector4i xy = GSVector4i::load((int)m_v.XYZ.u32[0]); - - xy = xy.insert16<3>(m_v.FOG.F); - xy = xy.upl16(); - xy -= context->XYOFFSET; - - GSVertexSW v; - - v.p = GSVector4(xy) * g_pos_scale; - - v.c = GSVector4(GSVector4i::load((int)m_v.RGBAQ.u32[0]).u8to32() << 7); - - if(tme) - { - float q; - - if(fst) - { - v.t = GSVector4(((GSVector4i)m_v.UV).upl16() << (16 - 4)); - q = 1.0f; - } - else - { - v.t = GSVector4(m_v.ST.S, m_v.ST.T); - v.t *= GSVector4(0x10000 << context->TEX0.TW, 0x10000 << context->TEX0.TH); - q = m_v.RGBAQ.Q; - } - - v.t = v.t.xyxy(GSVector4::load(q)); - } - - GSVertexSW& dst = m_vl.AddTail(); - - dst = v; - - dst.p.z = (float)min(m_v.XYZ.Z, 0xffffff00); // max value which can survive the uint32 => float => uint32 conversion - - int count = 0; - - if(GSVertexSW* v = DrawingKick(skip, count)) - { -if(!m_dump) -{ - GSVector4 pmin, pmax; - - switch(prim) - { - case GS_POINTLIST: - pmin = v[0].p; - pmax = v[0].p; - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - pmin = v[0].p.minv(v[1].p); - pmax = v[0].p.maxv(v[1].p); - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - pmin = v[0].p.minv(v[1].p).minv(v[2].p); - pmax = v[0].p.maxv(v[1].p).maxv(v[2].p); - break; - } - - GSVector4 scissor = context->scissor.ex; - - GSVector4 test = (pmax < scissor) | (pmin > scissor.zwxy()); - - switch(prim) - { - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - case GS_SPRITE: - test |= pmin.ceil() == pmax.ceil(); - break; - } - - switch(prim) - { - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - // are in line or just two of them are the same (cross product == 0) - GSVector4 tmp = (v[1].p - v[0].p) * (v[2].p - v[0].p).yxwz(); - test |= tmp == tmp.yxwz(); - break; - } - - if(test.mask() & 3) - { - return; - } -} - switch(prim) - { - case GS_POINTLIST: - break; - case GS_LINELIST: - case GS_LINESTRIP: - if(PRIM->IIP == 0) {v[0].c = v[1].c;} - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - if(PRIM->IIP == 0) {v[0].c = v[2].c; v[1].c = v[2].c;} - break; - case GS_SPRITE: - break; - } - - if(m_count < 30 && m_count >= 3) - { - GSVertexSW* v = &m_vertices[m_count - 3]; - - int tl = 0; - int br = 0; - - bool isquad = false; - - switch(prim) - { - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - case GS_TRIANGLELIST: - isquad = GSVertexSW::IsQuad(v, tl, br); - break; - } - - if(isquad) - { - m_count -= 3; - - if(m_count > 0) - { - tl += m_count; - br += m_count; - - Flush(); - } - - if(tl != 0) m_vertices[0] = m_vertices[tl]; - if(br != 1) m_vertices[1] = m_vertices[br]; - - m_count = 2; - - uint32 tmp = PRIM->PRIM; - PRIM->PRIM = GS_SPRITE; - - Flush(); - - PRIM->PRIM = tmp; - - m_perfmon.Put(GSPerfMon::Quad, 1); - - return; - } - } - - m_count += count; - } - } + void VertexKick(bool skip); }; diff --git a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp index 57f204f855..3efc752076 100644 --- a/plugins/GSdx/GSSetupPrimCodeGenerator.cpp +++ b/plugins/GSdx/GSSetupPrimCodeGenerator.cpp @@ -44,8 +44,6 @@ GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(GSScanlineEnvironment& env, u void GSSetupPrimCodeGenerator::Generate() { - const int params = 0; - if((m_en.z || m_en.f) && !m_sel.sprite || m_en.t || m_en.c && m_sel.iip) { for(int i = 0; i < 5; i++) diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index ab2f4ea660..65f7a6cf77 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -1213,7 +1213,7 @@ template void GSState::Transfer(uint8* mem, uint32 size) { GIFReg r; r.u64 = path.tag.PRIM; - (this->*m_fpGIFRegHandlers[GIF_A_D_REG_PRIM])(&r); + GIFRegHandlerPRIM(&r); } } } diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 925ddd0b9b..d03cf0ac0d 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -25,9 +25,6 @@ GSTextureFX10::GSTextureFX10() : m_dev(NULL) - , m_vb_max(0) - , m_vb_start(0) - , m_vb_count(0) { memset(&m_vs_cb_cache, 0, sizeof(m_vs_cb_cache)); memset(&m_ps_cb_cache, 0, sizeof(m_ps_cb_cache)); @@ -94,63 +91,7 @@ bool GSTextureFX10::Create(GSDevice10* dev) bool GSTextureFX10::SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMITIVE_TOPOLOGY prim) { - HRESULT hr; - - if(max(count * 3 / 2, 10000) > m_vb_max) - { - m_vb_old = m_vb; - m_vb = NULL; - m_vb_max = max(count * 2, 10000); - m_vb_start = 0; - m_vb_count = 0; - } - - if(!m_vb) - { - D3D10_BUFFER_DESC bd; - - memset(&bd, 0, sizeof(bd)); - - bd.Usage = D3D10_USAGE_DYNAMIC; - bd.ByteWidth = m_vb_max * sizeof(vertices[0]); - bd.BindFlags = D3D10_BIND_VERTEX_BUFFER; - bd.CPUAccessFlags = D3D10_CPU_ACCESS_WRITE; - - hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_vb); - - if(FAILED(hr)) return false; - } - - GSVertexHW10* v = NULL; - - int next = m_vb_start + m_vb_count; - - if(next + count <= m_vb_max) - { - if(SUCCEEDED(m_vb->Map(D3D10_MAP_WRITE_NO_OVERWRITE, 0, (void**)&v))) - { - memcpy(&v[next], vertices, count * sizeof(vertices[0])); - - m_vb->Unmap(); - } - - m_vb_start = next; - m_vb_count = count; - } - else - { - if(SUCCEEDED(m_vb->Map(D3D10_MAP_WRITE_DISCARD, 0, (void**)&v))) - { - memcpy(v, vertices, count * sizeof(vertices[0])); - - m_vb->Unmap(); - } - - m_vb_start = 0; - m_vb_count = count; - } - - m_dev->IASetVertexBuffer(m_vb, sizeof(vertices[0])); + m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); m_dev->IASetInputLayout(m_il); m_dev->IASetPrimitiveTopology(prim); @@ -589,5 +530,5 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, void GSTextureFX10::Draw() { - m_dev->DrawPrimitive(m_vb_count, m_vb_start); + m_dev->DrawPrimitive(); } diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 433b8565a7..db8c34c670 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -25,9 +25,6 @@ GSTextureFX9::GSTextureFX9() : m_dev(NULL) - , m_vb_max(0) - , m_vb_start(0) - , m_vb_count(0) { } @@ -89,57 +86,7 @@ GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim) { - HRESULT hr; - - if(max(count * 3 / 2, 10000) > m_vb_max) - { - m_vb_old = m_vb; - m_vb = NULL; - m_vb_max = max(count * 2, 10000); - m_vb_start = 0; - m_vb_count = 0; - } - - if(!m_vb) - { - hr = (*m_dev)->CreateVertexBuffer(m_vb_max * sizeof(vertices[0]), D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_vb, NULL); - - if(FAILED(hr)) return false; - } - - GSVertexHW9* v = NULL; - - int next = m_vb_start + m_vb_count; - int size = count * sizeof(vertices[0]); - - if(next + count <= m_vb_max) - { - int offset = next * sizeof(vertices[0]); - - if(SUCCEEDED(m_vb->Lock(offset, size, (void**)&v, D3DLOCK_NOOVERWRITE))) - { - memcpy(v, vertices, size); - - m_vb->Unlock(); - } - - m_vb_start = next; - m_vb_count = count; - } - else - { - if(SUCCEEDED(m_vb->Lock(0, size, (void**)&v, D3DLOCK_DISCARD))) - { - memcpy(v, vertices, size); - - m_vb->Unlock(); - } - - m_vb_start = 0; - m_vb_count = count; - } - - m_dev->IASetVertexBuffer(m_vb, sizeof(vertices[0])); + m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); m_dev->IASetInputLayout(m_il); m_dev->IASetPrimitiveTopology(prim); @@ -526,5 +473,5 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, void GSTextureFX9::Draw() { - m_dev->DrawPrimitive(m_vb_count, m_vb_start); + m_dev->DrawPrimitive(); } diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 831f72e1f6..65bb7bbfbf 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -155,12 +155,6 @@ private: hash_map m_ps_ss; hash_map m_om_dss; hash_map m_om_bs; - - CComPtr m_vb, m_vb_old; - int m_vb_max; - int m_vb_start; - int m_vb_count; - hash_map m_mskfix; GSTexture* CreateMskFix(uint32 size, uint32 msk, uint32 fix); diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index cd28cebad4..cfd2bc0203 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -1599,14 +1599,18 @@ public: #endif - #if _M_SSE >= 0x401 - static GSVector4i loadnt(const void* p) { - return GSVector4i(_mm_stream_load_si128((__m128i*)p)); - } + #if _M_SSE >= 0x401 - #endif + return GSVector4i(_mm_stream_load_si128((__m128i*)p)); + + #else + + return GSVector4i(_mm_load_si128((__m128i*)p)); + + #endif + } static GSVector4i loadl(const void* p) { @@ -1696,6 +1700,31 @@ public: #endif + static void storent(void* RESTRICT dst, const void* RESTRICT src, size_t size) + { + const GSVector4i* s = (const GSVector4i*)src; + GSVector4i* d = (GSVector4i*)dst; + + if(size == 0) return; + + size_t i = 0; + size_t j = size >> 6; + + for(; i < j; i++, s += 4, d += 4) + { + storent(&d[0], s[0]); + storent(&d[1], s[1]); + storent(&d[2], s[2]); + storent(&d[3], s[3]); + } + + size &= 63; + + if(size == 0) return; + + memcpy(d, s, size); + } + __forceinline static void transpose(GSVector4i& a, GSVector4i& b, GSVector4i& c, GSVector4i& d) { _MM_TRANSPOSE4_SI128(a.m, b.m, c.m, d.m); From 12e258ac16661fdd64c5160766b0e18eae40876b Mon Sep 17 00:00:00 2001 From: ramapcsx2 Date: Tue, 26 May 2009 14:03:29 +0000 Subject: [PATCH 132/296] mVU: Fix for a bug that would cause stall cycles to be counted 2 times. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1268 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index e8f1aaf0f8..74d77502a3 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -122,7 +122,7 @@ microVUt(void) mVUincCycles(int x) { } if (mVUregs.p) { calcCycles(mVUregs.p, x); - if (!mVUregs.p) { incP(); } + if (!mVUregs.p || (mVUregs.p && mVUregsTemp.p)) { incP(); } } calcCycles(mVUregs.r, x); calcCycles(mVUregs.xgkick, x); @@ -245,8 +245,8 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); mVUinfo = 0; - incCycles(1); startLoop(); + incCycles(1); mVUopU(); if (curI & _Ebit_) { branch = 1; } if (curI & _MDTbit_) { branch = 4; } From 5f1cc131c8da6817caf194ef3391dbd31b71f938 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 27 May 2009 01:49:48 +0000 Subject: [PATCH 133/296] Bugfix for Issue 241. Also increased the EErec's stack memory from 64k to 128k, since there's no reason to be frugal there anymore. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1269 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/ix86-32/iR5900-32.cpp | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 50626dfe4e..4a46b80a4c 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -75,7 +75,7 @@ u32 g_cpuHasConstReg = 0, g_cpuFlushedConstReg = 0; // Static Private Variables - R5900 Dynarec #define X86 -static const int RECSTACK_SIZE = 0x00010000; +static const int RECSTACK_SIZE = 0x00020000; static u8 *recMem = NULL; // the recompiled blocks will be here static u8* recStack = NULL; // stack mem @@ -349,7 +349,9 @@ int _flushUnusedConstReg() // u32* recAllocStackMem(int size, int align) { - recStackPtr += align - ((u32)recStackPtr % align); + jASSUME( align == 4 || align == 8 || align == 16 ); + + recStackPtr = (u8*) ( (((uptr)recStackPtr) + (align-1)) & ~(align-1) ); recStackPtr += size; return (u32*)(recStackPtr-size); } From 47bdb314738a289b3b51e97db13b5629911a882f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 27 May 2009 02:29:52 +0000 Subject: [PATCH 134/296] microVU: - Fixed a big bug with the help of sudonim and rama which greatly increases mVU's compatibility! git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1270 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 2 +- pcsx2/x86/microVU_Flags.inl | 5 +++-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index f1fc7e589b..7be750237c 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -322,7 +322,7 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { mVUinfo |= _swapOps; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } int curPC = iPC; - for (int i = mVUcount, j = 0; i > 1; i--, j++) { + for (int i = mVUcount, j = 0; i > 0; i--, j++) { incPC2(-2); if (doStatus) { mVUinfo |= _doMac; if (j >= 3) { break; } } } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 5f29d43c00..841a1414ce 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -82,9 +82,10 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking // Ensure last ~4+ instructions update mac flags + setCode(); for (int i = mVUcount; i > 0; i--, aCount++) { - if (doStatus) { if (__Mac) { mVUinfo |= _doMac; } if (aCount > 4) { break; } } - incPC2(-2); + if (doStatus) { if (__Mac) { mVUinfo |= _doMac; } if (aCount >= 4) { break; } } + incPC(-2); } // Status/Mac Flags Setup Code From b1eadad21668353e8dd925bc89f1846d2121894c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 27 May 2009 05:33:36 +0000 Subject: [PATCH 135/296] microVU: - Increased mVU run cycles (Fixes Kloana 2's Inf Loop messages) - Fixed some potential problems with the log code that would cause some games to crash. - Minor Changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1271 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU1micro.cpp | 2 +- pcsx2/x86/microVU_Compile.inl | 18 +++++++++++++++--- pcsx2/x86/microVU_Log.inl | 7 +++++++ pcsx2/x86/microVU_Misc.h | 6 +++--- pcsx2/x86/microVU_Misc.inl | 7 +++++-- 5 files changed, 31 insertions(+), 9 deletions(-) diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index a54bdd6bed..37c06e7682 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -279,7 +279,7 @@ namespace VU1micro FreezeXMMRegs(1); //FreezeMMXRegs(1); - runVUrec(VU1.VI[REG_TPC].UL, 300000 /*0x7fffffff*/, 1); + runVUrec(VU1.VI[REG_TPC].UL, 3000000 /*0x7fffffff*/, 1); //FreezeMMXRegs(0); FreezeXMMRegs(0); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 74d77502a3..8f7e3c1e2e 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -56,6 +56,8 @@ #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } #define doUpperOp() { mVUopU(); mVUdivSet(); } +#define doLowerOp() { incPC(-1); mVUopL(); incPC(1); } +#define doIbit() { if (curI & _Ibit_) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } //------------------------------------------------------------------ // Helper Functions @@ -274,8 +276,8 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { int x; for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { if (isEOB) { x = 0xffff; } - if (isNOP) { incPC(1); doUpperOp(); if (curI & _Ibit_) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } - else if (!swapOps) { incPC(1); doUpperOp(); incPC(-1); mVUopL(); incPC(1); } + if (isNOP) { incPC(1); doUpperOp(); doIbit(); } + else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } else { mVUopL(); incPC(1); doUpperOp(); } if (!isBdelay) { incPC(1); } @@ -364,12 +366,22 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { eBitTemination: mVUprint("mVUcompile ebit"); + int lStatus = findFlagInst(xStatus, 0x7fffffff); + int lMac = findFlagInst(xMac, 0x7fffffff); + int lClip = findFlagInst(xClip, 0x7fffffff); + mVUinfo = 0; incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; + if (doDivFlag) { + int flagReg; + getFlagReg(flagReg, lStatus); + AND32ItoR (flagReg, 0x0fcf); + OR32MtoR (flagReg, (uptr)&mVU->divFlag); + } // Do E-bit end stuff here mVUsetupRange(xPC - 8); - mVUendProgram(mVU->q, mVU->p, findFlagInst(xStatus, 0x7fffffff), findFlagInst(xMac, 0x7fffffff), findFlagInst(xClip, 0x7fffffff)); + mVUendProgram(mVU->q, mVU->p, lStatus, lMac, lClip); return thisPtr; } diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 0e86c23953..f685b36002 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -44,6 +44,9 @@ microVUt(void) __mVUdumpProgram(int progIndex) { bool bitX[7]; char str[30]; int delay = 0; + int bBranch = mVUbranch; + int bCode = mVU->code; + int bPC = iPC; mVUbranch = 0; sprintf(str, "%s\\microVU%d prog - %02d.html", LOGS_DIR, vuIndex, progIndex); @@ -116,6 +119,10 @@ microVUt(void) __mVUdumpProgram(int progIndex) { mVUlog("\n"); mVUlog("\n"); fclose(mVU->logFile); + mVUbranch = bBranch; + mVU->code = bCode; + iPC = bPC; + setCode(); } #endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index f5314d5910..94dc5a5603 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -291,9 +291,9 @@ declareAllVariables #endif // Speed Hacks (Set to 1 to turn On) -#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack (Very-Compatible) -#define CHECK_VU_FLAGHACK2 0 // Status Flag Speed Hack (More-Hacky) -#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack (Semi-Compatible) +#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack (Can cause Infinite loops, SPS, etc...) +#define CHECK_VU_FLAGHACK2 0 // Status Flag Speed Hack (Can cause Infinite loops, SPS, etc...) +#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack (Can cause SPS, Black Screens, etc...) // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index e0927813ca..bec3019248 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -302,12 +302,15 @@ microVUt(void) mVUrestoreRegs() { microVUt(void) mVUcheckSflag(int progIndex) { if (CHECK_VU_FLAGHACK) { microVU* mVU = mVUx; - mVUsFlagHack = 1; + int bFlagInfo = mVUflagInfo; + int bCode = mVU->code; + mVUsFlagHack = 1; for (u32 i = 0; i < mVU->progSize; i+=2) { mVU->code = mVU->prog.prog[progIndex].data[i]; mVUopL(); } - mVUflagInfo = 0; + mVUflagInfo = bFlagInfo; + mVU->code = bCode; mVU->prog.prog[progIndex].sFlagHack = mVUsFlagHack; } } From 3e12966919e4f0826093f2e2f7b53aa081ca2df1 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 27 May 2009 05:55:32 +0000 Subject: [PATCH 136/296] GSdx: Even more alpha test voodoo magic, it must be nailed now! And there is even an explanation. The tfx functions calculate At * Af >> 7, which means modulating by 0x80 should return At as the result. With the evil floating point pixel shader however 0x80 translates to 128/255 (0.502), not exactly 0.5, modulation as At' * Af' * 2 (' means 0 - 1.0 range) is not the same as with integers. At' = Af' = 0.502 At' * Af' * 2 = 0.504 If the alpha test happens to be "not equal to 0x80", then abs(0.504 - 128/255) < 0.5/255 will just miss. Solution is to re-scale those values to the integer range, do the calculations, and then back to float again, but in the end it just simplifies down to At' * Af' * 255/128, doh... At * Af >> 7 => ((At' * 255) * (Af' * 255) / 128) / 255 => At' * Af' * 255/128 At' = Af' = 0.502 0.502 * 0.502 * 255/128 = 0.502 (w00t!) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1272 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDevice10.cpp | 2 +- plugins/GSdx/GSDevice9.cpp | 2 +- plugins/GSdx/res/tfx10.fx | 8 ++++---- plugins/GSdx/res/tfx9.fx | 10 +++++----- 4 files changed, 11 insertions(+), 11 deletions(-) diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index bf22deb23d..a32bd1e7d4 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -511,7 +511,7 @@ void GSDevice10::IASetVertexBuffer(const void* vertices, size_t stride, size_t c { ASSERT(m_vertices.count == 0); - if(count > m_vertices.limit) + if(count * stride > m_vertices.limit * m_vertices.stride) { m_vertices.vb_old = m_vertices.vb; m_vertices.vb = NULL; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index e84029c303..7d4f12a940 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -648,7 +648,7 @@ void GSDevice9::IASetVertexBuffer(const void* vertices, size_t stride, size_t co { ASSERT(m_vertices.count == 0); - if(count > m_vertices.limit) + if(count * stride > m_vertices.limit * m_vertices.stride) { m_vertices.vb_old = m_vertices.vb; m_vertices.vb = NULL; diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 8f0fe77a9b..78d1c4403d 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -414,11 +414,11 @@ PS_OUTPUT ps_main(PS_INPUT input) { if(TCC == 0) { - c.rgb = c.rgb * t.rgb * 2; + c.rgb = c.rgb * t.rgb * 255.0f / 128; } else { - c = c * t * 2; + c = c * t * 255.0f / 128; } } else if(TFX == 1) @@ -434,7 +434,7 @@ PS_OUTPUT ps_main(PS_INPUT input) } else if(TFX == 2) { - c.rgb = c.rgb * t.rgb * 2 + c.a; + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; if(TCC == 1) { @@ -443,7 +443,7 @@ PS_OUTPUT ps_main(PS_INPUT input) } else if(TFX == 3) { - c.rgb = c.rgb * t.rgb * 2 + c.a; + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; if(TCC == 1) { diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index d825cf187b..384b85780d 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -241,11 +241,11 @@ float4 ps_main(PS_INPUT input) : COLOR { if(TCC == 0) { - c.rgb = c.rgb * t.rgb * 2; + c.rgb = c.rgb * t.rgb * 255.0f / 128; } else { - c = c * t * 2; + c = c * t * 255.0f / 128; } } else if(TFX == 1) @@ -261,7 +261,7 @@ float4 ps_main(PS_INPUT input) : COLOR } else if(TFX == 2) { - c.rgb = c.rgb * t.rgb * 2 + c.a; + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; if(TCC == 1) { @@ -270,7 +270,7 @@ float4 ps_main(PS_INPUT input) : COLOR } else if(TFX == 3) { - c.rgb = c.rgb * t.rgb * 2 + c.a; + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; if(TCC == 1) { @@ -292,7 +292,7 @@ float4 ps_main(PS_INPUT input) : COLOR } else if(ATST == 4) // e { - clip(0.6f / 256 - abs(c.a - AREF)); // FIXME: 0.5f is too small + clip(0.6f / 255 - abs(c.a - AREF)); // FIXME: 0.5f is too small } else if(ATST == 5 || ATST == 6) // ge, g { From 58c45c42ca6b26e0ce7e13277d894e23e07e85b3 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 27 May 2009 09:14:43 +0000 Subject: [PATCH 137/296] microVU is now selectable as a GUI option from the CPU dialog box. microVU speedhacks are also available in the speedhacks dialog. All the GUI stuff took me a few hours, so hopefully I didn't bug anything. Note to users: Please remember that microVU is a W.I.P. and will have bugs; but it also fixes some games Super VU has problems with. So have fun testing. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1273 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 31 +++++++------ pcsx2/Counters.cpp | 8 ++-- pcsx2/Linux/Pref.cpp | 15 +++--- pcsx2/VUmicro.h | 10 ++++ pcsx2/windows/CpuDlg.cpp | 36 ++++++++------- pcsx2/windows/HacksDlg.cpp | 16 +++++-- pcsx2/windows/ini.cpp | 13 ++++-- pcsx2/windows/pcsx2.rc | 73 ++++++++++++++++++----------- pcsx2/windows/resource.h | 8 +++- pcsx2/x86/iVU0micro.cpp | 85 +++++++++------------------------- pcsx2/x86/iVU1micro.cpp | 87 ++++++++++++----------------------- pcsx2/x86/microVU.cpp | 11 ++--- pcsx2/x86/microVU_Alloc.inl | 3 -- pcsx2/x86/microVU_Analyze.inl | 3 -- pcsx2/x86/microVU_Compile.inl | 2 - pcsx2/x86/microVU_Execute.inl | 2 - pcsx2/x86/microVU_Flags.inl | 2 - pcsx2/x86/microVU_Log.inl | 3 -- pcsx2/x86/microVU_Lower.inl | 2 - pcsx2/x86/microVU_Misc.h | 8 ++-- pcsx2/x86/microVU_Misc.inl | 5 +- pcsx2/x86/microVU_Tables.inl | 2 - pcsx2/x86/microVU_Upper.inl | 3 +- 23 files changed, 194 insertions(+), 234 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index d26587a84b..6e48fb03bf 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -44,21 +44,23 @@ extern SessionOverrideFlags g_Session; ////////////////////////////////////////////////////////////////////////// // Pcsx2 User Configuration Options! -//#define PCSX2_MICROVU // Use Micro VU recs instead of Zero VU Recs -//#define PCSX2_MICROVU_ // Fully enable Micro VU recs (temporary option for now) -#define PCSX2_GSMULTITHREAD 1 // uses multi-threaded gs -#define PCSX2_EEREC 0x10 -#define PCSX2_VU0REC 0x20 -#define PCSX2_VU1REC 0x40 -#define PCSX2_FRAMELIMIT_MASK 0xc00 -#define PCSX2_FRAMELIMIT_NORMAL 0x000 -#define PCSX2_FRAMELIMIT_LIMIT 0x400 -#define PCSX2_FRAMELIMIT_SKIP 0x800 +#define PCSX2_GSMULTITHREAD 0x0001 // Use Multi-Threaded GS +#define PCSX2_EEREC 0x0010 +#define PCSX2_VU0REC 0x0020 +#define PCSX2_VU1REC 0x0040 +#define PCSX2_FRAMELIMIT_MASK 0x0c00 +#define PCSX2_FRAMELIMIT_NORMAL 0x0000 +#define PCSX2_FRAMELIMIT_LIMIT 0x0400 +#define PCSX2_FRAMELIMIT_SKIP 0x0800 +#define PCSX2_MICROVU0 0x1000 // Use Micro VU0 recs instead of Zero VU0 Recs +#define PCSX2_MICROVU1 0x2000 // Use Micro VU1 recs instead of Zero VU1 Recs #define CHECK_FRAMELIMIT (Config.Options&PCSX2_FRAMELIMIT_MASK) //------------ CPU Options!!! --------------- -#define CHECK_MULTIGS (Config.Options&PCSX2_GSMULTITHREAD) +#define CHECK_MULTIGS (Config.Options&PCSX2_GSMULTITHREAD) +#define CHECK_MICROVU0 (Config.Options&PCSX2_MICROVU0) +#define CHECK_MICROVU1 (Config.Options&PCSX2_MICROVU1) #define CHECK_EEREC (!g_Session.ForceDisableEErec && Config.Options&PCSX2_EEREC) #define CHECK_VU0REC (!g_Session.ForceDisableVU0rec && Config.Options&PCSX2_VU0REC) #define CHECK_VU1REC (!g_Session.ForceDisableVU1rec && (Config.Options&PCSX2_VU1REC)) @@ -163,12 +165,15 @@ public: int Patch; int CustomFps; struct Hacks_t { - int EECycleRate; + int EECycleRate; bool IOPCycleDouble; bool WaitCycleExt; bool INTCSTATSlow; - int VUCycleSteal; bool IdleLoopFF; + int VUCycleSteal; + bool vuFlagHack1; + bool vuFlagHack2; + bool vuMinMax; bool ESCExits; // this is a hack!? } Hacks; int GameFixes; diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index 777d954cbb..d8af6b3f80 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -446,11 +446,9 @@ __forceinline bool rcntUpdate_vSync() // Accumulate hsync rounding errors: hsyncCounter.sCycle += vSyncInfo.hSyncError; - -#ifdef PCSX2_MICROVU - extern void mVUvsyncUpdate(); - mVUvsyncUpdate(); -#endif + + if (CHECK_MICROVU0) vsyncVUrec(0); + if (CHECK_MICROVU1) vsyncVUrec(1); # ifdef VSYNC_DEBUG vblankinc++; diff --git a/pcsx2/Linux/Pref.cpp b/pcsx2/Linux/Pref.cpp index 7db9dd2d4f..a6c9a79fdd 100644 --- a/pcsx2/Linux/Pref.cpp +++ b/pcsx2/Linux/Pref.cpp @@ -104,12 +104,15 @@ int LoadConfig() GetValuel("EECycleRate", Config.Hacks.EECycleRate); if (Config.Hacks.EECycleRate > 2) Config.Hacks.EECycleRate = 2; - GetValuel("IOPCycleDouble", Config.Hacks.IOPCycleDouble); - GetValuel("WaitCycleExt", Config.Hacks.WaitCycleExt); - GetValuel("INTCSTATSlow", Config.Hacks.INTCSTATSlow); - GetValuel("VUCycleSteal", Config.Hacks.VUCycleSteal); - GetValuel("IdleLoopFF", Config.Hacks.IdleLoopFF); - GetValuel("ESCExits", Config.Hacks.ESCExits); + GetValuel("IOPCycleDouble", Config.Hacks.IOPCycleDouble); + GetValuel("WaitCycleExt", Config.Hacks.WaitCycleExt); + GetValuel("INTCSTATSlow", Config.Hacks.INTCSTATSlow); + GetValuel("VUCycleSteal", Config.Hacks.VUCycleSteal); + GetValuel("vuFlagHack1", Config.Hacks.vuFlagHack1); + GetValuel("vuFlagHack2", Config.Hacks.vuFlagHack2); + GetValuel("vuMinMax", Config.Hacks.vuMinMax); + GetValuel("IdleLoopFF", Config.Hacks.IdleLoopFF); + GetValuel("ESCExits", Config.Hacks.ESCExits); if (Config.Hacks.VUCycleSteal < 0 || Config.Hacks.VUCycleSteal > 4) Config.Hacks.VUCycleSteal = 0; diff --git a/pcsx2/VUmicro.h b/pcsx2/VUmicro.h index 18bb18a384..0b60109f9d 100644 --- a/pcsx2/VUmicro.h +++ b/pcsx2/VUmicro.h @@ -72,6 +72,16 @@ void vuMicroMemReset(); // the CHECK_VU0REC / CHECK_VU1REC config options. void vuMicroCpuReset(); +///////////////////////////////////////////////////////////////// +// microVU Rec Stuff +// +extern void initVUrec(VURegs* vuRegs, const int vuIndex); +extern void closeVUrec(const int vuIndex); +extern void resetVUrec(const int vuIndex); +extern void clearVUrec(u32 addr, u32 size, const int vuIndex); +extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); +extern void vsyncVUrec(const int vuIndex); + ///////////////////////////////////////////////////////////////// // Everything else does stuff on a per-VU basis. // diff --git a/pcsx2/windows/CpuDlg.cpp b/pcsx2/windows/CpuDlg.cpp index d7a334385e..98f5e6b04d 100644 --- a/pcsx2/windows/CpuDlg.cpp +++ b/pcsx2/windows/CpuDlg.cpp @@ -52,15 +52,14 @@ BOOL CALLBACK CpuDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) Static_SetText(GetDlgItem(hW, IDC_CPUSPEEDNAME), _("CPU Speed")); Static_SetText(GetDlgItem(hW, IDC_FEATURESNAME), _("Features")); Static_SetText(GetDlgItem(hW, IDC_CPU_EEREC), _("EERec - EE/IOP recompiler (need MMX/SSE)")); - Static_SetText(GetDlgItem(hW, IDC_CPU_VUGROUP), _("VU Recompilers - All options are set by default")); - Static_SetText(GetDlgItem(hW, IDC_CPU_VU0REC), _("VU0rec - enable recompiler for VU0 unit")); - Static_SetText(GetDlgItem(hW, IDC_CPU_VU1REC), _("VU1rec - enable recompiler for VU1 unit")); + Static_SetText(GetDlgItem(hW, IDC_CPU_VUGROUP), _("VU Recompilers")); + Static_SetText(GetDlgItem(hW, IDC_CPU_VU0REC), _("VU0rec - Enable recompiler for VU0 unit.")); + Static_SetText(GetDlgItem(hW, IDC_CPU_VU1REC), _("VU1rec - Enable recompiler for VU1 unit.")); Static_SetText(GetDlgItem(hW, IDC_CPU_GSMULTI), _("Multi threaded GS mode (MTGS)\n(faster on dual core/HT procs, requires pcsx2 restart)")); Static_SetText(GetDlgItem(hW, IDC_FRAMELIMIT), _("Frame Limiting (F4 key switches the mode in-game!)")); Static_SetText(GetDlgItem(hW, IDC_CPU_FL_NORMAL), _("Normal - All frames are rendered as fast as possible.")); Static_SetText(GetDlgItem(hW, IDC_CPU_FL_LIMIT), _("Limit - Force frames to normal speeds if too fast.")); Static_SetText(GetDlgItem(hW, IDC_CPU_FL_SKIP), _("Frame Skip - In order to achieve normal speeds,\nsome frames are skipped (fast).\nFps displayed counts skipped frames too.")); - Static_SetText(GetDlgItem(hW, IDC_CPU_FL_SKIPVU), _("VU Skip - Same as 'Frame Skip', but tries to skip more.\nArtifacts might be present, but will be faster.")); Static_SetText(GetDlgItem(hW, IDC_CUSTOM_FPS), _("Custom FPS Limit (0=auto):")); Static_SetText(GetDlgItem(hW, IDC_FRAMESKIP_LABEL1), _("Skip Frames when slower than:\n(See Note 1)")); Static_SetText(GetDlgItem(hW, IDC_FRAMESKIP_LABEL2), _("Consecutive Frames before skipping:\n(See Note 2)")); @@ -86,13 +85,15 @@ BOOL CALLBACK CpuDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) // if(cpucaps.hasAMD64BitArchitecture) strcat(features,",x86-64"); SetDlgItemText(hW, IDC_FEATURESINPUT, features); - CheckDlgButton(hW, IDC_CPU_EEREC, !!(Config.Options&PCSX2_EEREC)); - CheckDlgButton(hW, IDC_CPU_VU0REC, !!(Config.Options&PCSX2_VU0REC)); - CheckDlgButton(hW, IDC_CPU_VU1REC, !!(Config.Options&PCSX2_VU1REC)); + CheckDlgButton(hW, IDC_CPU_EEREC, !!(Config.Options&PCSX2_EEREC)); + CheckDlgButton(hW, IDC_CPU_VU0REC, !!(Config.Options&PCSX2_VU0REC)); + CheckDlgButton(hW, IDC_CPU_VU1REC, !!(Config.Options&PCSX2_VU1REC)); + CheckDlgButton(hW, IDC_CPU_MVU0, !!(Config.Options&PCSX2_MICROVU0)); + CheckDlgButton(hW, IDC_CPU_MVU1, !!(Config.Options&PCSX2_MICROVU1)); - EnableDlgItem( hW, IDC_CPU_EEREC, !g_Session.ForceDisableEErec ); - EnableDlgItem( hW, IDC_CPU_VU0REC, !g_Session.ForceDisableVU0rec ); - EnableDlgItem( hW, IDC_CPU_VU1REC, !g_Session.ForceDisableVU1rec ); + EnableDlgItem( hW, IDC_CPU_EEREC, !g_Session.ForceDisableEErec ); + EnableDlgItem( hW, IDC_CPU_VU0REC, !g_Session.ForceDisableVU0rec ); + EnableDlgItem( hW, IDC_CPU_VU1REC, !g_Session.ForceDisableVU1rec ); CheckDlgButton(hW, IDC_CPU_GSMULTI, !!CHECK_MULTIGS); @@ -124,16 +125,19 @@ BOOL CALLBACK CpuDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) case IDOK: newopts = 0; - if( SendDlgItemMessage(hW,IDC_CPU_EEREC,BM_GETCHECK,0,0) ) newopts |= PCSX2_EEREC; + if( SendDlgItemMessage(hW,IDC_CPU_EEREC,BM_GETCHECK,0,0) ) newopts |= PCSX2_EEREC; - if( SendDlgItemMessage(hW,IDC_CPU_VU0REC,BM_GETCHECK,0,0) ) newopts |= PCSX2_VU0REC; - if( SendDlgItemMessage(hW,IDC_CPU_VU1REC,BM_GETCHECK,0,0) ) newopts |= PCSX2_VU1REC; + if( SendDlgItemMessage(hW,IDC_CPU_VU0REC,BM_GETCHECK,0,0) ) newopts |= PCSX2_VU0REC; + if( SendDlgItemMessage(hW,IDC_CPU_VU1REC,BM_GETCHECK,0,0) ) newopts |= PCSX2_VU1REC; + + if( SendDlgItemMessage(hW,IDC_CPU_MVU0,BM_GETCHECK,0,0) ) newopts |= PCSX2_MICROVU0; + if( SendDlgItemMessage(hW,IDC_CPU_MVU1,BM_GETCHECK,0,0) ) newopts |= PCSX2_MICROVU1; if( SendDlgItemMessage(hW,IDC_CPU_GSMULTI,BM_GETCHECK,0,0) ) newopts |= PCSX2_GSMULTITHREAD; - if( SendDlgItemMessage(hW,IDC_CPU_FL_NORMAL,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_NORMAL; - else if( SendDlgItemMessage(hW,IDC_CPU_FL_LIMIT,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_LIMIT; - else if( SendDlgItemMessage(hW,IDC_CPU_FL_SKIP,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_SKIP; + if( SendDlgItemMessage(hW,IDC_CPU_FL_NORMAL,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_NORMAL; + else if( SendDlgItemMessage(hW,IDC_CPU_FL_LIMIT,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_LIMIT; + else if( SendDlgItemMessage(hW,IDC_CPU_FL_SKIP,BM_GETCHECK,0,0) ) newopts |= PCSX2_FRAMELIMIT_SKIP; GetDlgItemText(hW, IDC_CUSTOMFPS, cfps, 20); Config.CustomFps = atoi(cfps); diff --git a/pcsx2/windows/HacksDlg.cpp b/pcsx2/windows/HacksDlg.cpp index a7a1efb74b..709a0d13e5 100644 --- a/pcsx2/windows/HacksDlg.cpp +++ b/pcsx2/windows/HacksDlg.cpp @@ -48,6 +48,9 @@ BOOL APIENTRY HacksProc(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) if(Config.Hacks.INTCSTATSlow) CheckDlgButton(hDlg, IDC_INTCSTATHACK, TRUE); if(Config.Hacks.IdleLoopFF) CheckDlgButton(hDlg, IDC_IDLELOOPFF, TRUE); if(Config.Hacks.ESCExits) CheckDlgButton(hDlg, IDC_ESCHACK, TRUE); + if(Config.Hacks.vuFlagHack1) CheckDlgButton(hDlg, IDC_VUHACK1, TRUE); + if(Config.Hacks.vuFlagHack2) CheckDlgButton(hDlg, IDC_VUHACK2, TRUE); + if(Config.Hacks.vuMinMax) CheckDlgButton(hDlg, IDC_VUHACK3, TRUE); SendDlgItemMessage(hDlg, IDC_VUCYCLE, TBM_SETRANGE, TRUE, MAKELONG(0, 4)); CheckVUCycleHack(hDlg, Config.Hacks.VUCycleSteal); @@ -87,11 +90,14 @@ BOOL APIENTRY HacksProc(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) } newhacks.IOPCycleDouble = !!IsDlgButtonChecked(hDlg, IDC_IOPSYNC); - newhacks.WaitCycleExt = !!IsDlgButtonChecked(hDlg, IDC_WAITCYCLES); - newhacks.INTCSTATSlow = !!IsDlgButtonChecked(hDlg, IDC_INTCSTATHACK); - newhacks.ESCExits = !!IsDlgButtonChecked(hDlg, IDC_ESCHACK); - newhacks.IdleLoopFF = !!IsDlgButtonChecked(hDlg, IDC_IDLELOOPFF); - newhacks.VUCycleSteal = SendDlgItemMessage(hDlg, IDC_VUCYCLE, TBM_GETPOS, 0, 0); + newhacks.WaitCycleExt = !!IsDlgButtonChecked(hDlg, IDC_WAITCYCLES); + newhacks.INTCSTATSlow = !!IsDlgButtonChecked(hDlg, IDC_INTCSTATHACK); + newhacks.ESCExits = !!IsDlgButtonChecked(hDlg, IDC_ESCHACK); + newhacks.vuFlagHack1 = !!IsDlgButtonChecked(hDlg, IDC_VUHACK1); + newhacks.vuFlagHack2 = !!IsDlgButtonChecked(hDlg, IDC_VUHACK2); + newhacks.vuMinMax = !!IsDlgButtonChecked(hDlg, IDC_VUHACK3); + newhacks.IdleLoopFF = !!IsDlgButtonChecked(hDlg, IDC_IDLELOOPFF); + newhacks.VUCycleSteal = SendDlgItemMessage(hDlg, IDC_VUCYCLE, TBM_GETPOS, 0, 0); CheckVUCycleHack(hDlg, newhacks.VUCycleSteal); EndDialog(hDlg, TRUE); diff --git a/pcsx2/windows/ini.cpp b/pcsx2/windows/ini.cpp index 81f25f82eb..93af72fd67 100644 --- a/pcsx2/windows/ini.cpp +++ b/pcsx2/windows/ini.cpp @@ -227,11 +227,14 @@ void IniFile::DoConfig( PcsxConfig& Conf ) Entry("EECycleRate", Config.Hacks.EECycleRate); if (Config.Hacks.EECycleRate > 2) Config.Hacks.EECycleRate = 2; - Entry("IOPCycleDouble", Config.Hacks.IOPCycleDouble); - Entry("WaitCycleExt", Config.Hacks.WaitCycleExt); - Entry("INTCSTATSlow", Config.Hacks.INTCSTATSlow); - Entry("VUCycleSteal", Config.Hacks.VUCycleSteal); - Entry("IdleLoopFF", Config.Hacks.IdleLoopFF); + Entry("IOPCycleDouble", Config.Hacks.IOPCycleDouble); + Entry("WaitCycleExt", Config.Hacks.WaitCycleExt); + Entry("INTCSTATSlow", Config.Hacks.INTCSTATSlow); + Entry("VUCycleSteal", Config.Hacks.VUCycleSteal); + Entry("vuFlagHack1", Config.Hacks.vuFlagHack1); + Entry("vuFlagHack2", Config.Hacks.vuFlagHack2); + Entry("vuMinMax", Config.Hacks.vuMinMax); + Entry("IdleLoopFF", Config.Hacks.IdleLoopFF); if (Conf.Hacks.VUCycleSteal < 0 || Conf.Hacks.VUCycleSteal > 4) Conf.Hacks.VUCycleSteal = 0; Entry("ESCExits", Config.Hacks.ESCExits); diff --git a/pcsx2/windows/pcsx2.rc b/pcsx2/windows/pcsx2.rc index e1013f35db..01972787ce 100644 --- a/pcsx2/windows/pcsx2.rc +++ b/pcsx2/windows/pcsx2.rc @@ -344,7 +344,7 @@ BEGIN CTEXT "Pcsx2 Official Website and Forums",IDC_LINK_WEBSITE,79,151,273,10,SS_NOPREFIX | SS_NOTIFY END -IDD_HACKS DIALOGEX 0, 0, 335, 263 +IDD_HACKS DIALOGEX 0, 0, 543, 229 STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_CAPTION | WS_SYSMENU CAPTION "PCSX2 Speed Hacks" FONT 8, "MS Shell Dlg", 400, 0, 0x1 @@ -352,29 +352,37 @@ BEGIN CONTROL "Default Cycle Rate",IDC_EESYNC_DEFAULT,"Button",BS_AUTORADIOBUTTON,13,44,87,10 CONTROL "Use x1.5 Cycle Rate",IDC_EESYNC1,"Button",BS_AUTORADIOBUTTON,13,79,87,10 CONTROL "Use x2 Cycle Rate",IDC_EESYNC2,"Button",BS_AUTORADIOBUTTON,13,113,83,10 - CONTROL "Enable IOP x2 Cycle Rate",IDC_IOPSYNC,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,174,88,98,10 - CONTROL "WaitCycles Sync Hack",IDC_WAITCYCLES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,174,127,90,10 - CONTROL "Escape Hack - Use Esc key to fully exit PCSX2.",IDC_ESCHACK, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,13,238,180,10 - DEFPUSHBUTTON "OK",IDOK,217,242,50,14 - PUSHBUTTON "Cancel",IDCANCEL,278,242,50,14 - CTEXT "These hacks will speed up emulation but reduce emulation compatibility or cause visual errors. If you have problems, disable all these and try again!",IDC_HACKDESC,18,7,286,19 - GROUPBOX "EmotionEngine (EE) Sync Hacks",IDC_STATIC,7,31,159,143 - GROUPBOX "Miscellaneous",IDC_STATIC,7,223,194,33 - LTEXT "Important: the X2 sync hack *will* cause choppy/skippy audio on many FMV movies.",IDC_STATIC,13,149,137,22 + CONTROL "Enable IOP x2 Cycle Rate",IDC_IOPSYNC,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,88,98,10 + CONTROL "WaitCycles Sync Hack",IDC_WAITCYCLES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,127,90,10 + CONTROL "Esc Hack - Use Esc key to fully exit PCSX2.",IDC_ESCHACK, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,13,197,152,10 + DEFPUSHBUTTON "OK",IDOK,424,208,50,14 + PUSHBUTTON "Cancel",IDCANCEL,486,208,50,14 + CTEXT "These hacks will speed up emulation but reduce emulation compatibility or cause visual errors.\nIf you have problems, disable all these and try again!",IDC_HACKDESC,7,7,529,19 + GROUPBOX "EmotionEngine (EE) Sync Hacks",IDC_STATIC,7,26,162,148 + GROUPBOX "Miscellaneous",IDC_STATIC,7,184,162,33 + LTEXT "Important: the X2 sync hack *will* cause choppy/skippy audio on many FMV movies.",IDC_STATIC,13,146,137,25 LTEXT "Big speedup! Works well with many games.",IDC_STATIC,25,124,125,19 LTEXT "Most compatible option - recommended for everyone with high-end machines.",IDC_STATIC,25,55,136,19 - LTEXT "Small speedup and works well with most games.",IDC_STATIC,186,100,134,22 - LTEXT "Small speedup. Works well with most games, but may cause certain games to crash or freeze up during bootup or stage changes.",IDC_STATIC,186,139,141,39 + LTEXT "Small speedup and works well with most games.",IDC_STATIC,200,100,134,22 + LTEXT "Small speedup. Works well with most games, but may cause certain games to crash or freeze up during bootup or stage changes.",IDC_STATIC,200,139,141,39 LTEXT "Moderate speedup and works well with most games.",IDC_STATIC,25,90,129,19 - CONTROL "INTC Sync Hack",IDC_INTCSTATHACK,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,174,43,127,10 - LTEXT "Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks).",IDC_STATIC,186,55,140,28 - CONTROL "",IDC_VUCYCLE,"msctls_trackbar32",TBS_BOTH | TBS_NOTICKS | WS_TABSTOP,13,178,36,15 - LTEXT "This space intentionally left blank",IDC_VUCYCLEDESC,25,196,142,24 - LTEXT "VU Cycle Stealing (experimental)",IDC_STATIC,49,183,105,8 + CONTROL "INTC Sync Hack",IDC_INTCSTATHACK,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,43,127,10 + LTEXT "Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks).",IDC_STATIC,200,55,140,28 + CONTROL "",IDC_VUCYCLE,"msctls_trackbar32",TBS_BOTH | TBS_NOTICKS | WS_TABSTOP,372,152,36,15 + LTEXT "This space intentionally left blank",IDC_VUCYCLEDESC,383,171,142,26 + LTEXT "VU Cycle Stealing (experimental)",IDC_STATIC,408,157,105,8 CONTROL "Idle Loop Fast-Forward (experimental)",IDC_IDLELOOPFF, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,174,179,139,10 - LTEXT "Speedup for a few games, including FFX with no known side effects. More later.",IDC_STATIC,186,193,142,26 + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,179,139,10 + LTEXT "Speedup for a few games, including FFX with no known side effects.",IDC_STATIC,200,193,144,23 + GROUPBOX "EmotionEngine (EE) Misc Hacks",IDC_STATIC,177,26,179,192 + GROUPBOX "Vertex Unit (VU) Hacks",IDC_STATIC,363,26,173,173 + CONTROL "Flag Hack 1 (microVU only)",IDC_VUHACK1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,42,155,12 + CONTROL "Flag Hack 2 (microVU only)",IDC_VUHACK2,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,79,154,12 + LTEXT "Big Speedup and moderately compatible. May cause SPS and Infinite loop problems.",IDC_STATIC,386,56,142,19 + LTEXT "Big Speedup and moderately compatible. SuperVU does something similar by default.",IDC_STATIC,385,92,144,20 + CONTROL "Min / Max Hack (microVU only)",IDC_VUHACK3,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,115,156,12 + LTEXT "Small Speedup. May cause SPS, hanging, and missing geometry.",IDC_STATIC,386,130,143,19 END @@ -395,12 +403,15 @@ BEGIN IDD_HACKS, DIALOG BEGIN LEFTMARGIN, 7 - RIGHTMARGIN, 328 + RIGHTMARGIN, 536 VERTGUIDE, 13 - VERTGUIDE, 174 - VERTGUIDE, 186 + VERTGUIDE, 177 + VERTGUIDE, 187 + VERTGUIDE, 200 + VERTGUIDE, 372 TOPMARGIN, 7 - BOTTOMMARGIN, 256 + BOTTOMMARGIN, 222 + HORZGUIDE, 26 END END #endif // APSTUDIO_INVOKED @@ -834,12 +845,12 @@ FONT 8, "MS Shell Dlg", 400, 0, 0x1 BEGIN CONTROL "EERec - EE/IOP recompiler (need MMX/SSE/SSE2)",IDC_CPU_EEREC, "Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,100,265,18 - CONTROL "VU0rec - enable recompiler for VU0 unit",IDC_CPU_VU0REC, + CONTROL "VU0rec - Enable recompiler for VU0 unit.",IDC_CPU_VU0REC, "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,14,127,219,18 - CONTROL "VU1rec - enable recompiler for VU1 unit",IDC_CPU_VU1REC, + CONTROL "VU1rec - Enable recompiler for VU1 unit.",IDC_CPU_VU1REC, "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,14,143,214,18 CONTROL "Multi threaded GS mode (MTGS)\n (faster on dual core/HT procs, requires pcsx2 restart)",IDC_CPU_GSMULTI, - "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,7,166,231,30 + "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,7,232,231,30 CONTROL "Normal - All frames are rendered as fast as possible.",IDC_CPU_FL_NORMAL, "Button",BS_AUTORADIOBUTTON | BS_MULTILINE | WS_GROUP,309,17,221,17 CONTROL "Limit - Force frames to normal speeds if too fast.\n (You can set a custom FPS limit below.)",IDC_CPU_FL_LIMIT, @@ -855,7 +866,7 @@ BEGIN LTEXT "",IDC_FAMILYINPUT,112,41,124,8 LTEXT "",IDC_FEATURESINPUT,111,79,124,8 LTEXT "",IDC_CPUSPEEDINPUT,111,61,124,8 - GROUPBOX "VU Recompilers - All options are set by default",IDC_CPU_VUGROUP,7,119,265,46 + GROUPBOX "VU Recompilers",IDC_CPU_VUGROUP,7,119,265,108 LTEXT "Features",IDC_FEATURESNAME,12,78,88,8 GROUPBOX "",IDC_STATIC,7,7,265,90 LTEXT "Custom FPS Limit (0=auto):",IDC_CUSTOM_FPS,327,103,124,12 @@ -871,6 +882,11 @@ BEGIN EDITTEXT IDC_CUSTOM_CONSECUTIVE_SKIP,456,157,53,13,ES_AUTOHSCROLL | ES_NUMBER LTEXT "Consecutive Frames to skip:\n(See Note 3)",IDC_FRAMESKIP_LABEL5,327,157,121,17 LTEXT "*Note 3: Will skip this number of frames before\n rendering the next sequence of frames. (0=default)\n (e.g. If set to 2, will skip 2 consecutive frames whenever its time\n to skip.)",IDC_FRAMESKIP_LABEL6,318,234,217,32 + GROUPBOX "Use microVU?",IDC_CPU_VUGROUP2,14,167,235,47 + CONTROL "microVU0 - Use microVU rec for VU0 recompilation.",IDC_CPU_MVU0, + "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,24,180,215,13 + CONTROL "microVU1 - Use microVU rec for VU1 recompilation.",IDC_CPU_MVU1, + "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,24,195,215,13 END @@ -894,6 +910,7 @@ BEGIN BEGIN LEFTMARGIN, 7 RIGHTMARGIN, 558 + VERTGUIDE, 24 TOPMARGIN, 7 END END diff --git a/pcsx2/windows/resource.h b/pcsx2/windows/resource.h index e519a3275a..e12f4547dd 100644 --- a/pcsx2/windows/resource.h +++ b/pcsx2/windows/resource.h @@ -176,8 +176,8 @@ #define IDC_CPU_FL_NORMAL 1265 #define IDC_CPU_FL_LIMIT 1266 #define IDC_CPU_FL_SKIP 1267 -#define IDC_CPU_FL_SKIPVU 1268 #define IDC_CPU_VUGROUP 1269 +#define IDC_CPU_VUGROUP2 1270 #define IDC_GROUPS 1272 #define IDC_PATCHES 1273 #define IDC_CUSTOMFPS 1275 @@ -275,7 +275,13 @@ #define IDC_SLIDER1 1327 #define IDC_VUCYCLE 1327 #define IDC_VUCYCLEDESC 1328 +#define IDC_VUHACK1 1329 #define IDC_IDLELOOPFF 1330 +#define IDC_VUHACK2 1331 +#define IDC_INTCSTATHACK4 1332 +#define IDC_VUHACK3 1332 +#define IDC_CPU_MVU0 1335 +#define IDC_CPU_MVU1 1336 #define IDC_CPULOG 1500 #define IDC_MEMLOG 1501 #define IDC_HWLOG 1502 diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index ccdfa0580d..cb85aa442b 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -23,81 +23,42 @@ #include "VUmicro.h" #include "iVUzerorec.h" -#ifndef PCSX2_MICROVU_ -namespace VU0micro -{ - void recAlloc() - { - SuperVUAlloc(0); - } - - void __fastcall recClear(u32 Addr, u32 Size) - { - SuperVUClear(Addr, Size, 0); // Size should be a multiple of 8 bytes! - } - - void recShutdown() - { - SuperVUDestroy( 0 ); - } - - static void recReset() - { - SuperVUReset(0); - - // this shouldn't be needed, but shouldn't hurt anything either. - x86FpuState = FPU_STATE; - } - - static void recStep() - { - } - - static void recExecuteBlock() - { - if((VU0.VI[REG_VPU_STAT].UL & 1) == 0) - return; - - FreezeXMMRegs(1); - SuperVUExecuteProgram(VU0.VI[ REG_TPC ].UL & 0xfff, 0); - FreezeXMMRegs(0); - } -} -#else - -extern void initVUrec(VURegs* vuRegs, const int vuIndex); -extern void closeVUrec(const int vuIndex); -extern void resetVUrec(const int vuIndex); -extern void clearVUrec(u32 addr, u32 size, const int vuIndex); -extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); +#define useMVU0 CHECK_MICROVU0 namespace VU0micro { - void recAlloc() { initVUrec(&VU0, 0); } - void __fastcall recClear(u32 Addr, u32 Size) { clearVUrec(Addr, Size, 0); } - void recShutdown() { closeVUrec(0); } - static void recReset() { resetVUrec(0); x86FpuState = FPU_STATE; } - static void recStep() {} - static void recExecuteBlock() + void recAlloc() { initVUrec(&VU0, 0); SuperVUAlloc(0); } + void recShutdown() { closeVUrec(0); SuperVUDestroy(0); } + + void __fastcall recClear(u32 Addr, u32 Size) { + if (useMVU0) clearVUrec(Addr, Size, 0); + else SuperVUClear(Addr, Size, 0); + } + + static void recReset() { + if (useMVU0) resetVUrec(0); + else SuperVUReset(0); + x86FpuState = FPU_STATE; + } + + static void recStep() {} + static void recExecuteBlock() { if ((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); - //FreezeMMXRegs(1); - runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); - //FreezeMMXRegs(0); + if (useMVU0) runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); + else SuperVUExecuteProgram(VU0.VI[REG_TPC].UL & 0xfff, 0); FreezeXMMRegs(0); } - } -#endif using namespace VU0micro; const VUmicroCpu recVU0 = { - recReset -, recStep -, recExecuteBlock -, recClear + recReset + , recStep + , recExecuteBlock + , recClear }; diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 37c06e7682..8e3f704a75 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -26,9 +26,7 @@ #include "VUmicro.h" #include "iVUzerorec.h" -#ifdef _DEBUG -extern u32 vudump; -#endif +#define useMVU1 CHECK_MICROVU1 int mVUdebugNow = 0; @@ -45,11 +43,8 @@ void VUtestPause() { runAmount++; if (runAmount < 654) return; -#ifndef PCSX2_MICROVU_ - SysPrintf("Super VU - Pass %d\n", runAmount); -#else - SysPrintf("Micro VU - Pass %d\n", runAmount); -#endif + if (useMVU1) SysPrintf("Micro VU - Pass %d\n", runAmount); + else SysPrintf("Super VU - Pass %d\n", runAmount); for (int i = 0; i < 32; i++) { SysPrintf("VF%02d = {%f, %f, %f, %f}\n", i, VU1.VF[i].F[0], VU1.VF[i].F[1], VU1.VF[i].F[2], VU1.VF[i].F[3]); @@ -88,17 +83,15 @@ void VUtestPause() { void VUtestPause() {} #endif +#ifdef _DEBUG +extern u32 vudump; +#endif #ifdef DEBUG_COMPARE2 #ifndef DEBUG_COMPARE #include #endif -extern void initVUrec(VURegs* vuRegs, const int vuIndex); -extern void closeVUrec(const int vuIndex); -extern void resetVUrec(const int vuIndex); -extern void clearVUrec(u32 addr, u32 size, const int vuIndex); -extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); PCSX2_ALIGNED16(u8 backVUregs[sizeof(VURegs)]); PCSX2_ALIGNED16(u8 cmpVUregs [sizeof(VURegs)]); @@ -253,21 +246,24 @@ namespace VU1micro } } #else -#ifdef PCSX2_MICROVU_ - -extern void initVUrec(VURegs* vuRegs, const int vuIndex); -extern void closeVUrec(const int vuIndex); -extern void resetVUrec(const int vuIndex); -extern void clearVUrec(u32 addr, u32 size, const int vuIndex); -extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); namespace VU1micro { - void recAlloc() { initVUrec(&VU1, 1); } - void __fastcall recClear(u32 Addr, u32 Size) { clearVUrec(Addr, Size, 1); } - void recShutdown() { closeVUrec(1); } - static void recReset() { resetVUrec(1); x86FpuState = FPU_STATE; } - static void recStep() {} + void recAlloc() { initVUrec(&VU1, 1); SuperVUAlloc(1); } + void recShutdown() { closeVUrec(1); SuperVUDestroy(1); } + + void __fastcall recClear(u32 Addr, u32 Size) { + if (useMVU1) clearVUrec(Addr, Size, 1); + else SuperVUClear(Addr, Size, 1); + } + + static void recReset() { + if (useMVU1) resetVUrec(1); + else SuperVUReset(1); + x86FpuState = FPU_STATE; + } + + static void recStep() {} static void recExecuteBlock() { if ((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; @@ -278,42 +274,19 @@ namespace VU1micro #endif FreezeXMMRegs(1); - //FreezeMMXRegs(1); - runVUrec(VU1.VI[REG_TPC].UL, 3000000 /*0x7fffffff*/, 1); - //FreezeMMXRegs(0); + if (useMVU1) runVUrec(VU1.VI[REG_TPC].UL, 3000000, 1); + else { + if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { + Console::Error("VU1 memory overflow!!: %x", params VU1.VI[REG_TPC].UL); + } + do { // while loop needed since not always will return finished + SuperVUExecuteProgram(VU1.VI[REG_TPC].UL & 0x3fff, 1); + } while( VU0.VI[REG_VPU_STAT].UL&0x100 ); + } FreezeXMMRegs(0); - VUtestPause(); } } -#else - -namespace VU1micro -{ - void recAlloc() { SuperVUAlloc(1); } - void __fastcall recClear(u32 Addr, u32 Size) { SuperVUClear(Addr, Size, 1); } - void recShutdown() { SuperVUDestroy(1); } - static void recReset() { SuperVUReset(1); x86FpuState = FPU_STATE; } - static void recStep() {} - - static void recExecuteBlock(void) - { - if((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) return; - if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { Console::Error("VU1 memory overflow!!: %x", params VU1.VI[REG_TPC].UL); } - assert((VU1.VI[ REG_TPC ].UL&7) == 0); - -#ifdef DEBUG_COMPARE - SysPrintf("(%08d) StartPC = 0x%04x\n", runAmount, VU1.VI[REG_TPC].UL); -#endif - - FreezeXMMRegs(1); - do { // while loop needed since not always will return finished - SuperVUExecuteProgram(VU1.VI[ REG_TPC ].UL & 0x3fff, 1); - } while( VU0.VI[ REG_VPU_STAT ].UL&0x100 ); - FreezeXMMRegs(0); - } -} -#endif #endif using namespace VU1micro; diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 8956d4b266..42e15ed33e 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -20,7 +20,6 @@ #include "PrecompiledHeader.h" #include "microVU.h" -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Micro VU - Global Variables @@ -193,7 +192,7 @@ microVUt(int) mVUfindLeastUsedProg() { // frame-based decrementing system in combination with a program-execution-based incrementing // system. In english: if last_used >= 2 it means the program has been used for the current // or prev frame. if it's 0, the program hasn't been used for a while. -microVUt(void) __mVUvsyncUpdate() { +microVUt(void) mVUvsyncUpdate() { microVU* mVU = mVUx; if (mVU->prog.total < mVU->prog.max) return; @@ -283,9 +282,7 @@ void runVUrec(u32 startPC, u32 cycles, const int vuIndex) { else startVU1(startPC, cycles); } -void mVUvsyncUpdate() { - __mVUvsyncUpdate<0>(); - __mVUvsyncUpdate<1>(); +void vsyncVUrec(const int vuIndex) { + if (!vuIndex) mVUvsyncUpdate<0>(); + else mVUvsyncUpdate<1>(); } - -#endif // PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index c0d9216f53..8556a6ec6a 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Micro VU - Pass 2 Functions @@ -769,5 +768,3 @@ microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { if (!_reg_ && (_fxf_ < 3)) { XOR32RtoR(GPRreg, GPRreg); } \ else { MOV32MtoR(GPRreg, (uptr)&mVU->regs->VF[_reg_].UL[0]); } \ } - -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 7be750237c..44eded0a77 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Micro VU - Pass 1 Functions @@ -384,5 +383,3 @@ microVUt(void) mVUanalyzeBranch2(int Is, int It) { if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is); analyzeVIreg1(It); } else { analyzeBranchVI(Is, _memReadIs); analyzeBranchVI(It, _memReadIt);} } - -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 8f7e3c1e2e..1a0a44e77d 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Helper Macros @@ -389,4 +388,3 @@ eBitTemination: void* __fastcall mVUcompileVU0(u32 startPC, uptr pState) { return mVUcompile<0>(startPC, pState); } void* __fastcall mVUcompileVU1(u32 startPC, uptr pState) { return mVUcompile<1>(startPC, pState); } -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 5c61f33ca6..456dcb8003 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Dispatcher Functions @@ -149,4 +148,3 @@ void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles) { return mVUexecute<1>(s void __fastcall mVUcleanUpVU0() { mVUcleanUp<0>(); } void __fastcall mVUcleanUpVU1() { mVUcleanUp<1>(); } -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 841a1414ce..4ac07fa77a 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet() { @@ -231,4 +230,3 @@ microVUt(void) mVUsetFlagInfo() { } } -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index f685b36002..bdbda985a1 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU // writes text directly to the microVU.txt, no newlines appended. microVUx(void) __mVULog(const char* fmt, ...) { @@ -124,5 +123,3 @@ microVUt(void) __mVUdumpProgram(int progIndex) { iPC = bPC; setCode(); } - -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index ad619cfb87..95a2784456 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Micro VU Micromode Lower instructions @@ -1258,4 +1257,3 @@ microVUf(void) mVU_JALR() { pass3 { mVUlog("JALR vi%02d, [vi%02d]", _Ft_, _Fs_); } } -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 94dc5a5603..8832690bd4 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -290,10 +290,10 @@ declareAllVariables #define mVUdumpProg 0&& #endif -// Speed Hacks (Set to 1 to turn On) -#define CHECK_VU_FLAGHACK 0 // Status Flag Speed Hack (Can cause Infinite loops, SPS, etc...) -#define CHECK_VU_FLAGHACK2 0 // Status Flag Speed Hack (Can cause Infinite loops, SPS, etc...) -#define CHECK_VU_MINMAXHACK 0 // Min/Max Speed Hack (Can cause SPS, Black Screens, etc...) +// Speed Hacks +#define CHECK_VU_FLAGHACK1 (u32)Config.Hacks.vuFlagHack1 // (Can cause Infinite loops, SPS, etc...) +#define CHECK_VU_FLAGHACK2 (u32)Config.Hacks.vuFlagHack2 // (Can cause Infinite loops, SPS, etc...) +#define CHECK_VU_MINMAXHACK (u32)Config.Hacks.vuMinMax // (Can cause SPS, Black Screens, etc...) // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index bec3019248..af3ff9608f 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Micro VU - Clamp Functions @@ -300,7 +299,7 @@ microVUt(void) mVUrestoreRegs() { // Reads entire microProgram and finds out if Status Flag is Used microVUt(void) mVUcheckSflag(int progIndex) { - if (CHECK_VU_FLAGHACK) { + if (CHECK_VU_FLAGHACK1) { microVU* mVU = mVUx; int bFlagInfo = mVUflagInfo; int bCode = mVU->code; @@ -370,5 +369,3 @@ void SSE_MIN2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { if (CHECK_VU_MINMAXHACK) { SSE_MINSS_XMM_to_XMM(to, from); } else { MIN_MAX_SS(to, from, 1); } } - -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Tables.inl b/pcsx2/x86/microVU_Tables.inl index 8d54233993..a6f5cbac61 100644 --- a/pcsx2/x86/microVU_Tables.inl +++ b/pcsx2/x86/microVU_Tables.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // Declarations @@ -265,4 +264,3 @@ microVUf(void) mVUunknown() { pass3 { mVUlog("Unknown", mVUgetCode); } } -#endif //PCSX2_MICROVU diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 3ecdb57fb7..160d044e89 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -17,7 +17,6 @@ */ #pragma once -#ifdef PCSX2_MICROVU //------------------------------------------------------------------ // mVUupdateFlags() - Updates status/mac flags @@ -643,4 +642,4 @@ microVUf(void) mVU_CLIP() { } pass3 { mVUlog("CLIP"); mVUlogCLIP(); } } -#endif //PCSX2_MICROVU + From 477f5fa674e227bdda66a06e717dd9da05e7d0fb Mon Sep 17 00:00:00 2001 From: ramapcsx2 Date: Wed, 27 May 2009 10:26:36 +0000 Subject: [PATCH 138/296] Minor changes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1274 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/pcsx2.rc | 4 ++-- pcsx2/x86/microVU.cpp | 2 +- pcsx2/x86/microVU_Flags.inl | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/pcsx2/windows/pcsx2.rc b/pcsx2/windows/pcsx2.rc index 01972787ce..ea89271c47 100644 --- a/pcsx2/windows/pcsx2.rc +++ b/pcsx2/windows/pcsx2.rc @@ -883,9 +883,9 @@ BEGIN LTEXT "Consecutive Frames to skip:\n(See Note 3)",IDC_FRAMESKIP_LABEL5,327,157,121,17 LTEXT "*Note 3: Will skip this number of frames before\n rendering the next sequence of frames. (0=default)\n (e.g. If set to 2, will skip 2 consecutive frames whenever its time\n to skip.)",IDC_FRAMESKIP_LABEL6,318,234,217,32 GROUPBOX "Use microVU?",IDC_CPU_VUGROUP2,14,167,235,47 - CONTROL "microVU0 - Use microVU rec for VU0 recompilation.",IDC_CPU_MVU0, + CONTROL "microVU0 - Use microVU rec for VU0 recompilation. ! Beta !",IDC_CPU_MVU0, "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,24,180,215,13 - CONTROL "microVU1 - Use microVU rec for VU1 recompilation.",IDC_CPU_MVU1, + CONTROL "microVU1 - Use microVU rec for VU1 recompilation. ! Beta !",IDC_CPU_MVU1, "Button",BS_AUTOCHECKBOX | BS_MULTILINE | WS_TABSTOP,24,195,215,13 END diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 42e15ed33e..a58200a911 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -159,7 +159,7 @@ microVUt(int) mVUfindLeastUsedProg() { mVUcacheProg(mVU->prog.total); // Cache Micro Program mVU->prog.prog[mVU->prog.total].used = 1; mVU->prog.prog[mVU->prog.total].last_used = 3; - Console::Notice("microVU%d: Cached MicroPrograms = %d", params vuIndex, mVU->prog.total+1); + DevCon::Notice("microVU%d: Cached MicroPrograms = %d", params vuIndex, mVU->prog.total+1); return mVU->prog.total; } else { diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 4ac07fa77a..060136f48b 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -52,7 +52,7 @@ microVUt(void) mVUstatusFlagOp() { } } iPC = curPC; - Console::Status("microVU%d: FSSET Optimization", params vuIndex); + DevCon::Notice("microVU%d: FSSET Optimization", params vuIndex); } int findFlagInst(int* fFlag, int cycles) { From e8ae911664b8b227f67b9df1a3dee7a48876d855 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Wed, 27 May 2009 13:10:43 +0000 Subject: [PATCH 139/296] Linux: Hack microVU into the Gui. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1275 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Defs.h | 15 +- pcsx2/Linux/CpuDlg.cpp | 8 +- pcsx2/Linux/HacksDlg.cpp | 6 + pcsx2/Linux/interface.c | 117 +++++- pcsx2/Linux/pcsx2.glade | 232 ++++++++++- pcsx2/x86/iVUzerorec.cpp | 5 + plugins/GSnull/GifTransfer.cpp | 4 - plugins/zeropad/Windows/win.cpp | 688 +++++++++++++++++--------------- 8 files changed, 732 insertions(+), 343 deletions(-) diff --git a/common/include/Pcsx2Defs.h b/common/include/Pcsx2Defs.h index 9060c0423c..60e7306de8 100644 --- a/common/include/Pcsx2Defs.h +++ b/common/include/Pcsx2Defs.h @@ -55,7 +55,7 @@ # else # define jBREAKPOINT() ((void) *(volatile char *) 0) # endif -# define jASSUME(exp) do { if(exp) ; else jBREAKPOINT() } while(0); +# define jASSUME(exp) do { if(exp) ; else jBREAKPOINT(); } while(0); # endif #endif @@ -94,6 +94,13 @@ default: \ #define __naked __declspec(naked) #define __unused /*unused*/ #define __noinline __declspec(noinline) + +// Don't know if there are Visual C++ equivalents of these. +#define __hot +#define __cold +#define likely(x) x +#define unlikely(x) x + #define CALLBACK __stdcall #else @@ -106,7 +113,7 @@ default: \ /* Test for GCC > 4.4.0; Should be adjusted when new versions come out */ #if GCC_VERSION >= 40400 #define THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 -#define __nooptimization __attribute__((optimize("O0")) +#define __nooptimization __attribute__((optimize("O0"))) #endif /* @@ -137,6 +144,10 @@ This theoretically unoptimizes. Not having much luck so far. #define _inline __inline__ __attribute__((unused)) #define __forceinline __attribute__((always_inline,unused)) #define __noinline __attribute__((noinline)) +#define __hot __attribute__((hot)) +#define __cold __attribute__((cold)) +#define likely(x) __builtin_expect(!!(x), 1) +#define unlikely(x) __builtin_expect(!!(x), 0) #endif #ifndef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0 diff --git a/pcsx2/Linux/CpuDlg.cpp b/pcsx2/Linux/CpuDlg.cpp index 4baca1577e..8982c63fc1 100644 --- a/pcsx2/Linux/CpuDlg.cpp +++ b/pcsx2/Linux/CpuDlg.cpp @@ -31,6 +31,10 @@ void OnCpu_Ok(GtkButton *button, gpointer user_data) newopts |= PCSX2_VU0REC; if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU1rec")))) newopts |= PCSX2_VU1REC; + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec")))) + newopts |= PCSX2_MICROVU0; + if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec")))) + newopts |= PCSX2_MICROVU1; if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_MTGS")))) newopts |= PCSX2_GSMULTITHREAD; if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitNormal")))) @@ -39,7 +43,7 @@ void OnCpu_Ok(GtkButton *button, gpointer user_data) newopts |= PCSX2_FRAMELIMIT_LIMIT; else if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitFS")))) newopts |= PCSX2_FRAMELIMIT_SKIP; - + Config.CustomFps = gtk_spin_button_get_value_as_int(GTK_SPIN_BUTTON(lookup_widget(CpuDlg, "CustomFPSLimit"))); Config.CustomFrameSkip = gtk_spin_button_get_value_as_int(GTK_SPIN_BUTTON(lookup_widget(CpuDlg, "FrameThreshold"))); Config.CustomConsecutiveFrames = gtk_spin_button_get_value_as_int(GTK_SPIN_BUTTON(lookup_widget(CpuDlg, "FramesBeforeSkipping"))); @@ -82,6 +86,8 @@ void OnConf_Cpu(GtkMenuItem *menuitem, gpointer user_data) gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_EERec")), !!CHECK_EEREC); gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU0rec")), !!CHECK_VU0REC); gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU1rec")), !!CHECK_VU1REC); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec")), !!CHECK_MICROVU0); + gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec")), !!CHECK_MICROVU1); gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_MTGS")), !!CHECK_MULTIGS); gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitNormal")), CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_NORMAL); gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitLimit")), CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_LIMIT); diff --git a/pcsx2/Linux/HacksDlg.cpp b/pcsx2/Linux/HacksDlg.cpp index 0177d60930..470425cbcf 100644 --- a/pcsx2/Linux/HacksDlg.cpp +++ b/pcsx2/Linux/HacksDlg.cpp @@ -79,6 +79,9 @@ void on_Speed_Hacks(GtkMenuItem *menuitem, gpointer user_data) set_checked(SpeedHacksDlg, "check_wait_cycles_sync_hack", Config.Hacks.WaitCycleExt); set_checked(SpeedHacksDlg, "check_intc_sync_hack", Config.Hacks.INTCSTATSlow); set_checked(SpeedHacksDlg, "check_idle_loop_fastforward", Config.Hacks.IdleLoopFF); + set_checked(SpeedHacksDlg, "check_microvu_flag_hack_1", Config.Hacks.vuFlagHack1); + set_checked(SpeedHacksDlg, "check_microvu_flag_hack_2", Config.Hacks.vuFlagHack2); + set_checked(SpeedHacksDlg, "check_microvu_min_max_hack", Config.Hacks.vuMinMax); gtk_range_set_value(vuScale, Config.Hacks.VUCycleSteal); on_vu_slider_changed(vuScale, NULL); @@ -101,6 +104,9 @@ void on_Speed_Hack_OK(GtkButton *button, gpointer user_data) newhacks.WaitCycleExt = is_checked(SpeedHacksDlg, "check_wait_cycles_sync_hack"); newhacks.INTCSTATSlow = is_checked(SpeedHacksDlg, "check_intc_sync_hack"); newhacks.IdleLoopFF = is_checked(SpeedHacksDlg, "check_idle_loop_fastforward"); + newhacks.vuFlagHack1 = is_checked(SpeedHacksDlg, "check_microvu_flag_hack_1"); + newhacks.vuFlagHack2 = is_checked(SpeedHacksDlg, "check_microvu_flag_hack_2"); + newhacks.vuMinMax = is_checked(SpeedHacksDlg, "check_microvu_min_max_hack"); newhacks.VUCycleSteal = gtk_range_get_value(GTK_RANGE(lookup_widget(SpeedHacksDlg, "VUCycleHackScale"))); newhacks.EECycleRate = gtk_range_get_value(GTK_RANGE(lookup_widget(SpeedHacksDlg, "EECycleHackScale"))); diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index 5c1fcf9ed7..0ed474d4b2 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -644,9 +644,19 @@ create_SpeedHacksDlg (void) GtkWidget *check_idle_loop_fastforward; GtkWidget *label110; GtkWidget *hseparator1; + GtkWidget *frame41; + GtkWidget *alignment36; + GtkWidget *vbox75; + GtkWidget *check_microvu_flag_hack_1; + GtkWidget *check_microvu_flag_hack_2; + GtkWidget *check_microvu_min_max_hack; + GtkWidget *label113; GtkWidget *dialog_action_area3; GtkWidget *button99; GtkWidget *button98; + GtkTooltips *tooltips; + + tooltips = gtk_tooltips_new (); SpeedHacksDlg = gtk_dialog_new (); gtk_widget_set_name (SpeedHacksDlg, "SpeedHacksDlg"); @@ -717,7 +727,7 @@ create_SpeedHacksDlg (void) gtk_box_pack_start (GTK_BOX (vbox61), label91, FALSE, FALSE, 0); gtk_label_set_line_wrap (GTK_LABEL (label91), TRUE); - label105 = gtk_label_new (_("EmotionEngine (EE) Sync Hacks")); + label105 = gtk_label_new (_("EmotionEngine (EE) Sync Hacks")); gtk_widget_set_name (label105, "label105"); gtk_widget_show (label105); gtk_frame_set_label_widget (GTK_FRAME (frame37), label105); @@ -752,7 +762,7 @@ create_SpeedHacksDlg (void) gtk_box_pack_start (GTK_BOX (vbox73), vu_cycle_stealing_label, FALSE, FALSE, 0); gtk_label_set_line_wrap (GTK_LABEL (vu_cycle_stealing_label), TRUE); - label111 = gtk_label_new (_("VU Cycle Stealing (Speedup for 3d geometry)")); + label111 = gtk_label_new (_("VU Cycle Stealing (Speedup for 3d geometry)")); gtk_widget_set_name (label111, "label111"); gtk_widget_show (label111); gtk_frame_set_label_widget (GTK_FRAME (frame39), label111); @@ -767,8 +777,9 @@ create_SpeedHacksDlg (void) gtk_widget_set_name (check_iop_cycle_rate, "check_iop_cycle_rate"); gtk_widget_show (check_iop_cycle_rate); gtk_box_pack_start (GTK_BOX (vbox60), check_iop_cycle_rate, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_iop_cycle_rate, _("Small speedup, and works well with most games."), NULL); - label96 = gtk_label_new (_("Small speedup, and works well with most games,")); + label96 = gtk_label_new (_("Small speedup, and works well with most games.")); gtk_widget_set_name (label96, "label96"); gtk_widget_show (label96); gtk_box_pack_start (GTK_BOX (vbox60), label96, FALSE, FALSE, 0); @@ -778,6 +789,7 @@ create_SpeedHacksDlg (void) gtk_widget_set_name (check_wait_cycles_sync_hack, "check_wait_cycles_sync_hack"); gtk_widget_show (check_wait_cycles_sync_hack); gtk_box_pack_start (GTK_BOX (vbox60), check_wait_cycles_sync_hack, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_wait_cycles_sync_hack, _("Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes."), NULL); label97 = gtk_label_new (_("Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes.")); gtk_widget_set_name (label97, "label97"); @@ -789,6 +801,7 @@ create_SpeedHacksDlg (void) gtk_widget_set_name (check_intc_sync_hack, "check_intc_sync_hack"); gtk_widget_show (check_intc_sync_hack); gtk_box_pack_start (GTK_BOX (vbox60), check_intc_sync_hack, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_intc_sync_hack, _("Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks)."), NULL); label101 = gtk_label_new (_("Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks).")); gtk_widget_set_name (label101, "label101"); @@ -805,8 +818,9 @@ create_SpeedHacksDlg (void) gtk_widget_set_name (check_idle_loop_fastforward, "check_idle_loop_fastforward"); gtk_widget_show (check_idle_loop_fastforward); gtk_box_pack_start (GTK_BOX (vbox71), check_idle_loop_fastforward, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_idle_loop_fastforward, _("Speedup for a few games, including FFX with no known side effects."), NULL); - label110 = gtk_label_new (_("Speedup for a few games, including FFX with no known side effects. More later.")); + label110 = gtk_label_new (_("Speedup for a few games, including FFX with no known side effects.")); gtk_widget_set_name (label110, "label110"); gtk_widget_show (label110); gtk_box_pack_start (GTK_BOX (vbox71), label110, FALSE, FALSE, 0); @@ -817,6 +831,46 @@ create_SpeedHacksDlg (void) gtk_widget_show (hseparator1); gtk_box_pack_start (GTK_BOX (vbox60), hseparator1, FALSE, FALSE, 0); + frame41 = gtk_frame_new (NULL); + gtk_widget_set_name (frame41, "frame41"); + gtk_widget_show (frame41); + gtk_box_pack_start (GTK_BOX (dialog_vbox3), frame41, TRUE, TRUE, 0); + + alignment36 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment36, "alignment36"); + gtk_widget_show (alignment36); + gtk_container_add (GTK_CONTAINER (frame41), alignment36); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment36), 0, 0, 12, 0); + + vbox75 = gtk_vbox_new (FALSE, 0); + gtk_widget_set_name (vbox75, "vbox75"); + gtk_widget_show (vbox75); + gtk_container_add (GTK_CONTAINER (alignment36), vbox75); + + check_microvu_flag_hack_1 = gtk_check_button_new_with_mnemonic (_("Flag Hack 1 (microVU only)")); + gtk_widget_set_name (check_microvu_flag_hack_1, "check_microvu_flag_hack_1"); + gtk_widget_show (check_microvu_flag_hack_1); + gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_flag_hack_1, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_microvu_flag_hack_1, _("Big Speedup and moderately compatible. May cause SPS and Infinite loop problems."), NULL); + + check_microvu_flag_hack_2 = gtk_check_button_new_with_mnemonic (_("Flag Hack 2 (microVU only)")); + gtk_widget_set_name (check_microvu_flag_hack_2, "check_microvu_flag_hack_2"); + gtk_widget_show (check_microvu_flag_hack_2); + gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_flag_hack_2, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_microvu_flag_hack_2, _("Big Speedup and moderately compatible. SuperVU does something similar by default."), NULL); + + check_microvu_min_max_hack = gtk_check_button_new_with_mnemonic (_("Min / Max Hack (microVU only)")); + gtk_widget_set_name (check_microvu_min_max_hack, "check_microvu_min_max_hack"); + gtk_widget_show (check_microvu_min_max_hack); + gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_min_max_hack, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_microvu_min_max_hack, _("Small Speedup. May cause SPS, hanging, and missing geometry."), NULL); + + label113 = gtk_label_new (_("Vertex Unit (VU) Hacks")); + gtk_widget_set_name (label113, "label113"); + gtk_widget_show (label113); + gtk_frame_set_label_widget (GTK_FRAME (frame41), label113); + gtk_label_set_use_markup (GTK_LABEL (label113), TRUE); + dialog_action_area3 = GTK_DIALOG (SpeedHacksDlg)->action_area; gtk_widget_set_name (dialog_action_area3, "dialog_action_area3"); gtk_widget_show (dialog_action_area3); @@ -882,9 +936,17 @@ create_SpeedHacksDlg (void) GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_idle_loop_fastforward, "check_idle_loop_fastforward"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label110, "label110"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, hseparator1, "hseparator1"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, frame41, "frame41"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, alignment36, "alignment36"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox75, "vbox75"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_flag_hack_1, "check_microvu_flag_hack_1"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_flag_hack_2, "check_microvu_flag_hack_2"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_min_max_hack, "check_microvu_min_max_hack"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label113, "label113"); GLADE_HOOKUP_OBJECT_NO_REF (SpeedHacksDlg, dialog_action_area3, "dialog_action_area3"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, button99, "button99"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, button98, "button98"); + GLADE_HOOKUP_OBJECT_NO_REF (SpeedHacksDlg, tooltips, "tooltips"); return SpeedHacksDlg; } @@ -3859,6 +3921,12 @@ create_CpuDlg (void) GtkWidget *GtkCheckButton_VU0rec; GtkWidget *GtkCheckButton_VU1rec; GtkWidget *label32; + GtkWidget *frame40; + GtkWidget *alignment35; + GtkWidget *vbox74; + GtkWidget *GtkCheckButton_microVU0rec; + GtkWidget *GtkCheckButton_microVU1rec; + GtkWidget *label112; GtkWidget *GtkCheckButton_MTGS; GtkWidget *vbox48; GtkWidget *frame9; @@ -3984,11 +4052,44 @@ create_CpuDlg (void) gtk_widget_show (GtkCheckButton_VU1rec); gtk_box_pack_start (GTK_BOX (vbox26), GtkCheckButton_VU1rec, FALSE, FALSE, 0); - label32 = gtk_label_new (_("VU Recompilers - All options are set by default")); + label32 = gtk_label_new (_("VU Recompilers")); gtk_widget_set_name (label32, "label32"); gtk_widget_show (label32); gtk_frame_set_label_widget (GTK_FRAME (frame6), label32); + frame40 = gtk_frame_new (NULL); + gtk_widget_set_name (frame40, "frame40"); + gtk_widget_show (frame40); + gtk_box_pack_start (GTK_BOX (vbox47), frame40, FALSE, FALSE, 0); + gtk_container_set_border_width (GTK_CONTAINER (frame40), 5); + + alignment35 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment35, "alignment35"); + gtk_widget_show (alignment35); + gtk_container_add (GTK_CONTAINER (frame40), alignment35); + gtk_container_set_border_width (GTK_CONTAINER (alignment35), 5); + + vbox74 = gtk_vbox_new (FALSE, 0); + gtk_widget_set_name (vbox74, "vbox74"); + gtk_widget_show (vbox74); + gtk_container_add (GTK_CONTAINER (alignment35), vbox74); + + GtkCheckButton_microVU0rec = gtk_check_button_new_with_mnemonic (_("microVU0 - Use microVU rec instead of VU0rec. ! Beta !")); + gtk_widget_set_name (GtkCheckButton_microVU0rec, "GtkCheckButton_microVU0rec"); + gtk_widget_show (GtkCheckButton_microVU0rec); + gtk_box_pack_start (GTK_BOX (vbox74), GtkCheckButton_microVU0rec, FALSE, FALSE, 0); + + GtkCheckButton_microVU1rec = gtk_check_button_new_with_mnemonic (_("microVU1 - Use microVU rec instead of VU1rec. ! Beta !")); + gtk_widget_set_name (GtkCheckButton_microVU1rec, "GtkCheckButton_microVU1rec"); + gtk_widget_show (GtkCheckButton_microVU1rec); + gtk_box_pack_start (GTK_BOX (vbox74), GtkCheckButton_microVU1rec, FALSE, FALSE, 0); + + label112 = gtk_label_new (_("microVU Recompilers - ! Beta !")); + gtk_widget_set_name (label112, "label112"); + gtk_widget_show (label112); + gtk_frame_set_label_widget (GTK_FRAME (frame40), label112); + gtk_label_set_use_markup (GTK_LABEL (label112), TRUE); + GtkCheckButton_MTGS = gtk_check_button_new_with_mnemonic (_("Multi threaded GS mode (MTGS)\n (faster on dual core/HT CPUs, requires pcsx2 restart)")); gtk_widget_set_name (GtkCheckButton_MTGS, "GtkCheckButton_MTGS"); gtk_widget_show (GtkCheckButton_MTGS); @@ -4194,6 +4295,12 @@ create_CpuDlg (void) GLADE_HOOKUP_OBJECT (CpuDlg, GtkCheckButton_VU0rec, "GtkCheckButton_VU0rec"); GLADE_HOOKUP_OBJECT (CpuDlg, GtkCheckButton_VU1rec, "GtkCheckButton_VU1rec"); GLADE_HOOKUP_OBJECT (CpuDlg, label32, "label32"); + GLADE_HOOKUP_OBJECT (CpuDlg, frame40, "frame40"); + GLADE_HOOKUP_OBJECT (CpuDlg, alignment35, "alignment35"); + GLADE_HOOKUP_OBJECT (CpuDlg, vbox74, "vbox74"); + GLADE_HOOKUP_OBJECT (CpuDlg, GtkCheckButton_microVU0rec, "GtkCheckButton_microVU0rec"); + GLADE_HOOKUP_OBJECT (CpuDlg, GtkCheckButton_microVU1rec, "GtkCheckButton_microVU1rec"); + GLADE_HOOKUP_OBJECT (CpuDlg, label112, "label112"); GLADE_HOOKUP_OBJECT (CpuDlg, GtkCheckButton_MTGS, "GtkCheckButton_MTGS"); GLADE_HOOKUP_OBJECT (CpuDlg, vbox48, "vbox48"); GLADE_HOOKUP_OBJECT (CpuDlg, frame9, "frame9"); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index 397d21bb1b..59dad5cd05 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -1394,7 +1394,7 @@ If you have problems, Disable all of these and try again. True - <b>EmotionEngine (EE) Sync Hacks</b> + EmotionEngine (EE) Sync Hacks False True GTK_JUSTIFY_LEFT @@ -1497,7 +1497,7 @@ If you have problems, Disable all of these and try again. True - <b>VU Cycle Stealing (Speedup for 3d geometry)</b> + VU Cycle Stealing (Speedup for 3d geometry) False True GTK_JUSTIFY_LEFT @@ -1540,6 +1540,7 @@ If you have problems, Disable all of these and try again. True + Small speedup, and works well with most games. True Enable IOP x2 Cycle Rate True @@ -1559,7 +1560,7 @@ If you have problems, Disable all of these and try again. True - Small speedup, and works well with most games, + Small speedup, and works well with most games. False False GTK_JUSTIFY_LEFT @@ -1585,6 +1586,7 @@ If you have problems, Disable all of these and try again. True + Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes. True WaitCycles Sync Hack True @@ -1630,6 +1632,7 @@ If you have problems, Disable all of these and try again. True + Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks). True INTC Sync Hack True @@ -1680,6 +1683,7 @@ If you have problems, Disable all of these and try again. True + Speedup for a few games, including FFX with no known side effects. True Idle Loop Fast-Forward (experimental) True @@ -1699,7 +1703,7 @@ If you have problems, Disable all of these and try again. True - Speedup for a few games, including FFX with no known side effects. More later. + Speedup for a few games, including FFX with no known side effects. False False GTK_JUSTIFY_LEFT @@ -1759,6 +1763,125 @@ If you have problems, Disable all of these and try again. False + + + + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + False + 0 + + + + True + Big Speedup and moderately compatible. May cause SPS and Infinite loop problems. + True + Flag Hack 1 (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Big Speedup and moderately compatible. SuperVU does something similar by default. + True + Flag Hack 2 (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Small Speedup. May cause SPS, hanging, and missing geometry. + True + Min / Max Hack (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + + + + + True + Vertex Unit (VU) Hacks + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + True + True + + @@ -6718,7 +6841,7 @@ Version x.x True - VU Recompilers - All options are set by default + VU Recompilers False False GTK_JUSTIFY_LEFT @@ -6745,6 +6868,105 @@ Version x.x + + + 5 + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + 5 + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 0 + 0 + + + + True + False + 0 + + + + True + True + microVU0 - Use microVU rec instead of VU0rec. ! Beta ! + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + True + microVU1 - Use microVU rec instead of VU1rec. ! Beta ! + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + + + + + True + microVU Recompilers - ! Beta ! + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + + + + 0 + False + False + + + True diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 4dbb9c6e3e..105f6a61e5 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -343,7 +343,12 @@ void SuperVUAlloc(int vuindex) { // upper 4 bits must be zero! // Changed "first try base" to 0xb800000, since 0x0c000000 liked to fail a lot. (air) + // With microVU as optional now, Linux fails with 0x0e000000, and prefers 0x0c000000. --arcum42 +#ifdef __LINUX__ + s_recVUMem = SysMmapEx(0x0c000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); +#else s_recVUMem = SysMmapEx(0x0e000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); +#endif if (s_recVUMem == NULL) { diff --git a/plugins/GSnull/GifTransfer.cpp b/plugins/GSnull/GifTransfer.cpp index 34e69fa3bb..1d81988b07 100644 --- a/plugins/GSnull/GifTransfer.cpp +++ b/plugins/GSnull/GifTransfer.cpp @@ -37,8 +37,6 @@ PCSX2_ALIGNED16( u8 g_RealGSMem[0x2000] ); static void RegHandlerSIGNAL(const u32* data) { - //MTGS_LOG("MTGS SIGNAL data %x_%x CSRw %x\n",data[0], data[1], CSRw); - GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); if ((CSRw & 0x1)) GSCSRr |= 1; // signal @@ -48,8 +46,6 @@ static void RegHandlerSIGNAL(const u32* data) static void RegHandlerFINISH(const u32* data) { - //MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); - if ((CSRw & 0x2)) GSCSRr |= 2; // finish if (!(GSIMR & 0x200) ) GSirq(); diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index 40e058721a..959f65fedc 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -29,142 +29,150 @@ using namespace std; -HINSTANCE hInst=NULL; +HINSTANCE hInst = NULL; static pthread_spinlock_t s_mutexStatus; static u32 s_keyPress[2], s_keyRelease[2]; extern u16 status[2]; extern string s_strIniPath; LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); -WNDPROC GSwndProc=NULL; -HWND GShwnd=NULL; +WNDPROC GSwndProc = NULL; +HWND GShwnd = NULL; extern keyEvent event; void SaveConfig() { - char *szTemp; - char szIniFile[256], szValue[256], szProf[256]; - int i, j; + char *szTemp; + char szIniFile[256], szValue[256], szProf[256]; + int i, j; - GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); - szTemp = strrchr(szIniFile, '\\'); + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); - if(!szTemp) return; - strcpy(szTemp, "\\inis\\zeropad.ini"); + if (!szTemp) return; + strcpy(szTemp, "\\inis\\zeropad.ini"); - for (j=0; j<2; j++) { - for (i=0; i=0x60 && key<=0x69) { - sprintf(buff, "NumPad %c", '0' + key - 0x60); - } - else sprintf(buff, "%c", key); - } - } - else if (key >= 0x1000 && key < 0x2000) - { - sprintf (buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); - } - else if (key >= 0x2000 && key < 0x3000) - { - static const char name[][4] = { "MIN", "MAX" }; - const int axis = (key & 0xff); - sprintf (buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); - if (index >= 17 && index <= 20) - buff[strlen (buff) -4] = '\0'; - } - else if (key >= 0x3000 && key < 0x4000) - { - static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; - const int pov = (key & 0xff); - sprintf (buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov /4, name[pov % 4]); - } + const int key = conf.keys[pad][index]; + char buff[16] = "NONE)"; + if (key < 0x100) + { + if (key == 0) + strcpy(buff, "NONE"); + else + { + if (key >= 0x60 && key <= 0x69) + { + sprintf(buff, "NumPad %c", '0' + key - 0x60); + } + else sprintf(buff, "%c", key); + } + } + else if (key >= 0x1000 && key < 0x2000) + { + sprintf(buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); + } + else if (key >= 0x2000 && key < 0x3000) + { + static const char name[][4] = { "MIN", "MAX" }; + const int axis = (key & 0xff); + sprintf(buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); + if (index >= 17 && index <= 20) + buff[strlen(buff) -4] = '\0'; + } + else if (key >= 0x3000 && key < 0x4000) + { + static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; + const int pov = (key & 0xff); + sprintf(buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov / 4, name[pov % 4]); + } - return buff; + return buff; } -BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { - HWND hWC; - TCITEM tcI; - int i,key, numkeys; - u8* pkeyboard; - static int disabled=0; - static int padn=0; +BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) +{ + HWND hWC; + TCITEM tcI; + int i, key, numkeys; + u8* pkeyboard; + static int disabled = 0; + static int padn = 0; - switch(uMsg) { - case WM_INITDIALOG: - LoadConfig(); - padn = 0; - if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); + switch (uMsg) + { + case WM_INITDIALOG: + LoadConfig(); + padn = 0; + if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); - for (i=0; i Date: Wed, 27 May 2009 14:27:50 +0000 Subject: [PATCH 140/296] mVU: Disabled breaking execution on bad programs since it's not needed anymore and costs speed. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1276 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU0micro.cpp | 1 + pcsx2/x86/iVU1micro.cpp | 2 ++ pcsx2/x86/microVU_Compile.inl | 3 ++- 3 files changed, 5 insertions(+), 1 deletion(-) diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index cb85aa442b..770e9f4d04 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -47,6 +47,7 @@ namespace VU0micro if ((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); + //Currently breaking mVU execution is disabled. Check mVUtestCycles() in microVU_Compile.inl if (useMVU0) runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); else SuperVUExecuteProgram(VU0.VI[REG_TPC].UL & 0xfff, 0); FreezeXMMRegs(0); diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 8e3f704a75..398a61769b 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -145,6 +145,7 @@ namespace VU1micro memcpy_fast((u8*)&VU1, (u8*)backVUregs, sizeof(VURegs)); memcpy_fast((u8*)VU1.Mem, (u8*)backVUmem, 0x4000); + //Currently breaking mVU execution is disabled. Check mVUtestCycles() in microVU_Compile.inl runVUrec(VU1.VI[REG_TPC].UL, 300000 /*0x7fffffff*/, 1); if ((memcmp((u8*)cmpVUregs, (u8*)&VU1, (16*32) + (16*16))) || (memcmp((u8*)cmpVUmem, (u8*)VU1.Mem, 0x4000))) { @@ -274,6 +275,7 @@ namespace VU1micro #endif FreezeXMMRegs(1); + //Currently breaking mVU execution is disabled. Check mVUtestCycles() in microVU_Compile.inl if (useMVU1) runVUrec(VU1.VI[REG_TPC].UL, 3000000, 1); else { if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 1a0a44e77d..1e743a057e 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -266,7 +266,8 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { // Sets Up Flag instances int xStatus[4], xMac[4], xClip[4]; int xCycles = mVUsetFlags(xStatus, xMac, xClip); - mVUtestCycles(); + + //mVUtestCycles(); //uncomment to re-enable breaking on bad programms, costs a few fps // Second Pass iPC = mVUstartPC; From a1cd98bc596e2ec29443255347124166a0f7f31a Mon Sep 17 00:00:00 2001 From: arcum42 Date: Wed, 27 May 2009 20:20:33 +0000 Subject: [PATCH 141/296] Linux: Revise the speed hack and cpu dialogs a bit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1277 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/CpuDlg.cpp | 66 ++++-- pcsx2/Linux/callbacks.h | 4 + pcsx2/Linux/interface.c | 100 ++++---- pcsx2/Linux/pcsx2.glade | 509 ++++++++++++++++++---------------------- 4 files changed, 315 insertions(+), 364 deletions(-) diff --git a/pcsx2/Linux/CpuDlg.cpp b/pcsx2/Linux/CpuDlg.cpp index 8982c63fc1..e9be522c1d 100644 --- a/pcsx2/Linux/CpuDlg.cpp +++ b/pcsx2/Linux/CpuDlg.cpp @@ -25,23 +25,18 @@ void OnCpu_Ok(GtkButton *button, gpointer user_data) { u32 newopts = 0; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_EERec")))) - newopts |= PCSX2_EEREC; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU0rec")))) - newopts |= PCSX2_VU0REC; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU1rec")))) - newopts |= PCSX2_VU1REC; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec")))) - newopts |= PCSX2_MICROVU0; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec")))) - newopts |= PCSX2_MICROVU1; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_MTGS")))) - newopts |= PCSX2_GSMULTITHREAD; - if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitNormal")))) + if is_checked(CpuDlg, "GtkCheckButton_EERec") newopts |= PCSX2_EEREC; + if is_checked(CpuDlg, "GtkCheckButton_VU0rec") newopts |= PCSX2_VU0REC; + if is_checked(CpuDlg, "GtkCheckButton_VU1rec") newopts |= PCSX2_VU1REC; + if is_checked(CpuDlg, "GtkCheckButton_microVU0rec") newopts |= PCSX2_MICROVU0; + if is_checked(CpuDlg, "GtkCheckButton_microVU1rec") newopts |= PCSX2_MICROVU1; + if is_checked(CpuDlg, "GtkCheckButton_MTGS") newopts |= PCSX2_GSMULTITHREAD; + + if is_checked(CpuDlg, "GtkRadioButton_LimitNormal") newopts |= PCSX2_FRAMELIMIT_NORMAL; - else if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitLimit")))) + else if is_checked(CpuDlg, "GtkRadioButton_LimitLimit") newopts |= PCSX2_FRAMELIMIT_LIMIT; - else if (gtk_toggle_button_get_active(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitFS")))) + else if is_checked(CpuDlg, "GtkRadioButton_LimitFS") newopts |= PCSX2_FRAMELIMIT_SKIP; Config.CustomFps = gtk_spin_button_get_value_as_int(GTK_SPIN_BUTTON(lookup_widget(CpuDlg, "CustomFPSLimit"))); @@ -76,6 +71,19 @@ void OnCpu_Ok(GtkButton *button, gpointer user_data) gtk_main_quit(); } +void DisableMicroVU() +{ + if is_checked(CpuDlg, "GtkCheckButton_VU0rec") + gtk_widget_set_sensitive(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec"), true); + else + gtk_widget_set_sensitive(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec"), false); + + if is_checked(CpuDlg, "GtkCheckButton_VU1rec") + gtk_widget_set_sensitive(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec"), true); + else + gtk_widget_set_sensitive(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec"), false); +} + void OnConf_Cpu(GtkMenuItem *menuitem, gpointer user_data) { char str[512]; @@ -83,16 +91,17 @@ void OnConf_Cpu(GtkMenuItem *menuitem, gpointer user_data) CpuDlg = create_CpuDlg(); gtk_window_set_title(GTK_WINDOW(CpuDlg), _("Configuration")); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_EERec")), !!CHECK_EEREC); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU0rec")), !!CHECK_VU0REC); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_VU1rec")), !!CHECK_VU1REC); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU0rec")), !!CHECK_MICROVU0); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_microVU1rec")), !!CHECK_MICROVU1); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkCheckButton_MTGS")), !!CHECK_MULTIGS); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitNormal")), CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_NORMAL); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitLimit")), CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_LIMIT); - gtk_toggle_button_set_state(GTK_TOGGLE_BUTTON(lookup_widget(CpuDlg, "GtkRadioButton_LimitFS")), CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_SKIP); - + set_checked(CpuDlg, "GtkCheckButton_EERec", !!CHECK_EEREC); + set_checked(CpuDlg, "GtkCheckButton_VU0rec", !!CHECK_VU0REC); + set_checked(CpuDlg, "GtkCheckButton_VU1rec", !!CHECK_VU1REC); + set_checked(CpuDlg, "GtkCheckButton_microVU0rec", !!CHECK_MICROVU0); + set_checked(CpuDlg, "GtkCheckButton_microVU1rec", !!CHECK_MICROVU1); + set_checked(CpuDlg, "GtkCheckButton_MTGS", !!CHECK_MULTIGS); + set_checked(CpuDlg, "GtkRadioButton_LimitNormal", CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_NORMAL); + set_checked(CpuDlg, "GtkRadioButton_LimitLimit", CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_LIMIT); + set_checked(CpuDlg, "GtkRadioButton_LimitFS", CHECK_FRAMELIMIT == PCSX2_FRAMELIMIT_SKIP); + DisableMicroVU(); + sprintf(str, "Cpu Vendor: %s", cpuinfo.x86ID); gtk_label_set_text(GTK_LABEL(lookup_widget(CpuDlg, "GtkLabel_CpuVendor")), str); sprintf(str, "Familly: %s", cpuinfo.x86Fam); @@ -116,4 +125,9 @@ void OnConf_Cpu(GtkMenuItem *menuitem, gpointer user_data) gtk_widget_show_all(CpuDlg); if (MainWindow) gtk_widget_set_sensitive(MainWindow, FALSE); gtk_main(); -} \ No newline at end of file +} + +void OnCpuCheckToggled(GtkToggleButton *togglebutton, gpointer user_data) +{ + DisableMicroVU(); +} diff --git a/pcsx2/Linux/callbacks.h b/pcsx2/Linux/callbacks.h index 7e20155d17..5cc7803634 100644 --- a/pcsx2/Linux/callbacks.h +++ b/pcsx2/Linux/callbacks.h @@ -261,3 +261,7 @@ OnMemcards_Ok (GtkButton *button, void On_Dialog_Close (GtkDialog *dialog, gpointer user_data); + +void +OnCpuCheckToggled (GtkToggleButton *togglebutton, + gpointer user_data); diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index 0ed474d4b2..94994df4bf 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -634,16 +634,14 @@ create_SpeedHacksDlg (void) GtkWidget *vu_cycle_stealing_label; GtkWidget *label111; GtkWidget *vbox60; + GtkWidget *frame42; + GtkWidget *alignment37; + GtkWidget *vbox76; GtkWidget *check_iop_cycle_rate; - GtkWidget *label96; GtkWidget *check_wait_cycles_sync_hack; - GtkWidget *label97; GtkWidget *check_intc_sync_hack; - GtkWidget *label101; - GtkWidget *vbox71; GtkWidget *check_idle_loop_fastforward; - GtkWidget *label110; - GtkWidget *hseparator1; + GtkWidget *label114; GtkWidget *frame41; GtkWidget *alignment36; GtkWidget *vbox75; @@ -773,68 +771,56 @@ create_SpeedHacksDlg (void) gtk_widget_show (vbox60); gtk_box_pack_start (GTK_BOX (hbox39), vbox60, FALSE, FALSE, 0); + frame42 = gtk_frame_new (NULL); + gtk_widget_set_name (frame42, "frame42"); + gtk_widget_show (frame42); + gtk_box_pack_start (GTK_BOX (vbox60), frame42, FALSE, FALSE, 0); + + alignment37 = gtk_alignment_new (0.5, 0.5, 1, 1); + gtk_widget_set_name (alignment37, "alignment37"); + gtk_widget_show (alignment37); + gtk_container_add (GTK_CONTAINER (frame42), alignment37); + gtk_alignment_set_padding (GTK_ALIGNMENT (alignment37), 0, 0, 12, 0); + + vbox76 = gtk_vbox_new (FALSE, 0); + gtk_widget_set_name (vbox76, "vbox76"); + gtk_widget_show (vbox76); + gtk_container_add (GTK_CONTAINER (alignment37), vbox76); + check_iop_cycle_rate = gtk_check_button_new_with_mnemonic (_("Enable IOP x2 Cycle Rate")); gtk_widget_set_name (check_iop_cycle_rate, "check_iop_cycle_rate"); gtk_widget_show (check_iop_cycle_rate); - gtk_box_pack_start (GTK_BOX (vbox60), check_iop_cycle_rate, FALSE, FALSE, 0); + gtk_box_pack_start (GTK_BOX (vbox76), check_iop_cycle_rate, FALSE, FALSE, 0); gtk_tooltips_set_tip (tooltips, check_iop_cycle_rate, _("Small speedup, and works well with most games."), NULL); - label96 = gtk_label_new (_("Small speedup, and works well with most games.")); - gtk_widget_set_name (label96, "label96"); - gtk_widget_show (label96); - gtk_box_pack_start (GTK_BOX (vbox60), label96, FALSE, FALSE, 0); - gtk_label_set_line_wrap (GTK_LABEL (label96), TRUE); - check_wait_cycles_sync_hack = gtk_check_button_new_with_mnemonic (_("WaitCycles Sync Hack")); gtk_widget_set_name (check_wait_cycles_sync_hack, "check_wait_cycles_sync_hack"); gtk_widget_show (check_wait_cycles_sync_hack); - gtk_box_pack_start (GTK_BOX (vbox60), check_wait_cycles_sync_hack, FALSE, FALSE, 0); + gtk_box_pack_start (GTK_BOX (vbox76), check_wait_cycles_sync_hack, FALSE, FALSE, 0); gtk_tooltips_set_tip (tooltips, check_wait_cycles_sync_hack, _("Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes."), NULL); - label97 = gtk_label_new (_("Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes.")); - gtk_widget_set_name (label97, "label97"); - gtk_widget_show (label97); - gtk_box_pack_start (GTK_BOX (vbox60), label97, FALSE, FALSE, 0); - gtk_label_set_line_wrap (GTK_LABEL (label97), TRUE); - check_intc_sync_hack = gtk_check_button_new_with_mnemonic (_("INTC Sync Hack")); gtk_widget_set_name (check_intc_sync_hack, "check_intc_sync_hack"); gtk_widget_show (check_intc_sync_hack); - gtk_box_pack_start (GTK_BOX (vbox60), check_intc_sync_hack, FALSE, FALSE, 0); + gtk_box_pack_start (GTK_BOX (vbox76), check_intc_sync_hack, FALSE, FALSE, 0); gtk_tooltips_set_tip (tooltips, check_intc_sync_hack, _("Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks)."), NULL); - label101 = gtk_label_new (_("Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks).")); - gtk_widget_set_name (label101, "label101"); - gtk_widget_show (label101); - gtk_box_pack_start (GTK_BOX (vbox60), label101, FALSE, FALSE, 0); - gtk_label_set_line_wrap (GTK_LABEL (label101), TRUE); - - vbox71 = gtk_vbox_new (FALSE, 0); - gtk_widget_set_name (vbox71, "vbox71"); - gtk_widget_show (vbox71); - gtk_box_pack_start (GTK_BOX (vbox60), vbox71, TRUE, TRUE, 0); - check_idle_loop_fastforward = gtk_check_button_new_with_mnemonic (_("Idle Loop Fast-Forward (experimental)")); gtk_widget_set_name (check_idle_loop_fastforward, "check_idle_loop_fastforward"); gtk_widget_show (check_idle_loop_fastforward); - gtk_box_pack_start (GTK_BOX (vbox71), check_idle_loop_fastforward, FALSE, FALSE, 0); + gtk_box_pack_start (GTK_BOX (vbox76), check_idle_loop_fastforward, FALSE, FALSE, 0); gtk_tooltips_set_tip (tooltips, check_idle_loop_fastforward, _("Speedup for a few games, including FFX with no known side effects."), NULL); - label110 = gtk_label_new (_("Speedup for a few games, including FFX with no known side effects.")); - gtk_widget_set_name (label110, "label110"); - gtk_widget_show (label110); - gtk_box_pack_start (GTK_BOX (vbox71), label110, FALSE, FALSE, 0); - gtk_label_set_line_wrap (GTK_LABEL (label110), TRUE); - - hseparator1 = gtk_hseparator_new (); - gtk_widget_set_name (hseparator1, "hseparator1"); - gtk_widget_show (hseparator1); - gtk_box_pack_start (GTK_BOX (vbox60), hseparator1, FALSE, FALSE, 0); + label114 = gtk_label_new (_("Misc Hacks")); + gtk_widget_set_name (label114, "label114"); + gtk_widget_show (label114); + gtk_frame_set_label_widget (GTK_FRAME (frame42), label114); + gtk_label_set_use_markup (GTK_LABEL (label114), TRUE); frame41 = gtk_frame_new (NULL); gtk_widget_set_name (frame41, "frame41"); gtk_widget_show (frame41); - gtk_box_pack_start (GTK_BOX (dialog_vbox3), frame41, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (vbox60), frame41, FALSE, FALSE, 0); alignment36 = gtk_alignment_new (0.5, 0.5, 1, 1); gtk_widget_set_name (alignment36, "alignment36"); @@ -926,16 +912,14 @@ create_SpeedHacksDlg (void) GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vu_cycle_stealing_label, "vu_cycle_stealing_label"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label111, "label111"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox60, "vbox60"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, frame42, "frame42"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, alignment37, "alignment37"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox76, "vbox76"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_iop_cycle_rate, "check_iop_cycle_rate"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label96, "label96"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_wait_cycles_sync_hack, "check_wait_cycles_sync_hack"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label97, "label97"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_intc_sync_hack, "check_intc_sync_hack"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label101, "label101"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox71, "vbox71"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_idle_loop_fastforward, "check_idle_loop_fastforward"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label110, "label110"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, hseparator1, "hseparator1"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label114, "label114"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, frame41, "frame41"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, alignment36, "alignment36"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox75, "vbox75"); @@ -4042,12 +4026,12 @@ create_CpuDlg (void) gtk_container_add (GTK_CONTAINER (frame6), vbox26); gtk_container_set_border_width (GTK_CONTAINER (vbox26), 5); - GtkCheckButton_VU0rec = gtk_check_button_new_with_mnemonic (_("VU0rec - enable recompiler for VU0 unit")); + GtkCheckButton_VU0rec = gtk_check_button_new_with_mnemonic (_("VU0rec - enable recompilers for VU0 unit")); gtk_widget_set_name (GtkCheckButton_VU0rec, "GtkCheckButton_VU0rec"); gtk_widget_show (GtkCheckButton_VU0rec); gtk_box_pack_start (GTK_BOX (vbox26), GtkCheckButton_VU0rec, FALSE, FALSE, 0); - GtkCheckButton_VU1rec = gtk_check_button_new_with_mnemonic (_("VU1rec - enable recompiler for VU1 unit")); + GtkCheckButton_VU1rec = gtk_check_button_new_with_mnemonic (_("VU1rec - enable recompilers for VU1 unit")); gtk_widget_set_name (GtkCheckButton_VU1rec, "GtkCheckButton_VU1rec"); gtk_widget_show (GtkCheckButton_VU1rec); gtk_box_pack_start (GTK_BOX (vbox26), GtkCheckButton_VU1rec, FALSE, FALSE, 0); @@ -4269,6 +4253,18 @@ create_CpuDlg (void) g_signal_connect ((gpointer) CpuDlg, "close", G_CALLBACK (On_Dialog_Close), NULL); + g_signal_connect ((gpointer) GtkCheckButton_VU0rec, "toggled", + G_CALLBACK (OnCpuCheckToggled), + NULL); + g_signal_connect ((gpointer) GtkCheckButton_VU1rec, "toggled", + G_CALLBACK (OnCpuCheckToggled), + NULL); + g_signal_connect ((gpointer) GtkCheckButton_microVU0rec, "toggled", + G_CALLBACK (OnCpuCheckToggled), + NULL); + g_signal_connect ((gpointer) GtkCheckButton_microVU1rec, "toggled", + G_CALLBACK (OnCpuCheckToggled), + NULL); g_signal_connect ((gpointer) button96, "clicked", G_CALLBACK (OnCpu_Ok), NULL); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index 59dad5cd05..392c391957 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -1538,176 +1538,122 @@ If you have problems, Disable all of these and try again. 0 - + True - Small speedup, and works well with most games. - True - Enable IOP x2 Cycle Rate - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Small speedup, and works well with most games. - False - False - GTK_JUSTIFY_LEFT - True - False - 0.5 - 0.5 - 0 - 0 - check_iop_cycle_rate - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - 0 - False - False - - - - - - True - Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes. - True - WaitCycles Sync Hack - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes. - False - False - GTK_JUSTIFY_LEFT - True - False - 0.5 - 0.5 - 0 - 0 - check_wait_cycles_sync_hack - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - 0 - False - False - - - - - - True - Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks). - True - INTC Sync Hack - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks). - False - False - GTK_JUSTIFY_LEFT - True - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - 0 - False - False - - - - - - True - False - 0 + 0 + 0.5 + GTK_SHADOW_ETCHED_IN - + True - Speedup for a few games, including FFX with no known side effects. - True - Idle Loop Fast-Forward (experimental) - True - GTK_RELIEF_NORMAL - True - False - False - True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + False + 0 + + + + True + Small speedup, and works well with most games. + True + Enable IOP x2 Cycle Rate + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Small speedup. Works well with most games, but it may cause certain games to crash, or freeze up during bootup or stage changes. + True + WaitCycles Sync Hack + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks). + True + INTC Sync Hack + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Speedup for a few games, including FFX with no known side effects. + True + Idle Loop Fast-Forward (experimental) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + - - 0 - False - False - - + True - Speedup for a few games, including FFX with no known side effects. + Misc Hacks False - False + True GTK_JUSTIFY_LEFT - True + False False 0.5 0.5 @@ -1719,22 +1665,128 @@ If you have problems, Disable all of these and try again. 0 - 0 - False - False + label_item 0 - True - True + False + False - + True + 0 + 0.5 + GTK_SHADOW_ETCHED_IN + + + + True + 0.5 + 0.5 + 1 + 1 + 0 + 0 + 12 + 0 + + + + True + False + 0 + + + + True + Big Speedup and moderately compatible. May cause SPS and Infinite loop problems. + True + Flag Hack 1 (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Big Speedup and moderately compatible. SuperVU does something similar by default. + True + Flag Hack 2 (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + True + Small Speedup. May cause SPS, hanging, and missing geometry. + True + Min / Max Hack (microVU only) + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + + + + + + + + + True + Vertex Unit (VU) Hacks + False + True + GTK_JUSTIFY_LEFT + False + False + 0.5 + 0.5 + 0 + 0 + PANGO_ELLIPSIZE_NONE + -1 + False + 0 + + + label_item + + 0 @@ -1763,125 +1815,6 @@ If you have problems, Disable all of these and try again. False - - - - True - 0 - 0.5 - GTK_SHADOW_ETCHED_IN - - - - True - 0.5 - 0.5 - 1 - 1 - 0 - 0 - 12 - 0 - - - - True - False - 0 - - - - True - Big Speedup and moderately compatible. May cause SPS and Infinite loop problems. - True - Flag Hack 1 (microVU only) - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Big Speedup and moderately compatible. SuperVU does something similar by default. - True - Flag Hack 2 (microVU only) - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Small Speedup. May cause SPS, hanging, and missing geometry. - True - Min / Max Hack (microVU only) - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - - - - - True - Vertex Unit (VU) Hacks - False - True - GTK_JUSTIFY_LEFT - False - False - 0.5 - 0.5 - 0 - 0 - PANGO_ELLIPSIZE_NONE - -1 - False - 0 - - - label_item - - - - - 0 - True - True - - @@ -6802,13 +6735,14 @@ Version x.x True True - VU0rec - enable recompiler for VU0 unit + VU0rec - enable recompilers for VU0 unit True GTK_RELIEF_NORMAL True False False True + 0 @@ -6821,13 +6755,14 @@ Version x.x True True - VU1rec - enable recompiler for VU1 unit + VU1rec - enable recompilers for VU1 unit True GTK_RELIEF_NORMAL True False False True + 0 @@ -6906,6 +6841,7 @@ Version x.x False False True + 0 @@ -6925,6 +6861,7 @@ Version x.x False False True + 0 From 50332c1cd13fdeefb0aaa172826032517d0e5a77 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 28 May 2009 02:27:08 +0000 Subject: [PATCH 142/296] microVU: minor optimizations/changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1278 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU0micro.cpp | 1 - pcsx2/x86/iVU1micro.cpp | 1 - pcsx2/x86/microVU_Compile.inl | 6 ++---- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Upper.inl | 12 ++++++------ 5 files changed, 9 insertions(+), 13 deletions(-) diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index 770e9f4d04..cb85aa442b 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -47,7 +47,6 @@ namespace VU0micro if ((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); - //Currently breaking mVU execution is disabled. Check mVUtestCycles() in microVU_Compile.inl if (useMVU0) runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); else SuperVUExecuteProgram(VU0.VI[REG_TPC].UL & 0xfff, 0); FreezeXMMRegs(0); diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 398a61769b..9b55671f75 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -275,7 +275,6 @@ namespace VU1micro #endif FreezeXMMRegs(1); - //Currently breaking mVU execution is disabled. Check mVUtestCycles() in microVU_Compile.inl if (useMVU1) runVUrec(VU1.VI[REG_TPC].UL, 3000000, 1); else { if (VU1.VI[REG_TPC].UL >= VU1.maxmicro) { diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 1e743a057e..8f0a1c1bbb 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -192,7 +192,7 @@ microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; iPC = mVUstartPC; mVUdebugNOW(0); - CMP32ItoM((uptr)&mVU->cycles, 0); + SUB32ItoM((uptr)&mVU->cycles, mVUcycles); u8* jmp8 = JG8(0); MOV32ItoR(gprT2, xPC); if (!vuIndex) CALLFunc((uptr)mVUwarning0); @@ -200,7 +200,6 @@ microVUt(void) mVUtestCycles() { MOV32ItoR(gprR, Roffset); // Restore gprR mVUendProgram(0, 0, sI, 0, cI); x86SetJ8(jmp8); - SUB32ItoM((uptr)&mVU->cycles, mVUcycles); } //------------------------------------------------------------------ @@ -266,8 +265,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { // Sets Up Flag instances int xStatus[4], xMac[4], xClip[4]; int xCycles = mVUsetFlags(xStatus, xMac, xClip); - - //mVUtestCycles(); //uncomment to re-enable breaking on bad programms, costs a few fps + mVUtestCycles(); // Second Pass iPC = mVUstartPC; diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 060136f48b..8aac0115a1 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -52,7 +52,7 @@ microVUt(void) mVUstatusFlagOp() { } } iPC = curPC; - DevCon::Notice("microVU%d: FSSET Optimization", params vuIndex); + DevCon::Status("microVU%d: FSSET Optimization", params vuIndex); } int findFlagInst(int* fFlag, int cycles) { diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 160d044e89..70e5470cb8 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -55,23 +55,23 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are + if (doStatus) pjmp = JZ8(0); // Skip if none are if (doMac) SHL32ItoR(mReg, 4 + ADD_XYZW); if (doStatus) OR32ItoR(sReg, 0x82); // SS, S flags - if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - x86SetJ8(pjmp); + if (_XYZW_SS && doStatus) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + if (doStatus) x86SetJ8(pjmp); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are + if (doStatus) pjmp = JZ8(0); // Skip if none are if (doMac) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } if (doStatus) { OR32ItoR(sReg, 0x41); } // ZS, Z flags - x86SetJ8(pjmp); + if (doStatus) x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ - if (_XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here + if (_XYZW_SS && doStatus) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here if (doMac) mVUallocMFLAGb(mReg, fmInstance); // Set Mac Flag } From 4bc698d698241d70224435922c9309c28990fefe Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 28 May 2009 02:57:01 +0000 Subject: [PATCH 143/296] GSdx: just fixing slow texture uploads with dx9... (odin sphere back to 40-50 fps) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1279 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTexture.h | 2 +- plugins/GSdx/GSTexture10.cpp | 9 +++++++- plugins/GSdx/GSTexture10.h | 2 +- plugins/GSdx/GSTexture7.cpp | 9 +++++++- plugins/GSdx/GSTexture7.h | 2 +- plugins/GSdx/GSTexture9.cpp | 4 ++-- plugins/GSdx/GSTexture9.h | 2 +- plugins/GSdx/GSTextureCache.cpp | 39 +++++++++++++++++++++++++-------- plugins/GSdx/GSTextureNull.h | 2 +- plugins/GSdx/GSUtil.cpp | 4 +++- 10 files changed, 56 insertions(+), 19 deletions(-) diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index d21d778425..26de1ca667 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -40,7 +40,7 @@ public: virtual int GetHeight() const = 0; virtual int GetFormat() const = 0; virtual bool Update(const GSVector4i& r, const void* data, int pitch) = 0; - virtual bool Map(uint8** bits, int& pitch) = 0; + virtual bool Map(uint8** bits, int& pitch, const GSVector4i* r = NULL) = 0; virtual void Unmap() = 0; virtual bool Save(const string& fn, bool dds = false) = 0; diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 051cbb07fd..830a1d5878 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -83,8 +83,15 @@ bool GSTexture10::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture10::Map(uint8** bits, int& pitch) +bool GSTexture10::Map(uint8** bits, int& pitch, const GSVector4i* r) { + if(r != NULL) + { + // ASSERT(0); // not implemented + + return false; + } + if(m_texture) { D3D10_MAPPED_TEXTURE2D map; diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index e01ae272f2..e11233d48f 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -44,7 +44,7 @@ public: int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch); + bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture7.cpp b/plugins/GSdx/GSTexture7.cpp index fd2d2dcbef..28c5fafe81 100644 --- a/plugins/GSdx/GSTexture7.cpp +++ b/plugins/GSdx/GSTexture7.cpp @@ -133,10 +133,17 @@ bool GSTexture7::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture7::Map(uint8** bits, int& pitch) +bool GSTexture7::Map(uint8** bits, int& pitch, const GSVector4i* r) { HRESULT hr; + if(r != NULL) + { + // ASSERT(0); // not implemented + + return false; + } + DDSURFACEDESC2 desc; if(SUCCEEDED(hr = m_system->Lock(NULL, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR, NULL))) diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index fafccbd863..21baa62d95 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -44,7 +44,7 @@ public: int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch); + bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index e0eaa212b6..f4eb298a0c 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -113,7 +113,7 @@ bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture9::Map(uint8** bits, int& pitch) +bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) { HRESULT hr; @@ -121,7 +121,7 @@ bool GSTexture9::Map(uint8** bits, int& pitch) { D3DLOCKED_RECT lr; - if(SUCCEEDED(hr = surface->LockRect(&lr, NULL, 0))) + if(SUCCEEDED(hr = surface->LockRect(&lr, (LPRECT)r, 0))) { *bits = (uint8*)lr.pBits; pitch = (int)lr.Pitch; diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index dff127d4ac..f8779e3222 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -43,7 +43,7 @@ public: int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch); + bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 90f717fd8a..27b789ee0f 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -698,20 +698,41 @@ void GSTextureCache::GSCachedTexture::Update() m_valid = m_valid.runion(r); - static uint8* bits = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - - int pitch = ((r.width() + 3) & ~3) * 4; - - if(m_renderer->m_psrr) + uint8* bits = NULL; + int pitch = 0; + + if(m_texture->Map(&bits, pitch, &r)) { - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource + + if(m_renderer->m_psrr) + { + m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + else + { + m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + + m_texture->Unmap(); } else { - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } + static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); + + pitch = ((r.width() + 3) & ~3) * 4; - m_texture->Update(r, bits, pitch); + if(m_renderer->m_psrr) + { + m_renderer->m_mem.ReadTextureNPNC(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + else + { + m_renderer->m_mem.ReadTextureNP(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + } + + m_texture->Update(r, buff, pitch); + } m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * m_bpp >> 3); } diff --git a/plugins/GSdx/GSTextureNull.h b/plugins/GSdx/GSTextureNull.h index cc094240a8..ec3ada8357 100644 --- a/plugins/GSdx/GSTextureNull.h +++ b/plugins/GSdx/GSTextureNull.h @@ -38,7 +38,7 @@ public: int GetHeight() const {return m_desc.h;} int GetFormat() const {return m_desc.format;} bool Update(const GSVector4i& r, const void* data, int pitch) {return true;} - bool Map(uint8** bits, int& pitch) {return true;} + bool Map(uint8** bits, int& pitch, const GSVector4i* r) {return false;} void Unmap() {} bool Save(const string& fn, bool dds = false) {return false;} }; diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index 5656f48c45..620464f6ef 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -174,7 +174,9 @@ bool GSUtil::IsDirect3D10Available() char* GSUtil::GetLibName() { - static string str = format("GSdx %d", SVN_REV); + static string str; + + str = format("GSdx %d", SVN_REV); if(SVN_MODS) str += "m"; From c3ae30f25030cfad7a6d9c2c38e21feae183e775 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 29 May 2009 00:32:51 +0000 Subject: [PATCH 144/296] microVU: - Implemented the tri-ace gamefix. Note: I've only tested SO3 so-far and it has some ugly graphical problems with mVU. SO3 does some evil VU stuff that it shouldn't be doing, but it does them anyways xD I know the cause of the problems, so I just have to code the solutions sometime in the future... P.S. SO3 looks really funny with the FlagHack1 speedhack! The characters turn into floating heads effectively turning the game into PacMan xD git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1281 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Misc.inl | 75 +++++++++++++++++++++++++++++++++++++ pcsx2/x86/microVU_Upper.inl | 2 +- 2 files changed, 76 insertions(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index af3ff9608f..e9c44a78bf 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -369,3 +369,78 @@ void SSE_MIN2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { if (CHECK_VU_MINMAXHACK) { SSE_MINSS_XMM_to_XMM(to, from); } else { MIN_MAX_SS(to, from, 1); } } + +// Warning: Modifies all vectors in 'to' and 'from', and Modifies xmmT1 and xmmT2 +void SSE_ADD2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + + if (!CHECK_VUADDSUBHACK) return; + u8 *localptr[8]; + + SSE_MOVAPS_XMM_to_XMM(xmmT1, to); + SSE_MOVAPS_XMM_to_XMM(xmmT2, from); + SSE2_MOVD_XMM_to_R(gprT2, to); + SHR32ItoR(gprT2, 23); + SSE2_MOVD_XMM_to_R(gprT1, from); + SHR32ItoR(gprT1, 23); + AND32ItoR(gprT2, 0xff); + AND32ItoR(gprT1, 0xff); + SUB32RtoR(gprT2, gprT1); //gprT2 = exponent difference + + CMP32ItoR(gprT2, 25); + localptr[0] = JGE8(0); + CMP32ItoR(gprT2, 0); + localptr[1] = JG8(0); + localptr[2] = JE8(0); + CMP32ItoR(gprT2, -25); + localptr[3] = JLE8(0); + NEG32R(gprT2); + DEC32R(gprT2); + MOV32ItoR(gprT1, 0xffffffff); + SHL32CLtoR(gprT1); + SSE2_PCMPEQB_XMM_to_XMM(to, to); + SSE2_MOVD_R_to_XMM(from, gprT1); + SSE_MOVSS_XMM_to_XMM(to, from); + SSE2_PCMPEQB_XMM_to_XMM(from, from); + localptr[4] = JMP8(0); + + x86SetJ8(localptr[0]); + MOV32ItoR(gprT1, 0x80000000); + SSE2_PCMPEQB_XMM_to_XMM(from, from); + SSE2_MOVD_R_to_XMM(to, gprT1); + SSE_MOVSS_XMM_to_XMM(from, to); + SSE2_PCMPEQB_XMM_to_XMM(to, to); + localptr[5] = JMP8(0); + + x86SetJ8(localptr[1]); + DEC32R(gprT2); + MOV32ItoR(gprT1, 0xffffffff); + SHL32CLtoR(gprT1); + SSE2_PCMPEQB_XMM_to_XMM(from, from); + SSE2_MOVD_R_to_XMM(to, gprT1); + SSE_MOVSS_XMM_to_XMM(from, to); + SSE2_PCMPEQB_XMM_to_XMM(to, to); + localptr[6] = JMP8(0); + + x86SetJ8(localptr[3]); + MOV32ItoR(gprT1, 0x80000000); + SSE2_PCMPEQB_XMM_to_XMM(to, to); + SSE2_MOVD_R_to_XMM(from, gprT1); + SSE_MOVSS_XMM_to_XMM(to, from); + SSE2_PCMPEQB_XMM_to_XMM(from, from); + localptr[7] = JMP8(0); + + x86SetJ8(localptr[2]); + x86SetJ8(localptr[4]); + x86SetJ8(localptr[5]); + x86SetJ8(localptr[6]); + x86SetJ8(localptr[7]); + + SSE_ANDPS_XMM_to_XMM(to, xmmT1); //to contains mask + SSE_ANDPS_XMM_to_XMM(from, xmmT2); //from contains mask + SSE_ADDSS_XMM_to_XMM(to, from); +} + +// Note: Wrapper function, Tri-Ace Games just need the SS implementation +void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { + SSE_ADDPS_XMM_to_XMM(to, from); +} diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 70e5470cb8..5a7c2c6383 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -481,7 +481,7 @@ microVUf(void) mVU_ABS() { pass3 { mVUlog("ABS"); mVUlogFtFs(); } } microVUf(void) mVU_ADD() { mVU_FMAC1 (ADD, "ADD"); } -microVUf(void) mVU_ADDi() { mVU_FMAC6 (ADD, "ADDi"); } +microVUf(void) mVU_ADDi() { mVU_FMAC6 (ADD2, "ADDi"); } microVUf(void) mVU_ADDq() { mVU_FMAC22(ADD, "ADDq"); } microVUf(void) mVU_ADDx() { mVU_FMAC3 (ADD, "ADDx"); } microVUf(void) mVU_ADDy() { mVU_FMAC3 (ADD, "ADDy"); } From 12ca3898d615284193a22049816f7b088d3cdd5f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 29 May 2009 03:02:21 +0000 Subject: [PATCH 145/296] microVU: - Added a 1-cycle xgkick delay, fixes SO3 graphic corruption and flickering. Theres still VU stuff the game does that isn't implemented, but the game seems to run fine w/o the implementation (only tested the beginning though). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1282 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 1 + pcsx2/x86/microVU_Analyze.inl | 6 ++++++ pcsx2/x86/microVU_Compile.inl | 7 ++++++- pcsx2/x86/microVU_Lower.inl | 18 ++++++++++++------ pcsx2/x86/microVU_Misc.h | 6 ++++++ 5 files changed, 31 insertions(+), 7 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index c4c0dcdda6..e956a4e3c5 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -117,6 +117,7 @@ struct microVU { u32 code; // Contains the current Instruction u32 divFlag; // 1 instance of I/D flags u32 VIbackup[2]; // Holds a backup of a VI reg if modified before a branch + u32 VIxgkick; // Holds a backup of a VI reg used for xgkick-delays u32 branch; // Holds branch compare result (IBxx) OR Holds address to Jump to (JALR/JR) u32 p; // Holds current P instance index u32 q; // Holds current Q instance index diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 44eded0a77..cb32980690 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -354,6 +354,12 @@ microVUt(void) mVUanalyzeXGkick(int Fs, int xCycles) { analyzeVIreg1(Fs); analyzeXGkick1(); analyzeXGkick2(xCycles); + // Note: Technically XGKICK should stall on the next instruction, + // this code stalls on the same instruction. The only case where this + // will be a problem with, is if you have very-specifically placed + // FMxxx or FSxxx opcodes checking flags near this instruction AND + // the XGKICK instruction stalls. No-game should be effected by + // this minor difference. } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 8f0a1c1bbb..69f43f6ff4 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -125,8 +125,11 @@ microVUt(void) mVUincCycles(int x) { calcCycles(mVUregs.p, x); if (!mVUregs.p || (mVUregs.p && mVUregsTemp.p)) { incP(); } } + if (mVUregs.xgkick) { + calcCycles(mVUregs.xgkick, x); + if (!mVUregs.xgkick) { mVUinfo |= _doXGKICK; } + } calcCycles(mVUregs.r, x); - calcCycles(mVUregs.xgkick, x); } microVUt(void) mVUsetCycles() { @@ -277,6 +280,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } else { mVUopL(); incPC(1); doUpperOp(); } + if (doXGKICK) { mVU_XGKICK_DELAY(); } if (!isBdelay) { incPC(1); } else { @@ -376,6 +380,7 @@ eBitTemination: AND32ItoR (flagReg, 0x0fcf); OR32MtoR (flagReg, (uptr)&mVU->divFlag); } + if (doXGKICK) { mVU_XGKICK_DELAY(); } // Do E-bit end stuff here mVUsetupRange(xPC - 8); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 95a2784456..11de47832c 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1113,18 +1113,24 @@ void __fastcall mVU_XGKICK__(u32 addr) { microVUf(void) mVU_XGKICK() { microVU* mVU = mVUx; - pass1 { mVUanalyzeXGkick(_Is_, 4); } + pass1 { mVUanalyzeXGkick(_Is_, mVU_XGKICK_CYCLES); } pass2 { mVUprint("XGkick"); - mVUallocVIa(gprT2, _Is_); // gprT2 = ECX for __fastcall - mVUbackupRegs(); - if (mtgsThread != NULL) CALLFunc((uptr)mVU_XGKICK_); - else CALLFunc((uptr)mVU_XGKICK__); - mVUrestoreRegs(); + mVUallocVIa(gprT1, _Is_); + MOV32RtoM((uptr)&mVU->VIxgkick, gprT1); } pass3 { mVUlog("XGKICK vi%02d", _Fs_); } } +microVUt(void) mVU_XGKICK_DELAY() { + microVU* mVU = mVUx; + MOV32MtoR(gprT2, (uptr)&mVU->VIxgkick); // gprT2 = ECX for __fastcall + mVUbackupRegs(); + if (mtgsThread != NULL) CALLFunc((uptr)mVU_XGKICK_); + else CALLFunc((uptr)mVU_XGKICK__); + mVUrestoreRegs(); +} + //------------------------------------------------------------------ // Branches/Jumps //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 8832690bd4..37efbff089 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -191,6 +191,7 @@ declareAllVariables #define _isEOB (1<<2) // End of Block #define _isBdelay (1<<3) // Cur Instruction in Branch Delay slot #define _isSflag (1<<4) // Cur Instruction uses status flag +#define _doXGKICK (1<<5) // Do XGKICK transfer on this instruction #define _writeQ (1<<6) #define _readQ (1<<6) // same as writeQ #define _writeP (1<<7) @@ -222,6 +223,7 @@ declareAllVariables #define isEOB (mVUinfo & (1<<2)) #define isBdelay (mVUinfo & (1<<3)) #define isSflag (mVUinfo & (1<<4)) +#define doXGKICK (mVUinfo & (1<<5)) #define readQ ((mVUinfo >> 6) & 1) // same as writeQ #define readP ((mVUinfo >> 7) & 1) // same as writeP #define writeQ (((mVUinfo >> 6) + 1) & 1) @@ -295,6 +297,10 @@ declareAllVariables #define CHECK_VU_FLAGHACK2 (u32)Config.Hacks.vuFlagHack2 // (Can cause Infinite loops, SPS, etc...) #define CHECK_VU_MINMAXHACK (u32)Config.Hacks.vuMinMax // (Can cause SPS, Black Screens, etc...) +// Unknown Data +#define mVU_XGKICK_CYCLES 1 // Its unknown at recompile time how long the xgkick transfer will take + // so give it a value that makes games happy :) (SO3 is fine at 1 cycle delay) + // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ uptr diff = ptr - start; \ From 949c02ed457dbf13c3847a16072300c070af3b99 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 29 May 2009 04:18:09 +0000 Subject: [PATCH 146/296] microVU: - Small Optimization... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1283 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Flags.inl | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 8aac0115a1..379903e412 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -184,23 +184,32 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { } } -microVUt(void) mVUflagPass(int startPC) { +#define shortBranch() { \ + if (branch == 3) { \ + mVUflagPass(aBranchAddr, (xCount - (mVUcount+1))); \ + mVUcount = 4; \ + } \ +} + +// Scan through instructions and check if flags are read (FSxxx, FMxxx, FCxxx opcodes) +microVUx(void) mVUflagPass(u32 startPC, u32 xCount) { microVU* mVU = mVUx; int oldPC = iPC; int oldCount = mVUcount; int oldBranch = mVUbranch; + int aBranchAddr; iPC = startPC / 4; mVUcount = 0; mVUbranch = 0; - for (int branch = 0; mVUcount < 4; mVUcount++) { + for (int branch = 0; mVUcount < xCount; mVUcount++) { incPC(1); if ( curI & _Ebit_ ) { branch = 1; } if ( curI & _MDTbit_ ) { branch = 4; } if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(); incPC(1); } - if (branch >= 2) { break; } + if (branch >= 2) { shortBranch(); break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { branch = 3; mVUbranch = 0; } + if (mVUbranch) { branch = (mVUbranch >= 9) ? 5 : 3; aBranchAddr = branchAddr; mVUbranch = 0; } incPC(1); } if (mVUcount < 4) { mVUflagInfo |= 0xfff; } @@ -214,17 +223,18 @@ microVUt(void) mVUflagPass(int startPC) { #define branchType2 else if (mVUbranch >= 9) // JR/JALR #define branchType3 else // Conditional Branch +// Checks if the first 4 instructions of a block will read flags microVUt(void) mVUsetFlagInfo() { microVU* mVU = mVUx; - branchType1 { incPC(-1); mVUflagPass(branchAddr); incPC(1); } + branchType1 { incPC(-1); mVUflagPass(branchAddr, 4); incPC(1); } branchType2 { mVUflagInfo |= 0xfff; } branchType3 { incPC(-1); - mVUflagPass(branchAddr); + mVUflagPass(branchAddr, 4); int backupFlagInfo = mVUflagInfo; mVUflagInfo = 0; incPC(4); // Branch Not Taken - mVUflagPass(xPC); + mVUflagPass(xPC, 4); incPC(-3); mVUflagInfo |= backupFlagInfo; } From 7fef12a4cb261de19a0d5d98e8d0f9ed67868dd0 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 29 May 2009 04:33:06 +0000 Subject: [PATCH 147/296] microVU: - fixed a bug from r1281 when the tri-ace gamefix is unchecked xD git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1284 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Misc.inl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index e9c44a78bf..0d7ebe631d 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -373,7 +373,7 @@ void SSE_MIN2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { // Warning: Modifies all vectors in 'to' and 'from', and Modifies xmmT1 and xmmT2 void SSE_ADD2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { - if (!CHECK_VUADDSUBHACK) return; + if (!CHECK_VUADDSUBHACK) { SSE_ADDSS_XMM_to_XMM(to, from); return; } u8 *localptr[8]; SSE_MOVAPS_XMM_to_XMM(xmmT1, to); From 65b90bce831e7e04a43b14d149a7c8aff140c038 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 29 May 2009 06:26:46 +0000 Subject: [PATCH 148/296] GSdx: there was a bug in unaligned 4/8 bit transfer, small texture glitches may be fixed (simpson game, car logos in gt4). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1285 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSCrc.h | 1 + plugins/GSdx/GSDevice9.cpp | 44 ++++++++++++++++++++++++++-------- plugins/GSdx/GSLocalMemory.cpp | 10 ++++---- plugins/GSdx/GSRendererHW.h | 24 +++++++++++++++++++ plugins/GSdx/GSState.cpp | 6 +++-- 6 files changed, 68 insertions(+), 18 deletions(-) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index b6b98060dc..a23939aeda 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -130,6 +130,7 @@ CRC::Game CRC::m_games[] = {0x045D77E9, TalesOfAbyss, US}, // undub {0xAA5EC3A3, TalesOfAbyss, JP}, {0xFB236A46, SonicUnleashed, US}, + {0x4C7BB3C8, SimpsonsGame, Unknown}, }; hash_map CRC::m_map; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index 731e495a09..1d68f1dc09 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -64,6 +64,7 @@ public: MajokkoALaMode2, TalesOfAbyss, SonicUnleashed, + SimpsonsGame, TitleCount, }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 7d4f12a940..8ccafc6a27 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -407,31 +407,55 @@ void GSDevice9::ClearRenderTarget(GSTexture* t, const GSVector4& c) ClearRenderTarget(t, (c * 255 + 0.5f).zyxw().rgba32()); } -void GSDevice9::ClearRenderTarget(GSTexture* t, uint32 c) +void GSDevice9::ClearRenderTarget(GSTexture* rt, uint32 c) { CComPtr surface; m_dev->GetRenderTarget(0, &surface); - m_dev->SetRenderTarget(0, *(GSTexture9*)t); + m_dev->SetRenderTarget(0, *(GSTexture9*)rt); m_dev->Clear(0, NULL, D3DCLEAR_TARGET, c, 0, 0); m_dev->SetRenderTarget(0, surface); } void GSDevice9::ClearDepth(GSTexture* t, float c) { - CComPtr surface; - m_dev->GetDepthStencilSurface(&surface); + GSTexture* rt = CreateRenderTarget(t->GetWidth(), t->GetHeight()); + + CComPtr rtsurface; + CComPtr dssurface; + + m_dev->GetRenderTarget(0, &rtsurface); + m_dev->GetDepthStencilSurface(&dssurface); + + m_dev->SetRenderTarget(0, *(GSTexture9*)rt); m_dev->SetDepthStencilSurface(*(GSTexture9*)t); + m_dev->Clear(0, NULL, D3DCLEAR_ZBUFFER, 0, c, 0); - m_dev->SetDepthStencilSurface(surface); + + m_dev->SetRenderTarget(0, rtsurface); + m_dev->SetDepthStencilSurface(dssurface); + + Recycle(rt); } void GSDevice9::ClearStencil(GSTexture* t, uint8 c) { - CComPtr surface; - m_dev->GetDepthStencilSurface(&surface); + GSTexture* rt = CreateRenderTarget(t->GetWidth(), t->GetHeight()); + + CComPtr rtsurface; + CComPtr dssurface; + + m_dev->GetRenderTarget(0, &rtsurface); + m_dev->GetDepthStencilSurface(&dssurface); + + m_dev->SetRenderTarget(0, *(GSTexture9*)rt); m_dev->SetDepthStencilSurface(*(GSTexture9*)t); + m_dev->Clear(0, NULL, D3DCLEAR_STENCIL, 0, 0, c); - m_dev->SetDepthStencilSurface(surface); + + m_dev->SetRenderTarget(0, rtsurface); + m_dev->SetDepthStencilSurface(dssurface); + + Recycle(rt); } GSTexture* GSDevice9::Create(int type, int w, int h, int format) @@ -492,7 +516,7 @@ GSTexture* GSDevice9::CreateRenderTarget(int w, int h, int format) GSTexture* GSDevice9::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D24S8); + return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D24S8); // D3DFMT_D32F_LOCKABLE } GSTexture* GSDevice9::CreateTexture(int w, int h, int format) @@ -661,7 +685,7 @@ void GSDevice9::IASetVertexBuffer(const void* vertices, size_t stride, size_t co { HRESULT hr; - hr = m_dev->CreateVertexBuffer(m_vertices.limit * stride, D3DUSAGE_DYNAMIC, 0, D3DPOOL_DEFAULT, &m_vertices.vb, NULL); + hr = m_dev->CreateVertexBuffer(m_vertices.limit * stride, D3DUSAGE_DYNAMIC | D3DUSAGE_WRITEONLY, 0, D3DPOOL_DEFAULT, &m_vertices.vb, NULL); if(FAILED(hr)) return; } diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 54346dabbe..31567fa8e3 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -813,12 +813,12 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, uint8* src, break; case PSM_PSMT8: ReadColumn8(y, dst, buff, 16); - memcpy(&buff[y2 * 16], &src[x], h2 * 16); + for(int i = 0, j = y2; i < h2; i++, j++) memcpy(&buff[j * 16], &src[i * srcpitch + x], 16); WriteColumn8(y, dst, buff, 16); break; case PSM_PSMT4: ReadColumn4(y, dst, buff, 16); - memcpy(&buff[y2 * 16], &src[x >> 1], h2 * 16); + for(int i = 0, j = y2; i < h2; i++, j++) memcpy(&buff[j * 16], &src[i * srcpitch + (x >> 1)], 16); WriteColumn4(y, dst, buff, 16); break; // TODO @@ -894,12 +894,12 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, uint8* src, break; case PSM_PSMT8: ReadColumn8(y, dst, buff, 16); - memcpy(&buff[0], &src[x], h * 16); + for(int i = 0; i < h; i++) memcpy(&buff[i * 16], &src[i * srcpitch + x], 16); WriteColumn8(y, dst, buff, 16); break; case PSM_PSMT4: ReadColumn4(y, dst, buff, 16); - memcpy(&buff[0], &src[x >> 1], h * 16); + for(int i = 0; i < h; i++) memcpy(&buff[i * 16], &src[i * srcpitch + (x >> 1)], 16); WriteColumn4(y, dst, buff, 16); break; // TODO @@ -2523,8 +2523,6 @@ HRESULT GSLocalMemory::SaveBMP(const string& fn, uint32 bp, uint32 bw, uint32 ps TEXA.TA0 = 0; TEXA.TA1 = 0x80; - // (this->*m_psm[TEX0.PSM].rtx)(GSVector4i(0, 0, w, h), bits, pitch, TEX0, TEXA); - readPixel rp = m_psm[psm].rp; uint8* p = (uint8*)bits; diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 98bcb59d65..d1e95de50d 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -572,6 +572,30 @@ protected: #pragma endregion + #pragma region Simpsons Game z buffer clear + + if(m_game.title == CRC::SimpsonsGame) + { + uint32 FBP = m_context->FRAME.Block(); + uint32 FBW = m_context->FRAME.FBW; + uint32 FPSM = m_context->FRAME.PSM; + + if(FBP == 0x01800 && FPSM == PSM_PSMZ24) + { + // instead of just simply drawing a full height 512x512 sprite to clear the z buffer, + // it uses a 512x256 sprite only, yet it is still able to fill the whole surface with zeros, + // how? by using a render target that overlaps with the lower half of the z buffer... + + m_dev->ClearDepth(ds, 0); + + return false; + } + + return true; + } + + #pragma endregion + return true; } diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 65f7a6cf77..d1d3ccad62 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -968,7 +968,7 @@ void GSState::Write(uint8* mem, int len) int w = m_env.TRXREG.RRW; int h = m_env.TRXREG.RRH; - // TRACE(_T("Write len=%d DBP=%05x DBW=%d DPSM=%d DSAX=%d DSAY=%d RRW=%d RRH=%d\n"), len, (int)m_env.BITBLTBUF.DBP, (int)m_env.BITBLTBUF.DBW, (int)m_env.BITBLTBUF.DPSM, dx, dy, w, h); + // TRACE(_T("Write len=%d DBP=%05x DBW=%d DPSM=%d DSAX=%d DSAY=%d RRW=%d RRH=%d\n"), len, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, m_env.TRXPOS.DSAX, m_env.TRXPOS.DSAY, m_env.TRXREG.RRW, m_env.TRXREG.RRH); if(!m_tr.Update(w, h, GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM].trbpp, len)) { @@ -1209,7 +1209,9 @@ template void GSState::Transfer(uint8* mem, uint32 size) { m_q = 1.0f; - if(path.tag.PRE && (path.tag.FLG & 2) == 0) + ASSERT(!(path.tag.PRE && path.tag.FLG == GIF_FLG_REGLIST)); + + if(path.tag.PRE && path.tag.FLG == GIF_FLG_PACKED) { GIFReg r; r.u64 = path.tag.PRIM; From 71480b31242cee2cb216f27797f96e4220ab2ed2 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 29 May 2009 06:33:56 +0000 Subject: [PATCH 149/296] Zeropad: A bunch of refactoring and cleanup. The Windows port probably won't compile right now. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1286 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 200 ++------------ plugins/zeropad/Linux/linux.h | 7 +- plugins/zeropad/Makefile.am | 2 +- plugins/zeropad/Windows/ZeroPAD_2008.vcproj | 16 ++ plugins/zeropad/Windows/gui.cpp | 110 ++++++++ plugins/zeropad/Windows/win.cpp | 268 ++----------------- plugins/zeropad/Windows/win.h | 33 +++ plugins/zeropad/joystick.cpp | 4 + plugins/zeropad/joystick.h | 1 + plugins/zeropad/keyboard.cpp | 277 ++++++++++++++++++++ plugins/zeropad/keyboard.h | 44 ++++ plugins/zeropad/zeropad.cpp | 30 +++ plugins/zeropad/zeropad.h | 9 +- 13 files changed, 568 insertions(+), 433 deletions(-) create mode 100644 plugins/zeropad/Windows/gui.cpp create mode 100644 plugins/zeropad/Windows/win.h create mode 100644 plugins/zeropad/keyboard.cpp create mode 100644 plugins/zeropad/keyboard.h diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 250e956fd1..55100bde96 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -19,8 +19,6 @@ #include "linux.h" Display *GSdsp; -static pthread_spinlock_t s_mutexStatus; -static u32 s_keyPress[2], s_keyRelease[2]; // thread safe static const char* s_pGuiKeyMap[] = { @@ -43,20 +41,14 @@ string GetLabelFromButton(const char* buttonname) s32 _PADopen(void *pDsp) { GSdsp = *(Display**)pDsp; - pthread_spin_init(&s_mutexStatus, PTHREAD_PROCESS_PRIVATE); - s_keyPress[0] = s_keyPress[1] = 0; - s_keyRelease[0] = s_keyRelease[1] = 0; - XAutoRepeatOff(GSdsp); - - JoystickInfo::EnumerateJoysticks(s_vjoysticks); + SetAutoRepeat(false); return 0; } void _PADclose() { - pthread_spin_destroy(&s_mutexStatus); - XAutoRepeatOn(GSdsp); + SetAutoRepeat(true); vector::iterator it = s_vjoysticks.begin(); @@ -70,16 +62,6 @@ void _PADclose() s_vjoysticks.clear(); } -void _PADupdate(int pad) -{ - pthread_spin_lock(&s_mutexStatus); - status[pad] |= s_keyRelease[pad]; - status[pad] &= ~s_keyPress[pad]; - s_keyRelease[pad] = 0; - s_keyPress[pad] = 0; - pthread_spin_unlock(&s_mutexStatus); -} - int _GetJoystickIdFromPAD(int pad) { // select the right joystick id @@ -99,7 +81,7 @@ int _GetJoystickIdFromPAD(int pad) } } - if ((joyid < 0) || (joyid >= (int)s_vjoysticks.size())) + if (JoystickIdWithinBounds(joyid)) { // get first unused joystick for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) @@ -113,85 +95,8 @@ int _GetJoystickIdFromPAD(int pad) void CALLBACK PADupdate(int pad) { - int i; - XEvent E; - int keyPress = 0, keyRelease = 0; - KeySym key; - - // keyboard input - while (XPending(GSdsp) > 0) - { - XNextEvent(GSdsp, &E); - switch (E.type) - { - case KeyPress: - key = XLookupKeysym((XKeyEvent *) & E, 0); - - i = FindKey(key, pad); - - // Analog controls. - if ((i > PAD_RY) && (i <= PAD_R_LEFT)) - { - switch (i) - { - case PAD_R_LEFT: - case PAD_R_UP: - case PAD_L_LEFT: - case PAD_L_UP: - Analog::ConfigurePad(Analog::AnalogToPad(i), pad, DEF_VALUE); - break; - case PAD_R_RIGHT: - case PAD_R_DOWN: - case PAD_L_RIGHT: - case PAD_L_DOWN: - Analog::ConfigurePad(Analog::AnalogToPad(i), pad, -DEF_VALUE); - break; - } - i += 0xff00; - } - - if (i != -1) - { - clear_bit(keyRelease, i); - set_bit(keyPress, i); - } - //PAD_LOG("Key pressed:%d\n", i); - - event.evt = KEYPRESS; - event.key = key; - break; - - case KeyRelease: - key = XLookupKeysym((XKeyEvent *) & E, 0); - - i = FindKey(key, pad); - - // Analog Controls. - if ((i > PAD_RY) && (i <= PAD_R_LEFT)) - { - Analog::ResetPad(Analog::AnalogToPad(i), pad); - i += 0xff00; - } - - if (i != -1) - { - clear_bit(keyPress, i); - set_bit(keyRelease, i); - } - - event.evt = KEYRELEASE; - event.key = key; - break; - - case FocusIn: - XAutoRepeatOff(GSdsp); - break; - - case FocusOut: - XAutoRepeatOn(GSdsp); - break; - } - } + // Poll keyboard. + PollForKeyboardInput(pad); // joystick info SDL_JoystickUpdate(); @@ -199,34 +104,25 @@ void CALLBACK PADupdate(int pad) for (int i = 0; i < PADKEYS; i++) { int key = conf.keys[PadEnum[pad][0]][i]; - JoystickInfo* pjoy = NULL; - - if (IS_JOYBUTTONS(key)) - { - int joyid = PAD_GETJOYID(key); - if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) + if (JoystickIdWithinBounds(PAD_GETJOYID(key))) + { + JoystickInfo* pjoy = s_vjoysticks[PAD_GETJOYID(key)]; + int pad = (pjoy)->GetPAD(); + + if (IS_JOYBUTTONS(key)) { - pjoy = s_vjoysticks[joyid]; - int pad = (pjoy)->GetPAD(); int value = SDL_JoystickGetButton((pjoy)->GetJoy(), PAD_GETJOYBUTTON(key)); - + if (value) - clear_bit(status[pad], i); // pressed + clear_bit(status[pad], i); // released else set_bit(status[pad], i); // pressed } - } - else if (IS_JOYSTICK(key)) - { - int joyid = PAD_GETJOYID(key); - - if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) + else if (IS_JOYSTICK(key)) { - pjoy = s_vjoysticks[joyid]; - int pad = (pjoy)->GetPAD(); int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - + switch (i) { case PAD_LX: @@ -240,18 +136,11 @@ void CALLBACK PADupdate(int pad) break; } } - } -#ifdef EXPERIMENTAL_POV_CODE - else if (IS_HAT(key)) - { - int joyid = PAD_GETJOYID(key); - - if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) + #ifdef EXPERIMENTAL_POV_CODE + else if (IS_HAT(key)) { - pjoy = s_vjoysticks[joyid]; - int pad = (pjoy)->GetPAD(); int value = SDL_JoystickGetHat((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); - + //PAD_LOG("Hat = %d for key %d\n", PAD_GETPOVDIR(key), key); if ((value != SDL_HAT_CENTERED) && (PAD_GETHATDIR(key) == value)) { @@ -270,16 +159,9 @@ void CALLBACK PADupdate(int pad) clear_bit(status[pad], i); } } - } -#endif - else if (IS_POV(key)) - { - int joyid = PAD_GETJOYID(key); - - if (joyid >= 0 && (joyid < (int)s_vjoysticks.size())) + #endif + else if (IS_POV(key)) { - pjoy = s_vjoysticks[joyid]; - int pad = (pjoy)->GetPAD(); int value = SDL_JoystickGetAxis((pjoy)->GetJoy(), PAD_GETJOYSTICK_AXIS(key)); if (PAD_GETPOVSIGN(key) && (value < -2048)) @@ -288,18 +170,9 @@ void CALLBACK PADupdate(int pad) clear_bit(status[pad], i); else set_bit(status[pad], i); - - } } } - - pthread_spin_lock(&s_mutexStatus); - s_keyPress[pad] |= keyPress; - s_keyPress[pad] &= ~keyRelease; - s_keyRelease[pad] |= keyRelease; - s_keyRelease[pad] &= ~keyPress; - pthread_spin_unlock(&s_mutexStatus); } void UpdateConf(int pad) @@ -324,7 +197,7 @@ void UpdateConf(int pad) if (IS_KEYBOARD(conf.keys[pad][i])) { - char* pstr = XKeysymToString(PAD_GETKEY(conf.keys[pad][i])); + char* pstr = KeysymToChar(PAD_GETKEY(conf.keys[pad][i])); if (pstr != NULL) tmp = pstr; } else if (IS_JOYBUTTONS(conf.keys[pad][i])) @@ -386,7 +259,7 @@ void UpdateConf(int pad) // check bounds int joyid = _GetJoystickIdFromPAD(pad); - if ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())) + if (JoystickIdWithinBounds(joyid)) gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), joyid); // select the combo else gtk_combo_box_set_active(GTK_COMBO_BOX(s_devicecombo), s_vjoysticks.size()); // no gamepad @@ -418,33 +291,6 @@ int GetLabelId(GtkWidget *label) return (int)(uptr)gtk_object_get_user_data(GTK_OBJECT(label)); } -bool PollKeyboard(char* &temp, u32 &pkey) -{ - GdkEvent *ev = gdk_event_get(); - - if (ev != NULL) - { - if (ev->type == GDK_KEY_PRESS) - { - - if (ev->key.keyval == GDK_Escape) - { - temp = "Unknown"; - pkey = NULL; - } - else - { - temp = XKeysymToString(ev->key.keyval); - pkey = ev->key.keyval; - } - - return true; - } - } - - return false; -} - void OnConf_Key(GtkButton *button, gpointer user_data) { GtkWidget* label = GetLabelWidget(button); @@ -465,7 +311,7 @@ void OnConf_Key(GtkButton *button, gpointer user_data) vector::iterator itjoy; char *tmp; - if (PollKeyboard(tmp, conf.keys[pad][key])) + if (PollX11Keyboard(tmp, conf.keys[pad][key])) { strcpy(str, tmp); captured = true; diff --git a/plugins/zeropad/Linux/linux.h b/plugins/zeropad/Linux/linux.h index 5a12f8c145..26de22f1b6 100644 --- a/plugins/zeropad/Linux/linux.h +++ b/plugins/zeropad/Linux/linux.h @@ -16,14 +16,15 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include "joystick.h" +#include "keyboard.h" +#include "zeropad.h" + #include #include #include #include -#include "joystick.h" - -#include "zeropad.h" extern "C" { diff --git a/plugins/zeropad/Makefile.am b/plugins/zeropad/Makefile.am index 3937ec3c90..8d383d794f 100644 --- a/plugins/zeropad/Makefile.am +++ b/plugins/zeropad/Makefile.am @@ -23,4 +23,4 @@ libZeroPAD_LDFLAGS+=-Wl,-soname,@ZEROPAD_SONAME@ libZeroPAD_LDADD=$(libZeroPAD_a_OBJECTS) libZeroPAD_a_SOURCES = joystick.cpp analog.cpp analog.h zeropad.cpp zeropad.h \ -Linux/gui.cpp Linux/linux.cpp Linux/support.c Linux/interface.c +Linux/gui.cpp Linux/linux.cpp Linux/support.c Linux/interface.c keyboard.cpp keyboard.h diff --git a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj index d5927633ca..48f0aaf2d6 100644 --- a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj +++ b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj @@ -246,6 +246,10 @@ RelativePath=".\win.cpp" > + + @@ -254,12 +258,20 @@ RelativePath="..\analog.cpp" > + + + + @@ -280,6 +292,10 @@ RelativePath="..\analog.h" > + + diff --git a/plugins/zeropad/Windows/gui.cpp b/plugins/zeropad/Windows/gui.cpp new file mode 100644 index 0000000000..a2a6a07fcf --- /dev/null +++ b/plugins/zeropad/Windows/gui.cpp @@ -0,0 +1,110 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include "win.h" + + void SaveConfig() +{ + char *szTemp; + char szIniFile[256], szValue[256], szProf[256]; + int i, j; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if (!szTemp) return; + strcpy(szTemp, "\\inis\\zeropad.ini"); + + for (j = 0; j < 2; j++) + { + for (i = 0; i < PADKEYS; i++) + { + sprintf(szProf, "%d_%d", j, i); + sprintf(szValue, "%d", conf.keys[j][i]); + WritePrivateProfileString("Interface", szProf, szValue, szIniFile); + } + } + + sprintf(szValue, "%u", conf.log); + WritePrivateProfileString("Interface", "Logging", szValue, szIniFile); +} + +void LoadConfig() +{ + FILE *fp; + char *szTemp; + char szIniFile[256], szValue[256], szProf[256]; + int i, j; + + memset(&conf, 0, sizeof(conf)); +#ifdef _WIN32 + conf.keys[0][0] = 'W'; // L2 + conf.keys[0][1] = 'O'; // R2 + conf.keys[0][2] = 'A'; // L1 + conf.keys[0][3] = ';'; // R1 + conf.keys[0][4] = 'I'; // TRIANGLE + conf.keys[0][5] = 'L'; // CIRCLE + conf.keys[0][6] = 'K'; // CROSS + conf.keys[0][7] = 'J'; // SQUARE + conf.keys[0][8] = 'V'; // SELECT + conf.keys[0][11] = 'N'; // START + conf.keys[0][12] = 'E'; // UP + conf.keys[0][13] = 'F'; // RIGHT + conf.keys[0][14] = 'D'; // DOWN + conf.keys[0][15] = 'S'; // LEFT +#endif + conf.log = 0; + + GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); + szTemp = strrchr(szIniFile, '\\'); + + if (!szTemp) return ; + strcpy(szTemp, "\\inis\\zeropad.ini"); + fp = fopen("inis\\zeropad.ini", "rt");//check if usbnull.ini really exists + if (!fp) + { + CreateDirectory("inis", NULL); + SaveConfig();//save and return + return ; + } + fclose(fp); + + for (j = 0; j < 2; j++) + { + for (i = 0; i < PADKEYS; i++) + { + sprintf(szProf, "%d_%d", j, i); + GetPrivateProfileString("Interface", szProf, NULL, szValue, 20, szIniFile); + conf.keys[j][i] = strtoul(szValue, NULL, 10); + } + } + + GetPrivateProfileString("Interface", "Logging", NULL, szValue, 20, szIniFile); + conf.log = strtoul(szValue, NULL, 10); +} + +void SysMessage(char *fmt, ...) +{ + va_list list; + char tmp[512]; + + va_start(list, fmt); + vsprintf(tmp, fmt, list); + va_end(list); + MessageBox(0, tmp, "PADwinKeyb Msg", 0); +} \ No newline at end of file diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index 959f65fedc..6bf8e79275 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -16,22 +16,11 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include -#include -#include -#include - -#include "resource.h" -#include "../zeropad.h" - -#include -#include +#include "win.h" using namespace std; HINSTANCE hInst = NULL; -static pthread_spinlock_t s_mutexStatus; -static u32 s_keyPress[2], s_keyRelease[2]; extern u16 status[2]; extern string s_strIniPath; @@ -41,105 +30,9 @@ HWND GShwnd = NULL; extern keyEvent event; - -void SaveConfig() -{ - char *szTemp; - char szIniFile[256], szValue[256], szProf[256]; - int i, j; - - GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); - szTemp = strrchr(szIniFile, '\\'); - - if (!szTemp) return; - strcpy(szTemp, "\\inis\\zeropad.ini"); - - for (j = 0; j < 2; j++) - { - for (i = 0; i < PADKEYS; i++) - { - sprintf(szProf, "%d_%d", j, i); - sprintf(szValue, "%d", conf.keys[j][i]); - WritePrivateProfileString("Interface", szProf, szValue, szIniFile); - } - } - - sprintf(szValue, "%u", conf.log); - WritePrivateProfileString("Interface", "Logging", szValue, szIniFile); -} - -void LoadConfig() -{ - FILE *fp; - char *szTemp; - char szIniFile[256], szValue[256], szProf[256]; - int i, j; - - memset(&conf, 0, sizeof(conf)); -#ifdef _WIN32 - conf.keys[0][0] = 'W'; // L2 - conf.keys[0][1] = 'O'; // R2 - conf.keys[0][2] = 'A'; // L1 - conf.keys[0][3] = ';'; // R1 - conf.keys[0][4] = 'I'; // TRIANGLE - conf.keys[0][5] = 'L'; // CIRCLE - conf.keys[0][6] = 'K'; // CROSS - conf.keys[0][7] = 'J'; // SQUARE - conf.keys[0][8] = 'V'; // SELECT - conf.keys[0][11] = 'N'; // START - conf.keys[0][12] = 'E'; // UP - conf.keys[0][13] = 'F'; // RIGHT - conf.keys[0][14] = 'D'; // DOWN - conf.keys[0][15] = 'S'; // LEFT -#endif - conf.log = 0; - - GetModuleFileName(GetModuleHandle((LPCSTR)hInst), szIniFile, 256); - szTemp = strrchr(szIniFile, '\\'); - - if (!szTemp) return ; - strcpy(szTemp, "\\inis\\zeropad.ini"); - fp = fopen("inis\\zeropad.ini", "rt");//check if usbnull.ini really exists - if (!fp) - { - CreateDirectory("inis", NULL); - SaveConfig();//save and return - return ; - } - fclose(fp); - - for (j = 0; j < 2; j++) - { - for (i = 0; i < PADKEYS; i++) - { - sprintf(szProf, "%d_%d", j, i); - GetPrivateProfileString("Interface", szProf, NULL, szValue, 20, szIniFile); - conf.keys[j][i] = strtoul(szValue, NULL, 10); - } - } - - GetPrivateProfileString("Interface", "Logging", NULL, szValue, 20, szIniFile); - conf.log = strtoul(szValue, NULL, 10); -} - -void SysMessage(char *fmt, ...) -{ - va_list list; - char tmp[512]; - - va_start(list, fmt); - vsprintf(tmp, fmt, list); - va_end(list); - MessageBox(0, tmp, "PADwinKeyb Msg", 0); -} - s32 _PADopen(void *pDsp) { - memset(&event, 0, sizeof(event)); LoadConfig(); - pthread_spin_init(&s_mutexStatus, PTHREAD_PROCESS_PRIVATE); - s_keyPress[0] = s_keyPress[1] = 0; - s_keyRelease[0] = s_keyRelease[1] = 0; if (GShwnd != NULL && GSwndProc != NULL) { @@ -162,167 +55,48 @@ void _PADclose() GSwndProc = NULL; GShwnd = NULL; } - pthread_spin_destroy(&s_mutexStatus); -} - -void _PADupdate(int pad) -{ - pthread_spin_lock(&s_mutexStatus); - status[pad] |= s_keyRelease[pad]; - status[pad] &= ~s_keyPress[pad]; - s_keyRelease[pad] = 0; - s_keyPress[pad] = 0; - pthread_spin_unlock(&s_mutexStatus); } +// Yes, let's not do anything when pcsx2 asks for us for an update. +// We certainly don't want to update the gamepad information... void CALLBACK PADupdate(int pad) { } -LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) -{ - int i, pad, keyPress[2] = {0}, keyRelease[2] = {0}; - static bool lbutton = false, rbutton = false; - - switch (msg) - { - case WM_KEYDOWN: - if (lParam & 0x40000000) - return TRUE; - - for (pad = 0; pad < 2; ++pad) - { - for (i = 0; i < PADKEYS; i++) - { - if (wParam == conf.keys[pad][i]) - { - keyPress[pad] |= (1 << i); - keyRelease[pad] &= ~(1 << i); - break; - } - } - } - - event.evt = KEYPRESS; - event.key = wParam; - break; - - case WM_KEYUP: - for (pad = 0; pad < 2; ++pad) - { - for (i = 0; i < PADKEYS; i++) - { - if (wParam == conf.keys[pad][i]) - { - keyPress[pad] &= ~(1 << i); - keyRelease[pad] |= (1 << i); - break; - } - } - } - - - event.evt = KEYRELEASE; - event.key = wParam; - break; - - case WM_LBUTTONDOWN: - lbutton = true; - break; - - case WM_LBUTTONUP: - g_lanalog[0].x = 0x80; - g_lanalog[0].y = 0x80; - g_lanalog[1].x = 0x80; - g_lanalog[1].y = 0x80; - lbutton = false; - break; - - case WM_RBUTTONDOWN: - rbutton = true; - break; - - case WM_RBUTTONUP: - g_ranalog[0].x = 0x80; - g_ranalog[0].y = 0x80; - g_ranalog[1].x = 0x80; - g_ranalog[1].y = 0x80; - rbutton = false; - break; - - case WM_MOUSEMOVE: - if (lbutton) - { - g_lanalog[0].x = LOWORD(lParam) & 254; - g_lanalog[0].y = HIWORD(lParam) & 254; - g_lanalog[1].x = LOWORD(lParam) & 254; - g_lanalog[1].y = HIWORD(lParam) & 254; - } - if (rbutton) - { - g_ranalog[0].x = LOWORD(lParam) & 254; - g_ranalog[0].y = HIWORD(lParam) & 254; - g_ranalog[1].x = LOWORD(lParam) & 254; - g_ranalog[1].y = HIWORD(lParam) & 254; - } - break; - - case WM_DESTROY: - case WM_QUIT: - event.evt = KEYPRESS; - event.key = VK_ESCAPE; - return GSwndProc(hWnd, msg, wParam, lParam); - - default: - return GSwndProc(hWnd, msg, wParam, lParam); - } - - pthread_spin_lock(&s_mutexStatus); - for (pad = 0; pad < 2; ++pad) - { - s_keyPress[pad] |= keyPress[pad]; - s_keyPress[pad] &= ~keyRelease[pad]; - s_keyRelease[pad] |= keyRelease[pad]; - s_keyRelease[pad] &= ~keyPress[pad]; - } - pthread_spin_unlock(&s_mutexStatus); - - return TRUE; -} - +// Most of the values look off in here by either a factor of 10 or 100. string GetKeyLabel(const int pad, const int index) { const int key = conf.keys[pad][index]; char buff[16] = "NONE)"; - if (key < 0x100) + if (key < 0x100) //IS_KEYBOARD; should be 0x10000 { if (key == 0) strcpy(buff, "NONE"); else { if (key >= 0x60 && key <= 0x69) - { sprintf(buff, "NumPad %c", '0' + key - 0x60); - } - else sprintf(buff, "%c", key); + else + sprintf(buff, "%c", key); } } - else if (key >= 0x1000 && key < 0x2000) + else if (key >= 0x1000 && key < 0x2000) //IS_JOYBUTTONS; 0x10000 - 0x20000 { sprintf(buff, "J%d_%d", (key & 0xfff) / 0x100, (key & 0xff) + 1); } - else if (key >= 0x2000 && key < 0x3000) + else if (key >= 0x2000 && key < 0x3000) // IS_JOYSTICK; 0x20000 - 0x30000 { static const char name[][4] = { "MIN", "MAX" }; const int axis = (key & 0xff); + sprintf(buff, "J%d_AXIS%d_%s", (key & 0xfff) / 0x100, axis / 2, name[axis % 2]); - if (index >= 17 && index <= 20) - buff[strlen(buff) -4] = '\0'; + if (index >= 17 && index <= 20) buff[strlen(buff) -4] = '\0'; } - else if (key >= 0x3000 && key < 0x4000) + else if (key >= 0x3000 && key < 0x4000) // IS_POV; 0x30000 - 0x50000 { - static const char name[][7] = { "FOWARD", "RIGHT", "BACK", "LEFT" }; + static const char name[][7] = { "FORWARD", "RIGHT", "BACK", "LEFT" }; const int pov = (key & 0xff); + sprintf(buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov / 4, name[pov % 4]); } @@ -343,7 +117,7 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) case WM_INITDIALOG: LoadConfig(); padn = 0; - if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); + if (conf.log) CheckDlgButton(hW, IDC_LOG, TRUE); for (i = 0; i < PADKEYS; i++) { @@ -369,6 +143,7 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { key = 0; //pkeyboard = SDL_GetKeyState(&numkeys); + // Yeah, there's no way this will work, given that it's disabled. for (int i = 0; i < numkeys; ++i) { if (pkeyboard[i]) @@ -382,9 +157,10 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { // check joystick } - - if (key != 0) + else { + // Get rid of the expired timer, try to configure the keys, fail horribly, + // and either crash or don't register a keypress. KillTimer(hW, 0x80); hWC = GetDlgItem(hW, disabled); conf.keys[padn][disabled-IDC_L2] = key; @@ -401,11 +177,13 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { if (LOWORD(wParam) == i) { + // A button was pressed if (disabled)//change selection EnableWindow(GetDlgItem(hW, disabled), TRUE); EnableWindow(GetDlgItem(hW, disabled = wParam), FALSE); + // Wait a silly amount of time for a keypress. SetTimer(hW, 0x80, 250, NULL); return TRUE; @@ -418,11 +196,13 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) KillTimer(hW, 0x80); EndDialog(hW, TRUE); return TRUE; + case IDOK: KillTimer(hW, 0x80); if (IsDlgButtonChecked(hW, IDC_LOG)) conf.log = 1; - else conf.log = 0; + else + conf.log = 0; SaveConfig(); EndDialog(hW, FALSE); return TRUE; diff --git a/plugins/zeropad/Windows/win.h b/plugins/zeropad/Windows/win.h new file mode 100644 index 0000000000..ed468e2c64 --- /dev/null +++ b/plugins/zeropad/Windows/win.h @@ -0,0 +1,33 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #include +#include +#include +#include + +#include "resource.h" +#include "../zeropad.h" + +#include +#include + +extern void SaveConfig(); +extern void LoadConfig(); +extern void SysMessage(char *fmt, ...); +extern LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); \ No newline at end of file diff --git a/plugins/zeropad/joystick.cpp b/plugins/zeropad/joystick.cpp index 579490b694..1878abb18f 100644 --- a/plugins/zeropad/joystick.cpp +++ b/plugins/zeropad/joystick.cpp @@ -69,6 +69,10 @@ const char *HatName(int value) return "Unknown"; } +bool JoystickIdWithinBounds(int joyid) +{ + return (!((joyid < 0) || (joyid >= (int)s_vjoysticks.size()))); +} // opens handles to all possible joysticks void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) { diff --git a/plugins/zeropad/joystick.h b/plugins/zeropad/joystick.h index 4f9d5b0052..1cecd8e23c 100644 --- a/plugins/zeropad/joystick.h +++ b/plugins/zeropad/joystick.h @@ -139,4 +139,5 @@ extern int s_selectedpad; extern vector s_vjoysticks; extern void UpdateJoysticks(); extern const char *HatName(int value); +extern bool JoystickIdWithinBounds(int joyid); #endif diff --git a/plugins/zeropad/keyboard.cpp b/plugins/zeropad/keyboard.cpp new file mode 100644 index 0000000000..bdcf1a6986 --- /dev/null +++ b/plugins/zeropad/keyboard.cpp @@ -0,0 +1,277 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + /* + * Theoretically, this header is for anything to do with keyboard input. + * Pragmatically, event handing's going in here too. + */ + + #include "keyboard.h" + + __forceinline int FindKey(int key, int pad) +{ + for (int p = 0; p < PADSUBKEYS; p++) + for (int i = 0; i < PADKEYS; i++) + if (key == conf.keys[(PadEnum[pad][p])][i]) return i; + return -1; +} + +char* KeysymToChar(int keysym) +{ + #ifdef __LINUX__ + return XKeysymToString(keysym); +#else + return; + #endif +} + +void PollForKeyboardInput(int pad) +{ + #ifdef __LINUX__ + PollForX11KeyboardInput(pad); +#endif +} + +void SetAutoRepeat(bool autorep) +{ + #ifdef __LINUX__ + if (autorep) + XAutoRepeatOn(GSdsp); + else + XAutoRepeatOff(GSdsp); +#endif +} + + #ifdef __LINUX__ +void PollForX11KeyboardInput(int pad) +{ + XEvent E; + KeySym key; + int keyPress = 0, keyRelease = 0; + int i; + + // keyboard input + while (XPending(GSdsp) > 0) + { + XNextEvent(GSdsp, &E); + switch (E.type) + { + case KeyPress: + key = XLookupKeysym((XKeyEvent *) & E, 0); + + i = FindKey(key, pad); + + // Analog controls. + if ((i > PAD_RY) && (i <= PAD_R_LEFT)) + { + switch (i) + { + case PAD_R_LEFT: + case PAD_R_UP: + case PAD_L_LEFT: + case PAD_L_UP: + Analog::ConfigurePad(Analog::AnalogToPad(i), pad, DEF_VALUE); + break; + case PAD_R_RIGHT: + case PAD_R_DOWN: + case PAD_L_RIGHT: + case PAD_L_DOWN: + Analog::ConfigurePad(Analog::AnalogToPad(i), pad, -DEF_VALUE); + break; + } + i += 0xff00; + } + + if (i != -1) + { + clear_bit(keyRelease, i); + set_bit(keyPress, i); + } + //PAD_LOG("Key pressed:%d\n", i); + + event.evt = KEYPRESS; + event.key = key; + break; + + case KeyRelease: + key = XLookupKeysym((XKeyEvent *) & E, 0); + + i = FindKey(key, pad); + + // Analog Controls. + if ((i > PAD_RY) && (i <= PAD_R_LEFT)) + { + Analog::ResetPad(Analog::AnalogToPad(i), pad); + i += 0xff00; + } + + if (i != -1) + { + clear_bit(keyPress, i); + set_bit(keyRelease, i); + } + + event.evt = KEYRELEASE; + event.key = key; + break; + + case FocusIn: + XAutoRepeatOff(GSdsp); + break; + + case FocusOut: + XAutoRepeatOn(GSdsp); + break; + } + } + + UpdateKeys(pad, keyPress, keyRelease); +} + +bool PollX11Keyboard(char* &temp, u32 &pkey) +{ + GdkEvent *ev = gdk_event_get(); + + if (ev != NULL) + { + if (ev->type == GDK_KEY_PRESS) + { + + if (ev->key.keyval == GDK_Escape) + { + temp = "Unknown"; + pkey = NULL; + } + else + { + temp = KeysymToChar(ev->key.keyval); + pkey = ev->key.keyval; + } + + return true; + } + } + + return false; +} +#else +LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) +{ + int i, pad, keyPress[2] = {0}, keyRelease[2] = {0}; + static bool lbutton = false, rbutton = false; + + switch (msg) + { + case WM_KEYDOWN: + if (lParam & 0x40000000) return TRUE; + + for (pad = 0; pad < 2; ++pad) + { + for (i = 0; i < PADKEYS; i++) + { + if (wParam == conf.keys[pad][i]) + { + set_bit(keyPress[pad], i); + clear_bit(keyRelease[pad], i); + break; + } + } + } + + event.evt = KEYPRESS; + event.key = wParam; + break; + + case WM_KEYUP: + for (pad = 0; pad < 2; ++pad) + { + for (i = 0; i < PADKEYS; i++) + { + if (wParam == conf.keys[pad][i]) + { + set_bit(keyRelease[pad], i); + clear_bit(keyPress[pad], i); + break; + } + } + } + + + event.evt = KEYRELEASE; + event.key = wParam; + break; + + /*case WM_LBUTTONDOWN: + lbutton = true; + break; + + case WM_LBUTTONUP: + g_lanalog[0].x = 0x80; + g_lanalog[0].y = 0x80; + g_lanalog[1].x = 0x80; + g_lanalog[1].y = 0x80; + lbutton = false; + break; + + case WM_RBUTTONDOWN: + rbutton = true; + break; + + case WM_RBUTTONUP: + g_ranalog[0].x = 0x80; + g_ranalog[0].y = 0x80; + g_ranalog[1].x = 0x80; + g_ranalog[1].y = 0x80; + rbutton = false; + break; + + case WM_MOUSEMOVE: + if (lbutton) + { + g_lanalog[0].x = LOWORD(lParam) & 254; + g_lanalog[0].y = HIWORD(lParam) & 254; + g_lanalog[1].x = LOWORD(lParam) & 254; + g_lanalog[1].y = HIWORD(lParam) & 254; + } + if (rbutton) + { + g_ranalog[0].x = LOWORD(lParam) & 254; + g_ranalog[0].y = HIWORD(lParam) & 254; + g_ranalog[1].x = LOWORD(lParam) & 254; + g_ranalog[1].y = HIWORD(lParam) & 254; + } + break;*/ + + case WM_DESTROY: + case WM_QUIT: + event.evt = KEYPRESS; + event.key = VK_ESCAPE; + return GSwndProc(hWnd, msg, wParam, lParam); + + default: + return GSwndProc(hWnd, msg, wParam, lParam); + } + + for (pad = 0; pad < 2; ++pad) + { + UpdateKeys(pad, keyPress, keyRelease) + } + + return TRUE; +} +#endif diff --git a/plugins/zeropad/keyboard.h b/plugins/zeropad/keyboard.h new file mode 100644 index 0000000000..471b8eddcd --- /dev/null +++ b/plugins/zeropad/keyboard.h @@ -0,0 +1,44 @@ +/* ZeroPAD - author: zerofrog(@gmail.com) + * Copyright (C) 2006-2007 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + + #ifndef __KEYBOARD_H__ + #define __KEYBOARD_H__ + +#include "zeropad.h" + +#ifdef __LINUX__ + +#include "Linux/linux.h" + +extern Display *GSdsp; +extern void PollForX11KeyboardInput(int pad); +extern bool PollX11Keyboard(char* &temp, u32 &pkey); + +#else + +extern WNDPROC GSwndProc = NULL; +extern HWND GShwnd = NULL; + +#endif + +extern char* KeysymToChar(int keysym); +extern void PollForKeyboardInput(int pad); +extern void SetAutoRepeat(bool autorep); +extern __forceinline int FindKey(int key, int pad); + + #endif \ No newline at end of file diff --git a/plugins/zeropad/zeropad.cpp b/plugins/zeropad/zeropad.cpp index 13582dfe27..493778a0fa 100644 --- a/plugins/zeropad/zeropad.cpp +++ b/plugins/zeropad/zeropad.cpp @@ -23,6 +23,7 @@ #include #include "zeropad.h" +#include "joystick.h" #ifndef _WIN32 @@ -122,6 +123,9 @@ int cmdLen; int ds2mode = 0; // DS Mode at start FILE *padLog = NULL; +pthread_spinlock_t s_mutexStatus; +u32 s_keyPress[2], s_keyRelease[2]; + static void InitLibraryName() { #ifdef _WIN32 @@ -262,14 +266,40 @@ s32 CALLBACK PADopen(void *pDsp) { memset(&event, 0, sizeof(event)); + pthread_spin_init(&s_mutexStatus, PTHREAD_PROCESS_PRIVATE); + s_keyPress[0] = s_keyPress[1] = 0; + s_keyRelease[0] = s_keyRelease[1] = 0; + + JoystickInfo::EnumerateJoysticks(s_vjoysticks); return _PADopen(pDsp); } void CALLBACK PADclose() { + pthread_spin_destroy(&s_mutexStatus); _PADclose(); } +void _PADupdate(int pad) +{ + pthread_spin_lock(&s_mutexStatus); + status[pad] |= s_keyRelease[pad]; + status[pad] &= ~s_keyPress[pad]; + s_keyRelease[pad] = 0; + s_keyPress[pad] = 0; + pthread_spin_unlock(&s_mutexStatus); +} + +void UpdateKeys(int pad, int keyPress, int keyRelease) +{ + pthread_spin_lock(&s_mutexStatus); + s_keyPress[pad] |= keyPress; + s_keyPress[pad] &= ~keyRelease; + s_keyRelease[pad] |= keyRelease; + s_keyRelease[pad] &= ~keyPress; + pthread_spin_unlock(&s_mutexStatus); +} + u32 CALLBACK PADquery() { return 3; // both diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index 8e3127b1f0..f22316ccc4 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -183,14 +183,7 @@ void LoadConfig(); void SaveConfig(); void SysMessage(char *fmt, ...); - -inline int FindKey(int key, int pad) -{ - for (int p = 0; p < PADSUBKEYS; p++) - for (int i = 0; i < PADKEYS; i++) - if (key == conf.keys[(PadEnum[pad][p])][i]) return i; - return -1; -} +void UpdateKeys(int pad, int keyPress, int keyRelease); #endif From 6d6e7fba361672fbc22057ec2488eb70d5d966eb Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 29 May 2009 08:32:54 +0000 Subject: [PATCH 150/296] microVU: - Changed 'recPass' from a template param to a function param. Halves MSVC linking-time in release modes, but still slow (from ~10 minutes to ~5 minutes to link on my PC). I'll continue working on this later... the next change will probably reduce Release link-times to 1~2 minutes (or less than 1 minute for you guys with macho-pc's ;p). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1287 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 4 +- pcsx2/x86/microVU_Compile.inl | 14 +- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 138 ++++++------- pcsx2/x86/microVU_Misc.h | 7 +- pcsx2/x86/microVU_Misc.inl | 2 +- pcsx2/x86/microVU_Tables.inl | 374 ++++++++++++++++------------------ pcsx2/x86/microVU_Upper.inl | 194 +++++++++--------- 8 files changed, 359 insertions(+), 376 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index e956a4e3c5..7a9c3326d3 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -147,8 +147,8 @@ void __fastcall mVUcleanUpVU0(); void __fastcall mVUcleanUpVU1(); void* __fastcall mVUcompileVU0(u32 startPC, uptr pState); void* __fastcall mVUcompileVU1(u32 startPC, uptr pState); -microVUf(void) mVUopU(); -microVUf(void) mVUopL(); +microVUf(void) mVUopU(mF); +microVUf(void) mVUopL(mF); // Private Functions microVUt(void) mVUclearProg(microVU* mVU, int progIndex); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 69f43f6ff4..c54ee08bb8 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -54,8 +54,8 @@ #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } -#define doUpperOp() { mVUopU(); mVUdivSet(); } -#define doLowerOp() { incPC(-1); mVUopL(); incPC(1); } +#define doUpperOp() { mVUopU(1); mVUdivSet(); } +#define doLowerOp() { incPC(-1); mVUopL(1); incPC(1); } #define doIbit() { if (curI & _Ibit_) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } //------------------------------------------------------------------ @@ -188,8 +188,8 @@ microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fC void __fastcall mVUwarning0(u32 PC) { Console::Error("microVU0 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } -void __fastcall mVUprintPC1(u32 PC) { Console::WriteLn("Block startPC [%04x]", params PC); } -void __fastcall mVUprintPC2(u32 PC) { Console::WriteLn("Block endPC [%04x]\n", params PC); } +void __fastcall mVUprintPC1(u32 PC) { Console::Write("Block PC [%04x] ", params PC); } +void __fastcall mVUprintPC2(u32 PC) { Console::Write("[%04x]\n", params PC); } microVUt(void) mVUtestCycles() { microVU* mVU = mVUx; @@ -250,11 +250,11 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUinfo = 0; startLoop(); incCycles(1); - mVUopU(); + mVUopU(0); if (curI & _Ebit_) { branch = 1; } if (curI & _MDTbit_) { branch = 4; } if (curI & _Ibit_) { mVUinfo |= _isNOP; } - else { incPC(-1); mVUopL(); incPC(1); } + else { incPC(-1); mVUopL(0); incPC(1); } mVUsetCycles(); if (mVU->p) { mVUinfo |= _readP; } if (mVU->q) { mVUinfo |= _readQ; } @@ -279,7 +279,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (isEOB) { x = 0xffff; } if (isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { mVUopL(); incPC(1); doUpperOp(); } + else { mVUopL(1); incPC(1); doUpperOp(); } if (doXGKICK) { mVU_XGKICK_DELAY(); } if (!isBdelay) { incPC(1); } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 379903e412..3097d6c87b 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -206,7 +206,7 @@ microVUx(void) mVUflagPass(u32 startPC, u32 xCount) { incPC(1); if ( curI & _Ebit_ ) { branch = 1; } if ( curI & _MDTbit_ ) { branch = 4; } - if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(); incPC(1); } + if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(3); incPC(1); } if (branch >= 2) { shortBranch(); break; } else if (branch == 1) { branch = 2; } if (mVUbranch) { branch = (mVUbranch >= 9) ? 5 : 3; aBranchAddr = branchAddr; mVUbranch = 0; } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 11de47832c..38598be5b9 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -42,7 +42,7 @@ x86SetJ8(aJump); \ } -microVUf(void) mVU_DIV() { +microVUf(void) mVU_DIV(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFDIV(_Fs_, _Fsf_, _Ft_, _Ftf_, 7); } pass2 { @@ -79,7 +79,7 @@ microVUf(void) mVU_DIV() { pass3 { mVUlog("DIV Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } -microVUf(void) mVU_SQRT() { +microVUf(void) mVU_SQRT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFDIV(0, 0, _Ft_, _Ftf_, 7); } pass2 { @@ -98,7 +98,7 @@ microVUf(void) mVU_SQRT() { pass3 { mVUlog("SQRT Q, vf%02d%s", _Ft_, _Ftf_String); } } -microVUf(void) mVU_RSQRT() { +microVUf(void) mVU_RSQRT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFDIV(_Fs_, _Fsf_, _Ft_, _Ftf_, 13); } pass2 { @@ -169,7 +169,7 @@ microVUt(void) mVU_EATAN_() { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); } -microVUf(void) mVU_EATAN() { +microVUf(void) mVU_EATAN(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 54); } pass2 { @@ -186,7 +186,7 @@ microVUf(void) mVU_EATAN() { pass3 { mVUlog("EATAN P"); } } -microVUf(void) mVU_EATANxy() { +microVUf(void) mVU_EATANxy(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 54); } pass2 { @@ -204,7 +204,7 @@ microVUf(void) mVU_EATANxy() { pass3 { mVUlog("EATANxy P"); } } -microVUf(void) mVU_EATANxz() { +microVUf(void) mVU_EATANxz(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 54); } pass2 { @@ -229,7 +229,7 @@ microVUf(void) mVU_EATANxz() { SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFt); \ } -microVUf(void) mVU_EEXP() { +microVUf(void) mVU_EEXP(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 44); } pass2 { @@ -278,7 +278,7 @@ microVUt(void) mVU_sumXYZ() { } } -microVUf(void) mVU_ELENG() { +microVUf(void) mVU_ELENG(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 18); } pass2 { @@ -291,7 +291,7 @@ microVUf(void) mVU_ELENG() { pass3 { mVUlog("ELENG P"); } } -microVUf(void) mVU_ERCPR() { +microVUf(void) mVU_ERCPR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 12); } pass2 { @@ -306,7 +306,7 @@ microVUf(void) mVU_ERCPR() { pass3 { mVUlog("ERCPR P"); } } -microVUf(void) mVU_ERLENG() { +microVUf(void) mVU_ERLENG(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 24); } pass2 { @@ -322,7 +322,7 @@ microVUf(void) mVU_ERLENG() { pass3 { mVUlog("ERLENG P"); } } -microVUf(void) mVU_ERSADD() { +microVUf(void) mVU_ERSADD(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 18); } pass2 { @@ -338,7 +338,7 @@ microVUf(void) mVU_ERSADD() { pass3 { mVUlog("ERSADD P"); } } -microVUf(void) mVU_ERSQRT() { +microVUf(void) mVU_ERSQRT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 18); } pass2 { @@ -353,7 +353,7 @@ microVUf(void) mVU_ERSQRT() { pass3 { mVUlog("ERSQRT P"); } } -microVUf(void) mVU_ESADD() { +microVUf(void) mVU_ESADD(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 11); } pass2 { @@ -372,7 +372,7 @@ microVUf(void) mVU_ESADD() { SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFs); \ } -microVUf(void) mVU_ESIN() { +microVUf(void) mVU_ESIN(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 29); } pass2 { @@ -399,7 +399,7 @@ microVUf(void) mVU_ESIN() { pass3 { mVUlog("ESIN P"); } } -microVUf(void) mVU_ESQRT() { +microVUf(void) mVU_ESQRT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 12); } pass2 { @@ -411,7 +411,7 @@ microVUf(void) mVU_ESQRT() { pass3 { mVUlog("ESQRT P"); } } -microVUf(void) mVU_ESUM() { +microVUf(void) mVU_ESUM(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeEFU2(_Fs_, 12); } pass2 { @@ -431,7 +431,7 @@ microVUf(void) mVU_ESUM() { // FCAND/FCEQ/FCGET/FCOR/FCSET //------------------------------------------------------------------ -microVUf(void) mVU_FCAND() { +microVUf(void) mVU_FCAND(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeCflag(1); } pass2 { @@ -445,7 +445,7 @@ microVUf(void) mVU_FCAND() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCEQ() { +microVUf(void) mVU_FCEQ(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeCflag(1); } pass2 { @@ -459,7 +459,7 @@ microVUf(void) mVU_FCEQ() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCGET() { +microVUf(void) mVU_FCGET(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeCflag(_It_); } pass2 { @@ -471,7 +471,7 @@ microVUf(void) mVU_FCGET() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCOR() { +microVUf(void) mVU_FCOR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeCflag(1); } pass2 { @@ -485,7 +485,7 @@ microVUf(void) mVU_FCOR() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCSET() { +microVUf(void) mVU_FCSET(mF) { microVU* mVU = mVUx; pass1 { mVUinfo |= _doClip; } pass2 { @@ -499,7 +499,7 @@ microVUf(void) mVU_FCSET() { // FMAND/FMEQ/FMOR //------------------------------------------------------------------ -microVUf(void) mVU_FMAND() { +microVUf(void) mVU_FMAND(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { @@ -512,7 +512,7 @@ microVUf(void) mVU_FMAND() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } -microVUf(void) mVU_FMEQ() { +microVUf(void) mVU_FMEQ(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { @@ -527,7 +527,7 @@ microVUf(void) mVU_FMEQ() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } -microVUf(void) mVU_FMOR() { +microVUf(void) mVU_FMOR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMflag(_Is_, _It_); } pass2 { @@ -544,7 +544,7 @@ microVUf(void) mVU_FMOR() { // FSAND/FSEQ/FSOR/FSSET //------------------------------------------------------------------ -microVUf(void) mVU_FSAND() { +microVUf(void) mVU_FSAND(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSflag(_It_); } pass2 { @@ -556,7 +556,7 @@ microVUf(void) mVU_FSAND() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSEQ() { +microVUf(void) mVU_FSEQ(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSflag(_It_); } pass2 { @@ -570,7 +570,7 @@ microVUf(void) mVU_FSEQ() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSOR() { +microVUf(void) mVU_FSOR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSflag(_It_); } pass2 { @@ -582,7 +582,7 @@ microVUf(void) mVU_FSOR() { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSSET() { +microVUf(void) mVU_FSSET(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFSSET(); } pass2 { @@ -600,7 +600,7 @@ microVUf(void) mVU_FSSET() { // IADD/IADDI/IADDIU/IAND/IOR/ISUB/ISUBIU //------------------------------------------------------------------ -microVUf(void) mVU_IADD() { +microVUf(void) mVU_IADD(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { @@ -615,7 +615,7 @@ microVUf(void) mVU_IADD() { pass3 { mVUlog("IADD vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_IADDI() { +microVUf(void) mVU_IADDI(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { @@ -626,7 +626,7 @@ microVUf(void) mVU_IADDI() { pass3 { mVUlog("IADDI vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm5_); } } -microVUf(void) mVU_IADDIU() { +microVUf(void) mVU_IADDIU(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { @@ -637,7 +637,7 @@ microVUf(void) mVU_IADDIU() { pass3 { mVUlog("IADDIU vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm15_); } } -microVUf(void) mVU_IAND() { +microVUf(void) mVU_IAND(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { @@ -651,7 +651,7 @@ microVUf(void) mVU_IAND() { pass3 { mVUlog("IAND vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_IOR() { +microVUf(void) mVU_IOR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { @@ -665,7 +665,7 @@ microVUf(void) mVU_IOR() { pass3 { mVUlog("IOR vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_ISUB() { +microVUf(void) mVU_ISUB(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } pass2 { @@ -684,7 +684,7 @@ microVUf(void) mVU_ISUB() { pass3 { mVUlog("ISUB vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_ISUBIU() { +microVUf(void) mVU_ISUBIU(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeIALU2(_Is_, _It_); } pass2 { @@ -699,7 +699,7 @@ microVUf(void) mVU_ISUBIU() { // MFIR/MFP/MOVE/MR32/MTIR //------------------------------------------------------------------ -microVUf(void) mVU_MFIR() { +microVUf(void) mVU_MFIR(mF) { microVU* mVU = mVUx; pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } pass2 { @@ -712,7 +712,7 @@ microVUf(void) mVU_MFIR() { pass3 { mVUlog("MFIR.%s vf%02d, vi%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MFP() { +microVUf(void) mVU_MFP(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMFP(_Ft_); } pass2 { @@ -722,7 +722,7 @@ microVUf(void) mVU_MFP() { pass3 { mVUlog("MFP.%s vf%02d, P", _XYZW_String, _Ft_); } } -microVUf(void) mVU_MOVE() { +microVUf(void) mVU_MOVE(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMOVE(_Fs_, _Ft_); } pass2 { @@ -732,7 +732,7 @@ microVUf(void) mVU_MOVE() { pass3 { mVUlog("MOVE.%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MR32() { +microVUf(void) mVU_MR32(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeMR32(_Fs_, _Ft_); } pass2 { @@ -743,7 +743,7 @@ microVUf(void) mVU_MR32() { pass3 { mVUlog("MR32.%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MTIR() { +microVUf(void) mVU_MTIR(mF) { microVU* mVU = mVUx; pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } pass2 { @@ -757,7 +757,7 @@ microVUf(void) mVU_MTIR() { // ILW/ILWR //------------------------------------------------------------------ -microVUf(void) mVU_ILW() { +microVUf(void) mVU_ILW(mF) { microVU* mVU = mVUx; pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { @@ -776,7 +776,7 @@ microVUf(void) mVU_ILW() { pass3 { mVUlog("ILW.%s vi%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_ILWR() { +microVUf(void) mVU_ILWR(mF) { microVU* mVU = mVUx; pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { @@ -798,7 +798,7 @@ microVUf(void) mVU_ILWR() { // ISW/ISWR //------------------------------------------------------------------ -microVUf(void) mVU_ISW() { +microVUf(void) mVU_ISW(mF) { microVU* mVU = mVUx; pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { @@ -824,7 +824,7 @@ microVUf(void) mVU_ISW() { pass3 { mVUlog("ISW.%s vi%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_ISWR() { +microVUf(void) mVU_ISWR(mF) { microVU* mVU = mVUx; pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { @@ -852,7 +852,7 @@ microVUf(void) mVU_ISWR() { // LQ/LQD/LQI //------------------------------------------------------------------ -microVUf(void) mVU_LQ() { +microVUf(void) mVU_LQ(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeLQ(_Ft_, _Is_, 0); } pass2 { @@ -871,7 +871,7 @@ microVUf(void) mVU_LQ() { pass3 { mVUlog("LQ.%s vf%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_LQD() { +microVUf(void) mVU_LQD(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } pass2 { @@ -893,7 +893,7 @@ microVUf(void) mVU_LQD() { pass3 { mVUlog("LQD.%s vf%02d, --vi%02d", _XYZW_String, _Ft_, _Is_); } } -microVUf(void) mVU_LQI() { +microVUf(void) mVU_LQI(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } pass2 { @@ -920,7 +920,7 @@ microVUf(void) mVU_LQI() { // SQ/SQD/SQI //------------------------------------------------------------------ -microVUf(void) mVU_SQ() { +microVUf(void) mVU_SQ(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSQ(_Fs_, _It_, 0); } pass2 { @@ -939,7 +939,7 @@ microVUf(void) mVU_SQ() { pass3 { mVUlog("SQ.%s vf%02d, vi%02d + %d", _XYZW_String, _Fs_, _Ft_, _Imm11_); } } -microVUf(void) mVU_SQD() { +microVUf(void) mVU_SQD(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } pass2 { @@ -959,7 +959,7 @@ microVUf(void) mVU_SQD() { pass3 { mVUlog("SQD.%s vf%02d, --vi%02d", _XYZW_String, _Fs_, _Ft_); } } -microVUf(void) mVU_SQI() { +microVUf(void) mVU_SQI(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } pass2 { @@ -984,7 +984,7 @@ microVUf(void) mVU_SQI() { // RINIT/RGET/RNEXT/RXOR //------------------------------------------------------------------ -microVUf(void) mVU_RINIT() { +microVUf(void) mVU_RINIT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeR1(_Fs_, _Fsf_); } pass2 { @@ -1009,14 +1009,14 @@ microVUt(void) mVU_RGET_(int Rreg) { } } -microVUf(void) mVU_RGET() { +microVUf(void) mVU_RGET(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeR2(_Ft_, 1); } pass2 { MOV32MtoR(gprT1, Rmem); mVU_RGET_(gprT1); } pass3 { mVUlog("RGET.%s vf%02d, R", _XYZW_String, _Ft_); } } -microVUf(void) mVU_RNEXT() { +microVUf(void) mVU_RNEXT(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeR2(_Ft_, 0); } pass2 { @@ -1041,7 +1041,7 @@ microVUf(void) mVU_RNEXT() { pass3 { mVUlog("RNEXT.%s vf%02d, R", _XYZW_String, _Ft_); } } -microVUf(void) mVU_RXOR() { +microVUf(void) mVU_RXOR(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeR1(_Fs_, _Fsf_); } pass2 { @@ -1058,13 +1058,13 @@ microVUf(void) mVU_RXOR() { // WaitP/WaitQ //------------------------------------------------------------------ -microVUf(void) mVU_WAITP() { +microVUf(void) mVU_WAITP(mF) { microVU* mVU = mVUx; pass1 { mVUstall = aMax(mVUstall, ((mVUregs.p) ? (mVUregs.p - 1) : 0)); } pass3 { mVUlog("WAITP"); } } -microVUf(void) mVU_WAITQ() { +microVUf(void) mVU_WAITQ(mF) { microVU* mVU = mVUx; pass1 { mVUstall = aMax(mVUstall, mVUregs.q); } pass3 { mVUlog("WAITQ"); } @@ -1074,7 +1074,7 @@ microVUf(void) mVU_WAITQ() { // XTOP/XITOP //------------------------------------------------------------------ -microVUf(void) mVU_XTOP() { +microVUf(void) mVU_XTOP(mF) { microVU* mVU = mVUx; pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { @@ -1084,7 +1084,7 @@ microVUf(void) mVU_XTOP() { pass3 { mVUlog("XTOP vi%02d", _Ft_); } } -microVUf(void) mVU_XITOP() { +microVUf(void) mVU_XITOP(mF) { microVU* mVU = mVUx; pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { @@ -1111,7 +1111,7 @@ void __fastcall mVU_XGKICK__(u32 addr) { GSGIFTRANSFER1((u32*)microVU1.regs->Mem, ((addr<<4)&0x3fff)); } -microVUf(void) mVU_XGKICK() { +microVUf(void) mVU_XGKICK(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeXGkick(_Is_, mVU_XGKICK_CYCLES); } pass2 { @@ -1142,13 +1142,13 @@ microVUt(void) mVU_XGKICK_DELAY() { pass4 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ } -microVUf(void) mVU_B() { +microVUf(void) mVU_B(mF) { microVU* mVU = mVUx; setBranchA(1, 0); pass3 { mVUlog("B [%04x]", branchAddr, branchAddr); } } -microVUf(void) mVU_BAL() { +microVUf(void) mVU_BAL(mF) { microVU* mVU = mVUx; setBranchA(2, _It_); pass1 { analyzeVIreg2(_It_, 1); } @@ -1159,7 +1159,7 @@ microVUf(void) mVU_BAL() { pass3 { mVUlog("BAL vi%02d [%04x]", _Ft_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBEQ() { +microVUf(void) mVU_IBEQ(mF) { microVU* mVU = mVUx; setBranchA(3, 0); pass1 { mVUanalyzeBranch2(_Is_, _It_); } @@ -1173,7 +1173,7 @@ microVUf(void) mVU_IBEQ() { pass3 { mVUlog("IBEQ vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBGEZ() { +microVUf(void) mVU_IBGEZ(mF) { microVU* mVU = mVUx; setBranchA(4, 0); pass1 { mVUanalyzeBranch1(_Is_); } @@ -1185,7 +1185,7 @@ microVUf(void) mVU_IBGEZ() { pass3 { mVUlog("IBGEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBGTZ() { +microVUf(void) mVU_IBGTZ(mF) { microVU* mVU = mVUx; setBranchA(5, 0); pass1 { mVUanalyzeBranch1(_Is_); } @@ -1197,7 +1197,7 @@ microVUf(void) mVU_IBGTZ() { pass3 { mVUlog("IBGTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBLEZ() { +microVUf(void) mVU_IBLEZ(mF) { microVU* mVU = mVUx; setBranchA(6, 0); pass1 { mVUanalyzeBranch1(_Is_); } @@ -1209,7 +1209,7 @@ microVUf(void) mVU_IBLEZ() { pass3 { mVUlog("IBLEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBLTZ() { +microVUf(void) mVU_IBLTZ(mF) { microVU* mVU = mVUx; setBranchA(7, 0); pass1 { mVUanalyzeBranch1(_Is_); } @@ -1221,7 +1221,7 @@ microVUf(void) mVU_IBLTZ() { pass3 { mVUlog("IBLTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBNE() { +microVUf(void) mVU_IBNE(mF) { microVU* mVU = mVUx; setBranchA(8, 0); pass1 { mVUanalyzeBranch2(_Is_, _It_); } @@ -1235,7 +1235,7 @@ microVUf(void) mVU_IBNE() { pass3 { mVUlog("IBNE vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_JR() { +microVUf(void) mVU_JR(mF) { microVU* mVU = mVUx; mVUbranch = 9; pass1 { analyzeVIreg1(_Is_); } @@ -1248,7 +1248,7 @@ microVUf(void) mVU_JR() { pass3 { mVUlog("JR [vi%02d]", _Fs_); } } -microVUf(void) mVU_JALR() { +microVUf(void) mVU_JALR(mF) { microVU* mVU = mVUx; mVUbranch = 10; pass1 { analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 1); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 37efbff089..695c9ac694 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -140,8 +140,11 @@ declareAllVariables #define mVUx (vuIndex ? µVU1 : µVU0) #define microVUt(aType) template __forceinline aType #define microVUx(aType) template aType -#define microVUf(aType) template aType -#define microVUq(aType) template __forceinline aType +#define microVUf(aType) template aType +#define microVUq(aType) template __forceinline aType + +// Function Params +#define mF int recPass // Define Passes #define pass1 if (recPass == 0) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 0d7ebe631d..8584169a41 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -306,7 +306,7 @@ microVUt(void) mVUcheckSflag(int progIndex) { mVUsFlagHack = 1; for (u32 i = 0; i < mVU->progSize; i+=2) { mVU->code = mVU->prog.prog[progIndex].data[i]; - mVUopL(); + mVUopL(3); } mVUflagInfo = bFlagInfo; mVU->code = bCode; diff --git a/pcsx2/x86/microVU_Tables.inl b/pcsx2/x86/microVU_Tables.inl index a6f5cbac61..64d3ac682d 100644 --- a/pcsx2/x86/microVU_Tables.inl +++ b/pcsx2/x86/microVU_Tables.inl @@ -23,180 +23,180 @@ //------------------------------------------------------------------ #define mVUgetCode (vuIndex ? microVU1.code : microVU0.code) -microVUf(void) mVU_UPPER_FD_00(); -microVUf(void) mVU_UPPER_FD_01(); -microVUf(void) mVU_UPPER_FD_10(); -microVUf(void) mVU_UPPER_FD_11(); -microVUf(void) mVULowerOP(); -microVUf(void) mVULowerOP_T3_00(); -microVUf(void) mVULowerOP_T3_01(); -microVUf(void) mVULowerOP_T3_10(); -microVUf(void) mVULowerOP_T3_11(); -microVUf(void) mVUunknown(); +microVUf(void) mVU_UPPER_FD_00(mF); +microVUf(void) mVU_UPPER_FD_01(mF); +microVUf(void) mVU_UPPER_FD_10(mF); +microVUf(void) mVU_UPPER_FD_11(mF); +microVUf(void) mVULowerOP(mF); +microVUf(void) mVULowerOP_T3_00(mF); +microVUf(void) mVULowerOP_T3_01(mF); +microVUf(void) mVULowerOP_T3_10(mF); +microVUf(void) mVULowerOP_T3_11(mF); +microVUf(void) mVUunknown(mF); //------------------------------------------------------------------ //------------------------------------------------------------------ // Opcode Tables //------------------------------------------------------------------ -#define microVU_LOWER_OPCODE(x, y) void (* mVULOWER_OPCODE##x##y [128])() = { \ - mVU_LQ , mVU_SQ , mVUunknown , mVUunknown, \ - mVU_ILW , mVU_ISW , mVUunknown , mVUunknown, \ - mVU_IADDIU , mVU_ISUBIU , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_FCEQ , mVU_FCSET , mVU_FCAND , mVU_FCOR, /* 0x10 */ \ - mVU_FSEQ , mVU_FSSET , mVU_FSAND , mVU_FSOR, \ - mVU_FMEQ , mVUunknown , mVU_FMAND , mVU_FMOR, \ - mVU_FCGET , mVUunknown , mVUunknown , mVUunknown, \ - mVU_B , mVU_BAL , mVUunknown , mVUunknown, /* 0x20 */ \ - mVU_JR , mVU_JALR , mVUunknown , mVUunknown, \ - mVU_IBEQ , mVU_IBNE , mVUunknown , mVUunknown, \ - mVU_IBLTZ , mVU_IBGTZ , mVU_IBLEZ , mVU_IBGEZ, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVULowerOP , mVUunknown , mVUunknown , mVUunknown, /* 0x40*/ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x50 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x60 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x70 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_LOWER_OPCODE(x) void (*mVULOWER_OPCODE##x [128])(mF) = { \ + mVU_LQ , mVU_SQ , mVUunknown , mVUunknown, \ + mVU_ILW , mVU_ISW , mVUunknown , mVUunknown, \ + mVU_IADDIU , mVU_ISUBIU , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_FCEQ , mVU_FCSET , mVU_FCAND , mVU_FCOR, /* 0x10 */ \ + mVU_FSEQ , mVU_FSSET , mVU_FSAND , mVU_FSOR, \ + mVU_FMEQ , mVUunknown , mVU_FMAND , mVU_FMOR, \ + mVU_FCGET , mVUunknown , mVUunknown , mVUunknown, \ + mVU_B , mVU_BAL , mVUunknown , mVUunknown, /* 0x20 */ \ + mVU_JR , mVU_JALR , mVUunknown , mVUunknown, \ + mVU_IBEQ , mVU_IBNE , mVUunknown , mVUunknown, \ + mVU_IBLTZ , mVU_IBGTZ , mVU_IBLEZ , mVU_IBGEZ, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP , mVUunknown , mVUunknown , mVUunknown, /* 0x40*/ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x50 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x60 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x70 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_LowerOP_T3_00_OPCODE(x, y) void (* mVULowerOP_T3_00_OPCODE##x##y [32])() = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_MOVE , mVU_LQI , mVU_DIV , mVU_MTIR, \ - mVU_RNEXT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_MFP , mVU_XTOP , mVU_XGKICK, \ - mVU_ESADD , mVU_EATANxy , mVU_ESQRT , mVU_ESIN, \ +#define microVU_LowerOP_T3_00_OPCODE(x) void (*mVULowerOP_T3_00_OPCODE##x [32])(mF) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MOVE , mVU_LQI , mVU_DIV , mVU_MTIR, \ + mVU_RNEXT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_MFP , mVU_XTOP , mVU_XGKICK, \ + mVU_ESADD , mVU_EATANxy, mVU_ESQRT , mVU_ESIN, \ }; -#define microVU_LowerOP_T3_01_OPCODE(x, y) void (* mVULowerOP_T3_01_OPCODE##x##y [32])() = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_MR32 , mVU_SQI , mVU_SQRT , mVU_MFIR, \ - mVU_RGET , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVU_XITOP , mVUunknown, \ - mVU_ERSADD , mVU_EATANxz , mVU_ERSQRT , mVU_EATAN, \ +#define microVU_LowerOP_T3_01_OPCODE(x) void (*mVULowerOP_T3_01_OPCODE##x [32])(mF) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MR32 , mVU_SQI , mVU_SQRT , mVU_MFIR, \ + mVU_RGET , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVU_XITOP , mVUunknown, \ + mVU_ERSADD , mVU_EATANxz, mVU_ERSQRT , mVU_EATAN, \ }; -#define microVU_LowerOP_T3_10_OPCODE(x, y) void (* mVULowerOP_T3_10_OPCODE##x##y [32])() = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_LQD , mVU_RSQRT , mVU_ILWR, \ - mVU_RINIT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_ELENG , mVU_ESUM , mVU_ERCPR , mVU_EEXP, \ +#define microVU_LowerOP_T3_10_OPCODE(x) void (*mVULowerOP_T3_10_OPCODE##x [32])(mF) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_LQD , mVU_RSQRT , mVU_ILWR, \ + mVU_RINIT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ELENG , mVU_ESUM , mVU_ERCPR , mVU_EEXP, \ }; -#define microVU_LowerOP_T3_11_OPCODE(x, y) void (* mVULowerOP_T3_11_OPCODE##x##y [32])() = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_SQD , mVU_WAITQ , mVU_ISWR, \ - mVU_RXOR , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_ERLENG , mVUunknown , mVU_WAITP , mVUunknown, \ +#define microVU_LowerOP_T3_11_OPCODE(x) void (*mVULowerOP_T3_11_OPCODE##x [32])(mF) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_SQD , mVU_WAITQ , mVU_ISWR, \ + mVU_RXOR , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ERLENG , mVUunknown , mVU_WAITP , mVUunknown, \ }; -#define microVU_LowerOP_OPCODE(x, y) void (* mVULowerOP_OPCODE##x##y [64])() = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x20 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_IADD , mVU_ISUB , mVU_IADDI , mVUunknown, /* 0x30 */ \ - mVU_IAND , mVU_IOR , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVULowerOP_T3_00, mVULowerOP_T3_01, mVULowerOP_T3_10, mVULowerOP_T3_11, \ +#define microVU_LowerOP_OPCODE(x) void (*mVULowerOP_OPCODE##x [64])(mF) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x20 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_IADD , mVU_ISUB , mVU_IADDI , mVUunknown, /* 0x30 */ \ + mVU_IAND , mVU_IOR , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP_T3_00, mVULowerOP_T3_01, mVULowerOP_T3_10, mVULowerOP_T3_11, \ }; -#define microVU_UPPER_OPCODE(x, y) void (* mVU_UPPER_OPCODE##x##y [64])() = { \ - mVU_ADDx , mVU_ADDy , mVU_ADDz , mVU_ADDw, \ - mVU_SUBx , mVU_SUBy , mVU_SUBz , mVU_SUBw, \ - mVU_MADDx , mVU_MADDy , mVU_MADDz , mVU_MADDw, \ - mVU_MSUBx , mVU_MSUBy , mVU_MSUBz , mVU_MSUBw, \ - mVU_MAXx , mVU_MAXy , mVU_MAXz , mVU_MAXw, /* 0x10 */ \ - mVU_MINIx , mVU_MINIy , mVU_MINIz , mVU_MINIw, \ - mVU_MULx , mVU_MULy , mVU_MULz , mVU_MULw, \ - mVU_MULq , mVU_MAXi , mVU_MULi , mVU_MINIi, \ - mVU_ADDq , mVU_MADDq , mVU_ADDi , mVU_MADDi, /* 0x20 */ \ - mVU_SUBq , mVU_MSUBq , mVU_SUBi , mVU_MSUBi, \ - mVU_ADD , mVU_MADD , mVU_MUL , mVU_MAX, \ - mVU_SUB , mVU_MSUB , mVU_OPMSUB , mVU_MINI, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_UPPER_FD_00, mVU_UPPER_FD_01, mVU_UPPER_FD_10, mVU_UPPER_FD_11, \ +#define microVU_UPPER_OPCODE(x) void (*mVU_UPPER_OPCODE##x [64])(mF) = { \ + mVU_ADDx , mVU_ADDy , mVU_ADDz , mVU_ADDw, \ + mVU_SUBx , mVU_SUBy , mVU_SUBz , mVU_SUBw, \ + mVU_MADDx , mVU_MADDy , mVU_MADDz , mVU_MADDw, \ + mVU_MSUBx , mVU_MSUBy , mVU_MSUBz , mVU_MSUBw, \ + mVU_MAXx , mVU_MAXy , mVU_MAXz , mVU_MAXw, /* 0x10 */ \ + mVU_MINIx , mVU_MINIy , mVU_MINIz , mVU_MINIw, \ + mVU_MULx , mVU_MULy , mVU_MULz , mVU_MULw, \ + mVU_MULq , mVU_MAXi , mVU_MULi , mVU_MINIi, \ + mVU_ADDq , mVU_MADDq , mVU_ADDi , mVU_MADDi, /* 0x20 */ \ + mVU_SUBq , mVU_MSUBq , mVU_SUBi , mVU_MSUBi, \ + mVU_ADD , mVU_MADD , mVU_MUL , mVU_MAX, \ + mVU_SUB , mVU_MSUB , mVU_OPMSUB , mVU_MINI, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_UPPER_FD_00, mVU_UPPER_FD_01, mVU_UPPER_FD_10, mVU_UPPER_FD_11, \ }; -#define microVU_UPPER_FD_00_TABLE(x, y) void (* mVU_UPPER_FD_00_TABLE##x##y [32])() = { \ - mVU_ADDAx , mVU_SUBAx , mVU_MADDAx , mVU_MSUBAx, \ - mVU_ITOF0 , mVU_FTOI0 , mVU_MULAx , mVU_MULAq, \ - mVU_ADDAq , mVU_SUBAq , mVU_ADDA , mVU_SUBA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_00_TABLE(x) void (*mVU_UPPER_FD_00_TABLE##x [32])(mF) = { \ + mVU_ADDAx , mVU_SUBAx , mVU_MADDAx , mVU_MSUBAx, \ + mVU_ITOF0 , mVU_FTOI0 , mVU_MULAx , mVU_MULAq, \ + mVU_ADDAq , mVU_SUBAq , mVU_ADDA , mVU_SUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_01_TABLE(x, y) void (* mVU_UPPER_FD_01_TABLE##x##y [32])() = { \ - mVU_ADDAy , mVU_SUBAy , mVU_MADDAy , mVU_MSUBAy, \ - mVU_ITOF4 , mVU_FTOI4 , mVU_MULAy , mVU_ABS, \ - mVU_MADDAq , mVU_MSUBAq , mVU_MADDA , mVU_MSUBA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_01_TABLE(x) void (* mVU_UPPER_FD_01_TABLE##x [32])(mF) = { \ + mVU_ADDAy , mVU_SUBAy , mVU_MADDAy , mVU_MSUBAy, \ + mVU_ITOF4 , mVU_FTOI4 , mVU_MULAy , mVU_ABS, \ + mVU_MADDAq , mVU_MSUBAq , mVU_MADDA , mVU_MSUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_10_TABLE(x, y) void (* mVU_UPPER_FD_10_TABLE##x##y [32])() = { \ - mVU_ADDAz , mVU_SUBAz , mVU_MADDAz , mVU_MSUBAz, \ - mVU_ITOF12 , mVU_FTOI12 , mVU_MULAz , mVU_MULAi, \ - mVU_ADDAi , mVU_SUBAi , mVU_MULA , mVU_OPMULA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_10_TABLE(x) void (* mVU_UPPER_FD_10_TABLE##x [32])(mF) = { \ + mVU_ADDAz , mVU_SUBAz , mVU_MADDAz , mVU_MSUBAz, \ + mVU_ITOF12 , mVU_FTOI12 , mVU_MULAz , mVU_MULAi, \ + mVU_ADDAi , mVU_SUBAi , mVU_MULA , mVU_OPMULA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_11_TABLE(x, y) void (* mVU_UPPER_FD_11_TABLE##x##y [32])() = { \ - mVU_ADDAw , mVU_SUBAw , mVU_MADDAw , mVU_MSUBAw, \ - mVU_ITOF15 , mVU_FTOI15 , mVU_MULAw , mVU_CLIP, \ - mVU_MADDAi , mVU_MSUBAi , mVUunknown , mVU_NOP, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_11_TABLE(x) void (* mVU_UPPER_FD_11_TABLE##x [32])(mF) = { \ + mVU_ADDAw , mVU_SUBAw , mVU_MADDAw , mVU_MSUBAw, \ + mVU_ITOF15 , mVU_FTOI15 , mVU_MULAw , mVU_CLIP, \ + mVU_MADDAi , mVU_MSUBAi , mVUunknown , mVU_NOP, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; //------------------------------------------------------------------ @@ -204,62 +204,42 @@ microVUf(void) mVUunknown(); //------------------------------------------------------------------ // Create Table Instances //------------------------------------------------------------------ -#define mVUcreateTable(x,y) \ -microVU_LOWER_OPCODE(x,y) \ -microVU_LowerOP_T3_00_OPCODE(x,y) \ -microVU_LowerOP_T3_01_OPCODE(x,y) \ -microVU_LowerOP_T3_10_OPCODE(x,y) \ -microVU_LowerOP_T3_11_OPCODE(x,y) \ -microVU_LowerOP_OPCODE(x,y) \ -microVU_UPPER_OPCODE(x,y) \ -microVU_UPPER_FD_00_TABLE(x,y) \ -microVU_UPPER_FD_01_TABLE(x,y) \ -microVU_UPPER_FD_10_TABLE(x,y) \ -microVU_UPPER_FD_11_TABLE(x,y) +#define mVUcreateTable(x) \ +microVU_LOWER_OPCODE(x) \ +microVU_LowerOP_T3_00_OPCODE(x) \ +microVU_LowerOP_T3_01_OPCODE(x) \ +microVU_LowerOP_T3_10_OPCODE(x) \ +microVU_LowerOP_T3_11_OPCODE(x) \ +microVU_LowerOP_OPCODE(x) \ +microVU_UPPER_OPCODE(x) \ +microVU_UPPER_FD_00_TABLE(x) \ +microVU_UPPER_FD_01_TABLE(x) \ +microVU_UPPER_FD_10_TABLE(x) \ +microVU_UPPER_FD_11_TABLE(x) -mVUcreateTable(0,0) -mVUcreateTable(0,1) -mVUcreateTable(0,2) -mVUcreateTable(0,3) -mVUcreateTable(1,0) -mVUcreateTable(1,1) -mVUcreateTable(1,2) -mVUcreateTable(1,3) +mVUcreateTable(0) +mVUcreateTable(1) //------------------------------------------------------------------ // Table Functions //------------------------------------------------------------------ #define doTableStuff(tableName, args) { \ - pass1 { \ - if (vuIndex) tableName##10[ args ](); \ - else tableName##00[ args ](); \ - } \ - pass2 { \ - if (vuIndex) tableName##11[ args ](); \ - else tableName##01[ args ](); \ - } \ - pass3 { \ - if (vuIndex) tableName##12[ args ](); \ - else tableName##02[ args ](); \ - } \ - pass4 { \ - if (vuIndex) tableName##13[ args ](); \ - else tableName##03[ args ](); \ - } \ + if (vuIndex) tableName##1[ args ](recPass); \ + else tableName##0[ args ](recPass); \ } -microVUf(void) mVU_UPPER_FD_00() { doTableStuff(mVU_UPPER_FD_00_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_01() { doTableStuff(mVU_UPPER_FD_01_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_10() { doTableStuff(mVU_UPPER_FD_10_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_11() { doTableStuff(mVU_UPPER_FD_11_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP() { doTableStuff(mVULowerOP_OPCODE, (mVUgetCode & 0x3f)); } -microVUf(void) mVULowerOP_T3_00() { doTableStuff(mVULowerOP_T3_00_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_01() { doTableStuff(mVULowerOP_T3_01_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_10() { doTableStuff(mVULowerOP_T3_10_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_11() { doTableStuff(mVULowerOP_T3_11_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVUopU() { doTableStuff(mVU_UPPER_OPCODE, (mVUgetCode & 0x3f)); } // Gets Upper Opcode -microVUf(void) mVUopL() { doTableStuff(mVULOWER_OPCODE, (mVUgetCode >> 25)); } // Gets Lower Opcode -microVUf(void) mVUunknown() { +microVUf(void) mVU_UPPER_FD_00(mF) { doTableStuff(mVU_UPPER_FD_00_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVU_UPPER_FD_01(mF) { doTableStuff(mVU_UPPER_FD_01_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVU_UPPER_FD_10(mF) { doTableStuff(mVU_UPPER_FD_10_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVU_UPPER_FD_11(mF) { doTableStuff(mVU_UPPER_FD_11_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVULowerOP(mF) { doTableStuff(mVULowerOP_OPCODE, (mVUgetCode & 0x3f)); } +microVUf(void) mVULowerOP_T3_00(mF) { doTableStuff(mVULowerOP_T3_00_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVULowerOP_T3_01(mF) { doTableStuff(mVULowerOP_T3_01_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVULowerOP_T3_10(mF) { doTableStuff(mVULowerOP_T3_10_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVULowerOP_T3_11(mF) { doTableStuff(mVULowerOP_T3_11_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +microVUf(void) mVUopU(mF) { doTableStuff(mVU_UPPER_OPCODE, (mVUgetCode & 0x3f)); } // Gets Upper Opcode +microVUf(void) mVUopL(mF) { doTableStuff(mVULOWER_OPCODE, (mVUgetCode >> 25)); } // Gets Lower Opcode +microVUf(void) mVUunknown(mF) { pass2 { SysPrintf("microVU%d: Unknown Micro VU opcode called (%x)\n", vuIndex, mVUgetCode); } pass3 { mVUlog("Unknown", mVUgetCode); } } diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 5a7c2c6383..ca1b31b4ba 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -469,7 +469,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX // Micro VU Micromode Upper instructions //------------------------------------------------------------------ -microVUf(void) mVU_ABS() { +microVUf(void) mVU_ABS(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } pass2 { @@ -480,92 +480,92 @@ microVUf(void) mVU_ABS() { } pass3 { mVUlog("ABS"); mVUlogFtFs(); } } -microVUf(void) mVU_ADD() { mVU_FMAC1 (ADD, "ADD"); } -microVUf(void) mVU_ADDi() { mVU_FMAC6 (ADD2, "ADDi"); } -microVUf(void) mVU_ADDq() { mVU_FMAC22(ADD, "ADDq"); } -microVUf(void) mVU_ADDx() { mVU_FMAC3 (ADD, "ADDx"); } -microVUf(void) mVU_ADDy() { mVU_FMAC3 (ADD, "ADDy"); } -microVUf(void) mVU_ADDz() { mVU_FMAC3 (ADD, "ADDz"); } -microVUf(void) mVU_ADDw() { mVU_FMAC3 (ADD, "ADDw"); } -microVUf(void) mVU_ADDA() { mVU_FMAC4 (ADD, "ADDA"); } -microVUf(void) mVU_ADDAi() { mVU_FMAC7 (ADD, "ADDAi"); } -microVUf(void) mVU_ADDAq() { mVU_FMAC23(ADD, "ADDAq"); } -microVUf(void) mVU_ADDAx() { mVU_FMAC5 (ADD, "ADDAx"); } -microVUf(void) mVU_ADDAy() { mVU_FMAC5 (ADD, "ADDAy"); } -microVUf(void) mVU_ADDAz() { mVU_FMAC5 (ADD, "ADDAz"); } -microVUf(void) mVU_ADDAw() { mVU_FMAC5 (ADD, "ADDAw"); } -microVUf(void) mVU_SUB() { mVU_FMAC1 (SUB, "SUB"); } -microVUf(void) mVU_SUBi() { mVU_FMAC6 (SUB, "SUBi"); } -microVUf(void) mVU_SUBq() { mVU_FMAC22(SUB, "SUBq"); } -microVUf(void) mVU_SUBx() { mVU_FMAC3 (SUB, "SUBx"); } -microVUf(void) mVU_SUBy() { mVU_FMAC3 (SUB, "SUBy"); } -microVUf(void) mVU_SUBz() { mVU_FMAC3 (SUB, "SUBz"); } -microVUf(void) mVU_SUBw() { mVU_FMAC3 (SUB, "SUBw"); } -microVUf(void) mVU_SUBA() { mVU_FMAC4 (SUB, "SUBA"); } -microVUf(void) mVU_SUBAi() { mVU_FMAC7 (SUB, "SUBAi"); } -microVUf(void) mVU_SUBAq() { mVU_FMAC23(SUB, "SUBAq"); } -microVUf(void) mVU_SUBAx() { mVU_FMAC5 (SUB, "SUBAx"); } -microVUf(void) mVU_SUBAy() { mVU_FMAC5 (SUB, "SUBAy"); } -microVUf(void) mVU_SUBAz() { mVU_FMAC5 (SUB, "SUBAz"); } -microVUf(void) mVU_SUBAw() { mVU_FMAC5 (SUB, "SUBAw"); } -microVUf(void) mVU_MUL() { mVU_FMAC1 (MUL, "MUL"); } -microVUf(void) mVU_MULi() { mVU_FMAC6 (MUL, "MULi"); } -microVUf(void) mVU_MULq() { mVU_FMAC22(MUL, "MULq"); } -microVUf(void) mVU_MULx() { mVU_FMAC3 (MUL, "MULx"); } -microVUf(void) mVU_MULy() { mVU_FMAC3 (MUL, "MULy"); } -microVUf(void) mVU_MULz() { mVU_FMAC3 (MUL, "MULz"); } -microVUf(void) mVU_MULw() { mVU_FMAC3 (MUL, "MULw"); } -microVUf(void) mVU_MULA() { mVU_FMAC4 (MUL, "MULA"); } -microVUf(void) mVU_MULAi() { mVU_FMAC7 (MUL, "MULAi"); } -microVUf(void) mVU_MULAq() { mVU_FMAC23(MUL, "MULAq"); } -microVUf(void) mVU_MULAx() { mVU_FMAC5 (MUL, "MULAx"); } -microVUf(void) mVU_MULAy() { mVU_FMAC5 (MUL, "MULAy"); } -microVUf(void) mVU_MULAz() { mVU_FMAC5 (MUL, "MULAz"); } -microVUf(void) mVU_MULAw() { mVU_FMAC5 (MUL, "MULAw"); } -microVUf(void) mVU_MADD() { mVU_FMAC8 (ADD, "MADD"); } -microVUf(void) mVU_MADDi() { mVU_FMAC12(ADD, "MADDi"); } -microVUf(void) mVU_MADDq() { mVU_FMAC24(ADD, "MADDq"); } -microVUf(void) mVU_MADDx() { mVU_FMAC10(ADD, "MADDx"); } -microVUf(void) mVU_MADDy() { mVU_FMAC10(ADD, "MADDy"); } -microVUf(void) mVU_MADDz() { mVU_FMAC10(ADD, "MADDz"); } -microVUf(void) mVU_MADDw() { mVU_FMAC10(ADD, "MADDw"); } -microVUf(void) mVU_MADDA() { mVU_FMAC14(ADD, "MADDA"); } -microVUf(void) mVU_MADDAi() { mVU_FMAC16(ADD, "MADDAi"); } -microVUf(void) mVU_MADDAq() { mVU_FMAC26(ADD, "MADDAq"); } -microVUf(void) mVU_MADDAx() { mVU_FMAC15(ADD, "MADDAx"); } -microVUf(void) mVU_MADDAy() { mVU_FMAC15(ADD, "MADDAy"); } -microVUf(void) mVU_MADDAz() { mVU_FMAC15(ADD, "MADDAz"); } -microVUf(void) mVU_MADDAw() { mVU_FMAC15(ADD, "MADDAw"); } -microVUf(void) mVU_MSUB() { mVU_FMAC9 (SUB, "MSUB"); } -microVUf(void) mVU_MSUBi() { mVU_FMAC13(SUB, "MSUBi"); } -microVUf(void) mVU_MSUBq() { mVU_FMAC25(SUB, "MSUBq"); } -microVUf(void) mVU_MSUBx() { mVU_FMAC11(SUB, "MSUBx"); } -microVUf(void) mVU_MSUBy() { mVU_FMAC11(SUB, "MSUBy"); } -microVUf(void) mVU_MSUBz() { mVU_FMAC11(SUB, "MSUBz"); } -microVUf(void) mVU_MSUBw() { mVU_FMAC11(SUB, "MSUBw"); } -microVUf(void) mVU_MSUBA() { mVU_FMAC14(SUB, "MSUBA"); } -microVUf(void) mVU_MSUBAi() { mVU_FMAC16(SUB, "MSUBAi"); } -microVUf(void) mVU_MSUBAq() { mVU_FMAC26(SUB, "MSUBAq"); } -microVUf(void) mVU_MSUBAx() { mVU_FMAC15(SUB, "MSUBAx"); } -microVUf(void) mVU_MSUBAy() { mVU_FMAC15(SUB, "MSUBAy"); } -microVUf(void) mVU_MSUBAz() { mVU_FMAC15(SUB, "MSUBAz"); } -microVUf(void) mVU_MSUBAw() { mVU_FMAC15(SUB, "MSUBAw"); } -microVUf(void) mVU_MAX() { mVU_FMAC27(MAX2, "MAX"); } -microVUf(void) mVU_MAXi() { mVU_FMAC28(MAX2, "MAXi"); } -microVUf(void) mVU_MAXx() { mVU_FMAC29(MAX2, "MAXx"); } -microVUf(void) mVU_MAXy() { mVU_FMAC29(MAX2, "MAXy"); } -microVUf(void) mVU_MAXz() { mVU_FMAC29(MAX2, "MAXz"); } -microVUf(void) mVU_MAXw() { mVU_FMAC29(MAX2, "MAXw"); } -microVUf(void) mVU_MINI() { mVU_FMAC27(MIN2, "MINI"); } -microVUf(void) mVU_MINIi() { mVU_FMAC28(MIN2, "MINIi"); } -microVUf(void) mVU_MINIx() { mVU_FMAC29(MIN2, "MINIx"); } -microVUf(void) mVU_MINIy() { mVU_FMAC29(MIN2, "MINIy"); } -microVUf(void) mVU_MINIz() { mVU_FMAC29(MIN2, "MINIz"); } -microVUf(void) mVU_MINIw() { mVU_FMAC29(MIN2, "MINIw"); } -microVUf(void) mVU_OPMULA() { mVU_FMAC18(MUL, "OPMULA"); } -microVUf(void) mVU_OPMSUB() { mVU_FMAC19(SUB, "OPMSUB"); } -microVUf(void) mVU_NOP() { pass3 { mVUlog("NOP"); } } -microVUq(void) mVU_FTOIx(uptr addr) { +microVUf(void) mVU_ADD(mF) { mVU_FMAC1 (ADD, "ADD"); } +microVUf(void) mVU_ADDi(mF) { mVU_FMAC6 (ADD2, "ADDi"); } +microVUf(void) mVU_ADDq(mF) { mVU_FMAC22(ADD, "ADDq"); } +microVUf(void) mVU_ADDx(mF) { mVU_FMAC3 (ADD, "ADDx"); } +microVUf(void) mVU_ADDy(mF) { mVU_FMAC3 (ADD, "ADDy"); } +microVUf(void) mVU_ADDz(mF) { mVU_FMAC3 (ADD, "ADDz"); } +microVUf(void) mVU_ADDw(mF) { mVU_FMAC3 (ADD, "ADDw"); } +microVUf(void) mVU_ADDA(mF) { mVU_FMAC4 (ADD, "ADDA"); } +microVUf(void) mVU_ADDAi(mF) { mVU_FMAC7 (ADD, "ADDAi"); } +microVUf(void) mVU_ADDAq(mF) { mVU_FMAC23(ADD, "ADDAq"); } +microVUf(void) mVU_ADDAx(mF) { mVU_FMAC5 (ADD, "ADDAx"); } +microVUf(void) mVU_ADDAy(mF) { mVU_FMAC5 (ADD, "ADDAy"); } +microVUf(void) mVU_ADDAz(mF) { mVU_FMAC5 (ADD, "ADDAz"); } +microVUf(void) mVU_ADDAw(mF) { mVU_FMAC5 (ADD, "ADDAw"); } +microVUf(void) mVU_SUB(mF) { mVU_FMAC1 (SUB, "SUB"); } +microVUf(void) mVU_SUBi(mF) { mVU_FMAC6 (SUB, "SUBi"); } +microVUf(void) mVU_SUBq(mF) { mVU_FMAC22(SUB, "SUBq"); } +microVUf(void) mVU_SUBx(mF) { mVU_FMAC3 (SUB, "SUBx"); } +microVUf(void) mVU_SUBy(mF) { mVU_FMAC3 (SUB, "SUBy"); } +microVUf(void) mVU_SUBz(mF) { mVU_FMAC3 (SUB, "SUBz"); } +microVUf(void) mVU_SUBw(mF) { mVU_FMAC3 (SUB, "SUBw"); } +microVUf(void) mVU_SUBA(mF) { mVU_FMAC4 (SUB, "SUBA"); } +microVUf(void) mVU_SUBAi(mF) { mVU_FMAC7 (SUB, "SUBAi"); } +microVUf(void) mVU_SUBAq(mF) { mVU_FMAC23(SUB, "SUBAq"); } +microVUf(void) mVU_SUBAx(mF) { mVU_FMAC5 (SUB, "SUBAx"); } +microVUf(void) mVU_SUBAy(mF) { mVU_FMAC5 (SUB, "SUBAy"); } +microVUf(void) mVU_SUBAz(mF) { mVU_FMAC5 (SUB, "SUBAz"); } +microVUf(void) mVU_SUBAw(mF) { mVU_FMAC5 (SUB, "SUBAw"); } +microVUf(void) mVU_MUL(mF) { mVU_FMAC1 (MUL, "MUL"); } +microVUf(void) mVU_MULi(mF) { mVU_FMAC6 (MUL, "MULi"); } +microVUf(void) mVU_MULq(mF) { mVU_FMAC22(MUL, "MULq"); } +microVUf(void) mVU_MULx(mF) { mVU_FMAC3 (MUL, "MULx"); } +microVUf(void) mVU_MULy(mF) { mVU_FMAC3 (MUL, "MULy"); } +microVUf(void) mVU_MULz(mF) { mVU_FMAC3 (MUL, "MULz"); } +microVUf(void) mVU_MULw(mF) { mVU_FMAC3 (MUL, "MULw"); } +microVUf(void) mVU_MULA(mF) { mVU_FMAC4 (MUL, "MULA"); } +microVUf(void) mVU_MULAi(mF) { mVU_FMAC7 (MUL, "MULAi"); } +microVUf(void) mVU_MULAq(mF) { mVU_FMAC23(MUL, "MULAq"); } +microVUf(void) mVU_MULAx(mF) { mVU_FMAC5 (MUL, "MULAx"); } +microVUf(void) mVU_MULAy(mF) { mVU_FMAC5 (MUL, "MULAy"); } +microVUf(void) mVU_MULAz(mF) { mVU_FMAC5 (MUL, "MULAz"); } +microVUf(void) mVU_MULAw(mF) { mVU_FMAC5 (MUL, "MULAw"); } +microVUf(void) mVU_MADD(mF) { mVU_FMAC8 (ADD, "MADD"); } +microVUf(void) mVU_MADDi(mF) { mVU_FMAC12(ADD, "MADDi"); } +microVUf(void) mVU_MADDq(mF) { mVU_FMAC24(ADD, "MADDq"); } +microVUf(void) mVU_MADDx(mF) { mVU_FMAC10(ADD, "MADDx"); } +microVUf(void) mVU_MADDy(mF) { mVU_FMAC10(ADD, "MADDy"); } +microVUf(void) mVU_MADDz(mF) { mVU_FMAC10(ADD, "MADDz"); } +microVUf(void) mVU_MADDw(mF) { mVU_FMAC10(ADD, "MADDw"); } +microVUf(void) mVU_MADDA(mF) { mVU_FMAC14(ADD, "MADDA"); } +microVUf(void) mVU_MADDAi(mF) { mVU_FMAC16(ADD, "MADDAi"); } +microVUf(void) mVU_MADDAq(mF) { mVU_FMAC26(ADD, "MADDAq"); } +microVUf(void) mVU_MADDAx(mF) { mVU_FMAC15(ADD, "MADDAx"); } +microVUf(void) mVU_MADDAy(mF) { mVU_FMAC15(ADD, "MADDAy"); } +microVUf(void) mVU_MADDAz(mF) { mVU_FMAC15(ADD, "MADDAz"); } +microVUf(void) mVU_MADDAw(mF) { mVU_FMAC15(ADD, "MADDAw"); } +microVUf(void) mVU_MSUB(mF) { mVU_FMAC9 (SUB, "MSUB"); } +microVUf(void) mVU_MSUBi(mF) { mVU_FMAC13(SUB, "MSUBi"); } +microVUf(void) mVU_MSUBq(mF) { mVU_FMAC25(SUB, "MSUBq"); } +microVUf(void) mVU_MSUBx(mF) { mVU_FMAC11(SUB, "MSUBx"); } +microVUf(void) mVU_MSUBy(mF) { mVU_FMAC11(SUB, "MSUBy"); } +microVUf(void) mVU_MSUBz(mF) { mVU_FMAC11(SUB, "MSUBz"); } +microVUf(void) mVU_MSUBw(mF) { mVU_FMAC11(SUB, "MSUBw"); } +microVUf(void) mVU_MSUBA(mF) { mVU_FMAC14(SUB, "MSUBA"); } +microVUf(void) mVU_MSUBAi(mF) { mVU_FMAC16(SUB, "MSUBAi"); } +microVUf(void) mVU_MSUBAq(mF) { mVU_FMAC26(SUB, "MSUBAq"); } +microVUf(void) mVU_MSUBAx(mF) { mVU_FMAC15(SUB, "MSUBAx"); } +microVUf(void) mVU_MSUBAy(mF) { mVU_FMAC15(SUB, "MSUBAy"); } +microVUf(void) mVU_MSUBAz(mF) { mVU_FMAC15(SUB, "MSUBAz"); } +microVUf(void) mVU_MSUBAw(mF) { mVU_FMAC15(SUB, "MSUBAw"); } +microVUf(void) mVU_MAX(mF) { mVU_FMAC27(MAX2, "MAX"); } +microVUf(void) mVU_MAXi(mF) { mVU_FMAC28(MAX2, "MAXi"); } +microVUf(void) mVU_MAXx(mF) { mVU_FMAC29(MAX2, "MAXx"); } +microVUf(void) mVU_MAXy(mF) { mVU_FMAC29(MAX2, "MAXy"); } +microVUf(void) mVU_MAXz(mF) { mVU_FMAC29(MAX2, "MAXz"); } +microVUf(void) mVU_MAXw(mF) { mVU_FMAC29(MAX2, "MAXw"); } +microVUf(void) mVU_MINI(mF) { mVU_FMAC27(MIN2, "MINI"); } +microVUf(void) mVU_MINIi(mF) { mVU_FMAC28(MIN2, "MINIi"); } +microVUf(void) mVU_MINIx(mF) { mVU_FMAC29(MIN2, "MINIx"); } +microVUf(void) mVU_MINIy(mF) { mVU_FMAC29(MIN2, "MINIy"); } +microVUf(void) mVU_MINIz(mF) { mVU_FMAC29(MIN2, "MINIz"); } +microVUf(void) mVU_MINIw(mF) { mVU_FMAC29(MIN2, "MINIw"); } +microVUf(void) mVU_OPMULA(mF) { mVU_FMAC18(MUL, "OPMULA"); } +microVUf(void) mVU_OPMSUB(mF) { mVU_FMAC19(SUB, "OPMSUB"); } +microVUf(void) mVU_NOP(mF) { pass3 { mVUlog("NOP"); } } +microVUq(void) mVU_FTOIx(uptr addr, int recPass) { microVU* mVU = mVUx; pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } pass2 { @@ -586,11 +586,11 @@ microVUq(void) mVU_FTOIx(uptr addr) { mVUallocFMAC2b(Ft); } } -microVUf(void) mVU_FTOI0() { mVU_FTOIx((uptr)0); pass3 { microVU* mVU = mVUx; mVUlog("FTOI0"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI4() { mVU_FTOIx((uptr)mVU_FTOI_4); pass3 { microVU* mVU = mVUx; mVUlog("FTOI4"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI12() { mVU_FTOIx((uptr)mVU_FTOI_12); pass3 { microVU* mVU = mVUx; mVUlog("FTOI12"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI15() { mVU_FTOIx((uptr)mVU_FTOI_15); pass3 { microVU* mVU = mVUx; mVUlog("FTOI15"); mVUlogFtFs(); } } -microVUq(void) mVU_ITOFx(uptr addr) { +microVUf(void) mVU_FTOI0(mF) { mVU_FTOIx((uptr)0, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI0"); mVUlogFtFs(); } } +microVUf(void) mVU_FTOI4(mF) { mVU_FTOIx((uptr)mVU_FTOI_4, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI4"); mVUlogFtFs(); } } +microVUf(void) mVU_FTOI12(mF) { mVU_FTOIx((uptr)mVU_FTOI_12, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI12"); mVUlogFtFs(); } } +microVUf(void) mVU_FTOI15(mF) { mVU_FTOIx((uptr)mVU_FTOI_15, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI15"); mVUlogFtFs(); } } +microVUq(void) mVU_ITOFx(uptr addr, int recPass) { microVU* mVU = mVUx; pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } pass2 { @@ -604,11 +604,11 @@ microVUq(void) mVU_ITOFx(uptr addr) { mVUallocFMAC2b(Ft); } } -microVUf(void) mVU_ITOF0() { mVU_ITOFx((uptr)0); pass3 { microVU* mVU = mVUx; mVUlog("ITOF0"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF4() { mVU_ITOFx((uptr)mVU_ITOF_4); pass3 { microVU* mVU = mVUx; mVUlog("ITOF4"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF12() { mVU_ITOFx((uptr)mVU_ITOF_12); pass3 { microVU* mVU = mVUx; mVUlog("ITOF12"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF15() { mVU_ITOFx((uptr)mVU_ITOF_15); pass3 { microVU* mVU = mVUx; mVUlog("ITOF15"); mVUlogFtFs(); } } -microVUf(void) mVU_CLIP() { +microVUf(void) mVU_ITOF0(mF) { mVU_ITOFx((uptr)0, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF0"); mVUlogFtFs(); } } +microVUf(void) mVU_ITOF4(mF) { mVU_ITOFx((uptr)mVU_ITOF_4, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF4"); mVUlogFtFs(); } } +microVUf(void) mVU_ITOF12(mF) { mVU_ITOFx((uptr)mVU_ITOF_12, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF12"); mVUlogFtFs(); } } +microVUf(void) mVU_ITOF15(mF) { mVU_ITOFx((uptr)mVU_ITOF_15, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF15"); mVUlogFtFs(); } } +microVUf(void) mVU_CLIP(mF) { microVU* mVU = mVUx; pass1 { mVUanalyzeFMAC4(_Fs_, _Ft_); } pass2 { From e384d001da6c8383242a600e2b673304de7d25c3 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 29 May 2009 08:39:32 +0000 Subject: [PATCH 151/296] Zeropad: Windows is back to compiling, though the config dialog is still broken. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1288 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Windows/win.cpp | 27 +++++++++++++-------------- plugins/zeropad/Windows/win.h | 3 ++- plugins/zeropad/analog.cpp | 8 ++++---- plugins/zeropad/joystick.h | 11 ++++++++--- plugins/zeropad/keyboard.cpp | 19 +++++++++++-------- plugins/zeropad/keyboard.h | 4 ++-- plugins/zeropad/zeropad.cpp | 3 ++- plugins/zeropad/zeropad.h | 4 ++++ 8 files changed, 46 insertions(+), 33 deletions(-) diff --git a/plugins/zeropad/Windows/win.cpp b/plugins/zeropad/Windows/win.cpp index 6bf8e79275..b396b11270 100644 --- a/plugins/zeropad/Windows/win.cpp +++ b/plugins/zeropad/Windows/win.cpp @@ -20,13 +20,12 @@ using namespace std; -HINSTANCE hInst = NULL; extern u16 status[2]; extern string s_strIniPath; -LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); WNDPROC GSwndProc = NULL; HWND GShwnd = NULL; +HINSTANCE hInst = NULL; extern keyEvent event; @@ -94,7 +93,7 @@ string GetKeyLabel(const int pad, const int index) } else if (key >= 0x3000 && key < 0x4000) // IS_POV; 0x30000 - 0x50000 { - static const char name[][7] = { "FORWARD", "RIGHT", "BACK", "LEFT" }; + static const char name[][8] = { "FORWARD", "RIGHT", "BACK", "LEFT" }; const int pov = (key & 0xff); sprintf(buff, "J%d_POV%d_%s", (key & 0xfff) / 0x100, pov / 4, name[pov % 4]); @@ -107,8 +106,9 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) { HWND hWC; TCITEM tcI; - int i, key, numkeys; - u8* pkeyboard; + int i, key; + int numkeys; + //u8* pkeyboard; static int disabled = 0; static int padn = 0; @@ -141,18 +141,19 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) case WM_TIMER: if (disabled) { - key = 0; //pkeyboard = SDL_GetKeyState(&numkeys); - // Yeah, there's no way this will work, given that it's disabled. + // Well, this doesn't exactly work either, but it's close... + numkeys = 256; for (int i = 0; i < numkeys; ++i) { - if (pkeyboard[i]) + if (GetAsyncKeyState(i)) { key = i; break; } } + if (key == 0) { // check joystick @@ -183,8 +184,7 @@ BOOL CALLBACK ConfigureDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) EnableWindow(GetDlgItem(hW, disabled = wParam), FALSE); - // Wait a silly amount of time for a keypress. - SetTimer(hW, 0x80, 250, NULL); + SetTimer(hW, 0x80, 2500, NULL); return TRUE; } @@ -250,10 +250,9 @@ BOOL CALLBACK AboutDlgProc(HWND hW, UINT uMsg, WPARAM wParam, LPARAM lParam) void CALLBACK PADconfigure() { - DialogBox(hInst, - MAKEINTRESOURCE(IDD_DIALOG1), - GetActiveWindow(), - (DLGPROC)ConfigureDlgProc); + INT_PTR ret; + + ret = DialogBoxParam(hInst,MAKEINTRESOURCE(IDD_DIALOG1),GetActiveWindow(),(DLGPROC)ConfigureDlgProc,1); } void CALLBACK PADabout() diff --git a/plugins/zeropad/Windows/win.h b/plugins/zeropad/Windows/win.h index ed468e2c64..0401f20031 100644 --- a/plugins/zeropad/Windows/win.h +++ b/plugins/zeropad/Windows/win.h @@ -30,4 +30,5 @@ extern void SaveConfig(); extern void LoadConfig(); extern void SysMessage(char *fmt, ...); -extern LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); \ No newline at end of file +extern LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam); +extern HINSTANCE hInst; \ No newline at end of file diff --git a/plugins/zeropad/analog.cpp b/plugins/zeropad/analog.cpp index 009f45c8db..14b8489b22 100644 --- a/plugins/zeropad/analog.cpp +++ b/plugins/zeropad/analog.cpp @@ -98,19 +98,19 @@ namespace Analog switch (padvalue) { case PAD_LX: - return (conf.options & PADOPTION_REVERTLX); + return ((conf.options & PADOPTION_REVERTLX) != 0); break; case PAD_RX: - return (conf.options & PADOPTION_REVERTRX); + return ((conf.options & PADOPTION_REVERTRX) != 0); break; case PAD_LY: - return (conf.options & PADOPTION_REVERTLY); + return ((conf.options & PADOPTION_REVERTLY) != 0); break; case PAD_RY: - return (conf.options & PADOPTION_REVERTRY); + return ((conf.options & PADOPTION_REVERTRY) != 0); break; default: diff --git a/plugins/zeropad/joystick.h b/plugins/zeropad/joystick.h index 1cecd8e23c..30d278c496 100644 --- a/plugins/zeropad/joystick.h +++ b/plugins/zeropad/joystick.h @@ -16,10 +16,15 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - #ifndef __JOYSTICK_H__ - #define __JOYSTICK_H__ - +#ifndef __JOYSTICK_H__ +#define __JOYSTICK_H__ + +#ifdef __LINUX__ #include +#else +#include +#endif + #include "zeropad.h" // holds all joystick info diff --git a/plugins/zeropad/keyboard.cpp b/plugins/zeropad/keyboard.cpp index bdcf1a6986..855bf89a1b 100644 --- a/plugins/zeropad/keyboard.cpp +++ b/plugins/zeropad/keyboard.cpp @@ -36,7 +36,10 @@ char* KeysymToChar(int keysym) #ifdef __LINUX__ return XKeysymToString(keysym); #else - return; + LPWORD temp; + + ToAscii((UINT) keysym, NULL, NULL, temp, NULL); + return (char*)temp; #endif } @@ -172,7 +175,7 @@ bool PollX11Keyboard(char* &temp, u32 &pkey) #else LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) { - int i, pad, keyPress[2] = {0}, keyRelease[2] = {0}; + int keyPress[2] = {0}, keyRelease[2] = {0}; static bool lbutton = false, rbutton = false; switch (msg) @@ -180,9 +183,9 @@ LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) case WM_KEYDOWN: if (lParam & 0x40000000) return TRUE; - for (pad = 0; pad < 2; ++pad) + for (int pad = 0; pad < 2; ++pad) { - for (i = 0; i < PADKEYS; i++) + for (int i = 0; i < PADKEYS; i++) { if (wParam == conf.keys[pad][i]) { @@ -198,9 +201,9 @@ LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) break; case WM_KEYUP: - for (pad = 0; pad < 2; ++pad) + for (int pad = 0; pad < 2; ++pad) { - for (i = 0; i < PADKEYS; i++) + for (int i = 0; i < PADKEYS; i++) { if (wParam == conf.keys[pad][i]) { @@ -267,9 +270,9 @@ LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) return GSwndProc(hWnd, msg, wParam, lParam); } - for (pad = 0; pad < 2; ++pad) + for (int pad = 0; pad < 2; ++pad) { - UpdateKeys(pad, keyPress, keyRelease) + UpdateKeys(pad, keyPress[pad], keyRelease[pad]); } return TRUE; diff --git a/plugins/zeropad/keyboard.h b/plugins/zeropad/keyboard.h index 471b8eddcd..1a61fe86c8 100644 --- a/plugins/zeropad/keyboard.h +++ b/plugins/zeropad/keyboard.h @@ -31,8 +31,8 @@ extern bool PollX11Keyboard(char* &temp, u32 &pkey); #else -extern WNDPROC GSwndProc = NULL; -extern HWND GShwnd = NULL; +extern WNDPROC GSwndProc; +extern HWND GShwnd; #endif diff --git a/plugins/zeropad/zeropad.cpp b/plugins/zeropad/zeropad.cpp index 493778a0fa..487fc18b64 100644 --- a/plugins/zeropad/zeropad.cpp +++ b/plugins/zeropad/zeropad.cpp @@ -23,7 +23,6 @@ #include #include "zeropad.h" -#include "joystick.h" #ifndef _WIN32 @@ -270,7 +269,9 @@ s32 CALLBACK PADopen(void *pDsp) s_keyPress[0] = s_keyPress[1] = 0; s_keyRelease[0] = s_keyRelease[1] = 0; +#ifdef __LINUX__ JoystickInfo::EnumerateJoysticks(s_vjoysticks); +#endif return _PADopen(pDsp); } diff --git a/plugins/zeropad/zeropad.h b/plugins/zeropad/zeropad.h index f22316ccc4..e74142fc4a 100644 --- a/plugins/zeropad/zeropad.h +++ b/plugins/zeropad/zeropad.h @@ -37,6 +37,7 @@ #include #include #include +#include using namespace std; #define PADdefs @@ -44,6 +45,9 @@ extern "C" { #include "PS2Edefs.h" } +#ifdef __LINUX__ +#include "joystick.h" +#endif #include "analog.h" #include "bitwise.h" From aa64a21e93045a5912d1604cd7c724b4dc9b68d5 Mon Sep 17 00:00:00 2001 From: refraction Date: Fri, 29 May 2009 18:44:42 +0000 Subject: [PATCH 152/296] Couple of fixes, one for Kingdom Hearts 2 Issue 239, others might help textures in the GT games (maybe!) Also prepared some code to add support for VU looping in MTGS mode, this will require a GS Spec update, so its commented out for now until i can sync an update with Gabest. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1289 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/GS.h | 4 +-- pcsx2/Gif.cpp | 26 +++++++++++-------- pcsx2/Hw.h | 17 ++++++++++++ pcsx2/MTGS.cpp | 47 +++++++++++++++++++++++++++++---- pcsx2/VUops.cpp | 28 +++++++++++++++++++- pcsx2/VifDma.cpp | 52 +++++++++++++++++++++++++------------ pcsx2/x86/iVUmicroLower.cpp | 17 ++++++++++-- pcsx2/x86/microVU_Lower.inl | 17 +++++++++++- 8 files changed, 170 insertions(+), 38 deletions(-) diff --git a/pcsx2/GS.h b/pcsx2/GS.h index 9f6f8f70f2..f70509cb8c 100644 --- a/pcsx2/GS.h +++ b/pcsx2/GS.h @@ -111,7 +111,7 @@ struct GIFPath return false; } } - + return true; } }; @@ -265,7 +265,7 @@ protected: // Processes a GIFtag & packet, and throws out some gsIRQs as needed. // Used to keep interrupts in sync with the EE, while the GS itself // runs potentially several frames behind. - u32 _gifTransferDummy( GIF_PATH pathidx, const u8 *pMem, u32 size ); + int _gifTransferDummy( GIF_PATH pathidx, const u8 *pMem, u32 size ); // Used internally by SendSimplePacket type functions uint _PrepForSimplePacket(); diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 892441607d..a30a2ad414 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -58,8 +58,6 @@ __forceinline void gsInterrupt() { return; } - if(Path3progress == 2) vif1Regs->stat &= ~VIF1_STAT_VGW; - if (gif->qwc > 0 || gspath3done == 0) { if (!(psHu32(DMAC_CTRL) & 0x1)) { Console::Notice("gs dma masked, re-scheduling..."); @@ -72,12 +70,14 @@ __forceinline void gsInterrupt() { return; } + vif1Regs->stat &= ~VIF1_STAT_VGW; gspath3done = 0; gscycles = 0; gif->chcr &= ~0x100; GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x4000; //FIFO empty - psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 + psHu32(GIF_STAT) &= ~GIF_STAT_P3Q; + //psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 psHu32(GIF_STAT)&= ~0x1F000000; // QFC=0 hwDmacIrq(DMAC_GIF); GIF_LOG("GIF DMA end"); @@ -86,7 +86,7 @@ __forceinline void gsInterrupt() { static u32 WRITERING_DMA(u32 *pMem, u32 qwc) { - psHu32(GIF_STAT) |= 0xE00; + psHu32(GIF_STAT) |= GIF_STAT_APATH3 | GIF_STAT_OPH; if( mtgsThread != NULL ) @@ -107,6 +107,7 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) memcpy_aligned(pgsmem, pMem, sizetoread<<4); mtgsThread->SendDataPacket(); + if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 return sizetoread; } else @@ -243,6 +244,7 @@ void GIFdma() if(Path3progress == 2/* && gif->qwc != 0*/) { + vif1Regs->stat &= ~VIF1_STAT_VGW; dmaGIFend(); return; } @@ -347,8 +349,9 @@ void dmaGIF() { return; } - gspath3done = 0; // For some reason this doesnt clear? So when the system starts the thread, we will clear it :) + gspath3done = 0; // For some reason this doesnt clear? So when the system starts the thread, we will clear it :) + psHu32(GIF_STAT) |= GIF_STAT_P3Q; GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x8000; //FIFO full psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [used to be 0xE00; // OPH=1 | APATH=3] @@ -376,7 +379,7 @@ void dmaGIF() { } //Halflife sets a QWC amount in chain mode, no tadr set. - if((gif->qwc > 0) && ((gif->chcr & 0x4) == 0x4)) gspath3done = 1; + if(gif->qwc > 0) gspath3done = 1; GIFdma(); } @@ -391,7 +394,7 @@ static __forceinline int mfifoGIFrbTransfer() { if ((gif->madr+mfifoqwc*16) > (psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR)+16)) { int s1 = ((psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR)+16) - gif->madr) >> 4; - + int s2 = (mfifoqwc - s1); // fixme - I don't think these should use WRITERING_DMA, since our source // isn't the DmaGetAddr(gif->madr) address that WRITERING_DMA expects. @@ -400,13 +403,14 @@ static __forceinline int mfifoGIFrbTransfer() { if (src == NULL) return -1; s1 = WRITERING_DMA(src, s1); - if(s1 == (((psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR)+16) - gif->madr) >> 4)) + if(s1 == (mfifoqwc - s2)) { /* and second copy 's2' bytes from 'maddr' to '&data[s1]' */ src = (u32*)PSM(psHu32(DMAC_RBOR)); if (src == NULL) return -1; - mfifoqwc = WRITERING_DMA(src, (mfifoqwc - s1)) + s1; - } + s2 = WRITERING_DMA(src, s2); + } else s2 = 0; + mfifoqwc = s1 + s2; } else @@ -579,10 +583,10 @@ void gifMFIFOInterrupt() if (!gifmfifoirq) gifqwc = 0; gifstate = GIF_STATE_READY; gif->chcr &= ~0x100; + vif1Regs->stat &= ~VIF1_STAT_VGW; hwDmacIrq(DMAC_GIF); GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x4000; //FIFO empty - psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 psHu32(GIF_STAT)&= ~0x1F000000; // QFC=0 } diff --git a/pcsx2/Hw.h b/pcsx2/Hw.h index 9a915ce5ba..480a22ef37 100644 --- a/pcsx2/Hw.h +++ b/pcsx2/Hw.h @@ -288,6 +288,23 @@ enum DMACIrqs #define VIF_STAT_ER1 (1<<13) #define VIF_STAT_FDR (1<<23) +//GIF_STAT + +#define GIF_STAT_M3R (1) //GIF_MODE Mask +#define GIF_STAT_M3P (1<<1) //VIF PATH3 Mask +#define GIF_STAT_IMT (1<<2) //Intermittent Transfer Mode +#define GIF_STAT_PSE (1<<3) //Temporary Transfer Stop +#define GIF_STAT_IP3 (1<<5) //Interrupted PATH3 +#define GIF_STAT_P3Q (1<<6) //PATH3 request Queued +#define GIF_STAT_P2Q (1<<7) //PATH2 request Queued +#define GIF_STAT_P1Q (1<<8) //PATH1 request Queued +#define GIF_STAT_OPH (1<<9) //Output Path (Outputting Data) +#define GIF_STAT_APATH1 (1<<10) //Data Transfer Path 1 (In progress) +#define GIF_STAT_APATH2 (2<<10) //Data Transfer Path 2 (In progress) +#define GIF_STAT_APATH3 (3<<10) //Data Transfer Path 3 (In progress) (Mask too) +#define GIF_STAT_DIR (1<<12) //Transfer Direction +#define GIF_STAT_FQC (31<<24) //QWC in GIF-FIFO + //DMA interrupts & masks enum DMAInter { diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 25560e85b0..659e517967 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -259,10 +259,11 @@ void mtgsThreadObject::Reset() // Used to keep interrupts in sync with the EE, while the GS itself // runs potentially several frames behind. // size - size of the packet in simd128's -__forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u8* pMem, u32 size ) +__forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u8* pMem, u32 size ) { GIFPath& path = m_path[pathidx]; - + /* bool path1loop = false; + int startval = size;*/ #ifdef PCSX2_GSRING_SAMPLING_STATS static uptr profStartPtr = 0; static uptr profEndPtr = 0; @@ -295,10 +296,25 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u } if( pathidx == 0 ) - { + { + // int transize = 0; // hack: if too much data for VU1, just ignore. // The GIF is evil : if nreg is 0, it's really 16. Otherwise it's the value in nreg. + /*const int numregs = path.tag.nreg ? path.tag.nreg : 16; + if(path.tag.flg < 2) + { + transize = (path.tag.nloop * numregs); + } + else transize = path.tag.nloop; + + if(transize > (path.tag.flg == 1 ? 0x800 : 0x400)) + { + //DevCon::Notice("Too much data"); + path.tag.nloop = 0; + if(path1loop == true)return ++size - 0x400; + else return ++size; + }*/ const int numregs = ((path.tag.nreg-1)&15)+1; if((path.tag.nloop * numregs) > (size * ((path.tag.flg == 1) ? 2 : 1))) @@ -374,7 +390,22 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u } } - if(pathidx == 0 || pathidx == 2) + if(pathidx == 0) + { + if(path.tag.eop && path.tag.nloop == 0) + { + break; + } + /*if((path.tag.nloop > 0 || (!path.tag.eop && path.tag.nloop == 0)) && size == 0) + { + if(path1loop == true) return size - 0x400; + //DevCon::Notice("Looping Nloop %x, Eop %x, FLG %x", params path.tag.nloop, path.tag.eop, path.tag.flg); + size = 0x400; + pMem -= 0x4000; + path1loop = true; + }*/ + } + if(pathidx == 2) { if(path.tag.eop && path.tag.nloop == 0) { @@ -386,6 +417,11 @@ __forceinline u32 mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u if(pathidx == 0) { + //If the XGKick has spun around the VU memory end address, we need to INCREASE the size sent. + /*if(path1loop == true) + { + return (size - 0x400); //This will cause a negative making eg. size(20) - retval(-30) = 50; + }*/ if(size == 0 && path.tag.nloop > 0) { path.tag.nloop = 0; @@ -533,8 +569,8 @@ int mtgsThreadObject::Callback() const u128* data = m_RingBuffer.GetPtr( m_RingPos+1 ); // make sure that tag>>16 is the MAX size readable - //GSgifTransfer1(((u32*)data) - 0x1000 + 4*qsize, 0x4000-qsize*16); GSgifTransfer1((u32*)(data - 0x400 + qsize), 0x4000-qsize*16); + //GSgifTransfer1((u32*)data, qsize); ringposinc += qsize; } break; @@ -863,6 +899,7 @@ int mtgsThreadObject::PrepDataPacket( GIF_PATH pathidx, const u8* srcdata, u32 s gif->madr += (size - retval) * 16; gif->qwc -= size - retval; } + //if(retval < 0) DevCon::Notice("Increasing size from %x to %x path %x", params size, size-retval, pathidx+1); size = size - retval; m_packet_size = size; size++; // takes into account our command qword. diff --git a/pcsx2/VUops.cpp b/pcsx2/VUops.cpp index 76a25daae6..dfcd5e5f1a 100644 --- a/pcsx2/VUops.cpp +++ b/pcsx2/VUops.cpp @@ -2061,7 +2061,33 @@ void _vuXGKICK(VURegs * VU) ptr = (u32*)GET_VU_MEM(VU, 0); memcpy(&tempmem[temp], ptr, ((VU->VI[_Is_].US[0]*16) & 0x3fff)); GSGIFTRANSFER1((u32*)&tempmem[0], 0); - } else*/ GSGIFTRANSFER1((u32*)VU->Mem, (VU->VI[_Is_].US[0]*16) & 0x3fff); + } else*/ + //DevCon::Notice("Addr %x", params VU->VI[_Is_].US[0] & 0x3fff); + if( mtgsThread != NULL ) + { + u32* data = (u32*)((u8*)VU->Mem + ((VU->VI[_Is_].US[0]*16) & 0x3fff)); + u32 size; + size = mtgsThread->PrepDataPacket( GIF_PATH_1, data, (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)) >> 4); + { + u8* pmem = mtgsThread->GetDataPacketPtr(); + + if((size << 4) > (u32)(0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff))) + { + //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params ((VU->VI[_Is_].US[0]*16) & 0x3fff) + (size << 4), (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)) >> 4, size - (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff) >> 4)); + memcpy_aligned(pmem, (u8*)VU->Mem+((VU->VI[_Is_].US[0]*16) & 0x3fff), 0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)); + size -= (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)) >> 4; + //DevCon::Notice("Size left %x", params size); + pmem += 0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff); + memcpy_aligned(pmem, (u8*)VU->Mem, size<<4); + } + else + { + memcpy_aligned(pmem, (u8*)VU->Mem+((VU->VI[_Is_].US[0]*16) & 0x3fff), size<<4); + } + mtgsThread->SendDataPacket(); + } + + }else GSGIFTRANSFER1((u32*)VU->Mem, (VU->VI[_Is_].US[0]*16) & 0x3fff); } void _vuXTOP(VURegs * VU) { diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index d1604902ae..8b249cfdf6 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -623,8 +623,8 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i if(tempsize > (u32)(VIFdmanum ? 0x4000 : 0x1000)) { - //DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); - if(vifRegs->cycle.cl == 1 && ((u32)(VIFdmanum ? 0x4000 : 0x1000) + ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) == tempsize + + if(((vifRegs->cycle.cl != vifRegs->cycle.wl) && ((u32)(VIFdmanum ? 0x4000 : 0x1000) + ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) == tempsize) || tempsize == (u32)(VIFdmanum ? 0x4000 : 0x1000)) { //Its a red herring! so ignore it! SSE unpacks will be much quicker @@ -632,6 +632,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i } else { + //DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); tempsize = size; size = 0; } @@ -1678,7 +1679,7 @@ static __forceinline void vif1UNPACK(u32 *data) return; } } - vif1FLUSH(); + //vif1FLUSH(); vl = (vif1.cmd) & 0x3; vn = (vif1.cmd >> 2) & 0x3; @@ -1890,7 +1891,7 @@ static int __fastcall Vif1TransDirectHL(u32 *data) } else { - psHu32(GIF_STAT) &= ~0x80; + psHu32(GIF_STAT) &= ~GIF_STAT_APATH2; ret = vif1.tag.size; vif1.tag.size = 0; vif1.cmd = 0; @@ -2057,7 +2058,7 @@ static void Vif1CMDFlush() // FLUSH/E/A if((vif1.cmd & 0x7f) == 0x13) { - if(Path3progress != 2 && gif->chcr & 0x100) // Gif is already transferring so wait for it. + if((Path3progress != 2 || !vif1Regs->mskpath3) && gif->chcr & 0x100) // Gif is already transferring so wait for it. { vif1Regs->stat |= VIF1_STAT_VGW; } @@ -2089,7 +2090,7 @@ static void Vif1CMDSTRowCol() // STROW / STCOL static void Vif1CMDMPGTransfer() // MPG { int vifNum; - vif1FLUSH(); + //vif1FLUSH(); vifNum = (u8)(vif1Regs->code >> 16); if (vifNum == 0) vifNum = 256; @@ -2113,15 +2114,15 @@ static void Vif1CMDDirectHL() // DIRECT/HL //FIXME: This should have timing in both cases, see note below. if((vif1.cmd & 0x7f) == 0x51) { - if(gif->chcr & 0x100 /*&& Path3progress == 0*/) //PATH3 is in image mode, so wait for end of transfer + if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 || Path3progress != 2)) //PATH3 is in image mode, so wait for end of transfer { //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); - if(vif1Regs->mskpath3)vif1Regs->stat |= VIF1_STAT_VGW; - else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( + /*if(vif1Regs->mskpath3)*/vif1Regs->stat |= VIF1_STAT_VGW; + //else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( } } - psHu32(GIF_STAT) |= 0x80; + psHu32(GIF_STAT) |= GIF_STAT_APATH2; } static void Vif1CMDNull() // invalid opcode @@ -2193,6 +2194,17 @@ int VIF1transfer(u32 *data, int size, int istag) while (vif1.vifpacketsize > 0) { + if((vif1.cmd & 0x7f) == 0x51) + { + if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 || Path3progress != 2)) //PATH3 is in image mode, so wait for end of transfer + { + //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); + /*if(vif1Regs->mskpath3)*/vif1Regs->stat |= VIF1_STAT_VGW; + //else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( + } + + } + if(vif1Regs->stat & VIF1_STAT_VGW) break; if (vif1.cmd) { vif1Regs->stat |= VIF1_STAT_VPS_T; //Decompression has started @@ -2350,7 +2362,7 @@ void vif1TransferFromMemory() } FreezeXMMRegs(0); - if (vif1Regs->mskpath3 == 0)vif1Regs->stat &= ~0x1f000000; + g_vifCycles += vif1ch->qwc * 2; vif1ch->madr += vif1ch->qwc * 16; // mgs3 scene changes vif1ch->qwc = 0; @@ -2401,6 +2413,7 @@ __forceinline void vif1SetupTransfer() case 1: //Normal (From memory) vif1.inprogress = 1; vif1.done = true; + g_vifCycles = 2; break; case 2: //Chain @@ -2513,7 +2526,12 @@ __forceinline void vif1Interrupt() } } - if (vif1.inprogress) _VIF1chain(); + if (vif1.inprogress) + { + _VIF1chain(); + CPU_INT(1, g_vifCycles); + return; + } if ((!vif1.done) || (vif1.inprogress & 0x1)) { @@ -2543,7 +2561,7 @@ __forceinline void vif1Interrupt() vif1ch->chcr &= ~0x100; g_vifCycles = 0; hwDmacIrq(DMAC_VIF1); - vif1Regs->stat &= ~0x1F000000; // FQC=0 + if(vif1ch->chcr & 0x1)vif1Regs->stat &= ~0x1F000000; // FQC=0 } void dmaVIF1() @@ -2557,8 +2575,7 @@ void dmaVIF1() g_vifCycles = 0; vif1.inprogress = 0; - vif1Regs->stat |= 0x10000000; // FQC=16 - + if (((psHu32(DMAC_CTRL) & 0xC) == 0x8)) // VIF MFIFO { //Console::WriteLn("VIFMFIFO\n"); @@ -2591,6 +2608,9 @@ void dmaVIF1() vif1.dmamode = 2; } + if(vif1.dmamode != 1)vif1Regs->stat |= 0x10000000; // FQC=16 + else vif1Regs->stat |= min((u16)16, vif1ch->qwc) << 24; // FQC=16 + // Chain Mode vif1.done = false; vif1Interrupt(); @@ -2705,7 +2725,7 @@ void vif1Write32(u32 mem, u32 value) vif1Regs->stat = (vif1Regs->stat & ~VIF1_STAT_FDR) | (value & VIF1_STAT_FDR); if (vif1Regs->stat & VIF1_STAT_FDR) { - vif1Regs->stat |= 0x01000000; + vif1Regs->stat |= 0x01000000; // FQC=1 - hack but it checks this is true before tranfer? (fatal frame) } else { diff --git a/pcsx2/x86/iVUmicroLower.cpp b/pcsx2/x86/iVUmicroLower.cpp index 14d3479b42..fcae6ca866 100644 --- a/pcsx2/x86/iVUmicroLower.cpp +++ b/pcsx2/x86/iVUmicroLower.cpp @@ -1977,7 +1977,7 @@ void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr) { u32 size; u32* data = (u32*)((u8*)pMem + (addr&0x3fff)); - + // fixme: The gifTagDummy function in the MTGS (called by PrepDataPacket) has a // hack that aborts the packet if it goes past the end of VU1 memory. // Chances are this should be a "loops around memory" situation, and the packet @@ -1989,7 +1989,20 @@ void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr) //if( size > 0 ) { u8* pmem = mtgsThread->GetDataPacketPtr(); - memcpy_aligned(pmem, (u8*)pMem+addr, size<<4); + + /* if((size << 4) > (0x4000-(addr&0x3fff))) + { + //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params (addr&0x3fff) + (size << 4), (0x4000-(addr&0x3fff)) >> 4, size - ((0x4000-(addr&0x3fff)) >> 4)); + memcpy_aligned(pmem, (u8*)pMem+addr, 0x4000-(addr&0x3fff)); + size -= (0x4000-(addr&0x3fff)) >> 4; + //DevCon::Notice("Size left %x", params size); + pmem += 0x4000-(addr&0x3fff); + memcpy_aligned(pmem, (u8*)pMem, size<<4); + } + else + {*/ + memcpy_aligned(pmem, (u8*)pMem+addr, size<<4); + // } mtgsThread->SendDataPacket(); } } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 38598be5b9..85bc94dce3 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1103,7 +1103,22 @@ void __fastcall mVU_XGKICK_(u32 addr) { u32 *data = (u32*)(microVU1.regs->Mem + addr); u32 size = mtgsThread->PrepDataPacket(GIF_PATH_1, data, (0x4000-addr) >> 4); u8 *pDest = mtgsThread->GetDataPacketPtr(); - memcpy_aligned(pDest, microVU1.regs->Mem + addr, size<<4); +/* + if((size << 4) > (0x4000-(addr&0x3fff))) + { + //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params (addr&0x3fff) + (size << 4), (0x4000-(addr&0x3fff)) >> 4, size - ((0x4000-(addr&0x3fff)) >> 4)); + memcpy_aligned(pDest, microVU1.regs->Mem + addr, 0x4000-(addr&0x3fff)); + size -= (0x4000-(addr&0x3fff)) >> 4; + //DevCon::Notice("Size left %x", params size); + pDest += 0x4000-(addr&0x3fff); + memcpy_aligned(pDest, microVU1.regs->Mem, size<<4); + } + else + { +*/ + memcpy_aligned(pDest, microVU1.regs->Mem + addr, size<<4); +// } + mtgsThread->SendDataPacket(); } From 1600dc1764a6d45e610440a7bcd6844aa3673332 Mon Sep 17 00:00:00 2001 From: refraction Date: Fri, 29 May 2009 18:58:06 +0000 Subject: [PATCH 153/296] From 1289 you can use MTGS mode in VUint, but the new code was enabled :P now disabled it so MTGS will work git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1290 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VUops.cpp | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/pcsx2/VUops.cpp b/pcsx2/VUops.cpp index dfcd5e5f1a..b257ae0ec2 100644 --- a/pcsx2/VUops.cpp +++ b/pcsx2/VUops.cpp @@ -2071,7 +2071,7 @@ void _vuXGKICK(VURegs * VU) { u8* pmem = mtgsThread->GetDataPacketPtr(); - if((size << 4) > (u32)(0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff))) + /* if((size << 4) > (u32)(0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff))) { //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params ((VU->VI[_Is_].US[0]*16) & 0x3fff) + (size << 4), (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)) >> 4, size - (0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff) >> 4)); memcpy_aligned(pmem, (u8*)VU->Mem+((VU->VI[_Is_].US[0]*16) & 0x3fff), 0x4000-((VU->VI[_Is_].US[0]*16) & 0x3fff)); @@ -2081,9 +2081,9 @@ void _vuXGKICK(VURegs * VU) memcpy_aligned(pmem, (u8*)VU->Mem, size<<4); } else - { + {*/ memcpy_aligned(pmem, (u8*)VU->Mem+((VU->VI[_Is_].US[0]*16) & 0x3fff), size<<4); - } + //} mtgsThread->SendDataPacket(); } From 51ab9a6888dfd2e61ffe7308ecf582ea0f576993 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 30 May 2009 04:16:45 +0000 Subject: [PATCH 154/296] microVU: - It took a while, but I managed to convert 100's of template functions to normal functions, reducing release link times by about ~50% again. pcsx2 takes about ~2 mins to link on release build, which is a lot more tolerable than ~10 minutes from a few days ago... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1291 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 26 +- pcsx2/x86/microVU.h | 18 +- pcsx2/x86/microVU_Alloc.inl | 318 ++++++++---------- pcsx2/x86/microVU_Analyze.inl | 69 ++-- pcsx2/x86/microVU_Compile.inl | 90 +++-- pcsx2/x86/microVU_Execute.inl | 18 +- pcsx2/x86/microVU_Flags.inl | 42 +-- pcsx2/x86/microVU_Log.inl | 7 +- pcsx2/x86/microVU_Lower.inl | 597 +++++++++++++++------------------- pcsx2/x86/microVU_Misc.h | 33 +- pcsx2/x86/microVU_Misc.inl | 37 +-- pcsx2/x86/microVU_Tables.inl | 332 ++++++++++--------- pcsx2/x86/microVU_Upper.inl | 435 ++++++++++++------------- 13 files changed, 924 insertions(+), 1098 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index a58200a911..4adcf883a9 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -35,7 +35,7 @@ declareAllVariables // Declares All Global Variables :D //------------------------------------------------------------------ // Only run this once per VU! ;) -microVUt(void) mVUinit(VURegs* vuRegsPtr) { +microVUf(void) mVUinit(VURegs* vuRegsPtr) { microVU* mVU = mVUx; mVU->regs = vuRegsPtr; @@ -71,8 +71,8 @@ microVUx(void) mVUreset() { // Setup Entrance/Exit Points x86SetPtr(mVU->cache); - mVUdispatcherA(); - mVUdispatcherB(); + mVUdispatcherA(mVU); + mVUdispatcherB(mVU); // Clear All Program Data memset(&mVU->prog, 0, sizeof(mVU->prog)); @@ -98,7 +98,7 @@ microVUx(void) mVUreset() { } // Free Allocated Resources -microVUt(void) mVUclose() { +microVUf(void) mVUclose() { microVU* mVU = mVUx; mVUprint((vuIndex) ? "microVU1: close" : "microVU0: close"); @@ -116,7 +116,7 @@ microVUt(void) mVUclose() { } // Clears Block Data in specified range -microVUt(void) mVUclear(u32 addr, u32 size) { +microVUf(void) mVUclear(u32 addr, u32 size) { microVU* mVU = mVUx; if (!mVU->prog.cleared) { memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); @@ -129,7 +129,7 @@ microVUt(void) mVUclear(u32 addr, u32 size) { //------------------------------------------------------------------ // Clears program data (Sets used to 1 because calling this function implies the program will be used at least once) -microVUt(void) mVUclearProg(int progIndex) { +microVUf(void) mVUclearProg(int progIndex) { microVU* mVU = mVUx; mVU->prog.prog[progIndex].used = 1; mVU->prog.prog[progIndex].last_used = 3; @@ -144,22 +144,22 @@ microVUt(void) mVUclearProg(int progIndex) { } // Caches Micro Program -microVUt(void) mVUcacheProg(int progIndex) { +microVUf(void) mVUcacheProg(int progIndex) { microVU* mVU = mVUx; memcpy_fast(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize); mVUdumpProg(progIndex); - mVUcheckSflag(progIndex); + mVUcheckSflag(mVU, progIndex); } // Finds the least used program, (if program list full clears and returns an old program; if not-full, returns free program) -microVUt(int) mVUfindLeastUsedProg() { +microVUf(int) mVUfindLeastUsedProg() { microVU* mVU = mVUx; if (mVU->prog.total < mVU->prog.max) { mVU->prog.total++; mVUcacheProg(mVU->prog.total); // Cache Micro Program mVU->prog.prog[mVU->prog.total].used = 1; mVU->prog.prog[mVU->prog.total].last_used = 3; - DevCon::Notice("microVU%d: Cached MicroPrograms = %d", params vuIndex, mVU->prog.total+1); + Console::Notice("microVU%d: Cached MicroPrograms = %d", params vuIndex, mVU->prog.total+1); return mVU->prog.total; } else { @@ -192,7 +192,7 @@ microVUt(int) mVUfindLeastUsedProg() { // frame-based decrementing system in combination with a program-execution-based incrementing // system. In english: if last_used >= 2 it means the program has been used for the current // or prev frame. if it's 0, the program hasn't been used for a while. -microVUt(void) mVUvsyncUpdate() { +microVUf(void) mVUvsyncUpdate() { microVU* mVU = mVUx; if (mVU->prog.total < mVU->prog.max) return; @@ -211,7 +211,7 @@ microVUt(void) mVUvsyncUpdate() { } // Compare Cached microProgram to mVU->regs->Micro -microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, bool cmpWholeProg) { +microVUf(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, bool cmpWholeProg) { microVU* mVU = mVUx; if (progUsed) { @@ -231,7 +231,7 @@ microVUt(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, b } // Searches for Cached Micro Program and sets prog.cur to it (returns 1 if program found, else returns 0) -microVUt(int) mVUsearchProg() { +microVUf(int) mVUsearchProg() { microVU* mVU = mVUx; if (mVU->prog.cleared) { // If cleared, we need to search for new program diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7a9c3326d3..c2445a93d4 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -137,24 +137,24 @@ extern void (*mVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); extern int mVUdebugNow; // Main Functions -microVUt(void) mVUinit(VURegs*); +microVUf(void) mVUinit(VURegs*); microVUx(void) mVUreset(); -microVUt(void) mVUclose(); -microVUt(void) mVUclear(u32, u32); +microVUf(void) mVUclose(); +microVUf(void) mVUclear(u32, u32); // Prototypes for Linux void __fastcall mVUcleanUpVU0(); void __fastcall mVUcleanUpVU1(); void* __fastcall mVUcompileVU0(u32 startPC, uptr pState); void* __fastcall mVUcompileVU1(u32 startPC, uptr pState); -microVUf(void) mVUopU(mF); -microVUf(void) mVUopL(mF); +mVUop(mVUopU); +mVUop(mVUopL); // Private Functions -microVUt(void) mVUclearProg(microVU* mVU, int progIndex); -microVUt(int) mVUfindLeastUsedProg(microVU* mVU); -microVUt(int) mVUsearchProg(); -microVUt(void) mVUcacheProg(int progIndex); +microVUf(void) mVUclearProg(microVU* mVU, int progIndex); +microVUf(int) mVUfindLeastUsedProg(microVU* mVU); +microVUf(int) mVUsearchProg(); +microVUf(void) mVUcacheProg(int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index 8556a6ec6a..0ab348f8be 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -26,23 +26,22 @@ // FMAC1 - Normal FMAC Opcodes //------------------------------------------------------------------ -#define getReg(reg, _reg_) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _X_Y_Z_W); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _X_Y_Z_W); \ +#define getReg(reg, _reg_) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _X_Y_Z_W); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _X_Y_Z_W); \ } -#define getZero(reg) { \ - if (_W) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], _X_Y_Z_W); } \ - else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ +#define getZero(reg) { \ + if (_W) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], _X_Y_Z_W); } \ + else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ } -#define getReg6(reg, _reg_) { \ - if (!_reg_) { getZero(reg); } \ - else { getReg(reg, _reg_); } \ +#define getReg6(reg, _reg_) { \ + if (!_reg_) { getZero(reg); } \ + else { getReg(reg, _reg_); } \ } -microVUt(void) mVUallocFMAC1a(int& Fd, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC1a(mV, int& Fd, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -51,61 +50,57 @@ microVUt(void) mVUallocFMAC1a(int& Fd, int& Fs, int& Ft) { else { getReg6(Ft, _Ft_); } } -microVUt(void) mVUallocFMAC1b(int& Fd) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC1b(mV, int& Fd) { if (!_Fd_) return; - if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmT1, _X_Y_Z_W); - mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 1); + if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmT1, _X_Y_Z_W); + mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 1); } //------------------------------------------------------------------ // FMAC2 - ABS/FTOI/ITOF Opcodes //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC2a(int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC2a(mV, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFs; getReg6(Fs, _Fs_); } -microVUt(void) mVUallocFMAC2b(int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC2b(mV, int& Ft) { if (!_Ft_) return; //if (CHECK_VU_OVERFLOW) mVUclamp1(Ft, xmmT1, _X_Y_Z_W); - mVUsaveReg(Ft, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUsaveReg(Ft, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } //------------------------------------------------------------------ // FMAC3 - BC(xyzw) FMAC Opcodes //------------------------------------------------------------------ -#define getReg3SS(reg, _reg_) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ +#define getReg3SS(reg, _reg_) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ } -#define getReg3(reg, _reg_) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ - mVUunpack_xyzw(reg, reg, 0); \ +#define getReg3(reg, _reg_) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _bc_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _bc_))); \ + mVUunpack_xyzw(reg, reg, 0); \ } -#define getZero3SS(reg) { \ - if (_bc_w) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], 1); } \ - else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ +#define getZero3SS(reg) { \ + if (_bc_w) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], 1); } \ + else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ } -#define getZero3(reg) { \ - if (_bc_w) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], 1); \ - mVUunpack_xyzw(reg, reg, 0); \ - } \ - else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ +#define getZero3(reg) { \ + if (_bc_w) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], 1); \ + mVUunpack_xyzw(reg, reg, 0); \ + } \ + else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ } -microVUt(void) mVUallocFMAC3a(int& Fd, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC3a(mV, int& Fd, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -124,26 +119,25 @@ microVUt(void) mVUallocFMAC3a(int& Fd, int& Fs, int& Ft) { } } -microVUt(void) mVUallocFMAC3b(int& Fd) { - mVUallocFMAC1b(Fd); +microVUt(void) mVUallocFMAC3b(mV, int& Fd) { + mVUallocFMAC1b(mVU, Fd); } //------------------------------------------------------------------ // FMAC4 - FMAC Opcodes Storing Result to ACC //------------------------------------------------------------------ -#define getReg4(reg, _reg_) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _xyzw_ACC); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _xyzw_ACC); \ +#define getReg4(reg, _reg_) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _xyzw_ACC); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, _xyzw_ACC); \ } -#define getZero4(reg) { \ - if (_W) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], _xyzw_ACC); } \ - else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ +#define getZero4(reg) { \ + if (_W) { mVUloadReg(reg, (uptr)&mVU->regs->VF[0].UL[0], _xyzw_ACC); } \ + else { SSE_XORPS_XMM_to_XMM(reg, reg); } \ } -microVUt(void) mVUallocFMAC4a(int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC4a(mV, int& ACC, int& Fs, int& Ft) { ACC = xmmACC; Fs = (_X_Y_Z_W == 15) ? xmmACC : xmmFs; Ft = xmmFt; @@ -162,9 +156,8 @@ microVUt(void) mVUallocFMAC4a(int& ACC, int& Fs, int& Ft) { } } -microVUt(void) mVUallocFMAC4b(int& ACC, int& Fs) { - microVU* mVU = mVUx; - if (CHECK_VU_OVERFLOW) mVUclamp1(Fs, xmmT1, _xyzw_ACC); +microVUt(void) mVUallocFMAC4b(mV, int& ACC, int& Fs) { + if (CHECK_VU_OVERFLOW) mVUclamp1(Fs, xmmT1, _xyzw_ACC); mVUmergeRegs(ACC, Fs, _X_Y_Z_W); } @@ -172,8 +165,7 @@ microVUt(void) mVUallocFMAC4b(int& ACC, int& Fs) { // FMAC5 - FMAC BC(xyzw) Opcodes Storing Result to ACC //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC5a(int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC5a(mV, int& ACC, int& Fs, int& Ft) { ACC = xmmACC; Fs = (_X_Y_Z_W == 15) ? xmmACC : xmmFs; Ft = xmmFt; @@ -192,23 +184,22 @@ microVUt(void) mVUallocFMAC5a(int& ACC, int& Fs, int& Ft) { } } -microVUt(void) mVUallocFMAC5b(int& ACC, int& Fs) { - mVUallocFMAC4b(ACC, Fs); +microVUt(void) mVUallocFMAC5b(mV, int& ACC, int& Fs) { + mVUallocFMAC4b(mVU, ACC, Fs); } //------------------------------------------------------------------ // FMAC6 - Normal FMAC Opcodes (I Reg) //------------------------------------------------------------------ -#define getIreg(reg, modXYZW) { \ - MOV32MtoR(gprT1, (uptr)&mVU->regs->VI[REG_I].UL); \ - SSE2_MOVD_R_to_XMM(reg, gprT1); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 8); \ - if (!((_XYZW_SS && modXYZW) || (_X_Y_Z_W == 8))) { mVUunpack_xyzw(reg, reg, 0); } \ +#define getIreg(reg, modXYZW) { \ + MOV32MtoR(gprT1, (uptr)&mVU->regs->VI[REG_I].UL); \ + SSE2_MOVD_R_to_XMM(reg, gprT1); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 8); \ + if (!((_XYZW_SS && modXYZW) || (_X_Y_Z_W == 8))) { mVUunpack_xyzw(reg, reg, 0); } \ } -microVUt(void) mVUallocFMAC6a(int& Fd, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC6a(mV, int& Fd, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -216,16 +207,15 @@ microVUt(void) mVUallocFMAC6a(int& Fd, int& Fs, int& Ft) { getReg6(Fs, _Fs_); } -microVUt(void) mVUallocFMAC6b(int& Fd) { - mVUallocFMAC1b(Fd); +microVUt(void) mVUallocFMAC6b(mV, int& Fd) { + mVUallocFMAC1b(mVU, Fd); } //------------------------------------------------------------------ // FMAC7 - FMAC Opcodes Storing Result to ACC (I Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC7a(int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC7a(mV, int& ACC, int& Fs, int& Ft) { ACC = xmmACC; Fs = (_X_Y_Z_W == 15) ? xmmACC : xmmFs; Ft = xmmFt; @@ -235,16 +225,15 @@ microVUt(void) mVUallocFMAC7a(int& ACC, int& Fs, int& Ft) { else { getReg4(Fs, _Fs_); } } -microVUt(void) mVUallocFMAC7b(int& ACC, int& Fs) { - mVUallocFMAC4b(ACC, Fs); +microVUt(void) mVUallocFMAC7b(mV, int& ACC, int& Fs) { + mVUallocFMAC4b(mVU, ACC, Fs); } //------------------------------------------------------------------ // FMAC8 - MADD FMAC Opcode Storing Result to Fd //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC8a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC8a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -264,19 +253,17 @@ microVUt(void) mVUallocFMAC8a(int& Fd, int& ACC, int& Fs, int& Ft) { } } -microVUt(void) mVUallocFMAC8b(int& Fd) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC8b(mV, int& Fd) { if (!_Fd_) return; - if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmT1, _xyzw_ACC); - mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 0); + if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmT1, _xyzw_ACC); + mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 0); } //------------------------------------------------------------------ // FMAC9 - MSUB FMAC Opcode Storing Result to Fd //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC9a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC9a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmT1; @@ -297,19 +284,17 @@ microVUt(void) mVUallocFMAC9a(int& Fd, int& ACC, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACC, xmmACC); } -microVUt(void) mVUallocFMAC9b(int& Fd) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC9b(mV, int& Fd) { if (!_Fd_) return; - if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmFt, _xyzw_ACC); - mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 0); + if (CHECK_VU_OVERFLOW) mVUclamp1(Fd, xmmFt, _xyzw_ACC); + mVUsaveReg(Fd, (uptr)&mVU->regs->VF[_Fd_].UL[0], _X_Y_Z_W, 0); } //------------------------------------------------------------------ // FMAC10 - MADD FMAC BC(xyzw) Opcode Storing Result to Fd //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC10a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC10a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -329,16 +314,15 @@ microVUt(void) mVUallocFMAC10a(int& Fd, int& ACC, int& Fs, int& Ft) { } } -microVUt(void) mVUallocFMAC10b(int& Fd) { - mVUallocFMAC8b(Fd); +microVUt(void) mVUallocFMAC10b(mV, int& Fd) { + mVUallocFMAC8b(mVU, Fd); } //------------------------------------------------------------------ // FMAC11 - MSUB FMAC BC(xyzw) Opcode Storing Result to Fd //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC11a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC11a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmT1; @@ -359,16 +343,15 @@ microVUt(void) mVUallocFMAC11a(int& Fd, int& ACC, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACC, xmmACC); } -microVUt(void) mVUallocFMAC11b(int& Fd) { - mVUallocFMAC9b(Fd); +microVUt(void) mVUallocFMAC11b(mV, int& Fd) { + mVUallocFMAC9b(mVU, Fd); } //------------------------------------------------------------------ // FMAC12 - MADD FMAC Opcode Storing Result to Fd (I Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC12a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC12a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -379,16 +362,15 @@ microVUt(void) mVUallocFMAC12a(int& Fd, int& ACC, int& Fs, int& Ft) { else { getReg4(Fs, _Fs_); } } -microVUt(void) mVUallocFMAC12b(int& Fd) { - mVUallocFMAC8b(Fd); +microVUt(void) mVUallocFMAC12b(mV, int& Fd) { + mVUallocFMAC8b(mVU, Fd); } //------------------------------------------------------------------ // FMAC13 - MSUB FMAC Opcode Storing Result to Fd (I Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC13a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC13a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmT1; @@ -400,16 +382,15 @@ microVUt(void) mVUallocFMAC13a(int& Fd, int& ACC, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACC, xmmACC); } -microVUt(void) mVUallocFMAC13b(int& Fd) { - mVUallocFMAC9b(Fd); +microVUt(void) mVUallocFMAC13b(mV, int& Fd) { + mVUallocFMAC9b(mVU, Fd); } //------------------------------------------------------------------ // FMAC14 - MADDA/MSUBA FMAC Opcodes //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC14a(int& ACCw, int& ACCr, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC14a(mV, int& ACCw, int& ACCr, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; ACCw = xmmACC; @@ -431,9 +412,8 @@ microVUt(void) mVUallocFMAC14a(int& ACCw, int& ACCr, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACCr, xmmACC); } -microVUt(void) mVUallocFMAC14b(int& ACCw, int& ACCr) { - microVU* mVU = mVUx; - if (CHECK_VU_OVERFLOW) mVUclamp1(ACCr, xmmFt, _xyzw_ACC); +microVUt(void) mVUallocFMAC14b(mV, int& ACCw, int& ACCr) { + if (CHECK_VU_OVERFLOW) mVUclamp1(ACCr, xmmFt, _xyzw_ACC); mVUmergeRegs(ACCw, ACCr, _X_Y_Z_W); } @@ -441,8 +421,7 @@ microVUt(void) mVUallocFMAC14b(int& ACCw, int& ACCr) { // FMAC15 - MADDA/MSUBA BC(xyzw) FMAC Opcode //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC15a(int& ACCw, int& ACCr, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC15a(mV, int& ACCw, int& ACCr, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; ACCw = xmmACC; @@ -464,16 +443,15 @@ microVUt(void) mVUallocFMAC15a(int& ACCw, int& ACCr, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACCr, xmmACC); } -microVUt(void) mVUallocFMAC15b(int& ACCw, int& ACCr) { - mVUallocFMAC14b(ACCw, ACCr); +microVUt(void) mVUallocFMAC15b(mV, int& ACCw, int& ACCr) { + mVUallocFMAC14b(mVU, ACCw, ACCr); } //------------------------------------------------------------------ // FMAC16 - MADDA/MSUBA FMAC Opcode (I Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC16a(int& ACCw, int& ACCr, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC16a(mV, int& ACCw, int& ACCr, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; ACCw = xmmACC; @@ -485,22 +463,21 @@ microVUt(void) mVUallocFMAC16a(int& ACCw, int& ACCr, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACCr, xmmACC); } -microVUt(void) mVUallocFMAC16b(int& ACCw, int& ACCr) { - mVUallocFMAC14b(ACCw, ACCr); +microVUt(void) mVUallocFMAC16b(mV, int& ACCw, int& ACCr) { + mVUallocFMAC14b(mVU, ACCw, ACCr); } //------------------------------------------------------------------ // FMAC17 - CLIP FMAC Opcode //------------------------------------------------------------------ -#define getReg9(reg, _reg_) { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], 1); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 1); \ - mVUunpack_xyzw(reg, reg, 0); \ +#define getReg9(reg, _reg_) { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], 1); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, 1); \ + mVUunpack_xyzw(reg, reg, 0); \ } -microVUt(void) mVUallocFMAC17a(int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC17a(mV, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; getReg6(Fs, _Fs_); @@ -511,8 +488,7 @@ microVUt(void) mVUallocFMAC17a(int& Fs, int& Ft) { // FMAC18 - OPMULA FMAC Opcode //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC18a(int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC18a(mV, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; ACC = xmmACC; @@ -527,16 +503,15 @@ microVUt(void) mVUallocFMAC18a(int& ACC, int& Fs, int& Ft) { SSE2_PSHUFD_XMM_to_XMM(Ft, Ft, 0xD2); // WYXZ } -microVUt(void) mVUallocFMAC18b(int& ACC, int& Fs) { - mVUallocFMAC4b(ACC, Fs); +microVUt(void) mVUallocFMAC18b(mV, int& ACC, int& Fs) { + mVUallocFMAC4b(mVU, ACC, Fs); } //------------------------------------------------------------------ // FMAC19 - OPMSUB FMAC Opcode //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC19a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC19a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmT1; @@ -553,21 +528,20 @@ microVUt(void) mVUallocFMAC19a(int& Fd, int& ACC, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACC, xmmACC); } -microVUt(void) mVUallocFMAC19b(int& Fd) { - mVUallocFMAC9b(Fd); +microVUt(void) mVUallocFMAC19b(mV, int& Fd) { + mVUallocFMAC9b(mVU, Fd); } //------------------------------------------------------------------ // FMAC22 - Normal FMAC Opcodes (Q Reg) //------------------------------------------------------------------ -#define getQreg(reg) { \ - mVUunpack_xyzw(reg, xmmPQ, readQ); \ - /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ +#define getQreg(reg) { \ + mVUunpack_xyzw(reg, xmmPQ, readQ); \ + /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ } -microVUt(void) mVUallocFMAC22a(int& Fd, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC22a(mV, int& Fd, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -575,16 +549,15 @@ microVUt(void) mVUallocFMAC22a(int& Fd, int& Fs, int& Ft) { getReg6(Fs, _Fs_); } -microVUt(void) mVUallocFMAC22b(int& Fd) { - mVUallocFMAC1b(Fd); +microVUt(void) mVUallocFMAC22b(mV, int& Fd) { + mVUallocFMAC1b(mVU, Fd); } //------------------------------------------------------------------ // FMAC23 - FMAC Opcodes Storing Result to ACC (Q Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC23a(int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC23a(mV, int& ACC, int& Fs, int& Ft) { ACC = xmmACC; Fs = (_X_Y_Z_W == 15) ? xmmACC : xmmFs; Ft = xmmFt; @@ -594,16 +567,15 @@ microVUt(void) mVUallocFMAC23a(int& ACC, int& Fs, int& Ft) { else { getReg4(Fs, _Fs_); } } -microVUt(void) mVUallocFMAC23b(int& ACC, int& Fs) { - mVUallocFMAC4b(ACC, Fs); +microVUt(void) mVUallocFMAC23b(mV, int& ACC, int& Fs) { + mVUallocFMAC4b(mVU, ACC, Fs); } //------------------------------------------------------------------ // FMAC24 - MADD FMAC Opcode Storing Result to Fd (Q Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC24a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC24a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmFs; @@ -614,16 +586,15 @@ microVUt(void) mVUallocFMAC24a(int& Fd, int& ACC, int& Fs, int& Ft) { else { getReg4(Fs, _Fs_); } } -microVUt(void) mVUallocFMAC24b(int& Fd) { - mVUallocFMAC8b(Fd); +microVUt(void) mVUallocFMAC24b(mV, int& Fd) { + mVUallocFMAC8b(mVU, Fd); } //------------------------------------------------------------------ // FMAC25 - MSUB FMAC Opcode Storing Result to Fd (Q Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC25a(int& Fd, int& ACC, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC25a(mV, int& Fd, int& ACC, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; Fd = xmmT1; @@ -635,16 +606,15 @@ microVUt(void) mVUallocFMAC25a(int& Fd, int& ACC, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACC, xmmACC); } -microVUt(void) mVUallocFMAC25b(int& Fd) { - mVUallocFMAC9b(Fd); +microVUt(void) mVUallocFMAC25b(mV, int& Fd) { + mVUallocFMAC9b(mVU, Fd); } //------------------------------------------------------------------ // FMAC26 - MADDA/MSUBA FMAC Opcode (Q Reg) //------------------------------------------------------------------ -microVUt(void) mVUallocFMAC26a(int& ACCw, int& ACCr, int& Fs, int& Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUallocFMAC26a(mV, int& ACCw, int& ACCr, int& Fs, int& Ft) { Fs = xmmFs; Ft = xmmFt; ACCw = xmmACC; @@ -656,8 +626,8 @@ microVUt(void) mVUallocFMAC26a(int& ACCw, int& ACCr, int& Fs, int& Ft) { SSE_MOVAPS_XMM_to_XMM(ACCr, xmmACC); } -microVUt(void) mVUallocFMAC26b(int& ACCw, int& ACCr) { - mVUallocFMAC14b(ACCw, ACCr); +microVUt(void) mVUallocFMAC26b(mV, int& ACCw, int& ACCr) { + mVUallocFMAC14b(mVU, ACCw, ACCr); } //------------------------------------------------------------------ @@ -688,24 +658,20 @@ microVUt(void) mVUallocSFLAGb(int reg, int fInstance) { MOV32RtoR(fInstance, reg); } -microVUt(void) mVUallocMFLAGa(int reg, int fInstance) { - microVU* mVU = mVUx; +microVUt(void) mVUallocMFLAGa(mV, int reg, int fInstance) { MOVZX32M16toR(reg, (uptr)&mVU->macFlag[fInstance]); } -microVUt(void) mVUallocMFLAGb(int reg, int fInstance) { - microVU* mVU = mVUx; +microVUt(void) mVUallocMFLAGb(mV, int reg, int fInstance) { //AND32ItoR(reg, 0xffff); MOV32RtoM((uptr)&mVU->macFlag[fInstance], reg); } -microVUt(void) mVUallocCFLAGa(int reg, int fInstance) { - microVU* mVU = mVUx; +microVUt(void) mVUallocCFLAGa(mV, int reg, int fInstance) { MOV32MtoR(reg, (uptr)&mVU->clipFlag[fInstance]); } -microVUt(void) mVUallocCFLAGb(int reg, int fInstance) { - microVU* mVU = mVUx; +microVUt(void) mVUallocCFLAGb(mV, int reg, int fInstance) { MOV32RtoM((uptr)&mVU->clipFlag[fInstance], reg); } @@ -713,18 +679,16 @@ microVUt(void) mVUallocCFLAGb(int reg, int fInstance) { // VI Reg Allocators //------------------------------------------------------------------ -microVUt(void) mVUallocVIa(int GPRreg, int _reg_) { - microVU* mVU = mVUx; +microVUt(void) mVUallocVIa(mV, int GPRreg, int _reg_) { if (!_reg_) { XOR32RtoR(GPRreg, GPRreg); } else if (isMMX(_reg_)) { MOVD32MMXtoR(GPRreg, mmVI(_reg_)); } else { MOVZX32Rm16toR(GPRreg, gprR, (_reg_ - 9) * 16); } } -microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { - microVU* mVU = mVUx; +microVUt(void) mVUallocVIb(mV, int GPRreg, int _reg_) { if (backupVI) { // Backs up reg to memory (used when VI is modified b4 a branch) MOV32RtoM((uptr)&mVU->VIbackup[1], GPRreg); - mVUallocVIa(GPRreg, _reg_); + mVUallocVIa(mVU, GPRreg, _reg_); MOV32RtoM((uptr)&mVU->VIbackup[0], GPRreg); MOV32MtoR(GPRreg, (uptr)&mVU->VIbackup[1]); } @@ -737,34 +701,34 @@ microVUt(void) mVUallocVIb(int GPRreg, int _reg_) { // P Reg Allocator //------------------------------------------------------------------ -#define getPreg(reg) { \ - mVUunpack_xyzw(reg, xmmPQ, (2 + readP)); \ - /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ +#define getPreg(reg) { \ + mVUunpack_xyzw(reg, xmmPQ, (2 + readP)); \ + /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ } //------------------------------------------------------------------ // Lower Instruction Allocator Helpers //------------------------------------------------------------------ -#define getReg5(reg, _reg_, _fxf_) { \ - if (!_reg_) { \ - if (_fxf_ < 3) { SSE_XORPS_XMM_to_XMM(reg, reg); } \ - else { mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], 1); } \ - } \ - else { \ - mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _fxf_))); \ - if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _fxf_))); \ - } \ +#define getReg5(reg, _reg_, _fxf_) { \ + if (!_reg_) { \ + if (_fxf_ < 3) { SSE_XORPS_XMM_to_XMM(reg, reg); } \ + else { mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], 1); } \ + } \ + else { \ + mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], (1 << (3 - _fxf_))); \ + if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT2, (1 << (3 - _fxf_))); \ + } \ } // Doesn't Clamp -#define getReg7(reg, _reg_) { \ - if (!_reg_) { getZero(reg); } \ - else { mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _X_Y_Z_W); } \ +#define getReg7(reg, _reg_) { \ + if (!_reg_) { getZero(reg); } \ + else { mVUloadReg(reg, (uptr)&mVU->regs->VF[_reg_].UL[0], _X_Y_Z_W); } \ } // VF to GPR -#define getReg8(GPRreg, _reg_, _fxf_) { \ - if (!_reg_ && (_fxf_ < 3)) { XOR32RtoR(GPRreg, GPRreg); } \ - else { MOV32MtoR(GPRreg, (uptr)&mVU->regs->VF[_reg_].UL[0]); } \ +#define getReg8(GPRreg, _reg_, _fxf_) { \ + if (!_reg_ && (_fxf_ < 3)) { XOR32RtoR(GPRreg, GPRreg); } \ + else { MOV32MtoR(GPRreg, (uptr)&mVU->regs->VF[_reg_].UL[0]); } \ } diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index cb32980690..0422662fba 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -61,8 +61,7 @@ } \ } -microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg1(Ft); @@ -73,8 +72,7 @@ microVUt(void) mVUanalyzeFMAC1(int Fd, int Fs, int Ft) { // FMAC2 - ABS/FTOI/ITOF Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeFMAC2(int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFMAC2(mV, int Fs, int Ft) { analyzeReg1(Fs); analyzeReg2(Ft, 0); } @@ -92,8 +90,7 @@ microVUt(void) mVUanalyzeFMAC2(int Fs, int Ft) { } \ } -microVUt(void) mVUanalyzeFMAC3(int Fd, int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { mVUinfo |= _doStatus; analyzeReg1(Fs); analyzeReg3(Ft); @@ -108,8 +105,7 @@ microVUt(void) mVUanalyzeFMAC3(int Fd, int Fs, int Ft) { if (reg) { mVUstall = aMax(mVUstall, aReg(reg).w); } \ } -microVUt(void) mVUanalyzeFMAC4(int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFMAC4(mV, int Fs, int Ft) { mVUinfo |= _doClip; analyzeReg1(Fs); analyzeReg4(Ft); @@ -123,16 +119,14 @@ microVUt(void) mVUanalyzeFMAC4(int Fs, int Ft) { #define analyzeVIreg2(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; mVUinfo |= _writesVI; mVU->VIbackup[0] = reg; } } #define analyzeVIreg3(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; } } -microVUt(void) mVUanalyzeIALU1(int Id, int Is, int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeIALU1(mV, int Id, int Is, int It) { if (!Id) { mVUinfo |= _isNOP; } analyzeVIreg1(Is); analyzeVIreg1(It); analyzeVIreg2(Id, 1); } -microVUt(void) mVUanalyzeIALU2(int Is, int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeIALU2(mV, int Is, int It) { if (!It) { mVUinfo |= _isNOP; } analyzeVIreg1(Is); analyzeVIreg2(It, 1); @@ -159,8 +153,7 @@ microVUt(void) mVUanalyzeIALU2(int Is, int It) { } \ } -microVUt(void) mVUanalyzeMR32(int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeMR32(mV, int Fs, int Ft) { if (!Ft) { mVUinfo |= _isNOP; } analyzeReg6(Fs); analyzeReg2(Ft, 1); @@ -191,8 +184,7 @@ microVUt(void) mVUanalyzeMR32(int Fs, int Ft) { #define analyzeQreg(x) { mVUregsTemp.q = x; mVUstall = aMax(mVUstall, mVUregs.q); } #define analyzePreg(x) { mVUregsTemp.p = x; mVUstall = aMax(mVUstall, ((mVUregs.p) ? (mVUregs.p - 1) : 0)); } -microVUt(void) mVUanalyzeFDIV(int Fs, int Fsf, int Ft, int Ftf, u8 xCycles) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFDIV(mV, int Fs, int Fsf, int Ft, int Ftf, u8 xCycles) { mVUprint("microVU: DIV Opcode"); analyzeReg5(Fs, Fsf); analyzeReg5(Ft, Ftf); @@ -203,15 +195,13 @@ microVUt(void) mVUanalyzeFDIV(int Fs, int Fsf, int Ft, int Ftf, u8 xCycles) { // EFU - EFU Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeEFU1(int Fs, int Fsf, u8 xCycles) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeEFU1(mV, int Fs, int Fsf, u8 xCycles) { mVUprint("microVU: EFU Opcode"); analyzeReg5(Fs, Fsf); analyzePreg(xCycles); } -microVUt(void) mVUanalyzeEFU2(int Fs, u8 xCycles) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeEFU2(mV, int Fs, u8 xCycles) { mVUprint("microVU: EFU Opcode"); analyzeReg1b(Fs); analyzePreg(xCycles); @@ -221,8 +211,7 @@ microVUt(void) mVUanalyzeEFU2(int Fs, u8 xCycles) { // MFP - MFP Opcode //------------------------------------------------------------------ -microVUt(void) mVUanalyzeMFP(int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeMFP(mV, int Ft) { if (!Ft) { mVUinfo |= _isNOP; } analyzeReg2(Ft, 1); } @@ -231,8 +220,7 @@ microVUt(void) mVUanalyzeMFP(int Ft) { // MOVE - MOVE Opcode //------------------------------------------------------------------ -microVUt(void) mVUanalyzeMOVE(int Fs, int Ft) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeMOVE(mV, int Fs, int Ft) { if (!Ft || (Ft == Fs)) { mVUinfo |= _isNOP; } analyzeReg1b(Fs); analyzeReg2(Ft, 1); @@ -243,8 +231,7 @@ microVUt(void) mVUanalyzeMOVE(int Fs, int Ft) { // LQx - LQ/LQD/LQI Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeLQ(int Ft, int Is, bool writeIs) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeLQ(mV, int Ft, int Is, bool writeIs) { analyzeVIreg1(Is); analyzeReg2(Ft, 1); if (!Ft) { mVUinfo |= (writeIs && Is) ? _noWriteVF : _isNOP; } @@ -255,8 +242,7 @@ microVUt(void) mVUanalyzeLQ(int Ft, int Is, bool writeIs) { // SQx - SQ/SQD/SQI Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeSQ(int Fs, int It, bool writeIt) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeSQ(mV, int Fs, int It, bool writeIt) { analyzeReg1b(Fs); analyzeVIreg1(It); if (writeIt) { analyzeVIreg2(It, 1); } @@ -268,14 +254,12 @@ microVUt(void) mVUanalyzeSQ(int Fs, int It, bool writeIt) { #define analyzeRreg() { mVUregsTemp.r = 1; } -microVUt(void) mVUanalyzeR1(int Fs, int Fsf) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeR1(mV, int Fs, int Fsf) { analyzeReg5(Fs, Fsf); analyzeRreg(); } -microVUt(void) mVUanalyzeR2(int Ft, bool canBeNOP) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { if (!Ft) { mVUinfo |= ((canBeNOP) ? _isNOP : _noWriteVF); } analyzeReg2(Ft, 1); analyzeRreg(); @@ -285,8 +269,7 @@ microVUt(void) mVUanalyzeR2(int Ft, bool canBeNOP) { // Sflag - Status Flag Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeSflag(int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeSflag(mV, int It) { if (!It) { mVUinfo |= _isNOP; } else { mVUinfo |= _swapOps; @@ -300,8 +283,7 @@ microVUt(void) mVUanalyzeSflag(int It) { analyzeVIreg3(It, 1); } -microVUt(void) mVUanalyzeFSSET() { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeFSSET(mV) { mVUinfo |= _isFSSET; // mVUinfo &= ~_doStatus; // Note: I'm not entirely sure if the non-sticky flags @@ -314,8 +296,7 @@ microVUt(void) mVUanalyzeFSSET() { // Mflag - Mac Flag Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeMflag(int Is, int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { if (!It) { mVUinfo |= _isNOP; } else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo |= _swapOps; @@ -335,8 +316,7 @@ microVUt(void) mVUanalyzeMflag(int Is, int It) { // Cflag - Clip Flag Opcodes //------------------------------------------------------------------ -microVUt(void) mVUanalyzeCflag(int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeCflag(mV, int It) { mVUinfo |= _swapOps; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } analyzeVIreg3(It, 1); @@ -349,8 +329,7 @@ microVUt(void) mVUanalyzeCflag(int It) { #define analyzeXGkick1() { mVUstall = aMax(mVUstall, mVUregs.xgkick); } #define analyzeXGkick2(x) { mVUregsTemp.xgkick = x; } -microVUt(void) mVUanalyzeXGkick(int Fs, int xCycles) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { analyzeVIreg1(Fs); analyzeXGkick1(); analyzeXGkick2(xCycles); @@ -378,14 +357,12 @@ microVUt(void) mVUanalyzeXGkick(int Fs, int xCycles) { } \ } -microVUt(void) mVUanalyzeBranch1(int Is) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeBranch1(mV, int Is) { if (mVUregs.VI[Is] || mVUstall) { analyzeVIreg1(Is); } else { analyzeBranchVI(Is, _memReadIs); } } -microVUt(void) mVUanalyzeBranch2(int Is, int It) { - microVU* mVU = mVUx; +microVUt(void) mVUanalyzeBranch2(mV, int Is, int It) { if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is); analyzeVIreg1(It); } else { analyzeBranchVI(Is, _memReadIs); analyzeBranchVI(It, _memReadIt);} } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index c54ee08bb8..857f001f72 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -23,7 +23,7 @@ //------------------------------------------------------------------ #define branchCase(JMPcc, nJMPcc) \ - mVUsetupBranch(xStatus, xMac, xClip, xCycles); \ + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ CMP16ItoM((uptr)&mVU->branch, 0); \ incPC2(1); \ if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = new microBlockManager(); } \ @@ -54,8 +54,8 @@ #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } -#define doUpperOp() { mVUopU(1); mVUdivSet(); } -#define doLowerOp() { incPC(-1); mVUopL(1); incPC(1); } +#define doUpperOp() { mVUopU(mVU, 1); mVUdivSet(mVU); } +#define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } #define doIbit() { if (curI & _Ibit_) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } //------------------------------------------------------------------ @@ -63,20 +63,19 @@ //------------------------------------------------------------------ // Used by mVUsetupRange -microVUt(void) mVUcheckIsSame() { - microVU* mVU = mVUx; +microVUt(void) mVUcheckIsSame(mV) { if (mVU->prog.isSame == -1) { mVU->prog.isSame = !memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); } if (mVU->prog.isSame == 0) { - mVUcacheProg(mVU->prog.cur); + if (!isVU1) mVUcacheProg<0>(mVU->prog.cur); + else mVUcacheProg<1>(mVU->prog.cur); } } // Sets up microProgram PC ranges based on whats been recompiled -microVUt(void) mVUsetupRange(u32 pc) { - microVU* mVU = mVUx; +microVUt(void) mVUsetupRange(mV, u32 pc) { if (mVUcurProg.range[0] == -1) { mVUcurProg.range[0] = (s32)pc; @@ -84,28 +83,26 @@ microVUt(void) mVUsetupRange(u32 pc) { } else if (mVUcurProg.range[0] > (s32)pc) { mVUcurProg.range[0] = (s32)pc; - mVUcheckIsSame(); + mVUcheckIsSame(mVU); } else if (mVUcurProg.range[1] < (s32)pc) { mVUcurProg.range[1] = (s32)pc; - mVUcheckIsSame(); + mVUcheckIsSame(mVU); } } // Recompiles Code for Proper Flags and Q/P regs on Block Linkings -microVUt(void) mVUsetupBranch(int* xStatus, int* xMac, int* xClip, int xCycles) { - microVU* mVU = mVUx; +microVUt(void) mVUsetupBranch(mV, int* xStatus, int* xMac, int* xClip, int xCycles) { mVUprint("mVUsetupBranch"); // Shuffle Flag Instances - mVUsetupFlags(xStatus, xMac, xClip, xCycles); + mVUsetupFlags(mVU, xStatus, xMac, xClip, xCycles); // Shuffle P/Q regs since every block starts at instance #0 if (mVU->p || mVU->q) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, shufflePQ); } } -microVUt(void) mVUincCycles(int x) { - microVU* mVU = mVUx; +microVUt(void) mVUincCycles(mV, int x) { mVUcycles += x; for (int z = 31; z > 0; z--) { calcCycles(mVUregs.VF[z].x, x); @@ -132,8 +129,7 @@ microVUt(void) mVUincCycles(int x) { calcCycles(mVUregs.r, x); } -microVUt(void) mVUsetCycles() { - microVU* mVU = mVUx; +microVUt(void) mVUsetCycles(mV) { incCycles(mVUstall); if (mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1] && mVUregsTemp.VFreg[0]) { // If upper Op && lower Op write to same VF reg mVUinfo |= (mVUregsTemp.r || mVUregsTemp.VI) ? _noWriteVF : _isNOP; // If lower Op doesn't modify anything else, then make it a NOP @@ -155,13 +151,12 @@ microVUt(void) mVUsetCycles() { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fClip) { - microVU* mVU = mVUx; +microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, int fClip) { // Save P/Q Regs if (qInst) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_Q].UL, xmmPQ); - if (vuIndex) { + if (isVU1) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, pInst ? 3 : 2); SSE_MOVSS_XMM_to_M32((uptr)&mVU->regs->VI[REG_P].UL, xmmPQ); } @@ -171,13 +166,13 @@ microVUt(void) mVUendProgram(int qInst, int pInst, int fStatus, int fMac, int fC getFlagReg(fStatus, fStatus); MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); } - mVUallocMFLAGa(gprT1, fMac); - mVUallocCFLAGa(gprT2, fClip); + mVUallocMFLAGa(mVU, gprT1, fMac); + mVUallocCFLAGa(mVU, gprT2, fClip); MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); // Clear 'is busy' Flags, Save PC, and Jump to Exit Point - AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (vuIndex ? ~0x100 : ~0x001)); // VBS0/VBS1 flag + AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (isVU1 ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); @@ -191,17 +186,16 @@ void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting void __fastcall mVUprintPC1(u32 PC) { Console::Write("Block PC [%04x] ", params PC); } void __fastcall mVUprintPC2(u32 PC) { Console::Write("[%04x]\n", params PC); } -microVUt(void) mVUtestCycles() { - microVU* mVU = mVUx; +microVUt(void) mVUtestCycles(mV) { iPC = mVUstartPC; mVUdebugNOW(0); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); u8* jmp8 = JG8(0); MOV32ItoR(gprT2, xPC); - if (!vuIndex) CALLFunc((uptr)mVUwarning0); - else CALLFunc((uptr)mVUwarning1); + if (!isVU1) CALLFunc((uptr)mVUwarning0); + else CALLFunc((uptr)mVUwarning1); MOV32ItoR(gprR, Roffset); // Restore gprR - mVUendProgram(0, 0, sI, 0, cI); + mVUendProgram(mVU, 0, 0, sI, 0, cI); x86SetJ8(jmp8); } @@ -209,7 +203,7 @@ microVUt(void) mVUtestCycles() { // Recompiler //------------------------------------------------------------------ -microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { +microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { microVU* mVU = mVUx; u8* thisPtr = x86Ptr; @@ -225,7 +219,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (pBlock) { return pBlock->x86ptrStart; } // Setup Program Bounds/Range - mVUsetupRange(startPC); + mVUsetupRange(mVU, startPC); // First Pass iPC = startPC / 4; @@ -250,25 +244,25 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUinfo = 0; startLoop(); incCycles(1); - mVUopU(0); + mVUopU(mVU, 0); if (curI & _Ebit_) { branch = 1; } if (curI & _MDTbit_) { branch = 4; } if (curI & _Ibit_) { mVUinfo |= _isNOP; } - else { incPC(-1); mVUopL(0); incPC(1); } - mVUsetCycles(); + else { incPC(-1); mVUopL(mVU, 0); incPC(1); } + mVUsetCycles(mVU); if (mVU->p) { mVUinfo |= _readP; } if (mVU->q) { mVUinfo |= _readQ; } if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(); branchEbit(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } + if (mVUbranch) { mVUsetFlagInfo(mVU); branchEbit(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } incPC(1); mVUcount++; } // Sets Up Flag instances int xStatus[4], xMac[4], xClip[4]; - int xCycles = mVUsetFlags(xStatus, xMac, xClip); - mVUtestCycles(); + int xCycles = mVUsetFlags(mVU, xStatus, xMac, xClip); + mVUtestCycles(mVU); // Second Pass iPC = mVUstartPC; @@ -279,14 +273,14 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (isEOB) { x = 0xffff; } if (isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { mVUopL(1); incPC(1); doUpperOp(); } - if (doXGKICK) { mVU_XGKICK_DELAY(); } + else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } + if (doXGKICK) { mVU_XGKICK_DELAY(mVU); } if (!isBdelay) { incPC(1); } else { microBlock* bBlock = NULL; u32* ajmp = 0; - mVUsetupRange(xPC); + mVUsetupRange(mVU, xPC); mVUdebugNOW(1); switch (mVUbranch) { @@ -302,7 +296,7 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { incPC(-3); // Go back to branch opcode (to get branch imm addr) if (eBitBranch) { iPC = branchAddr/4; goto eBitTemination; } // E-bit Was Set on Branch - mVUsetupBranch(xStatus, xMac, xClip, xCycles); + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); if (mVUblocks[branchAddr/8] == NULL) mVUblocks[branchAddr/8] = new microBlockManager(); @@ -317,15 +311,15 @@ microVUt(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUprint("mVUcompile JR/JALR"); memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - mVUsetupBranch(xStatus, xMac, xClip, xCycles); + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - mVUbackupRegs(); + mVUbackupRegs(mVU); MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) - if (!vuIndex) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) - else CALLFunc((uptr)mVUcompileVU1); - mVUrestoreRegs(); + if (!isVU1) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) + else CALLFunc((uptr)mVUcompileVU1); + mVUrestoreRegs(mVU); JMPR(gprT1); // Jump to rec-code address return thisPtr; } @@ -380,11 +374,11 @@ eBitTemination: AND32ItoR (flagReg, 0x0fcf); OR32MtoR (flagReg, (uptr)&mVU->divFlag); } - if (doXGKICK) { mVU_XGKICK_DELAY(); } + if (doXGKICK) { mVU_XGKICK_DELAY(mVU); } // Do E-bit end stuff here - mVUsetupRange(xPC - 8); - mVUendProgram(mVU->q, mVU->p, lStatus, lMac, lClip); + mVUsetupRange(mVU, xPC - 8); + mVUendProgram(mVU, mVU->q, mVU->p, lStatus, lMac, lClip); return thisPtr; } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 456dcb8003..0854b0bad3 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -23,13 +23,12 @@ //------------------------------------------------------------------ // Generates the code for entering recompiled blocks -microVUt(void) mVUdispatcherA() { - microVU* mVU = mVUx; +microVUt(void) mVUdispatcherA(mV) { mVU->startFunct = x86Ptr; // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. - if (!vuIndex) { CALLFunc((uptr)mVUexecuteVU0); } - else { CALLFunc((uptr)mVUexecuteVU1); } + if (!isVU1) { CALLFunc((uptr)mVUexecuteVU0); } + else { CALLFunc((uptr)mVUexecuteVU1); } // Backup cpu state PUSH32R(EBX); @@ -72,8 +71,7 @@ microVUt(void) mVUdispatcherA() { } // Generates the code to exit from recompiled blocks -microVUt(void) mVUdispatcherB() { - microVU* mVU = mVUx; +microVUt(void) mVUdispatcherB(mV) { mVU->exitFunct = x86Ptr; // Load EE's MXCSR state @@ -87,8 +85,8 @@ microVUt(void) mVUdispatcherB() { } // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. - if (!vuIndex) { CALLFunc((uptr)mVUcleanUpVU0); } - else { CALLFunc((uptr)mVUcleanUpVU1); } + if (!isVU1) { CALLFunc((uptr)mVUcleanUpVU0); } + else { CALLFunc((uptr)mVUcleanUpVU1); } // Restore cpu state POP32R(EDI); @@ -107,7 +105,7 @@ microVUt(void) mVUdispatcherB() { //------------------------------------------------------------------ // Executes for number of cycles -microVUt(void*) __fastcall mVUexecute(u32 startPC, u32 cycles) { +microVUx(void*) __fastcall mVUexecute(u32 startPC, u32 cycles) { microVU* mVU = mVUx; //mVUprint("microVU%x: startPC = 0x%x, cycles = 0x%x", params vuIndex, startPC, cycles); @@ -125,7 +123,7 @@ microVUt(void*) __fastcall mVUexecute(u32 startPC, u32 cycles) { // Cleanup Functions //------------------------------------------------------------------ -microVUt(void) mVUcleanUp() { +microVUx(void) mVUcleanUp() { microVU* mVU = mVUx; //mVUprint("microVU: Program exited successfully!"); //mVUprint("microVU: VF0 = {%x,%x,%x,%x}", params mVU->regs->VF[0].UL[0], mVU->regs->VF[0].UL[1], mVU->regs->VF[0].UL[2], mVU->regs->VF[0].UL[3]); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 3097d6c87b..09569e9240 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -19,8 +19,7 @@ #pragma once // Sets FDIV Flags at the proper time -microVUt(void) mVUdivSet() { - microVU* mVU = mVUx; +microVUt(void) mVUdivSet(mV) { int flagReg1, flagReg2; if (doDivFlag) { getFlagReg(flagReg1, fsInstance); @@ -31,8 +30,7 @@ microVUt(void) mVUdivSet() { } // Optimizes out unneeded status flag updates -microVUt(void) mVUstatusFlagOp() { - microVU* mVU = mVUx; +microVUt(void) mVUstatusFlagOp(mV) { int curPC = iPC; int i = mVUcount; bool runLoop = 1; @@ -52,7 +50,7 @@ microVUt(void) mVUstatusFlagOp() { } } iPC = curPC; - DevCon::Status("microVU%d: FSSET Optimization", params vuIndex); + DevCon::Status("microVU%d: FSSET Optimization", params getIndex); } int findFlagInst(int* fFlag, int cycles) { @@ -74,8 +72,7 @@ void sortFlag(int* fFlag, int* bFlag, int cycles) { #define sFlagCond ((doStatus && !mVUsFlagHack) || isFSSET || doDivFlag) // Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! -microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { - microVU* mVU = mVUx; +microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { int endPC = iPC; u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking @@ -120,9 +117,9 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { for (mVUcount = 0; mVUcount < xCount; mVUcount++) { if (isFSSET) { if (__Status) { // Don't Optimize out on the last ~4+ instructions - if ((xCount - mVUcount) > aCount) { mVUstatusFlagOp(); } + if ((xCount - mVUcount) > aCount) { mVUstatusFlagOp(mVU); } } - else mVUstatusFlagOp(); + else mVUstatusFlagOp(mVU); } cycles += mVUstall; @@ -151,8 +148,7 @@ microVUt(int) mVUsetFlags(int* xStatus, int* xMac, int* xClip) { #define shuffleClip ((bClip[3]<<6)|(bClip[2]<<4)|(bClip[1]<<2)|bClip[0]) // Recompiles Code for Proper Flags on Block Linkings -microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { - microVU* mVU = mVUx; +microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles) { if (__Status && !mVUflagHack) { int bStatus[4]; @@ -184,17 +180,16 @@ microVUt(void) mVUsetupFlags(int* xStatus, int* xMac, int* xClip, int cycles) { } } -#define shortBranch() { \ - if (branch == 3) { \ - mVUflagPass(aBranchAddr, (xCount - (mVUcount+1))); \ - mVUcount = 4; \ - } \ +#define shortBranch() { \ + if (branch == 3) { \ + mVUflagPass(mVU, aBranchAddr, (xCount - (mVUcount+1))); \ + mVUcount = 4; \ + } \ } // Scan through instructions and check if flags are read (FSxxx, FMxxx, FCxxx opcodes) -microVUx(void) mVUflagPass(u32 startPC, u32 xCount) { +void mVUflagPass(mV, u32 startPC, u32 xCount) { - microVU* mVU = mVUx; int oldPC = iPC; int oldCount = mVUcount; int oldBranch = mVUbranch; @@ -206,7 +201,7 @@ microVUx(void) mVUflagPass(u32 startPC, u32 xCount) { incPC(1); if ( curI & _Ebit_ ) { branch = 1; } if ( curI & _MDTbit_ ) { branch = 4; } - if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(3); incPC(1); } + if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(mVU, 3); incPC(1); } if (branch >= 2) { shortBranch(); break; } else if (branch == 1) { branch = 2; } if (mVUbranch) { branch = (mVUbranch >= 9) ? 5 : 3; aBranchAddr = branchAddr; mVUbranch = 0; } @@ -224,17 +219,16 @@ microVUx(void) mVUflagPass(u32 startPC, u32 xCount) { #define branchType3 else // Conditional Branch // Checks if the first 4 instructions of a block will read flags -microVUt(void) mVUsetFlagInfo() { - microVU* mVU = mVUx; - branchType1 { incPC(-1); mVUflagPass(branchAddr, 4); incPC(1); } +microVUt(void) mVUsetFlagInfo(mV) { + branchType1 { incPC(-1); mVUflagPass(mVU, branchAddr, 4); incPC(1); } branchType2 { mVUflagInfo |= 0xfff; } branchType3 { incPC(-1); - mVUflagPass(branchAddr, 4); + mVUflagPass(mVU, branchAddr, 4); int backupFlagInfo = mVUflagInfo; mVUflagInfo = 0; incPC(4); // Branch Not Taken - mVUflagPass(xPC, 4); + mVUflagPass(mVU, xPC, 4); incPC(-3); mVUflagInfo |= backupFlagInfo; } diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index bdbda985a1..9e5451a8fd 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -38,7 +38,7 @@ microVUx(void) __mVULog(const char* fmt, ...) { #define commaIf() { if (bitX[6]) { mVUlog(","); bitX[6] = 0; } } -microVUt(void) __mVUdumpProgram(int progIndex) { +microVUx(void) __mVUdumpProgram(int progIndex) { microVU* mVU = mVUx; bool bitX[7]; char str[30]; @@ -46,6 +46,7 @@ microVUt(void) __mVUdumpProgram(int progIndex) { int bBranch = mVUbranch; int bCode = mVU->code; int bPC = iPC; + int vuIndex = (mVU == µVU1) ? 1 : 0; mVUbranch = 0; sprintf(str, "%s\\microVU%d prog - %02d.html", LOGS_DIR, vuIndex, progIndex); @@ -88,7 +89,7 @@ microVUt(void) __mVUdumpProgram(int progIndex) { iPC = (i+1); mVUlog("", i*4); mVUlog("[%04x] (%08x) ", i*4, mVU->code); - mVUopU(); + mVUopU(mVU, 2); if (bitX[5]) { mVUlog(" ("); @@ -110,7 +111,7 @@ microVUt(void) __mVUdumpProgram(int progIndex) { } else { mVUlog("
\n[%04x] (%08x) ", i*4, mVU->code); - mVUopL(); + mVUopL(mVU, 2); mVUlog("\n\n

"); } } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 85bc94dce3..7da82eec97 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -42,9 +42,8 @@ x86SetJ8(aJump); \ } -microVUf(void) mVU_DIV(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFDIV(_Fs_, _Fsf_, _Ft_, _Ftf_, 7); } +mVUop(mVU_DIV) { + pass1 { mVUanalyzeFDIV(mVU, _Fs_, _Fsf_, _Ft_, _Ftf_, 7); } pass2 { u8 *ajmp, *bjmp, *cjmp, *djmp; getReg5(xmmFs, _Fs_, _Fsf_); @@ -69,7 +68,7 @@ microVUf(void) mVU_DIV(mF) { x86SetJ8(cjmp); MOV32ItoM((uptr)&mVU->divFlag, 0); // Clear I/D flags SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); + if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); @@ -79,9 +78,8 @@ microVUf(void) mVU_DIV(mF) { pass3 { mVUlog("DIV Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } -microVUf(void) mVU_SQRT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFDIV(0, 0, _Ft_, _Ftf_, 7); } +mVUop(mVU_SQRT) { + pass1 { mVUanalyzeFDIV(mVU, 0, 0, _Ft_, _Ftf_, 7); } pass2 { u8 *ajmp; getReg5(xmmFt, _Ft_, _Ftf_); @@ -98,9 +96,8 @@ microVUf(void) mVU_SQRT(mF) { pass3 { mVUlog("SQRT Q, vf%02d%s", _Ft_, _Ftf_String); } } -microVUf(void) mVU_RSQRT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFDIV(_Fs_, _Fsf_, _Ft_, _Ftf_, 13); } +mVUop(mVU_RSQRT) { + pass1 { mVUanalyzeFDIV(mVU, _Fs_, _Fsf_, _Ft_, _Ftf_, 13); } pass2 { u8 *ajmp, *bjmp, *cjmp, *djmp; getReg5(xmmFs, _Fs_, _Fsf_); @@ -127,7 +124,7 @@ microVUf(void) mVU_RSQRT(mF) { djmp = JMP8(0); x86SetJ8(ajmp); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); + if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); @@ -149,8 +146,7 @@ microVUf(void) mVU_RSQRT(mF) { SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFt); \ } -microVUt(void) mVU_EATAN_() { - microVU* mVU = mVUx; +microVUt(void) mVU_EATAN_(mV) { // ToDo: Can Be Optimized Further? (takes approximately (~115 cycles + mem access time) on a c2d) SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); @@ -169,9 +165,8 @@ microVUt(void) mVU_EATAN_() { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); } -microVUf(void) mVU_EATAN(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 54); } +mVUop(mVU_EATAN) { + pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 54); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -181,14 +176,13 @@ microVUf(void) mVU_EATAN(mF) { SSE_ADDSS_M32_to_XMM(xmmPQ, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); - mVU_EATAN_(); + mVU_EATAN_(mVU); } pass3 { mVUlog("EATAN P"); } } -microVUf(void) mVU_EATANxy(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 54); } +mVUop(mVU_EATANxy) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 54); } pass2 { getReg6(xmmFt, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFt, 0x01); @@ -199,14 +193,13 @@ microVUf(void) mVU_EATANxy(mF) { SSE_ADDSS_XMM_to_XMM(xmmFt, xmmPQ); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - mVU_EATAN_(); + mVU_EATAN_(mVU); } pass3 { mVUlog("EATANxy P"); } } -microVUf(void) mVU_EATANxz(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 54); } +mVUop(mVU_EATANxz) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 54); } pass2 { getReg6(xmmFt, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFt, 0x02); @@ -217,7 +210,7 @@ microVUf(void) mVU_EATANxz(mF) { SSE_ADDSS_XMM_to_XMM(xmmFt, xmmPQ); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmFt); - mVU_EATAN_(); + mVU_EATAN_(mVU); } pass3 { mVUlog("EATANxz P"); } } @@ -229,9 +222,8 @@ microVUf(void) mVU_EATANxz(mF) { SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFt); \ } -microVUf(void) mVU_EEXP(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 44); } +mVUop(mVU_EEXP) { + pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 44); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -278,22 +270,20 @@ microVUt(void) mVU_sumXYZ() { } } -microVUf(void) mVU_ELENG(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 18); } +mVUop(mVU_ELENG) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 18); } pass2 { getReg6(xmmFs, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance - mVU_sumXYZ(); + mVU_sumXYZ(); SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmPQ); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ELENG P"); } } -microVUf(void) mVU_ERCPR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 12); } +mVUop(mVU_ERCPR) { + pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 12); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -306,13 +296,12 @@ microVUf(void) mVU_ERCPR(mF) { pass3 { mVUlog("ERCPR P"); } } -microVUf(void) mVU_ERLENG(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 24); } +mVUop(mVU_ERLENG) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 24); } pass2 { getReg6(xmmFs, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance - mVU_sumXYZ(); + mVU_sumXYZ(); SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmPQ); SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); @@ -322,13 +311,12 @@ microVUf(void) mVU_ERLENG(mF) { pass3 { mVUlog("ERLENG P"); } } -microVUf(void) mVU_ERSADD(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 18); } +mVUop(mVU_ERSADD) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 18); } pass2 { getReg6(xmmFs, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance - mVU_sumXYZ(); + mVU_sumXYZ(); //SSE_RCPSS_XMM_to_XMM(xmmPQ, xmmPQ); // Lower Precision is bad? SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); @@ -338,9 +326,8 @@ microVUf(void) mVU_ERSADD(mF) { pass3 { mVUlog("ERSADD P"); } } -microVUf(void) mVU_ERSQRT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 18); } +mVUop(mVU_ERSQRT) { + pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 18); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -353,13 +340,12 @@ microVUf(void) mVU_ERSQRT(mF) { pass3 { mVUlog("ERSQRT P"); } } -microVUf(void) mVU_ESADD(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 11); } +mVUop(mVU_ESADD) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 11); } pass2 { getReg6(xmmFs, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance - mVU_sumXYZ(); + mVU_sumXYZ(); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ESADD P"); } @@ -372,9 +358,8 @@ microVUf(void) mVU_ESADD(mF) { SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmFs); \ } -microVUf(void) mVU_ESIN(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 29); } +mVUop(mVU_ESIN) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 29); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -399,9 +384,8 @@ microVUf(void) mVU_ESIN(mF) { pass3 { mVUlog("ESIN P"); } } -microVUf(void) mVU_ESQRT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU1(_Fs_, _Fsf_, 12); } +mVUop(mVU_ESQRT) { + pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 12); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -411,9 +395,8 @@ microVUf(void) mVU_ESQRT(mF) { pass3 { mVUlog("ESQRT P"); } } -microVUf(void) mVU_ESUM(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeEFU2(_Fs_, 12); } +mVUop(mVU_ESUM) { + pass1 { mVUanalyzeEFU2(mVU, _Fs_, 12); } pass2 { getReg6(xmmFs, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance @@ -431,66 +414,61 @@ microVUf(void) mVU_ESUM(mF) { // FCAND/FCEQ/FCGET/FCOR/FCSET //------------------------------------------------------------------ -microVUf(void) mVU_FCAND(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(1); } +mVUop(mVU_FCAND) { + pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, fvcInstance); AND32ItoR(gprT1, _Imm24_); ADD32ItoR(gprT1, 0xffffff); SHR32ItoR(gprT1, 24); - mVUallocVIb(gprT1, 1); + mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCAND vi01, $%x", _Imm24_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCEQ(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(1); } +mVUop(mVU_FCEQ) { + pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, fvcInstance); XOR32ItoR(gprT1, _Imm24_); SUB32ItoR(gprT1, 1); SHR32ItoR(gprT1, 31); - mVUallocVIb(gprT1, 1); + mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCEQ vi01, $%x", _Imm24_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCGET(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(_It_); } +mVUop(mVU_FCGET) { + pass1 { mVUanalyzeCflag(mVU, _It_); } pass2 { - mVUallocCFLAGa(gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, fvcInstance); AND32ItoR(gprT1, 0xfff); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FCGET vi%02d", _Ft_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCOR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeCflag(1); } +mVUop(mVU_FCOR) { + pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, fvcInstance); OR32ItoR(gprT1, _Imm24_); ADD32ItoR(gprT1, 1); // If 24 1's will make 25th bit 1, else 0 SHR32ItoR(gprT1, 24); // Get the 25th bit (also clears the rest of the garbage in the reg) - mVUallocVIb(gprT1, 1); + mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCOR vi01, $%x", _Imm24_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } } -microVUf(void) mVU_FCSET(mF) { - microVU* mVU = mVUx; +mVUop(mVU_FCSET) { pass1 { mVUinfo |= _doClip; } pass2 { MOV32ItoR(gprT1, _Imm24_); - mVUallocCFLAGb(gprT1, fcInstance); + mVUallocCFLAGb(mVU, gprT1, fcInstance); } pass3 { mVUlog("FCSET $%x", _Imm24_); } } @@ -499,42 +477,39 @@ microVUf(void) mVU_FCSET(mF) { // FMAND/FMEQ/FMOR //------------------------------------------------------------------ -microVUf(void) mVU_FMAND(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Is_, _It_); } +mVUop(mVU_FMAND) { + pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Is_); + mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocVIa(mVU, gprT2, _Is_); AND16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMAND vi%02d, vi%02d", _Ft_, _Fs_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } -microVUf(void) mVU_FMEQ(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Is_, _It_); } +mVUop(mVU_FMEQ) { + pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Is_); + mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocVIa(mVU, gprT2, _Is_); XOR32RtoR(gprT1, gprT2); SUB32ItoR(gprT1, 1); SHR32ItoR(gprT1, 31); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMEQ vi%02d, vi%02d", _Ft_, _Fs_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } } -microVUf(void) mVU_FMOR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMflag(_Is_, _It_); } +mVUop(mVU_FMOR) { + pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(gprT1, fvmInstance); - mVUallocVIa(gprT2, _Is_); + mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocVIa(mVU, gprT2, _Is_); OR16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMOR vi%02d, vi%02d", _Ft_, _Fs_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } @@ -544,47 +519,43 @@ microVUf(void) mVU_FMOR(mF) { // FSAND/FSEQ/FSOR/FSSET //------------------------------------------------------------------ -microVUf(void) mVU_FSAND(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_It_); } +mVUop(mVU_FSAND) { + pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, fvsInstance); AND16ItoR(gprT1, _Imm12_); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSEQ(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_It_); } +mVUop(mVU_FSEQ) { + pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, fvsInstance); XOR16ItoR(gprT1, _Imm12_); SUB16ItoR(gprT1, 1); SHR16ItoR(gprT1, 15); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSOR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSflag(_It_); } +mVUop(mVU_FSOR) { + pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, fvsInstance); OR16ItoR(gprT1, _Imm12_); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -microVUf(void) mVU_FSSET(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFSSET(); } +mVUop(mVU_FSSET) { + pass1 { mVUanalyzeFSSET(mVU); } pass2 { int flagReg1, flagReg2; getFlagReg(flagReg1, fsInstance); @@ -600,97 +571,90 @@ microVUf(void) mVU_FSSET(mF) { // IADD/IADDI/IADDIU/IAND/IOR/ISUB/ISUBIU //------------------------------------------------------------------ -microVUf(void) mVU_IADD(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } +mVUop(mVU_IADD) { + pass1 { mVUanalyzeIALU1(mVU, _Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); if (_It_ != _Is_) { - mVUallocVIa(gprT2, _It_); + mVUallocVIa(mVU, gprT2, _It_); ADD16RtoR(gprT1, gprT2); } else ADD16RtoR(gprT1, gprT1); - mVUallocVIb(gprT1, _Id_); + mVUallocVIb(mVU, gprT1, _Id_); } pass3 { mVUlog("IADD vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_IADDI(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Is_, _It_); } +mVUop(mVU_IADDI) { + pass1 { mVUanalyzeIALU2(mVU, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); ADD16ItoR(gprT1, _Imm5_); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("IADDI vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm5_); } } -microVUf(void) mVU_IADDIU(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Is_, _It_); } +mVUop(mVU_IADDIU) { + pass1 { mVUanalyzeIALU2(mVU, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); ADD16ItoR(gprT1, _Imm15_); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("IADDIU vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm15_); } } -microVUf(void) mVU_IAND(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } +mVUop(mVU_IAND) { + pass1 { mVUanalyzeIALU1(mVU, _Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); if (_It_ != _Is_) { - mVUallocVIa(gprT2, _It_); + mVUallocVIa(mVU, gprT2, _It_); AND32RtoR(gprT1, gprT2); } - mVUallocVIb(gprT1, _Id_); + mVUallocVIb(mVU, gprT1, _Id_); } pass3 { mVUlog("IAND vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_IOR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } +mVUop(mVU_IOR) { + pass1 { mVUanalyzeIALU1(mVU, _Id_, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); if (_It_ != _Is_) { - mVUallocVIa(gprT2, _It_); + mVUallocVIa(mVU, gprT2, _It_); OR32RtoR(gprT1, gprT2); } - mVUallocVIb(gprT1, _Id_); + mVUallocVIb(mVU, gprT1, _Id_); } pass3 { mVUlog("IOR vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_ISUB(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU1(_Id_, _Is_, _It_); } +mVUop(mVU_ISUB) { + pass1 { mVUanalyzeIALU1(mVU, _Id_, _Is_, _It_); } pass2 { if (_It_ != _Is_) { - mVUallocVIa(gprT1, _Is_); - mVUallocVIa(gprT2, _It_); + mVUallocVIa(mVU, gprT1, _Is_); + mVUallocVIa(mVU, gprT2, _It_); SUB16RtoR(gprT1, gprT2); - mVUallocVIb(gprT1, _Id_); + mVUallocVIb(mVU, gprT1, _Id_); } else if (!isMMX(_Id_)) { XOR32RtoR(gprT1, gprT1); - mVUallocVIb(gprT1, _Id_); + mVUallocVIb(mVU, gprT1, _Id_); } else { PXORRtoR(mmVI(_Id_), mmVI(_Id_)); } } pass3 { mVUlog("ISUB vi%02d, vi%02d, vi%02d", _Fd_, _Fs_, _Ft_); } } -microVUf(void) mVU_ISUBIU(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeIALU2(_Is_, _It_); } +mVUop(mVU_ISUBIU) { + pass1 { mVUanalyzeIALU2(mVU, _Is_, _It_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); SUB16ItoR(gprT1, _Imm15_); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("ISUBIU vi%02d, vi%02d, %d", _Ft_, _Fs_, _Imm15_); } } @@ -699,56 +663,51 @@ microVUf(void) mVU_ISUBIU(mF) { // MFIR/MFP/MOVE/MR32/MTIR //------------------------------------------------------------------ -microVUf(void) mVU_MFIR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_MFIR) { pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); MOVSX32R16toR(gprT1, gprT1); SSE2_MOVD_R_to_XMM(xmmT1, gprT1); - if (!_XYZW_SS) { mVUunpack_xyzw(xmmT1, xmmT1, 0); } - mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + if (!_XYZW_SS) { mVUunpack_xyzw(xmmT1, xmmT1, 0); } + mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } pass3 { mVUlog("MFIR.%s vf%02d, vi%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MFP(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMFP(_Ft_); } +mVUop(mVU_MFP) { + pass1 { mVUanalyzeMFP(mVU, _Ft_); } pass2 { getPreg(xmmFt); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } pass3 { mVUlog("MFP.%s vf%02d, P", _XYZW_String, _Ft_); } } -microVUf(void) mVU_MOVE(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMOVE(_Fs_, _Ft_); } +mVUop(mVU_MOVE) { + pass1 { mVUanalyzeMOVE(mVU, _Fs_, _Ft_); } pass2 { - mVUloadReg(xmmT1, (uptr)&mVU->regs->VF[_Fs_].UL[0], _X_Y_Z_W); - mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUloadReg(xmmT1, (uptr)&mVU->regs->VF[_Fs_].UL[0], _X_Y_Z_W); + mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } pass3 { mVUlog("MOVE.%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MR32(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeMR32(_Fs_, _Ft_); } +mVUop(mVU_MR32) { + pass1 { mVUanalyzeMR32(mVU, _Fs_, _Ft_); } pass2 { - mVUloadReg(xmmT1, (uptr)&mVU->regs->VF[_Fs_].UL[0], (_X_Y_Z_W == 8) ? 4 : 15); + mVUloadReg(xmmT1, (uptr)&mVU->regs->VF[_Fs_].UL[0], (_X_Y_Z_W == 8) ? 4 : 15); if (_X_Y_Z_W != 8) { SSE2_PSHUFD_XMM_to_XMM(xmmT1, xmmT1, 0x39); } - mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 0); + mVUsaveReg(xmmT1, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 0); } pass3 { mVUlog("MR32.%s vf%02d, vf%02d", _XYZW_String, _Ft_, _Fs_); } } -microVUf(void) mVU_MTIR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_MTIR) { pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->VF[_Fs_].UL[_Fsf_]); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("MTIR vi%02d, vf%02d%s", _Ft_, _Fs_, _Fsf_String); } } @@ -757,38 +716,36 @@ microVUf(void) mVU_MTIR(mF) { // ILW/ILWR //------------------------------------------------------------------ -microVUf(void) mVU_ILW(mF) { - microVU* mVU = mVUx; +mVUop(mVU_ILW) { pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + getVUmem(_Imm11_) + offsetSS); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } else { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); ADD32ItoR(gprT1, _Imm11_); - mVUaddrFix(gprT1); + mVUaddrFix(mVU, gprT1); MOVZX32Rm16toR(gprT1, gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } } pass3 { mVUlog("ILW.%s vi%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_ILWR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_ILWR) { pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } else { - mVUallocVIa(gprT1, _Is_); - mVUaddrFix(gprT1); + mVUallocVIa(mVU, gprT1, _Is_); + mVUaddrFix(mVU, gprT1); MOVZX32Rm16toR(gprT1, gprT1, (uptr)mVU->regs->Mem + offsetSS); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } } pass3 { mVUlog("ILWR.%s vi%02d, vi%02d", _XYZW_String, _Ft_, _Fs_); } @@ -798,23 +755,22 @@ microVUf(void) mVU_ILWR(mF) { // ISW/ISWR //------------------------------------------------------------------ -microVUf(void) mVU_ISW(mF) { - microVU* mVU = mVUx; +mVUop(mVU_ISW) { pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { if (!_Is_) { int imm = getVUmem(_Imm11_); - mVUallocVIa(gprT1, _It_); + mVUallocVIa(mVU, gprT1, _It_); if (_X) MOV32RtoM((uptr)mVU->regs->Mem + imm, gprT1); if (_Y) MOV32RtoM((uptr)mVU->regs->Mem + imm + 4, gprT1); if (_Z) MOV32RtoM((uptr)mVU->regs->Mem + imm + 8, gprT1); if (_W) MOV32RtoM((uptr)mVU->regs->Mem + imm + 12, gprT1); } else { - mVUallocVIa(gprT1, _Is_); - mVUallocVIa(gprT2, _It_); + mVUallocVIa(mVU, gprT1, _Is_); + mVUallocVIa(mVU, gprT2, _It_); ADD32ItoR(gprT1, _Imm11_); - mVUaddrFix(gprT1); + mVUaddrFix(mVU, gprT1); if (_X) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem); if (_Y) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem+4); if (_Z) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem+8); @@ -824,21 +780,20 @@ microVUf(void) mVU_ISW(mF) { pass3 { mVUlog("ISW.%s vi%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_ISWR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_ISWR) { pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } pass2 { if (!_Is_) { - mVUallocVIa(gprT1, _It_); + mVUallocVIa(mVU, gprT1, _It_); if (_X) MOV32RtoM((uptr)mVU->regs->Mem, gprT1); if (_Y) MOV32RtoM((uptr)mVU->regs->Mem+4, gprT1); if (_Z) MOV32RtoM((uptr)mVU->regs->Mem+8, gprT1); if (_W) MOV32RtoM((uptr)mVU->regs->Mem+12, gprT1); } else { - mVUallocVIa(gprT1, _Is_); - mVUallocVIa(gprT2, _It_); - mVUaddrFix(gprT1); + mVUallocVIa(mVU, gprT1, _Is_); + mVUallocVIa(mVU, gprT2, _It_); + mVUaddrFix(mVU, gprT1); if (_X) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem); if (_Y) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem+4); if (_Z) MOV32RtoRm(gprT1, gprT2, (uptr)mVU->regs->Mem+8); @@ -852,65 +807,62 @@ microVUf(void) mVU_ISWR(mF) { // LQ/LQD/LQI //------------------------------------------------------------------ -microVUf(void) mVU_LQ(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Is_, 0); } +mVUop(mVU_LQ) { + pass1 { mVUanalyzeLQ(mVU, _Ft_, _Is_, 0); } pass2 { if (!_Is_) { - mVUloadReg(xmmFt, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUloadReg(xmmFt, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); ADD32ItoR(gprT1, _Imm11_); - mVUaddrFix(gprT1); - mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUaddrFix(mVU, gprT1); + mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } } pass3 { mVUlog("LQ.%s vf%02d, vi%02d + %d", _XYZW_String, _Ft_, _Fs_, _Imm11_); } } -microVUf(void) mVU_LQD(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } +mVUop(mVU_LQD) { + pass1 { mVUanalyzeLQ(mVU, _Ft_, _Is_, 1); } pass2 { if (!_Is_ && !noWriteVF) { - mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); SUB16ItoR(gprT1, 1); - mVUallocVIb(gprT1, _Is_); + mVUallocVIb(mVU, gprT1, _Is_); if (!noWriteVF) { - mVUaddrFix(gprT1); - mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUaddrFix(mVU, gprT1); + mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } } } pass3 { mVUlog("LQD.%s vf%02d, --vi%02d", _XYZW_String, _Ft_, _Is_); } } -microVUf(void) mVU_LQI(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeLQ(_Ft_, _Is_, 1); } +mVUop(mVU_LQI) { + pass1 { mVUanalyzeLQ(mVU, _Ft_, _Is_, 1); } pass2 { if (!_Is_ && !noWriteVF) { - mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa((!noWriteVF) ? gprT1 : gprT2, _Is_); + mVUallocVIa(mVU, (!noWriteVF) ? gprT1 : gprT2, _Is_); if (!noWriteVF) { MOV32RtoR(gprT2, gprT1); - mVUaddrFix(gprT1); - mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); - mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); + mVUaddrFix(mVU, gprT1); + mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } ADD16ItoR(gprT2, 1); - mVUallocVIb(gprT2, _Is_); + mVUallocVIb(mVU, gprT2, _Is_); } } pass3 { mVUlog("LQI.%s vf%02d, vi%02d++", _XYZW_String, _Ft_, _Fs_); } @@ -920,61 +872,58 @@ microVUf(void) mVU_LQI(mF) { // SQ/SQD/SQI //------------------------------------------------------------------ -microVUf(void) mVU_SQ(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _It_, 0); } +mVUop(mVU_SQ) { + pass1 { mVUanalyzeSQ(mVU, _Fs_, _It_, 0); } pass2 { if (!_It_) { getReg7(xmmFs, _Fs_); - mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W, 1); + mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem + getVUmem(_Imm11_), _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _It_); + mVUallocVIa(mVU, gprT1, _It_); ADD32ItoR(gprT1, _Imm11_); - mVUaddrFix(gprT1); + mVUaddrFix(mVU, gprT1); getReg7(xmmFs, _Fs_); - mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); } } pass3 { mVUlog("SQ.%s vf%02d, vi%02d + %d", _XYZW_String, _Fs_, _Ft_, _Imm11_); } } -microVUf(void) mVU_SQD(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } +mVUop(mVU_SQD) { + pass1 { mVUanalyzeSQ(mVU, _Fs_, _It_, 1); } pass2 { if (!_It_) { getReg7(xmmFs, _Fs_); - mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); + mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _It_); + mVUallocVIa(mVU, gprT1, _It_); SUB16ItoR(gprT1, 1); - mVUallocVIb(gprT1, _It_); - mVUaddrFix(gprT1); + mVUallocVIb(mVU, gprT1, _It_); + mVUaddrFix(mVU, gprT1); getReg7(xmmFs, _Fs_); - mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); } } pass3 { mVUlog("SQD.%s vf%02d, --vi%02d", _XYZW_String, _Fs_, _Ft_); } } -microVUf(void) mVU_SQI(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeSQ(_Fs_, _It_, 1); } +mVUop(mVU_SQI) { + pass1 { mVUanalyzeSQ(mVU, _Fs_, _It_, 1); } pass2 { if (!_It_) { getReg7(xmmFs, _Fs_); - mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); + mVUsaveReg(xmmFs, (uptr)mVU->regs->Mem, _X_Y_Z_W, 1); } else { - mVUallocVIa(gprT1, _It_); + mVUallocVIa(mVU, gprT1, _It_); MOV32RtoR(gprT2, gprT1); - mVUaddrFix(gprT1); + mVUaddrFix(mVU, gprT1); getReg7(xmmFs, _Fs_); - mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); + mVUsaveReg2(xmmFs, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); ADD16ItoR(gprT2, 1); - mVUallocVIb(gprT2, _It_); + mVUallocVIb(mVU, gprT2, _It_); } } pass3 { mVUlog("SQI.%s vf%02d, vi%02d++", _XYZW_String, _Fs_, _Ft_); } @@ -984,9 +933,8 @@ microVUf(void) mVU_SQI(mF) { // RINIT/RGET/RNEXT/RXOR //------------------------------------------------------------------ -microVUf(void) mVU_RINIT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeR1(_Fs_, _Fsf_); } +mVUop(mVU_RINIT) { + pass1 { mVUanalyzeR1(mVU, _Fs_, _Fsf_); } pass2 { if (_Fs_ || (_Fsf_ == 3)) { getReg8(gprT1, _Fs_, _Fsf_); @@ -999,8 +947,7 @@ microVUf(void) mVU_RINIT(mF) { pass3 { mVUlog("RINIT R, vf%02d%s", _Fs_, _Fsf_String); } } -microVUt(void) mVU_RGET_(int Rreg) { - microVU* mVU = mVUx; +microVUt(void) mVU_RGET_(mV, int Rreg) { if (!noWriteVF) { if (_X) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[0], Rreg); if (_Y) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[1], Rreg); @@ -1009,16 +956,14 @@ microVUt(void) mVU_RGET_(int Rreg) { } } -microVUf(void) mVU_RGET(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeR2(_Ft_, 1); } - pass2 { MOV32MtoR(gprT1, Rmem); mVU_RGET_(gprT1); } +mVUop(mVU_RGET) { + pass1 { mVUanalyzeR2(mVU, _Ft_, 1); } + pass2 { MOV32MtoR(gprT1, Rmem); mVU_RGET_(mVU, gprT1); } pass3 { mVUlog("RGET.%s vf%02d, R", _XYZW_String, _Ft_); } } -microVUf(void) mVU_RNEXT(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeR2(_Ft_, 0); } +mVUop(mVU_RNEXT) { + pass1 { mVUanalyzeR2(mVU, _Ft_, 0); } pass2 { // algorithm from www.project-fao.org MOV32MtoR(gprR, Rmem); @@ -1035,15 +980,14 @@ microVUf(void) mVU_RNEXT(mF) { XOR32RtoR(gprR, gprT1); AND32ItoR(gprR, 0x007fffff); OR32ItoR (gprR, 0x3f800000); - mVU_RGET_(gprR); + mVU_RGET_(mVU, gprR); MOV32ItoR(gprR, Roffset); // Restore gprR } pass3 { mVUlog("RNEXT.%s vf%02d, R", _XYZW_String, _Ft_); } } -microVUf(void) mVU_RXOR(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeR1(_Fs_, _Fsf_); } +mVUop(mVU_RXOR) { + pass1 { mVUanalyzeR1(mVU, _Fs_, _Fsf_); } pass2 { if (_Fs_ || (_Fsf_ == 3)) { getReg8(gprT1, _Fs_, _Fsf_); @@ -1058,14 +1002,12 @@ microVUf(void) mVU_RXOR(mF) { // WaitP/WaitQ //------------------------------------------------------------------ -microVUf(void) mVU_WAITP(mF) { - microVU* mVU = mVUx; +mVUop(mVU_WAITP) { pass1 { mVUstall = aMax(mVUstall, ((mVUregs.p) ? (mVUregs.p - 1) : 0)); } pass3 { mVUlog("WAITP"); } } -microVUf(void) mVU_WAITQ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_WAITQ) { pass1 { mVUstall = aMax(mVUstall, mVUregs.q); } pass3 { mVUlog("WAITQ"); } } @@ -1074,22 +1016,20 @@ microVUf(void) mVU_WAITQ(mF) { // XTOP/XITOP //------------------------------------------------------------------ -microVUf(void) mVU_XTOP(mF) { - microVU* mVU = mVUx; +mVUop(mVU_XTOP) { pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->top); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("XTOP vi%02d", _Ft_); } } -microVUf(void) mVU_XITOP(mF) { - microVU* mVU = mVUx; +mVUop(mVU_XITOP) { pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->itop); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("XITOP vi%02d", _Ft_); } } @@ -1103,9 +1043,7 @@ void __fastcall mVU_XGKICK_(u32 addr) { u32 *data = (u32*)(microVU1.regs->Mem + addr); u32 size = mtgsThread->PrepDataPacket(GIF_PATH_1, data, (0x4000-addr) >> 4); u8 *pDest = mtgsThread->GetDataPacketPtr(); -/* - if((size << 4) > (0x4000-(addr&0x3fff))) - { +/* if((size << 4) > (0x4000-(addr&0x3fff))) { //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params (addr&0x3fff) + (size << 4), (0x4000-(addr&0x3fff)) >> 4, size - ((0x4000-(addr&0x3fff)) >> 4)); memcpy_aligned(pDest, microVU1.regs->Mem + addr, 0x4000-(addr&0x3fff)); size -= (0x4000-(addr&0x3fff)) >> 4; @@ -1113,12 +1051,9 @@ void __fastcall mVU_XGKICK_(u32 addr) { pDest += 0x4000-(addr&0x3fff); memcpy_aligned(pDest, microVU1.regs->Mem, size<<4); } - else - { -*/ + else { */ memcpy_aligned(pDest, microVU1.regs->Mem + addr, size<<4); // } - mtgsThread->SendDataPacket(); } @@ -1126,24 +1061,22 @@ void __fastcall mVU_XGKICK__(u32 addr) { GSGIFTRANSFER1((u32*)microVU1.regs->Mem, ((addr<<4)&0x3fff)); } -microVUf(void) mVU_XGKICK(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeXGkick(_Is_, mVU_XGKICK_CYCLES); } +mVUop(mVU_XGKICK) { + pass1 { mVUanalyzeXGkick(mVU, _Is_, mVU_XGKICK_CYCLES); } pass2 { mVUprint("XGkick"); - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->VIxgkick, gprT1); } pass3 { mVUlog("XGKICK vi%02d", _Fs_); } } -microVUt(void) mVU_XGKICK_DELAY() { - microVU* mVU = mVUx; +microVUt(void) mVU_XGKICK_DELAY(mV) { MOV32MtoR(gprT2, (uptr)&mVU->VIxgkick); // gprT2 = ECX for __fastcall - mVUbackupRegs(); + mVUbackupRegs(mVU); if (mtgsThread != NULL) CALLFunc((uptr)mVU_XGKICK_); else CALLFunc((uptr)mVU_XGKICK__); - mVUrestoreRegs(); + mVUrestoreRegs(mVU); } //------------------------------------------------------------------ @@ -1157,123 +1090,113 @@ microVUt(void) mVU_XGKICK_DELAY() { pass4 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ } -microVUf(void) mVU_B(mF) { - microVU* mVU = mVUx; +mVUop(mVU_B) { setBranchA(1, 0); pass3 { mVUlog("B [%04x]", branchAddr, branchAddr); } } -microVUf(void) mVU_BAL(mF) { - microVU* mVU = mVUx; +mVUop(mVU_BAL) { setBranchA(2, _It_); pass1 { analyzeVIreg2(_It_, 1); } pass2 { MOV32ItoR(gprT1, bSaveAddr); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("BAL vi%02d [%04x]", _Ft_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBEQ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBEQ) { setBranchA(3, 0); - pass1 { mVUanalyzeBranch2(_Is_, _It_); } + pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else { mVUallocVIa(gprT2, _It_); XOR32RtoR(gprT1, gprT2); } + else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBEQ vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBGEZ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBGEZ) { setBranchA(4, 0); - pass1 { mVUanalyzeBranch1(_Is_); } + pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBGEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBGTZ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBGTZ) { setBranchA(5, 0); - pass1 { mVUanalyzeBranch1(_Is_); } + pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBGTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBLEZ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBLEZ) { setBranchA(6, 0); - pass1 { mVUanalyzeBranch1(_Is_); } + pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBLEZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBLTZ(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBLTZ) { setBranchA(7, 0); - pass1 { mVUanalyzeBranch1(_Is_); } + pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBLTZ vi%02d [%04x]", _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_IBNE(mF) { - microVU* mVU = mVUx; +mVUop(mVU_IBNE) { setBranchA(8, 0); - pass1 { mVUanalyzeBranch2(_Is_, _It_); } + pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else mVUallocVIa(gprT1, _Is_); + else mVUallocVIa(mVU, gprT1, _Is_); if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); - else { mVUallocVIa(gprT2, _It_); XOR32RtoR(gprT1, gprT2); } + else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("IBNE vi%02d, vi%02d [%04x]", _Ft_, _Fs_, branchAddr, branchAddr); } } -microVUf(void) mVU_JR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_JR) { mVUbranch = 9; pass1 { analyzeVIreg1(_Is_); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); SHL32ItoR(gprT1, 3); - AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); + AND32ItoR(gprT1, isVU1 ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); } pass3 { mVUlog("JR [vi%02d]", _Fs_); } } -microVUf(void) mVU_JALR(mF) { - microVU* mVU = mVUx; +mVUop(mVU_JALR) { mVUbranch = 10; pass1 { analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 1); } pass2 { - mVUallocVIa(gprT1, _Is_); + mVUallocVIa(mVU, gprT1, _Is_); SHL32ItoR(gprT1, 3); - AND32ItoR(gprT1, vuIndex ? 0x3ff8 : 0xff8); + AND32ItoR(gprT1, isVU1 ? 0x3ff8 : 0xff8); MOV32RtoM((uptr)&mVU->branch, gprT1); MOV32ItoR(gprT1, bSaveAddr); - mVUallocVIb(gprT1, _It_); + mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("JALR vi%02d, [vi%02d]", _Ft_, _Fs_); } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 695c9ac694..00b0530c6c 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -105,7 +105,9 @@ declareAllVariables #define _Tbit_ (1<<27) #define _MDTbit_ 0 //( _Mbit_ | _Dbit_ | _Tbit_ ) // ToDo: Implement this stuff... -#define getVUmem(x) (((vuIndex == 1) ? (x & 0x3ff) : ((x >= 0x400) ? (x & 0x43f) : (x & 0xff))) * 16) +#define isVU1 (mVU == µVU1) +#define getIndex (isVU1 ? 1 : 0) +#define getVUmem(x) (((isVU1) ? (x & 0x3ff) : ((x >= 0x400) ? (x & 0x43f) : (x & 0xff))) * 16) #define offsetSS ((_X) ? (0) : ((_Y) ? (4) : ((_Z) ? 8: 12))) #define offsetReg ((_X) ? (0) : ((_Y) ? (1) : ((_Z) ? 2: 3))) @@ -136,15 +138,19 @@ declareAllVariables #define gprF2 6 // Status Flag 2 #define gprF3 7 // Status Flag 3 -// Template Stuff -#define mVUx (vuIndex ? µVU1 : µVU0) -#define microVUt(aType) template __forceinline aType -#define microVUx(aType) template aType -#define microVUf(aType) template aType -#define microVUq(aType) template __forceinline aType - // Function Params +#define mP microVU* mVU, int recPass +#define mV microVU* mVU #define mF int recPass +#define mX mVU, recPass + +// Function/Template Stuff +#define mVUx (vuIndex ? µVU1 : µVU0) +#define mVUop(opName) void opName (mP) +#define microVUt(aType) __forceinline aType +#define microVUx(aType) template aType +#define microVUf(aType) template __forceinline aType + // Define Passes #define pass1 if (recPass == 0) @@ -176,9 +182,9 @@ declareAllVariables #define setCode() { mVU->code = curI; } #define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } #define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } -#define incCycles(x) { mVUincCycles(x); } -#define bSaveAddr (((xPC + (2 * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) / 8) -#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((vuIndex) ? 0x3ff8:0xff8)) +#define incCycles(x) { mVUincCycles(mVU, x); } +#define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) +#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((isVU1) ? 0x3ff8 : 0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL @@ -288,7 +294,7 @@ declareAllVariables // Program Logging... #ifdef mVUlogProg -#define mVUlog __mVULog +#define mVUlog ((isVU1) ? __mVULog<1> : __mVULog<0>) #define mVUdumpProg __mVUdumpProgram #else #define mVUlog 0&& @@ -309,7 +315,8 @@ declareAllVariables uptr diff = ptr - start; \ if (diff >= limit) { \ Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); \ - mVUreset(); \ + if (!isVU1) mVUreset<0>(); \ + else mVUreset<1>(); \ } \ } diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 8584169a41..4d3ddc8fba 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -23,7 +23,7 @@ //------------------------------------------------------------------ // Used for Result Clamping -microVUx(void) mVUclamp1(int reg, int regT1, int xyzw) { +void mVUclamp1(int reg, int regT1, int xyzw) { switch (xyzw) { case 1: case 2: case 4: case 8: SSE_MINSS_XMM_to_XMM(reg, xmmMax); @@ -37,7 +37,7 @@ microVUx(void) mVUclamp1(int reg, int regT1, int xyzw) { } // Used for Operand Clamping -microVUx(void) mVUclamp2(int reg, int regT1, int xyzw) { +void mVUclamp2(int reg, int regT1, int xyzw) { if (CHECK_VU_SIGN_OVERFLOW) { switch (xyzw) { case 1: case 2: case 4: case 8: @@ -56,14 +56,14 @@ microVUx(void) mVUclamp2(int reg, int regT1, int xyzw) { break; } } - else mVUclamp1(reg, regT1, xyzw); + else mVUclamp1(reg, regT1, xyzw); } //------------------------------------------------------------------ // Micro VU - Misc Functions //------------------------------------------------------------------ -microVUx(void) mVUunpack_xyzw(int dstreg, int srcreg, int xyzw) { +void mVUunpack_xyzw(int dstreg, int srcreg, int xyzw) { switch ( xyzw ) { case 0: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x00); break; case 1: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x55); break; @@ -72,7 +72,7 @@ microVUx(void) mVUunpack_xyzw(int dstreg, int srcreg, int xyzw) { } } -microVUx(void) mVUloadReg(int reg, uptr offset, int xyzw) { +void mVUloadReg(int reg, uptr offset, int xyzw) { switch( xyzw ) { case 8: SSE_MOVSS_M32_to_XMM(reg, offset); break; // X case 4: SSE_MOVSS_M32_to_XMM(reg, offset+4); break; // Y @@ -82,7 +82,7 @@ microVUx(void) mVUloadReg(int reg, uptr offset, int xyzw) { } } -microVUx(void) mVUloadReg2(int reg, int gprReg, uptr offset, int xyzw) { +void mVUloadReg2(int reg, int gprReg, uptr offset, int xyzw) { switch( xyzw ) { case 8: SSE_MOVSS_Rm_to_XMM(reg, gprReg, offset); break; // X case 4: SSE_MOVSS_Rm_to_XMM(reg, gprReg, offset+4); break; // Y @@ -93,7 +93,7 @@ microVUx(void) mVUloadReg2(int reg, int gprReg, uptr offset, int xyzw) { } // Modifies the Source Reg! -microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { +void mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { /*SSE_MOVAPS_M128_to_XMM(xmmT2, offset); if (modXYZW && (xyzw == 8 || xyzw == 4 || xyzw == 2 || xyzw == 1)) { mVUunpack_xyzw(reg, reg, 0); @@ -135,13 +135,13 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { SSE_MOVHLPS_XMM_to_XMM(reg, reg); SSE_MOVSS_XMM_to_M32(offset+8, reg); break; // XYZ - case 4: if (!modXYZW) mVUunpack_xyzw(reg, reg, 1); + case 4: if (!modXYZW) mVUunpack_xyzw(reg, reg, 1); SSE_MOVSS_XMM_to_M32(offset+4, reg); break; // Y - case 2: if (!modXYZW) mVUunpack_xyzw(reg, reg, 2); + case 2: if (!modXYZW) mVUunpack_xyzw(reg, reg, 2); SSE_MOVSS_XMM_to_M32(offset+8, reg); break; // Z - case 1: if (!modXYZW) mVUunpack_xyzw(reg, reg, 3); + case 1: if (!modXYZW) mVUunpack_xyzw(reg, reg, 3); SSE_MOVSS_XMM_to_M32(offset+12, reg); break; // W case 8: SSE_MOVSS_XMM_to_M32(offset, reg); break; // X @@ -152,7 +152,7 @@ microVUx(void) mVUsaveReg(int reg, uptr offset, int xyzw, bool modXYZW) { } // Modifies the Source Reg! -microVUx(void) mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { +void mVUsaveReg2(int reg, int gprReg, u32 offset, int xyzw) { /*SSE_MOVAPSRmtoR(xmmT2, gprReg, offset); if (xyzw == 8 || xyzw == 4 || xyzw == 2 || xyzw == 1) { mVUunpack_xyzw(reg, reg, 0); @@ -262,8 +262,8 @@ void mVUmergeRegs(int dest, int src, int xyzw) { } // Transforms the Address in gprReg to valid VU0/VU1 Address -microVUt(void) mVUaddrFix(int gprReg) { - if (vuIndex) { +microVUt(void) mVUaddrFix(mV, int gprReg) { + if (mVU == µVU1) { AND32ItoR(gprReg, 0x3ff); // wrap around SHL32ItoR(gprReg, 4); } @@ -281,15 +281,13 @@ microVUt(void) mVUaddrFix(int gprReg) { } // Backup Volatile Regs (EAX, ECX, EDX, MM0~7, XMM0~7, are all volatile according to 32bit Win/Linux ABI) -microVUt(void) mVUbackupRegs() { - microVU* mVU = mVUx; +microVUt(void) mVUbackupRegs(mV) { SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->ACC.UL[0], xmmACC); SSE_MOVAPS_XMM_to_M128((uptr)&mVU->xmmPQb[0], xmmPQ); } // Restore Volatile Regs -microVUt(void) mVUrestoreRegs() { - microVU* mVU = mVUx; +microVUt(void) mVUrestoreRegs(mV) { SSE_MOVAPS_M128_to_XMM(xmmACC, (uptr)&mVU->regs->ACC.UL[0]); SSE_MOVAPS_M128_to_XMM(xmmPQ, (uptr)&mVU->xmmPQb[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); @@ -298,15 +296,14 @@ microVUt(void) mVUrestoreRegs() { } // Reads entire microProgram and finds out if Status Flag is Used -microVUt(void) mVUcheckSflag(int progIndex) { +microVUt(void) mVUcheckSflag(mV, int progIndex) { if (CHECK_VU_FLAGHACK1) { - microVU* mVU = mVUx; int bFlagInfo = mVUflagInfo; int bCode = mVU->code; mVUsFlagHack = 1; for (u32 i = 0; i < mVU->progSize; i+=2) { mVU->code = mVU->prog.prog[progIndex].data[i]; - mVUopL(3); + mVUopL(mVU, 3); } mVUflagInfo = bFlagInfo; mVU->code = bCode; diff --git a/pcsx2/x86/microVU_Tables.inl b/pcsx2/x86/microVU_Tables.inl index 64d3ac682d..2316c116f9 100644 --- a/pcsx2/x86/microVU_Tables.inl +++ b/pcsx2/x86/microVU_Tables.inl @@ -21,182 +21,182 @@ //------------------------------------------------------------------ // Declarations //------------------------------------------------------------------ -#define mVUgetCode (vuIndex ? microVU1.code : microVU0.code) +#define mVUgetCode (mVU->code) -microVUf(void) mVU_UPPER_FD_00(mF); -microVUf(void) mVU_UPPER_FD_01(mF); -microVUf(void) mVU_UPPER_FD_10(mF); -microVUf(void) mVU_UPPER_FD_11(mF); -microVUf(void) mVULowerOP(mF); -microVUf(void) mVULowerOP_T3_00(mF); -microVUf(void) mVULowerOP_T3_01(mF); -microVUf(void) mVULowerOP_T3_10(mF); -microVUf(void) mVULowerOP_T3_11(mF); -microVUf(void) mVUunknown(mF); +mVUop(mVU_UPPER_FD_00); +mVUop(mVU_UPPER_FD_01); +mVUop(mVU_UPPER_FD_10); +mVUop(mVU_UPPER_FD_11); +mVUop(mVULowerOP); +mVUop(mVULowerOP_T3_00); +mVUop(mVULowerOP_T3_01); +mVUop(mVULowerOP_T3_10); +mVUop(mVULowerOP_T3_11); +mVUop(mVUunknown); //------------------------------------------------------------------ //------------------------------------------------------------------ // Opcode Tables //------------------------------------------------------------------ -#define microVU_LOWER_OPCODE(x) void (*mVULOWER_OPCODE##x [128])(mF) = { \ - mVU_LQ , mVU_SQ , mVUunknown , mVUunknown, \ - mVU_ILW , mVU_ISW , mVUunknown , mVUunknown, \ - mVU_IADDIU , mVU_ISUBIU , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_FCEQ , mVU_FCSET , mVU_FCAND , mVU_FCOR, /* 0x10 */ \ - mVU_FSEQ , mVU_FSSET , mVU_FSAND , mVU_FSOR, \ - mVU_FMEQ , mVUunknown , mVU_FMAND , mVU_FMOR, \ - mVU_FCGET , mVUunknown , mVUunknown , mVUunknown, \ - mVU_B , mVU_BAL , mVUunknown , mVUunknown, /* 0x20 */ \ - mVU_JR , mVU_JALR , mVUunknown , mVUunknown, \ - mVU_IBEQ , mVU_IBNE , mVUunknown , mVUunknown, \ - mVU_IBLTZ , mVU_IBGTZ , mVU_IBLEZ , mVU_IBGEZ, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVULowerOP , mVUunknown , mVUunknown , mVUunknown, /* 0x40*/ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x50 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x60 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x70 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_LOWER_OPCODE(x) void (*mVULOWER_OPCODE##x [128])(mP) = { \ + mVU_LQ , mVU_SQ , mVUunknown , mVUunknown, \ + mVU_ILW , mVU_ISW , mVUunknown , mVUunknown, \ + mVU_IADDIU , mVU_ISUBIU , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_FCEQ , mVU_FCSET , mVU_FCAND , mVU_FCOR, /* 0x10 */ \ + mVU_FSEQ , mVU_FSSET , mVU_FSAND , mVU_FSOR, \ + mVU_FMEQ , mVUunknown , mVU_FMAND , mVU_FMOR, \ + mVU_FCGET , mVUunknown , mVUunknown , mVUunknown, \ + mVU_B , mVU_BAL , mVUunknown , mVUunknown, /* 0x20 */ \ + mVU_JR , mVU_JALR , mVUunknown , mVUunknown, \ + mVU_IBEQ , mVU_IBNE , mVUunknown , mVUunknown, \ + mVU_IBLTZ , mVU_IBGTZ , mVU_IBLEZ , mVU_IBGEZ, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP , mVUunknown , mVUunknown , mVUunknown, /* 0x40*/ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x50 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x60 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x70 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_LowerOP_T3_00_OPCODE(x) void (*mVULowerOP_T3_00_OPCODE##x [32])(mF) = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_MOVE , mVU_LQI , mVU_DIV , mVU_MTIR, \ - mVU_RNEXT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_MFP , mVU_XTOP , mVU_XGKICK, \ - mVU_ESADD , mVU_EATANxy, mVU_ESQRT , mVU_ESIN, \ +#define microVU_LowerOP_T3_00_OPCODE(x) void (*mVULowerOP_T3_00_OPCODE##x [32])(mP) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MOVE , mVU_LQI , mVU_DIV , mVU_MTIR, \ + mVU_RNEXT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_MFP , mVU_XTOP , mVU_XGKICK, \ + mVU_ESADD , mVU_EATANxy , mVU_ESQRT , mVU_ESIN, \ }; -#define microVU_LowerOP_T3_01_OPCODE(x) void (*mVULowerOP_T3_01_OPCODE##x [32])(mF) = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_MR32 , mVU_SQI , mVU_SQRT , mVU_MFIR, \ - mVU_RGET , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVU_XITOP , mVUunknown, \ - mVU_ERSADD , mVU_EATANxz, mVU_ERSQRT , mVU_EATAN, \ +#define microVU_LowerOP_T3_01_OPCODE(x) void (*mVULowerOP_T3_01_OPCODE##x [32])(mP) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_MR32 , mVU_SQI , mVU_SQRT , mVU_MFIR, \ + mVU_RGET , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVU_XITOP , mVUunknown, \ + mVU_ERSADD , mVU_EATANxz , mVU_ERSQRT , mVU_EATAN, \ }; -#define microVU_LowerOP_T3_10_OPCODE(x) void (*mVULowerOP_T3_10_OPCODE##x [32])(mF) = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_LQD , mVU_RSQRT , mVU_ILWR, \ - mVU_RINIT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_ELENG , mVU_ESUM , mVU_ERCPR , mVU_EEXP, \ +#define microVU_LowerOP_T3_10_OPCODE(x) void (*mVULowerOP_T3_10_OPCODE##x [32])(mP) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_LQD , mVU_RSQRT , mVU_ILWR, \ + mVU_RINIT , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ELENG , mVU_ESUM , mVU_ERCPR , mVU_EEXP, \ }; -#define microVU_LowerOP_T3_11_OPCODE(x) void (*mVULowerOP_T3_11_OPCODE##x [32])(mF) = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVU_SQD , mVU_WAITQ , mVU_ISWR, \ - mVU_RXOR , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_ERLENG , mVUunknown , mVU_WAITP , mVUunknown, \ +#define microVU_LowerOP_T3_11_OPCODE(x) void (*mVULowerOP_T3_11_OPCODE##x [32])(mP) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVU_SQD , mVU_WAITQ , mVU_ISWR, \ + mVU_RXOR , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_ERLENG , mVUunknown , mVU_WAITP , mVUunknown, \ }; -#define microVU_LowerOP_OPCODE(x) void (*mVULowerOP_OPCODE##x [64])(mF) = { \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x20 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_IADD , mVU_ISUB , mVU_IADDI , mVUunknown, /* 0x30 */ \ - mVU_IAND , mVU_IOR , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVULowerOP_T3_00, mVULowerOP_T3_01, mVULowerOP_T3_10, mVULowerOP_T3_11, \ +#define microVU_LowerOP_OPCODE(x) void (*mVULowerOP_OPCODE##x [64])(mP) = { \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x10 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x20 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_IADD , mVU_ISUB , mVU_IADDI , mVUunknown, /* 0x30 */ \ + mVU_IAND , mVU_IOR , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVULowerOP_T3_00, mVULowerOP_T3_01, mVULowerOP_T3_10, mVULowerOP_T3_11, \ }; -#define microVU_UPPER_OPCODE(x) void (*mVU_UPPER_OPCODE##x [64])(mF) = { \ - mVU_ADDx , mVU_ADDy , mVU_ADDz , mVU_ADDw, \ - mVU_SUBx , mVU_SUBy , mVU_SUBz , mVU_SUBw, \ - mVU_MADDx , mVU_MADDy , mVU_MADDz , mVU_MADDw, \ - mVU_MSUBx , mVU_MSUBy , mVU_MSUBz , mVU_MSUBw, \ - mVU_MAXx , mVU_MAXy , mVU_MAXz , mVU_MAXw, /* 0x10 */ \ - mVU_MINIx , mVU_MINIy , mVU_MINIz , mVU_MINIw, \ - mVU_MULx , mVU_MULy , mVU_MULz , mVU_MULw, \ - mVU_MULq , mVU_MAXi , mVU_MULi , mVU_MINIi, \ - mVU_ADDq , mVU_MADDq , mVU_ADDi , mVU_MADDi, /* 0x20 */ \ - mVU_SUBq , mVU_MSUBq , mVU_SUBi , mVU_MSUBi, \ - mVU_ADD , mVU_MADD , mVU_MUL , mVU_MAX, \ - mVU_SUB , mVU_MSUB , mVU_OPMSUB , mVU_MINI, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVU_UPPER_FD_00, mVU_UPPER_FD_01, mVU_UPPER_FD_10, mVU_UPPER_FD_11, \ +#define microVU_UPPER_OPCODE(x) void (*mVU_UPPER_OPCODE##x [64])(mP) = { \ + mVU_ADDx , mVU_ADDy , mVU_ADDz , mVU_ADDw, \ + mVU_SUBx , mVU_SUBy , mVU_SUBz , mVU_SUBw, \ + mVU_MADDx , mVU_MADDy , mVU_MADDz , mVU_MADDw, \ + mVU_MSUBx , mVU_MSUBy , mVU_MSUBz , mVU_MSUBw, \ + mVU_MAXx , mVU_MAXy , mVU_MAXz , mVU_MAXw, /* 0x10 */ \ + mVU_MINIx , mVU_MINIy , mVU_MINIz , mVU_MINIw, \ + mVU_MULx , mVU_MULy , mVU_MULz , mVU_MULw, \ + mVU_MULq , mVU_MAXi , mVU_MULi , mVU_MINIi, \ + mVU_ADDq , mVU_MADDq , mVU_ADDi , mVU_MADDi, /* 0x20 */ \ + mVU_SUBq , mVU_MSUBq , mVU_SUBi , mVU_MSUBi, \ + mVU_ADD , mVU_MADD , mVU_MUL , mVU_MAX, \ + mVU_SUB , mVU_MSUB , mVU_OPMSUB , mVU_MINI, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, /* 0x30 */ \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVU_UPPER_FD_00, mVU_UPPER_FD_01, mVU_UPPER_FD_10, mVU_UPPER_FD_11, \ }; -#define microVU_UPPER_FD_00_TABLE(x) void (*mVU_UPPER_FD_00_TABLE##x [32])(mF) = { \ - mVU_ADDAx , mVU_SUBAx , mVU_MADDAx , mVU_MSUBAx, \ - mVU_ITOF0 , mVU_FTOI0 , mVU_MULAx , mVU_MULAq, \ - mVU_ADDAq , mVU_SUBAq , mVU_ADDA , mVU_SUBA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_00_TABLE(x) void (*mVU_UPPER_FD_00_TABLE##x [32])(mP) = { \ + mVU_ADDAx , mVU_SUBAx , mVU_MADDAx , mVU_MSUBAx, \ + mVU_ITOF0 , mVU_FTOI0 , mVU_MULAx , mVU_MULAq, \ + mVU_ADDAq , mVU_SUBAq , mVU_ADDA , mVU_SUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_01_TABLE(x) void (* mVU_UPPER_FD_01_TABLE##x [32])(mF) = { \ - mVU_ADDAy , mVU_SUBAy , mVU_MADDAy , mVU_MSUBAy, \ - mVU_ITOF4 , mVU_FTOI4 , mVU_MULAy , mVU_ABS, \ - mVU_MADDAq , mVU_MSUBAq , mVU_MADDA , mVU_MSUBA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_01_TABLE(x) void (* mVU_UPPER_FD_01_TABLE##x [32])(mP) = { \ + mVU_ADDAy , mVU_SUBAy , mVU_MADDAy , mVU_MSUBAy, \ + mVU_ITOF4 , mVU_FTOI4 , mVU_MULAy , mVU_ABS, \ + mVU_MADDAq , mVU_MSUBAq , mVU_MADDA , mVU_MSUBA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_10_TABLE(x) void (* mVU_UPPER_FD_10_TABLE##x [32])(mF) = { \ - mVU_ADDAz , mVU_SUBAz , mVU_MADDAz , mVU_MSUBAz, \ - mVU_ITOF12 , mVU_FTOI12 , mVU_MULAz , mVU_MULAi, \ - mVU_ADDAi , mVU_SUBAi , mVU_MULA , mVU_OPMULA, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_10_TABLE(x) void (* mVU_UPPER_FD_10_TABLE##x [32])(mP) = { \ + mVU_ADDAz , mVU_SUBAz , mVU_MADDAz , mVU_MSUBAz, \ + mVU_ITOF12 , mVU_FTOI12 , mVU_MULAz , mVU_MULAi, \ + mVU_ADDAi , mVU_SUBAi , mVU_MULA , mVU_OPMULA, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; -#define microVU_UPPER_FD_11_TABLE(x) void (* mVU_UPPER_FD_11_TABLE##x [32])(mF) = { \ - mVU_ADDAw , mVU_SUBAw , mVU_MADDAw , mVU_MSUBAw, \ - mVU_ITOF15 , mVU_FTOI15 , mVU_MULAw , mVU_CLIP, \ - mVU_MADDAi , mVU_MSUBAi , mVUunknown , mVU_NOP, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ - mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ +#define microVU_UPPER_FD_11_TABLE(x) void (* mVU_UPPER_FD_11_TABLE##x [32])(mP) = { \ + mVU_ADDAw , mVU_SUBAw , mVU_MADDAw , mVU_MSUBAw, \ + mVU_ITOF15 , mVU_FTOI15 , mVU_MULAw , mVU_CLIP, \ + mVU_MADDAi , mVU_MSUBAi , mVUunknown , mVU_NOP, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ + mVUunknown , mVUunknown , mVUunknown , mVUunknown, \ }; //------------------------------------------------------------------ @@ -218,29 +218,27 @@ microVU_UPPER_FD_10_TABLE(x) \ microVU_UPPER_FD_11_TABLE(x) mVUcreateTable(0) -mVUcreateTable(1) //------------------------------------------------------------------ // Table Functions //------------------------------------------------------------------ -#define doTableStuff(tableName, args) { \ - if (vuIndex) tableName##1[ args ](recPass); \ - else tableName##0[ args ](recPass); \ +#define doTableStuff(tableName, args) { \ + tableName##0[ args ](mX); \ } -microVUf(void) mVU_UPPER_FD_00(mF) { doTableStuff(mVU_UPPER_FD_00_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_01(mF) { doTableStuff(mVU_UPPER_FD_01_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_10(mF) { doTableStuff(mVU_UPPER_FD_10_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVU_UPPER_FD_11(mF) { doTableStuff(mVU_UPPER_FD_11_TABLE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP(mF) { doTableStuff(mVULowerOP_OPCODE, (mVUgetCode & 0x3f)); } -microVUf(void) mVULowerOP_T3_00(mF) { doTableStuff(mVULowerOP_T3_00_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_01(mF) { doTableStuff(mVULowerOP_T3_01_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_10(mF) { doTableStuff(mVULowerOP_T3_10_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVULowerOP_T3_11(mF) { doTableStuff(mVULowerOP_T3_11_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } -microVUf(void) mVUopU(mF) { doTableStuff(mVU_UPPER_OPCODE, (mVUgetCode & 0x3f)); } // Gets Upper Opcode -microVUf(void) mVUopL(mF) { doTableStuff(mVULOWER_OPCODE, (mVUgetCode >> 25)); } // Gets Lower Opcode -microVUf(void) mVUunknown(mF) { - pass2 { SysPrintf("microVU%d: Unknown Micro VU opcode called (%x)\n", vuIndex, mVUgetCode); } +mVUop(mVU_UPPER_FD_00) { doTableStuff(mVU_UPPER_FD_00_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVU_UPPER_FD_01) { doTableStuff(mVU_UPPER_FD_01_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVU_UPPER_FD_10) { doTableStuff(mVU_UPPER_FD_10_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVU_UPPER_FD_11) { doTableStuff(mVU_UPPER_FD_11_TABLE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVULowerOP) { doTableStuff(mVULowerOP_OPCODE, (mVUgetCode & 0x3f)); } +mVUop(mVULowerOP_T3_00) { doTableStuff(mVULowerOP_T3_00_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVULowerOP_T3_01) { doTableStuff(mVULowerOP_T3_01_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVULowerOP_T3_10) { doTableStuff(mVULowerOP_T3_10_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVULowerOP_T3_11) { doTableStuff(mVULowerOP_T3_11_OPCODE, ((mVUgetCode >> 6) & 0x1f)); } +mVUop(mVUopU) { doTableStuff(mVU_UPPER_OPCODE, (mVUgetCode & 0x3f)); } // Gets Upper Opcode +mVUop(mVUopL) { doTableStuff(mVULOWER_OPCODE, (mVUgetCode >> 25)); } // Gets Lower Opcode +mVUop(mVUunknown) { + pass2 { SysPrintf("microVU%d: Unknown Micro VU opcode called (%x)\n", getIndex, mVUgetCode); } pass3 { mVUlog("Unknown", mVUgetCode); } } diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index ca1b31b4ba..30591004f4 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -27,8 +27,7 @@ #define SHIFT_XYZW(gprReg) { if (_XYZW_SS && modXYZW && !_W) { SHL32ItoR(gprReg, ADD_XYZW); } } // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations -microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modXYZW) { - microVU* mVU = mVUx; +microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool modXYZW) { int sReg, mReg = gprT1; static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; @@ -40,7 +39,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw if (doStatus) { getFlagReg(sReg, fsInstance); // Set sReg to valid GPR by Cur Flag Instance - mVUallocSFLAGa(sReg, fpsInstance); // Get Prev Status Flag + mVUallocSFLAGa(sReg, fpsInstance); // Get Prev Status Flag AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags } @@ -73,7 +72,7 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX if (_XYZW_SS && doStatus) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - if (doMac) mVUallocMFLAGb(mReg, fmInstance); // Set Mac Flag + if (doMac) mVUallocMFLAGb(mVU, mReg, fmInstance); // Set Mac Flag } //------------------------------------------------------------------ @@ -82,95 +81,88 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX // FMAC1 - Normal FMAC Opcodes #define mVU_FMAC1(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, Fs, Ft; \ - mVUallocFMAC1a(Fd, Fs, Ft); \ + mVUallocFMAC1a(mVU, Fd, Fs, Ft); \ if (_XYZW_SS) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ - mVUallocFMAC1b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ + mVUallocFMAC1b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogFt(); } \ } // FMAC3 - BC(xyzw) FMAC Opcodes #define mVU_FMAC3(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC3(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC3(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, Fs, Ft; \ - mVUallocFMAC3a(Fd, Fs, Ft); \ + mVUallocFMAC3a(mVU, Fd, Fs, Ft); \ if (_XYZW_SS) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ - mVUallocFMAC3b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ + mVUallocFMAC3b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogBC(); } \ } // FMAC4 - FMAC Opcodes Storing Result to ACC #define mVU_FMAC4(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, _Ft_); } \ pass2 { \ int ACC, Fs, Ft; \ - mVUallocFMAC4a(ACC, Fs, Ft); \ + mVUallocFMAC4a(mVU, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC4b(ACC, Fs); \ + mVUupdateFlags(mVU, Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC4b(mVU, ACC, Fs); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogFt(); } \ } // FMAC5 - FMAC BC(xyzw) Opcodes Storing Result to ACC #define mVU_FMAC5(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC3(0, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC3(mVU, 0, _Fs_, _Ft_); } \ pass2 { \ int ACC, Fs, Ft; \ - mVUallocFMAC5a(ACC, Fs, Ft); \ + mVUallocFMAC5a(mVU, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC5b(ACC, Fs); \ + mVUupdateFlags(mVU, Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC5b(mVU, ACC, Fs); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogBC(); } \ } // FMAC6 - Normal FMAC Opcodes (I Reg) #define mVU_FMAC6(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, Fs, Ft; \ - mVUallocFMAC6a(Fd, Fs, Ft); \ + mVUallocFMAC6a(mVU, Fd, Fs, Ft); \ if (_XYZW_SS) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ - mVUallocFMAC6b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ + mVUallocFMAC6b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogI(); } \ } // FMAC7 - FMAC Opcodes Storing Result to ACC (I Reg) #define mVU_FMAC7(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, 0); } \ pass2 { \ int ACC, Fs, Ft; \ - mVUallocFMAC7a(ACC, Fs, Ft); \ + mVUallocFMAC7a(mVU, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC7b(ACC, Fs); \ + mVUupdateFlags(mVU, Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC7b(mVU, ACC, Fs); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogI(); } \ } // FMAC8 - MADD FMAC Opcode Storing Result to Fd #define mVU_FMAC8(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC8a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC8a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(Fs, ACC); \ @@ -179,18 +171,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(Fs, ACC); \ } \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC8b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC8b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogFt(); } \ } // FMAC9 - MSUB FMAC Opcode Storing Result to Fd #define mVU_FMAC9(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC9a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC9a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACC, Fs); \ @@ -199,18 +190,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACC, Fs); \ } \ - mVUupdateFlags(Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC9b(Fd); \ + mVUupdateFlags(mVU, Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC9b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogFt(); } \ } // FMAC10 - MADD FMAC BC(xyzw) Opcode Storing Result to Fd #define mVU_FMAC10(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC3(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC3(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC10a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC10a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(Fs, ACC); \ @@ -219,18 +209,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(Fs, ACC); \ } \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC10b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC10b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogBC(); } \ } // FMAC11 - MSUB FMAC BC(xyzw) Opcode Storing Result to Fd #define mVU_FMAC11(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC3(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC3(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC11a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC11a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACC, Fs); \ @@ -239,18 +228,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACC, Fs); \ } \ - mVUupdateFlags(Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC11b(Fd); \ + mVUupdateFlags(mVU, Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC11b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogBC(); } \ } // FMAC12 - MADD FMAC Opcode Storing Result to Fd (I Reg) #define mVU_FMAC12(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC12a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC12a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(Fs, ACC); \ @@ -259,18 +247,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(Fs, ACC); \ } \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC12b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC12b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogI(); } \ } // FMAC13 - MSUB FMAC Opcode Storing Result to Fd (I Reg) #define mVU_FMAC13(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC13a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC13a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACC, Fs); \ @@ -279,18 +266,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACC, Fs); \ } \ - mVUupdateFlags(Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC13b(Fd); \ + mVUupdateFlags(mVU, Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC13b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogI(); } \ } // FMAC14 - MADDA/MSUBA FMAC Opcode #define mVU_FMAC14(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, _Ft_); } \ pass2 { \ int ACCw, ACCr, Fs, Ft; \ - mVUallocFMAC14a(ACCw, ACCr, Fs, Ft); \ + mVUallocFMAC14a(mVU, ACCw, ACCr, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACCr, Fs); \ @@ -299,18 +285,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACCr, Fs); \ } \ - mVUupdateFlags(ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC14b(ACCw, ACCr); \ + mVUupdateFlags(mVU, ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC14b(mVU, ACCw, ACCr); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogFt(); } \ } // FMAC15 - MADDA/MSUBA BC(xyzw) FMAC Opcode #define mVU_FMAC15(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC3(0, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC3(mVU, 0, _Fs_, _Ft_); } \ pass2 { \ int ACCw, ACCr, Fs, Ft; \ - mVUallocFMAC15a(ACCw, ACCr, Fs, Ft); \ + mVUallocFMAC15a(mVU, ACCw, ACCr, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACCr, Fs); \ @@ -319,18 +304,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACCr, Fs); \ } \ - mVUupdateFlags(ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC15b(ACCw, ACCr); \ + mVUupdateFlags(mVU, ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC15b(mVU, ACCw, ACCr); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogBC(); } \ } // FMAC16 - MADDA/MSUBA FMAC Opcode (I Reg) #define mVU_FMAC16(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, 0); } \ pass2 { \ int ACCw, ACCr, Fs, Ft; \ - mVUallocFMAC16a(ACCw, ACCr, Fs, Ft); \ + mVUallocFMAC16a(mVU, ACCw, ACCr, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACCr, Fs); \ @@ -339,73 +323,68 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACCr, Fs); \ } \ - mVUupdateFlags(ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC16b(ACCw, ACCr); \ + mVUupdateFlags(mVU, ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC16b(mVU, ACCw, ACCr); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogI(); } \ } // FMAC18 - OPMULA FMAC Opcode #define mVU_FMAC18(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, _Ft_); } \ pass2 { \ int ACC, Fs, Ft; \ - mVUallocFMAC18a(ACC, Fs, Ft); \ + mVUallocFMAC18a(mVU, ACC, Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC18b(ACC, Fs); \ + mVUupdateFlags(mVU, Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC18b(mVU, ACC, Fs); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogFt(); } \ } // FMAC19 - OPMSUB FMAC Opcode #define mVU_FMAC19(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, _Ft_); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, _Ft_); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC19a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC19a(mVU, Fd, ACC, Fs, Ft); \ SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACC, Fs); \ - mVUupdateFlags(Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC19b(Fd); \ + mVUupdateFlags(mVU, Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC19b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogFt(); } \ } // FMAC22 - Normal FMAC Opcodes (Q Reg) #define mVU_FMAC22(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, Fs, Ft; \ - mVUallocFMAC22a(Fd, Fs, Ft); \ + mVUallocFMAC22a(mVU, Fd, Fs, Ft); \ if (_XYZW_SS) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ - mVUallocFMAC22b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 1); \ + mVUallocFMAC22b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogQ(); } \ } // FMAC23 - FMAC Opcodes Storing Result to ACC (Q Reg) #define mVU_FMAC23(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, 0); } \ pass2 { \ int ACC, Fs, Ft; \ - mVUallocFMAC23a(ACC, Fs, Ft); \ + mVUallocFMAC23a(mVU, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) SSE_##operation##SS_XMM_to_XMM(Fs, Ft); \ else SSE_##operation##PS_XMM_to_XMM(Fs, Ft); \ - mVUupdateFlags(Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC23b(ACC, Fs); \ + mVUupdateFlags(mVU, Fs, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC23b(mVU, ACC, Fs); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogQ();} \ } // FMAC24 - MADD FMAC Opcode Storing Result to Fd (Q Reg) #define mVU_FMAC24(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC24a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC24a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(Fs, ACC); \ @@ -414,18 +393,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(Fs, ACC); \ } \ - mVUupdateFlags(Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC24b(Fd); \ + mVUupdateFlags(mVU, Fd, xmmT1, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC24b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogQ(); } \ } // FMAC25 - MSUB FMAC Opcode Storing Result to Fd (Q Reg) #define mVU_FMAC25(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(_Fd_, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, _Fd_, _Fs_, 0); } \ pass2 { \ int Fd, ACC, Fs, Ft; \ - mVUallocFMAC25a(Fd, ACC, Fs, Ft); \ + mVUallocFMAC25a(mVU, Fd, ACC, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACC, Fs); \ @@ -434,18 +412,17 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACC, Fs); \ } \ - mVUupdateFlags(Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC25b(Fd); \ + mVUupdateFlags(mVU, Fd, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC25b(mVU, Fd); \ } \ pass3 { mVUlog(OPname); mVUlogFd(); mVUlogQ(); } \ } // FMAC26 - MADDA/MSUBA FMAC Opcode (Q Reg) #define mVU_FMAC26(operation, OPname) { \ - microVU* mVU = mVUx; \ - pass1 { mVUanalyzeFMAC1(0, _Fs_, 0); } \ + pass1 { mVUanalyzeFMAC1(mVU, 0, _Fs_, 0); } \ pass2 { \ int ACCw, ACCr, Fs, Ft; \ - mVUallocFMAC26a(ACCw, ACCr, Fs, Ft); \ + mVUallocFMAC26a(mVU, ACCw, ACCr, Fs, Ft); \ if (_X_Y_Z_W == 8) { \ SSE_MULSS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##SS_XMM_to_XMM(ACCr, Fs); \ @@ -454,123 +431,121 @@ microVUt(void) mVUupdateFlags(int reg, int regT1, int regT2, int xyzw, bool modX SSE_MULPS_XMM_to_XMM(Fs, Ft); \ SSE_##operation##PS_XMM_to_XMM(ACCr, Fs); \ } \ - mVUupdateFlags(ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ - mVUallocFMAC26b(ACCw, ACCr); \ + mVUupdateFlags(mVU, ACCr, Fs, xmmT2, _X_Y_Z_W, 0); \ + mVUallocFMAC26b(mVU, ACCw, ACCr); \ } \ pass3 { mVUlog(OPname); mVUlogACC(); mVUlogQ(); } \ } // FMAC27~29 - MAX/MINI FMAC Opcodes -#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } -#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } -#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { microVU* mVU = mVUx; mVUinfo &= ~_doStatus; } } +#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } +#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } +#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } //------------------------------------------------------------------ // Micro VU Micromode Upper instructions //------------------------------------------------------------------ -microVUf(void) mVU_ABS(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } +mVUop(mVU_ABS) { + pass1 { mVUanalyzeFMAC2(mVU, _Fs_, _Ft_); } pass2 { int Fs, Ft; - mVUallocFMAC2a(Fs, Ft); + mVUallocFMAC2a(mVU, Fs, Ft); SSE_ANDPS_M128_to_XMM(Fs, (uptr)mVU_absclip); - mVUallocFMAC2b(Ft); + mVUallocFMAC2b(mVU, Ft); } pass3 { mVUlog("ABS"); mVUlogFtFs(); } } -microVUf(void) mVU_ADD(mF) { mVU_FMAC1 (ADD, "ADD"); } -microVUf(void) mVU_ADDi(mF) { mVU_FMAC6 (ADD2, "ADDi"); } -microVUf(void) mVU_ADDq(mF) { mVU_FMAC22(ADD, "ADDq"); } -microVUf(void) mVU_ADDx(mF) { mVU_FMAC3 (ADD, "ADDx"); } -microVUf(void) mVU_ADDy(mF) { mVU_FMAC3 (ADD, "ADDy"); } -microVUf(void) mVU_ADDz(mF) { mVU_FMAC3 (ADD, "ADDz"); } -microVUf(void) mVU_ADDw(mF) { mVU_FMAC3 (ADD, "ADDw"); } -microVUf(void) mVU_ADDA(mF) { mVU_FMAC4 (ADD, "ADDA"); } -microVUf(void) mVU_ADDAi(mF) { mVU_FMAC7 (ADD, "ADDAi"); } -microVUf(void) mVU_ADDAq(mF) { mVU_FMAC23(ADD, "ADDAq"); } -microVUf(void) mVU_ADDAx(mF) { mVU_FMAC5 (ADD, "ADDAx"); } -microVUf(void) mVU_ADDAy(mF) { mVU_FMAC5 (ADD, "ADDAy"); } -microVUf(void) mVU_ADDAz(mF) { mVU_FMAC5 (ADD, "ADDAz"); } -microVUf(void) mVU_ADDAw(mF) { mVU_FMAC5 (ADD, "ADDAw"); } -microVUf(void) mVU_SUB(mF) { mVU_FMAC1 (SUB, "SUB"); } -microVUf(void) mVU_SUBi(mF) { mVU_FMAC6 (SUB, "SUBi"); } -microVUf(void) mVU_SUBq(mF) { mVU_FMAC22(SUB, "SUBq"); } -microVUf(void) mVU_SUBx(mF) { mVU_FMAC3 (SUB, "SUBx"); } -microVUf(void) mVU_SUBy(mF) { mVU_FMAC3 (SUB, "SUBy"); } -microVUf(void) mVU_SUBz(mF) { mVU_FMAC3 (SUB, "SUBz"); } -microVUf(void) mVU_SUBw(mF) { mVU_FMAC3 (SUB, "SUBw"); } -microVUf(void) mVU_SUBA(mF) { mVU_FMAC4 (SUB, "SUBA"); } -microVUf(void) mVU_SUBAi(mF) { mVU_FMAC7 (SUB, "SUBAi"); } -microVUf(void) mVU_SUBAq(mF) { mVU_FMAC23(SUB, "SUBAq"); } -microVUf(void) mVU_SUBAx(mF) { mVU_FMAC5 (SUB, "SUBAx"); } -microVUf(void) mVU_SUBAy(mF) { mVU_FMAC5 (SUB, "SUBAy"); } -microVUf(void) mVU_SUBAz(mF) { mVU_FMAC5 (SUB, "SUBAz"); } -microVUf(void) mVU_SUBAw(mF) { mVU_FMAC5 (SUB, "SUBAw"); } -microVUf(void) mVU_MUL(mF) { mVU_FMAC1 (MUL, "MUL"); } -microVUf(void) mVU_MULi(mF) { mVU_FMAC6 (MUL, "MULi"); } -microVUf(void) mVU_MULq(mF) { mVU_FMAC22(MUL, "MULq"); } -microVUf(void) mVU_MULx(mF) { mVU_FMAC3 (MUL, "MULx"); } -microVUf(void) mVU_MULy(mF) { mVU_FMAC3 (MUL, "MULy"); } -microVUf(void) mVU_MULz(mF) { mVU_FMAC3 (MUL, "MULz"); } -microVUf(void) mVU_MULw(mF) { mVU_FMAC3 (MUL, "MULw"); } -microVUf(void) mVU_MULA(mF) { mVU_FMAC4 (MUL, "MULA"); } -microVUf(void) mVU_MULAi(mF) { mVU_FMAC7 (MUL, "MULAi"); } -microVUf(void) mVU_MULAq(mF) { mVU_FMAC23(MUL, "MULAq"); } -microVUf(void) mVU_MULAx(mF) { mVU_FMAC5 (MUL, "MULAx"); } -microVUf(void) mVU_MULAy(mF) { mVU_FMAC5 (MUL, "MULAy"); } -microVUf(void) mVU_MULAz(mF) { mVU_FMAC5 (MUL, "MULAz"); } -microVUf(void) mVU_MULAw(mF) { mVU_FMAC5 (MUL, "MULAw"); } -microVUf(void) mVU_MADD(mF) { mVU_FMAC8 (ADD, "MADD"); } -microVUf(void) mVU_MADDi(mF) { mVU_FMAC12(ADD, "MADDi"); } -microVUf(void) mVU_MADDq(mF) { mVU_FMAC24(ADD, "MADDq"); } -microVUf(void) mVU_MADDx(mF) { mVU_FMAC10(ADD, "MADDx"); } -microVUf(void) mVU_MADDy(mF) { mVU_FMAC10(ADD, "MADDy"); } -microVUf(void) mVU_MADDz(mF) { mVU_FMAC10(ADD, "MADDz"); } -microVUf(void) mVU_MADDw(mF) { mVU_FMAC10(ADD, "MADDw"); } -microVUf(void) mVU_MADDA(mF) { mVU_FMAC14(ADD, "MADDA"); } -microVUf(void) mVU_MADDAi(mF) { mVU_FMAC16(ADD, "MADDAi"); } -microVUf(void) mVU_MADDAq(mF) { mVU_FMAC26(ADD, "MADDAq"); } -microVUf(void) mVU_MADDAx(mF) { mVU_FMAC15(ADD, "MADDAx"); } -microVUf(void) mVU_MADDAy(mF) { mVU_FMAC15(ADD, "MADDAy"); } -microVUf(void) mVU_MADDAz(mF) { mVU_FMAC15(ADD, "MADDAz"); } -microVUf(void) mVU_MADDAw(mF) { mVU_FMAC15(ADD, "MADDAw"); } -microVUf(void) mVU_MSUB(mF) { mVU_FMAC9 (SUB, "MSUB"); } -microVUf(void) mVU_MSUBi(mF) { mVU_FMAC13(SUB, "MSUBi"); } -microVUf(void) mVU_MSUBq(mF) { mVU_FMAC25(SUB, "MSUBq"); } -microVUf(void) mVU_MSUBx(mF) { mVU_FMAC11(SUB, "MSUBx"); } -microVUf(void) mVU_MSUBy(mF) { mVU_FMAC11(SUB, "MSUBy"); } -microVUf(void) mVU_MSUBz(mF) { mVU_FMAC11(SUB, "MSUBz"); } -microVUf(void) mVU_MSUBw(mF) { mVU_FMAC11(SUB, "MSUBw"); } -microVUf(void) mVU_MSUBA(mF) { mVU_FMAC14(SUB, "MSUBA"); } -microVUf(void) mVU_MSUBAi(mF) { mVU_FMAC16(SUB, "MSUBAi"); } -microVUf(void) mVU_MSUBAq(mF) { mVU_FMAC26(SUB, "MSUBAq"); } -microVUf(void) mVU_MSUBAx(mF) { mVU_FMAC15(SUB, "MSUBAx"); } -microVUf(void) mVU_MSUBAy(mF) { mVU_FMAC15(SUB, "MSUBAy"); } -microVUf(void) mVU_MSUBAz(mF) { mVU_FMAC15(SUB, "MSUBAz"); } -microVUf(void) mVU_MSUBAw(mF) { mVU_FMAC15(SUB, "MSUBAw"); } -microVUf(void) mVU_MAX(mF) { mVU_FMAC27(MAX2, "MAX"); } -microVUf(void) mVU_MAXi(mF) { mVU_FMAC28(MAX2, "MAXi"); } -microVUf(void) mVU_MAXx(mF) { mVU_FMAC29(MAX2, "MAXx"); } -microVUf(void) mVU_MAXy(mF) { mVU_FMAC29(MAX2, "MAXy"); } -microVUf(void) mVU_MAXz(mF) { mVU_FMAC29(MAX2, "MAXz"); } -microVUf(void) mVU_MAXw(mF) { mVU_FMAC29(MAX2, "MAXw"); } -microVUf(void) mVU_MINI(mF) { mVU_FMAC27(MIN2, "MINI"); } -microVUf(void) mVU_MINIi(mF) { mVU_FMAC28(MIN2, "MINIi"); } -microVUf(void) mVU_MINIx(mF) { mVU_FMAC29(MIN2, "MINIx"); } -microVUf(void) mVU_MINIy(mF) { mVU_FMAC29(MIN2, "MINIy"); } -microVUf(void) mVU_MINIz(mF) { mVU_FMAC29(MIN2, "MINIz"); } -microVUf(void) mVU_MINIw(mF) { mVU_FMAC29(MIN2, "MINIw"); } -microVUf(void) mVU_OPMULA(mF) { mVU_FMAC18(MUL, "OPMULA"); } -microVUf(void) mVU_OPMSUB(mF) { mVU_FMAC19(SUB, "OPMSUB"); } -microVUf(void) mVU_NOP(mF) { pass3 { mVUlog("NOP"); } } -microVUq(void) mVU_FTOIx(uptr addr, int recPass) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } +mVUop(mVU_ADD) { mVU_FMAC1 (ADD, "ADD"); } +mVUop(mVU_ADDi) { mVU_FMAC6 (ADD2, "ADDi"); } +mVUop(mVU_ADDq) { mVU_FMAC22(ADD, "ADDq"); } +mVUop(mVU_ADDx) { mVU_FMAC3 (ADD, "ADDx"); } +mVUop(mVU_ADDy) { mVU_FMAC3 (ADD, "ADDy"); } +mVUop(mVU_ADDz) { mVU_FMAC3 (ADD, "ADDz"); } +mVUop(mVU_ADDw) { mVU_FMAC3 (ADD, "ADDw"); } +mVUop(mVU_ADDA) { mVU_FMAC4 (ADD, "ADDA"); } +mVUop(mVU_ADDAi) { mVU_FMAC7 (ADD, "ADDAi"); } +mVUop(mVU_ADDAq) { mVU_FMAC23(ADD, "ADDAq"); } +mVUop(mVU_ADDAx) { mVU_FMAC5 (ADD, "ADDAx"); } +mVUop(mVU_ADDAy) { mVU_FMAC5 (ADD, "ADDAy"); } +mVUop(mVU_ADDAz) { mVU_FMAC5 (ADD, "ADDAz"); } +mVUop(mVU_ADDAw) { mVU_FMAC5 (ADD, "ADDAw"); } +mVUop(mVU_SUB) { mVU_FMAC1 (SUB, "SUB"); } +mVUop(mVU_SUBi) { mVU_FMAC6 (SUB, "SUBi"); } +mVUop(mVU_SUBq) { mVU_FMAC22(SUB, "SUBq"); } +mVUop(mVU_SUBx) { mVU_FMAC3 (SUB, "SUBx"); } +mVUop(mVU_SUBy) { mVU_FMAC3 (SUB, "SUBy"); } +mVUop(mVU_SUBz) { mVU_FMAC3 (SUB, "SUBz"); } +mVUop(mVU_SUBw) { mVU_FMAC3 (SUB, "SUBw"); } +mVUop(mVU_SUBA) { mVU_FMAC4 (SUB, "SUBA"); } +mVUop(mVU_SUBAi) { mVU_FMAC7 (SUB, "SUBAi"); } +mVUop(mVU_SUBAq) { mVU_FMAC23(SUB, "SUBAq"); } +mVUop(mVU_SUBAx) { mVU_FMAC5 (SUB, "SUBAx"); } +mVUop(mVU_SUBAy) { mVU_FMAC5 (SUB, "SUBAy"); } +mVUop(mVU_SUBAz) { mVU_FMAC5 (SUB, "SUBAz"); } +mVUop(mVU_SUBAw) { mVU_FMAC5 (SUB, "SUBAw"); } +mVUop(mVU_MUL) { mVU_FMAC1 (MUL, "MUL"); } +mVUop(mVU_MULi) { mVU_FMAC6 (MUL, "MULi"); } +mVUop(mVU_MULq) { mVU_FMAC22(MUL, "MULq"); } +mVUop(mVU_MULx) { mVU_FMAC3 (MUL, "MULx"); } +mVUop(mVU_MULy) { mVU_FMAC3 (MUL, "MULy"); } +mVUop(mVU_MULz) { mVU_FMAC3 (MUL, "MULz"); } +mVUop(mVU_MULw) { mVU_FMAC3 (MUL, "MULw"); } +mVUop(mVU_MULA) { mVU_FMAC4 (MUL, "MULA"); } +mVUop(mVU_MULAi) { mVU_FMAC7 (MUL, "MULAi"); } +mVUop(mVU_MULAq) { mVU_FMAC23(MUL, "MULAq"); } +mVUop(mVU_MULAx) { mVU_FMAC5 (MUL, "MULAx"); } +mVUop(mVU_MULAy) { mVU_FMAC5 (MUL, "MULAy"); } +mVUop(mVU_MULAz) { mVU_FMAC5 (MUL, "MULAz"); } +mVUop(mVU_MULAw) { mVU_FMAC5 (MUL, "MULAw"); } +mVUop(mVU_MADD) { mVU_FMAC8 (ADD, "MADD"); } +mVUop(mVU_MADDi) { mVU_FMAC12(ADD, "MADDi"); } +mVUop(mVU_MADDq) { mVU_FMAC24(ADD, "MADDq"); } +mVUop(mVU_MADDx) { mVU_FMAC10(ADD, "MADDx"); } +mVUop(mVU_MADDy) { mVU_FMAC10(ADD, "MADDy"); } +mVUop(mVU_MADDz) { mVU_FMAC10(ADD, "MADDz"); } +mVUop(mVU_MADDw) { mVU_FMAC10(ADD, "MADDw"); } +mVUop(mVU_MADDA) { mVU_FMAC14(ADD, "MADDA"); } +mVUop(mVU_MADDAi) { mVU_FMAC16(ADD, "MADDAi"); } +mVUop(mVU_MADDAq) { mVU_FMAC26(ADD, "MADDAq"); } +mVUop(mVU_MADDAx) { mVU_FMAC15(ADD, "MADDAx"); } +mVUop(mVU_MADDAy) { mVU_FMAC15(ADD, "MADDAy"); } +mVUop(mVU_MADDAz) { mVU_FMAC15(ADD, "MADDAz"); } +mVUop(mVU_MADDAw) { mVU_FMAC15(ADD, "MADDAw"); } +mVUop(mVU_MSUB) { mVU_FMAC9 (SUB, "MSUB"); } +mVUop(mVU_MSUBi) { mVU_FMAC13(SUB, "MSUBi"); } +mVUop(mVU_MSUBq) { mVU_FMAC25(SUB, "MSUBq"); } +mVUop(mVU_MSUBx) { mVU_FMAC11(SUB, "MSUBx"); } +mVUop(mVU_MSUBy) { mVU_FMAC11(SUB, "MSUBy"); } +mVUop(mVU_MSUBz) { mVU_FMAC11(SUB, "MSUBz"); } +mVUop(mVU_MSUBw) { mVU_FMAC11(SUB, "MSUBw"); } +mVUop(mVU_MSUBA) { mVU_FMAC14(SUB, "MSUBA"); } +mVUop(mVU_MSUBAi) { mVU_FMAC16(SUB, "MSUBAi"); } +mVUop(mVU_MSUBAq) { mVU_FMAC26(SUB, "MSUBAq"); } +mVUop(mVU_MSUBAx) { mVU_FMAC15(SUB, "MSUBAx"); } +mVUop(mVU_MSUBAy) { mVU_FMAC15(SUB, "MSUBAy"); } +mVUop(mVU_MSUBAz) { mVU_FMAC15(SUB, "MSUBAz"); } +mVUop(mVU_MSUBAw) { mVU_FMAC15(SUB, "MSUBAw"); } +mVUop(mVU_MAX) { mVU_FMAC27(MAX2, "MAX"); } +mVUop(mVU_MAXi) { mVU_FMAC28(MAX2, "MAXi"); } +mVUop(mVU_MAXx) { mVU_FMAC29(MAX2, "MAXx"); } +mVUop(mVU_MAXy) { mVU_FMAC29(MAX2, "MAXy"); } +mVUop(mVU_MAXz) { mVU_FMAC29(MAX2, "MAXz"); } +mVUop(mVU_MAXw) { mVU_FMAC29(MAX2, "MAXw"); } +mVUop(mVU_MINI) { mVU_FMAC27(MIN2, "MINI"); } +mVUop(mVU_MINIi) { mVU_FMAC28(MIN2, "MINIi"); } +mVUop(mVU_MINIx) { mVU_FMAC29(MIN2, "MINIx"); } +mVUop(mVU_MINIy) { mVU_FMAC29(MIN2, "MINIy"); } +mVUop(mVU_MINIz) { mVU_FMAC29(MIN2, "MINIz"); } +mVUop(mVU_MINIw) { mVU_FMAC29(MIN2, "MINIw"); } +mVUop(mVU_OPMULA) { mVU_FMAC18(MUL, "OPMULA"); } +mVUop(mVU_OPMSUB) { mVU_FMAC19(SUB, "OPMSUB"); } +mVUop(mVU_NOP) { pass3 { mVUlog("NOP"); } } +void mVU_FTOIx(mP, uptr addr) { + pass1 { mVUanalyzeFMAC2(mVU, _Fs_, _Ft_); } pass2 { int Fs, Ft; - mVUallocFMAC2a(Fs, Ft); + mVUallocFMAC2a(mVU, Fs, Ft); // Note: For help understanding this algorithm see recVUMI_FTOI_Saturate() SSE_MOVAPS_XMM_to_XMM(xmmT1, Fs); @@ -583,38 +558,36 @@ microVUq(void) mVU_FTOIx(uptr addr, int recPass) { SSE_ANDPS_XMM_to_XMM(xmmT1, xmmFt); SSE2_PADDD_XMM_to_XMM(Fs, xmmT1); - mVUallocFMAC2b(Ft); + mVUallocFMAC2b(mVU, Ft); } } -microVUf(void) mVU_FTOI0(mF) { mVU_FTOIx((uptr)0, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI0"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI4(mF) { mVU_FTOIx((uptr)mVU_FTOI_4, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI4"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI12(mF) { mVU_FTOIx((uptr)mVU_FTOI_12, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI12"); mVUlogFtFs(); } } -microVUf(void) mVU_FTOI15(mF) { mVU_FTOIx((uptr)mVU_FTOI_15, recPass); pass3 { microVU* mVU = mVUx; mVUlog("FTOI15"); mVUlogFtFs(); } } -microVUq(void) mVU_ITOFx(uptr addr, int recPass) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFMAC2(_Fs_, _Ft_); } +mVUop(mVU_FTOI0) { mVU_FTOIx(mX, (uptr)0); pass3 { mVUlog("FTOI0"); mVUlogFtFs(); } } +mVUop(mVU_FTOI4) { mVU_FTOIx(mX, (uptr)mVU_FTOI_4); pass3 { mVUlog("FTOI4"); mVUlogFtFs(); } } +mVUop(mVU_FTOI12) { mVU_FTOIx(mX, (uptr)mVU_FTOI_12); pass3 { mVUlog("FTOI12"); mVUlogFtFs(); } } +mVUop(mVU_FTOI15) { mVU_FTOIx(mX, (uptr)mVU_FTOI_15); pass3 { mVUlog("FTOI15"); mVUlogFtFs(); } } +void mVU_ITOFx(mP, uptr addr) { + pass1 { mVUanalyzeFMAC2(mVU, _Fs_, _Ft_); } pass2 { int Fs, Ft; - mVUallocFMAC2a(Fs, Ft); + mVUallocFMAC2a(mVU, Fs, Ft); SSE2_CVTDQ2PS_XMM_to_XMM(Ft, Fs); if (addr) { SSE_MULPS_M128_to_XMM(Ft, addr); } //mVUclamp2(Ft, xmmT1, 15); // Clamp (not sure if this is needed) - mVUallocFMAC2b(Ft); + mVUallocFMAC2b(mVU, Ft); } } -microVUf(void) mVU_ITOF0(mF) { mVU_ITOFx((uptr)0, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF0"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF4(mF) { mVU_ITOFx((uptr)mVU_ITOF_4, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF4"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF12(mF) { mVU_ITOFx((uptr)mVU_ITOF_12, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF12"); mVUlogFtFs(); } } -microVUf(void) mVU_ITOF15(mF) { mVU_ITOFx((uptr)mVU_ITOF_15, recPass); pass3 { microVU* mVU = mVUx; mVUlog("ITOF15"); mVUlogFtFs(); } } -microVUf(void) mVU_CLIP(mF) { - microVU* mVU = mVUx; - pass1 { mVUanalyzeFMAC4(_Fs_, _Ft_); } +mVUop(mVU_ITOF0) { mVU_ITOFx(mX, (uptr)0); pass3 { mVUlog("ITOF0"); mVUlogFtFs(); } } +mVUop(mVU_ITOF4) { mVU_ITOFx(mX, (uptr)mVU_ITOF_4); pass3 { mVUlog("ITOF4"); mVUlogFtFs(); } } +mVUop(mVU_ITOF12) { mVU_ITOFx(mX, (uptr)mVU_ITOF_12); pass3 { mVUlog("ITOF12"); mVUlogFtFs(); } } +mVUop(mVU_ITOF15) { mVU_ITOFx(mX, (uptr)mVU_ITOF_15); pass3 { mVUlog("ITOF15"); mVUlogFtFs(); } } +mVUop(mVU_CLIP) { + pass1 { mVUanalyzeFMAC4(mVU, _Fs_, _Ft_); } pass2 { int Fs, Ft; - mVUallocFMAC17a(Fs, Ft); - mVUallocCFLAGa(gprT1, fpcInstance); + mVUallocFMAC17a(mVU, Fs, Ft); + mVUallocCFLAGa(mVU, gprT1, fpcInstance); SHL32ItoR(gprT1, 6); SSE_ANDPS_M128_to_XMM(Ft, (uptr)mVU_absclip); @@ -638,7 +611,7 @@ microVUf(void) mVU_CLIP(mF) { OR32RtoR (gprT1, gprT2); AND32ItoR(gprT1, 0xffffff); - mVUallocCFLAGb(gprT1, fcInstance); + mVUallocCFLAGb(mVU, gprT1, fcInstance); } pass3 { mVUlog("CLIP"); mVUlogCLIP(); } } From cf1f7eda8cb1669e12bc818672e6f8580d0ff127 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 30 May 2009 08:33:29 +0000 Subject: [PATCH 155/296] microVU: - Fixed a bug with my xgkick delay code from r1282. Fixes GT4. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1292 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 4 ++-- pcsx2/x86/microVU_Flags.inl | 3 +-- pcsx2/x86/microVU_Lower.inl | 22 ++++++++++++---------- 3 files changed, 15 insertions(+), 14 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 857f001f72..09d4b46933 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -274,7 +274,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } - if (doXGKICK) { mVU_XGKICK_DELAY(mVU); } + if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } if (!isBdelay) { incPC(1); } else { @@ -374,7 +374,7 @@ eBitTemination: AND32ItoR (flagReg, 0x0fcf); OR32MtoR (flagReg, (uptr)&mVU->divFlag); } - if (doXGKICK) { mVU_XGKICK_DELAY(mVU); } + if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } // Do E-bit end stuff here mVUsetupRange(mVU, xPC - 8); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 09569e9240..a049daac59 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -78,10 +78,9 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking // Ensure last ~4+ instructions update mac flags - setCode(); for (int i = mVUcount; i > 0; i--, aCount++) { if (doStatus) { if (__Mac) { mVUinfo |= _doMac; } if (aCount >= 4) { break; } } - incPC(-2); + incPC2(-2); } // Status/Mac Flags Setup Code diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 7da82eec97..2b461c674b 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1061,24 +1061,26 @@ void __fastcall mVU_XGKICK__(u32 addr) { GSGIFTRANSFER1((u32*)microVU1.regs->Mem, ((addr<<4)&0x3fff)); } +microVUt(void) mVU_XGKICK_DELAY(mV, bool memVI) { + mVUbackupRegs(mVU); + if (memVI) MOV32MtoR(gprT2, (uptr)&mVU->VIxgkick); + else mVUallocVIa(mVU, gprT2, _Is_); + if (mtgsThread) CALLFunc((uptr)mVU_XGKICK_); + else CALLFunc((uptr)mVU_XGKICK__); + mVUrestoreRegs(mVU); +} + mVUop(mVU_XGKICK) { pass1 { mVUanalyzeXGkick(mVU, _Is_, mVU_XGKICK_CYCLES); } pass2 { - mVUprint("XGkick"); - mVUallocVIa(mVU, gprT1, _Is_); + if (!mVU_XGKICK_CYCLES) { mVU_XGKICK_DELAY(mVU, 0); return; } + else if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); mVUinfo &= ~_doXGKICK; } + mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->VIxgkick, gprT1); } pass3 { mVUlog("XGKICK vi%02d", _Fs_); } } -microVUt(void) mVU_XGKICK_DELAY(mV) { - MOV32MtoR(gprT2, (uptr)&mVU->VIxgkick); // gprT2 = ECX for __fastcall - mVUbackupRegs(mVU); - if (mtgsThread != NULL) CALLFunc((uptr)mVU_XGKICK_); - else CALLFunc((uptr)mVU_XGKICK__); - mVUrestoreRegs(mVU); -} - //------------------------------------------------------------------ // Branches/Jumps //------------------------------------------------------------------ From ccd89d199bd339595a8d5b03dd6ca8a517e4e8cc Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 30 May 2009 11:50:05 +0000 Subject: [PATCH 156/296] microVU/Linux: Since __mVUdumpProgram is of type microVUx, vuIndex is already defined, and shouldn't be redefined. Fixes Linux compilation. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1293 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Log.inl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 9e5451a8fd..4ef6ccbadf 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -46,7 +46,7 @@ microVUx(void) __mVUdumpProgram(int progIndex) { int bBranch = mVUbranch; int bCode = mVU->code; int bPC = iPC; - int vuIndex = (mVU == µVU1) ? 1 : 0; + vuIndex = (mVU == µVU1) ? 1 : 0; mVUbranch = 0; sprintf(str, "%s\\microVU%d prog - %02d.html", LOGS_DIR, vuIndex, progIndex); From 187dffc5f45ae3c56393c48f2d09e8ed7932ca9d Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 30 May 2009 12:11:21 +0000 Subject: [PATCH 157/296] Zeropad: Fixes a minor mistake in r1286. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1294 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zeropad/Linux/linux.cpp | 2 +- plugins/zeropad/joystick.cpp | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/plugins/zeropad/Linux/linux.cpp b/plugins/zeropad/Linux/linux.cpp index 55100bde96..1cbe3ab0d6 100644 --- a/plugins/zeropad/Linux/linux.cpp +++ b/plugins/zeropad/Linux/linux.cpp @@ -81,7 +81,7 @@ int _GetJoystickIdFromPAD(int pad) } } - if (JoystickIdWithinBounds(joyid)) + if (!JoystickIdWithinBounds(joyid)) { // get first unused joystick for (joyid = 0; joyid < s_vjoysticks.size(); ++joyid) diff --git a/plugins/zeropad/joystick.cpp b/plugins/zeropad/joystick.cpp index 1878abb18f..1b3822d1cb 100644 --- a/plugins/zeropad/joystick.cpp +++ b/plugins/zeropad/joystick.cpp @@ -71,7 +71,7 @@ const char *HatName(int value) bool JoystickIdWithinBounds(int joyid) { - return (!((joyid < 0) || (joyid >= (int)s_vjoysticks.size()))); + return ((joyid >= 0) && (joyid < (int)s_vjoysticks.size())); } // opens handles to all possible joysticks void JoystickInfo::EnumerateJoysticks(vector& vjoysticks) From 4fc4a634b3cd0ecd087b67424ffb1ac46ac2dd13 Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 30 May 2009 16:57:15 +0000 Subject: [PATCH 158/296] Fixing up Fatal Frame, removed redundant path3 masking code from the olden days! git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1295 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/GS.h | 1 - pcsx2/Gif.cpp | 25 +++++-------------------- pcsx2/HwWrite.cpp | 4 ++-- pcsx2/VifDma.cpp | 22 ++++++---------------- pcsx2/x86/iVUzerorec.cpp | 4 ++-- 5 files changed, 15 insertions(+), 41 deletions(-) diff --git a/pcsx2/GS.h b/pcsx2/GS.h index f70509cb8c..a90d937815 100644 --- a/pcsx2/GS.h +++ b/pcsx2/GS.h @@ -27,7 +27,6 @@ #include "Common.h" #include "Threading.h" -#define GSPATH3FIX PCSX2_ALIGNED16( extern u8 g_RealGSMem[0x2000] ); #define GSCSRr *((u64*)(g_RealGSMem+0x1000)) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index a30a2ad414..039cc437b1 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -57,7 +57,8 @@ __forceinline void gsInterrupt() { //Console::WriteLn("Eh? why are you still interrupting! chcr %x, qwc %x, done = %x", params gif->chcr, gif->qwc, done); return; } - + if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + if (gif->qwc > 0 || gspath3done == 0) { if (!(psHu32(DMAC_CTRL) & 0x1)) { Console::Notice("gs dma masked, re-scheduling..."); @@ -74,6 +75,7 @@ __forceinline void gsInterrupt() { gspath3done = 0; gscycles = 0; gif->chcr &= ~0x100; + psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x4000; //FIFO empty psHu32(GIF_STAT) &= ~GIF_STAT_P3Q; @@ -107,7 +109,6 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) memcpy_aligned(pgsmem, pMem, sizetoread<<4); mtgsThread->SendDataPacket(); - if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 return sizetoread; } else @@ -120,11 +121,8 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) } int _GIFchain() { -#ifdef GSPATH3FIX + u32 qwc = ((psHu32(GIF_MODE) & 0x4) && (vif1Regs->mskpath3)) ? min(8, (int)gif->qwc) : min( gifsplit, (int)gif->qwc ); -#else - u32 qwc = gif->qwc; -#endif u32 *pMem; pMem = (u32*)dmaGetAddr(gif->madr); @@ -181,18 +179,6 @@ void GIFdma() } - -#ifndef GSPATH3FIX - if ( !(psHu32(GIF_MODE) & 0x4) ) { - if (vif1Regs->mskpath3 || psHu32(GIF_MODE) & 0x1) { - gif->chcr &= ~0x100; - psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 - hwDmacIrq(2); - return; - } - } -#endif - if ((psHu32(DMAC_CTRL) & 0xC0) == 0x80 && prevcycles != 0) { // STD == GIF Console::WriteLn("GS Stall Control Source = %x, Drain = %x\n MADR = %x, STADR = %x", params (psHu32(0xe000) >> 4) & 0x3, (psHu32(0xe000) >> 6) & 0x3, gif->madr, psHu32(DMAC_STADR)); @@ -216,7 +202,6 @@ void GIFdma() return; } -#ifdef GSPATH3FIX if (vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1)) { if(gif->qwc == 0) { if((gif->chcr & 0x10e) == 0x104) { @@ -253,7 +238,7 @@ void GIFdma() GIFdmaEnd(); return; } -#endif + // Transfer Dn_QWC from Dn_MADR to GIF if ((gif->chcr & 0xc) == 0 || gif->qwc > 0) { // Normal Mode diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index b9a346bd95..88ff3cba91 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -97,6 +97,7 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) //Its invalid for the hardware to write a DMA while it is active, not without Suspending the DMAC if((value & 0x100) && (psHu32(mem) & 0x100) == 0x100 && (psHu32(DMAC_CTRL) & 0x1) == 1) { DMA_LOG( "DMAExec32 Attempt to run DMA while one is already active mem = %x", mem ); + return; } // Upper 16bits of QWC should not be written since QWC is 16bits in size. @@ -1001,9 +1002,8 @@ void __fastcall hwWrite64_page_03( u32 mem, const mem64_t* srcval ) case GIF_MODE: { -#ifdef GSPATH3FIX Console::Status("GIFMODE64 %x", params value); -#endif + psHu64(GIF_MODE) = value; // set/clear bits 0 and 2 as per the GIF_MODE value. diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 8b249cfdf6..10bbfa4b0e 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -1891,7 +1891,7 @@ static int __fastcall Vif1TransDirectHL(u32 *data) } else { - psHu32(GIF_STAT) &= ~GIF_STAT_APATH2; + psHu32(GIF_STAT) &= ~(GIF_STAT_APATH2 | GIF_STAT_OPH); ret = vif1.tag.size; vif1.tag.size = 0; vif1.cmd = 0; @@ -2017,7 +2017,6 @@ static void Vif1CMDMskPath3() // MSKPATH3 vif1Regs->mskpath3 = (vif1Regs->code >> 15) & 0x1; //Console::WriteLn("VIF MSKPATH3 %x", params vif1Regs->mskpath3); -#ifdef GSPATH3FIX if ((vif1Regs->code >> 15) & 0x1) { psHu32(GIF_STAT) |= 0x2; @@ -2027,18 +2026,6 @@ static void Vif1CMDMskPath3() // MSKPATH3 Path3progress = 1; //Let the Gif know it can transfer again (making sure any vif stall isnt unset prematurely) psHu32(GIF_STAT) &= ~0x2; } -#else - if (vif1Regs->mskpath3) - { - if (gif->qwc) _GIFchain(); // Finish the transfer first - psHu32(GIF_STAT) |= 0x2; - } - else - { - psHu32(GIF_STAT) &= ~0x2; - if (gif->qwc) _GIFchain(); // Finish the transfer first - } -#endif vif1.cmd &= ~0x7f; } @@ -2122,7 +2109,7 @@ static void Vif1CMDDirectHL() // DIRECT/HL } } - psHu32(GIF_STAT) |= GIF_STAT_APATH2; + psHu32(GIF_STAT) |= (GIF_STAT_APATH2 | GIF_STAT_OPH); } static void Vif1CMDNull() // invalid opcode @@ -2263,6 +2250,8 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size == 0) break; } } + if(!vif1.cmd) vif1Regs->stat &= ~VIF1_STAT_VPS_D; + if(vif1Regs->stat & VIF1_STAT_VGW) break; } // End of Transfer loop @@ -2555,9 +2544,10 @@ __forceinline void vif1Interrupt() } #ifdef PCSX2_DEVBUILD if (vif1ch->qwc > 0) Console::WriteLn("VIF1 Ending with %x QWC left"); - if (vif1.cmd != 0) Console::WriteLn("vif1.cmd still set %x", params vif1.cmd); + if (vif1.cmd != 0) Console::WriteLn("vif1.cmd still set %x tag size %x", params vif1.cmd, vif1.tag.size); #endif + vif1Regs->stat &= ~VIF1_STAT_VPS; //Vif goes idle as the stall happened between commands; vif1ch->chcr &= ~0x100; g_vifCycles = 0; hwDmacIrq(DMAC_VIF1); diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 105f6a61e5..dc499e432f 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -4414,7 +4414,7 @@ void recVUMI_XGKICK_(VURegs *VU) CALLFunc((uptr)GSgifTransfer1); #endif } - psHu32(GIF_STAT) &= ~0x100; + psHu32(GIF_STAT) &= ~(GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) s_ScheduleXGKICK = 0; } @@ -4435,7 +4435,7 @@ void recVUMI_XGKICK(VURegs *VU, int info) SHL32ItoR(isreg, 4); AND32ItoR(isreg, 0x3fff); s_XGKICKReg = isreg; - psHu32(GIF_STAT) |= 0x100; + psHu32(GIF_STAT) |= (GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) From 0028fefcbc091aa1a7171226ec1085d5da234c76 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 30 May 2009 18:54:51 +0000 Subject: [PATCH 159/296] Various VifDma cleanups. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1296 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 146 +++++++++++++++++++++++++++-------------------- pcsx2/VifDma.h | 55 ++++++++++++++++++ 2 files changed, 138 insertions(+), 63 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 10bbfa4b0e..e24adc47d9 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -339,8 +339,7 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int const VIFUnpackFuncTable *ft; VURegs * VU; u8 *cdata = (u8*)data; - - + #ifdef _DEBUG u32 memsize = VIFdmanum ? 0x4000 : 0x1000; #endif @@ -453,7 +452,6 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int VIFUNPACK_LOG("Aligning packet done size = %d offset %d addr %x", size, vifRegs->offset, vif->tag.addr); } } - if (vif->cl != 0 || (size & 0xf)) //Check alignment for SSE unpacks { @@ -544,7 +542,6 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int return size>>2; } - static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned int VIFdmanum) { u32 *dest; @@ -554,9 +551,10 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i VURegs * VU; u8 *cdata = (u8*)data; u32 tempsize = 0; + const u32 memlimit = (VIFdmanum ? 0x4000 : 0x1000); #ifdef _DEBUG - u32 memsize = VIFdmanum ? 0x4000 : 0x1000; + u32 memsize = memlimit; #endif _mm_prefetch((char*)data, _MM_HINT_NTA); @@ -608,10 +606,10 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i #ifdef _DEBUG static int s_count = 0; #endif - if(v->addr >= (u32)(VIFdmanum ? 0x4000 : 0x1000)) + if (v->addr >= memlimit) { //DevCon::Notice("Overflown at the start"); - v->addr &= (u32)(VIFdmanum ? 0x3fff : 0xfff); + v->addr &= (memlimit - 1); dest = (u32*)(VU->Mem + v->addr); } @@ -620,12 +618,11 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) + (tempsize * 16); //Sanity Check (memory overflow) - if(tempsize > (u32)(VIFdmanum ? 0x4000 : 0x1000)) + if (tempsize > memlimit) { - - - if(((vifRegs->cycle.cl != vifRegs->cycle.wl) && ((u32)(VIFdmanum ? 0x4000 : 0x1000) + ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) == tempsize) - || tempsize == (u32)(VIFdmanum ? 0x4000 : 0x1000)) + if (((vifRegs->cycle.cl != vifRegs->cycle.wl) && + (memlimit + ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) == tempsize) || + (tempsize == memlimit)) { //Its a red herring! so ignore it! SSE unpacks will be much quicker tempsize = 0; @@ -636,8 +633,11 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i tempsize = size; size = 0; } - } else tempsize = 0; - + } + else + { + tempsize = 0; + } if (size >= ft->gsize) { @@ -673,7 +673,8 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i } #endif - if ((vifRegs->cycle.cl == 0) || (vifRegs->cycle.wl == 0) || ((vifRegs->cycle.cl == vifRegs->cycle.wl) && !(vifRegs->code & 0x10000000))) + if ((vifRegs->cycle.cl == 0) || (vifRegs->cycle.wl == 0) || + ((vifRegs->cycle.cl == vifRegs->cycle.wl) && !(vifRegs->code & 0x10000000))) { oldcycle = *(u32*) & vifRegs->cycle; vifRegs->cycle.cl = vifRegs->cycle.wl = 1; @@ -745,9 +746,9 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i while ((tempsize >= ft->gsize) && (vifRegs->num > 0)) { - if(v->addr >= (u32)(VIFdmanum ? 0x4000 : 0x1000)) + if(v->addr >= memlimit) { - v->addr &= (u32)(VIFdmanum ? 0x3fff : 0xfff); + v->addr &= (memlimit - 1); dest = (u32*)(VU->Mem + v->addr); } @@ -778,9 +779,9 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i vifRow[2] = vifRegs->r2; vifRow[3] = vifRegs->r3; } - if(v->addr >= (u32)(VIFdmanum ? 0x4000 : 0x1000)) + if(v->addr >= memlimit) { - v->addr &= (u32)(VIFdmanum ? 0x3fff : 0xfff); + v->addr &= (memlimit - 1); dest = (u32*)(VU->Mem + v->addr); } if(tempsize > 0) size = tempsize; @@ -828,11 +829,13 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i VIF_LOG("Out of Filling write data"); break; } + func(dest, (u32*)cdata, ft->qsize); cdata += ft->gsize; size -= ft->gsize; vif->cl++; vifRegs->num--; + if (vif->cl == vifRegs->cycle.wl) { vif->cl = 0; @@ -866,6 +869,7 @@ static void vuExecMicro(u32 addr, const u32 VIFdmanum) VU = &VU1; vif1FLUSH(); } + if (VU->vifRegs->itops > (VIFdmanum ? 0x3ffu : 0xffu)) Console::WriteLn("VIF%d ITOP overrun! %x", params VIFdmanum, VU->vifRegs->itops); @@ -1060,19 +1064,24 @@ static int __fastcall Vif0TransMPG(u32 *data) // MPG if (vif0.vifpacketsize < vif0.tag.size) { if((vif0.tag.addr + vif0.vifpacketsize) > 0x1000) DevCon::Notice("Vif0 MPG Split Overflow"); + vif0mpgTransfer(vif0.tag.addr, data, vif0.vifpacketsize); vif0.tag.addr += vif0.vifpacketsize << 2; vif0.tag.size -= vif0.vifpacketsize; + return vif0.vifpacketsize; } else { int ret; + if((vif0.tag.addr + vif0.tag.size) > 0x1000) DevCon::Notice("Vif0 MPG Overflow"); + vif0mpgTransfer(vif0.tag.addr, data, vif0.tag.size); ret = vif0.tag.size; vif0.tag.size = 0; vif0.cmd = 0; + return ret; } } @@ -1090,10 +1099,13 @@ static int __fastcall Vif0TransUnpack(u32 *data) // UNPACK vif0.tag.size -= vif0.vifpacketsize - VIFalign(data, &vif0.tag, vif0.vifpacketsize, VIF0dmanum); ret = ret - vif0.tag.size; data += ret; + if(vif0.vifpacketsize > 0) VIFunpack(data, &vif0.tag, vif0.vifpacketsize - ret, VIF0dmanum); + ProcessMemSkip((vif0.vifpacketsize - ret) << 2, (vif0.cmd & 0xf), VIF0dmanum); vif0.tag.size -= (vif0.vifpacketsize - ret); FreezeXMMRegs(0); + return vif0.vifpacketsize; } /* size is less that the total size, transfer is 'in pieces' */ @@ -1391,10 +1403,9 @@ int _chainVIF0() { VIF_LOG("dmaIrq Set\n"); - vif0.done = true; - return (vif0.done)?1:0; //End Transfer + vif0.done = true; //End Transfer } - return (vif0.done)?1:0; //Return Done + return (vif0.done) ? 1: 0; //Return Done } void vif0Interrupt() @@ -1414,6 +1425,7 @@ void vif0Interrupt() vif0ch->chcr &= ~0x100; return; } + if (vif0ch->qwc > 0 || vif0.irqoffset > 0) { if (vif0.stallontag) @@ -1448,6 +1460,7 @@ void vif0Interrupt() if (vif0ch->qwc > 0) Console::WriteLn("VIF0 Ending with QWC left"); if (vif0.cmd != 0) Console::WriteLn("vif0.cmd still set %x", params vif0.cmd); + vif0ch->chcr &= ~0x100; hwDmacIrq(DMAC_VIF0); vif0Regs->stat &= ~0xF000000; // FQC=0 @@ -1508,6 +1521,7 @@ void dmaVIF0() vif0.vifstalled = true; return; } + vif0.done = true; CPU_INT(0, g_vifCycles); return; @@ -1522,7 +1536,7 @@ void vif0Write32(u32 mem, u32 value) { switch (mem) { - case 0x10003830: // MARK + case VIF0_MARK: VIF_LOG("VIF0_MARK write32 0x%8.8x", value); /* Clear mark flag in VIF0_STAT and set mark with 'value' */ @@ -1530,7 +1544,7 @@ void vif0Write32(u32 mem, u32 value) vif0Regs->mark = value; break; - case 0x10003810: // FBRST + case VIF0_FBRST: VIF_LOG("VIF0_FBRST write32 0x%8.8x", value); if (value & 0x1) @@ -1546,6 +1560,7 @@ void vif0Write32(u32 mem, u32 value) vif0Regs->err = 0; vif0Regs->stat &= ~(0xF000000 | VIF0_STAT_INT | VIF0_STAT_VSS | VIF0_STAT_VIS | VIF0_STAT_VFS | VIF0_STAT_VPS); // FQC=0 } + if (value & 0x2) { /* Force Break the VIF */ @@ -1555,7 +1570,8 @@ void vif0Write32(u32 mem, u32 value) vif0Regs->stat &= ~VIF0_STAT_VPS; vif0.vifstalled = true; Console::WriteLn("vif0 force break"); - } + } + if (value & 0x4) { /* Stop VIF */ @@ -1565,6 +1581,7 @@ void vif0Write32(u32 mem, u32 value) vif0Regs->stat &= ~VIF0_STAT_VPS; vif0.vifstalled = true; } + if (value & 0x8) { bool cancel = false; @@ -1594,7 +1611,7 @@ void vif0Write32(u32 mem, u32 value) } break; - case 0x10003820: + case VIF0_ERR: // ERR VIF_LOG("VIF0_ERR write32 0x%8.8x", value); @@ -1604,11 +1621,11 @@ void vif0Write32(u32 mem, u32 value) default: Console::WriteLn("Unknown Vif0 write to %x", params mem); - if (mem >= 0x10003900 && mem < 0x10003980) + if (mem >= VIF0_R0 && mem < 0x10003980) // mem <= VIF0_C3? { assert((mem&0xf) == 0); - if (mem < 0x10003940) + if (mem < VIF0_C0) g_vifRow0[(mem>>4)&3] = value; else g_vifCol0[(mem>>4)&3] = value; @@ -1974,18 +1991,19 @@ static int __fastcall Vif1TransUnpack(u32 *data) //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// // Vif1 CMD Base Commands //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// -// fixme: Global variables should not have the same name as local variables, and this probably shouldn't be local anyways. -int transferred = 0; + static void Vif1CMDNop() // NOP { vif1.cmd &= ~0x7f; } + static void Vif1CMDSTCycl() // STCYCL { vif1Regs->cycle.cl = (u8)vif1Regs->code; vif1Regs->cycle.wl = (u8)(vif1Regs->code >> 8); vif1.cmd &= ~0x7f; } + static void Vif1CMDOffset() // OFFSET { vif1Regs->ofst = vif1Regs->code & 0x3ff; @@ -1993,11 +2011,13 @@ static void Vif1CMDOffset() // OFFSET vif1Regs->tops = vif1Regs->base; vif1.cmd &= ~0x7f; } + static void Vif1CMDBase() // BASE { vif1Regs->base = vif1Regs->code & 0x3ff; vif1.cmd &= ~0x7f; } + static void Vif1CMDITop() // ITOP { vif1Regs->itops = vif1Regs->code & 0x3ff; @@ -2010,10 +2030,8 @@ static void Vif1CMDSTMod() // STMOD vif1.cmd &= ~0x7f; } - static void Vif1CMDMskPath3() // MSKPATH3 { - vif1Regs->mskpath3 = (vif1Regs->code >> 15) & 0x1; //Console::WriteLn("VIF MSKPATH3 %x", params vif1Regs->mskpath3); @@ -2036,11 +2054,8 @@ static void Vif1CMDMark() // MARK vif1.cmd &= ~0x7f; } - - static void Vif1CMDFlush() // FLUSH/E/A { - vif1FLUSH(); if((vif1.cmd & 0x7f) == 0x13) @@ -2053,21 +2068,25 @@ static void Vif1CMDFlush() // FLUSH/E/A vif1.cmd &= ~0x7f; } + static void Vif1CMDMSCALF() //MSCAL/F { vif1FLUSH(); vuExecMicro((u16)(vif1Regs->code) << 3, VIF1dmanum); vif1.cmd &= ~0x7f; } + static void Vif1CMDMSCNT() // MSCNT { vuExecMicro(-1, VIF1dmanum); vif1.cmd &= ~0x7f; } + static void Vif1CMDSTMask() // STMASK { vif1.tag.size = 1; } + static void Vif1CMDSTRowCol() // STROW / STCOL { vif1.tag.addr = 0; @@ -2086,8 +2105,6 @@ static void Vif1CMDMPGTransfer() // MPG vif1.tag.size = vifNum * 2; } - - static void Vif1CMDDirectHL() // DIRECT/HL { int vifImm; @@ -2110,8 +2127,8 @@ static void Vif1CMDDirectHL() // DIRECT/HL } psHu32(GIF_STAT) |= (GIF_STAT_APATH2 | GIF_STAT_OPH); - } + static void Vif1CMDNull() // invalid opcode { // if ME1, then force the vif to interrupt @@ -2170,7 +2187,7 @@ void (*Vif1CMDTLB[82])() = int VIF1transfer(u32 *data, int size, int istag) { int ret; - transferred = vif1.vifstalled ? vif1.irqoffset : 0; // irqoffset necessary to add up the right qws, or else will spin (spiderman) + int transferred = vif1.vifstalled ? vif1.irqoffset : 0; // irqoffset necessary to add up the right qws, or else will spin (spiderman) VIF_LOG("VIF1transfer: size %x (vif1.cmd %x)", size, vif1.cmd); @@ -2191,7 +2208,9 @@ int VIF1transfer(u32 *data, int size, int istag) } } + if(vif1Regs->stat & VIF1_STAT_VGW) break; + if (vif1.cmd) { vif1Regs->stat |= VIF1_STAT_VPS_T; //Decompression has started @@ -2263,7 +2282,7 @@ int VIF1transfer(u32 *data, int size, int istag) { vif1.vifstalled = true; - if (((vif1Regs->code >> 24) & 0x7f) != 0x7)vif1Regs->stat |= VIF1_STAT_VIS; // Note: commenting this out fixes WALL-E + if (((vif1Regs->code >> 24) & 0x7f) != 0x7) vif1Regs->stat |= VIF1_STAT_VIS; // Note: commenting this out fixes WALL-E if (vif1ch->qwc == 0 && (vif1.irqoffset == 0 || istag == 1)) vif1.inprogress = 0; @@ -2326,6 +2345,7 @@ void vif1TransferFromMemory() // completely and execute the transfer there-after. FreezeXMMRegs(1); + if (GSreadFIFO2 == NULL) { for (size = vif1ch->qwc; size > 0; --size) @@ -2349,8 +2369,8 @@ void vif1TransferFromMemory() psHu64(0x5000) = pMem[2*vif1ch->qwc-2]; psHu64(0x5008) = pMem[2*vif1ch->qwc-1]; } + FreezeXMMRegs(0); - g_vifCycles += vif1ch->qwc * 2; vif1ch->madr += vif1ch->qwc * 16; // mgs3 scene changes @@ -2368,7 +2388,7 @@ int _VIF1chain() return 0; } - if (vif1.dmamode == 1) + if (vif1.dmamode == VIF_NORMAL_MEM_MODE) { vif1TransferFromMemory(); vif1.inprogress = 0; @@ -2398,14 +2418,14 @@ __forceinline void vif1SetupTransfer() { switch (vif1.dmamode) { - case 0: //Normal - case 1: //Normal (From memory) + case VIF_NORMAL_MODE: //Normal + case VIF_NORMAL_MEM_MODE: //Normal (From memory) vif1.inprogress = 1; vif1.done = true; g_vifCycles = 2; break; - case 2: //Chain + case VIF_CHAIN_MODE: //Chain int id; int ret; @@ -2490,7 +2510,6 @@ __forceinline void vif1Interrupt() } - if ((vif1ch->chcr & 0x100) == 0) Console::WriteLn("Vif1 running when CHCR == %x", params vif1ch->chcr); if (vif1.irq && vif1.tag.size == 0) @@ -2556,7 +2575,6 @@ __forceinline void vif1Interrupt() void dmaVIF1() { - VIF_LOG("dmaVIF1 chcr = %lx, madr = %lx, qwc = %lx\n" " tadr = %lx, asr0 = %lx, asr1 = %lx", vif1ch->chcr, vif1ch->madr, vif1ch->qwc, @@ -2564,7 +2582,6 @@ void dmaVIF1() g_vifCycles = 0; vif1.inprogress = 0; - if (((psHu32(DMAC_CTRL) & 0xC) == 0x8)) // VIF MFIFO { @@ -2581,7 +2598,6 @@ void dmaVIF1() } #endif - if (!(vif1ch->chcr & 0x4) || vif1ch->qwc > 0) // Normal Mode { @@ -2589,29 +2605,30 @@ void dmaVIF1() Console::WriteLn("DMA Stall Control on VIF1 normal"); if ((vif1ch->chcr & 0x1)) // to Memory - vif1.dmamode = 0; + vif1.dmamode = VIF_NORMAL_MODE; else - vif1.dmamode = 1; + vif1.dmamode = VIF_NORMAL_MEM_MODE; } else { - vif1.dmamode = 2; + vif1.dmamode = VIF_CHAIN_MODE; } - if(vif1.dmamode != 1)vif1Regs->stat |= 0x10000000; // FQC=16 - else vif1Regs->stat |= min((u16)16, vif1ch->qwc) << 24; // FQC=16 + if(vif1.dmamode != VIF_NORMAL_MEM_MODE) + vif1Regs->stat |= 0x10000000; // FQC=16 + else + vif1Regs->stat |= min((u16)16, vif1ch->qwc) << 24; // FQC=16 // Chain Mode vif1.done = false; vif1Interrupt(); } - void vif1Write32(u32 mem, u32 value) { switch (mem) { - case 0x10003c30: // MARK + case VIF1_MARK: VIF_LOG("VIF1_MARK write32 0x%8.8x", value); /* Clear mark flag in VIF1_STAT and set mark with 'value' */ @@ -2619,7 +2636,7 @@ void vif1Write32(u32 mem, u32 value) vif1Regs->mark = value; break; - case 0x10003c10: // FBRST + case VIF1_FBRST: // FBRST VIF_LOG("VIF1_FBRST write32 0x%8.8x", value); if (value & 0x1) @@ -2635,6 +2652,7 @@ void vif1Write32(u32 mem, u32 value) vif1.inprogress = 0; vif1Regs->stat &= ~(0x1F800000 | VIF1_STAT_INT | VIF1_STAT_VSS | VIF1_STAT_VIS | VIF1_STAT_VFS | VIF1_STAT_VPS); // FQC=0 } + if (value & 0x2) { /* Force Break the VIF */ @@ -2645,6 +2663,7 @@ void vif1Write32(u32 mem, u32 value) vif1.vifstalled = true; Console::WriteLn("vif1 force break"); } + if (value & 0x4) { /* Stop VIF */ @@ -2655,6 +2674,7 @@ void vif1Write32(u32 mem, u32 value) cpuRegs.interrupt &= ~((1 << 1) | (1 << 10)); //Stop all vif1 DMA's vif1.vifstalled = true; } + if (value & 0x8) { bool cancel = false; @@ -2690,15 +2710,15 @@ void vif1Write32(u32 mem, u32 value) } break; - case 0x10003c20: // ERR + case VIF1_ERR: // ERR VIF_LOG("VIF1_ERR write32 0x%8.8x", value); /* Set VIF1_ERR with 'value' */ vif1Regs->err = value; break; - case 0x10003c00: // STAT - VIF_LOG("VIF1_STAT write32 0x%8.8x", value); + case VIF1_STAT: // STAT + VIF_LOG("VIF1_STAT write32 0x%8.8x", value); #ifdef PCSX2_DEVBUILD /* Only FDR bit is writable, so mask the rest */ @@ -2726,16 +2746,16 @@ void vif1Write32(u32 mem, u32 value) } break; - case 0x10003c50: // MODE + case VIF1_MODE: // MODE vif1Regs->mode = value; break; default: Console::WriteLn("Unknown Vif1 write to %x", params mem); - if ((mem >= 0x10003d00) && (mem < 0x10003d80)) + if ((mem >= VIF1_R0) && (mem < 0x10003d80)) // mem <= VIF1_C3? { assert((mem&0xf) == 0); - if (mem < 0x10003d40) + if (mem < VIF1_C0) g_vifRow1[(mem>>4)&3] = value; else g_vifCol1[(mem>>4)&3] = value; diff --git a/pcsx2/VifDma.h b/pcsx2/VifDma.h index f210d49e2c..86a1d1f78d 100644 --- a/pcsx2/VifDma.h +++ b/pcsx2/VifDma.h @@ -18,6 +18,61 @@ #ifndef __VIFDMA_H__ #define __VIFDMA_H__ +enum VifModes +{ + VIF_NORMAL_MODE = 0, + VIF_NORMAL_MEM_MODE = 1, + VIF_CHAIN_MODE = 2 +}; + +enum VifMemoryLocations +{ + VIF0_STAT = 0x10003800, + VIF0_FBRST = 0x10003810, + VIF0_ERR = 0x10003820, + VIF0_MARK = 0x10003830, + VIF0_CYCLE = 0x10003840, + VIF0_MODE = 0x10003850, + VIF0_NUM = 0x10003860, + VIF0_MASK = 0x10003870, + VIF0_CODE = 0x10003880, + VIF0_ITOPS = 0x10003890, + VIF0_ITOP = 0x100038d0, + VIF0_TOP = 0x100038e0, + VIF0_R0 = 0x10003900, + VIF0_R1 = 0x10003910, + VIF0_R2 = 0x10003920, + VIF0_R3 = 0x10003930, + VIF0_C0 = 0x10003940, + VIF0_C1 = 0x10003950, + VIF0_C2 = 0x10003960, + VIF0_C3 = 0x10003970, + + VIF1_STAT = 0x10003c00, + VIF1_FBRST = 0x10003c10, + VIF1_ERR = 0x10003c20, + VIF1_MARK = 0x10003c30, + VIF1_CYCLE = 0x10003c40, + VIF1_MODE = 0x10003c50, + VIF1_NUM = 0x10003c60, + VIF1_MASK = 0x10003c70, + VIF1_CODE = 0x10003c80, + VIF1_ITOPS = 0x10003c90, + VIF1_BASE = 0x10003ca0, + VIF1_OFST = 0x10003cb0, + VIF1_TOPS = 0x10003cc0, + VIF1_ITOP = 0x10003cd0, + VIF1_TOP = 0x10003ce0, + VIF1_R0 = 0x10003d00, + VIF1_R1 = 0x10003d10, + VIF1_R2 = 0x10003d20, + VIF1_R3 = 0x10003d30, + VIF1_C0 = 0x10003d40, + VIF1_C1 = 0x10003d50, + VIF1_C2 = 0x10003d60, + VIF1_C3 = 0x10003d70 +}; + struct vifCode { u32 addr; u32 size; From 1694eecf2f75de8dfe81dbc77de56f6117550e0f Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 30 May 2009 22:25:23 +0000 Subject: [PATCH 160/296] Assorted cleanup involving VifDma, the Hw files, IopMem and Fifo. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1297 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/FiFo.cpp | 16 +-- pcsx2/Hw.h | 71 ++++++++++- pcsx2/HwRead.cpp | 55 ++++++--- pcsx2/HwWrite.cpp | 307 +++++++++++++++++++++++++++++----------------- pcsx2/IopDma.cpp | 14 +-- pcsx2/IopMem.cpp | 50 ++++---- pcsx2/Sif.cpp | 14 +-- pcsx2/VifDma.cpp | 81 ++++++------ pcsx2/VifDma.h | 48 -------- 9 files changed, 394 insertions(+), 262 deletions(-) diff --git a/pcsx2/FiFo.cpp b/pcsx2/FiFo.cpp index 92756317d5..2b0a022404 100644 --- a/pcsx2/FiFo.cpp +++ b/pcsx2/FiFo.cpp @@ -53,7 +53,7 @@ extern int FOreadpos; void __fastcall ReadFIFO_page_4(u32 mem, u64 *out) { - jASSUME( (mem >= 0x10004000) && (mem < 0x10005000) ); + jASSUME( (mem >= VIF0_FIFO) && (mem < VIF1_FIFO) ); VIF_LOG("ReadFIFO/VIF0 0x%08X", mem); //out[0] = psHu64(mem ); @@ -65,7 +65,7 @@ void __fastcall ReadFIFO_page_4(u32 mem, u64 *out) void __fastcall ReadFIFO_page_5(u32 mem, u64 *out) { - jASSUME( (mem >= 0x10005000) && (mem < 0x10006000) ); + jASSUME( (mem >= VIF1_FIFO) && (mem < GIF_FIFO) ); VIF_LOG("ReadFIFO/VIF1, addr=0x%08X", mem); @@ -87,7 +87,7 @@ void __fastcall ReadFIFO_page_5(u32 mem, u64 *out) void __fastcall ReadFIFO_page_6(u32 mem, u64 *out) { - jASSUME( (mem >= 0x10006000) && (mem < 0x10007000) ); + jASSUME( (mem >= GIF_FIFO) && (mem < IPUout_FIFO) ); DevCon::Notice( "ReadFIFO/GIF, addr=0x%x", params mem ); @@ -100,7 +100,7 @@ void __fastcall ReadFIFO_page_6(u32 mem, u64 *out) void __fastcall ReadFIFO_page_7(u32 mem, u64 *out) { - jASSUME( (mem >= 0x10007000) && (mem < 0x10008000) ); + jASSUME( (mem >= IPUout_FIFO) && (mem < D0_CHCR) ); // All addresses in this page map to 0x7000 and 0x7010: mem &= 0x10; @@ -125,7 +125,7 @@ void __fastcall ReadFIFO_page_7(u32 mem, u64 *out) void __fastcall WriteFIFO_page_4(u32 mem, const mem128_t *value) { - jASSUME( (mem >= 0x10004000) && (mem < 0x10005000) ); + jASSUME( (mem >= VIF0_FIFO) && (mem < VIF1_FIFO) ); VIF_LOG("WriteFIFO/VIF0, addr=0x%08X", mem); @@ -142,7 +142,7 @@ void __fastcall WriteFIFO_page_4(u32 mem, const mem128_t *value) void __fastcall WriteFIFO_page_5(u32 mem, const mem128_t *value) { - jASSUME( (mem >= 0x10005000) && (mem < 0x10006000) ); + jASSUME( (mem >= VIF1_FIFO) && (mem < GIF_FIFO) ); VIF_LOG("WriteFIFO/VIF1, addr=0x%08X", mem); @@ -164,7 +164,7 @@ void __fastcall WriteFIFO_page_5(u32 mem, const mem128_t *value) void __fastcall WriteFIFO_page_6(u32 mem, const mem128_t *value) { - jASSUME( (mem >= 0x10006000) && (mem < 0x10007000) ); + jASSUME( (mem >= GIF_FIFO) && (mem < IPUout_FIFO) ); GIF_LOG("WriteFIFO/GIF, addr=0x%08X", mem); //psHu64(mem ) = value[0]; @@ -192,7 +192,7 @@ void __fastcall WriteFIFO_page_6(u32 mem, const mem128_t *value) void __fastcall WriteFIFO_page_7(u32 mem, const mem128_t *value) { - jASSUME( (mem >= 0x10007000) && (mem < 0x10008000) ); + jASSUME( (mem >= IPUout_FIFO) && (mem < D0_CHCR) ); // All addresses in this page map to 0x7000 and 0x7010: mem &= 0x10; diff --git a/pcsx2/Hw.h b/pcsx2/Hw.h index 480a22ef37..0b0dc0cb0e 100644 --- a/pcsx2/Hw.h +++ b/pcsx2/Hw.h @@ -99,6 +99,55 @@ enum HWaddress GIF_CNT = 0x10003080, GIF_P3CNT = 0x10003090, GIF_P3TAG = 0x100030A0, + + // Vif Memory Locations + VIF0_STAT = 0x10003800, + VIF0_FBRST = 0x10003810, + VIF0_ERR = 0x10003820, + VIF0_MARK = 0x10003830, + VIF0_CYCLE = 0x10003840, + VIF0_MODE = 0x10003850, + VIF0_NUM = 0x10003860, + VIF0_MASK = 0x10003870, + VIF0_CODE = 0x10003880, + VIF0_ITOPS = 0x10003890, + VIF0_ITOP = 0x100038d0, + VIF0_TOP = 0x100038e0, + VIF0_R0 = 0x10003900, + VIF0_R1 = 0x10003910, + VIF0_R2 = 0x10003920, + VIF0_R3 = 0x10003930, + VIF0_C0 = 0x10003940, + VIF0_C1 = 0x10003950, + VIF0_C2 = 0x10003960, + VIF0_C3 = 0x10003970, + + VIF1_STAT = 0x10003c00, + VIF1_FBRST = 0x10003c10, + VIF1_ERR = 0x10003c20, + VIF1_MARK = 0x10003c30, + VIF1_CYCLE = 0x10003c40, + VIF1_MODE = 0x10003c50, + VIF1_NUM = 0x10003c60, + VIF1_MASK = 0x10003c70, + VIF1_CODE = 0x10003c80, + VIF1_ITOPS = 0x10003c90, + VIF1_BASE = 0x10003ca0, + VIF1_OFST = 0x10003cb0, + VIF1_TOPS = 0x10003cc0, + VIF1_ITOP = 0x10003cd0, + VIF1_TOP = 0x10003ce0, + VIF1_R0 = 0x10003d00, + VIF1_R1 = 0x10003d10, + VIF1_R2 = 0x10003d20, + VIF1_R3 = 0x10003d30, + VIF1_C0 = 0x10003d40, + VIF1_C1 = 0x10003d50, + VIF1_C2 = 0x10003d60, + VIF1_C3 = 0x10003d70, + + VIF0_FIFO = 0x10004000, + VIF1_FIFO = 0x10005000, GIF_FIFO = 0x10006000, IPUout_FIFO = 0x10007000, @@ -150,6 +199,7 @@ enum HWaddress D6_CHCR = 0x1000C400, D6_MADR = 0x1000C410, D6_QWC = 0x1000C420, + D6_TADR = 0x1000C430, //SIF2 D7_CHCR = 0x1000C800, @@ -161,6 +211,7 @@ enum HWaddress D8_MADR = 0x1000D010, D8_QWC = 0x1000D020, D8_SADR = 0x1000D080, + SPR1_CHCR = 0x1000D400, DMAC_CTRL = 0x1000E000, DMAC_STAT = 0x1000E010, @@ -173,10 +224,26 @@ enum HWaddress INTC_STAT = 0x1000F000, INTC_MASK = 0x1000F010, - SBUS_F220 = 0x1000F220, - SBUS_SMFLG = 0x1000F230, + SIO_LCR = 0x1000F100, + SIO_LSR = 0x1000F110, + SIO_IER = 0x1000F120, + SIO_ISR = 0x1000F130,// + SIO_FCR = 0x1000F140, + SIO_BGR = 0x1000F150, + SIO_TXFIFO = 0x1000F180, + SIO_RXFIFO = 0x1000F1C0, + + SBUS_F200 = 0x1000F200, //MSCOM + SBUS_F210 = 0x1000F210, //SMCOM + SBUS_F220 = 0x1000F220, //MSFLG + SBUS_F230 = 0x1000F230, //SMFLG SBUS_F240 = 0x1000F240, + SBUS_F250 = 0x1000F250, + SBUS_F260 = 0x1000F260, + MCH_RICM = 0x1000F430, + MCH_DRD = 0x1000F440, + DMAC_ENABLER = 0x1000F520, DMAC_ENABLEW = 0x1000F590, diff --git a/pcsx2/HwRead.cpp b/pcsx2/HwRead.cpp index 60a9eb6ab5..d1ca5c3b67 100644 --- a/pcsx2/HwRead.cpp +++ b/pcsx2/HwRead.cpp @@ -56,7 +56,7 @@ __forceinline mem8_t hwRead8(u32 mem) { u8 ret; - if( mem >= 0x10002000 && mem < 0x10008000 ) + if( mem >= IPU_CMD && mem < D0_CHCR ) DevCon::Notice("Unexpected hwRead8 from 0x%x", params mem); switch (mem) @@ -95,14 +95,23 @@ __forceinline mem8_t hwRead8(u32 mem) case 0x10001821: ret = (u8)(counters[3].target>>8); break; default: - if ((mem & 0xffffff0f) == 0x1000f200) + if ((mem & 0xffffff0f) == SBUS_F200) { - if(mem == 0x1000f260) ret = 0; - else if(mem == SBUS_F240) { - ret = psHu32(mem); - //psHu32(mem) &= ~0x4000; + switch (mem) + { + case SBUS_F240: + ret = psHu32(mem); + //psHu32(mem) &= ~0x4000; + break; + + case SBUS_F260: + ret = 0; + break; + + default: + ret = psHu32(mem); + break; } - else ret = psHu32(mem); return (u8)ret; } @@ -145,15 +154,23 @@ __forceinline mem16_t hwRead16(u32 mem) case RCNT3_TARGET: ret = (u16)counters[3].target; break; default: - if ((mem & 0xffffff0f) == 0x1000f200) + if ((mem & 0xffffff0f) == SBUS_F200) { - if(mem == 0x1000f260) ret = 0; - else if(mem == SBUS_F240) { - ret = psHu16(mem) | 0x0102; - psHu32(mem) &= ~0x4000; + switch (mem) + { + case SBUS_F240: + ret = psHu16(mem) | 0x0102; + psHu32(mem) &= ~0x4000; // not commented out like in bit mode? + break; + + case SBUS_F260: + ret = 0; + break; + + default: + ret = psHu32(mem); + break; } - else - ret = psHu32(mem); return (u16)ret; } ret = psHu16(mem); @@ -231,16 +248,16 @@ static __forceinline mem32_t __hwRead32_page_0F( u32 mem, bool intchack ) HW_LOG("INTC_MASK Read32, value=0x%x", psHu32(INTC_MASK)); break; - case 0xf130: // 0x1000f130 - case 0xf260: // 0x1000f260 SBUS? + case 0xf130: // SIO_ISR + case 0xf260: // SBUS_F260 case 0xf410: // 0x1000f410 case 0xf430: // MCH_RICM return 0; - case 0xf240: // 0x1000f240: SBUS + case 0xf240: // SBUS_F240 return psHu32(0xf240) | 0xF0000102; - case 0xf440: // 0x1000f440: MCH_DRD + case 0xf440: // MCH_DRD if( !((psHu32(0xf430) >> 6) & 0xF) ) { @@ -362,7 +379,7 @@ void __fastcall hwRead64_page_02(u32 mem, mem64_t* result ) void __fastcall hwRead64_generic_INTC_HACK(u32 mem, mem64_t* result ) { - if( mem == INTC_STAT ) IntCHackCheck(); + if (mem == INTC_STAT) IntCHackCheck(); *result = psHu64(mem); HW_LOG("Unknown Hardware Read 64 at %x",mem); diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index 88ff3cba91..f715b4e67b 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -43,7 +43,8 @@ static __forceinline void DmaExec8( void (*func)(), u32 mem, u8 value ) u32 qwcRegister = (mem | 0x20) & ~0x1; //Need to remove the lower bit else we end up clearing TADR //Its invalid for the hardware to write a DMA while it is active, not without Suspending the DMAC - if((value & 0x1) && (psHu8(mem) & 0x1) == 0x1 && (psHu32(DMAC_CTRL) & 0x1) == 1) { + if ((value & 0x1) && ((psHu8(mem) & 0x1) == 0x1) && ((psHu32(DMAC_CTRL) & 0x1) == 1)) + { DMA_LOG( "DMAExec8 Attempt to run DMA while one is already active mem = %x", mem ); } @@ -69,7 +70,8 @@ static __forceinline void DmaExec16( void (*func)(), u32 mem, u16 value ) u32 qwcRegister = mem | 0x20; //Its invalid for the hardware to write a DMA while it is active, not without Suspending the DMAC - if((value & 0x100) && (psHu32(mem) & 0x100) == 0x100 && (psHu32(DMAC_CTRL) & 0x1) == 1) { + if ((value & 0x100) && ((psHu32(mem) & 0x100) == 0x100) && ((psHu32(DMAC_CTRL) & 0x1) == 1)) + { DMA_LOG( "DMAExec16 Attempt to run DMA while one is already active mem = %x", mem); } @@ -95,7 +97,8 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) u32 qwcRegister = mem | 0x20; //Its invalid for the hardware to write a DMA while it is active, not without Suspending the DMAC - if((value & 0x100) && (psHu32(mem) & 0x100) == 0x100 && (psHu32(DMAC_CTRL) & 0x1) == 1) { + if ((value & 0x100) && ((psHu32(mem) & 0x100) == 0x100) && ((psHu32(DMAC_CTRL) & 0x1) == 1)) + { DMA_LOG( "DMAExec32 Attempt to run DMA while one is already active mem = %x", mem ); return; } @@ -110,8 +113,8 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) } /* Keep the old tag if in chain mode and hw doesnt set it*/ - if( (value & 0xc) == 0x4 && (value & 0xffff0000) == 0) - psHu32(mem) = (psHu32(mem) & 0xFFFF0000) | (u16)value; + if (((value & 0xc) == 0x4) && ((value & 0xffff0000) == 0)) + psHu32(mem) = (psHu32(mem) & 0xffff0000) | (u16)value; else /* Else (including Normal mode etc) write whatever the hardware sends*/ psHu32(mem) = (u32)value; @@ -127,14 +130,15 @@ char sio_buffer[1024]; int sio_count; u16 QueuedDMA = 0; -void hwWrite8(u32 mem, u8 value) { - - if( (mem>=0x10003800) && (mem<0x10004000) ) +void hwWrite8(u32 mem, u8 value) +{ + if ((mem >= VIF0_STAT) && (mem < VIF0_FIFO)) { u32 bytemod = mem & 0x3; u32 bitpos = 8 * bytemod; u32 newval = psHu8(mem) & (255UL << bitpos); - if( mem < 0x10003c00 ) + + if (mem < VIF1_STAT) vif0Write32( mem & ~0x3, newval | (value<= 0x10002000 && mem < 0x10008000 ) + if( mem >= IPU_CMD && mem < D0_CHCR ) DevCon::Notice( "hwWrite8 to 0x%x = 0x%x", params mem, value ); switch (mem) { @@ -168,14 +172,12 @@ void hwWrite8(u32 mem, u8 value) { case 0x10001811: rcntWmode(3, (counters[3].modeval & 0xff) | value << 8); break; case RCNT3_TARGET: rcntWtarget(3, value); break; - case 0x1000f180: + case SIO_TXFIFO: { - //bool flush = false; - // Terminate lines on CR or full buffers, and ignore \n's if the string contents // are empty (otherwise terminate on \n too!) - if( ( value == '\r' ) || ( sio_count == 1023 ) || - ( value == '\n' && sio_count != 0 ) ) + if (( value == '\r' ) || ( sio_count == 1023 ) || + ( value == '\n' && sio_count != 0 )) { sio_buffer[sio_count] = 0; Console::WriteLn( Color_Cyan, sio_buffer ); @@ -293,28 +295,48 @@ void hwWrite8(u32 mem, u8 value) { DmaExec8(dmaSPR1, mem, value); break; - case 0x1000f592: // DMAC_ENABLEW + case 0x1000f592: // DMAC_ENABLEW + 2 psHu8(0xf592) = value; psHu8(0xf522) = value; break; - case 0x1000f200: // SIF(?) + case SBUS_F200: // SIF(?) psHu8(mem) = value; break; - case 0x1000f240:// SIF(?) - if(!(value & 0x100)) - psHu32(mem) &= ~0x100; + case SBUS_F210: + psHu8(mem) = value; + break; + + case SBUS_F220: + psHu8(mem) = value; + break; + + case SBUS_F230: + psHu8(mem) = value; + break; + + case SBUS_F240:// SIF(?) + if (!(value & 0x100)) psHu32(mem) &= ~0x100; + break; + + case SBUS_F250: + psHu8(mem) = value; + break; + + case SBUS_F260: + psHu8(mem) = value; break; default: assert( (mem&0xff0f) != 0xf200 ); switch(mem&~3) { - case 0x1000f130: + case SIO_ISR: case 0x1000f410: - case 0x1000f430: + case MCH_RICM: break; + default: psHu8(mem) = value; } @@ -325,7 +347,7 @@ void hwWrite8(u32 mem, u8 value) { __forceinline void hwWrite16(u32 mem, u16 value) { - if( mem >= 0x10002000 && mem < 0x10008000 ) + if( mem >= IPU_CMD && mem < D0_CHCR ) Console::Notice( "hwWrite16 to %x", params mem ); switch(mem) @@ -374,22 +396,27 @@ __forceinline void hwWrite16(u32 mem, u16 value) HW_LOG("VIF1dma Madr %lx", value); psHu16(mem) = value;//dma1 madr break; + case D1_QWC: // dma1 - vif1 - qwc HW_LOG("VIF1dma QWC %lx", value); psHu16(mem) = value;//dma1 qwc break; + case D1_TADR: // dma1 - vif1 - tadr HW_LOG("VIF1dma TADR %lx", value); psHu16(mem) = value;//dma1 tadr break; + case D1_ASR0: // dma1 - vif1 - asr0 HW_LOG("VIF1dma ASR0 %lx", value); psHu16(mem) = value;//dma1 asr0 break; + case D1_ASR1: // dma1 - vif1 - asr1 HW_LOG("VIF1dma ASR1 %lx", value); psHu16(mem) = value;//dma1 asr1 break; + case D1_SADR: // dma1 - vif1 - sadr HW_LOG("VIF1dma SADR %lx", value); psHu16(mem) = value;//dma1 sadr @@ -412,22 +439,27 @@ __forceinline void hwWrite16(u32 mem, u16 value) psHu16(mem) = value;//dma2 madr HW_LOG("Hardware write DMA2_MADR 32bit at %x with value %x",mem,value); break; + case D2_QWC: psHu16(mem) = value;//dma2 qwc HW_LOG("Hardware write DMA2_QWC 32bit at %x with value %x",mem,value); break; + case D2_TADR: psHu16(mem) = value;//dma2 taddr HW_LOG("Hardware write DMA2_TADDR 32bit at %x with value %x",mem,value); break; + case D2_ASR0: psHu16(mem) = value;//dma2 asr0 HW_LOG("Hardware write DMA2_ASR0 32bit at %x with value %x",mem,value); break; + case D2_ASR1: psHu16(mem) = value;//dma2 asr1 HW_LOG("Hardware write DMA2_ASR1 32bit at %x with value %x",mem,value); break; + case D2_SADR: psHu16(mem) = value;//dma2 saddr HW_LOG("Hardware write DMA2_SADDR 32bit at %x with value %x",mem,value); @@ -449,14 +481,17 @@ __forceinline void hwWrite16(u32 mem, u16 value) psHu16(mem) = value;//dma2 madr HW_LOG("Hardware write IPU0DMA_MADR 32bit at %x with value %x",mem,value); break; + case D3_QWC: psHu16(mem) = value;//dma2 madr HW_LOG("Hardware write IPU0DMA_QWC 32bit at %x with value %x",mem,value); - break; + break; + case D3_TADR: psHu16(mem) = value;//dma2 tadr HW_LOG("Hardware write IPU0DMA_TADR 32bit at %x with value %x",mem,value); break; + case D3_SADR: psHu16(mem) = value;//dma2 saddr HW_LOG("Hardware write IPU0DMA_SADDR 32bit at %x with value %x",mem,value); @@ -478,14 +513,17 @@ __forceinline void hwWrite16(u32 mem, u16 value) psHu16(mem) = value;//dma2 madr HW_LOG("Hardware write IPU1DMA_MADR 32bit at %x with value %x",mem,value); break; + case D4_QWC: psHu16(mem) = value;//dma2 madr HW_LOG("Hardware write IPU1DMA_QWC 32bit at %x with value %x",mem,value); break; + case D4_TADR: psHu16(mem) = value;//dma2 tadr HW_LOG("Hardware write IPU1DMA_TADR 32bit at %x with value %x",mem,value); break; + case D4_SADR: psHu16(mem) = value;//dma2 saddr HW_LOG("Hardware write IPU1DMA_SADDR 32bit at %x with value %x",mem,value); @@ -502,9 +540,10 @@ __forceinline void hwWrite16(u32 mem, u16 value) DmaExec16(dmaSIF0, mem, value); break; - case 0x1000c002: + case 0x1000c002: // D5_CHCR + 2 //? break; + case D6_CHCR: // dma6 - sif1 DMA_LOG("SIF1dma %lx", value); if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) @@ -514,15 +553,24 @@ __forceinline void hwWrite16(u32 mem, u16 value) } DmaExec16(dmaSIF1, mem, value); break; - + + // Given the other values here, perhaps something like this is in order? + /*case 0x1000C402: // D6_CHCR + 2 + //? + break;*/ + #ifdef PCSX2_DEVBUILD - // No D6_MADR, and a TADR address that's not in the defines? + case D6_MADR: // dma6 - sif1 - madr + HW_LOG("SIF1dma MADR = %lx", value); + psHu16(mem) = value; + break; + case D6_QWC: // dma6 - sif1 - qwc HW_LOG("SIF1dma QWC = %lx", value); psHu16(mem) = value; break; - case 0x1000c430: // dma6 - sif1 - tadr + case D6_TADR: // dma6 - sif1 - tadr HW_LOG("SIF1dma TADR = %lx", value); psHu16(mem) = value; break; @@ -537,9 +585,11 @@ __forceinline void hwWrite16(u32 mem, u16 value) } DmaExec16(dmaSIF2, mem, value); break; - case 0x1000c802: + + case 0x1000c802: // D7_CHCR + 2 //? break; + case D8_CHCR: // dma8 - fromSPR DMA_LOG("fromSPRdma %lx", value); if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) @@ -550,7 +600,7 @@ __forceinline void hwWrite16(u32 mem, u16 value) DmaExec16(dmaSPR0, mem, value); break; - case 0x1000d400: // dma9 - toSPR + case SPR1_CHCR: // dma9 - toSPR DMA_LOG("toSPRdma %lx", value); if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -559,35 +609,48 @@ __forceinline void hwWrite16(u32 mem, u16 value) } DmaExec16(dmaSPR1, mem, value); break; - case 0x1000f592: // DMAC_ENABLEW + + case 0x1000f592: // DMAC_ENABLEW + 2 psHu16(0xf592) = value; psHu16(0xf522) = value; break; - case 0x1000f130: + + case SIO_ISR: case 0x1000f132: case 0x1000f410: case 0x1000f412: - case 0x1000f430: + case MCH_RICM: case 0x1000f432: break; - case 0x1000f200: + case SBUS_F200: + psHu16(mem) = value; + break; + + case SBUS_F210: psHu16(mem) = value; break; case SBUS_F220: psHu16(mem) |= value; break; - case SBUS_SMFLG: + + case SBUS_F230: psHu16(mem) &= ~value; break; + case SBUS_F240: - if(!(value & 0x100)) + if (!(value & 0x100)) psHu16(mem) &= ~0x100; else psHu16(mem) |= 0x100; break; - case 0x1000f260: + + case SBUS_F250: + psHu16(mem) = value; + break; + + case SBUS_F260: psHu16(mem) = 0; break; @@ -644,9 +707,9 @@ void __fastcall hwWrite32_page_02( u32 mem, u32 value ) // Page 3 contains writes to vif0 and vif1 registers, plus some GIF stuff! void __fastcall hwWrite32_page_03( u32 mem, u32 value ) { - if(mem>=0x10003800) + if (mem >= VIF0_STAT) { - if(mem<0x10003c00) + if(mem < VIF1_STAT) vif0Write32(mem, value); else vif1Write32(mem, value); @@ -657,6 +720,7 @@ void __fastcall hwWrite32_page_03( u32 mem, u32 value ) { case GIF_CTRL: psHu32(mem) = value & 0x8; + if (value & 0x1) gsGIFReset(); else if( value & 8 ) @@ -701,7 +765,7 @@ void __fastcall hwWrite32_page_0B( u32 mem, u32 value ) QueuedDMA |= 0x8; } DmaExec(dmaIPU0, mem, value); - return; + return; case D3_MADR: regName = "IPU0DMA_MADR"; break; case D3_QWC: regName = "IPU0DMA_QWC"; break; @@ -718,7 +782,7 @@ void __fastcall hwWrite32_page_0B( u32 mem, u32 value ) QueuedDMA |= 0x10; } DmaExec(dmaIPU1, mem, value); - return; + return; case D4_MADR: regName = "IPU1DMA_MADR"; break; case D4_QWC: regName = "IPU1DMA_QWC"; break; @@ -746,32 +810,39 @@ void __fastcall StartQueuedDMA() void __fastcall hwWrite32_page_0E( u32 mem, u32 value ) { - if( mem == DMAC_CTRL ) + switch (mem) { - HW_LOG("DMAC_CTRL Write 32bit %x", value); - //Check for DMAS that were started while the DMAC was disabled - if((psHu32(mem) & 0x1) == 0 && (value & 0x1) == 1) + case DMAC_CTRL: { + u32 oldvalue = psHu32(mem); + + HW_LOG("DMAC_CTRL Write 32bit %x", value); + psHu32(mem) = value; - if(QueuedDMA != 0) StartQueuedDMA(); - return; + //Check for DMAS that were started while the DMAC was disabled + if (((oldvalue & 0x1) == 0) && ((value & 0x1) == 1)) + { + if (QueuedDMA != 0) StartQueuedDMA(); + } + break; } + + case DMAC_STAT: + HW_LOG("DMAC_STAT Write 32bit %x", value); + + // lower 16 bits: clear on 1 + // upper 16 bits: reverse on 1 + + psHu16(0xe010) &= ~(value & 0xffff); + psHu16(0xe012) ^= (u16)(value >> 16); + + cpuTestDMACInts(); + break; + + default: + psHu32(mem) = value; + break; } - else if( mem == DMAC_STAT ) - { - HW_LOG("DMAC_STAT Write 32bit %x", value); - - // lower 16 bits: clear on 1 - // upper 16 bits: reverse on 1 - - psHu16(0xe010) &= ~(value & 0xffff); - psHu16(0xe012) ^= (u16)(value >> 16); - - cpuTestDMACInts(); - return; - } - - psHu32(mem) = value; } void __fastcall hwWrite32_page_0F( u32 mem, u32 value ) @@ -796,46 +867,50 @@ void __fastcall hwWrite32_page_0F( u32 mem, u32 value ) break; //------------------------------------------------------------------ - case HELPSWITCH(0x1000f430)://MCH_RICM: x:4|SA:12|x:5|SDEV:1|SOP:4|SBC:1|SDEV:5 + case HELPSWITCH(MCH_RICM)://MCH_RICM: x:4|SA:12|x:5|SDEV:1|SOP:4|SBC:1|SDEV:5 if ((((value >> 16) & 0xFFF) == 0x21) && (((value >> 6) & 0xF) == 1) && (((psHu32(0xf440) >> 7) & 1) == 0))//INIT & SRP=0 rdram_sdevid = 0; // if SIO repeater is cleared, reset sdevid psHu32(mem) = value & ~0x80000000; //kill the busy bit break; - case HELPSWITCH(0x1000f200): + case HELPSWITCH(SBUS_F200): psHu32(mem) = value; break; + case HELPSWITCH(SBUS_F220): psHu32(mem) |= value; break; - case HELPSWITCH(SBUS_SMFLG): + + case HELPSWITCH(SBUS_F230): psHu32(mem) &= ~value; break; + case HELPSWITCH(SBUS_F240): if(!(value & 0x100)) psHu32(mem) &= ~0x100; else psHu32(mem) |= 0x100; break; - case HELPSWITCH(0x1000f260): + + case HELPSWITCH(SBUS_F260): psHu32(mem) = 0; break; - case HELPSWITCH(0x1000f440)://MCH_DRD: + case HELPSWITCH(MCH_DRD)://MCH_DRD: psHu32(mem) = value; break; - case HELPSWITCH(DMAC_ENABLEW): // DMAC_ENABLEW + case HELPSWITCH(DMAC_ENABLEW): HW_LOG("DMAC_ENABLEW Write 32bit %lx", value); psHu32(0xf590) = value; psHu32(0xf520) = value; break; //------------------------------------------------------------------ - case HELPSWITCH(0x1000f130): + case HELPSWITCH(SIO_ISR): case HELPSWITCH(0x1000f410): HW_LOG("Unknown Hardware write 32 at %x with value %x (%x)", mem, value, cpuRegs.CP0.n.Status.val); - break; + break; default: psHu32(mem) = value; @@ -851,26 +926,31 @@ void __fastcall hwWrite32_generic( u32 mem, u32 value ) { case D0_CHCR: // dma0 - vif0 DMA_LOG("VIF0dma EXECUTE, value=0x%x", value); + if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) { DevCon::Notice("32 bit VIF0 DMA Start while DMAC Disabled\n"); QueuedDMA |= 0x1; } + DmaExec(dmaVIF0, mem, value); return; //------------------------------------------------------------------ case D1_CHCR: // dma1 - vif1 - chcr DMA_LOG("VIF1dma EXECUTE, value=0x%x", value); + if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) { DevCon::Notice("32 bit VIF1 DMA Start while DMAC Disabled\n"); QueuedDMA |= 0x2; } - if(value & 0x100) + + if (value & 0x100) { vif1.done = false; //This must be done here! some games (ala Crash of the Titans) pause the dma to start MFIFO - } + } + DmaExec(dmaVIF1, mem, value); return; @@ -921,9 +1001,9 @@ void __fastcall hwWrite32_generic( u32 mem, u32 value ) DmaExec(dmaSIF1, mem, value); return; - // Again, no MADR, and an undefined TADR. + case D6_MADR: regName = "SIF1dma MADR"; break; case D6_QWC: regName = "SIF1dma QWC"; break; - case 0x1000c430: regName = "SIF1dma TADR"; break; + case D6_TADR: regName = "SIF1dma TADR"; break; //------------------------------------------------------------------ case D7_CHCR: // dma7 - sif2 @@ -946,7 +1026,7 @@ void __fastcall hwWrite32_generic( u32 mem, u32 value ) DmaExec(dmaSPR0, mem, value); return; //------------------------------------------------------------------ - case 0x1000d400: // dma9 - toSPR + case SPR1_CHCR: // dma9 - toSPR DMA_LOG("SPR1dma EXECUTE (toSPR), value=0x%x", value); if ((value & 0x100) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -974,9 +1054,9 @@ void __fastcall hwWrite64_page_03( u32 mem, const mem64_t* srcval ) //hwWrite64( mem, *srcval ); return; const u64 value = *srcval; - if(mem>=0x10003800) + if (mem >= VIF0_STAT) { - if(mem<0x10003c00) + if (mem < VIF1_STAT) vif0Write32(mem, value); else vif1Write32(mem, value); @@ -997,23 +1077,23 @@ void __fastcall hwWrite64_page_03( u32 mem, const mem64_t* srcval ) else psHu32(GIF_STAT) &= ~8; } - - return; + break; case GIF_MODE: { + // set/clear bits 0 and 2 as per the GIF_MODE value. + const u32 bitmask = 0x1 | 0x4; + Console::Status("GIFMODE64 %x", params value); psHu64(GIF_MODE) = value; - - // set/clear bits 0 and 2 as per the GIF_MODE value. - const u32 bitmask = 0x1 | 0x4; psHu32(GIF_STAT) &= ~bitmask; psHu32(GIF_STAT) |= (u32)value & bitmask; + break; } case GIF_STAT: // stat is readonly - return; + break; } } @@ -1023,31 +1103,38 @@ void __fastcall hwWrite64_page_0E( u32 mem, const mem64_t* srcval ) const u64 value = *srcval; - if( mem == DMAC_CTRL ) + switch (mem) { - HW_LOG("DMAC_CTRL Write 64bit %x", value); - if((psHu32(mem) & 0x1) == 0 && (value & 0x1) == 1) + case DMAC_CTRL: { + u32 oldvalue = psHu32(mem); psHu64(mem) = value; - if(QueuedDMA != 0) StartQueuedDMA(); - return; + + HW_LOG("DMAC_CTRL Write 64bit %x", value); + + if (((oldvalue & 0x1) == 0) && ((value & 0x1) == 1)) + { + if (QueuedDMA != 0) StartQueuedDMA(); + } + break; } + + case DMAC_STAT: + HW_LOG("DMAC_STAT Write 64bit %x", value); + + // lower 16 bits: clear on 1 + // upper 16 bits: reverse on 1 + + psHu16(0xe010) &= ~(value & 0xffff); + psHu16(0xe012) ^= (u16)(value >> 16); + + cpuTestDMACInts(); + break; + + default: + psHu64(mem) = value; + break; } - else if( mem == DMAC_STAT ) - { - HW_LOG("DMAC_STAT Write 64bit %x", value); - - // lower 16 bits: clear on 1 - // upper 16 bits: reverse on 1 - - psHu16(0xe010) &= ~(value & 0xffff); - psHu16(0xe012) ^= (u16)(value >> 16); - - cpuTestDMACInts(); - return; - } - - psHu64(mem) = value; } void __fastcall hwWrite64_generic( u32 mem, const mem64_t* srcval ) @@ -1061,23 +1148,23 @@ void __fastcall hwWrite64_generic( u32 mem, const mem64_t* srcval ) case D2_CHCR: // dma2 - gif DMA_LOG("0x%8.8x hwWrite64: GSdma %x", cpuRegs.cycle, value); DmaExec(dmaGIF, mem, value); - break; + break; case INTC_STAT: HW_LOG("INTC_STAT Write 64bit %x", (u32)value); psHu32(INTC_STAT) &= ~value; //cpuTestINTCInts(); - break; + break; case INTC_MASK: HW_LOG("INTC_MASK Write 64bit %x", (u32)value); psHu32(INTC_MASK) ^= (u16)value; cpuTestINTCInts(); - break; + break; - case 0x1000f130: + case SIO_ISR: case 0x1000f410: - case 0x1000f430: + case MCH_RICM: break; case DMAC_ENABLEW: // DMAC_ENABLEW @@ -1118,9 +1205,9 @@ void __fastcall hwWrite128_generic(u32 mem, const mem128_t *srcval) psHu32(0xf520) = srcval[0]; break; - case 0x1000f130: + case SIO_ISR: case 0x1000f410: - case 0x1000f430: + case MCH_RICM: break; default: diff --git a/pcsx2/IopDma.cpp b/pcsx2/IopDma.cpp index 1e9231a99d..96ed3e3c7e 100644 --- a/pcsx2/IopDma.cpp +++ b/pcsx2/IopDma.cpp @@ -170,13 +170,13 @@ void psxDma9(u32 madr, u32 bcr, u32 chcr) SIF_LOG("IOP: dmaSIF0 chcr = %lx, madr = %lx, bcr = %lx, tadr = %lx", chcr, madr, bcr, HW_DMA9_TADR); iopsifbusy[0] = 1; - psHu32(0x1000F240) |= 0x2000; + psHu32(SBUS_F240) |= 0x2000; if (eesifbusy[0] == 1) { SIF0Dma(); - psHu32(0x1000F240) &= ~0x20; - psHu32(0x1000F240) &= ~0x2000; + psHu32(SBUS_F240) &= ~0x20; + psHu32(SBUS_F240) &= ~0x2000; } } @@ -185,15 +185,15 @@ void psxDma10(u32 madr, u32 bcr, u32 chcr) SIF_LOG("IOP: dmaSIF1 chcr = %lx, madr = %lx, bcr = %lx", chcr, madr, bcr); iopsifbusy[1] = 1; - psHu32(0x1000F240) |= 0x4000; + psHu32(SBUS_F240) |= 0x4000; if (eesifbusy[1] == 1) { FreezeXMMRegs(1); SIF1Dma(); - psHu32(0x1000F240) &= ~0x40; - psHu32(0x1000F240) &= ~0x100; - psHu32(0x1000F240) &= ~0x4000; + psHu32(SBUS_F240) &= ~0x40; + psHu32(SBUS_F240) &= ~0x100; + psHu32(SBUS_F240) &= ~0x4000; FreezeXMMRegs(0); } } diff --git a/pcsx2/IopMem.cpp b/pcsx2/IopMem.cpp index a9580a1cdb..55003d58b7 100644 --- a/pcsx2/IopMem.cpp +++ b/pcsx2/IopMem.cpp @@ -195,13 +195,13 @@ u16 iopMemRead16(u32 mem) switch(mem & 0xF0) { case 0x00: - ret= psHu16(0x1000F200); + ret= psHu16(SBUS_F200); break; case 0x10: - ret= psHu16(0x1000F210); + ret= psHu16(SBUS_F210); break; case 0x40: - ret= psHu16(0x1000F240) | 0x0002; + ret= psHu16(SBUS_F240) | 0x0002; break; case 0x60: ret = 0; @@ -261,19 +261,19 @@ u32 iopMemRead32(u32 mem) switch(mem & 0xF0) { case 0x00: - ret= psHu32(0x1000F200); + ret= psHu32(SBUS_F200); break; case 0x10: - ret= psHu32(0x1000F210); + ret= psHu32(SBUS_F210); break; case 0x20: - ret= psHu32(0x1000F220); + ret= psHu32(SBUS_F220); break; case 0x30: // EE Side - ret= psHu32(0x1000F230); + ret= psHu32(SBUS_F230); break; case 0x40: - ret= psHu32(0x1000F240) | 0xF0000002; + ret= psHu32(SBUS_F240) | 0xF0000002; break; case 0x60: ret = 0; @@ -405,7 +405,7 @@ void iopMemWrite16(u32 mem, u16 value) { case 0x10: // write to ps2 mem - psHu16(0x1000F210) = value; + psHu16(SBUS_F210) = value; return; case 0x40: { @@ -413,17 +413,19 @@ void iopMemWrite16(u32 mem, u16 value) // write to ps2 mem if(value & 0x20 || value & 0x80) { - psHu16(0x1000F240) &= ~0xF000; - psHu16(0x1000F240) |= 0x2000; + psHu16(SBUS_F240) &= ~0xF000; + psHu16(SBUS_F240) |= 0x2000; } - if(psHu16(0x1000F240) & temp) psHu16(0x1000F240) &= ~temp; - else psHu16(0x1000F240) |= temp; + if(psHu16(SBUS_F240) & temp) + psHu16(SBUS_F240) &= ~temp; + else + psHu16(SBUS_F240) |= temp; return; } case 0x60: - psHu32(0x1000F260) = 0; + psHu32(SBUS_F260) = 0; return; } @@ -490,36 +492,36 @@ void iopMemWrite32(u32 mem, u32 value) return; // this is the IOP, so read-only (do nothing) case 0x10: // IOP write path (EE/IOP readable) - psHu32(0x1000F210) = value; + psHu32(SBUS_F210) = value; return; case 0x20: // Bits cleared when written from IOP. - psHu32(0x1000F220) &= ~value; + psHu32(SBUS_F220) &= ~value; return; case 0x30: // bits set when written from IOP - psHu32(0x1000F230) |= value; + psHu32(SBUS_F230) |= value; return; case 0x40: // Control Register { u32 temp = value & 0xF0; - if(value & 0x20 || value & 0x80) + if (value & 0x20 || value & 0x80) { - psHu32(0x1000F240) &= ~0xF000; - psHu32(0x1000F240) |= 0x2000; + psHu32(SBUS_F240) &= ~0xF000; + psHu32(SBUS_F240) |= 0x2000; } - if(psHu32(0x1000F240) & temp) - psHu32(0x1000F240) &= ~temp; + if (psHu32(SBUS_F240) & temp) + psHu32(SBUS_F240) &= ~temp; else - psHu32(0x1000F240) |= temp; + psHu32(SBUS_F240) |= temp; return; } case 0x60: - psHu32(0x1000F260) = 0; + psHu32(SBUS_F260) = 0; return; } psxSu32(mem) = value; diff --git a/pcsx2/Sif.cpp b/pcsx2/Sif.cpp index 5fdaa6ca05..73d1c8d8f0 100644 --- a/pcsx2/Sif.cpp +++ b/pcsx2/Sif.cpp @@ -452,15 +452,15 @@ __forceinline void dmaSIF0() SIF_LOG("warning, sif0.fifoReadPos != sif0.fifoWritePos"); } - psHu32(0x1000F240) |= 0x2000; + psHu32(SBUS_F240) |= 0x2000; eesifbusy[0] = 1; if (iopsifbusy[0] == 1) { FreezeXMMRegs(1); hwIntcIrq(INTC_SBUS); SIF0Dma(); - psHu32(0x1000F240) &= ~0x20; - psHu32(0x1000F240) &= ~0x2000; + psHu32(SBUS_F240) &= ~0x20; + psHu32(SBUS_F240) &= ~0x2000; FreezeXMMRegs(0); } } @@ -475,15 +475,15 @@ __forceinline void dmaSIF1() SIF_LOG("warning, sif1.fifoReadPos != sif1.fifoWritePos"); } - psHu32(0x1000F240) |= 0x4000; + psHu32(SBUS_F240) |= 0x4000; eesifbusy[1] = 1; if (iopsifbusy[1] == 1) { FreezeXMMRegs(1); SIF1Dma(); - psHu32(0x1000F240) &= ~0x40; - psHu32(0x1000F240) &= ~0x100; - psHu32(0x1000F240) &= ~0x4000; + psHu32(SBUS_F240) &= ~0x40; + psHu32(SBUS_F240) &= ~0x100; + psHu32(SBUS_F240) &= ~0x4000; FreezeXMMRegs(0); } diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index e24adc47d9..e6ca05a339 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -351,7 +351,6 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int vifMaskRegs = g_vif0Masks; vif = &vif0; vifRow = g_vifRow0; - assert(v->addr < memsize); } else { @@ -360,8 +359,8 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int vifMaskRegs = g_vif1Masks; vif = &vif1; vifRow = g_vifRow1; - assert(v->addr < memsize); } + assert(v->addr < memsize); dest = (u32*)(VU->Mem + v->addr); @@ -1554,8 +1553,8 @@ void vif0Write32(u32 mem, u32 value) memzero_obj(vif0); vif0ch->qwc = 0; //? cpuRegs.interrupt &= ~1; //Stop all vif0 DMA's - psHu64(0x10004000) = 0; - psHu64(0x10004008) = 0; + psHu64(VIF0_FIFO) = 0; + psHu64(0x10004008) = 0; // VIF0_FIFO + 8 vif0.done = true; vif0Regs->err = 0; vif0Regs->stat &= ~(0xF000000 | VIF0_STAT_INT | VIF0_STAT_VSS | VIF0_STAT_VIS | VIF0_STAT_VFS | VIF0_STAT_VPS); // FQC=0 @@ -1619,22 +1618,25 @@ void vif0Write32(u32 mem, u32 value) vif0Regs->err = value; break; + case VIF0_R0: + case VIF0_R1: + case VIF0_R2: + case VIF0_R3: + assert((mem&0xf) == 0); + g_vifRow0[(mem>>4) & 3] = value; + break; + + case VIF0_C0: + case VIF0_C1: + case VIF0_C2: + case VIF0_C3: + assert((mem&0xf) == 0); + g_vifCol0[(mem>>4) & 3] = value; + break; + default: Console::WriteLn("Unknown Vif0 write to %x", params mem); - if (mem >= VIF0_R0 && mem < 0x10003980) // mem <= VIF0_C3? - { - assert((mem&0xf) == 0); - - if (mem < VIF0_C0) - g_vifRow0[(mem>>4)&3] = value; - else - g_vifCol0[(mem>>4)&3] = value; - - } - else - { - psHu32(mem) = value; - } + psHu32(mem) = value; break; } /* Other registers are read-only so do nothing for them */ @@ -1646,8 +1648,8 @@ void vif0Reset() memzero_obj(vif0); memzero_obj(*vif0Regs); SetNewMask(g_vif0Masks, g_vif0HasMask3, 0, 0xffffffff); - psHu64(0x10004000) = 0; - psHu64(0x10004008) = 0; + psHu64(VIF0_FIFO) = 0; + psHu64(0x10004008) = 0; // VIF0_FIFO + 8 vif0Regs->stat &= ~VIF0_STAT_VPS; vif0.done = true; vif0Regs->stat &= ~0xF000000; // FQC=0 @@ -1790,7 +1792,7 @@ static int __fastcall Vif1TransSTRow(u32 *data) pmem[0] = data[0]; pmem2[0] = data[0]; break; - jNO_DEFAULT; + jNO_DEFAULT; } vif1.tag.addr += ret; vif1.tag.size -= ret; @@ -2645,8 +2647,8 @@ void vif1Write32(u32 mem, u32 value) memzero_obj(vif1); cpuRegs.interrupt &= ~((1 << 1) | (1 << 10)); //Stop all vif1 DMA's vif1ch->qwc = 0; //? - psHu64(0x10005000) = 0; - psHu64(0x10005008) = 0; + psHu64(VIF1_FIFO) = 0; + psHu64(0x10005008) = 0; // VIF1_FIFO + 8 vif1.done = true; vif1Regs->err = 0; vif1.inprogress = 0; @@ -2750,20 +2752,25 @@ void vif1Write32(u32 mem, u32 value) vif1Regs->mode = value; break; + case VIF1_R0: + case VIF1_R1: + case VIF1_R2: + case VIF1_R3: + assert((mem&0xf) == 0); + g_vifRow1[(mem>>4) & 3] = value; + break; + + case VIF1_C0: + case VIF1_C1: + case VIF1_C2: + case VIF1_C3: + assert((mem&0xf) == 0); + g_vifCol1[(mem>>4) & 3] = value; + break; + default: Console::WriteLn("Unknown Vif1 write to %x", params mem); - if ((mem >= VIF1_R0) && (mem < 0x10003d80)) // mem <= VIF1_C3? - { - assert((mem&0xf) == 0); - if (mem < VIF1_C0) - g_vifRow1[(mem>>4)&3] = value; - else - g_vifCol1[(mem>>4)&3] = value; - } - else - { - psHu32(mem) = value; - } + psHu32(mem) = value; break; } @@ -2776,8 +2783,8 @@ void vif1Reset() memzero_obj(vif1); memzero_obj(*vif1Regs); SetNewMask(g_vif1Masks, g_vif1HasMask3, 0, 0xffffffff); - psHu64(0x10005000) = 0; - psHu64(0x10005008) = 0; + psHu64(VIF1_FIFO) = 0; + psHu64(0x10005008) = 0; // VIF1_FIFO + 8 vif1Regs->stat &= ~VIF1_STAT_VPS; vif1.done = true; cpuRegs.interrupt &= ~((1 << 1) | (1 << 10)); //Stop all vif1 DMA's diff --git a/pcsx2/VifDma.h b/pcsx2/VifDma.h index 86a1d1f78d..56c08d7d6f 100644 --- a/pcsx2/VifDma.h +++ b/pcsx2/VifDma.h @@ -25,54 +25,6 @@ enum VifModes VIF_CHAIN_MODE = 2 }; -enum VifMemoryLocations -{ - VIF0_STAT = 0x10003800, - VIF0_FBRST = 0x10003810, - VIF0_ERR = 0x10003820, - VIF0_MARK = 0x10003830, - VIF0_CYCLE = 0x10003840, - VIF0_MODE = 0x10003850, - VIF0_NUM = 0x10003860, - VIF0_MASK = 0x10003870, - VIF0_CODE = 0x10003880, - VIF0_ITOPS = 0x10003890, - VIF0_ITOP = 0x100038d0, - VIF0_TOP = 0x100038e0, - VIF0_R0 = 0x10003900, - VIF0_R1 = 0x10003910, - VIF0_R2 = 0x10003920, - VIF0_R3 = 0x10003930, - VIF0_C0 = 0x10003940, - VIF0_C1 = 0x10003950, - VIF0_C2 = 0x10003960, - VIF0_C3 = 0x10003970, - - VIF1_STAT = 0x10003c00, - VIF1_FBRST = 0x10003c10, - VIF1_ERR = 0x10003c20, - VIF1_MARK = 0x10003c30, - VIF1_CYCLE = 0x10003c40, - VIF1_MODE = 0x10003c50, - VIF1_NUM = 0x10003c60, - VIF1_MASK = 0x10003c70, - VIF1_CODE = 0x10003c80, - VIF1_ITOPS = 0x10003c90, - VIF1_BASE = 0x10003ca0, - VIF1_OFST = 0x10003cb0, - VIF1_TOPS = 0x10003cc0, - VIF1_ITOP = 0x10003cd0, - VIF1_TOP = 0x10003ce0, - VIF1_R0 = 0x10003d00, - VIF1_R1 = 0x10003d10, - VIF1_R2 = 0x10003d20, - VIF1_R3 = 0x10003d30, - VIF1_C0 = 0x10003d40, - VIF1_C1 = 0x10003d50, - VIF1_C2 = 0x10003d60, - VIF1_C3 = 0x10003d70 -}; - struct vifCode { u32 addr; u32 size; From eb3b31c5251cf8249e11074e1fc79e223f88fe8b Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 31 May 2009 05:45:59 +0000 Subject: [PATCH 161/296] microVU: - Gave the rec a better IR (intermediate representation) implementation. - Renamed microVU_Alloc to microVU_IR git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1298 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 8 +- pcsx2/x86/microVU.h | 4 +- pcsx2/x86/microVU_Alloc.inl | 6 +- pcsx2/x86/microVU_Analyze.inl | 72 ++++++----- pcsx2/x86/microVU_Compile.inl | 51 ++++---- pcsx2/x86/microVU_Flags.inl | 46 ++++--- pcsx2/x86/{microVU_Alloc.h => microVU_IR.h} | 49 ++++++- pcsx2/x86/microVU_Log.inl | 3 +- pcsx2/x86/microVU_Lower.inl | 134 ++++++++++---------- pcsx2/x86/microVU_Misc.h | 73 ++--------- pcsx2/x86/microVU_Misc.inl | 4 +- pcsx2/x86/microVU_Upper.inl | 50 ++++---- 12 files changed, 248 insertions(+), 252 deletions(-) rename pcsx2/x86/{microVU_Alloc.h => microVU_IR.h} (53%) diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index fa733283cd..5950b8d879 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -2038,10 +2038,6 @@ RelativePath="..\..\x86\microVU.h" > - - @@ -2062,6 +2058,10 @@ RelativePath="..\..\x86\microVU_Flags.inl" > + + diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index c2445a93d4..7fd6efc281 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -24,7 +24,7 @@ #include "VU.h" #include "GS.h" #include "ix86/ix86.h" -#include "microVU_Alloc.h" +#include "microVU_IR.h" #include "microVU_Misc.h" #define mMaxBlocks 32 // Max Blocks With Different Pipeline States (For n = 1, 2, 4, 8, 16, etc...) @@ -80,7 +80,7 @@ struct microProgram { u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache microBlockManager* block[progSize/2]; - microAllocInfo allocInfo; + microIR allocInfo; }; #define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index 0ab348f8be..eb0c5deda0 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -537,7 +537,7 @@ microVUt(void) mVUallocFMAC19b(mV, int& Fd) { //------------------------------------------------------------------ #define getQreg(reg) { \ - mVUunpack_xyzw(reg, xmmPQ, readQ); \ + mVUunpack_xyzw(reg, xmmPQ, mVUinfo.readQ); \ /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ } @@ -686,7 +686,7 @@ microVUt(void) mVUallocVIa(mV, int GPRreg, int _reg_) { } microVUt(void) mVUallocVIb(mV, int GPRreg, int _reg_) { - if (backupVI) { // Backs up reg to memory (used when VI is modified b4 a branch) + if (mVUlow.backupVI) { // Backs up reg to memory (used when VI is modified b4 a branch) MOV32RtoM((uptr)&mVU->VIbackup[1], GPRreg); mVUallocVIa(mVU, GPRreg, _reg_); MOV32RtoM((uptr)&mVU->VIbackup[0], GPRreg); @@ -702,7 +702,7 @@ microVUt(void) mVUallocVIb(mV, int GPRreg, int _reg_) { //------------------------------------------------------------------ #define getPreg(reg) { \ - mVUunpack_xyzw(reg, xmmPQ, (2 + readP)); \ + mVUunpack_xyzw(reg, xmmPQ, (2 + mVUinfo.readP)); \ /*if (CHECK_VU_EXTRA_OVERFLOW) mVUclamp2(reg, xmmT1, 15);*/ \ } diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 0422662fba..74c1cbcfc9 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -56,13 +56,13 @@ || (mVUregsTemp.VF[0].y && _Y) \ || (mVUregsTemp.VF[0].z && _Z) \ || (mVUregsTemp.VF[0].w && _W)) \ - { mVUinfo |= _swapOps; } \ + { mVUinfo.swapOps = 1; } \ } \ } \ } microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { - mVUinfo |= _doStatus; + sFLAG.doFlag = 1; analyzeReg1(Fs); analyzeReg1(Ft); analyzeReg2(Fd, 0); @@ -91,7 +91,7 @@ microVUt(void) mVUanalyzeFMAC2(mV, int Fs, int Ft) { } microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { - mVUinfo |= _doStatus; + sFLAG.doFlag = 1; analyzeReg1(Fs); analyzeReg3(Ft); analyzeReg2(Fd, 0); @@ -106,7 +106,7 @@ microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { } microVUt(void) mVUanalyzeFMAC4(mV, int Fs, int Ft) { - mVUinfo |= _doClip; + cFLAG.doFlag = 1; analyzeReg1(Fs); analyzeReg4(Ft); } @@ -116,18 +116,18 @@ microVUt(void) mVUanalyzeFMAC4(mV, int Fs, int Ft) { //------------------------------------------------------------------ #define analyzeVIreg1(reg) { if (reg) { mVUstall = aMax(mVUstall, mVUregs.VI[reg]); } } -#define analyzeVIreg2(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; mVUinfo |= _writesVI; mVU->VIbackup[0] = reg; } } +#define analyzeVIreg2(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; mVUlow.writesVI = 1; mVU->VIbackup[0] = reg; } } #define analyzeVIreg3(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; } } microVUt(void) mVUanalyzeIALU1(mV, int Id, int Is, int It) { - if (!Id) { mVUinfo |= _isNOP; } + if (!Id) { mVUlow.isNOP = 1; } analyzeVIreg1(Is); analyzeVIreg1(It); analyzeVIreg2(Id, 1); } microVUt(void) mVUanalyzeIALU2(mV, int Is, int It) { - if (!It) { mVUinfo |= _isNOP; } + if (!It) { mVUlow.isNOP = 1; } analyzeVIreg1(Is); analyzeVIreg2(It, 1); } @@ -148,13 +148,13 @@ microVUt(void) mVUanalyzeIALU2(mV, int Is, int It) { || (mVUregsTemp.VF[0].z && _Y) \ || (mVUregsTemp.VF[0].w && _Z) \ || (mVUregsTemp.VF[0].x && _W)) \ - { mVUinfo |= _swapOps; } \ + { mVUinfo.swapOps = 1; } \ } \ } \ } microVUt(void) mVUanalyzeMR32(mV, int Fs, int Ft) { - if (!Ft) { mVUinfo |= _isNOP; } + if (!Ft) { mVUlow.isNOP = 1; } analyzeReg6(Fs); analyzeReg2(Ft, 1); } @@ -176,7 +176,7 @@ microVUt(void) mVUanalyzeMR32(mV, int Fs, int Ft) { || (mVUregsTemp.VF[0].y && (fxf == 1)) \ || (mVUregsTemp.VF[0].z && (fxf == 2)) \ || (mVUregsTemp.VF[0].w && (fxf == 3))) \ - { mVUinfo |= _swapOps; } \ + { mVUinfo.swapOps = 1; } \ } \ } \ } @@ -212,7 +212,7 @@ microVUt(void) mVUanalyzeEFU2(mV, int Fs, u8 xCycles) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeMFP(mV, int Ft) { - if (!Ft) { mVUinfo |= _isNOP; } + if (!Ft) { mVUlow.isNOP = 1; } analyzeReg2(Ft, 1); } @@ -221,7 +221,7 @@ microVUt(void) mVUanalyzeMFP(mV, int Ft) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeMOVE(mV, int Fs, int Ft) { - if (!Ft || (Ft == Fs)) { mVUinfo |= _isNOP; } + if (!Ft || (Ft == Fs)) { mVUlow.isNOP = 1; } analyzeReg1b(Fs); analyzeReg2(Ft, 1); } @@ -234,7 +234,7 @@ microVUt(void) mVUanalyzeMOVE(mV, int Fs, int Ft) { microVUt(void) mVUanalyzeLQ(mV, int Ft, int Is, bool writeIs) { analyzeVIreg1(Is); analyzeReg2(Ft, 1); - if (!Ft) { mVUinfo |= (writeIs && Is) ? _noWriteVF : _isNOP; } + if (!Ft) { if (writeIs && Is) { mVUlow.noWriteVF = 1; } else { mVUlow.isNOP = 1; } } if (writeIs) { analyzeVIreg2(Is, 1); } } @@ -260,7 +260,7 @@ microVUt(void) mVUanalyzeR1(mV, int Fs, int Fsf) { } microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { - if (!Ft) { mVUinfo |= ((canBeNOP) ? _isNOP : _noWriteVF); } + if (!Ft) { if (canBeNOP) { mVUlow.isNOP = 1; } else { mVUlow.noWriteVF = 1; } } analyzeReg2(Ft, 1); analyzeRreg(); } @@ -270,13 +270,13 @@ microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeSflag(mV, int It) { - if (!It) { mVUinfo |= _isNOP; } + if (!It) { mVUlow.isNOP = 1; } else { - mVUinfo |= _swapOps; + mVUinfo.swapOps = 1; mVUsFlagHack = 0; // Don't Optimize Out Status Flags for this block if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf /*<< mVUcount*/; } - if (mVUcount >= 1) { incPC2(-2); mVUinfo |= _isSflag; incPC2(2); } - // Note: _isSflag is used for status flag optimizations. + if (mVUcount >= 1) { incPC2(-2); mVUlow.useSflag = 1; incPC2(2); } + // Note: useSflag is used for status flag optimizations when a FSSET instruction is called. // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } @@ -284,7 +284,7 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { } microVUt(void) mVUanalyzeFSSET(mV) { - mVUinfo |= _isFSSET; + mVUlow.isFSSET = 1; // mVUinfo &= ~_doStatus; // Note: I'm not entirely sure if the non-sticky flags // should be taken from the current upper instruction @@ -297,14 +297,14 @@ microVUt(void) mVUanalyzeFSSET(mV) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { - if (!It) { mVUinfo |= _isNOP; } + if (!It) { mVUlow.isNOP = 1; } else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) - mVUinfo |= _swapOps; + mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } int curPC = iPC; for (int i = mVUcount, j = 0; i > 0; i--, j++) { incPC2(-2); - if (doStatus) { mVUinfo |= _doMac; if (j >= 3) { break; } } + if (sFLAG.doFlag) { mFLAG.doFlag = 1; if (j >= 3) { break; } } } iPC = curPC; } @@ -317,7 +317,7 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeCflag(mV, int It) { - mVUinfo |= _swapOps; + mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } analyzeVIreg3(It, 1); } @@ -345,24 +345,26 @@ microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { // Branches - Branch Opcodes //------------------------------------------------------------------ -#define analyzeBranchVI(reg, infoVal) { \ - if (reg && (mVUcount > 0)) { /* Ensures branch is not first opcode in block */ \ - incPC2(-2); \ - if (writesVI && (reg == mVU->VIbackup[0])) { /* If prev Op modified VI reg */ \ - mVUinfo |= _backupVI; \ - incPC2(2); \ - mVUinfo |= infoVal; \ - } \ - else { incPC2(2); } \ - } \ +#define analyzeBranchVI(reg, infoVar) { \ + /* First ensure branch is not first opcode in block */ \ + if (reg && (mVUcount > 0)) { \ + incPC2(-2); \ + /* Check if prev Op modified VI reg */ \ + if (mVUlow.writesVI && (reg == mVU->VIbackup[0])) { \ + mVUlow.backupVI = 1; \ + incPC2(2); \ + infoVar = 1; \ + } \ + else { incPC2(2); } \ + } \ } microVUt(void) mVUanalyzeBranch1(mV, int Is) { if (mVUregs.VI[Is] || mVUstall) { analyzeVIreg1(Is); } - else { analyzeBranchVI(Is, _memReadIs); } + else { analyzeBranchVI(Is, mVUlow.memReadIs); } } microVUt(void) mVUanalyzeBranch2(mV, int Is, int It) { if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is); analyzeVIreg1(It); } - else { analyzeBranchVI(Is, _memReadIs); analyzeBranchVI(It, _memReadIt);} + else { analyzeBranchVI(Is, mVUlow.memReadIs); analyzeBranchVI(It, mVUlow.memReadIt);} } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 09d4b46933..689d8b82a4 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -36,7 +36,7 @@ #define branchWarning() { \ if (mVUbranch) { \ Console::Error("microVU%d Warning: Branch in E-bit/Branch delay slot! [%04x]", params vuIndex, xPC); \ - mVUinfo |= _isNOP; \ + mVUlow.isNOP = 1; \ } \ } @@ -49,14 +49,19 @@ } \ } -#define startLoop() { mVUdebug1(); mVUstall = 0; memset(&mVUregsTemp, 0, sizeof(mVUregsTemp)); } +#define startLoop() { \ + mVUdebug1(); \ + memset(&mVUinfo, 0, sizeof(mVUinfo)); \ + memset(&mVUregsTemp, 0, sizeof(mVUregsTemp)); \ +} + #define calcCycles(reg, x) { reg = ((reg > x) ? (reg - x) : 0); } #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } #define doUpperOp() { mVUopU(mVU, 1); mVUdivSet(mVU); } #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } -#define doIbit() { if (curI & _Ibit_) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } +#define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } //------------------------------------------------------------------ // Helper Functions @@ -114,7 +119,7 @@ microVUt(void) mVUincCycles(mV, int x) { calcCycles(mVUregs.VI[z], x); } if (mVUregs.q) { - if (mVUregs.q > 4) { calcCycles(mVUregs.q, x); if (mVUregs.q <= 4) { mVUinfo |= _doDivFlag; } } + if (mVUregs.q > 4) { calcCycles(mVUregs.q, x); if (mVUregs.q <= 4) { mVUinfo.doDivFlag = 1; } } else { calcCycles(mVUregs.q, x); } if (!mVUregs.q) { incQ(); } } @@ -124,7 +129,7 @@ microVUt(void) mVUincCycles(mV, int x) { } if (mVUregs.xgkick) { calcCycles(mVUregs.xgkick, x); - if (!mVUregs.xgkick) { mVUinfo |= _doXGKICK; } + if (!mVUregs.xgkick) { mVUinfo.doXGKICK = 1; } } calcCycles(mVUregs.r, x); } @@ -132,7 +137,8 @@ microVUt(void) mVUincCycles(mV, int x) { microVUt(void) mVUsetCycles(mV) { incCycles(mVUstall); if (mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1] && mVUregsTemp.VFreg[0]) { // If upper Op && lower Op write to same VF reg - mVUinfo |= (mVUregsTemp.r || mVUregsTemp.VI) ? _noWriteVF : _isNOP; // If lower Op doesn't modify anything else, then make it a NOP + if (mVUregsTemp.r || mVUregsTemp.VI) mVUlow.noWriteVF = 1; + else mVUlow.isNOP = 1; // If lower Op doesn't modify anything else, then make it a NOP } tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].x, mVUregsTemp.VF[0].x); tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].y, mVUregsTemp.VF[0].y); @@ -241,20 +247,21 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); - mVUinfo = 0; startLoop(); incCycles(1); mVUopU(mVU, 0); - if (curI & _Ebit_) { branch = 1; } + if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } if (curI & _MDTbit_) { branch = 4; } - if (curI & _Ibit_) { mVUinfo |= _isNOP; } + if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } else { incPC(-1); mVUopL(mVU, 0); incPC(1); } mVUsetCycles(mVU); - if (mVU->p) { mVUinfo |= _readP; } - if (mVU->q) { mVUinfo |= _readQ; } - if (branch >= 2) { mVUinfo |= _isEOB | ((branch == 3) ? _isBdelay : 0); mVUcount++; branchWarning(); break; } + mVUinfo.readQ = mVU->q; + mVUinfo.writeQ = !mVU->q; + mVUinfo.readP = mVU->p; + mVUinfo.writeP = !mVU->p; + if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(mVU); branchEbit(); branch = 3; mVUbranch = 0; mVUinfo |= _isBranch; } + if (mVUbranch) { mVUsetFlagInfo(mVU); branchEbit(); branch = 3; mVUbranch = 0; } incPC(1); mVUcount++; } @@ -270,13 +277,13 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbranch = 0; int x; for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { - if (isEOB) { x = 0xffff; } - if (isNOP) { incPC(1); doUpperOp(); doIbit(); } - else if (!swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } - if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } + if (mVUinfo.isEOB) { x = 0xffff; } + if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } + else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } + else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } + if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } - if (!isBdelay) { incPC(1); } + if (!mVUinfo.isBdelay) { incPC(1); } else { microBlock* bBlock = NULL; u32* ajmp = 0; @@ -365,16 +372,16 @@ eBitTemination: int lStatus = findFlagInst(xStatus, 0x7fffffff); int lMac = findFlagInst(xMac, 0x7fffffff); int lClip = findFlagInst(xClip, 0x7fffffff); - mVUinfo = 0; + memset(&mVUinfo, 0, sizeof(mVUinfo)); incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; - if (doDivFlag) { + if (mVUinfo.doDivFlag) { int flagReg; getFlagReg(flagReg, lStatus); AND32ItoR (flagReg, 0x0fcf); OR32MtoR (flagReg, (uptr)&mVU->divFlag); } - if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } + if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } // Do E-bit end stuff here mVUsetupRange(mVU, xPC - 8); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index a049daac59..fc63c3db9d 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -21,9 +21,9 @@ // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet(mV) { int flagReg1, flagReg2; - if (doDivFlag) { - getFlagReg(flagReg1, fsInstance); - if (!doStatus) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } + if (mVUinfo.doDivFlag) { + getFlagReg(flagReg1, sFLAG.write); + if (!sFLAG.doFlag) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } AND32ItoR(flagReg1, 0x0fcf); OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } @@ -34,19 +34,19 @@ microVUt(void) mVUstatusFlagOp(mV) { int curPC = iPC; int i = mVUcount; bool runLoop = 1; - if (doStatus) { mVUinfo |= _isSflag; } + if (sFLAG.doFlag) { mVUlow.useSflag = 1; } else { for (; i > 0; i--) { incPC2(-2); - if (isSflag) { runLoop = 0; break; } - if (doStatus) { mVUinfo |= _isSflag; break; } + if (mVUlow.useSflag) { runLoop = 0; break; } + if (sFLAG.doFlag) { mVUlow.useSflag = 1; break; } } } if (runLoop) { for (; i > 0; i--) { incPC2(-2); - if (isSflag) break; - mVUinfo &= ~_doStatus; + if (mVUlow.useSflag) break; + sFLAG.doFlag = 0; } } iPC = curPC; @@ -69,7 +69,7 @@ void sortFlag(int* fFlag, int* bFlag, int cycles) { } } -#define sFlagCond ((doStatus && !mVUsFlagHack) || isFSSET || doDivFlag) +#define sFlagCond ((sFLAG.doFlag && !mVUsFlagHack) || mVUlow.isFSSET || mVUinfo.doDivFlag) // Note: Flag handling is 'very' complex, it requires full knowledge of how microVU recs work, so don't touch! microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { @@ -77,9 +77,9 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { int endPC = iPC; u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking - // Ensure last ~4+ instructions update mac flags + // Ensure last ~4+ instructions update mac flags (if next block's first 4 instructions will read them) for (int i = mVUcount; i > 0; i--, aCount++) { - if (doStatus) { if (__Mac) { mVUinfo |= _doMac; } if (aCount >= 4) { break; } } + if (sFLAG.doFlag) { if (__Mac) { mFLAG.doFlag = 1; } if (aCount >= 4) { break; } } incPC2(-2); } @@ -114,7 +114,7 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { u32 xCount = mVUcount; // Backup count iPC = mVUstartPC; for (mVUcount = 0; mVUcount < xCount; mVUcount++) { - if (isFSSET) { + if (mVUlow.isFSSET) { if (__Status) { // Don't Optimize out on the last ~4+ instructions if ((xCount - mVUcount) > aCount) { mVUstatusFlagOp(mVU); } } @@ -122,17 +122,21 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { } cycles += mVUstall; - mVUinfo |= findFlagInst(xStatus, cycles) << 18; // _fvsInstance - mVUinfo |= findFlagInst(xMac, cycles) << 16; // _fvmInstance - mVUinfo |= findFlagInst(xClip, cycles) << 20; // _fvcInstance + sFLAG.read = findFlagInst(xStatus, cycles); + mFLAG.read = findFlagInst(xMac, cycles); + cFLAG.read = findFlagInst(xClip, cycles); + + sFLAG.write = xS; + mFLAG.write = xM; + cFLAG.write = xC; - mVUinfo |= xS << 12; // _fsInstance - mVUinfo |= xM << 10; // _fmInstance - mVUinfo |= xC << 14; // _fcInstance + sFLAG.lastWrite = (xS-1) & 3; + mFLAG.lastWrite = (xM-1) & 3; + cFLAG.lastWrite = (xC-1) & 3; - if (sFlagCond) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } - if (doMac) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } - if (doClip) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } + if (sFlagCond) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } + if (mFLAG.doFlag) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } + if (cFLAG.doFlag) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } cycles++; incPC2(2); diff --git a/pcsx2/x86/microVU_Alloc.h b/pcsx2/x86/microVU_IR.h similarity index 53% rename from pcsx2/x86/microVU_Alloc.h rename to pcsx2/x86/microVU_IR.h index c9fd35b4a3..a42a683a25 100644 --- a/pcsx2/x86/microVU_Alloc.h +++ b/pcsx2/x86/microVU_IR.h @@ -65,17 +65,58 @@ struct microBlock { u8* x86ptrStart; // Start of code }; +struct microUpperOp { + bool eBit; // Has E-bit set + bool iBit; // Has I-bit set +}; + +struct microLowerOp { + bool isNOP; // This instruction is a NOP + bool isFSSET; // This instruction is a FSSET + bool useSflag; // This instruction uses/reads Sflag + u32 branch; // Branch Type (0 = Not a Branch, 1 = B. 2 = BAL, 3~8 = Conditional Branches, 9 = JALR, 10 = JR) + bool noWriteVF; // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) + bool backupVI; // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) + bool memReadIs; // Read Is (VI reg) from memory (used by branches) + bool memReadIt; // Read If (VI reg) from memory (used by branches) + bool writesVI; // Current Instruction writes to VI (used by branches; note that flag-modifying opcodes shouldn't set this) +}; + +struct microFlagInst { + bool doFlag; // Update Flag on this Instruction + u8 write; // Points to the instance that should be written to (s-stage write) + u8 lastWrite; // Points to the instance that was last written to (most up-to-date flag) + u8 read; // Points to the instance that should be read by a lower instruction (t-stage read) +}; + +struct microOp { + u8 stall; // Info on how much current instruction stalled + bool isEOB; // Cur Instruction is last instruction in block (End of Block) + bool isBdelay; // Cur Instruction in Branch Delay slot + bool swapOps; // Run Lower Instruction before Upper Instruction + bool doXGKICK; // Do XGKICK transfer on this instruction + bool doDivFlag; // Transfer Div flag to Status Flag on this instruction + int readQ; // Q instance for reading + int writeQ; // Q instance for writing + int readP; // P instance for reading + int writeP; // P instance for writing + microFlagInst sFlag; // Status Flag Instance Info + microFlagInst mFlag; // Mac Flag Instance Info + microFlagInst cFlag; // Clip Flag Instance Info + microUpperOp uOp; // Upper Op Info + microLowerOp lOp; // Lower Op Info +}; + template -struct microAllocInfo { +struct microIR { microBlock* pBlock; // Pointer to a block in mVUblocks microBlock block; // Block/Pipeline info microTempRegInfo regsTemp; // Temp Pipeline info (used so that new pipeline info isn't conflicting between upper and lower instructions in the same cycle) - u8 branch; // 0 = No Branch, 1 = B. 2 = BAL, 3~8 = Conditional Branches, 9 = JALR, 10 = JR + microOp info[pSize/2]; // Info for Instructions in current block + u8 branch; u32 cycles; // Cycles for current block u32 count; // Number of VU 64bit instructions ran (starts at 0 for each block) u32 curPC; // Current PC u32 startPC; // Start PC for Cur Block u32 sFlagHack; // Optimize out all Status flag updates if microProgram doesn't use Status flags - u32 info[pSize/2]; // Info for Instructions in current block - u8 stall[pSize/2]; // Info on how much each instruction stalled (stores the max amount of cycles to stall for the current opcodes) }; diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 4ef6ccbadf..4b84ab61cc 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -46,8 +46,7 @@ microVUx(void) __mVUdumpProgram(int progIndex) { int bBranch = mVUbranch; int bCode = mVU->code; int bPC = iPC; - vuIndex = (mVU == µVU1) ? 1 : 0; - mVUbranch = 0; + mVUbranch = 0; sprintf(str, "%s\\microVU%d prog - %02d.html", LOGS_DIR, vuIndex, progIndex); mVU->logFile = fopen(str, "w"); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 2b461c674b..848bcb3d52 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -71,9 +71,9 @@ mVUop(mVU_DIV) { if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("DIV Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } @@ -89,9 +89,9 @@ mVUop(mVU_SQRT) { if (CHECK_VU_OVERFLOW) SSE_MINSS_XMM_to_XMM(xmmFt, xmmMax); // Clamp infinities (only need to do positive clamp since xmmFt is positive) SSE_SQRTSS_XMM_to_XMM(xmmFt, xmmFt); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFt); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("SQRT Q, vf%02d%s", _Ft_, _Ftf_String); } } @@ -127,9 +127,9 @@ mVUop(mVU_RSQRT) { if (CHECK_VU_OVERFLOW) mVUclamp1(xmmFs, xmmFt, 8); x86SetJ8(djmp); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - if (writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); + if (mVUinfo.writeQ) SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe1); } pass3 { mVUlog("RSQRT Q, vf%02d%s, vf%02d%s", _Fs_, _Fsf_String, _Ft_, _Ftf_String); } } @@ -162,14 +162,14 @@ microVUt(void) mVU_EATAN_(mV) { EATANhelper(mVU_T8); SSE_ADDSS_M32_to_XMM(xmmPQ, (uptr)mVU_Pi4); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); } mVUop(mVU_EATAN) { pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 54); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_SUBSS_M32_to_XMM(xmmFs, (uptr)mVU_one); @@ -186,7 +186,7 @@ mVUop(mVU_EATANxy) { pass2 { getReg6(xmmFt, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFt, 0x01); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_SUBSS_XMM_to_XMM(xmmFs, xmmFt); // y-x, not y-1? >< @@ -203,7 +203,7 @@ mVUop(mVU_EATANxz) { pass2 { getReg6(xmmFt, _Fs_); SSE2_PSHUFD_XMM_to_XMM(xmmFs, xmmFt, 0x02); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_SUBSS_XMM_to_XMM(xmmFs, xmmFt); @@ -226,7 +226,7 @@ mVUop(mVU_EEXP) { pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 44); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_MULSS_M32_to_XMM(xmmPQ, (uptr)mVU_E1); SSE_ADDSS_M32_to_XMM(xmmPQ, (uptr)mVU_one); @@ -249,7 +249,7 @@ mVUop(mVU_EEXP) { SSE_MOVSS_M32_to_XMM(xmmT1, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmT1, xmmPQ); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmT1); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("EEXP P"); } } @@ -274,10 +274,10 @@ mVUop(mVU_ELENG) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 18); } pass2 { getReg6(xmmFs, _Fs_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance mVU_sumXYZ(); SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmPQ); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ELENG P"); } } @@ -286,12 +286,12 @@ mVUop(mVU_ERCPR) { pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 12); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ERCPR P"); } } @@ -300,13 +300,13 @@ mVUop(mVU_ERLENG) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 24); } pass2 { getReg6(xmmFs, _Fs_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance mVU_sumXYZ(); SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmPQ); SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ERLENG P"); } } @@ -315,13 +315,13 @@ mVUop(mVU_ERSADD) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 18); } pass2 { getReg6(xmmFs, _Fs_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance mVU_sumXYZ(); //SSE_RCPSS_XMM_to_XMM(xmmPQ, xmmPQ); // Lower Precision is bad? SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ERSADD P"); } } @@ -330,12 +330,12 @@ mVUop(mVU_ERSQRT) { pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 18); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmFs); SSE_MOVSS_M32_to_XMM(xmmFs, (uptr)mVU_one); SSE_DIVSS_XMM_to_XMM(xmmFs, xmmPQ); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ERSQRT P"); } } @@ -344,9 +344,9 @@ mVUop(mVU_ESADD) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 11); } pass2 { getReg6(xmmFs, _Fs_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance mVU_sumXYZ(); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ESADD P"); } } @@ -362,7 +362,7 @@ mVUop(mVU_ESIN) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 29); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); //SSE_MULSS_M32_to_XMM(xmmPQ, (uptr)mVU_one); // Multiplying by 1 is redundant? SSE_MOVAPS_XMM_to_XMM(xmmFt, xmmFs); @@ -379,7 +379,7 @@ mVUop(mVU_ESIN) { SSE_MULSS_XMM_to_XMM(xmmT1, xmmFt); SSE_MULSS_M32_to_XMM(xmmT1, (uptr)mVU_S5); SSE_ADDSS_XMM_to_XMM(xmmPQ, xmmT1); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ESIN P"); } } @@ -388,9 +388,9 @@ mVUop(mVU_ESQRT) { pass1 { mVUanalyzeEFU1(mVU, _Fs_, _Fsf_, 12); } pass2 { getReg5(xmmFs, _Fs_, _Fsf_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE_SQRTSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ESQRT P"); } } @@ -399,13 +399,13 @@ mVUop(mVU_ESUM) { pass1 { mVUanalyzeEFU2(mVU, _Fs_, 12); } pass2 { getReg6(xmmFs, _Fs_); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip xmmPQ to get Valid P instance SSE2_PSHUFD_XMM_to_XMM(xmmFt, xmmFs, 0x1b); SSE_ADDPS_XMM_to_XMM(xmmFs, xmmFt); SSE2_PSHUFD_XMM_to_XMM(xmmFt, xmmFs, 0x01); SSE_ADDSS_XMM_to_XMM(xmmFs, xmmFt); SSE_MOVSS_XMM_to_XMM(xmmPQ, xmmFs); - SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, writeP ? 0x27 : 0xC6); // Flip back + SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, mVUinfo.writeP ? 0x27 : 0xC6); // Flip back } pass3 { mVUlog("ESUM P"); } } @@ -417,7 +417,7 @@ mVUop(mVU_ESUM) { mVUop(mVU_FCAND) { pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(mVU, gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, cFLAG.read); AND32ItoR(gprT1, _Imm24_); ADD32ItoR(gprT1, 0xffffff); SHR32ItoR(gprT1, 24); @@ -430,7 +430,7 @@ mVUop(mVU_FCAND) { mVUop(mVU_FCEQ) { pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(mVU, gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, cFLAG.read); XOR32ItoR(gprT1, _Imm24_); SUB32ItoR(gprT1, 1); SHR32ItoR(gprT1, 31); @@ -443,7 +443,7 @@ mVUop(mVU_FCEQ) { mVUop(mVU_FCGET) { pass1 { mVUanalyzeCflag(mVU, _It_); } pass2 { - mVUallocCFLAGa(mVU, gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, cFLAG.read); AND32ItoR(gprT1, 0xfff); mVUallocVIb(mVU, gprT1, _It_); } @@ -454,7 +454,7 @@ mVUop(mVU_FCGET) { mVUop(mVU_FCOR) { pass1 { mVUanalyzeCflag(mVU, 1); } pass2 { - mVUallocCFLAGa(mVU, gprT1, fvcInstance); + mVUallocCFLAGa(mVU, gprT1, cFLAG.read); OR32ItoR(gprT1, _Imm24_); ADD32ItoR(gprT1, 1); // If 24 1's will make 25th bit 1, else 0 SHR32ItoR(gprT1, 24); // Get the 25th bit (also clears the rest of the garbage in the reg) @@ -465,10 +465,10 @@ mVUop(mVU_FCOR) { } mVUop(mVU_FCSET) { - pass1 { mVUinfo |= _doClip; } + pass1 { cFLAG.doFlag = 1; } pass2 { MOV32ItoR(gprT1, _Imm24_); - mVUallocCFLAGb(mVU, gprT1, fcInstance); + mVUallocCFLAGb(mVU, gprT1, cFLAG.write); } pass3 { mVUlog("FCSET $%x", _Imm24_); } } @@ -480,7 +480,7 @@ mVUop(mVU_FCSET) { mVUop(mVU_FMAND) { pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocMFLAGa(mVU, gprT1, mFLAG.read); mVUallocVIa(mVU, gprT2, _Is_); AND16RtoR(gprT1, gprT2); mVUallocVIb(mVU, gprT1, _It_); @@ -492,7 +492,7 @@ mVUop(mVU_FMAND) { mVUop(mVU_FMEQ) { pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocMFLAGa(mVU, gprT1, mFLAG.read); mVUallocVIa(mVU, gprT2, _Is_); XOR32RtoR(gprT1, gprT2); SUB32ItoR(gprT1, 1); @@ -506,7 +506,7 @@ mVUop(mVU_FMEQ) { mVUop(mVU_FMOR) { pass1 { mVUanalyzeMflag(mVU, _Is_, _It_); } pass2 { - mVUallocMFLAGa(mVU, gprT1, fvmInstance); + mVUallocMFLAGa(mVU, gprT1, mFLAG.read); mVUallocVIa(mVU, gprT2, _Is_); OR16RtoR(gprT1, gprT2); mVUallocVIb(mVU, gprT1, _It_); @@ -522,7 +522,7 @@ mVUop(mVU_FMOR) { mVUop(mVU_FSAND) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, sFLAG.read); AND16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -533,7 +533,7 @@ mVUop(mVU_FSAND) { mVUop(mVU_FSEQ) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, sFLAG.read); XOR16ItoR(gprT1, _Imm12_); SUB16ItoR(gprT1, 1); SHR16ItoR(gprT1, 15); @@ -546,7 +546,7 @@ mVUop(mVU_FSEQ) { mVUop(mVU_FSOR) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, fvsInstance); + mVUallocSFLAGa(gprT1, sFLAG.read); OR16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -558,8 +558,8 @@ mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { int flagReg1, flagReg2; - getFlagReg(flagReg1, fsInstance); - if (!(doStatus||doDivFlag)) { getFlagReg(flagReg2, fpsInstance); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction + getFlagReg(flagReg1, sFLAG.write); + if (!(sFLAG.doFlag||mVUinfo.doDivFlag)) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction AND32ItoR(flagReg1, 0x03f); OR32ItoR (flagReg1, (_Imm12_ & 0xfc0)); } @@ -664,7 +664,7 @@ mVUop(mVU_ISUBIU) { //------------------------------------------------------------------ mVUop(mVU_MFIR) { - pass1 { if (!_Ft_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } + pass1 { if (!_Ft_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } pass2 { mVUallocVIa(mVU, gprT1, _Is_); MOVSX32R16toR(gprT1, gprT1); @@ -704,7 +704,7 @@ mVUop(mVU_MR32) { } mVUop(mVU_MTIR) { - pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->VF[_Fs_].UL[_Fsf_]); mVUallocVIb(mVU, gprT1, _It_); @@ -717,7 +717,7 @@ mVUop(mVU_MTIR) { //------------------------------------------------------------------ mVUop(mVU_ILW) { - pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + getVUmem(_Imm11_) + offsetSS); @@ -735,7 +735,7 @@ mVUop(mVU_ILW) { } mVUop(mVU_ILWR) { - pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + offsetSS); @@ -828,7 +828,7 @@ mVUop(mVU_LQ) { mVUop(mVU_LQD) { pass1 { mVUanalyzeLQ(mVU, _Ft_, _Is_, 1); } pass2 { - if (!_Is_ && !noWriteVF) { + if (!_Is_ && !mVUlow.noWriteVF) { mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } @@ -836,7 +836,7 @@ mVUop(mVU_LQD) { mVUallocVIa(mVU, gprT1, _Is_); SUB16ItoR(gprT1, 1); mVUallocVIb(mVU, gprT1, _Is_); - if (!noWriteVF) { + if (!mVUlow.noWriteVF) { mVUaddrFix(mVU, gprT1); mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); @@ -849,13 +849,13 @@ mVUop(mVU_LQD) { mVUop(mVU_LQI) { pass1 { mVUanalyzeLQ(mVU, _Ft_, _Is_, 1); } pass2 { - if (!_Is_ && !noWriteVF) { + if (!_Is_ && !mVUlow.noWriteVF) { mVUloadReg(xmmFt, (uptr)mVU->regs->Mem, _X_Y_Z_W); mVUsaveReg(xmmFt, (uptr)&mVU->regs->VF[_Ft_].UL[0], _X_Y_Z_W, 1); } else { - mVUallocVIa(mVU, (!noWriteVF) ? gprT1 : gprT2, _Is_); - if (!noWriteVF) { + mVUallocVIa(mVU, (!mVUlow.noWriteVF) ? gprT1 : gprT2, _Is_); + if (!mVUlow.noWriteVF) { MOV32RtoR(gprT2, gprT1); mVUaddrFix(mVU, gprT1); mVUloadReg2(xmmFt, gprT1, (uptr)mVU->regs->Mem, _X_Y_Z_W); @@ -948,7 +948,7 @@ mVUop(mVU_RINIT) { } microVUt(void) mVU_RGET_(mV, int Rreg) { - if (!noWriteVF) { + if (!mVUlow.noWriteVF) { if (_X) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[0], Rreg); if (_Y) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[1], Rreg); if (_Z) MOV32RtoM((uptr)&mVU->regs->VF[_Ft_].UL[2], Rreg); @@ -1017,7 +1017,7 @@ mVUop(mVU_WAITQ) { //------------------------------------------------------------------ mVUop(mVU_XTOP) { - pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->top); mVUallocVIb(mVU, gprT1, _It_); @@ -1026,7 +1026,7 @@ mVUop(mVU_XTOP) { } mVUop(mVU_XITOP) { - pass1 { if (!_It_) { mVUinfo |= _isNOP; } analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->itop); mVUallocVIb(mVU, gprT1, _It_); @@ -1073,8 +1073,8 @@ microVUt(void) mVU_XGKICK_DELAY(mV, bool memVI) { mVUop(mVU_XGKICK) { pass1 { mVUanalyzeXGkick(mVU, _Is_, mVU_XGKICK_CYCLES); } pass2 { - if (!mVU_XGKICK_CYCLES) { mVU_XGKICK_DELAY(mVU, 0); return; } - else if (doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); mVUinfo &= ~_doXGKICK; } + if (!mVU_XGKICK_CYCLES) { mVU_XGKICK_DELAY(mVU, 0); return; } + else if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); mVUinfo.doXGKICK = 0; } mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->VIxgkick, gprT1); } @@ -1086,7 +1086,7 @@ mVUop(mVU_XGKICK) { //------------------------------------------------------------------ #define setBranchA(x, _x_) { \ - pass1 { if (_Imm11_ == 1 && !_x_) { mVUinfo |= _isNOP; return; } mVUbranch = x; } \ + pass1 { if (_Imm11_ == 1 && !_x_) { mVUlow.isNOP = 1; return; } mVUbranch = x; } \ pass2 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ pass3 { mVUbranch = x; } \ pass4 { if (_Imm11_ == 1 && !_x_) { return; } mVUbranch = x; } \ @@ -1111,9 +1111,9 @@ mVUop(mVU_IBEQ) { setBranchA(3, 0); pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); - if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1124,7 +1124,7 @@ mVUop(mVU_IBGEZ) { setBranchA(4, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1135,7 +1135,7 @@ mVUop(mVU_IBGTZ) { setBranchA(5, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1146,7 +1146,7 @@ mVUop(mVU_IBLEZ) { setBranchA(6, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1157,7 +1157,7 @@ mVUop(mVU_IBLTZ) { setBranchA(7, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1168,9 +1168,9 @@ mVUop(mVU_IBNE) { setBranchA(8, 0); pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { - if (memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else mVUallocVIa(mVU, gprT1, _Is_); - if (memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 00b0530c6c..d36cfc7a8e 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -172,8 +172,15 @@ declareAllVariables #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC #define mVUsFlagHack mVUallocInfo.sFlagHack + #define mVUinfo mVUallocInfo.info[iPC / 2] -#define mVUstall mVUallocInfo.stall[iPC / 2] +#define mVUstall mVUinfo.stall +#define mVUup mVUinfo.uOp +#define mVUlow mVUinfo.lOp +#define sFLAG mVUinfo.sFlag +#define mFLAG mVUinfo.mFlag +#define cFLAG mVUinfo.cFlag + #define mVUstartPC mVUallocInfo.startPC #define mVUflagInfo mVUregs.needExactMatch #define mVUflagHack (mVUcurProg.sFlagHack) @@ -194,70 +201,6 @@ declareAllVariables #define __Mac (mVUflagInfo & (0xf<<4)) #define __Clip (mVUflagInfo & (0xf<<8)) -// Pass 1 uses these to set mVUinfo -#define _isNOP (1<<0) // Skip Lower Instruction -#define _isBranch (1<<1) // Cur Instruction is a Branch -#define _isEOB (1<<2) // End of Block -#define _isBdelay (1<<3) // Cur Instruction in Branch Delay slot -#define _isSflag (1<<4) // Cur Instruction uses status flag -#define _doXGKICK (1<<5) // Do XGKICK transfer on this instruction -#define _writeQ (1<<6) -#define _readQ (1<<6) // same as writeQ -#define _writeP (1<<7) -#define _readP (1<<7) // same as writeP -#define _doFlags (3<<8) -#define _doMac (1<<8) -#define _doStatus (1<<9) -#define _fmInstance (3<<10) // Mac Write Instance -#define _fsInstance (3<<12) // Status Write Instance -#define _fcInstance (3<<14) // Clip Write Instance -#define _fpsInstance (3<<12) // Prev.S. Write Instance -#define _fpcInstance (3<<14) // Prev.C. Write Instance -#define _fvmInstance (3<<16) // Mac Read Instance (at T-stage for lower instruction) -#define _fvsInstance (3<<18) // Status Read Instance (at T-stage for lower instruction) -#define _fvcInstance (3<<20) // Clip Read Instance (at T-stage for lower instruction) -#define _backupVI (1<<22) // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) -#define _memReadIs (1<<23) // Read Is (VI reg) from memory (used by branches) -#define _memReadIt (1<<24) // Read If (VI reg) from memory (used by branches) -#define _writesVI (1<<25) // Current Instruction writes to VI (used by branches; note that flag-modifying opcodes shouldn't set this) -#define _swapOps (1<<26) // Runs Lower Instruction Before Upper Instruction -#define _isFSSET (1<<27) // Cur Instruction is FSSET -#define _doDivFlag (1<<28) // Transfer Div flag to Status Flag -#define _doClip (1<<29) -#define _noWriteVF (1<<30) // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) - -// Pass 2 uses these to read mVUinfo -#define isNOP (mVUinfo & (1<<0)) -#define isBranch (mVUinfo & (1<<1)) -#define isEOB (mVUinfo & (1<<2)) -#define isBdelay (mVUinfo & (1<<3)) -#define isSflag (mVUinfo & (1<<4)) -#define doXGKICK (mVUinfo & (1<<5)) -#define readQ ((mVUinfo >> 6) & 1) // same as writeQ -#define readP ((mVUinfo >> 7) & 1) // same as writeP -#define writeQ (((mVUinfo >> 6) + 1) & 1) -#define writeP (((mVUinfo >> 7) + 1) & 1) -#define doFlags (mVUinfo & (3<<8)) -#define doMac (mVUinfo & (1<<8)) -#define doStatus (mVUinfo & (1<<9)) -#define fmInstance ((mVUinfo >> 10) & 3) -#define fsInstance ((mVUinfo >> 12) & 3) -#define fpsInstance ((((mVUinfo>>12) & 3) - 1) & 0x3) -#define fcInstance ((mVUinfo >> 14) & 3) -#define fpcInstance ((((mVUinfo>>14) & 3) - 1) & 0x3) -#define fvmInstance ((mVUinfo >> 16) & 3) -#define fvsInstance ((mVUinfo >> 18) & 3) -#define fvcInstance ((mVUinfo >> 20) & 3) -#define backupVI (mVUinfo & (1<<22)) -#define memReadIs (mVUinfo & (1<<23)) -#define memReadIt (mVUinfo & (1<<24)) -#define writesVI (mVUinfo & (1<<25)) -#define swapOps (mVUinfo & (1<<26)) -#define isFSSET (mVUinfo & (1<<27)) -#define doDivFlag (mVUinfo & (1<<28)) -#define doClip (mVUinfo & (1<<29)) -#define noWriteVF (mVUinfo & (1<<30)) - // Pass 3 Helper Macros #define _Fsf_String ((_Fsf_ == 3) ? "w" : ((_Fsf_ == 2) ? "z" : ((_Fsf_ == 1) ? "y" : "x"))) #define _Ftf_String ((_Ftf_ == 3) ? "w" : ((_Ftf_ == 2) ? "z" : ((_Ftf_ == 1) ? "y" : "x"))) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 4d3ddc8fba..f6d2c58b00 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -305,8 +305,8 @@ microVUt(void) mVUcheckSflag(mV, int progIndex) { mVU->code = mVU->prog.prog[progIndex].data[i]; mVUopL(mVU, 3); } - mVUflagInfo = bFlagInfo; - mVU->code = bCode; + mVUflagInfo = bFlagInfo; + mVU->code = bCode; mVU->prog.prog[progIndex].sFlagHack = mVUsFlagHack; } } diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 30591004f4..c2bb91e294 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -22,7 +22,7 @@ // mVUupdateFlags() - Updates status/mac flags //------------------------------------------------------------------ -#define AND_XYZW ((_XYZW_SS && modXYZW) ? (1) : (doMac ? (_X_Y_Z_W) : (flipMask[_X_Y_Z_W]))) +#define AND_XYZW ((_XYZW_SS && modXYZW) ? (1) : (mFLAG.doFlag ? (_X_Y_Z_W) : (flipMask[_X_Y_Z_W]))) #define ADD_XYZW ((_XYZW_SS && modXYZW) ? (_X ? 3 : (_Y ? 2 : (_Z ? 1 : 0))) : 0) #define SHIFT_XYZW(gprReg) { if (_XYZW_SS && modXYZW && !_W) { SHL32ItoR(gprReg, ADD_XYZW); } } @@ -32,14 +32,14 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; - //SysPrintf("doStatus = %d; doMac = %d\n", doStatus>>9, doMac>>8); - if (mVUsFlagHack) { mVUinfo &= ~_doStatus; } - if (!doFlags) return; - if (!doMac || (_XYZW_SS && modXYZW)) { regT1 = reg; } + //SysPrintf("Status = %d; Mac = %d\n", sFLAG.doFlag, mFLAG.doFlag); + if (mVUsFlagHack) { sFLAG.doFlag = 0; } + if (!sFLAG.doFlag && !mFLAG.doFlag) { return; } + if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw - if (doStatus) { - getFlagReg(sReg, fsInstance); // Set sReg to valid GPR by Cur Flag Instance - mVUallocSFLAGa(sReg, fpsInstance); // Get Prev Status Flag + if (sFLAG.doFlag) { + getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance + mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags } @@ -54,25 +54,25 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation - if (doStatus) pjmp = JZ8(0); // Skip if none are - if (doMac) SHL32ItoR(mReg, 4 + ADD_XYZW); - if (doStatus) OR32ItoR(sReg, 0x82); // SS, S flags - if (_XYZW_SS && doStatus) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - if (doStatus) x86SetJ8(pjmp); + if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); + if (sFLAG.doFlag) OR32ItoR(sReg, 0x82); // SS, S flags + if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation - if (doStatus) pjmp = JZ8(0); // Skip if none are - if (doMac) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (doStatus) { OR32ItoR(sReg, 0x41); } // ZS, Z flags - if (doStatus) x86SetJ8(pjmp); + if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } + if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags + if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ - if (_XYZW_SS && doStatus) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - - if (doMac) mVUallocMFLAGb(mVU, mReg, fmInstance); // Set Mac Flag + if (sFLAG.doFlag && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here + + if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag } //------------------------------------------------------------------ @@ -438,9 +438,9 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool } // FMAC27~29 - MAX/MINI FMAC Opcodes -#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } -#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } -#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { mVUinfo &= ~_doStatus; } } +#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } +#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } +#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } //------------------------------------------------------------------ // Micro VU Micromode Upper instructions @@ -587,7 +587,7 @@ mVUop(mVU_CLIP) { pass2 { int Fs, Ft; mVUallocFMAC17a(mVU, Fs, Ft); - mVUallocCFLAGa(mVU, gprT1, fpcInstance); + mVUallocCFLAGa(mVU, gprT1, cFLAG.lastWrite); SHL32ItoR(gprT1, 6); SSE_ANDPS_M128_to_XMM(Ft, (uptr)mVU_absclip); @@ -611,7 +611,7 @@ mVUop(mVU_CLIP) { OR32RtoR (gprT1, gprT2); AND32ItoR(gprT1, 0xffffff); - mVUallocCFLAGb(mVU, gprT1, fcInstance); + mVUallocCFLAGb(mVU, gprT1, cFLAG.write); } pass3 { mVUlog("CLIP"); mVUlogCLIP(); } } From e85124dff5f2952f9bbca0772d0174eb93ad907e Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 31 May 2009 10:37:20 +0000 Subject: [PATCH 162/296] microVU: - Fixed a bug from r1254. (Fixes weird circles on the logo of ffxii after selecting new-game) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1299 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Lower.inl | 1 + 1 file changed, 1 insertion(+) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 848bcb3d52..c7a4b7e0ca 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -980,6 +980,7 @@ mVUop(mVU_RNEXT) { XOR32RtoR(gprR, gprT1); AND32ItoR(gprR, 0x007fffff); OR32ItoR (gprR, 0x3f800000); + MOV32RtoM(Rmem, gprR); mVU_RGET_(mVU, gprR); MOV32ItoR(gprR, Roffset); // Restore gprR } From ebff9ec1c5323527c3175d6d87147a8f49f0809e Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 1 Jun 2009 01:21:01 +0000 Subject: [PATCH 163/296] microVU: - Tried a complex optimization with status flag updating. Didn't seem to work out as well as I thought. I need some benchmarks of mVU (w/o hacks) before this update and after this update to see how effective it is. If its not much faster I might just revert the change (because I'm sure no-one is going to understand how it works this way). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1300 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 7 +++-- pcsx2/x86/microVU_Alloc.inl | 23 +++------------- pcsx2/x86/microVU_Analyze.inl | 32 +++++++++++---------- pcsx2/x86/microVU_Compile.inl | 12 ++++---- pcsx2/x86/microVU_Execute.inl | 10 +++---- pcsx2/x86/microVU_Flags.inl | 37 ++++++++++++------------- pcsx2/x86/microVU_IR.h | 4 ++- pcsx2/x86/microVU_Lower.inl | 47 +++++++++++++++---------------- pcsx2/x86/microVU_Misc.h | 37 +++++++++++++++---------- pcsx2/x86/microVU_Misc.inl | 1 - pcsx2/x86/microVU_Upper.inl | 52 +++++++++++++++++++++-------------- 11 files changed, 134 insertions(+), 128 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7fd6efc281..7631cc9347 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -98,9 +98,10 @@ struct microProgManager { #define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) struct microVU { - PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) - PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) - PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ + PCSX2_ALIGNED16(u32 statusFlag[4]); // 4 instances of status flag (used in execution) + PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) + PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) + PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index eb0c5deda0..a04bc48008 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -634,28 +634,13 @@ microVUt(void) mVUallocFMAC26b(mV, int& ACCw, int& ACCr) { // Flag Allocators //------------------------------------------------------------------ -#define getFlagReg(regX, fInst) { \ - switch (fInst) { \ - case 0: regX = gprF0; break; \ - case 1: regX = gprF1; break; \ - case 2: regX = gprF2; break; \ - case 3: regX = gprF3; break; \ - default: \ - Console::Error("microVU: Flag Instance Error (fInst = %d)", params fInst); \ - regX = gprF0; \ - break; \ - } \ +microVUt(void) mVUallocSFLAGa(mV, int reg, int fInstance) { + MOVZX32M16toR(reg, (uptr)&mVU->statusFlag[fInstance]); } -microVUt(void) mVUallocSFLAGa(int reg, int fInstance) { - getFlagReg(fInstance, fInstance); - MOVZX32R16toR(reg, fInstance); -} - -microVUt(void) mVUallocSFLAGb(int reg, int fInstance) { - getFlagReg(fInstance, fInstance); +microVUt(void) mVUallocSFLAGb(mV, int reg, int fInstance) { //AND32ItoR(reg, 0xffff); - MOV32RtoR(fInstance, reg); + MOV32RtoM((uptr)&mVU->statusFlag[fInstance], reg); } microVUt(void) mVUallocMFLAGa(mV, int reg, int fInstance) { diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 74c1cbcfc9..8af22c363f 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -62,7 +62,8 @@ } microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { - sFLAG.doFlag = 1; + mVUup.doFlags = 1; + sFLAG.doSticky = 1; analyzeReg1(Fs); analyzeReg1(Ft); analyzeReg2(Fd, 0); @@ -91,7 +92,8 @@ microVUt(void) mVUanalyzeFMAC2(mV, int Fs, int Ft) { } microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { - sFLAG.doFlag = 1; + mVUup.doFlags = 1; + sFLAG.doSticky = 1; analyzeReg1(Fs); analyzeReg3(Ft); analyzeReg2(Fd, 0); @@ -269,6 +271,15 @@ microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { // Sflag - Status Flag Opcodes //------------------------------------------------------------------ +#define setFlagInst(xDoFlag) { \ + int curPC = iPC; \ + for (int i = mVUcount, j = 0; i > 0; i--, j++) { \ + incPC2(-2); \ + if (mVUup.doFlags) { xDoFlag = 1; if (j >= 3) { break; } } \ + } \ + iPC = curPC; \ +} + microVUt(void) mVUanalyzeSflag(mV, int It) { if (!It) { mVUlow.isNOP = 1; } else { @@ -279,17 +290,15 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { // Note: useSflag is used for status flag optimizations when a FSSET instruction is called. // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. + setFlagInst(sFLAG.doFlag); } analyzeVIreg3(It, 1); } microVUt(void) mVUanalyzeFSSET(mV) { - mVUlow.isFSSET = 1; - // mVUinfo &= ~_doStatus; - // Note: I'm not entirely sure if the non-sticky flags - // should be taken from the current upper instruction - // or if they should be taken from the previous instruction - // Uncomment the above line if the latter-case is true + mVUinfo.swapOps = 1; + mVUlow.isFSSET = 1; + sFLAG.doSticky = 0; } //------------------------------------------------------------------ @@ -301,12 +310,7 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } - int curPC = iPC; - for (int i = mVUcount, j = 0; i > 0; i--, j++) { - incPC2(-2); - if (sFLAG.doFlag) { mFLAG.doFlag = 1; if (j >= 3) { break; } } - } - iPC = curPC; + setFlagInst(mFLAG.doFlag); } analyzeVIreg1(Is); analyzeVIreg3(It, 1); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 689d8b82a4..f4674fc61c 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -59,7 +59,7 @@ #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } -#define doUpperOp() { mVUopU(mVU, 1); mVUdivSet(mVU); } +#define doUpperOp() { mVUdivSet(mVU); mVUopU(mVU, 1); } #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } #define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } @@ -169,8 +169,7 @@ microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, in // Save Flag Instances if (!mVUflagHack) { - getFlagReg(fStatus, fStatus); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprST); } mVUallocMFLAGa(mVU, gprT1, fMac); mVUallocCFLAGa(mVU, gprT2, fClip); @@ -200,7 +199,6 @@ microVUt(void) mVUtestCycles(mV) { MOV32ItoR(gprT2, xPC); if (!isVU1) CALLFunc((uptr)mVUwarning0); else CALLFunc((uptr)mVUwarning1); - MOV32ItoR(gprR, Roffset); // Restore gprR mVUendProgram(mVU, 0, 0, sI, 0, cI); x86SetJ8(jmp8); } @@ -322,7 +320,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbackupRegs(mVU); MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) - MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) + MOV32ItoR(gprT3, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) if (!isVU1) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) else CALLFunc((uptr)mVUcompileVU1); @@ -375,12 +373,12 @@ eBitTemination: memset(&mVUinfo, 0, sizeof(mVUinfo)); incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; - if (mVUinfo.doDivFlag) { + /*if (mVUinfo.doDivFlag) { int flagReg; getFlagReg(flagReg, lStatus); AND32ItoR (flagReg, 0x0fcf); OR32MtoR (flagReg, (uptr)&mVU->divFlag); - } + }*/ if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } // Do E-bit end stuff here diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 0854b0bad3..8c5bb00c6b 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -41,11 +41,11 @@ microVUt(void) mVUdispatcherA(mV) { // Load Regs MOV32ItoR(gprR, Roffset); // Load VI Reg Offset - MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); - AND32ItoR(gprF0, 0xffff); - MOV32RtoR(gprF1, gprF0); - MOV32RtoR(gprF2, gprF0); - MOV32RtoR(gprF3, gprF0); + MOV32MtoR(gprST, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); + /*AND32ItoR(gprST, 0xffff); + MOV32RtoR(gprF1, gprST); + MOV32RtoR(gprF2, gprST); + MOV32RtoR(gprF3, gprST);*/ SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index fc63c3db9d..70f50f05fa 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -20,12 +20,9 @@ // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet(mV) { - int flagReg1, flagReg2; if (mVUinfo.doDivFlag) { - getFlagReg(flagReg1, sFLAG.write); - if (!sFLAG.doFlag) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } - AND32ItoR(flagReg1, 0x0fcf); - OR32MtoR (flagReg1, (uptr)&mVU->divFlag); + AND32ItoR(gprST, 0xfffcffff); // Clear D/I bits + OR32MtoR (gprST, (uptr)&mVU->divFlag); // Set DS/IS/D/I bits } } @@ -34,18 +31,19 @@ microVUt(void) mVUstatusFlagOp(mV) { int curPC = iPC; int i = mVUcount; bool runLoop = 1; - if (sFLAG.doFlag) { mVUlow.useSflag = 1; } + if (mVUup.doFlags) { mVUlow.useSflag = 1; } else { for (; i > 0; i--) { incPC2(-2); if (mVUlow.useSflag) { runLoop = 0; break; } - if (sFLAG.doFlag) { mVUlow.useSflag = 1; break; } + if (mVUup.doFlags) { mVUlow.useSflag = 1; break; } } } if (runLoop) { for (; i > 0; i--) { incPC2(-2); if (mVUlow.useSflag) break; + sFLAG.doSticky = 0; sFLAG.doFlag = 0; } } @@ -79,7 +77,11 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { // Ensure last ~4+ instructions update mac flags (if next block's first 4 instructions will read them) for (int i = mVUcount; i > 0; i--, aCount++) { - if (sFLAG.doFlag) { if (__Mac) { mFLAG.doFlag = 1; } if (aCount >= 4) { break; } } + if (mVUup.doFlags) { + if (__Status) { sFLAG.doFlag = 1; } + if (__Mac) { mFLAG.doFlag = 1; } + if (aCount >= 4) { break; } + } incPC2(-2); } @@ -142,13 +144,13 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { incPC2(2); } - mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) | ((__Status) ? 0 : xS); + mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) /*| ((__Status) ? 0 : xS)*/; return cycles; } -#define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) -#define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) -#define shuffleClip ((bClip[3]<<6)|(bClip[2]<<4)|(bClip[1]<<2)|bClip[0]) +#define shuffleStatus ((bStatus[3]<<6)|(bStatus[2]<<4)|(bStatus[1]<<2)|bStatus[0]) +#define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) +#define shuffleClip ((bClip [3]<<6)|(bClip [2]<<4)|(bClip [1]<<2)|bClip [0]) // Recompiles Code for Proper Flags on Block Linkings microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles) { @@ -156,14 +158,9 @@ microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles if (__Status && !mVUflagHack) { int bStatus[4]; sortFlag(xStatus, bStatus, cycles); - MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); - MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); - MOV32RtoR(gprR, getFlagReg1(bStatus[2])); - MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); - MOV32RtoR(gprF0, gprT1); - MOV32RtoR(gprF1, gprT2); - MOV32RtoR(gprF2, gprR); - MOV32ItoR(gprR, Roffset); // Restore gprR + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->statusFlag); + SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleStatus); + SSE_MOVAPS_XMM_to_M128((uptr)mVU->statusFlag, xmmT1); } if (__Mac) { diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index a42a683a25..2ccd33b907 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -68,6 +68,7 @@ struct microBlock { struct microUpperOp { bool eBit; // Has E-bit set bool iBit; // Has I-bit set + bool doFlags; // This instruction updates Status/Mac Flags }; struct microLowerOp { @@ -83,7 +84,8 @@ struct microLowerOp { }; struct microFlagInst { - bool doFlag; // Update Flag on this Instruction + bool doSticky; // Update Sticky Flags (Status Flag Only) + bool doFlag; // Update Flag on this Instruction (For Status Flag, this means non-sticky bits) u8 write; // Points to the instance that should be written to (s-stage write) u8 lastWrite; // Points to the instance that was last written to (most up-to-date flag) u8 read; // Points to the instance that should be read by a lower instruction (t-stage read) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index c7a4b7e0ca..1bfd48e6e0 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -37,7 +37,7 @@ SSE_MOVMSKPS_XMM_to_R32(gprTemp, xmmReg); \ TEST32ItoR(gprTemp, 1); /* Check sign bit */ \ aJump = JZ8(0); /* Skip if positive */ \ - MOV32ItoM((uptr)&mVU->divFlag, 0x410); /* Set Invalid Flags */ \ + MOV32ItoM((uptr)&mVU->divFlag, 0x410000); /* Set Invalid Flags */ \ SSE_ANDPS_M128_to_XMM(xmmReg, (uptr)mVU_absclip); /* Abs(xmmReg) */ \ x86SetJ8(aJump); \ } @@ -54,10 +54,10 @@ mVUop(mVU_DIV) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero ajmp = JZ8(0); - MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x410000); // Set invalid flag (0/0) bjmp = JMP8(0); x86SetJ8(ajmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x820000); // Zero divide (only when not 0/0) x86SetJ8(bjmp); SSE_XORPS_XMM_to_XMM(xmmFs, xmmFt); @@ -112,10 +112,10 @@ mVUop(mVU_RSQRT) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero bjmp = JZ8(0); // Skip if none are - MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x410000); // Set invalid flag (0/0) cjmp = JMP8(0); x86SetJ8(bjmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide flag (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x820000); // Zero divide flag (only when not 0/0) x86SetJ8(cjmp); SSE_ANDPS_M128_to_XMM(xmmFs, (uptr)mVU_signbit); @@ -522,7 +522,7 @@ mVUop(mVU_FMOR) { mVUop(mVU_FSAND) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); + mVUallocSFLAGa(mVU, gprT1, sFLAG.read); AND16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -533,7 +533,7 @@ mVUop(mVU_FSAND) { mVUop(mVU_FSEQ) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); + mVUallocSFLAGa(mVU, gprT1, sFLAG.read); XOR16ItoR(gprT1, _Imm12_); SUB16ItoR(gprT1, 1); SHR16ItoR(gprT1, 15); @@ -546,7 +546,7 @@ mVUop(mVU_FSEQ) { mVUop(mVU_FSOR) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); + mVUallocSFLAGa(mVU, gprT1, sFLAG.read); OR16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -557,11 +557,13 @@ mVUop(mVU_FSOR) { mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { - int flagReg1, flagReg2; - getFlagReg(flagReg1, sFLAG.write); - if (!(sFLAG.doFlag||mVUinfo.doDivFlag)) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction - AND32ItoR(flagReg1, 0x03f); - OR32ItoR (flagReg1, (_Imm12_ & 0xfc0)); + int mask; + if (_Imm12_ & 0x800) mask |= 0x800000; + if (_Imm12_ & 0x400) mask |= 0x400000; + if (_Imm12_ & 0x080) mask |= 0x0000f0; + if (_Imm12_ & 0xc40) mask |= 0x00000f; + AND32ItoR(gprST, 0x30000); + if (mask) OR32ItoR(gprST, mask); } pass3 { mVUlog("FSSET $%x", _Imm12_); } pass4 { mVUsFlagHack = 0; } @@ -966,23 +968,22 @@ mVUop(mVU_RNEXT) { pass1 { mVUanalyzeR2(mVU, _Ft_, 0); } pass2 { // algorithm from www.project-fao.org - MOV32MtoR(gprR, Rmem); - MOV32RtoR(gprT1, gprR); + MOV32MtoR(gprT3, Rmem); + MOV32RtoR(gprT1, gprT3); SHR32ItoR(gprT1, 4); AND32ItoR(gprT1, 1); - MOV32RtoR(gprT2, gprR); + MOV32RtoR(gprT2, gprT3); SHR32ItoR(gprT2, 22); AND32ItoR(gprT2, 1); - SHL32ItoR(gprR, 1); + SHL32ItoR(gprT3, 1); XOR32RtoR(gprT1, gprT2); - XOR32RtoR(gprR, gprT1); - AND32ItoR(gprR, 0x007fffff); - OR32ItoR (gprR, 0x3f800000); - MOV32RtoM(Rmem, gprR); - mVU_RGET_(mVU, gprR); - MOV32ItoR(gprR, Roffset); // Restore gprR + XOR32RtoR(gprT3, gprT1); + AND32ItoR(gprT3, 0x007fffff); + OR32ItoR (gprT3, 0x3f800000); + MOV32RtoM(Rmem, gprT3); + mVU_RGET_(mVU, gprT3); } pass3 { mVUlog("RNEXT.%s vf%02d, R", _XYZW_String, _Ft_); } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index d36cfc7a8e..8d9ad3a747 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -131,12 +131,22 @@ declareAllVariables #define gprT1 0 // Temp Reg #define gprT2 1 // Temp Reg -#define gprR 2 // VI Reg Offset -#define gprF0 3 // Status Flag 0 +#define gprT3 2 // Temp Reg #define gprESP 4 // Don't use? -#define gprF1 5 // Status Flag 1 -#define gprF2 6 // Status Flag 2 -#define gprF3 7 // Status Flag 3 +#define gprT4 5 // Temp? +#define gprT5 6 // Temp? +#define gprR 7 // VI Reg Offset +#define gprST 3 // Status Sticky Flag +// gprST's Info is Stored as follows: +//----------------------------------------------------------------------------- +//|23 22 21 20||19 18 17 16||15 14 13 12||11 10 09 08||07 06 05 04||03 02 01 00| +//|DS|IS| || | D| I|| OS || US || SS || ZS | +//----------------------------------------------------------------------------- +// Storing Flags this way eliminates Jumps when updating sticky flags. +// +// When a Status Flag will be read, gprST is attached with +// the current status flag result in mVUupdateFlags. And the complete +// Status flag instance is stored in memory (mVU->statusFlag[instance]) // Function Params #define mP microVU* mVU, int recPass @@ -172,15 +182,13 @@ declareAllVariables #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC #define mVUsFlagHack mVUallocInfo.sFlagHack - -#define mVUinfo mVUallocInfo.info[iPC / 2] -#define mVUstall mVUinfo.stall -#define mVUup mVUinfo.uOp -#define mVUlow mVUinfo.lOp -#define sFLAG mVUinfo.sFlag -#define mFLAG mVUinfo.mFlag -#define cFLAG mVUinfo.cFlag - +#define mVUinfo mVUallocInfo.info[iPC/2] // IR info for current 64bit instruction +#define mVUstall mVUinfo.stall // Stall info for current instruction +#define mVUup mVUinfo.uOp // Upper Instruction Info +#define mVUlow mVUinfo.lOp // Lower Instruction Info +#define sFLAG mVUinfo.sFlag // Status Flag info for cur instruction +#define mFLAG mVUinfo.mFlag // Mac Flag info for cur instruction +#define cFLAG mVUinfo.cFlag // Clip Flag info for cur instruction #define mVUstartPC mVUallocInfo.startPC #define mVUflagInfo mVUregs.needExactMatch #define mVUflagHack (mVUcurProg.sFlagHack) @@ -268,6 +276,5 @@ declareAllVariables MOV32ItoR(gprT2, xPC); \ if (isEndPC) { CALLFunc((uptr)mVUprintPC2); } \ else { CALLFunc((uptr)mVUprintPC1); } \ - MOV32ItoR(gprR, Roffset); \ } \ } diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index f6d2c58b00..f7b0c1f3bc 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -292,7 +292,6 @@ microVUt(void) mVUrestoreRegs(mV) { SSE_MOVAPS_M128_to_XMM(xmmPQ, (uptr)&mVU->xmmPQb[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); SSE_MOVAPS_M128_to_XMM(xmmMin, (uptr)mVU_minvals); - MOV32ItoR(gprR, Roffset); // Restore gprR } // Reads entire microProgram and finds out if Status Flag is Used diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index c2bb91e294..578df21d49 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -28,21 +28,16 @@ // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool modXYZW) { - int sReg, mReg = gprT1; + int sReg = gprT3, mReg = gprT1; static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; //SysPrintf("Status = %d; Mac = %d\n", sFLAG.doFlag, mFLAG.doFlag); - if (mVUsFlagHack) { sFLAG.doFlag = 0; } - if (!sFLAG.doFlag && !mFLAG.doFlag) { return; } - if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } + if (mVUsFlagHack) { sFLAG.doSticky = 0; sFLAG.doFlag = 0; } + if (!mVUup.doFlags || (!sFLAG.doSticky && !sFLAG.doFlag && !mFLAG.doFlag)) { return; } + if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw - if (sFLAG.doFlag) { - getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance - mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag - AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags - } - + if (sFLAG.doFlag) { XOR32RtoR(sReg, sReg); } //-------------------------Check for Signed flags------------------------------ // The following code makes sure the Signed Bit isn't set with Negative Zero @@ -55,24 +50,41 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); - if (sFLAG.doFlag) OR32ItoR(sReg, 0x82); // SS, S flags - if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + if (mFLAG.doFlag || sFLAG.doSticky) SHL32ItoR(mReg, 4 + ADD_XYZW); + if (sFLAG.doFlag) OR32ItoR (sReg, 0x82); // SS, S flags + if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags + if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } + if (sFLAG.doSticky && !mFLAG.doFlag) { OR32RtoR(mReg, gprT2); } + if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ if (sFLAG.doFlag && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - - if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag + + if (sFLAG.doSticky) OR32RtoR(gprST, mReg); // Set Sticky Register (gprST) + if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag + if (sFLAG.doFlag) { // Attach Sticky Register With sReg + TEST32ItoR(gprST, 0x0f); + pjmp = JZ8(0); // Set Z bit? + OR32ItoR(sReg, 0x40); + x86SetJ8(pjmp); + TEST32ItoR(gprST, 0xf0); + pjmp = JZ8(0); // Set S bit? + OR32ItoR(sReg, 0x80); + x86SetJ8(pjmp); + MOV32RtoR(mReg, gprST); // Backup gprST + AND32ItoR(mReg, 0xc30000); // Get D/I Bits + SHR32ItoR(mReg, 12); // Shift D/I Bits to proper position + OR32RtoR (sReg, mReg); // Set D/I Bits + mVUallocSFLAGb(mVU, sReg, sFLAG.write); // Set Status Flag + } } //------------------------------------------------------------------ @@ -438,9 +450,9 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool } // FMAC27~29 - MAX/MINI FMAC Opcodes -#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } -#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } -#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } +#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { mVUup.doFlags = 0; } } +#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { mVUup.doFlags = 0; } } +#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { mVUup.doFlags = 0; } } //------------------------------------------------------------------ // Micro VU Micromode Upper instructions From 28f52a3f12aaef033eab0117182b11f07a52fc2f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 1 Jun 2009 04:58:06 +0000 Subject: [PATCH 164/296] microVU: - Backing up some changes before I revert... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1301 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 18 ++++++------- pcsx2/x86/microVU_Execute.inl | 5 +--- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 20 ++++++-------- pcsx2/x86/microVU_Misc.h | 10 ------- pcsx2/x86/microVU_Upper.inl | 49 ++++++++++++----------------------- 6 files changed, 35 insertions(+), 69 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index f4674fc61c..a388adb164 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -59,7 +59,6 @@ #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } -#define doUpperOp() { mVUdivSet(mVU); mVUopU(mVU, 1); } #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } #define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } @@ -275,10 +274,11 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbranch = 0; int x; for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { + mVUdivSet(mVU); if (mVUinfo.isEOB) { x = 0xffff; } - if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } - else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } + if (mVUlow.isNOP) { incPC(1); mVUopU(mVU, 1); doIbit(); } + else if (!mVUinfo.swapOps) { incPC(1); mVUopU(mVU, 1); doLowerOp(); } + else { mVUopL(mVU, 1); incPC(1); mVUopU(mVU, 1); } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } if (!mVUinfo.isBdelay) { incPC(1); } @@ -373,12 +373,10 @@ eBitTemination: memset(&mVUinfo, 0, sizeof(mVUinfo)); incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; - /*if (mVUinfo.doDivFlag) { - int flagReg; - getFlagReg(flagReg, lStatus); - AND32ItoR (flagReg, 0x0fcf); - OR32MtoR (flagReg, (uptr)&mVU->divFlag); - }*/ + if (mVUinfo.doDivFlag) { + AND32ItoR(gprST, 0xfcf); + OR32MtoR (gprST, (uptr)&mVU->divFlag); + } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } // Do E-bit end stuff here diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 8c5bb00c6b..f0b783b8e6 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -42,10 +42,7 @@ microVUt(void) mVUdispatcherA(mV) { // Load Regs MOV32ItoR(gprR, Roffset); // Load VI Reg Offset MOV32MtoR(gprST, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); - /*AND32ItoR(gprST, 0xffff); - MOV32RtoR(gprF1, gprST); - MOV32RtoR(gprF2, gprST); - MOV32RtoR(gprF3, gprST);*/ + AND32ItoR(gprST, 0xffff); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 70f50f05fa..2a9c84c67a 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -21,7 +21,7 @@ // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet(mV) { if (mVUinfo.doDivFlag) { - AND32ItoR(gprST, 0xfffcffff); // Clear D/I bits + AND32ItoR(gprST, 0xfcf); // Clear D/I bits OR32MtoR (gprST, (uptr)&mVU->divFlag); // Set DS/IS/D/I bits } } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 1bfd48e6e0..95671df2f7 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -37,7 +37,7 @@ SSE_MOVMSKPS_XMM_to_R32(gprTemp, xmmReg); \ TEST32ItoR(gprTemp, 1); /* Check sign bit */ \ aJump = JZ8(0); /* Skip if positive */ \ - MOV32ItoM((uptr)&mVU->divFlag, 0x410000); /* Set Invalid Flags */ \ + MOV32ItoM((uptr)&mVU->divFlag, 0x410); /* Set Invalid Flags */ \ SSE_ANDPS_M128_to_XMM(xmmReg, (uptr)mVU_absclip); /* Abs(xmmReg) */ \ x86SetJ8(aJump); \ } @@ -54,10 +54,10 @@ mVUop(mVU_DIV) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero ajmp = JZ8(0); - MOV32ItoM((uptr)&mVU->divFlag, 0x410000); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) bjmp = JMP8(0); x86SetJ8(ajmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820000); // Zero divide (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide (only when not 0/0) x86SetJ8(bjmp); SSE_XORPS_XMM_to_XMM(xmmFs, xmmFt); @@ -112,10 +112,10 @@ mVUop(mVU_RSQRT) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero bjmp = JZ8(0); // Skip if none are - MOV32ItoM((uptr)&mVU->divFlag, 0x410000); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) cjmp = JMP8(0); x86SetJ8(bjmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820000); // Zero divide flag (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide flag (only when not 0/0) x86SetJ8(cjmp); SSE_ANDPS_M128_to_XMM(xmmFs, (uptr)mVU_signbit); @@ -557,13 +557,9 @@ mVUop(mVU_FSOR) { mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { - int mask; - if (_Imm12_ & 0x800) mask |= 0x800000; - if (_Imm12_ & 0x400) mask |= 0x400000; - if (_Imm12_ & 0x080) mask |= 0x0000f0; - if (_Imm12_ & 0xc40) mask |= 0x00000f; - AND32ItoR(gprST, 0x30000); - if (mask) OR32ItoR(gprST, mask); + int imm = _Imm12_ & 0xfc0; + AND32ItoR(gprST, 0x3ff); + if (imm) OR32ItoR(gprST, imm); } pass3 { mVUlog("FSSET $%x", _Imm12_); } pass4 { mVUsFlagHack = 0; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 8d9ad3a747..d2cdc87956 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -137,16 +137,6 @@ declareAllVariables #define gprT5 6 // Temp? #define gprR 7 // VI Reg Offset #define gprST 3 // Status Sticky Flag -// gprST's Info is Stored as follows: -//----------------------------------------------------------------------------- -//|23 22 21 20||19 18 17 16||15 14 13 12||11 10 09 08||07 06 05 04||03 02 01 00| -//|DS|IS| || | D| I|| OS || US || SS || ZS | -//----------------------------------------------------------------------------- -// Storing Flags this way eliminates Jumps when updating sticky flags. -// -// When a Status Flag will be read, gprST is attached with -// the current status flag result in mVUupdateFlags. And the complete -// Status flag instance is stored in memory (mVU->statusFlag[instance]) // Function Params #define mP microVU* mVU, int recPass diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 578df21d49..b7fdc12b17 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -28,7 +28,7 @@ // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool modXYZW) { - int sReg = gprT3, mReg = gprT1; + int /*sReg = gprT3,*/ mReg = gprT1; static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; @@ -37,54 +37,39 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool if (!mVUup.doFlags || (!sFLAG.doSticky && !sFLAG.doFlag && !mFLAG.doFlag)) { return; } if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw - if (sFLAG.doFlag) { XOR32RtoR(sReg, sReg); } //-------------------------Check for Signed flags------------------------------ // The following code makes sure the Signed Bit isn't set with Negative Zero - SSE_XORPS_XMM_to_XMM(regT2, regT2); // Clear regT2 - SSE_CMPEQPS_XMM_to_XMM(regT2, regT1); // Set all F's if each vector is zero + SSE_XORPS_XMM_to_XMM(regT2, regT2); // Clear regT2 + SSE_CMPEQPS_XMM_to_XMM(regT2, regT1); // Set all F's if each vector is zero SSE_MOVMSKPS_XMM_to_R32(gprT2, regT2); // Used for Zero Flag Calculation SSE_ANDNPS_XMM_to_XMM(regT2, regT1); SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation - if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag || sFLAG.doSticky) SHL32ItoR(mReg, 4 + ADD_XYZW); - if (sFLAG.doFlag) OR32ItoR (sReg, 0x82); // SS, S flags - if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - if (sFLAG.doFlag) x86SetJ8(pjmp); + if (sFLAG.doSticky||sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); + if (sFLAG.doSticky) OR32ItoR (gprST, 0x82); // SS, S flags + else if (sFLAG.doFlag) OR32ItoR (gprST, 0x02); // S flag + if ((sFLAG.doSticky||sFLAG.doFlag) && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + if (sFLAG.doSticky||sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation - if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (sFLAG.doSticky && !mFLAG.doFlag) { OR32RtoR(mReg, gprT2); } - if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags - if (sFLAG.doFlag) x86SetJ8(pjmp); + if (sFLAG.doSticky||sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } + if (sFLAG.doSticky) { OR32ItoR(gprST, 0x41); } // ZS, Z flags + else if (sFLAG.doFlag) { OR32ItoR(gprST, 0x01); } // Z flag + if (sFLAG.doSticky||sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ - if (sFLAG.doFlag && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here + if ((sFLAG.doSticky||sFLAG.doFlag) && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - if (sFLAG.doSticky) OR32RtoR(gprST, mReg); // Set Sticky Register (gprST) - if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag - if (sFLAG.doFlag) { // Attach Sticky Register With sReg - TEST32ItoR(gprST, 0x0f); - pjmp = JZ8(0); // Set Z bit? - OR32ItoR(sReg, 0x40); - x86SetJ8(pjmp); - TEST32ItoR(gprST, 0xf0); - pjmp = JZ8(0); // Set S bit? - OR32ItoR(sReg, 0x80); - x86SetJ8(pjmp); - MOV32RtoR(mReg, gprST); // Backup gprST - AND32ItoR(mReg, 0xc30000); // Get D/I Bits - SHR32ItoR(mReg, 12); // Shift D/I Bits to proper position - OR32RtoR (sReg, mReg); // Set D/I Bits - mVUallocSFLAGb(mVU, sReg, sFLAG.write); // Set Status Flag - } + if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag + if (sFLAG.doFlag) mVUallocSFLAGb(mVU, gprST, mFLAG.write); // Set Status Flag } //------------------------------------------------------------------ From a336e6b277db32d5559867cbeeaff6328e015425 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 1 Jun 2009 05:09:28 +0000 Subject: [PATCH 165/296] microVU: - reverted to r1299 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1302 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 7 +++-- pcsx2/x86/microVU_Alloc.inl | 23 +++++++++++++--- pcsx2/x86/microVU_Analyze.inl | 32 ++++++++++------------ pcsx2/x86/microVU_Compile.inl | 20 ++++++++------ pcsx2/x86/microVU_Execute.inl | 7 +++-- pcsx2/x86/microVU_Flags.inl | 37 +++++++++++++------------ pcsx2/x86/microVU_IR.h | 4 +-- pcsx2/x86/microVU_Lower.inl | 33 ++++++++++++----------- pcsx2/x86/microVU_Misc.h | 27 ++++++++++--------- pcsx2/x86/microVU_Misc.inl | 1 + pcsx2/x86/microVU_Upper.inl | 51 ++++++++++++++++++----------------- 11 files changed, 135 insertions(+), 107 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7631cc9347..7fd6efc281 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -98,10 +98,9 @@ struct microProgManager { #define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) struct microVU { - PCSX2_ALIGNED16(u32 statusFlag[4]); // 4 instances of status flag (used in execution) - PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) - PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) - PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ + PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) + PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) + PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index a04bc48008..eb0c5deda0 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -634,13 +634,28 @@ microVUt(void) mVUallocFMAC26b(mV, int& ACCw, int& ACCr) { // Flag Allocators //------------------------------------------------------------------ -microVUt(void) mVUallocSFLAGa(mV, int reg, int fInstance) { - MOVZX32M16toR(reg, (uptr)&mVU->statusFlag[fInstance]); +#define getFlagReg(regX, fInst) { \ + switch (fInst) { \ + case 0: regX = gprF0; break; \ + case 1: regX = gprF1; break; \ + case 2: regX = gprF2; break; \ + case 3: regX = gprF3; break; \ + default: \ + Console::Error("microVU: Flag Instance Error (fInst = %d)", params fInst); \ + regX = gprF0; \ + break; \ + } \ } -microVUt(void) mVUallocSFLAGb(mV, int reg, int fInstance) { +microVUt(void) mVUallocSFLAGa(int reg, int fInstance) { + getFlagReg(fInstance, fInstance); + MOVZX32R16toR(reg, fInstance); +} + +microVUt(void) mVUallocSFLAGb(int reg, int fInstance) { + getFlagReg(fInstance, fInstance); //AND32ItoR(reg, 0xffff); - MOV32RtoM((uptr)&mVU->statusFlag[fInstance], reg); + MOV32RtoR(fInstance, reg); } microVUt(void) mVUallocMFLAGa(mV, int reg, int fInstance) { diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 8af22c363f..74c1cbcfc9 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -62,8 +62,7 @@ } microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { - mVUup.doFlags = 1; - sFLAG.doSticky = 1; + sFLAG.doFlag = 1; analyzeReg1(Fs); analyzeReg1(Ft); analyzeReg2(Fd, 0); @@ -92,8 +91,7 @@ microVUt(void) mVUanalyzeFMAC2(mV, int Fs, int Ft) { } microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { - mVUup.doFlags = 1; - sFLAG.doSticky = 1; + sFLAG.doFlag = 1; analyzeReg1(Fs); analyzeReg3(Ft); analyzeReg2(Fd, 0); @@ -271,15 +269,6 @@ microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { // Sflag - Status Flag Opcodes //------------------------------------------------------------------ -#define setFlagInst(xDoFlag) { \ - int curPC = iPC; \ - for (int i = mVUcount, j = 0; i > 0; i--, j++) { \ - incPC2(-2); \ - if (mVUup.doFlags) { xDoFlag = 1; if (j >= 3) { break; } } \ - } \ - iPC = curPC; \ -} - microVUt(void) mVUanalyzeSflag(mV, int It) { if (!It) { mVUlow.isNOP = 1; } else { @@ -290,15 +279,17 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { // Note: useSflag is used for status flag optimizations when a FSSET instruction is called. // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. - setFlagInst(sFLAG.doFlag); } analyzeVIreg3(It, 1); } microVUt(void) mVUanalyzeFSSET(mV) { - mVUinfo.swapOps = 1; - mVUlow.isFSSET = 1; - sFLAG.doSticky = 0; + mVUlow.isFSSET = 1; + // mVUinfo &= ~_doStatus; + // Note: I'm not entirely sure if the non-sticky flags + // should be taken from the current upper instruction + // or if they should be taken from the previous instruction + // Uncomment the above line if the latter-case is true } //------------------------------------------------------------------ @@ -310,7 +301,12 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } - setFlagInst(mFLAG.doFlag); + int curPC = iPC; + for (int i = mVUcount, j = 0; i > 0; i--, j++) { + incPC2(-2); + if (sFLAG.doFlag) { mFLAG.doFlag = 1; if (j >= 3) { break; } } + } + iPC = curPC; } analyzeVIreg1(Is); analyzeVIreg3(It, 1); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index a388adb164..689d8b82a4 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -59,6 +59,7 @@ #define tCycles(dest, src) { dest = aMax(dest, src); } #define incP() { mVU->p = (mVU->p+1) & 1; } #define incQ() { mVU->q = (mVU->q+1) & 1; } +#define doUpperOp() { mVUopU(mVU, 1); mVUdivSet(mVU); } #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } #define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } @@ -168,7 +169,8 @@ microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, in // Save Flag Instances if (!mVUflagHack) { - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprST); + getFlagReg(fStatus, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); } mVUallocMFLAGa(mVU, gprT1, fMac); mVUallocCFLAGa(mVU, gprT2, fClip); @@ -198,6 +200,7 @@ microVUt(void) mVUtestCycles(mV) { MOV32ItoR(gprT2, xPC); if (!isVU1) CALLFunc((uptr)mVUwarning0); else CALLFunc((uptr)mVUwarning1); + MOV32ItoR(gprR, Roffset); // Restore gprR mVUendProgram(mVU, 0, 0, sI, 0, cI); x86SetJ8(jmp8); } @@ -274,11 +277,10 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbranch = 0; int x; for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { - mVUdivSet(mVU); if (mVUinfo.isEOB) { x = 0xffff; } - if (mVUlow.isNOP) { incPC(1); mVUopU(mVU, 1); doIbit(); } - else if (!mVUinfo.swapOps) { incPC(1); mVUopU(mVU, 1); doLowerOp(); } - else { mVUopL(mVU, 1); incPC(1); mVUopU(mVU, 1); } + if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } + else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } + else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } if (!mVUinfo.isBdelay) { incPC(1); } @@ -320,7 +322,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUbackupRegs(mVU); MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) - MOV32ItoR(gprT3, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) + MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) if (!isVU1) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) else CALLFunc((uptr)mVUcompileVU1); @@ -374,8 +376,10 @@ eBitTemination: incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; if (mVUinfo.doDivFlag) { - AND32ItoR(gprST, 0xfcf); - OR32MtoR (gprST, (uptr)&mVU->divFlag); + int flagReg; + getFlagReg(flagReg, lStatus); + AND32ItoR (flagReg, 0x0fcf); + OR32MtoR (flagReg, (uptr)&mVU->divFlag); } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index f0b783b8e6..0854b0bad3 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -41,8 +41,11 @@ microVUt(void) mVUdispatcherA(mV) { // Load Regs MOV32ItoR(gprR, Roffset); // Load VI Reg Offset - MOV32MtoR(gprST, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); - AND32ItoR(gprST, 0xffff); + MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); + AND32ItoR(gprF0, 0xffff); + MOV32RtoR(gprF1, gprF0); + MOV32RtoR(gprF2, gprF0); + MOV32RtoR(gprF3, gprF0); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 2a9c84c67a..fc63c3db9d 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -20,9 +20,12 @@ // Sets FDIV Flags at the proper time microVUt(void) mVUdivSet(mV) { + int flagReg1, flagReg2; if (mVUinfo.doDivFlag) { - AND32ItoR(gprST, 0xfcf); // Clear D/I bits - OR32MtoR (gprST, (uptr)&mVU->divFlag); // Set DS/IS/D/I bits + getFlagReg(flagReg1, sFLAG.write); + if (!sFLAG.doFlag) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } + AND32ItoR(flagReg1, 0x0fcf); + OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } } @@ -31,19 +34,18 @@ microVUt(void) mVUstatusFlagOp(mV) { int curPC = iPC; int i = mVUcount; bool runLoop = 1; - if (mVUup.doFlags) { mVUlow.useSflag = 1; } + if (sFLAG.doFlag) { mVUlow.useSflag = 1; } else { for (; i > 0; i--) { incPC2(-2); if (mVUlow.useSflag) { runLoop = 0; break; } - if (mVUup.doFlags) { mVUlow.useSflag = 1; break; } + if (sFLAG.doFlag) { mVUlow.useSflag = 1; break; } } } if (runLoop) { for (; i > 0; i--) { incPC2(-2); if (mVUlow.useSflag) break; - sFLAG.doSticky = 0; sFLAG.doFlag = 0; } } @@ -77,11 +79,7 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { // Ensure last ~4+ instructions update mac flags (if next block's first 4 instructions will read them) for (int i = mVUcount; i > 0; i--, aCount++) { - if (mVUup.doFlags) { - if (__Status) { sFLAG.doFlag = 1; } - if (__Mac) { mFLAG.doFlag = 1; } - if (aCount >= 4) { break; } - } + if (sFLAG.doFlag) { if (__Mac) { mFLAG.doFlag = 1; } if (aCount >= 4) { break; } } incPC2(-2); } @@ -144,13 +142,13 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { incPC2(2); } - mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) /*| ((__Status) ? 0 : xS)*/; + mVUregs.flags = ((__Clip) ? 0 : (xC << 2)) | ((__Status) ? 0 : xS); return cycles; } -#define shuffleStatus ((bStatus[3]<<6)|(bStatus[2]<<4)|(bStatus[1]<<2)|bStatus[0]) -#define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) -#define shuffleClip ((bClip [3]<<6)|(bClip [2]<<4)|(bClip [1]<<2)|bClip [0]) +#define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) +#define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) +#define shuffleClip ((bClip[3]<<6)|(bClip[2]<<4)|(bClip[1]<<2)|bClip[0]) // Recompiles Code for Proper Flags on Block Linkings microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles) { @@ -158,9 +156,14 @@ microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles if (__Status && !mVUflagHack) { int bStatus[4]; sortFlag(xStatus, bStatus, cycles); - SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->statusFlag); - SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, shuffleStatus); - SSE_MOVAPS_XMM_to_M128((uptr)mVU->statusFlag, xmmT1); + MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); + MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); + MOV32RtoR(gprR, getFlagReg1(bStatus[2])); + MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, gprT1); + MOV32RtoR(gprF1, gprT2); + MOV32RtoR(gprF2, gprR); + MOV32ItoR(gprR, Roffset); // Restore gprR } if (__Mac) { diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 2ccd33b907..a42a683a25 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -68,7 +68,6 @@ struct microBlock { struct microUpperOp { bool eBit; // Has E-bit set bool iBit; // Has I-bit set - bool doFlags; // This instruction updates Status/Mac Flags }; struct microLowerOp { @@ -84,8 +83,7 @@ struct microLowerOp { }; struct microFlagInst { - bool doSticky; // Update Sticky Flags (Status Flag Only) - bool doFlag; // Update Flag on this Instruction (For Status Flag, this means non-sticky bits) + bool doFlag; // Update Flag on this Instruction u8 write; // Points to the instance that should be written to (s-stage write) u8 lastWrite; // Points to the instance that was last written to (most up-to-date flag) u8 read; // Points to the instance that should be read by a lower instruction (t-stage read) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 95671df2f7..c7a4b7e0ca 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -522,7 +522,7 @@ mVUop(mVU_FMOR) { mVUop(mVU_FSAND) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(mVU, gprT1, sFLAG.read); + mVUallocSFLAGa(gprT1, sFLAG.read); AND16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -533,7 +533,7 @@ mVUop(mVU_FSAND) { mVUop(mVU_FSEQ) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(mVU, gprT1, sFLAG.read); + mVUallocSFLAGa(gprT1, sFLAG.read); XOR16ItoR(gprT1, _Imm12_); SUB16ItoR(gprT1, 1); SHR16ItoR(gprT1, 15); @@ -546,7 +546,7 @@ mVUop(mVU_FSEQ) { mVUop(mVU_FSOR) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(mVU, gprT1, sFLAG.read); + mVUallocSFLAGa(gprT1, sFLAG.read); OR16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -557,9 +557,11 @@ mVUop(mVU_FSOR) { mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { - int imm = _Imm12_ & 0xfc0; - AND32ItoR(gprST, 0x3ff); - if (imm) OR32ItoR(gprST, imm); + int flagReg1, flagReg2; + getFlagReg(flagReg1, sFLAG.write); + if (!(sFLAG.doFlag||mVUinfo.doDivFlag)) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction + AND32ItoR(flagReg1, 0x03f); + OR32ItoR (flagReg1, (_Imm12_ & 0xfc0)); } pass3 { mVUlog("FSSET $%x", _Imm12_); } pass4 { mVUsFlagHack = 0; } @@ -964,22 +966,23 @@ mVUop(mVU_RNEXT) { pass1 { mVUanalyzeR2(mVU, _Ft_, 0); } pass2 { // algorithm from www.project-fao.org - MOV32MtoR(gprT3, Rmem); - MOV32RtoR(gprT1, gprT3); + MOV32MtoR(gprR, Rmem); + MOV32RtoR(gprT1, gprR); SHR32ItoR(gprT1, 4); AND32ItoR(gprT1, 1); - MOV32RtoR(gprT2, gprT3); + MOV32RtoR(gprT2, gprR); SHR32ItoR(gprT2, 22); AND32ItoR(gprT2, 1); - SHL32ItoR(gprT3, 1); + SHL32ItoR(gprR, 1); XOR32RtoR(gprT1, gprT2); - XOR32RtoR(gprT3, gprT1); - AND32ItoR(gprT3, 0x007fffff); - OR32ItoR (gprT3, 0x3f800000); - MOV32RtoM(Rmem, gprT3); - mVU_RGET_(mVU, gprT3); + XOR32RtoR(gprR, gprT1); + AND32ItoR(gprR, 0x007fffff); + OR32ItoR (gprR, 0x3f800000); + MOV32RtoM(Rmem, gprR); + mVU_RGET_(mVU, gprR); + MOV32ItoR(gprR, Roffset); // Restore gprR } pass3 { mVUlog("RNEXT.%s vf%02d, R", _XYZW_String, _Ft_); } } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index d2cdc87956..d36cfc7a8e 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -131,12 +131,12 @@ declareAllVariables #define gprT1 0 // Temp Reg #define gprT2 1 // Temp Reg -#define gprT3 2 // Temp Reg +#define gprR 2 // VI Reg Offset +#define gprF0 3 // Status Flag 0 #define gprESP 4 // Don't use? -#define gprT4 5 // Temp? -#define gprT5 6 // Temp? -#define gprR 7 // VI Reg Offset -#define gprST 3 // Status Sticky Flag +#define gprF1 5 // Status Flag 1 +#define gprF2 6 // Status Flag 2 +#define gprF3 7 // Status Flag 3 // Function Params #define mP microVU* mVU, int recPass @@ -172,13 +172,15 @@ declareAllVariables #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC #define mVUsFlagHack mVUallocInfo.sFlagHack -#define mVUinfo mVUallocInfo.info[iPC/2] // IR info for current 64bit instruction -#define mVUstall mVUinfo.stall // Stall info for current instruction -#define mVUup mVUinfo.uOp // Upper Instruction Info -#define mVUlow mVUinfo.lOp // Lower Instruction Info -#define sFLAG mVUinfo.sFlag // Status Flag info for cur instruction -#define mFLAG mVUinfo.mFlag // Mac Flag info for cur instruction -#define cFLAG mVUinfo.cFlag // Clip Flag info for cur instruction + +#define mVUinfo mVUallocInfo.info[iPC / 2] +#define mVUstall mVUinfo.stall +#define mVUup mVUinfo.uOp +#define mVUlow mVUinfo.lOp +#define sFLAG mVUinfo.sFlag +#define mFLAG mVUinfo.mFlag +#define cFLAG mVUinfo.cFlag + #define mVUstartPC mVUallocInfo.startPC #define mVUflagInfo mVUregs.needExactMatch #define mVUflagHack (mVUcurProg.sFlagHack) @@ -266,5 +268,6 @@ declareAllVariables MOV32ItoR(gprT2, xPC); \ if (isEndPC) { CALLFunc((uptr)mVUprintPC2); } \ else { CALLFunc((uptr)mVUprintPC1); } \ + MOV32ItoR(gprR, Roffset); \ } \ } diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index f7b0c1f3bc..f6d2c58b00 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -292,6 +292,7 @@ microVUt(void) mVUrestoreRegs(mV) { SSE_MOVAPS_M128_to_XMM(xmmPQ, (uptr)&mVU->xmmPQb[0]); SSE_MOVAPS_M128_to_XMM(xmmMax, (uptr)mVU_maxvals); SSE_MOVAPS_M128_to_XMM(xmmMin, (uptr)mVU_minvals); + MOV32ItoR(gprR, Roffset); // Restore gprR } // Reads entire microProgram and finds out if Status Flag is Used diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index b7fdc12b17..c2bb91e294 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -28,48 +28,51 @@ // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool modXYZW) { - int /*sReg = gprT3,*/ mReg = gprT1; + int sReg, mReg = gprT1; static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; //SysPrintf("Status = %d; Mac = %d\n", sFLAG.doFlag, mFLAG.doFlag); - if (mVUsFlagHack) { sFLAG.doSticky = 0; sFLAG.doFlag = 0; } - if (!mVUup.doFlags || (!sFLAG.doSticky && !sFLAG.doFlag && !mFLAG.doFlag)) { return; } - if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } + if (mVUsFlagHack) { sFLAG.doFlag = 0; } + if (!sFLAG.doFlag && !mFLAG.doFlag) { return; } + if (!mFLAG.doFlag || (_XYZW_SS && modXYZW)) { regT1 = reg; } else { SSE2_PSHUFD_XMM_to_XMM(regT1, reg, 0x1B); } // Flip wzyx to xyzw + if (sFLAG.doFlag) { + getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance + mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag + AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags + } + //-------------------------Check for Signed flags------------------------------ // The following code makes sure the Signed Bit isn't set with Negative Zero - SSE_XORPS_XMM_to_XMM(regT2, regT2); // Clear regT2 - SSE_CMPEQPS_XMM_to_XMM(regT2, regT1); // Set all F's if each vector is zero + SSE_XORPS_XMM_to_XMM(regT2, regT2); // Clear regT2 + SSE_CMPEQPS_XMM_to_XMM(regT2, regT1); // Set all F's if each vector is zero SSE_MOVMSKPS_XMM_to_R32(gprT2, regT2); // Used for Zero Flag Calculation SSE_ANDNPS_XMM_to_XMM(regT2, regT1); SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation - if (sFLAG.doSticky||sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); - if (sFLAG.doSticky) OR32ItoR (gprST, 0x82); // SS, S flags - else if (sFLAG.doFlag) OR32ItoR (gprST, 0x02); // S flag - if ((sFLAG.doSticky||sFLAG.doFlag) && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - if (sFLAG.doSticky||sFLAG.doFlag) x86SetJ8(pjmp); + if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); + if (sFLAG.doFlag) OR32ItoR(sReg, 0x82); // SS, S flags + if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation - if (sFLAG.doSticky||sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (sFLAG.doSticky) { OR32ItoR(gprST, 0x41); } // ZS, Z flags - else if (sFLAG.doFlag) { OR32ItoR(gprST, 0x01); } // Z flag - if (sFLAG.doSticky||sFLAG.doFlag) x86SetJ8(pjmp); + if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are + if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } + if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags + if (sFLAG.doFlag) x86SetJ8(pjmp); //-------------------------Write back flags------------------------------ - if ((sFLAG.doSticky||sFLAG.doFlag) && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - - if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag - if (sFLAG.doFlag) mVUallocSFLAGb(mVU, gprST, mFLAG.write); // Set Status Flag + if (sFLAG.doFlag && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here + + if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag } //------------------------------------------------------------------ @@ -435,9 +438,9 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool } // FMAC27~29 - MAX/MINI FMAC Opcodes -#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { mVUup.doFlags = 0; } } -#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { mVUup.doFlags = 0; } } -#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { mVUup.doFlags = 0; } } +#define mVU_FMAC27(operation, OPname) { mVU_FMAC1 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } +#define mVU_FMAC28(operation, OPname) { mVU_FMAC6 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } +#define mVU_FMAC29(operation, OPname) { mVU_FMAC3 (operation, OPname); pass1 { sFLAG.doFlag = 0; } } //------------------------------------------------------------------ // Micro VU Micromode Upper instructions From a8266a48d875ce7b79219da929ec9275ed8a526e Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 1 Jun 2009 22:29:27 +0000 Subject: [PATCH 166/296] microVU: - Non-functional change... just storing some IR info I'll need to use later. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1303 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 280 +++++++++++++++++++--------------- pcsx2/x86/microVU_IR.h | 23 ++- pcsx2/x86/microVU_Lower.inl | 22 +-- 3 files changed, 190 insertions(+), 135 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 74c1cbcfc9..9a07f3c2e5 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -23,35 +23,28 @@ //------------------------------------------------------------------ //------------------------------------------------------------------ -// FMAC1 - Normal FMAC Opcodes +// Helper Macros //------------------------------------------------------------------ -#define aReg(x) mVUregs.VF[x] +#define aReg(x) mVUregs.VF[x] #define bReg(x, y) mVUregsTemp.VFreg[y] = x; mVUregsTemp.VF[y] #define aMax(x, y) ((x > y) ? x : y) -#define analyzeReg1(reg) { \ - if (reg) { \ - if (_X) { mVUstall = aMax(mVUstall, aReg(reg).x); } \ - if (_Y) { mVUstall = aMax(mVUstall, aReg(reg).y); } \ - if (_Z) { mVUstall = aMax(mVUstall, aReg(reg).z); } \ - if (_W) { mVUstall = aMax(mVUstall, aReg(reg).w); } \ - } \ +// Read a VF reg by upper op +#define analyzeReg1(xReg, vfRead) { \ + if (xReg) { \ + if (_X) { mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; } \ + if (_Y) { mVUstall = aMax(mVUstall, aReg(xReg).y); vfRead.reg = xReg; vfRead.y = 1; } \ + if (_Z) { mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; } \ + if (_W) { mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; } \ + } \ } -#define analyzeReg2(reg, isLowOp) { \ - if (reg) { \ - if (_X) { bReg(reg, isLowOp).x = 4; } \ - if (_Y) { bReg(reg, isLowOp).y = 4; } \ - if (_Z) { bReg(reg, isLowOp).z = 4; } \ - if (_W) { bReg(reg, isLowOp).w = 4; } \ - } \ -} - -#define analyzeReg1b(reg) { \ - if (reg) { \ - analyzeReg1(reg); \ - if (mVUregsTemp.VFreg[0] == reg) { \ +// Read a VF reg by lower op +#define analyzeReg1b(xReg, vfRead) { \ + if (xReg) { \ + analyzeReg1(xReg, vfRead); \ + if (mVUregsTemp.VFreg[0] == xReg) { \ if ((mVUregsTemp.VF[0].x && _X) \ || (mVUregsTemp.VF[0].y && _Y) \ || (mVUregsTemp.VF[0].z && _Z) \ @@ -61,11 +54,115 @@ } \ } +// Write to a VF reg +#define analyzeReg2(xReg, vfWrite, isLowOp) { \ + if (xReg) { \ + if (_X) { bReg(xReg, isLowOp).x = 4; vfWrite.reg = xReg; vfWrite.x = 4; } \ + if (_Y) { bReg(xReg, isLowOp).y = 4; vfWrite.reg = xReg; vfWrite.y = 4; } \ + if (_Z) { bReg(xReg, isLowOp).z = 4; vfWrite.reg = xReg; vfWrite.z = 4; } \ + if (_W) { bReg(xReg, isLowOp).w = 4; vfWrite.reg = xReg; vfWrite.w = 4; } \ + } \ +} + +// Read a VF reg (BC opcodes) +#define analyzeReg3(xReg, vfRead) { \ + if (xReg) { \ + if (_bc_x) { mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; } \ + else if (_bc_y) { mVUstall = aMax(mVUstall, aReg(xReg).y); vfRead.reg = xReg; vfRead.y = 1; } \ + else if (_bc_z) { mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; } \ + else { mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; } \ + } \ +} + +// For Clip Opcode +#define analyzeReg4(xReg, vfRead) { \ + if (xReg) { \ + mVUstall = aMax(mVUstall, aReg(xReg).w); \ + vfRead.reg = xReg; vfRead.w = 1; \ + } \ +} + +// Read VF reg (FsF/FtF) +#define analyzeReg5(xReg, fxf, vfRead) { \ + if (xReg) { \ + switch (fxf) { \ + case 0: mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; break; \ + case 1: mVUstall = aMax(mVUstall, aReg(xReg).y); vfRead.reg = xReg; vfRead.y = 1; break; \ + case 2: mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; break; \ + case 3: mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; break; \ + } \ + if (mVUregsTemp.VFreg[0] == xReg) { \ + if ((mVUregsTemp.VF[0].x && (fxf == 0)) \ + || (mVUregsTemp.VF[0].y && (fxf == 1)) \ + || (mVUregsTemp.VF[0].z && (fxf == 2)) \ + || (mVUregsTemp.VF[0].w && (fxf == 3))) \ + { mVUinfo.swapOps = 1; } \ + } \ + } \ +} + +// Flips xyzw stalls to yzwx (MR32 Opcode) +#define analyzeReg6(xReg, vfRead) { \ + if (xReg) { \ + if (_X) { mVUstall = aMax(mVUstall, aReg(xReg).y); vfRead.reg = xReg; vfRead.y = 1; } \ + if (_Y) { mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; } \ + if (_Z) { mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; } \ + if (_W) { mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; } \ + if (mVUregsTemp.VFreg[0] == xReg) { \ + if ((mVUregsTemp.VF[0].y && _X) \ + || (mVUregsTemp.VF[0].z && _Y) \ + || (mVUregsTemp.VF[0].w && _Z) \ + || (mVUregsTemp.VF[0].x && _W)) \ + { mVUinfo.swapOps = 1; } \ + } \ + } \ +} + +// Reading a VI reg +#define analyzeVIreg1(xReg, viRead) { \ + if (xReg) { \ + mVUstall = aMax(mVUstall, mVUregs.VI[xReg]); \ + viRead.reg = xReg; viRead.used = 1; \ + } \ +} + +// Writing to a VI reg +#define analyzeVIreg2(xReg, viWrite, aCycles) { \ + if (xReg) { \ + mVUregsTemp.VIreg = xReg; \ + mVUregsTemp.VI = aCycles; \ + mVUlow.writesVI = 1; \ + mVU->VIbackup[0] = xReg; \ + viWrite.reg = xReg; \ + viWrite.used = aCycles; \ + } \ +} + +// Writing to a VI reg (FSxxx, FMxxx, FCxxx opcodes) +#define analyzeVIreg3(xReg, viWrite, aCycles) { \ + if (xReg) { \ + mVUregsTemp.VIreg = xReg; \ + mVUregsTemp.VI = aCycles; \ + viWrite.reg = xReg; \ + viWrite.used = aCycles; \ + } \ +} + +#define analyzeQreg(x) { mVUregsTemp.q = x; mVUstall = aMax(mVUstall, mVUregs.q); } +#define analyzePreg(x) { mVUregsTemp.p = x; mVUstall = aMax(mVUstall, ((mVUregs.p) ? (mVUregs.p - 1) : 0)); } +#define analyzeRreg() { mVUregsTemp.r = 1; } +#define analyzeXGkick1() { mVUstall = aMax(mVUstall, mVUregs.xgkick); } +#define analyzeXGkick2(x) { mVUregsTemp.xgkick = x; } + +//------------------------------------------------------------------ +// FMAC1 - Normal FMAC Opcodes +//------------------------------------------------------------------ + microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { sFLAG.doFlag = 1; - analyzeReg1(Fs); - analyzeReg1(Ft); - analyzeReg2(Fd, 0); + analyzeReg1(Fs, mVUup.VF_read[0]); + analyzeReg1(Ft, mVUup.VF_read[1]); + analyzeReg2(Fd, mVUup.VF_write, 0); } //------------------------------------------------------------------ @@ -73,121 +170,66 @@ microVUt(void) mVUanalyzeFMAC1(mV, int Fd, int Fs, int Ft) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeFMAC2(mV, int Fs, int Ft) { - analyzeReg1(Fs); - analyzeReg2(Ft, 0); + analyzeReg1(Fs, mVUup.VF_read[0]); + analyzeReg2(Ft, mVUup.VF_write, 0); } //------------------------------------------------------------------ // FMAC3 - BC(xyzw) FMAC Opcodes //------------------------------------------------------------------ -#define analyzeReg3(reg) { \ - if (reg) { \ - if (_bc_x) { mVUstall = aMax(mVUstall, aReg(reg).x); } \ - else if (_bc_y) { mVUstall = aMax(mVUstall, aReg(reg).y); } \ - else if (_bc_z) { mVUstall = aMax(mVUstall, aReg(reg).z); } \ - else { mVUstall = aMax(mVUstall, aReg(reg).w); } \ - } \ -} - microVUt(void) mVUanalyzeFMAC3(mV, int Fd, int Fs, int Ft) { sFLAG.doFlag = 1; - analyzeReg1(Fs); - analyzeReg3(Ft); - analyzeReg2(Fd, 0); + analyzeReg1(Fs, mVUup.VF_read[0]); + analyzeReg3(Ft, mVUup.VF_read[1]); + analyzeReg2(Fd, mVUup.VF_write, 0); } //------------------------------------------------------------------ // FMAC4 - Clip FMAC Opcode //------------------------------------------------------------------ -#define analyzeReg4(reg) { \ - if (reg) { mVUstall = aMax(mVUstall, aReg(reg).w); } \ -} - microVUt(void) mVUanalyzeFMAC4(mV, int Fs, int Ft) { cFLAG.doFlag = 1; - analyzeReg1(Fs); - analyzeReg4(Ft); + analyzeReg1(Fs, mVUup.VF_read[0]); + analyzeReg4(Ft, mVUup.VF_read[1]); } //------------------------------------------------------------------ // IALU - IALU Opcodes //------------------------------------------------------------------ -#define analyzeVIreg1(reg) { if (reg) { mVUstall = aMax(mVUstall, mVUregs.VI[reg]); } } -#define analyzeVIreg2(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; mVUlow.writesVI = 1; mVU->VIbackup[0] = reg; } } -#define analyzeVIreg3(reg, aCycles) { if (reg) { mVUregsTemp.VIreg = reg; mVUregsTemp.VI = aCycles; } } - microVUt(void) mVUanalyzeIALU1(mV, int Id, int Is, int It) { if (!Id) { mVUlow.isNOP = 1; } - analyzeVIreg1(Is); - analyzeVIreg1(It); - analyzeVIreg2(Id, 1); + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeVIreg1(It, mVUlow.VI_read[1]); + analyzeVIreg2(Id, mVUlow.VI_write, 1); } microVUt(void) mVUanalyzeIALU2(mV, int Is, int It) { if (!It) { mVUlow.isNOP = 1; } - analyzeVIreg1(Is); - analyzeVIreg2(It, 1); + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeVIreg2(It, mVUlow.VI_write, 1); } //------------------------------------------------------------------ // MR32 - MR32 Opcode //------------------------------------------------------------------ -// Flips xyzw stalls to yzwx -#define analyzeReg6(reg) { \ - if (reg) { \ - if (_X) { mVUstall = aMax(mVUstall, aReg(reg).y); } \ - if (_Y) { mVUstall = aMax(mVUstall, aReg(reg).z); } \ - if (_Z) { mVUstall = aMax(mVUstall, aReg(reg).w); } \ - if (_W) { mVUstall = aMax(mVUstall, aReg(reg).x); } \ - if (mVUregsTemp.VFreg[0] == reg) { \ - if ((mVUregsTemp.VF[0].y && _X) \ - || (mVUregsTemp.VF[0].z && _Y) \ - || (mVUregsTemp.VF[0].w && _Z) \ - || (mVUregsTemp.VF[0].x && _W)) \ - { mVUinfo.swapOps = 1; } \ - } \ - } \ -} - microVUt(void) mVUanalyzeMR32(mV, int Fs, int Ft) { if (!Ft) { mVUlow.isNOP = 1; } - analyzeReg6(Fs); - analyzeReg2(Ft, 1); + analyzeReg6(Fs, mVUlow.VF_read[0]); + analyzeReg2(Ft, mVUlow.VF_write, 1); } //------------------------------------------------------------------ // FDIV - DIV/SQRT/RSQRT Opcodes //------------------------------------------------------------------ -#define analyzeReg5(reg, fxf) { \ - if (reg) { \ - switch (fxf) { \ - case 0: mVUstall = aMax(mVUstall, aReg(reg).x); break; \ - case 1: mVUstall = aMax(mVUstall, aReg(reg).y); break; \ - case 2: mVUstall = aMax(mVUstall, aReg(reg).z); break; \ - case 3: mVUstall = aMax(mVUstall, aReg(reg).w); break; \ - } \ - if (mVUregsTemp.VFreg[0] == reg) { \ - if ((mVUregsTemp.VF[0].x && (fxf == 0)) \ - || (mVUregsTemp.VF[0].y && (fxf == 1)) \ - || (mVUregsTemp.VF[0].z && (fxf == 2)) \ - || (mVUregsTemp.VF[0].w && (fxf == 3))) \ - { mVUinfo.swapOps = 1; } \ - } \ - } \ -} - -#define analyzeQreg(x) { mVUregsTemp.q = x; mVUstall = aMax(mVUstall, mVUregs.q); } -#define analyzePreg(x) { mVUregsTemp.p = x; mVUstall = aMax(mVUstall, ((mVUregs.p) ? (mVUregs.p - 1) : 0)); } - microVUt(void) mVUanalyzeFDIV(mV, int Fs, int Fsf, int Ft, int Ftf, u8 xCycles) { mVUprint("microVU: DIV Opcode"); - analyzeReg5(Fs, Fsf); - analyzeReg5(Ft, Ftf); + analyzeReg5(Fs, Fsf, mVUlow.VF_read[0]); + analyzeReg5(Ft, Ftf, mVUlow.VF_read[1]); analyzeQreg(xCycles); } @@ -197,13 +239,13 @@ microVUt(void) mVUanalyzeFDIV(mV, int Fs, int Fsf, int Ft, int Ftf, u8 xCycles) microVUt(void) mVUanalyzeEFU1(mV, int Fs, int Fsf, u8 xCycles) { mVUprint("microVU: EFU Opcode"); - analyzeReg5(Fs, Fsf); + analyzeReg5(Fs, Fsf, mVUlow.VF_read[0]); analyzePreg(xCycles); } microVUt(void) mVUanalyzeEFU2(mV, int Fs, u8 xCycles) { mVUprint("microVU: EFU Opcode"); - analyzeReg1b(Fs); + analyzeReg1b(Fs, mVUlow.VF_read[0]); analyzePreg(xCycles); } @@ -213,7 +255,7 @@ microVUt(void) mVUanalyzeEFU2(mV, int Fs, u8 xCycles) { microVUt(void) mVUanalyzeMFP(mV, int Ft) { if (!Ft) { mVUlow.isNOP = 1; } - analyzeReg2(Ft, 1); + analyzeReg2(Ft, mVUlow.VF_write, 1); } //------------------------------------------------------------------ @@ -222,20 +264,19 @@ microVUt(void) mVUanalyzeMFP(mV, int Ft) { microVUt(void) mVUanalyzeMOVE(mV, int Fs, int Ft) { if (!Ft || (Ft == Fs)) { mVUlow.isNOP = 1; } - analyzeReg1b(Fs); - analyzeReg2(Ft, 1); + analyzeReg1b(Fs, mVUlow.VF_read[0]); + analyzeReg2 (Ft, mVUlow.VF_write, 1); } - //------------------------------------------------------------------ // LQx - LQ/LQD/LQI Opcodes //------------------------------------------------------------------ microVUt(void) mVUanalyzeLQ(mV, int Ft, int Is, bool writeIs) { - analyzeVIreg1(Is); - analyzeReg2(Ft, 1); + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeReg2 (Ft, mVUlow.VF_write, 1); if (!Ft) { if (writeIs && Is) { mVUlow.noWriteVF = 1; } else { mVUlow.isNOP = 1; } } - if (writeIs) { analyzeVIreg2(Is, 1); } + if (writeIs) { analyzeVIreg2(Is, mVUlow.VI_write, 1); } } //------------------------------------------------------------------ @@ -243,25 +284,23 @@ microVUt(void) mVUanalyzeLQ(mV, int Ft, int Is, bool writeIs) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeSQ(mV, int Fs, int It, bool writeIt) { - analyzeReg1b(Fs); - analyzeVIreg1(It); - if (writeIt) { analyzeVIreg2(It, 1); } + analyzeReg1b (Fs, mVUlow.VF_read[0]); + analyzeVIreg1(It, mVUlow.VI_read[0]); + if (writeIt) { analyzeVIreg2(It, mVUlow.VI_write, 1); } } //------------------------------------------------------------------ // R*** - R Reg Opcodes //------------------------------------------------------------------ -#define analyzeRreg() { mVUregsTemp.r = 1; } - microVUt(void) mVUanalyzeR1(mV, int Fs, int Fsf) { - analyzeReg5(Fs, Fsf); + analyzeReg5(Fs, Fsf, mVUlow.VF_read[0]); analyzeRreg(); } microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { if (!Ft) { if (canBeNOP) { mVUlow.isNOP = 1; } else { mVUlow.noWriteVF = 1; } } - analyzeReg2(Ft, 1); + analyzeReg2(Ft, mVUlow.VF_write, 1); analyzeRreg(); } @@ -280,7 +319,7 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } - analyzeVIreg3(It, 1); + analyzeVIreg3(It, mVUlow.VI_write, 1); } microVUt(void) mVUanalyzeFSSET(mV) { @@ -308,8 +347,8 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { } iPC = curPC; } - analyzeVIreg1(Is); - analyzeVIreg3(It, 1); + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeVIreg3(It, mVUlow.VI_write, 1); } //------------------------------------------------------------------ @@ -319,18 +358,15 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { microVUt(void) mVUanalyzeCflag(mV, int It) { mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } - analyzeVIreg3(It, 1); + analyzeVIreg3(It, mVUlow.VI_write, 1); } //------------------------------------------------------------------ // XGkick //------------------------------------------------------------------ -#define analyzeXGkick1() { mVUstall = aMax(mVUstall, mVUregs.xgkick); } -#define analyzeXGkick2(x) { mVUregsTemp.xgkick = x; } - microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { - analyzeVIreg1(Fs); + analyzeVIreg1(Fs, mVUlow.VI_read[0]); analyzeXGkick1(); analyzeXGkick2(xCycles); // Note: Technically XGKICK should stall on the next instruction, @@ -360,11 +396,11 @@ microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { } microVUt(void) mVUanalyzeBranch1(mV, int Is) { - if (mVUregs.VI[Is] || mVUstall) { analyzeVIreg1(Is); } + if (mVUregs.VI[Is] || mVUstall) { analyzeVIreg1(Is, mVUlow.VI_read[0]); } else { analyzeBranchVI(Is, mVUlow.memReadIs); } } microVUt(void) mVUanalyzeBranch2(mV, int Is, int It) { - if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is); analyzeVIreg1(It); } + if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is, mVUlow.VI_read[0]); analyzeVIreg1(It, mVUlow.VI_read[1]); } else { analyzeBranchVI(Is, mVUlow.memReadIs); analyzeBranchVI(It, mVUlow.memReadIt);} } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index a42a683a25..5400116e01 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -65,12 +65,31 @@ struct microBlock { u8* x86ptrStart; // Start of code }; +struct microVFreg { + u8 reg; // Reg Index + u8 x; // X vector read/written to? + u8 y; // Y vector read/written to? + u8 z; // Z vector read/written to? + u8 w; // W vector read/written to? +}; + +struct microVIreg { + u8 reg; // Reg Index + u8 used; // Reg is Used? (Read/Written) +}; + struct microUpperOp { - bool eBit; // Has E-bit set - bool iBit; // Has I-bit set + bool eBit; // Has E-bit set + bool iBit; // Has I-bit set + microVFreg VF_write; // VF Vectors written to by this instruction + microVFreg VF_read[2]; // VF Vectors read by this instruction }; struct microLowerOp { + microVFreg VF_write; // VF Vectors written to by this instruction + microVFreg VF_read[2]; // VF Vectors read by this instruction + microVIreg VI_write; // VI reg written to by this instruction + microVIreg VI_read[2]; // VI regs read by this instruction bool isNOP; // This instruction is a NOP bool isFSSET; // This instruction is a FSSET bool useSflag; // This instruction uses/reads Sflag diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index c7a4b7e0ca..3712a2f718 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -664,7 +664,7 @@ mVUop(mVU_ISUBIU) { //------------------------------------------------------------------ mVUop(mVU_MFIR) { - pass1 { if (!_Ft_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeReg2(_Ft_, 1); } + pass1 { if (!_Ft_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeReg2(_Ft_, mVUlow.VF_write, 1); } pass2 { mVUallocVIa(mVU, gprT1, _Is_); MOVSX32R16toR(gprT1, gprT1); @@ -704,7 +704,7 @@ mVUop(mVU_MR32) { } mVUop(mVU_MTIR) { - pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeReg5(_Fs_, _Fsf_); analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeReg5(_Fs_, _Fsf_, mVUlow.VF_read[0]); analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->VF[_Fs_].UL[_Fsf_]); mVUallocVIb(mVU, gprT1, _It_); @@ -717,7 +717,7 @@ mVUop(mVU_MTIR) { //------------------------------------------------------------------ mVUop(mVU_ILW) { - pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeVIreg2(_It_, mVUlow.VI_write, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + getVUmem(_Imm11_) + offsetSS); @@ -735,7 +735,7 @@ mVUop(mVU_ILW) { } mVUop(mVU_ILWR) { - pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 4); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeVIreg2(_It_, mVUlow.VI_write, 4); } pass2 { if (!_Is_) { MOVZX32M16toR(gprT1, (uptr)mVU->regs->Mem + offsetSS); @@ -756,7 +756,7 @@ mVUop(mVU_ILWR) { //------------------------------------------------------------------ mVUop(mVU_ISW) { - pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } + pass1 { analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeVIreg1(_It_, mVUlow.VI_read[1]); } pass2 { if (!_Is_) { int imm = getVUmem(_Imm11_); @@ -781,7 +781,7 @@ mVUop(mVU_ISW) { } mVUop(mVU_ISWR) { - pass1 { analyzeVIreg1(_Is_); analyzeVIreg1(_It_); } + pass1 { analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeVIreg1(_It_, mVUlow.VI_read[1]); } pass2 { if (!_Is_) { mVUallocVIa(mVU, gprT1, _It_); @@ -1018,7 +1018,7 @@ mVUop(mVU_WAITQ) { //------------------------------------------------------------------ mVUop(mVU_XTOP) { - pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->top); mVUallocVIb(mVU, gprT1, _It_); @@ -1027,7 +1027,7 @@ mVUop(mVU_XTOP) { } mVUop(mVU_XITOP) { - pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, 1); } + pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->itop); mVUallocVIb(mVU, gprT1, _It_); @@ -1100,7 +1100,7 @@ mVUop(mVU_B) { mVUop(mVU_BAL) { setBranchA(2, _It_); - pass1 { analyzeVIreg2(_It_, 1); } + pass1 { analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { MOV32ItoR(gprT1, bSaveAddr); mVUallocVIb(mVU, gprT1, _It_); @@ -1180,7 +1180,7 @@ mVUop(mVU_IBNE) { mVUop(mVU_JR) { mVUbranch = 9; - pass1 { analyzeVIreg1(_Is_); } + pass1 { analyzeVIreg1(_Is_, mVUlow.VI_read[0]); } pass2 { mVUallocVIa(mVU, gprT1, _Is_); SHL32ItoR(gprT1, 3); @@ -1192,7 +1192,7 @@ mVUop(mVU_JR) { mVUop(mVU_JALR) { mVUbranch = 10; - pass1 { analyzeVIreg1(_Is_); analyzeVIreg2(_It_, 1); } + pass1 { analyzeVIreg1(_Is_, mVUlow.VI_read[0]); analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { mVUallocVIa(mVU, gprT1, _Is_); SHL32ItoR(gprT1, 3); From 8c5b6ae749858d855d171ec607ce604490761400 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 2 Jun 2009 00:50:37 +0000 Subject: [PATCH 167/296] microVU: - Implemented some bizarre behavior where conditional branches will read old VI-reg values if the previous instructions continuously read+write to the VI reg. example from zerorec's comments: SQI vi5++ SQI vi5++ IBNE vi4, vi5 vi5's value in the branch should be the value before all the SQI's... more examples in iVUzerorec.cpp git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1304 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 65 ++++++++++++++++++----------------- pcsx2/x86/microVU_IR.h | 1 + 2 files changed, 34 insertions(+), 32 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 9a07f3c2e5..7319d4e4cf 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -29,6 +29,7 @@ #define aReg(x) mVUregs.VF[x] #define bReg(x, y) mVUregsTemp.VFreg[y] = x; mVUregsTemp.VF[y] #define aMax(x, y) ((x > y) ? x : y) +#define aMin(x, y) ((x < y) ? x : y) // Read a VF reg by upper op #define analyzeReg1(xReg, vfRead) { \ @@ -128,18 +129,6 @@ // Writing to a VI reg #define analyzeVIreg2(xReg, viWrite, aCycles) { \ - if (xReg) { \ - mVUregsTemp.VIreg = xReg; \ - mVUregsTemp.VI = aCycles; \ - mVUlow.writesVI = 1; \ - mVU->VIbackup[0] = xReg; \ - viWrite.reg = xReg; \ - viWrite.used = aCycles; \ - } \ -} - -// Writing to a VI reg (FSxxx, FMxxx, FCxxx opcodes) -#define analyzeVIreg3(xReg, viWrite, aCycles) { \ if (xReg) { \ mVUregsTemp.VIreg = xReg; \ mVUregsTemp.VI = aCycles; \ @@ -319,16 +308,13 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } - analyzeVIreg3(It, mVUlow.VI_write, 1); + mVUlow.readFlags = 1; + analyzeVIreg2(It, mVUlow.VI_write, 1); } microVUt(void) mVUanalyzeFSSET(mV) { mVUlow.isFSSET = 1; - // mVUinfo &= ~_doStatus; - // Note: I'm not entirely sure if the non-sticky flags - // should be taken from the current upper instruction - // or if they should be taken from the previous instruction - // Uncomment the above line if the latter-case is true + mVUlow.readFlags = 1; } //------------------------------------------------------------------ @@ -347,8 +333,9 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { } iPC = curPC; } + mVUlow.readFlags = 1; analyzeVIreg1(Is, mVUlow.VI_read[0]); - analyzeVIreg3(It, mVUlow.VI_write, 1); + analyzeVIreg2(It, mVUlow.VI_write, 1); } //------------------------------------------------------------------ @@ -357,8 +344,9 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { microVUt(void) mVUanalyzeCflag(mV, int It) { mVUinfo.swapOps = 1; + mVUlow.readFlags = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } - analyzeVIreg3(It, mVUlow.VI_write, 1); + analyzeVIreg2(It, mVUlow.VI_write, 1); } //------------------------------------------------------------------ @@ -381,18 +369,31 @@ microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { // Branches - Branch Opcodes //------------------------------------------------------------------ -#define analyzeBranchVI(reg, infoVar) { \ - /* First ensure branch is not first opcode in block */ \ - if (reg && (mVUcount > 0)) { \ - incPC2(-2); \ - /* Check if prev Op modified VI reg */ \ - if (mVUlow.writesVI && (reg == mVU->VIbackup[0])) { \ - mVUlow.backupVI = 1; \ - incPC2(2); \ - infoVar = 1; \ - } \ - else { incPC2(2); } \ - } \ +#define analyzeBranchVI(xReg, infoVar) { \ + if (xReg) { \ + int i; \ + int iEnd = aMin(4, mVUcount); \ + int bPC = iPC; \ + for (i = 0; i < iEnd; i++) { \ + incPC2(-2) \ + if ((mVUlow.VI_write.reg == xReg) && mVUlow.VI_write.used) { \ + if (mVUlow.readFlags) break; \ + if (i == 0) continue; \ + if (((mVUlow.VI_read[0].reg == xReg) && (mVUlow.VI_read[0].used)) \ + || ((mVUlow.VI_read[1].reg == xReg) && (mVUlow.VI_read[1].used))) \ + { continue; } \ + } \ + break; \ + } \ + if (i) { \ + DevCon::Status("microVU%d: Branch VI-Delay (%d)", params getIndex, i); \ + incPC2(2); \ + mVUlow.backupVI = 1; \ + iPC = bPC; \ + infoVar = 1; \ + } \ + iPC = bPC; \ + } \ } microVUt(void) mVUanalyzeBranch1(mV, int Is) { diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 5400116e01..6bce6c9775 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -99,6 +99,7 @@ struct microLowerOp { bool memReadIs; // Read Is (VI reg) from memory (used by branches) bool memReadIt; // Read If (VI reg) from memory (used by branches) bool writesVI; // Current Instruction writes to VI (used by branches; note that flag-modifying opcodes shouldn't set this) + bool readFlags; // Current Instruction reads Status, Mac, or Clip flags }; struct microFlagInst { From 074a87eadcf9f78a75655b18d1b1cd46dfe87fdb Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 3 Jun 2009 01:06:26 +0000 Subject: [PATCH 168/296] Improved the cpu speed/ghz detection. It should be a bit more accurate, and is faster (pcsx2 startup times are near instantaneous now). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1306 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/ix86/ix86_cpudetect.cpp | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/pcsx2/x86/ix86/ix86_cpudetect.cpp b/pcsx2/x86/ix86/ix86_cpudetect.cpp index 6e211619f3..4de1fd1123 100644 --- a/pcsx2/x86/ix86/ix86_cpudetect.cpp +++ b/pcsx2/x86/ix86/ix86_cpudetect.cpp @@ -196,20 +196,20 @@ s64 CPUSpeedHz( int time ) // Align the cpu execution to a timeGetTime boundary. // Without this the result could be skewed by up to several milliseconds. - - do { timeStart = timeGetTime( ); + + do { timeStart = GetCPUTicks(); } while( timeGetTime( ) == timeStart ); do { - timeStop = timeGetTime( ); + timeStop = GetCPUTicks( ); startTick = GetCPUTick( ); - } while( ( timeStop - timeStart ) < 1 ); + } while( ( timeStop - timeStart ) == 0 ); timeStart = timeStop; do { - timeStop = timeGetTime(); + timeStop = GetCPUTicks(); endTick = GetCPUTick(); } while( ( timeStop - timeStart ) < time ); @@ -364,10 +364,15 @@ void cpudetectInit() cpucaps.hasAMD64BitArchitecture = ( cpuinfo.x86EFlags >> 29 ) & 1; //64bit cpu cpucaps.has3DNOWInstructionExtensionsExt = ( cpuinfo.x86EFlags >> 30 ) & 1; //3dnow+ cpucaps.has3DNOWInstructionExtensions = ( cpuinfo.x86EFlags >> 31 ) & 1; //3dnow - cpucaps.hasStreamingSIMD4ExtensionsA = ( cpuinfo.x86EFlags2 >> 6 ) & 1; //INSERTQ / EXTRQ / MOVNT + cpucaps.hasStreamingSIMD4ExtensionsA = ( cpuinfo.x86EFlags2 >> 6 ) & 1; //INSERTQ / EXTRQ / MOVNT + InitCPUTicks(); + uint span = GetTickFrequency(); - cpuinfo.cpuspeed = (u32)(CPUSpeedHz( 400 ) / 400000 ); + if( (span % 1000) < 400 ) // helps minimize rounding errors + cpuinfo.cpuspeed = (u32)( CPUSpeedHz( span / 1000 ) / 1000 ); + else + cpuinfo.cpuspeed = (u32)( CPUSpeedHz( span / 500 ) / 2000 ); // --> SSE3 / SSSE3 / SSE4.1 / SSE 4.2 detection <-- From 510a23f1906091399a798a927f9ed95abd7ad456 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 3 Jun 2009 01:28:19 +0000 Subject: [PATCH 169/296] Quick cleanup of my prev commit -- changed some u32's to u64's, and deleted some code made obsolete by the new cpu detection. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1307 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/PrecompiledHeader.h | 7 ------- pcsx2/x86/ix86/ix86_cpudetect.cpp | 19 +++++++++---------- pcsx2/x86/ix86/ix86_legacy_internal.h | 1 - 3 files changed, 9 insertions(+), 18 deletions(-) diff --git a/pcsx2/PrecompiledHeader.h b/pcsx2/PrecompiledHeader.h index 8e91ef6769..c2de341fea 100644 --- a/pcsx2/PrecompiledHeader.h +++ b/pcsx2/PrecompiledHeader.h @@ -103,13 +103,6 @@ typedef int BOOL; # define __declspec(x) # endif -static __forceinline u32 timeGetTime() -{ - struct timeb t; - ftime(&t); - return (u32)(t.time*1000+t.millitm); -} - # ifndef strnicmp # define strnicmp strncasecmp # endif diff --git a/pcsx2/x86/ix86/ix86_cpudetect.cpp b/pcsx2/x86/ix86/ix86_cpudetect.cpp index 4de1fd1123..26a0dbb63c 100644 --- a/pcsx2/x86/ix86/ix86_cpudetect.cpp +++ b/pcsx2/x86/ix86/ix86_cpudetect.cpp @@ -131,7 +131,7 @@ extern s32 iCpuId( u32 cmd, u32 *regs ) #endif // _MSC_VER } -u64 GetCPUTick( void ) +u64 GetRdtsc( void ) { #if defined (_MSC_VER) && _MSC_VER >= 1400 @@ -152,7 +152,7 @@ u64 GetCPUTick( void ) ////////////////////////////////////////////////////////////////////////////////////////// // Note: This function doesn't support GCC/Linux. Looking online it seems the only -// way to simulate the Micrsoft SEH model is to use unix signals, and the 'sigaction' +// way to simulate the Microsoft SEH model is to use unix signals, and the 'sigaction' // function specifically. Maybe a project for a linux developer at a later date. :) #ifdef _MSC_VER static bool _test_instruction( void* pfnCall ) @@ -182,9 +182,9 @@ static char* bool_to_char( bool testcond ) ////////////////////////////////////////////////////////////////////////////////////////// // -s64 CPUSpeedHz( int time ) +s64 CPUSpeedHz( u64 time ) { - int timeStart, + u64 timeStart, timeStop; s64 startTick, endTick; @@ -194,23 +194,22 @@ s64 CPUSpeedHz( int time ) return 0; //check if function is supported } - // Align the cpu execution to a timeGetTime boundary. - // Without this the result could be skewed by up to several milliseconds. + // Align the cpu execution to a cpuTick boundary. do { timeStart = GetCPUTicks(); - } while( timeGetTime( ) == timeStart ); + } while( GetCPUTicks() == timeStart ); do { timeStop = GetCPUTicks( ); - startTick = GetCPUTick( ); + startTick = GetRdtsc( ); } while( ( timeStop - timeStart ) == 0 ); timeStart = timeStop; do { timeStop = GetCPUTicks(); - endTick = GetCPUTick(); + endTick = GetRdtsc(); } while( ( timeStop - timeStart ) < time ); @@ -367,7 +366,7 @@ void cpudetectInit() cpucaps.hasStreamingSIMD4ExtensionsA = ( cpuinfo.x86EFlags2 >> 6 ) & 1; //INSERTQ / EXTRQ / MOVNT InitCPUTicks(); - uint span = GetTickFrequency(); + u64 span = GetTickFrequency(); if( (span % 1000) < 400 ) // helps minimize rounding errors cpuinfo.cpuspeed = (u32)( CPUSpeedHz( span / 1000 ) / 1000 ); diff --git a/pcsx2/x86/ix86/ix86_legacy_internal.h b/pcsx2/x86/ix86/ix86_legacy_internal.h index 90ba4aef49..8bc97ef99f 100644 --- a/pcsx2/x86/ix86/ix86_legacy_internal.h +++ b/pcsx2/x86/ix86/ix86_legacy_internal.h @@ -49,5 +49,4 @@ extern void SibSB( uint ss, uint index, uint base ); extern void SET8R( int cc, int to ); extern u8* J8Rel( int cc, int to ); extern u32* J32Rel( int cc, u32 to ); -extern u64 GetCPUTick( void ); From d8fb66f143c97d5bc8f7486a3d4ca9da589905bd Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 3 Jun 2009 06:50:13 +0000 Subject: [PATCH 170/296] microVU: - Minor Changes/Cleanup Super VU: - Used the Linux SysMmapEx address for windows as well. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1308 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVUzerorec.cpp | 5 ----- pcsx2/x86/microVU.h | 2 +- pcsx2/x86/microVU_Alloc.inl | 7 +++---- pcsx2/x86/microVU_Analyze.inl | 14 ++++++++++---- pcsx2/x86/microVU_IR.h | 1 - pcsx2/x86/microVU_Lower.inl | 16 ++++++++-------- 6 files changed, 22 insertions(+), 23 deletions(-) diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index dc499e432f..15ad37ab9c 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -343,12 +343,7 @@ void SuperVUAlloc(int vuindex) { // upper 4 bits must be zero! // Changed "first try base" to 0xb800000, since 0x0c000000 liked to fail a lot. (air) - // With microVU as optional now, Linux fails with 0x0e000000, and prefers 0x0c000000. --arcum42 -#ifdef __LINUX__ s_recVUMem = SysMmapEx(0x0c000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); -#else - s_recVUMem = SysMmapEx(0x0e000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); -#endif if (s_recVUMem == NULL) { diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7fd6efc281..5b3f640b50 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -116,7 +116,7 @@ struct microVU { u8* exitFunct; // Ptr Function to the Exit code for recompiled programs u32 code; // Contains the current Instruction u32 divFlag; // 1 instance of I/D flags - u32 VIbackup[2]; // Holds a backup of a VI reg if modified before a branch + u32 VIbackup; // Holds a backup of a VI reg if modified before a branch u32 VIxgkick; // Holds a backup of a VI reg used for xgkick-delays u32 branch; // Holds branch compare result (IBxx) OR Holds address to Jump to (JALR/JR) u32 p; // Holds current P instance index diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index eb0c5deda0..cfd8ffac72 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -687,10 +687,9 @@ microVUt(void) mVUallocVIa(mV, int GPRreg, int _reg_) { microVUt(void) mVUallocVIb(mV, int GPRreg, int _reg_) { if (mVUlow.backupVI) { // Backs up reg to memory (used when VI is modified b4 a branch) - MOV32RtoM((uptr)&mVU->VIbackup[1], GPRreg); - mVUallocVIa(mVU, GPRreg, _reg_); - MOV32RtoM((uptr)&mVU->VIbackup[0], GPRreg); - MOV32MtoR(GPRreg, (uptr)&mVU->VIbackup[1]); + MOVZX32M16toR(gprR, (uptr)&mVU->regs->VI[_reg_].UL); + MOV32RtoM((uptr)&mVU->VIbackup, gprR); + MOV32ItoR(gprR, Roffset); } if (_reg_ == 0) { return; } else if (isMMX(_reg_)) { MOVD32RtoMMX(mmVI(_reg_), GPRreg); } diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 7319d4e4cf..ad6452c81d 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -372,12 +372,12 @@ microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { #define analyzeBranchVI(xReg, infoVar) { \ if (xReg) { \ int i; \ - int iEnd = aMin(4, mVUcount); \ + int iEnd = aMin(5, mVUcount); \ int bPC = iPC; \ for (i = 0; i < iEnd; i++) { \ incPC2(-2) \ if ((mVUlow.VI_write.reg == xReg) && mVUlow.VI_write.used) { \ - if (mVUlow.readFlags) break; \ + if (mVUlow.readFlags || i == 5) break; \ if (i == 0) continue; \ if (((mVUlow.VI_read[0].reg == xReg) && (mVUlow.VI_read[0].used)) \ || ((mVUlow.VI_read[1].reg == xReg) && (mVUlow.VI_read[1].used))) \ @@ -402,6 +402,12 @@ microVUt(void) mVUanalyzeBranch1(mV, int Is) { } microVUt(void) mVUanalyzeBranch2(mV, int Is, int It) { - if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { analyzeVIreg1(Is, mVUlow.VI_read[0]); analyzeVIreg1(It, mVUlow.VI_read[1]); } - else { analyzeBranchVI(Is, mVUlow.memReadIs); analyzeBranchVI(It, mVUlow.memReadIt);} + if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeVIreg1(It, mVUlow.VI_read[1]); + } + else { + analyzeBranchVI(Is, mVUlow.memReadIs); + analyzeBranchVI(It, mVUlow.memReadIt); + } } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 6bce6c9775..a109503651 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -98,7 +98,6 @@ struct microLowerOp { bool backupVI; // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) bool memReadIs; // Read Is (VI reg) from memory (used by branches) bool memReadIt; // Read If (VI reg) from memory (used by branches) - bool writesVI; // Current Instruction writes to VI (used by branches; note that flag-modifying opcodes shouldn't set this) bool readFlags; // Current Instruction reads Status, Mac, or Clip flags }; diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 3712a2f718..f7410f8d96 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1112,9 +1112,9 @@ mVUop(mVU_IBEQ) { setBranchA(3, 0); pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); - if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup); else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1125,7 +1125,7 @@ mVUop(mVU_IBGEZ) { setBranchA(4, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1136,7 +1136,7 @@ mVUop(mVU_IBGTZ) { setBranchA(5, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1147,7 +1147,7 @@ mVUop(mVU_IBLEZ) { setBranchA(6, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1158,7 +1158,7 @@ mVUop(mVU_IBLTZ) { setBranchA(7, 0); pass1 { mVUanalyzeBranch1(mVU, _Is_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); MOV32RtoM((uptr)&mVU->branch, gprT1); } @@ -1169,9 +1169,9 @@ mVUop(mVU_IBNE) { setBranchA(8, 0); pass1 { mVUanalyzeBranch2(mVU, _Is_, _It_); } pass2 { - if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIs) MOV32MtoR(gprT1, (uptr)&mVU->VIbackup); else mVUallocVIa(mVU, gprT1, _Is_); - if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup[0]); + if (mVUlow.memReadIt) XOR32MtoR(gprT1, (uptr)&mVU->VIbackup); else { mVUallocVIa(mVU, gprT2, _It_); XOR32RtoR(gprT1, gprT2); } MOV32RtoM((uptr)&mVU->branch, gprT1); } From 96f7911d263ed0d30e5b9150a86019b36871635f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 3 Jun 2009 09:22:25 +0000 Subject: [PATCH 171/296] microVU: - Second try at a status flag optimization. Still needs tweaking but too tired xD git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1309 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Alloc.inl | 31 +++++++++++++- pcsx2/x86/microVU_Compile.inl | 10 ++--- pcsx2/x86/microVU_Execute.inl | 16 +++++++- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 77 +++++++++++++++++++++++------------ pcsx2/x86/microVU_Misc.h | 3 ++ pcsx2/x86/microVU_Misc.inl | 2 + pcsx2/x86/microVU_Upper.inl | 22 ++++------ 8 files changed, 116 insertions(+), 47 deletions(-) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index cfd8ffac72..d14ba7a570 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -647,17 +647,44 @@ microVUt(void) mVUallocFMAC26b(mV, int& ACCw, int& ACCr) { } \ } +#define setBitSFLAG(bitTest, bitSet) { \ + TEST32ItoR(regT, bitTest); \ + pjmp = JZ8(0); \ + OR32ItoR(reg, bitSet); \ + x86SetJ8(pjmp); \ +} + +#define setBitFSEQ(bitX) { \ + TEST32ItoR(gprT1, bitX); \ + pjmp = JZ8(0); \ + OR32ItoR(gprT1, bitX); \ + x86SetJ8(pjmp); \ +} + microVUt(void) mVUallocSFLAGa(int reg, int fInstance) { getFlagReg(fInstance, fInstance); - MOVZX32R16toR(reg, fInstance); + MOV32RtoR(reg, fInstance); } microVUt(void) mVUallocSFLAGb(int reg, int fInstance) { getFlagReg(fInstance, fInstance); - //AND32ItoR(reg, 0xffff); MOV32RtoR(fInstance, reg); } +// Normalize Status Flag +microVUt(void) mVUallocSFLAGc(int reg, int regT, int fInstance) { + u8 *pjmp; + XOR32RtoR(reg, reg); + mVUallocSFLAGa(regT, fInstance); + setBitSFLAG(0x000f, 0x0001); // Z Bit + setBitSFLAG(0x00f0, 0x0002); // S Bit + setBitSFLAG(0x0f00, 0x0040); // ZS Bit + setBitSFLAG(0xf000, 0x0080); // SS Bit + AND32ItoR(regT, 0xffff0000); // DS/DI/OS/US/D/I/O/U Bits + SHR32ItoR(regT, 14); + OR32RtoR(reg, regT); +} + microVUt(void) mVUallocMFLAGa(mV, int reg, int fInstance) { MOVZX32M16toR(reg, (uptr)&mVU->macFlag[fInstance]); } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 689d8b82a4..5dd94ca18f 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -169,8 +169,8 @@ microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, in // Save Flag Instances if (!mVUflagHack) { - getFlagReg(fStatus, fStatus); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, fStatus); + mVUallocSFLAGc(gprT1, gprT2, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); } mVUallocMFLAGa(mVU, gprT1, fMac); mVUallocCFLAGa(mVU, gprT2, fClip); @@ -196,13 +196,13 @@ microVUt(void) mVUtestCycles(mV) { iPC = mVUstartPC; mVUdebugNOW(0); SUB32ItoM((uptr)&mVU->cycles, mVUcycles); - u8* jmp8 = JG8(0); + u32* jmp32 = JG32(0); MOV32ItoR(gprT2, xPC); if (!isVU1) CALLFunc((uptr)mVUwarning0); else CALLFunc((uptr)mVUwarning1); MOV32ItoR(gprR, Roffset); // Restore gprR mVUendProgram(mVU, 0, 0, sI, 0, cI); - x86SetJ8(jmp8); + x86SetJ32(jmp32); } //------------------------------------------------------------------ @@ -378,7 +378,7 @@ eBitTemination: if (mVUinfo.doDivFlag) { int flagReg; getFlagReg(flagReg, lStatus); - AND32ItoR (flagReg, 0x0fcf); + AND32ItoR (flagReg, 0xfff3ffff); OR32MtoR (flagReg, (uptr)&mVU->divFlag); } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 0854b0bad3..e3fe31d202 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -42,10 +42,24 @@ microVUt(void) mVUdispatcherA(mV) { // Load Regs MOV32ItoR(gprR, Roffset); // Load VI Reg Offset MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); - AND32ItoR(gprF0, 0xffff); + MOV32RtoR(gprF1, gprF0); + SHL32ItoR(gprF1, 3); + AND32ItoR(gprF1, 0x218); + MOV32RtoR(gprF2, gprF0); + SHL32ItoR(gprF2, 5); + AND32ItoR(gprF2, 0x1000); + OR32RtoR (gprF1, gprF2); + MOV32RtoR(gprF3, gprF0); + SHL32ItoR(gprF3, 14); + AND32ItoR(gprF3, 0x3cf0000); + OR32RtoR (gprF1, gprF3); + + MOV32RtoR(gprF0, gprF1); + MOV32RtoR(gprF2, gprF1); + MOV32RtoR(gprF3, gprF1); SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VI[REG_MAC_FLAG].UL); SSE_SHUFPS_XMM_to_XMM (xmmT1, xmmT1, 0); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index fc63c3db9d..7ed2e26245 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -24,7 +24,7 @@ microVUt(void) mVUdivSet(mV) { if (mVUinfo.doDivFlag) { getFlagReg(flagReg1, sFLAG.write); if (!sFLAG.doFlag) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } - AND32ItoR(flagReg1, 0x0fcf); + AND32ItoR(flagReg1, 0xfff3ffff); OR32MtoR (flagReg1, (uptr)&mVU->divFlag); } } diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index f7410f8d96..cef71c3874 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -37,7 +37,7 @@ SSE_MOVMSKPS_XMM_to_R32(gprTemp, xmmReg); \ TEST32ItoR(gprTemp, 1); /* Check sign bit */ \ aJump = JZ8(0); /* Skip if positive */ \ - MOV32ItoM((uptr)&mVU->divFlag, 0x410); /* Set Invalid Flags */ \ + MOV32ItoM((uptr)&mVU->divFlag, divI); /* Set Invalid Flags */ \ SSE_ANDPS_M128_to_XMM(xmmReg, (uptr)mVU_absclip); /* Abs(xmmReg) */ \ x86SetJ8(aJump); \ } @@ -54,10 +54,10 @@ mVUop(mVU_DIV) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero ajmp = JZ8(0); - MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, divI); // Set invalid flag (0/0) bjmp = JMP8(0); x86SetJ8(ajmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, divD); // Zero divide (only when not 0/0) x86SetJ8(bjmp); SSE_XORPS_XMM_to_XMM(xmmFs, xmmFt); @@ -112,10 +112,10 @@ mVUop(mVU_RSQRT) { testZero(xmmFs, xmmT1, gprT1); // Test if Fs is zero bjmp = JZ8(0); // Skip if none are - MOV32ItoM((uptr)&mVU->divFlag, 0x410); // Set invalid flag (0/0) + MOV32ItoM((uptr)&mVU->divFlag, divI); // Set invalid flag (0/0) cjmp = JMP8(0); x86SetJ8(bjmp); - MOV32ItoM((uptr)&mVU->divFlag, 0x820); // Zero divide flag (only when not 0/0) + MOV32ItoM((uptr)&mVU->divFlag, divD); // Zero divide flag (only when not 0/0) x86SetJ8(cjmp); SSE_ANDPS_M128_to_XMM(xmmFs, (uptr)mVU_signbit); @@ -522,7 +522,7 @@ mVUop(mVU_FMOR) { mVUop(mVU_FSAND) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); + mVUallocSFLAGc(gprT1, gprT2, sFLAG.read); AND16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -530,23 +530,10 @@ mVUop(mVU_FSAND) { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } -mVUop(mVU_FSEQ) { - pass1 { mVUanalyzeSflag(mVU, _It_); } - pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); - XOR16ItoR(gprT1, _Imm12_); - SUB16ItoR(gprT1, 1); - SHR16ItoR(gprT1, 15); - mVUallocVIb(mVU, gprT1, _It_); - } - pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } -} - mVUop(mVU_FSOR) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { - mVUallocSFLAGa(gprT1, sFLAG.read); + mVUallocSFLAGc(gprT1, gprT2, sFLAG.read); OR16ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } @@ -554,14 +541,54 @@ mVUop(mVU_FSOR) { pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } } +mVUop(mVU_FSEQ) { + pass1 { mVUanalyzeSflag(mVU, _It_); } + pass2 { + u8 *pjmp; + int imm = 0; + if (_Imm12_ & 0x0001) imm |= 0x000000f; // Z + if (_Imm12_ & 0x0002) imm |= 0x00000f0; // S + if (_Imm12_ & 0x0004) imm |= 0x0010000; // U + if (_Imm12_ & 0x0008) imm |= 0x0020000; // O + if (_Imm12_ & 0x0010) imm |= 0x0040000; // I + if (_Imm12_ & 0x0020) imm |= 0x0080000; // D + if (_Imm12_ & 0x0040) imm |= 0x0000f00; // ZS + if (_Imm12_ & 0x0080) imm |= 0x000f000; // SS + if (_Imm12_ & 0x0100) imm |= 0x0400000; // US + if (_Imm12_ & 0x0200) imm |= 0x0800000; // OS + if (_Imm12_ & 0x0400) imm |= 0x1000000; // IS + if (_Imm12_ & 0x0800) imm |= 0x2000000; // DS + + mVUallocSFLAGa(gprT1, sFLAG.read); + setBitFSEQ(0x000f); // Z bit + setBitFSEQ(0x00f0); // S bit + setBitFSEQ(0x0f00); // ZS bit + setBitFSEQ(0xf000); // SS bit + XOR32ItoR(gprT1, imm); + SUB32ItoR(gprT1, 1); + SHR32ItoR(gprT1, 31); + mVUallocVIb(mVU, gprT1, _It_); + } + pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } + pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } +} + mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { - int flagReg1, flagReg2; - getFlagReg(flagReg1, sFLAG.write); - if (!(sFLAG.doFlag||mVUinfo.doDivFlag)) { getFlagReg(flagReg2, sFLAG.lastWrite); MOV32RtoR(flagReg1, flagReg2); } // Get status result from last status setting instruction - AND32ItoR(flagReg1, 0x03f); - OR32ItoR (flagReg1, (_Imm12_ & 0xfc0)); + int sReg, imm = 0; + if (_Imm12_ & 0x0040) imm |= 0x0000f00; // ZS + if (_Imm12_ & 0x0080) imm |= 0x000f000; // SS + if (_Imm12_ & 0x0100) imm |= 0x0400000; // US + if (_Imm12_ & 0x0200) imm |= 0x0800000; // OS + if (_Imm12_ & 0x0400) imm |= 0x1000000; // IS + if (_Imm12_ & 0x0800) imm |= 0x2000000; // DS + getFlagReg(sReg, sFLAG.write); + if (!(sFLAG.doFlag || mVUinfo.doDivFlag)) { + mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag + } + AND32ItoR(sReg, 0xf00ff); // Keep Non-Sticky Bits + if (imm) OR32ItoR(sReg, imm); } pass3 { mVUlog("FSSET $%x", _Imm12_); } pass4 { mVUsFlagHack = 0; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index d36cfc7a8e..48d2ebcd59 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -105,6 +105,9 @@ declareAllVariables #define _Tbit_ (1<<27) #define _MDTbit_ 0 //( _Mbit_ | _Dbit_ | _Tbit_ ) // ToDo: Implement this stuff... +#define divI 0x1040000 +#define divD 0x2080000 + #define isVU1 (mVU == µVU1) #define getIndex (isVU1 ? 1 : 0) #define getVUmem(x) (((isVU1) ? (x & 0x3ff) : ((x >= 0x400) ? (x & 0x43f) : (x & 0xff))) * 16) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index f6d2c58b00..138683c9e9 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -300,6 +300,7 @@ microVUt(void) mVUcheckSflag(mV, int progIndex) { if (CHECK_VU_FLAGHACK1) { int bFlagInfo = mVUflagInfo; int bCode = mVU->code; + int bFlagHack = mVUsFlagHack; mVUsFlagHack = 1; for (u32 i = 0; i < mVU->progSize; i+=2) { mVU->code = mVU->prog.prog[progIndex].data[i]; @@ -308,6 +309,7 @@ microVUt(void) mVUcheckSflag(mV, int progIndex) { mVUflagInfo = bFlagInfo; mVU->code = bCode; mVU->prog.prog[progIndex].sFlagHack = mVUsFlagHack; + mVUsFlagHack = bFlagHack; } } diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index c2bb91e294..db0e92a51e 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -29,7 +29,6 @@ // Note: If modXYZW is true, then it adjusts XYZW for Single Scalar operations microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool modXYZW) { int sReg, mReg = gprT1; - static u8 *pjmp, *pjmp2; static const u16 flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; //SysPrintf("Status = %d; Mac = %d\n", sFLAG.doFlag, mFLAG.doFlag); @@ -40,7 +39,7 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool if (sFLAG.doFlag) { getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag - AND32ItoR(sReg, 0xff0); // Keep Sticky and D/I flags + AND32ItoR(sReg, 0xffffff00); // Keep Sticky and D/I flags } //-------------------------Check for Signed flags------------------------------ @@ -54,25 +53,22 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation - if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) SHL32ItoR(mReg, 4 + ADD_XYZW); - if (sFLAG.doFlag) OR32ItoR(sReg, 0x82); // SS, S flags - if (sFLAG.doFlag && _XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - if (sFLAG.doFlag) x86SetJ8(pjmp); + SHL32ItoR(mReg, 4 + ADD_XYZW); //-------------------------Check for Zero flags------------------------------ AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation - if (sFLAG.doFlag) pjmp = JZ8(0); // Skip if none are - if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); OR32RtoR(mReg, gprT2); } - if (sFLAG.doFlag) { OR32ItoR(sReg, 0x41); } // ZS, Z flags - if (sFLAG.doFlag) x86SetJ8(pjmp); + if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); } + OR32RtoR(mReg, gprT2); //-------------------------Write back flags------------------------------ - if (sFLAG.doFlag && _XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag + if (sFLAG.doFlag) { + OR32RtoR (sReg, mReg); + SHL32ItoR(mReg, 8); + OR32RtoR (sReg, mReg); + } } //------------------------------------------------------------------ From ffda8ead30584973a6a4cbd406960d1f4de675b5 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 3 Jun 2009 12:09:04 +0000 Subject: [PATCH 172/296] GSdx: mfc-free git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1310 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPU.cpp | 9 - plugins/GSdx/GPURenderer.cpp | 117 +++++++++- plugins/GSdx/GPURenderer.h | 244 +++++--------------- plugins/GSdx/GPURendererSW.cpp | 133 +++++------ plugins/GSdx/GPURendererSW.h | 12 +- plugins/GSdx/GPUSettingsDlg.cpp | 343 +++++++++++---------------- plugins/GSdx/GPUSettingsDlg.h | 43 +--- plugins/GSdx/GPUState.cpp | 2 +- plugins/GSdx/GS.cpp | 15 +- plugins/GSdx/GSCapture.cpp | 12 +- plugins/GSdx/GSCaptureDlg.cpp | 286 ++++++++++------------- plugins/GSdx/GSCaptureDlg.h | 38 +-- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSDevice.cpp | 14 +- plugins/GSdx/GSDevice.h | 7 +- plugins/GSdx/GSDevice10.cpp | 13 +- plugins/GSdx/GSDevice10.h | 2 +- plugins/GSdx/GSDevice7.cpp | 12 +- plugins/GSdx/GSDevice7.h | 2 +- plugins/GSdx/GSDevice9.cpp | 32 ++- plugins/GSdx/GSDevice9.h | 2 +- plugins/GSdx/GSDeviceNull.cpp | 4 +- plugins/GSdx/GSDeviceNull.h | 2 +- plugins/GSdx/GSDialog.cpp | 170 ++++++++++++++ plugins/GSdx/GSDialog.h | 54 +++++ plugins/GSdx/GSLocalMemory.cpp | 7 +- plugins/GSdx/GSRenderer.cpp | 65 +++--- plugins/GSdx/GSRenderer.h | 5 +- plugins/GSdx/GSRendererHW.h | 4 +- plugins/GSdx/GSRendererSW.h | 2 - plugins/GSdx/GSSetting.h | 22 +- plugins/GSdx/GSSettingsDlg.cpp | 372 +++++++++++------------------- plugins/GSdx/GSSettingsDlg.h | 53 +---- plugins/GSdx/GSState.cpp | 2 +- plugins/GSdx/GSTexture9.cpp | 5 +- plugins/GSdx/GSTextureCache10.cpp | 4 +- plugins/GSdx/GSTextureCache9.cpp | 4 +- plugins/GSdx/GSUtil.cpp | 10 +- plugins/GSdx/GSWnd.cpp | 138 +++++++++-- plugins/GSdx/GSWnd.h | 21 +- plugins/GSdx/GSdx.cpp | 96 +++----- plugins/GSdx/GSdx.h | 8 +- plugins/GSdx/GSdx.rc | 58 ++--- plugins/GSdx/GSdx_vs2008.vcproj | 24 +- plugins/GSdx/resource.h | 44 ++-- plugins/GSdx/stdafx.h | 61 ++--- 46 files changed, 1234 insertions(+), 1340 deletions(-) create mode 100644 plugins/GSdx/GSDialog.cpp create mode 100644 plugins/GSdx/GSDialog.h diff --git a/plugins/GSdx/GPU.cpp b/plugins/GSdx/GPU.cpp index d4f1501aa7..44ef8df348 100644 --- a/plugins/GSdx/GPU.cpp +++ b/plugins/GSdx/GPU.cpp @@ -87,8 +87,6 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) #ifdef _WINDOWS - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - s_hr = ::CoInitializeEx(NULL, COINIT_MULTITHREADED); if(!GSUtil::CheckDirectX()) @@ -126,11 +124,6 @@ EXPORT_C_(int32) GPUopen(HWND hWnd) EXPORT_C_(int32) GPUconfigure() { - -#ifdef _WINDOWS - - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - GPUSettingsDlg dlg; if(IDOK == dlg.DoModal()) @@ -139,8 +132,6 @@ EXPORT_C_(int32) GPUconfigure() GPUinit(); } -#endif - return 0; } diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index 5d04451f46..954732b2b1 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -49,23 +49,136 @@ GPURenderer::~GPURenderer() bool GPURenderer::Create(HWND hWnd) { + // TODO: move subclassing inside GSWnd::Attach + m_hWnd = hWnd; m_wndproc = (WNDPROC)GetWindowLongPtr(hWnd, GWLP_WNDPROC); SetWindowLongPtr(hWnd, GWLP_WNDPROC, (LONG_PTR)WndProc); + if(!m_wnd.Attach(m_hWnd)) + { + return false; + } + m_wnd2gpu[hWnd] = this; DWORD style = GetWindowLong(hWnd, GWL_STYLE); style |= WS_OVERLAPPEDWINDOW; SetWindowLong(hWnd, GWL_STYLE, style); - UpdateWindow(hWnd); - ShowWindow(hWnd, SW_SHOWNORMAL); + m_wnd.Show(); + + if(!m_dev->Create(&m_wnd, m_vsync)) + { + return false; + } + + Reset(); return true; } +bool GPURenderer::Merge() +{ + GSTexture* st[2] = {GetOutput(), NULL}; + + if(!st[0]) + { + return false; + } + + GSVector2i s = st[0]->GetSize(); + + GSVector4 sr[2]; + GSVector4 dr[2]; + + sr[0] = GSVector4(0, 0, 1, 1); + dr[0] = GSVector4(0, 0, s.x, s.y); + + m_dev->Merge(st, sr, dr, s, 1, 1, GSVector4(0, 0, 0, 1)); + + return true; +} + +void GPURenderer::VSync() +{ + GSPerfMonAutoTimer pmat(m_perfmon); + + m_perfmon.Put(GSPerfMon::Frame); + + // m_env.STATUS.LCF = ~m_env.STATUS.LCF; // ? + + if(!IsWindow(m_hWnd)) return; + + Flush(); + + if(!Merge()) return; + + // osd + + if((m_perfmon.GetFrame() & 0x1f) == 0) + { + m_perfmon.Update(); + + double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); + + GSVector4i r = m_env.GetDisplayRect(); + + int w = r.width() << m_scale.x; + int h = r.height() << m_scale.y; + + string s = format( + "%I64d | %d x %d | %.2f fps (%d%%) | %d/%d | %d%% CPU | %.2f | %.2f", + m_perfmon.GetFrame(), w, h, fps, (int)(100.0 * fps / m_env.GetFPS()), + (int)m_perfmon.Get(GSPerfMon::Prim), + (int)m_perfmon.Get(GSPerfMon::Draw), + m_perfmon.CPU(), + m_perfmon.Get(GSPerfMon::Swizzle) / 1024, + m_perfmon.Get(GSPerfMon::Unswizzle) / 1024 + ); + + double fillrate = m_perfmon.Get(GSPerfMon::Fillrate); + + if(fillrate > 0) + { + s = format("%s | %.2f mpps", s.c_str(), fps * fillrate / (1024 * 1024)); + } + + SetWindowText(m_hWnd, s.c_str()); + } + + if(m_dev->IsLost()) + { + ResetDevice(); + } + + GSVector4i r; + + GetClientRect(m_hWnd, r); + + m_dev->Present(r.fit(m_aspectratio), 0); +} + +bool GPURenderer::MakeSnapshot(const string& path) +{ + time_t t = time(NULL); + + char buff[16]; + + if(!strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) + { + return false; + } + + if(GSTexture* t = m_dev->GetCurrent()) + { + return t->Save(format("%s_%s.bmp", path.c_str(), buff)); + } + + return false; +} + LRESULT CALLBACK GPURenderer::WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) { map::iterator i = m_wnd2gpu.find(hWnd); diff --git a/plugins/GSdx/GPURenderer.h b/plugins/GSdx/GPURenderer.h index 852e3be0f6..45940ce8b9 100644 --- a/plugins/GSdx/GPURenderer.h +++ b/plugins/GSdx/GPURenderer.h @@ -27,31 +27,35 @@ class GPURenderer : public GPUState { + bool Merge(); + protected: + GSDevice* m_dev; int m_filter; int m_dither; int m_aspectratio; bool m_vsync; GSVector2i m_scale; + virtual void ResetDevice() {} + virtual GSTexture* GetOutput() = 0; + HWND m_hWnd; WNDPROC m_wndproc; static map m_wnd2gpu; + GSWnd m_wnd; static LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam); LRESULT OnMessage(UINT message, WPARAM wParam, LPARAM lParam); -public: - GSDevice* m_dev; - public: GPURenderer(GSDevice* dev); virtual ~GPURenderer(); virtual bool Create(HWND hWnd); - virtual void VSync() = 0; - virtual bool MakeSnapshot(const string& path) = 0; + virtual void VSync(); + virtual bool MakeSnapshot(const string& path); }; template @@ -71,66 +75,6 @@ protected: __super::Reset(); } - void VertexKick() - { - if(m_vl.GetCount() < (int)m_env.PRIM.VTX) - { - return; - } - - if(m_count > m_maxcount) - { - m_maxcount = max(10000, m_maxcount * 3/2); - m_vertices = (Vertex*)_aligned_realloc(m_vertices, sizeof(Vertex) * m_maxcount, 16); - m_maxcount -= 100; - } - - Vertex* v = &m_vertices[m_count]; - - int count = 0; - - switch(m_env.PRIM.TYPE) - { - case GPU_POLYGON: - m_vl.GetAt(0, v[0]); - m_vl.GetAt(1, v[1]); - m_vl.GetAt(2, v[2]); - m_vl.RemoveAll(); - count = 3; - break; - case GPU_LINE: - m_vl.GetAt(0, v[0]); - m_vl.GetAt(1, v[1]); - m_vl.RemoveAll(); - count = 2; - break; - case GPU_SPRITE: - m_vl.GetAt(0, v[0]); - m_vl.GetAt(1, v[1]); - m_vl.RemoveAll(); - count = 2; - break; - default: - ASSERT(0); - m_vl.RemoveAll(); - count = 0; - break; - } - - (this->*m_fpDrawingKickHandlers[m_env.PRIM.TYPE])(v, count); - - m_count += count; - } - - typedef void (GPURendererT::*DrawingKickHandler)(Vertex* v, int& count); - - DrawingKickHandler m_fpDrawingKickHandlers[4]; - - void DrawingKickNull(Vertex* v, int& count) - { - ASSERT(0); - } - void ResetPrim() { m_vl.RemoveAll(); @@ -164,149 +108,71 @@ protected: } } - virtual void ResetDevice() {} - virtual void Draw() = 0; - virtual GSTexture* GetOutput() = 0; - - bool Merge() + void GrowVertexBuffer() { - GSTexture* st[2] = {GetOutput(), NULL}; + m_maxcount = max(10000, m_maxcount * 3/2); + m_vertices = (Vertex*)_aligned_realloc(m_vertices, sizeof(Vertex) * m_maxcount, 16); + m_maxcount -= 100; + } - if(!st[0]) + __forceinline Vertex* DrawingKick(int& count) + { + count = (int)m_env.PRIM.VTX; + + if(m_vl.GetCount() < count) { - return false; + return NULL; } - GSVector2i s = st[0]->GetSize(); + if(m_count >= m_maxcount) + { + GrowVertexBuffer(); + } - GSVector4 sr[2]; - GSVector4 dr[2]; + Vertex* v = &m_vertices[m_count]; - sr[0] = GSVector4(0, 0, 1, 1); - dr[0] = GSVector4(0, 0, s.x, s.y); + switch(m_env.PRIM.TYPE) + { + case GPU_POLYGON: + m_vl.GetAt(0, v[0]); + m_vl.GetAt(1, v[1]); + m_vl.GetAt(2, v[2]); + m_vl.RemoveAll(); + break; + case GPU_LINE: + m_vl.GetAt(0, v[0]); + m_vl.GetAt(1, v[1]); + m_vl.RemoveAll(); + break; + case GPU_SPRITE: + m_vl.GetAt(0, v[0]); + m_vl.GetAt(1, v[1]); + m_vl.RemoveAll(); + break; + default: + ASSERT(0); + m_vl.RemoveAll(); + return NULL; + } - m_dev->Merge(st, sr, dr, s, 1, 1, GSVector4(0, 0, 0, 1)); - - return true; + return v; } + virtual void VertexKick() = 0; + + virtual void Draw() = 0; + public: GPURendererT(GSDevice* dev) : GPURenderer(dev) , m_count(0) - , m_maxcount(10000) + , m_maxcount(0) + , m_vertices(NULL) { - m_vertices = (Vertex*)_aligned_malloc(sizeof(Vertex) * m_maxcount, 16); - m_maxcount -= 100; - - for(int i = 0; i < countof(m_fpDrawingKickHandlers); i++) - { - m_fpDrawingKickHandlers[i] = &GPURendererT::DrawingKickNull; - } } virtual ~GPURendererT() { if(m_vertices) _aligned_free(m_vertices); } - - virtual bool Create(HWND hWnd) - { - if(!__super::Create(hWnd)) - { - return false; - } - - if(!m_dev->Create(hWnd, m_vsync)) - { - return false; - } - - Reset(); - - return true; - } - - virtual void VSync() - { - GSPerfMonAutoTimer pmat(m_perfmon); - - m_perfmon.Put(GSPerfMon::Frame); - - // m_env.STATUS.LCF = ~m_env.STATUS.LCF; // ? - - if(!IsWindow(m_hWnd)) return; - - Flush(); - - if(!Merge()) return; - - // osd - - static uint64 s_frame = 0; - static string s_stats; - - if(m_perfmon.GetFrame() - s_frame >= 30) - { - m_perfmon.Update(); - - s_frame = m_perfmon.GetFrame(); - - double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); - - GSVector4i r = m_env.GetDisplayRect(); - - int w = r.width() << m_scale.x; - int h = r.height() << m_scale.y; - - s_stats = format( - "%I64d | %d x %d | %.2f fps (%d%%) | %d/%d | %d%% CPU | %.2f | %.2f", - m_perfmon.GetFrame(), w, h, fps, (int)(100.0 * fps / m_env.GetFPS()), - (int)m_perfmon.Get(GSPerfMon::Prim), - (int)m_perfmon.Get(GSPerfMon::Draw), - m_perfmon.CPU(), - m_perfmon.Get(GSPerfMon::Swizzle) / 1024, - m_perfmon.Get(GSPerfMon::Unswizzle) / 1024 - ); - - double fillrate = m_perfmon.Get(GSPerfMon::Fillrate); - - if(fillrate > 0) - { - s_stats = format("%s | %.2f mpps", s_stats.c_str(), fps * fillrate / (1024 * 1024)); - } - - SetWindowText(m_hWnd, s_stats.c_str()); - } - - if(m_dev->IsLost()) - { - ResetDevice(); - } - - GSVector4i r; - - GetClientRect(m_hWnd, r); - - m_dev->Present(r.fit(m_aspectratio), 0); - } - - virtual bool MakeSnapshot(const string& path) - { - time_t t = time(NULL); - - char buff[16]; - - if(!strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) - { - return false; - } - - if(GSTexture* t = m_dev->GetCurrent()) - { - return t->Save(format("%s_%s.bmp", path.c_str(), buff)); - } - - return false; - } }; diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index 4fbec76f77..c71f72fdcb 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -28,10 +28,6 @@ GPURendererSW::GPURendererSW(GSDevice* dev) , m_texture(NULL) { m_rl.Create(this, theApp.GetConfig("swthreads", 1)); - - m_fpDrawingKickHandlers[GPU_POLYGON] = (DrawingKickHandler)&GPURendererSW::DrawingKickTriangle; - m_fpDrawingKickHandlers[GPU_LINE] = (DrawingKickHandler)&GPURendererSW::DrawingKickLine; - m_fpDrawingKickHandlers[GPU_SPRITE] = (DrawingKickHandler)&GPURendererSW::DrawingKickSprite; } GPURendererSW::~GPURendererSW() @@ -45,52 +41,46 @@ void GPURendererSW::ResetDevice() m_texture = NULL; } -GSVector4i GPURendererSW::GetScissor() +GSTexture* GPURendererSW::GetOutput() { - GSVector4i r; + GSVector4i r = m_env.GetDisplayRect(); - r.left = (int)m_env.DRAREATL.X << m_scale.x; - r.top = (int)m_env.DRAREATL.Y << m_scale.y; - r.right = min((int)(m_env.DRAREABR.X + 1) << m_scale.x, m_mem.GetWidth()); - r.bottom = min((int)(m_env.DRAREABR.Y + 1) << m_scale.y, m_mem.GetHeight()); + r.left <<= m_scale.x; + r.top <<= m_scale.y; + r.right <<= m_scale.x; + r.bottom <<= m_scale.y; - return r; -} - -void GPURendererSW::VertexKick() -{ - GSVertexSW& v = m_vl.AddTail(); - - // TODO: x/y + off.x/y should wrap around at +/-1024 - - int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.x; - int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.y; - - int s = m_v.UV.X; - int t = m_v.UV.Y; - - GSVector4 pt(x, y, s, t); - - v.p = pt.xyxy(GSVector4::zero()); - v.t = (pt.zwzw(GSVector4::zero()) + GSVector4(0.125f)) * 256.0f; - v.c = GSVector4(m_v.RGB.u32) * 128.0f; - - __super::VertexKick(); -} - -void GPURendererSW::DrawingKickTriangle(GSVertexSW* v, int& count) -{ // TODO -} + static uint32* buff = (uint32*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(uint32), 16); -void GPURendererSW::DrawingKickLine(GSVertexSW* v, int& count) -{ - // TODO -} + m_mem.ReadFrame32(r, buff, !!m_env.STATUS.ISRGB24); -void GPURendererSW::DrawingKickSprite(GSVertexSW* v, int& count) -{ - // TODO + int w = r.width(); + int h = r.height(); + + if(m_texture) + { + if(m_texture->GetWidth() != w || m_texture->GetHeight() != h) + { + delete m_texture; + + m_texture = NULL; + } + } + + if(!m_texture) + { + m_texture = m_dev->CreateTexture(w, h); + + if(!m_texture) + { + return NULL; + } + } + + m_texture->Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); + + return m_texture; } void GPURendererSW::Draw() @@ -137,11 +127,15 @@ void GPURendererSW::Draw() GSRasterizerData data; - data.scissor = GetScissor(); data.vertices = m_vertices; data.count = m_count; data.param = &p; + data.scissor.left = (int)m_env.DRAREATL.X << m_scale.x; + data.scissor.top = (int)m_env.DRAREATL.Y << m_scale.y; + data.scissor.right = min((int)(m_env.DRAREABR.X + 1) << m_scale.x, m_mem.GetWidth()); + data.scissor.bottom = min((int)(m_env.DRAREABR.Y + 1) << m_scale.y, m_mem.GetHeight()); + switch(env.PRIM.TYPE) { case GPU_POLYGON: data.primclass = GS_TRIANGLE_CLASS; break; @@ -185,45 +179,32 @@ void GPURendererSW::Draw() } } -GSTexture* GPURendererSW::GetOutput() +void GPURendererSW::VertexKick() { - GSVector4i r = m_env.GetDisplayRect(); + GSVertexSW& dst = m_vl.AddTail(); - r.left <<= m_scale.x; - r.top <<= m_scale.y; - r.right <<= m_scale.x; - r.bottom <<= m_scale.y; + // TODO: x/y + off.x/y should wrap around at +/-1024 - // TODO - static uint32* buff = (uint32*)_aligned_malloc(m_mem.GetWidth() * m_mem.GetHeight() * sizeof(uint32), 16); + int x = (int)(m_v.XY.X + m_env.DROFF.X) << m_scale.x; + int y = (int)(m_v.XY.Y + m_env.DROFF.Y) << m_scale.y; - m_mem.ReadFrame32(r, buff, !!m_env.STATUS.ISRGB24); + int s = m_v.UV.X; + int t = m_v.UV.Y; - int w = r.width(); - int h = r.height(); + GSVector4 pt(x, y, s, t); - if(m_texture) + dst.p = pt.xyxy(GSVector4::zero()); + dst.t = (pt.zwzw(GSVector4::zero()) + GSVector4(0.125f)) * 256.0f; + // dst.c = GSVector4(m_v.RGB.u32) * 128.0f; + dst.c = GSVector4(GSVector4i::load((int)m_v.RGB.u32).u8to32() << 7); + + int count = 0; + + if(GSVertexSW* v = DrawingKick(count)) { - if(m_texture->GetWidth() != w || m_texture->GetHeight() != h) - { - delete m_texture; + // TODO - m_texture = NULL; - } + m_count += count; } - - if(!m_texture) - { - m_texture = m_dev->CreateTexture(w, h); - - if(!m_texture) - { - return NULL; - } - } - - m_texture->Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); - - return m_texture; } diff --git a/plugins/GSdx/GPURendererSW.h b/plugins/GSdx/GPURendererSW.h index f797bc3080..aef5a140b1 100644 --- a/plugins/GSdx/GPURendererSW.h +++ b/plugins/GSdx/GPURendererSW.h @@ -31,17 +31,9 @@ protected: GSTexture* m_texture; void ResetDevice(); - - GSVector4i GetScissor(); - - void VertexKick(); - void DrawingKickTriangle(GSVertexSW* v, int& count); - void DrawingKickLine(GSVertexSW* v, int& count); - void DrawingKickSprite(GSVertexSW* v, int& count); - - void Draw(); - GSTexture* GetOutput(); + void VertexKick(); + void Draw(); public: GPURendererSW(GSDevice* dev); diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 96b9e5c28b..7ea1c5e289 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -23,112 +23,64 @@ #include "GSdx.h" #include "GSUtil.h" #include "GPUSettingsDlg.h" -#include -#include +#include "resource.h" GSSetting GPUSettingsDlg::g_renderers[] = { - {0, _T("Direct3D7 (Software)"), NULL}, - {1, _T("Direct3D9 (Software)"), NULL}, - {2, _T("Direct3D10 (Software)"), NULL}, -// {3, _T("Null (Null)"), NULL}, + {0, "Direct3D7 (Software)", NULL}, + {1, "Direct3D9 (Software)", NULL}, + {2, "Direct3D10 (Software)", NULL}, +// {3, "Null (Null)", NULL}, }; GSSetting GPUSettingsDlg::g_psversion[] = { - {D3DPS_VERSION(3, 0), _T("Pixel Shader 3.0"), NULL}, - {D3DPS_VERSION(2, 0), _T("Pixel Shader 2.0"), NULL}, - //{D3DPS_VERSION(1, 4), _T("Pixel Shader 1.4"), NULL}, - //{D3DPS_VERSION(1, 1), _T("Pixel Shader 1.1"), NULL}, - //{D3DPS_VERSION(0, 0), _T("Fixed Pipeline (bogus)"), NULL}, + {D3DPS_VERSION(3, 0), "Pixel Shader 3.0", NULL}, + {D3DPS_VERSION(2, 0), "Pixel Shader 2.0", NULL}, + //{D3DPS_VERSION(1, 4), "Pixel Shader 1.4", NULL}, + //{D3DPS_VERSION(1, 1), "Pixel Shader 1.1", NULL}, + //{D3DPS_VERSION(0, 0), "Fixed Pipeline (bogus)", NULL}, }; GSSetting GPUSettingsDlg::g_filter[] = { - {0, _T("Nearest"), NULL}, - {1, _T("Bilinear (polygons only)"), NULL}, - {2, _T("Bilinear"), NULL}, + {0, "Nearest", NULL}, + {1, "Bilinear (polygons only)", NULL}, + {2, "Bilinear", NULL}, }; GSSetting GPUSettingsDlg::g_dithering[] = { - {0, _T("Disabled"), NULL}, - {1, _T("Auto"), NULL}, + {0, "Disabled", NULL}, + {1, "Auto", NULL}, }; GSSetting GPUSettingsDlg::g_aspectratio[] = { - {0, _T("Stretch"), NULL}, - {1, _T("4:3"), NULL}, - {2, _T("16:9"), NULL}, + {0, "Stretch", NULL}, + {1, "4:3", NULL}, + {2, "16:9", NULL}, }; -GSSetting GPUSettingsDlg::g_internalresolution[] = +GSSetting GPUSettingsDlg::g_scale[] = { - {0 | (0 << 2), _T("H x 1 - V x 1"), NULL}, - {1 | (0 << 2), _T("H x 2 - V x 1"), NULL}, - {0 | (1 << 2), _T("H x 1 - V x 2"), NULL}, - {1 | (1 << 2), _T("H x 2 - V x 2"), NULL}, - {2 | (1 << 2), _T("H x 4 - V x 2"), NULL}, - {1 | (2 << 2), _T("H x 2 - V x 4"), NULL}, - {2 | (2 << 2), _T("H x 4 - V x 4"), NULL}, + {0 | (0 << 2), "H x 1 - V x 1", NULL}, + {1 | (0 << 2), "H x 2 - V x 1", NULL}, + {0 | (1 << 2), "H x 1 - V x 2", NULL}, + {1 | (1 << 2), "H x 2 - V x 2", NULL}, + {2 | (1 << 2), "H x 4 - V x 2", NULL}, + {1 | (2 << 2), "H x 2 - V x 4", NULL}, + {2 | (2 << 2), "H x 4 - V x 4", NULL}, }; -IMPLEMENT_DYNAMIC(GPUSettingsDlg, CDialog) - -GPUSettingsDlg::GPUSettingsDlg(CWnd* pParent /*=NULL*/) - : CDialog(GPUSettingsDlg::IDD, pParent) -{ - -} - -GPUSettingsDlg::~GPUSettingsDlg() +GPUSettingsDlg::GPUSettingsDlg() + : GSDialog(IDD_GPUCONFIG) { } -LRESULT GPUSettingsDlg::DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam) +void GPUSettingsDlg::OnInit() { - LRESULT ret = __super::DefWindowProc(message, wParam, lParam); - - if(message == WM_INITDIALOG) - { - SendMessage(WM_KICKIDLE); - } - - return ret; -} - -void GPUSettingsDlg::DoDataExchange(CDataExchange* pDX) -{ - __super::DoDataExchange(pDX); - DDX_Control(pDX, IDC_COMBO3, m_resolution); - DDX_Control(pDX, IDC_COMBO1, m_renderer); - DDX_Control(pDX, IDC_COMBO4, m_psversion); - DDX_Control(pDX, IDC_COMBO2, m_filter); - DDX_Control(pDX, IDC_COMBO5, m_dithering); - DDX_Control(pDX, IDC_COMBO6, m_aspectratio); - DDX_Control(pDX, IDC_COMBO7, m_internalresolution); - DDX_Control(pDX, IDC_SPIN3, m_swthreads); - DDX_Control(pDX, IDC_EDIT3, m_swthreadsedit); -} - -BEGIN_MESSAGE_MAP(GPUSettingsDlg, CDialog) - ON_MESSAGE_VOID(WM_KICKIDLE, OnKickIdle) - ON_UPDATE_COMMAND_UI(IDC_COMBO4, OnUpdateD3D9Options) - ON_UPDATE_COMMAND_UI(IDC_COMBO7, OnUpdateSWOptions) - ON_UPDATE_COMMAND_UI(IDC_SPIN3, OnUpdateSWOptions) - ON_UPDATE_COMMAND_UI(IDC_EDIT3, OnUpdateSWOptions) - ON_CBN_SELCHANGE(IDC_COMBO1, &GPUSettingsDlg::OnCbnSelchangeCombo1) -END_MESSAGE_MAP() - -void GPUSettingsDlg::OnKickIdle() -{ - UpdateDialogControls(this, false); -} - -BOOL GPUSettingsDlg::OnInitDialog() -{ - __super::OnInitDialog(); + __super::OnInit(); D3DCAPS9 caps; memset(&caps, 0, sizeof(caps)); @@ -136,163 +88,128 @@ BOOL GPUSettingsDlg::OnInitDialog() m_modes.clear(); - // windowed - { D3DDISPLAYMODE mode; memset(&mode, 0, sizeof(mode)); m_modes.push_back(mode); - int iItem = m_resolution.AddString(_T("Windowed")); - m_resolution.SetItemDataPtr(iItem, &m_modes.back()); - m_resolution.SetCurSel(iItem); - } + HWND hWnd = GetDlgItem(m_hWnd, IDC_RESOLUTION); - // fullscreen + ComboBoxAppend(hWnd, "Windowed", (LPARAM)&m_modes.back(), true); - if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) - { - uint32 ModeWidth = theApp.GetConfig("ModeWidth", 0); - uint32 ModeHeight = theApp.GetConfig("ModeHeight", 0); - uint32 ModeRefreshRate = theApp.GetConfig("ModeRefreshRate", 0); - - uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); - - for(uint32 i = 0; i < nModes; i++) + if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - D3DDISPLAYMODE mode; + uint32 w = theApp.GetConfig("ModeWidth", 0); + uint32 h = theApp.GetConfig("ModeHeight", 0); + uint32 hz = theApp.GetConfig("ModeRefreshRate", 0); - if(S_OK == d3d->EnumAdapterModes(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8, i, &mode)) + uint32 n = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); + + for(uint32 i = 0; i < n; i++) { - CString str; - str.Format(_T("%dx%d %dHz"), mode.Width, mode.Height, mode.RefreshRate); - int iItem = m_resolution.AddString(str); - - m_modes.push_back(mode); - m_resolution.SetItemDataPtr(iItem, &m_modes.back()); - - if(ModeWidth == mode.Width && ModeHeight == mode.Height && ModeRefreshRate == mode.RefreshRate) + if(S_OK == d3d->EnumAdapterModes(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8, i, &mode)) { - m_resolution.SetCurSel(iItem); + m_modes.push_back(mode); + + string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); + + ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } + + d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); + } + } + + { + bool isdx10avail = GSUtil::IsDirect3D10Available(); + + vector renderers; + + for(size_t i = 0; i < countof(g_renderers); i++) + { + if(i >= 3 && i <= 5 && !isdx10avail) continue; + + renderers.push_back(g_renderers[i]); } - d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); + HWND hWnd = GetDlgItem(m_hWnd, IDC_RENDERER); + + ComboBoxInit(hWnd, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); + + OnCommand(hWnd, IDC_RENDERER, CBN_SELCHANGE); } - bool isdx10avail = GSUtil::IsDirect3D10Available(); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_FILTER), g_filter, countof(g_filter), theApp.GetConfig("filter", 0)); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_DITHERING), g_dithering, countof(g_dithering), theApp.GetConfig("dithering", 1)); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_SCALE), g_scale, countof(g_scale), theApp.GetConfig("scale_x", 0) | (theApp.GetConfig("scale_y", 0) << 2)); - vector renderers; - - for(size_t i = 0; i < countof(g_renderers); i++) - { - if(i == 2 && !isdx10avail) continue; - - renderers.push_back(g_renderers[i]); - } - - GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, theApp.GetConfig("Renderer", 1)); - GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); - GSSetting::InitComboBox(g_filter, countof(g_filter), m_filter, theApp.GetConfig("filter", 0)); - GSSetting::InitComboBox(g_dithering, countof(g_dithering), m_dithering, theApp.GetConfig("dithering", 1)); - GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, theApp.GetConfig("AspectRatio", 1)); - GSSetting::InitComboBox(g_internalresolution, countof(g_internalresolution), m_internalresolution, theApp.GetConfig("scale_x", 0) | (theApp.GetConfig("scale_y", 0) << 2)); - - - OnCbnSelchangeCombo1(); - - // - - m_swthreads.SetRange(1, 16); - m_swthreads.SetPos(theApp.GetConfig("swthreads", 1)); - - // - - UpdateData(FALSE); - - return TRUE; // return TRUE unless you set the focus to a control - // EXCEPTION: OCX Property Pages should return FALSE + SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETRANGE, 0, MAKELPARAM(16, 1)); + SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("swthreads", 1), 0)); } -void GPUSettingsDlg::OnOK() +bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { - UpdateData(); - - if(m_resolution.GetCurSel() >= 0) + if(id == IDC_RENDERER && code == CBN_SELCHANGE) { - D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); + int item = (int)SendMessage(hWnd, CB_GETCURSEL, 0, 0); - theApp.SetConfig("ModeWidth", (int)mode.Width); - theApp.SetConfig("ModeHeight", (int)mode.Height); - theApp.SetConfig("ModeRefreshRate", (int)mode.RefreshRate); + if(item >= 0) + { + int i = (int)SendMessage(hWnd, CB_GETITEMDATA, item, 0); + + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), i == 1 ? SW_SHOW : SW_HIDE); + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), i == 2 ? SW_SHOW : SW_HIDE); + } + } + else if(id == IDOK) + { + INT_PTR data; + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RESOLUTION), data)) + { + const D3DDISPLAYMODE* mode = (D3DDISPLAYMODE*)data; + + theApp.SetConfig("ModeWidth", (int)mode->Width); + theApp.SetConfig("ModeHeight", (int)mode->Height); + theApp.SetConfig("ModeRefreshRate", (int)mode->RefreshRate); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), data)) + { + theApp.SetConfig("Renderer", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SHADER), data)) + { + theApp.SetConfig("PixelShaderVersion2", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_FILTER), data)) + { + theApp.SetConfig("filter", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_DITHERING), data)) + { + theApp.SetConfig("dithering", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), data)) + { + theApp.SetConfig("AspectRatio", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SCALE), data)) + { + theApp.SetConfig("scale_x", data & 3); + theApp.SetConfig("scale_y", (data >> 2) & 3); + } + + theApp.SetConfig("swthreads", (int)SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_GETPOS, 0, 0)); } - if(m_renderer.GetCurSel() >= 0) - { - theApp.SetConfig("Renderer", (int)m_renderer.GetItemData(m_renderer.GetCurSel())); - } - - if(m_psversion.GetCurSel() >= 0) - { - theApp.SetConfig("PixelShaderVersion2", (int)m_psversion.GetItemData(m_psversion.GetCurSel())); - } - - if(m_filter.GetCurSel() >= 0) - { - theApp.SetConfig("filter", (int)m_filter.GetItemData(m_filter.GetCurSel())); - } - - if(m_dithering.GetCurSel() >= 0) - { - theApp.SetConfig("dithering", (int)m_dithering.GetItemData(m_dithering.GetCurSel())); - } - - if(m_aspectratio.GetCurSel() >= 0) - { - theApp.SetConfig("AspectRatio", (int)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); - } - - if(m_internalresolution.GetCurSel() >= 0) - { - int value = (int)m_internalresolution.GetItemData(m_internalresolution.GetCurSel()); - - theApp.SetConfig("scale_x", value & 3); - theApp.SetConfig("scale_y", (value >> 2) & 3); - } - - theApp.SetConfig("swthreads", m_swthreads.GetPos()); - - __super::OnOK(); -} - -void GPUSettingsDlg::OnUpdateResolution(CCmdUI* pCmdUI) -{ - UpdateData(); - - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(i == 1); -} - -void GPUSettingsDlg::OnUpdateD3D9Options(CCmdUI* pCmdUI) -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(i == 1); -} - -void GPUSettingsDlg::OnUpdateSWOptions(CCmdUI* pCmdUI) -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(i >= 0 && i <= 2); -} - -void GPUSettingsDlg::OnCbnSelchangeCombo1() -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - GetDlgItem(IDC_LOGO9)->ShowWindow(i == 1 ? SW_SHOW : SW_HIDE); - GetDlgItem(IDC_LOGO10)->ShowWindow(i == 2 ? SW_SHOW : SW_HIDE); -} + return __super::OnCommand(hWnd, id, code); +} \ No newline at end of file diff --git a/plugins/GSdx/GPUSettingsDlg.h b/plugins/GSdx/GPUSettingsDlg.h index 04828cf892..517ffb6443 100644 --- a/plugins/GSdx/GPUSettingsDlg.h +++ b/plugins/GSdx/GPUSettingsDlg.h @@ -21,51 +21,24 @@ #pragma once +#include "GSDialog.h" #include "GSSetting.h" -#include "resource.h" -class GPUSettingsDlg : public CDialog +class GPUSettingsDlg : public GSDialog { - DECLARE_DYNAMIC(GPUSettingsDlg) - -private: list m_modes; +protected: + void OnInit(); + bool OnCommand(HWND hWnd, UINT id, UINT code); + public: - GPUSettingsDlg(CWnd* pParent = NULL); // standard constructor - virtual ~GPUSettingsDlg(); + GPUSettingsDlg(); static GSSetting g_renderers[]; static GSSetting g_psversion[]; static GSSetting g_filter[]; static GSSetting g_dithering[]; static GSSetting g_aspectratio[]; - static GSSetting g_internalresolution[]; - -// Dialog Data - enum { IDD = IDD_GPUCONFIG }; - CComboBox m_resolution; - CComboBox m_renderer; - CComboBox m_psversion; - CComboBox m_filter; - CComboBox m_dithering; - CComboBox m_aspectratio; - CComboBox m_internalresolution; - CSpinButtonCtrl m_swthreads; - CEdit m_swthreadsedit; - -protected: - virtual LRESULT DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam); - virtual void DoDataExchange(CDataExchange* pDX); // DDX/DDV support - virtual BOOL OnInitDialog(); - virtual void OnOK(); - - DECLARE_MESSAGE_MAP() - -public: - afx_msg void OnKickIdle(); - afx_msg void OnUpdateResolution(CCmdUI* pCmdUI); - afx_msg void OnUpdateD3D9Options(CCmdUI* pCmdUI); - afx_msg void OnUpdateSWOptions(CCmdUI* pCmdUI); - afx_msg void OnCbnSelchangeCombo1(); + static GSSetting g_scale[]; }; diff --git a/plugins/GSdx/GPUState.cpp b/plugins/GSdx/GPUState.cpp index 11e213c373..8fad682df1 100644 --- a/plugins/GSdx/GPUState.cpp +++ b/plugins/GSdx/GPUState.cpp @@ -175,7 +175,7 @@ void GPUState::ReadData(uint8* mem, uint32 size) { // ASSERT(0); - TRACE(_T("WARNING: ReadData\n")); + // printf"WARNING: ReadData\n"); // memset(&mem[remaining], 0, bytes - remaining); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 7b0aefabcc..3cc0cf8481 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -138,7 +138,7 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) s_gs->m_wnd.Show(); - *(HWND*)dsp = s_gs->m_wnd; + *(HWND*)dsp = (HWND)s_gs->m_wnd.GetHandle(); // if(mt) _mm_setcsr(MXCSR); @@ -147,12 +147,6 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) EXPORT_C_(INT32) GSopen(void* dsp, char* title, int mt) { -#ifdef _WINDOWS - - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - -#endif - int renderer; if(mt == 2) @@ -259,10 +253,6 @@ EXPORT_C_(int) GSfreeze(int mode, GSFreezeData* data) EXPORT_C GSconfigure() { -#ifdef _WINDOWS - - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - GSSettingsDlg dlg; if(IDOK == dlg.DoModal()) @@ -270,9 +260,6 @@ EXPORT_C GSconfigure() GSshutdown(); GSinit(); } - -#endif - } EXPORT_C_(INT32) GStest() diff --git a/plugins/GSdx/GSCapture.cpp b/plugins/GSdx/GSCapture.cpp index 7fca7cc0b8..cf42560cc0 100644 --- a/plugins/GSdx/GSCapture.cpp +++ b/plugins/GSdx/GSCapture.cpp @@ -403,8 +403,6 @@ bool GSCapture::BeginCapture(int fps) EndCapture(); - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - GSCaptureDlg dlg; if(IDOK != dlg.DoModal()) return false; @@ -412,6 +410,8 @@ bool GSCapture::BeginCapture(int fps) m_size.x = (dlg.m_width + 7) & ~7; m_size.y = (dlg.m_height + 7) & ~7; + wstring fn(dlg.m_filename.begin(), dlg.m_filename.end()); + // HRESULT hr; @@ -422,7 +422,7 @@ bool GSCapture::BeginCapture(int fps) if(FAILED(hr = m_graph.CoCreateInstance(CLSID_FilterGraph)) || FAILED(hr = cgb.CoCreateInstance(CLSID_CaptureGraphBuilder2)) || FAILED(hr = cgb->SetFiltergraph(m_graph)) - || FAILED(hr = cgb->SetOutputFileName(&MEDIASUBTYPE_Avi, CStringW(dlg.m_filename), &mux, NULL))) + || FAILED(hr = cgb->SetOutputFileName(&MEDIASUBTYPE_Avi, fn.c_str(), &mux, NULL))) { return false; } @@ -445,7 +445,8 @@ bool GSCapture::BeginCapture(int fps) { CFilterInfo fi; pBF->QueryFilterInfo(&fi); - printf("Filter [%p]: %s\n", pBF.p, CStringA(fi.achName)); + wstring s(fi.achName); + printf("Filter [%p]: %s\n", pBF.p, string(s.begin(), s.end()).c_str()); BeginEnumPins(pBF, pEP, pPin) { @@ -454,7 +455,8 @@ bool GSCapture::BeginCapture(int fps) CPinInfo pi; pPin->QueryPinInfo(&pi); - printf("- Pin [%p - %p]: %s (%s)\n", pPin.p, pPinTo.p, CStringA(pi.achName), pi.dir ? "out" : "in"); + wstring s(pi.achName); + printf("- Pin [%p - %p]: %s (%s)\n", pPin.p, pPinTo.p, string(s.begin(), s.end()).c_str(), pi.dir ? "out" : "in"); BeginEnumMediaTypes(pPin, pEMT, pmt) { diff --git a/plugins/GSdx/GSCaptureDlg.cpp b/plugins/GSdx/GSCaptureDlg.cpp index 9e3ae7c2e3..1d82fde7d8 100644 --- a/plugins/GSdx/GSCaptureDlg.cpp +++ b/plugins/GSdx/GSCaptureDlg.cpp @@ -20,230 +20,182 @@ */ #include "stdafx.h" -#include #include "GSdx.h" #include "GSCaptureDlg.h" -// GSCaptureDlg dialog - -IMPLEMENT_DYNAMIC(GSCaptureDlg, CDialog) -GSCaptureDlg::GSCaptureDlg(CWnd* pParent /*=NULL*/) - : CDialog(GSCaptureDlg::IDD, pParent) +GSCaptureDlg::GSCaptureDlg() + : GSDialog(IDD_CAPTURE) { m_width = theApp.GetConfig("CaptureWidth", 640); m_height = theApp.GetConfig("CaptureHeight", 480); - m_filename = theApp.GetConfig("CaptureFileName", "").c_str(); -} - -GSCaptureDlg::~GSCaptureDlg() -{ + m_filename = theApp.GetConfig("CaptureFileName", ""); } int GSCaptureDlg::GetSelCodec(Codec& c) { - int iSel = m_codeclist.GetCurSel(); + INT_PTR data = 0; - if(iSel < 0) return 0; - - Codec* codec = (Codec*)m_codeclist.GetItemDataPtr(iSel); - - if(codec == NULL) return 2; - - c = *codec; - - if(!c.filter) + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_CODECS), data)) { - c.moniker->BindToObject(NULL, NULL, __uuidof(IBaseFilter), (void**)&c.filter); + if(data == 0) return 2; - if(!c.filter) return 0; + c = *(Codec*)data; + + if(!c.filter) + { + c.moniker->BindToObject(NULL, NULL, __uuidof(IBaseFilter), (void**)&c.filter); + + if(!c.filter) return 0; + } + + return 1; } - return 1; + return 0; } -LRESULT GSCaptureDlg::DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam) +void GSCaptureDlg::OnInit() { - LRESULT ret = __super::DefWindowProc(message, wParam, lParam); + __super::OnInit(); - if(message == WM_INITDIALOG) SendMessage(WM_KICKIDLE); - - return(ret); -} - -void GSCaptureDlg::DoDataExchange(CDataExchange* pDX) -{ - __super::DoDataExchange(pDX); - - DDX_Text(pDX, IDC_EDIT1, m_filename); - DDX_Control(pDX, IDC_COMBO1, m_codeclist); - DDX_Text(pDX, IDC_EDIT2, m_width); - DDX_Text(pDX, IDC_EDIT4, m_height); -} - -BOOL GSCaptureDlg::OnInitDialog() -{ - __super::OnInitDialog(); + SetTextAsInt(IDC_WIDTH, m_width); + SetTextAsInt(IDC_HEIGHT, m_height); + SetText(IDC_FILENAME, m_filename.c_str()); m_codecs.clear(); - m_codeclist.ResetContent(); - m_codeclist.SetItemDataPtr(m_codeclist.AddString(_T("Uncompressed")), NULL); + string selected = theApp.GetConfig("CaptureVideoCodecDisplayName", ""); + + HWND hWnd = GetDlgItem(m_hWnd, IDC_CODECS); + + ComboBoxAppend(hWnd, "Uncompressed", 0, true); BeginEnumSysDev(CLSID_VideoCompressorCategory, moniker) { Codec c; + c.moniker = moniker; - LPOLESTR strName = NULL; - if(FAILED(moniker->GetDisplayName(NULL, NULL, &strName))) + wstring prefix; + + LPOLESTR str = NULL; + + if(FAILED(moniker->GetDisplayName(NULL, NULL, &str))) continue; - c.DisplayName = strName; - CoTaskMemFree(strName); + if(wcsstr(str, L"@device:dmo:")) prefix = L"(DMO) "; + else if(wcsstr(str, L"@device:sw:")) prefix = L"(DS) "; + else if(wcsstr(str, L"@device:cm:")) prefix = L"(VfW) "; + + c.DisplayName = str; + + CoTaskMemFree(str); CComPtr pPB; - moniker->BindToStorage(0, 0, IID_IPropertyBag, (void**)&pPB); + + if(FAILED(moniker->BindToStorage(0, 0, IID_IPropertyBag, (void**)&pPB))) + continue; CComVariant var; + if(FAILED(pPB->Read(CComBSTR(_T("FriendlyName")), &var, NULL))) continue; - c.FriendlyName = var.bstrVal; - - CStringW str = CStringW(c.DisplayName).MakeLower(); - CString prefix; - if(str.Find(L"@device:dmo:") == 0) prefix = _T("(DMO) "); - else if(str.Find(L"@device:sw:") == 0) prefix = _T("(DS) "); - else if(str.Find(L"@device:cm:") == 0) prefix = _T("(VfW) "); - c.FriendlyName = prefix + c.FriendlyName; + c.FriendlyName = prefix + var.bstrVal; m_codecs.push_back(c); - m_codeclist.SetItemDataPtr(m_codeclist.AddString(c.FriendlyName), &m_codecs.back()); + + string s(c.FriendlyName.begin(), c.FriendlyName.end()); + + ComboBoxAppend(hWnd, s.c_str(), (LPARAM)&m_codecs.back(), s == selected); } EndEnumSysDev +} - // - - CString DisplayNameToFind = theApp.GetConfig("CaptureVideoCodecDisplayName", "").c_str(); - - for(int i = 0; i < m_codeclist.GetCount(); i++) +bool GSCaptureDlg::OnCommand(HWND hWnd, UINT id, UINT code) +{ + if(id == IDC_BROWSE && code == BN_CLICKED) { - CString DisplayName; + char buff[MAX_PATH] = {0}; - Codec* codec = (Codec*)m_codeclist.GetItemDataPtr(i); + OPENFILENAME ofn; - if(codec) + memset(&ofn, 0, sizeof(ofn)); + + ofn.lStructSize = sizeof(ofn); + ofn.hwndOwner = m_hWnd; + ofn.lpstrFile = buff; + ofn.nMaxFile = countof(buff); + ofn.lpstrFilter = "Avi files (*.avi)\0*.avi\0"; + ofn.Flags = OFN_EXPLORER | OFN_ENABLESIZING | OFN_HIDEREADONLY | OFN_OVERWRITEPROMPT | OFN_PATHMUSTEXIST; + + strcpy(ofn.lpstrFile, m_filename.c_str()); + + if(GetSaveFileName(&ofn)) { - DisplayName = codec->DisplayName; + m_filename = ofn.lpstrFile; + + SetText(IDC_FILENAME, m_filename.c_str()); } - if(DisplayName == DisplayNameToFind) - { - m_codeclist.SetCurSel(i); - - break; - } + return true; } - - // - - UpdateData(FALSE); - - return TRUE; // return TRUE unless you set the focus to a control - // EXCEPTION: OCX Property Pages should return FALSE -} - -BEGIN_MESSAGE_MAP(GSCaptureDlg, CDialog) - ON_MESSAGE_VOID(WM_KICKIDLE, OnKickIdle) - ON_BN_CLICKED(IDC_BUTTON1, OnBnClickedButton1) - ON_BN_CLICKED(IDC_BUTTON2, OnBnClickedButton2) - ON_UPDATE_COMMAND_UI(IDC_BUTTON2, OnUpdateButton2) - ON_BN_CLICKED(IDOK, OnBnClickedOk) - ON_UPDATE_COMMAND_UI(IDOK, OnUpdateOK) -END_MESSAGE_MAP() - -// GSCaptureDlg message handlers - -void GSCaptureDlg::OnKickIdle() -{ - UpdateDialogControls(this, false); -} - -void GSCaptureDlg::OnBnClickedButton1() -{ - UpdateData(); - - DWORD flags = OFN_EXPLORER|OFN_ENABLESIZING|OFN_HIDEREADONLY|OFN_OVERWRITEPROMPT|OFN_PATHMUSTEXIST; - - CFileDialog fd(FALSE, _T("avi"), m_filename, flags, _T("Avi files (*.avi)|*.avi||"), this, 0); - - if(fd.DoModal() == IDOK) + else if(id == IDC_CONFIGURE && code == BN_CLICKED) { - m_filename = fd.GetPathName(); + Codec c; - UpdateData(FALSE); - } -} - -void GSCaptureDlg::OnBnClickedButton2() -{ - Codec c; - - if(GetSelCodec(c) != 1) return; - - if(CComQIPtr pSPP = c.filter) - { - CAUUID caGUID; - - memset(&caGUID, 0, sizeof(caGUID)); - - if(SUCCEEDED(pSPP->GetPages(&caGUID))) + if(GetSelCodec(c) == 1) { - IUnknown* lpUnk = NULL; - pSPP.QueryInterface(&lpUnk); - OleCreatePropertyFrame(m_hWnd, 0, 0, CStringW(c.FriendlyName), 1, (IUnknown**)&lpUnk, caGUID.cElems, caGUID.pElems, 0, 0, NULL); - lpUnk->Release(); + if(CComQIPtr pSPP = c.filter) + { + CAUUID caGUID; - if(caGUID.pElems) CoTaskMemFree(caGUID.pElems); + memset(&caGUID, 0, sizeof(caGUID)); + + if(SUCCEEDED(pSPP->GetPages(&caGUID))) + { + IUnknown* lpUnk = NULL; + pSPP.QueryInterface(&lpUnk); + OleCreatePropertyFrame(m_hWnd, 0, 0, c.FriendlyName.c_str(), 1, (IUnknown**)&lpUnk, caGUID.cElems, caGUID.pElems, 0, 0, NULL); + lpUnk->Release(); + + if(caGUID.pElems) CoTaskMemFree(caGUID.pElems); + } + } + else if(CComQIPtr pAMVfWCD = c.filter) + { + if(pAMVfWCD->ShowDialog(VfwCompressDialog_QueryConfig, NULL) == S_OK) + { + pAMVfWCD->ShowDialog(VfwCompressDialog_Config, m_hWnd); + } + } } + + return true; } - else if(CComQIPtr pAMVfWCD = c.filter) + else if(id == IDOK) { - if(pAMVfWCD->ShowDialog(VfwCompressDialog_QueryConfig, NULL) == S_OK) + m_width = GetTextAsInt(IDC_WIDTH); + m_height = GetTextAsInt(IDC_HEIGHT); + m_filename = GetText(IDC_FILENAME); + + Codec c; + + if(GetSelCodec(c) == 0) { - pAMVfWCD->ShowDialog(VfwCompressDialog_Config, m_hWnd); + return false; } + + m_enc = c.filter; + + theApp.SetConfig("CaptureWidth", m_width); + theApp.SetConfig("CaptureHeight", m_height); + theApp.SetConfig("CaptureFileName", m_filename.c_str()); + + wstring s = wstring(c.DisplayName.m_str); + + theApp.SetConfig("CaptureVideoCodecDisplayName", string(s.begin(), s.end()).c_str()); } -} -void GSCaptureDlg::OnUpdateButton2(CCmdUI* pCmdUI) -{ - pCmdUI->Enable(m_codeclist.GetCurSel() >= 0 && m_codeclist.GetItemDataPtr(m_codeclist.GetCurSel()) != NULL); -} - -void GSCaptureDlg::OnBnClickedOk() -{ - UpdateData(); - - Codec c; - - if(GetSelCodec(c) == 0) return; - - m_enc = c.filter; - - theApp.SetConfig("CaptureWidth", m_width); - theApp.SetConfig("CaptureHeight", m_height); - theApp.SetConfig("CaptureFileName", m_filename); - theApp.SetConfig("CaptureVideoCodecDisplayName", CString(c.DisplayName)); - - OnOK(); -} - -void GSCaptureDlg::OnUpdateOK(CCmdUI* pCmdUI) -{ - CString str; - - GetDlgItem(IDC_EDIT1)->GetWindowText(str); - - pCmdUI->Enable(!str.IsEmpty() && m_codeclist.GetCurSel() >= 0); -} + return __super::OnCommand(hWnd, id, code); +} \ No newline at end of file diff --git a/plugins/GSdx/GSCaptureDlg.h b/plugins/GSdx/GSCaptureDlg.h index 04d7215bd1..ceb0b04f64 100644 --- a/plugins/GSdx/GSCaptureDlg.h +++ b/plugins/GSdx/GSCaptureDlg.h @@ -21,21 +21,17 @@ #pragma once +#include "GSDialog.h" #include "resource.h" #include "baseclasses/streams.h" -// GSCaptureDlg dialog - -class GSCaptureDlg : public CDialog +class GSCaptureDlg : public GSDialog { - DECLARE_DYNAMIC(GSCaptureDlg) - -private: struct Codec { CComPtr moniker; CComPtr filter; - CString FriendlyName; + wstring FriendlyName; CComBSTR DisplayName; }; @@ -43,31 +39,15 @@ private: int GetSelCodec(Codec& c); -public: - GSCaptureDlg(CWnd* pParent = NULL); // standard constructor - virtual ~GSCaptureDlg(); - - CComPtr m_enc; - -// Dialog Data - enum { IDD = IDD_CAPTURE }; - CString m_filename; - CComboBox m_codeclist; - protected: - virtual LRESULT DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam); - virtual void DoDataExchange(CDataExchange* pDX); // DDX/DDV support - virtual BOOL OnInitDialog(); - - DECLARE_MESSAGE_MAP() + void OnInit(); + bool OnCommand(HWND hWnd, UINT id, UINT code); public: - afx_msg void OnKickIdle(); - afx_msg void OnBnClickedButton1(); - afx_msg void OnBnClickedButton2(); - afx_msg void OnUpdateButton2(CCmdUI* pCmdUI); - afx_msg void OnBnClickedOk(); - afx_msg void OnUpdateOK(CCmdUI* pCmdUI); + GSCaptureDlg(); + int m_width; int m_height; + string m_filename; + CComPtr m_enc; }; diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index a23939aeda..c0075a01d7 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -131,6 +131,7 @@ CRC::Game CRC::m_games[] = {0xAA5EC3A3, TalesOfAbyss, JP}, {0xFB236A46, SonicUnleashed, US}, {0x4C7BB3C8, SimpsonsGame, Unknown}, + {0x4C94B32C, SimpsonsGame, Unknown}, }; hash_map CRC::m_map; diff --git a/plugins/GSdx/GSDevice.cpp b/plugins/GSdx/GSDevice.cpp index 7a36bc98d5..7373028441 100644 --- a/plugins/GSdx/GSDevice.cpp +++ b/plugins/GSdx/GSDevice.cpp @@ -23,7 +23,7 @@ #include "GSDevice.h" GSDevice::GSDevice() - : m_hWnd(NULL) + : m_wnd(NULL) , m_backbuffer(NULL) , m_merge(NULL) , m_weavebob(NULL) @@ -36,9 +36,9 @@ GSDevice::~GSDevice() { } -bool GSDevice::Create(HWND hWnd, bool vsync) +bool GSDevice::Create(GSWnd* wnd, bool vsync) { - m_hWnd = hWnd; + m_wnd = wnd; m_vsync = vsync; return true; @@ -72,9 +72,7 @@ bool GSDevice::Reset(int w, int h, bool fs) void GSDevice::Present(const GSVector4i& r, int shader) { - GSVector4i cr; - - GetClientRect(m_hWnd, cr); + GSVector4i cr = m_wnd->GetClientRect(); if(m_backbuffer->GetWidth() != cr.width() || m_backbuffer->GetHeight() != cr.height()) { @@ -171,7 +169,7 @@ void GSDevice::Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVec m_current = m_merge; } -bool GSDevice::Interlace(const GSVector2i& ds, int field, int mode, float yoffset) +void GSDevice::Interlace(const GSVector2i& ds, int field, int mode, float yoffset) { if(!m_weavebob || !(m_weavebob->GetSize() == ds)) { @@ -212,6 +210,4 @@ bool GSDevice::Interlace(const GSVector2i& ds, int field, int mode, float yoffse { m_current = m_merge; } - - return true; } \ No newline at end of file diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index 3a45f8631c..a9229c8455 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -21,6 +21,7 @@ #pragma once +#include "GSWnd.h" #include "GSTexture.h" #include "GSVertex.h" #include "GSAlignedClass.h" @@ -52,7 +53,7 @@ class GSDevice : public GSAlignedClass<16> GSTexture* Fetch(int type, int w, int h, int format); protected: - HWND m_hWnd; + GSWnd* m_wnd; bool m_vsync; GSTexture* m_backbuffer; GSTexture* m_merge; @@ -72,7 +73,7 @@ public: void Recycle(GSTexture* t); - virtual bool Create(HWND hWnd, bool vsync); + virtual bool Create(GSWnd* wnd, bool vsync); virtual bool Reset(int w, int h, bool fs); virtual bool IsLost() {return false;} @@ -101,7 +102,7 @@ public: virtual bool IsCurrentRGBA() {return true;} void Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, const GSVector4& c); - bool Interlace(const GSVector2i& ds, int field, int mode, float yoffset); + void Interlace(const GSVector2i& ds, int field, int mode, float yoffset); virtual void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) {} virtual void OMSetRenderTargets(GSTexture* rt, GSTexture* ds) {}; diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index a32bd1e7d4..2a59eaa644 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -20,6 +20,7 @@ */ #include "stdafx.h" +#include "GSdx.h" #include "GSDevice10.h" #include "resource.h" @@ -55,9 +56,9 @@ GSDevice10::~GSDevice10() { } -bool GSDevice10::Create(HWND hWnd, bool vsync) +bool GSDevice10::Create(GSWnd* wnd, bool vsync) { - if(!__super::Create(hWnd, vsync)) + if(!__super::Create(wnd, vsync)) { return false; } @@ -80,7 +81,7 @@ bool GSDevice10::Create(HWND hWnd, bool vsync) //scd.BufferDesc.RefreshRate.Numerator = 60; //scd.BufferDesc.RefreshRate.Denominator = 1; scd.BufferUsage = DXGI_USAGE_RENDER_TARGET_OUTPUT; - scd.OutputWindow = hWnd; + scd.OutputWindow = (HWND)m_wnd->GetHandle(); scd.SampleDesc.Count = 1; scd.SampleDesc.Quality = 0; scd.Windowed = TRUE; @@ -748,7 +749,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "vs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "vs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -783,7 +784,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "gs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "gs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -811,7 +812,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M CComPtr shader, error; - hr = D3DX10CompileFromResource(AfxGetInstanceHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "ps_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "ps_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index 38793b42b6..d12bdd9bad 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -96,7 +96,7 @@ public: GSDevice10(); virtual ~GSDevice10(); - bool Create(HWND hWnd, bool vsync); + bool Create(GSWnd* wnd, bool vsync); bool Reset(int w, int h, bool fs); void Flip(); diff --git a/plugins/GSdx/GSDevice7.cpp b/plugins/GSdx/GSDevice7.cpp index f09d75b6bc..6995fedd5a 100644 --- a/plugins/GSdx/GSDevice7.cpp +++ b/plugins/GSdx/GSDevice7.cpp @@ -32,9 +32,9 @@ GSDevice7::~GSDevice7() { } -bool GSDevice7::Create(HWND hWnd, bool vsync) +bool GSDevice7::Create(GSWnd* wnd, bool vsync) { - if(!__super::Create(hWnd, vsync)) + if(!__super::Create(wnd, vsync)) { return false; } @@ -44,6 +44,8 @@ bool GSDevice7::Create(HWND hWnd, bool vsync) return false; } + HWND hWnd = (HWND)m_wnd->GetHandle(); + // TODO: fullscreen if(FAILED(m_dd->SetCooperativeLevel(hWnd, DDSCL_NORMAL))) @@ -135,9 +137,7 @@ void GSDevice7::Present(const GSVector4i& r, int shader) { HRESULT hr; - GSVector4i cr; - - GetClientRect(m_hWnd, cr); + GSVector4i cr = m_wnd->GetClientRect(); if(m_backbuffer->GetWidth() != cr.width() || m_backbuffer->GetHeight() != cr.height()) { @@ -163,7 +163,7 @@ void GSDevice7::Present(const GSVector4i& r, int shader) r2 = cr; - MapWindowPoints(m_hWnd, HWND_DESKTOP, (POINT*)&r2, 2); + MapWindowPoints((HWND)m_wnd->GetHandle(), HWND_DESKTOP, (POINT*)&r2, 2); if(m_vsync) { diff --git a/plugins/GSdx/GSDevice7.h b/plugins/GSdx/GSDevice7.h index ac47094df2..3e6f4c37af 100644 --- a/plugins/GSdx/GSDevice7.h +++ b/plugins/GSdx/GSDevice7.h @@ -39,7 +39,7 @@ public: GSDevice7(); virtual ~GSDevice7(); - bool Create(HWND hWnd, bool vsync); + bool Create(GSWnd* wnd, bool vsync); bool Reset(int w, int h, bool fs); void Present(const GSVector4i& r, int shader); diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 8ccafc6a27..443b1e56b5 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -61,9 +61,9 @@ GSDevice9::~GSDevice9() if(m_ps_cb) _aligned_free(m_ps_cb); } -bool GSDevice9::Create(HWND hWnd, bool vsync) +bool GSDevice9::Create(GSWnd* wnd, bool vsync) { - if(!__super::Create(hWnd, vsync)) + if(!__super::Create(wnd, vsync)) { return false; } @@ -118,13 +118,12 @@ bool GSDevice9::Create(HWND hWnd, bool vsync) if(psver > m_d3dcaps.PixelShaderVersion) { - CString str; - - str.Format(_T("Supported pixel shader version is too low!\n\nSupported: %d.%d\nSelected: %d.%d"), + string s = format( + "Supported pixel shader version is too low!\n\nSupported: %d.%d\nSelected: %d.%d", D3DSHADER_VERSION_MAJOR(m_d3dcaps.PixelShaderVersion), D3DSHADER_VERSION_MINOR(m_d3dcaps.PixelShaderVersion), D3DSHADER_VERSION_MAJOR(psver), D3DSHADER_VERSION_MINOR(psver)); - AfxMessageBox(str); + MessageBox(NULL, s.c_str(), "GSdx", MB_OK); return false; } @@ -251,7 +250,7 @@ bool GSDevice9::Reset(int w, int h, bool fs) memset(&m_pp, 0, sizeof(m_pp)); m_pp.Windowed = TRUE; - m_pp.hDeviceWindow = m_hWnd; + m_pp.hDeviceWindow = (HWND)m_wnd->GetHandle(); m_pp.SwapEffect = D3DSWAPEFFECT_FLIP; m_pp.BackBufferFormat = D3DFMT_X8R8G8B8; m_pp.BackBufferWidth = 1; @@ -276,16 +275,14 @@ bool GSDevice9::Reset(int w, int h, bool fs) m_pp.BackBufferHeight = mh; // m_pp.FullScreen_RefreshRateInHz = mrr; - ::SetWindowLong(m_hWnd, GWL_STYLE, ::GetWindowLong(m_hWnd, GWL_STYLE) & ~(WS_CAPTION|WS_THICKFRAME)); - ::SetWindowPos(m_hWnd, NULL, 0, 0, 0, 0, SWP_NOSIZE | SWP_NOMOVE | SWP_NOZORDER | SWP_NOACTIVATE); - ::SetMenu(m_hWnd, NULL); + m_wnd->HideFrame(); } if(!m_dev) { uint32 flags = D3DCREATE_MULTITHREADED | (m_d3dcaps.VertexProcessingCaps ? D3DCREATE_HARDWARE_VERTEXPROCESSING : D3DCREATE_SOFTWARE_VERTEXPROCESSING); - hr = m_d3d->CreateDevice(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, m_hWnd, flags, &m_pp, &m_dev); + hr = m_d3d->CreateDevice(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, (HWND)m_wnd->GetHandle(), flags, &m_pp, &m_dev); if(FAILED(hr)) return false; } @@ -516,7 +513,8 @@ GSTexture* GSDevice9::CreateRenderTarget(int w, int h, int format) GSTexture* GSDevice9::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D24S8); // D3DFMT_D32F_LOCKABLE + return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D24S8); + // return __super::CreateDepthStencil(w, h, format ? format : D3DFMT_D32F_LOCKABLE); } GSTexture* GSDevice9::CreateTexture(int w, int h, int format) @@ -946,15 +944,15 @@ static HRESULT LoadShader(uint32 id, LPCSTR& data, uint32& size) { CComPtr shader, error; - HRSRC hRes = FindResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), RT_RCDATA); + HRSRC hRes = FindResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), RT_RCDATA); if(!hRes) return E_FAIL; - size = SizeofResource(AfxGetResourceHandle(), hRes); + size = SizeofResource(theApp.GetModuleHandle(), hRes); if(size == 0) return E_FAIL; - HGLOBAL hResData = LoadResource(AfxGetResourceHandle(), hRes); + HGLOBAL hResData = LoadResource(theApp.GetModuleHandle(), hRes); if(!hResData) return E_FAIL; @@ -984,7 +982,7 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); LPCSTR data; uint32 size; @@ -1044,7 +1042,7 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(AfxGetResourceHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, flags, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, flags, &shader, &error, NULL); LPCSTR data; uint32 size; diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index a9a1123341..13be056227 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -132,7 +132,7 @@ public: GSDevice9(); virtual ~GSDevice9(); - bool Create(HWND hWnd, bool vsync); + bool Create(GSWnd* wnd, bool vsync); bool Reset(int w, int h, bool fs); bool IsLost(); diff --git a/plugins/GSdx/GSDeviceNull.cpp b/plugins/GSdx/GSDeviceNull.cpp index b9ec152c6c..75112a544f 100644 --- a/plugins/GSdx/GSDeviceNull.cpp +++ b/plugins/GSdx/GSDeviceNull.cpp @@ -22,9 +22,9 @@ #include "stdafx.h" #include "GSDeviceNull.h" -bool GSDeviceNull::Create(HWND hWnd, bool vsync) +bool GSDeviceNull::Create(GSWnd* wnd, bool vsync) { - if(!__super::Create(hWnd, vsync)) + if(!__super::Create(wnd, vsync)) { return false; } diff --git a/plugins/GSdx/GSDeviceNull.h b/plugins/GSdx/GSDeviceNull.h index 5b417529e2..3ce1c43885 100644 --- a/plugins/GSdx/GSDeviceNull.h +++ b/plugins/GSdx/GSDeviceNull.h @@ -35,6 +35,6 @@ private: public: GSDeviceNull() {} - bool Create(HWND hWnd, bool vsync); + bool Create(GSWnd* wnd, bool vsync); bool Reset(int w, int h, bool fs); }; diff --git a/plugins/GSdx/GSDialog.cpp b/plugins/GSdx/GSDialog.cpp new file mode 100644 index 0000000000..11c2101188 --- /dev/null +++ b/plugins/GSdx/GSDialog.cpp @@ -0,0 +1,170 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "StdAfx.h" +#include "GSdx.h" +#include "GSDialog.h" +#include "GSVector.h" + +GSDialog::GSDialog(UINT id) + : m_id(id) + , m_hWnd(NULL) +{ +} + +INT_PTR GSDialog::DoModal() +{ + return DialogBoxParam(theApp.GetModuleHandle(), MAKEINTRESOURCE(m_id), NULL, DialogProc, (LPARAM)this); +} + +INT_PTR CALLBACK GSDialog::DialogProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) +{ + GSDialog* dlg = NULL; + + if(message == WM_INITDIALOG) + { + dlg = (GSDialog*)lParam; + SetWindowLongPtr(hWnd, GWL_USERDATA, (LONG_PTR)dlg); + dlg->m_hWnd = hWnd; + + MONITORINFO mi; + mi.cbSize = sizeof(mi); + GetMonitorInfo(MonitorFromWindow(hWnd, MONITOR_DEFAULTTONEAREST), &mi); + + GSVector4i r; + GetWindowRect(hWnd, r); + + int x = (mi.rcWork.left + mi.rcWork.right - r.width()) / 2; + int y = (mi.rcWork.top + mi.rcWork.bottom - r.height()) / 2; + + SetWindowPos(hWnd, NULL, x, y, -1, -1, SWP_NOSIZE | SWP_NOZORDER | SWP_NOACTIVATE); + + dlg->OnInit(); + + return true; + } + + dlg = (GSDialog*)GetWindowLongPtr(hWnd, GWL_USERDATA); + + return dlg != NULL ? dlg->OnMessage(message, wParam, lParam) : FALSE; +} + +bool GSDialog::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) +{ + return message == WM_COMMAND ? OnCommand((HWND)lParam, LOWORD(wParam), HIWORD(wParam)) : false; +} + +bool GSDialog::OnCommand(HWND hWnd, UINT id, UINT code) +{ + if(id == IDOK || id == IDCANCEL) + { + EndDialog(m_hWnd, id); + + return true; + } + + return false; +} + +string GSDialog::GetText(UINT id) +{ + string s; + + char* buff = NULL; + + for(int size = 256, limit = 65536; size < limit; size <<= 1) + { + buff = new char[size]; + + if(GetDlgItemText(m_hWnd, id, buff, size)) + { + s = buff; + size = limit; + } + + delete [] buff; + } + + return s; +} + +int GSDialog::GetTextAsInt(UINT id) +{ + return atoi(GetText(id).c_str()); +} + +void GSDialog::SetText(UINT id, const char* str) +{ + SetDlgItemText(m_hWnd, id, str); +} + +void GSDialog::SetTextAsInt(UINT id, int i) +{ + char buff[32] = {0}; + itoa(i, buff, 10); + SetText(id, buff); +} + +void GSDialog::ComboBoxInit(HWND hWnd, const GSSetting* settings, int count, uint32 selid, uint32 maxid) +{ + for(int i = 0; i < count; i++) + { + if(settings[i].id <= maxid) + { + string str = settings[i].name; + + if(settings[i].note != NULL) + { + str = str + " (" + settings[i].note + ")"; + } + + ComboBoxAppend(hWnd, str.c_str(), (LPARAM)settings[i].id, settings[i].id == selid); + } + } +} + +int GSDialog::ComboBoxAppend(HWND hWnd, const char* str, LPARAM data, bool select) +{ + int item = (int)SendMessage(hWnd, CB_ADDSTRING, 0, (LPARAM)str); + + SendMessage(hWnd, CB_SETITEMDATA, item, (LPARAM)data); + + if(select) + { + SendMessage(hWnd, CB_SETCURSEL, item, 0); + } + + return item; +} + +bool GSDialog::ComboBoxGetSelData(HWND hWnd, INT_PTR& data) +{ + int item = SendMessage(hWnd, CB_GETCURSEL, 0, 0); + + if(item >= 0) + { + data = SendMessage(hWnd, CB_GETITEMDATA, item, 0); + + return true; + } + + return false; +} diff --git a/plugins/GSdx/GSDialog.h b/plugins/GSdx/GSDialog.h new file mode 100644 index 0000000000..269b1f8bdb --- /dev/null +++ b/plugins/GSdx/GSDialog.h @@ -0,0 +1,54 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSSetting.h" + +class GSDialog +{ + int m_id; + + static INT_PTR CALLBACK DialogProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam); + +protected: + HWND m_hWnd; + + virtual void OnInit() {} + virtual bool OnMessage(UINT message, WPARAM wParam, LPARAM lParam); + virtual bool OnCommand(HWND hWnd, UINT id, UINT code); + +public: + GSDialog(UINT id); + virtual ~GSDialog() {} + + INT_PTR DoModal(); + + string GetText(UINT id); + int GetTextAsInt(UINT id); + + void SetText(UINT id, const char* str); + void SetTextAsInt(UINT id, int i); + + static void ComboBoxInit(HWND hWnd, const GSSetting* settings, int count, uint32 selid, uint32 maxid = ~0); + static int ComboBoxAppend(HWND hWnd, const char* str, LPARAM data = 0, bool select = false); + static bool ComboBoxGetSelData(HWND hWnd, INT_PTR& data); +}; diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 31567fa8e3..67495d0aad 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -468,11 +468,8 @@ GSLocalMemory::~GSLocalMemory() for(hash_map::iterator i = m_omap.begin(); i != m_omap.end(); i++) { Offset* o = (*i).second; - - for(int i = 0; i < countof(o->col); i++) - { - _aligned_free(o->col); - } + + _aligned_free(o->col[0]); _aligned_free(o); } diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 53da363389..cd2dfc679a 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -56,7 +56,7 @@ bool GSRenderer::Create(const string& title) ASSERT(m_dev); - if(!m_dev->Create(m_wnd, m_vsync)) + if(!m_dev->Create(&m_wnd, m_vsync)) { return false; } @@ -243,10 +243,7 @@ bool GSRenderer::Merge(int field) int field2 = 1 - ((m_interlace - 1) & 1); int mode = (m_interlace - 1) >> 1; - if(!m_dev->Interlace(ds, field ^ field2, mode, tex[1]->m_scale.y)) - { - return false; - } + m_dev->Interlace(ds, field ^ field2, mode, tex[1]->m_scale.y); } } @@ -268,25 +265,20 @@ void GSRenderer::VSync(int field) // osd - static uint64 s_frame = 0; - static string s_stats; - - if(m_perfmon.GetFrame() - s_frame >= 30) + if((m_perfmon.GetFrame() & 0x1f) == 0) { m_perfmon.Update(); - s_frame = m_perfmon.GetFrame(); - double fps = 1000.0f / m_perfmon.Get(GSPerfMon::Frame); - string s = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; + string s2 = m_regs->SMODE2.INT ? (string("Interlaced ") + (m_regs->SMODE2.FFMD ? "(frame)" : "(field)")) : "Progressive"; GSVector4i r = GetDisplayRect(); - s_stats = format( + string s = format( "%I64d | %d x %d | %.2f fps (%d%%) | %s - %s | %s | %d/%d/%d | %d%% CPU | %.2f | %.2f", m_perfmon.GetFrame(), r.width(), r.height(), fps, (int)(100.0 * fps / GetFPS()), - s.c_str(), + s2.c_str(), GSSettingsDlg::g_interlace[m_interlace].name, GSSettingsDlg::g_aspectratio[m_aspectratio].name, (int)m_perfmon.Get(GSPerfMon::Quad), @@ -301,15 +293,15 @@ void GSRenderer::VSync(int field) if(fillrate > 0) { - s_stats += format(" | %.2f mpps", fps * fillrate / (1024 * 1024)); + s += format(" | %.2f mpps", fps * fillrate / (1024 * 1024)); } if(m_capture.IsCapturing()) { - s_stats += " | Recording..."; + s += " | Recording..."; } - m_wnd.SetWindowText(s_stats.c_str()); + m_wnd.SetWindowText(s.c_str()); } if(m_frameskip) @@ -324,11 +316,7 @@ void GSRenderer::VSync(int field) ResetDevice(); } - GSVector4i r; - - m_wnd.GetClientRect(r); - - m_dev->Present(r.fit(m_aspectratio), m_shader); + m_dev->Present(m_wnd.GetClientRect().fit(m_aspectratio), m_shader); // snapshot @@ -389,6 +377,23 @@ void GSRenderer::VSync(int field) } } +bool GSRenderer::MakeSnapshot(const string& path) +{ + if(m_snapshot.empty()) + { + time_t t = time(NULL); + + char buff[16]; + + if(strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) + { + m_snapshot = format("%s_%s", path.c_str(), buff); + } + } + + return true; +} + void GSRenderer::KeyEvent(GSKeyEventData* e) { if(e->type == KEYPRESS) @@ -422,19 +427,3 @@ void GSRenderer::KeyEvent(GSKeyEventData* e) } } -bool GSRenderer::MakeSnapshot(const string& path) -{ - if(m_snapshot.empty()) - { - time_t t = time(NULL); - - char buff[16]; - - if(strftime(buff, sizeof(buff), "%Y%m%d%H%M%S", localtime(&t))) - { - m_snapshot = format("%s_%s", path.c_str(), buff); - } - } - - return true; -} diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index b4d1f4c3d7..3e9bdd2308 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -63,10 +63,9 @@ public: virtual ~GSRenderer(); virtual bool Create(const string& title); - virtual void VSync(int field); - virtual void KeyEvent(GSKeyEventData* e); virtual bool MakeSnapshot(const string& path); + virtual void KeyEvent(GSKeyEventData* e); virtual void MinMaxUV(int w, int h, GSVector4i& r) { @@ -225,7 +224,7 @@ public: { } - ~GSRendererT() + virtual ~GSRendererT() { if(m_vertices) _aligned_free(m_vertices); } diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index d1e95de50d..9dcf987834 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -271,14 +271,14 @@ protected: void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { - TRACE(_T("[%d] InvalidateVideoMem %d,%d - %d,%d %05x (%d)\n"), (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.DBP, (int)BITBLTBUF.DPSM); + // printf("[%d] InvalidateVideoMem %d,%d - %d,%d %05x (%d)\n", (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.DBP, (int)BITBLTBUF.DPSM); m_tc->InvalidateVideoMem(BITBLTBUF, r); } void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { - TRACE(_T("[%d] InvalidateLocalMem %d,%d - %d,%d %05x (%d)\n"), (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.SBP, (int)BITBLTBUF.SPSM); + // printf("[%d] InvalidateLocalMem %d,%d - %d,%d %05x (%d)\n", (int)m_perfmon.GetFrame(), r.left, r.top, r.right, r.bottom, (int)BITBLTBUF.SBP, (int)BITBLTBUF.SPSM); m_tc->InvalidateLocalMem(BITBLTBUF, r); } diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index a3e083d799..0be46daee9 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -25,8 +25,6 @@ #include "GSTextureCacheSW.h" #include "GSDrawScanline.h" -extern const GSVector4 g_pos_scale; - class GSRendererSW : public GSRendererT { protected: diff --git a/plugins/GSdx/GSSetting.h b/plugins/GSdx/GSSetting.h index 0124dc014a..ced75eb7b9 100644 --- a/plugins/GSdx/GSSetting.h +++ b/plugins/GSdx/GSSetting.h @@ -23,23 +23,7 @@ struct GSSetting { - DWORD id; - const TCHAR* name; - const TCHAR* note; - - static void InitComboBox(const GSSetting* settings, int count, CComboBox& combobox, DWORD sel, DWORD maxid = ~0) - { - for(int i = 0; i < count; i++) - { - if(settings[i].id <= maxid) - { - CString str = settings[i].name; - if(settings[i].note != NULL) str = str + _T(" (") + settings[i].note + _T(")"); - int item = combobox.AddString(str); - combobox.SetItemData(item, settings[i].id); - if(settings[i].id == sel) combobox.SetCurSel(item); - } - } - } - + uint32 id; + const char* name; + const char* note; }; diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 928fa7ecb2..ca431f74a0 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -23,123 +23,55 @@ #include "GSdx.h" #include "GSSettingsDlg.h" #include "GSUtil.h" -#include -#include +#include "resource.h" GSSetting GSSettingsDlg::g_renderers[] = { - {0, _T("Direct3D9 (Hardware)"), NULL}, - {1, _T("Direct3D9 (Software)"), NULL}, - {2, _T("Direct3D9 (Null)"), NULL}, - {3, _T("Direct3D10 (Hardware)"), NULL}, - {4, _T("Direct3D10 (Software)"), NULL}, - {5, _T("Direct3D10 (Null)"), NULL}, - {6, _T("Null (Software)"), NULL}, - {7, _T("Null (Null)"), NULL}, + {0, "Direct3D9 (Hardware)", NULL}, + {1, "Direct3D9 (Software)", NULL}, + {2, "Direct3D9 (Null)", NULL}, + {3, "Direct3D10 (Hardware)", NULL}, + {4, "Direct3D10 (Software)", NULL}, + {5, "Direct3D10 (Null)", NULL}, + {6, "Null (Software)", NULL}, + {7, "Null (Null)", NULL}, }; GSSetting GSSettingsDlg::g_psversion[] = { - {D3DPS_VERSION(3, 0), _T("Pixel Shader 3.0"), NULL}, - {D3DPS_VERSION(2, 0), _T("Pixel Shader 2.0"), NULL}, - //{D3DPS_VERSION(1, 4), _T("Pixel Shader 1.4"), NULL}, - //{D3DPS_VERSION(1, 1), _T("Pixel Shader 1.1"), NULL}, - //{D3DPS_VERSION(0, 0), _T("Fixed Pipeline (bogus)"), NULL}, + {D3DPS_VERSION(3, 0), "Pixel Shader 3.0", NULL}, + {D3DPS_VERSION(2, 0), "Pixel Shader 2.0", NULL}, + //{D3DPS_VERSION(1, 4), "Pixel Shader 1.4", NULL}, + //{D3DPS_VERSION(1, 1), "Pixel Shader 1.1", NULL}, + //{D3DPS_VERSION(0, 0), "Fixed Pipeline (bogus)", NULL}, }; GSSetting GSSettingsDlg::g_interlace[] = { - {0, _T("None"), NULL}, - {1, _T("Weave tff"), _T("saw-tooth")}, - {2, _T("Weave bff"), _T("saw-tooth")}, - {3, _T("Bob tff"), _T("use blend if shaking")}, - {4, _T("Bob bff"), _T("use blend if shaking")}, - {5, _T("Blend tff"), _T("slight blur, 1/2 fps")}, - {6, _T("Blend bff"), _T("slight blur, 1/2 fps")}, + {0, "None", NULL}, + {1, "Weave tff", "saw-tooth"}, + {2, "Weave bff", "saw-tooth"}, + {3, "Bob tff", "use blend if shaking"}, + {4, "Bob bff", "use blend if shaking"}, + {5, "Blend tff", "slight blur, 1/2 fps"}, + {6, "Blend bff", "slight blur, 1/2 fps"}, }; GSSetting GSSettingsDlg::g_aspectratio[] = { - {0, _T("Stretch"), NULL}, - {1, _T("4:3"), NULL}, - {2, _T("16:9"), NULL}, + {0, "Stretch", NULL}, + {1, "4:3", NULL}, + {2, "16:9", NULL}, }; -IMPLEMENT_DYNAMIC(GSSettingsDlg, CDialog) -GSSettingsDlg::GSSettingsDlg(CWnd* pParent /*=NULL*/) - : CDialog(GSSettingsDlg::IDD, pParent) - , m_filter(1) - , m_nativeres(FALSE) - , m_vsync(FALSE) - , m_logz(FALSE) - , m_fba(TRUE) - , m_aa1(FALSE) - , m_blur(FALSE) +GSSettingsDlg::GSSettingsDlg() + : GSDialog(IDD_CONFIG) { } -GSSettingsDlg::~GSSettingsDlg() +void GSSettingsDlg::OnInit() { -} - -LRESULT GSSettingsDlg::DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam) -{ - LRESULT ret = __super::DefWindowProc(message, wParam, lParam); - - if(message == WM_INITDIALOG) - { - SendMessage(WM_KICKIDLE); - } - - return ret; -} - -void GSSettingsDlg::DoDataExchange(CDataExchange* pDX) -{ - __super::DoDataExchange(pDX); - DDX_Control(pDX, IDC_COMBO3, m_resolution); - DDX_Control(pDX, IDC_COMBO1, m_renderer); - DDX_Control(pDX, IDC_COMBO4, m_psversion); - DDX_Control(pDX, IDC_COMBO2, m_interlace); - DDX_Control(pDX, IDC_COMBO5, m_aspectratio); - DDX_Check(pDX, IDC_CHECK4, m_filter); - DDX_Control(pDX, IDC_SPIN1, m_resx); - DDX_Control(pDX, IDC_SPIN2, m_resy); - DDX_Control(pDX, IDC_SPIN3, m_swthreads); - DDX_Check(pDX, IDC_CHECK1, m_nativeres); - DDX_Control(pDX, IDC_EDIT1, m_resxedit); - DDX_Control(pDX, IDC_EDIT2, m_resyedit); - DDX_Control(pDX, IDC_EDIT3, m_swthreadsedit); - DDX_Check(pDX, IDC_CHECK2, m_vsync); - DDX_Check(pDX, IDC_CHECK5, m_logz); - DDX_Check(pDX, IDC_CHECK7, m_fba); - DDX_Check(pDX, IDC_CHECK8, m_aa1); - DDX_Check(pDX, IDC_CHECK9, m_blur); -} - -BEGIN_MESSAGE_MAP(GSSettingsDlg, CDialog) - ON_MESSAGE_VOID(WM_KICKIDLE, OnKickIdle) - ON_UPDATE_COMMAND_UI(IDC_SPIN1, OnUpdateResolution) - ON_UPDATE_COMMAND_UI(IDC_SPIN2, OnUpdateResolution) - ON_UPDATE_COMMAND_UI(IDC_EDIT1, OnUpdateResolution) - ON_UPDATE_COMMAND_UI(IDC_EDIT2, OnUpdateResolution) - ON_UPDATE_COMMAND_UI(IDC_COMBO4, OnUpdateD3D9Options) - ON_UPDATE_COMMAND_UI(IDC_CHECK3, OnUpdateD3D9Options) - ON_UPDATE_COMMAND_UI(IDC_CHECK5, OnUpdateD3D9Options) - ON_UPDATE_COMMAND_UI(IDC_CHECK7, OnUpdateD3D9Options) - ON_UPDATE_COMMAND_UI(IDC_SPIN3, OnUpdateSWOptions) - ON_UPDATE_COMMAND_UI(IDC_EDIT3, OnUpdateSWOptions) - ON_CBN_SELCHANGE(IDC_COMBO1, &GSSettingsDlg::OnCbnSelchangeCombo1) -END_MESSAGE_MAP() - -void GSSettingsDlg::OnKickIdle() -{ - UpdateDialogControls(this, false); -} - -BOOL GSSettingsDlg::OnInitDialog() -{ - __super::OnInitDialog(); + __super::OnInit(); D3DCAPS9 caps; memset(&caps, 0, sizeof(caps)); @@ -147,174 +79,140 @@ BOOL GSSettingsDlg::OnInitDialog() m_modes.clear(); - // windowed - { D3DDISPLAYMODE mode; memset(&mode, 0, sizeof(mode)); m_modes.push_back(mode); - int iItem = m_resolution.AddString(_T("Windowed")); - m_resolution.SetItemDataPtr(iItem, &m_modes.back()); - m_resolution.SetCurSel(iItem); - } + HWND hWnd = GetDlgItem(m_hWnd, IDC_RESOLUTION); - // fullscreen + ComboBoxAppend(hWnd, "Windowed", (LPARAM)&m_modes.back(), true); - if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) - { - uint32 ModeWidth = theApp.GetConfig("ModeWidth", 0); - uint32 ModeHeight = theApp.GetConfig("ModeHeight", 0); - uint32 ModeRefreshRate = theApp.GetConfig("ModeRefreshRate", 0); - - uint32 nModes = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); - - for(uint32 i = 0; i < nModes; i++) + if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { - D3DDISPLAYMODE mode; + uint32 w = theApp.GetConfig("ModeWidth", 0); + uint32 h = theApp.GetConfig("ModeHeight", 0); + uint32 hz = theApp.GetConfig("ModeRefreshRate", 0); - if(S_OK == d3d->EnumAdapterModes(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8, i, &mode)) + uint32 n = d3d->GetAdapterModeCount(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8); + + for(uint32 i = 0; i < n; i++) { - string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); - int iItem = m_resolution.AddString(str.c_str()); - - m_modes.push_back(mode); - m_resolution.SetItemDataPtr(iItem, &m_modes.back()); - - if(ModeWidth == mode.Width && ModeHeight == mode.Height && ModeRefreshRate == mode.RefreshRate) + if(S_OK == d3d->EnumAdapterModes(D3DADAPTER_DEFAULT, D3DFMT_X8R8G8B8, i, &mode)) { - m_resolution.SetCurSel(iItem); + m_modes.push_back(mode); + + string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); + + ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } + + d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); + } + } + + { + bool isdx10avail = GSUtil::IsDirect3D10Available(); + + vector renderers; + + for(size_t i = 0; i < countof(g_renderers); i++) + { + if(i >= 3 && i <= 5 && !isdx10avail) continue; + + renderers.push_back(g_renderers[i]); } - d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); + HWND hWnd = GetDlgItem(m_hWnd, IDC_RENDERER); + + ComboBoxInit(hWnd, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); + + OnCommand(hWnd, IDC_RENDERER, CBN_SELCHANGE); } - bool isdx10avail = GSUtil::IsDirect3D10Available(); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_INTERLACE), g_interlace, countof(g_interlace), theApp.GetConfig("Interlace", 0)); + ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); - vector renderers; + CheckDlgButton(m_hWnd, IDC_FILTER, theApp.GetConfig("filter", 1)); + CheckDlgButton(m_hWnd, IDC_VSYNC, theApp.GetConfig("vsync", 0)); + CheckDlgButton(m_hWnd, IDC_LOGZ, theApp.GetConfig("logz", 0)); + CheckDlgButton(m_hWnd, IDC_FBA, theApp.GetConfig("fba", 1)); + CheckDlgButton(m_hWnd, IDC_AA1, theApp.GetConfig("aa1", 0)); + CheckDlgButton(m_hWnd, IDC_BLUR, theApp.GetConfig("blur", 0)); + CheckDlgButton(m_hWnd, IDC_NATIVERES, theApp.GetConfig("nativeres", 0)); - for(size_t i = 0; i < countof(g_renderers); i++) - { - if(i >= 3 && i <= 5 && !isdx10avail) continue; + SendMessage(GetDlgItem(m_hWnd, IDC_RESX), UDM_SETRANGE, 0, MAKELPARAM(4096, 512)); + SendMessage(GetDlgItem(m_hWnd, IDC_RESX), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("resx", 1024), 0)); - renderers.push_back(g_renderers[i]); - } + SendMessage(GetDlgItem(m_hWnd, IDC_RESY), UDM_SETRANGE, 0, MAKELPARAM(4096, 512)); + SendMessage(GetDlgItem(m_hWnd, IDC_RESY), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("resy", 1024), 0)); - GSSetting::InitComboBox(&renderers[0], renderers.size(), m_renderer, theApp.GetConfig("Renderer", 0)); - GSSetting::InitComboBox(g_psversion, countof(g_psversion), m_psversion, theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); - GSSetting::InitComboBox(g_interlace, countof(g_interlace), m_interlace, theApp.GetConfig("Interlace", 0)); - GSSetting::InitComboBox(g_aspectratio, countof(g_aspectratio), m_aspectratio, theApp.GetConfig("AspectRatio", 1)); - - OnCbnSelchangeCombo1(); - - // - - m_filter = theApp.GetConfig("filter", 1); - m_vsync = !!theApp.GetConfig("vsync", 0); - m_logz = !!theApp.GetConfig("logz", 0); - m_fba = !!theApp.GetConfig("fba", 1); - m_aa1 = !!theApp.GetConfig("aa1", 0); - m_blur = !!theApp.GetConfig("blur", 0); - - m_resx.SetRange(512, 4096); - m_resy.SetRange(512, 4096); - m_resx.SetPos(theApp.GetConfig("resx", 1024)); - m_resy.SetPos(theApp.GetConfig("resy", 1024)); - m_nativeres = !!theApp.GetConfig("nativeres", 0); - - m_resx.EnableWindow(!m_nativeres); - m_resy.EnableWindow(!m_nativeres); - m_resxedit.EnableWindow(!m_nativeres); - m_resyedit.EnableWindow(!m_nativeres); - - m_swthreads.SetRange(1, 16); - m_swthreads.SetPos(theApp.GetConfig("swthreads", 1)); - - // - - UpdateData(FALSE); - - return TRUE; // return TRUE unless you set the focus to a control - // EXCEPTION: OCX Property Pages should return FALSE + SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETRANGE, 0, MAKELPARAM(16, 1)); + SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("swthreads", 1), 0)); } -void GSSettingsDlg::OnOK() +bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { - UpdateData(); - - if(m_resolution.GetCurSel() >= 0) + if(id == IDC_RENDERER && code == CBN_SELCHANGE) { - D3DDISPLAYMODE& mode = *(D3DDISPLAYMODE*)m_resolution.GetItemData(m_resolution.GetCurSel()); + int item = (int)SendMessage(hWnd, CB_GETCURSEL, 0, 0); - theApp.SetConfig("ModeWidth", (int)mode.Width); - theApp.SetConfig("ModeHeight", (int)mode.Height); - theApp.SetConfig("ModeRefreshRate", (int)mode.RefreshRate); + if(item >= 0) + { + int i = (int)SendMessage(hWnd, CB_GETITEMDATA, item, 0); + + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), i >= 0 && i <= 2 ? SW_SHOW : SW_HIDE); + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), i >= 3 && i <= 5 ? SW_SHOW : SW_HIDE); + } + } + else if(id == IDOK) + { + INT_PTR data; + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RESOLUTION), data)) + { + const D3DDISPLAYMODE* mode = (D3DDISPLAYMODE*)data; + + theApp.SetConfig("ModeWidth", (int)mode->Width); + theApp.SetConfig("ModeHeight", (int)mode->Height); + theApp.SetConfig("ModeRefreshRate", (int)mode->RefreshRate); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), data)) + { + theApp.SetConfig("Renderer", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SHADER), data)) + { + theApp.SetConfig("PixelShaderVersion2", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_INTERLACE), data)) + { + theApp.SetConfig("Interlace", (int)data); + } + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), data)) + { + theApp.SetConfig("AspectRatio", (int)data); + } + + theApp.SetConfig("filter", (int)IsDlgButtonChecked(m_hWnd, IDC_FILTER)); + theApp.SetConfig("vsync", (int)IsDlgButtonChecked(m_hWnd, IDC_VSYNC)); + theApp.SetConfig("logz", (int)IsDlgButtonChecked(m_hWnd, IDC_LOGZ)); + theApp.SetConfig("fba", (int)IsDlgButtonChecked(m_hWnd, IDC_FBA)); + theApp.SetConfig("aa1", (int)IsDlgButtonChecked(m_hWnd, IDC_AA1)); + theApp.SetConfig("blur", (int)IsDlgButtonChecked(m_hWnd, IDC_BLUR)); + theApp.SetConfig("nativeres", (int)IsDlgButtonChecked(m_hWnd, IDC_NATIVERES)); + + theApp.SetConfig("resx", (int)SendMessage(GetDlgItem(m_hWnd, IDC_RESX), UDM_GETPOS, 0, 0)); + theApp.SetConfig("resy", (int)SendMessage(GetDlgItem(m_hWnd, IDC_RESY), UDM_GETPOS, 0, 0)); + theApp.SetConfig("swthreads", (int)SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_GETPOS, 0, 0)); } - if(m_renderer.GetCurSel() >= 0) - { - theApp.SetConfig("Renderer", (uint32)m_renderer.GetItemData(m_renderer.GetCurSel())); - } - - if(m_psversion.GetCurSel() >= 0) - { - theApp.SetConfig("PixelShaderVersion2", (uint32)m_psversion.GetItemData(m_psversion.GetCurSel())); - } - - if(m_interlace.GetCurSel() >= 0) - { - theApp.SetConfig("Interlace", (uint32)m_interlace.GetItemData(m_interlace.GetCurSel())); - } - - if(m_aspectratio.GetCurSel() >= 0) - { - theApp.SetConfig("AspectRatio", (uint32)m_aspectratio.GetItemData(m_aspectratio.GetCurSel())); - } - - theApp.SetConfig("filter", m_filter); - theApp.SetConfig("vsync", m_vsync); - theApp.SetConfig("logz", m_logz); - theApp.SetConfig("fba", m_fba); - theApp.SetConfig("aa1", m_aa1); - theApp.SetConfig("blur", m_blur); - - theApp.SetConfig("resx", m_resx.GetPos()); - theApp.SetConfig("resy", m_resy.GetPos()); - theApp.SetConfig("swthreads", m_swthreads.GetPos()); - theApp.SetConfig("nativeres", m_nativeres); - - __super::OnOK(); + return __super::OnCommand(hWnd, id, code); } -void GSSettingsDlg::OnUpdateResolution(CCmdUI* pCmdUI) -{ - UpdateData(); - - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(!m_nativeres && (i == 0 || i == 3)); -} - -void GSSettingsDlg::OnUpdateD3D9Options(CCmdUI* pCmdUI) -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(i >= 0 && i <= 2); -} - -void GSSettingsDlg::OnUpdateSWOptions(CCmdUI* pCmdUI) -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - pCmdUI->Enable(i == 1 || i == 4 || i == 6); -} - -void GSSettingsDlg::OnCbnSelchangeCombo1() -{ - int i = (int)m_renderer.GetItemData(m_renderer.GetCurSel()); - - GetDlgItem(IDC_LOGO9)->ShowWindow(i >= 0 && i <= 2 ? SW_SHOW : SW_HIDE); - GetDlgItem(IDC_LOGO10)->ShowWindow(i >= 3 && i <= 5 ? SW_SHOW : SW_HIDE); -} diff --git a/plugins/GSdx/GSSettingsDlg.h b/plugins/GSdx/GSSettingsDlg.h index 1f48ce710d..f5f76f2503 100644 --- a/plugins/GSdx/GSSettingsDlg.h +++ b/plugins/GSdx/GSSettingsDlg.h @@ -21,59 +21,22 @@ #pragma once +#include "GSDialog.h" #include "GSSetting.h" -#include "resource.h" -class GSSettingsDlg : public CDialog +class GSSettingsDlg : public GSDialog { - DECLARE_DYNAMIC(GSSettingsDlg) - -private: list m_modes; +protected: + void OnInit(); + bool OnCommand(HWND hWnd, UINT id, UINT code); + public: - GSSettingsDlg(CWnd* pParent = NULL); // standard constructor - virtual ~GSSettingsDlg(); + GSSettingsDlg(); static GSSetting g_renderers[]; static GSSetting g_psversion[]; static GSSetting g_interlace[]; static GSSetting g_aspectratio[]; - -// Dialog Data - enum { IDD = IDD_CONFIG }; - CComboBox m_resolution; - CComboBox m_renderer; - CComboBox m_psversion; - CComboBox m_interlace; - CComboBox m_aspectratio; - int m_filter; - CSpinButtonCtrl m_resx; - CSpinButtonCtrl m_resy; - CSpinButtonCtrl m_swthreads; - BOOL m_nativeres; - CEdit m_resxedit; - CEdit m_resyedit; - CEdit m_swthreadsedit; - BOOL m_vsync; - BOOL m_logz; - BOOL m_fba; - BOOL m_aa1; - BOOL m_blur; - -protected: - virtual LRESULT DefWindowProc(UINT message, WPARAM wParam, LPARAM lParam); - virtual void DoDataExchange(CDataExchange* pDX); // DDX/DDV support - virtual BOOL OnInitDialog(); - virtual void OnOK(); - - DECLARE_MESSAGE_MAP() - -public: - afx_msg void OnKickIdle(); - afx_msg void OnUpdateResolution(CCmdUI* pCmdUI); - afx_msg void OnUpdateD3D9Options(CCmdUI* pCmdUI); - afx_msg void OnUpdateSWOptions(CCmdUI* pCmdUI); - afx_msg void OnCbnSelchangeCombo1(); -}; - +}; \ No newline at end of file diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index d1d3ccad62..672103be6e 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -596,7 +596,7 @@ void GSState::GIFRegHandlerPRMODECONT(GIFReg* r) PRIM = m_env.PRMODECONT.AC ? &m_env.PRIM : (GIFRegPRIM*)&m_env.PRMODE; - if(PRIM->PRIM == 7) TRACE(_T("Invalid PRMODECONT/PRIM\n")); + // if(PRIM->PRIM == 7) printf("Invalid PRMODECONT/PRIM\n"); m_context = &m_env.CTXT[PRIM->CTXT]; diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index f4eb298a0c..fbb0d9a966 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -36,8 +36,9 @@ GSTexture9::GSTexture9(IDirect3DSurface9* surface) if(m_desc.Type != D3DRTYPE_SURFACE) { - HRESULT hr = surface->GetContainer(__uuidof(IDirect3DTexture9), (void**)&m_texture); - ASSERT(SUCCEEDED(hr)); + surface->GetContainer(__uuidof(IDirect3DTexture9), (void**)&m_texture); + + ASSERT(m_texture != NULL); } } diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 7b27dea300..ebc31259f3 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -90,7 +90,7 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) return; } - TRACE(_T("GSRenderTarget::Read %d,%d - %d,%d (%08x)\n"), r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); + // printf("GSRenderTarget::Read %d,%d - %d,%d (%08x)\n", r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); // m_renderer->m_perfmon.Put(GSPerfMon::ReadRT, 1); @@ -202,7 +202,7 @@ bool GSTextureCache10::GSCachedTextureHW10::Create() switch(psm) { default: - TRACE(_T("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n"), m_TEX0.PSM, m_TEX0.CPSM); + // printf("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n", m_TEX0.PSM, m_TEX0.CPSM); case PSM_PSMCT32: m_bpp = 32; m_bpp2 = 0; diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index d83f118d27..2b98621de5 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -91,7 +91,7 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) return; } - TRACE(_T("GSRenderTarget::Read %d,%d - %d,%d (%08x)\n"), r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); + // printf("GSRenderTarget::Read %d,%d - %d,%d (%08x)\n", r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); // m_renderer->m_perfmon.Put(GSPerfMon::ReadRT, 1); @@ -201,7 +201,7 @@ bool GSTextureCache9::GSCachedTexture9::Create() switch(psm) { default: - TRACE(_T("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n"), m_TEX0.PSM, m_TEX0.CPSM); + // printf("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n", m_TEX0.PSM, m_TEX0.CPSM); case PSM_PSMCT32: m_bpp = 32; m_bpp2 = 0; diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index 620464f6ef..64ce305c5d 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -120,11 +120,11 @@ bool GSUtil::CheckDirectX() } else { - int res = AfxMessageBox(_T("You need to update directx, would you like to do it now?"), MB_YESNO); - - if(res == IDYES) + if(MessageBox(NULL, "You need to update directx, would you like to do it now?", "GSdx", MB_YESNO) == IDYES) { - ShellExecute(NULL, _T("open"), _T("http://www.microsoft.com/downloads/details.aspx?FamilyId=2DA43D38-DB71-4C1B-BC6A-9B6652CD92A3"), NULL, NULL, SW_SHOWNORMAL); + const char* url = "http://www.microsoft.com/downloads/details.aspx?FamilyId=2DA43D38-DB71-4C1B-BC6A-9B6652CD92A3"; + + ShellExecute(NULL, "open", url, NULL, NULL, SW_SHOWNORMAL); } return false; @@ -152,7 +152,7 @@ bool GSUtil::CheckSSE() { string s = format("This CPU does not support SSE %d.%02d", _M_SSE >> 8, _M_SSE & 0xff); - AfxMessageBox(s.c_str(), MB_OK); + MessageBox(NULL, s.c_str(), "GSdx", MB_OK); return false; } diff --git a/plugins/GSdx/GSWnd.cpp b/plugins/GSdx/GSWnd.cpp index bcb8ff9caa..aa9226c5f5 100644 --- a/plugins/GSdx/GSWnd.cpp +++ b/plugins/GSdx/GSWnd.cpp @@ -23,55 +23,149 @@ #include "GSdx.h" #include "GSWnd.h" -BEGIN_MESSAGE_MAP(GSWnd, CWnd) - ON_WM_CLOSE() -END_MESSAGE_MAP() - GSWnd::GSWnd() + : m_hWnd(NULL) { } GSWnd::~GSWnd() { - DestroyWindow(); +} + +LRESULT CALLBACK GSWnd::WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) +{ + GSWnd* wnd = NULL; + + if(message == WM_NCCREATE) + { + wnd = (GSWnd*)((LPCREATESTRUCT)lParam)->lpCreateParams; + + SetWindowLongPtr(hWnd, GWL_USERDATA, (LONG_PTR)wnd); + + wnd->m_hWnd = hWnd; + } + else + { + wnd = (GSWnd*)GetWindowLongPtr(hWnd, GWL_USERDATA); + } + + if(wnd == NULL) + { + return DefWindowProc(hWnd, message, wParam, lParam); + } + + return wnd->OnMessage(message, wParam, lParam); +} + +LRESULT GSWnd::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) +{ + switch(message) + { + case WM_CLOSE: + Hide(); + // DestroyWindow(m_hWnd); + return 0; + case WM_DESTROY: + PostQuitMessage(0); + return 0; + default: + break; + } + + return DefWindowProc(m_hWnd, message, wParam, lParam); } bool GSWnd::Create(const string& title) { - CRect r; + GSVector4i r; - GetDesktopWindow()->GetWindowRect(r); + GetWindowRect(GetDesktopWindow(), r); - CSize s(r.Width() / 3, r.Width() / 4); + int w = r.width() / 3; + int h = r.width() / 4; if(!GetSystemMetrics(SM_REMOTESESSION)) { - s.cx *= 2; - s.cy *= 2; + w *= 2; + h *= 2; } - r = CRect(r.CenterPoint() - CSize(s.cx / 2, s.cy / 2), s); + int x = (r.left + r.right - w) / 2; + int y = (r.top + r.bottom - h) / 2; - LPCTSTR wc = AfxRegisterWndClass(CS_VREDRAW|CS_HREDRAW|CS_DBLCLKS, theApp.LoadStandardCursor(IDC_ARROW), 0, 0); + WNDCLASS wc; - return !!CreateEx(0, wc, title.c_str(), WS_OVERLAPPEDWINDOW, r, NULL, 0); + memset(&wc, 0, sizeof(wc)); + + wc.style = CS_HREDRAW | CS_VREDRAW | CS_DBLCLKS; + wc.lpfnWndProc = WndProc; + wc.hInstance = theApp.GetModuleHandle(); + // TODO: wc.hIcon = ; + wc.hCursor = LoadCursor(NULL, IDC_ARROW); + wc.hbrBackground = (HBRUSH)GetStockObject(BLACK_BRUSH); + wc.lpszClassName = "GSWnd"; + + if(!RegisterClass(&wc)) + { + return false; + } + + DWORD style = WS_CLIPCHILDREN | WS_CLIPSIBLINGS | WS_OVERLAPPEDWINDOW | WS_BORDER; + + m_hWnd = CreateWindow(wc.lpszClassName, title.c_str(), style, x, y, w, h, NULL, NULL, wc.hInstance, (LPVOID)this); + + if(!m_hWnd) + { + return false; + } + + return true; +} + +bool GSWnd::Attach(HWND hWnd) +{ + // TODO: subclass + + m_hWnd = hWnd; + + return true; +} + +GSVector4i GSWnd::GetClientRect() +{ + GSVector4i r; + + ::GetClientRect(m_hWnd, r); + + return r; +} + +void GSWnd::SetWindowText(const char* title) +{ + ::SetWindowText(m_hWnd, title); } void GSWnd::Show() { - SetWindowPos(&wndTop, 0, 0, 0, 0, SWP_NOMOVE|SWP_NOSIZE); - SetForegroundWindow(); - ShowWindow(SW_SHOWNORMAL); + //SetWindowPos(&wndTop, 0, 0, 0, 0, SWP_NOMOVE|SWP_NOSIZE); + + SetForegroundWindow(m_hWnd); + + ShowWindow(m_hWnd, SW_SHOWNORMAL); + + UpdateWindow(m_hWnd); } void GSWnd::Hide() { - ShowWindow(SW_HIDE); + ShowWindow(m_hWnd, SW_HIDE); } -void GSWnd::OnClose() +void GSWnd::HideFrame() { - Hide(); - - PostMessage(WM_QUIT); -} \ No newline at end of file + SetWindowLong(m_hWnd, GWL_STYLE, GetWindowLong(m_hWnd, GWL_STYLE) & ~(WS_CAPTION|WS_THICKFRAME)); + + SetWindowPos(m_hWnd, NULL, 0, 0, 0, 0, SWP_NOSIZE | SWP_NOMOVE | SWP_NOZORDER | SWP_NOACTIVATE); + + SetMenu(m_hWnd, NULL); +} diff --git a/plugins/GSdx/GSWnd.h b/plugins/GSdx/GSWnd.h index 4de785f724..ce5a0bebcf 100644 --- a/plugins/GSdx/GSWnd.h +++ b/plugins/GSdx/GSWnd.h @@ -21,17 +21,30 @@ #pragma once -class GSWnd : public CWnd +#include "GSVector.h" + +class GSWnd { - DECLARE_MESSAGE_MAP() + HWND m_hWnd; + + static LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam); + virtual LRESULT OnMessage(UINT message, WPARAM wParam, LPARAM lParam); public: GSWnd(); virtual ~GSWnd(); - virtual bool Create(const string& title); + bool Create(const string& title); + bool Attach(HWND hWnd); + + void* GetHandle() {return m_hWnd;} + + GSVector4i GetClientRect(); + + void SetWindowText(const char* title); void Show(); void Hide(); - void OnClose(); + + void HideFrame(); }; diff --git a/plugins/GSdx/GSdx.cpp b/plugins/GSdx/GSdx.cpp index 339263ba7f..d5fe5c69d4 100644 --- a/plugins/GSdx/GSdx.cpp +++ b/plugins/GSdx/GSdx.cpp @@ -21,96 +21,58 @@ #include "stdafx.h" #include "GSdx.h" -#include -#include -// -// Note! -// -// If this DLL is dynamically linked against the MFC -// DLLs, any functions exported from this DLL which -// call into MFC must have the AFX_MANAGE_STATE macro -// added at the very beginning of the function. -// -// For example: -// -// extern "C" BOOL PASCAL EXPORT ExportedFunction() -// { -// AFX_MANAGE_STATE(AfxGetStaticModuleState()); -// // normal function body here -// } -// -// It is very important that this macro appear in each -// function, prior to any calls into MFC. This means that -// it must appear as the first statement within the -// function, even before any object variable declarations -// as their constructors may generate calls into the MFC -// DLL. -// -// Please see MFC Technical Notes 33 and 58 for additional -// details. -// +static HMODULE s_hModule; + +BOOL APIENTRY DllMain(HMODULE hModule, DWORD ul_reason_for_call, LPVOID lpReserved) +{ + switch(ul_reason_for_call) + { + case DLL_PROCESS_ATTACH: + s_hModule = hModule; + case DLL_THREAD_ATTACH: + case DLL_THREAD_DETACH: + case DLL_PROCESS_DETACH: + break; + } + + return TRUE; +} GSdxApp theApp; +const char* GSdxApp::m_ini = "inis/GSdx.ini"; +const char* GSdxApp::m_section = "Settings"; + GSdxApp::GSdxApp() { } -BOOL GSdxApp::InitInstance() +HMODULE GSdxApp::GetModuleHandle() { - __super::InitInstance(); - - SetRegistryKey(_T("Gabest")); - - CString str; - GetModuleFileName(AfxGetInstanceHandle(), str.GetBuffer(MAX_PATH), MAX_PATH); - str.ReleaseBuffer(); - - CPath path(str); - path.RenameExtension(_T(".ini")); - - CPath fn = path; - fn.StripPath(); - - path.RemoveFileSpec(); - path.Append(_T("..\\inis")); - CreateDirectory(path, NULL); - path.Append(fn); - - if(m_pszRegistryKey) - { - free((void*)m_pszRegistryKey); - } - - m_pszRegistryKey = NULL; - - if(m_pszProfileName) - { - free((void*)m_pszProfileName); - } - - m_pszProfileName = _tcsdup((LPCTSTR)path); - - return TRUE; + return s_hModule; } string GSdxApp::GetConfig(const char* entry, const char* value) { - return string(GetProfileString("Settings", entry, value)); + char buff[4096] = {0}; + GetPrivateProfileString(m_section, entry, value, buff, countof(buff), m_ini); + return string(buff); } void GSdxApp::SetConfig(const char* entry, const char* value) { - WriteProfileString("Settings", entry, value); + WritePrivateProfileString(m_section, entry, value, m_ini); } int GSdxApp::GetConfig(const char* entry, int value) { - return GetProfileInt("Settings", entry, value); + return GetPrivateProfileInt(m_section, entry, value, m_ini); } void GSdxApp::SetConfig(const char* entry, int value) { - WriteProfileInt("Settings", entry, value); + char buff[32] = {0}; + itoa(value, buff, 10); + SetConfig(entry, buff); } diff --git a/plugins/GSdx/GSdx.h b/plugins/GSdx/GSdx.h index e1abf14876..6b1f73b1ab 100644 --- a/plugins/GSdx/GSdx.h +++ b/plugins/GSdx/GSdx.h @@ -21,14 +21,16 @@ #pragma once -class GSdxApp : public CWinApp +class GSdxApp { -protected: - virtual BOOL InitInstance(); + static const char* m_ini; + static const char* m_section; public: GSdxApp(); + HMODULE GetModuleHandle(); + string GetConfig(const char* entry, const char* value); void SetConfig(const char* entry, const char* value); int GetConfig(const char* entry, int value); diff --git a/plugins/GSdx/GSdx.rc b/plugins/GSdx/GSdx.rc index 58ba4fc312..d26b46b7c4 100644 --- a/plugins/GSdx/GSdx.rc +++ b/plugins/GSdx/GSdx.rc @@ -90,31 +90,31 @@ BEGIN CONTROL 2022,IDC_LOGO10,"Static",SS_BITMAP,7,7,175,44 CONTROL 2021,IDC_LOGO9,"Static",SS_BITMAP,7,7,175,44 LTEXT "Resolution:",IDC_STATIC,7,59,37,8 - COMBOBOX IDC_COMBO3,71,57,111,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_RESOLUTION,71,57,111,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Renderer:",IDC_STATIC,7,74,34,8 - COMBOBOX IDC_COMBO1,71,72,111,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_RENDERER,71,72,111,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Shader:",IDC_STATIC,7,89,26,8 - COMBOBOX IDC_COMBO4,71,87,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_SHADER,71,87,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Interlacing (F5):",IDC_STATIC,7,105,53,8 - COMBOBOX IDC_COMBO2,71,102,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_INTERLACE,71,102,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Aspect Ratio (F6):",IDC_STATIC,7,120,60,8 - COMBOBOX IDC_COMBO5,71,117,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_ASPECTRATIO,71,117,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "D3D internal res:",IDC_STATIC,7,135,55,8 EDITTEXT IDC_EDIT1,71,132,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SPIN1,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,135,11,14 + CONTROL "",IDC_RESX,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,135,11,14 EDITTEXT IDC_EDIT2,109,132,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SPIN2,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,133,135,11,14 - CONTROL "Native",IDC_CHECK1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,149,134,33,10 + CONTROL "",IDC_RESY,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,133,135,11,14 + CONTROL "Native",IDC_NATIVERES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,149,134,33,10 LTEXT "SW rend. threads:",IDC_STATIC,7,149,60,8 EDITTEXT IDC_EDIT3,71,147,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SPIN3,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,150,11,14 - CONTROL "Texture filtering",IDC_CHECK4,"Button",BS_AUTO3STATE | WS_TABSTOP,7,167,67,10 - CONTROL "Logarithmic Z",IDC_CHECK5,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,168,58,10 - CONTROL "Wait vsync",IDC_CHECK2,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,180,51,10 - CONTROL "Alpha correction (FBA)",IDC_CHECK7,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,181,102,10 - CONTROL "Edge anti-aliasing (AA1, sw-mode only)",IDC_CHECK8, + CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,150,11,14 + CONTROL "Texture filtering",IDC_FILTER,"Button",BS_AUTO3STATE | WS_TABSTOP,7,167,67,10 + CONTROL "Logarithmic Z",IDC_LOGZ,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,168,58,10 + CONTROL "Wait vsync",IDC_VSYNC,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,180,51,10 + CONTROL "Alpha correction (FBA)",IDC_FBA,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,181,102,10 + CONTROL "Edge anti-aliasing (AA1, sw-mode only)",IDC_AA1, "Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,194,141,10 - CONTROL "Enable output merger blur effect",IDC_CHECK9,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,208,121,10 + CONTROL "Enable output merger blur effect",IDC_BLUR,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,208,121,10 DEFPUSHBUTTON "OK",IDOK,43,227,50,14 PUSHBUTTON "Cancel",IDCANCEL,96,227,50,14 END @@ -124,13 +124,13 @@ STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_CAPTION | WS_SYSM CAPTION "Capture settings" FONT 8, "MS Shell Dlg", 400, 0, 0x1 BEGIN - EDITTEXT IDC_EDIT1,7,7,207,14,ES_AUTOHSCROLL - PUSHBUTTON "Browse...",IDC_BUTTON1,222,7,50,14 - COMBOBOX IDC_COMBO1,7,27,207,122,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - PUSHBUTTON "Config...",IDC_BUTTON2,222,26,50,14 + EDITTEXT IDC_FILENAME,7,7,207,14,ES_AUTOHSCROLL + PUSHBUTTON "Browse...",IDC_BROWSE,222,7,50,14 + COMBOBOX IDC_CODECS,7,27,207,122,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + PUSHBUTTON "Config...",IDC_CONFIGURE,222,26,50,14 LTEXT "Size:",IDC_STATIC,6,50,16,8 - EDITTEXT IDC_EDIT2,30,47,31,14,ES_RIGHT | ES_AUTOHSCROLL | ES_NUMBER - EDITTEXT IDC_EDIT4,64,47,31,14,ES_RIGHT | ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_WIDTH,30,47,31,14,ES_RIGHT | ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_HEIGHT,64,47,31,14,ES_RIGHT | ES_AUTOHSCROLL | ES_NUMBER PUSHBUTTON "Cancel",IDCANCEL,169,47,50,14 DEFPUSHBUTTON "OK",IDOK,221,47,50,14 END @@ -142,25 +142,25 @@ FONT 8, "MS Shell Dlg", 400, 0, 0x1 BEGIN CONTROL 2021,IDC_LOGO9,"Static",SS_BITMAP,7,7,175,44 LTEXT "Resolution:",IDC_STATIC,7,59,37,8 - COMBOBOX IDC_COMBO3,78,57,104,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_RESOLUTION,78,57,104,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Renderer:",IDC_STATIC,7,74,34,8 - COMBOBOX IDC_COMBO1,78,72,104,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_RENDERER,78,72,104,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Shader:",IDC_STATIC,7,89,26,8 - COMBOBOX IDC_COMBO4,78,87,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_SHADER,78,87,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Texture Filter (Del):",IDC_STATIC,7,105,64,8 - COMBOBOX IDC_COMBO2,78,102,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_FILTER,78,102,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Dithering (End):",IDC_STATIC,7,120,52,8 - COMBOBOX IDC_COMBO5,78,117,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_DITHERING,78,117,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Aspect Ratio (PgDn):",IDC_STATIC,7,135,68,8 - COMBOBOX IDC_COMBO6,78,132,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_ASPECTRATIO,78,132,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Rendering Threads:",IDC_STATIC,7,165,64,8 EDITTEXT IDC_EDIT3,78,163,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SPIN3,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,169,11,14 + CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,169,11,14 DEFPUSHBUTTON "OK",IDOK,43,214,50,14 PUSHBUTTON "Cancel",IDCANCEL,96,214,50,14 CONTROL 2022,IDC_LOGO10,"Static",SS_BITMAP,7,7,175,44 LTEXT "Internal Resolution:",IDC_STATIC,7,150,64,8 - COMBOBOX IDC_COMBO7,78,147,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + COMBOBOX IDC_SCALE,78,147,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP END diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index bca134a194..e71c4955e4 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -23,7 +23,7 @@ Name="Debug|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;.\vsprops\debug.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > @@ -279,7 +279,7 @@ Name="Debug SSE2|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;.\vsprops\debug.vsprops;.\vsprops\sse2.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > @@ -532,7 +532,7 @@ Name="Release SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;.\vsprops\release.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" WholeProgramOptimization="1" > @@ -659,7 +659,7 @@ Name="Debug SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;.\vsprops\debug.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > @@ -1155,6 +1155,10 @@ RelativePath=".\GSDeviceNull.cpp" > + + @@ -1705,6 +1709,10 @@ RelativePath=".\GSDeviceNull.h" > + + diff --git a/plugins/GSdx/resource.h b/plugins/GSdx/resource.h index 48d4b60446..c5f074136f 100644 --- a/plugins/GSdx/resource.h +++ b/plugins/GSdx/resource.h @@ -2,40 +2,44 @@ // Microsoft Visual C++ generated include file. // Used by GSdx.rc // -#define IDC_CHECK1 2001 -#define IDC_CHECK2 2002 +#define IDC_NATIVERES 2001 +#define IDC_VSYNC 2002 #define IDC_CHECK3 2003 -#define IDC_CHECK5 2004 +#define IDC_LOGZ 2004 #define IDC_CHECK6 2005 -#define IDC_COMBO1 2006 -#define IDC_COMBO3 2007 -#define IDC_COMBO4 2008 +#define IDC_CODECS 2006 +#define IDC_RESOLUTION 2007 +#define IDC_SHADER 2008 #define IDC_EDIT1 2009 #define IDC_EDIT2 2010 -#define IDC_BUTTON1 2011 -#define IDC_CHECK8 2011 -#define IDC_BUTTON2 2012 +#define IDC_AA1 2011 #define IDC_EDIT3 2012 #define IDC_CUSTOM1 2013 #define IDC_CHECK4 2014 -#define IDC_COMBO2 2015 -#define IDC_COMBO5 2016 +#define IDC_FILTER 2015 +#define IDC_DITHERING 2016 #define IDC_RADIO1 2017 -#define IDC_COMBO6 2017 -#define IDC_SPIN1 2018 -#define IDC_COMBO7 2018 -#define IDC_SPIN2 2019 +#define IDC_RESX 2018 +#define IDC_RESY 2019 #define IDD_CONFIG 2020 -#define IDC_SPIN3 2020 #define IDB_LOGO9 2021 #define IDB_LOGO10 2022 -#define IDC_CHECK7 2023 +#define IDC_FBA 2023 #define IDC_LOGO9 2024 #define IDC_LOGO10 2025 #define IDD_CAPTURE 2026 -#define IDC_EDIT4 2027 #define IDD_GPUCONFIG 2027 -#define IDC_CHECK9 2028 +#define IDC_BLUR 2028 +#define IDC_RENDERER 2029 +#define IDC_INTERLACE 2030 +#define IDC_ASPECTRATIO 2031 +#define IDC_SWTHREADS 2032 +#define IDC_SCALE 2033 +#define IDC_BROWSE 2034 +#define IDC_FILENAME 2035 +#define IDC_WIDTH 2036 +#define IDC_HEIGHT 2037 +#define IDC_CONFIGURE 2038 #define IDR_CONVERT9_FX 10000 #define IDR_TFX9_FX 10001 #define IDR_MERGE9_FX 10002 @@ -51,7 +55,7 @@ #ifndef APSTUDIO_READONLY_SYMBOLS #define _APS_NEXT_RESOURCE_VALUE 10009 #define _APS_NEXT_COMMAND_VALUE 32771 -#define _APS_NEXT_CONTROL_VALUE 2029 +#define _APS_NEXT_CONTROL_VALUE 2039 #define _APS_NEXT_SYMED_VALUE 5000 #endif #endif diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index f982943160..e50189b796 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -4,39 +4,41 @@ #pragma once -#pragma warning(disable: 4996 4995 4324 4100 4101) +#pragma warning(disable: 4996 4995 4324 4100 4101 4201) #ifdef _WINDOWS -#ifndef VC_EXTRALEAN -#define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers -#endif +// The following macros define the minimum required platform. The minimum required platform +// is the earliest version of Windows, Internet Explorer etc. that has the necessary features to run +// your application. The macros work by enabling all features available on platform versions up to and +// including the version specified. // Modify the following defines if you have to target a platform prior to the ones specified below. // Refer to MSDN for the latest info on corresponding values for different platforms. -#ifndef WINVER // Allow use of features specific to Windows 95 and Windows NT 4 or later. -#define WINVER 0x0510 // Change this to the appropriate value to target Windows 98 and Windows 2000 or later. + +#ifndef WINVER // Specifies that the minimum required platform is Windows Vista. +#define WINVER 0x0600 // Change this to the appropriate value to target other versions of Windows. #endif -#ifndef _WIN32_WINNT // Allow use of features specific to Windows NT 4 or later. -#define _WIN32_WINNT 0x0400 // Change this to the appropriate value to target Windows 2000 or later. -#endif +#ifndef _WIN32_WINNT // Specifies that the minimum required platform is Windows Vista. +#define _WIN32_WINNT 0x0600 // Change this to the appropriate value to target other versions of Windows. +#endif -#ifndef _WIN32_WINDOWS // Allow use of features specific to Windows 98 or later. +#ifndef _WIN32_WINDOWS // Specifies that the minimum required platform is Windows 98. #define _WIN32_WINDOWS 0x0410 // Change this to the appropriate value to target Windows Me or later. #endif -#ifndef _WIN32_IE // Allow use of features specific to IE 4.0 or later. -#define _WIN32_IE 0x0400 // Change this to the appropriate value to target IE 5.0 or later. +#ifndef _WIN32_IE // Specifies that the minimum required platform is Internet Explorer 7.0. +#define _WIN32_IE 0x0700 // Change this to the appropriate value to target other versions of IE. #endif -#define _ATL_CSTRING_EXPLICIT_CONSTRUCTORS // some CString constructors will be explicit +#define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers -#include // MFC core and standard components -//#include // MFC extensions -#ifndef _AFX_NO_AFXCMN_SUPPORT -#include // MFC support for Windows Common Controls -#endif // _AFX_NO_AFXCMN_SUPPORT +#include +#include +#include +#include +#include #endif @@ -44,6 +46,7 @@ #include #include +#include #include #include @@ -77,21 +80,25 @@ typedef signed long long int64; #define ALIGN_STACK(n) __declspec(align(n)) int __dummy; #ifndef RESTRICT -#ifdef __INTEL_COMPILER - #define RESTRICT restrict -#elif _MSC_VER >= 1400 // TODO: gcc - #define RESTRICT __restrict -#else - #define RESTRICT + #ifdef __INTEL_COMPILER + #define RESTRICT restrict + #elif _MSC_VER >= 1400 // TODO: gcc + #define RESTRICT __restrict + #else + #define RESTRICT + #endif #endif + +#if defined(_DEBUG) && defined(_MSC_VER) + #define ASSERT assert +#else + #define ASSERT(exp) ((void)0) #endif #ifdef __x86_64__ -#define _M_AMD64 + #define _M_AMD64 #endif -extern "C" uint64 __rdtsc(); // TODO: gcc - // directx #include From 02167e90b3b9e36c9e131cd7bd2fdb0150897ff6 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 3 Jun 2009 19:22:50 +0000 Subject: [PATCH 173/296] GSdx: Streamlined several instances of CComPtr use in the Shader caches, as it was causing general slowdowns due to internal reference counters. Gives fairly significant speedups (6-15%) across most games and both DX9 and DX10 alike. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1311 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTextureFX10.cpp | 64 +++++++++++++--------------------- plugins/GSdx/GSTextureFX9.cpp | 57 +++++++++++------------------- 2 files changed, 44 insertions(+), 77 deletions(-) diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index d03cf0ac0d..6afc9b114d 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -100,15 +100,9 @@ bool GSTextureFX10::SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMI bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { - CComPtr vs; + hash_map >::const_iterator i = m_vs.find(sel); - hash_map >::iterator i = m_vs.find(sel); - - if(i != m_vs.end()) - { - vs = (*i).second; - } - else + if(i == m_vs.end()) { string str[5]; @@ -139,6 +133,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) }; CComPtr il; + CComPtr vs; m_dev->CompileShader(IDR_TFX10_FX, "vs_main", macro, &vs, layout, countof(layout), &il); @@ -148,6 +143,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) } m_vs[sel] = vs; + i = m_vs.find( sel ); } if(m_vs_cb_cache.Update(cb)) @@ -155,7 +151,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) (*m_dev)->UpdateSubresource(m_vs_cb, 0, NULL, cb, 0, 0); } - m_dev->VSSetShader(vs, m_vs_cb); + m_dev->VSSetShader((*i).second, m_vs_cb); return true; } @@ -164,11 +160,11 @@ bool GSTextureFX10::SetupGS(GSSelector sel) { HRESULT hr; - CComPtr gs; + ID3D10GeometryShader* gs = NULL; if(sel.prim > 0 && (sel.iip == 0 || sel.prim == 3)) // geometry shader works in every case, but not needed { - hash_map >::iterator i = m_gs.find(sel); + hash_map >::const_iterator i = m_gs.find(sel); if(i != m_gs.end()) { @@ -212,15 +208,9 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl { HRESULT hr; - CComPtr ps; + hash_map >::const_iterator i = m_ps.find(sel); - hash_map >::iterator i = m_ps.find(sel); - - if(i != m_ps.end()) - { - ps = (*i).second; - } - else + if(i == m_ps.end()) { string str[13]; @@ -256,9 +246,11 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl {NULL, NULL}, }; + CComPtr ps; hr = m_dev->CompileShader(IDR_TFX10_FX, "ps_main", macro, &ps); m_ps[sel] = ps; + i = m_ps.find(sel); } if(m_ps_cb_cache.Update(cb)) @@ -266,9 +258,9 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl (*m_dev)->UpdateSubresource(m_ps_cb, 0, NULL, cb, 0, 0); } - m_dev->PSSetShader(ps, m_ps_cb); + m_dev->PSSetShader((*i).second, m_ps_cb); - CComPtr ss0, ss1; + ID3D10SamplerState* ss0=NULL, *ss1=NULL; if(sel.tfx != 4) { @@ -277,7 +269,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl ssel.min = ssel.mag = 0; } - hash_map >::iterator i = m_ps_ss.find(ssel); + hash_map >::const_iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) { @@ -333,15 +325,9 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, { HRESULT hr; - CComPtr dss; + hash_map >::const_iterator i = m_om_dss.find(dssel); - hash_map >::iterator i = m_om_dss.find(dssel); - - if(i != m_om_dss.end()) - { - dss = (*i).second; - } - else + if(i == m_om_dss.end()) { D3D10_DEPTH_STENCIL_DESC dsd; @@ -377,22 +363,18 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, dsd.DepthFunc = ztst[dssel.ztst]; } + CComPtr dss; hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); m_om_dss[dssel] = dss; + i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState(dss, 1); + m_dev->OMSetDepthStencilState((*i).second, 1); - CComPtr bs; + hash_map >::const_iterator j = m_om_bs.find(bsel); - hash_map >::iterator j = m_om_bs.find(bsel); - - if(j != m_om_bs.end()) - { - bs = (*j).second; - } - else + if(j == m_om_bs.end()) { D3D10_BLEND_DESC bd; @@ -520,12 +502,14 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(bsel.wb) bd.RenderTargetWriteMask[0] |= D3D10_COLOR_WRITE_ENABLE_BLUE; if(bsel.wa) bd.RenderTargetWriteMask[0] |= D3D10_COLOR_WRITE_ENABLE_ALPHA; + CComPtr bs; hr = (*m_dev)->CreateBlendState(&bd, &bs); m_om_bs[bsel] = bs; + j = m_om_bs.find(bsel); } - m_dev->OMSetBlendState(bs, bf); + m_dev->OMSetBlendState((*j).second, bf); } void GSTextureFX10::Draw() diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index db8c34c670..be2aaeec2d 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -95,15 +95,9 @@ bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVET bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { - CComPtr vs; + hash_map >::const_iterator i = m_vs.find(sel); - hash_map >::iterator i = m_vs.find(sel); - - if(i != m_vs.end()) - { - vs = (*i).second; - } - else + if(i == m_vs.end()) { string str[4]; @@ -131,6 +125,7 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) }; CComPtr il; + CComPtr vs; m_dev->CompileShader(IDR_TFX9_FX, "vs_main", macro, &vs, layout, countof(layout), &il); @@ -140,9 +135,10 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) } m_vs[sel] = vs; + i = m_vs.find( sel ); } - m_dev->VSSetShader(vs, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); + m_dev->VSSetShader( (*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); return true; } @@ -185,15 +181,9 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample if(sel.wmt == 3) sel.wmt = 0; } - CComPtr ps; + hash_map >::const_iterator i = m_ps.find(sel); - hash_map >::iterator i = m_ps.find(sel); - - if(i != m_ps.end()) - { - ps = (*i).second; - } - else + if(i == m_ps.end()) { string str[12]; @@ -227,12 +217,14 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample {NULL, NULL}, }; + CComPtr ps; hr = m_dev->CompileShader(IDR_TFX9_FX, "ps_main", macro, &ps); m_ps[sel] = ps; + i = m_ps.find(sel); } - m_dev->PSSetShader(ps, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); + m_dev->PSSetShader((*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); Direct3DSamplerState9* ss = NULL; @@ -243,7 +235,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample ssel.min = ssel.mag = 0; } - hash_map::iterator i = m_ps_ss.find(ssel); + hash_map::const_iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) { @@ -286,13 +278,9 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, { Direct3DDepthStencilState9* dss = NULL; - hash_map::iterator i = m_om_dss.find(dssel); + hash_map::const_iterator i = m_om_dss.find(dssel); - if(i != m_om_dss.end()) - { - dss = (*i).second; - } - else + if(i == m_om_dss.end()) { dss = new Direct3DDepthStencilState9(); @@ -325,22 +313,16 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, } m_om_dss[dssel] = dss; + i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState(dss, 3); + m_dev->OMSetDepthStencilState((*i).second, 3); - Direct3DBlendState9* bs = NULL; - - hash_map::iterator j = m_om_bs.find(bsel); + hash_map::const_iterator j = m_om_bs.find(bsel); - if(j != m_om_bs.end()) + if(j == m_om_bs.end()) { - bs = (*j).second; - } - else - { - bs = new Direct3DBlendState9(); - + Direct3DBlendState9* bs = new Direct3DBlendState9(); memset(bs, 0, sizeof(*bs)); bs->BlendEnable = bsel.abe; @@ -466,9 +448,10 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(bsel.wa) bs->RenderTargetWriteMask |= D3DCOLORWRITEENABLE_ALPHA; m_om_bs[bsel] = bs; + j = m_om_bs.find(bsel); } - m_dev->OMSetBlendState(bs, 0x010101 * bf); + m_dev->OMSetBlendState((*j).second, 0x010101 * bf); } void GSTextureFX9::Draw() From db8feac8c3671c3e29b0403cb495e589f86e53de Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 3 Jun 2009 22:45:28 +0000 Subject: [PATCH 174/296] GSdx: the BeginScene/EndScene optimization for dx9 (Issue 256) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1312 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDevice.h | 16 ++++++++-------- plugins/GSdx/GSDevice9.cpp | 8 ++++++-- plugins/GSdx/GSTextureFX10.cpp | 9 ++++++++- plugins/GSdx/GSTextureFX9.cpp | 10 ++++++++-- 4 files changed, 30 insertions(+), 13 deletions(-) diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index a9229c8455..41502653f8 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -78,15 +78,15 @@ public: virtual bool IsLost() {return false;} virtual void Present(const GSVector4i& r, int shader); - virtual void Flip() {}; + virtual void Flip() {} - virtual void BeginScene() {}; - virtual void EndScene() {}; + virtual void BeginScene() {} + virtual void EndScene() {} - virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {}; - virtual void ClearRenderTarget(GSTexture* t, uint32 c) {}; - virtual void ClearDepth(GSTexture* t, float c) {}; - virtual void ClearStencil(GSTexture* t, uint8 c) {}; + virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {} + virtual void ClearRenderTarget(GSTexture* t, uint32 c) {} + virtual void ClearDepth(GSTexture* t, float c) {} + virtual void ClearStencil(GSTexture* t, uint8 c) {} virtual GSTexture* CreateRenderTarget(int w, int h, int format = 0); virtual GSTexture* CreateDepthStencil(int w, int h, int format = 0); @@ -105,5 +105,5 @@ public: void Interlace(const GSVector2i& ds, int field, int mode, float yoffset); virtual void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) {} - virtual void OMSetRenderTargets(GSTexture* rt, GSTexture* ds) {}; + virtual void OMSetRenderTargets(GSTexture* rt, GSTexture* ds) {} }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 443b1e56b5..af728d7b42 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -349,6 +349,8 @@ bool GSDevice9::IsLost() void GSDevice9::Flip() { + m_dev->EndScene(); + if(m_swapchain) { m_swapchain->Present(NULL, NULL, NULL, NULL, 0); @@ -357,11 +359,13 @@ void GSDevice9::Flip() { m_dev->Present(NULL, NULL, NULL, NULL); } + + m_dev->BeginScene(); } void GSDevice9::BeginScene() { - m_dev->BeginScene(); + // m_dev->BeginScene(); } void GSDevice9::DrawPrimitive() @@ -393,7 +397,7 @@ void GSDevice9::DrawPrimitive() void GSDevice9::EndScene() { - m_dev->EndScene(); + // m_dev->EndScene(); m_vertices.start += m_vertices.count; m_vertices.count = 0; diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 6afc9b114d..dd223a508e 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -143,7 +143,8 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) } m_vs[sel] = vs; - i = m_vs.find( sel ); + + i = m_vs.find(sel); } if(m_vs_cb_cache.Update(cb)) @@ -247,9 +248,11 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl }; CComPtr ps; + hr = m_dev->CompileShader(IDR_TFX10_FX, "ps_main", macro, &ps); m_ps[sel] = ps; + i = m_ps.find(sel); } @@ -364,9 +367,11 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, } CComPtr dss; + hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); m_om_dss[dssel] = dss; + i = m_om_dss.find(dssel); } @@ -503,9 +508,11 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(bsel.wa) bd.RenderTargetWriteMask[0] |= D3D10_COLOR_WRITE_ENABLE_ALPHA; CComPtr bs; + hr = (*m_dev)->CreateBlendState(&bd, &bs); m_om_bs[bsel] = bs; + j = m_om_bs.find(bsel); } diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index be2aaeec2d..5a28c42b3e 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -135,10 +135,11 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) } m_vs[sel] = vs; - i = m_vs.find( sel ); + + i = m_vs.find(sel); } - m_dev->VSSetShader( (*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); + m_dev->VSSetShader((*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); return true; } @@ -218,9 +219,11 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample }; CComPtr ps; + hr = m_dev->CompileShader(IDR_TFX9_FX, "ps_main", macro, &ps); m_ps[sel] = ps; + i = m_ps.find(sel); } @@ -313,6 +316,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, } m_om_dss[dssel] = dss; + i = m_om_dss.find(dssel); } @@ -323,6 +327,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(j == m_om_bs.end()) { Direct3DBlendState9* bs = new Direct3DBlendState9(); + memset(bs, 0, sizeof(*bs)); bs->BlendEnable = bsel.abe; @@ -448,6 +453,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, if(bsel.wa) bs->RenderTargetWriteMask |= D3DCOLORWRITEENABLE_ALPHA; m_om_bs[bsel] = bs; + j = m_om_bs.find(bsel); } From 07a4eb3491dfa76daed239acba5591c868e24a05 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 3 Jun 2009 23:51:23 +0000 Subject: [PATCH 175/296] microVU/pcsx2: - Removed flaghack1 since it didn't do much over flaghack2 (except break games :p). Flaghack2 is now renamed to "Status Flag Hack", and so-far we haven't found a single game it breaks, so compatibility is high. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1313 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 3 +-- pcsx2/Linux/HacksDlg.cpp | 16 +++++++--------- pcsx2/Linux/Pref.cpp | 3 +-- pcsx2/windows/HacksDlg.cpp | 18 ++++++++---------- pcsx2/windows/ini.cpp | 3 +-- pcsx2/windows/pcsx2.rc | 17 ++++++++--------- pcsx2/x86/microVU.cpp | 2 -- pcsx2/x86/microVU.h | 1 - pcsx2/x86/microVU_Compile.inl | 8 +++----- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_Lower.inl | 21 ++++++++++----------- pcsx2/x86/microVU_Misc.h | 7 +------ pcsx2/x86/microVU_Misc.inl | 18 ------------------ 13 files changed, 41 insertions(+), 78 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index 6e48fb03bf..05a693f28b 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -171,8 +171,7 @@ public: bool INTCSTATSlow; bool IdleLoopFF; int VUCycleSteal; - bool vuFlagHack1; - bool vuFlagHack2; + bool vuFlagHack; bool vuMinMax; bool ESCExits; // this is a hack!? } Hacks; diff --git a/pcsx2/Linux/HacksDlg.cpp b/pcsx2/Linux/HacksDlg.cpp index 470425cbcf..e9ebc80936 100644 --- a/pcsx2/Linux/HacksDlg.cpp +++ b/pcsx2/Linux/HacksDlg.cpp @@ -75,13 +75,12 @@ void on_Speed_Hacks(GtkMenuItem *menuitem, gpointer user_data) GtkRange *vuScale = GTK_RANGE(lookup_widget(SpeedHacksDlg, "VUCycleHackScale")); GtkRange *eeScale = GTK_RANGE(lookup_widget(SpeedHacksDlg, "EECycleHackScale")); - set_checked(SpeedHacksDlg, "check_iop_cycle_rate", Config.Hacks.IOPCycleDouble); - set_checked(SpeedHacksDlg, "check_wait_cycles_sync_hack", Config.Hacks.WaitCycleExt); - set_checked(SpeedHacksDlg, "check_intc_sync_hack", Config.Hacks.INTCSTATSlow); - set_checked(SpeedHacksDlg, "check_idle_loop_fastforward", Config.Hacks.IdleLoopFF); - set_checked(SpeedHacksDlg, "check_microvu_flag_hack_1", Config.Hacks.vuFlagHack1); - set_checked(SpeedHacksDlg, "check_microvu_flag_hack_2", Config.Hacks.vuFlagHack2); - set_checked(SpeedHacksDlg, "check_microvu_min_max_hack", Config.Hacks.vuMinMax); + set_checked(SpeedHacksDlg, "check_iop_cycle_rate", Config.Hacks.IOPCycleDouble); + set_checked(SpeedHacksDlg, "check_wait_cycles_sync_hack", Config.Hacks.WaitCycleExt); + set_checked(SpeedHacksDlg, "check_intc_sync_hack", Config.Hacks.INTCSTATSlow); + set_checked(SpeedHacksDlg, "check_idle_loop_fastforward", Config.Hacks.IdleLoopFF); + set_checked(SpeedHacksDlg, "check_microvu_flag_hack", Config.Hacks.vuFlagHack); + set_checked(SpeedHacksDlg, "check_microvu_min_max_hack", Config.Hacks.vuMinMax); gtk_range_set_value(vuScale, Config.Hacks.VUCycleSteal); on_vu_slider_changed(vuScale, NULL); @@ -104,8 +103,7 @@ void on_Speed_Hack_OK(GtkButton *button, gpointer user_data) newhacks.WaitCycleExt = is_checked(SpeedHacksDlg, "check_wait_cycles_sync_hack"); newhacks.INTCSTATSlow = is_checked(SpeedHacksDlg, "check_intc_sync_hack"); newhacks.IdleLoopFF = is_checked(SpeedHacksDlg, "check_idle_loop_fastforward"); - newhacks.vuFlagHack1 = is_checked(SpeedHacksDlg, "check_microvu_flag_hack_1"); - newhacks.vuFlagHack2 = is_checked(SpeedHacksDlg, "check_microvu_flag_hack_2"); + newhacks.vuFlagHack = is_checked(SpeedHacksDlg, "check_microvu_flag_hack"); newhacks.vuMinMax = is_checked(SpeedHacksDlg, "check_microvu_min_max_hack"); newhacks.VUCycleSteal = gtk_range_get_value(GTK_RANGE(lookup_widget(SpeedHacksDlg, "VUCycleHackScale"))); diff --git a/pcsx2/Linux/Pref.cpp b/pcsx2/Linux/Pref.cpp index a6c9a79fdd..ff06af2b48 100644 --- a/pcsx2/Linux/Pref.cpp +++ b/pcsx2/Linux/Pref.cpp @@ -108,8 +108,7 @@ int LoadConfig() GetValuel("WaitCycleExt", Config.Hacks.WaitCycleExt); GetValuel("INTCSTATSlow", Config.Hacks.INTCSTATSlow); GetValuel("VUCycleSteal", Config.Hacks.VUCycleSteal); - GetValuel("vuFlagHack1", Config.Hacks.vuFlagHack1); - GetValuel("vuFlagHack2", Config.Hacks.vuFlagHack2); + GetValuel("vuFlagHack", Config.Hacks.vuFlagHack); GetValuel("vuMinMax", Config.Hacks.vuMinMax); GetValuel("IdleLoopFF", Config.Hacks.IdleLoopFF); GetValuel("ESCExits", Config.Hacks.ESCExits); diff --git a/pcsx2/windows/HacksDlg.cpp b/pcsx2/windows/HacksDlg.cpp index 709a0d13e5..8f7bfe1a24 100644 --- a/pcsx2/windows/HacksDlg.cpp +++ b/pcsx2/windows/HacksDlg.cpp @@ -43,14 +43,13 @@ BOOL APIENTRY HacksProc(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) CheckRadioButton( hDlg, IDC_EESYNC_DEFAULT, IDC_EESYNC3, IDC_EESYNC_DEFAULT + Config.Hacks.EECycleRate ); - if(Config.Hacks.IOPCycleDouble) CheckDlgButton(hDlg, IDC_IOPSYNC, TRUE); - if(Config.Hacks.WaitCycleExt) CheckDlgButton(hDlg, IDC_WAITCYCLES, TRUE); - if(Config.Hacks.INTCSTATSlow) CheckDlgButton(hDlg, IDC_INTCSTATHACK, TRUE); - if(Config.Hacks.IdleLoopFF) CheckDlgButton(hDlg, IDC_IDLELOOPFF, TRUE); - if(Config.Hacks.ESCExits) CheckDlgButton(hDlg, IDC_ESCHACK, TRUE); - if(Config.Hacks.vuFlagHack1) CheckDlgButton(hDlg, IDC_VUHACK1, TRUE); - if(Config.Hacks.vuFlagHack2) CheckDlgButton(hDlg, IDC_VUHACK2, TRUE); - if(Config.Hacks.vuMinMax) CheckDlgButton(hDlg, IDC_VUHACK3, TRUE); + if(Config.Hacks.IOPCycleDouble) CheckDlgButton(hDlg, IDC_IOPSYNC, TRUE); + if(Config.Hacks.WaitCycleExt) CheckDlgButton(hDlg, IDC_WAITCYCLES, TRUE); + if(Config.Hacks.INTCSTATSlow) CheckDlgButton(hDlg, IDC_INTCSTATHACK, TRUE); + if(Config.Hacks.IdleLoopFF) CheckDlgButton(hDlg, IDC_IDLELOOPFF, TRUE); + if(Config.Hacks.ESCExits) CheckDlgButton(hDlg, IDC_ESCHACK, TRUE); + if(Config.Hacks.vuFlagHack) CheckDlgButton(hDlg, IDC_VUHACK1, TRUE); + if(Config.Hacks.vuMinMax) CheckDlgButton(hDlg, IDC_VUHACK3, TRUE); SendDlgItemMessage(hDlg, IDC_VUCYCLE, TBM_SETRANGE, TRUE, MAKELONG(0, 4)); CheckVUCycleHack(hDlg, Config.Hacks.VUCycleSteal); @@ -93,8 +92,7 @@ BOOL APIENTRY HacksProc(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) newhacks.WaitCycleExt = !!IsDlgButtonChecked(hDlg, IDC_WAITCYCLES); newhacks.INTCSTATSlow = !!IsDlgButtonChecked(hDlg, IDC_INTCSTATHACK); newhacks.ESCExits = !!IsDlgButtonChecked(hDlg, IDC_ESCHACK); - newhacks.vuFlagHack1 = !!IsDlgButtonChecked(hDlg, IDC_VUHACK1); - newhacks.vuFlagHack2 = !!IsDlgButtonChecked(hDlg, IDC_VUHACK2); + newhacks.vuFlagHack = !!IsDlgButtonChecked(hDlg, IDC_VUHACK1); newhacks.vuMinMax = !!IsDlgButtonChecked(hDlg, IDC_VUHACK3); newhacks.IdleLoopFF = !!IsDlgButtonChecked(hDlg, IDC_IDLELOOPFF); newhacks.VUCycleSteal = SendDlgItemMessage(hDlg, IDC_VUCYCLE, TBM_GETPOS, 0, 0); diff --git a/pcsx2/windows/ini.cpp b/pcsx2/windows/ini.cpp index 93af72fd67..2a80ad212a 100644 --- a/pcsx2/windows/ini.cpp +++ b/pcsx2/windows/ini.cpp @@ -231,8 +231,7 @@ void IniFile::DoConfig( PcsxConfig& Conf ) Entry("WaitCycleExt", Config.Hacks.WaitCycleExt); Entry("INTCSTATSlow", Config.Hacks.INTCSTATSlow); Entry("VUCycleSteal", Config.Hacks.VUCycleSteal); - Entry("vuFlagHack1", Config.Hacks.vuFlagHack1); - Entry("vuFlagHack2", Config.Hacks.vuFlagHack2); + Entry("vuFlagHack", Config.Hacks.vuFlagHack); Entry("vuMinMax", Config.Hacks.vuMinMax); Entry("IdleLoopFF", Config.Hacks.IdleLoopFF); if (Conf.Hacks.VUCycleSteal < 0 || Conf.Hacks.VUCycleSteal > 4) diff --git a/pcsx2/windows/pcsx2.rc b/pcsx2/windows/pcsx2.rc index ea89271c47..aeec13377b 100644 --- a/pcsx2/windows/pcsx2.rc +++ b/pcsx2/windows/pcsx2.rc @@ -369,20 +369,18 @@ BEGIN LTEXT "Moderate speedup and works well with most games.",IDC_STATIC,25,90,129,19 CONTROL "INTC Sync Hack",IDC_INTCSTATHACK,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,43,127,10 LTEXT "Huge speedup in many games, and a pretty high compatibility rate (some games still work better with EE sync hacks).",IDC_STATIC,200,55,140,28 - CONTROL "",IDC_VUCYCLE,"msctls_trackbar32",TBS_BOTH | TBS_NOTICKS | WS_TABSTOP,372,152,36,15 - LTEXT "This space intentionally left blank",IDC_VUCYCLEDESC,383,171,142,26 - LTEXT "VU Cycle Stealing (experimental)",IDC_STATIC,408,157,105,8 + CONTROL "",IDC_VUCYCLE,"msctls_trackbar32",TBS_BOTH | TBS_NOTICKS | WS_TABSTOP,372,141,36,15 + LTEXT "This space intentionally left blank",IDC_VUCYCLEDESC,383,160,142,26 + LTEXT "VU Cycle Stealing (experimental)",IDC_STATIC,408,146,105,8 CONTROL "Idle Loop Fast-Forward (experimental)",IDC_IDLELOOPFF, "Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,179,139,10 LTEXT "Speedup for a few games, including FFX with no known side effects.",IDC_STATIC,200,193,144,23 GROUPBOX "EmotionEngine (EE) Misc Hacks",IDC_STATIC,177,26,179,192 GROUPBOX "Vertex Unit (VU) Hacks",IDC_STATIC,363,26,173,173 - CONTROL "Flag Hack 1 (microVU only)",IDC_VUHACK1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,42,155,12 - CONTROL "Flag Hack 2 (microVU only)",IDC_VUHACK2,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,79,154,12 - LTEXT "Big Speedup and moderately compatible. May cause SPS and Infinite loop problems.",IDC_STATIC,386,56,142,19 - LTEXT "Big Speedup and moderately compatible. SuperVU does something similar by default.",IDC_STATIC,385,92,144,20 - CONTROL "Min / Max Hack (microVU only)",IDC_VUHACK3,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,115,156,12 - LTEXT "Small Speedup. May cause SPS, hanging, and missing geometry.",IDC_STATIC,386,130,143,19 + CONTROL "Status Flag Hack (microVU only)",IDC_VUHACK1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,42,155,12 + LTEXT "Big Speedup and high compatibility.\nSuperVU does something similar by default.",IDC_STATIC,383,55,144,20 + CONTROL "Min / Max Hack (microVU only)",IDC_VUHACK3,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,85,156,12 + LTEXT "Small Speedup. May cause SPS, hanging, graphical corruption, and missing geometry.",IDC_STATIC,383,99,143,28 END @@ -409,6 +407,7 @@ BEGIN VERTGUIDE, 187 VERTGUIDE, 200 VERTGUIDE, 372 + VERTGUIDE, 383 TOPMARGIN, 7 BOTTOMMARGIN, 222 HORZGUIDE, 26 diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 4adcf883a9..85980e3413 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -133,7 +133,6 @@ microVUf(void) mVUclearProg(int progIndex) { microVU* mVU = mVUx; mVU->prog.prog[progIndex].used = 1; mVU->prog.prog[progIndex].last_used = 3; - mVU->prog.prog[progIndex].sFlagHack = 0; mVU->prog.prog[progIndex].range[0] = -1; mVU->prog.prog[progIndex].range[1] = -1; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; @@ -148,7 +147,6 @@ microVUf(void) mVUcacheProg(int progIndex) { microVU* mVU = mVUx; memcpy_fast(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize); mVUdumpProg(progIndex); - mVUcheckSflag(mVU, progIndex); } // Finds the least used program, (if program list full clears and returns an old program; if not-full, returns free program) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 5b3f640b50..d7cc0d114d 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -74,7 +74,6 @@ struct microProgram { u32 data[progSize]; u32 used; // Number of times its been used u32 last_used; // Counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vSync) - u32 sFlagHack; // Optimize out Status Flag Updates if Program doesn't use Status Flags s32 range[2]; // The range of microMemory that has already been recompiled for the current program u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 5dd94ca18f..fded1265b0 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -168,10 +168,8 @@ microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, in } // Save Flag Instances - if (!mVUflagHack) { - mVUallocSFLAGc(gprT1, gprT2, fStatus); - MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); - } + mVUallocSFLAGc(gprT1, gprT2, fStatus); + MOV32RtoM((uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL, gprT1); mVUallocMFLAGa(mVU, gprT1, fMac); mVUallocCFLAGa(mVU, gprT2, fClip); MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); @@ -242,7 +240,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUpBlock = pBlock; mVUregs.flags = 0; mVUflagInfo = 0; - mVUsFlagHack = CHECK_VU_FLAGHACK2 | mVUflagHack; + mVUsFlagHack = CHECK_VU_FLAGHACK; bool eBitBranch = 0; // E-bit Set on Branch for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 7ed2e26245..94baea9b2f 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -153,7 +153,7 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { // Recompiles Code for Proper Flags on Block Linkings microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles) { - if (__Status && !mVUflagHack) { + if (__Status) { int bStatus[4]; sortFlag(xStatus, bStatus, cycles); MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index cef71c3874..234c742202 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -424,7 +424,7 @@ mVUop(mVU_FCAND) { mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCAND vi01, $%x", _Imm24_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } + pass4 { mVUflagInfo |= 0xf << 8; } } mVUop(mVU_FCEQ) { @@ -437,7 +437,7 @@ mVUop(mVU_FCEQ) { mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCEQ vi01, $%x", _Imm24_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } + pass4 { mVUflagInfo |= 0xf << 8; } } mVUop(mVU_FCGET) { @@ -448,7 +448,7 @@ mVUop(mVU_FCGET) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FCGET vi%02d", _Ft_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } + pass4 { mVUflagInfo |= 0xf << 8; } } mVUop(mVU_FCOR) { @@ -461,7 +461,7 @@ mVUop(mVU_FCOR) { mVUallocVIb(mVU, gprT1, 1); } pass3 { mVUlog("FCOR vi01, $%x", _Imm24_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 8); } + pass4 { mVUflagInfo |= 0xf << 8; } } mVUop(mVU_FCSET) { @@ -486,7 +486,7 @@ mVUop(mVU_FMAND) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMAND vi%02d, vi%02d", _Ft_, _Fs_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } + pass4 { mVUflagInfo |= 0xf << 4; } } mVUop(mVU_FMEQ) { @@ -500,7 +500,7 @@ mVUop(mVU_FMEQ) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMEQ vi%02d, vi%02d", _Ft_, _Fs_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } + pass4 { mVUflagInfo |= 0xf << 4; } } mVUop(mVU_FMOR) { @@ -512,7 +512,7 @@ mVUop(mVU_FMOR) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FMOR vi%02d, vi%02d", _Ft_, _Fs_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 4); } + pass4 { mVUflagInfo |= 0xf << 4; } } //------------------------------------------------------------------ @@ -527,7 +527,7 @@ mVUop(mVU_FSAND) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } + pass4 { mVUflagInfo |= 0xf; } } mVUop(mVU_FSOR) { @@ -538,7 +538,7 @@ mVUop(mVU_FSOR) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } + pass4 { mVUflagInfo |= 0xf; } } mVUop(mVU_FSEQ) { @@ -570,7 +570,7 @@ mVUop(mVU_FSEQ) { mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSEQ vi%02d, $%x", _Ft_, _Imm12_); } - pass4 { mVUflagInfo |= 0xf << (/*mVUcount +*/ 0); mVUsFlagHack = 0; } + pass4 { mVUflagInfo |= 0xf; } } mVUop(mVU_FSSET) { @@ -591,7 +591,6 @@ mVUop(mVU_FSSET) { if (imm) OR32ItoR(sReg, imm); } pass3 { mVUlog("FSSET $%x", _Imm12_); } - pass4 { mVUsFlagHack = 0; } } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 48d2ebcd59..09c57215bc 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -154,7 +154,6 @@ declareAllVariables #define microVUx(aType) template aType #define microVUf(aType) template __forceinline aType - // Define Passes #define pass1 if (recPass == 0) #define pass2 if (recPass == 1) @@ -175,7 +174,6 @@ declareAllVariables #define mVUregsTemp mVUallocInfo.regsTemp #define iPC mVUallocInfo.curPC #define mVUsFlagHack mVUallocInfo.sFlagHack - #define mVUinfo mVUallocInfo.info[iPC / 2] #define mVUstall mVUinfo.stall #define mVUup mVUinfo.uOp @@ -183,10 +181,8 @@ declareAllVariables #define sFLAG mVUinfo.sFlag #define mFLAG mVUinfo.mFlag #define cFLAG mVUinfo.cFlag - #define mVUstartPC mVUallocInfo.startPC #define mVUflagInfo mVUregs.needExactMatch -#define mVUflagHack (mVUcurProg.sFlagHack) #define xPC ((iPC / 2) * 8) #define curI ((u32*)mVU->regs->Micro)[iPC] //mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } @@ -248,8 +244,7 @@ declareAllVariables #endif // Speed Hacks -#define CHECK_VU_FLAGHACK1 (u32)Config.Hacks.vuFlagHack1 // (Can cause Infinite loops, SPS, etc...) -#define CHECK_VU_FLAGHACK2 (u32)Config.Hacks.vuFlagHack2 // (Can cause Infinite loops, SPS, etc...) +#define CHECK_VU_FLAGHACK (u32)Config.Hacks.vuFlagHack // (Can cause Infinite loops, SPS, etc...) #define CHECK_VU_MINMAXHACK (u32)Config.Hacks.vuMinMax // (Can cause SPS, Black Screens, etc...) // Unknown Data diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 138683c9e9..c780a55d70 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -295,24 +295,6 @@ microVUt(void) mVUrestoreRegs(mV) { MOV32ItoR(gprR, Roffset); // Restore gprR } -// Reads entire microProgram and finds out if Status Flag is Used -microVUt(void) mVUcheckSflag(mV, int progIndex) { - if (CHECK_VU_FLAGHACK1) { - int bFlagInfo = mVUflagInfo; - int bCode = mVU->code; - int bFlagHack = mVUsFlagHack; - mVUsFlagHack = 1; - for (u32 i = 0; i < mVU->progSize; i+=2) { - mVU->code = mVU->prog.prog[progIndex].data[i]; - mVUopL(mVU, 3); - } - mVUflagInfo = bFlagInfo; - mVU->code = bCode; - mVU->prog.prog[progIndex].sFlagHack = mVUsFlagHack; - mVUsFlagHack = bFlagHack; - } -} - static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK2[4]) = {0x00000000, 0x40000000, 0x00000000, 0x40000000}; From b792d0c50b16b6d2cabedc7168de3bfcdd2c6ec5 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 4 Jun 2009 00:15:05 +0000 Subject: [PATCH 176/296] GSdx: escape/resume fix git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1314 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTextureFX10.cpp | 3 ++- plugins/GSdx/GSWnd.cpp | 7 +++++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index dd223a508e..48cf096cf7 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -263,7 +263,8 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl m_dev->PSSetShader((*i).second, m_ps_cb); - ID3D10SamplerState* ss0=NULL, *ss1=NULL; + ID3D10SamplerState* ss0 = NULL; + ID3D10SamplerState* ss1 = NULL; if(sel.tfx != 4) { diff --git a/plugins/GSdx/GSWnd.cpp b/plugins/GSdx/GSWnd.cpp index aa9226c5f5..3f148dbc6b 100644 --- a/plugins/GSdx/GSWnd.cpp +++ b/plugins/GSdx/GSWnd.cpp @@ -105,9 +105,12 @@ bool GSWnd::Create(const string& title) wc.hbrBackground = (HBRUSH)GetStockObject(BLACK_BRUSH); wc.lpszClassName = "GSWnd"; - if(!RegisterClass(&wc)) + if(!GetClassInfo(wc.hInstance, wc.lpszClassName, &wc)) { - return false; + if(!RegisterClass(&wc)) + { + return false; + } } DWORD style = WS_CLIPCHILDREN | WS_CLIPSIBLINGS | WS_OVERLAPPEDWINDOW | WS_BORDER; From 21d96158718e065ffe04b5044a12f4a519f19c2f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 00:24:49 +0000 Subject: [PATCH 177/296] microVU: - Minor Changes. Shouldn't effect anything. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1315 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Alloc.inl | 8 ++++---- pcsx2/x86/microVU_Execute.inl | 8 ++++---- pcsx2/x86/microVU_Lower.inl | 26 +++++++++++++------------- 3 files changed, 21 insertions(+), 21 deletions(-) diff --git a/pcsx2/x86/microVU_Alloc.inl b/pcsx2/x86/microVU_Alloc.inl index d14ba7a570..61fd7c593a 100644 --- a/pcsx2/x86/microVU_Alloc.inl +++ b/pcsx2/x86/microVU_Alloc.inl @@ -676,10 +676,10 @@ microVUt(void) mVUallocSFLAGc(int reg, int regT, int fInstance) { u8 *pjmp; XOR32RtoR(reg, reg); mVUallocSFLAGa(regT, fInstance); - setBitSFLAG(0x000f, 0x0001); // Z Bit - setBitSFLAG(0x00f0, 0x0002); // S Bit - setBitSFLAG(0x0f00, 0x0040); // ZS Bit - setBitSFLAG(0xf000, 0x0080); // SS Bit + setBitSFLAG(0x0f00, 0x0001); // Z Bit + setBitSFLAG(0xf000, 0x0002); // S Bit + setBitSFLAG(0x000f, 0x0040); // ZS Bit + setBitSFLAG(0x00f0, 0x0080); // SS Bit AND32ItoR(regT, 0xffff0000); // DS/DI/OS/US/D/I/O/U Bits SHR32ItoR(regT, 14); OR32RtoR(reg, regT); diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index e3fe31d202..3f197ad740 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -44,12 +44,12 @@ microVUt(void) mVUdispatcherA(mV) { MOV32MtoR(gprF0, (uptr)&mVU->regs->VI[REG_STATUS_FLAG].UL); MOV32RtoR(gprF1, gprF0); - SHL32ItoR(gprF1, 3); - AND32ItoR(gprF1, 0x218); + SHR32ItoR(gprF1, 3); + AND32ItoR(gprF1, 0x18); MOV32RtoR(gprF2, gprF0); - SHL32ItoR(gprF2, 5); - AND32ItoR(gprF2, 0x1000); + SHL32ItoR(gprF2, 11); + AND32ItoR(gprF2, 0x1800); OR32RtoR (gprF1, gprF2); MOV32RtoR(gprF3, gprF0); diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index 234c742202..a91266f159 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -523,7 +523,7 @@ mVUop(mVU_FSAND) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { mVUallocSFLAGc(gprT1, gprT2, sFLAG.read); - AND16ItoR(gprT1, _Imm12_); + AND32ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSAND vi%02d, $%x", _Ft_, _Imm12_); } @@ -534,7 +534,7 @@ mVUop(mVU_FSOR) { pass1 { mVUanalyzeSflag(mVU, _It_); } pass2 { mVUallocSFLAGc(gprT1, gprT2, sFLAG.read); - OR16ItoR(gprT1, _Imm12_); + OR32ItoR(gprT1, _Imm12_); mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("FSOR vi%02d, $%x", _Ft_, _Imm12_); } @@ -546,24 +546,24 @@ mVUop(mVU_FSEQ) { pass2 { u8 *pjmp; int imm = 0; - if (_Imm12_ & 0x0001) imm |= 0x000000f; // Z - if (_Imm12_ & 0x0002) imm |= 0x00000f0; // S + if (_Imm12_ & 0x0001) imm |= 0x0000f00; // Z + if (_Imm12_ & 0x0002) imm |= 0x000f000; // S if (_Imm12_ & 0x0004) imm |= 0x0010000; // U if (_Imm12_ & 0x0008) imm |= 0x0020000; // O if (_Imm12_ & 0x0010) imm |= 0x0040000; // I if (_Imm12_ & 0x0020) imm |= 0x0080000; // D - if (_Imm12_ & 0x0040) imm |= 0x0000f00; // ZS - if (_Imm12_ & 0x0080) imm |= 0x000f000; // SS + if (_Imm12_ & 0x0040) imm |= 0x000000f; // ZS + if (_Imm12_ & 0x0080) imm |= 0x00000f0; // SS if (_Imm12_ & 0x0100) imm |= 0x0400000; // US if (_Imm12_ & 0x0200) imm |= 0x0800000; // OS if (_Imm12_ & 0x0400) imm |= 0x1000000; // IS if (_Imm12_ & 0x0800) imm |= 0x2000000; // DS mVUallocSFLAGa(gprT1, sFLAG.read); - setBitFSEQ(0x000f); // Z bit - setBitFSEQ(0x00f0); // S bit - setBitFSEQ(0x0f00); // ZS bit - setBitFSEQ(0xf000); // SS bit + setBitFSEQ(0x0f00); // Z bit + setBitFSEQ(0xf000); // S bit + setBitFSEQ(0x000f); // ZS bit + setBitFSEQ(0x00f0); // SS bit XOR32ItoR(gprT1, imm); SUB32ItoR(gprT1, 1); SHR32ItoR(gprT1, 31); @@ -577,8 +577,8 @@ mVUop(mVU_FSSET) { pass1 { mVUanalyzeFSSET(mVU); } pass2 { int sReg, imm = 0; - if (_Imm12_ & 0x0040) imm |= 0x0000f00; // ZS - if (_Imm12_ & 0x0080) imm |= 0x000f000; // SS + if (_Imm12_ & 0x0040) imm |= 0x000000f; // ZS + if (_Imm12_ & 0x0080) imm |= 0x00000f0; // SS if (_Imm12_ & 0x0100) imm |= 0x0400000; // US if (_Imm12_ & 0x0200) imm |= 0x0800000; // OS if (_Imm12_ & 0x0400) imm |= 0x1000000; // IS @@ -587,7 +587,7 @@ mVUop(mVU_FSSET) { if (!(sFLAG.doFlag || mVUinfo.doDivFlag)) { mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag } - AND32ItoR(sReg, 0xf00ff); // Keep Non-Sticky Bits + AND32ItoR(sReg, 0xfff00); // Keep Non-Sticky Bits if (imm) OR32ItoR(sReg, imm); } pass3 { mVUlog("FSSET $%x", _Imm12_); } From b65b439e6a405c4af9f8f9cfd22d4b7425b1bb45 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 00:50:38 +0000 Subject: [PATCH 178/296] microVU: - Further optimized status flag updating... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1316 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 18 ++++++++++++------ pcsx2/x86/microVU_Flags.inl | 8 ++++++-- pcsx2/x86/microVU_IR.h | 9 +++++---- pcsx2/x86/microVU_Upper.inl | 10 ++++++---- 4 files changed, 29 insertions(+), 16 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index ad6452c81d..0143321dc1 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -296,6 +296,15 @@ microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { //------------------------------------------------------------------ // Sflag - Status Flag Opcodes //------------------------------------------------------------------ +#define flagSet(xFLAG) { \ + int curPC = iPC; \ + for (int i = mVUcount, j = 0; i > 0; i--, j++) { \ + incPC2(-2); \ + if (sFLAG.doFlag) { xFLAG = 1; if (j >= 3) { break; } } \ + } \ + iPC = curPC; \ +} + microVUt(void) mVUanalyzeSflag(mV, int It) { if (!It) { mVUlow.isNOP = 1; } @@ -310,6 +319,7 @@ microVUt(void) mVUanalyzeSflag(mV, int It) { } mVUlow.readFlags = 1; analyzeVIreg2(It, mVUlow.VI_write, 1); + flagSet(sFLAG.doNonSticky); } microVUt(void) mVUanalyzeFSSET(mV) { @@ -326,16 +336,12 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) mVUinfo.swapOps = 1; if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } - int curPC = iPC; - for (int i = mVUcount, j = 0; i > 0; i--, j++) { - incPC2(-2); - if (sFLAG.doFlag) { mFLAG.doFlag = 1; if (j >= 3) { break; } } - } - iPC = curPC; + } mVUlow.readFlags = 1; analyzeVIreg1(Is, mVUlow.VI_read[0]); analyzeVIreg2(It, mVUlow.VI_write, 1); + flagSet(mFLAG.doFlag); } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 94baea9b2f..782595eeb3 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -77,9 +77,13 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { int endPC = iPC; u32 aCount = 1; // Amount of instructions needed to get valid mac flag instances for block linking - // Ensure last ~4+ instructions update mac flags (if next block's first 4 instructions will read them) + // Ensure last ~4+ instructions update mac/status flags (if next block's first 4 instructions will read them) for (int i = mVUcount; i > 0; i--, aCount++) { - if (sFLAG.doFlag) { if (__Mac) { mFLAG.doFlag = 1; } if (aCount >= 4) { break; } } + if (sFLAG.doFlag) { + if (__Mac) { mFLAG.doFlag = 1; } + if (__Status) { sFLAG.doNonSticky = 1; } + if (aCount >= 4) { break; } + } incPC2(-2); } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index a109503651..8652159143 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -102,10 +102,11 @@ struct microLowerOp { }; struct microFlagInst { - bool doFlag; // Update Flag on this Instruction - u8 write; // Points to the instance that should be written to (s-stage write) - u8 lastWrite; // Points to the instance that was last written to (most up-to-date flag) - u8 read; // Points to the instance that should be read by a lower instruction (t-stage read) + bool doFlag; // Update Flag on this Instruction + bool doNonSticky; // Update O,U,S,Z (non-sticky) bits on this Instruction (status flag only) + u8 write; // Points to the instance that should be written to (s-stage write) + u8 lastWrite; // Points to the instance that was last written to (most up-to-date flag) + u8 read; // Points to the instance that should be read by a lower instruction (t-stage read) }; struct microOp { diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index db0e92a51e..4fcbb95bc7 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -39,7 +39,7 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool if (sFLAG.doFlag) { getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag - AND32ItoR(sReg, 0xffffff00); // Keep Sticky and D/I flags + if (sFLAG.doNonSticky) AND32ItoR(sReg, 0xfffcff00); // Clear O,U,S,Z flags } //-------------------------Check for Signed flags------------------------------ @@ -59,15 +59,17 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); } - OR32RtoR(mReg, gprT2); + OR32RtoR(mReg, gprT2); //-------------------------Write back flags------------------------------ if (mFLAG.doFlag) mVUallocMFLAGb(mVU, mReg, mFLAG.write); // Set Mac Flag if (sFLAG.doFlag) { OR32RtoR (sReg, mReg); - SHL32ItoR(mReg, 8); - OR32RtoR (sReg, mReg); + if (sFLAG.doNonSticky) { + SHL32ItoR(mReg, 8); + OR32RtoR (sReg, mReg); + } } } From 5831b47c4e763c45e9a96ff1e2aebb608e64a5c0 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 02:05:57 +0000 Subject: [PATCH 179/296] microVU: - Fixed a bug from r1315. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1317 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Execute.inl | 7 +++---- pcsx2/x86/microVU_Upper.inl | 2 +- 2 files changed, 4 insertions(+), 5 deletions(-) diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 3f197ad740..39d1c85e3a 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -52,10 +52,9 @@ microVUt(void) mVUdispatcherA(mV) { AND32ItoR(gprF2, 0x1800); OR32RtoR (gprF1, gprF2); - MOV32RtoR(gprF3, gprF0); - SHL32ItoR(gprF3, 14); - AND32ItoR(gprF3, 0x3cf0000); - OR32RtoR (gprF1, gprF3); + SHL32ItoR(gprF0, 14); + AND32ItoR(gprF0, 0x3cf0000); + OR32RtoR (gprF1, gprF0); MOV32RtoR(gprF0, gprF1); MOV32RtoR(gprF2, gprF1); diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index 4fcbb95bc7..eee4fa4ab1 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -39,7 +39,7 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool if (sFLAG.doFlag) { getFlagReg(sReg, sFLAG.write); // Set sReg to valid GPR by Cur Flag Instance mVUallocSFLAGa(sReg, sFLAG.lastWrite); // Get Prev Status Flag - if (sFLAG.doNonSticky) AND32ItoR(sReg, 0xfffcff00); // Clear O,U,S,Z flags + if (sFLAG.doNonSticky) AND32ItoR(sReg, 0xfffc00ff); // Clear O,U,S,Z flags } //-------------------------Check for Signed flags------------------------------ From 4490d237dc645ed0f2af327cbc97141cd2b1244c Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 05:52:57 +0000 Subject: [PATCH 180/296] microVU: - Implemented m-bit functionality. Fixes Crash Twinsanity and most-likely other games... Note: If anyone has any games that have broken graphics when mVU prints "microVU0: M-bit set!" to console, then please notify me about it. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1318 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 24 +++++++++++++++--------- pcsx2/x86/microVU_Flags.inl | 2 +- pcsx2/x86/microVU_IR.h | 1 + pcsx2/x86/microVU_Misc.h | 8 ++++++-- 4 files changed, 23 insertions(+), 12 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index fded1265b0..0eb5a4be05 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -157,7 +157,7 @@ microVUt(void) mVUsetCycles(mV) { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, int fClip) { +microVUt(void) mVUendProgram(mV, bool clearIsBusy, int qInst, int pInst, int fStatus, int fMac, int fClip) { // Save P/Q Regs if (qInst) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } @@ -175,9 +175,12 @@ microVUt(void) mVUendProgram(mV, int qInst, int pInst, int fStatus, int fMac, in MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); - // Clear 'is busy' Flags, Save PC, and Jump to Exit Point - AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (isVU1 ? ~0x100 : ~0x001)); // VBS0/VBS1 flag - AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif + if (clearIsBusy) { // Clear 'is busy' Flags + AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (isVU1 ? ~0x100 : ~0x001)); // VBS0/VBS1 flag + AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif + } + + //Save PC, and Jump to Exit Point MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); } @@ -196,10 +199,10 @@ microVUt(void) mVUtestCycles(mV) { SUB32ItoM((uptr)&mVU->cycles, mVUcycles); u32* jmp32 = JG32(0); MOV32ItoR(gprT2, xPC); - if (!isVU1) CALLFunc((uptr)mVUwarning0); - else CALLFunc((uptr)mVUwarning1); + if (isVU1) CALLFunc((uptr)mVUwarning1); + //else CALLFunc((uptr)mVUwarning0); // VU0 is allowed early exit for COP2 Interlock Simulation MOV32ItoR(gprR, Roffset); // Restore gprR - mVUendProgram(mVU, 0, 0, sI, 0, cI); + mVUendProgram(mVU, (isVU1)?1:0, 0, 0, sI, 0, cI); x86SetJ32(jmp32); } @@ -249,7 +252,9 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { incCycles(1); mVUopU(mVU, 0); if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } - if (curI & _MDTbit_) { branch = 4; } + if (curI & _DTbit_) { branch = 4; } + if (curI & _Mbit_) { mVUup.mBit = 1; } + if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } else { incPC(-1); mVUopL(mVU, 0); incPC(1); } mVUsetCycles(mVU); @@ -276,6 +281,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { int x; for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { if (mVUinfo.isEOB) { x = 0xffff; } + if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } @@ -383,7 +389,7 @@ eBitTemination: // Do E-bit end stuff here mVUsetupRange(mVU, xPC - 8); - mVUendProgram(mVU, mVU->q, mVU->p, lStatus, lMac, lClip); + mVUendProgram(mVU, 1, mVU->q, mVU->p, lStatus, lMac, lClip); return thisPtr; } diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 782595eeb3..0658d78807 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -207,7 +207,7 @@ void mVUflagPass(mV, u32 startPC, u32 xCount) { for (int branch = 0; mVUcount < xCount; mVUcount++) { incPC(1); if ( curI & _Ebit_ ) { branch = 1; } - if ( curI & _MDTbit_ ) { branch = 4; } + if ( curI & _DTbit_ ) { branch = 4; } if (!(curI & _Ibit_) ) { incPC(-1); mVUopL(mVU, 3); incPC(1); } if (branch >= 2) { shortBranch(); break; } else if (branch == 1) { branch = 2; } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 8652159143..94b110779d 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -81,6 +81,7 @@ struct microVIreg { struct microUpperOp { bool eBit; // Has E-bit set bool iBit; // Has I-bit set + bool mBit; // Has M-bit set microVFreg VF_write; // VF Vectors written to by this instruction microVFreg VF_read[2]; // VF Vectors read by this instruction }; diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 09c57215bc..26f6cdd501 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -103,7 +103,7 @@ declareAllVariables #define _Mbit_ (1<<29) #define _Dbit_ (1<<28) #define _Tbit_ (1<<27) -#define _MDTbit_ 0 //( _Mbit_ | _Dbit_ | _Tbit_ ) // ToDo: Implement this stuff... +#define _DTbit_ 0 //( _Dbit_ | _Tbit_ ) // ToDo: Implement this stuff... #define divI 0x1040000 #define divD 0x2080000 @@ -231,7 +231,11 @@ declareAllVariables } #else #define mVUprint 0&& -#define mVUdebug1() {} +#define mVUdebug1() { \ + if (curI & _Mbit_) { Console::Status("microVU%d: M-bit set!", params getIndex); } \ + if (curI & _Dbit_) { DevCon::Status ("microVU%d: D-bit set!", params getIndex); } \ + if (curI & _Tbit_) { DevCon::Status ("microVU%d: T-bit set!", params getIndex); } \ +} #endif // Program Logging... From 0a6bc578603ea74cdc75a9d0b61f8b4722c9aa26 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 08:13:47 +0000 Subject: [PATCH 181/296] microVU: - Fixed a very rare case, where both upper and lower instructions read and write to each others regs. Example: ADD.xyzw VF1, VF2, VF3 MOVE.xyzw VF2, VF1 - Forgot to commit iVU0micro.cpp on my last update... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1319 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU0micro.cpp | 2 +- pcsx2/x86/microVU.h | 1 + pcsx2/x86/microVU_Analyze.inl | 38 ++++--------------------- pcsx2/x86/microVU_Compile.inl | 52 +++++++++++++++++++++++++++++++++-- pcsx2/x86/microVU_IR.h | 1 + 5 files changed, 57 insertions(+), 37 deletions(-) diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index cb85aa442b..9a3e4e5ce6 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -47,7 +47,7 @@ namespace VU0micro if ((VU0.VI[REG_VPU_STAT].UL & 1) == 0) return; FreezeXMMRegs(1); - if (useMVU0) runVUrec(VU0.VI[REG_TPC].UL, 50000, 0); + if (useMVU0) runVUrec(VU0.VI[REG_TPC].UL, 0x300, 0); else SuperVUExecuteProgram(VU0.VI[REG_TPC].UL & 0xfff, 0); FreezeXMMRegs(0); } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index d7cc0d114d..85d9e4163e 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -100,6 +100,7 @@ struct microVU { PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) PCSX2_ALIGNED16(u32 clipFlag[4]); // 4 instances of clip flag (used in execution) PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ + PCSX2_ALIGNED16(u32 xmmVFb[4]); // Backup for VF regs u32 index; // VU Index (VU0 or VU1) u32 microSize; // VU Micro Memory Size diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 0143321dc1..67169db9ed 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -31,7 +31,7 @@ #define aMax(x, y) ((x > y) ? x : y) #define aMin(x, y) ((x < y) ? x : y) -// Read a VF reg by upper op +// Read a VF reg #define analyzeReg1(xReg, vfRead) { \ if (xReg) { \ if (_X) { mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; } \ @@ -41,20 +41,6 @@ } \ } -// Read a VF reg by lower op -#define analyzeReg1b(xReg, vfRead) { \ - if (xReg) { \ - analyzeReg1(xReg, vfRead); \ - if (mVUregsTemp.VFreg[0] == xReg) { \ - if ((mVUregsTemp.VF[0].x && _X) \ - || (mVUregsTemp.VF[0].y && _Y) \ - || (mVUregsTemp.VF[0].z && _Z) \ - || (mVUregsTemp.VF[0].w && _W)) \ - { mVUinfo.swapOps = 1; } \ - } \ - } \ -} - // Write to a VF reg #define analyzeReg2(xReg, vfWrite, isLowOp) { \ if (xReg) { \ @@ -92,13 +78,6 @@ case 2: mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; break; \ case 3: mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; break; \ } \ - if (mVUregsTemp.VFreg[0] == xReg) { \ - if ((mVUregsTemp.VF[0].x && (fxf == 0)) \ - || (mVUregsTemp.VF[0].y && (fxf == 1)) \ - || (mVUregsTemp.VF[0].z && (fxf == 2)) \ - || (mVUregsTemp.VF[0].w && (fxf == 3))) \ - { mVUinfo.swapOps = 1; } \ - } \ } \ } @@ -109,13 +88,6 @@ if (_Y) { mVUstall = aMax(mVUstall, aReg(xReg).z); vfRead.reg = xReg; vfRead.z = 1; } \ if (_Z) { mVUstall = aMax(mVUstall, aReg(xReg).w); vfRead.reg = xReg; vfRead.w = 1; } \ if (_W) { mVUstall = aMax(mVUstall, aReg(xReg).x); vfRead.reg = xReg; vfRead.x = 1; } \ - if (mVUregsTemp.VFreg[0] == xReg) { \ - if ((mVUregsTemp.VF[0].y && _X) \ - || (mVUregsTemp.VF[0].z && _Y) \ - || (mVUregsTemp.VF[0].w && _Z) \ - || (mVUregsTemp.VF[0].x && _W)) \ - { mVUinfo.swapOps = 1; } \ - } \ } \ } @@ -234,7 +206,7 @@ microVUt(void) mVUanalyzeEFU1(mV, int Fs, int Fsf, u8 xCycles) { microVUt(void) mVUanalyzeEFU2(mV, int Fs, u8 xCycles) { mVUprint("microVU: EFU Opcode"); - analyzeReg1b(Fs, mVUlow.VF_read[0]); + analyzeReg1(Fs, mVUlow.VF_read[0]); analyzePreg(xCycles); } @@ -253,8 +225,8 @@ microVUt(void) mVUanalyzeMFP(mV, int Ft) { microVUt(void) mVUanalyzeMOVE(mV, int Fs, int Ft) { if (!Ft || (Ft == Fs)) { mVUlow.isNOP = 1; } - analyzeReg1b(Fs, mVUlow.VF_read[0]); - analyzeReg2 (Ft, mVUlow.VF_write, 1); + analyzeReg1(Fs, mVUlow.VF_read[0]); + analyzeReg2(Ft, mVUlow.VF_write, 1); } //------------------------------------------------------------------ @@ -273,7 +245,7 @@ microVUt(void) mVUanalyzeLQ(mV, int Ft, int Is, bool writeIs) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeSQ(mV, int Fs, int It, bool writeIt) { - analyzeReg1b (Fs, mVUlow.VF_read[0]); + analyzeReg1 (Fs, mVUlow.VF_read[0]); analyzeVIreg1(It, mVUlow.VI_read[0]); if (writeIt) { analyzeVIreg2(It, mVUlow.VI_write, 1); } } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 0eb5a4be05..592b76774a 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -49,6 +49,30 @@ } \ } +#define doBackupVF1() { \ + if (mVUinfo.backupVF && !mVUlow.noWriteVF) { \ + DevCon::Status("microVU%d: Backing Up VF Reg [%04x]", params getIndex, xPC); \ + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)&mVU->regs->VF[mVUlow.VF_write.reg].UL[0]); \ + SSE_MOVAPS_XMM_to_M128((uptr)mVU->xmmVFb, xmmT1); \ + } \ +} + +#define doBackupVF2() { \ + if (mVUinfo.backupVF && !mVUlow.noWriteVF) { \ + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->xmmVFb); \ + SSE_MOVAPS_M128_to_XMM(xmmT2, (uptr)&mVU->regs->VF[mVUlow.VF_write.reg].UL[0]); \ + SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VF[mVUlow.VF_write.reg].UL[0], xmmT1); \ + SSE_MOVAPS_XMM_to_M128((uptr)mVU->xmmVFb, xmmT2); \ + } \ +} + +#define doBackupVF3() { \ + if (mVUinfo.backupVF && !mVUlow.noWriteVF) { \ + SSE_MOVAPS_M128_to_XMM(xmmT1, (uptr)mVU->xmmVFb); \ + SSE_MOVAPS_XMM_to_M128((uptr)&mVU->regs->VF[mVUlow.VF_write.reg].UL[0], xmmT1); \ + } \ +} + #define startLoop() { \ mVUdebug1(); \ memset(&mVUinfo, 0, sizeof(mVUinfo)); \ @@ -61,6 +85,7 @@ #define incQ() { mVU->q = (mVU->q+1) & 1; } #define doUpperOp() { mVUopU(mVU, 1); mVUdivSet(mVU); } #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } +#define doSwapOp() { doBackupVF1(); mVUopL(mVU, 1); doBackupVF2(); incPC(1); doUpperOp(); doBackupVF3(); } #define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } //------------------------------------------------------------------ @@ -134,12 +159,34 @@ microVUt(void) mVUincCycles(mV, int x) { calcCycles(mVUregs.r, x); } +#define cmpVFregs(VFreg1, VFreg2, xVar) { \ + if (VFreg1.reg == VFreg2.reg) { \ + if ((VFreg1.x && VFreg2.x) \ + || (VFreg1.y && VFreg2.y) \ + || (VFreg1.z && VFreg2.z) \ + || (VFreg1.w && VFreg2.w)) \ + { xVar = 1; } \ + } \ +} + microVUt(void) mVUsetCycles(mV) { incCycles(mVUstall); - if (mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1] && mVUregsTemp.VFreg[0]) { // If upper Op && lower Op write to same VF reg + // If upper Op && lower Op write to same VF reg: + if ((mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1]) && mVUregsTemp.VFreg[0]) { if (mVUregsTemp.r || mVUregsTemp.VI) mVUlow.noWriteVF = 1; else mVUlow.isNOP = 1; // If lower Op doesn't modify anything else, then make it a NOP } + // If lower op reads a VF reg that upper Op writes to: + if ((mVUlow.VF_read[0].reg || mVUlow.VF_read[1].reg) && mVUup.VF_write.reg) { + cmpVFregs(mVUup.VF_write, mVUlow.VF_read[0], mVUinfo.swapOps); + cmpVFregs(mVUup.VF_write, mVUlow.VF_read[1], mVUinfo.swapOps); + } + // If above case is true, and upper op reads a VF reg that lower Op Writes to: + if (mVUinfo.swapOps && ((mVUup.VF_read[0].reg || mVUup.VF_read[1].reg) && mVUlow.VF_write.reg)) { + cmpVFregs(mVUlow.VF_write, mVUup.VF_read[0], mVUinfo.backupVF); + cmpVFregs(mVUlow.VF_write, mVUup.VF_read[1], mVUinfo.backupVF); + } + tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].x, mVUregsTemp.VF[0].x); tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].y, mVUregsTemp.VF[0].y); tCycles(mVUregs.VF[mVUregsTemp.VFreg[0]].z, mVUregsTemp.VF[0].z); @@ -254,7 +301,6 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } if (curI & _DTbit_) { branch = 4; } if (curI & _Mbit_) { mVUup.mBit = 1; } - if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } else { incPC(-1); mVUopL(mVU, 0); incPC(1); } mVUsetCycles(mVU); @@ -284,7 +330,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { mVUopL(mVU, 1); incPC(1); doUpperOp(); } + else { doSwapOp(); } if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } if (!mVUinfo.isBdelay) { incPC(1); } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 94b110779d..c3b6f979fb 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -115,6 +115,7 @@ struct microOp { bool isEOB; // Cur Instruction is last instruction in block (End of Block) bool isBdelay; // Cur Instruction in Branch Delay slot bool swapOps; // Run Lower Instruction before Upper Instruction + bool backupVF; // Backup mVUlow.VF_write.reg, and restore it before the Upper Instruction is called bool doXGKICK; // Do XGKICK transfer on this instruction bool doDivFlag; // Transfer Div flag to Status Flag on this instruction int readQ; // Q instance for reading From 7ee6eefb6870f9e44782b5ea74492378d554e868 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 4 Jun 2009 08:32:31 +0000 Subject: [PATCH 182/296] GSdx: config dialogs disable unrelated controls again git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1320 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPUSettingsDlg.cpp | 55 ++++++++++++++------------ plugins/GSdx/GPUSettingsDlg.h | 2 + plugins/GSdx/GSSettingsDlg.cpp | 68 +++++++++++++++++++++------------ plugins/GSdx/GSSettingsDlg.h | 2 + plugins/GSdx/GSdx.rc | 8 ++-- plugins/GSdx/resource.h | 6 +-- 6 files changed, 86 insertions(+), 55 deletions(-) diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 7ea1c5e289..9b159e3130 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -121,25 +121,18 @@ void GPUSettingsDlg::OnInit() } } + bool isdx10avail = GSUtil::IsDirect3D10Available(); + + vector renderers; + + for(size_t i = 0; i < countof(g_renderers); i++) { - bool isdx10avail = GSUtil::IsDirect3D10Available(); + if(i >= 3 && i <= 5 && !isdx10avail) continue; - vector renderers; - - for(size_t i = 0; i < countof(g_renderers); i++) - { - if(i >= 3 && i <= 5 && !isdx10avail) continue; - - renderers.push_back(g_renderers[i]); - } - - HWND hWnd = GetDlgItem(m_hWnd, IDC_RENDERER); - - ComboBoxInit(hWnd, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - - OnCommand(hWnd, IDC_RENDERER, CBN_SELCHANGE); + renderers.push_back(g_renderers[i]); } + ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); ComboBoxInit(GetDlgItem(m_hWnd, IDC_FILTER), g_filter, countof(g_filter), theApp.GetConfig("filter", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_DITHERING), g_dithering, countof(g_dithering), theApp.GetConfig("dithering", 1)); @@ -148,21 +141,15 @@ void GPUSettingsDlg::OnInit() SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETRANGE, 0, MAKELPARAM(16, 1)); SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("swthreads", 1), 0)); + + UpdateControls(); } bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { if(id == IDC_RENDERER && code == CBN_SELCHANGE) { - int item = (int)SendMessage(hWnd, CB_GETCURSEL, 0, 0); - - if(item >= 0) - { - int i = (int)SendMessage(hWnd, CB_GETITEMDATA, item, 0); - - ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), i == 1 ? SW_SHOW : SW_HIDE); - ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), i == 2 ? SW_SHOW : SW_HIDE); - } + UpdateControls(); } else if(id == IDOK) { @@ -212,4 +199,24 @@ bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) } return __super::OnCommand(hWnd, id, code); +} + +void GPUSettingsDlg::UpdateControls() +{ + INT_PTR i; + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), i)) + { + bool dx9 = i == 1; + bool dx10 = i == 2; + bool sw = i >= 0 && i <= 2; + + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); + + EnableWindow(GetDlgItem(m_hWnd, IDC_SHADER), dx9); + EnableWindow(GetDlgItem(m_hWnd, IDC_SCALE), sw); + EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS_EDIT), sw); + EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS), sw); + } } \ No newline at end of file diff --git a/plugins/GSdx/GPUSettingsDlg.h b/plugins/GSdx/GPUSettingsDlg.h index 517ffb6443..f450ce647d 100644 --- a/plugins/GSdx/GPUSettingsDlg.h +++ b/plugins/GSdx/GPUSettingsDlg.h @@ -28,6 +28,8 @@ class GPUSettingsDlg : public GSDialog { list m_modes; + void UpdateControls(); + protected: void OnInit(); bool OnCommand(HWND hWnd, UINT id, UINT code); diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index ca431f74a0..d4a2bdd627 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -112,25 +112,18 @@ void GSSettingsDlg::OnInit() } } + bool isdx10avail = GSUtil::IsDirect3D10Available(); + + vector renderers; + + for(size_t i = 0; i < countof(g_renderers); i++) { - bool isdx10avail = GSUtil::IsDirect3D10Available(); + if(i >= 3 && i <= 5 && !isdx10avail) continue; - vector renderers; - - for(size_t i = 0; i < countof(g_renderers); i++) - { - if(i >= 3 && i <= 5 && !isdx10avail) continue; - - renderers.push_back(g_renderers[i]); - } - - HWND hWnd = GetDlgItem(m_hWnd, IDC_RENDERER); - - ComboBoxInit(hWnd, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - - OnCommand(hWnd, IDC_RENDERER, CBN_SELCHANGE); + renderers.push_back(g_renderers[i]); } + ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); ComboBoxInit(GetDlgItem(m_hWnd, IDC_INTERLACE), g_interlace, countof(g_interlace), theApp.GetConfig("Interlace", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); @@ -151,21 +144,19 @@ void GSSettingsDlg::OnInit() SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETRANGE, 0, MAKELPARAM(16, 1)); SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("swthreads", 1), 0)); + + UpdateControls(); } bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { if(id == IDC_RENDERER && code == CBN_SELCHANGE) { - int item = (int)SendMessage(hWnd, CB_GETCURSEL, 0, 0); - - if(item >= 0) - { - int i = (int)SendMessage(hWnd, CB_GETITEMDATA, item, 0); - - ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), i >= 0 && i <= 2 ? SW_SHOW : SW_HIDE); - ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), i >= 3 && i <= 5 ? SW_SHOW : SW_HIDE); - } + UpdateControls(); + } + else if(id == IDC_NATIVERES && code == BN_CLICKED) + { + UpdateControls(); } else if(id == IDOK) { @@ -216,3 +207,32 @@ bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) return __super::OnCommand(hWnd, id, code); } +void GSSettingsDlg::UpdateControls() +{ + INT_PTR i; + + if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), i)) + { + bool dx9 = i >= 0 && i <= 2; + bool dx10 = i >= 3 && i <= 5; + bool hw = i == 0 || i == 3; + bool sw = i == 1 || i == 4 || i == 6; + bool native = !!IsDlgButtonChecked(m_hWnd, IDC_NATIVERES); + + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); + ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); + + EnableWindow(GetDlgItem(m_hWnd, IDC_RESOLUTION), dx9); + EnableWindow(GetDlgItem(m_hWnd, IDC_SHADER), dx9); + EnableWindow(GetDlgItem(m_hWnd, IDC_RESX), hw && !native); + EnableWindow(GetDlgItem(m_hWnd, IDC_RESX_EDIT), hw && !native); + EnableWindow(GetDlgItem(m_hWnd, IDC_RESY), hw && !native); + EnableWindow(GetDlgItem(m_hWnd, IDC_RESY_EDIT), hw && !native); + EnableWindow(GetDlgItem(m_hWnd, IDC_NATIVERES), hw); + EnableWindow(GetDlgItem(m_hWnd, IDC_LOGZ), dx9 && hw); + EnableWindow(GetDlgItem(m_hWnd, IDC_FBA), dx9 && hw); + EnableWindow(GetDlgItem(m_hWnd, IDC_AA1), sw); + EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS_EDIT), sw); + EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS), sw); + } +} \ No newline at end of file diff --git a/plugins/GSdx/GSSettingsDlg.h b/plugins/GSdx/GSSettingsDlg.h index f5f76f2503..2e3765a53a 100644 --- a/plugins/GSdx/GSSettingsDlg.h +++ b/plugins/GSdx/GSSettingsDlg.h @@ -28,6 +28,8 @@ class GSSettingsDlg : public GSDialog { list m_modes; + void UpdateControls(); + protected: void OnInit(); bool OnCommand(HWND hWnd, UINT id, UINT code); diff --git a/plugins/GSdx/GSdx.rc b/plugins/GSdx/GSdx.rc index d26b46b7c4..0850a8481b 100644 --- a/plugins/GSdx/GSdx.rc +++ b/plugins/GSdx/GSdx.rc @@ -100,13 +100,13 @@ BEGIN LTEXT "Aspect Ratio (F6):",IDC_STATIC,7,120,60,8 COMBOBOX IDC_ASPECTRATIO,71,117,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "D3D internal res:",IDC_STATIC,7,135,55,8 - EDITTEXT IDC_EDIT1,71,132,35,13,ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_RESX_EDIT,71,132,35,13,ES_AUTOHSCROLL | ES_NUMBER CONTROL "",IDC_RESX,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,135,11,14 - EDITTEXT IDC_EDIT2,109,132,35,13,ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_RESY_EDIT,109,132,35,13,ES_AUTOHSCROLL | ES_NUMBER CONTROL "",IDC_RESY,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,133,135,11,14 CONTROL "Native",IDC_NATIVERES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,149,134,33,10 LTEXT "SW rend. threads:",IDC_STATIC,7,149,60,8 - EDITTEXT IDC_EDIT3,71,147,35,13,ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_SWTHREADS_EDIT,71,147,35,13,ES_AUTOHSCROLL | ES_NUMBER CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,150,11,14 CONTROL "Texture filtering",IDC_FILTER,"Button",BS_AUTO3STATE | WS_TABSTOP,7,167,67,10 CONTROL "Logarithmic Z",IDC_LOGZ,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,168,58,10 @@ -154,7 +154,7 @@ BEGIN LTEXT "Aspect Ratio (PgDn):",IDC_STATIC,7,135,68,8 COMBOBOX IDC_ASPECTRATIO,78,132,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Rendering Threads:",IDC_STATIC,7,165,64,8 - EDITTEXT IDC_EDIT3,78,163,35,13,ES_AUTOHSCROLL | ES_NUMBER + EDITTEXT IDC_SWTHREADS_EDIT,78,163,35,13,ES_AUTOHSCROLL | ES_NUMBER CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,169,11,14 DEFPUSHBUTTON "OK",IDOK,43,214,50,14 PUSHBUTTON "Cancel",IDCANCEL,96,214,50,14 diff --git a/plugins/GSdx/resource.h b/plugins/GSdx/resource.h index c5f074136f..2083fa82ee 100644 --- a/plugins/GSdx/resource.h +++ b/plugins/GSdx/resource.h @@ -10,10 +10,10 @@ #define IDC_CODECS 2006 #define IDC_RESOLUTION 2007 #define IDC_SHADER 2008 -#define IDC_EDIT1 2009 -#define IDC_EDIT2 2010 +#define IDC_RESX_EDIT 2009 +#define IDC_RESY_EDIT 2010 #define IDC_AA1 2011 -#define IDC_EDIT3 2012 +#define IDC_SWTHREADS_EDIT 2012 #define IDC_CUSTOM1 2013 #define IDC_CHECK4 2014 #define IDC_FILTER 2015 From f8232c19ac8852d1bf26e5ff53bb19bbb909d917 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 08:55:45 +0000 Subject: [PATCH 183/296] microVU: - Minor Change... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1321 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 1 + 1 file changed, 1 insertion(+) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 592b76774a..bf5bd42951 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -101,6 +101,7 @@ microVUt(void) mVUcheckIsSame(mV) { if (mVU->prog.isSame == 0) { if (!isVU1) mVUcacheProg<0>(mVU->prog.cur); else mVUcacheProg<1>(mVU->prog.cur); + mVU->prog.isSame = 1; } } From e9ad2ffe7bf4087f16dc39ac8b32e7ec9a5b4c57 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 4 Jun 2009 09:30:35 +0000 Subject: [PATCH 184/296] GSdx: added post build step git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1322 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/vsprops/common.vsprops | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/GSdx/vsprops/common.vsprops b/plugins/GSdx/vsprops/common.vsprops index 782976e7bf..29fe0375bb 100644 --- a/plugins/GSdx/vsprops/common.vsprops +++ b/plugins/GSdx/vsprops/common.vsprops @@ -27,7 +27,7 @@ /> Date: Thu, 4 Jun 2009 17:50:56 +0000 Subject: [PATCH 185/296] Fix for Tekken Tag Issue 259 Removed the doubled up timings on SIF DMA cycles (should have been done when it was spotted really) Put in a fix for Scratchpad reads which cause a VTLB error. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1323 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 33 +++++++++++++++++++++++++-------- pcsx2/Hw.h | 4 ++-- pcsx2/IopCounters.cpp | 3 ++- pcsx2/Sif.cpp | 4 ++-- 4 files changed, 31 insertions(+), 13 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 039cc437b1..a5e44ff7b4 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -250,7 +250,7 @@ void GIFdma() if (((gif->qwc == 0) && (gif->chcr & 0xc) == 0)) gspath3done = 1; - else if(gif->qwc > 0) + else if(gif->qwc > 0 || gscycles) { GIFdmaEnd(); return; @@ -328,19 +328,20 @@ void dmaGIF() { //We used to addd wait time for the buffer to fill here, fixing some timing problems in path 3 masking //It takes the time of 24 QW for the BUS to become ready - The Punisher, And1 Streetball GIF_LOG("dmaGIFstart chcr = %lx, madr = %lx, qwc = %lx\n tadr = %lx, asr0 = %lx, asr1 = %lx", gif->chcr, gif->madr, gif->qwc, gif->tadr, gif->asr0, gif->asr1); - if ((psHu32(DMAC_CTRL) & 0xC) == 0xC ) { // GIF MFIFO - //Console::WriteLn("GIF MFIFO"); - gifMFIFOInterrupt(); - return; - } - + Path3progress = 2; gspath3done = 0; // For some reason this doesnt clear? So when the system starts the thread, we will clear it :) psHu32(GIF_STAT) |= GIF_STAT_P3Q; GSCSRr &= ~0xC000; //Clear FIFO stuff GSCSRr |= 0x8000; //FIFO full psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [used to be 0xE00; // OPH=1 | APATH=3] - Path3progress = 2; + + if ((psHu32(DMAC_CTRL) & 0xC) == 0xC ) { // GIF MFIFO + //Console::WriteLn("GIF MFIFO"); + gifMFIFOInterrupt(); + return; + } + if ((gif->qwc == 0) && ((gif->chcr & 0xc) != 0)){ u32 *ptag; @@ -536,11 +537,22 @@ void mfifoGIFtransfer(int qwc) { void gifMFIFOInterrupt() { + mfifocycles = 0; + if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + if (!(gif->chcr & 0x100)) { Console::WriteLn("WTF GIFMFIFO"); cpuRegs.interrupt &= ~(1 << 11); return ; } + + if(((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == 0) //Path2 gets priority in intermittent mode + { + //GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); + CPU_INT(11,mfifocycles); + return; + } + if((spr0->chcr & 0x100) && spr0->qwc == 0) { spr0->chcr &= ~0x100; @@ -566,6 +578,11 @@ void gifMFIFOInterrupt() #endif //if(gifqwc > 0) Console::WriteLn("GIF MFIFO ending with stuff in it %x", params gifqwc); if (!gifmfifoirq) gifqwc = 0; + + gspath3done = 0; + gscycles = 0; + psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + psHu32(GIF_STAT) &= ~GIF_STAT_P3Q; gifstate = GIF_STATE_READY; gif->chcr &= ~0x100; vif1Regs->stat &= ~VIF1_STAT_VGW; diff --git a/pcsx2/Hw.h b/pcsx2/Hw.h index 0b0dc0cb0e..7253727b1f 100644 --- a/pcsx2/Hw.h +++ b/pcsx2/Hw.h @@ -434,8 +434,8 @@ static __forceinline void *dmaGetAddr(u32 addr) { // if (addr & 0xf) { DMA_LOG("*PCSX2*: DMA address not 128bit aligned: %8.8x", addr); } - // teh sux why the f00k 0xE0000000 - if (addr & 0x80000000) return (void*)&psS[addr & 0x3ff0]; + // Need to check the physical address as well as just the "SPR" flag, as VTLB doesnt seem to handle it + if ((addr & 0x80000000) || (addr & 0x70000000) == 0x70000000) return (void*)&psS[addr & 0x3ff0]; ptr = (u8*)vtlb_GetPhyPtr(addr&0x1FFFFFF0); if (ptr == NULL) { diff --git a/pcsx2/IopCounters.cpp b/pcsx2/IopCounters.cpp index fb38ff3a94..5fa5212043 100644 --- a/pcsx2/IopCounters.cpp +++ b/pcsx2/IopCounters.cpp @@ -576,7 +576,7 @@ __forceinline void psxRcntWmode16( int index, u32 value ) { // gated counters are added up as per the h/vblank timers. // (the PIXEL alt source becomes a vsync gate) - + counter.mode |= IOPCNT_STOPPED; PSXCNT_LOG( "IOP Counter[%d] Gate Check set, value = 0x%04X", index, value ); if( index == 0 ) psxhblankgate |= 1; // fixme: these gate flags should be one var >_< @@ -621,6 +621,7 @@ __forceinline void psxRcntWmode32( int index, u32 value ) if(counter.mode & IOPCNT_ENABLE_GATE) { PSXCNT_LOG("IOP Counter[3] Gate Check set, value = %x", value); + counter.mode |= IOPCNT_STOPPED; psxvblankgate |= 1<<3; } else psxvblankgate &= ~(1<<3); diff --git a/pcsx2/Sif.cpp b/pcsx2/Sif.cpp index 73d1c8d8f0..940fa7a266 100644 --- a/pcsx2/Sif.cpp +++ b/pcsx2/Sif.cpp @@ -212,7 +212,7 @@ __forceinline void SIF0Dma() // Clearing handled by vtlb memory protection and manual blocks. //Cpu->Clear(sif0dma->madr, readSize*4); - cycles += readSize * BIAS; // fixme : BIAS is factored in below + cycles += readSize; // fixme : BIAS is factored in below sif0dma->qwc -= readSize; sif0dma->madr += readSize << 4; } @@ -364,7 +364,7 @@ __forceinline void SIF1Dma() SIF1write(data, qwTransfer << 2); sif1dma->madr += qwTransfer << 4; - cycles += qwTransfer * BIAS; // fixme : BIAS is factored in above + cycles += qwTransfer; // fixme : BIAS is factored in above sif1dma->qwc -= qwTransfer; } } From d1b5ee9f5fab7ffd071a9f5de60adfe07655dde1 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 4 Jun 2009 20:46:05 +0000 Subject: [PATCH 186/296] Couple of very minor changes for GIF/MTGS GIFTag reading git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1324 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 47 +++++++++++++++++++++++------------------------ pcsx2/MTGS.cpp | 37 ++++++++++++++++++++++++------------- 2 files changed, 47 insertions(+), 37 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index a5e44ff7b4..049ed0191a 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -93,7 +93,7 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) if( mtgsThread != NULL ) { - int sizetoread = (qwc)<<4; + int sizetoread = qwc; sizetoread = mtgsThread->PrepDataPacket( GIF_PATH_3, pMem, qwc ); u8* pgsmem = mtgsThread->GetDataPacketPtr(); @@ -122,7 +122,7 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) int _GIFchain() { - u32 qwc = ((psHu32(GIF_MODE) & 0x4) && (vif1Regs->mskpath3)) ? min(8, (int)gif->qwc) : min( gifsplit, (int)gif->qwc ); + u32 qwc = ((psHu32(GIF_MODE) & 0x4) || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : min( gifsplit, (int)gif->qwc ); u32 *pMem; pMem = (u32*)dmaGetAddr(gif->madr); @@ -301,27 +301,26 @@ void GIFdma() } prevcycles = 0; - if (!(vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1))) { - if (gspath3done == 0 && gif->qwc == 0) - { - - ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR - gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag - gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 - gif->madr = ptag[1]; + if (gspath3done == 0 && gif->qwc == 0) + { + + ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR + gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag + gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 + gif->madr = ptag[1]; + + gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); + if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag + GIF_LOG("dmaIrq Set"); + gspath3done = 1; + } + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28) & 0x7, gif->madr); + GIFdmaEnd(); + return; + + } else GIFdmaEnd(); + gscycles = 0; - gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); - if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag - GIF_LOG("dmaIrq Set"); - gspath3done = 1; - } - GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28) & 0x7, gif->madr); - GIFdmaEnd(); - return; - - } else GIFdmaEnd(); - gscycles = 0; - } } void dmaGIF() { @@ -372,7 +371,7 @@ void dmaGIF() { // called from only one location, so forceinline it: static __forceinline int mfifoGIFrbTransfer() { - u32 qwc = (psHu32(GIF_MODE) & 0x4 && vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; + u32 qwc = (psHu32(GIF_MODE) & 0x4 || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; int mfifoqwc = min(gifqwc, qwc); u32 *src; @@ -431,7 +430,7 @@ static __forceinline int mfifoGIFchain() { } else { - int mfifoqwc = (psHu32(GIF_MODE) & 0x4 && vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; + int mfifoqwc = (psHu32(GIF_MODE) & 0x4 || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; u32 *pMem = (u32*)dmaGetAddr(gif->madr); if (pMem == NULL) return -1; diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 659e517967..0cd782c5c0 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -390,13 +390,24 @@ __forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u } } - if(pathidx == 0) + if(path.tag.nloop == 0) { - if(path.tag.eop && path.tag.nloop == 0) + if(path.tag.eop) { - break; - } - /*if((path.tag.nloop > 0 || (!path.tag.eop && path.tag.nloop == 0)) && size == 0) + if(pathidx != 1) + { + break; + } + /*if((path.tag.nloop > 0 || (!path.tag.eop && path.tag.nloop == 0)) && size == 0) + { + if(path1loop == true) return size - 0x400; + //DevCon::Notice("Looping Nloop %x, Eop %x, FLG %x", params path.tag.nloop, path.tag.eop, path.tag.flg); + size = 0x400; + pMem -= 0x4000; + path1loop = true; + }*/ + } + /*else if(size == 0 && pathidx == 0) { if(path1loop == true) return size - 0x400; //DevCon::Notice("Looping Nloop %x, Eop %x, FLG %x", params path.tag.nloop, path.tag.eop, path.tag.flg); @@ -404,15 +415,15 @@ __forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u pMem -= 0x4000; path1loop = true; }*/ - } - if(pathidx == 2) + } + /*else if(size == 0 && pathidx == 0) { - if(path.tag.eop && path.tag.nloop == 0) - { - //if(pathidx == 2) GIF_LOG("BREAK PATH3 NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); - break; - } - } + if(path1loop == true) return size - 0x400; + //DevCon::Notice("Looping Nloop %x, Eop %x, FLG %x", params path.tag.nloop, path.tag.eop, path.tag.flg); + size = 0x400; + pMem -= 0x4000; + path1loop = true; + }*/ } if(pathidx == 0) From 2ffe0e81f7980b3ca2a69e4dd2379524da1e9151 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 4 Jun 2009 22:36:14 +0000 Subject: [PATCH 187/296] pcsx2: - Changed sVU/mVU allocation order and addresses to hopefully solve allocation problems a lot of people were having. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1325 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVU0micro.cpp | 4 ++-- pcsx2/x86/iVU1micro.cpp | 4 ++-- pcsx2/x86/iVUzerorec.cpp | 6 +++--- pcsx2/x86/microVU.cpp | 2 +- 4 files changed, 8 insertions(+), 8 deletions(-) diff --git a/pcsx2/x86/iVU0micro.cpp b/pcsx2/x86/iVU0micro.cpp index 9a3e4e5ce6..c4d4e993f2 100644 --- a/pcsx2/x86/iVU0micro.cpp +++ b/pcsx2/x86/iVU0micro.cpp @@ -27,8 +27,8 @@ namespace VU0micro { - void recAlloc() { initVUrec(&VU0, 0); SuperVUAlloc(0); } - void recShutdown() { closeVUrec(0); SuperVUDestroy(0); } + void recAlloc() { SuperVUAlloc(0); initVUrec(&VU0, 0); } + void recShutdown() { SuperVUDestroy(0); closeVUrec(0); } void __fastcall recClear(u32 Addr, u32 Size) { if (useMVU0) clearVUrec(Addr, Size, 0); diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 9b55671f75..4be95d592d 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -250,8 +250,8 @@ namespace VU1micro namespace VU1micro { - void recAlloc() { initVUrec(&VU1, 1); SuperVUAlloc(1); } - void recShutdown() { closeVUrec(1); SuperVUDestroy(1); } + void recAlloc() { SuperVUAlloc(1); initVUrec(&VU1, 1); } + void recShutdown() { SuperVUDestroy(1); closeVUrec(1);} void __fastcall recClear(u32 Addr, u32 Size) { if (useMVU1) clearVUrec(Addr, Size, 1); diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/iVUzerorec.cpp index 15ad37ab9c..9b3942d71e 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/iVUzerorec.cpp @@ -342,9 +342,9 @@ void SuperVUAlloc(int vuindex) if (s_recVUMem == NULL) { // upper 4 bits must be zero! - // Changed "first try base" to 0xb800000, since 0x0c000000 liked to fail a lot. (air) - s_recVUMem = SysMmapEx(0x0c000000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); - + // Changed "first try base" to 0xf1e0000, since 0x0c000000 liked to fail a lot. (cottonvibes) + s_recVUMem = SysMmapEx(0xf1e0000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); + if (s_recVUMem == NULL) { throw Exception::OutOfMemory( diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 85980e3413..33da63a585 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -47,7 +47,7 @@ microVUf(void) mVUinit(VURegs* vuRegsPtr) { memset(&mVU->prog, 0, sizeof(mVU->prog)); mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); - mVU->cache = SysMmapEx((vuIndex ? 0x0f240000 : 0x0e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); + mVU->cache = SysMmapEx((vuIndex ? 0x5f240000 : 0x5e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", (u32)mVU->cache)); mVUreset(); From 4097f01b35262c4c2a0d21252502f374e323438c Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 5 Jun 2009 00:20:31 +0000 Subject: [PATCH 188/296] Linux:Fixed up the speed hack dialog to show the correct speed hacks for microVU. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1326 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Linux/interface.c | 24 ++++++++---------------- pcsx2/Linux/pcsx2.glade | 28 ++++------------------------ 2 files changed, 12 insertions(+), 40 deletions(-) diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index 94994df4bf..75943ba474 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -645,8 +645,7 @@ create_SpeedHacksDlg (void) GtkWidget *frame41; GtkWidget *alignment36; GtkWidget *vbox75; - GtkWidget *check_microvu_flag_hack_1; - GtkWidget *check_microvu_flag_hack_2; + GtkWidget *check_microvu_flag_hack; GtkWidget *check_microvu_min_max_hack; GtkWidget *label113; GtkWidget *dialog_action_area3; @@ -833,23 +832,17 @@ create_SpeedHacksDlg (void) gtk_widget_show (vbox75); gtk_container_add (GTK_CONTAINER (alignment36), vbox75); - check_microvu_flag_hack_1 = gtk_check_button_new_with_mnemonic (_("Flag Hack 1 (microVU only)")); - gtk_widget_set_name (check_microvu_flag_hack_1, "check_microvu_flag_hack_1"); - gtk_widget_show (check_microvu_flag_hack_1); - gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_flag_hack_1, FALSE, FALSE, 0); - gtk_tooltips_set_tip (tooltips, check_microvu_flag_hack_1, _("Big Speedup and moderately compatible. May cause SPS and Infinite loop problems."), NULL); - - check_microvu_flag_hack_2 = gtk_check_button_new_with_mnemonic (_("Flag Hack 2 (microVU only)")); - gtk_widget_set_name (check_microvu_flag_hack_2, "check_microvu_flag_hack_2"); - gtk_widget_show (check_microvu_flag_hack_2); - gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_flag_hack_2, FALSE, FALSE, 0); - gtk_tooltips_set_tip (tooltips, check_microvu_flag_hack_2, _("Big Speedup and moderately compatible. SuperVU does something similar by default."), NULL); + check_microvu_flag_hack = gtk_check_button_new_with_mnemonic (_("Status Flag Hack (microVU only)")); + gtk_widget_set_name (check_microvu_flag_hack, "check_microvu_flag_hack"); + gtk_widget_show (check_microvu_flag_hack); + gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_flag_hack, FALSE, FALSE, 0); + gtk_tooltips_set_tip (tooltips, check_microvu_flag_hack, _("Big Speedup and high compatiblity. SuperVU does something similar by default."), NULL); check_microvu_min_max_hack = gtk_check_button_new_with_mnemonic (_("Min / Max Hack (microVU only)")); gtk_widget_set_name (check_microvu_min_max_hack, "check_microvu_min_max_hack"); gtk_widget_show (check_microvu_min_max_hack); gtk_box_pack_start (GTK_BOX (vbox75), check_microvu_min_max_hack, FALSE, FALSE, 0); - gtk_tooltips_set_tip (tooltips, check_microvu_min_max_hack, _("Small Speedup. May cause SPS, hanging, and missing geometry."), NULL); + gtk_tooltips_set_tip (tooltips, check_microvu_min_max_hack, _("Small Speedup. May cause SPS, hanging, graphical corruption, and missing geometry."), NULL); label113 = gtk_label_new (_("Vertex Unit (VU) Hacks")); gtk_widget_set_name (label113, "label113"); @@ -923,8 +916,7 @@ create_SpeedHacksDlg (void) GLADE_HOOKUP_OBJECT (SpeedHacksDlg, frame41, "frame41"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, alignment36, "alignment36"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, vbox75, "vbox75"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_flag_hack_1, "check_microvu_flag_hack_1"); - GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_flag_hack_2, "check_microvu_flag_hack_2"); + GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_flag_hack, "check_microvu_flag_hack"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, check_microvu_min_max_hack, "check_microvu_min_max_hack"); GLADE_HOOKUP_OBJECT (SpeedHacksDlg, label113, "label113"); GLADE_HOOKUP_OBJECT_NO_REF (SpeedHacksDlg, dialog_action_area3, "dialog_action_area3"); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index 392c391957..b71dc51c68 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -1702,31 +1702,11 @@ If you have problems, Disable all of these and try again. 0 - + True - Big Speedup and moderately compatible. May cause SPS and Infinite loop problems. + Big Speedup and high compatiblity. SuperVU does something similar by default. True - Flag Hack 1 (microVU only) - True - GTK_RELIEF_NORMAL - True - False - False - True - - - 0 - False - False - - - - - - True - Big Speedup and moderately compatible. SuperVU does something similar by default. - True - Flag Hack 2 (microVU only) + Status Flag Hack (microVU only) True GTK_RELIEF_NORMAL True @@ -1744,7 +1724,7 @@ If you have problems, Disable all of these and try again. True - Small Speedup. May cause SPS, hanging, and missing geometry. + Small Speedup. May cause SPS, hanging, graphical corruption, and missing geometry. True Min / Max Hack (microVU only) True From 154a886bf785a60b35008721959bdb79197b37ac Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Fri, 5 Jun 2009 00:42:27 +0000 Subject: [PATCH 189/296] Re-enabled Precompiled headers in release builds. Believe it or not this fixes crashes in the IPU's coroutine system that started happening in r1319. Yes, as you might recall, we *disabled* PCH on release builds back during Pcsx2-Playground because it was causing crashes in Coroutine. Now re-enabling PCH *fixes* the crashes in Coroutine. I'm not making this up. No, Ferrarinews, this revision *won't* improve GT4's graphics. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1327 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 100 ++++++++++++++++++--- 1 file changed, 87 insertions(+), 13 deletions(-) diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 5950b8d879..f3388e5078 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -224,7 +224,7 @@ SmallerTypeCheck="false" BufferSecurityCheck="false" EnableEnhancedInstructionSet="0" - UsePrecompiledHeader="0" + UsePrecompiledHeader="2" PrecompiledHeaderThrough="PrecompiledHeader.h" PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch" CompileAs="2" @@ -479,6 +479,15 @@ PrecompiledHeaderFile="$(IntDir)\win32.pch" /> + + + + + + + + + + + + + + + @@ -815,7 +861,7 @@ + + + @@ -1301,6 +1356,15 @@ PrecompiledHeaderFile="$(IntDir)\win32.pch" /> + + + + + + - - - @@ -2904,6 +2969,15 @@ PrecompiledHeaderFile="$(IntDir)\win32.pch" /> + + + Date: Fri, 5 Jun 2009 02:24:57 +0000 Subject: [PATCH 190/296] IPU1chain is now a function rather then a define. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1328 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IPU/IPU.cpp | 75 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 56 insertions(+), 19 deletions(-) diff --git a/pcsx2/IPU/IPU.cpp b/pcsx2/IPU/IPU.cpp index 9967133f71..03d488f778 100644 --- a/pcsx2/IPU/IPU.cpp +++ b/pcsx2/IPU/IPU.cpp @@ -174,6 +174,27 @@ void ipuShutdown() { } +void ReportIPU() +{ + Console::WriteLn("g_nDMATransfer = 0x%x.", params g_nDMATransfer); + Console::WriteLn("FIreadpos = 0x%x, FIwritepos = 0x%x.", params FIreadpos, FIwritepos); + Console::WriteLn("fifo_input = 0x%x.", params fifo_input); + Console::WriteLn("FOreadpos = 0x%x, FOwritepos = 0x%x.", params FOreadpos, FOwritepos); + Console::WriteLn("fifo_output = 0x%x.", params fifo_output); + Console::WriteLn("g_BP = 0x%x.", params g_BP); + Console::WriteLn("niq = 0x%x, iq = 0x%x.", params niq, iq); + Console::WriteLn("vqclut = 0x%x.", params vqclut); + Console::WriteLn("s_thresh = 0x%x.", params s_thresh); + Console::WriteLn("coded_block_pattern = 0x%x.", params coded_block_pattern); + Console::WriteLn("g_decoder = 0x%x.", params g_decoder); + Console::WriteLn("mpeg2_scan_norm = 0x%x, mpeg2_scan_alt = 0x%x.", params mpeg2_scan_norm, mpeg2_scan_alt); + Console::WriteLn("g_nCmdPos = 0x%x.", params g_nCmdPos); + Console::WriteLn("g_nCmdIndex = 0x%x.", params g_nCmdIndex); + Console::WriteLn("ipuCurCmd = 0x%x.", params ipuCurCmd); + Console::WriteLn("_readbits = 0x%x.", params _readbits); + Console::WriteLn("temp will equal 0x%x.", params readbits - _readbits); + Console::WriteLn(""); +} // fixme - ipuFreeze looks fairly broken. Should probably take a closer look at some point. void SaveState::ipuFreeze() @@ -1352,22 +1373,33 @@ int FIFOto_write(u32* pMem, int size) return firsttrans; } -// To do: convert this into a static inlined function. -#define IPU1chain() { \ - if (ipu1dma->qwc > 0) \ - { \ - int qwc = ipu1dma->qwc; \ - pMem = (u32*)dmaGetAddr(ipu1dma->madr); \ - if (pMem == NULL) { Console::Error("ipu1dma NULL!"); return totalqwc; } \ - qwc = FIFOto_write(pMem, qwc); \ - ipu1dma->madr += qwc<< 4; \ - ipu1dma->qwc -= qwc; \ - totalqwc += qwc; \ - if( ipu1dma->qwc > 0 ) { \ - g_nDMATransfer |= IPU_DMA_ACTV1; \ - return totalqwc; \ - } \ - } \ +static __forceinline s32 IPU1chain(u32* &pMem, int &totalqwc) +{ + if (ipu1dma->qwc > 0) + { + int qwc = ipu1dma->qwc; + + pMem = (u32*)dmaGetAddr(ipu1dma->madr); + + if (pMem == NULL) + { + Console::Error("ipu1dma NULL!"); + return totalqwc; + } + + qwc = FIFOto_write(pMem, qwc); + ipu1dma->madr += qwc<< 4; + ipu1dma->qwc -= qwc; + totalqwc += qwc; + + if (ipu1dma->qwc > 0) + { + g_nDMATransfer |= IPU_DMA_ACTV1; + return totalqwc; + } + } + + return -1; } extern void gsInterrupt(); @@ -1395,7 +1427,8 @@ int IPU1dma() // in kh, qwc == 0 when dma_actv1 is set if ((g_nDMATransfer & IPU_DMA_ACTV1) && ipu1dma->qwc > 0) { - IPU1chain(); + int temp = IPU1chain(pMem, totalqwc); + if (temp != -1) return temp; //Check TIE bit of CHCR and IRQ bit of tag if ((ipu1dma->chcr & 0x80) && (g_nDMATransfer&IPU_DMA_DOTIE1)) @@ -1467,7 +1500,10 @@ int IPU1dma() { IPU_LOG("dmaIPU1 Normal size=%d, addr=%lx, fifosize=%x", ipu1dma->qwc, ipu1dma->madr, 8 - g_BP.IFC); - IPU1chain(); + + int temp = IPU1chain(pMem, totalqwc); + if (temp != -1) return temp; + IPU_INT_TO((ipu1cycles + totalqwc)*BIAS); return totalqwc; } @@ -1534,7 +1570,8 @@ int IPU1dma() //Britney Dance beat does a blank NEXT tag, for some odd reason the fix doesnt work if after IPU1Chain O_o if ((ipu1dma->qwc == 0) && (!done) && !(g_nDMATransfer & IPU_DMA_DOTIE1)) IPU1dma(); - IPU1chain(); + int temp = IPU1chain(pMem, totalqwc); + if (temp != -1) return temp; if ((ipu1dma->chcr & 0x80) && (ptag[0]&0x80000000) && ipu1dma->qwc == 0) //Check TIE bit of CHCR and IRQ bit of tag { From c668c7c351c305bbb08ab9ca7c037f4affbfd436 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 5 Jun 2009 04:59:13 +0000 Subject: [PATCH 191/296] Revised Ipu1Chain, and tweaked Ipu1Dma. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1329 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IPU/IPU.cpp | 149 +++++++++++++++++++++------------------------- 1 file changed, 67 insertions(+), 82 deletions(-) diff --git a/pcsx2/IPU/IPU.cpp b/pcsx2/IPU/IPU.cpp index 03d488f778..8a183e91a2 100644 --- a/pcsx2/IPU/IPU.cpp +++ b/pcsx2/IPU/IPU.cpp @@ -1373,7 +1373,7 @@ int FIFOto_write(u32* pMem, int size) return firsttrans; } -static __forceinline s32 IPU1chain(u32* &pMem, int &totalqwc) +static __forceinline bool IPU1chain(u32* &pMem, int &totalqwc) { if (ipu1dma->qwc > 0) { @@ -1384,7 +1384,7 @@ static __forceinline s32 IPU1chain(u32* &pMem, int &totalqwc) if (pMem == NULL) { Console::Error("ipu1dma NULL!"); - return totalqwc; + return true; } qwc = FIFOto_write(pMem, qwc); @@ -1395,11 +1395,25 @@ static __forceinline s32 IPU1chain(u32* &pMem, int &totalqwc) if (ipu1dma->qwc > 0) { g_nDMATransfer |= IPU_DMA_ACTV1; - return totalqwc; + return true; } } - - return -1; + return false; +} + +static __forceinline bool IncreaseTadr(u32 tag) +{ + switch (tag & 0x70000000) + { + case 0x00000000: + ipu1dma->tadr += 16; + return true; + + case 0x70000000: + ipu1dma->tadr = ipu1dma->madr; + return true; + } + return false; } extern void gsInterrupt(); @@ -1427,95 +1441,76 @@ int IPU1dma() // in kh, qwc == 0 when dma_actv1 is set if ((g_nDMATransfer & IPU_DMA_ACTV1) && ipu1dma->qwc > 0) { - int temp = IPU1chain(pMem, totalqwc); - if (temp != -1) return temp; + if (IPU1chain(pMem, totalqwc)) return totalqwc; //Check TIE bit of CHCR and IRQ bit of tag - if ((ipu1dma->chcr & 0x80) && (g_nDMATransfer&IPU_DMA_DOTIE1)) + if ((ipu1dma->chcr & 0x80) && (g_nDMATransfer & IPU_DMA_DOTIE1)) { Console::WriteLn("IPU1 TIE"); - IPU_INT_TO(totalqwc*BIAS); + IPU_INT_TO(totalqwc * BIAS); g_nDMATransfer &= ~(IPU_DMA_ACTV1 | IPU_DMA_DOTIE1); g_nDMATransfer |= IPU_DMA_TIE1; return totalqwc; } - if ((ipu1dma->chcr&0xc) == 0) + if (!(ipu1dma->chcr & 0xc)) { - IPU_INT_TO(totalqwc*BIAS); + IPU_INT_TO(totalqwc * BIAS); return totalqwc; } else { u32 tag = ipu1dma->chcr; // upper bits describe current tag - if ((ipu1dma->chcr & 0x80) && (tag&0x80000000)) + if ((ipu1dma->chcr & 0x80) && (tag & 0x80000000)) { ptag = (u32*)dmaGetAddr(ipu1dma->tadr); - switch (tag&0x70000000) - { - case 0x00000000: - ipu1dma->tadr += 16; - break; - case 0x70000000: - ipu1dma->tadr = ipu1dma->madr; - break; - } + IncreaseTadr(tag); ipu1dma->chcr = (ipu1dma->chcr & 0xFFFF) | ((*ptag) & 0xFFFF0000); IPU_LOG("IPU dmaIrq Set"); - IPU_INT_TO(totalqwc*BIAS); + IPU_INT_TO(totalqwc * BIAS); g_nDMATransfer |= IPU_DMA_TIE1; return totalqwc; } - switch (tag&0x70000000) + if (IncreaseTadr(tag)) { - case 0x00000000: - ipu1dma->tadr += 16; - IPU_INT_TO((1 + totalqwc)*BIAS); - return totalqwc; - - case 0x70000000: - ipu1dma->tadr = ipu1dma->madr; - IPU_INT_TO((1 + totalqwc)*BIAS); - return totalqwc; + IPU_INT_TO((1 + totalqwc)*BIAS); + return totalqwc; } } g_nDMATransfer &= ~(IPU_DMA_ACTV1 | IPU_DMA_DOTIE1); } - - if ((ipu1dma->chcr & 0xc) == 0 && ipu1dma->qwc == 0) // Normal Mode + + if (((ipu1dma->chcr & 0xc) == 0) && (ipu1dma->qwc == 0)) // Normal Mode { //Console::WriteLn("ipu1 normal empty qwc?"); return totalqwc; } // Transfer Dn_QWC from Dn_MADR to GIF - - if ((ipu1dma->chcr & 0xc) == 0 || ipu1dma->qwc > 0) // Normal Mode + if (ipu1dma->qwc > 0) { IPU_LOG("dmaIPU1 Normal size=%d, addr=%lx, fifosize=%x", ipu1dma->qwc, ipu1dma->madr, 8 - g_BP.IFC); - int temp = IPU1chain(pMem, totalqwc); - if (temp != -1) return temp; + if (!IPU1chain(pMem, totalqwc)) IPU_INT_TO((ipu1cycles + totalqwc) * BIAS); - IPU_INT_TO((ipu1cycles + totalqwc)*BIAS); return totalqwc; } else { - // Chain Mode + // Chain Mode & ipu1dma->qwc is 0 ptag = (u32*)dmaGetAddr(ipu1dma->tadr); //Set memory pointer to TADR if (ptag == NULL) //Is ptag empty? { Console::Error("IPU1 BUSERR"); ipu1dma->chcr = (ipu1dma->chcr & 0xFFFF) | ((*ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 - psHu32(DMAC_STAT) |= 1 << 15; //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register return totalqwc; } @@ -1566,55 +1561,45 @@ int IPU1dma() g_nDMATransfer |= IPU_DMA_DOTIE1; else g_nDMATransfer &= ~IPU_DMA_DOTIE1; - - //Britney Dance beat does a blank NEXT tag, for some odd reason the fix doesnt work if after IPU1Chain O_o - if ((ipu1dma->qwc == 0) && (!done) && !(g_nDMATransfer & IPU_DMA_DOTIE1)) IPU1dma(); - - int temp = IPU1chain(pMem, totalqwc); - if (temp != -1) return temp; - - if ((ipu1dma->chcr & 0x80) && (ptag[0]&0x80000000) && ipu1dma->qwc == 0) //Check TIE bit of CHCR and IRQ bit of tag - { - Console::WriteLn("IPU1 TIE"); - - if (done) - { - ptag = (u32*)dmaGetAddr(ipu1dma->tadr); - - switch (ptag[0]&0x70000000) - { - case 0x00000000: - ipu1dma->tadr += 16; - break; - case 0x70000000: - ipu1dma->tadr = ipu1dma->madr; - break; - } - - ipu1dma->chcr = (ipu1dma->chcr & 0xFFFF) | ((*ptag) & 0xFFFF0000); - } - - IPU_INT_TO(ipu1cycles + totalqwc*BIAS); - g_nDMATransfer |= IPU_DMA_TIE1; - return totalqwc; - } - + if (ipu1dma->qwc == 0) { - switch (ptag[0]&0x70000000) + //if ((ipu1dma->chcr & 0x80) && (ptag[0] & 0x80000000)) //Check TIE bit of CHCR and IRQ bit of tag + if (g_nDMATransfer & IPU_DMA_DOTIE1) { - case 0x00000000: - ipu1dma->tadr += 16; - break; + Console::WriteLn("IPU1 TIE"); + + if (IPU1chain(pMem, totalqwc)) return totalqwc; + + if (done) + { + ptag = (u32*)dmaGetAddr(ipu1dma->tadr); - case 0x70000000: - ipu1dma->tadr = ipu1dma->madr; - break; + IncreaseTadr(ptag[0]); + + ipu1dma->chcr = (ipu1dma->chcr & 0xFFFF) | ((*ptag) & 0xFFFF0000); + } + + IPU_INT_TO(ipu1cycles + totalqwc * BIAS); // Should it be (ipu1cycles + totalqwc) * BIAS? + g_nDMATransfer |= IPU_DMA_TIE1; + return totalqwc; } + else + { + //Britney Dance beat does a blank NEXT tag, for some odd reason the fix doesnt work if after IPU1Chain O_o + if (!done) IPU1dma(); + if (IPU1chain(pMem, totalqwc)) return totalqwc; + } + + IncreaseTadr(ptag[0]); + } + else + { + if (IPU1chain(pMem, totalqwc)) return totalqwc; } } - IPU_INT_TO((ipu1cycles + totalqwc)*BIAS); + IPU_INT_TO((ipu1cycles + totalqwc) * BIAS); return totalqwc; } From 3f97423a30e16bed7bd51382035a3aad03039d17 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Fri, 5 Jun 2009 12:33:07 +0000 Subject: [PATCH 192/296] GSdx: Found three other instances of CComPtr that could be removed. (DX9 only) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1330 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTexture9.cpp | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index fbb0d9a966..6f8b1a5ff3 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -86,7 +86,7 @@ int GSTexture9::GetFormat() const bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) { - if(CComPtr surface = *this) + if(IDirect3DSurface9* surface = *this) { D3DLOCKED_RECT lr; @@ -118,7 +118,7 @@ bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) { HRESULT hr; - if(CComPtr surface = *this) + if(IDirect3DSurface9* surface = *this) { D3DLOCKED_RECT lr; @@ -136,7 +136,7 @@ bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) void GSTexture9::Unmap() { - if(CComPtr surface = *this) + if(IDirect3DSurface9* surface = *this) { surface->UnlockRect(); } From 01d1573fdcff0ce8be85bc2e712dfdee562278bd Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 5 Jun 2009 15:15:02 +0000 Subject: [PATCH 193/296] GSdx: cleaned up some obsolete code git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1331 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSLocalMemory.cpp | 100 ---------------------- plugins/GSdx/GSLocalMemory.h | 143 -------------------------------- plugins/GSdx/GSRendererHW.h | 10 +-- plugins/GSdx/GSRendererHW10.cpp | 6 +- plugins/GSdx/GSState.cpp | 46 ++++++++-- plugins/GSdx/GSTexture10.cpp | 14 ---- plugins/GSdx/GSTexture10.h | 4 - plugins/GSdx/GSTexture7.cpp | 10 --- plugins/GSdx/GSTexture7.h | 4 +- plugins/GSdx/GSTexture9.cpp | 38 ++------- plugins/GSdx/GSTexture9.h | 3 +- plugins/GSdx/GSVertex.h | 4 +- plugins/GSdx/GSVertexHW.h | 20 ++--- 13 files changed, 64 insertions(+), 338 deletions(-) diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 67495d0aad..004aae25ff 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -571,106 +571,6 @@ GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, cons return o; } -bool GSLocalMemory::FillRect(const GSVector4i& r, uint32 c, uint32 psm, uint32 bp, uint32 bw) -{ - const psm_t& tbl = m_psm[psm]; - - writePixel wp = tbl.wp; - pixelAddress bn = tbl.bn; - - int w = tbl.bs.x; - int h = tbl.bs.y; - int bpp = tbl.bpp; - - switch(bpp) - { - case 32: break; - case 16: c = (c & 0xffff) * 0x00010001; break; - case 8: c = (c & 0xff) * 0x01010101; break; - case 4: c = (c & 0xf) * 0x11111111; break; - } - - GSVector4i clip = r.ralign(tbl.bs); - - for(int y = r.top; y < clip.top; y++) - { - for(int x = r.left; x < r.right; x++) - { - (this->*wp)(x, y, c, bp, bw); - } - } - - for(int y = clip.bottom; y < r.bottom; y++) - { - for(int x = r.left; x < r.right; x++) - { - (this->*wp)(x, y, c, bp, bw); - } - } - - if(r.x < clip.left || clip.right < r.z) - { - for(int y = clip.top; y < clip.bottom; y += h) - { - for(int ys = y, ye = y + h; ys < ye; ys++) - { - for(int x = r.left; x < clip.left; x++) - { - (this->*wp)(x, ys, c, bp, bw); - } - - for(int x = clip.right; x < r.right; x++) - { - (this->*wp)(x, ys, c, bp, bw); - } - } - } - } - - GSVector4i c128(c); - - if(psm == PSM_PSMCT24 || psm == PSM_PSMZ24) - { - GSVector4i mask(0x00ffffff); - - for(int y = clip.top; y < clip.bottom; y += h) - { - for(int x = clip.left; x < clip.right; x += w) - { - GSVector4i* p = (GSVector4i*)&m_vm8[bn(x, y, bp, bw) << 8]; - - for(int i = 0; i < 16; i += 4) - { - p[i + 0] = p[i + 0].blend8(c128, mask); - p[i + 1] = p[i + 1].blend8(c128, mask); - p[i + 2] = p[i + 2].blend8(c128, mask); - p[i + 3] = p[i + 3].blend8(c128, mask); - } - } - } - } - else - { - for(int y = clip.top; y < clip.bottom; y += h) - { - for(int x = clip.left; x < clip.right; x += w) - { - GSVector4i* p = (GSVector4i*)&m_vm8[bn(x, y, bp, bw) << 8]; - - for(int i = 0; i < 16; i += 4) - { - p[i + 0] = c128; - p[i + 1] = c128; - p[i + 2] = c128; - p[i + 3] = c128; - } - } - } - } - - return true; -} - //////////////////// template diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index 4f111289a3..1a966ae603 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -753,149 +753,6 @@ public: // - __forceinline uint32 PixelAddressX(int PSM, int x, int y, uint32 bp, uint32 bw) - { - switch(PSM) - { - case PSM_PSMCT32: return PixelAddress32(x, y, bp, bw); - case PSM_PSMCT24: return PixelAddress32(x, y, bp, bw); - case PSM_PSMCT16: return PixelAddress16(x, y, bp, bw); - case PSM_PSMCT16S: return PixelAddress16S(x, y, bp, bw); - case PSM_PSMT8: return PixelAddress8(x, y, bp, bw); - case PSM_PSMT4: return PixelAddress4(x, y, bp, bw); - case PSM_PSMT8H: return PixelAddress32(x, y, bp, bw); - case PSM_PSMT4HL: return PixelAddress32(x, y, bp, bw); - case PSM_PSMT4HH: return PixelAddress32(x, y, bp, bw); - case PSM_PSMZ32: return PixelAddress32Z(x, y, bp, bw); - case PSM_PSMZ24: return PixelAddress32Z(x, y, bp, bw); - case PSM_PSMZ16: return PixelAddress16Z(x, y, bp, bw); - case PSM_PSMZ16S: return PixelAddress16SZ(x, y, bp, bw); - default: ASSERT(0); return PixelAddress32(x, y, bp, bw); - } - } - - __forceinline uint32 ReadPixelX(int PSM, uint32 addr) const - { - switch(PSM) - { - case PSM_PSMCT32: return ReadPixel32(addr); - case PSM_PSMCT24: return ReadPixel24(addr); - case PSM_PSMCT16: return ReadPixel16(addr); - case PSM_PSMCT16S: return ReadPixel16(addr); - case PSM_PSMT8: return ReadPixel8(addr); - case PSM_PSMT4: return ReadPixel4(addr); - case PSM_PSMT8H: return ReadPixel8H(addr); - case PSM_PSMT4HL: return ReadPixel4HL(addr); - case PSM_PSMT4HH: return ReadPixel4HH(addr); - case PSM_PSMZ32: return ReadPixel32(addr); - case PSM_PSMZ24: return ReadPixel24(addr); - case PSM_PSMZ16: return ReadPixel16(addr); - case PSM_PSMZ16S: return ReadPixel16(addr); - default: ASSERT(0); return ReadPixel32(addr); - } - } - - __forceinline uint32 ReadFrameX(int PSM, uint32 addr) const - { - switch(PSM) - { - case PSM_PSMCT32: return ReadPixel32(addr); - case PSM_PSMCT24: return ReadFrame24(addr); - case PSM_PSMCT16: return ReadFrame16(addr); - case PSM_PSMCT16S: return ReadFrame16(addr); - case PSM_PSMZ32: return ReadPixel32(addr); - case PSM_PSMZ24: return ReadFrame24(addr); - case PSM_PSMZ16: return ReadFrame16(addr); - case PSM_PSMZ16S: return ReadFrame16(addr); - default: ASSERT(0); return ReadPixel32(addr); - } - } - - __forceinline uint32 ReadTexelX(int PSM, uint32 addr, const GIFRegTEXA& TEXA) const - { - switch(PSM) - { - case PSM_PSMCT32: return ReadTexel32(addr, TEXA); - case PSM_PSMCT24: return ReadTexel24(addr, TEXA); - case PSM_PSMCT16: return ReadTexel16(addr, TEXA); - case PSM_PSMCT16S: return ReadTexel16(addr, TEXA); - case PSM_PSMT8: return ReadTexel8(addr, TEXA); - case PSM_PSMT4: return ReadTexel4(addr, TEXA); - case PSM_PSMT8H: return ReadTexel8H(addr, TEXA); - case PSM_PSMT4HL: return ReadTexel4HL(addr, TEXA); - case PSM_PSMT4HH: return ReadTexel4HH(addr, TEXA); - case PSM_PSMZ32: return ReadTexel32(addr, TEXA); - case PSM_PSMZ24: return ReadTexel24(addr, TEXA); - case PSM_PSMZ16: return ReadTexel16(addr, TEXA); - case PSM_PSMZ16S: return ReadTexel16(addr, TEXA); - default: ASSERT(0); return ReadTexel32(addr, TEXA); - } - } - - __forceinline uint32 ReadTexelX(int PSM, int x, int y, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const - { - switch(PSM) - { - case PSM_PSMCT32: return ReadTexel32(x, y, TEX0, TEXA); - case PSM_PSMCT24: return ReadTexel24(x, y, TEX0, TEXA); - case PSM_PSMCT16: return ReadTexel16(x, y, TEX0, TEXA); - case PSM_PSMCT16S: return ReadTexel16(x, y, TEX0, TEXA); - case PSM_PSMT8: return ReadTexel8(x, y, TEX0, TEXA); - case PSM_PSMT4: return ReadTexel4(x, y, TEX0, TEXA); - case PSM_PSMT8H: return ReadTexel8H(x, y, TEX0, TEXA); - case PSM_PSMT4HL: return ReadTexel4HL(x, y, TEX0, TEXA); - case PSM_PSMT4HH: return ReadTexel4HH(x, y, TEX0, TEXA); - case PSM_PSMZ32: return ReadTexel32Z(x, y, TEX0, TEXA); - case PSM_PSMZ24: return ReadTexel24Z(x, y, TEX0, TEXA); - case PSM_PSMZ16: return ReadTexel16Z(x, y, TEX0, TEXA); - case PSM_PSMZ16S: return ReadTexel16Z(x, y, TEX0, TEXA); - default: ASSERT(0); return ReadTexel32(x, y, TEX0, TEXA); - } - } - - __forceinline void WritePixelX(int PSM, uint32 addr, uint32 c) - { - switch(PSM) - { - case PSM_PSMCT32: WritePixel32(addr, c); break; - case PSM_PSMCT24: WritePixel24(addr, c); break; - case PSM_PSMCT16: WritePixel16(addr, c); break; - case PSM_PSMCT16S: WritePixel16(addr, c); break; - case PSM_PSMT8: WritePixel8(addr, c); break; - case PSM_PSMT4: WritePixel4(addr, c); break; - case PSM_PSMT8H: WritePixel8H(addr, c); break; - case PSM_PSMT4HL: WritePixel4HL(addr, c); break; - case PSM_PSMT4HH: WritePixel4HH(addr, c); break; - case PSM_PSMZ32: WritePixel32(addr, c); break; - case PSM_PSMZ24: WritePixel24(addr, c); break; - case PSM_PSMZ16: WritePixel16(addr, c); break; - case PSM_PSMZ16S: WritePixel16(addr, c); break; - default: ASSERT(0); WritePixel32(addr, c); break; - } - } - - __forceinline void WriteFrameX(int PSM, uint32 addr, uint32 c) - { - switch(PSM) - { - case PSM_PSMCT32: WritePixel32(addr, c); break; - case PSM_PSMCT24: WritePixel24(addr, c); break; - case PSM_PSMCT16: WriteFrame16(addr, c); break; - case PSM_PSMCT16S: WriteFrame16(addr, c); break; - case PSM_PSMZ32: WritePixel32(addr, c); break; - case PSM_PSMZ24: WritePixel24(addr, c); break; - case PSM_PSMZ16: WriteFrame16(addr, c); break; - case PSM_PSMZ16S: WriteFrame16(addr, c); break; - default: ASSERT(0); WritePixel32(addr, c); break; - } - } - - // FillRect - - bool FillRect(const GSVector4i& r, uint32 c, uint32 psm, uint32 bp, uint32 bw); - - // - template void WriteImageColumn(int l, int r, int y, int h, uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF); diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 9dcf987834..8155aaffca 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -77,10 +77,10 @@ protected: { for(int j = m_count - 3; i < j; i += 4) { - GSVector4 v0 = GSVector4(v[i + 0].m128[0]); - GSVector4 v1 = GSVector4(v[i + 1].m128[0]); - GSVector4 v2 = GSVector4(v[i + 2].m128[0]); - GSVector4 v3 = GSVector4(v[i + 3].m128[0]); + GSVector4 v0 = v[i + 0].vf[0]; + GSVector4 v1 = v[i + 1].vf[0]; + GSVector4 v2 = v[i + 2].vf[0]; + GSVector4 v3 = v[i + 3].vf[0]; minv = minv.minv((v0.minv(v1)).minv(v2.minv(v3))); maxv = maxv.maxv((v0.maxv(v1)).maxv(v2.maxv(v3))); @@ -88,7 +88,7 @@ protected: for(int j = m_count; i < j; i++) { - GSVector4 v0 = GSVector4(v[i + 0].m128[0]); + GSVector4 v0 = v[i + 0].vf[0]; minv = minv.minv(v0); maxv = maxv.maxv(v0); diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 5f085798a1..26a3ddee7c 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -75,8 +75,8 @@ void GSRendererHW10::VertexKick(bool skip) { GSVertexHW10& dst = m_vl.AddTail(); - dst.m128i[0] = m_v.m128i[0]; - dst.m128i[1] = m_v.m128i[1]; + dst.vi[0] = m_v.vi[0]; + dst.vi[1] = m_v.vi[1]; if(tme && fst) { @@ -513,7 +513,7 @@ void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) for(int i = 0, j = m_count; i < j; i++) { - GSVector4 p(GSVector4i(m_vertices[i].m128i[0]).uph16()); + GSVector4 p(m_vertices[i].vi[0].uph16()); pmin = p.minv(pmin); pmax = p.maxv(pmax); diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 672103be6e..9a2d347065 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -968,25 +968,53 @@ void GSState::Write(uint8* mem, int len) int w = m_env.TRXREG.RRW; int h = m_env.TRXREG.RRH; - // TRACE(_T("Write len=%d DBP=%05x DBW=%d DPSM=%d DSAX=%d DSAY=%d RRW=%d RRH=%d\n"), len, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, m_env.TRXPOS.DSAX, m_env.TRXPOS.DSAY, m_env.TRXREG.RRW, m_env.TRXREG.RRH); + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM]; - if(!m_tr.Update(w, h, GSLocalMemory::m_psm[m_env.BITBLTBUF.DPSM].trbpp, len)) + // printf("Write len=%d DBP=%05x DBW=%d DPSM=%d DSAX=%d DSAY=%d RRW=%d RRH=%d\n", len, m_env.BITBLTBUF.DBP, m_env.BITBLTBUF.DBW, m_env.BITBLTBUF.DPSM, m_env.TRXPOS.DSAX, m_env.TRXPOS.DSAY, m_env.TRXREG.RRW, m_env.TRXREG.RRH); + + if(!m_tr.Update(w, h, psm.trbpp, len)) { return; } - memcpy(&m_tr.buff[m_tr.end], mem, len); - - m_tr.end += len; - if(PRIM->TME && (m_env.BITBLTBUF.DBP == m_context->TEX0.TBP0 || m_env.BITBLTBUF.DBP == m_context->TEX0.CBP)) // TODO: hmmmm { FlushPrim(); } - if(m_tr.end >= m_tr.total) + if(m_tr.end == 0 && len >= m_tr.total) { - FlushWrite(); + // received all data in one piece, no need to buffer it + + // printf("%d >= %d\n", len, m_tr.total); + + (m_mem.*psm.wi)(m_tr.x, m_tr.y, mem, m_tr.total, m_env.BITBLTBUF, m_env.TRXPOS, m_env.TRXREG); + + m_tr.start = m_tr.end = m_tr.total; + + m_perfmon.Put(GSPerfMon::Swizzle, len); + + GSVector4i r; + + r.left = m_env.TRXPOS.DSAX; + r.top = m_env.TRXPOS.DSAY; + r.right = r.left + m_env.TRXREG.RRW; + r.bottom = r.top + m_env.TRXREG.RRH; + + InvalidateVideoMem(m_env.BITBLTBUF, r); + } + else + { + // printf("%d += %d (%d)\n", m_tr.end, len, m_tr.total); + + memcpy(&m_tr.buff[m_tr.end], mem, len); + + m_tr.end += len; + + if(m_tr.end >= m_tr.total) + { + FlushWrite(); + } } m_mem.m_clut.Invalidate(); @@ -1001,7 +1029,7 @@ void GSState::Read(uint8* mem, int len) int w = m_env.TRXREG.RRW; int h = m_env.TRXREG.RRH; - // TRACE(_T("Read len=%d SBP=%05x SBW=%d SPSM=%d SSAX=%d SSAY=%d RRW=%d RRH=%d\n"), len, (int)m_env.BITBLTBUF.SBP, (int)m_env.BITBLTBUF.SBW, (int)m_env.BITBLTBUF.SPSM, sx, sy, w, h); + // printf("Read len=%d SBP=%05x SBW=%d SPSM=%d SSAX=%d SSAY=%d RRW=%d RRH=%d\n", len, (int)m_env.BITBLTBUF.SBP, (int)m_env.BITBLTBUF.SBW, (int)m_env.BITBLTBUF.SPSM, sx, sy, w, h); if(!m_tr.Update(w, h, GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].trbpp, len)) { diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 830a1d5878..8b4de5b233 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -22,11 +22,6 @@ #include "stdafx.h" #include "GSTexture10.h" -GSTexture10::GSTexture10() -{ - memset(&m_desc, 0, sizeof(m_desc)); -} - GSTexture10::GSTexture10(ID3D10Texture2D* texture) : m_texture(texture) { @@ -36,10 +31,6 @@ GSTexture10::GSTexture10(ID3D10Texture2D* texture) m_texture->GetDesc(&m_desc); } -GSTexture10::~GSTexture10() -{ -} - GSTexture10::operator bool() { return !!m_texture; @@ -174,11 +165,6 @@ bool GSTexture10::Save(const string& fn, bool dds) return SUCCEEDED(D3DX10SaveTextureToFile(res, dds ? D3DX10_IFF_DDS : D3DX10_IFF_BMP, fn.c_str())); } -ID3D10Texture2D* GSTexture10::operator->() -{ - return m_texture; -} - GSTexture10::operator ID3D10Texture2D*() { return m_texture; diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index e11233d48f..a74b0ee8ac 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -33,9 +33,7 @@ class GSTexture10 : public GSTexture CComPtr m_dsv; public: - GSTexture10(); explicit GSTexture10(ID3D10Texture2D* texture); - virtual ~GSTexture10(); operator bool(); @@ -48,8 +46,6 @@ public: void Unmap(); bool Save(const string& fn, bool dds = false); - ID3D10Texture2D* operator->(); // TODO: remove direct access - operator ID3D10Texture2D*(); operator ID3D10ShaderResourceView*(); operator ID3D10RenderTargetView*(); diff --git a/plugins/GSdx/GSTexture7.cpp b/plugins/GSdx/GSTexture7.cpp index 28c5fafe81..6c3d23b916 100644 --- a/plugins/GSdx/GSTexture7.cpp +++ b/plugins/GSdx/GSTexture7.cpp @@ -22,12 +22,6 @@ #include "stdafx.h" #include "GSTexture7.h" -GSTexture7::GSTexture7() - : m_type(GSTexture::None) -{ - memset(&m_desc, 0, sizeof(m_desc)); -} - GSTexture7::GSTexture7(int type, IDirectDrawSurface7* system) : m_type(type) , m_system(system) @@ -51,10 +45,6 @@ GSTexture7::GSTexture7(int type, IDirectDrawSurface7* system, IDirectDrawSurface video->GetSurfaceDesc(&m_desc); } -GSTexture7::~GSTexture7() -{ -} - GSTexture7::operator bool() { return !!m_system; diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index 21baa62d95..99c1c73b65 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -32,10 +32,8 @@ class GSTexture7 : public GSTexture DDSURFACEDESC2 m_desc; public: - GSTexture7(); - explicit GSTexture7(int type, IDirectDrawSurface7* system); + GSTexture7(int type, IDirectDrawSurface7* system); GSTexture7(int type, IDirectDrawSurface7* system, IDirectDrawSurface7* video); - virtual ~GSTexture7(); operator bool(); diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index 6f8b1a5ff3..50799a6767 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -22,11 +22,6 @@ #include "stdafx.h" #include "GSTexture9.h" -GSTexture9::GSTexture9() -{ - memset(&m_desc, 0, sizeof(m_desc)); -} - GSTexture9::GSTexture9(IDirect3DSurface9* surface) { m_surface = surface; @@ -49,6 +44,8 @@ GSTexture9::GSTexture9(IDirect3DTexture9* texture) texture->GetDevice(&m_dev); texture->GetLevelDesc(0, &m_desc); texture->GetSurfaceLevel(0, &m_surface); + + ASSERT(m_surface != NULL); } GSTexture9::~GSTexture9() @@ -86,11 +83,11 @@ int GSTexture9::GetFormat() const bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) { - if(IDirect3DSurface9* surface = *this) + if(m_surface) { D3DLOCKED_RECT lr; - if(SUCCEEDED(surface->LockRect(&lr, r, 0))) + if(SUCCEEDED(m_surface->LockRect(&lr, r, 0))) { uint8* src = (uint8*)data; uint8* dst = (uint8*)lr.pBits; @@ -105,7 +102,7 @@ bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) memcpy(dst, src, bytes); } - surface->UnlockRect(); + m_surface->UnlockRect(); return true; } @@ -118,11 +115,11 @@ bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) { HRESULT hr; - if(IDirect3DSurface9* surface = *this) + if(m_surface) { D3DLOCKED_RECT lr; - if(SUCCEEDED(hr = surface->LockRect(&lr, (LPRECT)r, 0))) + if(SUCCEEDED(hr = m_surface->LockRect(&lr, (LPRECT)r, 0))) { *bits = (uint8*)lr.pBits; pitch = (int)lr.Pitch; @@ -136,9 +133,9 @@ bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) void GSTexture9::Unmap() { - if(IDirect3DSurface9* surface = *this) + if(m_surface) { - surface->UnlockRect(); + m_surface->UnlockRect(); } } @@ -196,29 +193,12 @@ bool GSTexture9::Save(const string& fn, bool dds) return false; } -IDirect3DTexture9* GSTexture9::operator->() -{ - return m_texture; -} - GSTexture9::operator IDirect3DSurface9*() { - if(m_texture && !m_surface) - { - m_texture->GetSurfaceLevel(0, &m_surface); - } - return m_surface; } GSTexture9::operator IDirect3DTexture9*() { - if(m_surface && !m_texture) - { - m_surface->GetContainer(__uuidof(IDirect3DTexture9), (void**)&m_texture); - - ASSERT(m_texture); - } - return m_texture; } diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index f8779e3222..5ee23c9370 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -31,7 +31,6 @@ class GSTexture9 : public GSTexture D3DSURFACE_DESC m_desc; public: - GSTexture9(); explicit GSTexture9(IDirect3DSurface9* surface); explicit GSTexture9(IDirect3DTexture9* texture); virtual ~GSTexture9(); @@ -47,7 +46,7 @@ public: void Unmap(); bool Save(const string& fn, bool dds = false); - IDirect3DTexture9* operator->(); // TODO: remove direct access + //IDirect3DTexture9* operator->(); // TODO: remove direct access operator IDirect3DSurface9*(); operator IDirect3DTexture9*(); diff --git a/plugins/GSdx/GSVertex.h b/plugins/GSdx/GSVertex.h index 9caff2d673..ee3bdc9874 100644 --- a/plugins/GSdx/GSVertex.h +++ b/plugins/GSdx/GSVertex.h @@ -40,8 +40,8 @@ __declspec(align(16)) struct GSVertex GIFRegFOG FOG; }; - struct {__m128i m128i[2];}; - struct {__m128 m128[2];}; + struct {GSVector4i vi[2];}; + struct {GSVector4 vf[2];}; }; GIFRegUV UV; diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index 7afaf65ac4..f43c718692 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -35,15 +35,11 @@ __declspec(align(16)) union GSVertexHW9 GSVector4 p; }; - #if _M_SSE >= 0x200 - - struct {__m128i m128i[2];}; - struct {__m128 m128[2];}; + struct {GSVector4i vi[2];}; + struct {GSVector4 vf[2];}; - GSVertexHW9& operator = (GSVertexHW9& v) {m128i[0] = v.m128i[0]; m128i[1] = v.m128i[1]; return *this;} + GSVertexHW9& operator = (GSVertexHW9& v) {vi[0] = v.vi[0]; vi[1] = v.vi[1]; return *this;} - #endif - float GetQ() {return p.w;} }; @@ -76,14 +72,10 @@ __declspec(align(16)) union GSVertexHW10 }; }; - #if _M_SSE >= 0x200 - - struct {__m128i m128i[2];}; - struct {__m128 m128[2];}; + struct {GSVector4i vi[2];}; + struct {GSVector4 vf[2];}; - GSVertexHW10& operator = (GSVertexHW10& v) {m128i[0] = v.m128i[0]; m128i[1] = v.m128i[1]; return *this;} - - #endif + GSVertexHW10& operator = (GSVertexHW10& v) {vi[0] = v.vi[0]; vi[1] = v.vi[1]; return *this;} float GetQ() {return q;} }; From c4531a5c31803a4b73afe4069e62832e3b176b84 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 6 Jun 2009 06:40:29 +0000 Subject: [PATCH 194/296] microVU: - Implemented a faster block-compare algorithm. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1334 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 10 +++--- pcsx2/x86/microVU_Compile.inl | 6 ++-- pcsx2/x86/microVU_IR.h | 17 +++++----- pcsx2/x86/microVU_Misc.h | 13 +++++++- pcsx2/x86/microVU_Misc.inl | 61 ++++++++++++++++++++++++++++++++++- 5 files changed, 90 insertions(+), 17 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 85d9e4163e..febcfbefbf 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -27,17 +27,18 @@ #include "microVU_IR.h" #include "microVU_Misc.h" + #define mMaxBlocks 32 // Max Blocks With Different Pipeline States (For n = 1, 2, 4, 8, 16, etc...) class microBlockManager { private: static const int MaxBlocks = mMaxBlocks - 1; int listSize; // Total Items - 1 int listI; // Index to Add new block - microBlock blockList[mMaxBlocks]; + microBlock* blockList; public: - microBlockManager() { reset(); } - ~microBlockManager() {} + microBlockManager() { blockList = (microBlock*)_aligned_malloc(sizeof(microBlock)*mMaxBlocks, 16); reset(); } + ~microBlockManager() { if (blockList) { _aligned_free(blockList); } } void reset() { listSize = -1; listI = -1; }; microBlock* add(microBlock* pBlock) { microBlock* thisBlock = search(&pBlock->pState); @@ -54,7 +55,7 @@ public: if (listSize < 0) return NULL; if (pState->needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) for (int i = 0; i <= listSize; i++) { - if (!memcmp(pState, &blockList[i].pState, sizeof(microRegInfo)/* - 4*/)) return &blockList[i]; + if (mVUquickSearch((void*)pState, (void*)&blockList[i].pState, sizeof(microRegInfo))) return &blockList[i]; } } else { // Can do Simple Search (Only Matches the Important Pipeline Stuff) @@ -158,6 +159,7 @@ microVUf(void) mVUcacheProg(int progIndex); void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles); void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles); +// recCall Function Pointer typedef void (__fastcall *mVUrecCall)(u32, u32); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index bf5bd42951..7c1ff506c6 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -171,7 +171,7 @@ microVUt(void) mVUincCycles(mV, int x) { } microVUt(void) mVUsetCycles(mV) { - incCycles(mVUstall); + mVUincCycles(mVU, mVUstall); // If upper Op && lower Op write to same VF reg: if ((mVUregsTemp.VFreg[0] == mVUregsTemp.VFreg[1]) && mVUregsTemp.VFreg[0]) { if (mVUregsTemp.r || mVUregsTemp.VI) mVUlow.noWriteVF = 1; @@ -297,7 +297,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); startLoop(); - incCycles(1); + mVUincCycles(mVU, 1); mVUopU(mVU, 0); if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } if (curI & _DTbit_) { branch = 4; } @@ -424,7 +424,7 @@ eBitTemination: int lMac = findFlagInst(xMac, 0x7fffffff); int lClip = findFlagInst(xClip, 0x7fffffff); memset(&mVUinfo, 0, sizeof(mVUinfo)); - incCycles(100); // Ensures Valid P/Q instances (And sets all cycle data to 0) + mVUincCycles(mVU, 100); // Ensures Valid P/Q instances (And sets all cycle data to 0) mVUcycles -= 100; if (mVUinfo.doDivFlag) { int flagReg; diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index c3b6f979fb..a5bcd8c6b8 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -32,15 +32,16 @@ union regInfo { #pragma pack(push, 1) #pragma warning(disable:4996) #endif -struct microRegInfo { - regInfo VF[32]; - u8 VI[32]; +__declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares + u32 needExactMatch; // If set, block needs an exact match of pipeline state u8 q; u8 p; u8 r; u8 xgkick; - u8 flags; // clip x2 :: status x2 - u32 needExactMatch; // If set, block needs an exact match of pipeline state (needs to be last 2 bytes in struct) + u8 VI[16]; + regInfo VF[32]; + u8 flags; // clip x2 :: status x2 + u8 padding[7]; // 160 bytes #if defined(_MSC_VER) }; #pragma pack(pop) @@ -59,9 +60,9 @@ struct microTempRegInfo { u8 xgkick; // Holds the cycle info for XGkick }; -struct microBlock { +__declspec(align(16)) struct microBlock { microRegInfo pState; // Detailed State of Pipeline - microRegInfo pStateEnd; // Detailed State of Pipeline at End of Block (needed by JR/JALR opcodes) + microRegInfo pStateEnd; // Detailed State of Pipeline at End of Block (needed by JR/JALR opcodes) u8* x86ptrStart; // Start of code }; @@ -91,10 +92,10 @@ struct microLowerOp { microVFreg VF_read[2]; // VF Vectors read by this instruction microVIreg VI_write; // VI reg written to by this instruction microVIreg VI_read[2]; // VI regs read by this instruction + u32 branch; // Branch Type (0 = Not a Branch, 1 = B. 2 = BAL, 3~8 = Conditional Branches, 9 = JALR, 10 = JR) bool isNOP; // This instruction is a NOP bool isFSSET; // This instruction is a FSSET bool useSflag; // This instruction uses/reads Sflag - u32 branch; // Branch Type (0 = Not a Branch, 1 = B. 2 = BAL, 3~8 = Conditional Branches, 9 = JALR, 10 = JR) bool noWriteVF; // Don't write back the result of a lower op to VF reg if upper op writes to same reg (or if VF = 0) bool backupVI; // Backup VI reg to memory if modified before branch (branch uses old VI value unless opcode is ILW or ILWR) bool memReadIs; // Read Is (VI reg) from memory (used by branches) diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 26f6cdd501..56bc34e72c 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -160,6 +160,18 @@ declareAllVariables #define pass3 if (recPass == 2) #define pass4 if (recPass == 3) +// Define mVUquickSearch +#if defined(_MSC_VER) +extern u32 __fastcall mVUsearchXMM(void *dest, void *src); +#define mVUquickSearch(dest, src, size) (mVUsearchXMM(dest, src) == 0xf) +#else +#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +// Note: GCC might not guarantee alignment on microRegInfo, +// so to be safe I'm using normal memcmp. If at least 8-byte +// alignment is guaranteed, the function below is faster. +// #define mVUquickSearch(dest, src, size) (!memcmp_mmx(dest, src, size)) +#endif + // Misc Macros... #define mVUprogI mVU->prog.prog[progIndex] #define mVUcurProg mVU->prog.prog[mVU->prog.cur] @@ -188,7 +200,6 @@ declareAllVariables #define setCode() { mVU->code = curI; } #define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } #define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } -#define incCycles(x) { mVUincCycles(mVU, x); } #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((isVU1) ? 0x3ff8 : 0xff8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index c780a55d70..7660b14479 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -60,7 +60,7 @@ void mVUclamp2(int reg, int regT1, int xyzw) { } //------------------------------------------------------------------ -// Micro VU - Misc Functions +// Micro VU - Reg Loading/Saving/Shuffling/Unpacking/Merging... //------------------------------------------------------------------ void mVUunpack_xyzw(int dstreg, int srcreg, int xyzw) { @@ -261,6 +261,10 @@ void mVUmergeRegs(int dest, int src, int xyzw) { } } +//------------------------------------------------------------------ +// Micro VU - Misc Functions +//------------------------------------------------------------------ + // Transforms the Address in gprReg to valid VU0/VU1 Address microVUt(void) mVUaddrFix(mV, int gprReg) { if (mVU == µVU1) { @@ -295,6 +299,10 @@ microVUt(void) mVUrestoreRegs(mV) { MOV32ItoR(gprR, Roffset); // Restore gprR } +//------------------------------------------------------------------ +// Micro VU - Custom SSE Instructions +//------------------------------------------------------------------ + static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK1[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; static const u32 PCSX2_ALIGNED16(MIN_MAX_MASK2[4]) = {0x00000000, 0x40000000, 0x00000000, 0x40000000}; @@ -425,3 +433,54 @@ void SSE_ADD2SS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { SSE_ADDPS_XMM_to_XMM(to, from); } + +//------------------------------------------------------------------ +// Micro VU - Custom Quick Search +//------------------------------------------------------------------ + +// Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) +static __declspec(naked) u32 __fastcall mVUsearchXMM(void *dest, void *src) { + + __asm { + movaps xmm0, [ecx] + pcmpeqd xmm0, [edx] + movaps xmm1, [ecx + 0x10] + pcmpeqd xmm1, [edx + 0x10] + pand xmm0, xmm1 + + movmskps eax, xmm0 + cmp eax, 0xf + jl exitPoint + + movaps xmm0, [ecx + 0x20] + pcmpeqd xmm0, [edx + 0x20] + movaps xmm1, [ecx + 0x30] + pcmpeqd xmm1, [edx + 0x30] + pand xmm0, xmm1 + + movaps xmm2, [ecx + 0x40] + pcmpeqd xmm2, [edx + 0x40] + movaps xmm3, [ecx + 0x50] + pcmpeqd xmm3, [edx + 0x50] + pand xmm2, xmm3 + + movaps xmm4, [ecx + 0x60] + pcmpeqd xmm4, [edx + 0x60] + movaps xmm5, [ecx + 0x70] + pcmpeqd xmm5, [edx + 0x70] + pand xmm4, xmm5 + + movaps xmm6, [ecx + 0x80] + pcmpeqd xmm6, [edx + 0x80] + movaps xmm7, [ecx + 0x90] + pcmpeqd xmm7, [edx + 0x90] + pand xmm6, xmm7 + + pand xmm0, xmm2 + pand xmm4, xmm6 + pand xmm0, xmm4 + movmskps eax, xmm0 +exitPoint: + ret + } +} From 3734d4735193dae7a89a7088ab6f7f181bb91d4a Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 6 Jun 2009 09:17:33 +0000 Subject: [PATCH 195/296] non-functional change [Win32/msvc only]: Small cleanup to pragma packing syntax. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1335 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IPU/mpeg2lib/Mpeg.cpp | 4 ++-- pcsx2/x86/microVU_IR.h | 5 +++-- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/pcsx2/IPU/mpeg2lib/Mpeg.cpp b/pcsx2/IPU/mpeg2lib/Mpeg.cpp index 85fb80e084..60639366b2 100644 --- a/pcsx2/IPU/mpeg2lib/Mpeg.cpp +++ b/pcsx2/IPU/mpeg2lib/Mpeg.cpp @@ -900,7 +900,7 @@ static void __fastcall slice_non_intra_DCT (decoder_t * const decoder, } #if defined(_MSC_VER) -#pragma pack(push, 1) +#pragma pack(1) #endif struct TGA_HEADER { @@ -922,7 +922,7 @@ struct TGA_HEADER // pixel data follows header #if defined(_MSC_VER) }; -#pragma pack(pop) +#pragma pack() #else } __attribute__((packed)); #endif diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index a5bcd8c6b8..740fd36bb5 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -29,7 +29,7 @@ union regInfo { }; #if defined(_MSC_VER) -#pragma pack(push, 1) +#pragma pack(1) #pragma warning(disable:4996) #endif __declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares @@ -44,7 +44,8 @@ __declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares u8 padding[7]; // 160 bytes #if defined(_MSC_VER) }; -#pragma pack(pop) +#pragma pack() +#pragma pack(show) #else } __attribute__((packed)); #endif From f7aae60371c945d272563b32eff9b90bb85d26d5 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sat, 6 Jun 2009 09:24:10 +0000 Subject: [PATCH 196/296] Fixed Linux, and merged a few minor changes I had lying around that I hadn't committed yet. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1336 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 10 +++----- pcsx2/HwRead.cpp | 32 +++++++++++------------- pcsx2/HwWrite.cpp | 46 ++++++++++++++++------------------ pcsx2/IPU/IPU.cpp | 23 ++++++++--------- pcsx2/IPU/IPU.h | 4 +++ pcsx2/SPR.cpp | 4 +-- pcsx2/Sif.cpp | 2 +- pcsx2/VU0.cpp | 2 +- pcsx2/Vif.cpp | 2 +- pcsx2/x86/microVU_Misc.inl | 51 ++++++++++++++++++++++++++++++++++++++ 10 files changed, 111 insertions(+), 65 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 049ed0191a..ac18745ca9 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -160,10 +160,7 @@ static __forceinline void dmaGIFend() // not to do the gif->qwc != 0 check. --arcum42 static __forceinline void GIFdmaEnd() { - //if (psHu32(GIF_MODE) & 0x4) CPU_INT(2, gscycles * BIAS); - /*else - CPU_INT(2, min( gifsplit, (int)gif->qwc ) * BIAS);*/ } void GIFdma() @@ -178,7 +175,6 @@ void GIFdma() return; } - if ((psHu32(DMAC_CTRL) & 0xC0) == 0x80 && prevcycles != 0) { // STD == GIF Console::WriteLn("GS Stall Control Source = %x, Drain = %x\n MADR = %x, STADR = %x", params (psHu32(0xe000) >> 4) & 0x3, (psHu32(0xe000) >> 6) & 0x3, gif->madr, psHu32(DMAC_STADR)); @@ -324,8 +320,8 @@ void GIFdma() } void dmaGIF() { - //We used to addd wait time for the buffer to fill here, fixing some timing problems in path 3 masking - //It takes the time of 24 QW for the BUS to become ready - The Punisher, And1 Streetball + //We used to add wait time for the buffer to fill here, fixing some timing problems in path 3 masking + //It takes the time of 24 QW for the BUS to become ready - The Punisher And Streetball GIF_LOG("dmaGIFstart chcr = %lx, madr = %lx, qwc = %lx\n tadr = %lx, asr0 = %lx, asr1 = %lx", gif->chcr, gif->madr, gif->qwc, gif->tadr, gif->asr0, gif->asr1); Path3progress = 2; @@ -555,7 +551,7 @@ void gifMFIFOInterrupt() if((spr0->chcr & 0x100) && spr0->qwc == 0) { spr0->chcr &= ~0x100; - hwDmacIrq(8); + hwDmacIrq(DMAC_FROM_SPR); } if(gifstate != GIF_STATE_STALL) { diff --git a/pcsx2/HwRead.cpp b/pcsx2/HwRead.cpp index d1ca5c3b67..d84b9a7da2 100644 --- a/pcsx2/HwRead.cpp +++ b/pcsx2/HwRead.cpp @@ -61,38 +61,37 @@ __forceinline mem8_t hwRead8(u32 mem) switch (mem) { - // Note: the values without defines = the defines + 1. case RCNT0_COUNT: ret = (u8)rcntRcount(0); break; case RCNT0_MODE: ret = (u8)counters[0].modeval; break; case RCNT0_TARGET: ret = (u8)counters[0].target; break; case RCNT0_HOLD: ret = (u8)counters[0].hold; break; - case 0x10000001: ret = (u8)(rcntRcount(0)>>8); break; - case 0x10000011: ret = (u8)(counters[0].modeval>>8); break; - case 0x10000021: ret = (u8)(counters[0].target>>8); break; - case 0x10000031: ret = (u8)(counters[0].hold>>8); break; + case RCNT0_COUNT + 1: ret = (u8)(rcntRcount(0)>>8); break; + case RCNT0_MODE + 1: ret = (u8)(counters[0].modeval>>8); break; + case RCNT0_TARGET + 1: ret = (u8)(counters[0].target>>8); break; + case RCNT0_HOLD + 1: ret = (u8)(counters[0].hold>>8); break; case RCNT1_COUNT: ret = (u8)rcntRcount(1); break; case RCNT1_MODE: ret = (u8)counters[1].modeval; break; case RCNT1_TARGET: ret = (u8)counters[1].target; break; case RCNT1_HOLD: ret = (u8)counters[1].hold; break; - case 0x10000801: ret = (u8)(rcntRcount(1)>>8); break; - case 0x10000811: ret = (u8)(counters[1].modeval>>8); break; - case 0x10000821: ret = (u8)(counters[1].target>>8); break; - case 0x10000831: ret = (u8)(counters[1].hold>>8); break; + case RCNT1_COUNT + 1: ret = (u8)(rcntRcount(1)>>8); break; + case RCNT1_MODE + 1: ret = (u8)(counters[1].modeval>>8); break; + case RCNT1_TARGET + 1: ret = (u8)(counters[1].target>>8); break; + case RCNT1_HOLD + 1: ret = (u8)(counters[1].hold>>8); break; case RCNT2_COUNT: ret = (u8)rcntRcount(2); break; case RCNT2_MODE: ret = (u8)counters[2].modeval; break; case RCNT2_TARGET: ret = (u8)counters[2].target; break; - case 0x10001001: ret = (u8)(rcntRcount(2)>>8); break; - case 0x10001011: ret = (u8)(counters[2].modeval>>8); break; - case 0x10001021: ret = (u8)(counters[2].target>>8); break; + case RCNT2_COUNT + 1: ret = (u8)(rcntRcount(2)>>8); break; + case RCNT2_MODE + 1: ret = (u8)(counters[2].modeval>>8); break; + case RCNT2_TARGET + 1: ret = (u8)(counters[2].target>>8); break; case RCNT3_COUNT: ret = (u8)rcntRcount(3); break; case RCNT3_MODE: ret = (u8)counters[3].modeval; break; case RCNT3_TARGET: ret = (u8)counters[3].target; break; - case 0x10001801: ret = (u8)(rcntRcount(3)>>8); break; - case 0x10001811: ret = (u8)(counters[3].modeval>>8); break; - case 0x10001821: ret = (u8)(counters[3].target>>8); break; + case RCNT3_COUNT + 1: ret = (u8)(rcntRcount(3)>>8); break; + case RCNT3_MODE + 1: ret = (u8)(counters[3].modeval>>8); break; + case RCNT3_TARGET + 1: ret = (u8)(counters[3].target>>8); break; default: if ((mem & 0xffffff0f) == SBUS_F200) @@ -225,7 +224,7 @@ mem32_t __fastcall hwRead32_page_01(u32 mem) // This is used internally to produce two inline versions, one with INTC_HACK, and one without. static __forceinline mem32_t __hwRead32_page_0F( u32 mem, bool intchack ) { - // *Performance Warning* This function is called -A-LOT. Be weary when making changes. It + // *Performance Warning* This function is called -A-LOT. Be wary when making changes. It // could impact FPS significantly. mem &= 0xffff; @@ -258,7 +257,6 @@ static __forceinline mem32_t __hwRead32_page_0F( u32 mem, bool intchack ) return psHu32(0xf240) | 0xF0000102; case 0xf440: // MCH_DRD - if( !((psHu32(0xf430) >> 6) & 0xF) ) { switch ((psHu32(0xf430)>>16) & 0xFFF) diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index f715b4e67b..43659dc34a 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -152,24 +152,24 @@ void hwWrite8(u32 mem, u8 value) switch (mem) { case RCNT0_COUNT: rcntWcount(0, value); break; case RCNT0_MODE: rcntWmode(0, (counters[0].modeval & 0xff00) | value); break; - case 0x10000011: rcntWmode(0, (counters[0].modeval & 0xff) | value << 8); break; + case RCNT0_MODE + 1: rcntWmode(0, (counters[0].modeval & 0xff) | value << 8); break; case RCNT0_TARGET: rcntWtarget(0, value); break; case RCNT0_HOLD: rcntWhold(0, value); break; case RCNT1_COUNT: rcntWcount(1, value); break; case RCNT1_MODE: rcntWmode(1, (counters[1].modeval & 0xff00) | value); break; - case 0x10000811: rcntWmode(1, (counters[1].modeval & 0xff) | value << 8); break; + case RCNT1_MODE + 1: rcntWmode(1, (counters[1].modeval & 0xff) | value << 8); break; case RCNT1_TARGET: rcntWtarget(1, value); break; case RCNT1_HOLD: rcntWhold(1, value); break; case RCNT2_COUNT: rcntWcount(2, value); break; case RCNT2_MODE: rcntWmode(2, (counters[2].modeval & 0xff00) | value); break; - case 0x10001011: rcntWmode(2, (counters[2].modeval & 0xff) | value << 8); break; + case RCNT2_MODE + 1: rcntWmode(2, (counters[2].modeval & 0xff) | value << 8); break; case RCNT2_TARGET: rcntWtarget(2, value); break; case RCNT3_COUNT: rcntWcount(3, value); break; case RCNT3_MODE: rcntWmode(3, (counters[3].modeval & 0xff00) | value); break; - case 0x10001811: rcntWmode(3, (counters[3].modeval & 0xff) | value << 8); break; + case RCNT3_MODE + 1: rcntWmode(3, (counters[3].modeval & 0xff) | value << 8); break; case RCNT3_TARGET: rcntWtarget(3, value); break; case SIO_TXFIFO: @@ -193,7 +193,7 @@ void hwWrite8(u32 mem, u8 value) //case 0x10003c02: //Tony Hawks Project 8 uses this // vif1Write32(mem & ~0x2, value << 16); // break; - case 0x10008001: // dma0 - vif0 + case D0_CHCR + 1: // dma0 - vif0 DMA_LOG("VIF0dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -203,7 +203,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaVIF0, mem, value); break; - case 0x10009001: // dma1 - vif1 + case D1_CHCR + 1: // dma1 - vif1 DMA_LOG("VIF1dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -214,7 +214,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaVIF1, mem, value); break; - case 0x1000a001: // dma2 - gif + case D2_CHCR + 1: // dma2 - gif DMA_LOG("GSdma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -224,7 +224,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaGIF, mem, value); break; - case 0x1000b001: // dma3 - fromIPU + case D3_CHCR + 1: // dma3 - fromIPU DMA_LOG("IPU0dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -234,7 +234,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaIPU0, mem, value); break; - case 0x1000b401: // dma4 - toIPU + case D4_CHCR + 1: // dma4 - toIPU DMA_LOG("IPU1dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -244,7 +244,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaIPU1, mem, value); break; - case 0x1000c001: // dma5 - sif0 + case D5_CHCR + 1: // dma5 - sif0 DMA_LOG("SIF0dma EXECUTE, value=0x%x", value); // if (value == 0) psxSu32(0x30) = 0x40000; if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) @@ -255,7 +255,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaSIF0, mem, value); break; - case 0x1000c401: // dma6 - sif1 + case D6_CHCR + 1: // dma6 - sif1 DMA_LOG("SIF1dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -265,7 +265,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaSIF1, mem, value); break; - case 0x1000c801: // dma7 - sif2 + case D7_CHCR + 1: // dma7 - sif2 DMA_LOG("SIF2dma EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -275,7 +275,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaSIF2, mem, value); break; - case 0x1000d001: // dma8 - fromSPR + case D8_CHCR + 1: // dma8 - fromSPR DMA_LOG("fromSPRdma8 EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -285,7 +285,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaSPR0, mem, value); break; - case 0x1000d401: // dma9 - toSPR + case SPR1_CHCR + 1: // dma9 - toSPR DMA_LOG("toSPRdma8 EXECUTE, value=0x%x", value); if ((value & 0x1) && !(psHu32(DMAC_CTRL) & 0x1)) { @@ -295,7 +295,7 @@ void hwWrite8(u32 mem, u8 value) DmaExec8(dmaSPR1, mem, value); break; - case 0x1000f592: // DMAC_ENABLEW + 2 + case DMAC_ENABLEW + 2: psHu8(0xf592) = value; psHu8(0xf522) = value; break; @@ -316,7 +316,7 @@ void hwWrite8(u32 mem, u8 value) psHu8(mem) = value; break; - case SBUS_F240:// SIF(?) + case SBUS_F240: if (!(value & 0x100)) psHu32(mem) &= ~0x100; break; @@ -540,7 +540,7 @@ __forceinline void hwWrite16(u32 mem, u16 value) DmaExec16(dmaSIF0, mem, value); break; - case 0x1000c002: // D5_CHCR + 2 + case D5_CHCR + 2: //? break; @@ -586,7 +586,7 @@ __forceinline void hwWrite16(u32 mem, u16 value) DmaExec16(dmaSIF2, mem, value); break; - case 0x1000c802: // D7_CHCR + 2 + case D7_CHCR + 2: //? break; @@ -610,17 +610,17 @@ __forceinline void hwWrite16(u32 mem, u16 value) DmaExec16(dmaSPR1, mem, value); break; - case 0x1000f592: // DMAC_ENABLEW + 2 + case DMAC_ENABLEW + 2: psHu16(0xf592) = value; psHu16(0xf522) = value; break; case SIO_ISR: - case 0x1000f132: + case SIO_ISR + 2: case 0x1000f410: - case 0x1000f412: + case 0x1000f410 + 2: case MCH_RICM: - case 0x1000f432: + case MCH_RICM + 2: break; case SBUS_F200: @@ -1139,8 +1139,6 @@ void __fastcall hwWrite64_page_0E( u32 mem, const mem64_t* srcval ) void __fastcall hwWrite64_generic( u32 mem, const mem64_t* srcval ) { - //hwWrite64( mem, *srcval ); return; - const u64 value = *srcval; switch (mem) diff --git a/pcsx2/IPU/IPU.cpp b/pcsx2/IPU/IPU.cpp index 8a183e91a2..a47e1ad681 100644 --- a/pcsx2/IPU/IPU.cpp +++ b/pcsx2/IPU/IPU.cpp @@ -241,7 +241,7 @@ void SaveState::ipuFreeze() bool ipuCanFreeze() { - return ipuCurCmd == 0xffffffff; + return (ipuCurCmd == 0xffffffff); } __forceinline u32 ipuRead32(u32 mem) @@ -256,7 +256,7 @@ __forceinline u32 ipuRead32(u32 mem) switch (mem) { - case 0x10: // IPU_CTRL + ipucase(IPU_CTRL): // IPU_CTRL ipuRegs->ctrl.IFC = g_BP.IFC; ipuRegs->ctrl.CBP = coded_block_pattern; @@ -265,7 +265,7 @@ __forceinline u32 ipuRead32(u32 mem) return ipuRegs->ctrl._u32; - case 0x20: // IPU_BP + ipucase(IPU_BP): // IPU_BP ipuRegs->ipubp = g_BP.BP & 0x7f; ipuRegs->ipubp |= g_BP.IFC << 8; @@ -290,20 +290,20 @@ __forceinline u64 ipuRead64(u32 mem) switch (mem) { - case 0x00: // IPU_CMD - if (ipuRegs->cmd.DATA&0xffffff) + ipucase(IPU_CMD): // IPU_CMD + if (ipuRegs->cmd.DATA & 0xffffff) IPU_LOG("Ipu read64: IPU_CMD=BUSY=%x, DATA=%08X", ipuRegs->cmd.BUSY ? 1 : 0, ipuRegs->cmd.DATA); break; - case 0x10: + ipucase(IPU_CTRL): DevCon::Notice("reading 64bit IPU ctrl"); break; - case 0x20: + ipucase(IPU_BP): DevCon::Notice("reading 64bit IPU top"); break; - case 0x30: // IPU_TOP + ipucase(IPU_TOP): // IPU_TOP IPU_LOG("Ipu read64: IPU_TOP=%x, bp = %d", ipuRegs->top, g_BP.BP); break; @@ -348,12 +348,12 @@ __forceinline void ipuWrite32(u32 mem, u32 value) switch (mem) { - case 0x00: // IPU_CMD + ipucase(IPU_CMD): // IPU_CMD IPU_LOG("Ipu write32: IPU_CMD=0x%08X", value); IPUCMD_WRITE(value); break; - case 0x10: // IPU_CTRL + ipucase(IPU_CTRL): // IPU_CTRL ipuRegs->ctrl._u32 = (value & 0x47f30000) | (ipuRegs->ctrl._u32 & 0x8000ffff); if (ipuRegs->ctrl.IDP == 3) { @@ -385,7 +385,7 @@ __forceinline void ipuWrite64(u32 mem, u64 value) switch (mem) { - case 0x00: + ipucase(IPU_CMD): IPU_LOG("Ipu write64: IPU_CMD=0x%08X", value); IPUCMD_WRITE((u32)value); break; @@ -788,7 +788,6 @@ __forceinline void IPU_INTERRUPT() //dma void IPUCMD_WRITE(u32 val) { - // don't process anything if currently busy if (ipuRegs->ctrl.BUSY) Console::WriteLn("IPU BUSY!"); // wait for thread diff --git a/pcsx2/IPU/IPU.h b/pcsx2/IPU/IPU.h index b644887550..c148bac652 100644 --- a/pcsx2/IPU/IPU.h +++ b/pcsx2/IPU/IPU.h @@ -41,6 +41,10 @@ #pragma pack(1) #endif + +#define ipumsk( src ) ( (src) & 0xff ) +#define ipucase( src ) case ipumsk(src) + // // Bitfield Structure // diff --git a/pcsx2/SPR.cpp b/pcsx2/SPR.cpp index 5c5291fad0..702cdb3306 100644 --- a/pcsx2/SPR.cpp +++ b/pcsx2/SPR.cpp @@ -241,7 +241,7 @@ void SPRFROMinterrupt() } if (spr0finished == 0) return; spr0->chcr &= ~0x100; - hwDmacIrq(8); + hwDmacIrq(DMAC_FROM_SPR); } @@ -418,7 +418,7 @@ void SPRTOinterrupt() _dmaSPR1(); if (spr1finished == 0) return; spr1->chcr &= ~0x100; - hwDmacIrq(9); + hwDmacIrq(DMAC_TO_SPR); } void SaveState::sprFreeze() diff --git a/pcsx2/Sif.cpp b/pcsx2/Sif.cpp index 940fa7a266..d6b9469d9d 100644 --- a/pcsx2/Sif.cpp +++ b/pcsx2/Sif.cpp @@ -495,7 +495,7 @@ __forceinline void dmaSIF2() sif2dma->chcr, sif2dma->madr, sif2dma->qwc); sif2dma->chcr &= ~0x100; - hwDmacIrq(7); + hwDmacIrq(DMAC_SIF2); Console::WriteLn("*PCSX2*: dmaSIF2"); } diff --git a/pcsx2/VU0.cpp b/pcsx2/VU0.cpp index 2b6c831f6d..a017527406 100644 --- a/pcsx2/VU0.cpp +++ b/pcsx2/VU0.cpp @@ -108,7 +108,7 @@ namespace OpcodeImpl // Asadr.Changed //TODO: check this - // HUH why ? doesn;t make any sense ... + // HUH why ? doesn't make any sense ... void SQC2() { u32 addr = _Imm_ + cpuRegs.GPR.r[_Rs_].UL[0]; //memWrite64(addr, VU0.VF[_Ft_].UD[0]); diff --git a/pcsx2/Vif.cpp b/pcsx2/Vif.cpp index 73e874c9f7..a3c7cfc4fa 100644 --- a/pcsx2/Vif.cpp +++ b/pcsx2/Vif.cpp @@ -547,7 +547,7 @@ void vifMFIFOInterrupt() if((spr0->chcr & 0x100) && spr0->qwc == 0) { spr0->chcr &= ~0x100; - hwDmacIrq(8); + hwDmacIrq(DMAC_FROM_SPR); } if (vif1.inprogress == 1) mfifo_VIF1chain(); diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 7660b14479..e2379dac43 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -438,6 +438,7 @@ void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { // Micro VU - Custom Quick Search //------------------------------------------------------------------ +#ifndef __LINUX__ // Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) static __declspec(naked) u32 __fastcall mVUsearchXMM(void *dest, void *src) { @@ -484,3 +485,53 @@ exitPoint: ret } } +#else +// Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) +static u32 __fastcall mVUsearchXMM(void *dest, void *src) +{ + __asm__ + ( + ".intel_syntax noprefix\n" + "movaps xmm0, [ecx]\n" + "pcmpeqd xmm0, [edx]\n" + "movaps xmm1, [ecx + 0x10]\n" + "pcmpeqd xmm1, [edx + 0x10]\n" + "pand xmm0, xmm1\n" + + "movmskps eax, xmm0\n" + "cmp eax, 0xf\n" + "jl exitPoint\n" + + "movaps xmm0, [ecx + 0x20]\n" + "pcmpeqd xmm0, [edx + 0x20]\n" + "movaps xmm1, [ecx + 0x30]\n" + "pcmpeqd xmm1, [edx + 0x30]\n" + "pand xmm0, xmm1\n" + + "movaps xmm2, [ecx + 0x40]\n" + "pcmpeqd xmm2, [edx + 0x40]\n" + "movaps xmm3, [ecx + 0x50]\n" + "pcmpeqd xmm3, [edx + 0x50]\n" + "pand xmm2, xmm3\n" + + "movaps xmm4, [ecx + 0x60]\n" + "pcmpeqd xmm4, [edx + 0x60]\n" + "movaps xmm5, [ecx + 0x70]\n" + "pcmpeqd xmm5, [edx + 0x70]\n" + "pand xmm4, xmm5\n" + + "movaps xmm6, [ecx + 0x80]\n" + "pcmpeqd xmm6, [edx + 0x80]\n" + "movaps xmm7, [ecx + 0x90]\n" + "pcmpeqd xmm7, [edx + 0x90]\n" + "pand xmm6, xmm7\n" + + "pand xmm0, xmm2\n" + "pand xmm4, xmm6\n" + "pand xmm0, xmm4\n" + "movmskps eax, xmm0\n" +"exitPoint:\n" + ".att_syntax\n" + ); +} +#endif From cdd68f345e215dbe5c7c23ffe7965bb2f9b461d6 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 6 Jun 2009 10:05:40 +0000 Subject: [PATCH 197/296] microVU: Improved struct alignment method using placement new syntax, reducing microBlockManager's heap allocation count from 2-per to 1-per. :) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1337 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 4 ++-- pcsx2/x86/microVU.h | 19 ++++++++++++++++--- pcsx2/x86/microVU_Compile.inl | 8 ++++---- pcsx2/x86/microVU_IR.h | 1 - 4 files changed, 22 insertions(+), 10 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 33da63a585..8685e1e147 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -62,7 +62,7 @@ microVUx(void) mVUreset() { // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { for (u32 j = 0; j < (mVU->progSize / 2); j++) { - safe_delete( mVU->prog.prog[i].block[j] ); + microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } } @@ -109,7 +109,7 @@ microVUf(void) mVUclose() { for (int i = 0; i <= mVU->prog.max; i++) { for (u32 j = 0; j < (mVU->progSize / 2); j++) { if (mVU->prog.prog[i].block[j]) { - safe_delete( mVU->prog.prog[i].block[j] ); + microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } } } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index febcfbefbf..a33cd8a5bf 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -32,13 +32,26 @@ class microBlockManager { private: static const int MaxBlocks = mMaxBlocks - 1; + microBlock blockList[mMaxBlocks]; // note: should always be first in the class to ensure xmm alignment int listSize; // Total Items - 1 int listI; // Index to Add new block - microBlock* blockList; public: - microBlockManager() { blockList = (microBlock*)_aligned_malloc(sizeof(microBlock)*mMaxBlocks, 16); reset(); } - ~microBlockManager() { if (blockList) { _aligned_free(blockList); } } + // our aligned replacement for 'new': + static microBlockManager* AlignedNew() { + microBlockManager* alloc = (microBlockManager*)_aligned_malloc(sizeof(microBlockManager), 16); + new (alloc) microBlockManager(); + return alloc; + } + + static void Delete( microBlockManager* dead ) { + if( dead == NULL ) return; + dead->~microBlockManager(); + _aligned_free( dead ); + } + + microBlockManager() { reset(); } + ~microBlockManager() { } void reset() { listSize = -1; listI = -1; }; microBlock* add(microBlock* pBlock) { microBlock* thisBlock = search(&pBlock->pState); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 7c1ff506c6..ee771e06ee 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -26,7 +26,7 @@ mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ CMP16ItoM((uptr)&mVU->branch, 0); \ incPC2(1); \ - if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = new microBlockManager(); } \ + if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ @@ -266,7 +266,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { startPC &= (vuIndex ? 0x3ff8 : 0xff8); if (mVUblocks[startPC/8] == NULL) { - mVUblocks[startPC/8] = new microBlockManager(); + mVUblocks[startPC/8] = microBlockManager::AlignedNew(); } // Searches for Existing Compiled Block (if found, then returns; else, compile) @@ -357,7 +357,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = new microBlockManager(); + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); // Check if branch-block has already been compiled pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); @@ -387,7 +387,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { incPC(-3); // Go back to branch opcode (to get branch imm addr) if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = new microBlockManager(); + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); // Check if branch-block has already been compiled pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 740fd36bb5..2fb0623027 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -45,7 +45,6 @@ __declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares #if defined(_MSC_VER) }; #pragma pack() -#pragma pack(show) #else } __attribute__((packed)); #endif From 743e93f3eec03f5f7bdfc7e4847be2438e5f1d38 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 6 Jun 2009 10:11:22 +0000 Subject: [PATCH 198/296] GSdx: should fix Issue 263, but I could not verify it git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1338 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRenderer.cpp | 2 +- plugins/GSdx/GSTexture.h | 2 ++ plugins/GSdx/GSTexture10.h | 1 + plugins/GSdx/GSTexture7.h | 1 + plugins/GSdx/GSTexture9.h | 3 +-- 5 files changed, 6 insertions(+), 3 deletions(-) diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index cd2dfc679a..1391efcd21 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -243,7 +243,7 @@ bool GSRenderer::Merge(int field) int field2 = 1 - ((m_interlace - 1) & 1); int mode = (m_interlace - 1) >> 1; - m_dev->Interlace(ds, field ^ field2, mode, tex[1]->m_scale.y); + m_dev->Interlace(ds, field ^ field2, mode, tex[1] ? tex[1]->m_scale.y : tex[0]->m_scale.y); } } diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index 26de1ca667..a85199a8f9 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -35,10 +35,12 @@ public: enum {None, RenderTarget, DepthStencil, Texture, Offscreen}; virtual operator bool() = 0; + virtual int GetType() const = 0; virtual int GetWidth() const = 0; virtual int GetHeight() const = 0; virtual int GetFormat() const = 0; + virtual bool Update(const GSVector4i& r, const void* data, int pitch) = 0; virtual bool Map(uint8** bits, int& pitch, const GSVector4i* r = NULL) = 0; virtual void Unmap() = 0; diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index a74b0ee8ac..5c8418ae6a 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -41,6 +41,7 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; + bool Update(const GSVector4i& r, const void* data, int pitch); bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index 99c1c73b65..8eb64d730c 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -41,6 +41,7 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; + bool Update(const GSVector4i& r, const void* data, int pitch); bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index 5ee23c9370..3f42c46bb1 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -41,13 +41,12 @@ public: int GetWidth() const; int GetHeight() const; int GetFormat() const; + bool Update(const GSVector4i& r, const void* data, int pitch); bool Map(uint8** bits, int& pitch, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); - //IDirect3DTexture9* operator->(); // TODO: remove direct access - operator IDirect3DSurface9*(); operator IDirect3DTexture9*(); }; From 8784c1cd4315801f8ad548652fc6d26de47ae38e Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 6 Jun 2009 10:17:07 +0000 Subject: [PATCH 199/296] GSdx: added vs2010 project files git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1339 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSdx_vs2010.vcxproj | 1494 +++++++++++++++++++++ plugins/GSdx/GSdx_vs2010.vcxproj.filters | 643 +++++++++ plugins/GSdx/vsprops/ProjectRootDir.props | 25 + plugins/GSdx/vsprops/common.props | 32 + plugins/GSdx/vsprops/debug.props | 15 + plugins/GSdx/vsprops/release.props | 23 + plugins/GSdx/vsprops/sse2.props | 11 + plugins/GSdx/vsprops/sse4.props | 11 + plugins/GSdx/vsprops/ssse3.props | 11 + 9 files changed, 2265 insertions(+) create mode 100644 plugins/GSdx/GSdx_vs2010.vcxproj create mode 100644 plugins/GSdx/GSdx_vs2010.vcxproj.filters create mode 100644 plugins/GSdx/vsprops/ProjectRootDir.props create mode 100644 plugins/GSdx/vsprops/common.props create mode 100644 plugins/GSdx/vsprops/debug.props create mode 100644 plugins/GSdx/vsprops/release.props create mode 100644 plugins/GSdx/vsprops/sse2.props create mode 100644 plugins/GSdx/vsprops/sse4.props create mode 100644 plugins/GSdx/vsprops/ssse3.props diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj b/plugins/GSdx/GSdx_vs2010.vcxproj new file mode 100644 index 0000000000..fb4053ff4d --- /dev/null +++ b/plugins/GSdx/GSdx_vs2010.vcxproj @@ -0,0 +1,1494 @@ + + + + Debug SSE2 + Win32 + + + Debug SSE2 + Win32 + + + Debug SSE2 + x64 + + + Debug SSE2 + x64 + + + Debug SSE4 + Win32 + + + Debug SSE4 + Win32 + + + Debug SSE4 + x64 + + + Debug SSE4 + x64 + + + Debug SSSE3 + Win32 + + + Debug SSSE3 + Win32 + + + Debug SSSE3 + x64 + + + Debug SSSE3 + x64 + + + Debug + Win32 + + + Debug + Win32 + + + Debug + x64 + + + Debug + x64 + + + Release SSE2 + Win32 + + + Release SSE2 + Win32 + + + Release SSE2 + x64 + + + Release SSE2 + x64 + + + Release SSE4 + Win32 + + + Release SSE4 + Win32 + + + Release SSE4 + x64 + + + Release SSE4 + x64 + + + Release SSSE3 + Win32 + + + Release SSSE3 + Win32 + + + Release SSSE3 + x64 + + + Release SSSE3 + x64 + + + Release + Win32 + + + Release + Win32 + + + Release + x64 + + + Release + x64 + + + + GSdx + {18E42F6F-3A62-41EE-B42F-79366C4F1E95} + GSdx + Win32Proj + + + + DynamicLibrary + false + MultiByte + + + DynamicLibrary + false + MultiByte + true + + + DynamicLibrary + Static + MultiByte + + + DynamicLibrary + Static + MultiByte + true + + + DynamicLibrary + false + MultiByte + + + DynamicLibrary + Static + MultiByte + false + + + DynamicLibrary + false + MultiByte + true + + + DynamicLibrary + Static + MultiByte + true + + + DynamicLibrary + false + MultiByte + true + + + DynamicLibrary + Static + MultiByte + true + + + DynamicLibrary + false + MultiByte + + + DynamicLibrary + Static + MultiByte + + + DynamicLibrary + false + MultiByte + + + DynamicLibrary + Static + MultiByte + + + DynamicLibrary + false + MultiByte + true + + + DynamicLibrary + Static + MultiByte + true + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + <_ProjectFileVersion>10.0.20506.1 + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + X64 + + + Use + + + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + X64 + + + Use + + + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + Use + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + Use + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + Use + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + Use + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + X64 + + + Use + + + + + MachineX64 + + + + + Use + + + .\GSdx.def + + + MachineX86 + + + + + X64 + + + Use + + + + + MachineX64 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + Create + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + \ No newline at end of file diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj.filters b/plugins/GSdx/GSdx_vs2010.vcxproj.filters new file mode 100644 index 0000000000..ee1b5c34d3 --- /dev/null +++ b/plugins/GSdx/GSdx_vs2010.vcxproj.filters @@ -0,0 +1,643 @@ + + + + + {14a43ae4-91f3-4405-a71c-c8f2d790e36b} + + + {93995380-89BD-4b04-88EB-625FBE52EBFB} + h;hpp;hxx;hm;inl;inc;xsd + + + {67DA6AB6-F800-4c08-8B7A-83BB121AAD01} + rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav + + + {e281bf48-1466-4355-b176-a117cc2dce40} + + + {4FC737F1-C7A5-4376-A066-2A32D752A2FF} + cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx + + + {62920db0-8592-425e-aa2f-9a2a7e85d851} + + + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + Source Files + + + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Baseclasses + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Header Files + + + Resource Files + + + Header Files + + + Header Files + + + Xbyak + + + Xbyak + + + Xbyak + + + Xbyak + + + + + Baseclasses + + + Baseclasses + + + Source Files + + + Shaders + + + Shaders + + + Resource Files + + + Shaders + + + Shaders + + + Resource Files + + + Resource Files + + + Shaders + + + Shaders + + + Shaders + + + Shaders + + + + + Resource Files + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/ProjectRootDir.props b/plugins/GSdx/vsprops/ProjectRootDir.props new file mode 100644 index 0000000000..fd31a6334d --- /dev/null +++ b/plugins/GSdx/vsprops/ProjectRootDir.props @@ -0,0 +1,25 @@ + + + $(ProjectDir). + $(ProjectRootDir)\..\.. + $(SvnRootDir)\common + plugins + + + <_PropertySheetDisplayName>ProjectRootDir + + + + $(ProjectRootDir) + + + $(SvnRootDir) + + + $(SvnCommonDir) + + + $(PcsxSubsection) + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/common.props b/plugins/GSdx/vsprops/common.props new file mode 100644 index 0000000000..488eba3520 --- /dev/null +++ b/plugins/GSdx/vsprops/common.props @@ -0,0 +1,32 @@ + + + <_PropertySheetDisplayName>common + $(SolutionDir)\bin\$(PcsxSubsection)\ + $(PlatformName)\$(Configuration)\ + + + + true + WIN32;_WINDOWS;_WIN32_WINNT=0x500;%(PreprocessorDefinitions) + Fast + false + Level4 + ProgramDatabase + 4995;4324;%(DisableSpecificWarnings) + + + JITProfiling.lib;d3d10.lib;d3dx10.lib;d3d9.lib;d3dx9.lib;ddraw.lib;dxguid.lib;winmm.lib;strmiids.lib;xinput.lib;%(AdditionalDependencies) + ./vtune;%(AdditionalLibraryDirectories) + d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll;%(DelayLoadDLLs) + true + Windows + false + + + .\postBuild.cmd "$(TargetPath)" "$(TargetName)" $(TargetExt) $(PcsxSubsection) + + + "$(SvnCommonDir)\vsprops\preBuild.cmd" "$(ProjectRootDir)" + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/debug.props b/plugins/GSdx/vsprops/debug.props new file mode 100644 index 0000000000..4eabec6606 --- /dev/null +++ b/plugins/GSdx/vsprops/debug.props @@ -0,0 +1,15 @@ + + + <_PropertySheetDisplayName>debug + true + + + + Disabled + _DEBUG;%(PreprocessorDefinitions) + true + StackFrameRuntimeCheck + MultiThreadedDebug + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/release.props b/plugins/GSdx/vsprops/release.props new file mode 100644 index 0000000000..d88644540e --- /dev/null +++ b/plugins/GSdx/vsprops/release.props @@ -0,0 +1,23 @@ + + + <_PropertySheetDisplayName>release + false + + + + MaxSpeed + AnySuitable + true + Speed + true + true + NDEBUG;_SECURE_SCL=0;%(PreprocessorDefinitions) + true + false + + + true + true + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/sse2.props b/plugins/GSdx/vsprops/sse2.props new file mode 100644 index 0000000000..593107b5a9 --- /dev/null +++ b/plugins/GSdx/vsprops/sse2.props @@ -0,0 +1,11 @@ + + + <_PropertySheetDisplayName>sse2 + + + + StreamingSIMDExtensions2 + _M_SSE=0x200;%(PreprocessorDefinitions) + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/sse4.props b/plugins/GSdx/vsprops/sse4.props new file mode 100644 index 0000000000..273b626cee --- /dev/null +++ b/plugins/GSdx/vsprops/sse4.props @@ -0,0 +1,11 @@ + + + <_PropertySheetDisplayName>sse4 + + + + _M_SSE=0x401;%(PreprocessorDefinitions) + StreamingSIMDExtensions2 + + + \ No newline at end of file diff --git a/plugins/GSdx/vsprops/ssse3.props b/plugins/GSdx/vsprops/ssse3.props new file mode 100644 index 0000000000..0d1131f6a8 --- /dev/null +++ b/plugins/GSdx/vsprops/ssse3.props @@ -0,0 +1,11 @@ + + + <_PropertySheetDisplayName>sse3 + + + + StreamingSIMDExtensions2 + _M_SSE=0x301;%(PreprocessorDefinitions) + + + \ No newline at end of file From 95ff8e3a98e2895a96ecb55e8eb6d654c481a7a8 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sat, 6 Jun 2009 11:00:30 +0000 Subject: [PATCH 200/296] GSdx: Applied normalization of UV in the DX9 texture repeat shader, allowing it to be re-enabled. Fixes massive slowdown in the Xenosaga 2 opening menu (Issue 262 / DX9 only) [note: collective fix uncovered by sudonim, feal87, and drk||raziel -- I'm just the patching boy] git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1340 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRendererHW9.cpp | 2 +- plugins/GSdx/res/tfx9.fx | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 851a02b50d..27be32f1d2 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -25,7 +25,7 @@ #include "resource.h" GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this), false) + : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this), true) { m_fba.enabled = !!theApp.GetConfig("fba", 1); m_logz = !!theApp.GetConfig("logz", 0); diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index 384b85780d..b17a07f822 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -116,12 +116,12 @@ sampler1D VMSKFIX : register(s3); float repeatu(float tc) { - return WMS == 3 ? tex1D(UMSKFIX, tc) : tc; + return WMS == 3 ? tex1D(UMSKFIX, tc*rWrH.x)* WH.x : tc; } float repeatv(float tc) { - return WMT == 3 ? tex1D(VMSKFIX, tc) : tc; + return WMT == 3 ? tex1D(VMSKFIX, tc*rWrH.y)* WH.y : tc; } float4 sample(float2 tc) From 4ed303566147c243a5da5fbf407d58798c9a6751 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 6 Jun 2009 17:53:34 +0000 Subject: [PATCH 201/296] GSdx: pixel shaders were reorganized, things might be broken :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1341 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRenderer.cpp | 3 +- plugins/GSdx/GSRenderer.h | 7 +- plugins/GSdx/GSRendererHW.h | 12 +- plugins/GSdx/GSRendererHW10.cpp | 7 +- plugins/GSdx/GSRendererHW9.cpp | 11 +- plugins/GSdx/GSTextureCache.cpp | 19 +- plugins/GSdx/GSTextureFX10.cpp | 5 - plugins/GSdx/GSTextureFX10.h | 9 +- plugins/GSdx/GSTextureFX9.cpp | 22 +- plugins/GSdx/GSTextureFX9.h | 9 +- plugins/GSdx/res/tfx10.fx | 384 ++++++++++++++++---------------- plugins/GSdx/res/tfx9.fx | 295 ++++++++++++++---------- 12 files changed, 414 insertions(+), 369 deletions(-) diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 1391efcd21..0096c85607 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -22,11 +22,10 @@ #include "StdAfx.h" #include "GSRenderer.h" -GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr) +GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev) : GSState(base, mt, irq) , m_dev(dev) , m_shader(0) - , m_psrr(psrr) { m_interlace = theApp.GetConfig("interlace", 0); m_aspectratio = theApp.GetConfig("aspectratio", 1); diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 3e9bdd2308..97ad72df2c 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -51,7 +51,6 @@ protected: public: GSWnd m_wnd; GSDevice* m_dev; - bool m_psrr; int s_n; bool s_dump; @@ -59,7 +58,7 @@ public: bool s_savez; public: - GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr = true); + GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev); virtual ~GSRenderer(); virtual bool Create(const string& title); @@ -216,8 +215,8 @@ protected: virtual void Draw() = 0; public: - GSRendererT(uint8* base, bool mt, void (*irq)(), GSDevice* dev, bool psrr = true) - : GSRenderer(base, mt, irq, dev, psrr) + GSRendererT(uint8* base, bool mt, void (*irq)(), GSDevice* dev) + : GSRenderer(base, mt, irq, dev) , m_count(0) , m_maxcount(0) , m_vertices(NULL) diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 8155aaffca..aea77663f3 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -179,8 +179,8 @@ protected: vr.z = v.z; break; case CLAMP_REGION_REPEAT: - if(m_psrr) {vr.x = maxu; vr.z = vr.x + (minu + 1);} - //else {vr.x = 0; vr.z = w;} + vr.x = maxu; + vr.z = vr.x + (minu + 1); break; default: __assume(0); @@ -202,8 +202,8 @@ protected: vr.w = v.w; break; case CLAMP_REGION_REPEAT: - if(m_psrr) {vr.y = maxv; vr.w = vr.y + (minv + 1);} - //else {r.y = 0; r.w = w;} + vr.y = maxv; + vr.w = vr.y + (minv + 1); break; default: __assume(0); @@ -691,8 +691,8 @@ protected: } public: - GSRendererHW(uint8* base, bool mt, void (*irq)(), GSDevice* dev, GSTextureCache* tc, bool psrr) - : GSRendererT(base, mt, irq, dev, psrr) + GSRendererHW(uint8* base, bool mt, void (*irq)(), GSDevice* dev, GSTextureCache* tc) + : GSRendererT(base, mt, irq, dev) , m_tc(tc) , m_width(1024) , m_height(1024) diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 26a3ddee7c..69bf582df4 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -25,7 +25,7 @@ #include "resource.h" GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice10(), new GSTextureCache10(this), true) + : GSRendererHW(base, mt, irq, new GSDevice10(), new GSTextureCache10(this)) { InitVertexKick(); } @@ -361,7 +361,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tau = 0; break; case 2: - ps_cb.MINU = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_cb.MINU = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); ps_cb.MAXU = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_ssel.tau = 0; break; @@ -383,7 +383,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tav = 0; break; case 2: - ps_cb.MINV = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_cb.MINV = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); ps_cb.MAXV = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_ssel.tav = 0; break; @@ -401,6 +401,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_cb.WH = GSVector2(w, h); ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); + ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else { diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 27be32f1d2..3b90ca47a8 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -25,7 +25,7 @@ #include "resource.h" GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this), true) + : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this)) { m_fba.enabled = !!theApp.GetConfig("fba", 1); m_logz = !!theApp.GetConfig("logz", 0); @@ -335,7 +335,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_ssel.tau = 0; break; case 2: - ps_cb.MINU = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_cb.MINU = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); ps_cb.MAXU = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_ssel.tau = 0; break; @@ -357,7 +357,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_ssel.tav = 0; break; case 2: - ps_cb.MINV = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_cb.MINV = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); ps_cb.MAXV = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_ssel.tav = 0; break; @@ -375,6 +375,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_cb.WH = GSVector2(w, h); ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); + ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else { @@ -393,7 +394,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); m_tfx.SetupIA(m_vertices, m_count, topology); m_tfx.SetupVS(vs_sel, &vs_cb); - m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL, m_psrr); + m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); m_tfx.SetupRS(w, h, scissor); // draw @@ -411,7 +412,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_sel.atst = iatst[ps_sel.atst]; - m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel, m_psrr); + m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel); bool z = om_dssel.zwe; bool r = om_bsel.wr; diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 27b789ee0f..43a6587019 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -266,7 +266,6 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() { if(TEX0.PSM == t->m_TEX0.PSM && TEX0.TBW == t->m_TEX0.TBW && TEX0.TW == t->m_TEX0.TW && TEX0.TH == t->m_TEX0.TH - && (m_renderer->m_psrr || (CLAMP.WMS != 3 && t->m_CLAMP.WMS != 3 && CLAMP.WMT != 3 && t->m_CLAMP.WMT != 3 || CLAMP.u64 == t->m_CLAMP.u64)) && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) { m_tex.splice(m_tex.begin(), m_tex, i); @@ -705,14 +704,7 @@ void GSTextureCache::GSCachedTexture::Update() { // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource - if(m_renderer->m_psrr) - { - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - else - { - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } + m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); m_texture->Unmap(); } @@ -722,14 +714,7 @@ void GSTextureCache::GSCachedTexture::Update() pitch = ((r.width() + 3) & ~3) * 4; - if(m_renderer->m_psrr) - { - m_renderer->m_mem.ReadTextureNPNC(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } - else - { - m_renderer->m_mem.ReadTextureNP(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); - } + m_renderer->m_mem.ReadTextureNPNC(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); m_texture->Update(r, buff, pitch); } diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 48cf096cf7..0dec170dcc 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -268,11 +268,6 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(sel.tfx != 4) { - if(sel.bpp >= 3 || sel.wms >= 3 || sel.wmt >= 3) - { - ssel.min = ssel.mag = 0; - } - hash_map >::const_iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index f87306972d..cc8653ca7e 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -79,12 +79,12 @@ public: { GSVector4 FogColor; float MINU; - float MAXU; float MINV; + float MAXU; float MAXV; uint32 UMSK; - uint32 UFIX; uint32 VMSK; + uint32 UFIX; uint32 VFIX; float TA0; float TA1; @@ -92,6 +92,7 @@ public: float _pad[1]; GSVector2 WH; GSVector2 rWrH; + GSVector4 HalfTexel; struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} @@ -105,14 +106,16 @@ public: GSVector4i b2 = b[2]; GSVector4i b3 = b[3]; GSVector4i b4 = b[4]; + GSVector4i b5 = b[5]; - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4)).alltrue()) + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) { a[0] = b0; a[1] = b1; a[2] = b2; a[3] = b3; a[4] = b4; + a[5] = b5; return true; } diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 5a28c42b3e..7abf2a550c 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -144,11 +144,11 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) return true; } -bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal, bool psrr) +bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) { m_dev->PSSetShaderResources(tex, pal); - if(tex && psrr && (sel.wms == 3 || sel.wmt == 3)) + if(tex && (sel.wms == 3 || sel.wmt == 3)) { if(sel.wms == 3) { @@ -167,21 +167,15 @@ bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSampler } } - UpdatePS(sel, cb, ssel, psrr); + UpdatePS(sel, cb, ssel); return true; } -void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, bool psrr) +void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) { HRESULT hr; - if(!psrr) - { - if(sel.wms == 3) sel.wms = 0; - if(sel.wmt == 3) sel.wmt = 0; - } - hash_map >::const_iterator i = m_ps.find(sel); if(i == m_ps.end()) @@ -233,10 +227,10 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample if(sel.tfx != 4) { - if(sel.bpp >= 3 || sel.wms >= 3 || sel.wmt >= 3) - { - ssel.min = ssel.mag = 0; - } + bool b = sel.bpp < 3 && sel.wms < 3 && sel.wmt < 3; + + ssel.min = b; + ssel.mag = b; hash_map::const_iterator i = m_ps_ss.find(ssel); diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 65bb7bbfbf..54fbb258ca 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -55,12 +55,12 @@ public: { GSVector4 FogColor; float MINU; - float MAXU; float MINV; + float MAXU; float MAXV; uint32 UMSK; - uint32 UFIX; uint32 VMSK; + uint32 UFIX; uint32 VFIX; float TA0; float TA1; @@ -68,6 +68,7 @@ public: float _pad[1]; GSVector2 WH; GSVector2 rWrH; + GSVector4 HalfTexel; }; union PSSelector @@ -166,8 +167,8 @@ public: bool SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim); bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal, bool psrr); - void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, bool psrr); + bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); + void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 78d1c4403d..dd172431d8 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -155,28 +155,25 @@ void gs_main(line VS_OUTPUT input[2], inout TriangleStream stream) #endif -Texture2D Texture; -Texture2D Palette; +Texture2D Texture; +Texture2D Palette; SamplerState TextureSampler; SamplerState PaletteSampler; cbuffer cb1 { float4 FogColor; - float MINU; - float MAXU; - float MINV; - float MAXV; - uint UMSK; - uint UFIX; - uint VMSK; - uint VFIX; + float2 MINUV; + float2 MAXUV; + uint2 UVMSK; + uint2 UVFIX; float TA0; float TA1; float AREF; float _pad; float2 WH; float2 rWrH; + float4 HalfTexel; }; struct PS_INPUT @@ -194,8 +191,8 @@ struct PS_OUTPUT #ifndef FST #define FST 0 -#define WMS 3 -#define WMT 3 +#define WMS 0 +#define WMT 0 #define BPP 0 #define AEM 0 #define TFX 0 @@ -225,191 +222,185 @@ float4 Extract16(uint i) return f; } -int repeatu(int tc) +int2 wrapu(float2 f, int2 i) { - return WMS == 3 ? ((tc & UMSK) | UFIX) : tc; -} - -int repeatv(int tc) -{ - return WMT == 3 ? ((tc & VMSK) | VFIX) : tc; -} - -float4 sample(float2 tc) -{ - float4 t; - - // if(WMS >= 2 || WMT >= 2) - if(WMS >= 3 || WMT >= 3) + if(WMS == 0) { - int4 itc = tc.xyxy * WH.xyxy; - - float4 tc01; - - tc01.x = repeatu(itc.x); - tc01.y = repeatv(itc.y); - tc01.z = repeatu(itc.z + 1); - tc01.w = repeatv(itc.w + 1); - - tc01 *= rWrH.xyxy; + i = frac(f) * WH.xx; + } + else if(WMS == 1) + { + i = saturate(f) * WH.xx; + } + else if(WMS == 2) + { + i = clamp(f, MINUV.xx, MAXUV.xx) * WH.xx; + } + else if(WMS == 3) + { + i = (i & UVMSK.xx) | UVFIX.xx; + } + + return i; +} - float4 t00 = Texture.Sample(TextureSampler, tc01.xy); - float4 t01 = Texture.Sample(TextureSampler, tc01.zy); - float4 t10 = Texture.Sample(TextureSampler, tc01.xw); - float4 t11 = Texture.Sample(TextureSampler, tc01.zw); +int2 wrapv(float2 f, int2 i) +{ + if(WMT == 0) + { + i = frac(f) * WH.yy; + } + else if(WMT == 1) + { + i = saturate(f) * WH.yy; + } + else if(WMT == 2) + { + i = clamp(f, MINUV.yy, MAXUV.yy) * WH.yy; + } + else if(WMT == 3) + { + i = (i & UVMSK.yy) | UVFIX.yy; + } + + return i; +} - float2 dd = frac(tc * WH); +int4 wrapuv(float4 f, int4 i) +{ + if(WMT == 0) + { + i = frac(f) * WH.xyxy; + } + else if(WMT == 1) + { + i = saturate(f) * WH.xyxy; + } + else if(WMT == 2) + { + i = clamp(f, MINUV.xyxy, MAXUV.xyxy) * WH.xyxy; + } + else if(WMT == 3) + { + i = (i & UVMSK.xyxy) | UVFIX.xyxy; + } - t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); + return i; +} + +int4 wrap(float4 uv, int4 iuv) +{ + if(WMS == WMT) + { + iuv = wrapuv(uv, iuv); } else + { + iuv.xz = wrapu(uv.xz, iuv.xz); + iuv.yw = wrapv(uv.yw, iuv.yw); + } + + return iuv; +} + +float4 sample(float2 tc, float w) +{ + if(FST == 0) + { + tc /= w; + } + + float4 t; +/* + if(BPP < 3 && WMS < 2 && WMT < 2) { t = Texture.Sample(TextureSampler, tc); } +*/ + if(BPP < 3 && WMS < 3 && WMT < 3) + { + if(WMS == 2 && WMT == 2) tc = clamp(tc, MINUV.xy, MAXUV.xy); + else if(WMS == 2) tc.x = clamp(tc.x, MINUV.x, MAXUV.x); + else if(WMT == 2) tc.y = clamp(tc.y, MINUV.y, MAXUV.y); + + t = Texture.Sample(TextureSampler, tc); + } + else + { + float4 uv = tc.xyxy + HalfTexel; + float4 uv2 = uv * WH.xyxy; + float2 dd = frac(uv2.xy); + + int4 iuv = wrap(uv, uv2); + + float4 t00, t01, t10, t11; + + if(BPP == 3) // 8HP + 32-bit palette + { + float4 a; + + a.x = Texture.Load(int3(iuv.xy, 0)).a; + a.y = Texture.Load(int3(iuv.zy, 0)).a; + a.z = Texture.Load(int3(iuv.xw, 0)).a; + a.w = Texture.Load(int3(iuv.zw, 0)).a; + + t00 = Palette.Load(a.x); + t01 = Palette.Load(a.y); + t10 = Palette.Load(a.z); + t11 = Palette.Load(a.w); + } + else if(BPP == 4) // 8HP + 16-bit palette + { + // TODO: yuck, just pre-convert the palette to 32-bit + } + else if(BPP == 5) // 16P + { + float4 r; + + r.x = Texture.Load(int3(iuv.xy, 0)).r; + r.y = Texture.Load(int3(iuv.zy, 0)).r; + r.z = Texture.Load(int3(iuv.xw, 0)).r; + r.w = Texture.Load(int3(iuv.zw, 0)).r; + + uint4 i = r * 65535; + + t00 = Extract16(i.x); + t01 = Extract16(i.y); + t10 = Extract16(i.z); + t11 = Extract16(i.w); + } + else + { + t00 = Texture.Load(int3(iuv.xy, 0)); + t01 = Texture.Load(int3(iuv.zy, 0)); + t10 = Texture.Load(int3(iuv.xw, 0)); + t11 = Texture.Load(int3(iuv.zw, 0)); + } + + t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); + } + + if(BPP == 1) // 24 + { + t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; + } + else if(BPP == 2 || BPP == 5) // 16 || 16P + { + if(BPP == 5) + { + t = Normalize16(t); + } + + // a bit incompatible with up-scaling because the 1 bit alpha is interpolated + + t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; + } return t; } -float4 sample8hp(float2 tc) +float4 tfx(float4 t, float4 c) { - float4 tc01; - - // if(WMS >= 2 || WMT >= 2) - if(WMS >= 3 || WMT >= 3) - { - int4 itc = tc.xyxy * WH.xyxy; - - tc01.x = repeatu(itc.x); - tc01.y = repeatv(itc.y); - tc01.z = repeatu(itc.z + 1); - tc01.w = repeatv(itc.w + 1); - - tc01 *= rWrH.xyxy; - } - else - { - tc01.x = tc.x; - tc01.y = tc.y; - tc01.z = tc.x + rWrH.x; - tc01.w = tc.y + rWrH.y; - } - - float4 t; - - t.x = Texture.Sample(TextureSampler, tc01.xy).a; - t.y = Texture.Sample(TextureSampler, tc01.zy).a; - t.z = Texture.Sample(TextureSampler, tc01.xw).a; - t.w = Texture.Sample(TextureSampler, tc01.zw).a; - - float4 t00 = Palette.Sample(PaletteSampler, t.x); - float4 t01 = Palette.Sample(PaletteSampler, t.y); - float4 t10 = Palette.Sample(PaletteSampler, t.z); - float4 t11 = Palette.Sample(PaletteSampler, t.w); - - float2 dd = frac(tc * WH); - - return lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); -} - -float4 sample16p(float2 tc) -{ - float4 t; - - float4 tc01; - - // if(WMS >= 2 || WMT >= 2) - if(WMS >= 3 || WMT >= 3) - { - int4 itc = tc.xyxy * WH.xyxy; - - tc01.x = repeatu(itc.x); - tc01.y = repeatv(itc.y); - tc01.z = repeatu(itc.z + 1); - tc01.w = repeatv(itc.w + 1); - - tc01 *= rWrH.xyxy; - } - else - { - tc01.x = tc.x; - tc01.y = tc.y; - tc01.z = tc.x + rWrH.x; - tc01.w = tc.y + rWrH.y; - } - - t.x = Texture.Sample(TextureSampler, tc01.xy).r; - t.y = Texture.Sample(TextureSampler, tc01.zy).r; - t.z = Texture.Sample(TextureSampler, tc01.xw).r; - t.w = Texture.Sample(TextureSampler, tc01.zw).r; - - uint4 i = t * 65535; - - float4 t00 = Extract16(i.x); - float4 t01 = Extract16(i.y); - float4 t10 = Extract16(i.z); - float4 t11 = Extract16(i.w); - - float2 dd = frac(tc * WH); - - return Normalize16(lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y)); -} - -PS_OUTPUT ps_main(PS_INPUT input) -{ - float2 tc = input.t.xy; - - if(FST == 0) - { - tc /= input.t.w; - } - - tc -= rWrH / 2; - - if(WMS == 2) - { - tc.x = clamp(tc.x, MINU, MAXU); - } - - if(WMT == 2) - { - tc.y = clamp(tc.y, MINV, MAXV); - } - - float4 t; - - if(BPP == 0) // 32 - { - t = sample(tc); - } - else if(BPP == 1) // 24 - { - t = sample(tc); - - t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; - } - else if(BPP == 2) // 16 - { - t = sample(tc); - - t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - } - else if(BPP == 3) // 8HP / 32-bit palette - { - t = sample8hp(tc); - } - else if(BPP == 4) // 8HP / 16-bit palette - { - // TODO: yuck, just pre-convert the palette to 32-bit - } - else if(BPP == 5) // 16P - { - t = sample16p(tc); - - t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - } - - float4 c = input.c; - if(TFX == 0) { if(TCC == 0) @@ -435,7 +426,7 @@ PS_OUTPUT ps_main(PS_INPUT input) else if(TFX == 2) { c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - + if(TCC == 1) { c.a += t.a; @@ -444,15 +435,18 @@ PS_OUTPUT ps_main(PS_INPUT input) else if(TFX == 3) { c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - + if(TCC == 1) { c.a = t.a; } } + + return saturate(c); +} - c = saturate(c); - +void atst(float4 c) +{ if(ATE == 1) { if(ATST == 0) @@ -476,12 +470,28 @@ PS_OUTPUT ps_main(PS_INPUT input) clip(abs(c.a - AREF) - 0.4f / 255); // FIXME: 0.5f is too much } } +} +float4 fog(float4 c, float f) +{ if(FOG == 1) { - c.rgb = lerp(FogColor.rgb, c.rgb, input.t.z); + c.rgb = lerp(FogColor.rgb, c.rgb, f); } + return c; +} + +PS_OUTPUT ps_main(PS_INPUT input) +{ + float4 t = sample(input.t.xy, input.t.w); + + float4 c = tfx(t, input.c); + + atst(c); + + c = fog(c, input.t.z); + if(CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes { c.rgb = 1; diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index b17a07f822..ebe822ffee 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -71,22 +71,19 @@ VS_OUTPUT vs_main(VS_INPUT input) return output; } -float4 ps_params[5]; +float4 ps_params[6]; #define FogColor ps_params[0].bgra -#define MINU ps_params[1].x -#define MAXU ps_params[1].y -#define MINV ps_params[1].z -#define MAXV ps_params[1].w -#define UMSK ps_params[2].x -#define UFIX ps_params[2].y -#define VMSK ps_params[2].z -#define VFIX ps_params[2].w +#define MINUV ps_params[1].xy +#define MAXUV ps_params[1].zw +#define UVMSK ps_params[2].xy +#define UVFIX ps_params[2].zw #define TA0 ps_params[3].x #define TA1 ps_params[3].y #define AREF ps_params[3].z #define WH ps_params[4].xy #define rWrH ps_params[4].zw +#define HalfTexel ps_params[5] struct PS_INPUT { @@ -114,129 +111,170 @@ sampler1D Palette : register(s1); sampler1D UMSKFIX : register(s2); sampler1D VMSKFIX : register(s3); -float repeatu(float tc) +float2 wrapu(float2 f) { - return WMS == 3 ? tex1D(UMSKFIX, tc*rWrH.x)* WH.x : tc; -} - -float repeatv(float tc) -{ - return WMT == 3 ? tex1D(VMSKFIX, tc*rWrH.y)* WH.y : tc; -} - -float4 sample(float2 tc) -{ - float4 t; - - // if(WMS >= 2 || WMT >= 2) - if(WMS >= 3 || WMT >= 3) + if(WMS == 0) { - tc -= rWrH / 2; - - int4 itc = tc.xyxy * WH.xyxy; - - float4 tc01; - - tc01.x = repeatu(itc.x); - tc01.y = repeatv(itc.y); - tc01.z = repeatu(itc.z + 1); - tc01.w = repeatv(itc.w + 1); + f = frac(f); + } + else if(WMS == 1) + { + f = saturate(f); + } + else if(WMS == 2) + { + f = clamp(f, MINUV.xx, MAXUV.xx); + } + else if(WMS == 3) + { + f.x = tex1D(UMSKFIX, f.x); + f.y = tex1D(UMSKFIX, f.y); + } - tc01 *= rWrH.xyxy; + return f; +} - float4 t00 = tex2D(Texture, tc01.xy); - float4 t01 = tex2D(Texture, tc01.zy); - float4 t10 = tex2D(Texture, tc01.xw); - float4 t11 = tex2D(Texture, tc01.zw); +float2 wrapv(float2 f) +{ + if(WMS == 0) + { + f = frac(f); + } + else if(WMS == 1) + { + f = saturate(f); + } + else if(WMS == 2) + { + f = clamp(f, MINUV.yy, MAXUV.yy); + } + else if(WMS == 3) + { + f.x = tex1D(VMSKFIX, f.x); + f.y = tex1D(VMSKFIX, f.y); + } - float2 dd = frac(tc * WH); + return f; +} - t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); +float4 wrapuv(float4 f) +{ + if(WMS == 0) + { + f = frac(f); + } + else if(WMS == 1) + { + f = saturate(f); + } + else if(WMS == 2) + { + f = clamp(f, MINUV.xyxy, MAXUV.xyxy); + } + else if(WMS == 3) + { + f.x = tex1D(UMSKFIX, f.x); + f.y = tex1D(VMSKFIX, f.y); + f.z = tex1D(UMSKFIX, f.z); + f.w = tex1D(VMSKFIX, f.w); + } + + return f; +} + +float4 wrap(float4 uv) +{ + if(WMS == WMT) + { + uv = wrapuv(uv); } else + { + uv.xz = wrapu(uv.xz); + uv.yw = wrapv(uv.yw); + } + + return uv; +} + +float4 sample(float2 tc, float w) +{ + if(FST == 0) + { + tc /= w; + } + + float4 t; +/* + if(BPP < 3 && WMS < 2 && WMT < 2) { t = tex2D(Texture, tc); } +*/ + if(BPP < 3 && WMS < 3 && WMT < 3) + { + if(WMS == 2 && WMT == 2) tc = clamp(tc, MINUV.xy, MAXUV.xy); + else if(WMS == 2) tc.x = clamp(tc.x, MINUV.x, MAXUV.x); + else if(WMT == 2) tc.y = clamp(tc.y, MINUV.y, MAXUV.y); + + t = tex2D(Texture, tc); + } + else + { + float4 uv = tc.xyxy + HalfTexel; + float2 dd = frac(uv.xy * WH); + + uv = wrap(uv); + + float4 t00, t01, t10, t11; + + if(BPP == 3) // 8HP ln + { + float4 a; + + a.x = tex2D(Texture, uv.xy).a; + a.y = tex2D(Texture, uv.zy).a; + a.z = tex2D(Texture, uv.xw).a; + a.w = tex2D(Texture, uv.zw).a; + + if(RT == 1) a *= 0.5; + + t00 = tex1D(Palette, a.x); + t01 = tex1D(Palette, a.y); + t10 = tex1D(Palette, a.z); + t11 = tex1D(Palette, a.w); + } + else + { + t00 = tex2D(Texture, uv.xy); + t01 = tex2D(Texture, uv.zy); + t10 = tex2D(Texture, uv.xw); + t11 = tex2D(Texture, uv.zw); + } + + t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); + } + + if(BPP == 0) // 32 + { + if(RT == 1) t.a *= 0.5; + } + else if(BPP == 1) // 24 + { + t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; + } + else if(BPP == 2) // 16 + { + // a bit incompatible with up-scaling because the 1 bit alpha is interpolated + + t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; + } return t; } -float4 sample8hp(float2 tc) +float4 tfx(float4 t, float4 c) { - tc -= rWrH / 2; - - float4 tc01; - - tc01.x = tc.x; - tc01.y = tc.y; - tc01.z = tc.x + rWrH.x; - tc01.w = tc.y + rWrH.y; - - float4 t; - - t.x = tex2D(Texture, tc01.xy).a; - t.y = tex2D(Texture, tc01.zy).a; - t.z = tex2D(Texture, tc01.xw).a; - t.w = tex2D(Texture, tc01.zw).a; - - if(RT == 1) t *= 0.5; - - float4 t00 = tex1D(Palette, t.x); - float4 t01 = tex1D(Palette, t.y); - float4 t10 = tex1D(Palette, t.z); - float4 t11 = tex1D(Palette, t.w); - - float2 dd = frac(tc * WH); - - return lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); -} - -float4 ps_main(PS_INPUT input) : COLOR -{ - float2 tc = input.t.xy; - - if(FST == 0) - { - tc /= input.t.w; - } - - if(WMS == 2) - { - tc.x = clamp(tc.x, MINU, MAXU); - } - - if(WMT == 2) - { - tc.y = clamp(tc.y, MINV, MAXV); - } - - float4 t; - - if(BPP == 0) // 32 - { - t = sample(tc); - - if(RT == 1) t.a *= 0.5; - } - else if(BPP == 1) // 24 - { - t = sample(tc); - - t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; - } - else if(BPP == 2) // 16 - { - t = sample(tc); - - t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - } - else if(BPP == 3) // 8HP ln - { - t = sample8hp(tc); - } - - float4 c = input.c; - if(TFX == 0) { if(TCC == 0) @@ -277,9 +315,12 @@ float4 ps_main(PS_INPUT input) : COLOR c.a = t.a; } } + + return saturate(c); +} - c = saturate(c); - +void atst(float4 c) +{ if(ATE == 1) { if(ATST == 0) @@ -303,12 +344,28 @@ float4 ps_main(PS_INPUT input) : COLOR clip(abs(c.a - AREF) - 0.4f / 255); // FIXME: 0.5f is too much } } +} +float4 fog(float4 c, float f) +{ if(FOG == 1) { - c.rgb = lerp(FogColor.rgb, c.rgb, input.t.z); + c.rgb = lerp(FogColor.rgb, c.rgb, f); } + return c; +} + +float4 ps_main(PS_INPUT input) : COLOR +{ + float4 t = sample(input.t.xy, input.t.w); + + float4 c = tfx(t, input.c); + + atst(c); + + c = fog(c, input.t.z); + if(CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes { c.rgb = 1; From 96fb298371832bcdac8f0cafe2208dea076e8c28 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 7 Jun 2009 07:13:54 +0000 Subject: [PATCH 202/296] Linux: Make the build system even more hackish, to avoid issues with AC_OUTPUT and automake 1.11. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1342 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/3rdparty | 1 + pcsx2/configure.ac | 5 +---- 2 files changed, 2 insertions(+), 4 deletions(-) create mode 120000 pcsx2/3rdparty diff --git a/pcsx2/3rdparty b/pcsx2/3rdparty new file mode 120000 index 0000000000..2781723b7c --- /dev/null +++ b/pcsx2/3rdparty @@ -0,0 +1 @@ +../3rdparty/ \ No newline at end of file diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index d6e41960b1..7b7a745f92 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -31,9 +31,6 @@ customcflags=$enableval,customcflags=no) AC_ARG_ENABLE(optimization, AC_HELP_STRING([--disable-optimization], [No optimizations when compiling]), optimization=$enableval,optimization=yes) -AC_ARG_ENABLE(unsupportedspeedfreak, AC_HELP_STRING([--enable-unsupportedspeedfreak], [Set the compiler to do unreasonable optimizations that will probably make all your games crash.]), -unsupportedspeedfreak=$enableval,unsupportedspeedfreak=no) - dnl Check for debug build AC_MSG_CHECKING(debug build) AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), @@ -191,7 +188,7 @@ AC_OUTPUT([ tinyxml/Makefile x86/Makefile x86/ix86/Makefile - ../3rdparty/zlib/Makefile + 3rdparty/zlib/Makefile ]) dnl bindir = pcsx2exe From b61f7cc4d18e2dd426b088d2d50e1066595759b8 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 7 Jun 2009 09:42:10 +0000 Subject: [PATCH 203/296] microVU: - Implemented E-bit Conditional Branches and Jumps. Should fix infinite loop problems in games like Zone of the Enders, but it hasn't been tested yet since I don't have the game xD. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1343 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 14 ++-- pcsx2/x86/microVU_Compile.inl | 141 +++++++++++++++++++--------------- 2 files changed, 86 insertions(+), 69 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index a33cd8a5bf..0dab141a4b 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -32,27 +32,27 @@ class microBlockManager { private: static const int MaxBlocks = mMaxBlocks - 1; - microBlock blockList[mMaxBlocks]; // note: should always be first in the class to ensure xmm alignment + microBlock blockList[mMaxBlocks]; // Should always be first in the class to ensure 16-byte alignment int listSize; // Total Items - 1 int listI; // Index to Add new block public: - // our aligned replacement for 'new': + // Aligned replacement for 'new' static microBlockManager* AlignedNew() { microBlockManager* alloc = (microBlockManager*)_aligned_malloc(sizeof(microBlockManager), 16); new (alloc) microBlockManager(); return alloc; } - - static void Delete( microBlockManager* dead ) { - if( dead == NULL ) return; + // Use instead of normal 'delete' + static void Delete(microBlockManager* dead) { + if (dead == NULL) return; dead->~microBlockManager(); _aligned_free( dead ); } microBlockManager() { reset(); } - ~microBlockManager() { } - void reset() { listSize = -1; listI = -1; }; + ~microBlockManager() {} + void reset() { listSize = -1; listI = -1; }; microBlock* add(microBlock* pBlock) { microBlock* thisBlock = search(&pBlock->pState); if (!thisBlock) { diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index ee771e06ee..30ac835a2b 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -22,15 +22,30 @@ // Helper Macros //------------------------------------------------------------------ -#define branchCase(JMPcc, nJMPcc) \ - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ - CMP16ItoM((uptr)&mVU->branch, 0); \ - incPC2(1); \ - if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ - bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ - incPC2(-1); \ - if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ - else { ajmp = JMPcc((uptr)0); } \ +#define branchCase(JMPcc, nJMPcc, ebitJMP) \ + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ + CMP16ItoM((uptr)&mVU->branch, 0); \ + if (mVUup.eBit) { /* Conditional Branch With E-Bit Set */ \ + mVUendProgram(mVU, 2, xStatus, xMac, xClip); \ + u8* eJMP = ebitJMP(0); \ + incPC(1); /* Set PC to First instruction of Non-Taken Side */ \ + MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); \ + JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); \ + x86SetJ8(eJMP); \ + incPC(-4); /* Go Back to Branch Opcode to get branchAddr */ \ + iPC = branchAddr/4; \ + MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); \ + JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); \ + return thisPtr; \ + } \ + else { /* Normal Conditional Branch */ \ + incPC2(1); /* Check if Branch Non-Taken Side has already been recompiled */ \ + if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ + bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ + incPC2(-1); \ + if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ + else { ajmp = JMPcc((uptr)0); } \ + } \ break #define branchWarning() { \ @@ -40,15 +55,6 @@ } \ } -#define branchEbit() { \ - if (branch == 2) { \ - if (!((mVUbranch == 1) || (mVUbranch == 2))) { \ - Console::Error("microVU%d Warning: Jump with E-bit not Implemented [%04x]", params vuIndex, xPC); \ - } \ - else { eBitBranch = 1; } \ - } \ -} - #define doBackupVF1() { \ if (mVUinfo.backupVF && !mVUlow.noWriteVF) { \ DevCon::Status("microVU%d: Backing Up VF Reg [%04x]", params getIndex, xPC); \ @@ -205,7 +211,31 @@ microVUt(void) mVUsetCycles(mV) { tCycles(mVUregs.xgkick, mVUregsTemp.xgkick); } -microVUt(void) mVUendProgram(mV, bool clearIsBusy, int qInst, int pInst, int fStatus, int fMac, int fClip) { +#define sI ((mVUpBlock->pState.needExactMatch & 0x000f) ? 0 : ((mVUpBlock->pState.flags >> 0) & 3)) +#define cI ((mVUpBlock->pState.needExactMatch & 0x0f00) ? 0 : ((mVUpBlock->pState.flags >> 2) & 3)) + +microVUt(void) mVUendProgram(mV, int isEbit, int* xStatus, int* xMac, int* xClip) { + + int fStatus = (isEbit) ? findFlagInst(xStatus, 0x7fffffff) : sI; + int fMac = (isEbit) ? findFlagInst(xMac, 0x7fffffff) : 0; + int fClip = (isEbit) ? findFlagInst(xClip, 0x7fffffff) : cI; + int qInst = 0; + int pInst = 0; + + if (isEbit) { + mVUprint("mVUcompile ebit"); + memset(&mVUinfo, 0, sizeof(mVUinfo)); + mVUincCycles(mVU, 100); // Ensures Valid P/Q instances (And sets all cycle data to 0) + mVUcycles -= 100; + qInst = mVU->q; + pInst = mVU->p; + if (mVUinfo.doDivFlag) { + sFLAG.doFlag = 1; + sFLAG.write = fStatus; + mVUdivSet(mVU); + } + if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } + } // Save P/Q Regs if (qInst) { SSE2_PSHUFD_XMM_to_XMM(xmmPQ, xmmPQ, 0xe5); } @@ -223,19 +253,18 @@ microVUt(void) mVUendProgram(mV, bool clearIsBusy, int qInst, int pInst, int fSt MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); - if (clearIsBusy) { // Clear 'is busy' Flags + if (isEbit || !isVU1) { // Clear 'is busy' Flags AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (isVU1 ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif } - //Save PC, and Jump to Exit Point - MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); - JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); + if (isEbit != 2) { // Save PC, and Jump to Exit Point + mVUsetupRange(mVU, xPC); + MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); + JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); + } } -#define sI ((mVUpBlock->pState.needExactMatch & 0x000f) ? 0 : ((mVUpBlock->pState.flags >> 0) & 3)) -#define cI ((mVUpBlock->pState.needExactMatch & 0x0f00) ? 0 : ((mVUpBlock->pState.flags >> 2) & 3)) - void __fastcall mVUwarning0(u32 PC) { Console::Error("microVU0 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUprintPC1(u32 PC) { Console::Write("Block PC [%04x] ", params PC); } @@ -250,7 +279,7 @@ microVUt(void) mVUtestCycles(mV) { if (isVU1) CALLFunc((uptr)mVUwarning1); //else CALLFunc((uptr)mVUwarning0); // VU0 is allowed early exit for COP2 Interlock Simulation MOV32ItoR(gprR, Roffset); // Restore gprR - mVUendProgram(mVU, (isVU1)?1:0, 0, 0, sI, 0, cI); + mVUendProgram(mVU, 0, NULL, NULL, NULL); x86SetJ32(jmp32); } @@ -292,7 +321,6 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUregs.flags = 0; mVUflagInfo = 0; mVUsFlagHack = CHECK_VU_FLAGHACK; - bool eBitBranch = 0; // E-bit Set on Branch for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { incPC(1); @@ -311,7 +339,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUinfo.writeP = !mVU->p; if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(mVU); branchEbit(); branch = 3; mVUbranch = 0; } + if (mVUbranch) { mVUsetFlagInfo(mVU); branch = 3; mVUbranch = 0; } incPC(1); mVUcount++; } @@ -342,18 +370,18 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { mVUdebugNOW(1); switch (mVUbranch) { - case 3: branchCase(JE32, JNE32); // IBEQ - case 4: branchCase(JGE32, JNGE32); // IBGEZ - case 5: branchCase(JG32, JNG32); // IBGTZ - case 6: branchCase(JLE32, JNLE32); // IBLEQ - case 7: branchCase(JL32, JNL32); // IBLTZ - case 8: branchCase(JNE32, JE32); // IBNEQ + case 3: branchCase(JE32, JNE32, JE8); // IBEQ + case 4: branchCase(JGE32, JNGE32, JGE8); // IBGEZ + case 5: branchCase(JG32, JNG32, JG8); // IBGTZ + case 6: branchCase(JLE32, JNLE32, JLE8); // IBLEQ + case 7: branchCase(JL32, JNL32, JL8); // IBLTZ + case 8: branchCase(JNE32, JE32, JNE8); // IBNEQ case 1: case 2: // B/BAL mVUprint("mVUcompile B/BAL"); incPC(-3); // Go back to branch opcode (to get branch imm addr) - if (eBitBranch) { iPC = branchAddr/4; goto eBitTemination; } // E-bit Was Set on Branch + if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); if (mVUblocks[branchAddr/8] == NULL) @@ -368,6 +396,16 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { case 9: case 10: // JR/JALR mVUprint("mVUcompile JR/JALR"); + incPC(-3); // Go back to jump opcode + + if (mVUup.eBit) { // E-bit Jump + mVUendProgram(mVU, 2, xStatus, xMac, xClip); + MOV32MtoR(gprT1, (uptr)&mVU->branch); + MOV32RtoM((uptr)&mVU->regs->VI[REG_TPC].UL, gprT1); + JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); + return thisPtr; + } + memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); @@ -375,8 +413,8 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) - if (!isVU1) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) - else CALLFunc((uptr)mVUcompileVU1); + if (!vuIndex) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) + else CALLFunc((uptr)mVUcompileVU1); mVUrestoreRegs(mVU); JMPR(gprT1); // Jump to rec-code address return thisPtr; @@ -397,8 +435,7 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { } else { uptr jumpAddr; - u32 bPC = iPC; // mVUcompile can modify iPC, mVUregs, and mVUflagInfo, so back them up - u32 bFlagInfo = mVUflagInfo; + u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs so back them up memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); incPC2(1); // Get PC for branch not-taken @@ -406,7 +443,6 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { else mVUcompileVU1(xPC, (uptr)&mVUregs); iPC = bPC; - mVUflagInfo = bFlagInfo; incPC(-3); // Go back to branch opcode (to get branch imm addr) if (!vuIndex) jumpAddr = (uptr)mVUcompileVU0(branchAddr, (uptr)&pBlock->pStateEnd); else jumpAddr = (uptr)mVUcompileVU1(branchAddr, (uptr)&pBlock->pStateEnd); @@ -417,27 +453,8 @@ microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { } if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } -eBitTemination: - - mVUprint("mVUcompile ebit"); - int lStatus = findFlagInst(xStatus, 0x7fffffff); - int lMac = findFlagInst(xMac, 0x7fffffff); - int lClip = findFlagInst(xClip, 0x7fffffff); - memset(&mVUinfo, 0, sizeof(mVUinfo)); - mVUincCycles(mVU, 100); // Ensures Valid P/Q instances (And sets all cycle data to 0) - mVUcycles -= 100; - if (mVUinfo.doDivFlag) { - int flagReg; - getFlagReg(flagReg, lStatus); - AND32ItoR (flagReg, 0xfff3ffff); - OR32MtoR (flagReg, (uptr)&mVU->divFlag); - } - if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } - - // Do E-bit end stuff here - mVUsetupRange(mVU, xPC - 8); - mVUendProgram(mVU, 1, mVU->q, mVU->p, lStatus, lMac, lClip); - + // E-bit End + mVUendProgram(mVU, 1, xStatus, xMac, xClip); return thisPtr; } From 2ca4a603b3fe60167ca6d6a0457bb3eb8f869a79 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 7 Jun 2009 12:04:39 +0000 Subject: [PATCH 204/296] microVU: - Rewrote the custom compare function to use the emitter instead of inline asm. - Set Linux builds to use the function. Note: If this revision causes microVU to crash on Linux, it means GCC isn't guaranteeing 16-byte alignment on microRegInfo and microBlock structs. So it'll need to use normal memcmp instead (see microVU_Misc.h) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1345 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 3 +- pcsx2/x86/microVU_IR.h | 19 ++++-- pcsx2/x86/microVU_Misc.h | 17 ++--- pcsx2/x86/microVU_Misc.inl | 131 ++++++++++++------------------------- 4 files changed, 64 insertions(+), 106 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 8685e1e147..a2f65ed38f 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -49,7 +49,8 @@ microVUf(void) mVUinit(VURegs* vuRegsPtr) { mVU->cache = SysMmapEx((vuIndex ? 0x5f240000 : 0x5e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", (u32)mVU->cache)); - + + mVUemitSearch(); mVUreset(); } diff --git a/pcsx2/x86/microVU_IR.h b/pcsx2/x86/microVU_IR.h index 2fb0623027..08a794100f 100644 --- a/pcsx2/x86/microVU_IR.h +++ b/pcsx2/x86/microVU_IR.h @@ -32,6 +32,7 @@ union regInfo { #pragma pack(1) #pragma warning(disable:4996) #endif + __declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares u32 needExactMatch; // If set, block needs an exact match of pipeline state u8 q; @@ -44,6 +45,16 @@ __declspec(align(16)) struct microRegInfo { // Ordered for Faster Compares u8 padding[7]; // 160 bytes #if defined(_MSC_VER) }; +#else +} __attribute__((packed)); +#endif + +__declspec(align(16)) struct microBlock { + microRegInfo pState; // Detailed State of Pipeline + microRegInfo pStateEnd; // Detailed State of Pipeline at End of Block (needed by JR/JALR opcodes) + u8* x86ptrStart; // Start of code +#if defined(_MSC_VER) +}; #pragma pack() #else } __attribute__((packed)); @@ -60,12 +71,6 @@ struct microTempRegInfo { u8 xgkick; // Holds the cycle info for XGkick }; -__declspec(align(16)) struct microBlock { - microRegInfo pState; // Detailed State of Pipeline - microRegInfo pStateEnd; // Detailed State of Pipeline at End of Block (needed by JR/JALR opcodes) - u8* x86ptrStart; // Start of code -}; - struct microVFreg { u8 reg; // Reg Index u8 x; // X vector read/written to? @@ -132,8 +137,8 @@ struct microOp { template struct microIR { - microBlock* pBlock; // Pointer to a block in mVUblocks microBlock block; // Block/Pipeline info + microBlock* pBlock; // Pointer to a block in mVUblocks microTempRegInfo regsTemp; // Temp Pipeline info (used so that new pipeline info isn't conflicting between upper and lower instructions in the same cycle) microOp info[pSize/2]; // Info for Instructions in current block u8 branch; diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 56bc34e72c..682c5407c6 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -161,16 +161,13 @@ declareAllVariables #define pass4 if (recPass == 3) // Define mVUquickSearch -#if defined(_MSC_VER) -extern u32 __fastcall mVUsearchXMM(void *dest, void *src); -#define mVUquickSearch(dest, src, size) (mVUsearchXMM(dest, src) == 0xf) -#else -#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) -// Note: GCC might not guarantee alignment on microRegInfo, -// so to be safe I'm using normal memcmp. If at least 8-byte -// alignment is guaranteed, the function below is faster. -// #define mVUquickSearch(dest, src, size) (!memcmp_mmx(dest, src, size)) -#endif +extern u8 mVUsearchXMM[0x1000]; +typedef u32 (__fastcall *mVUCall)(void*, void*); +#define mVUquickSearch(dest, src, size) ((((mVUCall)((void*)mVUsearchXMM))(dest, src)) == 0xf) +// Note: If GCC builds crash with above function, it means +// that they're not guaranteeing 16-byte alignment on the structs +// being compared. So use this function instead: +// #define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) // Misc Macros... #define mVUprogI mVU->prog.prog[progIndex] diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index e2379dac43..ca60c20b33 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -438,100 +438,55 @@ void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { // Micro VU - Custom Quick Search //------------------------------------------------------------------ -#ifndef __LINUX__ +PCSX2_ALIGNED(0x1000, static u8 mVUsearchXMM[0x1000]); + // Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) -static __declspec(naked) u32 __fastcall mVUsearchXMM(void *dest, void *src) { +void mVUemitSearch() { + using namespace x86Emitter; + HostSys::MemProtect(mVUsearchXMM, 0x1000, Protect_ReadWrite, false); + memset_8<0xcc,0x1000>(mVUsearchXMM); + xSetPtr(mVUsearchXMM); - __asm { - movaps xmm0, [ecx] - pcmpeqd xmm0, [edx] - movaps xmm1, [ecx + 0x10] - pcmpeqd xmm1, [edx + 0x10] - pand xmm0, xmm1 + xMOVAPS (xmm0, ptr32[ecx]); + xPCMP.EQD(xmm0, ptr32[edx]); + xMOVAPS (xmm1, ptr32[ecx + 0x10]); + xPCMP.EQD(xmm1, ptr32[edx + 0x10]); + xPAND (xmm0, xmm1); - movmskps eax, xmm0 - cmp eax, 0xf - jl exitPoint + xMOVMSKPS(eax, xmm0); + xCMP (eax, 0xf); + xForwardJL8 exitPoint; - movaps xmm0, [ecx + 0x20] - pcmpeqd xmm0, [edx + 0x20] - movaps xmm1, [ecx + 0x30] - pcmpeqd xmm1, [edx + 0x30] - pand xmm0, xmm1 + xMOVAPS (xmm0, ptr32[ecx + 0x20]); + xPCMP.EQD(xmm0, ptr32[edx + 0x20]); + xMOVAPS (xmm1, ptr32[ecx + 0x30]); + xPCMP.EQD(xmm1, ptr32[edx + 0x30]); + xPAND (xmm0, xmm1); - movaps xmm2, [ecx + 0x40] - pcmpeqd xmm2, [edx + 0x40] - movaps xmm3, [ecx + 0x50] - pcmpeqd xmm3, [edx + 0x50] - pand xmm2, xmm3 + xMOVAPS (xmm2, ptr32[ecx + 0x40]); + xPCMP.EQD(xmm2, ptr32[edx + 0x40]); + xMOVAPS (xmm3, ptr32[ecx + 0x50]); + xPCMP.EQD(xmm3, ptr32[edx + 0x50]); + xPAND (xmm2, xmm3); - movaps xmm4, [ecx + 0x60] - pcmpeqd xmm4, [edx + 0x60] - movaps xmm5, [ecx + 0x70] - pcmpeqd xmm5, [edx + 0x70] - pand xmm4, xmm5 + xMOVAPS (xmm4, ptr32[ecx + 0x60]); + xPCMP.EQD(xmm4, ptr32[edx + 0x60]); + xMOVAPS (xmm5, ptr32[ecx + 0x70]); + xPCMP.EQD(xmm5, ptr32[edx + 0x70]); + xPAND (xmm4, xmm5); - movaps xmm6, [ecx + 0x80] - pcmpeqd xmm6, [edx + 0x80] - movaps xmm7, [ecx + 0x90] - pcmpeqd xmm7, [edx + 0x90] - pand xmm6, xmm7 + xMOVAPS (xmm6, ptr32[ecx + 0x80]); + xPCMP.EQD(xmm6, ptr32[edx + 0x80]); + xMOVAPS (xmm7, ptr32[ecx + 0x90]); + xPCMP.EQD(xmm7, ptr32[edx + 0x90]); + xPAND (xmm6, xmm7); - pand xmm0, xmm2 - pand xmm4, xmm6 - pand xmm0, xmm4 - movmskps eax, xmm0 -exitPoint: - ret - } + xPAND (xmm0, xmm2); + xPAND (xmm4, xmm6); + xPAND (xmm0, xmm4); + xMOVMSKPS(eax, xmm0); + + exitPoint.SetTarget(); + xRET(); + HostSys::MemProtect(mVUsearchXMM, 0x1000, Protect_ReadOnly, true ); } -#else -// Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) -static u32 __fastcall mVUsearchXMM(void *dest, void *src) -{ - __asm__ - ( - ".intel_syntax noprefix\n" - "movaps xmm0, [ecx]\n" - "pcmpeqd xmm0, [edx]\n" - "movaps xmm1, [ecx + 0x10]\n" - "pcmpeqd xmm1, [edx + 0x10]\n" - "pand xmm0, xmm1\n" - - "movmskps eax, xmm0\n" - "cmp eax, 0xf\n" - "jl exitPoint\n" - - "movaps xmm0, [ecx + 0x20]\n" - "pcmpeqd xmm0, [edx + 0x20]\n" - "movaps xmm1, [ecx + 0x30]\n" - "pcmpeqd xmm1, [edx + 0x30]\n" - "pand xmm0, xmm1\n" - - "movaps xmm2, [ecx + 0x40]\n" - "pcmpeqd xmm2, [edx + 0x40]\n" - "movaps xmm3, [ecx + 0x50]\n" - "pcmpeqd xmm3, [edx + 0x50]\n" - "pand xmm2, xmm3\n" - - "movaps xmm4, [ecx + 0x60]\n" - "pcmpeqd xmm4, [edx + 0x60]\n" - "movaps xmm5, [ecx + 0x70]\n" - "pcmpeqd xmm5, [edx + 0x70]\n" - "pand xmm4, xmm5\n" - - "movaps xmm6, [ecx + 0x80]\n" - "pcmpeqd xmm6, [edx + 0x80]\n" - "movaps xmm7, [ecx + 0x90]\n" - "pcmpeqd xmm7, [edx + 0x90]\n" - "pand xmm6, xmm7\n" - - "pand xmm0, xmm2\n" - "pand xmm4, xmm6\n" - "pand xmm0, xmm4\n" - "movmskps eax, xmm0\n" -"exitPoint:\n" - ".att_syntax\n" - ); -} -#endif From 67b2d2bfc2a36f1305b36357eb9753b251909c64 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 7 Jun 2009 21:08:21 +0000 Subject: [PATCH 205/296] Linux: Not sure if this is neccessarily an alignment issue, but Kingdom Hearts 1 crashes on newgame with the new compare function, so we'll have it use memcmp for the moment. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1348 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Misc.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 682c5407c6..2c97c3f820 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -163,11 +163,14 @@ declareAllVariables // Define mVUquickSearch extern u8 mVUsearchXMM[0x1000]; typedef u32 (__fastcall *mVUCall)(void*, void*); +#ifndef __LINUX__ #define mVUquickSearch(dest, src, size) ((((mVUCall)((void*)mVUsearchXMM))(dest, src)) == 0xf) // Note: If GCC builds crash with above function, it means // that they're not guaranteeing 16-byte alignment on the structs // being compared. So use this function instead: -// #define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +#else +#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +#endif // Misc Macros... #define mVUprogI mVU->prog.prog[progIndex] From e538ecd7d8df00ae651d6bcdeba5375e84014834 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Mon, 8 Jun 2009 16:17:08 +0000 Subject: [PATCH 206/296] * Fixed CDVD startup code sot hat the CRC for games is more reliably detected when booting through Run->Execute. * Minor bugfix to MTGS thread startup code; wasn't copying the register state properly (probably didn't matter either way) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1349 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/CDVD/CDVD.cpp | 18 +++++++++++++++--- pcsx2/MTGS.cpp | 2 +- pcsx2/Plugins.cpp | 8 ++++---- 3 files changed, 20 insertions(+), 8 deletions(-) diff --git a/pcsx2/CDVD/CDVD.cpp b/pcsx2/CDVD/CDVD.cpp index 675a070ec7..3bd4b07bc4 100644 --- a/pcsx2/CDVD/CDVD.cpp +++ b/pcsx2/CDVD/CDVD.cpp @@ -402,7 +402,7 @@ void cdvdReadKey(u8 arg0, u16 arg1, u32 arg2, u8* key) { s32 cdvdGetToc(void* toc) { s32 ret = CDVDgetTOC(toc); - if (ret == -1) ret = 0x80; + if (ret == -1) ret = 0x80; return ret; } @@ -569,13 +569,25 @@ void cdvdNewDiskCB() { cdvd.Type = CDVDgetDiskType(); + char str[g_MaxPath]; + int result = GetPS2ElfName(str); + if (cdvd.Type == CDVD_TYPE_PS2CD) { - char str[g_MaxPath]; // Does the SYSTEM.CNF file only say "BOOT="? PS1 CD then. - if(GetPS2ElfName(str) == 1) cdvd.Type = CDVD_TYPE_PSCD; + if(result == 1) cdvd.Type = CDVD_TYPE_PSCD; } + + // Now's a good time to reload the ELF info... + if( ElfCRC == 0 ) + { + ElfCRC = loadElfCRC( str ); + ElfApplyPatches(); + LoadGameSpecificSettings(); + GSsetGameCRC( ElfCRC, 0 ); + } + } void mechaDecryptBytes( u32 madr, int size ) diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 0cd782c5c0..55195caf5e 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -511,7 +511,7 @@ int mtgsThreadObject::Callback() { Console::WriteLn("MTGS > Thread Started, Opening GS Plugin..."); - memcpy_aligned( m_gsMem, PS2MEM_GS, sizeof(m_gsMem) ); + memcpy_aligned( m_gsMem, PS2MEM_GS, sizeof(PS2MEM_GS) ); GSsetBaseMem( m_gsMem ); GSirqCallback( NULL ); diff --git a/pcsx2/Plugins.cpp b/pcsx2/Plugins.cpp index dcd9709a4b..0c78f6cc41 100644 --- a/pcsx2/Plugins.cpp +++ b/pcsx2/Plugins.cpp @@ -737,14 +737,11 @@ int OpenPlugins(const char* pTitleFilename) if (ret != 0) { if (g_Startup.BootMode != BootMode_Elf) { Msgbox::Alert("Error Opening CDVD Plugin"); goto OpenError; } - else { Console::Notice("Running ELF File Without CDVD Plugin Support!"); cdvdElf = 1; goto skipOpenCDVD; } + else { Console::Notice("Running ELF File Without CDVD Plugin Support!"); cdvdElf = 1; } } OpenStatus.CDVD = true; - cdvdNewDiskCB(); } -skipOpenCDVD: - if( !OpenStatus.GS ) { ret = gsOpen(); if (ret != 0) { Msgbox::Alert("Error Opening GS Plugin"); goto OpenError; } @@ -816,6 +813,9 @@ skipOpenCDVD: OpenStatus.FW = true; } + if( !cdvdElf ) + cdvdNewDiskCB(); + #ifndef _WIN32 //chdir(MAIN_DIR); #endif From 70cd55b332dbd792d99c7eb1290a14366b80730d Mon Sep 17 00:00:00 2001 From: arcum42 Date: Mon, 8 Jun 2009 21:24:18 +0000 Subject: [PATCH 207/296] Linux: Use memcmp_mmx instead of memcmp; Add a workaround to a segfault in iVif on some games. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1350 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/iVif.cpp | 77 +++++++++++++++++----------------------- pcsx2/x86/microVU_Misc.h | 3 +- 2 files changed, 35 insertions(+), 45 deletions(-) diff --git a/pcsx2/x86/iVif.cpp b/pcsx2/x86/iVif.cpp index c068ecd8e1..a55fea26c4 100644 --- a/pcsx2/x86/iVif.cpp +++ b/pcsx2/x86/iVif.cpp @@ -25,8 +25,6 @@ #include #include -//#define USE_OLD_IVIF_CODE - // sse2 highly optimized vif (~200 separate functions are built) zerofrog(@gmail.com) extern u32 g_vif1Masks[48], g_vif0Masks[48]; extern u32 g_vif1HasMask3[4], g_vif0HasMask3[4]; @@ -55,7 +53,10 @@ extern u8 s_maskwrite[256]; extern "C" PCSX2_ALIGNED16(u32 s_TempDecompress[4]) = {0}; -#if defined(_MSC_VER) || !defined(USE_OLD_IVIF_CODE) +#ifdef __LINUX__ +static void __forceinline UseOldMaskCode(u32* &vif1masks, u32 &mask); +#endif + void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) { u32 i; @@ -66,7 +67,13 @@ void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) prev |= s_maskwrite[mask&0xff]; hasmask[i] = prev; - if( (mask&0xff) != (oldmask&0xff) ) { + if ((mask&0xff) != (oldmask&0xff)) +#ifdef __LINUX__ + if (mask == 0) // Temporary workaround for a bug causing a segfault. + UseOldMaskCode(vif1masks, mask); + else +#endif + { __m128i r0, r1, r2, r3; r0 = _mm_load_si128((__m128i*)&s_maskarr[mask&15][0]); r2 = _mm_unpackhi_epi16(r0, r0); @@ -90,47 +97,29 @@ void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) FreezeXMMRegs(0); } -#else // gcc -// After some experimentation, I'm putting the old code back in for now for testing purposes, as the -// other version reliably SegFaults when loading YuGiOh: Duelist of the Roses on Linux (when setting -// r0 to _mm_load_si128, when vif1masks=0x846f670, hasmask=0x846f454, mask=0, and oldmask=5). -// Seems to work everywhere else. It'll stay disabled for now, but it's easier for me to fiddle with if its in here. -arcum42 -void __fastcall SetNewMask(u32* vif1masks, u32* hasmask, u32 mask, u32 oldmask) +#ifdef __LINUX__ +static void __forceinline UseOldMaskCode(u32* &vif1masks, u32 &mask) { - u32 i; - u32 prev = 0; - FreezeXMMRegs(1); + u8* p0 = (u8*)&s_maskarr[mask&15][0]; + u8* p1 = (u8*)&s_maskarr[(mask>>4)&15][0]; - for(i = 0; i < 4; ++i, mask >>= 8, oldmask >>= 8, vif1masks += 16) { - - prev |= s_maskwrite[mask&0xff];//((mask&3)==3)||((mask&0xc)==0xc)||((mask&0x30)==0x30)||((mask&0xc0)==0xc0); - hasmask[i] = prev; - - if( (mask&0xff) != (oldmask&0xff) ) { - u8* p0 = (u8*)&s_maskarr[mask&15][0]; - u8* p1 = (u8*)&s_maskarr[(mask>>4)&15][0]; - - __asm__(".intel_syntax noprefix\n" - "movaps xmm0, [%0]\n" - "movaps xmm1, [%1]\n" - "movaps xmm2, xmm0\n" - "punpcklwd xmm0, xmm0\n" - "punpckhwd xmm2, xmm2\n" - "movaps xmm3, xmm1\n" - "punpcklwd xmm1, xmm1\n" - "punpckhwd xmm3, xmm3\n" - "movq [%2], xmm0\n" - "movq [%2+8], xmm1\n" - "movhps [%2+16], xmm0\n" - "movhps [%2+24], xmm1\n" - "movq [%2+32], xmm2\n" - "movq [%2+40], xmm3\n" - "movhps [%2+48], xmm2\n" - "movhps [%2+56], xmm3\n" - ".att_syntax\n" : : "r"(p0), "r"(p1), "r"(vif1masks) ); - } - } - FreezeXMMRegs(0); + __asm__(".intel_syntax noprefix\n" + "movaps xmm0, [%0]\n" + "movaps xmm1, [%1]\n" + "movaps xmm2, xmm0\n" + "punpcklwd xmm0, xmm0\n" + "punpckhwd xmm2, xmm2\n" + "movaps xmm3, xmm1\n" + "punpcklwd xmm1, xmm1\n" + "punpckhwd xmm3, xmm3\n" + "movq [%2], xmm0\n" + "movq [%2+8], xmm1\n" + "movhps [%2+16], xmm0\n" + "movhps [%2+24], xmm1\n" + "movq [%2+32], xmm2\n" + "movq [%2+40], xmm3\n" + "movhps [%2+48], xmm2\n" + "movhps [%2+56], xmm3\n" + ".att_syntax\n" : : "r"(p0), "r"(p1), "r"(vif1masks) ); } - #endif diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 2c97c3f820..abe3b9d12b 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -169,7 +169,8 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); // that they're not guaranteeing 16-byte alignment on the structs // being compared. So use this function instead: #else -#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +//#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +#define mVUquickSearch(dest, src, size) (!memcmp_mmx(dest, src, size)) #endif // Misc Macros... From b678bb129c1073b645e116a1f930a8e62c634eea Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 9 Jun 2009 10:13:28 +0000 Subject: [PATCH 208/296] GSdx: texture filtering checkbox fixed, changed texture lookup a bit, removed more unused code git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1351 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDirtyRect.cpp | 4 +- plugins/GSdx/GSLocalMemory.cpp | 336 +----------------------------- plugins/GSdx/GSLocalMemory.h | 9 +- plugins/GSdx/GSRendererHW10.cpp | 12 +- plugins/GSdx/GSRendererHW9.cpp | 12 +- plugins/GSdx/GSRendererSW.cpp | 6 +- plugins/GSdx/GSTextureCache.cpp | 32 +-- plugins/GSdx/GSTextureCache.h | 3 +- plugins/GSdx/GSTextureCache10.cpp | 9 +- plugins/GSdx/GSTextureCache9.cpp | 9 +- plugins/GSdx/GSTextureCacheSW.cpp | 3 +- plugins/GSdx/GSTextureFX10.cpp | 10 +- plugins/GSdx/GSTextureFX10.h | 51 ++--- plugins/GSdx/GSTextureFX9.cpp | 16 +- plugins/GSdx/GSTextureFX9.h | 21 +- plugins/GSdx/GSdx_vs2008.vcproj | 8 + plugins/GSdx/res/tfx10.fx | 32 +-- plugins/GSdx/res/tfx9.fx | 43 ++-- 18 files changed, 145 insertions(+), 471 deletions(-) diff --git a/plugins/GSdx/GSDirtyRect.cpp b/plugins/GSdx/GSDirtyRect.cpp index 4276d0d47e..422fad5f22 100644 --- a/plugins/GSdx/GSDirtyRect.cpp +++ b/plugins/GSdx/GSDirtyRect.cpp @@ -75,7 +75,9 @@ GSVector4i GSDirtyRectList::GetDirtyRectAndClear(const GIFRegTEX0& TEX0, const G clear(); - return r.rintersect(GSVector4i(0, 0, size.x, size.y)); + GSVector2i bs = GSLocalMemory::m_psm[TEX0.PSM].bs; + + return r.ralign(bs).rintersect(GSVector4i(0, 0, size.x, size.y)); } return GSVector4i::zero(); diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 004aae25ff..a13bdef939 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -1720,26 +1720,7 @@ void GSLocalMemory::ReadTextureBlock16SZ(uint32 bp, uint8* dst, int dstpitch, co /////////////////// -void GSLocalMemory::ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) -{ - readTexture rtx = m_psm[TEX0.PSM].rtx; - readTexel rt = m_psm[TEX0.PSM].rt; - GSVector2i bs = m_psm[TEX0.PSM].bs; - - if(r.width() < bs.x || r.height() < bs.y - || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) - || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1)) - || CLAMP.WMS == 3 || CLAMP.WMT == 3) - { - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); - } - else - { - (this->*rtx)(r, dst, dstpitch, TEX0, TEXA); - } -} - -void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) { readTexture rtx = m_psm[TEX0.PSM].rtx; readTexel rt = m_psm[TEX0.PSM].rt; @@ -1749,7 +1730,7 @@ void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1))) { - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, rt, rtx); } else { @@ -1936,50 +1917,7 @@ void GSLocalMemory::ReadTexture16SZNP(const GSVector4i& r, uint8* dst, int dstpi /////////////////// -void GSLocalMemory::ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) -{ - readTexture rtx = m_psm[TEX0.PSM].rtxNP; - readTexel rt = m_psm[TEX0.PSM].rtNP; - GSVector2i bs = m_psm[TEX0.PSM].bs; - - if(r.width() < bs.x || r.height() < bs.y - || (r.left & (bs.x - 1)) || (r.top & (bs.y - 1)) - || (r.right & (bs.x - 1)) || (r.bottom & (bs.y - 1)) - || CLAMP.WMS == 3 || CLAMP.WMT == 3) - { - uint32 psm = TEX0.PSM; - - switch(psm) - { - case PSM_PSMT8: - case PSM_PSMT8H: - case PSM_PSMT4: - case PSM_PSMT4HL: - case PSM_PSMT4HH: - psm = TEX0.CPSM; - break; - } - - switch(psm) - { - default: - case PSM_PSMCT32: - case PSM_PSMCT24: - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - ReadTexture(r, dst, dstpitch, TEX0, TEXA, CLAMP, rt, rtx); - break; - } - } - else - { - (this->*rtx)(r, dst, dstpitch, TEX0, TEXA); - } -} - -void GSLocalMemory::ReadTextureNPNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP) +void GSLocalMemory::ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) { readTexture rtx = m_psm[TEX0.PSM].rtxNP; readTexel rt = m_psm[TEX0.PSM].rtNP; @@ -2007,11 +1945,11 @@ void GSLocalMemory::ReadTextureNPNC(const GSVector4i& r, uint8* dst, int dstpitc default: case PSM_PSMCT32: case PSM_PSMCT24: - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, rt, rtx); break; case PSM_PSMCT16: case PSM_PSMCT16S: - ReadTextureNC(r, dst, dstpitch, TEX0, TEXA, rt, rtx); + ReadTexture(r, dst, dstpitch, TEX0, TEXA, rt, rtx); break; } } @@ -2106,266 +2044,9 @@ void GSLocalMemory::ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, co // template -void GSLocalMemory::ReadTexture(const GSVector4i& r2, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx) +void GSLocalMemory::ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx) { - // TODO: this is a mess, make it more simple - - GSVector4i r = r2; - - uint32 wms = CLAMP.WMS, wmt = CLAMP.WMT; - uint32 minu = CLAMP.MINU, maxu = CLAMP.MAXU; - uint32 minv = CLAMP.MINV, maxv = CLAMP.MAXV; - - GSVector2i bs = m_psm[TEX0.PSM].bs; - - int bsxm = bs.x - 1; - int bsym = bs.y - 1; - - if(wms == 3 || wmt == 3) - { - if(wms == 3 && wmt == 3) - { - int w = minu + 1; - int h = minv + 1; - - w = (w + bsxm) & ~bsxm; - h = (h + bsym) & ~bsym; - - if(w % bs.x == 0 && maxu % bs.x == 0 && h % bs.y == 0 && maxv % bs.y == 0) - { -// printf("!!! 1 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - - T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - - (this->*rtx)(GSVector4i(maxu, maxv, maxu + w, maxv + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); - - dst -= r.left * sizeof(T); - -// int left = (r.left + minu) & ~minu; -// int right = r.right & ~minu; - - for(int y = r.top; y < r.bottom; y++, dst += dstpitch) - { - T* src = &buff[(y & minv) * w]; - - int x = r.left; -/* - for(; x < left; x++) - { - ((T*)dst)[x] = src[x & minu]; - } - - for(; x < right; x += minu + 1) - { - memcpy(&((T*)dst)[x], src, sizeof(T) * (minu + 1)); - } -*/ - for(; x < r.right; x++) - { - ((T*)dst)[x] = src[x & minu]; - } - } - - _aligned_free(buff); - - return; - } - } - - if(wms == 2) - { - int left = r.left; - r.left = min(r.right, max(r.left, (int)minu)); - r.right = max(r.left, min(r.right, (int)maxu + 1)); - dst += (r.left - left) * sizeof(T); - } - - if(wmt == 2) - { - int top = r.top; - r.top = min(r.bottom, max(r.top, (int)minv)); - r.bottom = max(r.top, min(r.bottom, (int)maxv + 1)); - dst += (r.top - top) * dstpitch; - } - - if(wms == 3 && wmt != 3) - { - int w = ((minu + 1) + bsxm) & ~bsxm; - - if(w % bs.x == 0 && maxu % bs.x == 0) - { -// printf("!!! 2 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - int top = r.top & ~bsym; - int bottom = (r.bottom + bsym) & ~bsym; - - int h = bottom - top; - - T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - - (this->*rtx)(GSVector4i(maxu, top, maxu + w, top + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); - - dst -= r.left * sizeof(T); - -// int left = (r.left + minu) & ~minu; -// int right = r.right & ~minu; - - for(int y = r.top; y < r.bottom; y++, dst += dstpitch) - { - T* src = &buff[(y - top) * w]; - - int x = r.left; -/* - for(; x < left; x++) - { - ((T*)dst)[x] = src[x & minu]; - } - - for(; x < right; x += minu + 1) - { - memcpy(&((T*)dst)[x], src, sizeof(T) * (minu + 1)); - } -*/ - for(; x < r.right; x++) - { - ((T*)dst)[x] = src[x & minu]; - } - } - - _aligned_free(buff); - - return; - } - } - - if(wms != 3 && wmt == 3) - { - int h = (minv + 1 + bsym) & ~bsym; - - if(h % bs.y == 0 && maxv % bs.y == 0) - { -// printf("!!! 3 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - int left = r.left & ~bsxm; - int right = (r.right + bsxm) & ~bsxm; - - int w = right - left; - - T* buff = (T*)_aligned_malloc(w * h * sizeof(T), 16); - - (this->*rtx)(GSVector4i(left, maxv, left + w, maxv + h), (uint8*)buff, w * sizeof(T), TEX0, TEXA); - - for(int y = r.top; y < r.bottom; y++, dst += dstpitch) - { - T* src = &buff[(y & minv) * w + (r.left - left)]; - - memcpy(dst, src, sizeof(T) * r.width()); - } - - _aligned_free(buff); - - return; - } - } - - switch(wms) - { - default: for(int x = r.left; x < r.right; x++) m_xtbl[x] = x; break; - case 3: for(int x = r.left; x < r.right; x++) m_xtbl[x] = (x & minu) | maxu; break; - } - - switch(wmt) - { - default: for(int y = r.top; y < r.bottom; y++) m_ytbl[y] = y; break; - case 3: for(int y = r.top; y < r.bottom; y++) m_ytbl[y] = (y & minv) | maxv; break; - } - -// printf("!!! 4 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - - for(int y = r.top; y < r.bottom; y++, dst += dstpitch) - for(int x = r.left, i = 0; x < r.right; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(m_xtbl[x], m_ytbl[y], TEX0, TEXA); - } - else - { - // find a block-aligned rect that fits between r and the region clamped area (if any) - - GSVector4i r1 = r; - GSVector4i r2 = r; - - r1.left = (r1.left + bsxm) & ~bsxm; - r1.top = (r1.top + bsym) & ~bsym; - r1.right = r1.right & ~bsxm; - r1.bottom = r1.bottom & ~bsym; - - if(wms == 2 && minu < maxu) - { - r2.left = minu & ~bsxm; - r2.right = (maxu + bsxm) & ~bsxm; - } - - if(wmt == 2 && minv < maxv) - { - r2.top = minv & ~bsym; - r2.bottom = (maxv + bsym) & ~bsym; - } - - GSVector4i cr = r1.rintersect(r2); - - bool aligned = ((DWORD_PTR)(dst + (cr.left - r.left) * sizeof(T)) & 0xf) == 0; - - if(cr.rempty() || !aligned) - { - // TODO: expand r to block size, read into temp buffer, copy to r (like above) - -if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); - -// printf("!!! 5 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - - for(int y = r.top; y < r.bottom; y++, dst += dstpitch) - for(int x = r.left, i = 0; x < r.right; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - } - else - { -// printf("!!! 6 wms = %d, wmt = %d, %3x %3x %3x %3x, %d %d - %d %d\n", wms, wmt, minu, maxu, minv, maxv, r.left, r.top, r.right, r.bottom); - - for(int y = r.top; y < cr.top; y++, dst += dstpitch) - for(int x = r.left, i = 0; x < r.right; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - - if(!cr.rempty()) - { - (this->*rtx)(cr, dst + (cr.left - r.left) * sizeof(T), dstpitch, TEX0, TEXA); - } - - for(int y = cr.top; y < cr.bottom; y++, dst += dstpitch) - { - for(int x = r.left, i = 0; x < cr.left; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - for(int x = cr.right, i = x - r.left; x < r.right; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - } - - for(int y = cr.bottom; y < r.bottom; y++, dst += dstpitch) - for(int x = r.left, i = 0; x < r.right; x++, i++) - ((T*)dst)[i] = (T)(this->*rt)(x, y, TEX0, TEXA); - } - } -} - -template -void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx) -{ - GSVector2i bs = m_psm[TEX0.PSM].bs; - - int bsxm = bs.x - 1; - int bsym = bs.y - 1; - - GSVector4i cr; - - cr.left = (r.left + bsxm) & ~bsxm; - cr.top = (r.top + bsym) & ~bsym; - cr.right = r.right & ~bsxm; - cr.bottom = r.bottom & ~bsym; + GSVector4i cr = r.ralign(m_psm[TEX0.PSM].bs); bool aligned = ((DWORD_PTR)(dst + (cr.left - r.left) * sizeof(T)) & 0xf) == 0; @@ -2373,7 +2054,8 @@ void GSLocalMemory::ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, { // TODO: expand r to block size, read into temp buffer, copy to r (like above) -if(!aligned) printf("unaligned memory pointer passed to ReadTexture\n"); +if(!aligned) +printf("unaligned memory pointer passed to ReadTexture\n"); for(int y = r.top; y < r.bottom; y++, dst += dstpitch) for(int x = r.left, i = 0; x < r.right; x++, i++) diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index 1a966ae603..f690fafa2a 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -795,8 +795,7 @@ public: void ReadTexture16Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture16SZ(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - void ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA); void ReadTextureBlock32(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; void ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const; @@ -824,8 +823,7 @@ public: void ReadTexture16ZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; void ReadTexture16SZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const; - void ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); - void ReadTextureNPNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP); + void ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA); // pal ? 8 : 32 @@ -845,8 +843,7 @@ public: static uint32 m_xtbl[1024], m_ytbl[1024]; - template void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GIFRegCLAMP& CLAMP, readTexel rt, readTexture rtx); - template void ReadTextureNC(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx); + template void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx); // diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 69bf582df4..4423d582e5 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -324,28 +324,27 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; ps_sel.fba = context->FBA.FBA; ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; + ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; GSTextureFX10::PSSamplerSelector ps_ssel; - ps_ssel.min = m_filter == 2 ? (context->TEX1.MMIN & 1) : m_filter; - ps_ssel.mag = m_filter == 2 ? (context->TEX1.MMAG & 1) : m_filter; ps_ssel.tau = 0; ps_ssel.tav = 0; + ps_ssel.ltf = ps_sel.ltf; GSTextureFX10::PSConstantBuffer ps_cb; - ps_cb.FogColor = GSVector4(env.FOGCOL.FCR, env.FOGCOL.FCG, env.FOGCOL.FCB, 0) / 255.0f; + ps_cb.FogColorAREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; ps_cb.TA0 = (float)(int)env.TEXA.TA0 / 255; ps_cb.TA1 = (float)(int)env.TEXA.TA1 / 255; - ps_cb.AREF = (float)(int)context->TEST.AREF / 255; if(context->TEST.ATST == 2 || context->TEST.ATST == 5) { - ps_cb.AREF -= 0.9f/256; + ps_cb.FogColorAREF.a -= 0.9f / 255; } else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) { - ps_cb.AREF += 0.9f/256; + ps_cb.FogColorAREF.a += 0.9f / 255; } if(tex) @@ -400,7 +399,6 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache float h = (float)tex->m_texture->GetHeight(); ps_cb.WH = GSVector2(w, h); - ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 3b90ca47a8..b4b0877352 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -298,28 +298,27 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_sel.fog = PRIM->FGE; ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; ps_sel.rt = tex && tex->m_rendered; + ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; GSTextureFX9::PSSamplerSelector ps_ssel; - ps_ssel.min = m_filter == 2 ? (context->TEX1.MMIN & 1) : m_filter; - ps_ssel.mag = m_filter == 2 ? (context->TEX1.MMAG & 1) : m_filter; ps_ssel.tau = 0; ps_ssel.tav = 0; + ps_ssel.ltf = ps_sel.ltf; GSTextureFX9::PSConstantBuffer ps_cb; - ps_cb.FogColor = GSVector4(env.FOGCOL.FCR, env.FOGCOL.FCG, env.FOGCOL.FCB, 0) / 255.0f; + ps_cb.FogColorAREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; ps_cb.TA0 = (float)(int)env.TEXA.TA0 / 255; ps_cb.TA1 = (float)(int)env.TEXA.TA1 / 255; - ps_cb.AREF = (float)(int)context->TEST.AREF / 255; if(context->TEST.ATST == 2 || context->TEST.ATST == 5) { - ps_cb.AREF -= 0.9f/256; + ps_cb.FogColorAREF.a -= 0.9f / 255; } else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) { - ps_cb.AREF += 0.9f/256; + ps_cb.FogColorAREF.a += 0.9f / 255; } if(tex) @@ -374,7 +373,6 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: float h = (float)tex->m_texture->GetHeight(); ps_cb.WH = GSVector2(w, h); - ps_cb.rWrH = GSVector2(1.0f / w, 1.0f / h); ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index 2b78701155..d9e0d732b6 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -114,15 +114,11 @@ GSTexture* GSRendererSW::GetOutput(int i) } } - GIFRegCLAMP CLAMP; - - CLAMP.WMS = CLAMP.WMT = 1; - // TODO static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); static int pitch = 1024 * 4; - m_mem.ReadTexture(r, buff, pitch, TEX0, m_env.TEXA, CLAMP); + m_mem.ReadTexture(r, buff, pitch, TEX0, m_env.TEXA); m_texture[i]->Update(r, buff, pitch); diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 43a6587019..07d10931f8 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -260,21 +260,21 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) { - t = *i; + GSCachedTexture* t2 = *i; - if(GSUtil::HasSharedBits(t->m_TEX0.TBP0, t->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + if((((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) & 0xffefffff) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW (PSM & ~1) TW TH { - if(TEX0.PSM == t->m_TEX0.PSM && TEX0.TBW == t->m_TEX0.TBW - && TEX0.TW == t->m_TEX0.TW && TEX0.TH == t->m_TEX0.TH - && (pal == 0 || TEX0.CPSM == t->m_TEX0.CPSM && GSVector4i::compare(t->m_clut, clut, pal * sizeof(clut[0])))) - { - m_tex.splice(m_tex.begin(), m_tex, i); - - break; - } + continue; } - t = NULL; + if(!(pal == 0 || t2->m_TEX0.CPSM == TEX0.CPSM && GSVector4i::compare(t2->m_clut, clut, pal * sizeof(clut[0])))) + { + continue; + } + + t = t2; + + m_tex.splice(m_tex.begin(), m_tex, i); } if(t == NULL) @@ -367,6 +367,8 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() t->Update(); + m_tex_used = true; + return t; } @@ -595,9 +597,11 @@ void GSTextureCache::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const void GSTextureCache::IncAge() { - RecycleByAge(m_tex, 2); + RecycleByAge(m_tex, m_tex_used ? 2 : 30); RecycleByAge(m_rt); RecycleByAge(m_ds); + + m_tex_used = false; } // GSTextureCache::GSSurface @@ -704,7 +708,7 @@ void GSTextureCache::GSCachedTexture::Update() { // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource - m_renderer->m_mem.ReadTextureNPNC(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); m_texture->Unmap(); } @@ -714,7 +718,7 @@ void GSTextureCache::GSCachedTexture::Update() pitch = ((r.width() + 3) & ~3) * 4; - m_renderer->m_mem.ReadTextureNPNC(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA, m_renderer->m_context->CLAMP); + m_renderer->m_mem.ReadTextureNP(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); m_texture->Update(r, buff, pitch); } diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 66daec1834..34bef246ca 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -72,7 +72,6 @@ public: bool GetDirtyRect(GSVector4i& r); public: - GIFRegCLAMP m_CLAMP; uint32* m_clut; // * GSVector4i m_valid; int m_bpp; @@ -96,6 +95,8 @@ protected: list m_ds; list m_tex; + bool m_tex_used; + template void RecycleByAge(list& l, int maxage = 60) { for(list::iterator i = l.begin(); i != l.end(); ) diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index ebc31259f3..e3c902d1e1 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -55,12 +55,7 @@ void GSTextureCache10::GSRenderTargetHW10::Update() TEXA.TA0 = 0; TEXA.TA1 = 0x80; - GIFRegCLAMP CLAMP; - - CLAMP.WMS = 0; - CLAMP.WMT = 0; - - m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA, CLAMP); + m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA); // s->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); @@ -182,7 +177,6 @@ bool GSTextureCache10::GSCachedTextureHW10::Create() // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); m_TEX0 = m_renderer->m_context->TEX0; - m_CLAMP = m_renderer->m_context->CLAMP; uint32 psm = m_TEX0.PSM; @@ -240,7 +234,6 @@ bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); m_TEX0 = m_renderer->m_context->TEX0; - m_CLAMP = m_renderer->m_context->CLAMP; m_rendered = true; int tw = 1 << m_TEX0.TW; diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index 2b98621de5..3c65349fdf 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -57,12 +57,7 @@ void GSTextureCache9::GSRenderTarget9::Update() TEXA.TA0 = 0; TEXA.TA1 = 0x80; - GIFRegCLAMP CLAMP; - - CLAMP.WMS = 0; - CLAMP.WMT = 0; - - m_renderer->m_mem.ReadTexture(r, bits, pitch, m_TEX0, TEXA, CLAMP); + m_renderer->m_mem.ReadTexture(r, bits, pitch, m_TEX0, TEXA); t->Unmap(); @@ -181,7 +176,6 @@ bool GSTextureCache9::GSCachedTexture9::Create() // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); m_TEX0 = m_renderer->m_context->TEX0; - m_CLAMP = m_renderer->m_context->CLAMP; uint32 psm = m_TEX0.PSM; @@ -239,7 +233,6 @@ bool GSTextureCache9::GSCachedTexture9::Create(GSRenderTarget* rt) // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); m_TEX0 = m_renderer->m_context->TEX0; - m_CLAMP = m_renderer->m_context->CLAMP; m_rendered = true; int tw = 1 << m_TEX0.TW; diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index 524521e5ec..ca7144a438 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -46,8 +46,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE { GSTexture* t2 = (*i).first; - // if(t2->m_TEX0.TBP0 != TEX0.TBP0 || t2->m_TEX0.TBW != TEX0.TBW || t2->m_TEX0.PSM != TEX0.PSM || t2->m_TEX0.TW != TEX0.TW || t2->m_TEX0.TH != TEX0.TH) - if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) + if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH { continue; } diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 0dec170dcc..60d79f75dc 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -213,7 +213,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(i == m_ps.end()) { - string str[13]; + string str[14]; str[0] = format("%d", sel.fst); str[1] = format("%d", sel.wms); @@ -228,6 +228,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl str[10] = format("%d", sel.clr1); str[11] = format("%d", sel.fba); str[12] = format("%d", sel.aout); + str[13] = format("%d", sel.ltf); D3D10_SHADER_MACRO macro[] = { @@ -244,6 +245,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl {"CLR1", str[10].c_str()}, {"FBA", str[11].c_str()}, {"AOUT", str[12].c_str()}, + {"LTF", str[13].c_str()}, {NULL, NULL}, }; @@ -280,11 +282,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl memset(&sd, 0, sizeof(sd)); - sd.Filter = D3D10_ENCODE_BASIC_FILTER( - (ssel.min ? D3D10_FILTER_TYPE_LINEAR : D3D10_FILTER_TYPE_POINT), - (ssel.mag ? D3D10_FILTER_TYPE_LINEAR : D3D10_FILTER_TYPE_POINT), - D3D10_FILTER_TYPE_POINT, - false); + sd.Filter = ssel.ltf ? D3D10_FILTER_MIN_MAG_LINEAR_MIP_POINT : D3D10_FILTER_MIN_MAG_MIP_POINT; sd.AddressU = ssel.tau ? D3D10_TEXTURE_ADDRESS_WRAP : D3D10_TEXTURE_ADDRESS_CLAMP; sd.AddressV = ssel.tav ? D3D10_TEXTURE_ADDRESS_WRAP : D3D10_TEXTURE_ADDRESS_CLAMP; diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index cc8653ca7e..3377acd1fd 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -77,7 +77,11 @@ public: __declspec(align(16)) struct PSConstantBuffer { - GSVector4 FogColor; + GSVector4 FogColorAREF; + GSVector4 HalfTexel; + GSVector2 WH; + float TA0; + float TA1; float MINU; float MINV; float MAXU; @@ -86,13 +90,6 @@ public: uint32 VMSK; uint32 UFIX; uint32 VFIX; - float TA0; - float TA1; - float AREF; - float _pad[1]; - GSVector2 WH; - GSVector2 rWrH; - GSVector4 HalfTexel; struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} @@ -106,16 +103,14 @@ public: GSVector4i b2 = b[2]; GSVector4i b3 = b[3]; GSVector4i b4 = b[4]; - GSVector4i b5 = b[5]; - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4)).alltrue()) { a[0] = b0; a[1] = b1; a[2] = b2; a[3] = b3; a[4] = b4; - a[5] = b5; return true; } @@ -124,6 +119,19 @@ public: } }; + union GSSelector + { + struct + { + uint32 iip:1; + uint32 prim:2; + }; + + uint32 key; + + operator uint32() {return key & 0x7;} + }; + union PSSelector { struct @@ -141,24 +149,12 @@ public: uint32 clr1:1; uint32 fba:1; uint32 aout:1; + uint32 ltf:1; }; uint32 key; - operator uint32() {return key & 0x1fffff;} - }; - - union GSSelector - { - struct - { - uint32 iip:1; - uint32 prim:2; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} + operator uint32() {return key & 0x3fffff;} }; union PSSamplerSelector @@ -167,13 +163,12 @@ public: { uint32 tau:1; uint32 tav:1; - uint32 min:1; - uint32 mag:1; + uint32 ltf:1; }; uint32 key; - operator uint32() {return key & 0xf;} + operator uint32() {return key & 0x7;} }; union OMDepthStencilSelector diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 7abf2a550c..31f27b0983 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -180,7 +180,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample if(i == m_ps.end()) { - string str[12]; + string str[13]; str[0] = format("%d", sel.fst); str[1] = format("%d", sel.wms); @@ -194,6 +194,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample str[9] = format("%d", sel.fog); str[10] = format("%d", sel.clr1); str[11] = format("%d", sel.rt); + str[12] = format("%d", sel.ltf); D3DXMACRO macro[] = { @@ -209,6 +210,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample {"FOG", str[9].c_str()}, {"CLR1", str[10].c_str()}, {"RT", str[11].c_str()}, + {"LTF", str[12].c_str()}, {NULL, NULL}, }; @@ -227,10 +229,10 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample if(sel.tfx != 4) { - bool b = sel.bpp < 3 && sel.wms < 3 && sel.wmt < 3; - - ssel.min = b; - ssel.mag = b; + if(!(sel.bpp < 3 && sel.wms < 3 && sel.wmt < 3)) + { + ssel.ltf = 0; + } hash_map::const_iterator i = m_ps_ss.find(ssel); @@ -244,8 +246,8 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample memset(ss, 0, sizeof(*ss)); - ss->FilterMin[0] = ssel.min ? D3DTEXF_LINEAR : D3DTEXF_POINT; - ss->FilterMag[0] = ssel.mag ? D3DTEXF_LINEAR : D3DTEXF_POINT; + ss->FilterMin[0] = ssel.ltf ? D3DTEXF_LINEAR : D3DTEXF_POINT; + ss->FilterMag[0] = ssel.ltf ? D3DTEXF_LINEAR : D3DTEXF_POINT; ss->FilterMin[1] = D3DTEXF_POINT; ss->FilterMag[1] = D3DTEXF_POINT; diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 54fbb258ca..e3e1b5823e 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -53,7 +53,11 @@ public: struct PSConstantBuffer { - GSVector4 FogColor; + GSVector4 FogColorAREF; + GSVector4 HalfTexel; + GSVector2 WH; + float TA0; + float TA1; float MINU; float MINV; float MAXU; @@ -62,13 +66,6 @@ public: uint32 VMSK; uint32 UFIX; uint32 VFIX; - float TA0; - float TA1; - float AREF; - float _pad[1]; - GSVector2 WH; - GSVector2 rWrH; - GSVector4 HalfTexel; }; union PSSelector @@ -87,11 +84,12 @@ public: uint32 fog:1; uint32 clr1:1; uint32 rt:1; + uint32 ltf:1; }; uint32 key; - operator uint32() {return key & 0xfffff;} + operator uint32() {return key & 0x1fffff;} }; union PSSamplerSelector @@ -100,13 +98,12 @@ public: { uint32 tau:1; uint32 tav:1; - uint32 min:1; - uint32 mag:1; + uint32 ltf:1; }; uint32 key; - operator uint32() {return key & 0xf;} + operator uint32() {return key & 0x7;} }; union OMDepthStencilSelector diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index e71c4955e4..ba86150a75 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1398,6 +1398,14 @@ + + + Date: Tue, 9 Jun 2009 14:15:39 +0000 Subject: [PATCH 209/296] They are vector units, not vertex units. (Yeah, silly commit, I know :p ) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1352 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/pcsx2.rc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/windows/pcsx2.rc b/pcsx2/windows/pcsx2.rc index aeec13377b..4c2e8fc59e 100644 --- a/pcsx2/windows/pcsx2.rc +++ b/pcsx2/windows/pcsx2.rc @@ -376,7 +376,7 @@ BEGIN "Button",BS_AUTOCHECKBOX | WS_TABSTOP,187,179,139,10 LTEXT "Speedup for a few games, including FFX with no known side effects.",IDC_STATIC,200,193,144,23 GROUPBOX "EmotionEngine (EE) Misc Hacks",IDC_STATIC,177,26,179,192 - GROUPBOX "Vertex Unit (VU) Hacks",IDC_STATIC,363,26,173,173 + GROUPBOX "Vector Unit (VU) Hacks",IDC_STATIC,363,26,173,173 CONTROL "Status Flag Hack (microVU only)",IDC_VUHACK1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,42,155,12 LTEXT "Big Speedup and high compatibility.\nSuperVU does something similar by default.",IDC_STATIC,383,55,144,20 CONTROL "Min / Max Hack (microVU only)",IDC_VUHACK3,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,372,85,156,12 From a7f2130595e769a09418706d41c01d475afc035f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 10 Jun 2009 04:18:06 +0000 Subject: [PATCH 210/296] microVU: - Small Optimization... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1353 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Flags.inl | 57 ++++++++++++++++++++++++++++++------- pcsx2/x86/microVU_Misc.h | 1 + 2 files changed, 48 insertions(+), 10 deletions(-) diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 0658d78807..121038b3ed 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -62,11 +62,16 @@ int findFlagInst(int* fFlag, int cycles) { } // Setup Last 4 instances of Status/Mac/Clip flags (needed for accurate block linking) -void sortFlag(int* fFlag, int* bFlag, int cycles) { +int sortFlag(int* fFlag, int* bFlag, int cycles) { + int lFlag = -5; + int x = 0; for (int i = 0; i < 4; i++) { bFlag[i] = findFlagInst(fFlag, cycles); + if (lFlag != bFlag[i]) { x++; } + lFlag = bFlag[i]; cycles++; } + return x; // Returns the number of Valid Flag Instances } #define sFlagCond ((sFLAG.doFlag && !mVUsFlagHack) || mVUlow.isFSSET || mVUinfo.doDivFlag) @@ -151,6 +156,9 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { } #define getFlagReg1(x) ((x == 3) ? gprF3 : ((x == 2) ? gprF2 : ((x == 1) ? gprF1 : gprF0))) +#define getFlagReg2(x) ((bStatus[0] == x) ? getFlagReg1(x) : gprT1) +#define getFlagReg3(x) ((gFlag == x) ? gprT1 : getFlagReg1(x)) +#define getFlagReg4(x) ((gFlag == x) ? gprT1 : gprT2) #define shuffleMac ((bMac [3]<<6)|(bMac [2]<<4)|(bMac [1]<<2)|bMac [0]) #define shuffleClip ((bClip[3]<<6)|(bClip[2]<<4)|(bClip[1]<<2)|bClip[0]) @@ -159,15 +167,44 @@ microVUt(void) mVUsetupFlags(mV, int* xStatus, int* xMac, int* xClip, int cycles if (__Status) { int bStatus[4]; - sortFlag(xStatus, bStatus, cycles); - MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); - MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); - MOV32RtoR(gprR, getFlagReg1(bStatus[2])); - MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); - MOV32RtoR(gprF0, gprT1); - MOV32RtoR(gprF1, gprT2); - MOV32RtoR(gprF2, gprR); - MOV32ItoR(gprR, Roffset); // Restore gprR + int sortRegs = sortFlag(xStatus, bStatus, cycles); + // DevCon::Status("sortRegs = %d", params sortRegs); + // Note: Emitter will optimize out mov(reg1, reg1) cases... + // There 'is' still room for small optimizations but the + // sorting algorithm would be really complex and not really + // a noticeable improvement... (Most common cases are 1 & 2) + if (sortRegs == 1) { + MOV32RtoR(gprF0, getFlagReg1(bStatus[0])); + MOV32RtoR(gprF1, getFlagReg1(bStatus[1])); + MOV32RtoR(gprF2, getFlagReg1(bStatus[2])); + MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); + } + else if (sortRegs == 2) { + MOV32RtoR(gprT1, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, getFlagReg1(bStatus[0])); + MOV32RtoR(gprF1, getFlagReg2(bStatus[1])); + MOV32RtoR(gprF2, getFlagReg2(bStatus[2])); + MOV32RtoR(gprF3, gprT1); + } + else if (sortRegs == 3) { + int gFlag = (bStatus[0] == bStatus[1]) ? bStatus[2] : bStatus[1]; + MOV32RtoR(gprT1, getFlagReg1(gFlag)); + MOV32RtoR(gprT2, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, getFlagReg1(bStatus[0])); + MOV32RtoR(gprF1, getFlagReg3(bStatus[1])); + MOV32RtoR(gprF2, getFlagReg4(bStatus[2])); + MOV32RtoR(gprF3, gprT2); + } + else { + MOV32RtoR(gprT1, getFlagReg1(bStatus[0])); + MOV32RtoR(gprT2, getFlagReg1(bStatus[1])); + MOV32RtoR(gprR, getFlagReg1(bStatus[2])); + MOV32RtoR(gprF3, getFlagReg1(bStatus[3])); + MOV32RtoR(gprF0, gprT1); + MOV32RtoR(gprF1, gprT2); + MOV32RtoR(gprF2, gprR); + MOV32ItoR(gprR, Roffset); // Restore gprR + } } if (__Mac) { diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index abe3b9d12b..685356beca 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -66,6 +66,7 @@ declareAllVariables //------------------------------------------------------------------ // Helper Macros //------------------------------------------------------------------ + #define _Ft_ ((mVU->code >> 16) & 0x1F) // The ft part of the instruction register #define _Fs_ ((mVU->code >> 11) & 0x1F) // The fs part of the instruction register #define _Fd_ ((mVU->code >> 6) & 0x1F) // The fd part of the instruction register From 4e20305fe2adea218d0c8d50b38d3dad5db26f07 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 10 Jun 2009 17:03:23 +0000 Subject: [PATCH 211/296] GSdx: dx10 hw mode sharper again git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1354 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSRendererHW10.cpp | 37 ++++--- plugins/GSdx/GSRendererHW9.cpp | 32 +++--- plugins/GSdx/GSState.cpp | 2 +- plugins/GSdx/GSTextureFX10.cpp | 5 + plugins/GSdx/GSTextureFX10.h | 14 +-- plugins/GSdx/GSTextureFX9.cpp | 4 +- plugins/GSdx/GSTextureFX9.h | 11 +- plugins/GSdx/res/tfx10.fx | 178 ++++++++++++-------------------- plugins/GSdx/res/tfx9.fx | 150 +++++++++++---------------- 10 files changed, 182 insertions(+), 252 deletions(-) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index c0075a01d7..c8d5a44f42 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -115,6 +115,7 @@ CRC::Game CRC::m_games[] = {0xA61A4C6D, GodOfWar, Unknown}, {0xE23D532B, GodOfWar, Unknown}, {0xDF1AF973, GodOfWar, Unknown}, + {0xD6385328, GodOfWar, Unknown}, {0x2F123FD8, GodOfWar2, RU}, {0x2F123FD8, GodOfWar2, US}, {0x44A8A22A, GodOfWar2, EU}, diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 4423d582e5..f300ac44dd 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -287,8 +287,10 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); - float ox = (float)(int)context->XYOFFSET.OFX; - float oy = (float)(int)context->XYOFFSET.OFY; + float ox = (float)((int)context->XYOFFSET.OFX - 8); + float oy = (float)((int)context->XYOFFSET.OFY - 8); + + // dx10 pixel center is different => offset - 8 (half pixel) vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); vs_cb.VertexOffset = GSVector4(ox * sx + 1, -(oy * sy + 1), 0.0f, -1.0f); @@ -351,6 +353,11 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache { ps_sel.bpp = tex->m_bpp2; + int w = tex->m_texture->GetWidth(); + int h = tex->m_texture->GetHeight(); + + ps_cb.WH = GSVector2i(w, h); + switch(context->CLAMP.WMS) { case 0: @@ -360,13 +367,15 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tau = 0; break; case 2: - ps_cb.MINU = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); - ps_cb.MAXU = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); + ps_cb.MinMax.x = (int)context->CLAMP.MINU * w / (1 << context->TEX0.TW); + ps_cb.MinMax.z = (int)context->CLAMP.MAXU * w / (1 << context->TEX0.TW); + ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_ssel.tau = 0; break; case 3: - ps_cb.UMSK = context->CLAMP.MINU; - ps_cb.UFIX = context->CLAMP.MAXU; + ps_cb.MskFix.x = context->CLAMP.MINU; + ps_cb.MskFix.z = context->CLAMP.MAXU; ps_ssel.tau = 1; break; default: @@ -382,24 +391,20 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tav = 0; break; case 2: - ps_cb.MINV = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); - ps_cb.MAXV = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); + ps_cb.MinMax.y = (int)context->CLAMP.MINV * h / (1 << context->TEX0.TH); + ps_cb.MinMax.w = (int)context->CLAMP.MAXV * h / (1 << context->TEX0.TH); + ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_ssel.tav = 0; break; case 3: - ps_cb.VMSK = context->CLAMP.MINV; - ps_cb.VFIX = context->CLAMP.MAXV; + ps_cb.MskFix.y = context->CLAMP.MINV; + ps_cb.MskFix.w = context->CLAMP.MAXV; ps_ssel.tav = 1; break; default: __assume(0); } - - float w = (float)tex->m_texture->GetWidth(); - float h = (float)tex->m_texture->GetHeight(); - - ps_cb.WH = GSVector2(w, h); - ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else { diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index b4b0877352..f88143c334 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -325,6 +325,12 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: { ps_sel.bpp = tex->m_bpp2; + int w = tex->m_texture->GetWidth(); + int h = tex->m_texture->GetHeight(); + + ps_cb.WH = GSVector2(w, h); + ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); + switch(context->CLAMP.WMS) { case 0: @@ -334,13 +340,15 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_ssel.tau = 0; break; case 2: - ps_cb.MINU = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); - ps_cb.MAXU = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); + ps_cb.MinMax.x = (float)(int)context->CLAMP.MINU / (1 << context->TEX0.TW); + ps_cb.MinMax.z = (float)(int)context->CLAMP.MAXU / (1 << context->TEX0.TW); + ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_ssel.tau = 0; break; case 3: - ps_cb.UMSK = context->CLAMP.MINU; - ps_cb.UFIX = context->CLAMP.MAXU; + ps_cb.MskFix.x = context->CLAMP.MINU; + ps_cb.MskFix.z = context->CLAMP.MAXU; ps_ssel.tau = 1; break; default: @@ -356,24 +364,20 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_ssel.tav = 0; break; case 2: - ps_cb.MINV = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); - ps_cb.MAXV = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); + ps_cb.MinMax.y = (float)(int)context->CLAMP.MINV / (1 << context->TEX0.TH); + ps_cb.MinMax.w = (float)(int)context->CLAMP.MAXV / (1 << context->TEX0.TH); + ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_ssel.tav = 0; break; case 3: - ps_cb.VMSK = context->CLAMP.MINV; - ps_cb.VFIX = context->CLAMP.MAXV; + ps_cb.MskFix.y = context->CLAMP.MINV; + ps_cb.MskFix.w = context->CLAMP.MAXV; ps_ssel.tav = 1; break; default: __assume(0); } - - float w = (float)tex->m_texture->GetWidth(); - float h = (float)tex->m_texture->GetHeight(); - - ps_cb.WH = GSVector2(w, h); - ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); } else { diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 9a2d347065..1458fd7290 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -1237,7 +1237,7 @@ template void GSState::Transfer(uint8* mem, uint32 size) { m_q = 1.0f; - ASSERT(!(path.tag.PRE && path.tag.FLG == GIF_FLG_REGLIST)); + // ASSERT(!(path.tag.PRE && path.tag.FLG == GIF_FLG_REGLIST)); // kingdom hearts if(path.tag.PRE && path.tag.FLG == GIF_FLG_PACKED) { diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 60d79f75dc..b00de4e1d0 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -270,6 +270,11 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(sel.tfx != 4) { + if(!(sel.bpp < 3 && sel.wms < 3 && sel.wmt < 3)) + { + ssel.ltf = 0; + } + hash_map >::const_iterator i = m_ps_ss.find(ssel); if(i != m_ps_ss.end()) diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 3377acd1fd..14e36c03e7 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -78,18 +78,12 @@ public: __declspec(align(16)) struct PSConstantBuffer { GSVector4 FogColorAREF; - GSVector4 HalfTexel; - GSVector2 WH; + GSVector2i WH; float TA0; float TA1; - float MINU; - float MINV; - float MAXU; - float MAXV; - uint32 UMSK; - uint32 VMSK; - uint32 UFIX; - uint32 VFIX; + GSVector4i MinMax; + GSVector4 MinMaxF; + GSVector4i MskFix; struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 31f27b0983..f0ed275778 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -152,7 +152,7 @@ bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSampler { if(sel.wms == 3) { - if(GSTexture* t = CreateMskFix(tex->GetWidth(), cb->UMSK, cb->UFIX)) + if(GSTexture* t = CreateMskFix(tex->GetWidth(), cb->MskFix.x, cb->MskFix.z)) { (*m_dev)->SetTexture(2, *(GSTexture9*)t); } @@ -160,7 +160,7 @@ bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSampler if(sel.wmt == 3) { - if(GSTexture* t = CreateMskFix(tex->GetHeight(), cb->VMSK, cb->VFIX)) + if(GSTexture* t = CreateMskFix(tex->GetHeight(), cb->MskFix.y, cb->MskFix.w)) { (*m_dev)->SetTexture(3, *(GSTexture9*)t); } diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index e3e1b5823e..c657a2b06d 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -58,14 +58,9 @@ public: GSVector2 WH; float TA0; float TA1; - float MINU; - float MINV; - float MAXU; - float MAXV; - uint32 UMSK; - uint32 VMSK; - uint32 UFIX; - uint32 VFIX; + GSVector4 MinMax; + GSVector4 MinMaxF; + GSVector4i MskFix; }; union PSSelector diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 72f9688732..a16b2a8286 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -48,16 +48,7 @@ VS_OUTPUT vs_main(VS_INPUT input) VS_OUTPUT output; - float4 p = float4(input.p, input.z, 0); - - // FIXME: - // A litte accuracy problem in many games where the screen is copied in columns and the sides have a half pixel gap for some reason - // (probably to avoid page breaks with linear filtering). That half pixel coordinate gets multiplied by 2 (VertexScale) and occasionally - // ends on .9999999, which the rasterizer floors to 1 less pixel we need, leaving a visible gap after drawing. - - p += float4(0.5f, 0.5f, 0, 0); // add 1/32 pixel - - output.p = p * VertexScale - VertexOffset; + output.p = float4(input.p, input.z, 0) * VertexScale - VertexOffset; if(VS_TME == 1) { @@ -164,14 +155,12 @@ cbuffer cb1 { float3 FogColor; float AREF; - float4 HalfTexel; - float2 WH; + uint2 WH; float TA0; float TA1; - float2 MINUV; - float2 MAXUV; - uint2 UVMSK; - uint2 UVFIX; + uint4 MinMax; + float4 MinMaxF; + uint4 MskFix; }; struct PS_INPUT @@ -189,14 +178,14 @@ struct PS_OUTPUT #ifndef FST #define FST 0 -#define WMS 1 -#define WMT 1 +#define WMS 3 +#define WMT 3 #define BPP 0 #define AEM 0 #define TFX 0 #define TCC 1 #define ATE 1 -#define ATST 2 +#define ATST 4 #define FOG 1 #define CLR1 0 #define FBA 0 @@ -221,85 +210,56 @@ float4 Extract16(uint i) return f; } -int2 wrapu(float2 f, int2 i) -{ - if(WMS == 0) - { - i = frac(f) * WH.xx; - } - else if(WMS == 1) - { - i = saturate(f) * WH.xx; - } - else if(WMS == 2) - { - i = clamp(f, MINUV.xx, MAXUV.xx) * WH.xx; - } - else if(WMS == 3) - { - i = (i & UVMSK.xx) | UVFIX.xx; - } - - return i; -} - -int2 wrapv(float2 f, int2 i) -{ - if(WMT == 0) - { - i = frac(f) * WH.yy; - } - else if(WMT == 1) - { - i = saturate(f) * WH.yy; - } - else if(WMT == 2) - { - i = clamp(f, MINUV.yy, MAXUV.yy) * WH.yy; - } - else if(WMT == 3) - { - i = (i & UVMSK.yy) | UVFIX.yy; - } - - return i; -} - -int4 wrapuv(float4 f, int4 i) -{ - if(WMT == 0) - { - i = frac(f) * WH.xyxy; - } - else if(WMT == 1) - { - i = saturate(f) * WH.xyxy; - } - else if(WMT == 2) - { - i = clamp(f, MINUV.xyxy, MAXUV.xyxy) * WH.xyxy; - } - else if(WMT == 3) - { - i = (i & UVMSK.xyxy) | UVFIX.xyxy; - } - - return i; -} - -int4 wrap(float4 uv, int4 iuv) +int4 wrapuv(int4 uv) { if(WMS == WMT) { - iuv = wrapuv(uv, iuv); + switch(WMS) + { + case 0: uv &= WH.xyxy - 1; break; + case 1: uv = clamp(uv, 0, WH.xyxy); break; + case 2: uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); break; + case 3: uv = (uv & MskFix.xyxy) | MskFix.zwzw; break; + } } else { - iuv.xz = wrapu(uv.xz, iuv.xz); - iuv.yw = wrapv(uv.yw, iuv.yw); + switch(WMS) + { + case 0: uv.xz &= WH.xx - 1; break; + case 1: uv.xz = clamp(uv.xz, 0, WH.xx); break; + case 2: uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); break; + case 3: uv.xz = (uv.xz & MskFix.xx) | MskFix.zz; break; + } + + switch(WMT) + { + case 0: uv.yw &= WH.yy - 1; break; + case 1: uv.yw = clamp(uv.yw, 0, WH.yy); break; + case 2: uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); break; + case 3: uv.yw = (uv.yw & MskFix.yy) | MskFix.ww; break; + } } - return iuv; + return uv; +} + +float2 clampuv(float2 tc) +{ + if(WMS == 2 && WMT == 2) + { + tc = clamp(tc, MinMaxF.xy, MinMaxF.zw); + } + else if(WMS == 2) + { + tc.x = clamp(tc.x, MinMaxF.x, MinMaxF.z); + } + else if(WMT == 2) + { + tc.y = clamp(tc.y, MinMaxF.y, MinMaxF.w); + } + + return tc; } float4 sample(float2 tc, float w) @@ -318,19 +278,13 @@ float4 sample(float2 tc, float w) */ if(BPP < 3 && WMS < 3 && WMT < 3) { - if(WMS == 2 && WMT == 2) tc = clamp(tc, MINUV.xy, MAXUV.xy); - else if(WMS == 2) tc.x = clamp(tc.x, MINUV.x, MAXUV.x); - else if(WMT == 2) tc.y = clamp(tc.y, MINUV.y, MAXUV.y); - - t = Texture.Sample(TextureSampler, tc); + t = Texture.Sample(TextureSampler, clampuv(tc)); } else { - float4 uv = tc.xyxy + HalfTexel; - float4 uv2 = uv * WH.xyxy; - float2 dd = frac(uv2.xy); - - int4 iuv = wrap(uv, uv2); + float4 tc2 = (tc * WH).xyxy + float4(-0.499, -0.499, 0.501, 0.501); + float2 dd = frac(tc2.xy); + int4 uv = wrapuv((int4)tc2); float4 t00, t01, t10, t11; @@ -338,10 +292,10 @@ float4 sample(float2 tc, float w) { float4 a; - a.x = Texture.Load(int3(iuv.xy, 0)).a; - a.y = Texture.Load(int3(iuv.zy, 0)).a; - a.z = Texture.Load(int3(iuv.xw, 0)).a; - a.w = Texture.Load(int3(iuv.zw, 0)).a; + a.x = Texture.Load(int3(uv.xy, 0)).a; + a.y = Texture.Load(int3(uv.zy, 0)).a; + a.z = Texture.Load(int3(uv.xw, 0)).a; + a.w = Texture.Load(int3(uv.zw, 0)).a; t00 = Palette.Load(a.x); t01 = Palette.Load(a.y); @@ -356,10 +310,10 @@ float4 sample(float2 tc, float w) { float4 r; - r.x = Texture.Load(int3(iuv.xy, 0)).r; - r.y = Texture.Load(int3(iuv.zy, 0)).r; - r.z = Texture.Load(int3(iuv.xw, 0)).r; - r.w = Texture.Load(int3(iuv.zw, 0)).r; + r.x = Texture.Load(int3(uv.xy, 0)).r; + r.y = Texture.Load(int3(uv.zy, 0)).r; + r.z = Texture.Load(int3(uv.xw, 0)).r; + r.w = Texture.Load(int3(uv.zw, 0)).r; uint4 i = r * 65535; @@ -370,10 +324,10 @@ float4 sample(float2 tc, float w) } else { - t00 = Texture.Load(int3(iuv.xy, 0)); - t01 = Texture.Load(int3(iuv.zy, 0)); - t10 = Texture.Load(int3(iuv.xw, 0)); - t11 = Texture.Load(int3(iuv.zw, 0)); + t00 = Texture.Load(int3(uv.xy, 0)); + t01 = Texture.Load(int3(uv.zy, 0)); + t10 = Texture.Load(int3(uv.xw, 0)); + t11 = Texture.Load(int3(uv.zw, 0)); } if(LTF) diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index 0c2956423a..9c1011f735 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -71,7 +71,7 @@ VS_OUTPUT vs_main(VS_INPUT input) return output; } -float4 ps_params[4]; +float4 ps_params[5]; #define FogColor ps_params[0].bgr #define AREF ps_params[0].a @@ -79,8 +79,8 @@ float4 ps_params[4]; #define WH ps_params[2].xy #define TA0 ps_params[2].z #define TA1 ps_params[2].w -#define MINUV ps_params[3].xy -#define MAXUV ps_params[3].zw +#define MinMax ps_params[3] +#define MinMaxF ps_params[4] struct PS_INPUT { @@ -90,8 +90,8 @@ struct PS_INPUT #ifndef FST #define FST 0 -#define WMS 1 -#define WMT 1 +#define WMS 3 +#define WMT 3 #define BPP 0 #define AEM 0 #define TFX 0 @@ -109,92 +109,69 @@ sampler1D Palette : register(s1); sampler1D UMSKFIX : register(s2); sampler1D VMSKFIX : register(s3); -float2 wrapu(float2 f) -{ - if(WMS == 0) - { - f = frac(f); - } - else if(WMS == 1) - { - f = saturate(f); - } - else if(WMS == 2) - { - f = clamp(f, MINUV.xx, MAXUV.xx); - } - else if(WMS == 3) - { - f.x = tex1D(UMSKFIX, f.x); - f.y = tex1D(UMSKFIX, f.y); - } - - return f; -} - -float2 wrapv(float2 f) -{ - if(WMS == 0) - { - f = frac(f); - } - else if(WMS == 1) - { - f = saturate(f); - } - else if(WMS == 2) - { - f = clamp(f, MINUV.yy, MAXUV.yy); - } - else if(WMS == 3) - { - f.x = tex1D(VMSKFIX, f.x); - f.y = tex1D(VMSKFIX, f.y); - } - - return f; -} - -float4 wrapuv(float4 f) -{ - if(WMS == 0) - { - f = frac(f); - } - else if(WMS == 1) - { - f = saturate(f); - } - else if(WMS == 2) - { - f = clamp(f, MINUV.xyxy, MAXUV.xyxy); - } - else if(WMS == 3) - { - f.x = tex1D(UMSKFIX, f.x); - f.y = tex1D(VMSKFIX, f.y); - f.z = tex1D(UMSKFIX, f.z); - f.w = tex1D(VMSKFIX, f.w); - } - - return f; -} - -float4 wrap(float4 uv) +float4 wrapuv(float4 uv) { if(WMS == WMT) { - uv = wrapuv(uv); + switch(WMS) + { + case 0: uv = frac(uv); break; + case 1: uv = saturate(uv); break; + case 2: uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); break; + case 3: + uv.x = tex1D(UMSKFIX, uv.x); + uv.y = tex1D(VMSKFIX, uv.y); + uv.z = tex1D(UMSKFIX, uv.z); + uv.w = tex1D(VMSKFIX, uv.w); + break; + } } else { - uv.xz = wrapu(uv.xz); - uv.yw = wrapv(uv.yw); + switch(WMS) + { + case 0: uv.xz = frac(uv.xz); break; + case 1: uv.xz = saturate(uv.xz); break; + case 2: uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); break; + case 3: + uv.x = tex1D(UMSKFIX, uv.x); + uv.z = tex1D(UMSKFIX, uv.z); + break; + } + + switch(WMT) + { + case 0: uv.yw = frac(uv.yw); break; + case 1: uv.yw = saturate(uv.yw); break; + case 2: uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); break; + case 3: + uv.y = tex1D(VMSKFIX, uv.y); + uv.w = tex1D(VMSKFIX, uv.w); + break; + } } return uv; } +float2 clampuv(float2 tc) +{ + if(WMS == 2 && WMT == 2) + { + tc = clamp(tc, MinMaxF.xy, MinMaxF.zw); + } + else if(WMS == 2) + { + tc.x = clamp(tc.x, MinMaxF.x, MinMaxF.z); + } + else if(WMT == 2) + { + tc.y = clamp(tc.y, MinMaxF.y, MinMaxF.w); + } + + return tc; +} + float4 sample(float2 tc, float w) { if(FST == 0) @@ -211,18 +188,13 @@ float4 sample(float2 tc, float w) */ if(BPP < 3 && WMS < 3 && WMT < 3) { - if(WMS == 2 && WMT == 2) tc = clamp(tc, MINUV.xy, MAXUV.xy); - else if(WMS == 2) tc.x = clamp(tc.x, MINUV.x, MAXUV.x); - else if(WMT == 2) tc.y = clamp(tc.y, MINUV.y, MAXUV.y); - - t = tex2D(Texture, tc); + t = tex2D(Texture, clampuv(tc)); } else { - float4 uv = tc.xyxy + HalfTexel; - float2 dd = frac(uv.xy * WH); - - uv = wrap(uv); + float4 tc2 = tc.xyxy + HalfTexel; + float2 dd = frac(tc2.xy * WH); + float4 uv = wrapuv(tc2); float4 t00, t01, t10, t11; From bd229032c07bef0c3fbd8ccb295c48a4f858064c Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 11 Jun 2009 00:08:23 +0000 Subject: [PATCH 212/296] GSdx: fix for genji git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1355 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSCrc.h | 1 + plugins/GSdx/GSState.cpp | 17 +++++++++++++++++ 3 files changed, 19 insertions(+) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index c8d5a44f42..b55978635e 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -133,6 +133,7 @@ CRC::Game CRC::m_games[] = {0xFB236A46, SonicUnleashed, US}, {0x4C7BB3C8, SimpsonsGame, Unknown}, {0x4C94B32C, SimpsonsGame, Unknown}, + {0xD71B57F4, Genji, Unknown}, }; hash_map CRC::m_map; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index 1d68f1dc09..cc64037f8f 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -65,6 +65,7 @@ public: TalesOfAbyss, SonicUnleashed, SimpsonsGame, + Genji, TitleCount, }; diff --git a/plugins/GSdx/GSState.cpp b/plugins/GSdx/GSState.cpp index 1458fd7290..197f12d81e 100644 --- a/plugins/GSdx/GSState.cpp +++ b/plugins/GSdx/GSState.cpp @@ -2195,6 +2195,22 @@ bool GSC_SonicUnleashed(const GSFrameInfo& fi, int& skip) return true; } +bool GSC_Genji(const GSFrameInfo& fi, int& skip) +{ + if(skip == 0) + { + if(fi.TME && fi.FBP == 0x01500 && fi.FPSM == PSM_PSMCT16 && fi.TBP0 == 0x00e00 && fi.TPSM == PSM_PSMZ16) + { + skip = 6; // + } + } + else + { + } + + return true; +} + bool GSState::IsBadFrame(int& skip) { GSFrameInfo fi; @@ -2241,6 +2257,7 @@ bool GSState::IsBadFrame(int& skip) map[CRC::Onimusha3] = GSC_Onimusha3; map[CRC::TalesOfAbyss] = GSC_TalesOfAbyss; map[CRC::SonicUnleashed] = GSC_SonicUnleashed; + map[CRC::Genji] = GSC_Genji; } // TODO: just set gsc in SetGameCRC once From f8a94b15ff81f944f7b46c0df8577a44bf4c514e Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 11 Jun 2009 09:02:10 +0000 Subject: [PATCH 213/296] pcsx2: - Cleanup and organizational changes of VU related stuff. - Separated Super VU, microVU, and Interpreter files. - Renamed all Super VU related files with an 'sVU' prefix. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1356 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Docs/ChangeLog.txt | 2 +- pcsx2/Misc.cpp | 2 +- pcsx2/SaveState.cpp | 2 +- pcsx2/System.cpp | 2 +- pcsx2/VU0micro.cpp | 2 +- pcsx2/VU1micro.cpp | 4 +- pcsx2/VUflags.h | 8 +- pcsx2/VUmicro.h | 7 +- pcsx2/VUmicroMem.cpp | 5 +- pcsx2/VUops.h | 6 +- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 112 +- pcsx2/x86/iCOP2.cpp | 2 +- pcsx2/x86/iCore.h | 2 +- pcsx2/x86/iVU0micro.cpp | 2 +- pcsx2/x86/iVU1micro.cpp | 5 +- pcsx2/x86/ix86-32/iR5900-32.cpp | 4 +- pcsx2/x86/ix86-32/iR5900Templates.cpp | 4 +- pcsx2/x86/{iVUops.h => sVU_Debug.h} | 116 +- .../x86/{iVUmicroLower.cpp => sVU_Lower.cpp} | 4018 +++---- pcsx2/x86/{iVUmicro.cpp => sVU_Micro.cpp} | 3480 +++--- pcsx2/x86/{iVUmicro.h => sVU_Micro.h} | 574 +- .../x86/{iVUmicroUpper.cpp => sVU_Upper.cpp} | 6140 +++++------ pcsx2/x86/{iVUzerorec.cpp => sVU_zerorec.cpp} | 9364 ++++++++--------- pcsx2/x86/{iVUzerorec.h => sVU_zerorec.h} | 97 +- 24 files changed, 11972 insertions(+), 11988 deletions(-) rename pcsx2/x86/{iVUops.h => sVU_Debug.h} (95%) rename pcsx2/x86/{iVUmicroLower.cpp => sVU_Lower.cpp} (96%) rename pcsx2/x86/{iVUmicro.cpp => sVU_Micro.cpp} (96%) rename pcsx2/x86/{iVUmicro.h => sVU_Micro.h} (97%) rename pcsx2/x86/{iVUmicroUpper.cpp => sVU_Upper.cpp} (97%) rename pcsx2/x86/{iVUzerorec.cpp => sVU_zerorec.cpp} (96%) rename pcsx2/x86/{iVUzerorec.h => sVU_zerorec.h} (85%) diff --git a/pcsx2/Docs/ChangeLog.txt b/pcsx2/Docs/ChangeLog.txt index 4c9c3ea7f2..182d60802b 100644 --- a/pcsx2/Docs/ChangeLog.txt +++ b/pcsx2/Docs/ChangeLog.txt @@ -221,7 +221,7 @@ v0.9.1: * 05/08/05:[+] Added faster Cd/DVD reads to the Advanced options (Refraction) [-] The previous DVD/CD change by ref was just a hack, so i commited the old version again - [-] I've commented out RECOMPILE_VUMI_LOADSTORE on iVUmicro.h, instead + [-] I've commented out RECOMPILE_VUMI_LOADSTORE on sVU_Micro.h, instead of commenting it out in iVU1micro.c as ref did on last update, that way is better until i fix the real problem [*] Improved vuDIV for .x to .x cases diff --git a/pcsx2/Misc.cpp b/pcsx2/Misc.cpp index 77992cdc70..d3799e17e2 100644 --- a/pcsx2/Misc.cpp +++ b/pcsx2/Misc.cpp @@ -33,7 +33,7 @@ #include "VUmicro.h" #include "VU.h" #include "iCore.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" #include "BaseblockEx.h" // included for devbuild block dumping (which may or may not work anymore?) #include "GS.h" diff --git a/pcsx2/SaveState.cpp b/pcsx2/SaveState.cpp index 569e4e045a..ce42bf2577 100644 --- a/pcsx2/SaveState.cpp +++ b/pcsx2/SaveState.cpp @@ -25,7 +25,7 @@ #include "VUmicro.h" #include "VU.h" #include "iCore.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" #include "GS.h" #include "COP0.h" diff --git a/pcsx2/System.cpp b/pcsx2/System.cpp index aa1170d3b2..eff13a3209 100644 --- a/pcsx2/System.cpp +++ b/pcsx2/System.cpp @@ -25,7 +25,7 @@ #include "iR5900.h" #include "R3000A.h" #include "IopMem.h" -#include "iVUzerorec.h" // for SuperVUReset +#include "sVU_zerorec.h" // for SuperVUReset #include "R5900Exceptions.h" diff --git a/pcsx2/VU0micro.cpp b/pcsx2/VU0micro.cpp index 772d8ecda3..1502e2cdb2 100644 --- a/pcsx2/VU0micro.cpp +++ b/pcsx2/VU0micro.cpp @@ -32,7 +32,7 @@ #include "VUflags.h" #include "VUops.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" using namespace R5900; diff --git a/pcsx2/VU1micro.cpp b/pcsx2/VU1micro.cpp index dfc3151127..8ebb1b2272 100644 --- a/pcsx2/VU1micro.cpp +++ b/pcsx2/VU1micro.cpp @@ -27,8 +27,8 @@ #include "VU.h" #include "VUops.h" #include "VUmicro.h" -#include "iVUmicro.h" -#include "iVUzerorec.h" +#include "sVU_Micro.h" +#include "sVU_zerorec.h" VURegs* g_pVU1; diff --git a/pcsx2/VUflags.h b/pcsx2/VUflags.h index 4b04a2c012..e0eaad2acf 100644 --- a/pcsx2/VUflags.h +++ b/pcsx2/VUflags.h @@ -16,9 +16,7 @@ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ -#ifndef __VUFLAGS_H__ -#define __VUFLAGS_H__ - +#pragma once #include "VU.h" void vuUpdateDI(VURegs * VU); @@ -33,7 +31,3 @@ __forceinline void VU_MACz_CLEAR(VURegs * VU); __forceinline void VU_MACw_CLEAR(VURegs * VU); void VU_STAT_UPDATE(VURegs * VU); - -#endif - - diff --git a/pcsx2/VUmicro.h b/pcsx2/VUmicro.h index 0b60109f9d..922d74328a 100644 --- a/pcsx2/VUmicro.h +++ b/pcsx2/VUmicro.h @@ -16,9 +16,7 @@ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ -#ifndef __VUMICRO_H__ -#define __VUMICRO_H__ - +#pragma once #include "VU.h" struct VUmicroCpu @@ -1315,6 +1313,3 @@ void (*PREFIX##_LOWER_OPCODE[128])(_VURegsNum *VUregsn) = { \ #endif #include "VUops.h" - -#endif - diff --git a/pcsx2/VUmicroMem.cpp b/pcsx2/VUmicroMem.cpp index 59ed7cd653..907d15a8ae 100644 --- a/pcsx2/VUmicroMem.cpp +++ b/pcsx2/VUmicroMem.cpp @@ -18,13 +18,10 @@ #include "PrecompiledHeader.h" -#include - #include "Common.h" #include "R5900.h" #include "VUmicro.h" - -#include "iVUzerorec.h" +#include "sVU_zerorec.h" // The following CpuVU objects are value types instead of handles or pointers because they are // modified on the fly to implement VU1 Skip. diff --git a/pcsx2/VUops.h b/pcsx2/VUops.h index 2512002169..c19e021279 100644 --- a/pcsx2/VUops.h +++ b/pcsx2/VUops.h @@ -16,9 +16,7 @@ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ -#ifndef __VU1OPS_H__ -#define __VU1OPS_H__ - +#pragma once #include "VU.h" extern __forceinline u32 VU_MAC_UPDATE( int shift, VURegs * VU, float f); @@ -395,5 +393,3 @@ void _vuRegsEEXP(VURegs * VU, _VURegsNum *VUregsn); void _vuRegsXITOP(VURegs * VU, _VURegsNum *VUregsn); void _vuRegsXGKICK(VURegs * VU, _VURegsNum *VUregsn); void _vuRegsXTOP(VURegs * VU, _VURegsNum *VUregsn); - -#endif diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index f3388e5078..5d2ca8cf13 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -2025,26 +2025,10 @@ RelativePath="..\..\VU0micro.cpp" > - - - - - - - - @@ -2053,14 +2037,6 @@ RelativePath="..\..\VUmicroMem.cpp" > - - - - @@ -2072,34 +2048,6 @@ RelativePath="..\..\x86\iVU1micro.cpp" > - - - - - - - - - - - - - - @@ -2160,6 +2108,66 @@ > + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - #include "Common.h" -#include "R5900OpcodeTables.h" #include "iR5900.h" #include "VUmicro.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" #define useMVU1 CHECK_MICROVU1 diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 4a46b80a4c..48212413e8 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -36,11 +36,11 @@ #include "iMMI.h" #include "iFPU.h" #include "iCOP0.h" -#include "iVUmicro.h" +#include "sVU_Micro.h" #include "VU.h" #include "VUmicro.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" #include "vtlb.h" #include "SamplProf.h" diff --git a/pcsx2/x86/ix86-32/iR5900Templates.cpp b/pcsx2/x86/ix86-32/iR5900Templates.cpp index 2149848dfc..b2633bcd6d 100644 --- a/pcsx2/x86/ix86-32/iR5900Templates.cpp +++ b/pcsx2/x86/ix86-32/iR5900Templates.cpp @@ -26,11 +26,11 @@ #include "iMMI.h" #include "iFPU.h" #include "iCOP0.h" -#include "iVUmicro.h" +#include "sVU_Micro.h" #include "VU.h" #include "VUmicro.h" -#include "iVUzerorec.h" +#include "sVU_zerorec.h" #include "vtlb.h" diff --git a/pcsx2/x86/iVUops.h b/pcsx2/x86/sVU_Debug.h similarity index 95% rename from pcsx2/x86/iVUops.h rename to pcsx2/x86/sVU_Debug.h index e369935c19..9c6c023e9a 100644 --- a/pcsx2/x86/iVUops.h +++ b/pcsx2/x86/sVU_Debug.h @@ -1,57 +1,59 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA - */ - -#define REC_VUOP(VU, f) { \ - _freeXMMregs(/*&VU*/); \ - _freeMMXregs(); \ - SetFPUstate();) \ - MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ - CALLFunc((uptr)VU##MI_##f); \ -} - -#define REC_VUOPs(VU, f) { \ - _freeXMMregs(); \ - _freeMMXregs(); \ - SetFPUstate();) \ - if (VU==&VU1) { \ - MOV32ItoM((uptr)&VU1.code, (u32)VU1.code); \ - CALLFunc((uptr)VU1MI_##f); \ - } \ - else { \ - MOV32ItoM((uptr)&VU0.code, (u32)VU0.code); \ - CALLFunc((uptr)VU0MI_##f); \ - } \ -} - -#define REC_VUOPFLAGS(VU, f) { \ - _freeXMMregs(/*&VU*/); \ - _freeMMXregs(); \ - SetFPUstate(); \ - MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ - CALLFunc((uptr)VU##MI_##f); \ -} - -#define REC_VUBRANCH(VU, f) { \ - _freeXMMregs(/*&VU*/); \ - _freeMMXregs(); \ - SetFPUstate(); \ - MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ - MOV32ItoM((uptr)&VU.VI[REG_TPC].UL, (u32)pc); \ - CALLFunc((uptr)VU##MI_##f); \ - branch = 1; \ -} +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +#pragma once + +#define REC_VUOP(VU, f) { \ + _freeXMMregs(/*&VU*/); \ + _freeMMXregs(); \ + SetFPUstate();) \ + MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ + CALLFunc((uptr)VU##MI_##f); \ +} + +#define REC_VUOPs(VU, f) { \ + _freeXMMregs(); \ + _freeMMXregs(); \ + SetFPUstate();) \ + if (VU==&VU1) { \ + MOV32ItoM((uptr)&VU1.code, (u32)VU1.code); \ + CALLFunc((uptr)VU1MI_##f); \ + } \ + else { \ + MOV32ItoM((uptr)&VU0.code, (u32)VU0.code); \ + CALLFunc((uptr)VU0MI_##f); \ + } \ +} + +#define REC_VUOPFLAGS(VU, f) { \ + _freeXMMregs(/*&VU*/); \ + _freeMMXregs(); \ + SetFPUstate(); \ + MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ + CALLFunc((uptr)VU##MI_##f); \ +} + +#define REC_VUBRANCH(VU, f) { \ + _freeXMMregs(/*&VU*/); \ + _freeMMXregs(); \ + SetFPUstate(); \ + MOV32ItoM((uptr)&VU.code, (u32)VU.code); \ + MOV32ItoM((uptr)&VU.VI[REG_TPC].UL, (u32)pc); \ + CALLFunc((uptr)VU##MI_##f); \ + branch = 1; \ +} diff --git a/pcsx2/x86/iVUmicroLower.cpp b/pcsx2/x86/sVU_Lower.cpp similarity index 96% rename from pcsx2/x86/iVUmicroLower.cpp rename to pcsx2/x86/sVU_Lower.cpp index fcae6ca866..26d6e76ee2 100644 --- a/pcsx2/x86/iVUmicroLower.cpp +++ b/pcsx2/x86/sVU_Lower.cpp @@ -1,2009 +1,2009 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include "PrecompiledHeader.h" - -#include "Common.h" -#include "GS.h" -#include "R5900OpcodeTables.h" -#include "iR5900.h" -#include "iMMI.h" -#include "iFPU.h" -#include "iCOP0.h" -#include "VUmicro.h" -#include "VUflags.h" -#include "iVUmicro.h" -#include "iVUops.h" -#include "iVUzerorec.h" -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// Helper Macros -//------------------------------------------------------------------ -#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register -#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register -#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register -#define _It_ (_Ft_ & 15) -#define _Is_ (_Fs_ & 15) -#define _Id_ (_Fd_ & 15) - -#define _X (( VU->code>>24) & 0x1) -#define _Y (( VU->code>>23) & 0x1) -#define _Z (( VU->code>>22) & 0x1) -#define _W (( VU->code>>21) & 0x1) - -#define _XYZW_SS (_X+_Y+_Z+_W==1) - -#define _Fsf_ (( VU->code >> 21) & 0x03) -#define _Ftf_ (( VU->code >> 23) & 0x03) - -#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) -#define _UImm11_ (s32)(VU->code & 0x7ff) - -#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] -#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] -#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] -#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] - -#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] -#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] -#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] - -#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) - -#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] -#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] -#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] -#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] - -#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) - - -static const PCSX2_ALIGNED16(u32 VU_ONE[4]) = {0x3f800000, 0xffffffff, 0xffffffff, 0xffffffff}; -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// *VU Lower Instructions!* -// -// Note: * = Checked for errors by cottonvibes -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// DIV* -//------------------------------------------------------------------ -PCSX2_ALIGNED16(u64 DIV_TEMP_XMM[2]); -void recVUMI_DIV(VURegs *VU, int info) -{ - u8 *pjmp, *pjmp1; - u32 *ajmp32, *bjmp32; - - //Console::WriteLn("recVUMI_DIV()"); - AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags - - // FT can be zero here! so we need to check if its zero and set the correct flag. - SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // Clear EEREC_TEMP - SSE_CMPEQPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); // Set all F's if each vector is zero - - SSE_MOVMSKPS_XMM_to_R32( EAX, EEREC_TEMP); // Move the sign bits of the previous calculation - - AND32ItoR( EAX, (1<<_Ftf_) ); // Grab "Is Zero" bits from the previous calculation - ajmp32 = JZ32(0); // Skip if none are - - SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // Clear EEREC_TEMP - SSE_CMPEQPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // Set all F's if each vector is zero - SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); // Move the sign bits of the previous calculation - - AND32ItoR( EAX, (1<<_Fsf_) ); // Grab "Is Zero" bits from the previous calculation - pjmp = JZ8(0); - OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410 ); // Set invalid flag (0/0) - pjmp1 = JMP8(0); - x86SetJ8(pjmp); - OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x820 ); // Zero divide (only when not 0/0) - x86SetJ8(pjmp1); - - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - - _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); - SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); - - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)&g_maxvals[0]); // If division by zero, then EEREC_TEMP = +/- fmax - - bjmp32 = JMP32(0); - - x86SetJ32(ajmp32); - - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat5_useEAX(EEREC_S, EEREC_TEMP, (1 << (3-_Fsf_))); - vuFloat5_useEAX(EEREC_T, EEREC_TEMP, (1 << (3-_Ftf_))); - } - - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - - _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); - - vuFloat_useEAX(info, EEREC_TEMP, 0x8); - - x86SetJ32(bjmp32); - - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SQRT* -//------------------------------------------------------------------ -void recVUMI_SQRT( VURegs *VU, int info ) -{ - u8* pjmp; - //Console::WriteLn("recVUMI_SQRT()"); - - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, _Ftf_); - AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags - - /* Check for negative sqrt */ - SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); - AND32ItoR(EAX, 1); //Check sign - pjmp = JZ8(0); //Skip if none are - OR32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410); // Invalid Flag - Negative number sqrt - x86SetJ8(pjmp); - - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // Do a cardinal sqrt - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp infinities (only need to do positive clamp since EEREC_TEMP is positive) - SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// RSQRT* -//------------------------------------------------------------------ -PCSX2_ALIGNED16(u64 RSQRT_TEMP_XMM[2]); -void recVUMI_RSQRT(VURegs *VU, int info) -{ - u8 *ajmp8, *bjmp8; - u8 *qjmp1, *qjmp2; - int t1reg, t1boolean; - //Console::WriteLn("recVUMI_RSQRT()"); - - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, _Ftf_); - AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags - - /* Check for negative divide */ - SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); - AND32ItoR(EAX, 1); //Check sign - ajmp8 = JZ8(0); //Skip if none are - OR32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410); // Invalid Flag - Negative number sqrt - x86SetJ8(ajmp8); - - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // Do a cardinal sqrt - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp Infinities to Fmax - SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - t1reg = _vuGetTempXMMreg(info); - if( t1reg < 0 ) { - for (t1reg = 0; ( (t1reg == EEREC_TEMP) || (t1reg == EEREC_S) ); t1reg++) - ; // Makes t1reg not be EEREC_TEMP or EEREC_S. - SSE_MOVAPS_XMM_to_M128( (uptr)&RSQRT_TEMP_XMM[0], t1reg ); // backup data in t1reg to a temp address - t1boolean = 1; - } - else t1boolean = 0; - - // Ft can still be zero here! so we need to check if its zero and set the correct flag. - SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg - SSE_CMPEQSS_XMM_to_XMM(t1reg, EEREC_TEMP); // Set all F's if each vector is zero - - SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation - - AND32ItoR( EAX, 0x01 ); // Grab "Is Zero" bits from the previous calculation - ajmp8 = JZ8(0); // Skip if none are - - //check for 0/0 - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - - SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear EEREC_TEMP - SSE_CMPEQPS_XMM_to_XMM(t1reg, EEREC_TEMP); // Set all F's if each vector is zero - SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation - - AND32ItoR( EAX, 0x01 ); // Grab "Is Zero" bits from the previous calculation - qjmp1 = JZ8(0); - OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410 ); // Set invalid flag (0/0) - qjmp2 = JMP8(0); - x86SetJ8(qjmp1); - OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x820 ); // Zero divide (only when not 0/0) - x86SetJ8(qjmp2); - - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)&g_maxvals[0]); // If division by zero, then EEREC_TEMP = +/- fmax - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); - bjmp8 = JMP8(0); - x86SetJ8(ajmp8); - - _unpackVFSS_xyzw(t1reg, EEREC_S, _Fsf_); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat_useEAX(info, t1reg, 0x8); // Clamp Infinities - SSE_DIVSS_XMM_to_XMM(t1reg, EEREC_TEMP); - vuFloat_useEAX(info, t1reg, 0x8); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), t1reg); - - x86SetJ8(bjmp8); - - if (t1boolean) SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)&RSQRT_TEMP_XMM[0] ); // restore t1reg data - else _freeXMMreg(t1reg); // free t1reg -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// _addISIMMtoIT() - Used in IADDI, IADDIU, and ISUBIU instructions -//------------------------------------------------------------------ -void _addISIMMtoIT(VURegs *VU, s16 imm, int info) -{ - int isreg = -1, itreg; - if (_It_ == 0) return; - - if( _Is_ == 0 ) { - itreg = ALLOCVI(_It_, MODE_WRITE); - MOV32ItoR(itreg, imm&0xffff); - return; - } - - ADD_VI_NEEDED(_It_); - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_WRITE); - - if ( _It_ == _Is_ ) { - if (imm != 0 ) ADD16ItoR(itreg, imm); - } - else { - if( imm ) { - LEA32RtoR(itreg, isreg, imm); - MOVZX32R16toR(itreg, itreg); - } - else MOV32RtoR(itreg, isreg); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// IADDI -//------------------------------------------------------------------ -void recVUMI_IADDI(VURegs *VU, int info) -{ - s16 imm; - - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_IADDI"); - imm = ( VU->code >> 6 ) & 0x1f; - imm = ( imm & 0x10 ? 0xfff0 : 0) | ( imm & 0xf ); - _addISIMMtoIT(VU, imm, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// IADDIU -//------------------------------------------------------------------ -void recVUMI_IADDIU(VURegs *VU, int info) -{ - s16 imm; - - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_IADDIU"); - imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); - _addISIMMtoIT(VU, imm, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// IADD -//------------------------------------------------------------------ -void recVUMI_IADD( VURegs *VU, int info ) -{ - int idreg, isreg = -1, itreg = -1; - if ( _Id_ == 0 ) return; - //Console::WriteLn("recVUMI_IADD"); - if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { - idreg = ALLOCVI(_Id_, MODE_WRITE); - XOR32RtoR(idreg, idreg); - return; - } - - ADD_VI_NEEDED(_Is_); - ADD_VI_NEEDED(_It_); - idreg = ALLOCVI(_Id_, MODE_WRITE); - - if ( _Is_ == 0 ) - { - if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { - if( idreg != itreg ) MOV32RtoR(idreg, itreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); - } - else if ( _It_ == 0 ) - { - if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { - if( idreg != isreg ) MOV32RtoR(idreg, isreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); - } - else { - //ADD_VI_NEEDED(_It_); - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_READ); - - if( idreg == isreg ) ADD32RtoR(idreg, itreg); - else if( idreg == itreg ) ADD32RtoR(idreg, isreg); - else LEA32RRtoR(idreg, isreg, itreg); - MOVZX32R16toR(idreg, idreg); // needed since don't know if idreg's upper bits are 0 - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// IAND -//------------------------------------------------------------------ -void recVUMI_IAND( VURegs *VU, int info ) -{ - int idreg, isreg = -1, itreg = -1; - if ( _Id_ == 0 ) return; - //Console::WriteLn("recVUMI_IAND"); - if ( ( _Is_ == 0 ) || ( _It_ == 0 ) ) { - idreg = ALLOCVI(_Id_, MODE_WRITE); - XOR32RtoR(idreg, idreg); - return; - } - - ADD_VI_NEEDED(_Is_); - ADD_VI_NEEDED(_It_); - idreg = ALLOCVI(_Id_, MODE_WRITE); - - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_READ); - - if( idreg == isreg ) AND16RtoR(idreg, itreg); - else if( idreg == itreg ) AND16RtoR(idreg, isreg); - else { - MOV32RtoR(idreg, itreg); - AND32RtoR(idreg, isreg); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// IOR -//------------------------------------------------------------------ -void recVUMI_IOR( VURegs *VU, int info ) -{ - int idreg, isreg = -1, itreg = -1; - if ( _Id_ == 0 ) return; - //Console::WriteLn("recVUMI_IOR"); - if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { - idreg = ALLOCVI(_Id_, MODE_WRITE); - XOR32RtoR(idreg, idreg); - return; - } - - ADD_VI_NEEDED(_Is_); - ADD_VI_NEEDED(_It_); - idreg = ALLOCVI(_Id_, MODE_WRITE); - - if ( _Is_ == 0 ) - { - if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { - if( idreg != itreg ) MOV32RtoR(idreg, itreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); - } - else if ( _It_ == 0 ) - { - if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { - if( idreg != isreg ) MOV32RtoR(idreg, isreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); - } - else - { - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_READ); - - if( idreg == isreg ) OR16RtoR(idreg, itreg); - else if( idreg == itreg ) OR16RtoR(idreg, isreg); - else { - MOV32RtoR(idreg, isreg); - OR32RtoR(idreg, itreg); - } - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ISUB -//------------------------------------------------------------------ -void recVUMI_ISUB( VURegs *VU, int info ) -{ - int idreg, isreg = -1, itreg = -1; - if ( _Id_ == 0 ) return; - //Console::WriteLn("recVUMI_ISUB"); - if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { - idreg = ALLOCVI(_Id_, MODE_WRITE); - XOR32RtoR(idreg, idreg); - return; - } - - ADD_VI_NEEDED(_Is_); - ADD_VI_NEEDED(_It_); - idreg = ALLOCVI(_Id_, MODE_WRITE); - - if ( _Is_ == 0 ) - { - if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { - if( idreg != itreg ) MOV32RtoR(idreg, itreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); - NEG16R(idreg); - } - else if ( _It_ == 0 ) - { - if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { - if( idreg != isreg ) MOV32RtoR(idreg, isreg); - } - else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); - } - else - { - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_READ); - - if( idreg == isreg ) SUB16RtoR(idreg, itreg); - else if( idreg == itreg ) { - SUB16RtoR(idreg, isreg); - NEG16R(idreg); - } - else { - MOV32RtoR(idreg, isreg); - SUB16RtoR(idreg, itreg); - } - } -} -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// ISUBIU -//------------------------------------------------------------------ -void recVUMI_ISUBIU( VURegs *VU, int info ) -{ - s16 imm; - - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_ISUBIU"); - imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); - imm = -imm; - _addISIMMtoIT(VU, imm, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MOVE* -//------------------------------------------------------------------ -void recVUMI_MOVE( VURegs *VU, int info ) -{ - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - //Console::WriteLn("recVUMI_MOVE"); - if (_X_Y_Z_W == 0x8) SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_S); - else if (_X_Y_Z_W == 0xf) SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MFIR* -//------------------------------------------------------------------ -void recVUMI_MFIR( VURegs *VU, int info ) -{ - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - //Console::WriteLn("recVUMI_MFIR"); - _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, 1); - - if( _XYZW_SS ) { - SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); - _vuFlipRegSS(VU, EEREC_T); - SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); - SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_T); - } - else if (_X_Y_Z_W != 0xf) { - SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); - SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - SSE2_MOVD_M32_to_XMM(EEREC_T, VU_VI_ADDR(_Is_, 1)-2); - SSE2_PSRAD_I8_to_XMM(EEREC_T, 16); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MTIR* -//------------------------------------------------------------------ -void recVUMI_MTIR( VURegs *VU, int info ) -{ - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_MTIR"); - _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, 2); - - if( _Fsf_ == 0 ) { - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_S); - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_TEMP); - } - - AND32ItoM(VU_VI_ADDR(_It_, 0), 0xffff); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MR32* -//------------------------------------------------------------------ -void recVUMI_MR32( VURegs *VU, int info ) -{ - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - //Console::WriteLn("recVUMI_MR32"); - if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x39); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x39); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// _loadEAX() -// -// NOTE: If x86reg < 0, reads directly from offset -//------------------------------------------------------------------ -void _loadEAX(VURegs *VU, int x86reg, uptr offset, int info) -{ - assert( offset < 0x80000000 ); - - if( x86reg >= 0 ) { - switch(_X_Y_Z_W) { - case 3: // ZW - SSE_MOVHPS_Rm_to_XMM(EEREC_T, x86reg, offset+8); - break; - case 6: // YZ - SSE_SHUFPS_Rm_to_XMM(EEREC_T, x86reg, offset, 0x9c); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x78); - break; - - case 8: // X - SSE_MOVSS_Rm_to_XMM(EEREC_TEMP, x86reg, offset); - SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - break; - case 9: // XW - SSE_SHUFPS_Rm_to_XMM(EEREC_T, x86reg, offset, 0xc9); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xd2); - break; - case 12: // XY - SSE_MOVLPS_Rm_to_XMM(EEREC_T, x86reg, offset); - break; - case 15: - if( VU == &VU1 ) SSE_MOVAPSRmtoR(EEREC_T, x86reg, offset); - else SSE_MOVUPSRmtoR(EEREC_T, x86reg, offset); - break; - default: - if( VU == &VU1 ) SSE_MOVAPSRmtoR(EEREC_TEMP, x86reg, offset); - else SSE_MOVUPSRmtoR(EEREC_TEMP, x86reg, offset); - - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - break; - } - } - else { - switch(_X_Y_Z_W) { - case 3: // ZW - SSE_MOVHPS_M64_to_XMM(EEREC_T, offset+8); - break; - case 6: // YZ - SSE_SHUFPS_M128_to_XMM(EEREC_T, offset, 0x9c); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x78); - break; - case 8: // X - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, offset); - SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - break; - case 9: // XW - SSE_SHUFPS_M128_to_XMM(EEREC_T, offset, 0xc9); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xd2); - break; - case 12: // XY - SSE_MOVLPS_M64_to_XMM(EEREC_T, offset); - break; - case 15: - if( VU == &VU1 ) SSE_MOVAPS_M128_to_XMM(EEREC_T, offset); - else SSE_MOVUPS_M128_to_XMM(EEREC_T, offset); - break; - default: - if( VU == &VU1 ) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, offset); - else SSE_MOVUPS_M128_to_XMM(EEREC_TEMP, offset); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - break; - } - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// recVUTransformAddr() -//------------------------------------------------------------------ -int recVUTransformAddr(int x86reg, VURegs* VU, int vireg, int imm) -{ - u8* pjmp[2]; - if( x86reg == EAX ) { - if (imm) ADD32ItoR(x86reg, imm); - } - else { - if( imm ) LEA32RtoR(EAX, x86reg, imm); - else MOV32RtoR(EAX, x86reg); - } - - if( VU == &VU1 ) { - AND32ItoR(EAX, 0x3ff); // wrap around - SHL32ItoR(EAX, 4); - } - else { - - // VU0 has a somewhat interesting memory mapping: - // if addr >= 0x4000, reads VU1's VF regs and VI regs - // if addr < 0x4000, wrap around at 0x1000 - - CMP32ItoR(EAX, 0x400); - pjmp[0] = JL8(0); // if addr >= 0x4000, reads VU1's VF regs and VI regs - AND32ItoR(EAX, 0x43f); - pjmp[1] = JMP8(0); - x86SetJ8(pjmp[0]); - AND32ItoR(EAX, 0xff); // if addr < 0x4000, wrap around - x86SetJ8(pjmp[1]); - - SHL32ItoR(EAX, 4); // multiply by 16 (shift left by 4) - } - - return EAX; -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// LQ -//------------------------------------------------------------------ -void recVUMI_LQ(VURegs *VU, int info) -{ - s16 imm; - if ( _Ft_ == 0 ) return; - //Console::WriteLn("recVUMI_LQ"); - imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); - if (_Is_ == 0) { - _loadEAX(VU, -1, (uptr)GET_VU_MEM(VU, (u32)imm*16), info); - } - else { - int isreg = ALLOCVI(_Is_, MODE_READ); - _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem, info); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// LQD -//------------------------------------------------------------------ -void recVUMI_LQD( VURegs *VU, int info ) -{ - int isreg; - //Console::WriteLn("recVUMI_LQD"); - if ( _Is_ != 0 ) { - isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); - SUB16ItoR( isreg, 1 ); - } - - if ( _Ft_ == 0 ) return; - - if ( _Is_ == 0 ) _loadEAX(VU, -1, (uptr)VU->Mem, info); - else _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// LQI -//------------------------------------------------------------------ -void recVUMI_LQI(VURegs *VU, int info) -{ - int isreg; - //Console::WriteLn("recVUMI_LQI"); - if ( _Ft_ == 0 ) { - if( _Is_ != 0 ) { - if( (isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_WRITE|MODE_READ)) >= 0 ) { - ADD16ItoR(isreg, 1); - } - else { - ADD16ItoM( VU_VI_ADDR( _Is_, 0 ), 1 ); - } - } - return; - } - - if (_Is_ == 0) { - _loadEAX(VU, -1, (uptr)VU->Mem, info); - } - else { - isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); - _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); - ADD16ItoR( isreg, 1 ); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// _saveEAX() -//------------------------------------------------------------------ -void _saveEAX(VURegs *VU, int x86reg, uptr offset, int info) -{ - assert( offset < 0x80000000 ); - - if ( _Fs_ == 0 ) { - if ( _XYZW_SS ) { - u32 c = _W ? 0x3f800000 : 0; - if ( x86reg >= 0 ) MOV32ItoRm(x86reg, c, offset+(_W?12:(_Z?8:(_Y?4:0)))); - else MOV32ItoM(offset+(_W?12:(_Z?8:(_Y?4:0))), c); - } - else { - - // (this is one of my test cases for the new emitter --air) - using namespace x86Emitter; - xAddressReg thisreg( x86reg ); - - if ( _X ) xMOV(ptr32[thisreg+offset], 0x00000000); - if ( _Y ) xMOV(ptr32[thisreg+offset+4], 0x00000000); - if ( _Z ) xMOV(ptr32[thisreg+offset+8], 0x00000000); - if ( _W ) xMOV(ptr32[thisreg+offset+12], 0x3f800000); - } - return; - } - - switch ( _X_Y_Z_W ) { - case 1: // W - SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x27); - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); - else SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); - break; - case 2: // Z - SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); - else SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); - break; - case 3: // ZW - if ( x86reg >= 0 ) SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_S, offset+8); - else SSE_MOVHPS_XMM_to_M64(offset+8, EEREC_S); - break; - case 4: // Y - SSE2_PSHUFLW_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x4e); - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); - else SSE_MOVSS_XMM_to_M32(offset+4, EEREC_TEMP); - break; - case 5: // YW - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xB1); - SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if ( x86reg >= 0 ) { - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset+4); - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); - } - else { - SSE_MOVSS_XMM_to_M32(offset+4, EEREC_S); - SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); - } - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xB1); - break; - case 6: // YZ - SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0xc9); - if ( x86reg >= 0 ) SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); - else SSE_MOVLPS_XMM_to_M64(offset+4, EEREC_TEMP); - break; - case 7: // YZW - SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x93); //ZYXW - if ( x86reg >= 0 ) { - SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); - } - else { - SSE_MOVHPS_XMM_to_M64(offset+4, EEREC_TEMP); - SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); - } - break; - case 8: // X - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); - else SSE_MOVSS_XMM_to_M32(offset, EEREC_S); - break; - case 9: // XW - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); - else SSE_MOVSS_XMM_to_M32(offset, EEREC_S); - - SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0xff); //WWWW - - if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); - else SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); - - break; - case 10: //XZ - SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if ( x86reg >= 0 ) { - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); - } - else { - SSE_MOVSS_XMM_to_M32(offset, EEREC_S); - SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); - } - break; - case 11: //XZW - if ( x86reg >= 0 ) { - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); - SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_S, offset+8); - } - else { - SSE_MOVSS_XMM_to_M32(offset, EEREC_S); - SSE_MOVHPS_XMM_to_M64(offset+8, EEREC_S); - } - break; - case 12: // XY - if ( x86reg >= 0 ) SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_S, offset+0); - else SSE_MOVLPS_XMM_to_M64(offset, EEREC_S); - break; - case 13: // XYW - SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x4b); //YXZW - if ( x86reg >= 0 ) { - SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+0); - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); - } - else { - SSE_MOVHPS_XMM_to_M64(offset, EEREC_TEMP); - SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); - } - break; - case 14: // XYZ - SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if ( x86reg >= 0 ) { - SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_S, offset+0); - SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); - } - else { - SSE_MOVLPS_XMM_to_M64(offset, EEREC_S); - SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); - } - break; - case 15: // XYZW - if ( VU == &VU1 ) { - if( x86reg >= 0 ) SSE_MOVAPSRtoRm(x86reg, EEREC_S, offset+0); - else SSE_MOVAPS_XMM_to_M128(offset, EEREC_S); - } - else { - if( x86reg >= 0 ) SSE_MOVUPSRtoRm(x86reg, EEREC_S, offset+0); - else { - if( offset & 15 ) SSE_MOVUPS_XMM_to_M128(offset, EEREC_S); - else SSE_MOVAPS_XMM_to_M128(offset, EEREC_S); - } - } - break; - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SQ -//------------------------------------------------------------------ -void recVUMI_SQ(VURegs *VU, int info) -{ - s16 imm; - //Console::WriteLn("recVUMI_SQ"); - imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); - if ( _It_ == 0 ) _saveEAX(VU, -1, (uptr)GET_VU_MEM(VU, (int)imm * 16), info); - else { - int itreg = ALLOCVI(_It_, MODE_READ); - _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, imm), (uptr)VU->Mem, info); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SQD -//------------------------------------------------------------------ -void recVUMI_SQD(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_SQD"); - if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); - else { - int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); - SUB16ItoR( itreg, 1 ); - _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SQI -//------------------------------------------------------------------ -void recVUMI_SQI(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_SQI"); - if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); - else { - int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); - _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); - ADD16ItoR( itreg, 1 ); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ILW -//------------------------------------------------------------------ -void recVUMI_ILW(VURegs *VU, int info) -{ - int itreg; - s16 imm, off; - - if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; - //Console::WriteLn("recVUMI_ILW"); - imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); - if (_X) off = 0; - else if (_Y) off = 4; - else if (_Z) off = 8; - else if (_W) off = 12; - - ADD_VI_NEEDED(_Is_); - itreg = ALLOCVI(_It_, MODE_WRITE); - - if ( _Is_ == 0 ) { - MOVZX32M16toR( itreg, (uptr)GET_VU_MEM(VU, (int)imm * 16 + off) ); - } - else { - int isreg = ALLOCVI(_Is_, MODE_READ); - MOV32RmtoR(itreg, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem + off); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ISW -//------------------------------------------------------------------ -void recVUMI_ISW( VURegs *VU, int info ) -{ - s16 imm; - //Console::WriteLn("recVUMI_ISW"); - imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); - - if (_Is_ == 0) { - uptr off = (uptr)GET_VU_MEM(VU, (int)imm * 16); - int itreg = ALLOCVI(_It_, MODE_READ); - - if (_X) MOV32RtoM(off, itreg); - if (_Y) MOV32RtoM(off+4, itreg); - if (_Z) MOV32RtoM(off+8, itreg); - if (_W) MOV32RtoM(off+12, itreg); - } - else { - int x86reg, isreg, itreg; - - ADD_VI_NEEDED(_It_); - isreg = ALLOCVI(_Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_READ); - - x86reg = recVUTransformAddr(isreg, VU, _Is_, imm); - - if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); - if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); - if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); - if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ILWR -//------------------------------------------------------------------ -void recVUMI_ILWR( VURegs *VU, int info ) -{ - int off, itreg; - - if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; - //Console::WriteLn("recVUMI_ILWR"); - if (_X) off = 0; - else if (_Y) off = 4; - else if (_Z) off = 8; - else if (_W) off = 12; - - ADD_VI_NEEDED(_Is_); - itreg = ALLOCVI(_It_, MODE_WRITE); - - if ( _Is_ == 0 ) { - MOVZX32M16toR( itreg, (uptr)VU->Mem + off ); - } - else { - int isreg = ALLOCVI(_Is_, MODE_READ); - MOVZX32Rm16toR(itreg, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem + off); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ISWR -//------------------------------------------------------------------ -void recVUMI_ISWR( VURegs *VU, int info ) -{ - int itreg; - //Console::WriteLn("recVUMI_ISWR"); - ADD_VI_NEEDED(_Is_); - itreg = ALLOCVI(_It_, MODE_READ); - - if (_Is_ == 0) { - if (_X) MOV32RtoM((uptr)VU->Mem, itreg); - if (_Y) MOV32RtoM((uptr)VU->Mem+4, itreg); - if (_Z) MOV32RtoM((uptr)VU->Mem+8, itreg); - if (_W) MOV32RtoM((uptr)VU->Mem+12, itreg); - } - else { - int x86reg; - int isreg = ALLOCVI(_Is_, MODE_READ); - x86reg = recVUTransformAddr(isreg, VU, _Is_, 0); - - if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); - if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); - if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); - if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// RINIT* -//------------------------------------------------------------------ -void recVUMI_RINIT(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_RINIT()"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode & MODE_NOFLUSH) ) { - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 2); - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)s_mask); - SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)VU_ONE); - SSE_MOVSS_XMM_to_M32(VU_REGR_ADDR, EEREC_TEMP); - } - else { - int rreg = ALLOCVI(REG_R, MODE_WRITE); - - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - MOV32MtoR( rreg, VU_VFx_ADDR( _Fs_ ) + 4 * _Fsf_ ); - AND32ItoR( rreg, 0x7fffff ); - OR32ItoR( rreg, 0x7f << 23 ); - - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// RGET* -//------------------------------------------------------------------ -void recVUMI_RGET(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_RGET()"); - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); - - if (_X_Y_Z_W != 0xf) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_REGR_ADDR); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_T, VU_REGR_ADDR); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// RNEXT* -//------------------------------------------------------------------ -void recVUMI_RNEXT( VURegs *VU, int info ) -{ - int rreg, x86temp0, x86temp1; - //Console::WriteLn("recVUMI_RNEXT()"); - - rreg = ALLOCVI(REG_R, MODE_WRITE|MODE_READ); - - x86temp0 = ALLOCTEMPX86(0); - x86temp1 = ALLOCTEMPX86(0); - - // code from www.project-fao.org - //MOV32MtoR(rreg, VU_REGR_ADDR); - MOV32RtoR(x86temp0, rreg); - SHR32ItoR(x86temp0, 4); - AND32ItoR(x86temp0, 1); - - MOV32RtoR(x86temp1, rreg); - SHR32ItoR(x86temp1, 22); - AND32ItoR(x86temp1, 1); - - SHL32ItoR(rreg, 1); - XOR32RtoR(x86temp0, x86temp1); - XOR32RtoR(rreg, x86temp0); - AND32ItoR(rreg, 0x7fffff); - OR32ItoR(rreg, 0x3f800000); - - _freeX86reg(x86temp0); - _freeX86reg(x86temp1); - - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) { - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); - return; - } - - recVUMI_RGET(VU, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// RXOR* -//------------------------------------------------------------------ -void recVUMI_RXOR( VURegs *VU, int info ) -{ - //Console::WriteLn("recVUMI_RXOR()"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode & MODE_NOFLUSH) ) { - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - - SSE_XORPS_M128_to_XMM(EEREC_TEMP, VU_REGR_ADDR); - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)s_mask); - SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)s_fones); - SSE_MOVSS_XMM_to_M32(VU_REGR_ADDR, EEREC_TEMP); - } - else { - int rreg = ALLOCVI(REG_R, MODE_WRITE|MODE_READ); - - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - XOR32MtoR( rreg, VU_VFx_ADDR( _Fs_ ) + 4 * _Fsf_ ); - AND32ItoR( rreg, 0x7fffff ); - OR32ItoR ( rreg, 0x3f800000 ); - - _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// WAITQ -//------------------------------------------------------------------ -void recVUMI_WAITQ( VURegs *VU, int info ) -{ - //Console::WriteLn("recVUMI_WAITQ"); -// if( info & PROCESS_VU_SUPER ) { -// //CALLFunc(waitqfn); -// SuperVUFlush(0, 1); -// } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FSAND -//------------------------------------------------------------------ -void recVUMI_FSAND( VURegs *VU, int info ) -{ - int itreg; - u16 imm; - //Console::WriteLn("recVUMI_FSAND"); - imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - if(_It_ == 0) return; - - itreg = ALLOCVI(_It_, MODE_WRITE); - MOV32MtoR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - AND32ItoR( itreg, imm ); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FSEQ -//------------------------------------------------------------------ -void recVUMI_FSEQ( VURegs *VU, int info ) -{ - int itreg; - u16 imm; - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_FSEQ"); - imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - - itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); - - MOVZX32M16toR( EAX, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - XOR32RtoR(itreg, itreg); - CMP16ItoR(EAX, imm); - SETE8R(itreg); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FSOR -//------------------------------------------------------------------ -void recVUMI_FSOR( VURegs *VU, int info ) -{ - int itreg; - u32 imm; - if(_It_ == 0) return; - //Console::WriteLn("recVUMI_FSOR"); - imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); - - itreg = ALLOCVI(_It_, MODE_WRITE); - - MOVZX32M16toR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); - OR32ItoR( itreg, imm ); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FSSET -//------------------------------------------------------------------ -void recVUMI_FSSET(VURegs *VU, int info) -{ - u32 writeaddr = VU_VI_ADDR(REG_STATUS_FLAG, 0); - u32 prevaddr = VU_VI_ADDR(REG_STATUS_FLAG, 2); - - u16 imm = 0; - //Console::WriteLn("recVUMI_FSSET"); - imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7FF); - - // keep the low 6 bits ONLY if the upper instruction is an fmac instruction (otherwise rewrite) - metal gear solid 3 - //if( (info & PROCESS_VU_SUPER) && VUREC_FMAC ) { - MOV32MtoR(EAX, prevaddr); - AND32ItoR(EAX, 0x3f); - if ((imm&0xfc0) != 0) OR32ItoR(EAX, imm & 0xFC0); - MOV32RtoM(writeaddr ? writeaddr : prevaddr, EAX); - //} - //else { - // MOV32ItoM(writeaddr ? writeaddr : prevaddr, imm&0xfc0); - //} -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FMAND -//------------------------------------------------------------------ -void recVUMI_FMAND( VURegs *VU, int info ) -{ - int isreg, itreg; - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_FMAND"); - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); - - if( isreg >= 0 ) { - if( itreg != isreg ) MOV32RtoR(itreg, isreg); - } - else MOVZX32M16toR(itreg, VU_VI_ADDR(_Is_, 1)); - - AND16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FMEQ -//------------------------------------------------------------------ -void recVUMI_FMEQ( VURegs *VU, int info ) -{ - int itreg, isreg; - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_FMEQ"); - if( _It_ == _Is_ ) { - itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG - - CMP16MtoR(itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); - SETE8R(EAX); - MOVZX32R8toR(itreg, EAX); - } - else { - ADD_VI_NEEDED(_Is_); - itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); - isreg = ALLOCVI(_Is_, MODE_READ); - - XOR32RtoR(itreg, itreg); - - CMP16MtoR(isreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); - SETE8R(itreg); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FMOR -//------------------------------------------------------------------ -void recVUMI_FMOR( VURegs *VU, int info ) -{ - int isreg, itreg; - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_FMOR"); - if( _Is_ == 0 ) { - itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); - MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); - } - else if( _It_ == _Is_ ) { - itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG); - OR16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); - } - else { - isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); - itreg = ALLOCVI(_It_, MODE_WRITE); - - MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); - - if( isreg >= 0 ) - OR16RtoR( itreg, isreg ); - else - OR16MtoR( itreg, VU_VI_ADDR(_Is_, 1) ); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FCAND -//------------------------------------------------------------------ -void recVUMI_FCAND( VURegs *VU, int info ) -{ - int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); - //Console::WriteLn("recVUMI_FCAND"); - MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); - XOR32RtoR( itreg, itreg ); - AND32ItoR( EAX, VU->code & 0xFFFFFF ); - - SETNZ8R(itreg); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FCEQ -//------------------------------------------------------------------ -void recVUMI_FCEQ( VURegs *VU, int info ) -{ - int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); - //Console::WriteLn("recVUMI_FCEQ"); - MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); - AND32ItoR( EAX, 0xffffff ); - XOR32RtoR( itreg, itreg ); - CMP32ItoR( EAX, VU->code&0xffffff ); - - SETE8R(itreg); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FCOR -//------------------------------------------------------------------ -void recVUMI_FCOR( VURegs *VU, int info ) -{ - int itreg; - //Console::WriteLn("recVUMI_FCOR"); - itreg = ALLOCVI(1, MODE_WRITE); - MOV32MtoR( itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); - OR32ItoR ( itreg, VU->code ); - AND32ItoR( itreg, 0xffffff ); - ADD32ItoR( itreg, 1 ); // If 24 1's will make 25th bit 1, else 0 - SHR32ItoR( itreg, 24 ); // Get the 25th bit (also clears the rest of the garbage in the reg) -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FCSET -//------------------------------------------------------------------ -void recVUMI_FCSET( VURegs *VU, int info ) -{ - u32 addr = VU_VI_ADDR(REG_CLIP_FLAG, 0); - //Console::WriteLn("recVUMI_FCSET"); - MOV32ItoM(addr ? addr : VU_VI_ADDR(REG_CLIP_FLAG, 2), VU->code&0xffffff ); - - if( !(info & (PROCESS_VU_SUPER|PROCESS_VU_COP2)) ) - MOV32ItoM( VU_VI_ADDR(REG_CLIP_FLAG, 1), VU->code&0xffffff ); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FCGET -//------------------------------------------------------------------ -void recVUMI_FCGET( VURegs *VU, int info ) -{ - int itreg; - if(_It_ == 0) return; - //Console::WriteLn("recVUMI_FCGET"); - itreg = ALLOCVI(_It_, MODE_WRITE); - - MOV32MtoR(itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1)); - AND32ItoR(itreg, 0x0fff); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// _recbranchAddr() -// -// NOTE: Due to static var dependencies, several SuperVU branch instructions -// are still located in iVUzerorec.cpp. -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// MFP* -//------------------------------------------------------------------ -void recVUMI_MFP(VURegs *VU, int info) -{ - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - //Console::WriteLn("recVUMI_MFP"); - if( _XYZW_SS ) { - _vuFlipRegSS(VU, EEREC_T); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 1)); - SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_T); - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 1)); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_T, VU_VI_ADDR(REG_P, 1)); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// WAITP -//------------------------------------------------------------------ -static PCSX2_ALIGNED16(float s_tempmem[4]); -void recVUMI_WAITP(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_WAITP"); -// if( info & PROCESS_VU_SUPER ) -// SuperVUFlush(1, 1); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// vuSqSumXYZ()* -// -// NOTE: In all EFU insts, EEREC_D is a temp reg -//------------------------------------------------------------------ -void vuSqSumXYZ(int regd, int regs, int regtemp) // regd.x = x ^ 2 + y ^ 2 + z ^ 2 -{ - //Console::WriteLn("VU: SUMXYZ"); - if( cpucaps.hasStreamingSIMD4Extensions ) - { - SSE_MOVAPS_XMM_to_XMM(regd, regs); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat2(regd, regtemp, 0xf); - SSE4_DPPS_XMM_to_XMM(regd, regd, 0x71); - } - else - { - SSE_MOVAPS_XMM_to_XMM(regtemp, regs); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat2(regtemp, regd, 0xf); - SSE_MULPS_XMM_to_XMM(regtemp, regtemp); // xyzw ^ 2 - - if( cpucaps.hasStreamingSIMD3Extensions ) { - SSE3_HADDPS_XMM_to_XMM(regd, regtemp); - SSE_ADDPS_XMM_to_XMM(regd, regtemp); // regd.z = x ^ 2 + y ^ 2 + z ^ 2 - SSE_MOVHLPS_XMM_to_XMM(regd, regd); // regd.x = regd.z - } - else { - SSE_MOVSS_XMM_to_XMM(regd, regtemp); - SSE2_PSHUFLW_XMM_to_XMM(regtemp, regtemp, 0x4e); // wzyx -> wzxy - SSE_ADDSS_XMM_to_XMM(regd, regtemp); // x ^ 2 + y ^ 2 - SSE_SHUFPS_XMM_to_XMM(regtemp, regtemp, 0xD2); // wzxy -> wxyz - SSE_ADDSS_XMM_to_XMM(regd, regtemp); // x ^ 2 + y ^ 2 + z ^ 2 - } - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ESADD* -//------------------------------------------------------------------ -void recVUMI_ESADD( VURegs *VU, int info) -{ - //Console::WriteLn("VU: ESADD"); - assert( VU == &VU1 ); - if( EEREC_TEMP == EEREC_D ) { // special code to reset P ( FixMe: don't know if this is still needed! (cottonvibes) ) - Console::Notice("ESADD: Resetting P reg!!!\n"); - MOV32ItoM(VU_VI_ADDR(REG_P, 0), 0); - return; - } - vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_D); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ERSADD* -//------------------------------------------------------------------ -void recVUMI_ERSADD( VURegs *VU, int info ) -{ - //Console::WriteLn("VU: ERSADD"); - assert( VU == &VU1 ); - vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); - // don't use RCPSS (very bad precision) - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ELENG* -//------------------------------------------------------------------ -void recVUMI_ELENG( VURegs *VU, int info ) -{ - //Console::WriteLn("VU: ELENG"); - assert( VU == &VU1 ); - vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive - SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_D); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ERLENG* -//------------------------------------------------------------------ -void recVUMI_ERLENG( VURegs *VU, int info ) -{ - //Console::WriteLn("VU: ERLENG"); - assert( VU == &VU1 ); - vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive - SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D); // regd <- sqrt(x^2 + y^2 + z^2) - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); // temp = 1 / sqrt(x^2 + y^2 + z^2) - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// EATANxy -//------------------------------------------------------------------ -void recVUMI_EATANxy( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - //Console::WriteLn("recVUMI_EATANxy"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { - SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); - FLD32((uptr)&s_tempmem[0]); - FLD32((uptr)&s_tempmem[1]); - } - else { - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - FLD32((uptr)&VU->VF[_Fs_].UL[0]); - FLD32((uptr)&VU->VF[_Fs_].UL[1]); - } - - FPATAN(); - FSTP32(VU_VI_ADDR(REG_P, 0)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// EATANxz -//------------------------------------------------------------------ -void recVUMI_EATANxz( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - //Console::WriteLn("recVUMI_EATANxz"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { - SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); - FLD32((uptr)&s_tempmem[0]); - FLD32((uptr)&s_tempmem[2]); - } - else { - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - FLD32((uptr)&VU->VF[_Fs_].UL[0]); - FLD32((uptr)&VU->VF[_Fs_].UL[2]); - } - FPATAN(); - FSTP32(VU_VI_ADDR(REG_P, 0)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ESUM* -//------------------------------------------------------------------ -void recVUMI_ESUM( VURegs *VU, int info ) -{ - //Console::WriteLn("VU: ESUM"); - assert( VU == &VU1 ); - - if( cpucaps.hasStreamingSIMD3Extensions ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat_useEAX(info, EEREC_TEMP, 0xf); - SSE3_HADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - SSE3_HADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - } - else { - SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // z, w, z, w - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // z+x, w+y, z+z, w+w - SSE_UNPCKLPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // z+x, z+x, w+y, w+y - SSE_MOVHLPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); // w+y, w+y, w+y, w+y - SSE_ADDSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); // x+y+z+w, w+y, w+y, w+y - } - - vuFloat_useEAX(info, EEREC_TEMP, 8); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ERCPR* -//------------------------------------------------------------------ -void recVUMI_ERCPR( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - //Console::WriteLn("VU1: ERCPR"); - - // don't use RCPSS (very bad precision) - switch ( _Fsf_ ) { - case 0: //0001 - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - break; - case 1: //0010 - SSE2_PSHUFLW_XMM_to_XMM(EEREC_S, EEREC_S, 0x4e); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE2_PSHUFLW_XMM_to_XMM(EEREC_S, EEREC_S, 0x4e); - break; - case 2: //0100 - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xc6); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xc6); - break; - case 3: //1000 - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0x27); - if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 - SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0x27); - break; - } - - vuFloat_useEAX(info, EEREC_TEMP, 8); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ESQRT* -//------------------------------------------------------------------ -void recVUMI_ESQRT( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - - //Console::WriteLn("VU1: ESQRT"); - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // abs(x) - if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp - SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ERSQRT* -//------------------------------------------------------------------ -void recVUMI_ERSQRT( VURegs *VU, int info ) -{ - int t1reg = _vuGetTempXMMreg(info); - - assert( VU == &VU1 ); - //Console::WriteLn("VU1: ERSQRT"); - - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // abs(x) - SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp Infinities to Fmax - SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // SQRT(abs(x)) - - if( t1reg >= 0 ) - { - SSE_MOVSS_M32_to_XMM(t1reg, (uptr)VU_ONE); - SSE_DIVSS_XMM_to_XMM(t1reg, EEREC_TEMP); - vuFloat_useEAX(info, t1reg, 8); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), t1reg); - _freeXMMreg(t1reg); - } - else - { - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); - SSE_DIVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 0)); - vuFloat_useEAX(info, EEREC_TEMP, 8); - SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ESIN -//------------------------------------------------------------------ -void recVUMI_ESIN( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - - //Console::WriteLn("recVUMI_ESIN"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { - switch(_Fsf_) { - case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); - case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); - default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); - } - FLD32((uptr)&s_tempmem[_Fsf_]); - } - else { - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - FLD32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); - } - - FSIN(); - FSTP32(VU_VI_ADDR(REG_P, 0)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// EATAN -//------------------------------------------------------------------ -void recVUMI_EATAN( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - - //Console::WriteLn("recVUMI_EATAN"); - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { - switch(_Fsf_) { - case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); - case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); - default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); - } - FLD32((uptr)&s_tempmem[_Fsf_]); - } - else { - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - } - - FLD1(); - FLD32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); - FPATAN(); - FSTP32(VU_VI_ADDR(REG_P, 0)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// EEXP -//------------------------------------------------------------------ -void recVUMI_EEXP( VURegs *VU, int info ) -{ - assert( VU == &VU1 ); - //Console::WriteLn("recVUMI_EEXP"); - FLDL2E(); - - if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { - switch(_Fsf_) { - case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); - case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); - default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); - } - FMUL32((uptr)&s_tempmem[_Fsf_]); - } - else { - if( xmmregs[EEREC_S].mode & MODE_WRITE ) { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); - xmmregs[EEREC_S].mode &= ~MODE_WRITE; - } - - FMUL32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); - } - - // basically do 2^(log_2(e) * val) - FLD(0); - FRNDINT(); - FXCH(1); - FSUB32Rto0(1); - F2XM1(); - FLD1(); - FADD320toR(1); - FSCALE(); - FSTP(1); - - FSTP32(VU_VI_ADDR(REG_P, 0)); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// XITOP -//------------------------------------------------------------------ -void recVUMI_XITOP( VURegs *VU, int info ) -{ - int itreg; - if (_It_ == 0) return; - //Console::WriteLn("recVUMI_XITOP"); - itreg = ALLOCVI(_It_, MODE_WRITE); - MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->itop ); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// XTOP -//------------------------------------------------------------------ -void recVUMI_XTOP( VURegs *VU, int info ) -{ - int itreg; - if ( _It_ == 0 ) return; - //Console::WriteLn("recVUMI_XTOP"); - itreg = ALLOCVI(_It_, MODE_WRITE); - MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->top ); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// VU1XGKICK_MTGSTransfer() - Called by ivuZerorec.cpp -//------------------------------------------------------------------ -void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr) -{ - u32 size; - u32* data = (u32*)((u8*)pMem + (addr&0x3fff)); - - // fixme: The gifTagDummy function in the MTGS (called by PrepDataPacket) has a - // hack that aborts the packet if it goes past the end of VU1 memory. - // Chances are this should be a "loops around memory" situation, and the packet - // should be continued starting at addr zero (0). - - size = mtgsThread->PrepDataPacket( GIF_PATH_1, data, (0x4000-(addr&0x3fff)) >> 4); - jASSUME( size > 0 ); - - //if( size > 0 ) - { - u8* pmem = mtgsThread->GetDataPacketPtr(); - - /* if((size << 4) > (0x4000-(addr&0x3fff))) - { - //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params (addr&0x3fff) + (size << 4), (0x4000-(addr&0x3fff)) >> 4, size - ((0x4000-(addr&0x3fff)) >> 4)); - memcpy_aligned(pmem, (u8*)pMem+addr, 0x4000-(addr&0x3fff)); - size -= (0x4000-(addr&0x3fff)) >> 4; - //DevCon::Notice("Size left %x", params size); - pmem += 0x4000-(addr&0x3fff); - memcpy_aligned(pmem, (u8*)pMem, size<<4); - } - else - {*/ - memcpy_aligned(pmem, (u8*)pMem+addr, size<<4); - // } - mtgsThread->SendDataPacket(); - } -} -//------------------------------------------------------------------ +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "PrecompiledHeader.h" + +#include "Common.h" +#include "GS.h" +#include "R5900OpcodeTables.h" +#include "iR5900.h" +#include "iMMI.h" +#include "iFPU.h" +#include "iCOP0.h" +#include "VUmicro.h" +#include "VUflags.h" +#include "sVU_Micro.h" +#include "sVU_Debug.h" +#include "sVU_zerorec.h" +//------------------------------------------------------------------ + +//------------------------------------------------------------------ +// Helper Macros +//------------------------------------------------------------------ +#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register +#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register +#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) + +#define _X (( VU->code>>24) & 0x1) +#define _Y (( VU->code>>23) & 0x1) +#define _Z (( VU->code>>22) & 0x1) +#define _W (( VU->code>>21) & 0x1) + +#define _XYZW_SS (_X+_Y+_Z+_W==1) + +#define _Fsf_ (( VU->code >> 21) & 0x03) +#define _Ftf_ (( VU->code >> 23) & 0x03) + +#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) +#define _UImm11_ (s32)(VU->code & 0x7ff) + +#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] +#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] +#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] +#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] + +#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] +#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] +#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] + +#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) + +#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] +#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] +#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] +#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] + +#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) + + +static const PCSX2_ALIGNED16(u32 VU_ONE[4]) = {0x3f800000, 0xffffffff, 0xffffffff, 0xffffffff}; +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// *VU Lower Instructions!* +// +// Note: * = Checked for errors by cottonvibes +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// DIV* +//------------------------------------------------------------------ +PCSX2_ALIGNED16(u64 DIV_TEMP_XMM[2]); +void recVUMI_DIV(VURegs *VU, int info) +{ + u8 *pjmp, *pjmp1; + u32 *ajmp32, *bjmp32; + + //Console::WriteLn("recVUMI_DIV()"); + AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags + + // FT can be zero here! so we need to check if its zero and set the correct flag. + SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // Clear EEREC_TEMP + SSE_CMPEQPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); // Set all F's if each vector is zero + + SSE_MOVMSKPS_XMM_to_R32( EAX, EEREC_TEMP); // Move the sign bits of the previous calculation + + AND32ItoR( EAX, (1<<_Ftf_) ); // Grab "Is Zero" bits from the previous calculation + ajmp32 = JZ32(0); // Skip if none are + + SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // Clear EEREC_TEMP + SSE_CMPEQPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // Set all F's if each vector is zero + SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); // Move the sign bits of the previous calculation + + AND32ItoR( EAX, (1<<_Fsf_) ); // Grab "Is Zero" bits from the previous calculation + pjmp = JZ8(0); + OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410 ); // Set invalid flag (0/0) + pjmp1 = JMP8(0); + x86SetJ8(pjmp); + OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x820 ); // Zero divide (only when not 0/0) + x86SetJ8(pjmp1); + + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + + _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); + SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); + + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)&g_maxvals[0]); // If division by zero, then EEREC_TEMP = +/- fmax + + bjmp32 = JMP32(0); + + x86SetJ32(ajmp32); + + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat5_useEAX(EEREC_S, EEREC_TEMP, (1 << (3-_Fsf_))); + vuFloat5_useEAX(EEREC_T, EEREC_TEMP, (1 << (3-_Ftf_))); + } + + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + + _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + _vuFlipRegSS_xyzw(EEREC_T, _Ftf_); + + vuFloat_useEAX(info, EEREC_TEMP, 0x8); + + x86SetJ32(bjmp32); + + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SQRT* +//------------------------------------------------------------------ +void recVUMI_SQRT( VURegs *VU, int info ) +{ + u8* pjmp; + //Console::WriteLn("recVUMI_SQRT()"); + + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, _Ftf_); + AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags + + /* Check for negative sqrt */ + SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); + AND32ItoR(EAX, 1); //Check sign + pjmp = JZ8(0); //Skip if none are + OR32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410); // Invalid Flag - Negative number sqrt + x86SetJ8(pjmp); + + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // Do a cardinal sqrt + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp infinities (only need to do positive clamp since EEREC_TEMP is positive) + SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// RSQRT* +//------------------------------------------------------------------ +PCSX2_ALIGNED16(u64 RSQRT_TEMP_XMM[2]); +void recVUMI_RSQRT(VURegs *VU, int info) +{ + u8 *ajmp8, *bjmp8; + u8 *qjmp1, *qjmp2; + int t1reg, t1boolean; + //Console::WriteLn("recVUMI_RSQRT()"); + + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, _Ftf_); + AND32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0xFCF); // Clear D/I flags + + /* Check for negative divide */ + SSE_MOVMSKPS_XMM_to_R32(EAX, EEREC_TEMP); + AND32ItoR(EAX, 1); //Check sign + ajmp8 = JZ8(0); //Skip if none are + OR32ItoM(VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410); // Invalid Flag - Negative number sqrt + x86SetJ8(ajmp8); + + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // Do a cardinal sqrt + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp Infinities to Fmax + SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + t1reg = _vuGetTempXMMreg(info); + if( t1reg < 0 ) { + for (t1reg = 0; ( (t1reg == EEREC_TEMP) || (t1reg == EEREC_S) ); t1reg++) + ; // Makes t1reg not be EEREC_TEMP or EEREC_S. + SSE_MOVAPS_XMM_to_M128( (uptr)&RSQRT_TEMP_XMM[0], t1reg ); // backup data in t1reg to a temp address + t1boolean = 1; + } + else t1boolean = 0; + + // Ft can still be zero here! so we need to check if its zero and set the correct flag. + SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg + SSE_CMPEQSS_XMM_to_XMM(t1reg, EEREC_TEMP); // Set all F's if each vector is zero + + SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation + + AND32ItoR( EAX, 0x01 ); // Grab "Is Zero" bits from the previous calculation + ajmp8 = JZ8(0); // Skip if none are + + //check for 0/0 + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + + SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear EEREC_TEMP + SSE_CMPEQPS_XMM_to_XMM(t1reg, EEREC_TEMP); // Set all F's if each vector is zero + SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation + + AND32ItoR( EAX, 0x01 ); // Grab "Is Zero" bits from the previous calculation + qjmp1 = JZ8(0); + OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x410 ); // Set invalid flag (0/0) + qjmp2 = JMP8(0); + x86SetJ8(qjmp1); + OR32ItoM( VU_VI_ADDR(REG_STATUS_FLAG, 2), 0x820 ); // Zero divide (only when not 0/0) + x86SetJ8(qjmp2); + + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)&g_maxvals[0]); // If division by zero, then EEREC_TEMP = +/- fmax + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), EEREC_TEMP); + bjmp8 = JMP8(0); + x86SetJ8(ajmp8); + + _unpackVFSS_xyzw(t1reg, EEREC_S, _Fsf_); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat_useEAX(info, t1reg, 0x8); // Clamp Infinities + SSE_DIVSS_XMM_to_XMM(t1reg, EEREC_TEMP); + vuFloat_useEAX(info, t1reg, 0x8); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_Q, 0), t1reg); + + x86SetJ8(bjmp8); + + if (t1boolean) SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)&RSQRT_TEMP_XMM[0] ); // restore t1reg data + else _freeXMMreg(t1reg); // free t1reg +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// _addISIMMtoIT() - Used in IADDI, IADDIU, and ISUBIU instructions +//------------------------------------------------------------------ +void _addISIMMtoIT(VURegs *VU, s16 imm, int info) +{ + int isreg = -1, itreg; + if (_It_ == 0) return; + + if( _Is_ == 0 ) { + itreg = ALLOCVI(_It_, MODE_WRITE); + MOV32ItoR(itreg, imm&0xffff); + return; + } + + ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE); + + if ( _It_ == _Is_ ) { + if (imm != 0 ) ADD16ItoR(itreg, imm); + } + else { + if( imm ) { + LEA32RtoR(itreg, isreg, imm); + MOVZX32R16toR(itreg, itreg); + } + else MOV32RtoR(itreg, isreg); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// IADDI +//------------------------------------------------------------------ +void recVUMI_IADDI(VURegs *VU, int info) +{ + s16 imm; + + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_IADDI"); + imm = ( VU->code >> 6 ) & 0x1f; + imm = ( imm & 0x10 ? 0xfff0 : 0) | ( imm & 0xf ); + _addISIMMtoIT(VU, imm, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// IADDIU +//------------------------------------------------------------------ +void recVUMI_IADDIU(VURegs *VU, int info) +{ + s16 imm; + + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_IADDIU"); + imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); + _addISIMMtoIT(VU, imm, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// IADD +//------------------------------------------------------------------ +void recVUMI_IADD( VURegs *VU, int info ) +{ + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; + //Console::WriteLn("recVUMI_IADD"); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); + return; + } + + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); + + if ( _Is_ == 0 ) + { + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); + } + else if ( _It_ == 0 ) + { + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); + } + else { + //ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); + + if( idreg == isreg ) ADD32RtoR(idreg, itreg); + else if( idreg == itreg ) ADD32RtoR(idreg, isreg); + else LEA32RRtoR(idreg, isreg, itreg); + MOVZX32R16toR(idreg, idreg); // needed since don't know if idreg's upper bits are 0 + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// IAND +//------------------------------------------------------------------ +void recVUMI_IAND( VURegs *VU, int info ) +{ + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; + //Console::WriteLn("recVUMI_IAND"); + if ( ( _Is_ == 0 ) || ( _It_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); + return; + } + + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); + + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); + + if( idreg == isreg ) AND16RtoR(idreg, itreg); + else if( idreg == itreg ) AND16RtoR(idreg, isreg); + else { + MOV32RtoR(idreg, itreg); + AND32RtoR(idreg, isreg); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// IOR +//------------------------------------------------------------------ +void recVUMI_IOR( VURegs *VU, int info ) +{ + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; + //Console::WriteLn("recVUMI_IOR"); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); + return; + } + + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); + + if ( _Is_ == 0 ) + { + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); + } + else if ( _It_ == 0 ) + { + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); + } + else + { + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); + + if( idreg == isreg ) OR16RtoR(idreg, itreg); + else if( idreg == itreg ) OR16RtoR(idreg, isreg); + else { + MOV32RtoR(idreg, isreg); + OR32RtoR(idreg, itreg); + } + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ISUB +//------------------------------------------------------------------ +void recVUMI_ISUB( VURegs *VU, int info ) +{ + int idreg, isreg = -1, itreg = -1; + if ( _Id_ == 0 ) return; + //Console::WriteLn("recVUMI_ISUB"); + if ( ( _It_ == 0 ) && ( _Is_ == 0 ) ) { + idreg = ALLOCVI(_Id_, MODE_WRITE); + XOR32RtoR(idreg, idreg); + return; + } + + ADD_VI_NEEDED(_Is_); + ADD_VI_NEEDED(_It_); + idreg = ALLOCVI(_Id_, MODE_WRITE); + + if ( _Is_ == 0 ) + { + if( (itreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, MODE_READ)) >= 0 ) { + if( idreg != itreg ) MOV32RtoR(idreg, itreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_It_, 1)); + NEG16R(idreg); + } + else if ( _It_ == 0 ) + { + if( (isreg = _checkX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, MODE_READ)) >= 0 ) { + if( idreg != isreg ) MOV32RtoR(idreg, isreg); + } + else MOVZX32M16toR(idreg, VU_VI_ADDR(_Is_, 1)); + } + else + { + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); + + if( idreg == isreg ) SUB16RtoR(idreg, itreg); + else if( idreg == itreg ) { + SUB16RtoR(idreg, isreg); + NEG16R(idreg); + } + else { + MOV32RtoR(idreg, isreg); + SUB16RtoR(idreg, itreg); + } + } +} +//------------------------------------------------------------------ + +//------------------------------------------------------------------ +// ISUBIU +//------------------------------------------------------------------ +void recVUMI_ISUBIU( VURegs *VU, int info ) +{ + s16 imm; + + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_ISUBIU"); + imm = ( ( VU->code >> 10 ) & 0x7800 ) | ( VU->code & 0x7ff ); + imm = -imm; + _addISIMMtoIT(VU, imm, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MOVE* +//------------------------------------------------------------------ +void recVUMI_MOVE( VURegs *VU, int info ) +{ + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + //Console::WriteLn("recVUMI_MOVE"); + if (_X_Y_Z_W == 0x8) SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_S); + else if (_X_Y_Z_W == 0xf) SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MFIR* +//------------------------------------------------------------------ +void recVUMI_MFIR( VURegs *VU, int info ) +{ + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + //Console::WriteLn("recVUMI_MFIR"); + _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _Is_, 1); + + if( _XYZW_SS ) { + SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); + _vuFlipRegSS(VU, EEREC_T); + SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); + SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_T); + } + else if (_X_Y_Z_W != 0xf) { + SSE2_MOVD_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(_Is_, 1)-2); + SSE2_PSRAD_I8_to_XMM(EEREC_TEMP, 16); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + SSE2_MOVD_M32_to_XMM(EEREC_T, VU_VI_ADDR(_Is_, 1)-2); + SSE2_PSRAD_I8_to_XMM(EEREC_T, 16); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MTIR* +//------------------------------------------------------------------ +void recVUMI_MTIR( VURegs *VU, int info ) +{ + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_MTIR"); + _deleteX86reg(X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), _It_, 2); + + if( _Fsf_ == 0 ) { + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_S); + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(_It_, 0), EEREC_TEMP); + } + + AND32ItoM(VU_VI_ADDR(_It_, 0), 0xffff); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MR32* +//------------------------------------------------------------------ +void recVUMI_MR32( VURegs *VU, int info ) +{ + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + //Console::WriteLn("recVUMI_MR32"); + if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x39); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x39); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// _loadEAX() +// +// NOTE: If x86reg < 0, reads directly from offset +//------------------------------------------------------------------ +void _loadEAX(VURegs *VU, int x86reg, uptr offset, int info) +{ + assert( offset < 0x80000000 ); + + if( x86reg >= 0 ) { + switch(_X_Y_Z_W) { + case 3: // ZW + SSE_MOVHPS_Rm_to_XMM(EEREC_T, x86reg, offset+8); + break; + case 6: // YZ + SSE_SHUFPS_Rm_to_XMM(EEREC_T, x86reg, offset, 0x9c); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x78); + break; + + case 8: // X + SSE_MOVSS_Rm_to_XMM(EEREC_TEMP, x86reg, offset); + SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + break; + case 9: // XW + SSE_SHUFPS_Rm_to_XMM(EEREC_T, x86reg, offset, 0xc9); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xd2); + break; + case 12: // XY + SSE_MOVLPS_Rm_to_XMM(EEREC_T, x86reg, offset); + break; + case 15: + if( VU == &VU1 ) SSE_MOVAPSRmtoR(EEREC_T, x86reg, offset); + else SSE_MOVUPSRmtoR(EEREC_T, x86reg, offset); + break; + default: + if( VU == &VU1 ) SSE_MOVAPSRmtoR(EEREC_TEMP, x86reg, offset); + else SSE_MOVUPSRmtoR(EEREC_TEMP, x86reg, offset); + + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + break; + } + } + else { + switch(_X_Y_Z_W) { + case 3: // ZW + SSE_MOVHPS_M64_to_XMM(EEREC_T, offset+8); + break; + case 6: // YZ + SSE_SHUFPS_M128_to_XMM(EEREC_T, offset, 0x9c); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0x78); + break; + case 8: // X + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, offset); + SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + break; + case 9: // XW + SSE_SHUFPS_M128_to_XMM(EEREC_T, offset, 0xc9); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xd2); + break; + case 12: // XY + SSE_MOVLPS_M64_to_XMM(EEREC_T, offset); + break; + case 15: + if( VU == &VU1 ) SSE_MOVAPS_M128_to_XMM(EEREC_T, offset); + else SSE_MOVUPS_M128_to_XMM(EEREC_T, offset); + break; + default: + if( VU == &VU1 ) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, offset); + else SSE_MOVUPS_M128_to_XMM(EEREC_TEMP, offset); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + break; + } + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// recVUTransformAddr() +//------------------------------------------------------------------ +int recVUTransformAddr(int x86reg, VURegs* VU, int vireg, int imm) +{ + u8* pjmp[2]; + if( x86reg == EAX ) { + if (imm) ADD32ItoR(x86reg, imm); + } + else { + if( imm ) LEA32RtoR(EAX, x86reg, imm); + else MOV32RtoR(EAX, x86reg); + } + + if( VU == &VU1 ) { + AND32ItoR(EAX, 0x3ff); // wrap around + SHL32ItoR(EAX, 4); + } + else { + + // VU0 has a somewhat interesting memory mapping: + // if addr >= 0x4000, reads VU1's VF regs and VI regs + // if addr < 0x4000, wrap around at 0x1000 + + CMP32ItoR(EAX, 0x400); + pjmp[0] = JL8(0); // if addr >= 0x4000, reads VU1's VF regs and VI regs + AND32ItoR(EAX, 0x43f); + pjmp[1] = JMP8(0); + x86SetJ8(pjmp[0]); + AND32ItoR(EAX, 0xff); // if addr < 0x4000, wrap around + x86SetJ8(pjmp[1]); + + SHL32ItoR(EAX, 4); // multiply by 16 (shift left by 4) + } + + return EAX; +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// LQ +//------------------------------------------------------------------ +void recVUMI_LQ(VURegs *VU, int info) +{ + s16 imm; + if ( _Ft_ == 0 ) return; + //Console::WriteLn("recVUMI_LQ"); + imm = (VU->code & 0x400) ? (VU->code & 0x3ff) | 0xfc00 : (VU->code & 0x3ff); + if (_Is_ == 0) { + _loadEAX(VU, -1, (uptr)GET_VU_MEM(VU, (u32)imm*16), info); + } + else { + int isreg = ALLOCVI(_Is_, MODE_READ); + _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem, info); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// LQD +//------------------------------------------------------------------ +void recVUMI_LQD( VURegs *VU, int info ) +{ + int isreg; + //Console::WriteLn("recVUMI_LQD"); + if ( _Is_ != 0 ) { + isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); + SUB16ItoR( isreg, 1 ); + } + + if ( _Ft_ == 0 ) return; + + if ( _Is_ == 0 ) _loadEAX(VU, -1, (uptr)VU->Mem, info); + else _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// LQI +//------------------------------------------------------------------ +void recVUMI_LQI(VURegs *VU, int info) +{ + int isreg; + //Console::WriteLn("recVUMI_LQI"); + if ( _Ft_ == 0 ) { + if( _Is_ != 0 ) { + if( (isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_WRITE|MODE_READ)) >= 0 ) { + ADD16ItoR(isreg, 1); + } + else { + ADD16ItoM( VU_VI_ADDR( _Is_, 0 ), 1 ); + } + } + return; + } + + if (_Is_ == 0) { + _loadEAX(VU, -1, (uptr)VU->Mem, info); + } + else { + isreg = ALLOCVI(_Is_, MODE_READ|MODE_WRITE); + _loadEAX(VU, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem, info); + ADD16ItoR( isreg, 1 ); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// _saveEAX() +//------------------------------------------------------------------ +void _saveEAX(VURegs *VU, int x86reg, uptr offset, int info) +{ + assert( offset < 0x80000000 ); + + if ( _Fs_ == 0 ) { + if ( _XYZW_SS ) { + u32 c = _W ? 0x3f800000 : 0; + if ( x86reg >= 0 ) MOV32ItoRm(x86reg, c, offset+(_W?12:(_Z?8:(_Y?4:0)))); + else MOV32ItoM(offset+(_W?12:(_Z?8:(_Y?4:0))), c); + } + else { + + // (this is one of my test cases for the new emitter --air) + using namespace x86Emitter; + xAddressReg thisreg( x86reg ); + + if ( _X ) xMOV(ptr32[thisreg+offset], 0x00000000); + if ( _Y ) xMOV(ptr32[thisreg+offset+4], 0x00000000); + if ( _Z ) xMOV(ptr32[thisreg+offset+8], 0x00000000); + if ( _W ) xMOV(ptr32[thisreg+offset+12], 0x3f800000); + } + return; + } + + switch ( _X_Y_Z_W ) { + case 1: // W + SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x27); + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); + else SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); + break; + case 2: // Z + SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); + else SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); + break; + case 3: // ZW + if ( x86reg >= 0 ) SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_S, offset+8); + else SSE_MOVHPS_XMM_to_M64(offset+8, EEREC_S); + break; + case 4: // Y + SSE2_PSHUFLW_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x4e); + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); + else SSE_MOVSS_XMM_to_M32(offset+4, EEREC_TEMP); + break; + case 5: // YW + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xB1); + SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if ( x86reg >= 0 ) { + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset+4); + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); + } + else { + SSE_MOVSS_XMM_to_M32(offset+4, EEREC_S); + SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); + } + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xB1); + break; + case 6: // YZ + SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0xc9); + if ( x86reg >= 0 ) SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); + else SSE_MOVLPS_XMM_to_M64(offset+4, EEREC_TEMP); + break; + case 7: // YZW + SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x93); //ZYXW + if ( x86reg >= 0 ) { + SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+4); + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); + } + else { + SSE_MOVHPS_XMM_to_M64(offset+4, EEREC_TEMP); + SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); + } + break; + case 8: // X + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); + else SSE_MOVSS_XMM_to_M32(offset, EEREC_S); + break; + case 9: // XW + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); + else SSE_MOVSS_XMM_to_M32(offset, EEREC_S); + + SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0xff); //WWWW + + if ( x86reg >= 0 ) SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); + else SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); + + break; + case 10: //XZ + SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if ( x86reg >= 0 ) { + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); + } + else { + SSE_MOVSS_XMM_to_M32(offset, EEREC_S); + SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); + } + break; + case 11: //XZW + if ( x86reg >= 0 ) { + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_S, offset); + SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_S, offset+8); + } + else { + SSE_MOVSS_XMM_to_M32(offset, EEREC_S); + SSE_MOVHPS_XMM_to_M64(offset+8, EEREC_S); + } + break; + case 12: // XY + if ( x86reg >= 0 ) SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_S, offset+0); + else SSE_MOVLPS_XMM_to_M64(offset, EEREC_S); + break; + case 13: // XYW + SSE2_PSHUFD_XMM_to_XMM(EEREC_TEMP, EEREC_S, 0x4b); //YXZW + if ( x86reg >= 0 ) { + SSE_MOVHPS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+0); + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+12); + } + else { + SSE_MOVHPS_XMM_to_M64(offset, EEREC_TEMP); + SSE_MOVSS_XMM_to_M32(offset+12, EEREC_TEMP); + } + break; + case 14: // XYZ + SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if ( x86reg >= 0 ) { + SSE_MOVLPS_XMM_to_Rm(x86reg, EEREC_S, offset+0); + SSE_MOVSS_XMM_to_Rm(x86reg, EEREC_TEMP, offset+8); + } + else { + SSE_MOVLPS_XMM_to_M64(offset, EEREC_S); + SSE_MOVSS_XMM_to_M32(offset+8, EEREC_TEMP); + } + break; + case 15: // XYZW + if ( VU == &VU1 ) { + if( x86reg >= 0 ) SSE_MOVAPSRtoRm(x86reg, EEREC_S, offset+0); + else SSE_MOVAPS_XMM_to_M128(offset, EEREC_S); + } + else { + if( x86reg >= 0 ) SSE_MOVUPSRtoRm(x86reg, EEREC_S, offset+0); + else { + if( offset & 15 ) SSE_MOVUPS_XMM_to_M128(offset, EEREC_S); + else SSE_MOVAPS_XMM_to_M128(offset, EEREC_S); + } + } + break; + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SQ +//------------------------------------------------------------------ +void recVUMI_SQ(VURegs *VU, int info) +{ + s16 imm; + //Console::WriteLn("recVUMI_SQ"); + imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); + if ( _It_ == 0 ) _saveEAX(VU, -1, (uptr)GET_VU_MEM(VU, (int)imm * 16), info); + else { + int itreg = ALLOCVI(_It_, MODE_READ); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, imm), (uptr)VU->Mem, info); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SQD +//------------------------------------------------------------------ +void recVUMI_SQD(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_SQD"); + if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); + else { + int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); + SUB16ItoR( itreg, 1 ); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SQI +//------------------------------------------------------------------ +void recVUMI_SQI(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_SQI"); + if (_It_ == 0) _saveEAX(VU, -1, (uptr)VU->Mem, info); + else { + int itreg = ALLOCVI(_It_, MODE_READ|MODE_WRITE); + _saveEAX(VU, recVUTransformAddr(itreg, VU, _It_, 0), (uptr)VU->Mem, info); + ADD16ItoR( itreg, 1 ); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ILW +//------------------------------------------------------------------ +void recVUMI_ILW(VURegs *VU, int info) +{ + int itreg; + s16 imm, off; + + if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; + //Console::WriteLn("recVUMI_ILW"); + imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); + if (_X) off = 0; + else if (_Y) off = 4; + else if (_Z) off = 8; + else if (_W) off = 12; + + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE); + + if ( _Is_ == 0 ) { + MOVZX32M16toR( itreg, (uptr)GET_VU_MEM(VU, (int)imm * 16 + off) ); + } + else { + int isreg = ALLOCVI(_Is_, MODE_READ); + MOV32RmtoR(itreg, recVUTransformAddr(isreg, VU, _Is_, imm), (uptr)VU->Mem + off); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ISW +//------------------------------------------------------------------ +void recVUMI_ISW( VURegs *VU, int info ) +{ + s16 imm; + //Console::WriteLn("recVUMI_ISW"); + imm = ( VU->code & 0x400) ? ( VU->code & 0x3ff) | 0xfc00 : ( VU->code & 0x3ff); + + if (_Is_ == 0) { + uptr off = (uptr)GET_VU_MEM(VU, (int)imm * 16); + int itreg = ALLOCVI(_It_, MODE_READ); + + if (_X) MOV32RtoM(off, itreg); + if (_Y) MOV32RtoM(off+4, itreg); + if (_Z) MOV32RtoM(off+8, itreg); + if (_W) MOV32RtoM(off+12, itreg); + } + else { + int x86reg, isreg, itreg; + + ADD_VI_NEEDED(_It_); + isreg = ALLOCVI(_Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_READ); + + x86reg = recVUTransformAddr(isreg, VU, _Is_, imm); + + if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); + if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); + if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); + if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ILWR +//------------------------------------------------------------------ +void recVUMI_ILWR( VURegs *VU, int info ) +{ + int off, itreg; + + if ( ( _It_ == 0 ) || ( _X_Y_Z_W == 0 ) ) return; + //Console::WriteLn("recVUMI_ILWR"); + if (_X) off = 0; + else if (_Y) off = 4; + else if (_Z) off = 8; + else if (_W) off = 12; + + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE); + + if ( _Is_ == 0 ) { + MOVZX32M16toR( itreg, (uptr)VU->Mem + off ); + } + else { + int isreg = ALLOCVI(_Is_, MODE_READ); + MOVZX32Rm16toR(itreg, recVUTransformAddr(isreg, VU, _Is_, 0), (uptr)VU->Mem + off); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ISWR +//------------------------------------------------------------------ +void recVUMI_ISWR( VURegs *VU, int info ) +{ + int itreg; + //Console::WriteLn("recVUMI_ISWR"); + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_READ); + + if (_Is_ == 0) { + if (_X) MOV32RtoM((uptr)VU->Mem, itreg); + if (_Y) MOV32RtoM((uptr)VU->Mem+4, itreg); + if (_Z) MOV32RtoM((uptr)VU->Mem+8, itreg); + if (_W) MOV32RtoM((uptr)VU->Mem+12, itreg); + } + else { + int x86reg; + int isreg = ALLOCVI(_Is_, MODE_READ); + x86reg = recVUTransformAddr(isreg, VU, _Is_, 0); + + if (_X) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem); + if (_Y) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+4); + if (_Z) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+8); + if (_W) MOV32RtoRm(x86reg, itreg, (uptr)VU->Mem+12); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// RINIT* +//------------------------------------------------------------------ +void recVUMI_RINIT(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_RINIT()"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode & MODE_NOFLUSH) ) { + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 2); + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)s_mask); + SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)VU_ONE); + SSE_MOVSS_XMM_to_M32(VU_REGR_ADDR, EEREC_TEMP); + } + else { + int rreg = ALLOCVI(REG_R, MODE_WRITE); + + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + MOV32MtoR( rreg, VU_VFx_ADDR( _Fs_ ) + 4 * _Fsf_ ); + AND32ItoR( rreg, 0x7fffff ); + OR32ItoR( rreg, 0x7f << 23 ); + + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// RGET* +//------------------------------------------------------------------ +void recVUMI_RGET(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_RGET()"); + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); + + if (_X_Y_Z_W != 0xf) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_REGR_ADDR); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_T, VU_REGR_ADDR); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// RNEXT* +//------------------------------------------------------------------ +void recVUMI_RNEXT( VURegs *VU, int info ) +{ + int rreg, x86temp0, x86temp1; + //Console::WriteLn("recVUMI_RNEXT()"); + + rreg = ALLOCVI(REG_R, MODE_WRITE|MODE_READ); + + x86temp0 = ALLOCTEMPX86(0); + x86temp1 = ALLOCTEMPX86(0); + + // code from www.project-fao.org + //MOV32MtoR(rreg, VU_REGR_ADDR); + MOV32RtoR(x86temp0, rreg); + SHR32ItoR(x86temp0, 4); + AND32ItoR(x86temp0, 1); + + MOV32RtoR(x86temp1, rreg); + SHR32ItoR(x86temp1, 22); + AND32ItoR(x86temp1, 1); + + SHL32ItoR(rreg, 1); + XOR32RtoR(x86temp0, x86temp1); + XOR32RtoR(rreg, x86temp0); + AND32ItoR(rreg, 0x7fffff); + OR32ItoR(rreg, 0x3f800000); + + _freeX86reg(x86temp0); + _freeX86reg(x86temp1); + + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) { + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); + return; + } + + recVUMI_RGET(VU, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// RXOR* +//------------------------------------------------------------------ +void recVUMI_RXOR( VURegs *VU, int info ) +{ + //Console::WriteLn("recVUMI_RXOR()"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode & MODE_NOFLUSH) ) { + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + + SSE_XORPS_M128_to_XMM(EEREC_TEMP, VU_REGR_ADDR); + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)s_mask); + SSE_ORPS_M128_to_XMM(EEREC_TEMP, (uptr)s_fones); + SSE_MOVSS_XMM_to_M32(VU_REGR_ADDR, EEREC_TEMP); + } + else { + int rreg = ALLOCVI(REG_R, MODE_WRITE|MODE_READ); + + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + XOR32MtoR( rreg, VU_VFx_ADDR( _Fs_ ) + 4 * _Fsf_ ); + AND32ItoR( rreg, 0x7fffff ); + OR32ItoR ( rreg, 0x3f800000 ); + + _deleteX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), REG_R, 1); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// WAITQ +//------------------------------------------------------------------ +void recVUMI_WAITQ( VURegs *VU, int info ) +{ + //Console::WriteLn("recVUMI_WAITQ"); +// if( info & PROCESS_VU_SUPER ) { +// //CALLFunc(waitqfn); +// SuperVUFlush(0, 1); +// } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FSAND +//------------------------------------------------------------------ +void recVUMI_FSAND( VURegs *VU, int info ) +{ + int itreg; + u16 imm; + //Console::WriteLn("recVUMI_FSAND"); + imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); + if(_It_ == 0) return; + + itreg = ALLOCVI(_It_, MODE_WRITE); + MOV32MtoR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); + AND32ItoR( itreg, imm ); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FSEQ +//------------------------------------------------------------------ +void recVUMI_FSEQ( VURegs *VU, int info ) +{ + int itreg; + u16 imm; + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_FSEQ"); + imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); + + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); + + MOVZX32M16toR( EAX, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); + XOR32RtoR(itreg, itreg); + CMP16ItoR(EAX, imm); + SETE8R(itreg); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FSOR +//------------------------------------------------------------------ +void recVUMI_FSOR( VURegs *VU, int info ) +{ + int itreg; + u32 imm; + if(_It_ == 0) return; + //Console::WriteLn("recVUMI_FSOR"); + imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7ff); + + itreg = ALLOCVI(_It_, MODE_WRITE); + + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_STATUS_FLAG, 1) ); + OR32ItoR( itreg, imm ); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FSSET +//------------------------------------------------------------------ +void recVUMI_FSSET(VURegs *VU, int info) +{ + u32 writeaddr = VU_VI_ADDR(REG_STATUS_FLAG, 0); + u32 prevaddr = VU_VI_ADDR(REG_STATUS_FLAG, 2); + + u16 imm = 0; + //Console::WriteLn("recVUMI_FSSET"); + imm = (((VU->code >> 21 ) & 0x1) << 11) | (VU->code & 0x7FF); + + // keep the low 6 bits ONLY if the upper instruction is an fmac instruction (otherwise rewrite) - metal gear solid 3 + //if( (info & PROCESS_VU_SUPER) && VUREC_FMAC ) { + MOV32MtoR(EAX, prevaddr); + AND32ItoR(EAX, 0x3f); + if ((imm&0xfc0) != 0) OR32ItoR(EAX, imm & 0xFC0); + MOV32RtoM(writeaddr ? writeaddr : prevaddr, EAX); + //} + //else { + // MOV32ItoM(writeaddr ? writeaddr : prevaddr, imm&0xfc0); + //} +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FMAND +//------------------------------------------------------------------ +void recVUMI_FMAND( VURegs *VU, int info ) +{ + int isreg, itreg; + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_FMAND"); + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); + + if( isreg >= 0 ) { + if( itreg != isreg ) MOV32RtoR(itreg, isreg); + } + else MOVZX32M16toR(itreg, VU_VI_ADDR(_Is_, 1)); + + AND16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FMEQ +//------------------------------------------------------------------ +void recVUMI_FMEQ( VURegs *VU, int info ) +{ + int itreg, isreg; + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_FMEQ"); + if( _It_ == _Is_ ) { + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG + + CMP16MtoR(itreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); + SETE8R(EAX); + MOVZX32R8toR(itreg, EAX); + } + else { + ADD_VI_NEEDED(_Is_); + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_8BITREG); + isreg = ALLOCVI(_Is_, MODE_READ); + + XOR32RtoR(itreg, itreg); + + CMP16MtoR(isreg, VU_VI_ADDR(REG_MAC_FLAG, 1)); + SETE8R(itreg); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FMOR +//------------------------------------------------------------------ +void recVUMI_FMOR( VURegs *VU, int info ) +{ + int isreg, itreg; + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_FMOR"); + if( _Is_ == 0 ) { + itreg = ALLOCVI(_It_, MODE_WRITE);//|MODE_8BITREG); + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + } + else if( _It_ == _Is_ ) { + itreg = ALLOCVI(_It_, MODE_WRITE|MODE_READ);//|MODE_8BITREG); + OR16MtoR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + } + else { + isreg = _checkX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), _Is_, MODE_READ); + itreg = ALLOCVI(_It_, MODE_WRITE); + + MOVZX32M16toR( itreg, VU_VI_ADDR(REG_MAC_FLAG, 1) ); + + if( isreg >= 0 ) + OR16RtoR( itreg, isreg ); + else + OR16MtoR( itreg, VU_VI_ADDR(_Is_, 1) ); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FCAND +//------------------------------------------------------------------ +void recVUMI_FCAND( VURegs *VU, int info ) +{ + int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); + //Console::WriteLn("recVUMI_FCAND"); + MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); + XOR32RtoR( itreg, itreg ); + AND32ItoR( EAX, VU->code & 0xFFFFFF ); + + SETNZ8R(itreg); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FCEQ +//------------------------------------------------------------------ +void recVUMI_FCEQ( VURegs *VU, int info ) +{ + int itreg = ALLOCVI(1, MODE_WRITE|MODE_8BITREG); + //Console::WriteLn("recVUMI_FCEQ"); + MOV32MtoR( EAX, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); + AND32ItoR( EAX, 0xffffff ); + XOR32RtoR( itreg, itreg ); + CMP32ItoR( EAX, VU->code&0xffffff ); + + SETE8R(itreg); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FCOR +//------------------------------------------------------------------ +void recVUMI_FCOR( VURegs *VU, int info ) +{ + int itreg; + //Console::WriteLn("recVUMI_FCOR"); + itreg = ALLOCVI(1, MODE_WRITE); + MOV32MtoR( itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1) ); + OR32ItoR ( itreg, VU->code ); + AND32ItoR( itreg, 0xffffff ); + ADD32ItoR( itreg, 1 ); // If 24 1's will make 25th bit 1, else 0 + SHR32ItoR( itreg, 24 ); // Get the 25th bit (also clears the rest of the garbage in the reg) +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FCSET +//------------------------------------------------------------------ +void recVUMI_FCSET( VURegs *VU, int info ) +{ + u32 addr = VU_VI_ADDR(REG_CLIP_FLAG, 0); + //Console::WriteLn("recVUMI_FCSET"); + MOV32ItoM(addr ? addr : VU_VI_ADDR(REG_CLIP_FLAG, 2), VU->code&0xffffff ); + + if( !(info & (PROCESS_VU_SUPER|PROCESS_VU_COP2)) ) + MOV32ItoM( VU_VI_ADDR(REG_CLIP_FLAG, 1), VU->code&0xffffff ); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FCGET +//------------------------------------------------------------------ +void recVUMI_FCGET( VURegs *VU, int info ) +{ + int itreg; + if(_It_ == 0) return; + //Console::WriteLn("recVUMI_FCGET"); + itreg = ALLOCVI(_It_, MODE_WRITE); + + MOV32MtoR(itreg, VU_VI_ADDR(REG_CLIP_FLAG, 1)); + AND32ItoR(itreg, 0x0fff); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// _recbranchAddr() +// +// NOTE: Due to static var dependencies, several SuperVU branch instructions +// are still located in iVUzerorec.cpp. +//------------------------------------------------------------------ + +//------------------------------------------------------------------ +// MFP* +//------------------------------------------------------------------ +void recVUMI_MFP(VURegs *VU, int info) +{ + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + //Console::WriteLn("recVUMI_MFP"); + if( _XYZW_SS ) { + _vuFlipRegSS(VU, EEREC_T); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 1)); + SSE_MOVSS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_T); + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 1)); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_T, VU_VI_ADDR(REG_P, 1)); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// WAITP +//------------------------------------------------------------------ +static PCSX2_ALIGNED16(float s_tempmem[4]); +void recVUMI_WAITP(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_WAITP"); +// if( info & PROCESS_VU_SUPER ) +// SuperVUFlush(1, 1); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// vuSqSumXYZ()* +// +// NOTE: In all EFU insts, EEREC_D is a temp reg +//------------------------------------------------------------------ +void vuSqSumXYZ(int regd, int regs, int regtemp) // regd.x = x ^ 2 + y ^ 2 + z ^ 2 +{ + //Console::WriteLn("VU: SUMXYZ"); + if( cpucaps.hasStreamingSIMD4Extensions ) + { + SSE_MOVAPS_XMM_to_XMM(regd, regs); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat2(regd, regtemp, 0xf); + SSE4_DPPS_XMM_to_XMM(regd, regd, 0x71); + } + else + { + SSE_MOVAPS_XMM_to_XMM(regtemp, regs); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat2(regtemp, regd, 0xf); + SSE_MULPS_XMM_to_XMM(regtemp, regtemp); // xyzw ^ 2 + + if( cpucaps.hasStreamingSIMD3Extensions ) { + SSE3_HADDPS_XMM_to_XMM(regd, regtemp); + SSE_ADDPS_XMM_to_XMM(regd, regtemp); // regd.z = x ^ 2 + y ^ 2 + z ^ 2 + SSE_MOVHLPS_XMM_to_XMM(regd, regd); // regd.x = regd.z + } + else { + SSE_MOVSS_XMM_to_XMM(regd, regtemp); + SSE2_PSHUFLW_XMM_to_XMM(regtemp, regtemp, 0x4e); // wzyx -> wzxy + SSE_ADDSS_XMM_to_XMM(regd, regtemp); // x ^ 2 + y ^ 2 + SSE_SHUFPS_XMM_to_XMM(regtemp, regtemp, 0xD2); // wzxy -> wxyz + SSE_ADDSS_XMM_to_XMM(regd, regtemp); // x ^ 2 + y ^ 2 + z ^ 2 + } + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ESADD* +//------------------------------------------------------------------ +void recVUMI_ESADD( VURegs *VU, int info) +{ + //Console::WriteLn("VU: ESADD"); + assert( VU == &VU1 ); + if( EEREC_TEMP == EEREC_D ) { // special code to reset P ( FixMe: don't know if this is still needed! (cottonvibes) ) + Console::Notice("ESADD: Resetting P reg!!!\n"); + MOV32ItoM(VU_VI_ADDR(REG_P, 0), 0); + return; + } + vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_D); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ERSADD* +//------------------------------------------------------------------ +void recVUMI_ERSADD( VURegs *VU, int info ) +{ + //Console::WriteLn("VU: ERSADD"); + assert( VU == &VU1 ); + vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); + // don't use RCPSS (very bad precision) + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ELENG* +//------------------------------------------------------------------ +void recVUMI_ELENG( VURegs *VU, int info ) +{ + //Console::WriteLn("VU: ELENG"); + assert( VU == &VU1 ); + vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive + SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_D); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ERLENG* +//------------------------------------------------------------------ +void recVUMI_ERLENG( VURegs *VU, int info ) +{ + //Console::WriteLn("VU: ERLENG"); + assert( VU == &VU1 ); + vuSqSumXYZ(EEREC_D, EEREC_S, EEREC_TEMP); + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_D, (uptr)g_maxvals); // Only need to do positive clamp since (x ^ 2 + y ^ 2 + z ^ 2) is positive + SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D); // regd <- sqrt(x^2 + y^2 + z^2) + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); // temp = 1 / sqrt(x^2 + y^2 + z^2) + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// EATANxy +//------------------------------------------------------------------ +void recVUMI_EATANxy( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + //Console::WriteLn("recVUMI_EATANxy"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { + SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); + FLD32((uptr)&s_tempmem[0]); + FLD32((uptr)&s_tempmem[1]); + } + else { + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + FLD32((uptr)&VU->VF[_Fs_].UL[0]); + FLD32((uptr)&VU->VF[_Fs_].UL[1]); + } + + FPATAN(); + FSTP32(VU_VI_ADDR(REG_P, 0)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// EATANxz +//------------------------------------------------------------------ +void recVUMI_EATANxz( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + //Console::WriteLn("recVUMI_EATANxz"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { + SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); + FLD32((uptr)&s_tempmem[0]); + FLD32((uptr)&s_tempmem[2]); + } + else { + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + FLD32((uptr)&VU->VF[_Fs_].UL[0]); + FLD32((uptr)&VU->VF[_Fs_].UL[2]); + } + FPATAN(); + FSTP32(VU_VI_ADDR(REG_P, 0)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ESUM* +//------------------------------------------------------------------ +void recVUMI_ESUM( VURegs *VU, int info ) +{ + //Console::WriteLn("VU: ESUM"); + assert( VU == &VU1 ); + + if( cpucaps.hasStreamingSIMD3Extensions ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat_useEAX(info, EEREC_TEMP, 0xf); + SSE3_HADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + SSE3_HADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + } + else { + SSE_MOVHLPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // z, w, z, w + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); // z+x, w+y, z+z, w+w + SSE_UNPCKLPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // z+x, z+x, w+y, w+y + SSE_MOVHLPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); // w+y, w+y, w+y, w+y + SSE_ADDSS_XMM_to_XMM(EEREC_TEMP, EEREC_D); // x+y+z+w, w+y, w+y, w+y + } + + vuFloat_useEAX(info, EEREC_TEMP, 8); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ERCPR* +//------------------------------------------------------------------ +void recVUMI_ERCPR( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + //Console::WriteLn("VU1: ERCPR"); + + // don't use RCPSS (very bad precision) + switch ( _Fsf_ ) { + case 0: //0001 + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + break; + case 1: //0010 + SSE2_PSHUFLW_XMM_to_XMM(EEREC_S, EEREC_S, 0x4e); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE2_PSHUFLW_XMM_to_XMM(EEREC_S, EEREC_S, 0x4e); + break; + case 2: //0100 + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xc6); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0xc6); + break; + case 3: //1000 + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0x27); + if (CHECK_VU_EXTRA_OVERFLOW) vuFloat5_useEAX(EEREC_S, EEREC_TEMP, 8); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); // temp <- 1 + SSE_DIVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SHUFPS_XMM_to_XMM(EEREC_S, EEREC_S, 0x27); + break; + } + + vuFloat_useEAX(info, EEREC_TEMP, 8); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ESQRT* +//------------------------------------------------------------------ +void recVUMI_ESQRT( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + + //Console::WriteLn("VU1: ESQRT"); + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // abs(x) + if (CHECK_VU_OVERFLOW) SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Only need to do positive clamp + SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ERSQRT* +//------------------------------------------------------------------ +void recVUMI_ERSQRT( VURegs *VU, int info ) +{ + int t1reg = _vuGetTempXMMreg(info); + + assert( VU == &VU1 ); + //Console::WriteLn("VU1: ERSQRT"); + + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_S, _Fsf_); + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)const_clip); // abs(x) + SSE_MINSS_M32_to_XMM(EEREC_TEMP, (uptr)g_maxvals); // Clamp Infinities to Fmax + SSE_SQRTSS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // SQRT(abs(x)) + + if( t1reg >= 0 ) + { + SSE_MOVSS_M32_to_XMM(t1reg, (uptr)VU_ONE); + SSE_DIVSS_XMM_to_XMM(t1reg, EEREC_TEMP); + vuFloat_useEAX(info, t1reg, 8); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), t1reg); + _freeXMMreg(t1reg); + } + else + { + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)VU_ONE); + SSE_DIVSS_M32_to_XMM(EEREC_TEMP, VU_VI_ADDR(REG_P, 0)); + vuFloat_useEAX(info, EEREC_TEMP, 8); + SSE_MOVSS_XMM_to_M32(VU_VI_ADDR(REG_P, 0), EEREC_TEMP); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ESIN +//------------------------------------------------------------------ +void recVUMI_ESIN( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + + //Console::WriteLn("recVUMI_ESIN"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { + switch(_Fsf_) { + case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); + case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); + default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); + } + FLD32((uptr)&s_tempmem[_Fsf_]); + } + else { + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + FLD32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); + } + + FSIN(); + FSTP32(VU_VI_ADDR(REG_P, 0)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// EATAN +//------------------------------------------------------------------ +void recVUMI_EATAN( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + + //Console::WriteLn("recVUMI_EATAN"); + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { + switch(_Fsf_) { + case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); + case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); + default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); + } + FLD32((uptr)&s_tempmem[_Fsf_]); + } + else { + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + } + + FLD1(); + FLD32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); + FPATAN(); + FSTP32(VU_VI_ADDR(REG_P, 0)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// EEXP +//------------------------------------------------------------------ +void recVUMI_EEXP( VURegs *VU, int info ) +{ + assert( VU == &VU1 ); + //Console::WriteLn("recVUMI_EEXP"); + FLDL2E(); + + if( (xmmregs[EEREC_S].mode & MODE_WRITE) && (xmmregs[EEREC_S].mode&MODE_NOFLUSH) ) { + switch(_Fsf_) { + case 0: SSE_MOVSS_XMM_to_M32((uptr)s_tempmem, EEREC_S); + case 1: SSE_MOVLPS_XMM_to_M64((uptr)s_tempmem, EEREC_S); + default: SSE_MOVHPS_XMM_to_M64((uptr)&s_tempmem[2], EEREC_S); + } + FMUL32((uptr)&s_tempmem[_Fsf_]); + } + else { + if( xmmregs[EEREC_S].mode & MODE_WRITE ) { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[_Fs_], EEREC_S); + xmmregs[EEREC_S].mode &= ~MODE_WRITE; + } + + FMUL32((uptr)&VU->VF[_Fs_].UL[_Fsf_]); + } + + // basically do 2^(log_2(e) * val) + FLD(0); + FRNDINT(); + FXCH(1); + FSUB32Rto0(1); + F2XM1(); + FLD1(); + FADD320toR(1); + FSCALE(); + FSTP(1); + + FSTP32(VU_VI_ADDR(REG_P, 0)); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// XITOP +//------------------------------------------------------------------ +void recVUMI_XITOP( VURegs *VU, int info ) +{ + int itreg; + if (_It_ == 0) return; + //Console::WriteLn("recVUMI_XITOP"); + itreg = ALLOCVI(_It_, MODE_WRITE); + MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->itop ); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// XTOP +//------------------------------------------------------------------ +void recVUMI_XTOP( VURegs *VU, int info ) +{ + int itreg; + if ( _It_ == 0 ) return; + //Console::WriteLn("recVUMI_XTOP"); + itreg = ALLOCVI(_It_, MODE_WRITE); + MOVZX32M16toR( itreg, (uptr)&VU->vifRegs->top ); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// VU1XGKICK_MTGSTransfer() - Called by ivuZerorec.cpp +//------------------------------------------------------------------ +void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr) +{ + u32 size; + u32* data = (u32*)((u8*)pMem + (addr&0x3fff)); + + // fixme: The gifTagDummy function in the MTGS (called by PrepDataPacket) has a + // hack that aborts the packet if it goes past the end of VU1 memory. + // Chances are this should be a "loops around memory" situation, and the packet + // should be continued starting at addr zero (0). + + size = mtgsThread->PrepDataPacket( GIF_PATH_1, data, (0x4000-(addr&0x3fff)) >> 4); + jASSUME( size > 0 ); + + //if( size > 0 ) + { + u8* pmem = mtgsThread->GetDataPacketPtr(); + + /* if((size << 4) > (0x4000-(addr&0x3fff))) + { + //DevCon::Notice("addr + Size = 0x%x, transferring %x then doing %x", params (addr&0x3fff) + (size << 4), (0x4000-(addr&0x3fff)) >> 4, size - ((0x4000-(addr&0x3fff)) >> 4)); + memcpy_aligned(pmem, (u8*)pMem+addr, 0x4000-(addr&0x3fff)); + size -= (0x4000-(addr&0x3fff)) >> 4; + //DevCon::Notice("Size left %x", params size); + pmem += 0x4000-(addr&0x3fff); + memcpy_aligned(pmem, (u8*)pMem, size<<4); + } + else + {*/ + memcpy_aligned(pmem, (u8*)pMem+addr, size<<4); + // } + mtgsThread->SendDataPacket(); + } +} +//------------------------------------------------------------------ diff --git a/pcsx2/x86/iVUmicro.cpp b/pcsx2/x86/sVU_Micro.cpp similarity index 96% rename from pcsx2/x86/iVUmicro.cpp rename to pcsx2/x86/sVU_Micro.cpp index 7738a52da9..89ddd307e2 100644 --- a/pcsx2/x86/iVUmicro.cpp +++ b/pcsx2/x86/sVU_Micro.cpp @@ -1,1740 +1,1740 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include "PrecompiledHeader.h" - -#include "Common.h" -#include "GS.h" -#include "R5900OpcodeTables.h" -#include "iR5900.h" -#include "iMMI.h" -#include "iFPU.h" -#include "iCOP0.h" -#include "VUmicro.h" -#include "VUflags.h" -#include "iVUmicro.h" -#include "iVUops.h" -#include "iVUzerorec.h" - -#ifdef _WIN32 -#pragma warning(disable:4244) -#pragma warning(disable:4761) -#endif -//------------------------------------------------------------------ - -// fixme - VUmicro should really use its own static vars for pc and branch. -// Sharing with the EE's copies of pc and branch is not cool! (air) - -//------------------------------------------------------------------ -// Helper Macros -//------------------------------------------------------------------ -#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register -#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register -#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register -#define _It_ (_Ft_ & 15) -#define _Is_ (_Fs_ & 15) -#define _Id_ (_Fd_ & 15) - -#define _X (( VU->code>>24) & 0x1) -#define _Y (( VU->code>>23) & 0x1) -#define _Z (( VU->code>>22) & 0x1) -#define _W (( VU->code>>21) & 0x1) - -#define _XYZW_SS (_X+_Y+_Z+_W==1) - -#define _Fsf_ (( VU->code >> 21) & 0x03) -#define _Ftf_ (( VU->code >> 23) & 0x03) - -#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) -#define _UImm11_ (s32)(VU->code & 0x7ff) - -#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] -#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] -#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] -#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] - -#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] -#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] -#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] - -#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) - -#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] -#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] -#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] -#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] - -#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// Global Variables -//------------------------------------------------------------------ -int vucycle; - -PCSX2_ALIGNED16(const float s_fones[8]) = {1.0f, 1.0f, 1.0f, 1.0f, -1.0f, -1.0f, -1.0f, -1.0f}; -PCSX2_ALIGNED16(const u32 s_mask[4]) = {0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff}; -PCSX2_ALIGNED16(const u32 s_expmask[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; -PCSX2_ALIGNED16(const u32 g_minvals[4]) = {0xff7fffff, 0xff7fffff, 0xff7fffff, 0xff7fffff}; -PCSX2_ALIGNED16(const u32 g_maxvals[4]) = {0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff}; -PCSX2_ALIGNED16(const u32 const_clip[8]) = {0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, - 0x80000000, 0x80000000, 0x80000000, 0x80000000}; -PCSX2_ALIGNED(64, const u32 g_ones[4]) = {0x00000001, 0x00000001, 0x00000001, 0x00000001}; -PCSX2_ALIGNED16(const u32 g_minvals_XYZW[16][4]) = -{ - { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //0000 - { 0xffffffff, 0xffffffff, 0xffffffff, 0xff7fffff }, //0001 - { 0xffffffff, 0xffffffff, 0xff7fffff, 0xffffffff }, //0010 - { 0xffffffff, 0xffffffff, 0xff7fffff, 0xff7fffff }, //0011 - { 0xffffffff, 0xff7fffff, 0xffffffff, 0xffffffff }, //0100 - { 0xffffffff, 0xff7fffff, 0xffffffff, 0xff7fffff }, //0101 - { 0xffffffff, 0xff7fffff, 0xff7fffff, 0xffffffff }, //0110 - { 0xffffffff, 0xff7fffff, 0xff7fffff, 0xff7fffff }, //0111 - { 0xff7fffff, 0xffffffff, 0xffffffff, 0xffffffff }, //1000 - { 0xff7fffff, 0xffffffff, 0xffffffff, 0xff7fffff }, //1001 - { 0xff7fffff, 0xffffffff, 0xff7fffff, 0xffffffff }, //1010 - { 0xff7fffff, 0xffffffff, 0xff7fffff, 0xff7fffff }, //1011 - { 0xff7fffff, 0xff7fffff, 0xffffffff, 0xffffffff }, //1100 - { 0xff7fffff, 0xff7fffff, 0xffffffff, 0xff7fffff }, //1101 - { 0xff7fffff, 0xff7fffff, 0xff7fffff, 0xffffffff }, //1110 - { 0xff7fffff, 0xff7fffff, 0xff7fffff, 0xff7fffff }, //1111 -}; -PCSX2_ALIGNED16(const u32 g_maxvals_XYZW[16][4])= -{ - { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //0000 - { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7f7fffff }, //0001 - { 0x7fffffff, 0x7fffffff, 0x7f7fffff, 0x7fffffff }, //0010 - { 0x7fffffff, 0x7fffffff, 0x7f7fffff, 0x7f7fffff }, //0011 - { 0x7fffffff, 0x7f7fffff, 0x7fffffff, 0x7fffffff }, //0100 - { 0x7fffffff, 0x7f7fffff, 0x7fffffff, 0x7f7fffff }, //0101 - { 0x7fffffff, 0x7f7fffff, 0x7f7fffff, 0x7fffffff }, //0110 - { 0x7fffffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff }, //0111 - { 0x7f7fffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //1000 - { 0x7f7fffff, 0x7fffffff, 0x7fffffff, 0x7f7fffff }, //1001 - { 0x7f7fffff, 0x7fffffff, 0x7f7fffff, 0x7fffffff }, //1010 - { 0x7f7fffff, 0x7fffffff, 0x7f7fffff, 0x7f7fffff }, //1011 - { 0x7f7fffff, 0x7f7fffff, 0x7fffffff, 0x7fffffff }, //1100 - { 0x7f7fffff, 0x7f7fffff, 0x7fffffff, 0x7f7fffff }, //1101 - { 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7fffffff }, //1110 - { 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff }, //1111 -}; -//------------------------------------------------------------------ - -//------------------------------------------------------------------ -// VU Pipeline/Test Stalls/Analyzing Functions -//------------------------------------------------------------------ -void _recvuFMACflush(VURegs * VU, bool intermediate) { - int i; - - for (i=0; i<8; i++) { - if (VU->fmac[i].enable == 0) continue; - - if( intermediate ) { - if ((vucycle - VU->fmac[i].sCycle) > VU->fmac[i].Cycle) { -// VUM_LOG("flushing FMAC pipe[%d]", i); - VU->fmac[i].enable = 0; - } - } - else { - if ((vucycle - VU->fmac[i].sCycle) >= VU->fmac[i].Cycle) { -// VUM_LOG("flushing FMAC pipe[%d]", i); - VU->fmac[i].enable = 0; - } - } - } -} - -void _recvuFDIVflush(VURegs * VU, bool intermediate) { - if (VU->fdiv.enable == 0) return; - - if( intermediate ) { - if ((vucycle - VU->fdiv.sCycle) > VU->fdiv.Cycle) { -// Console::WriteLn("flushing FDIV pipe"); - VU->fdiv.enable = 0; - } - } - else { - if ((vucycle - VU->fdiv.sCycle) >= VU->fdiv.Cycle) { -// Console::WriteLn("flushing FDIV pipe"); - VU->fdiv.enable = 0; - } - } -} - -void _recvuEFUflush(VURegs * VU, bool intermediate) { - if (VU->efu.enable == 0) return; - - if( intermediate ) { - if ((vucycle - VU->efu.sCycle) > VU->efu.Cycle) { -// Console::WriteLn("flushing FDIV pipe"); - VU->efu.enable = 0; - } - } - else { - if ((vucycle - VU->efu.sCycle) >= VU->efu.Cycle) { -// Console::WriteLn("flushing FDIV pipe"); - VU->efu.enable = 0; - } - } -} - -void _recvuIALUflush(VURegs * VU, bool intermediate) { - int i; - - for (i=0; i<8; i++) { - if (VU->ialu[i].enable == 0) continue; - - if( intermediate ) { - if ((vucycle - VU->ialu[i].sCycle) > VU->ialu[i].Cycle) { -// VUM_LOG("flushing IALU pipe[%d]", i); - VU->ialu[i].enable = 0; - } - } - else { - if ((vucycle - VU->ialu[i].sCycle) >= VU->ialu[i].Cycle) { -// VUM_LOG("flushing IALU pipe[%d]", i); - VU->ialu[i].enable = 0; - } - } - } -} - -void _recvuTestPipes(VURegs * VU, bool intermediate) { // intermediate = true if called by upper FMAC stall detection - _recvuFMACflush(VU, intermediate); - _recvuFDIVflush(VU, intermediate); - _recvuEFUflush(VU, intermediate); - _recvuIALUflush(VU, intermediate); -} - -void _recvuFMACTestStall(VURegs * VU, int reg, int xyzw) { - int cycle; - int i; - u32 mask = 0; - - for (i=0; i<8; i++) { - if (VU->fmac[i].enable == 0) continue; - if (VU->fmac[i].reg == reg && (VU->fmac[i].xyzw & xyzw)) break; - } - - if (i == 8) return; - - // do a perchannel delay - // old code -// cycle = VU->fmac[i].Cycle - (vucycle - VU->fmac[i].sCycle); - - // new code - mask = 4; // w -// if( VU->fmac[i].xyzw & 1 ) mask = 4; // w -// else if( VU->fmac[i].xyzw & 2 ) mask = 3; // z -// else if( VU->fmac[i].xyzw & 4 ) mask = 2; // y -// else { -// assert(VU->fmac[i].xyzw & 8 ); -// mask = 1; // x -// } - -// mask = 0; -// if( VU->fmac[i].xyzw & 1 ) mask++; // w -// else if( VU->fmac[i].xyzw & 2 ) mask++; // z -// else if( VU->fmac[i].xyzw & 4 ) mask++; // y -// else if( VU->fmac[i].xyzw & 8 ) mask++; // x - - assert( (int)VU->fmac[i].sCycle < (int)vucycle ); - cycle = 0; - if( vucycle - VU->fmac[i].sCycle < mask ) - cycle = mask - (vucycle - VU->fmac[i].sCycle); - - VU->fmac[i].enable = 0; - vucycle+= cycle; - _recvuTestPipes(VU, true); // for lower instructions -} - -void _recvuIALUTestStall(VURegs * VU, int reg) { - int cycle; - int i; - u32 latency; - - for (i=0; i<8; i++) { - if (VU->ialu[i].enable == 0) continue; - if (VU->ialu[i].reg == reg) break; - } - - if (i == 8) return; - - latency = VU->ialu[i].Cycle + 1; - cycle = 0; - if( vucycle - VU->ialu[i].sCycle < latency ) - cycle = latency - (vucycle - VU->ialu[i].sCycle); - - VU->ialu[i].enable = 0; - vucycle+= cycle; - _recvuTestPipes(VU, true); -} - -void _recvuFMACAdd(VURegs * VU, int reg, int xyzw) { - int i; - - /* find a free fmac pipe */ - for (i=0; i<8; i++) { - if (VU->fmac[i].enable == 1) continue; - break; - } - - if (i==8) Console::Error("*PCSX2*: error , out of fmacs"); -// VUM_LOG("adding FMAC pipe[%d]; reg %d", i, reg); - - VU->fmac[i].enable = 1; - VU->fmac[i].sCycle = vucycle; - VU->fmac[i].Cycle = 3; - VU->fmac[i].xyzw = xyzw; - VU->fmac[i].reg = reg; -} - -void _recvuFDIVAdd(VURegs * VU, int cycles) { -// Console::WriteLn("adding FDIV pipe"); - VU->fdiv.enable = 1; - VU->fdiv.sCycle = vucycle; - VU->fdiv.Cycle = cycles; -} - -void _recvuEFUAdd(VURegs * VU, int cycles) { -// Console::WriteLn("adding EFU pipe"); - VU->efu.enable = 1; - VU->efu.sCycle = vucycle; - VU->efu.Cycle = cycles; -} - -void _recvuIALUAdd(VURegs * VU, int reg, int cycles) { - int i; - - /* find a free ialu pipe */ - for (i=0; i<8; i++) { - if (VU->ialu[i].enable == 1) continue; - break; - } - - if (i==8) Console::Error("*PCSX2*: error , out of ialus"); - - VU->ialu[i].enable = 1; - VU->ialu[i].sCycle = vucycle; - VU->ialu[i].Cycle = cycles; - VU->ialu[i].reg = reg; -} - -void _recvuTestIALUStalls(VURegs * VU, _VURegsNum *VUregsn) { - - int VIread0 = 0, VIread1 = 0; // max 2 integer registers are read simulataneously - int i; - - for(i=0;i<16;i++) { // find used integer(vi00-vi15) registers - if( (VUregsn->VIread >> i) & 1 ) { - if( VIread0 ) VIread1 = i; - else VIread0 = i; - } - } - - if( VIread0 ) _recvuIALUTestStall(VU, VIread0); - if( VIread1 ) _recvuIALUTestStall(VU, VIread1); -} - -void _recvuAddIALUStalls(VURegs * VU, _VURegsNum *VUregsn) { - if (VUregsn->VIwrite && VUregsn->cycles) { - int VIWrite0 = 0; - int i; - - for(i=0;i<16;i++) { // find used(vi00-vi15) registers - if( (VUregsn->VIwrite >> i) & 1 ) { - VIWrite0 = i; - } - } - if( VIWrite0 ) _recvuIALUAdd(VU, VIWrite0, VUregsn->cycles); - } -} - -void _recvuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn, bool upper) { - - if( VUregsn->VFread0 && (VUregsn->VFread0 == VUregsn->VFread1) ) { - _recvuFMACTestStall(VU, VUregsn->VFread0, VUregsn->VFr0xyzw|VUregsn->VFr1xyzw); - } - else { - if (VUregsn->VFread0) _recvuFMACTestStall(VU, VUregsn->VFread0, VUregsn->VFr0xyzw); - if (VUregsn->VFread1) _recvuFMACTestStall(VU, VUregsn->VFread1, VUregsn->VFr1xyzw); - } - - if( !upper && VUregsn->VIread ) _recvuTestIALUStalls(VU, VUregsn); // for lower instructions which read integer reg -} - -void _recvuAddFMACStalls(VURegs * VU, _VURegsNum *VUregsn) { - - if (VUregsn->VFwrite) _recvuFMACAdd(VU, VUregsn->VFwrite, VUregsn->VFwxyzw); - else if (VUregsn->VIwrite & (1 << REG_CLIP_FLAG)) _recvuFMACAdd(VU, -REG_CLIP_FLAG, 0); // REG_CLIP_FLAG pipe - else _recvuFMACAdd(VU, 0, 0); // cause no data dependency with fp registers -} - -void _recvuFlushFDIV(VURegs * VU) { - int cycle; - - if (VU->fdiv.enable == 0) return; - - cycle = VU->fdiv.Cycle + 1 - (vucycle - VU->fdiv.sCycle); //VU->fdiv.Cycle contains the latency minus 1 (6 or 12) -// Console::WriteLn("waiting FDIV pipe %d", params cycle); - VU->fdiv.enable = 0; - vucycle+= cycle; -} - -void _recvuFlushEFU(VURegs * VU) { - int cycle; - - if (VU->efu.enable == 0) return; - - cycle = VU->efu.Cycle - (vucycle - VU->efu.sCycle); -// Console::WriteLn("waiting FDIV pipe %d", params cycle); - VU->efu.enable = 0; - vucycle+= cycle; -} - -void _recvuTestFDIVStalls(VURegs * VU, _VURegsNum *VUregsn) { - _recvuTestFMACStalls(VU,VUregsn, false); - _recvuFlushFDIV(VU); -} - -void _recvuTestEFUStalls(VURegs * VU, _VURegsNum *VUregsn) { - _recvuTestFMACStalls(VU,VUregsn, false); - _recvuFlushEFU(VU); -} - -void _recvuAddFDIVStalls(VURegs * VU, _VURegsNum *VUregsn) { -// _vuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn); - if (VUregsn->VIwrite & (1 << REG_Q)) { - _recvuFDIVAdd(VU, VUregsn->cycles); - } -} - -void _recvuAddEFUStalls(VURegs * VU, _VURegsNum *VUregsn) { -// _vuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn); - if (VUregsn->VIwrite & (1 << REG_P)) { - _recvuEFUAdd(VU, VUregsn->cycles); - } -} - -void _recvuTestUpperStalls(VURegs * VU, _VURegsNum *VUregsn) { - switch (VUregsn->pipe) { - case VUPIPE_FMAC: _recvuTestFMACStalls(VU, VUregsn, true); break; - } -} - -void _recvuTestLowerStalls(VURegs * VU, _VURegsNum *VUregsn) { - switch (VUregsn->pipe) { - case VUPIPE_FMAC: _recvuTestFMACStalls(VU, VUregsn, false); break; - case VUPIPE_FDIV: _recvuTestFDIVStalls(VU, VUregsn); break; - case VUPIPE_EFU: _recvuTestEFUStalls(VU, VUregsn); break; - case VUPIPE_IALU: _recvuTestIALUStalls(VU, VUregsn); break; - case VUPIPE_BRANCH: _recvuTestIALUStalls(VU, VUregsn); break; - } -} - -void _recvuAddUpperStalls(VURegs * VU, _VURegsNum *VUregsn) { - switch (VUregsn->pipe) { - case VUPIPE_FMAC: _recvuAddFMACStalls(VU, VUregsn); break; - } -} - -void _recvuAddLowerStalls(VURegs * VU, _VURegsNum *VUregsn) { - switch (VUregsn->pipe) { - case VUPIPE_FMAC: _recvuAddFMACStalls(VU, VUregsn); break; - case VUPIPE_FDIV: _recvuAddFDIVStalls(VU, VUregsn); break; - case VUPIPE_EFU: _recvuAddEFUStalls(VU, VUregsn); break; - case VUPIPE_IALU: _recvuAddIALUStalls(VU, VUregsn); break; // note: only ILW and ILWR cause stall in IALU pipe - } -} - -void SuperVUAnalyzeOp(VURegs *VU, _vuopinfo *info, _VURegsNum* pCodeRegs) -{ - _VURegsNum* lregs; - _VURegsNum* uregs; - int *ptr; - - lregs = pCodeRegs; - uregs = pCodeRegs+1; - - ptr = (int*)&VU->Micro[pc]; - pc += 8; - - if (ptr[1] & 0x40000000) { // EOP - branch |= 8; - } - - VU->code = ptr[1]; - if (VU == &VU1) VU1regs_UPPER_OPCODE[VU->code & 0x3f](uregs); - else VU0regs_UPPER_OPCODE[VU->code & 0x3f](uregs); - - _recvuTestUpperStalls(VU, uregs); - switch(VU->code & 0x3f) { - case 0x10: case 0x11: case 0x12: case 0x13: - case 0x14: case 0x15: case 0x16: case 0x17: - case 0x1d: case 0x1f: - case 0x2b: case 0x2f: - break; - - case 0x3c: - switch ((VU->code >> 6) & 0x1f) { - case 0x4: case 0x5: - break; - default: - info->statusflag = 4; - info->macflag = 4; - break; - } - break; - case 0x3d: - switch ((VU->code >> 6) & 0x1f) { - case 0x4: case 0x5: case 0x7: - break; - default: - info->statusflag = 4; - info->macflag = 4; - break; - } - break; - case 0x3e: - switch ((VU->code >> 6) & 0x1f) { - case 0x4: case 0x5: - break; - default: - info->statusflag = 4; - info->macflag = 4; - break; - } - break; - case 0x3f: - switch ((VU->code >> 6) & 0x1f) { - case 0x4: case 0x5: case 0x7: case 0xb: - break; - default: - info->statusflag = 4; - info->macflag = 4; - break; - } - break; - - default: - info->statusflag = 4; - info->macflag = 4; - break; - } - - if (uregs->VIread & (1 << REG_Q)) { info->q |= 2; } - if (uregs->VIread & (1 << REG_P)) { info->p |= 2; assert( VU == &VU1 ); } - - // check upper flags - if (ptr[1] & 0x80000000) { // I flag - info->cycle = vucycle; - memzero_obj(*lregs); - } - else { - - VU->code = ptr[0]; - if (VU == &VU1) VU1regs_LOWER_OPCODE[VU->code >> 25](lregs); - else VU0regs_LOWER_OPCODE[VU->code >> 25](lregs); - - _recvuTestLowerStalls(VU, lregs); - info->cycle = vucycle; - - if (lregs->pipe == VUPIPE_BRANCH) { - branch |= 1; - } - - if (lregs->VIwrite & (1 << REG_Q)) { - info->q |= 4; - info->cycles = lregs->cycles; - info->pqinst = (VU->code&2)>>1; // rsqrt is 2 - } - else if (lregs->pipe == VUPIPE_FDIV) { - info->q |= 8|1; - info->pqinst = 0; - } - - if (lregs->VIwrite & (1 << REG_P)) { - assert( VU == &VU1 ); - info->p |= 4; - info->cycles = lregs->cycles; - - switch( VU->code & 0xff ) { - case 0xfd: info->pqinst = 0; break; //eatan - case 0x7c: info->pqinst = 0; break; //eatanxy - case 0x7d: info->pqinst = 0; break; //eatanzy - case 0xfe: info->pqinst = 1; break; //eexp - case 0xfc: info->pqinst = 2; break; //esin - case 0x3f: info->pqinst = 3; break; //erleng - case 0x3e: info->pqinst = 4; break; //eleng - case 0x3d: info->pqinst = 4; break; //ersadd - case 0xbd: info->pqinst = 4; break; //ersqrt - case 0xbe: info->pqinst = 5; break; //ercpr - case 0xbc: info->pqinst = 5; break; //esqrt - case 0x7e: info->pqinst = 5; break; //esum - case 0x3c: info->pqinst = 6; break; //esadd - default: assert(0); - } - } - else if (lregs->pipe == VUPIPE_EFU) { - info->p |= 8|1; - } - - if (lregs->VIread & (1 << REG_STATUS_FLAG)) info->statusflag|= VUOP_READ; - if (lregs->VIread & (1 << REG_MAC_FLAG)) info->macflag|= VUOP_READ; - - if (lregs->VIwrite & (1 << REG_STATUS_FLAG)) info->statusflag|= VUOP_WRITE; - if (lregs->VIwrite & (1 << REG_MAC_FLAG)) info->macflag|= VUOP_WRITE; - - if (lregs->VIread & (1 << REG_Q)) { info->q |= 2; } - if (lregs->VIread & (1 << REG_P)) { info->p |= 2; assert( VU == &VU1 ); } - - _recvuAddLowerStalls(VU, lregs); - } - - _recvuAddUpperStalls(VU, uregs); - _recvuTestPipes(VU, false); - - vucycle++; -} - -int eeVURecompileCode(VURegs *VU, _VURegsNum* regs) -{ - int info = 0; - int vfread0=-1, vfread1 = -1, vfwrite = -1, vfacc = -1, vftemp=-1; - - assert( regs != NULL ); - - if( regs->VFread0 ) _addNeededVFtoXMMreg(regs->VFread0); - if( regs->VFread1 ) _addNeededVFtoXMMreg(regs->VFread1); - if( regs->VFwrite ) _addNeededVFtoXMMreg(regs->VFwrite); - if( regs->VIread & (1<VIread & (1<VFread0 ) vfread0 = _allocVFtoXMMreg(VU, -1, regs->VFread0, MODE_READ); - else if( regs->VIread & (1<VFread1 ) vfread1 = _allocVFtoXMMreg(VU, -1, regs->VFread1, MODE_READ); - else if( (regs->VIread & (1<VFr1xyzw != 0xff) vfread1 = _allocVFtoXMMreg(VU, -1, 0, MODE_READ); - - if( regs->VIread & (1<VIwrite&(1<VIwrite & (1<VFwxyzw != 0xf?MODE_READ:0)); - } - - if( regs->VFwrite ) { - assert( !(regs->VIwrite&(1<VFwrite, MODE_WRITE|(regs->VFwxyzw != 0xf?MODE_READ:0)); - } - - if( vfacc>= 0 ) info |= PROCESS_EE_SET_ACC(vfacc); - if( vfwrite >= 0 ) { - if( regs->VFwrite == _Ft_ && vfread1 < 0 ) { - info |= PROCESS_EE_SET_T(vfwrite); - } - else { - assert( regs->VFwrite == _Fd_ ); - info |= PROCESS_EE_SET_D(vfwrite); - } - } - - if( vfread0 >= 0 ) info |= PROCESS_EE_SET_S(vfread0); - if( vfread1 >= 0 ) info |= PROCESS_EE_SET_T(vfread1); - - vftemp = _allocTempXMMreg(XMMT_FPS, -1); - info |= PROCESS_VU_SET_TEMP(vftemp); - - if( regs->VIwrite & (1 << REG_CLIP_FLAG) ) { - // CLIP inst, need two extra temp registers, put it EEREC_D and EEREC_ACC - int t1reg = _allocTempXMMreg(XMMT_FPS, -1); - int t2reg = _allocTempXMMreg(XMMT_FPS, -1); - - info |= PROCESS_EE_SET_D(t1reg); - info |= PROCESS_EE_SET_ACC(t2reg); - - _freeXMMreg(t1reg); // don't need - _freeXMMreg(t2reg); // don't need - } - else if( regs->VIwrite & (1<VI[reg].UL; - - if( read != 1 ) { - if( reg == REG_MAC_FLAG ) return (uptr)&VU->macflag; - if( reg == REG_CLIP_FLAG ) return (uptr)&VU->clipflag; - if( reg == REG_STATUS_FLAG ) return (uptr)&VU->statusflag; - if( reg == REG_Q ) return (uptr)&VU->q; - if( reg == REG_P ) return (uptr)&VU->p; - } - - return (uptr)&VU->VI[reg].UL; -} - -// gets a temp reg that is not EEREC_TEMP -int _vuGetTempXMMreg(int info) -{ - int t1reg = -1; - - if( _hasFreeXMMreg() ) { - t1reg = _allocTempXMMreg(XMMT_FPS, -1); - - if( t1reg == EEREC_TEMP ) { - if( _hasFreeXMMreg() ) { - int t = _allocTempXMMreg(XMMT_FPS, -1); - _freeXMMreg(t1reg); - t1reg = t; - } - else { - _freeXMMreg(t1reg); - t1reg = -1; - } - } - } - - return t1reg; -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// Misc VU Reg Flipping/Merging Functions -//------------------------------------------------------------------ -void _unpackVF_xyzw(int dstreg, int srcreg, int xyzw) -{ - switch (xyzw) { - case 0: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x00); break; - case 1: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x55); break; - case 2: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0xaa); break; - case 3: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0xff); break; - } -} - -void _unpackVFSS_xyzw(int dstreg, int srcreg, int xyzw) -{ - switch (xyzw) { - case 0: SSE_MOVSS_XMM_to_XMM(dstreg, srcreg); break; - case 1: if ( cpucaps.hasStreamingSIMD4Extensions ) SSE4_INSERTPS_XMM_to_XMM(dstreg, srcreg, _MM_MK_INSERTPS_NDX(1, 0, 0)); - else SSE2_PSHUFLW_XMM_to_XMM(dstreg, srcreg, 0xee); - break; - case 2: SSE_MOVHLPS_XMM_to_XMM(dstreg, srcreg); break; - case 3: if ( cpucaps.hasStreamingSIMD4Extensions ) SSE4_INSERTPS_XMM_to_XMM(dstreg, srcreg, _MM_MK_INSERTPS_NDX(3, 0, 0)); - else { SSE_MOVHLPS_XMM_to_XMM(dstreg, srcreg); SSE2_PSHUFLW_XMM_to_XMM(dstreg, dstreg, 0xee); } - break; - } -} - -void _vuFlipRegSS(VURegs * VU, int reg) -{ - assert( _XYZW_SS ); - if( _Y ) SSE2_PSHUFLW_XMM_to_XMM(reg, reg, 0x4e); - else if( _Z ) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc6); - else if( _W ) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x27); -} - -void _vuFlipRegSS_xyzw(int reg, int xyzw) -{ - switch ( xyzw ) { - case 1: SSE2_PSHUFLW_XMM_to_XMM(reg, reg, 0x4e); break; - case 2: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc6); break; - case 3: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x27); break; - } -} - -void _vuMoveSS(VURegs * VU, int dstreg, int srcreg) -{ - assert( _XYZW_SS ); - if( _Y ) _unpackVFSS_xyzw(dstreg, srcreg, 1); - else if( _Z ) _unpackVFSS_xyzw(dstreg, srcreg, 2); - else if( _W ) _unpackVFSS_xyzw(dstreg, srcreg, 3); - else _unpackVFSS_xyzw(dstreg, srcreg, 0); -} - -// 1 - src, 0 - dest wzyx -void VU_MERGE0(int dest, int src) { // 0000s -} -void VU_MERGE1(int dest, int src) { // 1000 - SSE_MOVHLPS_XMM_to_XMM(src, dest); - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xc4); -} -void VU_MERGE1b(int dest, int src) { // 1000s - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); -} -void VU_MERGE2(int dest, int src) { // 0100 - SSE_MOVHLPS_XMM_to_XMM(src, dest); - SSE_SHUFPS_XMM_to_XMM(dest, src, 0x64); -} -void VU_MERGE2b(int dest, int src) { // 0100s - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); -} -void VU_MERGE3(int dest, int src) { // 1100s - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); -} -void VU_MERGE4(int dest, int src) { // 0010 - SSE_MOVSS_XMM_to_XMM(src, dest); - SSE2_MOVSD_XMM_to_XMM(dest, src); -} -void VU_MERGE4b(int dest, int src) { // 0010s - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); -} -void VU_MERGE5(int dest, int src) { // 1010 - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xd8); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xd8); -} -void VU_MERGE5b(int dest, int src) { // 1010s - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); -} -void VU_MERGE6(int dest, int src) { // 0110 - SSE_SHUFPS_XMM_to_XMM(dest, src, 0x9c); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x78); -} -void VU_MERGE6b(int dest, int src) { // 0110s - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); -} -void VU_MERGE7(int dest, int src) { // 1110 - SSE_MOVSS_XMM_to_XMM(src, dest); - SSE_MOVAPS_XMM_to_XMM(dest, src); -} -void VU_MERGE7b(int dest, int src) { // 1110s - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); -} -void VU_MERGE8(int dest, int src) { // 0001s - SSE_MOVSS_XMM_to_XMM(dest, src); -} -void VU_MERGE9(int dest, int src) { // 1001 - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xc9); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xd2); -} -void VU_MERGE9b(int dest, int src) { // 1001s - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); -} -void VU_MERGE10(int dest, int src) { // 0101 - SSE_SHUFPS_XMM_to_XMM(dest, src, 0x8d); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x72); -} -void VU_MERGE10b(int dest, int src) { // 0101s - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); -} -void VU_MERGE11(int dest, int src) { // 1101s - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); -} -void VU_MERGE12(int dest, int src) { // 0011 - SSE2_MOVSD_XMM_to_XMM(dest, src); -} -void VU_MERGE13(int dest, int src) { // 1011 - SSE_MOVHLPS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, dest, 0x64); - SSE_MOVAPS_XMM_to_XMM(dest, src); -} -void VU_MERGE13b(int dest, int src) { // 1011s - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); -} -void VU_MERGE14(int dest, int src) { // 0111 - SSE_MOVHLPS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, dest, 0xc4); - SSE_MOVAPS_XMM_to_XMM(dest, src); -} -void VU_MERGE14b(int dest, int src) { // 0111s - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); - SSE_MOVSS_XMM_to_XMM(dest, src); - SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); - SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); -} -void VU_MERGE15(int dest, int src) { // 1111s - SSE_MOVAPS_XMM_to_XMM(dest, src); -} - -typedef void (*VUMERGEFN)(int dest, int src); - -static VUMERGEFN s_VuMerge[16] = { - VU_MERGE0, VU_MERGE1, VU_MERGE2, VU_MERGE3, - VU_MERGE4, VU_MERGE5, VU_MERGE6, VU_MERGE7, - VU_MERGE8, VU_MERGE9, VU_MERGE10, VU_MERGE11, - VU_MERGE12, VU_MERGE13, VU_MERGE14, VU_MERGE15 }; - -static VUMERGEFN s_VuMerge2[16] = { - VU_MERGE0, VU_MERGE1b, VU_MERGE2b, VU_MERGE3, - VU_MERGE4b, VU_MERGE5b, VU_MERGE6b, VU_MERGE7b, - VU_MERGE8, VU_MERGE9b, VU_MERGE10b, VU_MERGE11, - VU_MERGE12, VU_MERGE13b, VU_MERGE14b, VU_MERGE15 }; - -// Modifies the Source Reg! -void VU_MERGE_REGS_CUSTOM(int dest, int src, int xyzw) { - xyzw &= 0xf; - if ( (dest != src) && (xyzw != 0) ) { - if ( cpucaps.hasStreamingSIMD4Extensions && (xyzw != 0x8) && (xyzw != 0xf) ) { - xyzw = ((xyzw & 1) << 3) | ((xyzw & 2) << 1) | ((xyzw & 4) >> 1) | ((xyzw & 8) >> 3); - SSE4_BLENDPS_XMM_to_XMM(dest, src, xyzw); - } - else s_VuMerge[xyzw](dest, src); - } -} -// Doesn't Modify the Source Reg! (ToDo: s_VuMerge2() has room for optimization) -void VU_MERGE_REGS_SAFE(int dest, int src, int xyzw) { - xyzw &= 0xf; - if ( (dest != src) && (xyzw != 0) ) { - if ( cpucaps.hasStreamingSIMD4Extensions && (xyzw != 0x8) && (xyzw != 0xf) ) { - xyzw = ((xyzw & 1) << 3) | ((xyzw & 2) << 1) | ((xyzw & 4) >> 1) | ((xyzw & 8) >> 3); - SSE4_BLENDPS_XMM_to_XMM(dest, src, xyzw); - } - else s_VuMerge2[xyzw](dest, src); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// Misc VU Reg Clamping/Overflow Functions -//------------------------------------------------------------------ -#define CLAMP_NORMAL_SSE4(n) \ - SSE_MOVAPS_XMM_to_XMM(regTemp, regd);\ - SSE4_PMINUD_M128_to_XMM(regd, (uptr)&g_minvals_XYZW[n][0]);\ - SSE2_PSUBD_XMM_to_XMM(regTemp, regd);\ - SSE2_PCMPGTD_M128_to_XMM(regTemp, (uptr)&g_ones[0]);\ - SSE4_PMINSD_M128_to_XMM(regd, (uptr)&g_maxvals_XYZW[n][0]);\ - SSE2_PSLLD_I8_to_XMM(regTemp, 31);\ - SSE_XORPS_XMM_to_XMM(regd, regTemp); - -#define CLAMP_SIGN_SSE4(n) \ - SSE4_PMINSD_M128_to_XMM(regd, (uptr)&g_maxvals_XYZW[n][0]);\ - SSE4_PMINUD_M128_to_XMM(regd, (uptr)&g_minvals_XYZW[n][0]); - -void vFloat0(int regd, int regTemp) { } //0000 -void vFloat1(int regd, int regTemp) { //1000 - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); -} -void vFloat1c(int regd, int regTemp) { //1000 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(1); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat2(int regd, int regTemp) { //0100 - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); -} -void vFloat2c(int regd, int regTemp) { //0100 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(2); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat3(int regd, int regTemp) { //1100 - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); -} -void vFloat3b(int regd, int regTemp) { //1100 //regTemp is Modified - SSE2_MOVSD_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE2_MOVSD_XMM_to_XMM(regd, regTemp); -} -void vFloat3c(int regd, int regTemp) { //1100 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(3); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat4(int regd, int regTemp) { //0010 - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); -} -void vFloat4c(int regd, int regTemp) { //0010 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(4); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat5(int regd, int regTemp) { //1010 - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); -} -void vFloat5b(int regd, int regTemp) { //1010 //regTemp is Modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_NORMAL_SSE4(5); - } - else { - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); - } -} -void vFloat5c(int regd, int regTemp) { //1010 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(5); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat6(int regd, int regTemp) { //0110 - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); -} -void vFloat6b(int regd, int regTemp) { //0110 //regTemp is Modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_NORMAL_SSE4(6); - } - else { - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); - } -} -void vFloat6c(int regd, int regTemp) { //0110 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(6); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat7(int regd, int regTemp) { //1110 - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x39); -} -void vFloat7_useEAX(int regd, int regTemp) { //1110 //EAX is Modified - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - if ( cpucaps.hasStreamingSIMD4Extensions ) - SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); - else { - SSE_PINSRW_R32_to_XMM(regd, EAX, 0); - SHR32ItoR(EAX, 16); - SSE_PINSRW_R32_to_XMM(regd, EAX, 1); - } -} -void vFloat7b(int regd, int regTemp) { //1110 //regTemp is Modified - SSE_MOVSS_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_MOVSS_XMM_to_XMM(regd, regTemp); -} -void vFloat7c(int regd, int regTemp) { //1110 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(7); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x39); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat7c_useEAX(int regd, int regTemp) { //1110 //EAX is Modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(7); - } - else { - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - SSE2_MOVD_R_to_XMM(regTemp, EAX); - SSE_MOVSS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat8(int regd, int regTemp) { //0001 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); -} -void vFloat8c(int regd, int regTemp) { //0001 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(8); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat9(int regd, int regTemp) { //1001 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); -} -void vFloat9b(int regd, int regTemp) { //1001 //regTemp is Modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_NORMAL_SSE4(9); - } - else { - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - } -} -void vFloat9c(int regd, int regTemp) { //1001 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(9); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat10(int regd, int regTemp) { //0101 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); -} -void vFloat10b(int regd, int regTemp) { //0101 //regTemp is Modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_NORMAL_SSE4(10); - } - else { - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - } -} -void vFloat10c(int regd, int regTemp) { //0101 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(10); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat11(int regd, int regTemp) { //1101 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); -} -void vFloat11_useEAX(int regd, int regTemp) { //1101 //EAX is Modified - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - if ( cpucaps.hasStreamingSIMD4Extensions ) - SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); - else { - SSE_PINSRW_R32_to_XMM(regd, EAX, 0); - SHR32ItoR(EAX, 16); - SSE_PINSRW_R32_to_XMM(regd, EAX, 1); - } - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); -} -void vFloat11b(int regd, int regTemp) { //1101 //regTemp is Modified - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_MOVSS_XMM_to_XMM(regTemp, regd); - SSE2_MOVSD_XMM_to_XMM(regd, regTemp); -} -void vFloat11c(int regd, int regTemp) { //1101 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(11); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat11c_useEAX(int regd, int regTemp) { //1101 // EAX is modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(11); - } - else { - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - SSE2_MOVD_R_to_XMM(regTemp, EAX); - SSE_MOVSS_XMM_to_XMM(regd, regTemp); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - } -} -void vFloat12(int regd, int regTemp) { //0011 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); -} -void vFloat12b(int regd, int regTemp) { //0011 //regTemp is Modified - SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE2_PUNPCKLQDQ_XMM_to_XMM(regd, regTemp); -} -void vFloat12c(int regd, int regTemp) { //0011 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(12); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat13(int regd, int regTemp) { //1011 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); -} -void vFloat13_useEAX(int regd, int regTemp) { //1011 // EAX is modified - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - if ( cpucaps.hasStreamingSIMD4Extensions ) - SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); - else { - SSE_PINSRW_R32_to_XMM(regd, EAX, 0); - SHR32ItoR(EAX, 16); - SSE_PINSRW_R32_to_XMM(regd, EAX, 1); - } - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); -} -void vFloat13b(int regd, int regTemp) { //1011 //regTemp is Modified - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); - SSE_SHUFPS_XMM_to_XMM(regd, regTemp, 0x64); -} -void vFloat13c(int regd, int regTemp) { //1011 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(13); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat13c_useEAX(int regd, int regTemp) { //1011 // EAX is modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(13); - } - else { - SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0xc6); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - SSE2_MOVD_R_to_XMM(regTemp, EAX); - SSE_MOVSS_XMM_to_XMM(regd, regTemp); - SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0xc6); - } -} -void vFloat14(int regd, int regTemp) { //0111 - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); -} -void vFloat14_useEAX(int regd, int regTemp) { //0111 // EAX is modified - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - if ( cpucaps.hasStreamingSIMD4Extensions ) - SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); - else { - SSE_PINSRW_R32_to_XMM(regd, EAX, 0); - SHR32ItoR(EAX, 16); - SSE_PINSRW_R32_to_XMM(regd, EAX, 1); - } - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); -} -void vFloat14b(int regd, int regTemp) { //0111 //regTemp is Modified - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); - SSE_SHUFPS_XMM_to_XMM(regd, regTemp, 0xc4); -} -void vFloat14c(int regd, int regTemp) { //0111 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(14); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); - SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} -void vFloat14c_useEAX(int regd, int regTemp) { //0111 // EAX is modified - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(14); - } - else { - SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0x27); - SSE2_MOVD_XMM_to_R(EAX, regd); - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - SSE2_MOVD_R_to_XMM(regTemp, EAX); - SSE_MOVSS_XMM_to_XMM(regd, regTemp); - SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0x27); - } -} -void vFloat15(int regd, int regTemp) { //1111 - SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); - SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); -} -void vFloat15c(int regd, int regTemp) { //1111 - if ( cpucaps.hasStreamingSIMD4Extensions ) { - CLAMP_SIGN_SSE4(15); - } - else { - SSE_MOVAPS_XMM_to_XMM(regTemp, regd); - SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); - SSE_MINPS_M128_to_XMM(regd, (uptr)&g_maxvals[0]); - SSE_MAXPS_M128_to_XMM(regd, (uptr)&g_minvals[0]); - SSE_ORPS_XMM_to_XMM(regd, regTemp); - } -} - -vFloat vFloats1[16] = { //regTemp is not modified - vFloat0, vFloat1, vFloat2, vFloat3, - vFloat4, vFloat5, vFloat6, vFloat7, - vFloat8, vFloat9, vFloat10, vFloat11, - vFloat12, vFloat13, vFloat14, vFloat15 }; - -vFloat vFloats1_useEAX[16] = { //regTemp is not modified but EAX is used - vFloat0, vFloat1, vFloat2, vFloat3, - vFloat4, vFloat5, vFloat6, vFloat7_useEAX, - vFloat8, vFloat9, vFloat10, vFloat11_useEAX, - vFloat12, vFloat13_useEAX, vFloat14_useEAX, vFloat15 }; - -vFloat vFloats2[16] = { //regTemp is modified - vFloat0, vFloat1, vFloat2, vFloat3b, - vFloat4, vFloat5b, vFloat6b, vFloat7b, - vFloat8, vFloat9b, vFloat10b, vFloat11b, - vFloat12b, vFloat13b, vFloat14b, vFloat15 }; - -vFloat vFloats4[16] = { //regTemp is modified - vFloat0, vFloat1c, vFloat2c, vFloat3c, - vFloat4c, vFloat5c, vFloat6c, vFloat7c, - vFloat8c, vFloat9c, vFloat10c, vFloat11c, - vFloat12c, vFloat13c, vFloat14c, vFloat15c }; - -vFloat vFloats4_useEAX[16] = { //regTemp is modified and EAX is used - vFloat0, vFloat1c, vFloat2c, vFloat3c, - vFloat4c, vFloat5c, vFloat6c, vFloat7c_useEAX, - vFloat8c, vFloat9c, vFloat10c, vFloat11c_useEAX, - vFloat12c, vFloat13c_useEAX, vFloat14c_useEAX, vFloat15c }; - -//------------------------------------------------------------------ -// Clamping Functions (wrapper for vFloat* functions) -// vuFloat : "normal" clamping -// vuFloat_useEAX : "normal" clamping (faster but EAX is modified) -// vuFloat2 : "normal" clamping (fastest but regTemp is modified) -// vuFloat3 : "preserve sign" clamping for pointer -// vuFloat4 : "preserve sign" clamping (regTemp is modified; *FASTEST* on SSE4 CPUs) -// vuFloat4_useEAX : "preserve sign" clamping (faster but regTemp and EAX are modified) -// vuFloat5 : wrapper function for vuFloat2 and vuFloat4 -// vuFloat5_useEAX : wrapper function for vuFloat2 and vuFloat4_useEAX -// vuFloatExtra : for debugging -// -// Notice 1: vuFloat*_useEAX may be slower on AMD CPUs, which have independent execution pipeline for -// vector and scalar instructions (need checks) -// Notice 2: recVUMI_MUL_xyzw_toD and recVUMI_MADD_xyzw_toD use vFloats directly! -//------------------------------------------------------------------ - -// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (doesn't use any temp regs) -void vuFloat( int info, int regd, int XYZW) { - if( CHECK_VU_OVERFLOW ) { - /*if ( (XYZW != 0) && (XYZW != 8) && (XYZW != 0xF) ) { - int t1reg = _vuGetTempXMMreg(info); - if (t1reg >= 0) { - vuFloat2( regd, t1reg, XYZW ); - _freeXMMreg( t1reg ); - return; - } - }*/ - //vuFloatExtra(regd, XYZW); - vFloats1[XYZW](regd, regd); - } -} - -// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (uses EAX as a temp register; faster but **destroys EAX**) -void vuFloat_useEAX( int info, int regd, int XYZW) { - if( CHECK_VU_OVERFLOW ) { - vFloats1_useEAX[XYZW](regd, regd); - } -} - -// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (uses a temp reg) -void vuFloat2(int regd, int regTemp, int XYZW) { - if( CHECK_VU_OVERFLOW ) { - //vuFloatExtra(regd, XYZW); - vFloats2[XYZW](regd, regTemp); - } -} - -// Clamps +/-NaN and +/-Inf to +/-fMax (uses a temp reg) -void vuFloat4(int regd, int regTemp, int XYZW) { - if( CHECK_VU_OVERFLOW ) { - vFloats4[XYZW](regd, regTemp); - } -} - -// Clamps +/-NaN and +/-Inf to +/-fMax (uses a temp reg, and uses EAX as a temp register; faster but **destroys EAX**) -void vuFloat4_useEAX(int regd, int regTemp, int XYZW) { - if( CHECK_VU_OVERFLOW ) { - vFloats4_useEAX[XYZW](regd, regTemp); - } -} - -// Uses vuFloat4 or vuFloat2 depending on the CHECK_VU_SIGN_OVERFLOW setting -void vuFloat5(int regd, int regTemp, int XYZW) { - if (CHECK_VU_SIGN_OVERFLOW) { - vuFloat4(regd, regTemp, XYZW); - } - else vuFloat2(regd, regTemp, XYZW); -} - -// Uses vuFloat4_useEAX or vuFloat2 depending on the CHECK_VU_SIGN_OVERFLOW setting (uses EAX as a temp register; faster but **destoroyes EAX**) -void vuFloat5_useEAX(int regd, int regTemp, int XYZW) { - if (CHECK_VU_SIGN_OVERFLOW) { - vuFloat4_useEAX(regd, regTemp, XYZW); - } - else vuFloat2(regd, regTemp, XYZW); -} - -// Clamps +/-infs to +/-fMax, and +/-NaNs to +/-fMax -void vuFloat3(uptr x86ptr) { - u8* pjmp; - - if( CHECK_VU_OVERFLOW ) { - CMP32ItoM(x86ptr, 0x7f800000 ); - pjmp = JL8(0); // Signed Comparison - MOV32ItoM(x86ptr, 0x7f7fffff ); - x86SetJ8(pjmp); - - CMP32ItoM(x86ptr, 0xff800000 ); - pjmp = JB8(0); // Unsigned Comparison - MOV32ItoM(x86ptr, 0xff7fffff ); - x86SetJ8(pjmp); - } -} - -PCSX2_ALIGNED16(u64 vuFloatData[2]); -PCSX2_ALIGNED16(u64 vuFloatData2[2]); -// Makes NaN == 0, Infinities stay the same; Very Slow - Use only for debugging -void vuFloatExtra( int regd, int XYZW) { - int t1reg = (regd == 0) ? (regd + 1) : (regd - 1); - int t2reg = (regd <= 1) ? (regd + 2) : (regd - 2); - SSE_MOVAPS_XMM_to_M128( (uptr)vuFloatData, t1reg ); - SSE_MOVAPS_XMM_to_M128( (uptr)vuFloatData2, t2reg ); - - SSE_XORPS_XMM_to_XMM(t1reg, t1reg); - SSE_CMPORDPS_XMM_to_XMM(t1reg, regd); - SSE_MOVAPS_XMM_to_XMM(t2reg, regd); - SSE_ANDPS_XMM_to_XMM(t2reg, t1reg); - VU_MERGE_REGS_CUSTOM(regd, t2reg, XYZW); - - SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)vuFloatData ); - SSE_MOVAPS_M128_to_XMM( t2reg, (uptr)vuFloatData2 ); -} - -static PCSX2_ALIGNED16(u32 tempRegX[]) = {0x00000000, 0x00000000, 0x00000000, 0x00000000}; - -// Called by testWhenOverflow() function -void testPrintOverflow() { - tempRegX[0] &= 0xff800000; - tempRegX[1] &= 0xff800000; - tempRegX[2] &= 0xff800000; - tempRegX[3] &= 0xff800000; - if ( (tempRegX[0] == 0x7f800000) || (tempRegX[1] == 0x7f800000) || (tempRegX[2] == 0x7f800000) || (tempRegX[3] == 0x7f800000) ) - Console::Notice( "VU OVERFLOW!: Changing to +Fmax!!!!!!!!!!!!" ); - if ( (tempRegX[0] == 0xff800000) || (tempRegX[1] == 0xff800000) || (tempRegX[2] == 0xff800000) || (tempRegX[3] == 0xff800000) ) - Console::Notice( "VU OVERFLOW!: Changing to -Fmax!!!!!!!!!!!!" ); -} - -// Outputs to the console when overflow has occured. -void testWhenOverflow(int info, int regd, int t0reg) { - SSE_MOVAPS_XMM_to_M128((uptr)tempRegX, regd); - CALLFunc((uptr)testPrintOverflow); -} +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "PrecompiledHeader.h" + +#include "Common.h" +#include "GS.h" +#include "R5900OpcodeTables.h" +#include "iR5900.h" +#include "iMMI.h" +#include "iFPU.h" +#include "iCOP0.h" +#include "VUmicro.h" +#include "VUflags.h" +#include "sVU_Micro.h" +#include "sVU_Debug.h" +#include "sVU_zerorec.h" + +#ifdef _WIN32 +#pragma warning(disable:4244) +#pragma warning(disable:4761) +#endif +//------------------------------------------------------------------ + +// fixme - VUmicro should really use its own static vars for pc and branch. +// Sharing with the EE's copies of pc and branch is not cool! (air) + +//------------------------------------------------------------------ +// Helper Macros +//------------------------------------------------------------------ +#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register +#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register +#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) + +#define _X (( VU->code>>24) & 0x1) +#define _Y (( VU->code>>23) & 0x1) +#define _Z (( VU->code>>22) & 0x1) +#define _W (( VU->code>>21) & 0x1) + +#define _XYZW_SS (_X+_Y+_Z+_W==1) + +#define _Fsf_ (( VU->code >> 21) & 0x03) +#define _Ftf_ (( VU->code >> 23) & 0x03) + +#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) +#define _UImm11_ (s32)(VU->code & 0x7ff) + +#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] +#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] +#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] +#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] + +#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] +#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] +#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] + +#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) + +#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] +#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] +#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] +#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] + +#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// Global Variables +//------------------------------------------------------------------ +int vucycle; + +PCSX2_ALIGNED16(const float s_fones[8]) = {1.0f, 1.0f, 1.0f, 1.0f, -1.0f, -1.0f, -1.0f, -1.0f}; +PCSX2_ALIGNED16(const u32 s_mask[4]) = {0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff}; +PCSX2_ALIGNED16(const u32 s_expmask[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; +PCSX2_ALIGNED16(const u32 g_minvals[4]) = {0xff7fffff, 0xff7fffff, 0xff7fffff, 0xff7fffff}; +PCSX2_ALIGNED16(const u32 g_maxvals[4]) = {0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff}; +PCSX2_ALIGNED16(const u32 const_clip[8]) = {0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, + 0x80000000, 0x80000000, 0x80000000, 0x80000000}; +PCSX2_ALIGNED(64, const u32 g_ones[4]) = {0x00000001, 0x00000001, 0x00000001, 0x00000001}; +PCSX2_ALIGNED16(const u32 g_minvals_XYZW[16][4]) = +{ + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //0000 + { 0xffffffff, 0xffffffff, 0xffffffff, 0xff7fffff }, //0001 + { 0xffffffff, 0xffffffff, 0xff7fffff, 0xffffffff }, //0010 + { 0xffffffff, 0xffffffff, 0xff7fffff, 0xff7fffff }, //0011 + { 0xffffffff, 0xff7fffff, 0xffffffff, 0xffffffff }, //0100 + { 0xffffffff, 0xff7fffff, 0xffffffff, 0xff7fffff }, //0101 + { 0xffffffff, 0xff7fffff, 0xff7fffff, 0xffffffff }, //0110 + { 0xffffffff, 0xff7fffff, 0xff7fffff, 0xff7fffff }, //0111 + { 0xff7fffff, 0xffffffff, 0xffffffff, 0xffffffff }, //1000 + { 0xff7fffff, 0xffffffff, 0xffffffff, 0xff7fffff }, //1001 + { 0xff7fffff, 0xffffffff, 0xff7fffff, 0xffffffff }, //1010 + { 0xff7fffff, 0xffffffff, 0xff7fffff, 0xff7fffff }, //1011 + { 0xff7fffff, 0xff7fffff, 0xffffffff, 0xffffffff }, //1100 + { 0xff7fffff, 0xff7fffff, 0xffffffff, 0xff7fffff }, //1101 + { 0xff7fffff, 0xff7fffff, 0xff7fffff, 0xffffffff }, //1110 + { 0xff7fffff, 0xff7fffff, 0xff7fffff, 0xff7fffff }, //1111 +}; +PCSX2_ALIGNED16(const u32 g_maxvals_XYZW[16][4])= +{ + { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //0000 + { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7f7fffff }, //0001 + { 0x7fffffff, 0x7fffffff, 0x7f7fffff, 0x7fffffff }, //0010 + { 0x7fffffff, 0x7fffffff, 0x7f7fffff, 0x7f7fffff }, //0011 + { 0x7fffffff, 0x7f7fffff, 0x7fffffff, 0x7fffffff }, //0100 + { 0x7fffffff, 0x7f7fffff, 0x7fffffff, 0x7f7fffff }, //0101 + { 0x7fffffff, 0x7f7fffff, 0x7f7fffff, 0x7fffffff }, //0110 + { 0x7fffffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff }, //0111 + { 0x7f7fffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //1000 + { 0x7f7fffff, 0x7fffffff, 0x7fffffff, 0x7f7fffff }, //1001 + { 0x7f7fffff, 0x7fffffff, 0x7f7fffff, 0x7fffffff }, //1010 + { 0x7f7fffff, 0x7fffffff, 0x7f7fffff, 0x7f7fffff }, //1011 + { 0x7f7fffff, 0x7f7fffff, 0x7fffffff, 0x7fffffff }, //1100 + { 0x7f7fffff, 0x7f7fffff, 0x7fffffff, 0x7f7fffff }, //1101 + { 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7fffffff }, //1110 + { 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff }, //1111 +}; +//------------------------------------------------------------------ + +//------------------------------------------------------------------ +// VU Pipeline/Test Stalls/Analyzing Functions +//------------------------------------------------------------------ +void _recvuFMACflush(VURegs * VU, bool intermediate) { + int i; + + for (i=0; i<8; i++) { + if (VU->fmac[i].enable == 0) continue; + + if( intermediate ) { + if ((vucycle - VU->fmac[i].sCycle) > VU->fmac[i].Cycle) { +// VUM_LOG("flushing FMAC pipe[%d]", i); + VU->fmac[i].enable = 0; + } + } + else { + if ((vucycle - VU->fmac[i].sCycle) >= VU->fmac[i].Cycle) { +// VUM_LOG("flushing FMAC pipe[%d]", i); + VU->fmac[i].enable = 0; + } + } + } +} + +void _recvuFDIVflush(VURegs * VU, bool intermediate) { + if (VU->fdiv.enable == 0) return; + + if( intermediate ) { + if ((vucycle - VU->fdiv.sCycle) > VU->fdiv.Cycle) { +// Console::WriteLn("flushing FDIV pipe"); + VU->fdiv.enable = 0; + } + } + else { + if ((vucycle - VU->fdiv.sCycle) >= VU->fdiv.Cycle) { +// Console::WriteLn("flushing FDIV pipe"); + VU->fdiv.enable = 0; + } + } +} + +void _recvuEFUflush(VURegs * VU, bool intermediate) { + if (VU->efu.enable == 0) return; + + if( intermediate ) { + if ((vucycle - VU->efu.sCycle) > VU->efu.Cycle) { +// Console::WriteLn("flushing FDIV pipe"); + VU->efu.enable = 0; + } + } + else { + if ((vucycle - VU->efu.sCycle) >= VU->efu.Cycle) { +// Console::WriteLn("flushing FDIV pipe"); + VU->efu.enable = 0; + } + } +} + +void _recvuIALUflush(VURegs * VU, bool intermediate) { + int i; + + for (i=0; i<8; i++) { + if (VU->ialu[i].enable == 0) continue; + + if( intermediate ) { + if ((vucycle - VU->ialu[i].sCycle) > VU->ialu[i].Cycle) { +// VUM_LOG("flushing IALU pipe[%d]", i); + VU->ialu[i].enable = 0; + } + } + else { + if ((vucycle - VU->ialu[i].sCycle) >= VU->ialu[i].Cycle) { +// VUM_LOG("flushing IALU pipe[%d]", i); + VU->ialu[i].enable = 0; + } + } + } +} + +void _recvuTestPipes(VURegs * VU, bool intermediate) { // intermediate = true if called by upper FMAC stall detection + _recvuFMACflush(VU, intermediate); + _recvuFDIVflush(VU, intermediate); + _recvuEFUflush(VU, intermediate); + _recvuIALUflush(VU, intermediate); +} + +void _recvuFMACTestStall(VURegs * VU, int reg, int xyzw) { + int cycle; + int i; + u32 mask = 0; + + for (i=0; i<8; i++) { + if (VU->fmac[i].enable == 0) continue; + if (VU->fmac[i].reg == reg && (VU->fmac[i].xyzw & xyzw)) break; + } + + if (i == 8) return; + + // do a perchannel delay + // old code +// cycle = VU->fmac[i].Cycle - (vucycle - VU->fmac[i].sCycle); + + // new code + mask = 4; // w +// if( VU->fmac[i].xyzw & 1 ) mask = 4; // w +// else if( VU->fmac[i].xyzw & 2 ) mask = 3; // z +// else if( VU->fmac[i].xyzw & 4 ) mask = 2; // y +// else { +// assert(VU->fmac[i].xyzw & 8 ); +// mask = 1; // x +// } + +// mask = 0; +// if( VU->fmac[i].xyzw & 1 ) mask++; // w +// else if( VU->fmac[i].xyzw & 2 ) mask++; // z +// else if( VU->fmac[i].xyzw & 4 ) mask++; // y +// else if( VU->fmac[i].xyzw & 8 ) mask++; // x + + assert( (int)VU->fmac[i].sCycle < (int)vucycle ); + cycle = 0; + if( vucycle - VU->fmac[i].sCycle < mask ) + cycle = mask - (vucycle - VU->fmac[i].sCycle); + + VU->fmac[i].enable = 0; + vucycle+= cycle; + _recvuTestPipes(VU, true); // for lower instructions +} + +void _recvuIALUTestStall(VURegs * VU, int reg) { + int cycle; + int i; + u32 latency; + + for (i=0; i<8; i++) { + if (VU->ialu[i].enable == 0) continue; + if (VU->ialu[i].reg == reg) break; + } + + if (i == 8) return; + + latency = VU->ialu[i].Cycle + 1; + cycle = 0; + if( vucycle - VU->ialu[i].sCycle < latency ) + cycle = latency - (vucycle - VU->ialu[i].sCycle); + + VU->ialu[i].enable = 0; + vucycle+= cycle; + _recvuTestPipes(VU, true); +} + +void _recvuFMACAdd(VURegs * VU, int reg, int xyzw) { + int i; + + /* find a free fmac pipe */ + for (i=0; i<8; i++) { + if (VU->fmac[i].enable == 1) continue; + break; + } + + if (i==8) Console::Error("*PCSX2*: error , out of fmacs"); +// VUM_LOG("adding FMAC pipe[%d]; reg %d", i, reg); + + VU->fmac[i].enable = 1; + VU->fmac[i].sCycle = vucycle; + VU->fmac[i].Cycle = 3; + VU->fmac[i].xyzw = xyzw; + VU->fmac[i].reg = reg; +} + +void _recvuFDIVAdd(VURegs * VU, int cycles) { +// Console::WriteLn("adding FDIV pipe"); + VU->fdiv.enable = 1; + VU->fdiv.sCycle = vucycle; + VU->fdiv.Cycle = cycles; +} + +void _recvuEFUAdd(VURegs * VU, int cycles) { +// Console::WriteLn("adding EFU pipe"); + VU->efu.enable = 1; + VU->efu.sCycle = vucycle; + VU->efu.Cycle = cycles; +} + +void _recvuIALUAdd(VURegs * VU, int reg, int cycles) { + int i; + + /* find a free ialu pipe */ + for (i=0; i<8; i++) { + if (VU->ialu[i].enable == 1) continue; + break; + } + + if (i==8) Console::Error("*PCSX2*: error , out of ialus"); + + VU->ialu[i].enable = 1; + VU->ialu[i].sCycle = vucycle; + VU->ialu[i].Cycle = cycles; + VU->ialu[i].reg = reg; +} + +void _recvuTestIALUStalls(VURegs * VU, _VURegsNum *VUregsn) { + + int VIread0 = 0, VIread1 = 0; // max 2 integer registers are read simulataneously + int i; + + for(i=0;i<16;i++) { // find used integer(vi00-vi15) registers + if( (VUregsn->VIread >> i) & 1 ) { + if( VIread0 ) VIread1 = i; + else VIread0 = i; + } + } + + if( VIread0 ) _recvuIALUTestStall(VU, VIread0); + if( VIread1 ) _recvuIALUTestStall(VU, VIread1); +} + +void _recvuAddIALUStalls(VURegs * VU, _VURegsNum *VUregsn) { + if (VUregsn->VIwrite && VUregsn->cycles) { + int VIWrite0 = 0; + int i; + + for(i=0;i<16;i++) { // find used(vi00-vi15) registers + if( (VUregsn->VIwrite >> i) & 1 ) { + VIWrite0 = i; + } + } + if( VIWrite0 ) _recvuIALUAdd(VU, VIWrite0, VUregsn->cycles); + } +} + +void _recvuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn, bool upper) { + + if( VUregsn->VFread0 && (VUregsn->VFread0 == VUregsn->VFread1) ) { + _recvuFMACTestStall(VU, VUregsn->VFread0, VUregsn->VFr0xyzw|VUregsn->VFr1xyzw); + } + else { + if (VUregsn->VFread0) _recvuFMACTestStall(VU, VUregsn->VFread0, VUregsn->VFr0xyzw); + if (VUregsn->VFread1) _recvuFMACTestStall(VU, VUregsn->VFread1, VUregsn->VFr1xyzw); + } + + if( !upper && VUregsn->VIread ) _recvuTestIALUStalls(VU, VUregsn); // for lower instructions which read integer reg +} + +void _recvuAddFMACStalls(VURegs * VU, _VURegsNum *VUregsn) { + + if (VUregsn->VFwrite) _recvuFMACAdd(VU, VUregsn->VFwrite, VUregsn->VFwxyzw); + else if (VUregsn->VIwrite & (1 << REG_CLIP_FLAG)) _recvuFMACAdd(VU, -REG_CLIP_FLAG, 0); // REG_CLIP_FLAG pipe + else _recvuFMACAdd(VU, 0, 0); // cause no data dependency with fp registers +} + +void _recvuFlushFDIV(VURegs * VU) { + int cycle; + + if (VU->fdiv.enable == 0) return; + + cycle = VU->fdiv.Cycle + 1 - (vucycle - VU->fdiv.sCycle); //VU->fdiv.Cycle contains the latency minus 1 (6 or 12) +// Console::WriteLn("waiting FDIV pipe %d", params cycle); + VU->fdiv.enable = 0; + vucycle+= cycle; +} + +void _recvuFlushEFU(VURegs * VU) { + int cycle; + + if (VU->efu.enable == 0) return; + + cycle = VU->efu.Cycle - (vucycle - VU->efu.sCycle); +// Console::WriteLn("waiting FDIV pipe %d", params cycle); + VU->efu.enable = 0; + vucycle+= cycle; +} + +void _recvuTestFDIVStalls(VURegs * VU, _VURegsNum *VUregsn) { + _recvuTestFMACStalls(VU,VUregsn, false); + _recvuFlushFDIV(VU); +} + +void _recvuTestEFUStalls(VURegs * VU, _VURegsNum *VUregsn) { + _recvuTestFMACStalls(VU,VUregsn, false); + _recvuFlushEFU(VU); +} + +void _recvuAddFDIVStalls(VURegs * VU, _VURegsNum *VUregsn) { +// _vuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn); + if (VUregsn->VIwrite & (1 << REG_Q)) { + _recvuFDIVAdd(VU, VUregsn->cycles); + } +} + +void _recvuAddEFUStalls(VURegs * VU, _VURegsNum *VUregsn) { +// _vuTestFMACStalls(VURegs * VU, _VURegsNum *VUregsn); + if (VUregsn->VIwrite & (1 << REG_P)) { + _recvuEFUAdd(VU, VUregsn->cycles); + } +} + +void _recvuTestUpperStalls(VURegs * VU, _VURegsNum *VUregsn) { + switch (VUregsn->pipe) { + case VUPIPE_FMAC: _recvuTestFMACStalls(VU, VUregsn, true); break; + } +} + +void _recvuTestLowerStalls(VURegs * VU, _VURegsNum *VUregsn) { + switch (VUregsn->pipe) { + case VUPIPE_FMAC: _recvuTestFMACStalls(VU, VUregsn, false); break; + case VUPIPE_FDIV: _recvuTestFDIVStalls(VU, VUregsn); break; + case VUPIPE_EFU: _recvuTestEFUStalls(VU, VUregsn); break; + case VUPIPE_IALU: _recvuTestIALUStalls(VU, VUregsn); break; + case VUPIPE_BRANCH: _recvuTestIALUStalls(VU, VUregsn); break; + } +} + +void _recvuAddUpperStalls(VURegs * VU, _VURegsNum *VUregsn) { + switch (VUregsn->pipe) { + case VUPIPE_FMAC: _recvuAddFMACStalls(VU, VUregsn); break; + } +} + +void _recvuAddLowerStalls(VURegs * VU, _VURegsNum *VUregsn) { + switch (VUregsn->pipe) { + case VUPIPE_FMAC: _recvuAddFMACStalls(VU, VUregsn); break; + case VUPIPE_FDIV: _recvuAddFDIVStalls(VU, VUregsn); break; + case VUPIPE_EFU: _recvuAddEFUStalls(VU, VUregsn); break; + case VUPIPE_IALU: _recvuAddIALUStalls(VU, VUregsn); break; // note: only ILW and ILWR cause stall in IALU pipe + } +} + +void SuperVUAnalyzeOp(VURegs *VU, _vuopinfo *info, _VURegsNum* pCodeRegs) +{ + _VURegsNum* lregs; + _VURegsNum* uregs; + int *ptr; + + lregs = pCodeRegs; + uregs = pCodeRegs+1; + + ptr = (int*)&VU->Micro[pc]; + pc += 8; + + if (ptr[1] & 0x40000000) { // EOP + branch |= 8; + } + + VU->code = ptr[1]; + if (VU == &VU1) VU1regs_UPPER_OPCODE[VU->code & 0x3f](uregs); + else VU0regs_UPPER_OPCODE[VU->code & 0x3f](uregs); + + _recvuTestUpperStalls(VU, uregs); + switch(VU->code & 0x3f) { + case 0x10: case 0x11: case 0x12: case 0x13: + case 0x14: case 0x15: case 0x16: case 0x17: + case 0x1d: case 0x1f: + case 0x2b: case 0x2f: + break; + + case 0x3c: + switch ((VU->code >> 6) & 0x1f) { + case 0x4: case 0x5: + break; + default: + info->statusflag = 4; + info->macflag = 4; + break; + } + break; + case 0x3d: + switch ((VU->code >> 6) & 0x1f) { + case 0x4: case 0x5: case 0x7: + break; + default: + info->statusflag = 4; + info->macflag = 4; + break; + } + break; + case 0x3e: + switch ((VU->code >> 6) & 0x1f) { + case 0x4: case 0x5: + break; + default: + info->statusflag = 4; + info->macflag = 4; + break; + } + break; + case 0x3f: + switch ((VU->code >> 6) & 0x1f) { + case 0x4: case 0x5: case 0x7: case 0xb: + break; + default: + info->statusflag = 4; + info->macflag = 4; + break; + } + break; + + default: + info->statusflag = 4; + info->macflag = 4; + break; + } + + if (uregs->VIread & (1 << REG_Q)) { info->q |= 2; } + if (uregs->VIread & (1 << REG_P)) { info->p |= 2; assert( VU == &VU1 ); } + + // check upper flags + if (ptr[1] & 0x80000000) { // I flag + info->cycle = vucycle; + memzero_obj(*lregs); + } + else { + + VU->code = ptr[0]; + if (VU == &VU1) VU1regs_LOWER_OPCODE[VU->code >> 25](lregs); + else VU0regs_LOWER_OPCODE[VU->code >> 25](lregs); + + _recvuTestLowerStalls(VU, lregs); + info->cycle = vucycle; + + if (lregs->pipe == VUPIPE_BRANCH) { + branch |= 1; + } + + if (lregs->VIwrite & (1 << REG_Q)) { + info->q |= 4; + info->cycles = lregs->cycles; + info->pqinst = (VU->code&2)>>1; // rsqrt is 2 + } + else if (lregs->pipe == VUPIPE_FDIV) { + info->q |= 8|1; + info->pqinst = 0; + } + + if (lregs->VIwrite & (1 << REG_P)) { + assert( VU == &VU1 ); + info->p |= 4; + info->cycles = lregs->cycles; + + switch( VU->code & 0xff ) { + case 0xfd: info->pqinst = 0; break; //eatan + case 0x7c: info->pqinst = 0; break; //eatanxy + case 0x7d: info->pqinst = 0; break; //eatanzy + case 0xfe: info->pqinst = 1; break; //eexp + case 0xfc: info->pqinst = 2; break; //esin + case 0x3f: info->pqinst = 3; break; //erleng + case 0x3e: info->pqinst = 4; break; //eleng + case 0x3d: info->pqinst = 4; break; //ersadd + case 0xbd: info->pqinst = 4; break; //ersqrt + case 0xbe: info->pqinst = 5; break; //ercpr + case 0xbc: info->pqinst = 5; break; //esqrt + case 0x7e: info->pqinst = 5; break; //esum + case 0x3c: info->pqinst = 6; break; //esadd + default: assert(0); + } + } + else if (lregs->pipe == VUPIPE_EFU) { + info->p |= 8|1; + } + + if (lregs->VIread & (1 << REG_STATUS_FLAG)) info->statusflag|= VUOP_READ; + if (lregs->VIread & (1 << REG_MAC_FLAG)) info->macflag|= VUOP_READ; + + if (lregs->VIwrite & (1 << REG_STATUS_FLAG)) info->statusflag|= VUOP_WRITE; + if (lregs->VIwrite & (1 << REG_MAC_FLAG)) info->macflag|= VUOP_WRITE; + + if (lregs->VIread & (1 << REG_Q)) { info->q |= 2; } + if (lregs->VIread & (1 << REG_P)) { info->p |= 2; assert( VU == &VU1 ); } + + _recvuAddLowerStalls(VU, lregs); + } + + _recvuAddUpperStalls(VU, uregs); + _recvuTestPipes(VU, false); + + vucycle++; +} + +int eeVURecompileCode(VURegs *VU, _VURegsNum* regs) +{ + int info = 0; + int vfread0=-1, vfread1 = -1, vfwrite = -1, vfacc = -1, vftemp=-1; + + assert( regs != NULL ); + + if( regs->VFread0 ) _addNeededVFtoXMMreg(regs->VFread0); + if( regs->VFread1 ) _addNeededVFtoXMMreg(regs->VFread1); + if( regs->VFwrite ) _addNeededVFtoXMMreg(regs->VFwrite); + if( regs->VIread & (1<VIread & (1<VFread0 ) vfread0 = _allocVFtoXMMreg(VU, -1, regs->VFread0, MODE_READ); + else if( regs->VIread & (1<VFread1 ) vfread1 = _allocVFtoXMMreg(VU, -1, regs->VFread1, MODE_READ); + else if( (regs->VIread & (1<VFr1xyzw != 0xff) vfread1 = _allocVFtoXMMreg(VU, -1, 0, MODE_READ); + + if( regs->VIread & (1<VIwrite&(1<VIwrite & (1<VFwxyzw != 0xf?MODE_READ:0)); + } + + if( regs->VFwrite ) { + assert( !(regs->VIwrite&(1<VFwrite, MODE_WRITE|(regs->VFwxyzw != 0xf?MODE_READ:0)); + } + + if( vfacc>= 0 ) info |= PROCESS_EE_SET_ACC(vfacc); + if( vfwrite >= 0 ) { + if( regs->VFwrite == _Ft_ && vfread1 < 0 ) { + info |= PROCESS_EE_SET_T(vfwrite); + } + else { + assert( regs->VFwrite == _Fd_ ); + info |= PROCESS_EE_SET_D(vfwrite); + } + } + + if( vfread0 >= 0 ) info |= PROCESS_EE_SET_S(vfread0); + if( vfread1 >= 0 ) info |= PROCESS_EE_SET_T(vfread1); + + vftemp = _allocTempXMMreg(XMMT_FPS, -1); + info |= PROCESS_VU_SET_TEMP(vftemp); + + if( regs->VIwrite & (1 << REG_CLIP_FLAG) ) { + // CLIP inst, need two extra temp registers, put it EEREC_D and EEREC_ACC + int t1reg = _allocTempXMMreg(XMMT_FPS, -1); + int t2reg = _allocTempXMMreg(XMMT_FPS, -1); + + info |= PROCESS_EE_SET_D(t1reg); + info |= PROCESS_EE_SET_ACC(t2reg); + + _freeXMMreg(t1reg); // don't need + _freeXMMreg(t2reg); // don't need + } + else if( regs->VIwrite & (1<VI[reg].UL; + + if( read != 1 ) { + if( reg == REG_MAC_FLAG ) return (uptr)&VU->macflag; + if( reg == REG_CLIP_FLAG ) return (uptr)&VU->clipflag; + if( reg == REG_STATUS_FLAG ) return (uptr)&VU->statusflag; + if( reg == REG_Q ) return (uptr)&VU->q; + if( reg == REG_P ) return (uptr)&VU->p; + } + + return (uptr)&VU->VI[reg].UL; +} + +// gets a temp reg that is not EEREC_TEMP +int _vuGetTempXMMreg(int info) +{ + int t1reg = -1; + + if( _hasFreeXMMreg() ) { + t1reg = _allocTempXMMreg(XMMT_FPS, -1); + + if( t1reg == EEREC_TEMP ) { + if( _hasFreeXMMreg() ) { + int t = _allocTempXMMreg(XMMT_FPS, -1); + _freeXMMreg(t1reg); + t1reg = t; + } + else { + _freeXMMreg(t1reg); + t1reg = -1; + } + } + } + + return t1reg; +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// Misc VU Reg Flipping/Merging Functions +//------------------------------------------------------------------ +void _unpackVF_xyzw(int dstreg, int srcreg, int xyzw) +{ + switch (xyzw) { + case 0: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x00); break; + case 1: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0x55); break; + case 2: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0xaa); break; + case 3: SSE2_PSHUFD_XMM_to_XMM(dstreg, srcreg, 0xff); break; + } +} + +void _unpackVFSS_xyzw(int dstreg, int srcreg, int xyzw) +{ + switch (xyzw) { + case 0: SSE_MOVSS_XMM_to_XMM(dstreg, srcreg); break; + case 1: if ( cpucaps.hasStreamingSIMD4Extensions ) SSE4_INSERTPS_XMM_to_XMM(dstreg, srcreg, _MM_MK_INSERTPS_NDX(1, 0, 0)); + else SSE2_PSHUFLW_XMM_to_XMM(dstreg, srcreg, 0xee); + break; + case 2: SSE_MOVHLPS_XMM_to_XMM(dstreg, srcreg); break; + case 3: if ( cpucaps.hasStreamingSIMD4Extensions ) SSE4_INSERTPS_XMM_to_XMM(dstreg, srcreg, _MM_MK_INSERTPS_NDX(3, 0, 0)); + else { SSE_MOVHLPS_XMM_to_XMM(dstreg, srcreg); SSE2_PSHUFLW_XMM_to_XMM(dstreg, dstreg, 0xee); } + break; + } +} + +void _vuFlipRegSS(VURegs * VU, int reg) +{ + assert( _XYZW_SS ); + if( _Y ) SSE2_PSHUFLW_XMM_to_XMM(reg, reg, 0x4e); + else if( _Z ) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc6); + else if( _W ) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x27); +} + +void _vuFlipRegSS_xyzw(int reg, int xyzw) +{ + switch ( xyzw ) { + case 1: SSE2_PSHUFLW_XMM_to_XMM(reg, reg, 0x4e); break; + case 2: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0xc6); break; + case 3: SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x27); break; + } +} + +void _vuMoveSS(VURegs * VU, int dstreg, int srcreg) +{ + assert( _XYZW_SS ); + if( _Y ) _unpackVFSS_xyzw(dstreg, srcreg, 1); + else if( _Z ) _unpackVFSS_xyzw(dstreg, srcreg, 2); + else if( _W ) _unpackVFSS_xyzw(dstreg, srcreg, 3); + else _unpackVFSS_xyzw(dstreg, srcreg, 0); +} + +// 1 - src, 0 - dest wzyx +void VU_MERGE0(int dest, int src) { // 0000s +} +void VU_MERGE1(int dest, int src) { // 1000 + SSE_MOVHLPS_XMM_to_XMM(src, dest); + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xc4); +} +void VU_MERGE1b(int dest, int src) { // 1000s + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); +} +void VU_MERGE2(int dest, int src) { // 0100 + SSE_MOVHLPS_XMM_to_XMM(src, dest); + SSE_SHUFPS_XMM_to_XMM(dest, src, 0x64); +} +void VU_MERGE2b(int dest, int src) { // 0100s + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); +} +void VU_MERGE3(int dest, int src) { // 1100s + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); +} +void VU_MERGE4(int dest, int src) { // 0010 + SSE_MOVSS_XMM_to_XMM(src, dest); + SSE2_MOVSD_XMM_to_XMM(dest, src); +} +void VU_MERGE4b(int dest, int src) { // 0010s + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); +} +void VU_MERGE5(int dest, int src) { // 1010 + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xd8); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xd8); +} +void VU_MERGE5b(int dest, int src) { // 1010s + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); +} +void VU_MERGE6(int dest, int src) { // 0110 + SSE_SHUFPS_XMM_to_XMM(dest, src, 0x9c); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x78); +} +void VU_MERGE6b(int dest, int src) { // 0110s + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); +} +void VU_MERGE7(int dest, int src) { // 1110 + SSE_MOVSS_XMM_to_XMM(src, dest); + SSE_MOVAPS_XMM_to_XMM(dest, src); +} +void VU_MERGE7b(int dest, int src) { // 1110s + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); +} +void VU_MERGE8(int dest, int src) { // 0001s + SSE_MOVSS_XMM_to_XMM(dest, src); +} +void VU_MERGE9(int dest, int src) { // 1001 + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xc9); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xd2); +} +void VU_MERGE9b(int dest, int src) { // 1001s + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); +} +void VU_MERGE10(int dest, int src) { // 0101 + SSE_SHUFPS_XMM_to_XMM(dest, src, 0x8d); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x72); +} +void VU_MERGE10b(int dest, int src) { // 0101s + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); +} +void VU_MERGE11(int dest, int src) { // 1101s + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(dest, src, 0xe4); +} +void VU_MERGE12(int dest, int src) { // 0011 + SSE2_MOVSD_XMM_to_XMM(dest, src); +} +void VU_MERGE13(int dest, int src) { // 1011 + SSE_MOVHLPS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, dest, 0x64); + SSE_MOVAPS_XMM_to_XMM(dest, src); +} +void VU_MERGE13b(int dest, int src) { // 1011s + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0x27); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0x27); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); +} +void VU_MERGE14(int dest, int src) { // 0111 + SSE_MOVHLPS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, dest, 0xc4); + SSE_MOVAPS_XMM_to_XMM(dest, src); +} +void VU_MERGE14b(int dest, int src) { // 0111s + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xE1); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xE1); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); + SSE_MOVSS_XMM_to_XMM(dest, src); + SSE_SHUFPS_XMM_to_XMM(src, src, 0xC6); + SSE_SHUFPS_XMM_to_XMM(dest, dest, 0xC6); +} +void VU_MERGE15(int dest, int src) { // 1111s + SSE_MOVAPS_XMM_to_XMM(dest, src); +} + +typedef void (*VUMERGEFN)(int dest, int src); + +static VUMERGEFN s_VuMerge[16] = { + VU_MERGE0, VU_MERGE1, VU_MERGE2, VU_MERGE3, + VU_MERGE4, VU_MERGE5, VU_MERGE6, VU_MERGE7, + VU_MERGE8, VU_MERGE9, VU_MERGE10, VU_MERGE11, + VU_MERGE12, VU_MERGE13, VU_MERGE14, VU_MERGE15 }; + +static VUMERGEFN s_VuMerge2[16] = { + VU_MERGE0, VU_MERGE1b, VU_MERGE2b, VU_MERGE3, + VU_MERGE4b, VU_MERGE5b, VU_MERGE6b, VU_MERGE7b, + VU_MERGE8, VU_MERGE9b, VU_MERGE10b, VU_MERGE11, + VU_MERGE12, VU_MERGE13b, VU_MERGE14b, VU_MERGE15 }; + +// Modifies the Source Reg! +void VU_MERGE_REGS_CUSTOM(int dest, int src, int xyzw) { + xyzw &= 0xf; + if ( (dest != src) && (xyzw != 0) ) { + if ( cpucaps.hasStreamingSIMD4Extensions && (xyzw != 0x8) && (xyzw != 0xf) ) { + xyzw = ((xyzw & 1) << 3) | ((xyzw & 2) << 1) | ((xyzw & 4) >> 1) | ((xyzw & 8) >> 3); + SSE4_BLENDPS_XMM_to_XMM(dest, src, xyzw); + } + else s_VuMerge[xyzw](dest, src); + } +} +// Doesn't Modify the Source Reg! (ToDo: s_VuMerge2() has room for optimization) +void VU_MERGE_REGS_SAFE(int dest, int src, int xyzw) { + xyzw &= 0xf; + if ( (dest != src) && (xyzw != 0) ) { + if ( cpucaps.hasStreamingSIMD4Extensions && (xyzw != 0x8) && (xyzw != 0xf) ) { + xyzw = ((xyzw & 1) << 3) | ((xyzw & 2) << 1) | ((xyzw & 4) >> 1) | ((xyzw & 8) >> 3); + SSE4_BLENDPS_XMM_to_XMM(dest, src, xyzw); + } + else s_VuMerge2[xyzw](dest, src); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// Misc VU Reg Clamping/Overflow Functions +//------------------------------------------------------------------ +#define CLAMP_NORMAL_SSE4(n) \ + SSE_MOVAPS_XMM_to_XMM(regTemp, regd);\ + SSE4_PMINUD_M128_to_XMM(regd, (uptr)&g_minvals_XYZW[n][0]);\ + SSE2_PSUBD_XMM_to_XMM(regTemp, regd);\ + SSE2_PCMPGTD_M128_to_XMM(regTemp, (uptr)&g_ones[0]);\ + SSE4_PMINSD_M128_to_XMM(regd, (uptr)&g_maxvals_XYZW[n][0]);\ + SSE2_PSLLD_I8_to_XMM(regTemp, 31);\ + SSE_XORPS_XMM_to_XMM(regd, regTemp); + +#define CLAMP_SIGN_SSE4(n) \ + SSE4_PMINSD_M128_to_XMM(regd, (uptr)&g_maxvals_XYZW[n][0]);\ + SSE4_PMINUD_M128_to_XMM(regd, (uptr)&g_minvals_XYZW[n][0]); + +void vFloat0(int regd, int regTemp) { } //0000 +void vFloat1(int regd, int regTemp) { //1000 + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); +} +void vFloat1c(int regd, int regTemp) { //1000 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(1); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat2(int regd, int regTemp) { //0100 + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); +} +void vFloat2c(int regd, int regTemp) { //0100 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(2); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat3(int regd, int regTemp) { //1100 + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); +} +void vFloat3b(int regd, int regTemp) { //1100 //regTemp is Modified + SSE2_MOVSD_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE2_MOVSD_XMM_to_XMM(regd, regTemp); +} +void vFloat3c(int regd, int regTemp) { //1100 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(3); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat4(int regd, int regTemp) { //0010 + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); +} +void vFloat4c(int regd, int regTemp) { //0010 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(4); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat5(int regd, int regTemp) { //1010 + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); +} +void vFloat5b(int regd, int regTemp) { //1010 //regTemp is Modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_NORMAL_SSE4(5); + } + else { + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); + } +} +void vFloat5c(int regd, int regTemp) { //1010 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(5); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat6(int regd, int regTemp) { //0110 + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); +} +void vFloat6b(int regd, int regTemp) { //0110 //regTemp is Modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_NORMAL_SSE4(6); + } + else { + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); + } +} +void vFloat6c(int regd, int regTemp) { //0110 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(6); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat7(int regd, int regTemp) { //1110 + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x39); +} +void vFloat7_useEAX(int regd, int regTemp) { //1110 //EAX is Modified + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + if ( cpucaps.hasStreamingSIMD4Extensions ) + SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); + else { + SSE_PINSRW_R32_to_XMM(regd, EAX, 0); + SHR32ItoR(EAX, 16); + SSE_PINSRW_R32_to_XMM(regd, EAX, 1); + } +} +void vFloat7b(int regd, int regTemp) { //1110 //regTemp is Modified + SSE_MOVSS_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_MOVSS_XMM_to_XMM(regd, regTemp); +} +void vFloat7c(int regd, int regTemp) { //1110 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(7); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x39); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat7c_useEAX(int regd, int regTemp) { //1110 //EAX is Modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(7); + } + else { + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + SSE2_MOVD_R_to_XMM(regTemp, EAX); + SSE_MOVSS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat8(int regd, int regTemp) { //0001 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); +} +void vFloat8c(int regd, int regTemp) { //0001 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(8); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat9(int regd, int regTemp) { //1001 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); +} +void vFloat9b(int regd, int regTemp) { //1001 //regTemp is Modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_NORMAL_SSE4(9); + } + else { + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + } +} +void vFloat9c(int regd, int regTemp) { //1001 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(9); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat10(int regd, int regTemp) { //0101 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); +} +void vFloat10b(int regd, int regTemp) { //0101 //regTemp is Modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_NORMAL_SSE4(10); + } + else { + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + } +} +void vFloat10c(int regd, int regTemp) { //0101 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(10); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat11(int regd, int regTemp) { //1101 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); +} +void vFloat11_useEAX(int regd, int regTemp) { //1101 //EAX is Modified + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + if ( cpucaps.hasStreamingSIMD4Extensions ) + SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); + else { + SSE_PINSRW_R32_to_XMM(regd, EAX, 0); + SHR32ItoR(EAX, 16); + SSE_PINSRW_R32_to_XMM(regd, EAX, 1); + } + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); +} +void vFloat11b(int regd, int regTemp) { //1101 //regTemp is Modified + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_MOVSS_XMM_to_XMM(regTemp, regd); + SSE2_MOVSD_XMM_to_XMM(regd, regTemp); +} +void vFloat11c(int regd, int regTemp) { //1101 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(11); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x36); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat11c_useEAX(int regd, int regTemp) { //1101 // EAX is modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(11); + } + else { + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + SSE2_MOVD_R_to_XMM(regTemp, EAX); + SSE_MOVSS_XMM_to_XMM(regd, regTemp); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + } +} +void vFloat12(int regd, int regTemp) { //0011 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); +} +void vFloat12b(int regd, int regTemp) { //0011 //regTemp is Modified + SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE2_PUNPCKLQDQ_XMM_to_XMM(regd, regTemp); +} +void vFloat12c(int regd, int regTemp) { //0011 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(12); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat13(int regd, int regTemp) { //1011 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); +} +void vFloat13_useEAX(int regd, int regTemp) { //1011 // EAX is modified + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + if ( cpucaps.hasStreamingSIMD4Extensions ) + SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); + else { + SSE_PINSRW_R32_to_XMM(regd, EAX, 0); + SHR32ItoR(EAX, 16); + SSE_PINSRW_R32_to_XMM(regd, EAX, 1); + } + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); +} +void vFloat13b(int regd, int regTemp) { //1011 //regTemp is Modified + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); + SSE_SHUFPS_XMM_to_XMM(regd, regTemp, 0x64); +} +void vFloat13c(int regd, int regTemp) { //1011 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(13); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x2d); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat13c_useEAX(int regd, int regTemp) { //1011 // EAX is modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(13); + } + else { + SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0xc6); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + SSE2_MOVD_R_to_XMM(regTemp, EAX); + SSE_MOVSS_XMM_to_XMM(regd, regTemp); + SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0xc6); + } +} +void vFloat14(int regd, int regTemp) { //0111 + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); +} +void vFloat14_useEAX(int regd, int regTemp) { //0111 // EAX is modified + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + if ( cpucaps.hasStreamingSIMD4Extensions ) + SSE4_PINSRD_R32_to_XMM(regd, EAX, 0x00); + else { + SSE_PINSRW_R32_to_XMM(regd, EAX, 0); + SHR32ItoR(EAX, 16); + SSE_PINSRW_R32_to_XMM(regd, EAX, 1); + } + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x27); +} +void vFloat14b(int regd, int regTemp) { //0111 //regTemp is Modified + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_MOVHLPS_XMM_to_XMM(regTemp, regd); + SSE_SHUFPS_XMM_to_XMM(regd, regTemp, 0xc4); +} +void vFloat14c(int regd, int regTemp) { //0111 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(14); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE2_PSHUFLW_XMM_to_XMM(regd, regd, 0x4e); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc6); + SSE_MINSS_M32_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXSS_M32_to_XMM(regd, (uptr)g_minvals); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0xc9); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} +void vFloat14c_useEAX(int regd, int regTemp) { //0111 // EAX is modified + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(14); + } + else { + SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0x27); + SSE2_MOVD_XMM_to_R(EAX, regd); + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + SSE2_MOVD_R_to_XMM(regTemp, EAX); + SSE_MOVSS_XMM_to_XMM(regd, regTemp); + SSE2_PSHUFD_XMM_to_XMM(regd, regd, 0x27); + } +} +void vFloat15(int regd, int regTemp) { //1111 + SSE_MINPS_M128_to_XMM(regd, (uptr)g_maxvals); + SSE_MAXPS_M128_to_XMM(regd, (uptr)g_minvals); +} +void vFloat15c(int regd, int regTemp) { //1111 + if ( cpucaps.hasStreamingSIMD4Extensions ) { + CLAMP_SIGN_SSE4(15); + } + else { + SSE_MOVAPS_XMM_to_XMM(regTemp, regd); + SSE_ANDPS_M128_to_XMM(regTemp, (uptr)&const_clip[4]); + SSE_MINPS_M128_to_XMM(regd, (uptr)&g_maxvals[0]); + SSE_MAXPS_M128_to_XMM(regd, (uptr)&g_minvals[0]); + SSE_ORPS_XMM_to_XMM(regd, regTemp); + } +} + +vFloat vFloats1[16] = { //regTemp is not modified + vFloat0, vFloat1, vFloat2, vFloat3, + vFloat4, vFloat5, vFloat6, vFloat7, + vFloat8, vFloat9, vFloat10, vFloat11, + vFloat12, vFloat13, vFloat14, vFloat15 }; + +vFloat vFloats1_useEAX[16] = { //regTemp is not modified but EAX is used + vFloat0, vFloat1, vFloat2, vFloat3, + vFloat4, vFloat5, vFloat6, vFloat7_useEAX, + vFloat8, vFloat9, vFloat10, vFloat11_useEAX, + vFloat12, vFloat13_useEAX, vFloat14_useEAX, vFloat15 }; + +vFloat vFloats2[16] = { //regTemp is modified + vFloat0, vFloat1, vFloat2, vFloat3b, + vFloat4, vFloat5b, vFloat6b, vFloat7b, + vFloat8, vFloat9b, vFloat10b, vFloat11b, + vFloat12b, vFloat13b, vFloat14b, vFloat15 }; + +vFloat vFloats4[16] = { //regTemp is modified + vFloat0, vFloat1c, vFloat2c, vFloat3c, + vFloat4c, vFloat5c, vFloat6c, vFloat7c, + vFloat8c, vFloat9c, vFloat10c, vFloat11c, + vFloat12c, vFloat13c, vFloat14c, vFloat15c }; + +vFloat vFloats4_useEAX[16] = { //regTemp is modified and EAX is used + vFloat0, vFloat1c, vFloat2c, vFloat3c, + vFloat4c, vFloat5c, vFloat6c, vFloat7c_useEAX, + vFloat8c, vFloat9c, vFloat10c, vFloat11c_useEAX, + vFloat12c, vFloat13c_useEAX, vFloat14c_useEAX, vFloat15c }; + +//------------------------------------------------------------------ +// Clamping Functions (wrapper for vFloat* functions) +// vuFloat : "normal" clamping +// vuFloat_useEAX : "normal" clamping (faster but EAX is modified) +// vuFloat2 : "normal" clamping (fastest but regTemp is modified) +// vuFloat3 : "preserve sign" clamping for pointer +// vuFloat4 : "preserve sign" clamping (regTemp is modified; *FASTEST* on SSE4 CPUs) +// vuFloat4_useEAX : "preserve sign" clamping (faster but regTemp and EAX are modified) +// vuFloat5 : wrapper function for vuFloat2 and vuFloat4 +// vuFloat5_useEAX : wrapper function for vuFloat2 and vuFloat4_useEAX +// vuFloatExtra : for debugging +// +// Notice 1: vuFloat*_useEAX may be slower on AMD CPUs, which have independent execution pipeline for +// vector and scalar instructions (need checks) +// Notice 2: recVUMI_MUL_xyzw_toD and recVUMI_MADD_xyzw_toD use vFloats directly! +//------------------------------------------------------------------ + +// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (doesn't use any temp regs) +void vuFloat( int info, int regd, int XYZW) { + if( CHECK_VU_OVERFLOW ) { + /*if ( (XYZW != 0) && (XYZW != 8) && (XYZW != 0xF) ) { + int t1reg = _vuGetTempXMMreg(info); + if (t1reg >= 0) { + vuFloat2( regd, t1reg, XYZW ); + _freeXMMreg( t1reg ); + return; + } + }*/ + //vuFloatExtra(regd, XYZW); + vFloats1[XYZW](regd, regd); + } +} + +// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (uses EAX as a temp register; faster but **destroys EAX**) +void vuFloat_useEAX( int info, int regd, int XYZW) { + if( CHECK_VU_OVERFLOW ) { + vFloats1_useEAX[XYZW](regd, regd); + } +} + +// Clamps +/-NaN to +fMax and +/-Inf to +/-fMax (uses a temp reg) +void vuFloat2(int regd, int regTemp, int XYZW) { + if( CHECK_VU_OVERFLOW ) { + //vuFloatExtra(regd, XYZW); + vFloats2[XYZW](regd, regTemp); + } +} + +// Clamps +/-NaN and +/-Inf to +/-fMax (uses a temp reg) +void vuFloat4(int regd, int regTemp, int XYZW) { + if( CHECK_VU_OVERFLOW ) { + vFloats4[XYZW](regd, regTemp); + } +} + +// Clamps +/-NaN and +/-Inf to +/-fMax (uses a temp reg, and uses EAX as a temp register; faster but **destroys EAX**) +void vuFloat4_useEAX(int regd, int regTemp, int XYZW) { + if( CHECK_VU_OVERFLOW ) { + vFloats4_useEAX[XYZW](regd, regTemp); + } +} + +// Uses vuFloat4 or vuFloat2 depending on the CHECK_VU_SIGN_OVERFLOW setting +void vuFloat5(int regd, int regTemp, int XYZW) { + if (CHECK_VU_SIGN_OVERFLOW) { + vuFloat4(regd, regTemp, XYZW); + } + else vuFloat2(regd, regTemp, XYZW); +} + +// Uses vuFloat4_useEAX or vuFloat2 depending on the CHECK_VU_SIGN_OVERFLOW setting (uses EAX as a temp register; faster but **destoroyes EAX**) +void vuFloat5_useEAX(int regd, int regTemp, int XYZW) { + if (CHECK_VU_SIGN_OVERFLOW) { + vuFloat4_useEAX(regd, regTemp, XYZW); + } + else vuFloat2(regd, regTemp, XYZW); +} + +// Clamps +/-infs to +/-fMax, and +/-NaNs to +/-fMax +void vuFloat3(uptr x86ptr) { + u8* pjmp; + + if( CHECK_VU_OVERFLOW ) { + CMP32ItoM(x86ptr, 0x7f800000 ); + pjmp = JL8(0); // Signed Comparison + MOV32ItoM(x86ptr, 0x7f7fffff ); + x86SetJ8(pjmp); + + CMP32ItoM(x86ptr, 0xff800000 ); + pjmp = JB8(0); // Unsigned Comparison + MOV32ItoM(x86ptr, 0xff7fffff ); + x86SetJ8(pjmp); + } +} + +PCSX2_ALIGNED16(u64 vuFloatData[2]); +PCSX2_ALIGNED16(u64 vuFloatData2[2]); +// Makes NaN == 0, Infinities stay the same; Very Slow - Use only for debugging +void vuFloatExtra( int regd, int XYZW) { + int t1reg = (regd == 0) ? (regd + 1) : (regd - 1); + int t2reg = (regd <= 1) ? (regd + 2) : (regd - 2); + SSE_MOVAPS_XMM_to_M128( (uptr)vuFloatData, t1reg ); + SSE_MOVAPS_XMM_to_M128( (uptr)vuFloatData2, t2reg ); + + SSE_XORPS_XMM_to_XMM(t1reg, t1reg); + SSE_CMPORDPS_XMM_to_XMM(t1reg, regd); + SSE_MOVAPS_XMM_to_XMM(t2reg, regd); + SSE_ANDPS_XMM_to_XMM(t2reg, t1reg); + VU_MERGE_REGS_CUSTOM(regd, t2reg, XYZW); + + SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)vuFloatData ); + SSE_MOVAPS_M128_to_XMM( t2reg, (uptr)vuFloatData2 ); +} + +static PCSX2_ALIGNED16(u32 tempRegX[]) = {0x00000000, 0x00000000, 0x00000000, 0x00000000}; + +// Called by testWhenOverflow() function +void testPrintOverflow() { + tempRegX[0] &= 0xff800000; + tempRegX[1] &= 0xff800000; + tempRegX[2] &= 0xff800000; + tempRegX[3] &= 0xff800000; + if ( (tempRegX[0] == 0x7f800000) || (tempRegX[1] == 0x7f800000) || (tempRegX[2] == 0x7f800000) || (tempRegX[3] == 0x7f800000) ) + Console::Notice( "VU OVERFLOW!: Changing to +Fmax!!!!!!!!!!!!" ); + if ( (tempRegX[0] == 0xff800000) || (tempRegX[1] == 0xff800000) || (tempRegX[2] == 0xff800000) || (tempRegX[3] == 0xff800000) ) + Console::Notice( "VU OVERFLOW!: Changing to -Fmax!!!!!!!!!!!!" ); +} + +// Outputs to the console when overflow has occured. +void testWhenOverflow(int info, int regd, int t0reg) { + SSE_MOVAPS_XMM_to_M128((uptr)tempRegX, regd); + CALLFunc((uptr)testPrintOverflow); +} diff --git a/pcsx2/x86/iVUmicro.h b/pcsx2/x86/sVU_Micro.h similarity index 97% rename from pcsx2/x86/iVUmicro.h rename to pcsx2/x86/sVU_Micro.h index 777a7dc063..214a1c425c 100644 --- a/pcsx2/x86/iVUmicro.h +++ b/pcsx2/x86/sVU_Micro.h @@ -1,288 +1,286 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2003 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#ifndef __IVUMICRO_H__ -#define __IVUMICRO_H__ - -#include "VUmicro.h" - -extern u32 vudump; - -#define VU0_MEMSIZE 0x1000 -#define VU1_MEMSIZE 0x4000 - -void recResetVU0(); -void recExecuteVU0Block(); -void recClearVU0( u32 Addr, u32 Size ); - -void recVU1Init(); -void recVU1Shutdown(); -void recResetVU1(); -void recExecuteVU1Block(); -void recClearVU1( u32 Addr, u32 Size ); - - -u32 GetVIAddr(VURegs * VU, int reg, int read, int info); // returns the correct VI addr -void recUpdateFlags(VURegs * VU, int reg, int info); - -void _recvuTestPipes(VURegs * VU); -void _recvuFlushFDIV(VURegs * VU); -void _recvuTestUpperStalls(VURegs * VU, _VURegsNum *VUregsn); -void _recvuTestLowerStalls(VURegs * VU, _VURegsNum *VUregsn); -void _recvuAddUpperStalls(VURegs * VU, _VURegsNum *VUregsn); -void _recvuAddLowerStalls(VURegs * VU, _VURegsNum *VUregsn); - -#define VUOP_READ 2 -#define VUOP_WRITE 4 - -// save on mem -struct _vuopinfo { - int cycle; - int cycles; - u8 statusflag; - u8 macflag; - u8 clipflag; - u8 dummy; - u8 q; - u8 p; - u16 pqinst; // bit of instruction specifying index (srec only) -}; - -void SuperVUAnalyzeOp(VURegs *VU, _vuopinfo *info, _VURegsNum* pCodeRegs); -int eeVURecompileCode(VURegs *VU, _VURegsNum* regs); // allocates all the necessary regs and returns the indices -void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr); // used for MTGS in XGKICK - -extern int vucycle; -typedef void (*vFloat)(int regd, int regTemp); -extern vFloat vFloats1[16]; -extern vFloat vFloats1_useEAX[16]; -extern vFloat vFloats2[16]; -extern vFloat vFloats4[16]; -extern vFloat vFloats4_useEAX[16]; -PCSX2_ALIGNED16_EXTERN(const float s_fones[8]); -PCSX2_ALIGNED16_EXTERN(const u32 s_mask[4]); -PCSX2_ALIGNED16_EXTERN(const u32 s_expmask[4]); -PCSX2_ALIGNED16_EXTERN(const u32 g_minvals[4]); -PCSX2_ALIGNED16_EXTERN(const u32 g_maxvals[4]); -PCSX2_ALIGNED16_EXTERN(const u32 const_clip[8]); - -u32 GetVIAddr(VURegs * VU, int reg, int read, int info); -int _vuGetTempXMMreg(int info); -void vuFloat(int info, int regd, int XYZW); -void vuFloat_useEAX(int regd, int regTemp, int XYZW); -void vuFloat2(int regd, int regTemp, int XYZW); -void vuFloat3(uptr x86ptr); -void vuFloat4(int regd, int regTemp, int XYZW); -void vuFloat4_useEAX(int regd, int regTemp, int XYZW); -void vuFloat5(int regd, int regTemp, int XYZW); -void vuFloat5_useEAX(int regd, int regTemp, int XYZW); -void _vuFlipRegSS(VURegs * VU, int reg); -void _vuFlipRegSS_xyzw(int reg, int xyzw); -void _vuMoveSS(VURegs * VU, int dstreg, int srcreg); -void _unpackVF_xyzw(int dstreg, int srcreg, int xyzw); -void _unpackVFSS_xyzw(int dstreg, int srcreg, int xyzw); -void VU_MERGE_REGS_CUSTOM(int dest, int src, int xyzw); -void VU_MERGE_REGS_SAFE(int dest, int src, int xyzw); -#define VU_MERGE_REGS(dest, src) { \ - VU_MERGE_REGS_CUSTOM(dest, src, _X_Y_Z_W); \ -} - -// use for allocating vi regs -#define ALLOCTEMPX86(mode) _allocX86reg(-1, X86TYPE_TEMP, 0, ((info&PROCESS_VU_SUPER)?0:MODE_NOFRAME)|mode) -#define ALLOCVI(vi, mode) _allocX86reg(-1, X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), vi, ((info&PROCESS_VU_SUPER)?0:MODE_NOFRAME)|mode) -#define ADD_VI_NEEDED(vi) _addNeededX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), vi); - -#define SWAP(x, y) *(u32*)&y ^= *(u32*)&x ^= *(u32*)&y ^= *(u32*)&x; - -/***************************************** - VU Micromode Upper instructions -*****************************************/ - -void recVUMI_ABS(VURegs *vuRegs, int info); -void recVUMI_ADD(VURegs *vuRegs, int info); -void recVUMI_ADDi(VURegs *vuRegs, int info); -void recVUMI_ADDq(VURegs *vuRegs, int info); -void recVUMI_ADDx(VURegs *vuRegs, int info); -void recVUMI_ADDy(VURegs *vuRegs, int info); -void recVUMI_ADDz(VURegs *vuRegs, int info); -void recVUMI_ADDw(VURegs *vuRegs, int info); -void recVUMI_ADDA(VURegs *vuRegs, int info); -void recVUMI_ADDAi(VURegs *vuRegs, int info); -void recVUMI_ADDAq(VURegs *vuRegs, int info); -void recVUMI_ADDAx(VURegs *vuRegs, int info); -void recVUMI_ADDAy(VURegs *vuRegs, int info); -void recVUMI_ADDAz(VURegs *vuRegs, int info); -void recVUMI_ADDAw(VURegs *vuRegs, int info); -void recVUMI_SUB(VURegs *vuRegs, int info); -void recVUMI_SUBi(VURegs *vuRegs, int info); -void recVUMI_SUBq(VURegs *vuRegs, int info); -void recVUMI_SUBx(VURegs *vuRegs, int info); -void recVUMI_SUBy(VURegs *vuRegs, int info); -void recVUMI_SUBz(VURegs *vuRegs, int info); -void recVUMI_SUBw(VURegs *vuRegs, int info); -void recVUMI_SUBA(VURegs *vuRegs, int info); -void recVUMI_SUBAi(VURegs *vuRegs, int info); -void recVUMI_SUBAq(VURegs *vuRegs, int info); -void recVUMI_SUBAx(VURegs *vuRegs, int info); -void recVUMI_SUBAy(VURegs *vuRegs, int info); -void recVUMI_SUBAz(VURegs *vuRegs, int info); -void recVUMI_SUBAw(VURegs *vuRegs, int info); -void recVUMI_MUL(VURegs *vuRegs, int info); -void recVUMI_MULi(VURegs *vuRegs, int info); -void recVUMI_MULq(VURegs *vuRegs, int info); -void recVUMI_MULx(VURegs *vuRegs, int info); -void recVUMI_MULy(VURegs *vuRegs, int info); -void recVUMI_MULz(VURegs *vuRegs, int info); -void recVUMI_MULw(VURegs *vuRegs, int info); -void recVUMI_MULA(VURegs *vuRegs, int info); -void recVUMI_MULAi(VURegs *vuRegs, int info); -void recVUMI_MULAq(VURegs *vuRegs, int info); -void recVUMI_MULAx(VURegs *vuRegs, int info); -void recVUMI_MULAy(VURegs *vuRegs, int info); -void recVUMI_MULAz(VURegs *vuRegs, int info); -void recVUMI_MULAw(VURegs *vuRegs, int info); -void recVUMI_MADD(VURegs *vuRegs, int info); -void recVUMI_MADDi(VURegs *vuRegs, int info); -void recVUMI_MADDq(VURegs *vuRegs, int info); -void recVUMI_MADDx(VURegs *vuRegs, int info); -void recVUMI_MADDy(VURegs *vuRegs, int info); -void recVUMI_MADDz(VURegs *vuRegs, int info); -void recVUMI_MADDw(VURegs *vuRegs, int info); -void recVUMI_MADDA(VURegs *vuRegs, int info); -void recVUMI_MADDAi(VURegs *vuRegs, int info); -void recVUMI_MADDAq(VURegs *vuRegs, int info); -void recVUMI_MADDAx(VURegs *vuRegs, int info); -void recVUMI_MADDAy(VURegs *vuRegs, int info); -void recVUMI_MADDAz(VURegs *vuRegs, int info); -void recVUMI_MADDAw(VURegs *vuRegs, int info); -void recVUMI_MSUB(VURegs *vuRegs, int info); -void recVUMI_MSUBi(VURegs *vuRegs, int info); -void recVUMI_MSUBq(VURegs *vuRegs, int info); -void recVUMI_MSUBx(VURegs *vuRegs, int info); -void recVUMI_MSUBy(VURegs *vuRegs, int info); -void recVUMI_MSUBz(VURegs *vuRegs, int info); -void recVUMI_MSUBw(VURegs *vuRegs, int info); -void recVUMI_MSUBA(VURegs *vuRegs, int info); -void recVUMI_MSUBAi(VURegs *vuRegs, int info); -void recVUMI_MSUBAq(VURegs *vuRegs, int info); -void recVUMI_MSUBAx(VURegs *vuRegs, int info); -void recVUMI_MSUBAy(VURegs *vuRegs, int info); -void recVUMI_MSUBAz(VURegs *vuRegs, int info); -void recVUMI_MSUBAw(VURegs *vuRegs, int info); -void recVUMI_MAX(VURegs *vuRegs, int info); -void recVUMI_MAXi(VURegs *vuRegs, int info); -void recVUMI_MAXx(VURegs *vuRegs, int info); -void recVUMI_MAXy(VURegs *vuRegs, int info); -void recVUMI_MAXz(VURegs *vuRegs, int info); -void recVUMI_MAXw(VURegs *vuRegs, int info); -void recVUMI_MINI(VURegs *vuRegs, int info); -void recVUMI_MINIi(VURegs *vuRegs, int info); -void recVUMI_MINIx(VURegs *vuRegs, int info); -void recVUMI_MINIy(VURegs *vuRegs, int info); -void recVUMI_MINIz(VURegs *vuRegs, int info); -void recVUMI_MINIw(VURegs *vuRegs, int info); -void recVUMI_OPMULA(VURegs *vuRegs, int info); -void recVUMI_OPMSUB(VURegs *vuRegs, int info); -void recVUMI_NOP(VURegs *vuRegs, int info); -void recVUMI_FTOI0(VURegs *vuRegs, int info); -void recVUMI_FTOI4(VURegs *vuRegs, int info); -void recVUMI_FTOI12(VURegs *vuRegs, int info); -void recVUMI_FTOI15(VURegs *vuRegs, int info); -void recVUMI_ITOF0(VURegs *vuRegs, int info); -void recVUMI_ITOF4(VURegs *vuRegs, int info); -void recVUMI_ITOF12(VURegs *vuRegs, int info); -void recVUMI_ITOF15(VURegs *vuRegs, int info); -void recVUMI_CLIP(VURegs *vuRegs, int info); - -/***************************************** - VU Micromode Lower instructions -*****************************************/ - -void recVUMI_DIV(VURegs *vuRegs, int info); -void recVUMI_SQRT(VURegs *vuRegs, int info); -void recVUMI_RSQRT(VURegs *vuRegs, int info); -void recVUMI_IADD(VURegs *vuRegs, int info); -void recVUMI_IADDI(VURegs *vuRegs, int info); -void recVUMI_IADDIU(VURegs *vuRegs, int info); -void recVUMI_IAND(VURegs *vuRegs, int info); -void recVUMI_IOR(VURegs *vuRegs, int info); -void recVUMI_ISUB(VURegs *vuRegs, int info); -void recVUMI_ISUBIU(VURegs *vuRegs, int info); -void recVUMI_MOVE(VURegs *vuRegs, int info); -void recVUMI_MFIR(VURegs *vuRegs, int info); -void recVUMI_MTIR(VURegs *vuRegs, int info); -void recVUMI_MR32(VURegs *vuRegs, int info); -void recVUMI_LQ(VURegs *vuRegs, int info); -void recVUMI_LQD(VURegs *vuRegs, int info); -void recVUMI_LQI(VURegs *vuRegs, int info); -void recVUMI_SQ(VURegs *vuRegs, int info); -void recVUMI_SQD(VURegs *vuRegs, int info); -void recVUMI_SQI(VURegs *vuRegs, int info); -void recVUMI_ILW(VURegs *vuRegs, int info); -void recVUMI_ISW(VURegs *vuRegs, int info); -void recVUMI_ILWR(VURegs *vuRegs, int info); -void recVUMI_ISWR(VURegs *vuRegs, int info); -void recVUMI_LOI(VURegs *vuRegs, int info); -void recVUMI_RINIT(VURegs *vuRegs, int info); -void recVUMI_RGET(VURegs *vuRegs, int info); -void recVUMI_RNEXT(VURegs *vuRegs, int info); -void recVUMI_RXOR(VURegs *vuRegs, int info); -void recVUMI_WAITQ(VURegs *vuRegs, int info); -void recVUMI_FSAND(VURegs *vuRegs, int info); -void recVUMI_FSEQ(VURegs *vuRegs, int info); -void recVUMI_FSOR(VURegs *vuRegs, int info); -void recVUMI_FSSET(VURegs *vuRegs, int info); -void recVUMI_FMAND(VURegs *vuRegs, int info); -void recVUMI_FMEQ(VURegs *vuRegs, int info); -void recVUMI_FMOR(VURegs *vuRegs, int info); -void recVUMI_FCAND(VURegs *vuRegs, int info); -void recVUMI_FCEQ(VURegs *vuRegs, int info); -void recVUMI_FCOR(VURegs *vuRegs, int info); -void recVUMI_FCSET(VURegs *vuRegs, int info); -void recVUMI_FCGET(VURegs *vuRegs, int info); -void recVUMI_IBEQ(VURegs *vuRegs, int info); -void recVUMI_IBGEZ(VURegs *vuRegs, int info); -void recVUMI_IBGTZ(VURegs *vuRegs, int info); -void recVUMI_IBLTZ(VURegs *vuRegs, int info); -void recVUMI_IBLEZ(VURegs *vuRegs, int info); -void recVUMI_IBNE(VURegs *vuRegs, int info); -void recVUMI_B(VURegs *vuRegs, int info); -void recVUMI_BAL(VURegs *vuRegs, int info); -void recVUMI_JR(VURegs *vuRegs, int info); -void recVUMI_JALR(VURegs *vuRegs, int info); -void recVUMI_MFP(VURegs *vuRegs, int info); -void recVUMI_WAITP(VURegs *vuRegs, int info); -void recVUMI_ESADD(VURegs *vuRegs, int info); -void recVUMI_ERSADD(VURegs *vuRegs, int info); -void recVUMI_ELENG(VURegs *vuRegs, int info); -void recVUMI_ERLENG(VURegs *vuRegs, int info); -void recVUMI_EATANxy(VURegs *vuRegs, int info); -void recVUMI_EATANxz(VURegs *vuRegs, int info); -void recVUMI_ESUM(VURegs *vuRegs, int info); -void recVUMI_ERCPR(VURegs *vuRegs, int info); -void recVUMI_ESQRT(VURegs *vuRegs, int info); -void recVUMI_ERSQRT(VURegs *vuRegs, int info); -void recVUMI_ESIN(VURegs *vuRegs, int info); -void recVUMI_EATAN(VURegs *vuRegs, int info); -void recVUMI_EEXP(VURegs *vuRegs, int info); -void recVUMI_XGKICK(VURegs *vuRegs, int info); -void recVUMI_XTOP(VURegs *vuRegs, int info); -void recVUMI_XITOP(VURegs *vuRegs, int info); -void recVUMI_XTOP( VURegs *VU , int info); - -#endif /* __IVUMICRO_H__ */ +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2003 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#pragma once + +#include "VUmicro.h" + +extern u32 vudump; + +#define VU0_MEMSIZE 0x1000 +#define VU1_MEMSIZE 0x4000 + +void recResetVU0(); +void recExecuteVU0Block(); +void recClearVU0( u32 Addr, u32 Size ); + +void recVU1Init(); +void recVU1Shutdown(); +void recResetVU1(); +void recExecuteVU1Block(); +void recClearVU1( u32 Addr, u32 Size ); + + +u32 GetVIAddr(VURegs * VU, int reg, int read, int info); // returns the correct VI addr +void recUpdateFlags(VURegs * VU, int reg, int info); + +void _recvuTestPipes(VURegs * VU); +void _recvuFlushFDIV(VURegs * VU); +void _recvuTestUpperStalls(VURegs * VU, _VURegsNum *VUregsn); +void _recvuTestLowerStalls(VURegs * VU, _VURegsNum *VUregsn); +void _recvuAddUpperStalls(VURegs * VU, _VURegsNum *VUregsn); +void _recvuAddLowerStalls(VURegs * VU, _VURegsNum *VUregsn); + +#define VUOP_READ 2 +#define VUOP_WRITE 4 + +// save on mem +struct _vuopinfo { + int cycle; + int cycles; + u8 statusflag; + u8 macflag; + u8 clipflag; + u8 dummy; + u8 q; + u8 p; + u16 pqinst; // bit of instruction specifying index (srec only) +}; + +void SuperVUAnalyzeOp(VURegs *VU, _vuopinfo *info, _VURegsNum* pCodeRegs); +int eeVURecompileCode(VURegs *VU, _VURegsNum* regs); // allocates all the necessary regs and returns the indices +void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr); // used for MTGS in XGKICK + +extern int vucycle; +typedef void (*vFloat)(int regd, int regTemp); +extern vFloat vFloats1[16]; +extern vFloat vFloats1_useEAX[16]; +extern vFloat vFloats2[16]; +extern vFloat vFloats4[16]; +extern vFloat vFloats4_useEAX[16]; +PCSX2_ALIGNED16_EXTERN(const float s_fones[8]); +PCSX2_ALIGNED16_EXTERN(const u32 s_mask[4]); +PCSX2_ALIGNED16_EXTERN(const u32 s_expmask[4]); +PCSX2_ALIGNED16_EXTERN(const u32 g_minvals[4]); +PCSX2_ALIGNED16_EXTERN(const u32 g_maxvals[4]); +PCSX2_ALIGNED16_EXTERN(const u32 const_clip[8]); + +u32 GetVIAddr(VURegs * VU, int reg, int read, int info); +int _vuGetTempXMMreg(int info); +void vuFloat(int info, int regd, int XYZW); +void vuFloat_useEAX(int regd, int regTemp, int XYZW); +void vuFloat2(int regd, int regTemp, int XYZW); +void vuFloat3(uptr x86ptr); +void vuFloat4(int regd, int regTemp, int XYZW); +void vuFloat4_useEAX(int regd, int regTemp, int XYZW); +void vuFloat5(int regd, int regTemp, int XYZW); +void vuFloat5_useEAX(int regd, int regTemp, int XYZW); +void _vuFlipRegSS(VURegs * VU, int reg); +void _vuFlipRegSS_xyzw(int reg, int xyzw); +void _vuMoveSS(VURegs * VU, int dstreg, int srcreg); +void _unpackVF_xyzw(int dstreg, int srcreg, int xyzw); +void _unpackVFSS_xyzw(int dstreg, int srcreg, int xyzw); +void VU_MERGE_REGS_CUSTOM(int dest, int src, int xyzw); +void VU_MERGE_REGS_SAFE(int dest, int src, int xyzw); +#define VU_MERGE_REGS(dest, src) { \ + VU_MERGE_REGS_CUSTOM(dest, src, _X_Y_Z_W); \ +} + +// use for allocating vi regs +#define ALLOCTEMPX86(mode) _allocX86reg(-1, X86TYPE_TEMP, 0, ((info&PROCESS_VU_SUPER)?0:MODE_NOFRAME)|mode) +#define ALLOCVI(vi, mode) _allocX86reg(-1, X86TYPE_VI|((VU==&VU1)?X86TYPE_VU1:0), vi, ((info&PROCESS_VU_SUPER)?0:MODE_NOFRAME)|mode) +#define ADD_VI_NEEDED(vi) _addNeededX86reg(X86TYPE_VI|(VU==&VU1?X86TYPE_VU1:0), vi); + +#define SWAP(x, y) *(u32*)&y ^= *(u32*)&x ^= *(u32*)&y ^= *(u32*)&x; + +/***************************************** + VU Micromode Upper instructions +*****************************************/ + +void recVUMI_ABS(VURegs *vuRegs, int info); +void recVUMI_ADD(VURegs *vuRegs, int info); +void recVUMI_ADDi(VURegs *vuRegs, int info); +void recVUMI_ADDq(VURegs *vuRegs, int info); +void recVUMI_ADDx(VURegs *vuRegs, int info); +void recVUMI_ADDy(VURegs *vuRegs, int info); +void recVUMI_ADDz(VURegs *vuRegs, int info); +void recVUMI_ADDw(VURegs *vuRegs, int info); +void recVUMI_ADDA(VURegs *vuRegs, int info); +void recVUMI_ADDAi(VURegs *vuRegs, int info); +void recVUMI_ADDAq(VURegs *vuRegs, int info); +void recVUMI_ADDAx(VURegs *vuRegs, int info); +void recVUMI_ADDAy(VURegs *vuRegs, int info); +void recVUMI_ADDAz(VURegs *vuRegs, int info); +void recVUMI_ADDAw(VURegs *vuRegs, int info); +void recVUMI_SUB(VURegs *vuRegs, int info); +void recVUMI_SUBi(VURegs *vuRegs, int info); +void recVUMI_SUBq(VURegs *vuRegs, int info); +void recVUMI_SUBx(VURegs *vuRegs, int info); +void recVUMI_SUBy(VURegs *vuRegs, int info); +void recVUMI_SUBz(VURegs *vuRegs, int info); +void recVUMI_SUBw(VURegs *vuRegs, int info); +void recVUMI_SUBA(VURegs *vuRegs, int info); +void recVUMI_SUBAi(VURegs *vuRegs, int info); +void recVUMI_SUBAq(VURegs *vuRegs, int info); +void recVUMI_SUBAx(VURegs *vuRegs, int info); +void recVUMI_SUBAy(VURegs *vuRegs, int info); +void recVUMI_SUBAz(VURegs *vuRegs, int info); +void recVUMI_SUBAw(VURegs *vuRegs, int info); +void recVUMI_MUL(VURegs *vuRegs, int info); +void recVUMI_MULi(VURegs *vuRegs, int info); +void recVUMI_MULq(VURegs *vuRegs, int info); +void recVUMI_MULx(VURegs *vuRegs, int info); +void recVUMI_MULy(VURegs *vuRegs, int info); +void recVUMI_MULz(VURegs *vuRegs, int info); +void recVUMI_MULw(VURegs *vuRegs, int info); +void recVUMI_MULA(VURegs *vuRegs, int info); +void recVUMI_MULAi(VURegs *vuRegs, int info); +void recVUMI_MULAq(VURegs *vuRegs, int info); +void recVUMI_MULAx(VURegs *vuRegs, int info); +void recVUMI_MULAy(VURegs *vuRegs, int info); +void recVUMI_MULAz(VURegs *vuRegs, int info); +void recVUMI_MULAw(VURegs *vuRegs, int info); +void recVUMI_MADD(VURegs *vuRegs, int info); +void recVUMI_MADDi(VURegs *vuRegs, int info); +void recVUMI_MADDq(VURegs *vuRegs, int info); +void recVUMI_MADDx(VURegs *vuRegs, int info); +void recVUMI_MADDy(VURegs *vuRegs, int info); +void recVUMI_MADDz(VURegs *vuRegs, int info); +void recVUMI_MADDw(VURegs *vuRegs, int info); +void recVUMI_MADDA(VURegs *vuRegs, int info); +void recVUMI_MADDAi(VURegs *vuRegs, int info); +void recVUMI_MADDAq(VURegs *vuRegs, int info); +void recVUMI_MADDAx(VURegs *vuRegs, int info); +void recVUMI_MADDAy(VURegs *vuRegs, int info); +void recVUMI_MADDAz(VURegs *vuRegs, int info); +void recVUMI_MADDAw(VURegs *vuRegs, int info); +void recVUMI_MSUB(VURegs *vuRegs, int info); +void recVUMI_MSUBi(VURegs *vuRegs, int info); +void recVUMI_MSUBq(VURegs *vuRegs, int info); +void recVUMI_MSUBx(VURegs *vuRegs, int info); +void recVUMI_MSUBy(VURegs *vuRegs, int info); +void recVUMI_MSUBz(VURegs *vuRegs, int info); +void recVUMI_MSUBw(VURegs *vuRegs, int info); +void recVUMI_MSUBA(VURegs *vuRegs, int info); +void recVUMI_MSUBAi(VURegs *vuRegs, int info); +void recVUMI_MSUBAq(VURegs *vuRegs, int info); +void recVUMI_MSUBAx(VURegs *vuRegs, int info); +void recVUMI_MSUBAy(VURegs *vuRegs, int info); +void recVUMI_MSUBAz(VURegs *vuRegs, int info); +void recVUMI_MSUBAw(VURegs *vuRegs, int info); +void recVUMI_MAX(VURegs *vuRegs, int info); +void recVUMI_MAXi(VURegs *vuRegs, int info); +void recVUMI_MAXx(VURegs *vuRegs, int info); +void recVUMI_MAXy(VURegs *vuRegs, int info); +void recVUMI_MAXz(VURegs *vuRegs, int info); +void recVUMI_MAXw(VURegs *vuRegs, int info); +void recVUMI_MINI(VURegs *vuRegs, int info); +void recVUMI_MINIi(VURegs *vuRegs, int info); +void recVUMI_MINIx(VURegs *vuRegs, int info); +void recVUMI_MINIy(VURegs *vuRegs, int info); +void recVUMI_MINIz(VURegs *vuRegs, int info); +void recVUMI_MINIw(VURegs *vuRegs, int info); +void recVUMI_OPMULA(VURegs *vuRegs, int info); +void recVUMI_OPMSUB(VURegs *vuRegs, int info); +void recVUMI_NOP(VURegs *vuRegs, int info); +void recVUMI_FTOI0(VURegs *vuRegs, int info); +void recVUMI_FTOI4(VURegs *vuRegs, int info); +void recVUMI_FTOI12(VURegs *vuRegs, int info); +void recVUMI_FTOI15(VURegs *vuRegs, int info); +void recVUMI_ITOF0(VURegs *vuRegs, int info); +void recVUMI_ITOF4(VURegs *vuRegs, int info); +void recVUMI_ITOF12(VURegs *vuRegs, int info); +void recVUMI_ITOF15(VURegs *vuRegs, int info); +void recVUMI_CLIP(VURegs *vuRegs, int info); + +/***************************************** + VU Micromode Lower instructions +*****************************************/ + +void recVUMI_DIV(VURegs *vuRegs, int info); +void recVUMI_SQRT(VURegs *vuRegs, int info); +void recVUMI_RSQRT(VURegs *vuRegs, int info); +void recVUMI_IADD(VURegs *vuRegs, int info); +void recVUMI_IADDI(VURegs *vuRegs, int info); +void recVUMI_IADDIU(VURegs *vuRegs, int info); +void recVUMI_IAND(VURegs *vuRegs, int info); +void recVUMI_IOR(VURegs *vuRegs, int info); +void recVUMI_ISUB(VURegs *vuRegs, int info); +void recVUMI_ISUBIU(VURegs *vuRegs, int info); +void recVUMI_MOVE(VURegs *vuRegs, int info); +void recVUMI_MFIR(VURegs *vuRegs, int info); +void recVUMI_MTIR(VURegs *vuRegs, int info); +void recVUMI_MR32(VURegs *vuRegs, int info); +void recVUMI_LQ(VURegs *vuRegs, int info); +void recVUMI_LQD(VURegs *vuRegs, int info); +void recVUMI_LQI(VURegs *vuRegs, int info); +void recVUMI_SQ(VURegs *vuRegs, int info); +void recVUMI_SQD(VURegs *vuRegs, int info); +void recVUMI_SQI(VURegs *vuRegs, int info); +void recVUMI_ILW(VURegs *vuRegs, int info); +void recVUMI_ISW(VURegs *vuRegs, int info); +void recVUMI_ILWR(VURegs *vuRegs, int info); +void recVUMI_ISWR(VURegs *vuRegs, int info); +void recVUMI_LOI(VURegs *vuRegs, int info); +void recVUMI_RINIT(VURegs *vuRegs, int info); +void recVUMI_RGET(VURegs *vuRegs, int info); +void recVUMI_RNEXT(VURegs *vuRegs, int info); +void recVUMI_RXOR(VURegs *vuRegs, int info); +void recVUMI_WAITQ(VURegs *vuRegs, int info); +void recVUMI_FSAND(VURegs *vuRegs, int info); +void recVUMI_FSEQ(VURegs *vuRegs, int info); +void recVUMI_FSOR(VURegs *vuRegs, int info); +void recVUMI_FSSET(VURegs *vuRegs, int info); +void recVUMI_FMAND(VURegs *vuRegs, int info); +void recVUMI_FMEQ(VURegs *vuRegs, int info); +void recVUMI_FMOR(VURegs *vuRegs, int info); +void recVUMI_FCAND(VURegs *vuRegs, int info); +void recVUMI_FCEQ(VURegs *vuRegs, int info); +void recVUMI_FCOR(VURegs *vuRegs, int info); +void recVUMI_FCSET(VURegs *vuRegs, int info); +void recVUMI_FCGET(VURegs *vuRegs, int info); +void recVUMI_IBEQ(VURegs *vuRegs, int info); +void recVUMI_IBGEZ(VURegs *vuRegs, int info); +void recVUMI_IBGTZ(VURegs *vuRegs, int info); +void recVUMI_IBLTZ(VURegs *vuRegs, int info); +void recVUMI_IBLEZ(VURegs *vuRegs, int info); +void recVUMI_IBNE(VURegs *vuRegs, int info); +void recVUMI_B(VURegs *vuRegs, int info); +void recVUMI_BAL(VURegs *vuRegs, int info); +void recVUMI_JR(VURegs *vuRegs, int info); +void recVUMI_JALR(VURegs *vuRegs, int info); +void recVUMI_MFP(VURegs *vuRegs, int info); +void recVUMI_WAITP(VURegs *vuRegs, int info); +void recVUMI_ESADD(VURegs *vuRegs, int info); +void recVUMI_ERSADD(VURegs *vuRegs, int info); +void recVUMI_ELENG(VURegs *vuRegs, int info); +void recVUMI_ERLENG(VURegs *vuRegs, int info); +void recVUMI_EATANxy(VURegs *vuRegs, int info); +void recVUMI_EATANxz(VURegs *vuRegs, int info); +void recVUMI_ESUM(VURegs *vuRegs, int info); +void recVUMI_ERCPR(VURegs *vuRegs, int info); +void recVUMI_ESQRT(VURegs *vuRegs, int info); +void recVUMI_ERSQRT(VURegs *vuRegs, int info); +void recVUMI_ESIN(VURegs *vuRegs, int info); +void recVUMI_EATAN(VURegs *vuRegs, int info); +void recVUMI_EEXP(VURegs *vuRegs, int info); +void recVUMI_XGKICK(VURegs *vuRegs, int info); +void recVUMI_XTOP(VURegs *vuRegs, int info); +void recVUMI_XITOP(VURegs *vuRegs, int info); +void recVUMI_XTOP( VURegs *VU , int info); + diff --git a/pcsx2/x86/iVUmicroUpper.cpp b/pcsx2/x86/sVU_Upper.cpp similarity index 97% rename from pcsx2/x86/iVUmicroUpper.cpp rename to pcsx2/x86/sVU_Upper.cpp index 61393c43fe..6e06ccdf08 100644 --- a/pcsx2/x86/iVUmicroUpper.cpp +++ b/pcsx2/x86/sVU_Upper.cpp @@ -1,3070 +1,3070 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include "PrecompiledHeader.h" - -#include "Common.h" -#include "GS.h" -#include "R5900OpcodeTables.h" -#include "iR5900.h" -#include "iMMI.h" -#include "iFPU.h" -#include "iCOP0.h" -#include "VUmicro.h" -#include "VUflags.h" -#include "iVUmicro.h" -#include "iVUops.h" -#include "iVUzerorec.h" -//------------------------------------------------------------------ -#define MINMAXFIX 1 -//------------------------------------------------------------------ -// Helper Macros -//------------------------------------------------------------------ -#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register -#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register -#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register - -#define _X (( VU->code>>24) & 0x1) -#define _Y (( VU->code>>23) & 0x1) -#define _Z (( VU->code>>22) & 0x1) -#define _W (( VU->code>>21) & 0x1) - -#define _XYZW_SS (_X+_Y+_Z+_W==1) - -#define _Fsf_ (( VU->code >> 21) & 0x03) -#define _Ftf_ (( VU->code >> 23) & 0x03) - -#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) -#define _UImm11_ (s32)(VU->code & 0x7ff) - -#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] -#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] -#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] -#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] - -#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] -#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] -#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] - -#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) - -#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] -#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] -#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] -#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] - -#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// Global Variables -//------------------------------------------------------------------ -static const PCSX2_ALIGNED16(int SSEmovMask[ 16 ][ 4 ]) = -{ - { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, - { 0x00000000, 0x00000000, 0x00000000, 0xFFFFFFFF }, - { 0x00000000, 0x00000000, 0xFFFFFFFF, 0x00000000 }, - { 0x00000000, 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF }, - { 0x00000000, 0xFFFFFFFF, 0x00000000, 0x00000000 }, - { 0x00000000, 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF }, - { 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000 }, - { 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF }, - { 0xFFFFFFFF, 0x00000000, 0x00000000, 0x00000000 }, - { 0xFFFFFFFF, 0x00000000, 0x00000000, 0xFFFFFFFF }, - { 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF, 0x00000000 }, - { 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF }, - { 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000, 0x00000000 }, - { 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF }, - { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000 }, - { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF } -}; - -static const PCSX2_ALIGNED16(u32 const_abs_table[16][4]) = -{ - { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //0000 - { 0xffffffff, 0xffffffff, 0xffffffff, 0x7fffffff }, //0001 - { 0xffffffff, 0xffffffff, 0x7fffffff, 0xffffffff }, //0010 - { 0xffffffff, 0xffffffff, 0x7fffffff, 0x7fffffff }, //0011 - { 0xffffffff, 0x7fffffff, 0xffffffff, 0xffffffff }, //0100 - { 0xffffffff, 0x7fffffff, 0xffffffff, 0x7fffffff }, //0101 - { 0xffffffff, 0x7fffffff, 0x7fffffff, 0xffffffff }, //0110 - { 0xffffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //0111 - { 0x7fffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //1000 - { 0x7fffffff, 0xffffffff, 0xffffffff, 0x7fffffff }, //1001 - { 0x7fffffff, 0xffffffff, 0x7fffffff, 0xffffffff }, //1010 - { 0x7fffffff, 0xffffffff, 0x7fffffff, 0x7fffffff }, //1011 - { 0x7fffffff, 0x7fffffff, 0xffffffff, 0xffffffff }, //1100 - { 0x7fffffff, 0x7fffffff, 0xffffffff, 0x7fffffff }, //1101 - { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0xffffffff }, //1110 - { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //1111 -}; - -static const PCSX2_ALIGNED16(float recMult_float_to_int4[4]) = { 16.0, 16.0, 16.0, 16.0 }; -static const PCSX2_ALIGNED16(float recMult_float_to_int12[4]) = { 4096.0, 4096.0, 4096.0, 4096.0 }; -static const PCSX2_ALIGNED16(float recMult_float_to_int15[4]) = { 32768.0, 32768.0, 32768.0, 32768.0 }; - -static const PCSX2_ALIGNED16(float recMult_int_to_float4[4]) = { 0.0625f, 0.0625f, 0.0625f, 0.0625f }; -static const PCSX2_ALIGNED16(float recMult_int_to_float12[4]) = { 0.000244140625, 0.000244140625, 0.000244140625, 0.000244140625 }; -static const PCSX2_ALIGNED16(float recMult_int_to_float15[4]) = { 0.000030517578125, 0.000030517578125, 0.000030517578125, 0.000030517578125 }; - -static const PCSX2_ALIGNED16(u32 VU_Underflow_Mask1[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; -static const PCSX2_ALIGNED16(u32 VU_Underflow_Mask2[4]) = {0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff}; -static const PCSX2_ALIGNED16(u32 VU_Zero_Mask[4]) = {0x00000000, 0x00000000, 0x00000000, 0x00000000}; -static const PCSX2_ALIGNED16(u32 VU_Zero_Helper_Mask[4]) = {0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff}; -static const PCSX2_ALIGNED16(u32 VU_Signed_Zero_Mask[4]) = {0x80000000, 0x80000000, 0x80000000, 0x80000000}; -static const PCSX2_ALIGNED16(u32 VU_Pos_Infinity[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; -static const PCSX2_ALIGNED16(u32 VU_Neg_Infinity[4]) = {0xff800000, 0xff800000, 0xff800000, 0xff800000}; -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// recUpdateFlags() - Computes the flags for the Upper Opcodes -// -// Note: Computes under/overflow flags if CHECK_VU_EXTRA_FLAGS is 1 -//------------------------------------------------------------------ -PCSX2_ALIGNED16(u64 TEMPXMMData[2]); -void recUpdateFlags(VURegs * VU, int reg, int info) -{ - static u8 *pjmp, *pjmp2; - static u32 *pjmp32; - static u32 macaddr, stataddr, prevstataddr; - static int x86macflag, x86statflag, x86temp; - static int t1reg, t1regBoolean; - static const int flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; - - if( !(info & PROCESS_VU_UPDATEFLAGS) ) { - if (CHECK_VU_EXTRA_OVERFLOW) { - if (reg != EEREC_TEMP) vuFloat2(reg, EEREC_TEMP, _X_Y_Z_W); - else vuFloat_useEAX(info, reg, _X_Y_Z_W); - } - return; - } - - //Console::WriteLn ("recUpdateFlags"); - - macaddr = VU_VI_ADDR(REG_MAC_FLAG, 0); - stataddr = VU_VI_ADDR(REG_STATUS_FLAG, 0); // write address - prevstataddr = VU_VI_ADDR(REG_STATUS_FLAG, 2); // previous address - - if( stataddr == 0 ) stataddr = prevstataddr; - if( macaddr == 0 ) { - Console::WriteLn( "VU ALLOCATION WARNING: Using Mac Flag Previous Address!" ); - macaddr = VU_VI_ADDR(REG_MAC_FLAG, 2); - } - - x86macflag = ALLOCTEMPX86(0); - x86statflag = ALLOCTEMPX86(0); - - if (reg == EEREC_TEMP) { - t1reg = _vuGetTempXMMreg(info); - if (t1reg < 0) { - //Console::WriteLn( "VU ALLOCATION ERROR: Temp reg can't be allocated!!!!" ); - t1reg = (reg == 0) ? 1 : 0; // Make t1reg != reg - SSE_MOVAPS_XMM_to_M128( (uptr)TEMPXMMData, t1reg ); // Backup data to temp address - t1regBoolean = 1; - } - else t1regBoolean = 0; - } - else { - t1reg = EEREC_TEMP; - t1regBoolean = 2; - } - - SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x1B); // Flip wzyx to xyzw - MOV32MtoR(x86statflag, prevstataddr); // Load the previous status in to x86statflag - AND16ItoR(x86statflag, 0xff0); // Keep Sticky and D/I flags - - - if (CHECK_VU_EXTRA_FLAGS) { // Checks all flags - - x86temp = ALLOCTEMPX86(0); - - //-------------------------Check for Overflow flags------------------------------ - - //SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg - //SSE_CMPUNORDPS_XMM_to_XMM(t1reg, reg); // If reg == NaN then set Vector to 0xFFFFFFFF - - //SSE_MOVAPS_XMM_to_XMM(t1reg, reg); - //SSE_MINPS_M128_to_XMM(t1reg, (uptr)g_maxvals); - //SSE_MAXPS_M128_to_XMM(t1reg, (uptr)g_minvals); - //SSE_CMPNEPS_XMM_to_XMM(t1reg, reg); // If they're not equal, then overflow has occured - - SSE_MOVAPS_XMM_to_XMM(t1reg, reg); - SSE_ANDPS_M128_to_XMM(t1reg, (uptr)VU_Zero_Helper_Mask); - SSE_CMPEQPS_M128_to_XMM(t1reg, (uptr)VU_Pos_Infinity); // If infinity, then overflow has occured (NaN's don't report as overflow) - - SSE_MOVMSKPS_XMM_to_R32(x86macflag, t1reg); // Move the sign bits of the previous calculation - - AND16ItoR(x86macflag, _X_Y_Z_W ); // Grab "Has Overflowed" bits from the previous calculation (also make sure we're only grabbing from the XYZW being modified) - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x208); // OS, O flags - SHL16ItoR(x86macflag, 12); - if (_XYZW_SS) pjmp32 = JMP32(0); // Skip Underflow Check - x86SetJ8(pjmp); - - //-------------------------Check for Underflow flags------------------------------ - - SSE_MOVAPS_XMM_to_XMM(t1reg, reg); // t1reg <- reg - - SSE_ANDPS_M128_to_XMM(t1reg, (uptr)&VU_Underflow_Mask1[ 0 ]); - SSE_CMPEQPS_M128_to_XMM(t1reg, (uptr)&VU_Zero_Mask[ 0 ]); // If (t1reg == zero exponent) then set Vector to 0xFFFFFFFF - - SSE_ANDPS_XMM_to_XMM(t1reg, reg); - SSE_ANDPS_M128_to_XMM(t1reg, (uptr)&VU_Underflow_Mask2[ 0 ]); - SSE_CMPNEPS_M128_to_XMM(t1reg, (uptr)&VU_Zero_Mask[ 0 ]); // If (t1reg != zero mantisa) then set Vector to 0xFFFFFFFF - - SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation - - AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Has Underflowed" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x104); // US, U flags - SHL16ItoR(EAX, 8); - OR32RtoR(x86macflag, EAX); - x86SetJ8(pjmp); - - //-------------------------Optional Code: Denormals Are Zero------------------------------ - if (CHECK_VU_UNDERFLOW) { // Sets underflow/denormals to zero - SSE_ANDNPS_XMM_to_XMM(t1reg, reg); // t1reg = !t1reg & reg (t1reg = denormals are positive zero) - VU_MERGE_REGS_SAFE(t1reg, reg, (15 - flipMask[_X_Y_Z_W])); // Send t1reg the vectors that shouldn't be modified (since reg was flipped, we need a mask to get the unmodified vectors) - // Now we have Denormals are Positive Zero in t1reg; the next two lines take Signed Zero into account - SSE_ANDPS_M128_to_XMM(reg, (uptr)&VU_Signed_Zero_Mask[ 0 ]); // Only keep the sign bit for each vector - SSE_ORPS_XMM_to_XMM(reg, t1reg); // Denormals are Signed Zero, and unmodified vectors stay the same! - } - - if (_XYZW_SS) x86SetJ32(pjmp32); // If we skipped the Underflow Flag Checking (when we had an Overflow), return here - - vuFloat2(reg, t1reg, flipMask[_X_Y_Z_W]); // Clamp overflowed vectors that were modified (remember reg's vectors have been flipped, so have to use a flipmask) - - //-------------------------Check for Signed flags------------------------------ - - // The following code makes sure the Signed Bit isn't set with Negative Zero - SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg - SSE_CMPEQPS_XMM_to_XMM(t1reg, reg); // Set all F's if each vector is zero - SSE_MOVMSKPS_XMM_to_R32(x86temp, t1reg); // Used for Zero Flag Calculation - SSE_ANDNPS_XMM_to_XMM(t1reg, reg); - - SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the t1reg - - AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Is Signed" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x82); // SS, S flags - SHL16ItoR(EAX, 4); - OR32RtoR(x86macflag, EAX); - if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - x86SetJ8(pjmp); - - //-------------------------Check for Zero flags------------------------------ - - AND16ItoR(x86temp, _X_Y_Z_W ); // Grab "Is Zero" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x41); // ZS, Z flags - OR32RtoR(x86macflag, x86temp); - x86SetJ8(pjmp); - - _freeX86reg(x86temp); - } - else { // Only Checks for Sign and Zero Flags - - vuFloat2(reg, t1reg, flipMask[_X_Y_Z_W]); // Clamp overflowed vectors that were modified (remember reg's vectors have been flipped, so have to use a flipmask) - - //-------------------------Check for Signed flags------------------------------ - - // The following code makes sure the Signed Bit isn't set with Negative Zero - SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg - SSE_CMPEQPS_XMM_to_XMM(t1reg, reg); // Set all F's if each vector is zero - SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Used for Zero Flag Calculation - SSE_ANDNPS_XMM_to_XMM(t1reg, reg); - - SSE_MOVMSKPS_XMM_to_R32(x86macflag, t1reg); // Move the sign bits of the t1reg - - AND16ItoR(x86macflag, _X_Y_Z_W ); // Grab "Is Signed" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x82); // SS, S flags - SHL16ItoR(x86macflag, 4); - if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking - x86SetJ8(pjmp); - - //-------------------------Check for Zero flags------------------------------ - - AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Is Zero" bits from the previous calculation - pjmp = JZ8(0); // Skip if none are - OR16ItoR(x86statflag, 0x41); // ZS, Z flags - OR32RtoR(x86macflag, EAX); - x86SetJ8(pjmp); - } - //-------------------------Finally: Send the Flags to the Mac Flag Address------------------------------ - - if (_XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here - - if (t1regBoolean == 2) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x1B); // Flip back reg to wzyx (have to do this because reg != EEREC_TEMP) - else if (t1regBoolean == 1) SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)TEMPXMMData ); // Restore data from temo address - else _freeXMMreg(t1reg); // Free temp reg - - MOV16RtoM(macaddr, x86macflag); - MOV16RtoM(stataddr, x86statflag); - - _freeX86reg(x86macflag); - _freeX86reg(x86statflag); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// Custom VU ADD/SUB routines by Nneeve -// -// Note: See FPU_ADD_SUB() for more info on what this is doing. -//------------------------------------------------------------------ -static PCSX2_ALIGNED16(u32 VU_addsuband[2][4]); -static PCSX2_ALIGNED16(u32 VU_addsub_reg[2][4]); - -static u32 tempECX; - -void VU_ADD_SUB(u32 regd, u32 regt, int is_sub, int info) -{ - u8 *localptr[4][8]; - - MOV32RtoM((uptr)&tempECX, ECX); - - int temp1 = ECX; //receives regd - int temp2 = ALLOCTEMPX86(0); - - if (temp2 == ECX) - { - temp2 = ALLOCTEMPX86(0); - _freeX86reg(ECX); - } - - SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[0][0], regd); - SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[1][0], regt); - - SSE2_PCMPEQB_XMM_to_XMM(regd, regd); - SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsuband[0][0], regd); - SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsuband[1][0], regd); - SSE_MOVAPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); - - SSE2_PSLLD_I8_to_XMM(regd, 1); - SSE2_PSLLD_I8_to_XMM(regt, 1); - - SSE2_PSRLD_I8_to_XMM(regd, 24); - SSE2_PSRLD_I8_to_XMM(regt, 24); - - SSE2_PSUBD_XMM_to_XMM(regd, regt); - -#define PERFORM(i) \ - \ - SSE_PEXTRW_XMM_to_R32(temp1, regd, i*2); \ - MOVSX32R16toR(temp1, temp1); \ - CMP32ItoR(temp1, 25);\ - localptr[i][0] = JGE8(0);\ - CMP32ItoR(temp1, 0);\ - localptr[i][1] = JG8(0);\ - localptr[i][2] = JE8(0);\ - CMP32ItoR(temp1, -25);\ - localptr[i][3] = JLE8(0);\ - \ - NEG32R(temp1); \ - DEC32R(temp1);\ - MOV32ItoR(temp2, 0xffffffff); \ - SHL32CLtoR(temp2); \ - MOV32RtoM((uptr)&VU_addsuband[0][i], temp2);\ - localptr[i][4] = JMP8(0);\ - \ - x86SetJ8(localptr[i][0]);\ - MOV32ItoM((uptr)&VU_addsuband[1][i], 0x80000000);\ - localptr[i][5] = JMP8(0);\ - \ - x86SetJ8(localptr[i][1]);\ - DEC32R(temp1);\ - MOV32ItoR(temp2, 0xffffffff);\ - SHL32CLtoR(temp2); \ - MOV32RtoM((uptr)&VU_addsuband[1][i], temp2);\ - localptr[i][6] = JMP8(0);\ - \ - x86SetJ8(localptr[i][3]);\ - MOV32ItoM((uptr)&VU_addsuband[0][i], 0x80000000);\ - localptr[i][7] = JMP8(0);\ - \ - x86SetJ8(localptr[i][2]);\ - \ - x86SetJ8(localptr[i][4]);\ - x86SetJ8(localptr[i][5]);\ - x86SetJ8(localptr[i][6]);\ - x86SetJ8(localptr[i][7]); - - PERFORM(0); - PERFORM(1); - PERFORM(2); - PERFORM(3); -#undef PERFORM - - SSE_MOVAPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); - SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); - - SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsuband[0][0]); - SSE_ANDPS_M128_to_XMM(regt, (uptr)&VU_addsuband[1][0]); - - if (is_sub) SSE_SUBPS_XMM_to_XMM(regd, regt); - else SSE_ADDPS_XMM_to_XMM(regd, regt); - - SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); - - _freeX86reg(temp2); - - MOV32MtoR(ECX, (uptr)&tempECX); -} - -void VU_ADD_SUB_SS(u32 regd, u32 regt, int is_sub, int is_mem, int info) -{ - u8 *localptr[8]; - u32 addrt = regt; //for case is_mem - - MOV32RtoM((uptr)&tempECX, ECX); - - int temp1 = ECX; //receives regd - int temp2 = ALLOCTEMPX86(0); - - if (temp2 == ECX) - { - temp2 = ALLOCTEMPX86(0); - _freeX86reg(ECX); - } - - SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[0][0], regd); - if (!is_mem) SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[1][0], regt); - - SSE2_MOVD_XMM_to_R(temp1, regd); - SHR32ItoR(temp1, 23); - - if (is_mem) { - MOV32MtoR(temp2, addrt); - MOV32RtoM((uptr)&VU_addsub_reg[1][0], temp2); - SHR32ItoR(temp2, 23); - } - else { - SSE2_MOVD_XMM_to_R(temp2, regt); - SHR32ItoR(temp2, 23); - } - - AND32ItoR(temp1, 0xff); - AND32ItoR(temp2, 0xff); - - SUB32RtoR(temp1, temp2); //temp1 = exponent difference - - CMP32ItoR(temp1, 25); - localptr[0] = JGE8(0); - CMP32ItoR(temp1, 0); - localptr[1] = JG8(0); - localptr[2] = JE8(0); - CMP32ItoR(temp1, -25); - localptr[3] = JLE8(0); - - NEG32R(temp1); - DEC32R(temp1); - MOV32ItoR(temp2, 0xffffffff); - SHL32CLtoR(temp2); - SSE2_PCMPEQB_XMM_to_XMM(regd, regd); - if (is_mem) { - SSE_PINSRW_R32_to_XMM(regd, temp2, 0); - SHR32ItoR(temp2, 16); - SSE_PINSRW_R32_to_XMM(regd, temp2, 1); - } - else { - SSE2_MOVD_R_to_XMM(regt, temp2); - SSE_MOVSS_XMM_to_XMM(regd, regt); - SSE2_PCMPEQB_XMM_to_XMM(regt, regt); - } - localptr[4] = JMP8(0); - - x86SetJ8(localptr[0]); - MOV32ItoR(temp2, 0x80000000); - if (is_mem) - AND32RtoM((uptr)&VU_addsub_reg[1][0], temp2); - else { - SSE2_PCMPEQB_XMM_to_XMM(regt, regt); - SSE2_MOVD_R_to_XMM(regd, temp2); - SSE_MOVSS_XMM_to_XMM(regt, regd); - } - SSE2_PCMPEQB_XMM_to_XMM(regd, regd); - localptr[5] = JMP8(0); - - x86SetJ8(localptr[1]); - DEC32R(temp1); - MOV32ItoR(temp2, 0xffffffff); - SHL32CLtoR(temp2); - if (is_mem) - AND32RtoM((uptr)&VU_addsub_reg[1][0], temp2); - else { - SSE2_PCMPEQB_XMM_to_XMM(regt, regt); - SSE2_MOVD_R_to_XMM(regd, temp2); - SSE_MOVSS_XMM_to_XMM(regt, regd); - } - SSE2_PCMPEQB_XMM_to_XMM(regd, regd); - localptr[6] = JMP8(0); - - x86SetJ8(localptr[3]); - MOV32ItoR(temp2, 0x80000000); - SSE2_PCMPEQB_XMM_to_XMM(regd, regd); - if (is_mem) { - SSE_PINSRW_R32_to_XMM(regd, temp2, 0); - SHR32ItoR(temp2, 16); - SSE_PINSRW_R32_to_XMM(regd, temp2, 1); - } - else { - SSE2_MOVD_R_to_XMM(regt, temp2); - SSE_MOVSS_XMM_to_XMM(regd, regt); - SSE2_PCMPEQB_XMM_to_XMM(regt, regt); - } - localptr[7] = JMP8(0); - - x86SetJ8(localptr[2]); - x86SetJ8(localptr[4]); - x86SetJ8(localptr[5]); - x86SetJ8(localptr[6]); - x86SetJ8(localptr[7]); - - if (is_mem) - { - SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); //regd contains mask - - if (is_sub) SSE_SUBSS_M32_to_XMM(regd, (uptr)&VU_addsub_reg[1][0]); - else SSE_ADDSS_M32_to_XMM(regd, (uptr)&VU_addsub_reg[1][0]); - } - else - { - SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); //regd contains mask - SSE_ANDPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); //regt contains mask - - if (is_sub) SSE_SUBSS_XMM_to_XMM(regd, regt); - else SSE_ADDSS_XMM_to_XMM(regd, regt); - - SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); - } - - _freeX86reg(temp2); - - MOV32MtoR(ECX, (uptr)&tempECX); -} - -void SSE_ADDPS_XMM_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB(regd, regt, 0, info); - } - else SSE_ADDPS_XMM_to_XMM(regd, regt); -} -void SSE_SUBPS_XMM_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB(regd, regt, 1, info); - } - else SSE_SUBPS_XMM_to_XMM(regd, regt); -} -void SSE_ADDSS_XMM_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB_SS(regd, regt, 0, 0, info); - } - else SSE_ADDSS_XMM_to_XMM(regd, regt); -} -void SSE_SUBSS_XMM_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB_SS(regd, regt, 1, 0, info); - } - else SSE_SUBSS_XMM_to_XMM(regd, regt); -} -void SSE_ADDSS_M32_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB_SS(regd, regt, 0, 1, info); - } - else SSE_ADDSS_M32_to_XMM(regd, regt); -} -void SSE_SUBSS_M32_to_XMM_custom(int info, int regd, int regt) { - if (CHECK_VUADDSUBHACK) { - VU_ADD_SUB_SS(regd, regt, 1, 1, info); - } - else SSE_SUBSS_M32_to_XMM(regd, regt); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// *VU Upper Instructions!* -// -// Note: * = Checked for errors by cottonvibes -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ABS* -//------------------------------------------------------------------ -void recVUMI_ABS(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_ABS()"); - if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; - - if ((_X_Y_Z_W == 0x8) || (_X_Y_Z_W == 0xf)) { - VU_MERGE_REGS(EEREC_T, EEREC_S); - SSE_ANDPS_M128_to_XMM(EEREC_T, (uptr)&const_abs_table[ _X_Y_Z_W ][ 0 ] ); - } - else { // Use a temp reg because VU_MERGE_REGS() modifies source reg! - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_abs_table[ _X_Y_Z_W ][ 0 ] ); - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ADD*, ADD_iq*, ADD_xyzw* -//------------------------------------------------------------------ -PCSX2_ALIGNED16(float s_two[4]) = {0,0,0,2}; -void recVUMI_ADD(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_ADD()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; // Don't do anything and just clear flags - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - - if ( _Fs_ == 0 && _Ft_ == 0 ) { // if adding VF00 with VF00, then the result is always 0,0,0,2 - if ( _X_Y_Z_W != 0xf ) { - SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)s_two); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)s_two); - } - else { - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - if( _X_Y_Z_W == 8 ) { // If only adding x, then we can do a Scalar Add - if (EEREC_D == EEREC_S) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); - else if (EEREC_D == EEREC_T) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { // If xyzw != 1111, then we have to use a temp reg - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { // All xyzw being modified (xyzw == 1111) - if (EEREC_D == EEREC_S) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_T); - else if (EEREC_D == EEREC_T) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_ADD_iq(VURegs *VU, uptr addr, int info) -{ - //Console::WriteLn("recVUMI_ADD_iq()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - } - - if ( _XYZW_SS ) { - if ( EEREC_D == EEREC_TEMP ) { - _vuFlipRegSS(VU, EEREC_S); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_S); - _vuFlipRegSS(VU, EEREC_D); // have to flip over EEREC_D for computing flags! - } - else if ( EEREC_D == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_D); - SSE_ADDSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_D); - } - else { - if ( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDSS_M32_to_XMM_custom(info, EEREC_D, addr); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_ADDPS_XMM_to_XMM_custom(info, EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - } - } - else { - if ( (_X_Y_Z_W != 0xf) || (EEREC_D == EEREC_S) || (EEREC_D == EEREC_TEMP) ) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - } - - if (_X_Y_Z_W != 0xf) { - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if ( EEREC_D == EEREC_TEMP ) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); - else if ( EEREC_D == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - else { - SSE_MOVSS_M32_to_XMM(EEREC_D, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); - SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_ADD_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn("recVUMI_ADD_xyzw()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - } - - if ( _Ft_ == 0 && xyzw < 3 ) { // just move since adding zero - if ( _X_Y_Z_W == 0x8 ) { VU_MERGE_REGS(EEREC_D, EEREC_S); } - else if ( _X_Y_Z_W != 0xf ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - else if ( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP) ) { - if ( xyzw == 0 ) { - if ( EEREC_D == EEREC_T ) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - else if( _Fs_ == 0 && !_W ) { // just move - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if ( _X_Y_Z_W != 0xf ) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if( EEREC_D == EEREC_TEMP ) { _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); } - else if( EEREC_D == EEREC_S ) { _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); } - else { _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_ADDi(VURegs *VU, int info) { recVUMI_ADD_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_ADDq(VURegs *VU, int info) { recVUMI_ADD_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_ADDx(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 0, info); } -void recVUMI_ADDy(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 1, info); } -void recVUMI_ADDz(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 2, info); } -void recVUMI_ADDw(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ADDA*, ADDA_iq*, ADDA_xyzw* -//------------------------------------------------------------------ -void recVUMI_ADDA(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_ADDA()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - - if( _X_Y_Z_W == 8 ) { - if (EEREC_ACC == EEREC_S) SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); // Can this case happen? (cottonvibes) - else if (EEREC_ACC == EEREC_T) SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_S); // Can this case happen? - else { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - else { - if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_T); // Can this case happen? - else if( EEREC_ACC == EEREC_T ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); // Can this case happen? - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_ADDA_iq(VURegs *VU, uptr addr, int info) -{ - //Console::WriteLn("recVUMI_ADDA_iq()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - } - - if( _XYZW_SS ) { - assert( EEREC_ACC != EEREC_TEMP ); - if( EEREC_ACC == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_ACC); - SSE_ADDSS_M32_to_XMM(EEREC_ACC, addr); - _vuFlipRegSS(VU, EEREC_ACC); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_ADDSS_M32_to_XMM(EEREC_ACC, addr); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - } - } - else { - if( _X_Y_Z_W != 0xf || EEREC_ACC == EEREC_S ) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - } - - if (_X_Y_Z_W != 0xf) { - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - else { - if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - else { - SSE_MOVSS_M32_to_XMM(EEREC_ACC, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_ACC, EEREC_ACC, 0x00); - SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_ADDA_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn("recVUMI_ADDA_xyzw()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - } - - if( _X_Y_Z_W == 8 ) { - assert( EEREC_ACC != EEREC_T ); - if( xyzw == 0 ) { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - if( _Fs_ == 0 ) { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - else { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - } - } - else { - if( _X_Y_Z_W != 0xf || EEREC_ACC == EEREC_S ) - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - - if (_X_Y_Z_W != 0xf) { - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - else { - if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - else { - _unpackVF_xyzw(EEREC_ACC, EEREC_T, xyzw); - SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_ADDAi(VURegs *VU, int info) { recVUMI_ADDA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_ADDAq(VURegs *VU, int info) { recVUMI_ADDA_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_ADDAx(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 0, info); } -void recVUMI_ADDAy(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 1, info); } -void recVUMI_ADDAz(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 2, info); } -void recVUMI_ADDAw(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SUB*, SUB_iq*, SUB_xyzw* -//------------------------------------------------------------------ -void recVUMI_SUB(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_SUB()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - - if( EEREC_S == EEREC_T ) { - if (_X_Y_Z_W != 0xf) SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&SSEmovMask[15-_X_Y_Z_W][0]); - else SSE_XORPS_XMM_to_XMM(EEREC_D, EEREC_D); - } - else if( _X_Y_Z_W == 8 ) { - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - if (EEREC_D == EEREC_S) { - if (_Ft_) SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - else if (EEREC_D == EEREC_T) { - if (_Ft_) { - SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - } - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - if (_Ft_) SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else { - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if( ( _Ft_ > 0 ) || _W ) SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if (EEREC_D == EEREC_S) SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_T); - else if (EEREC_D == EEREC_T) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_SUB_iq(VURegs *VU, uptr addr, int info) -{ - //Console::WriteLn("recVUMI_SUB_iq()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - } - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - - if( _XYZW_SS ) { - if( EEREC_D == EEREC_TEMP ) { - _vuFlipRegSS(VU, EEREC_S); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_S); - _vuFlipRegSS(VU, EEREC_D); - } - else if( EEREC_D == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_D); - SSE_SUBSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_D); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBSS_M32_to_XMM(EEREC_D, addr); - } - else { - _vuMoveSS(VU, EEREC_TEMP, EEREC_S); - _vuFlipRegSS(VU, EEREC_D); - SSE_SUBSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_D); - } - } - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - VU_MERGE_REGS(EEREC_D, t1reg); - _freeXMMreg(t1reg); - } - else { - // negate - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - } - else { - if( EEREC_D == EEREC_TEMP ) { - SSE_XORPS_M128_to_XMM(EEREC_D, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_SUB_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn("recVUMI_SUB_xyzw()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - } - - if ( _X_Y_Z_W == 8 ) { - if ( (xyzw == 0) && (_Ft_ == _Fs_) ) { - SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&SSEmovMask[7][0]); - } - else if ( EEREC_D == EEREC_TEMP ) { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - if ( (_Ft_ > 0) || (xyzw == 3) ) { - _vuFlipRegSS_xyzw(EEREC_T, xyzw); - SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); - _vuFlipRegSS_xyzw(EEREC_T, xyzw); - } - } - else { - if ( (_Ft_ > 0) || (xyzw == 3) ) { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - else { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - VU_MERGE_REGS(EEREC_D, t1reg); - _freeXMMreg(t1reg); - } - else { - // negate - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - } - else { - if( EEREC_D == EEREC_TEMP ) { - SSE_XORPS_M128_to_XMM(EEREC_D, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_SUBi(VURegs *VU, int info) { recVUMI_SUB_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_SUBq(VURegs *VU, int info) { recVUMI_SUB_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_SUBx(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 0, info); } -void recVUMI_SUBy(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 1, info); } -void recVUMI_SUBz(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 2, info); } -void recVUMI_SUBw(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// SUBA*, SUBA_iq, SUBA_xyzw -//------------------------------------------------------------------ -void recVUMI_SUBA(VURegs *VU, int info) -{ - //Console::WriteLn("recVUMI_SUBA()"); - if ( _X_Y_Z_W == 0 ) goto flagUpdate; - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - - if( EEREC_S == EEREC_T ) { - if (_X_Y_Z_W != 0xf) SSE_ANDPS_M128_to_XMM(EEREC_ACC, (uptr)&SSEmovMask[15-_X_Y_Z_W][0]); - else SSE_XORPS_XMM_to_XMM(EEREC_ACC, EEREC_ACC); - } - else if( _X_Y_Z_W == 8 ) { - if (EEREC_ACC == EEREC_S) SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); - else if (EEREC_ACC == EEREC_T) { - SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - else { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - else { - if( EEREC_ACC == EEREC_S ) SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_T); - else if( EEREC_ACC == EEREC_T ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - } -flagUpdate: - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_SUBA_iq(VURegs *VU, uptr addr, int info) -{ - //Console::WriteLn ("recVUMI_SUBA_iq"); - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - } - - if( _XYZW_SS ) { - if( EEREC_ACC == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_ACC); - SSE_SUBSS_M32_to_XMM(EEREC_ACC, addr); - _vuFlipRegSS(VU, EEREC_ACC); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBSS_M32_to_XMM(EEREC_ACC, addr); - } - else { - _vuMoveSS(VU, EEREC_TEMP, EEREC_S); - _vuFlipRegSS(VU, EEREC_ACC); - SSE_SUBSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_ACC); - } - } - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - VU_MERGE_REGS(EEREC_ACC, t1reg); - _freeXMMreg(t1reg); - } - else { - // negate - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - } - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_SUBA_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn ("recVUMI_SUBA_xyzw"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - } - - if( _X_Y_Z_W == 8 ) { - if( xyzw == 0 ) { - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - } - else { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - VU_MERGE_REGS(EEREC_ACC, t1reg); - _freeXMMreg(t1reg); - } - else { - // negate - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); - } - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); - SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); - } - } - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_SUBAi(VURegs *VU, int info) { recVUMI_SUBA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_SUBAq(VURegs *VU, int info) { recVUMI_SUBA_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_SUBAx(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 0, info); } -void recVUMI_SUBAy(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 1, info); } -void recVUMI_SUBAz(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 2, info); } -void recVUMI_SUBAw(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MUL -//------------------------------------------------------------------ -void recVUMI_MUL_toD(VURegs *VU, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MUL_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - } - - if (_X_Y_Z_W == 1 && (_Ft_ == 0 || _Fs_==0) ) { // W - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, _Ft_ ? EEREC_T : EEREC_S); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else if( _Fd_ == _Fs_ && _Fs_ == _Ft_ && _XYZW_SS ) { - _vuFlipRegSS(VU, EEREC_D); - SSE_MULSS_XMM_to_XMM(EEREC_D, EEREC_D); - _vuFlipRegSS(VU, EEREC_D); - } - else if( _X_Y_Z_W == 8 ) { - if (regd == EEREC_S) SSE_MULSS_XMM_to_XMM(regd, EEREC_T); - else if (regd == EEREC_T) SSE_MULSS_XMM_to_XMM(regd, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_XMM_to_XMM(regd, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - else if (regd == EEREC_T) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - else { - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); - SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - } - } -} - -void recVUMI_MUL_iq_toD(VURegs *VU, uptr addr, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MUL_iq_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - } - - if( _XYZW_SS ) { - if( regd == EEREC_TEMP ) { - _vuFlipRegSS(VU, EEREC_S); - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_M32_to_XMM(regd, addr); - _vuFlipRegSS(VU, EEREC_S); - _vuFlipRegSS(VU, regd); - } - else if( regd == EEREC_S ) { - _vuFlipRegSS(VU, regd); - SSE_MULSS_M32_to_XMM(regd, addr); - _vuFlipRegSS(VU, regd); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_M32_to_XMM(regd, addr); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - } - } - else { - if( _X_Y_Z_W != 0xf || regd == EEREC_TEMP || regd == EEREC_S ) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - } - - if (_X_Y_Z_W != 0xf) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_TEMP ) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - else if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); - else { - SSE_MOVSS_M32_to_XMM(regd, addr); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x00); - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - } - } - } -} - -void recVUMI_MUL_xyzw_toD(VURegs *VU, int xyzw, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MUL_xyzw_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - } - if (_Fs_) { // This is needed for alot of games; so always clamp this operand - if (CHECK_VU_SIGN_OVERFLOW) vFloats4_useEAX[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set - else vFloats2[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set - } - if( _Ft_ == 0 ) { - if( xyzw < 3 ) { - if (_X_Y_Z_W != 0xf) { - SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else SSE_XORPS_XMM_to_XMM(regd, regd); - } - else { - assert(xyzw==3); - if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); - } - } - else if( _X_Y_Z_W == 8 ) { - if( regd == EEREC_TEMP ) { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MULSS_XMM_to_XMM(regd, EEREC_S); - } - else { - if( xyzw == 0 ) { - if( regd == EEREC_T ) { - SSE_MULSS_XMM_to_XMM(regd, EEREC_S); - } - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_XMM_to_XMM(regd, EEREC_T); - } - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_XMM_to_XMM(regd, EEREC_TEMP); - } - } - } - else { - if( _X_Y_Z_W != 0xf || regd == EEREC_TEMP || regd == EEREC_S ) - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - - if (_X_Y_Z_W != 0xf) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_TEMP ) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - else if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); - else { - _unpackVF_xyzw(regd, EEREC_T, xyzw); - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - } - } - } -} - -void recVUMI_MUL(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MUL"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MUL_toD(VU, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MUL_iq(VURegs *VU, int addr, int info) -{ - //Console::WriteLn ("recVUMI_MUL_iq"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MUL_iq_toD(VU, addr, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); - // spacefisherman needs overflow checking on MULi.z -} - -void recVUMI_MUL_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn ("recVUMI_MUL_xyzw"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MUL_xyzw_toD(VU, xyzw, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MULi(VURegs *VU, int info) { recVUMI_MUL_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MULq(VURegs *VU, int info) { recVUMI_MUL_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_MULx(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 0, info); } -void recVUMI_MULy(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 1, info); } -void recVUMI_MULz(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 2, info); } -void recVUMI_MULw(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MULA -//------------------------------------------------------------------ -void recVUMI_MULA( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MULA"); - recVUMI_MUL_toD(VU, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MULA_iq(VURegs *VU, int addr, int info) -{ - //Console::WriteLn ("recVUMI_MULA_iq"); - recVUMI_MUL_iq_toD(VU, addr, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MULA_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn ("recVUMI_MULA_xyzw"); - recVUMI_MUL_xyzw_toD(VU, xyzw, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MULAi(VURegs *VU, int info) { recVUMI_MULA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MULAq(VURegs *VU, int info) { recVUMI_MULA_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_MULAx(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 0, info); } -void recVUMI_MULAy(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 1, info); } -void recVUMI_MULAz(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 2, info); } -void recVUMI_MULAw(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MADD -//------------------------------------------------------------------ -void recVUMI_MADD_toD(VURegs *VU, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MADD_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - } - - - if( _X_Y_Z_W == 8 ) { - if( regd == EEREC_ACC ) { - SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if (regd == EEREC_T) { - SSE_MULSS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - else if (regd == EEREC_S) { - SSE_MULSS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_ACC ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if (regd == EEREC_T) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else if (regd == EEREC_S) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); - SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - } -} - -void recVUMI_MADD_iq_toD(VURegs *VU, uptr addr, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MADD_iq_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - vuFloat3(addr); - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - } - - if( _X_Y_Z_W == 8 ) { - if( _Fs_ == 0 ) { - // do nothing if regd == ACC (ACCx <= ACCx + 0.0 * *addr) - if( regd != EEREC_ACC ) { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); - } - return; - } - - if( regd == EEREC_ACC ) { - assert( EEREC_TEMP < iREGCNT_XMM ); - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if( regd == EEREC_S ) { - SSE_MULSS_M32_to_XMM(regd, addr); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); - SSE_MULSS_M32_to_XMM(regd, addr); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - } - else { - if( _Fs_ == 0 ) { - if( regd == EEREC_ACC ) { // ACCxyz is unchanged, ACCw <= ACCw + *addr - if( _W ) { // if _W is zero, do nothing - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); // { *addr, 0, 0, 0 } - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x27); // { 0, 0, 0, *addr } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); // { ACCx, ACCy, ACCz, ACCw + *addr } - } - } - else { // DESTxyz <= ACCxyz, DESTw <= ACCw + *addr - if( _W ) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); // { *addr, 0, 0, 0 } - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x27); // { 0, 0, 0, *addr } - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); // { ACCx, ACCy, ACCz, ACCw + *addr } - } - else SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - - return; - } - - if( _X_Y_Z_W != 0xf || regd == EEREC_ACC || regd == EEREC_TEMP || regd == EEREC_S ) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - } - - if (_X_Y_Z_W != 0xf) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_ACC ) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if( regd == EEREC_S ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else if( regd == EEREC_TEMP ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - SSE_MOVSS_M32_to_XMM(regd, addr); - SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x00); - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - } - } -} - -void recVUMI_MADD_xyzw_toD(VURegs *VU, int xyzw, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MADD_xyzw_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - } - if (_Fs_) { // This is needed for alot of games; so always clamp this operand - if (CHECK_VU_SIGN_OVERFLOW) vFloats4_useEAX[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set - else vFloats2[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set - } - if( _Ft_ == 0 ) { - - if( xyzw == 3 ) { - // just add - if( _X_Y_Z_W == 8 ) { - if( regd == EEREC_S ) SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); - SSE_ADDSS_XMM_to_XMM(regd, EEREC_S); - } - } - else { - if( _X_Y_Z_W != 0xf ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_S ) SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - else { - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_ADDPS_XMM_to_XMM(regd, EEREC_S); - } - } - } - } - else { - // just move acc to regd - if( _X_Y_Z_W != 0xf ) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); - } - - return; - } - - if( _X_Y_Z_W == 8 ) { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - - if( regd == EEREC_ACC ) { - SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if( regd == EEREC_S ) { - SSE_MULSS_XMM_to_XMM(regd, EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - else if( regd == EEREC_TEMP ) { - SSE_MULSS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); - SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } - SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); - } - } - else { - if( _X_Y_Z_W != 0xf || regd == EEREC_ACC || regd == EEREC_TEMP || regd == EEREC_S ) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - } - - if (_X_Y_Z_W != 0xf) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - - VU_MERGE_REGS(regd, EEREC_TEMP); - } - else { - if( regd == EEREC_ACC ) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if( regd == EEREC_S ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else if( regd == EEREC_TEMP ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - else { - _unpackVF_xyzw(regd, EEREC_T, xyzw); - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); - } - } - } -} - -void recVUMI_MADD(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MADD"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MADD_toD(VU, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MADD_iq(VURegs *VU, int addr, int info) -{ - //Console::WriteLn ("recVUMI_MADD_iq"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MADD_iq_toD(VU, addr, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MADD_xyzw(VURegs *VU, int xyzw, int info) -{ - //Console::WriteLn ("recVUMI_MADD_xyzw"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MADD_xyzw_toD(VU, xyzw, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); - // super bust-a-move arrows needs overflow clamping -} - -void recVUMI_MADDi(VURegs *VU, int info) { recVUMI_MADD_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MADDq(VURegs *VU, int info) { recVUMI_MADD_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_MADDx(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 0, info); } -void recVUMI_MADDy(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 1, info); } -void recVUMI_MADDz(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 2, info); } -void recVUMI_MADDw(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MADDA -//------------------------------------------------------------------ -void recVUMI_MADDA( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MADDA"); - recVUMI_MADD_toD(VU, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAi( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAi"); - recVUMI_MADD_iq_toD( VU, VU_VI_ADDR(REG_I, 1), EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAq( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAq "); - recVUMI_MADD_iq_toD( VU, VU_REGQ_ADDR, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAx( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAx"); - recVUMI_MADD_xyzw_toD(VU, 0, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAy( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAy"); - recVUMI_MADD_xyzw_toD(VU, 1, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAz( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAz"); - recVUMI_MADD_xyzw_toD(VU, 2, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MADDAw( VURegs *VU , int info) -{ - //Console::WriteLn ("recVUMI_MADDAw"); - recVUMI_MADD_xyzw_toD(VU, 3, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MSUB -//------------------------------------------------------------------ -void recVUMI_MSUB_toD(VURegs *VU, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MSUB_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - } - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_ACC); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - VU_MERGE_REGS(regd, t1reg); - _freeXMMreg(t1reg); - } - else { - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - } - else { - if( regd == EEREC_S ) { - assert( regd != EEREC_ACC ); - SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); - } - else if( regd == EEREC_T ) { - assert( regd != EEREC_ACC ); - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); - } - else if( regd == EEREC_TEMP ) { - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); - SSE_MULPS_XMM_to_XMM(regd, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); - } - } -} - -void recVUMI_MSUB_temp_toD(VURegs *VU, int regd, int info) -{ - //Console::WriteLn ("recVUMI_MSUB_temp_toD"); - - if (_X_Y_Z_W != 0xf) { - int t1reg = _vuGetTempXMMreg(info); - - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - - if( t1reg >= 0 ) { - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_ACC); - SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); - - if ( regd != EEREC_TEMP ) { VU_MERGE_REGS(regd, t1reg); } - else SSE_MOVAPS_XMM_to_XMM(regd, t1reg); - - _freeXMMreg(t1reg); - } - else { - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - VU_MERGE_REGS(regd, EEREC_TEMP); - } - } - else { - if( regd == EEREC_ACC ) { - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); - } - else if( regd == EEREC_S ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); - } - else if( regd == EEREC_TEMP ) { - SSE_MULPS_XMM_to_XMM(regd, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); - } - else { - SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } - SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); - } - } -} - -void recVUMI_MSUB_iq_toD(VURegs *VU, int regd, int addr, int info) -{ - //Console::WriteLn ("recVUMI_MSUB_iq_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - vuFloat3(addr); - } - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - recVUMI_MSUB_temp_toD(VU, regd, info); -} - -void recVUMI_MSUB_xyzw_toD(VURegs *VU, int regd, int xyzw, int info) -{ - //Console::WriteLn ("recVUMI_MSUB_xyzw_toD"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 1 << (3 - xyzw)); - vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); - } - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - recVUMI_MSUB_temp_toD(VU, regd, info); -} - -void recVUMI_MSUB(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MSUB"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_toD(VU, EEREC_D, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MSUB_iq(VURegs *VU, int addr, int info) -{ - //Console::WriteLn ("recVUMI_MSUB_iq"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_iq_toD(VU, EEREC_D, addr, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MSUBi(VURegs *VU, int info) { recVUMI_MSUB_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MSUBq(VURegs *VU, int info) { recVUMI_MSUB_iq(VU, VU_REGQ_ADDR, info); } -void recVUMI_MSUBx(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MSUBx"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 0, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MSUBy(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MSUBy"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 1, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MSUBz(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MSUBz"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 2, info); - recUpdateFlags(VU, EEREC_D, info); -} - -void recVUMI_MSUBw(VURegs *VU, int info) -{ - //Console::WriteLn ("recVUMI_MSUBw"); - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 3, info); - recUpdateFlags(VU, EEREC_D, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MSUBA -//------------------------------------------------------------------ -void recVUMI_MSUBA( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBA"); - recVUMI_MSUB_toD(VU, EEREC_ACC, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAi( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAi "); - recVUMI_MSUB_iq_toD( VU, EEREC_ACC, VU_VI_ADDR(REG_I, 1), info ); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAq( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAq"); - recVUMI_MSUB_iq_toD( VU, EEREC_ACC, VU_REGQ_ADDR, info ); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAx( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAx"); - recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 0, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAy( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAy"); - recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 1, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAz( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAz "); - recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 2, info); - recUpdateFlags(VU, EEREC_ACC, info); -} - -void recVUMI_MSUBAw( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_MSUBAw"); - recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 3, info); - recUpdateFlags(VU, EEREC_ACC, info); -} -//------------------------------------------------------------------ - - -static const u32 PCSX2_ALIGNED16(special_mask[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; -static const u32 PCSX2_ALIGNED16(special_mask2[4]) = {0, 0x40000000, 0, 0x40000000}; - -u32 PCSX2_ALIGNED16(temp_loc[4]); -u32 PCSX2_ALIGNED16(temp_loc2[4]); - -//MAX/MINI are non-arithmetic operations that implicitly support numbers with the EXP field being 0 ("denormals"). -// -//As such, they are sometimes used for integer move and (positive!) integer max/min, knowing that integers that -//represent denormals will not be flushed to 0. -// -//As such, this implementation performs a non-arithmetic operation that supports "denormals" and "infs/nans". -//There might be an easier way to do it but here, MAX/MIN is performed with PMAXPD/PMINPD. -//Fake double-precision numbers are constructed by copying the sign of the original numbers, clearing the upper 32 bits, -//setting the 62nd bit to 1 (to ensure double-precision number is "normalized") and having the lower 32bits -//being the same as the original number. - -void MINMAXlogical(VURegs *VU, int info, int min, int mode, uptr addr = 0, int xyzw = 0) -//mode1 = iq, mode2 = xyzw, mode0 = normal -{ - int t1regbool = 0; - int t1reg = _vuGetTempXMMreg(info); - if (t1reg < 0) - { - t1regbool = 1; - for (t1reg = 0; ( (t1reg == EEREC_D) || (t1reg == EEREC_S) || (mode != 1 && t1reg == EEREC_T) - || (t1reg == EEREC_TEMP) ); t1reg++); // Find unused reg (For first temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)temp_loc, t1reg); // Backup t1reg XMM reg - } - int t2regbool = -1; - int t2reg = EEREC_TEMP; - if (EEREC_TEMP == EEREC_D || EEREC_TEMP == EEREC_S || (mode != 1 && EEREC_TEMP == EEREC_T)) - { - t2regbool = 0; - t2reg = _vuGetTempXMMreg(info); - if (t2reg < 0) - { - t2regbool = 1; - for (t2reg = 0; ( (t2reg == EEREC_D) || (t2reg == EEREC_S) || (mode != 1 && t2reg == EEREC_T) || - (t2reg == t1reg) || (t2reg == EEREC_TEMP) ); t2reg++); // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)temp_loc2, t2reg); // Backup t2reg XMM reg - } - } - - if (_X || _Y) - { - SSE2_PSHUFD_XMM_to_XMM(t1reg, EEREC_S, 0x50); - SSE2_PAND_M128_to_XMM(t1reg, (uptr)special_mask); - SSE2_POR_M128_to_XMM(t1reg, (uptr)special_mask2); - if (mode == 0) - SSE2_PSHUFD_XMM_to_XMM(t2reg, EEREC_T, 0x50); - else if (mode == 1) - { - SSE2_MOVD_M32_to_XMM(t2reg, addr); - SSE2_PSHUFD_XMM_to_XMM(t2reg, t2reg, 0x00); - } - else if (mode == 2) - _unpackVF_xyzw(t2reg, EEREC_T, xyzw); - SSE2_PAND_M128_to_XMM(t2reg, (uptr)special_mask); - SSE2_POR_M128_to_XMM(t2reg, (uptr)special_mask2); - if (min) - SSE2_MINPD_XMM_to_XMM(t1reg, t2reg); - else - SSE2_MAXPD_XMM_to_XMM(t1reg, t2reg); - SSE2_PSHUFD_XMM_to_XMM(t1reg, t1reg, 0x88); - VU_MERGE_REGS_CUSTOM(EEREC_D, t1reg, 0xc & _X_Y_Z_W); - } - - if (_Z || _W) - { - SSE2_PSHUFD_XMM_to_XMM(t1reg, EEREC_S, 0xfa); - SSE2_PAND_M128_to_XMM(t1reg, (uptr)special_mask); - SSE2_POR_M128_to_XMM(t1reg, (uptr)special_mask2); - if (mode == 0) - SSE2_PSHUFD_XMM_to_XMM(t2reg, EEREC_T, 0xfa); - else if (mode == 1) - { - SSE2_MOVD_M32_to_XMM(t2reg, addr); - SSE2_PSHUFD_XMM_to_XMM(t2reg, t2reg, 0x00); - } - else if (mode == 2) - _unpackVF_xyzw(t2reg, EEREC_T, xyzw); - SSE2_PAND_M128_to_XMM(t2reg, (uptr)special_mask); - SSE2_POR_M128_to_XMM(t2reg, (uptr)special_mask2); - if (min) - SSE2_MINPD_XMM_to_XMM(t1reg, t2reg); - else - SSE2_MAXPD_XMM_to_XMM(t1reg, t2reg); - SSE2_PSHUFD_XMM_to_XMM(t1reg, t1reg, 0x88); - VU_MERGE_REGS_CUSTOM(EEREC_D, t1reg, 0x3 & _X_Y_Z_W); - } - - if (t1regbool == 0) - _freeXMMreg(t1reg); - else if (t1regbool == 1) - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)temp_loc); // Restore t1reg XMM reg - if (t2regbool == 0) - _freeXMMreg(t2reg); - else if (t2regbool == 1) - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)temp_loc2); // Restore t2reg XMM reg -} - -//------------------------------------------------------------------ -// MAX -//------------------------------------------------------------------ - -void recVUMI_MAX(VURegs *VU, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MAX"); - - if (MINMAXFIX) - MINMAXlogical(VU, info, 0, 0); - else - { - - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - - if( _X_Y_Z_W == 8 ) { - if (EEREC_D == EEREC_S) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); - else if (EEREC_D == EEREC_T) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if( EEREC_D == EEREC_S ) SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_T); - else if( EEREC_D == EEREC_T ) SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - } -} - -void recVUMI_MAX_iq(VURegs *VU, uptr addr, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MAX_iq"); - - if (MINMAXFIX) - MINMAXlogical(VU, info, 0, 1, addr); - else - { - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - vuFloat3(addr); - - if( _XYZW_SS ) { - if( EEREC_D == EEREC_TEMP ) { - _vuFlipRegSS(VU, EEREC_S); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_S); - - // have to flip over EEREC_D if computing flags! - //if( (info & PROCESS_VU_UPDATEFLAGS) ) - _vuFlipRegSS(VU, EEREC_D); - } - else if( EEREC_D == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_D); - SSE_MAXSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_D); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXSS_M32_to_XMM(EEREC_D, addr); - } - else { - _vuMoveSS(VU, EEREC_TEMP, EEREC_S); - _vuFlipRegSS(VU, EEREC_D); - SSE_MAXSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_D); - } - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if(EEREC_D == EEREC_S) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_D, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); - SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - } -} - -void recVUMI_MAX_xyzw(VURegs *VU, int xyzw, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MAX_xyzw"); - - if (_Fs_ == 0 && _Ft_ == 0) - { - if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { - if( xyzw < 3 ) { - SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)s_fones); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - else if (_X_Y_Z_W != 0xf) { - if( xyzw < 3 ) { - if( _X_Y_Z_W & 1 ) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&VU->VF[0].UL[0]); // w included, so insert the whole reg - else SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // w not included, can zero out - } - else SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)s_fones); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - //If VF0.w isnt chosen as the constant, then its going to be MAX( 0, VF0 ), so the result is VF0 - if( xyzw < 3 ) { SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)&VU->VF[0].UL[0]); } - else SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)s_fones); - } - return; - } - - if (MINMAXFIX) - MINMAXlogical(VU, info, 0, 2, 0, xyzw); - else - { - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - - if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { - if( xyzw == 0 ) { - if( EEREC_D == EEREC_S ) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); - else if( EEREC_D == EEREC_T ) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - else if (_X_Y_Z_W != 0xf) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if (EEREC_D == EEREC_S) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); - SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - } -} - -void recVUMI_MAXi(VURegs *VU, int info) { recVUMI_MAX_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MAXx(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 0, info); } -void recVUMI_MAXy(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 1, info); } -void recVUMI_MAXz(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 2, info); } -void recVUMI_MAXw(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// MINI -//------------------------------------------------------------------ -void recVUMI_MINI(VURegs *VU, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MINI"); - - if (MINMAXFIX) - MINMAXlogical(VU, info, 1, 0); - else - { - - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); - - if( _X_Y_Z_W == 8 ) { - if (EEREC_D == EEREC_S) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); - else if (EEREC_D == EEREC_T) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if( EEREC_D == EEREC_S ) { - //ClampUnordered(EEREC_T, EEREC_TEMP, 0); // need for GT4 vu0rec - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_T); - } - else if( EEREC_D == EEREC_T ) { - //ClampUnordered(EEREC_S, EEREC_TEMP, 0); // need for GT4 vu0rec - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - } -} - -void recVUMI_MINI_iq(VURegs *VU, uptr addr, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MINI_iq"); - - if (MINMAXFIX) - MINMAXlogical(VU, info, 1, 1, addr); - else - { - - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - vuFloat3(addr); - - if( _XYZW_SS ) { - if( EEREC_D == EEREC_TEMP ) { - _vuFlipRegSS(VU, EEREC_S); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_S); - - // have to flip over EEREC_D if computing flags! - //if( (info & PROCESS_VU_UPDATEFLAGS) ) - _vuFlipRegSS(VU, EEREC_D); - } - else if( EEREC_D == EEREC_S ) { - _vuFlipRegSS(VU, EEREC_D); - SSE_MINSS_M32_to_XMM(EEREC_D, addr); - _vuFlipRegSS(VU, EEREC_D); - } - else { - if( _X ) { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINSS_M32_to_XMM(EEREC_D, addr); - } - else { - _vuMoveSS(VU, EEREC_TEMP, EEREC_S); - _vuFlipRegSS(VU, EEREC_D); - SSE_MINSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - _vuFlipRegSS(VU, EEREC_D); - } - } - } - else if (_X_Y_Z_W != 0xf) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if(EEREC_D == EEREC_S) { - SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - SSE_MOVSS_M32_to_XMM(EEREC_D, addr); - SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - } -} - -void recVUMI_MINI_xyzw(VURegs *VU, int xyzw, int info) -{ - if ( _Fd_ == 0 ) return; - //Console::WriteLn ("recVUMI_MINI_xyzw"); - - if (_Fs_ == 0 && _Ft_ == 0) - { - if( _X_Y_Z_W == 0xf ) - { - //If VF0.w is the constant, the result will match VF0, else its all 0's - if(xyzw == 3) SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)&VU->VF[0].UL[0]); - else SSE_XORPS_XMM_to_XMM(EEREC_D, EEREC_D); - } - else - { - //If VF0.w is the constant, the result will match VF0, else its all 0's - if(xyzw == 3) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&VU->VF[0].UL[0]); - else SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - return; - } - if (MINMAXFIX) - MINMAXlogical(VU, info, 1, 2, 0, xyzw); - else - { - if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping - if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); - - if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { - if( xyzw == 0 ) { - if( EEREC_D == EEREC_S ) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); - else if( EEREC_D == EEREC_T ) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_S); - else { - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); - } - } - else { - _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); - SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - } - else if (_X_Y_Z_W != 0xf) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - VU_MERGE_REGS(EEREC_D, EEREC_TEMP); - } - else { - if (EEREC_D == EEREC_S) { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); - } - else { - _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); - SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); - } - } - } -} - -void recVUMI_MINIi(VURegs *VU, int info) { recVUMI_MINI_iq(VU, VU_VI_ADDR(REG_I, 1), info); } -void recVUMI_MINIx(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 0, info); } -void recVUMI_MINIy(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 1, info); } -void recVUMI_MINIz(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 2, info); } -void recVUMI_MINIw(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 3, info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// OPMULA -//------------------------------------------------------------------ -void recVUMI_OPMULA( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_OPMULA"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, 0xE); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 0xE); - } - - SSE_MOVAPS_XMM_to_XMM( EEREC_TEMP, EEREC_S ); - SSE_SHUFPS_XMM_to_XMM( EEREC_T, EEREC_T, 0xD2 ); // EEREC_T = WYXZ - SSE_SHUFPS_XMM_to_XMM( EEREC_TEMP, EEREC_TEMP, 0xC9 ); // EEREC_TEMP = WXZY - SSE_MULPS_XMM_to_XMM( EEREC_TEMP, EEREC_T ); - - VU_MERGE_REGS_CUSTOM(EEREC_ACC, EEREC_TEMP, 14); - - // revert EEREC_T - if( EEREC_T != EEREC_ACC ) - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xC9); - - recUpdateFlags(VU, EEREC_ACC, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// OPMSUB -//------------------------------------------------------------------ -void recVUMI_OPMSUB( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_OPMSUB"); - if (CHECK_VU_EXTRA_OVERFLOW) { - if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, 0xE); - if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 0xE); - } - - if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xD2); // EEREC_T = WYXZ - SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0xC9); // EEREC_TEMP = WXZY - SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); - - // negate and add - SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); - SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); - VU_MERGE_REGS_CUSTOM(EEREC_D, EEREC_TEMP, 14); - - // revert EEREC_T - if( EEREC_T != EEREC_D ) SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xC9); - - recUpdateFlags(VU, EEREC_D, info); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// NOP -//------------------------------------------------------------------ -void recVUMI_NOP( VURegs *VU, int info ) -{ - //Console::WriteLn ("recVUMI_NOP"); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// recVUMI_FTOI_Saturate() - Saturates result from FTOI Instructions -//------------------------------------------------------------------ -static const PCSX2_ALIGNED16(int rec_const_0x8000000[4]) = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 }; - -void recVUMI_FTOI_Saturate(int rec_s, int rec_t, int rec_tmp1, int rec_tmp2) -{ - //Console::WriteLn ("recVUMI_FTOI_Saturate"); - //Duplicate the xor'd sign bit to the whole value - //FFFF FFFF for positive, 0 for negative - SSE_MOVAPS_XMM_to_XMM(rec_tmp1, rec_s); - SSE2_PXOR_M128_to_XMM(rec_tmp1, (uptr)&const_clip[4]); - SSE2_PSRAD_I8_to_XMM(rec_tmp1, 31); - - //Create mask: 0 where !=8000 0000 - SSE_MOVAPS_XMM_to_XMM(rec_tmp2, rec_t); - SSE2_PCMPEQD_M128_to_XMM(rec_tmp2, (uptr)&const_clip[4]); - - //AND the mask w/ the edit values - SSE_ANDPS_XMM_to_XMM(rec_tmp1, rec_tmp2); - - //if v==8000 0000 && positive -> 8000 0000 + FFFF FFFF -> 7FFF FFFF - //if v==8000 0000 && negative -> 8000 0000 + 0 -> 8000 0000 - //if v!=8000 0000 -> v+0 (masked from the and) - - //Add the values as needed - SSE2_PADDD_XMM_to_XMM(rec_t, rec_tmp1); -} -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// FTOI 0/4/12/15 -//------------------------------------------------------------------ -static PCSX2_ALIGNED16(float FTIO_Temp1[4]) = { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; -static PCSX2_ALIGNED16(float FTIO_Temp2[4]) = { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; -void recVUMI_FTOI0(VURegs *VU, int info) -{ - int t1reg, t2reg; // Temp XMM regs - - if ( _Ft_ == 0 ) return; - - //Console::WriteLn ("recVUMI_FTOI0"); - - if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg (For first temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg - } - - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - if (EEREC_T != EEREC_S) { - SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); - vuFloat_useEAX( info, EEREC_T, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_T, EEREC_T); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - } - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg (For first temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg - } - - SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - } - } -} - -void recVUMI_FTOIX(VURegs *VU, int addr, int info) -{ - int t1reg, t2reg; // Temp XMM regs - - if ( _Ft_ == 0 ) return; - - //Console::WriteLn ("recVUMI_FTOIX"); - if (_X_Y_Z_W != 0xf) { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); - vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg (For first temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg - } - - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - } - else { - if (EEREC_T != EEREC_S) { - SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); - SSE_MULPS_M128_to_XMM(EEREC_T, addr); - vuFloat_useEAX( info, EEREC_T, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_T, EEREC_T); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - } - } - else { - SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); - vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) - SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); - - t1reg = _vuGetTempXMMreg(info); - - if( t1reg >= 0 ) { // If theres a temp XMM reg available - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg - _freeXMMreg(t1reg); // Free temp reg - } - else { // No temp reg available - for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) - ; // Find unused reg (For first temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg - - for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) - ; // Find unused reg (For second temp reg) - SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg - - recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result - - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg - SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg - } - - SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_TEMP); - } - } -} - -void recVUMI_FTOI4( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int4[0], info); } -void recVUMI_FTOI12( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int12[0], info); } -void recVUMI_FTOI15( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int15[0], info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// ITOF 0/4/12/15 -//------------------------------------------------------------------ -void recVUMI_ITOF0( VURegs *VU, int info ) -{ - if ( _Ft_ == 0 ) return; - - //Console::WriteLn ("recVUMI_ITOF0"); - if (_X_Y_Z_W != 0xf) { - SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - vuFloat_useEAX( info, EEREC_TEMP, 15); // Clamp infinities - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - xmmregs[EEREC_T].mode |= MODE_WRITE; - } - else { - SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_T, EEREC_S); - vuFloat2(EEREC_T, EEREC_TEMP, 15); // Clamp infinities - } -} - -void recVUMI_ITOFX(VURegs *VU, int addr, int info) -{ - if ( _Ft_ == 0 ) return; - - //Console::WriteLn ("recVUMI_ITOFX"); - if (_X_Y_Z_W != 0xf) { - SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_TEMP, EEREC_S); - SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); - vuFloat_useEAX( info, EEREC_TEMP, 15); // Clamp infinities - VU_MERGE_REGS(EEREC_T, EEREC_TEMP); - xmmregs[EEREC_T].mode |= MODE_WRITE; - } - else { - SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_T, EEREC_S); - SSE_MULPS_M128_to_XMM(EEREC_T, addr); - vuFloat2(EEREC_T, EEREC_TEMP, 15); // Clamp infinities - } -} - -void recVUMI_ITOF4( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float4[0], info); } -void recVUMI_ITOF12( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float12[0], info); } -void recVUMI_ITOF15( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float15[0], info); } -//------------------------------------------------------------------ - - -//------------------------------------------------------------------ -// CLIP -//------------------------------------------------------------------ -void recVUMI_CLIP(VURegs *VU, int info) -{ - int t1reg = EEREC_D; - int t2reg = EEREC_ACC; - int x86temp1, x86temp2; - - u32 clipaddr = VU_VI_ADDR(REG_CLIP_FLAG, 0); - u32 prevclipaddr = VU_VI_ADDR(REG_CLIP_FLAG, 2); - - if( clipaddr == 0 ) { // battle star has a clip right before fcset - Console::WriteLn("skipping vu clip"); - return; - } - - //Flush the clip flag before processing, incase of double clip commands (GoW) - - if( prevclipaddr != (uptr)&VU->VI[REG_CLIP_FLAG] ) { - MOV32MtoR(EAX, prevclipaddr); - MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); - } - - assert( clipaddr != 0 ); - assert( t1reg != t2reg && t1reg != EEREC_TEMP && t2reg != EEREC_TEMP ); - - x86temp1 = ALLOCTEMPX86(MODE_8BITREG); - x86temp2 = ALLOCTEMPX86(MODE_8BITREG); - - //if ( (x86temp1 == 0) || (x86temp2 == 0) ) Console::Error("VU CLIP Allocation Error: EAX being allocated!"); - - _freeXMMreg(t1reg); // These should have been freed at allocation in eeVURecompileCode() - _freeXMMreg(t2reg); // but if they've been used since then, then free them. (just doing this incase :p (cottonvibes)) - - if( _Ft_ == 0 ) { - SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&s_fones[0]); // all 1s - SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)&s_fones[4]); - } - else { - _unpackVF_xyzw(EEREC_TEMP, EEREC_T, 3); - SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[0]); - SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_TEMP); - SSE_ORPS_M128_to_XMM(t1reg, (uptr)&const_clip[4]); - } - - MOV32MtoR(EAX, prevclipaddr); - - SSE_CMPNLEPS_XMM_to_XMM(t1reg, EEREC_S); //-w, -z, -y, -x - SSE_CMPLTPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); //+w, +z, +y, +x - - SHL32ItoR(EAX, 6); - - SSE_MOVAPS_XMM_to_XMM(t2reg, EEREC_TEMP); //t2 = +w, +z, +y, +x - SSE_UNPCKLPS_XMM_to_XMM(EEREC_TEMP, t1reg); //EEREC_TEMP = -y,+y,-x,+x - SSE_UNPCKHPS_XMM_to_XMM(t2reg, t1reg); //t2reg = -w,+w,-z,+z - SSE_MOVMSKPS_XMM_to_R32(x86temp2, EEREC_TEMP); // -y,+y,-x,+x - SSE_MOVMSKPS_XMM_to_R32(x86temp1, t2reg); // -w,+w,-z,+z - - AND8ItoR(x86temp1, 0x3); - SHL8ItoR(x86temp1, 4); - OR8RtoR(EAX, x86temp1); - AND8ItoR(x86temp2, 0xf); - OR8RtoR(EAX, x86temp2); - AND32ItoR(EAX, 0xffffff); - - MOV32RtoM(clipaddr, EAX); - - if (( !(info & (PROCESS_VU_SUPER|PROCESS_VU_COP2)) ) ) //Instantly update the flag if its called from elsewhere (unlikely, but ok) - MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); - - _freeX86reg(x86temp1); - _freeX86reg(x86temp2); -} +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "PrecompiledHeader.h" + +#include "Common.h" +#include "GS.h" +#include "R5900OpcodeTables.h" +#include "iR5900.h" +#include "iMMI.h" +#include "iFPU.h" +#include "iCOP0.h" +#include "VUmicro.h" +#include "VUflags.h" +#include "sVU_Micro.h" +#include "sVU_Debug.h" +#include "sVU_zerorec.h" +//------------------------------------------------------------------ +#define MINMAXFIX 1 +//------------------------------------------------------------------ +// Helper Macros +//------------------------------------------------------------------ +#define _Ft_ (( VU->code >> 16) & 0x1F) // The rt part of the instruction register +#define _Fs_ (( VU->code >> 11) & 0x1F) // The rd part of the instruction register +#define _Fd_ (( VU->code >> 6) & 0x1F) // The sa part of the instruction register + +#define _X (( VU->code>>24) & 0x1) +#define _Y (( VU->code>>23) & 0x1) +#define _Z (( VU->code>>22) & 0x1) +#define _W (( VU->code>>21) & 0x1) + +#define _XYZW_SS (_X+_Y+_Z+_W==1) + +#define _Fsf_ (( VU->code >> 21) & 0x03) +#define _Ftf_ (( VU->code >> 23) & 0x03) + +#define _Imm11_ (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) +#define _UImm11_ (s32)(VU->code & 0x7ff) + +#define VU_VFx_ADDR(x) (uptr)&VU->VF[x].UL[0] +#define VU_VFy_ADDR(x) (uptr)&VU->VF[x].UL[1] +#define VU_VFz_ADDR(x) (uptr)&VU->VF[x].UL[2] +#define VU_VFw_ADDR(x) (uptr)&VU->VF[x].UL[3] + +#define VU_REGR_ADDR (uptr)&VU->VI[REG_R] +#define VU_REGQ_ADDR (uptr)&VU->VI[REG_Q] +#define VU_REGMAC_ADDR (uptr)&VU->VI[REG_MAC_FLAG] + +#define VU_VI_ADDR(x, read) GetVIAddr(VU, x, read, info) + +#define VU_ACCx_ADDR (uptr)&VU->ACC.UL[0] +#define VU_ACCy_ADDR (uptr)&VU->ACC.UL[1] +#define VU_ACCz_ADDR (uptr)&VU->ACC.UL[2] +#define VU_ACCw_ADDR (uptr)&VU->ACC.UL[3] + +#define _X_Y_Z_W ((( VU->code >> 21 ) & 0xF ) ) +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// Global Variables +//------------------------------------------------------------------ +static const PCSX2_ALIGNED16(int SSEmovMask[ 16 ][ 4 ]) = +{ + { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, + { 0x00000000, 0x00000000, 0x00000000, 0xFFFFFFFF }, + { 0x00000000, 0x00000000, 0xFFFFFFFF, 0x00000000 }, + { 0x00000000, 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF }, + { 0x00000000, 0xFFFFFFFF, 0x00000000, 0x00000000 }, + { 0x00000000, 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF }, + { 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000 }, + { 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF }, + { 0xFFFFFFFF, 0x00000000, 0x00000000, 0x00000000 }, + { 0xFFFFFFFF, 0x00000000, 0x00000000, 0xFFFFFFFF }, + { 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF, 0x00000000 }, + { 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF, 0xFFFFFFFF }, + { 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000, 0x00000000 }, + { 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000, 0xFFFFFFFF }, + { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000 }, + { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF } +}; + +static const PCSX2_ALIGNED16(u32 const_abs_table[16][4]) = +{ + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //0000 + { 0xffffffff, 0xffffffff, 0xffffffff, 0x7fffffff }, //0001 + { 0xffffffff, 0xffffffff, 0x7fffffff, 0xffffffff }, //0010 + { 0xffffffff, 0xffffffff, 0x7fffffff, 0x7fffffff }, //0011 + { 0xffffffff, 0x7fffffff, 0xffffffff, 0xffffffff }, //0100 + { 0xffffffff, 0x7fffffff, 0xffffffff, 0x7fffffff }, //0101 + { 0xffffffff, 0x7fffffff, 0x7fffffff, 0xffffffff }, //0110 + { 0xffffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //0111 + { 0x7fffffff, 0xffffffff, 0xffffffff, 0xffffffff }, //1000 + { 0x7fffffff, 0xffffffff, 0xffffffff, 0x7fffffff }, //1001 + { 0x7fffffff, 0xffffffff, 0x7fffffff, 0xffffffff }, //1010 + { 0x7fffffff, 0xffffffff, 0x7fffffff, 0x7fffffff }, //1011 + { 0x7fffffff, 0x7fffffff, 0xffffffff, 0xffffffff }, //1100 + { 0x7fffffff, 0x7fffffff, 0xffffffff, 0x7fffffff }, //1101 + { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0xffffffff }, //1110 + { 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff }, //1111 +}; + +static const PCSX2_ALIGNED16(float recMult_float_to_int4[4]) = { 16.0, 16.0, 16.0, 16.0 }; +static const PCSX2_ALIGNED16(float recMult_float_to_int12[4]) = { 4096.0, 4096.0, 4096.0, 4096.0 }; +static const PCSX2_ALIGNED16(float recMult_float_to_int15[4]) = { 32768.0, 32768.0, 32768.0, 32768.0 }; + +static const PCSX2_ALIGNED16(float recMult_int_to_float4[4]) = { 0.0625f, 0.0625f, 0.0625f, 0.0625f }; +static const PCSX2_ALIGNED16(float recMult_int_to_float12[4]) = { 0.000244140625, 0.000244140625, 0.000244140625, 0.000244140625 }; +static const PCSX2_ALIGNED16(float recMult_int_to_float15[4]) = { 0.000030517578125, 0.000030517578125, 0.000030517578125, 0.000030517578125 }; + +static const PCSX2_ALIGNED16(u32 VU_Underflow_Mask1[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; +static const PCSX2_ALIGNED16(u32 VU_Underflow_Mask2[4]) = {0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff}; +static const PCSX2_ALIGNED16(u32 VU_Zero_Mask[4]) = {0x00000000, 0x00000000, 0x00000000, 0x00000000}; +static const PCSX2_ALIGNED16(u32 VU_Zero_Helper_Mask[4]) = {0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff}; +static const PCSX2_ALIGNED16(u32 VU_Signed_Zero_Mask[4]) = {0x80000000, 0x80000000, 0x80000000, 0x80000000}; +static const PCSX2_ALIGNED16(u32 VU_Pos_Infinity[4]) = {0x7f800000, 0x7f800000, 0x7f800000, 0x7f800000}; +static const PCSX2_ALIGNED16(u32 VU_Neg_Infinity[4]) = {0xff800000, 0xff800000, 0xff800000, 0xff800000}; +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// recUpdateFlags() - Computes the flags for the Upper Opcodes +// +// Note: Computes under/overflow flags if CHECK_VU_EXTRA_FLAGS is 1 +//------------------------------------------------------------------ +PCSX2_ALIGNED16(u64 TEMPXMMData[2]); +void recUpdateFlags(VURegs * VU, int reg, int info) +{ + static u8 *pjmp, *pjmp2; + static u32 *pjmp32; + static u32 macaddr, stataddr, prevstataddr; + static int x86macflag, x86statflag, x86temp; + static int t1reg, t1regBoolean; + static const int flipMask[16] = {0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15}; + + if( !(info & PROCESS_VU_UPDATEFLAGS) ) { + if (CHECK_VU_EXTRA_OVERFLOW) { + if (reg != EEREC_TEMP) vuFloat2(reg, EEREC_TEMP, _X_Y_Z_W); + else vuFloat_useEAX(info, reg, _X_Y_Z_W); + } + return; + } + + //Console::WriteLn ("recUpdateFlags"); + + macaddr = VU_VI_ADDR(REG_MAC_FLAG, 0); + stataddr = VU_VI_ADDR(REG_STATUS_FLAG, 0); // write address + prevstataddr = VU_VI_ADDR(REG_STATUS_FLAG, 2); // previous address + + if( stataddr == 0 ) stataddr = prevstataddr; + if( macaddr == 0 ) { + Console::WriteLn( "VU ALLOCATION WARNING: Using Mac Flag Previous Address!" ); + macaddr = VU_VI_ADDR(REG_MAC_FLAG, 2); + } + + x86macflag = ALLOCTEMPX86(0); + x86statflag = ALLOCTEMPX86(0); + + if (reg == EEREC_TEMP) { + t1reg = _vuGetTempXMMreg(info); + if (t1reg < 0) { + //Console::WriteLn( "VU ALLOCATION ERROR: Temp reg can't be allocated!!!!" ); + t1reg = (reg == 0) ? 1 : 0; // Make t1reg != reg + SSE_MOVAPS_XMM_to_M128( (uptr)TEMPXMMData, t1reg ); // Backup data to temp address + t1regBoolean = 1; + } + else t1regBoolean = 0; + } + else { + t1reg = EEREC_TEMP; + t1regBoolean = 2; + } + + SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x1B); // Flip wzyx to xyzw + MOV32MtoR(x86statflag, prevstataddr); // Load the previous status in to x86statflag + AND16ItoR(x86statflag, 0xff0); // Keep Sticky and D/I flags + + + if (CHECK_VU_EXTRA_FLAGS) { // Checks all flags + + x86temp = ALLOCTEMPX86(0); + + //-------------------------Check for Overflow flags------------------------------ + + //SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg + //SSE_CMPUNORDPS_XMM_to_XMM(t1reg, reg); // If reg == NaN then set Vector to 0xFFFFFFFF + + //SSE_MOVAPS_XMM_to_XMM(t1reg, reg); + //SSE_MINPS_M128_to_XMM(t1reg, (uptr)g_maxvals); + //SSE_MAXPS_M128_to_XMM(t1reg, (uptr)g_minvals); + //SSE_CMPNEPS_XMM_to_XMM(t1reg, reg); // If they're not equal, then overflow has occured + + SSE_MOVAPS_XMM_to_XMM(t1reg, reg); + SSE_ANDPS_M128_to_XMM(t1reg, (uptr)VU_Zero_Helper_Mask); + SSE_CMPEQPS_M128_to_XMM(t1reg, (uptr)VU_Pos_Infinity); // If infinity, then overflow has occured (NaN's don't report as overflow) + + SSE_MOVMSKPS_XMM_to_R32(x86macflag, t1reg); // Move the sign bits of the previous calculation + + AND16ItoR(x86macflag, _X_Y_Z_W ); // Grab "Has Overflowed" bits from the previous calculation (also make sure we're only grabbing from the XYZW being modified) + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x208); // OS, O flags + SHL16ItoR(x86macflag, 12); + if (_XYZW_SS) pjmp32 = JMP32(0); // Skip Underflow Check + x86SetJ8(pjmp); + + //-------------------------Check for Underflow flags------------------------------ + + SSE_MOVAPS_XMM_to_XMM(t1reg, reg); // t1reg <- reg + + SSE_ANDPS_M128_to_XMM(t1reg, (uptr)&VU_Underflow_Mask1[ 0 ]); + SSE_CMPEQPS_M128_to_XMM(t1reg, (uptr)&VU_Zero_Mask[ 0 ]); // If (t1reg == zero exponent) then set Vector to 0xFFFFFFFF + + SSE_ANDPS_XMM_to_XMM(t1reg, reg); + SSE_ANDPS_M128_to_XMM(t1reg, (uptr)&VU_Underflow_Mask2[ 0 ]); + SSE_CMPNEPS_M128_to_XMM(t1reg, (uptr)&VU_Zero_Mask[ 0 ]); // If (t1reg != zero mantisa) then set Vector to 0xFFFFFFFF + + SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the previous calculation + + AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Has Underflowed" bits from the previous calculation + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x104); // US, U flags + SHL16ItoR(EAX, 8); + OR32RtoR(x86macflag, EAX); + x86SetJ8(pjmp); + + //-------------------------Optional Code: Denormals Are Zero------------------------------ + if (CHECK_VU_UNDERFLOW) { // Sets underflow/denormals to zero + SSE_ANDNPS_XMM_to_XMM(t1reg, reg); // t1reg = !t1reg & reg (t1reg = denormals are positive zero) + VU_MERGE_REGS_SAFE(t1reg, reg, (15 - flipMask[_X_Y_Z_W])); // Send t1reg the vectors that shouldn't be modified (since reg was flipped, we need a mask to get the unmodified vectors) + // Now we have Denormals are Positive Zero in t1reg; the next two lines take Signed Zero into account + SSE_ANDPS_M128_to_XMM(reg, (uptr)&VU_Signed_Zero_Mask[ 0 ]); // Only keep the sign bit for each vector + SSE_ORPS_XMM_to_XMM(reg, t1reg); // Denormals are Signed Zero, and unmodified vectors stay the same! + } + + if (_XYZW_SS) x86SetJ32(pjmp32); // If we skipped the Underflow Flag Checking (when we had an Overflow), return here + + vuFloat2(reg, t1reg, flipMask[_X_Y_Z_W]); // Clamp overflowed vectors that were modified (remember reg's vectors have been flipped, so have to use a flipmask) + + //-------------------------Check for Signed flags------------------------------ + + // The following code makes sure the Signed Bit isn't set with Negative Zero + SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg + SSE_CMPEQPS_XMM_to_XMM(t1reg, reg); // Set all F's if each vector is zero + SSE_MOVMSKPS_XMM_to_R32(x86temp, t1reg); // Used for Zero Flag Calculation + SSE_ANDNPS_XMM_to_XMM(t1reg, reg); + + SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Move the sign bits of the t1reg + + AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Is Signed" bits from the previous calculation + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x82); // SS, S flags + SHL16ItoR(EAX, 4); + OR32RtoR(x86macflag, EAX); + if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + x86SetJ8(pjmp); + + //-------------------------Check for Zero flags------------------------------ + + AND16ItoR(x86temp, _X_Y_Z_W ); // Grab "Is Zero" bits from the previous calculation + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x41); // ZS, Z flags + OR32RtoR(x86macflag, x86temp); + x86SetJ8(pjmp); + + _freeX86reg(x86temp); + } + else { // Only Checks for Sign and Zero Flags + + vuFloat2(reg, t1reg, flipMask[_X_Y_Z_W]); // Clamp overflowed vectors that were modified (remember reg's vectors have been flipped, so have to use a flipmask) + + //-------------------------Check for Signed flags------------------------------ + + // The following code makes sure the Signed Bit isn't set with Negative Zero + SSE_XORPS_XMM_to_XMM(t1reg, t1reg); // Clear t1reg + SSE_CMPEQPS_XMM_to_XMM(t1reg, reg); // Set all F's if each vector is zero + SSE_MOVMSKPS_XMM_to_R32(EAX, t1reg); // Used for Zero Flag Calculation + SSE_ANDNPS_XMM_to_XMM(t1reg, reg); + + SSE_MOVMSKPS_XMM_to_R32(x86macflag, t1reg); // Move the sign bits of the t1reg + + AND16ItoR(x86macflag, _X_Y_Z_W ); // Grab "Is Signed" bits from the previous calculation + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x82); // SS, S flags + SHL16ItoR(x86macflag, 4); + if (_XYZW_SS) pjmp2 = JMP8(0); // If negative and not Zero, we can skip the Zero Flag checking + x86SetJ8(pjmp); + + //-------------------------Check for Zero flags------------------------------ + + AND16ItoR(EAX, _X_Y_Z_W ); // Grab "Is Zero" bits from the previous calculation + pjmp = JZ8(0); // Skip if none are + OR16ItoR(x86statflag, 0x41); // ZS, Z flags + OR32RtoR(x86macflag, EAX); + x86SetJ8(pjmp); + } + //-------------------------Finally: Send the Flags to the Mac Flag Address------------------------------ + + if (_XYZW_SS) x86SetJ8(pjmp2); // If we skipped the Zero Flag Checking, return here + + if (t1regBoolean == 2) SSE_SHUFPS_XMM_to_XMM(reg, reg, 0x1B); // Flip back reg to wzyx (have to do this because reg != EEREC_TEMP) + else if (t1regBoolean == 1) SSE_MOVAPS_M128_to_XMM( t1reg, (uptr)TEMPXMMData ); // Restore data from temo address + else _freeXMMreg(t1reg); // Free temp reg + + MOV16RtoM(macaddr, x86macflag); + MOV16RtoM(stataddr, x86statflag); + + _freeX86reg(x86macflag); + _freeX86reg(x86statflag); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// Custom VU ADD/SUB routines by Nneeve +// +// Note: See FPU_ADD_SUB() for more info on what this is doing. +//------------------------------------------------------------------ +static PCSX2_ALIGNED16(u32 VU_addsuband[2][4]); +static PCSX2_ALIGNED16(u32 VU_addsub_reg[2][4]); + +static u32 tempECX; + +void VU_ADD_SUB(u32 regd, u32 regt, int is_sub, int info) +{ + u8 *localptr[4][8]; + + MOV32RtoM((uptr)&tempECX, ECX); + + int temp1 = ECX; //receives regd + int temp2 = ALLOCTEMPX86(0); + + if (temp2 == ECX) + { + temp2 = ALLOCTEMPX86(0); + _freeX86reg(ECX); + } + + SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[0][0], regd); + SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[1][0], regt); + + SSE2_PCMPEQB_XMM_to_XMM(regd, regd); + SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsuband[0][0], regd); + SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsuband[1][0], regd); + SSE_MOVAPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); + + SSE2_PSLLD_I8_to_XMM(regd, 1); + SSE2_PSLLD_I8_to_XMM(regt, 1); + + SSE2_PSRLD_I8_to_XMM(regd, 24); + SSE2_PSRLD_I8_to_XMM(regt, 24); + + SSE2_PSUBD_XMM_to_XMM(regd, regt); + +#define PERFORM(i) \ + \ + SSE_PEXTRW_XMM_to_R32(temp1, regd, i*2); \ + MOVSX32R16toR(temp1, temp1); \ + CMP32ItoR(temp1, 25);\ + localptr[i][0] = JGE8(0);\ + CMP32ItoR(temp1, 0);\ + localptr[i][1] = JG8(0);\ + localptr[i][2] = JE8(0);\ + CMP32ItoR(temp1, -25);\ + localptr[i][3] = JLE8(0);\ + \ + NEG32R(temp1); \ + DEC32R(temp1);\ + MOV32ItoR(temp2, 0xffffffff); \ + SHL32CLtoR(temp2); \ + MOV32RtoM((uptr)&VU_addsuband[0][i], temp2);\ + localptr[i][4] = JMP8(0);\ + \ + x86SetJ8(localptr[i][0]);\ + MOV32ItoM((uptr)&VU_addsuband[1][i], 0x80000000);\ + localptr[i][5] = JMP8(0);\ + \ + x86SetJ8(localptr[i][1]);\ + DEC32R(temp1);\ + MOV32ItoR(temp2, 0xffffffff);\ + SHL32CLtoR(temp2); \ + MOV32RtoM((uptr)&VU_addsuband[1][i], temp2);\ + localptr[i][6] = JMP8(0);\ + \ + x86SetJ8(localptr[i][3]);\ + MOV32ItoM((uptr)&VU_addsuband[0][i], 0x80000000);\ + localptr[i][7] = JMP8(0);\ + \ + x86SetJ8(localptr[i][2]);\ + \ + x86SetJ8(localptr[i][4]);\ + x86SetJ8(localptr[i][5]);\ + x86SetJ8(localptr[i][6]);\ + x86SetJ8(localptr[i][7]); + + PERFORM(0); + PERFORM(1); + PERFORM(2); + PERFORM(3); +#undef PERFORM + + SSE_MOVAPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); + SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); + + SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsuband[0][0]); + SSE_ANDPS_M128_to_XMM(regt, (uptr)&VU_addsuband[1][0]); + + if (is_sub) SSE_SUBPS_XMM_to_XMM(regd, regt); + else SSE_ADDPS_XMM_to_XMM(regd, regt); + + SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); + + _freeX86reg(temp2); + + MOV32MtoR(ECX, (uptr)&tempECX); +} + +void VU_ADD_SUB_SS(u32 regd, u32 regt, int is_sub, int is_mem, int info) +{ + u8 *localptr[8]; + u32 addrt = regt; //for case is_mem + + MOV32RtoM((uptr)&tempECX, ECX); + + int temp1 = ECX; //receives regd + int temp2 = ALLOCTEMPX86(0); + + if (temp2 == ECX) + { + temp2 = ALLOCTEMPX86(0); + _freeX86reg(ECX); + } + + SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[0][0], regd); + if (!is_mem) SSE_MOVAPS_XMM_to_M128((uptr)&VU_addsub_reg[1][0], regt); + + SSE2_MOVD_XMM_to_R(temp1, regd); + SHR32ItoR(temp1, 23); + + if (is_mem) { + MOV32MtoR(temp2, addrt); + MOV32RtoM((uptr)&VU_addsub_reg[1][0], temp2); + SHR32ItoR(temp2, 23); + } + else { + SSE2_MOVD_XMM_to_R(temp2, regt); + SHR32ItoR(temp2, 23); + } + + AND32ItoR(temp1, 0xff); + AND32ItoR(temp2, 0xff); + + SUB32RtoR(temp1, temp2); //temp1 = exponent difference + + CMP32ItoR(temp1, 25); + localptr[0] = JGE8(0); + CMP32ItoR(temp1, 0); + localptr[1] = JG8(0); + localptr[2] = JE8(0); + CMP32ItoR(temp1, -25); + localptr[3] = JLE8(0); + + NEG32R(temp1); + DEC32R(temp1); + MOV32ItoR(temp2, 0xffffffff); + SHL32CLtoR(temp2); + SSE2_PCMPEQB_XMM_to_XMM(regd, regd); + if (is_mem) { + SSE_PINSRW_R32_to_XMM(regd, temp2, 0); + SHR32ItoR(temp2, 16); + SSE_PINSRW_R32_to_XMM(regd, temp2, 1); + } + else { + SSE2_MOVD_R_to_XMM(regt, temp2); + SSE_MOVSS_XMM_to_XMM(regd, regt); + SSE2_PCMPEQB_XMM_to_XMM(regt, regt); + } + localptr[4] = JMP8(0); + + x86SetJ8(localptr[0]); + MOV32ItoR(temp2, 0x80000000); + if (is_mem) + AND32RtoM((uptr)&VU_addsub_reg[1][0], temp2); + else { + SSE2_PCMPEQB_XMM_to_XMM(regt, regt); + SSE2_MOVD_R_to_XMM(regd, temp2); + SSE_MOVSS_XMM_to_XMM(regt, regd); + } + SSE2_PCMPEQB_XMM_to_XMM(regd, regd); + localptr[5] = JMP8(0); + + x86SetJ8(localptr[1]); + DEC32R(temp1); + MOV32ItoR(temp2, 0xffffffff); + SHL32CLtoR(temp2); + if (is_mem) + AND32RtoM((uptr)&VU_addsub_reg[1][0], temp2); + else { + SSE2_PCMPEQB_XMM_to_XMM(regt, regt); + SSE2_MOVD_R_to_XMM(regd, temp2); + SSE_MOVSS_XMM_to_XMM(regt, regd); + } + SSE2_PCMPEQB_XMM_to_XMM(regd, regd); + localptr[6] = JMP8(0); + + x86SetJ8(localptr[3]); + MOV32ItoR(temp2, 0x80000000); + SSE2_PCMPEQB_XMM_to_XMM(regd, regd); + if (is_mem) { + SSE_PINSRW_R32_to_XMM(regd, temp2, 0); + SHR32ItoR(temp2, 16); + SSE_PINSRW_R32_to_XMM(regd, temp2, 1); + } + else { + SSE2_MOVD_R_to_XMM(regt, temp2); + SSE_MOVSS_XMM_to_XMM(regd, regt); + SSE2_PCMPEQB_XMM_to_XMM(regt, regt); + } + localptr[7] = JMP8(0); + + x86SetJ8(localptr[2]); + x86SetJ8(localptr[4]); + x86SetJ8(localptr[5]); + x86SetJ8(localptr[6]); + x86SetJ8(localptr[7]); + + if (is_mem) + { + SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); //regd contains mask + + if (is_sub) SSE_SUBSS_M32_to_XMM(regd, (uptr)&VU_addsub_reg[1][0]); + else SSE_ADDSS_M32_to_XMM(regd, (uptr)&VU_addsub_reg[1][0]); + } + else + { + SSE_ANDPS_M128_to_XMM(regd, (uptr)&VU_addsub_reg[0][0]); //regd contains mask + SSE_ANDPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); //regt contains mask + + if (is_sub) SSE_SUBSS_XMM_to_XMM(regd, regt); + else SSE_ADDSS_XMM_to_XMM(regd, regt); + + SSE_MOVAPS_M128_to_XMM(regt, (uptr)&VU_addsub_reg[1][0]); + } + + _freeX86reg(temp2); + + MOV32MtoR(ECX, (uptr)&tempECX); +} + +void SSE_ADDPS_XMM_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB(regd, regt, 0, info); + } + else SSE_ADDPS_XMM_to_XMM(regd, regt); +} +void SSE_SUBPS_XMM_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB(regd, regt, 1, info); + } + else SSE_SUBPS_XMM_to_XMM(regd, regt); +} +void SSE_ADDSS_XMM_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB_SS(regd, regt, 0, 0, info); + } + else SSE_ADDSS_XMM_to_XMM(regd, regt); +} +void SSE_SUBSS_XMM_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB_SS(regd, regt, 1, 0, info); + } + else SSE_SUBSS_XMM_to_XMM(regd, regt); +} +void SSE_ADDSS_M32_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB_SS(regd, regt, 0, 1, info); + } + else SSE_ADDSS_M32_to_XMM(regd, regt); +} +void SSE_SUBSS_M32_to_XMM_custom(int info, int regd, int regt) { + if (CHECK_VUADDSUBHACK) { + VU_ADD_SUB_SS(regd, regt, 1, 1, info); + } + else SSE_SUBSS_M32_to_XMM(regd, regt); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// *VU Upper Instructions!* +// +// Note: * = Checked for errors by cottonvibes +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ABS* +//------------------------------------------------------------------ +void recVUMI_ABS(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_ABS()"); + if ( (_Ft_ == 0) || (_X_Y_Z_W == 0) ) return; + + if ((_X_Y_Z_W == 0x8) || (_X_Y_Z_W == 0xf)) { + VU_MERGE_REGS(EEREC_T, EEREC_S); + SSE_ANDPS_M128_to_XMM(EEREC_T, (uptr)&const_abs_table[ _X_Y_Z_W ][ 0 ] ); + } + else { // Use a temp reg because VU_MERGE_REGS() modifies source reg! + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_abs_table[ _X_Y_Z_W ][ 0 ] ); + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ADD*, ADD_iq*, ADD_xyzw* +//------------------------------------------------------------------ +PCSX2_ALIGNED16(float s_two[4]) = {0,0,0,2}; +void recVUMI_ADD(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_ADD()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; // Don't do anything and just clear flags + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + + if ( _Fs_ == 0 && _Ft_ == 0 ) { // if adding VF00 with VF00, then the result is always 0,0,0,2 + if ( _X_Y_Z_W != 0xf ) { + SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)s_two); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)s_two); + } + else { + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + if( _X_Y_Z_W == 8 ) { // If only adding x, then we can do a Scalar Add + if (EEREC_D == EEREC_S) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); + else if (EEREC_D == EEREC_T) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { // If xyzw != 1111, then we have to use a temp reg + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { // All xyzw being modified (xyzw == 1111) + if (EEREC_D == EEREC_S) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_T); + else if (EEREC_D == EEREC_T) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_ADD_iq(VURegs *VU, uptr addr, int info) +{ + //Console::WriteLn("recVUMI_ADD_iq()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + } + + if ( _XYZW_SS ) { + if ( EEREC_D == EEREC_TEMP ) { + _vuFlipRegSS(VU, EEREC_S); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_S); + _vuFlipRegSS(VU, EEREC_D); // have to flip over EEREC_D for computing flags! + } + else if ( EEREC_D == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_D); + SSE_ADDSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_D); + } + else { + if ( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDSS_M32_to_XMM_custom(info, EEREC_D, addr); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_ADDPS_XMM_to_XMM_custom(info, EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + } + } + else { + if ( (_X_Y_Z_W != 0xf) || (EEREC_D == EEREC_S) || (EEREC_D == EEREC_TEMP) ) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + } + + if (_X_Y_Z_W != 0xf) { + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if ( EEREC_D == EEREC_TEMP ) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); + else if ( EEREC_D == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + else { + SSE_MOVSS_M32_to_XMM(EEREC_D, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); + SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_ADD_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn("recVUMI_ADD_xyzw()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + } + + if ( _Ft_ == 0 && xyzw < 3 ) { // just move since adding zero + if ( _X_Y_Z_W == 0x8 ) { VU_MERGE_REGS(EEREC_D, EEREC_S); } + else if ( _X_Y_Z_W != 0xf ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + else if ( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP) ) { + if ( xyzw == 0 ) { + if ( EEREC_D == EEREC_T ) SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + else if( _Fs_ == 0 && !_W ) { // just move + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if ( _X_Y_Z_W != 0xf ) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if( EEREC_D == EEREC_TEMP ) { _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); } + else if( EEREC_D == EEREC_S ) { _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); } + else { _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_ADDi(VURegs *VU, int info) { recVUMI_ADD_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_ADDq(VURegs *VU, int info) { recVUMI_ADD_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_ADDx(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 0, info); } +void recVUMI_ADDy(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 1, info); } +void recVUMI_ADDz(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 2, info); } +void recVUMI_ADDw(VURegs *VU, int info) { recVUMI_ADD_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ADDA*, ADDA_iq*, ADDA_xyzw* +//------------------------------------------------------------------ +void recVUMI_ADDA(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_ADDA()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + + if( _X_Y_Z_W == 8 ) { + if (EEREC_ACC == EEREC_S) SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); // Can this case happen? (cottonvibes) + else if (EEREC_ACC == EEREC_T) SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_S); // Can this case happen? + else { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + else { + if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_T); // Can this case happen? + else if( EEREC_ACC == EEREC_T ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); // Can this case happen? + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_ADDA_iq(VURegs *VU, uptr addr, int info) +{ + //Console::WriteLn("recVUMI_ADDA_iq()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + } + + if( _XYZW_SS ) { + assert( EEREC_ACC != EEREC_TEMP ); + if( EEREC_ACC == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_ACC); + SSE_ADDSS_M32_to_XMM(EEREC_ACC, addr); + _vuFlipRegSS(VU, EEREC_ACC); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_ADDSS_M32_to_XMM(EEREC_ACC, addr); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + } + } + else { + if( _X_Y_Z_W != 0xf || EEREC_ACC == EEREC_S ) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + } + + if (_X_Y_Z_W != 0xf) { + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + else { + if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + else { + SSE_MOVSS_M32_to_XMM(EEREC_ACC, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_ACC, EEREC_ACC, 0x00); + SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_ADDA_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn("recVUMI_ADDA_xyzw()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + } + + if( _X_Y_Z_W == 8 ) { + assert( EEREC_ACC != EEREC_T ); + if( xyzw == 0 ) { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + if( _Fs_ == 0 ) { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + else { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_ADDSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + } + } + else { + if( _X_Y_Z_W != 0xf || EEREC_ACC == EEREC_S ) + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + + if (_X_Y_Z_W != 0xf) { + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + else { + if( EEREC_ACC == EEREC_S ) SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + else { + _unpackVF_xyzw(EEREC_ACC, EEREC_T, xyzw); + SSE_ADDPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_ADDAi(VURegs *VU, int info) { recVUMI_ADDA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_ADDAq(VURegs *VU, int info) { recVUMI_ADDA_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_ADDAx(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 0, info); } +void recVUMI_ADDAy(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 1, info); } +void recVUMI_ADDAz(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 2, info); } +void recVUMI_ADDAw(VURegs *VU, int info) { recVUMI_ADDA_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SUB*, SUB_iq*, SUB_xyzw* +//------------------------------------------------------------------ +void recVUMI_SUB(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_SUB()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + + if( EEREC_S == EEREC_T ) { + if (_X_Y_Z_W != 0xf) SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&SSEmovMask[15-_X_Y_Z_W][0]); + else SSE_XORPS_XMM_to_XMM(EEREC_D, EEREC_D); + } + else if( _X_Y_Z_W == 8 ) { + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + if (EEREC_D == EEREC_S) { + if (_Ft_) SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + else if (EEREC_D == EEREC_T) { + if (_Ft_) { + SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + } + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + if (_Ft_) SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else { + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if( ( _Ft_ > 0 ) || _W ) SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if (EEREC_D == EEREC_S) SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_T); + else if (EEREC_D == EEREC_T) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_SUB_iq(VURegs *VU, uptr addr, int info) +{ + //Console::WriteLn("recVUMI_SUB_iq()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + } + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + + if( _XYZW_SS ) { + if( EEREC_D == EEREC_TEMP ) { + _vuFlipRegSS(VU, EEREC_S); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_S); + _vuFlipRegSS(VU, EEREC_D); + } + else if( EEREC_D == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_D); + SSE_SUBSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_D); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBSS_M32_to_XMM(EEREC_D, addr); + } + else { + _vuMoveSS(VU, EEREC_TEMP, EEREC_S); + _vuFlipRegSS(VU, EEREC_D); + SSE_SUBSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_D); + } + } + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + VU_MERGE_REGS(EEREC_D, t1reg); + _freeXMMreg(t1reg); + } + else { + // negate + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + } + else { + if( EEREC_D == EEREC_TEMP ) { + SSE_XORPS_M128_to_XMM(EEREC_D, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_SUB_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn("recVUMI_SUB_xyzw()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if ( !_Fd_ ) info = (info & ~PROCESS_EE_SET_D(0xf)) | PROCESS_EE_SET_D(EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + } + + if ( _X_Y_Z_W == 8 ) { + if ( (xyzw == 0) && (_Ft_ == _Fs_) ) { + SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&SSEmovMask[7][0]); + } + else if ( EEREC_D == EEREC_TEMP ) { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + if ( (_Ft_ > 0) || (xyzw == 3) ) { + _vuFlipRegSS_xyzw(EEREC_T, xyzw); + SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_T); + _vuFlipRegSS_xyzw(EEREC_T, xyzw); + } + } + else { + if ( (_Ft_ > 0) || (xyzw == 3) ) { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + else { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + VU_MERGE_REGS(EEREC_D, t1reg); + _freeXMMreg(t1reg); + } + else { + // negate + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + } + else { + if( EEREC_D == EEREC_TEMP ) { + SSE_XORPS_M128_to_XMM(EEREC_D, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_SUBi(VURegs *VU, int info) { recVUMI_SUB_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_SUBq(VURegs *VU, int info) { recVUMI_SUB_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_SUBx(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 0, info); } +void recVUMI_SUBy(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 1, info); } +void recVUMI_SUBz(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 2, info); } +void recVUMI_SUBw(VURegs *VU, int info) { recVUMI_SUB_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// SUBA*, SUBA_iq, SUBA_xyzw +//------------------------------------------------------------------ +void recVUMI_SUBA(VURegs *VU, int info) +{ + //Console::WriteLn("recVUMI_SUBA()"); + if ( _X_Y_Z_W == 0 ) goto flagUpdate; + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + + if( EEREC_S == EEREC_T ) { + if (_X_Y_Z_W != 0xf) SSE_ANDPS_M128_to_XMM(EEREC_ACC, (uptr)&SSEmovMask[15-_X_Y_Z_W][0]); + else SSE_XORPS_XMM_to_XMM(EEREC_ACC, EEREC_ACC); + } + else if( _X_Y_Z_W == 8 ) { + if (EEREC_ACC == EEREC_S) SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); + else if (EEREC_ACC == EEREC_T) { + SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + else { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + else { + if( EEREC_ACC == EEREC_S ) SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_T); + else if( EEREC_ACC == EEREC_T ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + } +flagUpdate: + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_SUBA_iq(VURegs *VU, uptr addr, int info) +{ + //Console::WriteLn ("recVUMI_SUBA_iq"); + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + } + + if( _XYZW_SS ) { + if( EEREC_ACC == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_ACC); + SSE_SUBSS_M32_to_XMM(EEREC_ACC, addr); + _vuFlipRegSS(VU, EEREC_ACC); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBSS_M32_to_XMM(EEREC_ACC, addr); + } + else { + _vuMoveSS(VU, EEREC_TEMP, EEREC_S); + _vuFlipRegSS(VU, EEREC_ACC); + SSE_SUBSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_ACC); + } + } + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + VU_MERGE_REGS(EEREC_ACC, t1reg); + _freeXMMreg(t1reg); + } + else { + // negate + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + } + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_SUBA_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn ("recVUMI_SUBA_xyzw"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + } + + if( _X_Y_Z_W == 8 ) { + if( xyzw == 0 ) { + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_T); + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBSS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + } + else { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_S); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + VU_MERGE_REGS(EEREC_ACC, t1reg); + _freeXMMreg(t1reg); + } + else { + // negate + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_ACC, EEREC_TEMP); + } + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_ACC, EEREC_S); + SSE_SUBPS_XMM_to_XMM(EEREC_ACC, EEREC_TEMP); + } + } + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_SUBAi(VURegs *VU, int info) { recVUMI_SUBA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_SUBAq(VURegs *VU, int info) { recVUMI_SUBA_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_SUBAx(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 0, info); } +void recVUMI_SUBAy(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 1, info); } +void recVUMI_SUBAz(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 2, info); } +void recVUMI_SUBAw(VURegs *VU, int info) { recVUMI_SUBA_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MUL +//------------------------------------------------------------------ +void recVUMI_MUL_toD(VURegs *VU, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MUL_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + } + + if (_X_Y_Z_W == 1 && (_Ft_ == 0 || _Fs_==0) ) { // W + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, _Ft_ ? EEREC_T : EEREC_S); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else if( _Fd_ == _Fs_ && _Fs_ == _Ft_ && _XYZW_SS ) { + _vuFlipRegSS(VU, EEREC_D); + SSE_MULSS_XMM_to_XMM(EEREC_D, EEREC_D); + _vuFlipRegSS(VU, EEREC_D); + } + else if( _X_Y_Z_W == 8 ) { + if (regd == EEREC_S) SSE_MULSS_XMM_to_XMM(regd, EEREC_T); + else if (regd == EEREC_T) SSE_MULSS_XMM_to_XMM(regd, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_XMM_to_XMM(regd, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + else if (regd == EEREC_T) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + else { + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); + SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + } + } +} + +void recVUMI_MUL_iq_toD(VURegs *VU, uptr addr, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MUL_iq_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + } + + if( _XYZW_SS ) { + if( regd == EEREC_TEMP ) { + _vuFlipRegSS(VU, EEREC_S); + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_M32_to_XMM(regd, addr); + _vuFlipRegSS(VU, EEREC_S); + _vuFlipRegSS(VU, regd); + } + else if( regd == EEREC_S ) { + _vuFlipRegSS(VU, regd); + SSE_MULSS_M32_to_XMM(regd, addr); + _vuFlipRegSS(VU, regd); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_M32_to_XMM(regd, addr); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + } + } + else { + if( _X_Y_Z_W != 0xf || regd == EEREC_TEMP || regd == EEREC_S ) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + } + + if (_X_Y_Z_W != 0xf) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_TEMP ) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + else if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); + else { + SSE_MOVSS_M32_to_XMM(regd, addr); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x00); + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + } + } + } +} + +void recVUMI_MUL_xyzw_toD(VURegs *VU, int xyzw, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MUL_xyzw_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + } + if (_Fs_) { // This is needed for alot of games; so always clamp this operand + if (CHECK_VU_SIGN_OVERFLOW) vFloats4_useEAX[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set + else vFloats2[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set + } + if( _Ft_ == 0 ) { + if( xyzw < 3 ) { + if (_X_Y_Z_W != 0xf) { + SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else SSE_XORPS_XMM_to_XMM(regd, regd); + } + else { + assert(xyzw==3); + if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); + } + } + else if( _X_Y_Z_W == 8 ) { + if( regd == EEREC_TEMP ) { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MULSS_XMM_to_XMM(regd, EEREC_S); + } + else { + if( xyzw == 0 ) { + if( regd == EEREC_T ) { + SSE_MULSS_XMM_to_XMM(regd, EEREC_S); + } + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_XMM_to_XMM(regd, EEREC_T); + } + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_XMM_to_XMM(regd, EEREC_TEMP); + } + } + } + else { + if( _X_Y_Z_W != 0xf || regd == EEREC_TEMP || regd == EEREC_S ) + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + + if (_X_Y_Z_W != 0xf) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_TEMP ) SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + else if (regd == EEREC_S) SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); + else { + _unpackVF_xyzw(regd, EEREC_T, xyzw); + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + } + } + } +} + +void recVUMI_MUL(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MUL"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MUL_toD(VU, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MUL_iq(VURegs *VU, int addr, int info) +{ + //Console::WriteLn ("recVUMI_MUL_iq"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MUL_iq_toD(VU, addr, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); + // spacefisherman needs overflow checking on MULi.z +} + +void recVUMI_MUL_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn ("recVUMI_MUL_xyzw"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MUL_xyzw_toD(VU, xyzw, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MULi(VURegs *VU, int info) { recVUMI_MUL_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MULq(VURegs *VU, int info) { recVUMI_MUL_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_MULx(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 0, info); } +void recVUMI_MULy(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 1, info); } +void recVUMI_MULz(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 2, info); } +void recVUMI_MULw(VURegs *VU, int info) { recVUMI_MUL_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MULA +//------------------------------------------------------------------ +void recVUMI_MULA( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MULA"); + recVUMI_MUL_toD(VU, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MULA_iq(VURegs *VU, int addr, int info) +{ + //Console::WriteLn ("recVUMI_MULA_iq"); + recVUMI_MUL_iq_toD(VU, addr, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MULA_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn ("recVUMI_MULA_xyzw"); + recVUMI_MUL_xyzw_toD(VU, xyzw, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MULAi(VURegs *VU, int info) { recVUMI_MULA_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MULAq(VURegs *VU, int info) { recVUMI_MULA_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_MULAx(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 0, info); } +void recVUMI_MULAy(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 1, info); } +void recVUMI_MULAz(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 2, info); } +void recVUMI_MULAw(VURegs *VU, int info) { recVUMI_MULA_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MADD +//------------------------------------------------------------------ +void recVUMI_MADD_toD(VURegs *VU, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MADD_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + } + + + if( _X_Y_Z_W == 8 ) { + if( regd == EEREC_ACC ) { + SSE_MOVSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if (regd == EEREC_T) { + SSE_MULSS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + else if (regd == EEREC_S) { + SSE_MULSS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_ACC ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if (regd == EEREC_T) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else if (regd == EEREC_S) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); + SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + } +} + +void recVUMI_MADD_iq_toD(VURegs *VU, uptr addr, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MADD_iq_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + vuFloat3(addr); + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + } + + if( _X_Y_Z_W == 8 ) { + if( _Fs_ == 0 ) { + // do nothing if regd == ACC (ACCx <= ACCx + 0.0 * *addr) + if( regd != EEREC_ACC ) { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); + } + return; + } + + if( regd == EEREC_ACC ) { + assert( EEREC_TEMP < iREGCNT_XMM ); + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if( regd == EEREC_S ) { + SSE_MULSS_M32_to_XMM(regd, addr); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_S); + SSE_MULSS_M32_to_XMM(regd, addr); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + } + else { + if( _Fs_ == 0 ) { + if( regd == EEREC_ACC ) { // ACCxyz is unchanged, ACCw <= ACCw + *addr + if( _W ) { // if _W is zero, do nothing + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); // { *addr, 0, 0, 0 } + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x27); // { 0, 0, 0, *addr } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); // { ACCx, ACCy, ACCz, ACCw + *addr } + } + } + else { // DESTxyz <= ACCxyz, DESTw <= ACCw + *addr + if( _W ) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); // { *addr, 0, 0, 0 } + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x27); // { 0, 0, 0, *addr } + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); // { ACCx, ACCy, ACCz, ACCw + *addr } + } + else SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + + return; + } + + if( _X_Y_Z_W != 0xf || regd == EEREC_ACC || regd == EEREC_TEMP || regd == EEREC_S ) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + } + + if (_X_Y_Z_W != 0xf) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_ACC ) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if( regd == EEREC_S ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else if( regd == EEREC_TEMP ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + SSE_MOVSS_M32_to_XMM(regd, addr); + SSE_SHUFPS_XMM_to_XMM(regd, regd, 0x00); + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + } + } +} + +void recVUMI_MADD_xyzw_toD(VURegs *VU, int xyzw, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MADD_xyzw_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + } + if (_Fs_) { // This is needed for alot of games; so always clamp this operand + if (CHECK_VU_SIGN_OVERFLOW) vFloats4_useEAX[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set + else vFloats2[_X_Y_Z_W]( EEREC_S, EEREC_TEMP ); // Always clamp EEREC_S, regardless if CHECK_VU_OVERFLOW is set + } + if( _Ft_ == 0 ) { + + if( xyzw == 3 ) { + // just add + if( _X_Y_Z_W == 8 ) { + if( regd == EEREC_S ) SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); + SSE_ADDSS_XMM_to_XMM(regd, EEREC_S); + } + } + else { + if( _X_Y_Z_W != 0xf ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_S ) SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + else { + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_ADDPS_XMM_to_XMM(regd, EEREC_S); + } + } + } + } + else { + // just move acc to regd + if( _X_Y_Z_W != 0xf ) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); + } + + return; + } + + if( _X_Y_Z_W == 8 ) { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + + if( regd == EEREC_ACC ) { + SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if( regd == EEREC_S ) { + SSE_MULSS_XMM_to_XMM(regd, EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + else if( regd == EEREC_TEMP ) { + SSE_MULSS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + SSE_MOVSS_XMM_to_XMM(regd, EEREC_ACC); + SSE_MULSS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, 8); } + SSE_ADDSS_XMM_to_XMM(regd, EEREC_TEMP); + } + } + else { + if( _X_Y_Z_W != 0xf || regd == EEREC_ACC || regd == EEREC_TEMP || regd == EEREC_S ) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + } + + if (_X_Y_Z_W != 0xf) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + + VU_MERGE_REGS(regd, EEREC_TEMP); + } + else { + if( regd == EEREC_ACC ) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if( regd == EEREC_S ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else if( regd == EEREC_TEMP ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + else { + _unpackVF_xyzw(regd, EEREC_T, xyzw); + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_ADDPS_XMM_to_XMM(regd, EEREC_ACC); + } + } + } +} + +void recVUMI_MADD(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MADD"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MADD_toD(VU, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MADD_iq(VURegs *VU, int addr, int info) +{ + //Console::WriteLn ("recVUMI_MADD_iq"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MADD_iq_toD(VU, addr, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MADD_xyzw(VURegs *VU, int xyzw, int info) +{ + //Console::WriteLn ("recVUMI_MADD_xyzw"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MADD_xyzw_toD(VU, xyzw, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); + // super bust-a-move arrows needs overflow clamping +} + +void recVUMI_MADDi(VURegs *VU, int info) { recVUMI_MADD_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MADDq(VURegs *VU, int info) { recVUMI_MADD_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_MADDx(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 0, info); } +void recVUMI_MADDy(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 1, info); } +void recVUMI_MADDz(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 2, info); } +void recVUMI_MADDw(VURegs *VU, int info) { recVUMI_MADD_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MADDA +//------------------------------------------------------------------ +void recVUMI_MADDA( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MADDA"); + recVUMI_MADD_toD(VU, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAi( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAi"); + recVUMI_MADD_iq_toD( VU, VU_VI_ADDR(REG_I, 1), EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAq( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAq "); + recVUMI_MADD_iq_toD( VU, VU_REGQ_ADDR, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAx( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAx"); + recVUMI_MADD_xyzw_toD(VU, 0, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAy( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAy"); + recVUMI_MADD_xyzw_toD(VU, 1, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAz( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAz"); + recVUMI_MADD_xyzw_toD(VU, 2, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MADDAw( VURegs *VU , int info) +{ + //Console::WriteLn ("recVUMI_MADDAw"); + recVUMI_MADD_xyzw_toD(VU, 3, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MSUB +//------------------------------------------------------------------ +void recVUMI_MSUB_toD(VURegs *VU, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MSUB_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + } + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_ACC); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + VU_MERGE_REGS(regd, t1reg); + _freeXMMreg(t1reg); + } + else { + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + } + else { + if( regd == EEREC_S ) { + assert( regd != EEREC_ACC ); + SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); + } + else if( regd == EEREC_T ) { + assert( regd != EEREC_ACC ); + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); + } + else if( regd == EEREC_TEMP ) { + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_S); + SSE_MULPS_XMM_to_XMM(regd, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); + } + } +} + +void recVUMI_MSUB_temp_toD(VURegs *VU, int regd, int info) +{ + //Console::WriteLn ("recVUMI_MSUB_temp_toD"); + + if (_X_Y_Z_W != 0xf) { + int t1reg = _vuGetTempXMMreg(info); + + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + + if( t1reg >= 0 ) { + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_ACC); + SSE_SUBPS_XMM_to_XMM(t1reg, EEREC_TEMP); + + if ( regd != EEREC_TEMP ) { VU_MERGE_REGS(regd, t1reg); } + else SSE_MOVAPS_XMM_to_XMM(regd, t1reg); + + _freeXMMreg(t1reg); + } + else { + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + VU_MERGE_REGS(regd, EEREC_TEMP); + } + } + else { + if( regd == EEREC_ACC ) { + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); + } + else if( regd == EEREC_S ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_TEMP); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); + } + else if( regd == EEREC_TEMP ) { + SSE_MULPS_XMM_to_XMM(regd, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, regd, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_XORPS_M128_to_XMM(regd, (uptr)&const_clip[4]); + } + else { + SSE_MOVAPS_XMM_to_XMM(regd, EEREC_ACC); + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + if (CHECK_VU_EXTRA_OVERFLOW) { vuFloat_useEAX( info, EEREC_TEMP, _X_Y_Z_W ); } + SSE_SUBPS_XMM_to_XMM(regd, EEREC_TEMP); + } + } +} + +void recVUMI_MSUB_iq_toD(VURegs *VU, int regd, int addr, int info) +{ + //Console::WriteLn ("recVUMI_MSUB_iq_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + vuFloat3(addr); + } + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + recVUMI_MSUB_temp_toD(VU, regd, info); +} + +void recVUMI_MSUB_xyzw_toD(VURegs *VU, int regd, int xyzw, int info) +{ + //Console::WriteLn ("recVUMI_MSUB_xyzw_toD"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 1 << (3 - xyzw)); + vuFloat5_useEAX( EEREC_ACC, EEREC_TEMP, _X_Y_Z_W ); + } + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + recVUMI_MSUB_temp_toD(VU, regd, info); +} + +void recVUMI_MSUB(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MSUB"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_toD(VU, EEREC_D, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MSUB_iq(VURegs *VU, int addr, int info) +{ + //Console::WriteLn ("recVUMI_MSUB_iq"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_iq_toD(VU, EEREC_D, addr, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MSUBi(VURegs *VU, int info) { recVUMI_MSUB_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MSUBq(VURegs *VU, int info) { recVUMI_MSUB_iq(VU, VU_REGQ_ADDR, info); } +void recVUMI_MSUBx(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MSUBx"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 0, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MSUBy(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MSUBy"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 1, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MSUBz(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MSUBz"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 2, info); + recUpdateFlags(VU, EEREC_D, info); +} + +void recVUMI_MSUBw(VURegs *VU, int info) +{ + //Console::WriteLn ("recVUMI_MSUBw"); + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + recVUMI_MSUB_xyzw_toD(VU, EEREC_D, 3, info); + recUpdateFlags(VU, EEREC_D, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MSUBA +//------------------------------------------------------------------ +void recVUMI_MSUBA( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBA"); + recVUMI_MSUB_toD(VU, EEREC_ACC, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAi( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAi "); + recVUMI_MSUB_iq_toD( VU, EEREC_ACC, VU_VI_ADDR(REG_I, 1), info ); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAq( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAq"); + recVUMI_MSUB_iq_toD( VU, EEREC_ACC, VU_REGQ_ADDR, info ); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAx( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAx"); + recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 0, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAy( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAy"); + recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 1, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAz( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAz "); + recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 2, info); + recUpdateFlags(VU, EEREC_ACC, info); +} + +void recVUMI_MSUBAw( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_MSUBAw"); + recVUMI_MSUB_xyzw_toD(VU, EEREC_ACC, 3, info); + recUpdateFlags(VU, EEREC_ACC, info); +} +//------------------------------------------------------------------ + + +static const u32 PCSX2_ALIGNED16(special_mask[4]) = {0xffffffff, 0x80000000, 0xffffffff, 0x80000000}; +static const u32 PCSX2_ALIGNED16(special_mask2[4]) = {0, 0x40000000, 0, 0x40000000}; + +u32 PCSX2_ALIGNED16(temp_loc[4]); +u32 PCSX2_ALIGNED16(temp_loc2[4]); + +//MAX/MINI are non-arithmetic operations that implicitly support numbers with the EXP field being 0 ("denormals"). +// +//As such, they are sometimes used for integer move and (positive!) integer max/min, knowing that integers that +//represent denormals will not be flushed to 0. +// +//As such, this implementation performs a non-arithmetic operation that supports "denormals" and "infs/nans". +//There might be an easier way to do it but here, MAX/MIN is performed with PMAXPD/PMINPD. +//Fake double-precision numbers are constructed by copying the sign of the original numbers, clearing the upper 32 bits, +//setting the 62nd bit to 1 (to ensure double-precision number is "normalized") and having the lower 32bits +//being the same as the original number. + +void MINMAXlogical(VURegs *VU, int info, int min, int mode, uptr addr = 0, int xyzw = 0) +//mode1 = iq, mode2 = xyzw, mode0 = normal +{ + int t1regbool = 0; + int t1reg = _vuGetTempXMMreg(info); + if (t1reg < 0) + { + t1regbool = 1; + for (t1reg = 0; ( (t1reg == EEREC_D) || (t1reg == EEREC_S) || (mode != 1 && t1reg == EEREC_T) + || (t1reg == EEREC_TEMP) ); t1reg++); // Find unused reg (For first temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)temp_loc, t1reg); // Backup t1reg XMM reg + } + int t2regbool = -1; + int t2reg = EEREC_TEMP; + if (EEREC_TEMP == EEREC_D || EEREC_TEMP == EEREC_S || (mode != 1 && EEREC_TEMP == EEREC_T)) + { + t2regbool = 0; + t2reg = _vuGetTempXMMreg(info); + if (t2reg < 0) + { + t2regbool = 1; + for (t2reg = 0; ( (t2reg == EEREC_D) || (t2reg == EEREC_S) || (mode != 1 && t2reg == EEREC_T) || + (t2reg == t1reg) || (t2reg == EEREC_TEMP) ); t2reg++); // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)temp_loc2, t2reg); // Backup t2reg XMM reg + } + } + + if (_X || _Y) + { + SSE2_PSHUFD_XMM_to_XMM(t1reg, EEREC_S, 0x50); + SSE2_PAND_M128_to_XMM(t1reg, (uptr)special_mask); + SSE2_POR_M128_to_XMM(t1reg, (uptr)special_mask2); + if (mode == 0) + SSE2_PSHUFD_XMM_to_XMM(t2reg, EEREC_T, 0x50); + else if (mode == 1) + { + SSE2_MOVD_M32_to_XMM(t2reg, addr); + SSE2_PSHUFD_XMM_to_XMM(t2reg, t2reg, 0x00); + } + else if (mode == 2) + _unpackVF_xyzw(t2reg, EEREC_T, xyzw); + SSE2_PAND_M128_to_XMM(t2reg, (uptr)special_mask); + SSE2_POR_M128_to_XMM(t2reg, (uptr)special_mask2); + if (min) + SSE2_MINPD_XMM_to_XMM(t1reg, t2reg); + else + SSE2_MAXPD_XMM_to_XMM(t1reg, t2reg); + SSE2_PSHUFD_XMM_to_XMM(t1reg, t1reg, 0x88); + VU_MERGE_REGS_CUSTOM(EEREC_D, t1reg, 0xc & _X_Y_Z_W); + } + + if (_Z || _W) + { + SSE2_PSHUFD_XMM_to_XMM(t1reg, EEREC_S, 0xfa); + SSE2_PAND_M128_to_XMM(t1reg, (uptr)special_mask); + SSE2_POR_M128_to_XMM(t1reg, (uptr)special_mask2); + if (mode == 0) + SSE2_PSHUFD_XMM_to_XMM(t2reg, EEREC_T, 0xfa); + else if (mode == 1) + { + SSE2_MOVD_M32_to_XMM(t2reg, addr); + SSE2_PSHUFD_XMM_to_XMM(t2reg, t2reg, 0x00); + } + else if (mode == 2) + _unpackVF_xyzw(t2reg, EEREC_T, xyzw); + SSE2_PAND_M128_to_XMM(t2reg, (uptr)special_mask); + SSE2_POR_M128_to_XMM(t2reg, (uptr)special_mask2); + if (min) + SSE2_MINPD_XMM_to_XMM(t1reg, t2reg); + else + SSE2_MAXPD_XMM_to_XMM(t1reg, t2reg); + SSE2_PSHUFD_XMM_to_XMM(t1reg, t1reg, 0x88); + VU_MERGE_REGS_CUSTOM(EEREC_D, t1reg, 0x3 & _X_Y_Z_W); + } + + if (t1regbool == 0) + _freeXMMreg(t1reg); + else if (t1regbool == 1) + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)temp_loc); // Restore t1reg XMM reg + if (t2regbool == 0) + _freeXMMreg(t2reg); + else if (t2regbool == 1) + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)temp_loc2); // Restore t2reg XMM reg +} + +//------------------------------------------------------------------ +// MAX +//------------------------------------------------------------------ + +void recVUMI_MAX(VURegs *VU, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MAX"); + + if (MINMAXFIX) + MINMAXlogical(VU, info, 0, 0); + else + { + + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + + if( _X_Y_Z_W == 8 ) { + if (EEREC_D == EEREC_S) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); + else if (EEREC_D == EEREC_T) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if( EEREC_D == EEREC_S ) SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_T); + else if( EEREC_D == EEREC_T ) SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + } +} + +void recVUMI_MAX_iq(VURegs *VU, uptr addr, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MAX_iq"); + + if (MINMAXFIX) + MINMAXlogical(VU, info, 0, 1, addr); + else + { + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + vuFloat3(addr); + + if( _XYZW_SS ) { + if( EEREC_D == EEREC_TEMP ) { + _vuFlipRegSS(VU, EEREC_S); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_S); + + // have to flip over EEREC_D if computing flags! + //if( (info & PROCESS_VU_UPDATEFLAGS) ) + _vuFlipRegSS(VU, EEREC_D); + } + else if( EEREC_D == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_D); + SSE_MAXSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_D); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXSS_M32_to_XMM(EEREC_D, addr); + } + else { + _vuMoveSS(VU, EEREC_TEMP, EEREC_S); + _vuFlipRegSS(VU, EEREC_D); + SSE_MAXSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_D); + } + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if(EEREC_D == EEREC_S) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_D, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); + SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + } +} + +void recVUMI_MAX_xyzw(VURegs *VU, int xyzw, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MAX_xyzw"); + + if (_Fs_ == 0 && _Ft_ == 0) + { + if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { + if( xyzw < 3 ) { + SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, (uptr)s_fones); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + else if (_X_Y_Z_W != 0xf) { + if( xyzw < 3 ) { + if( _X_Y_Z_W & 1 ) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&VU->VF[0].UL[0]); // w included, so insert the whole reg + else SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); // w not included, can zero out + } + else SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)s_fones); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + //If VF0.w isnt chosen as the constant, then its going to be MAX( 0, VF0 ), so the result is VF0 + if( xyzw < 3 ) { SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)&VU->VF[0].UL[0]); } + else SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)s_fones); + } + return; + } + + if (MINMAXFIX) + MINMAXlogical(VU, info, 0, 2, 0, xyzw); + else + { + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + + if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { + if( xyzw == 0 ) { + if( EEREC_D == EEREC_S ) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); + else if( EEREC_D == EEREC_T ) SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MAXSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + else if (_X_Y_Z_W != 0xf) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MAXPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if (EEREC_D == EEREC_S) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); + SSE_MAXPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + } +} + +void recVUMI_MAXi(VURegs *VU, int info) { recVUMI_MAX_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MAXx(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 0, info); } +void recVUMI_MAXy(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 1, info); } +void recVUMI_MAXz(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 2, info); } +void recVUMI_MAXw(VURegs *VU, int info) { recVUMI_MAX_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// MINI +//------------------------------------------------------------------ +void recVUMI_MINI(VURegs *VU, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MINI"); + + if (MINMAXFIX) + MINMAXlogical(VU, info, 1, 0); + else + { + + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, _X_Y_Z_W ); + + if( _X_Y_Z_W == 8 ) { + if (EEREC_D == EEREC_S) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); + else if (EEREC_D == EEREC_T) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if( EEREC_D == EEREC_S ) { + //ClampUnordered(EEREC_T, EEREC_TEMP, 0); // need for GT4 vu0rec + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_T); + } + else if( EEREC_D == EEREC_T ) { + //ClampUnordered(EEREC_S, EEREC_TEMP, 0); // need for GT4 vu0rec + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + } +} + +void recVUMI_MINI_iq(VURegs *VU, uptr addr, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MINI_iq"); + + if (MINMAXFIX) + MINMAXlogical(VU, info, 1, 1, addr); + else + { + + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + vuFloat3(addr); + + if( _XYZW_SS ) { + if( EEREC_D == EEREC_TEMP ) { + _vuFlipRegSS(VU, EEREC_S); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_S); + + // have to flip over EEREC_D if computing flags! + //if( (info & PROCESS_VU_UPDATEFLAGS) ) + _vuFlipRegSS(VU, EEREC_D); + } + else if( EEREC_D == EEREC_S ) { + _vuFlipRegSS(VU, EEREC_D); + SSE_MINSS_M32_to_XMM(EEREC_D, addr); + _vuFlipRegSS(VU, EEREC_D); + } + else { + if( _X ) { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINSS_M32_to_XMM(EEREC_D, addr); + } + else { + _vuMoveSS(VU, EEREC_TEMP, EEREC_S); + _vuFlipRegSS(VU, EEREC_D); + SSE_MINSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + _vuFlipRegSS(VU, EEREC_D); + } + } + } + else if (_X_Y_Z_W != 0xf) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if(EEREC_D == EEREC_S) { + SSE_MOVSS_M32_to_XMM(EEREC_TEMP, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0x00); + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + SSE_MOVSS_M32_to_XMM(EEREC_D, addr); + SSE_SHUFPS_XMM_to_XMM(EEREC_D, EEREC_D, 0x00); + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + } +} + +void recVUMI_MINI_xyzw(VURegs *VU, int xyzw, int info) +{ + if ( _Fd_ == 0 ) return; + //Console::WriteLn ("recVUMI_MINI_xyzw"); + + if (_Fs_ == 0 && _Ft_ == 0) + { + if( _X_Y_Z_W == 0xf ) + { + //If VF0.w is the constant, the result will match VF0, else its all 0's + if(xyzw == 3) SSE_MOVAPS_M128_to_XMM(EEREC_D, (uptr)&VU->VF[0].UL[0]); + else SSE_XORPS_XMM_to_XMM(EEREC_D, EEREC_D); + } + else + { + //If VF0.w is the constant, the result will match VF0, else its all 0's + if(xyzw == 3) SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&VU->VF[0].UL[0]); + else SSE_XORPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + return; + } + if (MINMAXFIX) + MINMAXlogical(VU, info, 1, 2, 0, xyzw); + else + { + if (_Fs_) vuFloat4_useEAX( EEREC_S, EEREC_TEMP, _X_Y_Z_W ); // Always do Preserved Sign Clamping + if (_Ft_) vuFloat4_useEAX( EEREC_T, EEREC_TEMP, ( 1 << (3 - xyzw) ) ); + + if( _X_Y_Z_W == 8 && (EEREC_D != EEREC_TEMP)) { + if( xyzw == 0 ) { + if( EEREC_D == EEREC_S ) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); + else if( EEREC_D == EEREC_T ) SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_S); + else { + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_T); + } + } + else { + _unpackVFSS_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S); + SSE_MINSS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + } + else if (_X_Y_Z_W != 0xf) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MINPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + VU_MERGE_REGS(EEREC_D, EEREC_TEMP); + } + else { + if (EEREC_D == EEREC_S) { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, xyzw); + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_TEMP); + } + else { + _unpackVF_xyzw(EEREC_D, EEREC_T, xyzw); + SSE_MINPS_XMM_to_XMM(EEREC_D, EEREC_S); + } + } + } +} + +void recVUMI_MINIi(VURegs *VU, int info) { recVUMI_MINI_iq(VU, VU_VI_ADDR(REG_I, 1), info); } +void recVUMI_MINIx(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 0, info); } +void recVUMI_MINIy(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 1, info); } +void recVUMI_MINIz(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 2, info); } +void recVUMI_MINIw(VURegs *VU, int info) { recVUMI_MINI_xyzw(VU, 3, info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// OPMULA +//------------------------------------------------------------------ +void recVUMI_OPMULA( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_OPMULA"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, 0xE); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 0xE); + } + + SSE_MOVAPS_XMM_to_XMM( EEREC_TEMP, EEREC_S ); + SSE_SHUFPS_XMM_to_XMM( EEREC_T, EEREC_T, 0xD2 ); // EEREC_T = WYXZ + SSE_SHUFPS_XMM_to_XMM( EEREC_TEMP, EEREC_TEMP, 0xC9 ); // EEREC_TEMP = WXZY + SSE_MULPS_XMM_to_XMM( EEREC_TEMP, EEREC_T ); + + VU_MERGE_REGS_CUSTOM(EEREC_ACC, EEREC_TEMP, 14); + + // revert EEREC_T + if( EEREC_T != EEREC_ACC ) + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xC9); + + recUpdateFlags(VU, EEREC_ACC, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// OPMSUB +//------------------------------------------------------------------ +void recVUMI_OPMSUB( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_OPMSUB"); + if (CHECK_VU_EXTRA_OVERFLOW) { + if (_Fs_) vuFloat5_useEAX( EEREC_S, EEREC_TEMP, 0xE); + if (_Ft_) vuFloat5_useEAX( EEREC_T, EEREC_TEMP, 0xE); + } + + if( !_Fd_ ) info |= PROCESS_EE_SET_D(EEREC_TEMP); + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xD2); // EEREC_T = WYXZ + SSE_SHUFPS_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP, 0xC9); // EEREC_TEMP = WXZY + SSE_MULPS_XMM_to_XMM(EEREC_TEMP, EEREC_T); + + // negate and add + SSE_XORPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[4]); + SSE_ADDPS_XMM_to_XMM(EEREC_TEMP, EEREC_ACC); + VU_MERGE_REGS_CUSTOM(EEREC_D, EEREC_TEMP, 14); + + // revert EEREC_T + if( EEREC_T != EEREC_D ) SSE_SHUFPS_XMM_to_XMM(EEREC_T, EEREC_T, 0xC9); + + recUpdateFlags(VU, EEREC_D, info); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// NOP +//------------------------------------------------------------------ +void recVUMI_NOP( VURegs *VU, int info ) +{ + //Console::WriteLn ("recVUMI_NOP"); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// recVUMI_FTOI_Saturate() - Saturates result from FTOI Instructions +//------------------------------------------------------------------ +static const PCSX2_ALIGNED16(int rec_const_0x8000000[4]) = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 }; + +void recVUMI_FTOI_Saturate(int rec_s, int rec_t, int rec_tmp1, int rec_tmp2) +{ + //Console::WriteLn ("recVUMI_FTOI_Saturate"); + //Duplicate the xor'd sign bit to the whole value + //FFFF FFFF for positive, 0 for negative + SSE_MOVAPS_XMM_to_XMM(rec_tmp1, rec_s); + SSE2_PXOR_M128_to_XMM(rec_tmp1, (uptr)&const_clip[4]); + SSE2_PSRAD_I8_to_XMM(rec_tmp1, 31); + + //Create mask: 0 where !=8000 0000 + SSE_MOVAPS_XMM_to_XMM(rec_tmp2, rec_t); + SSE2_PCMPEQD_M128_to_XMM(rec_tmp2, (uptr)&const_clip[4]); + + //AND the mask w/ the edit values + SSE_ANDPS_XMM_to_XMM(rec_tmp1, rec_tmp2); + + //if v==8000 0000 && positive -> 8000 0000 + FFFF FFFF -> 7FFF FFFF + //if v==8000 0000 && negative -> 8000 0000 + 0 -> 8000 0000 + //if v!=8000 0000 -> v+0 (masked from the and) + + //Add the values as needed + SSE2_PADDD_XMM_to_XMM(rec_t, rec_tmp1); +} +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// FTOI 0/4/12/15 +//------------------------------------------------------------------ +static PCSX2_ALIGNED16(float FTIO_Temp1[4]) = { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; +static PCSX2_ALIGNED16(float FTIO_Temp2[4]) = { 0x00000000, 0x00000000, 0x00000000, 0x00000000 }; +void recVUMI_FTOI0(VURegs *VU, int info) +{ + int t1reg, t2reg; // Temp XMM regs + + if ( _Ft_ == 0 ) return; + + //Console::WriteLn ("recVUMI_FTOI0"); + + if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg (For first temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg + } + + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + if (EEREC_T != EEREC_S) { + SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); + vuFloat_useEAX( info, EEREC_T, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_T, EEREC_T); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + } + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg (For first temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg + } + + SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + } + } +} + +void recVUMI_FTOIX(VURegs *VU, int addr, int info) +{ + int t1reg, t2reg; // Temp XMM regs + + if ( _Ft_ == 0 ) return; + + //Console::WriteLn ("recVUMI_FTOIX"); + if (_X_Y_Z_W != 0xf) { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); + vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg (For first temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg + } + + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + } + else { + if (EEREC_T != EEREC_S) { + SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_S); + SSE_MULPS_M128_to_XMM(EEREC_T, addr); + vuFloat_useEAX( info, EEREC_T, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_T, EEREC_T); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_T, EEREC_TEMP, t1reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + } + } + else { + SSE_MOVAPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); + vuFloat_useEAX( info, EEREC_TEMP, 0xf ); // Clamp Infs and NaNs to pos/neg fmax (NaNs always to positive fmax) + SSE2_CVTTPS2DQ_XMM_to_XMM(EEREC_TEMP, EEREC_TEMP); + + t1reg = _vuGetTempXMMreg(info); + + if( t1reg >= 0 ) { // If theres a temp XMM reg available + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg)); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t2reg); // Backup XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp1); // Restore XMM reg + _freeXMMreg(t1reg); // Free temp reg + } + else { // No temp reg available + for (t1reg = 0; ( (t1reg == EEREC_S) || (t1reg == EEREC_T) || (t1reg == EEREC_TEMP) ); t1reg++) + ; // Find unused reg (For first temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp1, t1reg); // Backup t1reg XMM reg + + for (t2reg = 0; ( (t2reg == EEREC_S) || (t2reg == EEREC_T) || (t2reg == EEREC_TEMP) || (t2reg == t1reg) ); t2reg++) + ; // Find unused reg (For second temp reg) + SSE_MOVAPS_XMM_to_M128((uptr)FTIO_Temp2, t2reg); // Backup t2reg XMM reg + + recVUMI_FTOI_Saturate(EEREC_S, EEREC_TEMP, t1reg, t2reg); // Saturate if Float->Int conversion returned illegal result + + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)FTIO_Temp1); // Restore t1reg XMM reg + SSE_MOVAPS_M128_to_XMM(t2reg, (uptr)FTIO_Temp2); // Restore t2reg XMM reg + } + + SSE_MOVAPS_XMM_to_XMM(EEREC_T, EEREC_TEMP); + } + } +} + +void recVUMI_FTOI4( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int4[0], info); } +void recVUMI_FTOI12( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int12[0], info); } +void recVUMI_FTOI15( VURegs *VU, int info ) { recVUMI_FTOIX(VU, (uptr)&recMult_float_to_int15[0], info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// ITOF 0/4/12/15 +//------------------------------------------------------------------ +void recVUMI_ITOF0( VURegs *VU, int info ) +{ + if ( _Ft_ == 0 ) return; + + //Console::WriteLn ("recVUMI_ITOF0"); + if (_X_Y_Z_W != 0xf) { + SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + vuFloat_useEAX( info, EEREC_TEMP, 15); // Clamp infinities + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + xmmregs[EEREC_T].mode |= MODE_WRITE; + } + else { + SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_T, EEREC_S); + vuFloat2(EEREC_T, EEREC_TEMP, 15); // Clamp infinities + } +} + +void recVUMI_ITOFX(VURegs *VU, int addr, int info) +{ + if ( _Ft_ == 0 ) return; + + //Console::WriteLn ("recVUMI_ITOFX"); + if (_X_Y_Z_W != 0xf) { + SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_TEMP, EEREC_S); + SSE_MULPS_M128_to_XMM(EEREC_TEMP, addr); + vuFloat_useEAX( info, EEREC_TEMP, 15); // Clamp infinities + VU_MERGE_REGS(EEREC_T, EEREC_TEMP); + xmmregs[EEREC_T].mode |= MODE_WRITE; + } + else { + SSE2_CVTDQ2PS_XMM_to_XMM(EEREC_T, EEREC_S); + SSE_MULPS_M128_to_XMM(EEREC_T, addr); + vuFloat2(EEREC_T, EEREC_TEMP, 15); // Clamp infinities + } +} + +void recVUMI_ITOF4( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float4[0], info); } +void recVUMI_ITOF12( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float12[0], info); } +void recVUMI_ITOF15( VURegs *VU, int info ) { recVUMI_ITOFX(VU, (uptr)&recMult_int_to_float15[0], info); } +//------------------------------------------------------------------ + + +//------------------------------------------------------------------ +// CLIP +//------------------------------------------------------------------ +void recVUMI_CLIP(VURegs *VU, int info) +{ + int t1reg = EEREC_D; + int t2reg = EEREC_ACC; + int x86temp1, x86temp2; + + u32 clipaddr = VU_VI_ADDR(REG_CLIP_FLAG, 0); + u32 prevclipaddr = VU_VI_ADDR(REG_CLIP_FLAG, 2); + + if( clipaddr == 0 ) { // battle star has a clip right before fcset + Console::WriteLn("skipping vu clip"); + return; + } + + //Flush the clip flag before processing, incase of double clip commands (GoW) + + if( prevclipaddr != (uptr)&VU->VI[REG_CLIP_FLAG] ) { + MOV32MtoR(EAX, prevclipaddr); + MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); + } + + assert( clipaddr != 0 ); + assert( t1reg != t2reg && t1reg != EEREC_TEMP && t2reg != EEREC_TEMP ); + + x86temp1 = ALLOCTEMPX86(MODE_8BITREG); + x86temp2 = ALLOCTEMPX86(MODE_8BITREG); + + //if ( (x86temp1 == 0) || (x86temp2 == 0) ) Console::Error("VU CLIP Allocation Error: EAX being allocated!"); + + _freeXMMreg(t1reg); // These should have been freed at allocation in eeVURecompileCode() + _freeXMMreg(t2reg); // but if they've been used since then, then free them. (just doing this incase :p (cottonvibes)) + + if( _Ft_ == 0 ) { + SSE_MOVAPS_M128_to_XMM(EEREC_TEMP, (uptr)&s_fones[0]); // all 1s + SSE_MOVAPS_M128_to_XMM(t1reg, (uptr)&s_fones[4]); + } + else { + _unpackVF_xyzw(EEREC_TEMP, EEREC_T, 3); + SSE_ANDPS_M128_to_XMM(EEREC_TEMP, (uptr)&const_clip[0]); + SSE_MOVAPS_XMM_to_XMM(t1reg, EEREC_TEMP); + SSE_ORPS_M128_to_XMM(t1reg, (uptr)&const_clip[4]); + } + + MOV32MtoR(EAX, prevclipaddr); + + SSE_CMPNLEPS_XMM_to_XMM(t1reg, EEREC_S); //-w, -z, -y, -x + SSE_CMPLTPS_XMM_to_XMM(EEREC_TEMP, EEREC_S); //+w, +z, +y, +x + + SHL32ItoR(EAX, 6); + + SSE_MOVAPS_XMM_to_XMM(t2reg, EEREC_TEMP); //t2 = +w, +z, +y, +x + SSE_UNPCKLPS_XMM_to_XMM(EEREC_TEMP, t1reg); //EEREC_TEMP = -y,+y,-x,+x + SSE_UNPCKHPS_XMM_to_XMM(t2reg, t1reg); //t2reg = -w,+w,-z,+z + SSE_MOVMSKPS_XMM_to_R32(x86temp2, EEREC_TEMP); // -y,+y,-x,+x + SSE_MOVMSKPS_XMM_to_R32(x86temp1, t2reg); // -w,+w,-z,+z + + AND8ItoR(x86temp1, 0x3); + SHL8ItoR(x86temp1, 4); + OR8RtoR(EAX, x86temp1); + AND8ItoR(x86temp2, 0xf); + OR8RtoR(EAX, x86temp2); + AND32ItoR(EAX, 0xffffff); + + MOV32RtoM(clipaddr, EAX); + + if (( !(info & (PROCESS_VU_SUPER|PROCESS_VU_COP2)) ) ) //Instantly update the flag if its called from elsewhere (unlikely, but ok) + MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); + + _freeX86reg(x86temp1); + _freeX86reg(x86temp2); +} diff --git a/pcsx2/x86/iVUzerorec.cpp b/pcsx2/x86/sVU_zerorec.cpp similarity index 96% rename from pcsx2/x86/iVUzerorec.cpp rename to pcsx2/x86/sVU_zerorec.cpp index 9b3942d71e..c02a7366fd 100644 --- a/pcsx2/x86/iVUzerorec.cpp +++ b/pcsx2/x86/sVU_zerorec.cpp @@ -1,4682 +1,4682 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA - */ - -// Super VU recompiler - author: zerofrog(@gmail.com) - -#include "PrecompiledHeader.h" - -#include -#include -#include -#include -#include - -#ifndef _WIN32 -#include -#endif - -#include "Common.h" - -#include "GS.h" -#include "R5900.h" -#include "VU.h" -#include "iR5900.h" - -#include "iVUzerorec.h" -#include "SamplProf.h" -#include "NakedAsm.h" - -using namespace std; - -// temporary externs -extern void iDumpVU0Registers(); -extern void iDumpVU1Registers(); - -// SuperVURec optimization options, uncomment only for debugging purposes -#define SUPERVU_CACHING // vu programs are saved and queried via memcompare (should be no reason to disable this) -#define SUPERVU_WRITEBACKS // don't flush the writebacks after every block -#define SUPERVU_X86CACHING // use x86reg caching (faster) (not really. rather lots slower :p (rama) ) -#define SUPERVU_VIBRANCHDELAY // when integers are modified right before a branch that uses the integer, - // the old integer value is used in the branch, fixes kh2 - -#define SUPERVU_PROPAGATEFLAGS // the correct behavior of VUs, for some reason superman breaks gfx with it on... - -// registers won't be flushed at block boundaries (faster) (nothing noticable speed-wise, causes SPS in Ratchet and clank (Nneeve) ) -#ifndef _DEBUG -//#define SUPERVU_INTERCACHING -#endif - -#define SUPERVU_CHECKCONDITION 0 // has to be 0!! - -#define VU_EXESIZE 0x00800000 - -#define _Imm11_ (s32)( (vucode & 0x400) ? (0xfffffc00 | (vucode & 0x3ff)) : (vucode & 0x3ff) ) -#define _UImm11_ (s32)(vucode & 0x7ff) - -#define _Ft_ ((VU->code >> 16) & 0x1F) // The rt part of the instruction register -#define _Fs_ ((VU->code >> 11) & 0x1F) // The rd part of the instruction register -#define _Fd_ ((VU->code >> 6) & 0x1F) // The sa part of the instruction register -#define _It_ (_Ft_ & 15) -#define _Is_ (_Fs_ & 15) -#define _Id_ (_Fd_ & 15) - -static const u32 QWaitTimes[] = { 6, 12 }; -static const u32 PWaitTimes[] = { 53, 43, 28, 23, 17, 11, 10 }; - -static u32 s_vuInfo; // info passed into rec insts - -static const u32 s_MemSize[2] = {VU0_MEMSIZE, VU1_MEMSIZE}; -static u8* s_recVUMem = NULL, *s_recVUPtr = NULL; - -// tables which are defined at the bottom of this massive file. -extern void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info); -extern void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info); - -#define INST_Q_READ 0x0001 // flush Q -#define INST_P_READ 0x0002 // flush P -#define INST_BRANCH_DELAY 0x0004 -#define INST_CLIP_WRITE 0x0040 // inst writes CLIP in the future -#define INST_STATUS_WRITE 0x0080 -#define INST_MAC_WRITE 0x0100 -#define INST_Q_WRITE 0x0200 -#define INST_CACHE_VI 0x0400 // write old vi value to s_VIBranchDelay - -// Let's tempt fate by defining two different constants with almost identical names -#define INST_DUMMY_ 0x8000 -#define INST_DUMMY 0x83c0 - -#define VFFREE_INVALID0 0x80000000 // (vffree[i]&0xf) is invalid - -//#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); ++(it)) - -#ifdef _DEBUG -u32 s_vucount = 0; - -static u32 g_vu1lastrec = 0, skipparent = -1; -static u32 s_svulast = 0, s_vufnheader; -static u32 badaddrs[][2] = {0, 0xffff}; -#endif - -union VURecRegs -{ - struct - { - u16 reg; - u16 type; - }; - u32 id; -}; - -#define SUPERVU_XGKICKDELAY 1 // yes this is needed as default (wipeout) - -class VuBaseBlock; - -struct VuFunctionHeader -{ - struct RANGE - { - RANGE() : pmem(NULL) {} - - u16 start, size; - void* pmem; // all the mem - }; - - VuFunctionHeader() : pprogfunc(NULL), startpc(0xffffffff) {} - ~VuFunctionHeader() - { - for (vector::iterator it = ranges.begin(); it != ranges.end(); ++it) - { - free(it->pmem); - } - } - - // returns true if the checksum for the current mem is the same as this fn - bool IsSame(void* pmem); - - u32 startpc; - void* pprogfunc; - - vector ranges; -}; - -struct VuBlockHeader -{ - VuBaseBlock* pblock; - u32 delay; -}; - -// one vu inst (lower and upper) -class VuInstruction -{ - public: - VuInstruction() - { - memzero_obj(*this); - nParentPc = -1; - vicached = -1; - } - - int nParentPc; // used for syncing with flag writes, -1 for no parent - - _vuopinfo info; - - _VURegsNum regs[2]; // [0] - lower, [1] - upper - u32 livevars[2]; // live variables right before this inst, [0] - inst, [1] - float - u32 addvars[2]; // live variables to add - u32 usedvars[2]; // set if var is used in the future including vars used in this inst - u32 keepvars[2]; - u16 pqcycles; // the number of cycles to stall if function writes to the regs - u16 type; // INST_ - - u32 pClipWrite, pMACWrite, pStatusWrite; // addrs to write the flags - u32 vffree[2]; - s8 vfwrite[2], vfread0[2], vfread1[2], vfacc[2]; - s8 vfflush[2]; // extra flush regs - s8 vicached; // if >= 0, then use the cached integer s_VIBranchDelay - VuInstruction *pPrevInst; - - int SetCachedRegs(int upper, u32 vuxyz); - void Recompile(list::iterator& itinst, u32 vuxyz); -}; - -enum BlockType -{ - BLOCKTYPE_EOP = 0x01, // at least one of the children of the block contains eop (or the block itself) - BLOCKTYPE_FUNCTION = 0x02, - BLOCKTYPE_HASEOP = 0x04, // last inst of block is an eop - BLOCKTYPE_MACFLAGS = 0x08, - BLOCKTYPE_ANALYZED = 0x40, - BLOCKTYPE_IGNORE = 0x80, // special for recursive fns - BLOCKTYPE_ANALYZEDPARENT = 0x100 -}; - -// base block used when recompiling -class VuBaseBlock -{ - public: - typedef list LISTBLOCKS; - - VuBaseBlock(); - - // returns true if the leads to a EOP (ALL VU blocks must ret true) - void AssignVFRegs(); - void AssignVIRegs(int parent); - - list::iterator GetInstIterAtPc(int instpc); - void GetInstsAtPc(int instpc, list& listinsts); - - void Recompile(); - - u16 type; // BLOCKTYPE_ - u16 id; - u16 startpc; - u16 endpc; // first inst not in block - void* pcode; // x86 code pointer - void* pendcode; // end of the x86 code pointer - int cycles; - list insts; - list parents; - LISTBLOCKS blocks; // blocks branches to - u32* pChildJumps[4]; // addrs that need to be filled with the children's start addrs - // if highest bit is set, addr needs to be relational - u32 vuxyz; // corresponding bit is set if reg's xyz channels are used only - u32 vuxy; // corresponding bit is set if reg's xyz channels are used only - - _xmmregs startregs[iREGCNT_XMM], endregs[iREGCNT_XMM]; - int nStartx86, nEndx86; // indices into s_vecRegArray - - int allocX86Regs; - int prevFlagsOutOfBlock; -}; - -struct WRITEBACK -{ - WRITEBACK() : nParentPc(0), cycle(0) //, pStatusWrite(NULL), pMACWrite(NULL) - { - viwrite[0] = viwrite[1] = 0; - viread[0] = viread[1] = 0; - } - - void InitInst(VuInstruction* pinst, int cycle) const - { - u32 write = viwrite[0] | viwrite[1]; - pinst->type = ((write & (1 << REG_CLIP_FLAG)) ? INST_CLIP_WRITE : 0) | - ((write & (1 << REG_MAC_FLAG)) ? INST_MAC_WRITE : 0) | - ((write & (1 << REG_STATUS_FLAG)) ? INST_STATUS_WRITE : 0) | - ((write & (1 << REG_Q)) ? INST_Q_WRITE : 0); - pinst->nParentPc = nParentPc; - pinst->info.cycle = cycle; - for (int i = 0; i < 2; ++i) - { - pinst->regs[i].VIwrite = viwrite[i]; - pinst->regs[i].VIread = viread[i]; - } - } - - static int SortWritebacks(const WRITEBACK& w1, const WRITEBACK& w2) - { - return w1.cycle < w2.cycle; - } - - int nParentPc; - int cycle; - u32 viwrite[2]; - u32 viread[2]; -}; - -struct VUPIPELINES -{ - fmacPipe fmac[8]; - fdivPipe fdiv; - efuPipe efu; - ialuPipe ialu[8]; - list< WRITEBACK > listWritebacks; -}; - -VuBaseBlock::VuBaseBlock() -{ - type = 0; - endpc = 0; - cycles = 0; - pcode = NULL; - id = 0; - memzero_obj(pChildJumps); - memzero_obj(startregs); - memzero_obj(endregs); - allocX86Regs = nStartx86 = nEndx86 = -1; - prevFlagsOutOfBlock = 0; -} - -#define SUPERVU_STACKSIZE 0x1000 - -static list s_listVUHeaders[2]; -static list* s_plistCachedHeaders[2] = {NULL, NULL}; -static VuFunctionHeader** recVUHeaders[2] = {NULL, NULL}; -static VuBlockHeader* recVUBlocks[2] = {NULL, NULL}; -static u8* recVUStack = NULL, *recVUStackPtr = NULL; -static vector<_x86regs> s_vecRegArray(128); - -static VURegs* VU = NULL; -static list s_listBlocks; -static u32 s_vu = 0; -static u32 s_UnconditionalDelay = 0; // 1 if there are two sequential branches and the last is unconditional -static u32 g_nLastBlockExecuted = 0; - -static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex); -static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const VUPIPELINES& pipes); -static void SuperVUInitLiveness(VuBaseBlock* pblock); -static void SuperVULivenessAnalysis(); -static void SuperVUEliminateDeadCode(); -static void SuperVUAssignRegs(); - -//void SuperVUFreeXMMreg(int xmmreg, int xmmtype, int reg); -#define SuperVUFreeXMMreg 0&& -void SuperVUFreeXMMregs(u32* livevars); - -static u32* SuperVUStaticAlloc(u32 size); -static void SuperVURecompile(); - -// allocate VU resources -void SuperVUAlloc(int vuindex) -{ - // The old -1 crap has been depreciated on this function. Please - // specify either 0 or 1, thanks. - jASSUME(vuindex >= 0); - - // upper 4 bits must be zero! - if (s_recVUMem == NULL) - { - // upper 4 bits must be zero! - // Changed "first try base" to 0xf1e0000, since 0x0c000000 liked to fail a lot. (cottonvibes) - s_recVUMem = SysMmapEx(0xf1e0000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); - - if (s_recVUMem == NULL) - { - throw Exception::OutOfMemory( - fmt_string("SuperVU Error > failed to allocate recompiler memory (addr: 0x%x)", (u32)s_recVUMem) - ); - } - - ProfilerRegisterSource("VURec", s_recVUMem, VU_EXESIZE); - - if (recVUStack == NULL) recVUStack = new u8[SUPERVU_STACKSIZE * 4]; - } - - if (vuindex >= 0) - { - jASSUME(s_recVUMem != NULL); - - if (recVUHeaders[vuindex] == NULL) - recVUHeaders[vuindex] = new VuFunctionHeader* [s_MemSize[vuindex] / 8]; - if (recVUBlocks[vuindex] == NULL) - recVUBlocks[vuindex] = new VuBlockHeader[s_MemSize[vuindex] / 8]; - if (s_plistCachedHeaders[vuindex] == NULL) - s_plistCachedHeaders[vuindex] = new list[s_MemSize[vuindex] / 8]; - } -} - -void DestroyCachedHeaders(int vuindex, int j) -{ - list::iterator it = s_plistCachedHeaders[vuindex][j].begin(); - - while (it != s_plistCachedHeaders[vuindex][j].end()) - { - delete *it; - it++; - } - - s_plistCachedHeaders[vuindex][j].clear(); -} - -void DestroyVUHeaders(int vuindex) -{ - list::iterator it = s_listVUHeaders[vuindex].begin(); - - while (it != s_listVUHeaders[vuindex].end()) - { - delete *it; - it++; - } - - s_listVUHeaders[vuindex].clear(); -} - -// destroy VU resources -void SuperVUDestroy(int vuindex) -{ - list::iterator it; - - if (vuindex < 0) - { - SuperVUDestroy(0); - SuperVUDestroy(1); - ProfilerTerminateSource("VURec"); - SafeSysMunmap(s_recVUMem, VU_EXESIZE); - safe_delete_array(recVUStack); - } - else - { - safe_delete_array(recVUHeaders[vuindex]); - safe_delete_array(recVUBlocks[vuindex]); - - if (s_plistCachedHeaders[vuindex] != NULL) - { - for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) - { - DestroyCachedHeaders(vuindex, j); - } - safe_delete_array(s_plistCachedHeaders[vuindex]); - } - DestroyVUHeaders(vuindex); - } -} - -// reset VU -void SuperVUReset(int vuindex) -{ -#ifdef _DEBUG - s_vucount = 0; -#endif - - if (s_recVUMem == NULL) - return; - - //jASSUME( s_recVUMem != NULL ); - - if (vuindex < 0) - { - DbgCon::Status("SuperVU reset > Resetting recompiler memory and structures."); - - // Does this cause problems on VU recompiler resets? It could, if the VU works like - // the EE used to, and actually tries to re-enter the recBlock after issuing a clear. (air) - - //memset_8<0xcd, VU_EXESIZE>(s_recVUMem); - memzero_ptr(recVUStack); - - s_recVUPtr = s_recVUMem; - } - else - { - DbgCon::Status("SuperVU reset [VU%d] > Resetting the recs and junk", params vuindex); - list::iterator it; - if (recVUHeaders[vuindex]) memset(recVUHeaders[vuindex], 0, sizeof(VuFunctionHeader*) * (s_MemSize[vuindex] / 8)); - if (recVUBlocks[vuindex]) memset(recVUBlocks[vuindex], 0, sizeof(VuBlockHeader) * (s_MemSize[vuindex] / 8)); - - if (s_plistCachedHeaders[vuindex] != NULL) - { - for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) - { - DestroyCachedHeaders(vuindex, j); - } - } - DestroyVUHeaders(vuindex); - } -} - -// clear the block and any joining blocks -void __fastcall SuperVUClear(u32 startpc, u32 size, int vuindex) -{ - vector::iterator itrange; - list::iterator it = s_listVUHeaders[vuindex].begin(); - u32 endpc = startpc + (size + (8 - (size & 7))); // Adding this code to ensure size is always a multiple of 8, it can be simplified to startpc+size if size is always a multiple of 8 (cottonvibes) - while (it != s_listVUHeaders[vuindex].end()) - { - - // for every fn, check if it has code in the range - for(itrange = (*it)->ranges.begin(); itrange != (*it)->ranges.end(); itrange++) - { - if (startpc < (u32)itrange->start + itrange->size && itrange->start < endpc) - break; - } - - if (itrange != (*it)->ranges.end()) - { - recVUHeaders[vuindex][(*it)->startpc/8] = NULL; -#ifdef SUPERVU_CACHING - list* plist = &s_plistCachedHeaders[vuindex][(*it)->startpc/8]; - plist->push_back(*it); - if (plist->size() > 30) - { - // list is too big, delete - //Console::Notice("Performance warning: deleting cached VU program!"); - delete plist->front(); - plist->pop_front(); - } - it = s_listVUHeaders[vuindex].erase(it); -#else - delete *it; - it = s_listVUHeaders[vuindex].erase(it); -#endif - } - else ++it; - } -} - -static VuFunctionHeader* s_pFnHeader = NULL; -static VuBaseBlock* s_pCurBlock = NULL; -static VuInstruction* s_pCurInst = NULL; -static u32 s_StatusRead = 0, s_MACRead = 0, s_ClipRead = 0; // read addrs -static u32 s_PrevStatusWrite = 0, s_PrevMACWrite = 0, s_PrevClipWrite = 0, s_PrevIWrite = 0; -static u32 s_WriteToReadQ = 0; - -static u32 s_VIBranchDelay = 0; //Value of register to use in a vi branch delayed situation - - -u32 s_TotalVUCycles; // total cycles since start of program execution - - -u32 SuperVUGetVIAddr(int reg, int read) -{ - assert(s_pCurInst != NULL); - - switch (reg) - { - case REG_STATUS_FLAG: - { - u32 addr = (read == 2) ? s_PrevStatusWrite : (read ? s_StatusRead : s_pCurInst->pStatusWrite); - assert(!read || addr != 0); - return addr; - } - case REG_MAC_FLAG: - { - u32 addr = (read == 2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); - return addr; - } - case REG_CLIP_FLAG: - { - u32 addr = (read == 2) ? s_PrevClipWrite : (read ? s_ClipRead : s_pCurInst->pClipWrite); - assert(!read || addr != 0); - return addr; - } - case REG_Q: - return (read || s_WriteToReadQ) ? (uptr)&VU->VI[REG_Q] : (uptr)&VU->q; - case REG_P: - return read ? (uptr)&VU->VI[REG_P] : (uptr)&VU->p; - case REG_I: - return s_PrevIWrite; - } - -#ifdef SUPERVU_VIBRANCHDELAY - if ((read != 0) && (s_pCurInst->regs[0].pipe == VUPIPE_BRANCH) && (s_pCurInst->vicached >= 0) && (s_pCurInst->vicached == reg)) - { - return (uptr)&s_VIBranchDelay; // test for branch delays - } -#endif - - return (uptr)&VU->VI[reg]; -} - -void SuperVUDumpBlock(list& blocks, int vuindex) -{ - FILE *f; - char str[256]; - u32 *mem; - u32 i; - - Path::CreateDirectory("dumps"); - string filename(Path::Combine("dumps", fmt_string("svu%cdump%.4X.txt", s_vu ? '0' : '1', s_pFnHeader->startpc))); - //Console::WriteLn( "dump1 %x => %s", params s_pFnHeader->startpc, filename ); - - f = fopen(filename.c_str(), "w"); - - fprintf(f, "Format: upper_inst lower_inst\ntype f:vf_live_vars vf_used_vars i:vi_live_vars vi_used_vars inst_cycle pq_inst\n"); - fprintf(f, "Type: %.2x - qread, %.2x - pread, %.2x - clip_write, %.2x - status_write\n" - "%.2x - mac_write, %.2x -qflush\n", - INST_Q_READ, INST_P_READ, INST_CLIP_WRITE, INST_STATUS_WRITE, INST_MAC_WRITE, INST_Q_WRITE); - fprintf(f, "XMM: Upper: read0 read1 write acc temp; Lower: read0 read1 write acc temp\n\n"); - - list::iterator itblock; - list::iterator itinst; - VuBaseBlock::LISTBLOCKS::iterator itchild; - - for(itblock = blocks.begin(); itblock != blocks.end(); itblock++) - { - fprintf(f, "block:%c %x-%x; children: ", ((*itblock)->type&BLOCKTYPE_HASEOP) ? '*' : ' ', - (*itblock)->startpc, (*itblock)->endpc - 8); - - for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) - { - fprintf(f, "%x ", (*itchild)->startpc); - } - fprintf(f, "; vuxyz = %x, vuxy = %x\n", (*itblock)->vuxyz&(*itblock)->insts.front().usedvars[1], - (*itblock)->vuxy&(*itblock)->insts.front().usedvars[1]); - - itinst = (*itblock)->insts.begin(); - i = (*itblock)->startpc; - while (itinst != (*itblock)->insts.end()) - { - assert(i <= (*itblock)->endpc); - if (itinst->type & INST_DUMMY) - { - if (itinst->nParentPc >= 0 && !(itinst->type&INST_DUMMY_)) - { - // search for the parent - fprintf(f, "writeback 0x%x (%x)\n", itinst->type, itinst->nParentPc); - } - } - else - { - mem = (u32*) & VU->Micro[i]; - char* pstr = disVU1MicroUF(mem[1], i + 4); - fprintf(f, "%.4x: %-40s", i, pstr); - if (mem[1] & 0x80000000) fprintf(f, " I=%f(%.8x)\n", *(float*)mem, mem[0]); - else fprintf(f, "%s\n", disVU1MicroLF(mem[0], i)); - i += 8; - } - - ++itinst; - } - - fprintf(f, "\n"); - - _x86regs* pregs; - if ((*itblock)->nStartx86 >= 0 || (*itblock)->nEndx86 >= 0) - { - fprintf(f, "X86: AX CX DX BX SP BP SI DI\n"); - } - - if ((*itblock)->nStartx86 >= 0) - { - pregs = &s_vecRegArray[(*itblock)->nStartx86]; - fprintf(f, "STR: "); - for (i = 0; i < iREGCNT_GPR; ++i) - { - if (pregs[i].inuse) - fprintf(f, "%.2d ", pregs[i].reg); - else - fprintf(f, "-1 "); - } - fprintf(f, "\n"); - } - - if ((*itblock)->nEndx86 >= 0) - { - fprintf(f, "END: "); - pregs = &s_vecRegArray[(*itblock)->nEndx86]; - for (i = 0; i < iREGCNT_GPR; ++i) - { - if (pregs[i].inuse) - fprintf(f, "%.2d ", pregs[i].reg); - else - fprintf(f, "-1 "); - } - fprintf(f, "\n"); - } - - itinst = (*itblock)->insts.begin(); - for (i = (*itblock)->startpc; i < (*itblock)->endpc; ++itinst) - { - - if (itinst->type & INST_DUMMY) - { - } - else - { - sprintf(str, "%.4x:%x f:%.8x_%.8x", i, itinst->type, itinst->livevars[1], itinst->usedvars[1]); - fprintf(f, "%-46s i:%.8x_%.8x c:%d pq:%d\n", str, - itinst->livevars[0], itinst->usedvars[0], (int)itinst->info.cycle, (int)itinst->pqcycles); - - sprintf(str, "XMM r0:%d r1:%d w:%d a:%d t:%x;", - itinst->vfread0[1], itinst->vfread1[1], itinst->vfwrite[1], itinst->vfacc[1], itinst->vffree[1]); - fprintf(f, "%-46s r0:%d r1:%d w:%d a:%d t:%x\n", str, - itinst->vfread0[0], itinst->vfread1[0], itinst->vfwrite[0], itinst->vfacc[0], itinst->vffree[0]); - i += 8; - } - } - -#ifdef __LINUX__ - // dump the asm - if ((*itblock)->pcode != NULL) - { - char command[255]; - FILE* fasm = fopen("mydump1", "wb"); - //Console::WriteLn("writing: %x, %x", params (*itblock)->startpc, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode); - fwrite((*itblock)->pcode, 1, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode, fasm); - fclose(fasm); - sprintf(command, "objdump -D --target=binary --architecture=i386 -M intel mydump1 > tempdump"); - system(command); - fasm = fopen("tempdump", "r"); - // read all of it and write it to f - fseek(fasm, 0, SEEK_END); - vector vbuffer(ftell(fasm)); - fseek(fasm, 0, SEEK_SET); - fread(&vbuffer[0], vbuffer.size(), 1, fasm); - - fprintf(f, "\n\n"); - fwrite(&vbuffer[0], vbuffer.size(), 1, f); - fclose(fasm); - } -#endif - - fprintf(f, "\n---------------\n"); - } - - fclose(f); -} - -// uncomment to count svu exec time -//#define SUPERVU_COUNT - -// Private methods -void* SuperVUGetProgram(u32 startpc, int vuindex) -{ - assert(startpc < s_MemSize[vuindex]); - assert((startpc % 8) == 0); - assert(recVUHeaders[vuindex] != NULL); - VuFunctionHeader** pheader = &recVUHeaders[vuindex][startpc/8]; - - if (*pheader == NULL) - { -#ifdef _DEBUG -// if( vuindex ) VU1.VI[REG_TPC].UL = startpc; -// else VU0.VI[REG_TPC].UL = startpc; -// __Log("VU: %x\n", startpc); -// iDumpVU1Registers(); -// vudump |= 2; -#endif - - // measure run time - //QueryPerformanceCounter(&svubase); - -#ifdef SUPERVU_CACHING - void* pmem = (vuindex & 1) ? VU1.Micro : VU0.Micro; - // check if program exists in cache - list::iterator it; - for(it = s_plistCachedHeaders[vuindex][startpc/8].begin(); it != s_plistCachedHeaders[vuindex][startpc/8].end(); it++) - { - if ((*it)->IsSame(pmem)) - { - // found, transfer to regular lists - void* pfn = (*it)->pprogfunc; - recVUHeaders[vuindex][startpc/8] = *it; - s_listVUHeaders[vuindex].push_back(*it); - s_plistCachedHeaders[vuindex][startpc/8].erase(it); - return pfn; - } - } -#endif - - *pheader = SuperVURecompileProgram(startpc, vuindex); - - if (*pheader == NULL) - { - assert(s_TotalVUCycles > 0); - if (vuindex) - VU1.VI[REG_TPC].UL = startpc; - else - VU0.VI[REG_TPC].UL = startpc; - - return (void*)SuperVUEndProgram; - } - - //QueryPerformanceCounter(&svufinal); - //svutime += (u32)(svufinal.QuadPart-svubase.QuadPart); - - assert((*pheader)->pprogfunc != NULL); - } - //else assert( (*pheader)->IsSame((vuindex&1) ? VU1.Micro : VU0.Micro) ); - - assert((*pheader)->startpc == startpc); - - return (*pheader)->pprogfunc; -} - -bool VuFunctionHeader::IsSame(void* pmem) -{ -#ifdef SUPERVU_CACHING - vector::iterator it; - for(it = ranges.begin(); it != ranges.end(); it++) - { - if (memcmp_mmx((u8*)pmem + it->start, it->pmem, it->size)) - return false; - } -#endif - return true; -} - -list::iterator VuBaseBlock::GetInstIterAtPc(int instpc) -{ - assert(instpc >= 0); - - u32 curpc = startpc; - list::iterator it; - for (it = insts.begin(); it != insts.end(); ++it) - { - if (it->type & INST_DUMMY) continue; - if (curpc == instpc) break; - curpc += 8; - } - - if (it != insts.end()) return it; - - assert(0); - return insts.begin(); -} - -void VuBaseBlock::GetInstsAtPc(int instpc, list& listinsts) -{ - assert(instpc >= 0); - - listinsts.clear(); - - u32 curpc = startpc; - list::iterator it; - for (it = insts.begin(); it != insts.end(); ++it) - { - if (it->type & INST_DUMMY) continue; - if (curpc == instpc) break; - curpc += 8; - } - - if (it != insts.end()) - { - listinsts.push_back(&(*it)); - return; - } - - // look for the pc in other blocks - for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) - { - if (*itblock == this) continue; - - if (instpc >= (*itblock)->startpc && instpc < (*itblock)->endpc) - { - listinsts.push_back(&(*(*itblock)->GetInstIterAtPc(instpc))); - } - } - - assert(listinsts.size() > 0); -} - -static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) -{ - assert(vuindex < 2); - assert(s_recVUPtr != NULL); - //Console::WriteLn("svu%c rec: %x", params '0'+vuindex, startpc); - - // if recPtr reached the mem limit reset whole mem - if (((uptr)s_recVUPtr - (uptr)s_recVUMem) >= VU_EXESIZE - 0x40000) - { - //Console::WriteLn("SuperVU reset mem"); - SuperVUReset(0); - SuperVUReset(1); - SuperVUReset(-1); - if (s_TotalVUCycles > 0) - { - // already executing, so return NULL - return NULL; - } - } - - list::iterator itblock; - - s_vu = vuindex; - VU = s_vu ? &VU1 : &VU0; - s_pFnHeader = new VuFunctionHeader(); - s_listVUHeaders[vuindex].push_back(s_pFnHeader); - s_pFnHeader->startpc = startpc; - - memset(recVUBlocks[s_vu], 0, sizeof(VuBlockHeader) * (s_MemSize[s_vu] / 8)); - - // analyze the global graph - s_listBlocks.clear(); - VUPIPELINES pipes; - memzero_obj(pipes.fmac); - memzero_obj(pipes.fdiv); - memzero_obj(pipes.efu); - memzero_obj(pipes.ialu); - SuperVUBuildBlocks(NULL, startpc, pipes); - - // fill parents - VuBaseBlock::LISTBLOCKS::iterator itchild; - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) - { - (*itchild)->parents.push_back(*itblock); - } - - //(*itblock)->type &= ~(BLOCKTYPE_IGNORE|BLOCKTYPE_ANALYZED); - } - - assert(s_listBlocks.front()->startpc == startpc); - s_listBlocks.front()->type |= BLOCKTYPE_FUNCTION; - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - SuperVUInitLiveness(*itblock); - } - - SuperVULivenessAnalysis(); - SuperVUEliminateDeadCode(); - SuperVUAssignRegs(); - -#ifdef _DEBUG - if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); -#endif - - // code generation - x86SetPtr(s_recVUPtr); - branch = 0; - - SuperVURecompile(); - - s_recVUPtr = x86Ptr; - - // set the function's range - VuFunctionHeader::RANGE r; - s_pFnHeader->ranges.reserve(s_listBlocks.size()); - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - r.start = (*itblock)->startpc; - r.size = (*itblock)->endpc - (*itblock)->startpc; -#ifdef SUPERVU_CACHING - //memxor_mmx(r.checksum, &VU->Micro[r.start], r.size); - r.pmem = malloc(r.size); - memcpy_fast(r.pmem, &VU->Micro[r.start], r.size); -#endif - s_pFnHeader->ranges.push_back(r); - } - -#if defined(_DEBUG) && defined(__LINUX__) - // dump at the end to capture the actual code - if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); -#endif - - // destroy - for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) - { - delete *itblock; - } - s_listBlocks.clear(); - - assert(s_recVUPtr < s_recVUMem + VU_EXESIZE); - - return s_pFnHeader; -} - -static int _recbranchAddr(u32 vucode) -{ - s32 bpc = pc + (_Imm11_ << 3); - /* - if ( bpc < 0 ) { - Console::WriteLn("zerorec branch warning: bpc < 0 ( %x ); Using unsigned imm11", params bpc); - bpc = pc + (_UImm11_ << 3); - }*/ - bpc &= (s_MemSize[s_vu] - 1); - - return bpc; -} - -// return inst that flushes everything -static VuInstruction SuperVUFlushInst() -{ - VuInstruction inst; - // don't need to raed q/p - inst.type = INST_DUMMY_;//|INST_Q_READ|INST_P_READ; - return inst; -} - -void SuperVUAddWritebacks(VuBaseBlock* pblock, const list& listWritebacks) -{ -#ifdef SUPERVU_WRITEBACKS - // regardless of repetition, add the pipes (for selfloops) - list::const_iterator itwriteback = listWritebacks.begin(); - list::iterator itinst = pblock->insts.begin(), itinst2; - - while (itwriteback != listWritebacks.end()) - { - if (itinst != pblock->insts.end() && (itinst->info.cycle < itwriteback->cycle || (itinst->type&INST_DUMMY))) - { - ++itinst; - continue; - } - - itinst2 = pblock->insts.insert(itinst, VuInstruction()); - itwriteback->InitInst(&(*itinst2), vucycle); - ++itwriteback; - } -#endif -} - -#ifdef SUPERVU_VIBRANCHDELAY -static VuInstruction* getDelayInst(VuInstruction* pInst) -{ - // check for the N cycle branch delay - // example of 2 cycles delay (monster house) : - // sqi vi05 - // sqi vi05 - // ibeq vi05, vi03 - // The ibeq should read the vi05 before the first sqi - - //more info: - - // iaddiu vi01, 0, 1 - // ibeq vi01, 0 <- reads vi01 before the iaddiu - - // iaddiu vi01, 0, 1 - // iaddiu vi01, vi01, 1 - // iaddiu vi01, vi01, 1 - // ibeq vi01, 0 <- reads vi01 before the last two iaddiu's (so the value read is 1) - - // ilw vi02, addr - // iaddiu vi01, 0, 1 - // ibeq vi01, vi02 <- reads current values of both vi01 and vi02 because the branch instruction stalls - - int delay = 1; - VuInstruction* pDelayInst = NULL; - VuInstruction* pTargetInst = pInst->pPrevInst; - while (1) - { - if (pTargetInst != NULL - && pTargetInst->info.cycle + delay == pInst->info.cycle - && (pTargetInst->regs[0].pipe == VUPIPE_IALU || pTargetInst->regs[0].pipe == VUPIPE_FMAC) - && ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) - && (delay == 1 || ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) == ((pTargetInst->regs[0].VIwrite & pInst->pPrevInst->regs[0].VIread) & 0xffff)) - && !(pTargetInst->regs[0].VIread&((1 << REG_STATUS_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_CLIP_FLAG)))) - { - pDelayInst = pTargetInst; - pTargetInst = pTargetInst->pPrevInst; - delay++; - if (delay == 5) //maximum delay is 4 (length of the pipeline) - { - DevCon::WriteLn("supervu: cycle branch delay maximum (4) is reached"); - break; - } - } - else break; - } - if (delay > 1) DevCon::WriteLn("supervu: %d cycle branch delay detected: %x %x", params delay - 1, pc, s_pFnHeader->startpc); - return pDelayInst; -} -#endif - -static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const VUPIPELINES& pipes) -{ - // check if block already exists - //Console::WriteLn("startpc %x", params startpc); - startpc &= (s_vu ? 0x3fff : 0xfff); - VuBlockHeader* pbh = &recVUBlocks[s_vu][startpc/8]; - - if (pbh->pblock != NULL) - { - - VuBaseBlock* pblock = pbh->pblock; - list::iterator itinst; - - if (pblock->startpc == startpc) - { - SuperVUAddWritebacks(pblock, pipes.listWritebacks); - return pblock; - } - - // have to divide the blocks, pnewblock is first block - assert(startpc > pblock->startpc); - assert(startpc < pblock->endpc); - - u32 dummyinst = (startpc - pblock->startpc) >> 3; - - // count inst non-dummy insts - itinst = pblock->insts.begin(); - int cycleoff = 0; - - while (dummyinst > 0) - { - if (itinst->type & INST_DUMMY) - ++itinst; - else - { - cycleoff = itinst->info.cycle; - ++itinst; - --dummyinst; - } - } - - // NOTE: still leaves insts with their writebacks in different blocks - while (itinst->type & INST_DUMMY) - ++itinst; - - // the difference in cycles between dummy insts (naruto utlimate ninja) - int cyclediff = 0; - if (parent == pblock) - cyclediff = itinst->info.cycle - cycleoff; - cycleoff = itinst->info.cycle; - - // new block - VuBaseBlock* pnewblock = new VuBaseBlock(); - s_listBlocks.push_back(pnewblock); - - pnewblock->startpc = startpc; - pnewblock->endpc = pblock->endpc; - pnewblock->cycles = pblock->cycles - cycleoff + cyclediff; - - pnewblock->blocks.splice(pnewblock->blocks.end(), pblock->blocks); - pnewblock->insts.splice(pnewblock->insts.end(), pblock->insts, itinst, pblock->insts.end()); - pnewblock->type = pblock->type; - - // any writebacks in the next 3 cycles also belong to original block -// for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); ) { -// if( (itinst->type & INST_DUMMY) && itinst->nParentPc >= 0 && itinst->nParentPc < (int)startpc ) { -// -// if( !(itinst->type & INST_Q_WRITE) ) -// pblock->insts.push_back(*itinst); -// itinst = pnewblock->insts.erase(itinst); -// continue; -// } -// -// ++itinst; -// } - - pbh = &recVUBlocks[s_vu][startpc/8]; - for (u32 inst = startpc; inst < pblock->endpc; inst += 8) - { - if (pbh->pblock == pblock) - pbh->pblock = pnewblock; - ++pbh; - } - - for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); itinst++) - { - itinst->info.cycle -= cycleoff; - } - - SuperVUAddWritebacks(pnewblock, pipes.listWritebacks); - - // old block - pblock->blocks.push_back(pnewblock); - pblock->endpc = startpc; - pblock->cycles = cycleoff; - pblock->type &= BLOCKTYPE_MACFLAGS; - //pblock->insts.push_back(SuperVUFlushInst()); //don't need - - return pnewblock; - } - - VuBaseBlock* pblock = new VuBaseBlock(); - s_listBlocks.push_back(pblock); - - int i = 0; - branch = 0; - pc = startpc; - pblock->startpc = startpc; - - // clear stalls (might be a prob) - memcpy(VU->fmac, pipes.fmac, sizeof(pipes.fmac)); - memcpy(&VU->fdiv, &pipes.fdiv, sizeof(pipes.fdiv)); - memcpy(&VU->efu, &pipes.efu, sizeof(pipes.efu)); - memcpy(VU->ialu, pipes.ialu, sizeof(pipes.ialu)); -// memset(VU->fmac, 0, sizeof(VU->fmac)); -// memset(&VU->fdiv, 0, sizeof(VU->fdiv)); -// memset(&VU->efu, 0, sizeof(VU->efu)); - - vucycle = 0; - - u8 macflags = 0; - - list< WRITEBACK > listWritebacks; - list< WRITEBACK >::iterator itwriteback; - list::iterator itinst; - u32 hasSecondBranch = 0; - u32 needFullStatusFlag = 0; - -#ifdef SUPERVU_WRITEBACKS - listWritebacks = pipes.listWritebacks; -#endif - - // first analysis pass for status flags - while (1) - { - u32* ptr = (u32*) & VU->Micro[pc]; - pc += 8; - int prevbranch = branch; - - if (ptr[1] & 0x40000000) - branch = 1; - - if (!(ptr[1] & 0x80000000)) // not I - { - switch (ptr[0] >> 25) - { - case 0x24: // jr - case 0x25: // jalr - case 0x20: // B - case 0x21: // BAL - case 0x28: // IBEQ - case 0x2f: // IBGEZ - case 0x2d: // IBGTZ - case 0x2e: // IBLEZ - case 0x2c: // IBLTZ - case 0x29: // IBNE - branch = 1; - break; - - case 0x14: // fseq - case 0x17: // fsor - //needFullStatusFlag = 2; - break; - - case 0x16: // fsand - if ((ptr[0]&0xc0)) - { - // sometimes full sticky bits are needed (simple series 2000 - oane chapara) - //Console::WriteLn("needSticky: %x-%x", params s_pFnHeader->startpc, startpc); - needFullStatusFlag = 2; - } - break; - } - } - - if (prevbranch) - break; - - if (pc >= s_MemSize[s_vu]) - { - Console::Error("inf vu0 prog %x", params startpc); - break; - } - } - - // second full pass - pc = startpc; - branch = 0; - VuInstruction* pprevinst = NULL, *pinst = NULL; - - while (1) - { - - if (pc == s_MemSize[s_vu]) - { - branch |= 8; - break; - } - - if (!branch && pbh->pblock != NULL) - { - pblock->blocks.push_back(pbh->pblock); - break; - } - - int prevbranch = branch; - - if (!prevbranch) - { - pbh->pblock = pblock; - } - else assert(prevbranch || pbh->pblock == NULL); - - pblock->insts.push_back(VuInstruction()); - - pprevinst = pinst; - pinst = &pblock->insts.back(); - pinst->pPrevInst = pprevinst; - SuperVUAnalyzeOp(VU, &pinst->info, pinst->regs); - -#ifdef SUPERVU_VIBRANCHDELAY - if (pinst->regs[0].pipe == VUPIPE_BRANCH && pblock->insts.size() > 1) - { - - VuInstruction* pdelayinst = getDelayInst(pinst); - if (pdelayinst) - { - pdelayinst->type |= INST_CACHE_VI; - - // find the correct register - u32 mask = pdelayinst->regs[0].VIwrite & pinst->regs[0].VIread; - for (int i = 0; i < 16; ++i) - { - if (mask & (1 << i)) - { - pdelayinst->vicached = i; - break; - } - } - - pinst->vicached = pdelayinst->vicached; - } - } -#endif - - if (prevbranch) - { - if (pinst->regs[0].pipe == VUPIPE_BRANCH) - hasSecondBranch = 1; - pinst->type |= INST_BRANCH_DELAY; - } - - // check write back - for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end();) - { - if (pinst->info.cycle >= itwriteback->cycle) - { - itinst = pblock->insts.insert(--pblock->insts.end(), VuInstruction()); - itwriteback->InitInst(&(*itinst), pinst->info.cycle); - itwriteback = listWritebacks.erase(itwriteback); - } - else ++itwriteback; - } - - // add new writebacks - WRITEBACK w; - const u32 allflags = (1 << REG_CLIP_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG); - for (int j = 0; j < 2; ++j) w.viwrite[j] = pinst->regs[j].VIwrite & allflags; - - if (pinst->info.macflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_MAC_FLAG); - if (pinst->info.statusflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_STATUS_FLAG); - - if ((pinst->info.macflag | pinst->info.statusflag) & VUOP_READ) - macflags = 1; - if (pinst->regs[0].VIread & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) - macflags = 1; - -// if( pinst->regs[1].pipe == VUPIPE_FMAC && (pinst->regs[1].VFwrite==0&&!(pinst->regs[1].VIwrite&(1<regs[0].VIread |= (1<VIwrite |= lregs->VIwrite & (1<info.statusflag&VUOP_WRITE) && !(pinst->regs[0].VIwrite&(1 << REG_STATUS_FLAG))) && needFullStatusFlag) - { - // don't read if first inst - if (needFullStatusFlag == 1) - w.viread[1] |= (1 << REG_STATUS_FLAG); - else --needFullStatusFlag; - } - - for (int j = 0; j < 2; ++j) - { - w.viread[j] |= pinst->regs[j].VIread & allflags; - - if ((pinst->regs[j].VIread&(1 << REG_STATUS_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_STATUS_FLAG))) - { - // don't need the read anymore - pinst->regs[j].VIread &= ~(1 << REG_STATUS_FLAG); - } - if ((pinst->regs[j].VIread&(1 << REG_MAC_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_MAC_FLAG))) - { - // don't need the read anymore - pinst->regs[j].VIread &= ~(1 << REG_MAC_FLAG); - } - - pinst->regs[j].VIwrite &= ~allflags; - } - - if (pinst->info.macflag & VUOP_READ) w.viread[1] |= 1 << REG_MAC_FLAG; - if (pinst->info.statusflag & VUOP_READ) w.viread[1] |= 1 << REG_STATUS_FLAG; - - w.nParentPc = pc - 8; - w.cycle = pinst->info.cycle + 4; - listWritebacks.push_back(w); - } - - if (pinst->info.q&VUOP_READ) pinst->type |= INST_Q_READ; - if (pinst->info.p&VUOP_READ) pinst->type |= INST_P_READ; - - if (pinst->info.q&VUOP_WRITE) - { - pinst->pqcycles = QWaitTimes[pinst->info.pqinst] + 1; - - memset(&w, 0, sizeof(w)); - w.nParentPc = pc - 8; - w.cycle = pinst->info.cycle + pinst->pqcycles; - w.viwrite[0] = 1 << REG_Q; - listWritebacks.push_back(w); - } - if (pinst->info.p&VUOP_WRITE) - pinst->pqcycles = PWaitTimes[pinst->info.pqinst] + 1; - - if (prevbranch) - { - break; - } - - // make sure there is always a branch - // sensible soccer overflows on vu0, so increase the limit... - if ((s_vu == 1 && i >= 0x799) || (s_vu == 0 && i >= 0x201)) - { - Console::Error("VuRec base block doesn't terminate!"); - assert(0); - break; - } - - i++; - pbh++; - } - - if (macflags) - pblock->type |= BLOCKTYPE_MACFLAGS; - - pblock->endpc = pc; - u32 lastpc = pc; - - pblock->cycles = vucycle; - -#ifdef SUPERVU_WRITEBACKS - if (!branch || (branch&8)) -#endif - { - // flush writebacks - if (listWritebacks.size() > 0) - { - listWritebacks.sort(WRITEBACK::SortWritebacks); - for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) - { - if (itwriteback->viwrite[0] & (1 << REG_Q)) - { - // ignore all Q writebacks - continue; - } - - pblock->insts.push_back(VuInstruction()); - itwriteback->InitInst(&pblock->insts.back(), vucycle); - } - - listWritebacks.clear(); - } - } - - if (!branch) return pblock; - - if (branch & 8) - { - // what if also a jump? - pblock->type |= BLOCKTYPE_EOP | BLOCKTYPE_HASEOP; - - // add an instruction to flush p and q (if written) - pblock->insts.push_back(SuperVUFlushInst()); - return pblock; - } - - // it is a (cond) branch or a jump - u32 vucode = *(u32*)(VU->Micro + lastpc - 16); - int bpc = _recbranchAddr(vucode) - 8; - - VUPIPELINES newpipes; - memcpy(newpipes.fmac, VU->fmac, sizeof(newpipes.fmac)); - memcpy(&newpipes.fdiv, &VU->fdiv, sizeof(newpipes.fdiv)); - memcpy(&newpipes.efu, &VU->efu, sizeof(newpipes.efu)); - memcpy(newpipes.ialu, VU->ialu, sizeof(newpipes.ialu)); - - for (i = 0; i < 8; ++i) newpipes.fmac[i].sCycle -= vucycle; - newpipes.fdiv.sCycle -= vucycle; - newpipes.efu.sCycle -= vucycle; - for (i = 0; i < 8; ++i) newpipes.ialu[i].sCycle -= vucycle; - - if (listWritebacks.size() > 0) - { - // flush all when jumping, send down the pipe when in branching - bool bFlushWritebacks = (vucode >> 25) == 0x24 || (vucode >> 25) == 0x25;//||(vucode>>25)==0x20||(vucode>>25)==0x21; - - listWritebacks.sort(WRITEBACK::SortWritebacks); - for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) - { - if (itwriteback->viwrite[0] & (1 << REG_Q)) - { - // ignore all Q writebacks - continue; - } - - if (itwriteback->cycle < vucycle || bFlushWritebacks) - { - pblock->insts.push_back(VuInstruction()); - itwriteback->InitInst(&pblock->insts.back(), vucycle); - } - else - { - newpipes.listWritebacks.push_back(*itwriteback); - newpipes.listWritebacks.back().cycle -= vucycle; - } - } - } - - if (newpipes.listWritebacks.size() > 0) // other blocks might read the mac flags - pblock->type |= BLOCKTYPE_MACFLAGS; - - u32 firstbranch = vucode >> 25; - switch (firstbranch) - { - case 0x24: // jr - pblock->type |= BLOCKTYPE_EOP; // jump out of procedure, since not returning, set EOP - pblock->insts.push_back(SuperVUFlushInst()); - firstbranch = 0xff; //Non-Conditional Jump - break; - - case 0x25: // jalr - { - // linking, so will return to procedure - pblock->insts.push_back(SuperVUFlushInst()); - - VuBaseBlock* pjumpblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert(pblock != NULL); - - pblock->blocks.push_back(pjumpblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } - case 0x20: // B - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert(pblock != NULL); - - pblock->blocks.push_back(pbranchblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } - case 0x21: // BAL - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert(pblock != NULL); - pblock->blocks.push_back(pbranchblock); - firstbranch = 0xff; //Non-Conditional Jump - break; - } - case 0x28: // IBEQ - case 0x2f: // IBGEZ - case 0x2d: // IBGTZ - case 0x2e: // IBLEZ - case 0x2c: // IBLTZ - case 0x29: // IBNE - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - assert(pblock != NULL); - pblock->blocks.push_back(pbranchblock); - - // if has a second branch that is B or BAL, skip this - u32 secondbranch = (*(u32*)(VU->Micro + lastpc - 8)) >> 25; - if (!hasSecondBranch || (secondbranch != 0x21 && secondbranch != 0x20)) - { - pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); - - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - } - - break; - } - default: - assert(pblock->blocks.size() == 1); - break; - } - - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - -#ifdef SUPERVU_VIBRANCHDELAY -///////////////////////////////////////////////////////////////////////////////////////////////////////////////////// -///// NOTE! This could still be a hack for KH2/GoW, but until we know how it properly works, this will do for now./// -///////////////////////////////////////////////////////////////////////////////////////////////////////////////////// - - if (hasSecondBranch && firstbranch != 0xff) //check the previous jump was conditional and there is a second branch - { -#else - if (hasSecondBranch) - { -#endif - - u32 vucode = *(u32*)(VU->Micro + lastpc - 8); - pc = lastpc; - int bpc = _recbranchAddr(vucode); - - switch (vucode >> 25) - { - case 0x24: // jr - Console::Error("svurec bad jr jump!"); - assert(0); - break; - - case 0x25: // jalr - { - Console::Error("svurec bad jalr jump!"); - assert(0); - break; - } - case 0x20: // B - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - - pblock->blocks.push_back(pbranchblock); - break; - } - case 0x21: // BAL - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // replace instead of pushing a new block - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - break; - } - case 0x28: // IBEQ - case 0x2f: // IBGEZ - case 0x2d: // IBGTZ - case 0x2e: // IBLEZ - case 0x2c: // IBLTZ - case 0x29: // IBNE - { - VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); - - // update pblock since could have changed - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - - // only add the block if the previous branch doesn't include the next instruction (ie, if a direct jump) - if (firstbranch == 0x24 || firstbranch == 0x25 || firstbranch == 0x20 || firstbranch == 0x21) - { - pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); - - pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; - pblock->blocks.push_back(pbranchblock); - } - - break; - } - - jNO_DEFAULT; - } - } - - return recVUBlocks[s_vu][startpc/8].pblock; -} - -static void SuperVUInitLiveness(VuBaseBlock* pblock) -{ - list::iterator itinst, itnext; - - assert(pblock->insts.size() > 0); - - for (itinst = pblock->insts.begin(); itinst != pblock->insts.end(); ++itinst) - { - - if (itinst->type & INST_DUMMY_) - { - itinst->addvars[0] = itinst->addvars[1] = 0xffffffff; - itinst->livevars[0] = itinst->livevars[1] = 0xffffffff; - itinst->keepvars[0] = itinst->keepvars[1] = 0xffffffff; - itinst->usedvars[0] = itinst->usedvars[1] = 0; - } - else - { - itinst->addvars[0] = itinst->regs[0].VIread | itinst->regs[1].VIread; - itinst->addvars[1] = (itinst->regs[0].VFread0 ? (1 << itinst->regs[0].VFread0) : 0) | - (itinst->regs[0].VFread1 ? (1 << itinst->regs[0].VFread1) : 0) | - (itinst->regs[1].VFread0 ? (1 << itinst->regs[1].VFread0) : 0) | - (itinst->regs[1].VFread1 ? (1 << itinst->regs[1].VFread1) : 0); - - // vf0 is not handled by VFread - if (!itinst->regs[0].VFread0 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; - if (!itinst->regs[1].VFread0 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; - if (!itinst->regs[0].VFread1 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[0].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; - if (!itinst->regs[1].VFread1 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[1].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; - - - u32 vfwrite = 0; - if (itinst->regs[0].VFwrite != 0) - { - if (itinst->regs[0].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[0].VFwrite; - else vfwrite |= 1 << itinst->regs[0].VFwrite; - } - if (itinst->regs[1].VFwrite != 0) - { - if (itinst->regs[1].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[1].VFwrite; - else vfwrite |= 1 << itinst->regs[1].VFwrite; - } - if ((itinst->regs[1].VIwrite & (1 << REG_ACC_FLAG)) && itinst->regs[1].VFwxyzw != 0xf) - itinst->addvars[1] |= 1 << REG_ACC_FLAG; - - u32 viwrite = (itinst->regs[0].VIwrite | itinst->regs[1].VIwrite); - - itinst->usedvars[0] = itinst->addvars[0] | viwrite; - itinst->usedvars[1] = itinst->addvars[1] | vfwrite; - -// itinst->addvars[0] &= ~viwrite; -// itinst->addvars[1] &= ~vfwrite; - itinst->keepvars[0] = ~viwrite; - itinst->keepvars[1] = ~vfwrite; - } - } - - itinst = --pblock->insts.end(); - while (itinst != pblock->insts.begin()) - { - itnext = itinst; - --itnext; - - itnext->usedvars[0] |= itinst->usedvars[0]; - itnext->usedvars[1] |= itinst->usedvars[1]; - - itinst = itnext; - } -} - -u32 COMPUTE_LIVE(u32 R, u32 K, u32 L) -{ - u32 live = R | ((L) & (K)); - // special process mac and status flags - // only propagate liveness if doesn't write to the flag - if (!(L&(1 << REG_STATUS_FLAG)) && !(K&(1 << REG_STATUS_FLAG))) - live &= ~(1 << REG_STATUS_FLAG); - if (!(L&(1 << REG_MAC_FLAG)) && !(K&(1 << REG_MAC_FLAG))) - live &= ~(1 << REG_MAC_FLAG); - return live;//|(1<::reverse_iterator itblock; - list::iterator itinst, itnext; - VuBaseBlock::LISTBLOCKS::iterator itchild; - - u32 livevars[2]; - - do - { - changed = FALSE; - for (itblock = s_listBlocks.rbegin(); itblock != s_listBlocks.rend(); ++itblock) - { - - u32 newlive; - VuBaseBlock* pb = *itblock; - - // the last inst relies on the neighbor's insts - itinst = --pb->insts.end(); - - if (pb->blocks.size() > 0) - { - livevars[0] = 0; - livevars[1] = 0; - for (itchild = pb->blocks.begin(); itchild != pb->blocks.end(); ++itchild) - { - VuInstruction& front = (*itchild)->insts.front(); - livevars[0] |= front.livevars[0]; - livevars[1] |= front.livevars[1]; - } - - newlive = COMPUTE_LIVE(itinst->addvars[0], itinst->keepvars[0], livevars[0]); - - // should propagate status flags whose parent insts are not in this block -// if( itinst->nParentPc >= 0 && (itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) -// newlive |= livevars[0]&((1<livevars[0] != newlive) - { - changed = TRUE; - itinst->livevars[0] = newlive; - } - - newlive = COMPUTE_LIVE(itinst->addvars[1], itinst->keepvars[1], livevars[1]); - if (itinst->livevars[1] != newlive) - { - changed = TRUE; - itinst->livevars[1] = newlive; - } - } - - while (itinst != pb->insts.begin()) - { - - itnext = itinst; - --itnext; - - newlive = COMPUTE_LIVE(itnext->addvars[0], itnext->keepvars[0], itinst->livevars[0]); - - // should propagate status flags whose parent insts are not in this block -// if( itnext->nParentPc >= 0 && (itnext->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) && !(itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) -// newlive |= itinst->livevars[0]&((1<livevars[0] != newlive) - { - changed = TRUE; - itnext->livevars[0] = newlive; - itnext->livevars[1] = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); - } - else - { - newlive = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); - if (itnext->livevars[1] != newlive) - { - changed = TRUE; - itnext->livevars[1] = newlive; - } - } - - itinst = itnext; - } - -// if( (livevars[0] | itinst->livevars[0]) != itinst->livevars[0] ) { -// changed = TRUE; -// itinst->livevars[0] |= livevars[0]; -// } -// if( (livevars[1] | itinst->livevars[1]) != itinst->livevars[1] ) { -// changed = TRUE; -// itinst->livevars[1] |= livevars[1]; -// } -// -// while( itinst != pb->insts.begin() ) { -// -// itnext = itinst; --itnext; -// if( (itnext->livevars[0] | (itinst->livevars[0] & itnext->keepvars[0])) != itnext->livevars[0] ) { -// changed = TRUE; -// itnext->livevars[0] |= itinst->livevars[0] & itnext->keepvars[0]; -// itnext->livevars[1] |= itinst->livevars[1] & itnext->keepvars[1]; -// } -// else if( (itnext->livevars[1] | (itinst->livevars[1] & itnext->keepvars[1])) != itnext->livevars[1] ) { -// changed = TRUE; -// itnext->livevars[1] |= itinst->livevars[1] & itnext->keepvars[1]; -// } -// -// itinst = itnext; -// } - } - - } - while (changed); -} - -static void SuperVUEliminateDeadCode() -{ - list::iterator itblock; - VuBaseBlock::LISTBLOCKS::iterator itchild; - list::iterator itinst, itnext; - list listParents; - list::iterator itparent; - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - -#ifdef _DEBUG - u32 startpc = (*itblock)->startpc; - u32 curpc = startpc; -#endif - - itnext = (*itblock)->insts.begin(); - itinst = itnext++; - while (itnext != (*itblock)->insts.end()) - { - if (itinst->type & (INST_CLIP_WRITE | INST_MAC_WRITE | INST_STATUS_WRITE)) - { - u32 live0 = itnext->livevars[0]; - if (itinst->nParentPc >= 0 && itnext->nParentPc >= 0 && itinst->nParentPc != itnext->nParentPc) // superman returns - { - // take the live vars from the next next inst - list::iterator itnextnext = itnext; - ++itnextnext; - if (itnextnext != (*itblock)->insts.end()) - { - live0 = itnextnext->livevars[0]; - } - } - - itinst->regs[0].VIwrite &= live0; - itinst->regs[1].VIwrite &= live0; - - u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; - - (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); - int removetype = 0; - - for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) - { - VuInstruction* parent = *itparent; - - if (viwrite & (1 << REG_CLIP_FLAG)) - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); - } - else - removetype |= INST_CLIP_WRITE; - - if (parent->info.macflag && (itinst->type & INST_MAC_WRITE)) - { - if (!(viwrite&(1 << REG_MAC_FLAG))) - { - //parent->info.macflag = 0; - // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); -#endif - // if VUPIPE_FMAC and destination is vf00, probably need to keep the mac flag - if (parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) - { - parent->regs[0].VIwrite |= ((1 << REG_MAC_FLAG)); - parent->regs[1].VIwrite |= ((1 << REG_MAC_FLAG)); - } - else - removetype |= INST_MAC_WRITE; - } - else - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); - } - } - else removetype |= INST_MAC_WRITE; - - if (parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) - { - if (!(viwrite&(1 << REG_STATUS_FLAG))) - { - //parent->info.statusflag = 0; - // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); -#endif - if (parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) - { - parent->regs[0].VIwrite |= ((1 << REG_STATUS_FLAG)); - parent->regs[1].VIwrite |= ((1 << REG_STATUS_FLAG)); - } - else - removetype |= INST_STATUS_WRITE; - } - else - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); - } - } - else removetype |= INST_STATUS_WRITE; - } - - itinst->type &= ~removetype; - if (itinst->type == 0) - { - itnext = (*itblock)->insts.erase(itinst); - itinst = itnext++; - continue; - } - } -#ifdef _DEBUG - else - { - curpc += 8; - } -#endif - itinst = itnext; - ++itnext; - } - - if (itinst->type & INST_DUMMY) - { - // last inst with the children - u32 mask = 0; - for (itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); ++itchild) - { - mask |= (*itchild)->insts.front().livevars[0]; - } - itinst->regs[0].VIwrite &= mask; - itinst->regs[1].VIwrite &= mask; - u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; - - if (itinst->nParentPc >= 0) - { - - (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); - int removetype = 0; - - for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) - { - VuInstruction* parent = *itparent; - - if (viwrite & (1 << REG_CLIP_FLAG)) - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); - } - else removetype |= INST_CLIP_WRITE; - - if (parent->info.macflag && (itinst->type & INST_MAC_WRITE)) - { - if (!(viwrite&(1 << REG_MAC_FLAG))) - { - //parent->info.macflag = 0; -#ifndef SUPERVU_WRITEBACKS - assert(!(parent->regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); -#endif - removetype |= INST_MAC_WRITE; - } - else - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); - } - } - else removetype |= INST_MAC_WRITE; - - if (parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) - { - if (!(viwrite&(1 << REG_STATUS_FLAG))) - { - //parent->info.statusflag = 0; -#ifndef SUPERVU_WRITEBACKS - assert(!(parent->regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); -#endif - removetype |= INST_STATUS_WRITE; - } - else - { - parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); - parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); - } - } - else removetype |= INST_STATUS_WRITE; - } - - itinst->type &= ~removetype; - if (itinst->type == 0) - { - (*itblock)->insts.erase(itinst); - } - } - } - } -} - -void VuBaseBlock::AssignVFRegs() -{ - int i; - VuBaseBlock::LISTBLOCKS::iterator itchild; - list::iterator itblock; - list::iterator itinst, itnext, itinst2; - - // init the start regs - if (type & BLOCKTYPE_ANALYZED) return; // nothing changed - memcpy(xmmregs, startregs, sizeof(xmmregs)); - - if (type & BLOCKTYPE_ANALYZED) - { - // check if changed - for (i = 0; i < iREGCNT_XMM; ++i) - { - if (xmmregs[i].inuse != startregs[i].inuse) - break; - if (xmmregs[i].inuse && (xmmregs[i].reg != startregs[i].reg || xmmregs[i].type != startregs[i].type)) - break; - } - - if (i == iREGCNT_XMM) return; // nothing changed - } - - u8* oldX86 = x86Ptr; - - for(itinst = insts.begin(); itinst != insts.end(); itinst++) - { - - if (itinst->type & INST_DUMMY) continue; - - // reserve, go from upper to lower - int lastwrite = -1; - - for (i = 1; i >= 0; --i) - { - _VURegsNum* regs = itinst->regs + i; - - - // redo the counters so that the proper regs are released - for (int j = 0; j < iREGCNT_XMM; ++j) - { - if (xmmregs[j].inuse) - { - if (xmmregs[j].type == XMMTYPE_VFREG) - { - int count = 0; - itinst2 = itinst; - - if (i) - { - if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg) - { - itinst2 = insts.end(); - break; - } - else - { - ++count; - ++itinst2; - } - } - - while (itinst2 != insts.end()) - { - if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg || - itinst2->regs[1].VFread0 == xmmregs[j].reg || itinst2->regs[1].VFread1 == xmmregs[j].reg || itinst2->regs[1].VFwrite == xmmregs[j].reg) - break; - - ++count; - ++itinst2; - } - xmmregs[j].counter = 1000 - count; - } - else - { - assert(xmmregs[j].type == XMMTYPE_ACC); - - int count = 0; - itinst2 = itinst; - - if (i) ++itinst2; // acc isn't used in lower insts - - while (itinst2 != insts.end()) - { - assert(!((itinst2->regs[0].VIread | itinst2->regs[0].VIwrite) & (1 << REG_ACC_FLAG))); - - if ((itinst2->regs[1].VIread | itinst2->regs[1].VIwrite) & (1 << REG_ACC_FLAG)) - break; - - ++count; - ++itinst2; - } - - xmmregs[j].counter = 1000 - count; - } - } - } - - if (regs->VFread0) _addNeededVFtoXMMreg(regs->VFread0); - if (regs->VFread1) _addNeededVFtoXMMreg(regs->VFread1); - if (regs->VFwrite) _addNeededVFtoXMMreg(regs->VFwrite); - if (regs->VIread & (1 << REG_ACC_FLAG)) _addNeededACCtoXMMreg(); - if (regs->VIread & (1 << REG_VF0_FLAG)) _addNeededVFtoXMMreg(0); - - // alloc - itinst->vfread0[i] = itinst->vfread1[i] = itinst->vfwrite[i] = itinst->vfacc[i] = -1; - itinst->vfflush[i] = -1; - - if (regs->VFread0) - itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); - else if (regs->VIread & (1 << REG_VF0_FLAG)) - itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); - - if (regs->VFread1) - itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); - else if ((regs->VIread & (1 << REG_VF0_FLAG)) && regs->VFr1xyzw != 0xff) - itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); - - if (regs->VIread & (1 << REG_ACC_FLAG)) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); - - int reusereg = -1; // 0 - VFwrite, 1 - VFAcc - - if (regs->VFwrite) - { - assert(!(regs->VIwrite&(1 << REG_ACC_FLAG))); - - if (regs->VFwxyzw == 0xf) - { - itinst->vfwrite[i] = _checkXMMreg(XMMTYPE_VFREG, regs->VFwrite, 0); - if (itinst->vfwrite[i] < 0) reusereg = 0; - } - else - { - itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); - } - } - else if (regs->VIwrite & (1 << REG_ACC_FLAG)) - { - - if (regs->VFwxyzw == 0xf) - { - itinst->vfacc[i] = _checkXMMreg(XMMTYPE_ACC, 0, 0); - if (itinst->vfacc[i] < 0) reusereg = 1; - } - else - { - itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); - } - } - - if (reusereg >= 0) - { - // reuse - itnext = itinst; - itnext++; - - u8 type = reusereg ? XMMTYPE_ACC : XMMTYPE_VFREG; - u8 reg = reusereg ? 0 : regs->VFwrite; - - if (itinst->vfacc[i] >= 0 && lastwrite != itinst->vfacc[i] && - (itnext == insts.end() || ((regs->VIread&(1 << REG_ACC_FLAG)) && (!(itnext->usedvars[0]&(1 << REG_ACC_FLAG)) || !(itnext->livevars[0]&(1 << REG_ACC_FLAG)))))) - { - - assert(reusereg == 0); - if (itnext == insts.end() || (itnext->livevars[0]&(1 << REG_ACC_FLAG))) _freeXMMreg(itinst->vfacc[i]); - xmmregs[itinst->vfacc[i]].inuse = 1; - xmmregs[itinst->vfacc[i]].reg = reg; - xmmregs[itinst->vfacc[i]].type = type; - xmmregs[itinst->vfacc[i]].mode = 0; - itinst->vfwrite[i] = itinst->vfacc[i]; - } - else if (itinst->vfread0[i] >= 0 && lastwrite != itinst->vfread0[i] && - (itnext == insts.end() || (regs->VFread0 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread0)) || !(itnext->livevars[1]&(1 << regs->VFread0)))))) - { - - if (itnext == insts.end() || (itnext->livevars[1]®s->VFread0)) _freeXMMreg(itinst->vfread0[i]); - - xmmregs[itinst->vfread0[i]].inuse = 1; - xmmregs[itinst->vfread0[i]].reg = reg; - xmmregs[itinst->vfread0[i]].type = type; - xmmregs[itinst->vfread0[i]].mode = 0; - - if (reusereg) - itinst->vfacc[i] = itinst->vfread0[i]; - else - itinst->vfwrite[i] = itinst->vfread0[i]; - } - else if (itinst->vfread1[i] >= 0 && lastwrite != itinst->vfread1[i] && - (itnext == insts.end() || (regs->VFread1 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread1)) || !(itnext->livevars[1]&(1 << regs->VFread1)))))) - { - - if (itnext == insts.end() || (itnext->livevars[1]®s->VFread1)) _freeXMMreg(itinst->vfread1[i]); - - xmmregs[itinst->vfread1[i]].inuse = 1; - xmmregs[itinst->vfread1[i]].reg = reg; - xmmregs[itinst->vfread1[i]].type = type; - xmmregs[itinst->vfread1[i]].mode = 0; - if (reusereg) - itinst->vfacc[i] = itinst->vfread1[i]; - else - itinst->vfwrite[i] = itinst->vfread1[i]; - } - else - { - if (reusereg) - itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); - else - itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); - } - } - - if (itinst->vfwrite[i] >= 0) lastwrite = itinst->vfwrite[i]; - else if (itinst->vfacc[i] >= 0) lastwrite = itinst->vfacc[i]; - - // always alloc at least 1 temp reg - int free0 = (i || regs->VFwrite || regs->VFread0 || regs->VFread1 || (regs->VIwrite & (1 << REG_ACC_FLAG)) || (regs->VIread & (1 << REG_VF0_FLAG))) - ? _allocTempXMMreg(XMMT_FPS, -1) : -1; - int free1 = 0, free2 = 0; - - if (i == 0 && itinst->vfwrite[1] >= 0 && (itinst->vfread0[0] == itinst->vfwrite[1] || itinst->vfread1[0] == itinst->vfwrite[1])) - { - itinst->vfflush[i] = _allocTempXMMreg(XMMT_FPS, -1); - } - - if (i == 1 && (regs->VIwrite & (1 << REG_CLIP_FLAG))) - { - // CLIP inst, need two extra regs - if (free0 < 0) free0 = _allocTempXMMreg(XMMT_FPS, -1); - - free1 = _allocTempXMMreg(XMMT_FPS, -1); - free2 = _allocTempXMMreg(XMMT_FPS, -1); - _freeXMMreg(free1); - _freeXMMreg(free2); - } - else if (regs->VIwrite & (1 << REG_P)) - { - // EFU inst, need extra reg - free1 = _allocTempXMMreg(XMMT_FPS, -1); - if (free0 == -1) free0 = free1; - _freeXMMreg(free1); - } - - if (itinst->vfflush[i] >= 0) _freeXMMreg(itinst->vfflush[i]); - if (free0 >= 0) _freeXMMreg(free0); - - itinst->vffree[i] = (free0 & 0xf) | (free1 << 8) | (free2 << 16); - if (free0 == -1) itinst->vffree[i] |= VFFREE_INVALID0; - - _clearNeededXMMregs(); - } - } - - assert(x86Ptr == oldX86); - u32 analyzechildren = !(type & BLOCKTYPE_ANALYZED); - type |= BLOCKTYPE_ANALYZED; - - //memset(endregs, 0, sizeof(endregs)); - - if (analyzechildren) - { - for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) - { - (*itchild)->AssignVFRegs(); - } - } -} - -struct MARKOVBLANKET -{ - list parents; - list children; -}; - -static MARKOVBLANKET s_markov; - -void VuBaseBlock::AssignVIRegs(int parent) -{ - const int maxregs = 6; - - if (parent) - { - if ((type&BLOCKTYPE_ANALYZEDPARENT)) - return; - - type |= BLOCKTYPE_ANALYZEDPARENT; - s_markov.parents.push_back(this); - for (LISTBLOCKS::iterator it = blocks.begin(); it != blocks.end(); ++it) - { - (*it)->AssignVIRegs(0); - } - return; - } - - if ((type&BLOCKTYPE_ANALYZED)) - return; - - // child - assert(allocX86Regs == -1); - allocX86Regs = s_vecRegArray.size(); - s_vecRegArray.resize(allocX86Regs + iREGCNT_GPR); - - _x86regs* pregs = &s_vecRegArray[allocX86Regs]; - memset(pregs, 0, sizeof(_x86regs)*iREGCNT_GPR); - - assert(parents.size() > 0); - - list::iterator itparent; - u32 usedvars = insts.front().usedvars[0]; - u32 livevars = insts.front().livevars[0]; - - if (parents.size() > 0) - { - u32 usedvars2 = 0xffffffff; - - for(itparent = parents.begin(); itparent != parents.end(); itparent++) - { - usedvars2 &= (*itparent)->insts.front().usedvars[0]; - } - - usedvars |= usedvars2; - } - - usedvars &= livevars; - - // currently order doesn't matter - int num = 0; - - if (usedvars) - { - for (int i = 1; i < 16; ++i) - { - if (usedvars & (1 << i)) - { - pregs[num].inuse = 1; - pregs[num].reg = i; - - livevars &= ~(1 << i); - - if (++num >= maxregs) break; - } - } - } - - if (num < maxregs) - { - livevars &= ~usedvars; - livevars &= insts.back().usedvars[0]; - - if (livevars) - { - for (int i = 1; i < 16; ++i) - { - if (livevars & (1 << i)) - { - pregs[num].inuse = 1; - pregs[num].reg = i; - - if (++num >= maxregs) break; - } - } - } - } - - s_markov.children.push_back(this); - type |= BLOCKTYPE_ANALYZED; - - for(itparent = parents.begin(); itparent != parents.end(); itparent++) - { - (*itparent)->AssignVIRegs(1); - } -} - -static void SuperVUAssignRegs() -{ - list::iterator itblock, itblock2; - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - (*itblock)->type &= ~BLOCKTYPE_ANALYZED; - } - s_listBlocks.front()->AssignVFRegs(); - - // VI assignments, find markov blanket for each node in the graph - // then allocate regs based on the commonly used ones -#ifdef SUPERVU_X86CACHING - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - (*itblock)->type &= ~(BLOCKTYPE_ANALYZED | BLOCKTYPE_ANALYZEDPARENT); - } - s_vecRegArray.resize(0); - u8 usedregs[16]; - - // note: first block always has to start with no alloc regs - bool bfirst = true; - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - - if (!((*itblock)->type & BLOCKTYPE_ANALYZED)) - { - - if ((*itblock)->parents.size() == 0) - { - (*itblock)->type |= BLOCKTYPE_ANALYZED; - bfirst = false; - continue; - } - - s_markov.children.clear(); - s_markov.parents.clear(); - (*itblock)->AssignVIRegs(0); - - // assign the regs - int regid = s_vecRegArray.size(); - s_vecRegArray.resize(regid + iREGCNT_GPR); - - _x86regs* mergedx86 = &s_vecRegArray[regid]; - memset(mergedx86, 0, sizeof(_x86regs)*iREGCNT_GPR); - - if (!bfirst) - { - *(u32*)usedregs = *((u32*)usedregs + 1) = *((u32*)usedregs + 2) = *((u32*)usedregs + 3) = 0; - - for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) - { - assert((*itblock2)->allocX86Regs >= 0); - _x86regs* pregs = &s_vecRegArray[(*itblock2)->allocX86Regs]; - for (int i = 0; i < iREGCNT_GPR; ++i) - { - if (pregs[i].inuse && pregs[i].reg < 16) - { - //assert( pregs[i].reg < 16); - usedregs[pregs[i].reg]++; - } - } - } - - int num = 1; - for (int i = 0; i < 16; ++i) - { - if (usedregs[i] == s_markov.children.size()) - { - // use - mergedx86[num].inuse = 1; - mergedx86[num].reg = i; - mergedx86[num].type = (s_vu ? X86TYPE_VU1 : 0) | X86TYPE_VI; - mergedx86[num].mode = MODE_READ; - if (++num >= iREGCNT_GPR) - break; - if (num == ESP) - ++num; - } - } - - for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) - { - assert((*itblock2)->nStartx86 == -1); - (*itblock2)->nStartx86 = regid; - } - - for(itblock2 = s_markov.parents.begin(); itblock2 != s_markov.parents.end(); itblock2++) - { - assert((*itblock2)->nEndx86 == -1); - (*itblock2)->nEndx86 = regid; - } - } - - bfirst = false; - } - } -#endif -} - -////////////////// -// Recompilation -////////////////// - -// cycles in which the last Q,P regs were finished (written to VU->VI[]) -// the write occurs before the instruction is executed at that cycle -// compare with s_TotalVUCycles -// if less than 0, already flushed -int s_writeQ, s_writeP; - -// declare the saved registers -uptr s_vu1esp, s_callstack;//, s_vu1esp -uptr s_vu1ebp, s_vuebx, s_vuedi, s_vu1esi; - -static int s_recWriteQ, s_recWriteP; // wait times during recompilation -static int s_needFlush; // first bit - Q, second bit - P, third bit - Q has been written, fourth bit - P has been written - -static int s_JumpX86; -static int s_ScheduleXGKICK = 0, s_XGKICKReg = -1; - -void recVUMI_XGKICK_(VURegs *VU); - -void SuperVUCleanupProgram(u32 startpc, int vuindex) -{ -#ifdef SUPERVU_COUNT - QueryPerformanceCounter(&svufinal); - svutime += (u32)(svufinal.QuadPart - svubase.QuadPart); -#endif - - assert(s_vu1esp == 0); - - VU = vuindex ? &VU1 : &VU0; - VU->cycle += s_TotalVUCycles; - - //VU cycle stealing hack, 3000 cycle maximum so it doesn't get out of hand - if (s_TotalVUCycles < 3000) - cpuRegs.cycle += s_TotalVUCycles * Config.Hacks.VUCycleSteal; - else - cpuRegs.cycle += 3000 * Config.Hacks.VUCycleSteal; - - if ((int)s_writeQ > 0) VU->VI[REG_Q] = VU->q; - if ((int)s_writeP > 0) - { - assert(VU == &VU1); - VU1.VI[REG_P] = VU1.p; // only VU1 - } - - //memset(recVUStack, 0, SUPERVU_STACKSIZE * 4); - - // Could clear allocation info to prevent possibly bad data being used in other parts of pcsx2; - // not doing this because it's slow and not needed (rama) - // _initXMMregs(); - // _initMMXregs(); - // _initX86regs(); -} - -#if defined(_MSC_VER) - -// entry point of all vu programs from emulator calls -__declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) -{ - __asm - { - mov eax, dword ptr [esp] - mov s_TotalVUCycles, 0 // necessary to be here! - add esp, 4 - mov s_callstack, eax - call SuperVUGetProgram - - // save cpu state - mov s_vu1ebp, ebp - mov s_vu1esi, esi // have to save even in Release - mov s_vuedi, edi // have to save even in Release - mov s_vuebx, ebx - } -#ifdef _DEBUG - __asm - { - mov s_vu1esp, esp - } -#endif - - __asm - { - //stmxcsr s_ssecsr - ldmxcsr g_sseVUMXCSR - - // init vars - mov s_writeQ, 0xffffffff - mov s_writeP, 0xffffffff - - jmp eax - } -} - -// exit point of all vu programs -__declspec(naked) static void SuperVUEndProgram() -{ - __asm - { - // restore cpu state - ldmxcsr g_sseMXCSR - - mov ebp, s_vu1ebp - mov esi, s_vu1esi - mov edi, s_vuedi - mov ebx, s_vuebx - } - -#ifdef _DEBUG - __asm - { - sub s_vu1esp, esp - } -#endif - - __asm - { - call SuperVUCleanupProgram - jmp s_callstack // so returns correctly - } -} - -#endif - -// Flushes P/Q regs -void SuperVUFlush(int p, int wait) -{ - u8* pjmp[3]; - if (!(s_needFlush&(1 << p))) return; - - int recwait = p ? s_recWriteP : s_recWriteQ; - if (!wait && s_pCurInst->info.cycle < recwait) return; - - if (recwait == 0) - { - // write didn't happen this block - MOV32MtoR(EAX, p ? (uptr)&s_writeP : (uptr)&s_writeQ); - OR32RtoR(EAX, EAX); - pjmp[0] = JS8(0); - - if (s_pCurInst->info.cycle) SUB32ItoR(EAX, s_pCurInst->info.cycle); - - // if writeQ <= total+offset - if (!wait) // only write back if time is up - { - CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); - pjmp[1] = JG8(0); - } - else - { - // add (writeQ-total-offset) to s_TotalVUCycles - // necessary? - CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); - pjmp[2] = JLE8(0); - MOV32RtoM((uptr)&s_TotalVUCycles, EAX); - x86SetJ8(pjmp[2]); - } - } - else if (wait && s_pCurInst->info.cycle < recwait) - { - ADD32ItoM((uptr)&s_TotalVUCycles, recwait); - } - - MOV32MtoR(EAX, SuperVUGetVIAddr(p ? REG_P : REG_Q, 0)); - MOV32ItoM(p ? (uptr)&s_writeP : (uptr)&s_writeQ, 0x80000000); - MOV32RtoM(SuperVUGetVIAddr(p ? REG_P : REG_Q, 1), EAX); - - if (recwait == 0) - { - if (!wait) x86SetJ8(pjmp[1]); - x86SetJ8(pjmp[0]); - } - - if (wait || (!p && recwait == 0 && s_pCurInst->info.cycle >= 12) || (!p && recwait > 0 && s_pCurInst->info.cycle >= recwait)) - s_needFlush &= ~(1 << p); -} - -// executed only once per program -static u32* SuperVUStaticAlloc(u32 size) -{ - assert(recVUStackPtr + size <= recVUStack + SUPERVU_STACKSIZE); - // always zero - if (size == 4) *(u32*)recVUStackPtr = 0; - else memset(recVUStackPtr, 0, size); - recVUStackPtr += size; - return (u32*)(recVUStackPtr - size); -} - -static void SuperVURecompile() -{ - // save cpu state - recVUStackPtr = recVUStack; - - _initXMMregs(); - - list::iterator itblock; - - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - (*itblock)->type &= ~BLOCKTYPE_ANALYZED; - } - - s_listBlocks.front()->Recompile(); - - // make sure everything compiled - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - assert(((*itblock)->type & BLOCKTYPE_ANALYZED) && (*itblock)->pcode != NULL); - } - - // link all blocks - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - VuBaseBlock::LISTBLOCKS::iterator itchild; - - assert((*itblock)->blocks.size() <= ArraySize((*itblock)->pChildJumps)); - - int i = 0; - for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) - { - - if ((u32)(uptr)(*itblock)->pChildJumps[i] == 0xffffffff) - continue; - - if ((*itblock)->pChildJumps[i] == NULL) - { - VuBaseBlock* pchild = *itchild; - - if (pchild->type & BLOCKTYPE_HASEOP) - { - assert(pchild->blocks.size() == 0); - - AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag - AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); - - MOV32ItoM((uptr)&VU->VI[REG_TPC], pchild->endpc); - JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); - } - // only other case is when there are two branches - else - { - assert((*itblock)->insts.back().regs[0].pipe == VUPIPE_BRANCH); - } - - continue; - } - - if ((u32)(uptr)(*itblock)->pChildJumps[i] & 0x80000000) - { - // relative - assert((uptr)(*itblock)->pChildJumps[i] <= 0xffffffff); - (*itblock)->pChildJumps[i] = (u32*)((uptr)(*itblock)->pChildJumps[i] & 0x7fffffff); - *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode - ((uptr)(*itblock)->pChildJumps[i] + 4); - } - else - { - *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode; - } - - ++i; - } - } - - s_pFnHeader->pprogfunc = s_listBlocks.front()->pcode; -} - -// debug - - -u32 s_saveecx, s_saveedx, s_saveebx, s_saveesi, s_saveedi, s_saveebp; -u32 g_curdebugvu; - -//float vuDouble(u32 f); - -#if defined(_MSC_VER) -__declspec(naked) static void svudispfn() -{ - __asm - { - mov g_curdebugvu, eax - mov s_saveecx, ecx - mov s_saveedx, edx - mov s_saveebx, ebx - mov s_saveesi, esi - mov s_saveedi, edi - mov s_saveebp, ebp - } -#else - -void svudispfntemp() -{ -#endif - -#ifdef _DEBUG - static u32 i; - - if (((vudump&8) && g_curdebugvu) || ((vudump&0x80) && !g_curdebugvu)) //&& g_vu1lastrec != g_vu1last ) { - { - - if (skipparent != g_vu1lastrec) - { - for (i = 0; i < ArraySize(badaddrs); ++i) - { - if (s_svulast == badaddrs[i][1] && g_vu1lastrec == badaddrs[i][0]) - break; - } - - if (i == ArraySize(badaddrs)) - { - //static int curesp; - //__asm mov curesp, esp - //Console::WriteLn("tVU: %x %x %x", s_svulast, s_vucount, s_vufnheader); - if (g_curdebugvu) iDumpVU1Registers(); - else iDumpVU0Registers(); - s_vucount++; - } - } - - g_vu1lastrec = s_svulast; - } -#endif - -#if defined(_MSC_VER) - __asm - { - mov ecx, s_saveecx - mov edx, s_saveedx - mov ebx, s_saveebx - mov esi, s_saveesi - mov edi, s_saveedi - mov ebp, s_saveebp - ret - } -#endif -} - -// frees all regs taking into account the livevars -void SuperVUFreeXMMregs(u32* livevars) -{ - for (int i = 0; i < iREGCNT_XMM; ++i) - { - if (xmmregs[i].inuse) - { - // same reg - if ((xmmregs[i].mode & MODE_WRITE)) - { - -#ifdef SUPERVU_INTERCACHING - if (xmmregs[i].type == XMMTYPE_VFREG) - { - if (!(livevars[1] & (1 << xmmregs[i].reg))) continue; - } - else if (xmmregs[i].type == XMMTYPE_ACC) - { - if (!(livevars[0] & (1 << REG_ACC_FLAG))) continue; - } -#endif - - if (xmmregs[i].mode & MODE_VUXYZ) - { - // ALWAYS update - u32 addr = xmmregs[i].type == XMMTYPE_VFREG ? (uptr) & VU->VF[xmmregs[i].reg] : (uptr) & VU->ACC; - - if (xmmregs[i].mode & MODE_VUZ) - { - SSE_MOVHPS_XMM_to_M64(addr, (x86SSERegType)i); - SSE_SHUFPS_M128_to_XMM((x86SSERegType)i, addr, 0xc4); - } - else - { - SSE_MOVHPS_M64_to_XMM((x86SSERegType)i, addr + 8); - } - - xmmregs[i].mode &= ~MODE_VUXYZ; - } - - _freeXMMreg(i); - } - } - } - - //_freeXMMregs(); -} - -void SuperVUTestVU0Condition(u32 incstack) -{ - if (s_vu && !SUPERVU_CHECKCONDITION) return; // vu0 only - - CMP32ItoM((uptr)&s_TotalVUCycles, 512); // sometimes games spin on vu0, so be careful with this value - // woody hangs if too high - - if (incstack) - { - u8* ptr = JB8(0); - - ADD32ItoR(ESP, incstack); - //CALLFunc((u32)timeout); - JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); - - x86SetJ8(ptr); - } - else JAE32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 6)); -} - -void VuBaseBlock::Recompile() -{ - if (type & BLOCKTYPE_ANALYZED) return; - - x86Align(16); - pcode = x86Ptr; - -#ifdef _DEBUG - MOV32ItoM((uptr)&s_vufnheader, s_pFnHeader->startpc); - MOV32ItoM((uptr)&VU->VI[REG_TPC], startpc); - MOV32ItoM((uptr)&s_svulast, startpc); - - list::iterator itparent; - for (itparent = parents.begin(); itparent != parents.end(); ++itparent) - { - if ((*itparent)->blocks.size() == 1 && (*itparent)->blocks.front()->startpc == startpc && - ((*itparent)->insts.size() < 2 || (----(*itparent)->insts.end())->regs[0].pipe != VUPIPE_BRANCH)) - { - MOV32ItoM((uptr)&skipparent, (*itparent)->startpc); - break; - } - } - - if (itparent == parents.end()) MOV32ItoM((uptr)&skipparent, -1); - - MOV32ItoR(EAX, s_vu); - CALLFunc((uptr)svudispfn); -#endif - - s_pCurBlock = this; - s_needFlush = 3; - pc = startpc; - branch = 0; - s_recWriteQ = s_recWriteP = 0; - s_XGKICKReg = -1; - s_ScheduleXGKICK = 0; - - s_ClipRead = s_PrevClipWrite = (uptr) & VU->VI[REG_CLIP_FLAG]; - s_StatusRead = s_PrevStatusWrite = (uptr) & VU->VI[REG_STATUS_FLAG]; - s_MACRead = s_PrevMACWrite = (uptr) & VU->VI[REG_MAC_FLAG]; - s_PrevIWrite = (uptr) & VU->VI[REG_I]; - s_JumpX86 = 0; - s_UnconditionalDelay = 0; - - memcpy(xmmregs, startregs, sizeof(xmmregs)); -#ifdef SUPERVU_X86CACHING - if (nStartx86 >= 0) - memcpy(x86regs, &s_vecRegArray[nStartx86], sizeof(x86regs)); - else - _initX86regs(); -#else - _initX86regs(); -#endif - - list::iterator itinst; - for(itinst = insts.begin(); itinst != insts.end(); itinst++) - { - s_pCurInst = &(*itinst); - if (s_JumpX86 > 0) - { - if (!x86regs[s_JumpX86].inuse) - { - // load - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_READ); - } - x86regs[s_JumpX86].needed = 1; - } - - if (s_ScheduleXGKICK && s_XGKICKReg > 0) - { - assert(x86regs[s_XGKICKReg].inuse); - x86regs[s_XGKICKReg].needed = 1; - } - itinst->Recompile(itinst, vuxyz); - - if (s_ScheduleXGKICK > 0) - { - if (s_ScheduleXGKICK-- == 1) - { - recVUMI_XGKICK_(VU); - } - } - } - assert(pc == endpc); - assert(s_ScheduleXGKICK == 0); - - // flush flags - if (s_PrevClipWrite != (uptr)&VU->VI[REG_CLIP_FLAG]) - { - MOV32MtoR(EAX, s_PrevClipWrite); - MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); - } - if (s_PrevStatusWrite != (uptr)&VU->VI[REG_STATUS_FLAG]) - { - MOV32MtoR(EAX, s_PrevStatusWrite); - MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); - } - if (s_PrevMACWrite != (uptr)&VU->VI[REG_MAC_FLAG]) - { - MOV32MtoR(EAX, s_PrevMACWrite); - MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); - } -// if( s_StatusRead != (uptr)&VU->VI[REG_STATUS_FLAG] ) { -// // only lower 8 bits valid! -// MOVZX32M8toR(EAX, s_StatusRead); -// MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); -// } -// if( s_MACRead != (uptr)&VU->VI[REG_MAC_FLAG] ) { -// // only lower 8 bits valid! -// MOVZX32M8toR(EAX, s_MACRead); -// MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); -// } - if (s_PrevIWrite != (uptr)&VU->VI[REG_I]) - { - MOV32ItoM((uptr)&VU->VI[REG_I], *(u32*)s_PrevIWrite); // never changes - } - - ADD32ItoM((uptr)&s_TotalVUCycles, cycles); - - // compute branches, jumps, eop - if (type & BLOCKTYPE_HASEOP) - { - // end - _freeXMMregs(); - _freeX86regs(); - AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag - AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); - - if (!branch) MOV32ItoM((uptr)&VU->VI[REG_TPC], endpc); - - JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); - } - else - { - - u32 livevars[2] = {0}; - - list::iterator lastinst = GetInstIterAtPc(endpc - 8); - lastinst++; - - if (lastinst != insts.end()) - { - livevars[0] = lastinst->livevars[0]; - livevars[1] = lastinst->livevars[1]; - } - else - { - // take from children - if (blocks.size() > 0) - { - LISTBLOCKS::iterator itchild; - for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) - { - livevars[0] |= (*itchild)->insts.front().livevars[0]; - livevars[1] |= (*itchild)->insts.front().livevars[1]; - } - } - else - { - livevars[0] = ~0; - livevars[1] = ~0; - } - } - - SuperVUFreeXMMregs(livevars); - - // get rid of any writes, otherwise _freeX86regs will write - x86regs[s_JumpX86].mode &= ~MODE_WRITE; - - if (branch == 1) - { - if (!x86regs[s_JumpX86].inuse) - { - assert(x86regs[s_JumpX86].type == X86TYPE_VUJUMP); - s_JumpX86 = 0xffffffff; // notify to jump from g_recWriteback - } - } - - // align VI regs -#ifdef SUPERVU_X86CACHING - if (nEndx86 >= 0) - { - _x86regs* endx86 = &s_vecRegArray[nEndx86]; - for (int i = 0; i < iREGCNT_GPR; ++i) - { - if (endx86[i].inuse) - { - - if (s_JumpX86 == i && x86regs[s_JumpX86].inuse) - { - x86regs[s_JumpX86].inuse = 0; - x86regs[EAX].inuse = 1; - MOV32RtoR(EAX, s_JumpX86); - s_JumpX86 = EAX; - } - - if (x86regs[i].inuse) - { - if (x86regs[i].type == endx86[i].type && x86regs[i].reg == endx86[i].reg) - { - _freeX86reg(i); - // will continue to use it - continue; - } - -#ifdef SUPERVU_INTERCACHING - if (x86regs[i].type == (X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0))) - { - if (livevars[0] & (1 << x86regs[i].reg)) - _freeX86reg(i); - else - x86regs[i].inuse = 0; - } - else -#endif - { - _freeX86reg(i); - } - } - - // realloc - _allocX86reg(i, endx86[i].type, endx86[i].reg, MODE_READ); - if (x86regs[i].mode & MODE_WRITE) - { - _freeX86reg(i); - x86regs[i].inuse = 1; - } - } - else _freeX86reg(i); - } - } - else _freeX86regs(); -#else - _freeX86regs(); -#endif - - // store the last block executed - MOV32ItoM((uptr)&g_nLastBlockExecuted, s_pCurBlock->startpc); - - switch (branch) - { - case 1: // branch, esi has new prog - - SuperVUTestVU0Condition(0); - - if (s_JumpX86 == 0xffffffff) - JMP32M((uptr)&g_recWriteback); - else - JMPR(s_JumpX86); - - break; - case 4: // jalr - pChildJumps[0] = (u32*)0xffffffff; - // fall through - - case 0x10: // jump, esi has new vupc - { - _freeXMMregs(); - _freeX86regs(); - - SuperVUTestVU0Condition(8); - - // already onto stack - CALLFunc((uptr)SuperVUGetProgram); - ADD32ItoR(ESP, 8); - JMPR(EAX); - break; - } - - case 0x13: // jr with uncon branch, uncond branch takes precendence (dropship) - { -// s32 delta = (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) << 3; -// ADD32ItoRmOffset(ESP, delta, 0); - - ADD32ItoR(ESP, 8); // restore - pChildJumps[0] = (u32*)((uptr)JMP32(0) | 0x80000000); - break; - } - case 0: - case 3: // unconditional branch - pChildJumps[s_UnconditionalDelay] = (u32*)((uptr)JMP32(0) | 0x80000000); - break; - - default: - DevCon::Error("Bad branch %x\n", params branch); - assert(0); - break; - } - } - - pendcode = x86Ptr; - type |= BLOCKTYPE_ANALYZED; - - LISTBLOCKS::iterator itchild; - for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) - { - (*itchild)->Recompile(); - } -} - -#define GET_VUXYZMODE(reg) 0//((vuxyz&(1<<(reg)))?MODE_VUXYZ:0) - -int VuInstruction::SetCachedRegs(int upper, u32 vuxyz) -{ - if (vfread0[upper] >= 0) - { - SuperVUFreeXMMreg(vfread0[upper], XMMTYPE_VFREG, regs[upper].VFread0); - _allocVFtoXMMreg(VU, vfread0[upper], regs[upper].VFread0, MODE_READ | GET_VUXYZMODE(regs[upper].VFread0)); - } - if (vfread1[upper] >= 0) - { - SuperVUFreeXMMreg(vfread1[upper], XMMTYPE_VFREG, regs[upper].VFread1); - _allocVFtoXMMreg(VU, vfread1[upper], regs[upper].VFread1, MODE_READ | GET_VUXYZMODE(regs[upper].VFread1)); - } - if (vfacc[upper] >= 0 && (regs[upper].VIread&(1 << REG_ACC_FLAG))) - { - SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); - _allocACCtoXMMreg(VU, vfacc[upper], MODE_READ); - } - if (vfwrite[upper] >= 0) - { - assert(regs[upper].VFwrite > 0); - SuperVUFreeXMMreg(vfwrite[upper], XMMTYPE_VFREG, regs[upper].VFwrite); - _allocVFtoXMMreg(VU, vfwrite[upper], regs[upper].VFwrite, - MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0) | GET_VUXYZMODE(regs[upper].VFwrite)); - } - if (vfacc[upper] >= 0 && (regs[upper].VIwrite&(1 << REG_ACC_FLAG))) - { - SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); - _allocACCtoXMMreg(VU, vfacc[upper], MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0)); - } - - int info = PROCESS_VU_SUPER; - if (vfread0[upper] >= 0) info |= PROCESS_EE_SET_S(vfread0[upper]); - if (vfread1[upper] >= 0) info |= PROCESS_EE_SET_T(vfread1[upper]); - if (vfacc[upper] >= 0) info |= PROCESS_VU_SET_ACC(vfacc[upper]); - if (vfwrite[upper] >= 0) - { - if (regs[upper].VFwrite == _Ft_ && vfread1[upper] < 0) - { - info |= PROCESS_EE_SET_T(vfwrite[upper]); - } - else - { - assert(regs[upper].VFwrite == _Fd_); - info |= PROCESS_EE_SET_D(vfwrite[upper]); - } - } - - if (!(vffree[upper]&VFFREE_INVALID0)) - { - SuperVUFreeXMMreg(vffree[upper]&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, vffree[upper]&0xf); - } - info |= PROCESS_VU_SET_TEMP(vffree[upper] & 0xf); - - if (vfflush[upper] >= 0) - { - SuperVUFreeXMMreg(vfflush[upper], XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, vfflush[upper]); - } - - if (upper && (regs[upper].VIwrite & (1 << REG_CLIP_FLAG))) - { - // CLIP inst, need two extra temp registers, put it EEREC_D and EEREC_ACC - assert(vfwrite[upper] == -1); - SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); - info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); - - SuperVUFreeXMMreg((vffree[upper] >> 16)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 16)&0xf); - info |= PROCESS_EE_SET_ACC((vffree[upper] >> 16) & 0xf); - - _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore - _freeXMMreg((vffree[upper] >> 16)&0xf); // don't need anymore - } - else if (regs[upper].VIwrite & (1 << REG_P)) - { - SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); - _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); - info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); - _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore - } - - if (vfflush[upper] >= 0) _freeXMMreg(vfflush[upper]); - if (!(vffree[upper]&VFFREE_INVALID0)) - _freeXMMreg(vffree[upper]&0xf); // don't need anymore - - if ((regs[0].VIwrite | regs[1].VIwrite) & ((1 << REG_STATUS_FLAG) | (1 << REG_MAC_FLAG))) - info |= PROCESS_VU_UPDATEFLAGS; - - return info; -} - -void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) -{ - //static PCSX2_ALIGNED16(VECTOR _VF); - //static PCSX2_ALIGNED16(VECTOR _VFc); - u32 *ptr; - u8* pjmp; - int vfregstore = 0; - - assert(s_pCurInst == this); - s_WriteToReadQ = 0; - - ptr = (u32*) & VU->Micro[ pc ]; - - if (type & INST_Q_READ) - SuperVUFlush(0, (ptr[0] == 0x800003bf) || !!(regs[0].VIwrite & (1 << REG_Q))); - if (type & INST_P_READ) - SuperVUFlush(1, (ptr[0] == 0x800007bf) || !!(regs[0].VIwrite & (1 << REG_P))); - - if (type & INST_DUMMY) - { - - // find nParentPc - VuInstruction* pparentinst = NULL; - - // if true, will check if parent block was executed before getting the results of the flags (superman returns) - int nParentCheckForExecution = -1; - -// int badaddrs[] = { -// 0x60,0x68,0x70,0x60,0x68,0x70,0x88,0x90,0x98,0x98,0xa8,0xb8,0x88,0x90, -// 0x4a8,0x4a8,0x398,0x3a0,0x3a8,0xa0 -// }; - -#ifdef SUPERVU_PROPAGATEFLAGS - if (nParentPc != -1 && (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc)) - { - -// if( !s_vu ) { -// for(int j = 0; j < ARRAYSIZE(badaddrs); ++j) { -// if( badaddrs[j] == nParentPc ) -// goto NoParent; -// } -// } - - list::iterator itblock; - for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) - { - if (nParentPc >= (*itblock)->startpc && nParentPc < (*itblock)->endpc) - { - pparentinst = &(*(*itblock)->GetInstIterAtPc(nParentPc)); - //if( !s_vu ) SysPrintf("%x ", nParentPc); - if (find(s_pCurBlock->parents.begin(), s_pCurBlock->parents.end(), *itblock) != s_pCurBlock->parents.end()) - nParentCheckForExecution = (*itblock)->startpc; - break; - } - } - - assert(pparentinst != NULL); - } -#endif - - if (type & INST_CLIP_WRITE) - { - if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) - { - - if (!CHECK_VUCLIPFLAGHACK && pparentinst != NULL) - { - - if (nParentCheckForExecution >= 0) - { - if (pparentinst->pClipWrite == 0) - pparentinst->pClipWrite = (uptr)SuperVUStaticAlloc(4); - - if (s_ClipRead == 0) - s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; - - CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); - u8* jptr = JNE8(0); - CMP32ItoM((uptr)&s_ClipRead, (uptr)&VU->VI[REG_CLIP_FLAG]); - u8* jptr2 = JE8(0); - MOV32MtoR(EAX, pparentinst->pClipWrite); - MOV32RtoM(s_ClipRead, EAX); - x86SetJ8(jptr); - x86SetJ8(jptr2); - } - } - else s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; - } - else - { - s_ClipRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pClipWrite; - if (s_ClipRead == 0) Console::WriteLn("super ClipRead allocation error!"); - } - } - - // before modifying, check if they will ever be read - if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) - { - - u8 outofblock = 0; - if (type & INST_STATUS_WRITE) - { - - if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) - { - - // reading from out of this block, so already flushed to mem - if (pparentinst != NULL) //&& pparentinst->pStatusWrite != NULL ) { - { - - // might not have processed it yet, so reserve a mem loc - if (pparentinst->pStatusWrite == 0) - { - pparentinst->pStatusWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pparentinst->pStatusWrite, 0); - } - -// if( s_pCurBlock->prevFlagsOutOfBlock && s_StatusRead != NULL ) { -// // or instead since don't now which parent we came from -// MOV32MtoR(EAX, pparentinst->pStatusWrite); -// OR32RtoM(s_StatusRead, EAX); -// MOV32ItoM(pparentinst->pStatusWrite, 0); -// } - - if (nParentCheckForExecution >= 0) - { - - // don't now which parent we came from, so have to check -// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); -// if( s_StatusRead != NULL ) -// MOV32MtoR(EAX, s_StatusRead); -// else -// MOV32MtoR(EAX, (uptr)&VU->VI[REG_STATUS_FLAG]); -// s_StatusRead = tempstatus; - - if (s_StatusRead == 0) - s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; - - CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); - u8* jptr = JNE8(0); - MOV32MtoR(EAX, pparentinst->pStatusWrite); - MOV32ItoM(pparentinst->pStatusWrite, 0); - MOV32RtoM(s_StatusRead, EAX); - x86SetJ8(jptr); - } - else - { - uptr tempstatus = (uptr)SuperVUStaticAlloc(4); - MOV32MtoR(EAX, pparentinst->pStatusWrite); - MOV32RtoM(tempstatus, EAX); - MOV32ItoM(pparentinst->pStatusWrite, 0); - s_StatusRead = tempstatus; - } - - outofblock = 2; - } - else - s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; - } - else - { - s_StatusRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pStatusWrite; - if (s_StatusRead == 0) Console::WriteLn("super StatusRead allocation error!"); -// if( pc >= (u32)s_pCurBlock->endpc-8 ) { -// // towards the end, so variable might be leaded to another block (silent hill 4) -// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); -// MOV32MtoR(EAX, s_StatusRead); -// MOV32RtoM(tempstatus, EAX); -// MOV32ItoM(s_StatusRead, 0); -// s_StatusRead = tempstatus; -// } - } - } - if (type & INST_MAC_WRITE) - { - - if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) - { - // reading from out of this block, so already flushed to mem - - if (pparentinst != NULL) //&& pparentinst->pMACWrite != NULL ) { - { - // necessary for (katamari) - // towards the end, so variable might be leaked to another block (silent hill 4) - - // might not have processed it yet, so reserve a mem loc - if (pparentinst->pMACWrite == 0) - { - pparentinst->pMACWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pparentinst->pMACWrite, 0); - } - -// if( s_pCurBlock->prevFlagsOutOfBlock && s_MACRead != NULL ) { -// // or instead since don't now which parent we came from -// MOV32MtoR(EAX, pparentinst->pMACWrite); -// OR32RtoM(s_MACRead, EAX); -// MOV32ItoM(pparentinst->pMACWrite, 0); -// } - if (nParentCheckForExecution >= 0) - { - - // don't now which parent we came from, so have to check -// uptr tempmac = (uptr)SuperVUStaticAlloc(4); -// if( s_MACRead != NULL ) -// MOV32MtoR(EAX, s_MACRead); -// else -// MOV32MtoR(EAX, (uptr)&VU->VI[REG_MAC_FLAG]); -// s_MACRead = tempmac; - - if (s_MACRead == 0) s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; - - CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); - u8* jptr = JNE8(0); - MOV32MtoR(EAX, pparentinst->pMACWrite); - MOV32ItoM(pparentinst->pMACWrite, 0); - MOV32RtoM(s_MACRead, EAX); - x86SetJ8(jptr); - } - else - { - uptr tempMAC = (uptr)SuperVUStaticAlloc(4); - MOV32MtoR(EAX, pparentinst->pMACWrite); - MOV32RtoM(tempMAC, EAX); - MOV32ItoM(pparentinst->pMACWrite, 0); - s_MACRead = tempMAC; - } - - outofblock = 2; - } - else - s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; - -// if( pc >= (u32)s_pCurBlock->endpc-8 ) { -// // towards the end, so variable might be leaked to another block (silent hill 4) -// uptr tempMAC = (uptr)SuperVUStaticAlloc(4); -// MOV32MtoR(EAX, s_MACRead); -// MOV32RtoM(tempMAC, EAX); -// MOV32ItoM(s_MACRead, 0); -// s_MACRead = tempMAC; -// } - } - else - { - s_MACRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pMACWrite; - } - } - - s_pCurBlock->prevFlagsOutOfBlock = outofblock; - } - else if (pparentinst != NULL) - { - // make sure to reset the mac and status flags! (katamari) - if (pparentinst->pStatusWrite) - MOV32ItoM(pparentinst->pStatusWrite, 0); - if (pparentinst->pMACWrite) - MOV32ItoM(pparentinst->pMACWrite, 0); - } - - assert(s_ClipRead != 0); - assert(s_MACRead != 0); - assert(s_StatusRead != 0); - - return; - } - - s_pCurBlock->prevFlagsOutOfBlock = 0; - -#ifdef _DEBUG - MOV32ItoR(EAX, pc); -#endif - - assert(!(type & (INST_CLIP_WRITE | INST_STATUS_WRITE | INST_MAC_WRITE))); - pc += 8; - - list::const_iterator itinst2; - - if ((regs[0].VIwrite | regs[1].VIwrite) & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) - { - if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) - { - if (pMACWrite == 0) - { - pMACWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pMACWrite, 0); - } - if (pStatusWrite == 0) - { - pStatusWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pStatusWrite, 0); - } - } - else - { - assert(s_StatusRead == (uptr)&VU->VI[REG_STATUS_FLAG]); - assert(s_MACRead == (uptr)&VU->VI[REG_MAC_FLAG]); - pMACWrite = s_MACRead; - pStatusWrite = s_StatusRead; - } - } - - if ((pClipWrite == 0) && ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_CLIP_FLAG))) - { - pClipWrite = (uptr)SuperVUStaticAlloc(4); - //MOV32ItoM(pClipWrite, 0); - } - -#ifdef SUPERVU_X86CACHING - // redo the counters so that the proper regs are released - for (int j = 0; j < iREGCNT_GPR; ++j) - { - if (x86regs[j].inuse && X86_ISVI(x86regs[j].type)) - { - int count = 0; - itinst2 = itinst; - - while (itinst2 != s_pCurBlock->insts.end()) - { - if ((itinst2->regs[0].VIread | itinst2->regs[0].VIwrite | itinst2->regs[1].VIread | itinst2->regs[1].VIwrite) && (1 << x86regs[j].reg)) - break; - - ++count; - ++itinst2; - } - - x86regs[j].counter = 1000 - count; - } - } -#endif - - if (s_vu == 0 && (ptr[1] & 0x20000000)) // M flag - { - OR8ItoM((uptr)&VU->flags, VUFLAG_MFLAGSET); - } - if (ptr[1] & 0x10000000) // D flag - { - TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x400 : 0x004); - u8* ptr = JZ8(0); - OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x200 : 0x002); - _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); - x86SetJ8(ptr); - } - if (ptr[1] & 0x08000000) // T flag - { - TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x800 : 0x008); - u8* ptr = JZ8(0); - OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x400 : 0x004); - _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); - x86SetJ8(ptr); - } - - // check upper flags - if (ptr[1] & 0x80000000) // I flag - { - - assert(!(regs[0].VIwrite & ((1 << REG_Q) | (1 << REG_P)))); - - VU->code = ptr[1]; - s_vuInfo = SetCachedRegs(1, vuxyz); - if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; - if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; - - recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); - - s_PrevIWrite = (uptr)ptr; - _clearNeededXMMregs(); - _clearNeededX86regs(); - } - else - { - if (regs[0].VIwrite & (1 << REG_Q)) - { - - // search for all the insts between this inst and writeback - itinst2 = itinst; - ++itinst2; - u32 cacheq = (itinst2 == s_pCurBlock->insts.end()); - u32* codeptr2 = ptr + 2; - - while (itinst2 != s_pCurBlock->insts.end()) - { - if (!(itinst2->type & INST_DUMMY) && ((itinst2->regs[0].VIwrite&(1 << REG_Q)) || codeptr2[0] == 0x800003bf)) // waitq, or fdiv inst - { - break; - } - if ((itinst2->type & INST_Q_WRITE) && itinst2->nParentPc == pc - 8) - { - break; - } - if (itinst2->type & INST_Q_READ) - { - cacheq = 1; - break; - } - if (itinst2->type & INST_DUMMY) - { - ++itinst2; - continue; - } - codeptr2 += 2; - ++itinst2; - } - - if (itinst2 == s_pCurBlock->insts.end()) - cacheq = 1; - - int x86temp = -1; - if (cacheq) - x86temp = _allocX86reg(-1, X86TYPE_TEMP, 0, 0); - - // new is written so flush old - // if type & INST_Q_READ, already flushed - if (!(type & INST_Q_READ) && s_recWriteQ == 0) MOV32MtoR(EAX, (uptr)&s_writeQ); - - if (cacheq) - MOV32MtoR(x86temp, (uptr)&s_TotalVUCycles); - - if (!(type & INST_Q_READ)) - { - if (s_recWriteQ == 0) - { - OR32RtoR(EAX, EAX); - pjmp = JS8(0); - MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); - MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); - x86SetJ8(pjmp); - } - else if (s_needFlush & 1) - { - MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); - MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); - s_needFlush &= ~1; - } - } - - // write new Q - if (cacheq) - { - assert(s_pCurInst->pqcycles > 1); - ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); - MOV32RtoM((uptr)&s_writeQ, x86temp); - s_needFlush |= 1; - } - else - { - // won't be writing back - s_WriteToReadQ = 1; - s_needFlush &= ~1; - MOV32ItoM((uptr)&s_writeQ, 0x80000001); - } - - s_recWriteQ = s_pCurInst->info.cycle + s_pCurInst->pqcycles; - - if (x86temp >= 0) - _freeX86reg(x86temp); - } - - if (regs[0].VIwrite & (1 << REG_P)) - { - int x86temp = _allocX86reg(-1, X86TYPE_TEMP, 0, 0); - - // new is written so flush old - if (!(type & INST_P_READ) && s_recWriteP == 0) - MOV32MtoR(EAX, (uptr)&s_writeP); - MOV32MtoR(x86temp, (uptr)&s_TotalVUCycles); - - if (!(type & INST_P_READ)) - { - if (s_recWriteP == 0) - { - OR32RtoR(EAX, EAX); - pjmp = JS8(0); - MOV32MtoR(EAX, SuperVUGetVIAddr(REG_P, 0)); - MOV32RtoM(SuperVUGetVIAddr(REG_P, 1), EAX); - x86SetJ8(pjmp); - } - else if (s_needFlush & 2) - { - MOV32MtoR(EAX, SuperVUGetVIAddr(REG_P, 0)); - MOV32RtoM(SuperVUGetVIAddr(REG_P, 1), EAX); - s_needFlush &= ~2; - } - } - - // write new P - assert(s_pCurInst->pqcycles > 1); - ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); - MOV32RtoM((uptr)&s_writeP, x86temp); - s_needFlush |= 2; - - s_recWriteP = s_pCurInst->info.cycle + s_pCurInst->pqcycles; - - _freeX86reg(x86temp); - } - - // waitq - if (ptr[0] == 0x800003bf) SuperVUFlush(0, 1); - // waitp - if (ptr[0] == 0x800007bf) SuperVUFlush(1, 1); - -#ifdef PCSX2_DEVBUILD - if (regs[1].VIread & regs[0].VIwrite & ~((1 << REG_Q) | (1 << REG_P) | (1 << REG_VF0_FLAG) | (1 << REG_ACC_FLAG))) - { - Console::Notice("*PCSX2*: Warning, VI write to the same reg %x in both lower/upper cycle %x", params regs[1].VIread & regs[0].VIwrite, s_pCurBlock->startpc); - } -#endif - - u32 modewrite = 0; - if (vfwrite[1] >= 0 && xmmregs[vfwrite[1]].inuse && xmmregs[vfwrite[1]].type == XMMTYPE_VFREG && xmmregs[vfwrite[1]].reg == regs[1].VFwrite) - modewrite = xmmregs[vfwrite[1]].mode & MODE_WRITE; - - VU->code = ptr[1]; - s_vuInfo = SetCachedRegs(1, vuxyz); - - if (vfwrite[1] >= 0) - { - assert(regs[1].VFwrite > 0); - - if (vfwrite[0] == vfwrite[1]) - { - //Console::WriteLn("*PCSX2*: Warning, VF write to the same reg in both lower/upper cycle %x", params s_pCurBlock->startpc); - } - - if (vfread0[0] == vfwrite[1] || vfread1[0] == vfwrite[1]) - { - assert(regs[0].VFread0 == regs[1].VFwrite || regs[0].VFread1 == regs[1].VFwrite); - assert(vfflush[0] >= 0); - if (modewrite) - { - SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[regs[1].VFwrite], (x86SSERegType)vfwrite[1]); - } - vfregstore = 1; - } - } - - if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; - if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; - - recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); - _clearNeededXMMregs(); - _clearNeededX86regs(); - - // necessary because status can be set by both upper and lower - if (regs[1].VIwrite & (1 << REG_STATUS_FLAG)) - { - assert(pStatusWrite != 0); - s_PrevStatusWrite = pStatusWrite; - } - - VU->code = ptr[0]; - s_vuInfo = SetCachedRegs(0, vuxyz); - - if (vfregstore) - { - // load - SSE_MOVAPS_M128_to_XMM(vfflush[0], (uptr)&VU->VF[regs[1].VFwrite]); - - assert(xmmregs[vfwrite[1]].mode & MODE_WRITE); - - // replace with vfflush - if (_Fs_ == regs[1].VFwrite) - { - s_vuInfo &= ~PROCESS_EE_SET_S(0xf); - s_vuInfo |= PROCESS_EE_SET_S(vfflush[0]); - } - if (_Ft_ == regs[1].VFwrite) - { - s_vuInfo &= ~PROCESS_EE_SET_T(0xf); - s_vuInfo |= PROCESS_EE_SET_T(vfflush[0]); - } - - xmmregs[vfflush[0]].mode |= MODE_NOFLUSH | MODE_WRITE; // so that lower inst doesn't flush - } - - // notify vuinsts that upper inst is a fmac - if (regs[1].pipe == VUPIPE_FMAC) - s_vuInfo |= PROCESS_VU_SET_FMAC(); - - if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; - if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; - -#ifdef SUPERVU_VIBRANCHDELAY - if (type & INST_CACHE_VI) - { - assert(vicached >= 0); - int cachedreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), vicached, MODE_READ); - MOV32RtoM((uptr)&s_VIBranchDelay, cachedreg); - } -#endif - - // check if inst before branch and the write is the same as the read in the branch (wipeout) -// int oldreg=0; -// if( pc == s_pCurBlock->endpc-16 ) { -// itinst2 = itinst; ++itinst2; -// if( itinst2->regs[0].pipe == VUPIPE_BRANCH && (itinst->regs[0].VIwrite&itinst2->regs[0].VIread) ) { -// -// CALLFunc((u32)branchfn); -// assert( itinst->regs[0].VIwrite & 0xffff ); -// Console::WriteLn("vi write before branch"); -// for(s_CacheVIReg = 0; s_CacheVIReg < 16; ++s_CacheVIReg) { -// if( itinst->regs[0].VIwrite & (1<endpc-8 && s_CacheVIReg >= 0 ) { -// assert( s_CacheVIX86 > 0 && x86regs[s_CacheVIX86].inuse && x86regs[s_CacheVIX86].reg == s_CacheVIReg && x86regs[s_CacheVIX86].type == X86TYPE_VITEMP ); -// -// oldreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), s_CacheVIReg, MODE_READ); -// x86regs[s_CacheVIX86].needed = 1; -// assert( x86regs[oldreg].mode & MODE_WRITE ); -// -// x86regs[s_CacheVIX86].type = X86TYPE_VI|(s_vu?X86TYPE_VU1:0); -// x86regs[oldreg].type = X86TYPE_VITEMP; -// } - - recVU_LOWER_OPCODE[ VU->code >> 25 ](VU, s_vuInfo); - -// if( pc == s_pCurBlock->endpc-8 && s_CacheVIReg >= 0 ) { -// // revert -// x86regs[s_CacheVIX86].inuse = 0; -// x86regs[oldreg].type = X86TYPE_VI|(s_vu?X86TYPE_VU1:0); -// } - - _clearNeededXMMregs(); - _clearNeededX86regs(); - } - - // clip is always written so ok - if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_CLIP_FLAG)) - { - assert(pClipWrite != 0); - s_PrevClipWrite = pClipWrite; - } - - if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_STATUS_FLAG)) - { - assert(pStatusWrite != 0); - s_PrevStatusWrite = pStatusWrite; - } - - if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_MAC_FLAG)) - { - assert(pMACWrite != 0); - s_PrevMACWrite = pMACWrite; - } -} - -/////////////////////////////////// -// Super VU Recompilation Tables // -/////////////////////////////////// - -void recVUMI_BranchHandle() -{ - int bpc = _recbranchAddr(VU->code); - int curjump = 0; - - if (s_pCurInst->type & INST_BRANCH_DELAY) - { - assert((branch&0x17) != 0x10 && (branch&0x17) != 4); // no jump handlig for now - - if ((branch & 0x7) == 3) - { - // previous was a direct jump - curjump = 1; - } - else if (branch & 1) curjump = 2; - } - - assert(s_JumpX86 > 0); - - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); - MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR - s_pCurBlock->pChildJumps[curjump] = (u32*)x86Ptr - 1; - - if (!(s_pCurInst->type & INST_BRANCH_DELAY)) - { - j8Ptr[1] = JMP8(0); - x86SetJ8(j8Ptr[ 0 ]); - - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), pc + 8); - MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR - s_pCurBlock->pChildJumps[curjump+1] = (u32*)x86Ptr - 1; - - x86SetJ8(j8Ptr[ 1 ]); - } - else - x86SetJ8(j8Ptr[ 0 ]); - - branch |= 1; -} - -// supervu specific insts -void recVUMI_IBQ_prep() -{ - int isreg, itreg; - - if (_Is_ == 0) - { -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) - { - itreg = -1; - } - else -#endif - { - itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); - } - - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - - if (itreg >= 0) - { - CMP16ItoR(itreg, 0); - } - else CMP16ItoM(SuperVUGetVIAddr(_It_, 1), 0); - } - else if (_It_ == 0) - { -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - - if (isreg >= 0) - { - CMP16ItoR(isreg, 0); - } - else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); - - } - else - { - _addNeededX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_); - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) - { - itreg = -1; - - if (isreg <= 0) - { - // allocate fsreg - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ | MODE_WRITE); - MOV32MtoR(isreg, SuperVUGetVIAddr(_Is_, 1)); - } - else - isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - } - else -#endif - { - itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); - } - - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - - if (isreg >= 0) - { - if (itreg >= 0) - { - CMP16RtoR(isreg, itreg); - } - else CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); - } - else if (itreg >= 0) - { - CMP16MtoR(itreg, SuperVUGetVIAddr(_Is_, 1)); - } - else - { - isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); - } - } -} - -void recVUMI_IBEQ(VURegs* vuu, s32 info) -{ - recVUMI_IBQ_prep(); - j8Ptr[ 0 ] = JNE8(0); - recVUMI_BranchHandle(); -} - -void recVUMI_IBGEZ(VURegs* vuu, s32 info) -{ - int isreg; - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - - if (isreg >= 0) - { - TEST16RtoR(isreg, isreg); - j8Ptr[ 0 ] = JS8(0); - } - else - { - CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); - j8Ptr[ 0 ] = JL8(0); - } - - recVUMI_BranchHandle(); -} - -void recVUMI_IBGTZ(VURegs* vuu, s32 info) -{ - int isreg; - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - - if (isreg >= 0) - { - CMP16ItoR(isreg, 0); - j8Ptr[ 0 ] = JLE8(0); - } - else - { - CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); - j8Ptr[ 0 ] = JLE8(0); - } - recVUMI_BranchHandle(); -} - -void recVUMI_IBLEZ(VURegs* vuu, s32 info) -{ - int isreg; - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - - if (isreg >= 0) - { - CMP16ItoR(isreg, 0); - j8Ptr[ 0 ] = JG8(0); - } - else - { - CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); - j8Ptr[ 0 ] = JG8(0); - } - recVUMI_BranchHandle(); -} - -void recVUMI_IBLTZ(VURegs* vuu, s32 info) -{ - int isreg; - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - -#ifdef SUPERVU_VIBRANCHDELAY - if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) - { - isreg = -1; - } - else -#endif - { - isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - } - - if (isreg >= 0) - { - TEST16RtoR(isreg, isreg); - j8Ptr[ 0 ] = JNS8(0); - } - else - { - CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); - j8Ptr[ 0 ] = JGE8(0); - } - recVUMI_BranchHandle(); -} - -void recVUMI_IBNE(VURegs* vuu, s32 info) -{ - recVUMI_IBQ_prep(); - j8Ptr[ 0 ] = JE8(0); - recVUMI_BranchHandle(); -} - -void recVUMI_B(VURegs* vuu, s32 info) -{ - // supervu will take care of the rest - int bpc = _recbranchAddr(VU->code); - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); - - // loops to self, so check condition - if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) - { - SuperVUTestVU0Condition(0); - } - - if (s_pCurBlock->blocks.size() > 1) - { - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - MOV32ItoR(s_JumpX86, 1); - s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; - s_UnconditionalDelay = 1; - } - - branch |= 3; -} - -void recVUMI_BAL(VURegs* vuu, s32 info) -{ - int bpc = _recbranchAddr(VU->code); - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) - MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); - - // loops to self, so check condition - if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) - { - SuperVUTestVU0Condition(0); - } - - if (_It_) - { - _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); - MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); - } - - if (s_pCurBlock->blocks.size() > 1) - { - s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); - MOV32ItoR(s_JumpX86, 1); - s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; - s_UnconditionalDelay = 1; - } - - branch |= 3; -} - -void recVUMI_JR(VURegs* vuu, s32 info) -{ - int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - LEA32RStoR(EAX, isreg, 3); - - //Mask the address to something valid - if (vuu == &VU0) - AND32ItoR(EAX, 0xfff); - else - AND32ItoR(EAX, 0x3fff); - - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); - - if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) - { - PUSH32I(s_vu); - PUSH32R(EAX); - } - branch |= 0x10; // 0x08 is reserved -} - -void recVUMI_JALR(VURegs* vuu, s32 info) -{ - _addNeededX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_); - - int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - LEA32RStoR(EAX, isreg, 3); - - //Mask the address to something valid - if (vuu == &VU0) - AND32ItoR(EAX, 0xfff); - else - AND32ItoR(EAX, 0x3fff); - - if (_It_) - { - _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); - MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); - } - - if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); - - if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) - { - PUSH32I(s_vu); - PUSH32R(EAX); - } - - branch |= 4; -} - -#ifdef PCSX2_DEVBUILD -void vu1xgkick(u32* pMem, u32 addr) -{ - assert(addr < 0x4000); -#ifdef SUPERVU_COUNT - StopSVUCounter(); -#endif - - GSGIFTRANSFER1(pMem, addr); - -#ifdef SUPERVU_COUNT - StartSVUCounter(); -#endif -} -#endif - -void recVUMI_XGKICK_(VURegs *VU) -{ - assert(s_XGKICKReg > 0 && x86regs[s_XGKICKReg].inuse && x86regs[s_XGKICKReg].type == X86TYPE_VITEMP); - - x86regs[s_XGKICKReg].inuse = 0; // so free doesn't flush - _freeX86regs(); - _freeXMMregs(); - - PUSH32R(s_XGKICKReg); - PUSH32I((uptr)VU->Mem); - - //CALLFunc((u32)countfn); - - if (mtgsThread != NULL) - { - CALLFunc((uptr)VU1XGKICK_MTGSTransfer); - ADD32ItoR(ESP, 8); - } - else - { -#ifdef PCSX2_DEVBUILD - CALLFunc((uptr)vu1xgkick); - ADD32ItoR(ESP, 8); -#else - CALLFunc((uptr)GSgifTransfer1); -#endif - } - psHu32(GIF_STAT) &= ~(GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) - s_ScheduleXGKICK = 0; -} - -void recVUMI_XGKICK(VURegs *VU, int info) -{ - if (s_ScheduleXGKICK) - { - // second xgkick, so launch the first - recVUMI_XGKICK_(VU); - } - - int isreg = _allocX86reg(X86ARG2, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); - _freeX86reg(isreg); // flush - x86regs[isreg].inuse = 1; - x86regs[isreg].type = X86TYPE_VITEMP; - x86regs[isreg].needed = 1; - x86regs[isreg].mode = MODE_WRITE | MODE_READ; - SHL32ItoR(isreg, 4); - AND32ItoR(isreg, 0x3fff); - s_XGKICKReg = isreg; - psHu32(GIF_STAT) |= (GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) - - - if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) - { - recVUMI_XGKICK_(VU); - } - else - { - // Erementar Gerad hack. - // Corrects the color of some graphics on a game that has somewhat scrambled graphics either way, and only works with ZeroGS. Not even ZZOgl. :) - if (g_VUGameFixes & VUFIX_XGKICKDELAY2) - s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc - pc) / 8); - else - s_ScheduleXGKICK = 2; - } -} - -void recVU_UPPER_FD_00(VURegs* VU, s32 info); -void recVU_UPPER_FD_01(VURegs* VU, s32 info); -void recVU_UPPER_FD_10(VURegs* VU, s32 info); -void recVU_UPPER_FD_11(VURegs* VU, s32 info); -void recVULowerOP(VURegs* VU, s32 info); -void recVULowerOP_T3_00(VURegs* VU, s32 info); -void recVULowerOP_T3_01(VURegs* VU, s32 info); -void recVULowerOP_T3_10(VURegs* VU, s32 info); -void recVULowerOP_T3_11(VURegs* VU, s32 info); -void recVUunknown(VURegs* VU, s32 info); - -void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info) = -{ - recVUMI_LQ , recVUMI_SQ , recVUunknown , recVUunknown, - recVUMI_ILW , recVUMI_ISW , recVUunknown , recVUunknown, - recVUMI_IADDIU, recVUMI_ISUBIU, recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_FCEQ , recVUMI_FCSET , recVUMI_FCAND, recVUMI_FCOR, /* 0x10 */ - recVUMI_FSEQ , recVUMI_FSSET , recVUMI_FSAND, recVUMI_FSOR, - recVUMI_FMEQ , recVUunknown , recVUMI_FMAND, recVUMI_FMOR, - recVUMI_FCGET , recVUunknown , recVUunknown , recVUunknown, - recVUMI_B , recVUMI_BAL , recVUunknown , recVUunknown, /* 0x20 */ - recVUMI_JR , recVUMI_JALR , recVUunknown , recVUunknown, - recVUMI_IBEQ , recVUMI_IBNE , recVUunknown , recVUunknown, - recVUMI_IBLTZ , recVUMI_IBGTZ , recVUMI_IBLEZ, recVUMI_IBGEZ, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVULowerOP , recVUunknown , recVUunknown , recVUunknown, /* 0x40*/ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x50 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x60 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x70 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, -}; - -void (*recVULowerOP_T3_00_OPCODE[32])(VURegs* VU, s32 info) = -{ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_MOVE , recVUMI_LQI , recVUMI_DIV , recVUMI_MTIR, - recVUMI_RNEXT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_MFP , recVUMI_XTOP , recVUMI_XGKICK, - recVUMI_ESADD , recVUMI_EATANxy, recVUMI_ESQRT, recVUMI_ESIN, -}; - -void (*recVULowerOP_T3_01_OPCODE[32])(VURegs* VU, s32 info) = -{ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_MR32 , recVUMI_SQI , recVUMI_SQRT , recVUMI_MFIR, - recVUMI_RGET , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUMI_XITOP, recVUunknown, - recVUMI_ERSADD, recVUMI_EATANxz, recVUMI_ERSQRT, recVUMI_EATAN, -}; - -void (*recVULowerOP_T3_10_OPCODE[32])(VURegs* VU, s32 info) = -{ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_LQD , recVUMI_RSQRT, recVUMI_ILWR, - recVUMI_RINIT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_ELENG , recVUMI_ESUM , recVUMI_ERCPR, recVUMI_EEXP, -}; - -void (*recVULowerOP_T3_11_OPCODE[32])(VURegs* VU, s32 info) = -{ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUMI_SQD , recVUMI_WAITQ, recVUMI_ISWR, - recVUMI_RXOR , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_ERLENG, recVUunknown , recVUMI_WAITP, recVUunknown, -}; - -void (*recVULowerOP_OPCODE[64])(VURegs* VU, s32 info) = -{ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x20 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUMI_IADD , recVUMI_ISUB , recVUMI_IADDI, recVUunknown, /* 0x30 */ - recVUMI_IAND , recVUMI_IOR , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVULowerOP_T3_00, recVULowerOP_T3_01, recVULowerOP_T3_10, recVULowerOP_T3_11, -}; - -void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info) = -{ - recVUMI_ADDx , recVUMI_ADDy , recVUMI_ADDz , recVUMI_ADDw, - recVUMI_SUBx , recVUMI_SUBy , recVUMI_SUBz , recVUMI_SUBw, - recVUMI_MADDx , recVUMI_MADDy , recVUMI_MADDz , recVUMI_MADDw, - recVUMI_MSUBx , recVUMI_MSUBy , recVUMI_MSUBz , recVUMI_MSUBw, - recVUMI_MAXx , recVUMI_MAXy , recVUMI_MAXz , recVUMI_MAXw, /* 0x10 */ - recVUMI_MINIx , recVUMI_MINIy , recVUMI_MINIz , recVUMI_MINIw, - recVUMI_MULx , recVUMI_MULy , recVUMI_MULz , recVUMI_MULw, - recVUMI_MULq , recVUMI_MAXi , recVUMI_MULi , recVUMI_MINIi, - recVUMI_ADDq , recVUMI_MADDq , recVUMI_ADDi , recVUMI_MADDi, /* 0x20 */ - recVUMI_SUBq , recVUMI_MSUBq , recVUMI_SUBi , recVUMI_MSUBi, - recVUMI_ADD , recVUMI_MADD , recVUMI_MUL , recVUMI_MAX, - recVUMI_SUB , recVUMI_MSUB , recVUMI_OPMSUB, recVUMI_MINI, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVUunknown , recVUunknown , recVUunknown , recVUunknown, - recVU_UPPER_FD_00, recVU_UPPER_FD_01, recVU_UPPER_FD_10, recVU_UPPER_FD_11, -}; - -void (*recVU_UPPER_FD_00_TABLE[32])(VURegs* VU, s32 info) = -{ - recVUMI_ADDAx, recVUMI_SUBAx , recVUMI_MADDAx, recVUMI_MSUBAx, - recVUMI_ITOF0, recVUMI_FTOI0, recVUMI_MULAx , recVUMI_MULAq , - recVUMI_ADDAq, recVUMI_SUBAq, recVUMI_ADDA , recVUMI_SUBA , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_01_TABLE[32])(VURegs* VU, s32 info) = -{ - recVUMI_ADDAy , recVUMI_SUBAy , recVUMI_MADDAy, recVUMI_MSUBAy, - recVUMI_ITOF4 , recVUMI_FTOI4 , recVUMI_MULAy , recVUMI_ABS , - recVUMI_MADDAq, recVUMI_MSUBAq, recVUMI_MADDA , recVUMI_MSUBA , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_10_TABLE[32])(VURegs* VU, s32 info) = -{ - recVUMI_ADDAz , recVUMI_SUBAz , recVUMI_MADDAz, recVUMI_MSUBAz, - recVUMI_ITOF12, recVUMI_FTOI12, recVUMI_MULAz , recVUMI_MULAi , - recVUMI_ADDAi, recVUMI_SUBAi , recVUMI_MULA , recVUMI_OPMULA, - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void (*recVU_UPPER_FD_11_TABLE[32])(VURegs* VU, s32 info) = -{ - recVUMI_ADDAw , recVUMI_SUBAw , recVUMI_MADDAw, recVUMI_MSUBAw, - recVUMI_ITOF15, recVUMI_FTOI15, recVUMI_MULAw , recVUMI_CLIP , - recVUMI_MADDAi, recVUMI_MSUBAi, recVUunknown , recVUMI_NOP , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , - recVUunknown , recVUunknown , recVUunknown , recVUunknown , -}; - -void recVU_UPPER_FD_00(VURegs* VU, s32 info) -{ - recVU_UPPER_FD_00_TABLE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVU_UPPER_FD_01(VURegs* VU, s32 info) -{ - recVU_UPPER_FD_01_TABLE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVU_UPPER_FD_10(VURegs* VU, s32 info) -{ - recVU_UPPER_FD_10_TABLE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVU_UPPER_FD_11(VURegs* VU, s32 info) -{ - recVU_UPPER_FD_11_TABLE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVULowerOP(VURegs* VU, s32 info) -{ - recVULowerOP_OPCODE[ VU->code & 0x3f ](VU, info); -} - -void recVULowerOP_T3_00(VURegs* VU, s32 info) -{ - recVULowerOP_T3_00_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVULowerOP_T3_01(VURegs* VU, s32 info) -{ - recVULowerOP_T3_01_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVULowerOP_T3_10(VURegs* VU, s32 info) -{ - recVULowerOP_T3_10_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVULowerOP_T3_11(VURegs* VU, s32 info) -{ - recVULowerOP_T3_11_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); -} - -void recVUunknown(VURegs* VU, s32 info) -{ - Console::Notice("Unknown SVU micromode opcode called"); -} +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +// Super VU recompiler - author: zerofrog(@gmail.com) + +#include "PrecompiledHeader.h" + +#include +#include +#include +#include +#include + +#ifndef _WIN32 +#include +#endif + +#include "Common.h" + +#include "GS.h" +#include "R5900.h" +#include "VU.h" +#include "iR5900.h" + +#include "sVU_zerorec.h" +#include "SamplProf.h" +#include "NakedAsm.h" + +using namespace std; + +// temporary externs +extern void iDumpVU0Registers(); +extern void iDumpVU1Registers(); + +// SuperVURec optimization options, uncomment only for debugging purposes +#define SUPERVU_CACHING // vu programs are saved and queried via memcompare (should be no reason to disable this) +#define SUPERVU_WRITEBACKS // don't flush the writebacks after every block +#define SUPERVU_X86CACHING // use x86reg caching (faster) (not really. rather lots slower :p (rama) ) +#define SUPERVU_VIBRANCHDELAY // when integers are modified right before a branch that uses the integer, + // the old integer value is used in the branch, fixes kh2 + +#define SUPERVU_PROPAGATEFLAGS // the correct behavior of VUs, for some reason superman breaks gfx with it on... + +// registers won't be flushed at block boundaries (faster) (nothing noticable speed-wise, causes SPS in Ratchet and clank (Nneeve) ) +#ifndef _DEBUG +//#define SUPERVU_INTERCACHING +#endif + +#define SUPERVU_CHECKCONDITION 0 // has to be 0!! + +#define VU_EXESIZE 0x00800000 + +#define _Imm11_ (s32)( (vucode & 0x400) ? (0xfffffc00 | (vucode & 0x3ff)) : (vucode & 0x3ff) ) +#define _UImm11_ (s32)(vucode & 0x7ff) + +#define _Ft_ ((VU->code >> 16) & 0x1F) // The rt part of the instruction register +#define _Fs_ ((VU->code >> 11) & 0x1F) // The rd part of the instruction register +#define _Fd_ ((VU->code >> 6) & 0x1F) // The sa part of the instruction register +#define _It_ (_Ft_ & 15) +#define _Is_ (_Fs_ & 15) +#define _Id_ (_Fd_ & 15) + +static const u32 QWaitTimes[] = { 6, 12 }; +static const u32 PWaitTimes[] = { 53, 43, 28, 23, 17, 11, 10 }; + +static u32 s_vuInfo; // info passed into rec insts + +static const u32 s_MemSize[2] = {VU0_MEMSIZE, VU1_MEMSIZE}; +static u8* s_recVUMem = NULL, *s_recVUPtr = NULL; + +// tables which are defined at the bottom of this massive file. +extern void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info); +extern void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info); + +#define INST_Q_READ 0x0001 // flush Q +#define INST_P_READ 0x0002 // flush P +#define INST_BRANCH_DELAY 0x0004 +#define INST_CLIP_WRITE 0x0040 // inst writes CLIP in the future +#define INST_STATUS_WRITE 0x0080 +#define INST_MAC_WRITE 0x0100 +#define INST_Q_WRITE 0x0200 +#define INST_CACHE_VI 0x0400 // write old vi value to s_VIBranchDelay + +// Let's tempt fate by defining two different constants with almost identical names +#define INST_DUMMY_ 0x8000 +#define INST_DUMMY 0x83c0 + +#define VFFREE_INVALID0 0x80000000 // (vffree[i]&0xf) is invalid + +//#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); ++(it)) + +#ifdef _DEBUG +u32 s_vucount = 0; + +static u32 g_vu1lastrec = 0, skipparent = -1; +static u32 s_svulast = 0, s_vufnheader; +static u32 badaddrs[][2] = {0, 0xffff}; +#endif + +union VURecRegs +{ + struct + { + u16 reg; + u16 type; + }; + u32 id; +}; + +#define SUPERVU_XGKICKDELAY 1 // yes this is needed as default (wipeout) + +class VuBaseBlock; + +struct VuFunctionHeader +{ + struct RANGE + { + RANGE() : pmem(NULL) {} + + u16 start, size; + void* pmem; // all the mem + }; + + VuFunctionHeader() : pprogfunc(NULL), startpc(0xffffffff) {} + ~VuFunctionHeader() + { + for (vector::iterator it = ranges.begin(); it != ranges.end(); ++it) + { + free(it->pmem); + } + } + + // returns true if the checksum for the current mem is the same as this fn + bool IsSame(void* pmem); + + u32 startpc; + void* pprogfunc; + + vector ranges; +}; + +struct VuBlockHeader +{ + VuBaseBlock* pblock; + u32 delay; +}; + +// one vu inst (lower and upper) +class VuInstruction +{ + public: + VuInstruction() + { + memzero_obj(*this); + nParentPc = -1; + vicached = -1; + } + + int nParentPc; // used for syncing with flag writes, -1 for no parent + + _vuopinfo info; + + _VURegsNum regs[2]; // [0] - lower, [1] - upper + u32 livevars[2]; // live variables right before this inst, [0] - inst, [1] - float + u32 addvars[2]; // live variables to add + u32 usedvars[2]; // set if var is used in the future including vars used in this inst + u32 keepvars[2]; + u16 pqcycles; // the number of cycles to stall if function writes to the regs + u16 type; // INST_ + + u32 pClipWrite, pMACWrite, pStatusWrite; // addrs to write the flags + u32 vffree[2]; + s8 vfwrite[2], vfread0[2], vfread1[2], vfacc[2]; + s8 vfflush[2]; // extra flush regs + s8 vicached; // if >= 0, then use the cached integer s_VIBranchDelay + VuInstruction *pPrevInst; + + int SetCachedRegs(int upper, u32 vuxyz); + void Recompile(list::iterator& itinst, u32 vuxyz); +}; + +enum BlockType +{ + BLOCKTYPE_EOP = 0x01, // at least one of the children of the block contains eop (or the block itself) + BLOCKTYPE_FUNCTION = 0x02, + BLOCKTYPE_HASEOP = 0x04, // last inst of block is an eop + BLOCKTYPE_MACFLAGS = 0x08, + BLOCKTYPE_ANALYZED = 0x40, + BLOCKTYPE_IGNORE = 0x80, // special for recursive fns + BLOCKTYPE_ANALYZEDPARENT = 0x100 +}; + +// base block used when recompiling +class VuBaseBlock +{ + public: + typedef list LISTBLOCKS; + + VuBaseBlock(); + + // returns true if the leads to a EOP (ALL VU blocks must ret true) + void AssignVFRegs(); + void AssignVIRegs(int parent); + + list::iterator GetInstIterAtPc(int instpc); + void GetInstsAtPc(int instpc, list& listinsts); + + void Recompile(); + + u16 type; // BLOCKTYPE_ + u16 id; + u16 startpc; + u16 endpc; // first inst not in block + void* pcode; // x86 code pointer + void* pendcode; // end of the x86 code pointer + int cycles; + list insts; + list parents; + LISTBLOCKS blocks; // blocks branches to + u32* pChildJumps[4]; // addrs that need to be filled with the children's start addrs + // if highest bit is set, addr needs to be relational + u32 vuxyz; // corresponding bit is set if reg's xyz channels are used only + u32 vuxy; // corresponding bit is set if reg's xyz channels are used only + + _xmmregs startregs[iREGCNT_XMM], endregs[iREGCNT_XMM]; + int nStartx86, nEndx86; // indices into s_vecRegArray + + int allocX86Regs; + int prevFlagsOutOfBlock; +}; + +struct WRITEBACK +{ + WRITEBACK() : nParentPc(0), cycle(0) //, pStatusWrite(NULL), pMACWrite(NULL) + { + viwrite[0] = viwrite[1] = 0; + viread[0] = viread[1] = 0; + } + + void InitInst(VuInstruction* pinst, int cycle) const + { + u32 write = viwrite[0] | viwrite[1]; + pinst->type = ((write & (1 << REG_CLIP_FLAG)) ? INST_CLIP_WRITE : 0) | + ((write & (1 << REG_MAC_FLAG)) ? INST_MAC_WRITE : 0) | + ((write & (1 << REG_STATUS_FLAG)) ? INST_STATUS_WRITE : 0) | + ((write & (1 << REG_Q)) ? INST_Q_WRITE : 0); + pinst->nParentPc = nParentPc; + pinst->info.cycle = cycle; + for (int i = 0; i < 2; ++i) + { + pinst->regs[i].VIwrite = viwrite[i]; + pinst->regs[i].VIread = viread[i]; + } + } + + static int SortWritebacks(const WRITEBACK& w1, const WRITEBACK& w2) + { + return w1.cycle < w2.cycle; + } + + int nParentPc; + int cycle; + u32 viwrite[2]; + u32 viread[2]; +}; + +struct VUPIPELINES +{ + fmacPipe fmac[8]; + fdivPipe fdiv; + efuPipe efu; + ialuPipe ialu[8]; + list< WRITEBACK > listWritebacks; +}; + +VuBaseBlock::VuBaseBlock() +{ + type = 0; + endpc = 0; + cycles = 0; + pcode = NULL; + id = 0; + memzero_obj(pChildJumps); + memzero_obj(startregs); + memzero_obj(endregs); + allocX86Regs = nStartx86 = nEndx86 = -1; + prevFlagsOutOfBlock = 0; +} + +#define SUPERVU_STACKSIZE 0x1000 + +static list s_listVUHeaders[2]; +static list* s_plistCachedHeaders[2] = {NULL, NULL}; +static VuFunctionHeader** recVUHeaders[2] = {NULL, NULL}; +static VuBlockHeader* recVUBlocks[2] = {NULL, NULL}; +static u8* recVUStack = NULL, *recVUStackPtr = NULL; +static vector<_x86regs> s_vecRegArray(128); + +static VURegs* VU = NULL; +static list s_listBlocks; +static u32 s_vu = 0; +static u32 s_UnconditionalDelay = 0; // 1 if there are two sequential branches and the last is unconditional +static u32 g_nLastBlockExecuted = 0; + +static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex); +static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const VUPIPELINES& pipes); +static void SuperVUInitLiveness(VuBaseBlock* pblock); +static void SuperVULivenessAnalysis(); +static void SuperVUEliminateDeadCode(); +static void SuperVUAssignRegs(); + +//void SuperVUFreeXMMreg(int xmmreg, int xmmtype, int reg); +#define SuperVUFreeXMMreg 0&& +void SuperVUFreeXMMregs(u32* livevars); + +static u32* SuperVUStaticAlloc(u32 size); +static void SuperVURecompile(); + +// allocate VU resources +void SuperVUAlloc(int vuindex) +{ + // The old -1 crap has been depreciated on this function. Please + // specify either 0 or 1, thanks. + jASSUME(vuindex >= 0); + + // upper 4 bits must be zero! + if (s_recVUMem == NULL) + { + // upper 4 bits must be zero! + // Changed "first try base" to 0xf1e0000, since 0x0c000000 liked to fail a lot. (cottonvibes) + s_recVUMem = SysMmapEx(0xf1e0000, VU_EXESIZE, 0x10000000, "SuperVUAlloc"); + + if (s_recVUMem == NULL) + { + throw Exception::OutOfMemory( + fmt_string("SuperVU Error > failed to allocate recompiler memory (addr: 0x%x)", (u32)s_recVUMem) + ); + } + + ProfilerRegisterSource("VURec", s_recVUMem, VU_EXESIZE); + + if (recVUStack == NULL) recVUStack = new u8[SUPERVU_STACKSIZE * 4]; + } + + if (vuindex >= 0) + { + jASSUME(s_recVUMem != NULL); + + if (recVUHeaders[vuindex] == NULL) + recVUHeaders[vuindex] = new VuFunctionHeader* [s_MemSize[vuindex] / 8]; + if (recVUBlocks[vuindex] == NULL) + recVUBlocks[vuindex] = new VuBlockHeader[s_MemSize[vuindex] / 8]; + if (s_plistCachedHeaders[vuindex] == NULL) + s_plistCachedHeaders[vuindex] = new list[s_MemSize[vuindex] / 8]; + } +} + +void DestroyCachedHeaders(int vuindex, int j) +{ + list::iterator it = s_plistCachedHeaders[vuindex][j].begin(); + + while (it != s_plistCachedHeaders[vuindex][j].end()) + { + delete *it; + it++; + } + + s_plistCachedHeaders[vuindex][j].clear(); +} + +void DestroyVUHeaders(int vuindex) +{ + list::iterator it = s_listVUHeaders[vuindex].begin(); + + while (it != s_listVUHeaders[vuindex].end()) + { + delete *it; + it++; + } + + s_listVUHeaders[vuindex].clear(); +} + +// destroy VU resources +void SuperVUDestroy(int vuindex) +{ + list::iterator it; + + if (vuindex < 0) + { + SuperVUDestroy(0); + SuperVUDestroy(1); + ProfilerTerminateSource("VURec"); + SafeSysMunmap(s_recVUMem, VU_EXESIZE); + safe_delete_array(recVUStack); + } + else + { + safe_delete_array(recVUHeaders[vuindex]); + safe_delete_array(recVUBlocks[vuindex]); + + if (s_plistCachedHeaders[vuindex] != NULL) + { + for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) + { + DestroyCachedHeaders(vuindex, j); + } + safe_delete_array(s_plistCachedHeaders[vuindex]); + } + DestroyVUHeaders(vuindex); + } +} + +// reset VU +void SuperVUReset(int vuindex) +{ +#ifdef _DEBUG + s_vucount = 0; +#endif + + if (s_recVUMem == NULL) + return; + + //jASSUME( s_recVUMem != NULL ); + + if (vuindex < 0) + { + DbgCon::Status("SuperVU reset > Resetting recompiler memory and structures."); + + // Does this cause problems on VU recompiler resets? It could, if the VU works like + // the EE used to, and actually tries to re-enter the recBlock after issuing a clear. (air) + + //memset_8<0xcd, VU_EXESIZE>(s_recVUMem); + memzero_ptr(recVUStack); + + s_recVUPtr = s_recVUMem; + } + else + { + DbgCon::Status("SuperVU reset [VU%d] > Resetting the recs and junk", params vuindex); + list::iterator it; + if (recVUHeaders[vuindex]) memset(recVUHeaders[vuindex], 0, sizeof(VuFunctionHeader*) * (s_MemSize[vuindex] / 8)); + if (recVUBlocks[vuindex]) memset(recVUBlocks[vuindex], 0, sizeof(VuBlockHeader) * (s_MemSize[vuindex] / 8)); + + if (s_plistCachedHeaders[vuindex] != NULL) + { + for (u32 j = 0; j < s_MemSize[vuindex] / 8; ++j) + { + DestroyCachedHeaders(vuindex, j); + } + } + DestroyVUHeaders(vuindex); + } +} + +// clear the block and any joining blocks +void __fastcall SuperVUClear(u32 startpc, u32 size, int vuindex) +{ + vector::iterator itrange; + list::iterator it = s_listVUHeaders[vuindex].begin(); + u32 endpc = startpc + (size + (8 - (size & 7))); // Adding this code to ensure size is always a multiple of 8, it can be simplified to startpc+size if size is always a multiple of 8 (cottonvibes) + while (it != s_listVUHeaders[vuindex].end()) + { + + // for every fn, check if it has code in the range + for(itrange = (*it)->ranges.begin(); itrange != (*it)->ranges.end(); itrange++) + { + if (startpc < (u32)itrange->start + itrange->size && itrange->start < endpc) + break; + } + + if (itrange != (*it)->ranges.end()) + { + recVUHeaders[vuindex][(*it)->startpc/8] = NULL; +#ifdef SUPERVU_CACHING + list* plist = &s_plistCachedHeaders[vuindex][(*it)->startpc/8]; + plist->push_back(*it); + if (plist->size() > 30) + { + // list is too big, delete + //Console::Notice("Performance warning: deleting cached VU program!"); + delete plist->front(); + plist->pop_front(); + } + it = s_listVUHeaders[vuindex].erase(it); +#else + delete *it; + it = s_listVUHeaders[vuindex].erase(it); +#endif + } + else ++it; + } +} + +static VuFunctionHeader* s_pFnHeader = NULL; +static VuBaseBlock* s_pCurBlock = NULL; +static VuInstruction* s_pCurInst = NULL; +static u32 s_StatusRead = 0, s_MACRead = 0, s_ClipRead = 0; // read addrs +static u32 s_PrevStatusWrite = 0, s_PrevMACWrite = 0, s_PrevClipWrite = 0, s_PrevIWrite = 0; +static u32 s_WriteToReadQ = 0; + +static u32 s_VIBranchDelay = 0; //Value of register to use in a vi branch delayed situation + + +u32 s_TotalVUCycles; // total cycles since start of program execution + + +u32 SuperVUGetVIAddr(int reg, int read) +{ + assert(s_pCurInst != NULL); + + switch (reg) + { + case REG_STATUS_FLAG: + { + u32 addr = (read == 2) ? s_PrevStatusWrite : (read ? s_StatusRead : s_pCurInst->pStatusWrite); + assert(!read || addr != 0); + return addr; + } + case REG_MAC_FLAG: + { + u32 addr = (read == 2) ? s_PrevMACWrite : (read ? s_MACRead : s_pCurInst->pMACWrite); + return addr; + } + case REG_CLIP_FLAG: + { + u32 addr = (read == 2) ? s_PrevClipWrite : (read ? s_ClipRead : s_pCurInst->pClipWrite); + assert(!read || addr != 0); + return addr; + } + case REG_Q: + return (read || s_WriteToReadQ) ? (uptr)&VU->VI[REG_Q] : (uptr)&VU->q; + case REG_P: + return read ? (uptr)&VU->VI[REG_P] : (uptr)&VU->p; + case REG_I: + return s_PrevIWrite; + } + +#ifdef SUPERVU_VIBRANCHDELAY + if ((read != 0) && (s_pCurInst->regs[0].pipe == VUPIPE_BRANCH) && (s_pCurInst->vicached >= 0) && (s_pCurInst->vicached == reg)) + { + return (uptr)&s_VIBranchDelay; // test for branch delays + } +#endif + + return (uptr)&VU->VI[reg]; +} + +void SuperVUDumpBlock(list& blocks, int vuindex) +{ + FILE *f; + char str[256]; + u32 *mem; + u32 i; + + Path::CreateDirectory("dumps"); + string filename(Path::Combine("dumps", fmt_string("svu%cdump%.4X.txt", s_vu ? '0' : '1', s_pFnHeader->startpc))); + //Console::WriteLn( "dump1 %x => %s", params s_pFnHeader->startpc, filename ); + + f = fopen(filename.c_str(), "w"); + + fprintf(f, "Format: upper_inst lower_inst\ntype f:vf_live_vars vf_used_vars i:vi_live_vars vi_used_vars inst_cycle pq_inst\n"); + fprintf(f, "Type: %.2x - qread, %.2x - pread, %.2x - clip_write, %.2x - status_write\n" + "%.2x - mac_write, %.2x -qflush\n", + INST_Q_READ, INST_P_READ, INST_CLIP_WRITE, INST_STATUS_WRITE, INST_MAC_WRITE, INST_Q_WRITE); + fprintf(f, "XMM: Upper: read0 read1 write acc temp; Lower: read0 read1 write acc temp\n\n"); + + list::iterator itblock; + list::iterator itinst; + VuBaseBlock::LISTBLOCKS::iterator itchild; + + for(itblock = blocks.begin(); itblock != blocks.end(); itblock++) + { + fprintf(f, "block:%c %x-%x; children: ", ((*itblock)->type&BLOCKTYPE_HASEOP) ? '*' : ' ', + (*itblock)->startpc, (*itblock)->endpc - 8); + + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { + fprintf(f, "%x ", (*itchild)->startpc); + } + fprintf(f, "; vuxyz = %x, vuxy = %x\n", (*itblock)->vuxyz&(*itblock)->insts.front().usedvars[1], + (*itblock)->vuxy&(*itblock)->insts.front().usedvars[1]); + + itinst = (*itblock)->insts.begin(); + i = (*itblock)->startpc; + while (itinst != (*itblock)->insts.end()) + { + assert(i <= (*itblock)->endpc); + if (itinst->type & INST_DUMMY) + { + if (itinst->nParentPc >= 0 && !(itinst->type&INST_DUMMY_)) + { + // search for the parent + fprintf(f, "writeback 0x%x (%x)\n", itinst->type, itinst->nParentPc); + } + } + else + { + mem = (u32*) & VU->Micro[i]; + char* pstr = disVU1MicroUF(mem[1], i + 4); + fprintf(f, "%.4x: %-40s", i, pstr); + if (mem[1] & 0x80000000) fprintf(f, " I=%f(%.8x)\n", *(float*)mem, mem[0]); + else fprintf(f, "%s\n", disVU1MicroLF(mem[0], i)); + i += 8; + } + + ++itinst; + } + + fprintf(f, "\n"); + + _x86regs* pregs; + if ((*itblock)->nStartx86 >= 0 || (*itblock)->nEndx86 >= 0) + { + fprintf(f, "X86: AX CX DX BX SP BP SI DI\n"); + } + + if ((*itblock)->nStartx86 >= 0) + { + pregs = &s_vecRegArray[(*itblock)->nStartx86]; + fprintf(f, "STR: "); + for (i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse) + fprintf(f, "%.2d ", pregs[i].reg); + else + fprintf(f, "-1 "); + } + fprintf(f, "\n"); + } + + if ((*itblock)->nEndx86 >= 0) + { + fprintf(f, "END: "); + pregs = &s_vecRegArray[(*itblock)->nEndx86]; + for (i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse) + fprintf(f, "%.2d ", pregs[i].reg); + else + fprintf(f, "-1 "); + } + fprintf(f, "\n"); + } + + itinst = (*itblock)->insts.begin(); + for (i = (*itblock)->startpc; i < (*itblock)->endpc; ++itinst) + { + + if (itinst->type & INST_DUMMY) + { + } + else + { + sprintf(str, "%.4x:%x f:%.8x_%.8x", i, itinst->type, itinst->livevars[1], itinst->usedvars[1]); + fprintf(f, "%-46s i:%.8x_%.8x c:%d pq:%d\n", str, + itinst->livevars[0], itinst->usedvars[0], (int)itinst->info.cycle, (int)itinst->pqcycles); + + sprintf(str, "XMM r0:%d r1:%d w:%d a:%d t:%x;", + itinst->vfread0[1], itinst->vfread1[1], itinst->vfwrite[1], itinst->vfacc[1], itinst->vffree[1]); + fprintf(f, "%-46s r0:%d r1:%d w:%d a:%d t:%x\n", str, + itinst->vfread0[0], itinst->vfread1[0], itinst->vfwrite[0], itinst->vfacc[0], itinst->vffree[0]); + i += 8; + } + } + +#ifdef __LINUX__ + // dump the asm + if ((*itblock)->pcode != NULL) + { + char command[255]; + FILE* fasm = fopen("mydump1", "wb"); + //Console::WriteLn("writing: %x, %x", params (*itblock)->startpc, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode); + fwrite((*itblock)->pcode, 1, (uptr)(*itblock)->pendcode - (uptr)(*itblock)->pcode, fasm); + fclose(fasm); + sprintf(command, "objdump -D --target=binary --architecture=i386 -M intel mydump1 > tempdump"); + system(command); + fasm = fopen("tempdump", "r"); + // read all of it and write it to f + fseek(fasm, 0, SEEK_END); + vector vbuffer(ftell(fasm)); + fseek(fasm, 0, SEEK_SET); + fread(&vbuffer[0], vbuffer.size(), 1, fasm); + + fprintf(f, "\n\n"); + fwrite(&vbuffer[0], vbuffer.size(), 1, f); + fclose(fasm); + } +#endif + + fprintf(f, "\n---------------\n"); + } + + fclose(f); +} + +// uncomment to count svu exec time +//#define SUPERVU_COUNT + +// Private methods +void* SuperVUGetProgram(u32 startpc, int vuindex) +{ + assert(startpc < s_MemSize[vuindex]); + assert((startpc % 8) == 0); + assert(recVUHeaders[vuindex] != NULL); + VuFunctionHeader** pheader = &recVUHeaders[vuindex][startpc/8]; + + if (*pheader == NULL) + { +#ifdef _DEBUG +// if( vuindex ) VU1.VI[REG_TPC].UL = startpc; +// else VU0.VI[REG_TPC].UL = startpc; +// __Log("VU: %x\n", startpc); +// iDumpVU1Registers(); +// vudump |= 2; +#endif + + // measure run time + //QueryPerformanceCounter(&svubase); + +#ifdef SUPERVU_CACHING + void* pmem = (vuindex & 1) ? VU1.Micro : VU0.Micro; + // check if program exists in cache + list::iterator it; + for(it = s_plistCachedHeaders[vuindex][startpc/8].begin(); it != s_plistCachedHeaders[vuindex][startpc/8].end(); it++) + { + if ((*it)->IsSame(pmem)) + { + // found, transfer to regular lists + void* pfn = (*it)->pprogfunc; + recVUHeaders[vuindex][startpc/8] = *it; + s_listVUHeaders[vuindex].push_back(*it); + s_plistCachedHeaders[vuindex][startpc/8].erase(it); + return pfn; + } + } +#endif + + *pheader = SuperVURecompileProgram(startpc, vuindex); + + if (*pheader == NULL) + { + assert(s_TotalVUCycles > 0); + if (vuindex) + VU1.VI[REG_TPC].UL = startpc; + else + VU0.VI[REG_TPC].UL = startpc; + + return (void*)SuperVUEndProgram; + } + + //QueryPerformanceCounter(&svufinal); + //svutime += (u32)(svufinal.QuadPart-svubase.QuadPart); + + assert((*pheader)->pprogfunc != NULL); + } + //else assert( (*pheader)->IsSame((vuindex&1) ? VU1.Micro : VU0.Micro) ); + + assert((*pheader)->startpc == startpc); + + return (*pheader)->pprogfunc; +} + +bool VuFunctionHeader::IsSame(void* pmem) +{ +#ifdef SUPERVU_CACHING + vector::iterator it; + for(it = ranges.begin(); it != ranges.end(); it++) + { + if (memcmp_mmx((u8*)pmem + it->start, it->pmem, it->size)) + return false; + } +#endif + return true; +} + +list::iterator VuBaseBlock::GetInstIterAtPc(int instpc) +{ + assert(instpc >= 0); + + u32 curpc = startpc; + list::iterator it; + for (it = insts.begin(); it != insts.end(); ++it) + { + if (it->type & INST_DUMMY) continue; + if (curpc == instpc) break; + curpc += 8; + } + + if (it != insts.end()) return it; + + assert(0); + return insts.begin(); +} + +void VuBaseBlock::GetInstsAtPc(int instpc, list& listinsts) +{ + assert(instpc >= 0); + + listinsts.clear(); + + u32 curpc = startpc; + list::iterator it; + for (it = insts.begin(); it != insts.end(); ++it) + { + if (it->type & INST_DUMMY) continue; + if (curpc == instpc) break; + curpc += 8; + } + + if (it != insts.end()) + { + listinsts.push_back(&(*it)); + return; + } + + // look for the pc in other blocks + for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) + { + if (*itblock == this) continue; + + if (instpc >= (*itblock)->startpc && instpc < (*itblock)->endpc) + { + listinsts.push_back(&(*(*itblock)->GetInstIterAtPc(instpc))); + } + } + + assert(listinsts.size() > 0); +} + +static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) +{ + assert(vuindex < 2); + assert(s_recVUPtr != NULL); + //Console::WriteLn("svu%c rec: %x", params '0'+vuindex, startpc); + + // if recPtr reached the mem limit reset whole mem + if (((uptr)s_recVUPtr - (uptr)s_recVUMem) >= VU_EXESIZE - 0x40000) + { + //Console::WriteLn("SuperVU reset mem"); + SuperVUReset(0); + SuperVUReset(1); + SuperVUReset(-1); + if (s_TotalVUCycles > 0) + { + // already executing, so return NULL + return NULL; + } + } + + list::iterator itblock; + + s_vu = vuindex; + VU = s_vu ? &VU1 : &VU0; + s_pFnHeader = new VuFunctionHeader(); + s_listVUHeaders[vuindex].push_back(s_pFnHeader); + s_pFnHeader->startpc = startpc; + + memset(recVUBlocks[s_vu], 0, sizeof(VuBlockHeader) * (s_MemSize[s_vu] / 8)); + + // analyze the global graph + s_listBlocks.clear(); + VUPIPELINES pipes; + memzero_obj(pipes.fmac); + memzero_obj(pipes.fdiv); + memzero_obj(pipes.efu); + memzero_obj(pipes.ialu); + SuperVUBuildBlocks(NULL, startpc, pipes); + + // fill parents + VuBaseBlock::LISTBLOCKS::iterator itchild; + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { + (*itchild)->parents.push_back(*itblock); + } + + //(*itblock)->type &= ~(BLOCKTYPE_IGNORE|BLOCKTYPE_ANALYZED); + } + + assert(s_listBlocks.front()->startpc == startpc); + s_listBlocks.front()->type |= BLOCKTYPE_FUNCTION; + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + SuperVUInitLiveness(*itblock); + } + + SuperVULivenessAnalysis(); + SuperVUEliminateDeadCode(); + SuperVUAssignRegs(); + +#ifdef _DEBUG + if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); +#endif + + // code generation + x86SetPtr(s_recVUPtr); + branch = 0; + + SuperVURecompile(); + + s_recVUPtr = x86Ptr; + + // set the function's range + VuFunctionHeader::RANGE r; + s_pFnHeader->ranges.reserve(s_listBlocks.size()); + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + r.start = (*itblock)->startpc; + r.size = (*itblock)->endpc - (*itblock)->startpc; +#ifdef SUPERVU_CACHING + //memxor_mmx(r.checksum, &VU->Micro[r.start], r.size); + r.pmem = malloc(r.size); + memcpy_fast(r.pmem, &VU->Micro[r.start], r.size); +#endif + s_pFnHeader->ranges.push_back(r); + } + +#if defined(_DEBUG) && defined(__LINUX__) + // dump at the end to capture the actual code + if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); +#endif + + // destroy + for (list::iterator itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); ++itblock) + { + delete *itblock; + } + s_listBlocks.clear(); + + assert(s_recVUPtr < s_recVUMem + VU_EXESIZE); + + return s_pFnHeader; +} + +static int _recbranchAddr(u32 vucode) +{ + s32 bpc = pc + (_Imm11_ << 3); + /* + if ( bpc < 0 ) { + Console::WriteLn("zerorec branch warning: bpc < 0 ( %x ); Using unsigned imm11", params bpc); + bpc = pc + (_UImm11_ << 3); + }*/ + bpc &= (s_MemSize[s_vu] - 1); + + return bpc; +} + +// return inst that flushes everything +static VuInstruction SuperVUFlushInst() +{ + VuInstruction inst; + // don't need to raed q/p + inst.type = INST_DUMMY_;//|INST_Q_READ|INST_P_READ; + return inst; +} + +void SuperVUAddWritebacks(VuBaseBlock* pblock, const list& listWritebacks) +{ +#ifdef SUPERVU_WRITEBACKS + // regardless of repetition, add the pipes (for selfloops) + list::const_iterator itwriteback = listWritebacks.begin(); + list::iterator itinst = pblock->insts.begin(), itinst2; + + while (itwriteback != listWritebacks.end()) + { + if (itinst != pblock->insts.end() && (itinst->info.cycle < itwriteback->cycle || (itinst->type&INST_DUMMY))) + { + ++itinst; + continue; + } + + itinst2 = pblock->insts.insert(itinst, VuInstruction()); + itwriteback->InitInst(&(*itinst2), vucycle); + ++itwriteback; + } +#endif +} + +#ifdef SUPERVU_VIBRANCHDELAY +static VuInstruction* getDelayInst(VuInstruction* pInst) +{ + // check for the N cycle branch delay + // example of 2 cycles delay (monster house) : + // sqi vi05 + // sqi vi05 + // ibeq vi05, vi03 + // The ibeq should read the vi05 before the first sqi + + //more info: + + // iaddiu vi01, 0, 1 + // ibeq vi01, 0 <- reads vi01 before the iaddiu + + // iaddiu vi01, 0, 1 + // iaddiu vi01, vi01, 1 + // iaddiu vi01, vi01, 1 + // ibeq vi01, 0 <- reads vi01 before the last two iaddiu's (so the value read is 1) + + // ilw vi02, addr + // iaddiu vi01, 0, 1 + // ibeq vi01, vi02 <- reads current values of both vi01 and vi02 because the branch instruction stalls + + int delay = 1; + VuInstruction* pDelayInst = NULL; + VuInstruction* pTargetInst = pInst->pPrevInst; + while (1) + { + if (pTargetInst != NULL + && pTargetInst->info.cycle + delay == pInst->info.cycle + && (pTargetInst->regs[0].pipe == VUPIPE_IALU || pTargetInst->regs[0].pipe == VUPIPE_FMAC) + && ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) + && (delay == 1 || ((pTargetInst->regs[0].VIwrite & pInst->regs[0].VIread) & 0xffff) == ((pTargetInst->regs[0].VIwrite & pInst->pPrevInst->regs[0].VIread) & 0xffff)) + && !(pTargetInst->regs[0].VIread&((1 << REG_STATUS_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_CLIP_FLAG)))) + { + pDelayInst = pTargetInst; + pTargetInst = pTargetInst->pPrevInst; + delay++; + if (delay == 5) //maximum delay is 4 (length of the pipeline) + { + DevCon::WriteLn("supervu: cycle branch delay maximum (4) is reached"); + break; + } + } + else break; + } + if (delay > 1) DevCon::WriteLn("supervu: %d cycle branch delay detected: %x %x", params delay - 1, pc, s_pFnHeader->startpc); + return pDelayInst; +} +#endif + +static VuBaseBlock* SuperVUBuildBlocks(VuBaseBlock* parent, u32 startpc, const VUPIPELINES& pipes) +{ + // check if block already exists + //Console::WriteLn("startpc %x", params startpc); + startpc &= (s_vu ? 0x3fff : 0xfff); + VuBlockHeader* pbh = &recVUBlocks[s_vu][startpc/8]; + + if (pbh->pblock != NULL) + { + + VuBaseBlock* pblock = pbh->pblock; + list::iterator itinst; + + if (pblock->startpc == startpc) + { + SuperVUAddWritebacks(pblock, pipes.listWritebacks); + return pblock; + } + + // have to divide the blocks, pnewblock is first block + assert(startpc > pblock->startpc); + assert(startpc < pblock->endpc); + + u32 dummyinst = (startpc - pblock->startpc) >> 3; + + // count inst non-dummy insts + itinst = pblock->insts.begin(); + int cycleoff = 0; + + while (dummyinst > 0) + { + if (itinst->type & INST_DUMMY) + ++itinst; + else + { + cycleoff = itinst->info.cycle; + ++itinst; + --dummyinst; + } + } + + // NOTE: still leaves insts with their writebacks in different blocks + while (itinst->type & INST_DUMMY) + ++itinst; + + // the difference in cycles between dummy insts (naruto utlimate ninja) + int cyclediff = 0; + if (parent == pblock) + cyclediff = itinst->info.cycle - cycleoff; + cycleoff = itinst->info.cycle; + + // new block + VuBaseBlock* pnewblock = new VuBaseBlock(); + s_listBlocks.push_back(pnewblock); + + pnewblock->startpc = startpc; + pnewblock->endpc = pblock->endpc; + pnewblock->cycles = pblock->cycles - cycleoff + cyclediff; + + pnewblock->blocks.splice(pnewblock->blocks.end(), pblock->blocks); + pnewblock->insts.splice(pnewblock->insts.end(), pblock->insts, itinst, pblock->insts.end()); + pnewblock->type = pblock->type; + + // any writebacks in the next 3 cycles also belong to original block +// for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); ) { +// if( (itinst->type & INST_DUMMY) && itinst->nParentPc >= 0 && itinst->nParentPc < (int)startpc ) { +// +// if( !(itinst->type & INST_Q_WRITE) ) +// pblock->insts.push_back(*itinst); +// itinst = pnewblock->insts.erase(itinst); +// continue; +// } +// +// ++itinst; +// } + + pbh = &recVUBlocks[s_vu][startpc/8]; + for (u32 inst = startpc; inst < pblock->endpc; inst += 8) + { + if (pbh->pblock == pblock) + pbh->pblock = pnewblock; + ++pbh; + } + + for(itinst = pnewblock->insts.begin(); itinst != pnewblock->insts.end(); itinst++) + { + itinst->info.cycle -= cycleoff; + } + + SuperVUAddWritebacks(pnewblock, pipes.listWritebacks); + + // old block + pblock->blocks.push_back(pnewblock); + pblock->endpc = startpc; + pblock->cycles = cycleoff; + pblock->type &= BLOCKTYPE_MACFLAGS; + //pblock->insts.push_back(SuperVUFlushInst()); //don't need + + return pnewblock; + } + + VuBaseBlock* pblock = new VuBaseBlock(); + s_listBlocks.push_back(pblock); + + int i = 0; + branch = 0; + pc = startpc; + pblock->startpc = startpc; + + // clear stalls (might be a prob) + memcpy(VU->fmac, pipes.fmac, sizeof(pipes.fmac)); + memcpy(&VU->fdiv, &pipes.fdiv, sizeof(pipes.fdiv)); + memcpy(&VU->efu, &pipes.efu, sizeof(pipes.efu)); + memcpy(VU->ialu, pipes.ialu, sizeof(pipes.ialu)); +// memset(VU->fmac, 0, sizeof(VU->fmac)); +// memset(&VU->fdiv, 0, sizeof(VU->fdiv)); +// memset(&VU->efu, 0, sizeof(VU->efu)); + + vucycle = 0; + + u8 macflags = 0; + + list< WRITEBACK > listWritebacks; + list< WRITEBACK >::iterator itwriteback; + list::iterator itinst; + u32 hasSecondBranch = 0; + u32 needFullStatusFlag = 0; + +#ifdef SUPERVU_WRITEBACKS + listWritebacks = pipes.listWritebacks; +#endif + + // first analysis pass for status flags + while (1) + { + u32* ptr = (u32*) & VU->Micro[pc]; + pc += 8; + int prevbranch = branch; + + if (ptr[1] & 0x40000000) + branch = 1; + + if (!(ptr[1] & 0x80000000)) // not I + { + switch (ptr[0] >> 25) + { + case 0x24: // jr + case 0x25: // jalr + case 0x20: // B + case 0x21: // BAL + case 0x28: // IBEQ + case 0x2f: // IBGEZ + case 0x2d: // IBGTZ + case 0x2e: // IBLEZ + case 0x2c: // IBLTZ + case 0x29: // IBNE + branch = 1; + break; + + case 0x14: // fseq + case 0x17: // fsor + //needFullStatusFlag = 2; + break; + + case 0x16: // fsand + if ((ptr[0]&0xc0)) + { + // sometimes full sticky bits are needed (simple series 2000 - oane chapara) + //Console::WriteLn("needSticky: %x-%x", params s_pFnHeader->startpc, startpc); + needFullStatusFlag = 2; + } + break; + } + } + + if (prevbranch) + break; + + if (pc >= s_MemSize[s_vu]) + { + Console::Error("inf vu0 prog %x", params startpc); + break; + } + } + + // second full pass + pc = startpc; + branch = 0; + VuInstruction* pprevinst = NULL, *pinst = NULL; + + while (1) + { + + if (pc == s_MemSize[s_vu]) + { + branch |= 8; + break; + } + + if (!branch && pbh->pblock != NULL) + { + pblock->blocks.push_back(pbh->pblock); + break; + } + + int prevbranch = branch; + + if (!prevbranch) + { + pbh->pblock = pblock; + } + else assert(prevbranch || pbh->pblock == NULL); + + pblock->insts.push_back(VuInstruction()); + + pprevinst = pinst; + pinst = &pblock->insts.back(); + pinst->pPrevInst = pprevinst; + SuperVUAnalyzeOp(VU, &pinst->info, pinst->regs); + +#ifdef SUPERVU_VIBRANCHDELAY + if (pinst->regs[0].pipe == VUPIPE_BRANCH && pblock->insts.size() > 1) + { + + VuInstruction* pdelayinst = getDelayInst(pinst); + if (pdelayinst) + { + pdelayinst->type |= INST_CACHE_VI; + + // find the correct register + u32 mask = pdelayinst->regs[0].VIwrite & pinst->regs[0].VIread; + for (int i = 0; i < 16; ++i) + { + if (mask & (1 << i)) + { + pdelayinst->vicached = i; + break; + } + } + + pinst->vicached = pdelayinst->vicached; + } + } +#endif + + if (prevbranch) + { + if (pinst->regs[0].pipe == VUPIPE_BRANCH) + hasSecondBranch = 1; + pinst->type |= INST_BRANCH_DELAY; + } + + // check write back + for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end();) + { + if (pinst->info.cycle >= itwriteback->cycle) + { + itinst = pblock->insts.insert(--pblock->insts.end(), VuInstruction()); + itwriteback->InitInst(&(*itinst), pinst->info.cycle); + itwriteback = listWritebacks.erase(itwriteback); + } + else ++itwriteback; + } + + // add new writebacks + WRITEBACK w; + const u32 allflags = (1 << REG_CLIP_FLAG) | (1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG); + for (int j = 0; j < 2; ++j) w.viwrite[j] = pinst->regs[j].VIwrite & allflags; + + if (pinst->info.macflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_MAC_FLAG); + if (pinst->info.statusflag & VUOP_WRITE) w.viwrite[1] |= (1 << REG_STATUS_FLAG); + + if ((pinst->info.macflag | pinst->info.statusflag) & VUOP_READ) + macflags = 1; + if (pinst->regs[0].VIread & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) + macflags = 1; + +// if( pinst->regs[1].pipe == VUPIPE_FMAC && (pinst->regs[1].VFwrite==0&&!(pinst->regs[1].VIwrite&(1<regs[0].VIread |= (1<VIwrite |= lregs->VIwrite & (1<info.statusflag&VUOP_WRITE) && !(pinst->regs[0].VIwrite&(1 << REG_STATUS_FLAG))) && needFullStatusFlag) + { + // don't read if first inst + if (needFullStatusFlag == 1) + w.viread[1] |= (1 << REG_STATUS_FLAG); + else --needFullStatusFlag; + } + + for (int j = 0; j < 2; ++j) + { + w.viread[j] |= pinst->regs[j].VIread & allflags; + + if ((pinst->regs[j].VIread&(1 << REG_STATUS_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_STATUS_FLAG))) + { + // don't need the read anymore + pinst->regs[j].VIread &= ~(1 << REG_STATUS_FLAG); + } + if ((pinst->regs[j].VIread&(1 << REG_MAC_FLAG)) && (pinst->regs[j].VIwrite&(1 << REG_MAC_FLAG))) + { + // don't need the read anymore + pinst->regs[j].VIread &= ~(1 << REG_MAC_FLAG); + } + + pinst->regs[j].VIwrite &= ~allflags; + } + + if (pinst->info.macflag & VUOP_READ) w.viread[1] |= 1 << REG_MAC_FLAG; + if (pinst->info.statusflag & VUOP_READ) w.viread[1] |= 1 << REG_STATUS_FLAG; + + w.nParentPc = pc - 8; + w.cycle = pinst->info.cycle + 4; + listWritebacks.push_back(w); + } + + if (pinst->info.q&VUOP_READ) pinst->type |= INST_Q_READ; + if (pinst->info.p&VUOP_READ) pinst->type |= INST_P_READ; + + if (pinst->info.q&VUOP_WRITE) + { + pinst->pqcycles = QWaitTimes[pinst->info.pqinst] + 1; + + memset(&w, 0, sizeof(w)); + w.nParentPc = pc - 8; + w.cycle = pinst->info.cycle + pinst->pqcycles; + w.viwrite[0] = 1 << REG_Q; + listWritebacks.push_back(w); + } + if (pinst->info.p&VUOP_WRITE) + pinst->pqcycles = PWaitTimes[pinst->info.pqinst] + 1; + + if (prevbranch) + { + break; + } + + // make sure there is always a branch + // sensible soccer overflows on vu0, so increase the limit... + if ((s_vu == 1 && i >= 0x799) || (s_vu == 0 && i >= 0x201)) + { + Console::Error("VuRec base block doesn't terminate!"); + assert(0); + break; + } + + i++; + pbh++; + } + + if (macflags) + pblock->type |= BLOCKTYPE_MACFLAGS; + + pblock->endpc = pc; + u32 lastpc = pc; + + pblock->cycles = vucycle; + +#ifdef SUPERVU_WRITEBACKS + if (!branch || (branch&8)) +#endif + { + // flush writebacks + if (listWritebacks.size() > 0) + { + listWritebacks.sort(WRITEBACK::SortWritebacks); + for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) + { + if (itwriteback->viwrite[0] & (1 << REG_Q)) + { + // ignore all Q writebacks + continue; + } + + pblock->insts.push_back(VuInstruction()); + itwriteback->InitInst(&pblock->insts.back(), vucycle); + } + + listWritebacks.clear(); + } + } + + if (!branch) return pblock; + + if (branch & 8) + { + // what if also a jump? + pblock->type |= BLOCKTYPE_EOP | BLOCKTYPE_HASEOP; + + // add an instruction to flush p and q (if written) + pblock->insts.push_back(SuperVUFlushInst()); + return pblock; + } + + // it is a (cond) branch or a jump + u32 vucode = *(u32*)(VU->Micro + lastpc - 16); + int bpc = _recbranchAddr(vucode) - 8; + + VUPIPELINES newpipes; + memcpy(newpipes.fmac, VU->fmac, sizeof(newpipes.fmac)); + memcpy(&newpipes.fdiv, &VU->fdiv, sizeof(newpipes.fdiv)); + memcpy(&newpipes.efu, &VU->efu, sizeof(newpipes.efu)); + memcpy(newpipes.ialu, VU->ialu, sizeof(newpipes.ialu)); + + for (i = 0; i < 8; ++i) newpipes.fmac[i].sCycle -= vucycle; + newpipes.fdiv.sCycle -= vucycle; + newpipes.efu.sCycle -= vucycle; + for (i = 0; i < 8; ++i) newpipes.ialu[i].sCycle -= vucycle; + + if (listWritebacks.size() > 0) + { + // flush all when jumping, send down the pipe when in branching + bool bFlushWritebacks = (vucode >> 25) == 0x24 || (vucode >> 25) == 0x25;//||(vucode>>25)==0x20||(vucode>>25)==0x21; + + listWritebacks.sort(WRITEBACK::SortWritebacks); + for (itwriteback = listWritebacks.begin(); itwriteback != listWritebacks.end(); ++itwriteback) + { + if (itwriteback->viwrite[0] & (1 << REG_Q)) + { + // ignore all Q writebacks + continue; + } + + if (itwriteback->cycle < vucycle || bFlushWritebacks) + { + pblock->insts.push_back(VuInstruction()); + itwriteback->InitInst(&pblock->insts.back(), vucycle); + } + else + { + newpipes.listWritebacks.push_back(*itwriteback); + newpipes.listWritebacks.back().cycle -= vucycle; + } + } + } + + if (newpipes.listWritebacks.size() > 0) // other blocks might read the mac flags + pblock->type |= BLOCKTYPE_MACFLAGS; + + u32 firstbranch = vucode >> 25; + switch (firstbranch) + { + case 0x24: // jr + pblock->type |= BLOCKTYPE_EOP; // jump out of procedure, since not returning, set EOP + pblock->insts.push_back(SuperVUFlushInst()); + firstbranch = 0xff; //Non-Conditional Jump + break; + + case 0x25: // jalr + { + // linking, so will return to procedure + pblock->insts.push_back(SuperVUFlushInst()); + + VuBaseBlock* pjumpblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); + + pblock->blocks.push_back(pjumpblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } + case 0x20: // B + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); + + pblock->blocks.push_back(pbranchblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } + case 0x21: // BAL + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); + pblock->blocks.push_back(pbranchblock); + firstbranch = 0xff; //Non-Conditional Jump + break; + } + case 0x28: // IBEQ + case 0x2f: // IBGEZ + case 0x2d: // IBGTZ + case 0x2e: // IBLEZ + case 0x2c: // IBLTZ + case 0x29: // IBNE + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + assert(pblock != NULL); + pblock->blocks.push_back(pbranchblock); + + // if has a second branch that is B or BAL, skip this + u32 secondbranch = (*(u32*)(VU->Micro + lastpc - 8)) >> 25; + if (!hasSecondBranch || (secondbranch != 0x21 && secondbranch != 0x20)) + { + pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + } + + break; + } + default: + assert(pblock->blocks.size() == 1); + break; + } + + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + +#ifdef SUPERVU_VIBRANCHDELAY +///////////////////////////////////////////////////////////////////////////////////////////////////////////////////// +///// NOTE! This could still be a hack for KH2/GoW, but until we know how it properly works, this will do for now./// +///////////////////////////////////////////////////////////////////////////////////////////////////////////////////// + + if (hasSecondBranch && firstbranch != 0xff) //check the previous jump was conditional and there is a second branch + { +#else + if (hasSecondBranch) + { +#endif + + u32 vucode = *(u32*)(VU->Micro + lastpc - 8); + pc = lastpc; + int bpc = _recbranchAddr(vucode); + + switch (vucode >> 25) + { + case 0x24: // jr + Console::Error("svurec bad jr jump!"); + assert(0); + break; + + case 0x25: // jalr + { + Console::Error("svurec bad jalr jump!"); + assert(0); + break; + } + case 0x20: // B + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + + pblock->blocks.push_back(pbranchblock); + break; + } + case 0x21: // BAL + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // replace instead of pushing a new block + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + break; + } + case 0x28: // IBEQ + case 0x2f: // IBGEZ + case 0x2d: // IBGTZ + case 0x2e: // IBLEZ + case 0x2c: // IBLTZ + case 0x29: // IBNE + { + VuBaseBlock* pbranchblock = SuperVUBuildBlocks(pblock, bpc, newpipes); + + // update pblock since could have changed + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + + // only add the block if the previous branch doesn't include the next instruction (ie, if a direct jump) + if (firstbranch == 0x24 || firstbranch == 0x25 || firstbranch == 0x20 || firstbranch == 0x21) + { + pbranchblock = SuperVUBuildBlocks(pblock, lastpc, newpipes); + + pblock = recVUBlocks[s_vu][lastpc/8-2].pblock; + pblock->blocks.push_back(pbranchblock); + } + + break; + } + + jNO_DEFAULT; + } + } + + return recVUBlocks[s_vu][startpc/8].pblock; +} + +static void SuperVUInitLiveness(VuBaseBlock* pblock) +{ + list::iterator itinst, itnext; + + assert(pblock->insts.size() > 0); + + for (itinst = pblock->insts.begin(); itinst != pblock->insts.end(); ++itinst) + { + + if (itinst->type & INST_DUMMY_) + { + itinst->addvars[0] = itinst->addvars[1] = 0xffffffff; + itinst->livevars[0] = itinst->livevars[1] = 0xffffffff; + itinst->keepvars[0] = itinst->keepvars[1] = 0xffffffff; + itinst->usedvars[0] = itinst->usedvars[1] = 0; + } + else + { + itinst->addvars[0] = itinst->regs[0].VIread | itinst->regs[1].VIread; + itinst->addvars[1] = (itinst->regs[0].VFread0 ? (1 << itinst->regs[0].VFread0) : 0) | + (itinst->regs[0].VFread1 ? (1 << itinst->regs[0].VFread1) : 0) | + (itinst->regs[1].VFread0 ? (1 << itinst->regs[1].VFread0) : 0) | + (itinst->regs[1].VFread1 ? (1 << itinst->regs[1].VFread1) : 0); + + // vf0 is not handled by VFread + if (!itinst->regs[0].VFread0 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; + if (!itinst->regs[1].VFread0 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG))) itinst->addvars[1] |= 1; + if (!itinst->regs[0].VFread1 && (itinst->regs[0].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[0].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; + if (!itinst->regs[1].VFread1 && (itinst->regs[1].VIread & (1 << REG_VF0_FLAG)) && itinst->regs[1].VFr1xyzw != 0xff) itinst->addvars[1] |= 1; + + + u32 vfwrite = 0; + if (itinst->regs[0].VFwrite != 0) + { + if (itinst->regs[0].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[0].VFwrite; + else vfwrite |= 1 << itinst->regs[0].VFwrite; + } + if (itinst->regs[1].VFwrite != 0) + { + if (itinst->regs[1].VFwxyzw != 0xf) itinst->addvars[1] |= 1 << itinst->regs[1].VFwrite; + else vfwrite |= 1 << itinst->regs[1].VFwrite; + } + if ((itinst->regs[1].VIwrite & (1 << REG_ACC_FLAG)) && itinst->regs[1].VFwxyzw != 0xf) + itinst->addvars[1] |= 1 << REG_ACC_FLAG; + + u32 viwrite = (itinst->regs[0].VIwrite | itinst->regs[1].VIwrite); + + itinst->usedvars[0] = itinst->addvars[0] | viwrite; + itinst->usedvars[1] = itinst->addvars[1] | vfwrite; + +// itinst->addvars[0] &= ~viwrite; +// itinst->addvars[1] &= ~vfwrite; + itinst->keepvars[0] = ~viwrite; + itinst->keepvars[1] = ~vfwrite; + } + } + + itinst = --pblock->insts.end(); + while (itinst != pblock->insts.begin()) + { + itnext = itinst; + --itnext; + + itnext->usedvars[0] |= itinst->usedvars[0]; + itnext->usedvars[1] |= itinst->usedvars[1]; + + itinst = itnext; + } +} + +u32 COMPUTE_LIVE(u32 R, u32 K, u32 L) +{ + u32 live = R | ((L) & (K)); + // special process mac and status flags + // only propagate liveness if doesn't write to the flag + if (!(L&(1 << REG_STATUS_FLAG)) && !(K&(1 << REG_STATUS_FLAG))) + live &= ~(1 << REG_STATUS_FLAG); + if (!(L&(1 << REG_MAC_FLAG)) && !(K&(1 << REG_MAC_FLAG))) + live &= ~(1 << REG_MAC_FLAG); + return live;//|(1<::reverse_iterator itblock; + list::iterator itinst, itnext; + VuBaseBlock::LISTBLOCKS::iterator itchild; + + u32 livevars[2]; + + do + { + changed = FALSE; + for (itblock = s_listBlocks.rbegin(); itblock != s_listBlocks.rend(); ++itblock) + { + + u32 newlive; + VuBaseBlock* pb = *itblock; + + // the last inst relies on the neighbor's insts + itinst = --pb->insts.end(); + + if (pb->blocks.size() > 0) + { + livevars[0] = 0; + livevars[1] = 0; + for (itchild = pb->blocks.begin(); itchild != pb->blocks.end(); ++itchild) + { + VuInstruction& front = (*itchild)->insts.front(); + livevars[0] |= front.livevars[0]; + livevars[1] |= front.livevars[1]; + } + + newlive = COMPUTE_LIVE(itinst->addvars[0], itinst->keepvars[0], livevars[0]); + + // should propagate status flags whose parent insts are not in this block +// if( itinst->nParentPc >= 0 && (itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) +// newlive |= livevars[0]&((1<livevars[0] != newlive) + { + changed = TRUE; + itinst->livevars[0] = newlive; + } + + newlive = COMPUTE_LIVE(itinst->addvars[1], itinst->keepvars[1], livevars[1]); + if (itinst->livevars[1] != newlive) + { + changed = TRUE; + itinst->livevars[1] = newlive; + } + } + + while (itinst != pb->insts.begin()) + { + + itnext = itinst; + --itnext; + + newlive = COMPUTE_LIVE(itnext->addvars[0], itnext->keepvars[0], itinst->livevars[0]); + + // should propagate status flags whose parent insts are not in this block +// if( itnext->nParentPc >= 0 && (itnext->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) && !(itinst->type & (INST_STATUS_WRITE|INST_MAC_WRITE)) ) +// newlive |= itinst->livevars[0]&((1<livevars[0] != newlive) + { + changed = TRUE; + itnext->livevars[0] = newlive; + itnext->livevars[1] = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); + } + else + { + newlive = COMPUTE_LIVE(itnext->addvars[1], itnext->keepvars[1], itinst->livevars[1]); + if (itnext->livevars[1] != newlive) + { + changed = TRUE; + itnext->livevars[1] = newlive; + } + } + + itinst = itnext; + } + +// if( (livevars[0] | itinst->livevars[0]) != itinst->livevars[0] ) { +// changed = TRUE; +// itinst->livevars[0] |= livevars[0]; +// } +// if( (livevars[1] | itinst->livevars[1]) != itinst->livevars[1] ) { +// changed = TRUE; +// itinst->livevars[1] |= livevars[1]; +// } +// +// while( itinst != pb->insts.begin() ) { +// +// itnext = itinst; --itnext; +// if( (itnext->livevars[0] | (itinst->livevars[0] & itnext->keepvars[0])) != itnext->livevars[0] ) { +// changed = TRUE; +// itnext->livevars[0] |= itinst->livevars[0] & itnext->keepvars[0]; +// itnext->livevars[1] |= itinst->livevars[1] & itnext->keepvars[1]; +// } +// else if( (itnext->livevars[1] | (itinst->livevars[1] & itnext->keepvars[1])) != itnext->livevars[1] ) { +// changed = TRUE; +// itnext->livevars[1] |= itinst->livevars[1] & itnext->keepvars[1]; +// } +// +// itinst = itnext; +// } + } + + } + while (changed); +} + +static void SuperVUEliminateDeadCode() +{ + list::iterator itblock; + VuBaseBlock::LISTBLOCKS::iterator itchild; + list::iterator itinst, itnext; + list listParents; + list::iterator itparent; + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + +#ifdef _DEBUG + u32 startpc = (*itblock)->startpc; + u32 curpc = startpc; +#endif + + itnext = (*itblock)->insts.begin(); + itinst = itnext++; + while (itnext != (*itblock)->insts.end()) + { + if (itinst->type & (INST_CLIP_WRITE | INST_MAC_WRITE | INST_STATUS_WRITE)) + { + u32 live0 = itnext->livevars[0]; + if (itinst->nParentPc >= 0 && itnext->nParentPc >= 0 && itinst->nParentPc != itnext->nParentPc) // superman returns + { + // take the live vars from the next next inst + list::iterator itnextnext = itnext; + ++itnextnext; + if (itnextnext != (*itblock)->insts.end()) + { + live0 = itnextnext->livevars[0]; + } + } + + itinst->regs[0].VIwrite &= live0; + itinst->regs[1].VIwrite &= live0; + + u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; + + (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); + int removetype = 0; + + for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) + { + VuInstruction* parent = *itparent; + + if (viwrite & (1 << REG_CLIP_FLAG)) + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); + } + else + removetype |= INST_CLIP_WRITE; + + if (parent->info.macflag && (itinst->type & INST_MAC_WRITE)) + { + if (!(viwrite&(1 << REG_MAC_FLAG))) + { + //parent->info.macflag = 0; + // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); +#endif + // if VUPIPE_FMAC and destination is vf00, probably need to keep the mac flag + if (parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) + { + parent->regs[0].VIwrite |= ((1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= ((1 << REG_MAC_FLAG)); + } + else + removetype |= INST_MAC_WRITE; + } + else + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); + } + } + else removetype |= INST_MAC_WRITE; + + if (parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) + { + if (!(viwrite&(1 << REG_STATUS_FLAG))) + { + //parent->info.statusflag = 0; + // parent->regs[0].VIwrite &= ~(1<regs[1].VIwrite &= ~(1<regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); +#endif + if (parent->regs[1].pipe == VUPIPE_FMAC && (parent->regs[1].VFwrite == 0 && !(parent->regs[1].VIwrite&(1 << REG_ACC_FLAG)))) + { + parent->regs[0].VIwrite |= ((1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= ((1 << REG_STATUS_FLAG)); + } + else + removetype |= INST_STATUS_WRITE; + } + else + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); + } + } + else removetype |= INST_STATUS_WRITE; + } + + itinst->type &= ~removetype; + if (itinst->type == 0) + { + itnext = (*itblock)->insts.erase(itinst); + itinst = itnext++; + continue; + } + } +#ifdef _DEBUG + else + { + curpc += 8; + } +#endif + itinst = itnext; + ++itnext; + } + + if (itinst->type & INST_DUMMY) + { + // last inst with the children + u32 mask = 0; + for (itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); ++itchild) + { + mask |= (*itchild)->insts.front().livevars[0]; + } + itinst->regs[0].VIwrite &= mask; + itinst->regs[1].VIwrite &= mask; + u32 viwrite = itinst->regs[0].VIwrite | itinst->regs[1].VIwrite; + + if (itinst->nParentPc >= 0) + { + + (*itblock)->GetInstsAtPc(itinst->nParentPc, listParents); + int removetype = 0; + + for(itparent = listParents.begin(); itparent != listParents.end(); itparent++) + { + VuInstruction* parent = *itparent; + + if (viwrite & (1 << REG_CLIP_FLAG)) + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_CLIP_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_CLIP_FLAG)); + } + else removetype |= INST_CLIP_WRITE; + + if (parent->info.macflag && (itinst->type & INST_MAC_WRITE)) + { + if (!(viwrite&(1 << REG_MAC_FLAG))) + { + //parent->info.macflag = 0; +#ifndef SUPERVU_WRITEBACKS + assert(!(parent->regs[0].VIwrite & (1 << REG_MAC_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_MAC_FLAG))); +#endif + removetype |= INST_MAC_WRITE; + } + else + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_MAC_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_MAC_FLAG)); + } + } + else removetype |= INST_MAC_WRITE; + + if (parent->info.statusflag && (itinst->type & INST_STATUS_WRITE)) + { + if (!(viwrite&(1 << REG_STATUS_FLAG))) + { + //parent->info.statusflag = 0; +#ifndef SUPERVU_WRITEBACKS + assert(!(parent->regs[0].VIwrite & (1 << REG_STATUS_FLAG)) && !(parent->regs[1].VIwrite & (1 << REG_STATUS_FLAG))); +#endif + removetype |= INST_STATUS_WRITE; + } + else + { + parent->regs[0].VIwrite |= (itinst->regs[0].VIwrite & (1 << REG_STATUS_FLAG)); + parent->regs[1].VIwrite |= (itinst->regs[1].VIwrite & (1 << REG_STATUS_FLAG)); + } + } + else removetype |= INST_STATUS_WRITE; + } + + itinst->type &= ~removetype; + if (itinst->type == 0) + { + (*itblock)->insts.erase(itinst); + } + } + } + } +} + +void VuBaseBlock::AssignVFRegs() +{ + int i; + VuBaseBlock::LISTBLOCKS::iterator itchild; + list::iterator itblock; + list::iterator itinst, itnext, itinst2; + + // init the start regs + if (type & BLOCKTYPE_ANALYZED) return; // nothing changed + memcpy(xmmregs, startregs, sizeof(xmmregs)); + + if (type & BLOCKTYPE_ANALYZED) + { + // check if changed + for (i = 0; i < iREGCNT_XMM; ++i) + { + if (xmmregs[i].inuse != startregs[i].inuse) + break; + if (xmmregs[i].inuse && (xmmregs[i].reg != startregs[i].reg || xmmregs[i].type != startregs[i].type)) + break; + } + + if (i == iREGCNT_XMM) return; // nothing changed + } + + u8* oldX86 = x86Ptr; + + for(itinst = insts.begin(); itinst != insts.end(); itinst++) + { + + if (itinst->type & INST_DUMMY) continue; + + // reserve, go from upper to lower + int lastwrite = -1; + + for (i = 1; i >= 0; --i) + { + _VURegsNum* regs = itinst->regs + i; + + + // redo the counters so that the proper regs are released + for (int j = 0; j < iREGCNT_XMM; ++j) + { + if (xmmregs[j].inuse) + { + if (xmmregs[j].type == XMMTYPE_VFREG) + { + int count = 0; + itinst2 = itinst; + + if (i) + { + if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg) + { + itinst2 = insts.end(); + break; + } + else + { + ++count; + ++itinst2; + } + } + + while (itinst2 != insts.end()) + { + if (itinst2->regs[0].VFread0 == xmmregs[j].reg || itinst2->regs[0].VFread1 == xmmregs[j].reg || itinst2->regs[0].VFwrite == xmmregs[j].reg || + itinst2->regs[1].VFread0 == xmmregs[j].reg || itinst2->regs[1].VFread1 == xmmregs[j].reg || itinst2->regs[1].VFwrite == xmmregs[j].reg) + break; + + ++count; + ++itinst2; + } + xmmregs[j].counter = 1000 - count; + } + else + { + assert(xmmregs[j].type == XMMTYPE_ACC); + + int count = 0; + itinst2 = itinst; + + if (i) ++itinst2; // acc isn't used in lower insts + + while (itinst2 != insts.end()) + { + assert(!((itinst2->regs[0].VIread | itinst2->regs[0].VIwrite) & (1 << REG_ACC_FLAG))); + + if ((itinst2->regs[1].VIread | itinst2->regs[1].VIwrite) & (1 << REG_ACC_FLAG)) + break; + + ++count; + ++itinst2; + } + + xmmregs[j].counter = 1000 - count; + } + } + } + + if (regs->VFread0) _addNeededVFtoXMMreg(regs->VFread0); + if (regs->VFread1) _addNeededVFtoXMMreg(regs->VFread1); + if (regs->VFwrite) _addNeededVFtoXMMreg(regs->VFwrite); + if (regs->VIread & (1 << REG_ACC_FLAG)) _addNeededACCtoXMMreg(); + if (regs->VIread & (1 << REG_VF0_FLAG)) _addNeededVFtoXMMreg(0); + + // alloc + itinst->vfread0[i] = itinst->vfread1[i] = itinst->vfwrite[i] = itinst->vfacc[i] = -1; + itinst->vfflush[i] = -1; + + if (regs->VFread0) + itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, regs->VFread0, 0); + else if (regs->VIread & (1 << REG_VF0_FLAG)) + itinst->vfread0[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + + if (regs->VFread1) + itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, regs->VFread1, 0); + else if ((regs->VIread & (1 << REG_VF0_FLAG)) && regs->VFr1xyzw != 0xff) + itinst->vfread1[i] = _allocVFtoXMMreg(VU, -1, 0, 0); + + if (regs->VIread & (1 << REG_ACC_FLAG)) itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + + int reusereg = -1; // 0 - VFwrite, 1 - VFAcc + + if (regs->VFwrite) + { + assert(!(regs->VIwrite&(1 << REG_ACC_FLAG))); + + if (regs->VFwxyzw == 0xf) + { + itinst->vfwrite[i] = _checkXMMreg(XMMTYPE_VFREG, regs->VFwrite, 0); + if (itinst->vfwrite[i] < 0) reusereg = 0; + } + else + { + itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); + } + } + else if (regs->VIwrite & (1 << REG_ACC_FLAG)) + { + + if (regs->VFwxyzw == 0xf) + { + itinst->vfacc[i] = _checkXMMreg(XMMTYPE_ACC, 0, 0); + if (itinst->vfacc[i] < 0) reusereg = 1; + } + else + { + itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + } + } + + if (reusereg >= 0) + { + // reuse + itnext = itinst; + itnext++; + + u8 type = reusereg ? XMMTYPE_ACC : XMMTYPE_VFREG; + u8 reg = reusereg ? 0 : regs->VFwrite; + + if (itinst->vfacc[i] >= 0 && lastwrite != itinst->vfacc[i] && + (itnext == insts.end() || ((regs->VIread&(1 << REG_ACC_FLAG)) && (!(itnext->usedvars[0]&(1 << REG_ACC_FLAG)) || !(itnext->livevars[0]&(1 << REG_ACC_FLAG)))))) + { + + assert(reusereg == 0); + if (itnext == insts.end() || (itnext->livevars[0]&(1 << REG_ACC_FLAG))) _freeXMMreg(itinst->vfacc[i]); + xmmregs[itinst->vfacc[i]].inuse = 1; + xmmregs[itinst->vfacc[i]].reg = reg; + xmmregs[itinst->vfacc[i]].type = type; + xmmregs[itinst->vfacc[i]].mode = 0; + itinst->vfwrite[i] = itinst->vfacc[i]; + } + else if (itinst->vfread0[i] >= 0 && lastwrite != itinst->vfread0[i] && + (itnext == insts.end() || (regs->VFread0 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread0)) || !(itnext->livevars[1]&(1 << regs->VFread0)))))) + { + + if (itnext == insts.end() || (itnext->livevars[1]®s->VFread0)) _freeXMMreg(itinst->vfread0[i]); + + xmmregs[itinst->vfread0[i]].inuse = 1; + xmmregs[itinst->vfread0[i]].reg = reg; + xmmregs[itinst->vfread0[i]].type = type; + xmmregs[itinst->vfread0[i]].mode = 0; + + if (reusereg) + itinst->vfacc[i] = itinst->vfread0[i]; + else + itinst->vfwrite[i] = itinst->vfread0[i]; + } + else if (itinst->vfread1[i] >= 0 && lastwrite != itinst->vfread1[i] && + (itnext == insts.end() || (regs->VFread1 > 0 && (!(itnext->usedvars[1]&(1 << regs->VFread1)) || !(itnext->livevars[1]&(1 << regs->VFread1)))))) + { + + if (itnext == insts.end() || (itnext->livevars[1]®s->VFread1)) _freeXMMreg(itinst->vfread1[i]); + + xmmregs[itinst->vfread1[i]].inuse = 1; + xmmregs[itinst->vfread1[i]].reg = reg; + xmmregs[itinst->vfread1[i]].type = type; + xmmregs[itinst->vfread1[i]].mode = 0; + if (reusereg) + itinst->vfacc[i] = itinst->vfread1[i]; + else + itinst->vfwrite[i] = itinst->vfread1[i]; + } + else + { + if (reusereg) + itinst->vfacc[i] = _allocACCtoXMMreg(VU, -1, 0); + else + itinst->vfwrite[i] = _allocVFtoXMMreg(VU, -1, regs->VFwrite, 0); + } + } + + if (itinst->vfwrite[i] >= 0) lastwrite = itinst->vfwrite[i]; + else if (itinst->vfacc[i] >= 0) lastwrite = itinst->vfacc[i]; + + // always alloc at least 1 temp reg + int free0 = (i || regs->VFwrite || regs->VFread0 || regs->VFread1 || (regs->VIwrite & (1 << REG_ACC_FLAG)) || (regs->VIread & (1 << REG_VF0_FLAG))) + ? _allocTempXMMreg(XMMT_FPS, -1) : -1; + int free1 = 0, free2 = 0; + + if (i == 0 && itinst->vfwrite[1] >= 0 && (itinst->vfread0[0] == itinst->vfwrite[1] || itinst->vfread1[0] == itinst->vfwrite[1])) + { + itinst->vfflush[i] = _allocTempXMMreg(XMMT_FPS, -1); + } + + if (i == 1 && (regs->VIwrite & (1 << REG_CLIP_FLAG))) + { + // CLIP inst, need two extra regs + if (free0 < 0) free0 = _allocTempXMMreg(XMMT_FPS, -1); + + free1 = _allocTempXMMreg(XMMT_FPS, -1); + free2 = _allocTempXMMreg(XMMT_FPS, -1); + _freeXMMreg(free1); + _freeXMMreg(free2); + } + else if (regs->VIwrite & (1 << REG_P)) + { + // EFU inst, need extra reg + free1 = _allocTempXMMreg(XMMT_FPS, -1); + if (free0 == -1) free0 = free1; + _freeXMMreg(free1); + } + + if (itinst->vfflush[i] >= 0) _freeXMMreg(itinst->vfflush[i]); + if (free0 >= 0) _freeXMMreg(free0); + + itinst->vffree[i] = (free0 & 0xf) | (free1 << 8) | (free2 << 16); + if (free0 == -1) itinst->vffree[i] |= VFFREE_INVALID0; + + _clearNeededXMMregs(); + } + } + + assert(x86Ptr == oldX86); + u32 analyzechildren = !(type & BLOCKTYPE_ANALYZED); + type |= BLOCKTYPE_ANALYZED; + + //memset(endregs, 0, sizeof(endregs)); + + if (analyzechildren) + { + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { + (*itchild)->AssignVFRegs(); + } + } +} + +struct MARKOVBLANKET +{ + list parents; + list children; +}; + +static MARKOVBLANKET s_markov; + +void VuBaseBlock::AssignVIRegs(int parent) +{ + const int maxregs = 6; + + if (parent) + { + if ((type&BLOCKTYPE_ANALYZEDPARENT)) + return; + + type |= BLOCKTYPE_ANALYZEDPARENT; + s_markov.parents.push_back(this); + for (LISTBLOCKS::iterator it = blocks.begin(); it != blocks.end(); ++it) + { + (*it)->AssignVIRegs(0); + } + return; + } + + if ((type&BLOCKTYPE_ANALYZED)) + return; + + // child + assert(allocX86Regs == -1); + allocX86Regs = s_vecRegArray.size(); + s_vecRegArray.resize(allocX86Regs + iREGCNT_GPR); + + _x86regs* pregs = &s_vecRegArray[allocX86Regs]; + memset(pregs, 0, sizeof(_x86regs)*iREGCNT_GPR); + + assert(parents.size() > 0); + + list::iterator itparent; + u32 usedvars = insts.front().usedvars[0]; + u32 livevars = insts.front().livevars[0]; + + if (parents.size() > 0) + { + u32 usedvars2 = 0xffffffff; + + for(itparent = parents.begin(); itparent != parents.end(); itparent++) + { + usedvars2 &= (*itparent)->insts.front().usedvars[0]; + } + + usedvars |= usedvars2; + } + + usedvars &= livevars; + + // currently order doesn't matter + int num = 0; + + if (usedvars) + { + for (int i = 1; i < 16; ++i) + { + if (usedvars & (1 << i)) + { + pregs[num].inuse = 1; + pregs[num].reg = i; + + livevars &= ~(1 << i); + + if (++num >= maxregs) break; + } + } + } + + if (num < maxregs) + { + livevars &= ~usedvars; + livevars &= insts.back().usedvars[0]; + + if (livevars) + { + for (int i = 1; i < 16; ++i) + { + if (livevars & (1 << i)) + { + pregs[num].inuse = 1; + pregs[num].reg = i; + + if (++num >= maxregs) break; + } + } + } + } + + s_markov.children.push_back(this); + type |= BLOCKTYPE_ANALYZED; + + for(itparent = parents.begin(); itparent != parents.end(); itparent++) + { + (*itparent)->AssignVIRegs(1); + } +} + +static void SuperVUAssignRegs() +{ + list::iterator itblock, itblock2; + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + } + s_listBlocks.front()->AssignVFRegs(); + + // VI assignments, find markov blanket for each node in the graph + // then allocate regs based on the commonly used ones +#ifdef SUPERVU_X86CACHING + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~(BLOCKTYPE_ANALYZED | BLOCKTYPE_ANALYZEDPARENT); + } + s_vecRegArray.resize(0); + u8 usedregs[16]; + + // note: first block always has to start with no alloc regs + bool bfirst = true; + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + + if (!((*itblock)->type & BLOCKTYPE_ANALYZED)) + { + + if ((*itblock)->parents.size() == 0) + { + (*itblock)->type |= BLOCKTYPE_ANALYZED; + bfirst = false; + continue; + } + + s_markov.children.clear(); + s_markov.parents.clear(); + (*itblock)->AssignVIRegs(0); + + // assign the regs + int regid = s_vecRegArray.size(); + s_vecRegArray.resize(regid + iREGCNT_GPR); + + _x86regs* mergedx86 = &s_vecRegArray[regid]; + memset(mergedx86, 0, sizeof(_x86regs)*iREGCNT_GPR); + + if (!bfirst) + { + *(u32*)usedregs = *((u32*)usedregs + 1) = *((u32*)usedregs + 2) = *((u32*)usedregs + 3) = 0; + + for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) + { + assert((*itblock2)->allocX86Regs >= 0); + _x86regs* pregs = &s_vecRegArray[(*itblock2)->allocX86Regs]; + for (int i = 0; i < iREGCNT_GPR; ++i) + { + if (pregs[i].inuse && pregs[i].reg < 16) + { + //assert( pregs[i].reg < 16); + usedregs[pregs[i].reg]++; + } + } + } + + int num = 1; + for (int i = 0; i < 16; ++i) + { + if (usedregs[i] == s_markov.children.size()) + { + // use + mergedx86[num].inuse = 1; + mergedx86[num].reg = i; + mergedx86[num].type = (s_vu ? X86TYPE_VU1 : 0) | X86TYPE_VI; + mergedx86[num].mode = MODE_READ; + if (++num >= iREGCNT_GPR) + break; + if (num == ESP) + ++num; + } + } + + for(itblock2 = s_markov.children.begin(); itblock2 != s_markov.children.end(); itblock2++) + { + assert((*itblock2)->nStartx86 == -1); + (*itblock2)->nStartx86 = regid; + } + + for(itblock2 = s_markov.parents.begin(); itblock2 != s_markov.parents.end(); itblock2++) + { + assert((*itblock2)->nEndx86 == -1); + (*itblock2)->nEndx86 = regid; + } + } + + bfirst = false; + } + } +#endif +} + +////////////////// +// Recompilation +////////////////// + +// cycles in which the last Q,P regs were finished (written to VU->VI[]) +// the write occurs before the instruction is executed at that cycle +// compare with s_TotalVUCycles +// if less than 0, already flushed +int s_writeQ, s_writeP; + +// declare the saved registers +uptr s_vu1esp, s_callstack;//, s_vu1esp +uptr s_vu1ebp, s_vuebx, s_vuedi, s_vu1esi; + +static int s_recWriteQ, s_recWriteP; // wait times during recompilation +static int s_needFlush; // first bit - Q, second bit - P, third bit - Q has been written, fourth bit - P has been written + +static int s_JumpX86; +static int s_ScheduleXGKICK = 0, s_XGKICKReg = -1; + +void recVUMI_XGKICK_(VURegs *VU); + +void SuperVUCleanupProgram(u32 startpc, int vuindex) +{ +#ifdef SUPERVU_COUNT + QueryPerformanceCounter(&svufinal); + svutime += (u32)(svufinal.QuadPart - svubase.QuadPart); +#endif + + assert(s_vu1esp == 0); + + VU = vuindex ? &VU1 : &VU0; + VU->cycle += s_TotalVUCycles; + + //VU cycle stealing hack, 3000 cycle maximum so it doesn't get out of hand + if (s_TotalVUCycles < 3000) + cpuRegs.cycle += s_TotalVUCycles * Config.Hacks.VUCycleSteal; + else + cpuRegs.cycle += 3000 * Config.Hacks.VUCycleSteal; + + if ((int)s_writeQ > 0) VU->VI[REG_Q] = VU->q; + if ((int)s_writeP > 0) + { + assert(VU == &VU1); + VU1.VI[REG_P] = VU1.p; // only VU1 + } + + //memset(recVUStack, 0, SUPERVU_STACKSIZE * 4); + + // Could clear allocation info to prevent possibly bad data being used in other parts of pcsx2; + // not doing this because it's slow and not needed (rama) + // _initXMMregs(); + // _initMMXregs(); + // _initX86regs(); +} + +#if defined(_MSC_VER) + +// entry point of all vu programs from emulator calls +__declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) +{ + __asm + { + mov eax, dword ptr [esp] + mov s_TotalVUCycles, 0 // necessary to be here! + add esp, 4 + mov s_callstack, eax + call SuperVUGetProgram + + // save cpu state + mov s_vu1ebp, ebp + mov s_vu1esi, esi // have to save even in Release + mov s_vuedi, edi // have to save even in Release + mov s_vuebx, ebx + } +#ifdef _DEBUG + __asm + { + mov s_vu1esp, esp + } +#endif + + __asm + { + //stmxcsr s_ssecsr + ldmxcsr g_sseVUMXCSR + + // init vars + mov s_writeQ, 0xffffffff + mov s_writeP, 0xffffffff + + jmp eax + } +} + +// exit point of all vu programs +__declspec(naked) static void SuperVUEndProgram() +{ + __asm + { + // restore cpu state + ldmxcsr g_sseMXCSR + + mov ebp, s_vu1ebp + mov esi, s_vu1esi + mov edi, s_vuedi + mov ebx, s_vuebx + } + +#ifdef _DEBUG + __asm + { + sub s_vu1esp, esp + } +#endif + + __asm + { + call SuperVUCleanupProgram + jmp s_callstack // so returns correctly + } +} + +#endif + +// Flushes P/Q regs +void SuperVUFlush(int p, int wait) +{ + u8* pjmp[3]; + if (!(s_needFlush&(1 << p))) return; + + int recwait = p ? s_recWriteP : s_recWriteQ; + if (!wait && s_pCurInst->info.cycle < recwait) return; + + if (recwait == 0) + { + // write didn't happen this block + MOV32MtoR(EAX, p ? (uptr)&s_writeP : (uptr)&s_writeQ); + OR32RtoR(EAX, EAX); + pjmp[0] = JS8(0); + + if (s_pCurInst->info.cycle) SUB32ItoR(EAX, s_pCurInst->info.cycle); + + // if writeQ <= total+offset + if (!wait) // only write back if time is up + { + CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); + pjmp[1] = JG8(0); + } + else + { + // add (writeQ-total-offset) to s_TotalVUCycles + // necessary? + CMP32MtoR(EAX, (uptr)&s_TotalVUCycles); + pjmp[2] = JLE8(0); + MOV32RtoM((uptr)&s_TotalVUCycles, EAX); + x86SetJ8(pjmp[2]); + } + } + else if (wait && s_pCurInst->info.cycle < recwait) + { + ADD32ItoM((uptr)&s_TotalVUCycles, recwait); + } + + MOV32MtoR(EAX, SuperVUGetVIAddr(p ? REG_P : REG_Q, 0)); + MOV32ItoM(p ? (uptr)&s_writeP : (uptr)&s_writeQ, 0x80000000); + MOV32RtoM(SuperVUGetVIAddr(p ? REG_P : REG_Q, 1), EAX); + + if (recwait == 0) + { + if (!wait) x86SetJ8(pjmp[1]); + x86SetJ8(pjmp[0]); + } + + if (wait || (!p && recwait == 0 && s_pCurInst->info.cycle >= 12) || (!p && recwait > 0 && s_pCurInst->info.cycle >= recwait)) + s_needFlush &= ~(1 << p); +} + +// executed only once per program +static u32* SuperVUStaticAlloc(u32 size) +{ + assert(recVUStackPtr + size <= recVUStack + SUPERVU_STACKSIZE); + // always zero + if (size == 4) *(u32*)recVUStackPtr = 0; + else memset(recVUStackPtr, 0, size); + recVUStackPtr += size; + return (u32*)(recVUStackPtr - size); +} + +static void SuperVURecompile() +{ + // save cpu state + recVUStackPtr = recVUStack; + + _initXMMregs(); + + list::iterator itblock; + + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + (*itblock)->type &= ~BLOCKTYPE_ANALYZED; + } + + s_listBlocks.front()->Recompile(); + + // make sure everything compiled + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + assert(((*itblock)->type & BLOCKTYPE_ANALYZED) && (*itblock)->pcode != NULL); + } + + // link all blocks + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + VuBaseBlock::LISTBLOCKS::iterator itchild; + + assert((*itblock)->blocks.size() <= ArraySize((*itblock)->pChildJumps)); + + int i = 0; + for(itchild = (*itblock)->blocks.begin(); itchild != (*itblock)->blocks.end(); itchild++) + { + + if ((u32)(uptr)(*itblock)->pChildJumps[i] == 0xffffffff) + continue; + + if ((*itblock)->pChildJumps[i] == NULL) + { + VuBaseBlock* pchild = *itchild; + + if (pchild->type & BLOCKTYPE_HASEOP) + { + assert(pchild->blocks.size() == 0); + + AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag + AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); + + MOV32ItoM((uptr)&VU->VI[REG_TPC], pchild->endpc); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); + } + // only other case is when there are two branches + else + { + assert((*itblock)->insts.back().regs[0].pipe == VUPIPE_BRANCH); + } + + continue; + } + + if ((u32)(uptr)(*itblock)->pChildJumps[i] & 0x80000000) + { + // relative + assert((uptr)(*itblock)->pChildJumps[i] <= 0xffffffff); + (*itblock)->pChildJumps[i] = (u32*)((uptr)(*itblock)->pChildJumps[i] & 0x7fffffff); + *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode - ((uptr)(*itblock)->pChildJumps[i] + 4); + } + else + { + *(*itblock)->pChildJumps[i] = (uptr)(*itchild)->pcode; + } + + ++i; + } + } + + s_pFnHeader->pprogfunc = s_listBlocks.front()->pcode; +} + +// debug + + +u32 s_saveecx, s_saveedx, s_saveebx, s_saveesi, s_saveedi, s_saveebp; +u32 g_curdebugvu; + +//float vuDouble(u32 f); + +#if defined(_MSC_VER) +__declspec(naked) static void svudispfn() +{ + __asm + { + mov g_curdebugvu, eax + mov s_saveecx, ecx + mov s_saveedx, edx + mov s_saveebx, ebx + mov s_saveesi, esi + mov s_saveedi, edi + mov s_saveebp, ebp + } +#else + +void svudispfntemp() +{ +#endif + +#ifdef _DEBUG + static u32 i; + + if (((vudump&8) && g_curdebugvu) || ((vudump&0x80) && !g_curdebugvu)) //&& g_vu1lastrec != g_vu1last ) { + { + + if (skipparent != g_vu1lastrec) + { + for (i = 0; i < ArraySize(badaddrs); ++i) + { + if (s_svulast == badaddrs[i][1] && g_vu1lastrec == badaddrs[i][0]) + break; + } + + if (i == ArraySize(badaddrs)) + { + //static int curesp; + //__asm mov curesp, esp + //Console::WriteLn("tVU: %x %x %x", s_svulast, s_vucount, s_vufnheader); + if (g_curdebugvu) iDumpVU1Registers(); + else iDumpVU0Registers(); + s_vucount++; + } + } + + g_vu1lastrec = s_svulast; + } +#endif + +#if defined(_MSC_VER) + __asm + { + mov ecx, s_saveecx + mov edx, s_saveedx + mov ebx, s_saveebx + mov esi, s_saveesi + mov edi, s_saveedi + mov ebp, s_saveebp + ret + } +#endif +} + +// frees all regs taking into account the livevars +void SuperVUFreeXMMregs(u32* livevars) +{ + for (int i = 0; i < iREGCNT_XMM; ++i) + { + if (xmmregs[i].inuse) + { + // same reg + if ((xmmregs[i].mode & MODE_WRITE)) + { + +#ifdef SUPERVU_INTERCACHING + if (xmmregs[i].type == XMMTYPE_VFREG) + { + if (!(livevars[1] & (1 << xmmregs[i].reg))) continue; + } + else if (xmmregs[i].type == XMMTYPE_ACC) + { + if (!(livevars[0] & (1 << REG_ACC_FLAG))) continue; + } +#endif + + if (xmmregs[i].mode & MODE_VUXYZ) + { + // ALWAYS update + u32 addr = xmmregs[i].type == XMMTYPE_VFREG ? (uptr) & VU->VF[xmmregs[i].reg] : (uptr) & VU->ACC; + + if (xmmregs[i].mode & MODE_VUZ) + { + SSE_MOVHPS_XMM_to_M64(addr, (x86SSERegType)i); + SSE_SHUFPS_M128_to_XMM((x86SSERegType)i, addr, 0xc4); + } + else + { + SSE_MOVHPS_M64_to_XMM((x86SSERegType)i, addr + 8); + } + + xmmregs[i].mode &= ~MODE_VUXYZ; + } + + _freeXMMreg(i); + } + } + } + + //_freeXMMregs(); +} + +void SuperVUTestVU0Condition(u32 incstack) +{ + if (s_vu && !SUPERVU_CHECKCONDITION) return; // vu0 only + + CMP32ItoM((uptr)&s_TotalVUCycles, 512); // sometimes games spin on vu0, so be careful with this value + // woody hangs if too high + + if (incstack) + { + u8* ptr = JB8(0); + + ADD32ItoR(ESP, incstack); + //CALLFunc((u32)timeout); + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); + + x86SetJ8(ptr); + } + else JAE32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 6)); +} + +void VuBaseBlock::Recompile() +{ + if (type & BLOCKTYPE_ANALYZED) return; + + x86Align(16); + pcode = x86Ptr; + +#ifdef _DEBUG + MOV32ItoM((uptr)&s_vufnheader, s_pFnHeader->startpc); + MOV32ItoM((uptr)&VU->VI[REG_TPC], startpc); + MOV32ItoM((uptr)&s_svulast, startpc); + + list::iterator itparent; + for (itparent = parents.begin(); itparent != parents.end(); ++itparent) + { + if ((*itparent)->blocks.size() == 1 && (*itparent)->blocks.front()->startpc == startpc && + ((*itparent)->insts.size() < 2 || (----(*itparent)->insts.end())->regs[0].pipe != VUPIPE_BRANCH)) + { + MOV32ItoM((uptr)&skipparent, (*itparent)->startpc); + break; + } + } + + if (itparent == parents.end()) MOV32ItoM((uptr)&skipparent, -1); + + MOV32ItoR(EAX, s_vu); + CALLFunc((uptr)svudispfn); +#endif + + s_pCurBlock = this; + s_needFlush = 3; + pc = startpc; + branch = 0; + s_recWriteQ = s_recWriteP = 0; + s_XGKICKReg = -1; + s_ScheduleXGKICK = 0; + + s_ClipRead = s_PrevClipWrite = (uptr) & VU->VI[REG_CLIP_FLAG]; + s_StatusRead = s_PrevStatusWrite = (uptr) & VU->VI[REG_STATUS_FLAG]; + s_MACRead = s_PrevMACWrite = (uptr) & VU->VI[REG_MAC_FLAG]; + s_PrevIWrite = (uptr) & VU->VI[REG_I]; + s_JumpX86 = 0; + s_UnconditionalDelay = 0; + + memcpy(xmmregs, startregs, sizeof(xmmregs)); +#ifdef SUPERVU_X86CACHING + if (nStartx86 >= 0) + memcpy(x86regs, &s_vecRegArray[nStartx86], sizeof(x86regs)); + else + _initX86regs(); +#else + _initX86regs(); +#endif + + list::iterator itinst; + for(itinst = insts.begin(); itinst != insts.end(); itinst++) + { + s_pCurInst = &(*itinst); + if (s_JumpX86 > 0) + { + if (!x86regs[s_JumpX86].inuse) + { + // load + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_READ); + } + x86regs[s_JumpX86].needed = 1; + } + + if (s_ScheduleXGKICK && s_XGKICKReg > 0) + { + assert(x86regs[s_XGKICKReg].inuse); + x86regs[s_XGKICKReg].needed = 1; + } + itinst->Recompile(itinst, vuxyz); + + if (s_ScheduleXGKICK > 0) + { + if (s_ScheduleXGKICK-- == 1) + { + recVUMI_XGKICK_(VU); + } + } + } + assert(pc == endpc); + assert(s_ScheduleXGKICK == 0); + + // flush flags + if (s_PrevClipWrite != (uptr)&VU->VI[REG_CLIP_FLAG]) + { + MOV32MtoR(EAX, s_PrevClipWrite); + MOV32RtoM((uptr)&VU->VI[REG_CLIP_FLAG], EAX); + } + if (s_PrevStatusWrite != (uptr)&VU->VI[REG_STATUS_FLAG]) + { + MOV32MtoR(EAX, s_PrevStatusWrite); + MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); + } + if (s_PrevMACWrite != (uptr)&VU->VI[REG_MAC_FLAG]) + { + MOV32MtoR(EAX, s_PrevMACWrite); + MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); + } +// if( s_StatusRead != (uptr)&VU->VI[REG_STATUS_FLAG] ) { +// // only lower 8 bits valid! +// MOVZX32M8toR(EAX, s_StatusRead); +// MOV32RtoM((uptr)&VU->VI[REG_STATUS_FLAG], EAX); +// } +// if( s_MACRead != (uptr)&VU->VI[REG_MAC_FLAG] ) { +// // only lower 8 bits valid! +// MOVZX32M8toR(EAX, s_MACRead); +// MOV32RtoM((uptr)&VU->VI[REG_MAC_FLAG], EAX); +// } + if (s_PrevIWrite != (uptr)&VU->VI[REG_I]) + { + MOV32ItoM((uptr)&VU->VI[REG_I], *(u32*)s_PrevIWrite); // never changes + } + + ADD32ItoM((uptr)&s_TotalVUCycles, cycles); + + // compute branches, jumps, eop + if (type & BLOCKTYPE_HASEOP) + { + // end + _freeXMMregs(); + _freeX86regs(); + AND32ItoM((uptr)&VU0.VI[ REG_VPU_STAT ].UL, s_vu ? ~0x100 : ~0x001); // E flag + AND32ItoM((uptr)&VU->vifRegs->stat, ~0x4); + + if (!branch) MOV32ItoM((uptr)&VU->VI[REG_TPC], endpc); + + JMP32((uptr)SuperVUEndProgram - ((uptr)x86Ptr + 5)); + } + else + { + + u32 livevars[2] = {0}; + + list::iterator lastinst = GetInstIterAtPc(endpc - 8); + lastinst++; + + if (lastinst != insts.end()) + { + livevars[0] = lastinst->livevars[0]; + livevars[1] = lastinst->livevars[1]; + } + else + { + // take from children + if (blocks.size() > 0) + { + LISTBLOCKS::iterator itchild; + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { + livevars[0] |= (*itchild)->insts.front().livevars[0]; + livevars[1] |= (*itchild)->insts.front().livevars[1]; + } + } + else + { + livevars[0] = ~0; + livevars[1] = ~0; + } + } + + SuperVUFreeXMMregs(livevars); + + // get rid of any writes, otherwise _freeX86regs will write + x86regs[s_JumpX86].mode &= ~MODE_WRITE; + + if (branch == 1) + { + if (!x86regs[s_JumpX86].inuse) + { + assert(x86regs[s_JumpX86].type == X86TYPE_VUJUMP); + s_JumpX86 = 0xffffffff; // notify to jump from g_recWriteback + } + } + + // align VI regs +#ifdef SUPERVU_X86CACHING + if (nEndx86 >= 0) + { + _x86regs* endx86 = &s_vecRegArray[nEndx86]; + for (int i = 0; i < iREGCNT_GPR; ++i) + { + if (endx86[i].inuse) + { + + if (s_JumpX86 == i && x86regs[s_JumpX86].inuse) + { + x86regs[s_JumpX86].inuse = 0; + x86regs[EAX].inuse = 1; + MOV32RtoR(EAX, s_JumpX86); + s_JumpX86 = EAX; + } + + if (x86regs[i].inuse) + { + if (x86regs[i].type == endx86[i].type && x86regs[i].reg == endx86[i].reg) + { + _freeX86reg(i); + // will continue to use it + continue; + } + +#ifdef SUPERVU_INTERCACHING + if (x86regs[i].type == (X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0))) + { + if (livevars[0] & (1 << x86regs[i].reg)) + _freeX86reg(i); + else + x86regs[i].inuse = 0; + } + else +#endif + { + _freeX86reg(i); + } + } + + // realloc + _allocX86reg(i, endx86[i].type, endx86[i].reg, MODE_READ); + if (x86regs[i].mode & MODE_WRITE) + { + _freeX86reg(i); + x86regs[i].inuse = 1; + } + } + else _freeX86reg(i); + } + } + else _freeX86regs(); +#else + _freeX86regs(); +#endif + + // store the last block executed + MOV32ItoM((uptr)&g_nLastBlockExecuted, s_pCurBlock->startpc); + + switch (branch) + { + case 1: // branch, esi has new prog + + SuperVUTestVU0Condition(0); + + if (s_JumpX86 == 0xffffffff) + JMP32M((uptr)&g_recWriteback); + else + JMPR(s_JumpX86); + + break; + case 4: // jalr + pChildJumps[0] = (u32*)0xffffffff; + // fall through + + case 0x10: // jump, esi has new vupc + { + _freeXMMregs(); + _freeX86regs(); + + SuperVUTestVU0Condition(8); + + // already onto stack + CALLFunc((uptr)SuperVUGetProgram); + ADD32ItoR(ESP, 8); + JMPR(EAX); + break; + } + + case 0x13: // jr with uncon branch, uncond branch takes precendence (dropship) + { +// s32 delta = (s32)(VU->code & 0x400 ? 0xfffffc00 | (VU->code & 0x3ff) : VU->code & 0x3ff) << 3; +// ADD32ItoRmOffset(ESP, delta, 0); + + ADD32ItoR(ESP, 8); // restore + pChildJumps[0] = (u32*)((uptr)JMP32(0) | 0x80000000); + break; + } + case 0: + case 3: // unconditional branch + pChildJumps[s_UnconditionalDelay] = (u32*)((uptr)JMP32(0) | 0x80000000); + break; + + default: + DevCon::Error("Bad branch %x\n", params branch); + assert(0); + break; + } + } + + pendcode = x86Ptr; + type |= BLOCKTYPE_ANALYZED; + + LISTBLOCKS::iterator itchild; + for(itchild = blocks.begin(); itchild != blocks.end(); itchild++) + { + (*itchild)->Recompile(); + } +} + +#define GET_VUXYZMODE(reg) 0//((vuxyz&(1<<(reg)))?MODE_VUXYZ:0) + +int VuInstruction::SetCachedRegs(int upper, u32 vuxyz) +{ + if (vfread0[upper] >= 0) + { + SuperVUFreeXMMreg(vfread0[upper], XMMTYPE_VFREG, regs[upper].VFread0); + _allocVFtoXMMreg(VU, vfread0[upper], regs[upper].VFread0, MODE_READ | GET_VUXYZMODE(regs[upper].VFread0)); + } + if (vfread1[upper] >= 0) + { + SuperVUFreeXMMreg(vfread1[upper], XMMTYPE_VFREG, regs[upper].VFread1); + _allocVFtoXMMreg(VU, vfread1[upper], regs[upper].VFread1, MODE_READ | GET_VUXYZMODE(regs[upper].VFread1)); + } + if (vfacc[upper] >= 0 && (regs[upper].VIread&(1 << REG_ACC_FLAG))) + { + SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); + _allocACCtoXMMreg(VU, vfacc[upper], MODE_READ); + } + if (vfwrite[upper] >= 0) + { + assert(regs[upper].VFwrite > 0); + SuperVUFreeXMMreg(vfwrite[upper], XMMTYPE_VFREG, regs[upper].VFwrite); + _allocVFtoXMMreg(VU, vfwrite[upper], regs[upper].VFwrite, + MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0) | GET_VUXYZMODE(regs[upper].VFwrite)); + } + if (vfacc[upper] >= 0 && (regs[upper].VIwrite&(1 << REG_ACC_FLAG))) + { + SuperVUFreeXMMreg(vfacc[upper], XMMTYPE_ACC, 0); + _allocACCtoXMMreg(VU, vfacc[upper], MODE_WRITE | (regs[upper].VFwxyzw != 0xf ? MODE_READ : 0)); + } + + int info = PROCESS_VU_SUPER; + if (vfread0[upper] >= 0) info |= PROCESS_EE_SET_S(vfread0[upper]); + if (vfread1[upper] >= 0) info |= PROCESS_EE_SET_T(vfread1[upper]); + if (vfacc[upper] >= 0) info |= PROCESS_VU_SET_ACC(vfacc[upper]); + if (vfwrite[upper] >= 0) + { + if (regs[upper].VFwrite == _Ft_ && vfread1[upper] < 0) + { + info |= PROCESS_EE_SET_T(vfwrite[upper]); + } + else + { + assert(regs[upper].VFwrite == _Fd_); + info |= PROCESS_EE_SET_D(vfwrite[upper]); + } + } + + if (!(vffree[upper]&VFFREE_INVALID0)) + { + SuperVUFreeXMMreg(vffree[upper]&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, vffree[upper]&0xf); + } + info |= PROCESS_VU_SET_TEMP(vffree[upper] & 0xf); + + if (vfflush[upper] >= 0) + { + SuperVUFreeXMMreg(vfflush[upper], XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, vfflush[upper]); + } + + if (upper && (regs[upper].VIwrite & (1 << REG_CLIP_FLAG))) + { + // CLIP inst, need two extra temp registers, put it EEREC_D and EEREC_ACC + assert(vfwrite[upper] == -1); + SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); + info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); + + SuperVUFreeXMMreg((vffree[upper] >> 16)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 16)&0xf); + info |= PROCESS_EE_SET_ACC((vffree[upper] >> 16) & 0xf); + + _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore + _freeXMMreg((vffree[upper] >> 16)&0xf); // don't need anymore + } + else if (regs[upper].VIwrite & (1 << REG_P)) + { + SuperVUFreeXMMreg((vffree[upper] >> 8)&0xf, XMMTYPE_TEMP, 0); + _allocTempXMMreg(XMMT_FPS, (vffree[upper] >> 8)&0xf); + info |= PROCESS_EE_SET_D((vffree[upper] >> 8) & 0xf); + _freeXMMreg((vffree[upper] >> 8)&0xf); // don't need anymore + } + + if (vfflush[upper] >= 0) _freeXMMreg(vfflush[upper]); + if (!(vffree[upper]&VFFREE_INVALID0)) + _freeXMMreg(vffree[upper]&0xf); // don't need anymore + + if ((regs[0].VIwrite | regs[1].VIwrite) & ((1 << REG_STATUS_FLAG) | (1 << REG_MAC_FLAG))) + info |= PROCESS_VU_UPDATEFLAGS; + + return info; +} + +void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) +{ + //static PCSX2_ALIGNED16(VECTOR _VF); + //static PCSX2_ALIGNED16(VECTOR _VFc); + u32 *ptr; + u8* pjmp; + int vfregstore = 0; + + assert(s_pCurInst == this); + s_WriteToReadQ = 0; + + ptr = (u32*) & VU->Micro[ pc ]; + + if (type & INST_Q_READ) + SuperVUFlush(0, (ptr[0] == 0x800003bf) || !!(regs[0].VIwrite & (1 << REG_Q))); + if (type & INST_P_READ) + SuperVUFlush(1, (ptr[0] == 0x800007bf) || !!(regs[0].VIwrite & (1 << REG_P))); + + if (type & INST_DUMMY) + { + + // find nParentPc + VuInstruction* pparentinst = NULL; + + // if true, will check if parent block was executed before getting the results of the flags (superman returns) + int nParentCheckForExecution = -1; + +// int badaddrs[] = { +// 0x60,0x68,0x70,0x60,0x68,0x70,0x88,0x90,0x98,0x98,0xa8,0xb8,0x88,0x90, +// 0x4a8,0x4a8,0x398,0x3a0,0x3a8,0xa0 +// }; + +#ifdef SUPERVU_PROPAGATEFLAGS + if (nParentPc != -1 && (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc)) + { + +// if( !s_vu ) { +// for(int j = 0; j < ARRAYSIZE(badaddrs); ++j) { +// if( badaddrs[j] == nParentPc ) +// goto NoParent; +// } +// } + + list::iterator itblock; + for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) + { + if (nParentPc >= (*itblock)->startpc && nParentPc < (*itblock)->endpc) + { + pparentinst = &(*(*itblock)->GetInstIterAtPc(nParentPc)); + //if( !s_vu ) SysPrintf("%x ", nParentPc); + if (find(s_pCurBlock->parents.begin(), s_pCurBlock->parents.end(), *itblock) != s_pCurBlock->parents.end()) + nParentCheckForExecution = (*itblock)->startpc; + break; + } + } + + assert(pparentinst != NULL); + } +#endif + + if (type & INST_CLIP_WRITE) + { + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { + + if (!CHECK_VUCLIPFLAGHACK && pparentinst != NULL) + { + + if (nParentCheckForExecution >= 0) + { + if (pparentinst->pClipWrite == 0) + pparentinst->pClipWrite = (uptr)SuperVUStaticAlloc(4); + + if (s_ClipRead == 0) + s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; + + CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); + u8* jptr = JNE8(0); + CMP32ItoM((uptr)&s_ClipRead, (uptr)&VU->VI[REG_CLIP_FLAG]); + u8* jptr2 = JE8(0); + MOV32MtoR(EAX, pparentinst->pClipWrite); + MOV32RtoM(s_ClipRead, EAX); + x86SetJ8(jptr); + x86SetJ8(jptr2); + } + } + else s_ClipRead = (uptr) & VU->VI[REG_CLIP_FLAG]; + } + else + { + s_ClipRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pClipWrite; + if (s_ClipRead == 0) Console::WriteLn("super ClipRead allocation error!"); + } + } + + // before modifying, check if they will ever be read + if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) + { + + u8 outofblock = 0; + if (type & INST_STATUS_WRITE) + { + + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { + + // reading from out of this block, so already flushed to mem + if (pparentinst != NULL) //&& pparentinst->pStatusWrite != NULL ) { + { + + // might not have processed it yet, so reserve a mem loc + if (pparentinst->pStatusWrite == 0) + { + pparentinst->pStatusWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pparentinst->pStatusWrite, 0); + } + +// if( s_pCurBlock->prevFlagsOutOfBlock && s_StatusRead != NULL ) { +// // or instead since don't now which parent we came from +// MOV32MtoR(EAX, pparentinst->pStatusWrite); +// OR32RtoM(s_StatusRead, EAX); +// MOV32ItoM(pparentinst->pStatusWrite, 0); +// } + + if (nParentCheckForExecution >= 0) + { + + // don't now which parent we came from, so have to check +// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); +// if( s_StatusRead != NULL ) +// MOV32MtoR(EAX, s_StatusRead); +// else +// MOV32MtoR(EAX, (uptr)&VU->VI[REG_STATUS_FLAG]); +// s_StatusRead = tempstatus; + + if (s_StatusRead == 0) + s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; + + CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); + u8* jptr = JNE8(0); + MOV32MtoR(EAX, pparentinst->pStatusWrite); + MOV32ItoM(pparentinst->pStatusWrite, 0); + MOV32RtoM(s_StatusRead, EAX); + x86SetJ8(jptr); + } + else + { + uptr tempstatus = (uptr)SuperVUStaticAlloc(4); + MOV32MtoR(EAX, pparentinst->pStatusWrite); + MOV32RtoM(tempstatus, EAX); + MOV32ItoM(pparentinst->pStatusWrite, 0); + s_StatusRead = tempstatus; + } + + outofblock = 2; + } + else + s_StatusRead = (uptr) & VU->VI[REG_STATUS_FLAG]; + } + else + { + s_StatusRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pStatusWrite; + if (s_StatusRead == 0) Console::WriteLn("super StatusRead allocation error!"); +// if( pc >= (u32)s_pCurBlock->endpc-8 ) { +// // towards the end, so variable might be leaded to another block (silent hill 4) +// uptr tempstatus = (uptr)SuperVUStaticAlloc(4); +// MOV32MtoR(EAX, s_StatusRead); +// MOV32RtoM(tempstatus, EAX); +// MOV32ItoM(s_StatusRead, 0); +// s_StatusRead = tempstatus; +// } + } + } + if (type & INST_MAC_WRITE) + { + + if (nParentPc < s_pCurBlock->startpc || nParentPc >= (int)pc) + { + // reading from out of this block, so already flushed to mem + + if (pparentinst != NULL) //&& pparentinst->pMACWrite != NULL ) { + { + // necessary for (katamari) + // towards the end, so variable might be leaked to another block (silent hill 4) + + // might not have processed it yet, so reserve a mem loc + if (pparentinst->pMACWrite == 0) + { + pparentinst->pMACWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pparentinst->pMACWrite, 0); + } + +// if( s_pCurBlock->prevFlagsOutOfBlock && s_MACRead != NULL ) { +// // or instead since don't now which parent we came from +// MOV32MtoR(EAX, pparentinst->pMACWrite); +// OR32RtoM(s_MACRead, EAX); +// MOV32ItoM(pparentinst->pMACWrite, 0); +// } + if (nParentCheckForExecution >= 0) + { + + // don't now which parent we came from, so have to check +// uptr tempmac = (uptr)SuperVUStaticAlloc(4); +// if( s_MACRead != NULL ) +// MOV32MtoR(EAX, s_MACRead); +// else +// MOV32MtoR(EAX, (uptr)&VU->VI[REG_MAC_FLAG]); +// s_MACRead = tempmac; + + if (s_MACRead == 0) s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; + + CMP32ItoM((uptr)&g_nLastBlockExecuted, nParentCheckForExecution); + u8* jptr = JNE8(0); + MOV32MtoR(EAX, pparentinst->pMACWrite); + MOV32ItoM(pparentinst->pMACWrite, 0); + MOV32RtoM(s_MACRead, EAX); + x86SetJ8(jptr); + } + else + { + uptr tempMAC = (uptr)SuperVUStaticAlloc(4); + MOV32MtoR(EAX, pparentinst->pMACWrite); + MOV32RtoM(tempMAC, EAX); + MOV32ItoM(pparentinst->pMACWrite, 0); + s_MACRead = tempMAC; + } + + outofblock = 2; + } + else + s_MACRead = (uptr) & VU->VI[REG_MAC_FLAG]; + +// if( pc >= (u32)s_pCurBlock->endpc-8 ) { +// // towards the end, so variable might be leaked to another block (silent hill 4) +// uptr tempMAC = (uptr)SuperVUStaticAlloc(4); +// MOV32MtoR(EAX, s_MACRead); +// MOV32RtoM(tempMAC, EAX); +// MOV32ItoM(s_MACRead, 0); +// s_MACRead = tempMAC; +// } + } + else + { + s_MACRead = s_pCurBlock->GetInstIterAtPc(nParentPc)->pMACWrite; + } + } + + s_pCurBlock->prevFlagsOutOfBlock = outofblock; + } + else if (pparentinst != NULL) + { + // make sure to reset the mac and status flags! (katamari) + if (pparentinst->pStatusWrite) + MOV32ItoM(pparentinst->pStatusWrite, 0); + if (pparentinst->pMACWrite) + MOV32ItoM(pparentinst->pMACWrite, 0); + } + + assert(s_ClipRead != 0); + assert(s_MACRead != 0); + assert(s_StatusRead != 0); + + return; + } + + s_pCurBlock->prevFlagsOutOfBlock = 0; + +#ifdef _DEBUG + MOV32ItoR(EAX, pc); +#endif + + assert(!(type & (INST_CLIP_WRITE | INST_STATUS_WRITE | INST_MAC_WRITE))); + pc += 8; + + list::const_iterator itinst2; + + if ((regs[0].VIwrite | regs[1].VIwrite) & ((1 << REG_MAC_FLAG) | (1 << REG_STATUS_FLAG))) + { + if (s_pCurBlock->type & BLOCKTYPE_MACFLAGS) + { + if (pMACWrite == 0) + { + pMACWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pMACWrite, 0); + } + if (pStatusWrite == 0) + { + pStatusWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pStatusWrite, 0); + } + } + else + { + assert(s_StatusRead == (uptr)&VU->VI[REG_STATUS_FLAG]); + assert(s_MACRead == (uptr)&VU->VI[REG_MAC_FLAG]); + pMACWrite = s_MACRead; + pStatusWrite = s_StatusRead; + } + } + + if ((pClipWrite == 0) && ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_CLIP_FLAG))) + { + pClipWrite = (uptr)SuperVUStaticAlloc(4); + //MOV32ItoM(pClipWrite, 0); + } + +#ifdef SUPERVU_X86CACHING + // redo the counters so that the proper regs are released + for (int j = 0; j < iREGCNT_GPR; ++j) + { + if (x86regs[j].inuse && X86_ISVI(x86regs[j].type)) + { + int count = 0; + itinst2 = itinst; + + while (itinst2 != s_pCurBlock->insts.end()) + { + if ((itinst2->regs[0].VIread | itinst2->regs[0].VIwrite | itinst2->regs[1].VIread | itinst2->regs[1].VIwrite) && (1 << x86regs[j].reg)) + break; + + ++count; + ++itinst2; + } + + x86regs[j].counter = 1000 - count; + } + } +#endif + + if (s_vu == 0 && (ptr[1] & 0x20000000)) // M flag + { + OR8ItoM((uptr)&VU->flags, VUFLAG_MFLAGSET); + } + if (ptr[1] & 0x10000000) // D flag + { + TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x400 : 0x004); + u8* ptr = JZ8(0); + OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x200 : 0x002); + _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); + x86SetJ8(ptr); + } + if (ptr[1] & 0x08000000) // T flag + { + TEST32ItoM((uptr)&VU0.VI[REG_FBRST].UL, s_vu ? 0x800 : 0x008); + u8* ptr = JZ8(0); + OR32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, s_vu ? 0x400 : 0x004); + _callFunctionArg1((uptr)hwIntcIrq, MEM_CONSTTAG, s_vu ? INTC_VU1 : INTC_VU0); + x86SetJ8(ptr); + } + + // check upper flags + if (ptr[1] & 0x80000000) // I flag + { + + assert(!(regs[0].VIwrite & ((1 << REG_Q) | (1 << REG_P)))); + + VU->code = ptr[1]; + s_vuInfo = SetCachedRegs(1, vuxyz); + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; + + recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); + + s_PrevIWrite = (uptr)ptr; + _clearNeededXMMregs(); + _clearNeededX86regs(); + } + else + { + if (regs[0].VIwrite & (1 << REG_Q)) + { + + // search for all the insts between this inst and writeback + itinst2 = itinst; + ++itinst2; + u32 cacheq = (itinst2 == s_pCurBlock->insts.end()); + u32* codeptr2 = ptr + 2; + + while (itinst2 != s_pCurBlock->insts.end()) + { + if (!(itinst2->type & INST_DUMMY) && ((itinst2->regs[0].VIwrite&(1 << REG_Q)) || codeptr2[0] == 0x800003bf)) // waitq, or fdiv inst + { + break; + } + if ((itinst2->type & INST_Q_WRITE) && itinst2->nParentPc == pc - 8) + { + break; + } + if (itinst2->type & INST_Q_READ) + { + cacheq = 1; + break; + } + if (itinst2->type & INST_DUMMY) + { + ++itinst2; + continue; + } + codeptr2 += 2; + ++itinst2; + } + + if (itinst2 == s_pCurBlock->insts.end()) + cacheq = 1; + + int x86temp = -1; + if (cacheq) + x86temp = _allocX86reg(-1, X86TYPE_TEMP, 0, 0); + + // new is written so flush old + // if type & INST_Q_READ, already flushed + if (!(type & INST_Q_READ) && s_recWriteQ == 0) MOV32MtoR(EAX, (uptr)&s_writeQ); + + if (cacheq) + MOV32MtoR(x86temp, (uptr)&s_TotalVUCycles); + + if (!(type & INST_Q_READ)) + { + if (s_recWriteQ == 0) + { + OR32RtoR(EAX, EAX); + pjmp = JS8(0); + MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); + MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); + x86SetJ8(pjmp); + } + else if (s_needFlush & 1) + { + MOV32MtoR(EAX, SuperVUGetVIAddr(REG_Q, 0)); + MOV32RtoM(SuperVUGetVIAddr(REG_Q, 1), EAX); + s_needFlush &= ~1; + } + } + + // write new Q + if (cacheq) + { + assert(s_pCurInst->pqcycles > 1); + ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); + MOV32RtoM((uptr)&s_writeQ, x86temp); + s_needFlush |= 1; + } + else + { + // won't be writing back + s_WriteToReadQ = 1; + s_needFlush &= ~1; + MOV32ItoM((uptr)&s_writeQ, 0x80000001); + } + + s_recWriteQ = s_pCurInst->info.cycle + s_pCurInst->pqcycles; + + if (x86temp >= 0) + _freeX86reg(x86temp); + } + + if (regs[0].VIwrite & (1 << REG_P)) + { + int x86temp = _allocX86reg(-1, X86TYPE_TEMP, 0, 0); + + // new is written so flush old + if (!(type & INST_P_READ) && s_recWriteP == 0) + MOV32MtoR(EAX, (uptr)&s_writeP); + MOV32MtoR(x86temp, (uptr)&s_TotalVUCycles); + + if (!(type & INST_P_READ)) + { + if (s_recWriteP == 0) + { + OR32RtoR(EAX, EAX); + pjmp = JS8(0); + MOV32MtoR(EAX, SuperVUGetVIAddr(REG_P, 0)); + MOV32RtoM(SuperVUGetVIAddr(REG_P, 1), EAX); + x86SetJ8(pjmp); + } + else if (s_needFlush & 2) + { + MOV32MtoR(EAX, SuperVUGetVIAddr(REG_P, 0)); + MOV32RtoM(SuperVUGetVIAddr(REG_P, 1), EAX); + s_needFlush &= ~2; + } + } + + // write new P + assert(s_pCurInst->pqcycles > 1); + ADD32ItoR(x86temp, s_pCurInst->info.cycle + s_pCurInst->pqcycles); + MOV32RtoM((uptr)&s_writeP, x86temp); + s_needFlush |= 2; + + s_recWriteP = s_pCurInst->info.cycle + s_pCurInst->pqcycles; + + _freeX86reg(x86temp); + } + + // waitq + if (ptr[0] == 0x800003bf) SuperVUFlush(0, 1); + // waitp + if (ptr[0] == 0x800007bf) SuperVUFlush(1, 1); + +#ifdef PCSX2_DEVBUILD + if (regs[1].VIread & regs[0].VIwrite & ~((1 << REG_Q) | (1 << REG_P) | (1 << REG_VF0_FLAG) | (1 << REG_ACC_FLAG))) + { + Console::Notice("*PCSX2*: Warning, VI write to the same reg %x in both lower/upper cycle %x", params regs[1].VIread & regs[0].VIwrite, s_pCurBlock->startpc); + } +#endif + + u32 modewrite = 0; + if (vfwrite[1] >= 0 && xmmregs[vfwrite[1]].inuse && xmmregs[vfwrite[1]].type == XMMTYPE_VFREG && xmmregs[vfwrite[1]].reg == regs[1].VFwrite) + modewrite = xmmregs[vfwrite[1]].mode & MODE_WRITE; + + VU->code = ptr[1]; + s_vuInfo = SetCachedRegs(1, vuxyz); + + if (vfwrite[1] >= 0) + { + assert(regs[1].VFwrite > 0); + + if (vfwrite[0] == vfwrite[1]) + { + //Console::WriteLn("*PCSX2*: Warning, VF write to the same reg in both lower/upper cycle %x", params s_pCurBlock->startpc); + } + + if (vfread0[0] == vfwrite[1] || vfread1[0] == vfwrite[1]) + { + assert(regs[0].VFread0 == regs[1].VFwrite || regs[0].VFread1 == regs[1].VFwrite); + assert(vfflush[0] >= 0); + if (modewrite) + { + SSE_MOVAPS_XMM_to_M128((uptr)&VU->VF[regs[1].VFwrite], (x86SSERegType)vfwrite[1]); + } + vfregstore = 1; + } + } + + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; + + recVU_UPPER_OPCODE[ VU->code & 0x3f ](VU, s_vuInfo); + _clearNeededXMMregs(); + _clearNeededX86regs(); + + // necessary because status can be set by both upper and lower + if (regs[1].VIwrite & (1 << REG_STATUS_FLAG)) + { + assert(pStatusWrite != 0); + s_PrevStatusWrite = pStatusWrite; + } + + VU->code = ptr[0]; + s_vuInfo = SetCachedRegs(0, vuxyz); + + if (vfregstore) + { + // load + SSE_MOVAPS_M128_to_XMM(vfflush[0], (uptr)&VU->VF[regs[1].VFwrite]); + + assert(xmmregs[vfwrite[1]].mode & MODE_WRITE); + + // replace with vfflush + if (_Fs_ == regs[1].VFwrite) + { + s_vuInfo &= ~PROCESS_EE_SET_S(0xf); + s_vuInfo |= PROCESS_EE_SET_S(vfflush[0]); + } + if (_Ft_ == regs[1].VFwrite) + { + s_vuInfo &= ~PROCESS_EE_SET_T(0xf); + s_vuInfo |= PROCESS_EE_SET_T(vfflush[0]); + } + + xmmregs[vfflush[0]].mode |= MODE_NOFLUSH | MODE_WRITE; // so that lower inst doesn't flush + } + + // notify vuinsts that upper inst is a fmac + if (regs[1].pipe == VUPIPE_FMAC) + s_vuInfo |= PROCESS_VU_SET_FMAC(); + + if (s_JumpX86 > 0) x86regs[s_JumpX86].needed = 1; + if (s_ScheduleXGKICK && s_XGKICKReg > 0) x86regs[s_XGKICKReg].needed = 1; + +#ifdef SUPERVU_VIBRANCHDELAY + if (type & INST_CACHE_VI) + { + assert(vicached >= 0); + int cachedreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), vicached, MODE_READ); + MOV32RtoM((uptr)&s_VIBranchDelay, cachedreg); + } +#endif + + // check if inst before branch and the write is the same as the read in the branch (wipeout) +// int oldreg=0; +// if( pc == s_pCurBlock->endpc-16 ) { +// itinst2 = itinst; ++itinst2; +// if( itinst2->regs[0].pipe == VUPIPE_BRANCH && (itinst->regs[0].VIwrite&itinst2->regs[0].VIread) ) { +// +// CALLFunc((u32)branchfn); +// assert( itinst->regs[0].VIwrite & 0xffff ); +// Console::WriteLn("vi write before branch"); +// for(s_CacheVIReg = 0; s_CacheVIReg < 16; ++s_CacheVIReg) { +// if( itinst->regs[0].VIwrite & (1<endpc-8 && s_CacheVIReg >= 0 ) { +// assert( s_CacheVIX86 > 0 && x86regs[s_CacheVIX86].inuse && x86regs[s_CacheVIX86].reg == s_CacheVIReg && x86regs[s_CacheVIX86].type == X86TYPE_VITEMP ); +// +// oldreg = _allocX86reg(-1, X86TYPE_VI|(s_vu?X86TYPE_VU1:0), s_CacheVIReg, MODE_READ); +// x86regs[s_CacheVIX86].needed = 1; +// assert( x86regs[oldreg].mode & MODE_WRITE ); +// +// x86regs[s_CacheVIX86].type = X86TYPE_VI|(s_vu?X86TYPE_VU1:0); +// x86regs[oldreg].type = X86TYPE_VITEMP; +// } + + recVU_LOWER_OPCODE[ VU->code >> 25 ](VU, s_vuInfo); + +// if( pc == s_pCurBlock->endpc-8 && s_CacheVIReg >= 0 ) { +// // revert +// x86regs[s_CacheVIX86].inuse = 0; +// x86regs[oldreg].type = X86TYPE_VI|(s_vu?X86TYPE_VU1:0); +// } + + _clearNeededXMMregs(); + _clearNeededX86regs(); + } + + // clip is always written so ok + if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_CLIP_FLAG)) + { + assert(pClipWrite != 0); + s_PrevClipWrite = pClipWrite; + } + + if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_STATUS_FLAG)) + { + assert(pStatusWrite != 0); + s_PrevStatusWrite = pStatusWrite; + } + + if ((regs[0].VIwrite | regs[1].VIwrite) & (1 << REG_MAC_FLAG)) + { + assert(pMACWrite != 0); + s_PrevMACWrite = pMACWrite; + } +} + +/////////////////////////////////// +// Super VU Recompilation Tables // +/////////////////////////////////// + +void recVUMI_BranchHandle() +{ + int bpc = _recbranchAddr(VU->code); + int curjump = 0; + + if (s_pCurInst->type & INST_BRANCH_DELAY) + { + assert((branch&0x17) != 0x10 && (branch&0x17) != 4); // no jump handlig for now + + if ((branch & 0x7) == 3) + { + // previous was a direct jump + curjump = 1; + } + else if (branch & 1) curjump = 2; + } + + assert(s_JumpX86 > 0); + + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR + s_pCurBlock->pChildJumps[curjump] = (u32*)x86Ptr - 1; + + if (!(s_pCurInst->type & INST_BRANCH_DELAY)) + { + j8Ptr[1] = JMP8(0); + x86SetJ8(j8Ptr[ 0 ]); + + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), pc + 8); + MOV32ItoR(s_JumpX86, 1); // use 1 to disable optimization to XOR + s_pCurBlock->pChildJumps[curjump+1] = (u32*)x86Ptr - 1; + + x86SetJ8(j8Ptr[ 1 ]); + } + else + x86SetJ8(j8Ptr[ 0 ]); + + branch |= 1; +} + +// supervu specific insts +void recVUMI_IBQ_prep() +{ + int isreg, itreg; + + if (_Is_ == 0) + { +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) + { + itreg = -1; + } + else +#endif + { + itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); + } + + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + + if (itreg >= 0) + { + CMP16ItoR(itreg, 0); + } + else CMP16ItoM(SuperVUGetVIAddr(_It_, 1), 0); + } + else if (_It_ == 0) + { +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + + if (isreg >= 0) + { + CMP16ItoR(isreg, 0); + } + else CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0); + + } + else + { + _addNeededX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_); + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _It_) + { + itreg = -1; + + if (isreg <= 0) + { + // allocate fsreg + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = _allocX86reg(-1, X86TYPE_TEMP, 0, MODE_READ | MODE_WRITE); + MOV32MtoR(isreg, SuperVUGetVIAddr(_Is_, 1)); + } + else + isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + } + else +#endif + { + itreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _It_, MODE_READ); + } + + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + + if (isreg >= 0) + { + if (itreg >= 0) + { + CMP16RtoR(isreg, itreg); + } + else CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); + } + else if (itreg >= 0) + { + CMP16MtoR(itreg, SuperVUGetVIAddr(_Is_, 1)); + } + else + { + isreg = _allocX86reg(-1, X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + CMP16MtoR(isreg, SuperVUGetVIAddr(_It_, 1)); + } + } +} + +void recVUMI_IBEQ(VURegs* vuu, s32 info) +{ + recVUMI_IBQ_prep(); + j8Ptr[ 0 ] = JNE8(0); + recVUMI_BranchHandle(); +} + +void recVUMI_IBGEZ(VURegs* vuu, s32 info) +{ + int isreg; + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + + if (isreg >= 0) + { + TEST16RtoR(isreg, isreg); + j8Ptr[ 0 ] = JS8(0); + } + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JL8(0); + } + + recVUMI_BranchHandle(); +} + +void recVUMI_IBGTZ(VURegs* vuu, s32 info) +{ + int isreg; + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + + if (isreg >= 0) + { + CMP16ItoR(isreg, 0); + j8Ptr[ 0 ] = JLE8(0); + } + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JLE8(0); + } + recVUMI_BranchHandle(); +} + +void recVUMI_IBLEZ(VURegs* vuu, s32 info) +{ + int isreg; + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + + if (isreg >= 0) + { + CMP16ItoR(isreg, 0); + j8Ptr[ 0 ] = JG8(0); + } + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JG8(0); + } + recVUMI_BranchHandle(); +} + +void recVUMI_IBLTZ(VURegs* vuu, s32 info) +{ + int isreg; + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + +#ifdef SUPERVU_VIBRANCHDELAY + if (s_pCurInst->vicached >= 0 && s_pCurInst->vicached == _Is_) + { + isreg = -1; + } + else +#endif + { + isreg = _checkX86reg(X86TYPE_VI | (VU == &VU1 ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + } + + if (isreg >= 0) + { + TEST16RtoR(isreg, isreg); + j8Ptr[ 0 ] = JNS8(0); + } + else + { + CMP16ItoM(SuperVUGetVIAddr(_Is_, 1), 0x0); + j8Ptr[ 0 ] = JGE8(0); + } + recVUMI_BranchHandle(); +} + +void recVUMI_IBNE(VURegs* vuu, s32 info) +{ + recVUMI_IBQ_prep(); + j8Ptr[ 0 ] = JE8(0); + recVUMI_BranchHandle(); +} + +void recVUMI_B(VURegs* vuu, s32 info) +{ + // supervu will take care of the rest + int bpc = _recbranchAddr(VU->code); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + + // loops to self, so check condition + if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) + { + SuperVUTestVU0Condition(0); + } + + if (s_pCurBlock->blocks.size() > 1) + { + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + MOV32ItoR(s_JumpX86, 1); + s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; + s_UnconditionalDelay = 1; + } + + branch |= 3; +} + +void recVUMI_BAL(VURegs* vuu, s32 info) +{ + int bpc = _recbranchAddr(VU->code); + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0 || SUPERVU_CHECKCONDITION) + MOV32ItoM(SuperVUGetVIAddr(REG_TPC, 0), bpc); + + // loops to self, so check condition + if (bpc == s_pCurBlock->startpc && (s_vu == 0 || SUPERVU_CHECKCONDITION)) + { + SuperVUTestVU0Condition(0); + } + + if (_It_) + { + _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); + MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); + } + + if (s_pCurBlock->blocks.size() > 1) + { + s_JumpX86 = _allocX86reg(-1, X86TYPE_VUJUMP, 0, MODE_WRITE); + MOV32ItoR(s_JumpX86, 1); + s_pCurBlock->pChildJumps[(s_pCurInst->type & INST_BRANCH_DELAY)?1:0] = (u32*)x86Ptr - 1; + s_UnconditionalDelay = 1; + } + + branch |= 3; +} + +void recVUMI_JR(VURegs* vuu, s32 info) +{ + int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + LEA32RStoR(EAX, isreg, 3); + + //Mask the address to something valid + if (vuu == &VU0) + AND32ItoR(EAX, 0xfff); + else + AND32ItoR(EAX, 0x3fff); + + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); + + if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) + { + PUSH32I(s_vu); + PUSH32R(EAX); + } + branch |= 0x10; // 0x08 is reserved +} + +void recVUMI_JALR(VURegs* vuu, s32 info) +{ + _addNeededX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_); + + int isreg = _allocX86reg(-1, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + LEA32RStoR(EAX, isreg, 3); + + //Mask the address to something valid + if (vuu == &VU0) + AND32ItoR(EAX, 0xfff); + else + AND32ItoR(EAX, 0x3fff); + + if (_It_) + { + _deleteX86reg(X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _It_, 2); + MOV16ItoM(SuperVUGetVIAddr(_It_, 0), (pc + 8) >> 3); + } + + if ((s_pCurBlock->type & BLOCKTYPE_HASEOP) || s_vu == 0) MOV32RtoM(SuperVUGetVIAddr(REG_TPC, 0), EAX); + + if (!(s_pCurBlock->type & BLOCKTYPE_HASEOP)) + { + PUSH32I(s_vu); + PUSH32R(EAX); + } + + branch |= 4; +} + +#ifdef PCSX2_DEVBUILD +void vu1xgkick(u32* pMem, u32 addr) +{ + assert(addr < 0x4000); +#ifdef SUPERVU_COUNT + StopSVUCounter(); +#endif + + GSGIFTRANSFER1(pMem, addr); + +#ifdef SUPERVU_COUNT + StartSVUCounter(); +#endif +} +#endif + +void recVUMI_XGKICK_(VURegs *VU) +{ + assert(s_XGKICKReg > 0 && x86regs[s_XGKICKReg].inuse && x86regs[s_XGKICKReg].type == X86TYPE_VITEMP); + + x86regs[s_XGKICKReg].inuse = 0; // so free doesn't flush + _freeX86regs(); + _freeXMMregs(); + + PUSH32R(s_XGKICKReg); + PUSH32I((uptr)VU->Mem); + + //CALLFunc((u32)countfn); + + if (mtgsThread != NULL) + { + CALLFunc((uptr)VU1XGKICK_MTGSTransfer); + ADD32ItoR(ESP, 8); + } + else + { +#ifdef PCSX2_DEVBUILD + CALLFunc((uptr)vu1xgkick); + ADD32ItoR(ESP, 8); +#else + CALLFunc((uptr)GSgifTransfer1); +#endif + } + psHu32(GIF_STAT) &= ~(GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) + s_ScheduleXGKICK = 0; +} + +void recVUMI_XGKICK(VURegs *VU, int info) +{ + if (s_ScheduleXGKICK) + { + // second xgkick, so launch the first + recVUMI_XGKICK_(VU); + } + + int isreg = _allocX86reg(X86ARG2, X86TYPE_VI | (s_vu ? X86TYPE_VU1 : 0), _Is_, MODE_READ); + _freeX86reg(isreg); // flush + x86regs[isreg].inuse = 1; + x86regs[isreg].type = X86TYPE_VITEMP; + x86regs[isreg].needed = 1; + x86regs[isreg].mode = MODE_WRITE | MODE_READ; + SHL32ItoR(isreg, 4); + AND32ItoR(isreg, 0x3fff); + s_XGKICKReg = isreg; + psHu32(GIF_STAT) |= (GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) + + + if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) + { + recVUMI_XGKICK_(VU); + } + else + { + // Erementar Gerad hack. + // Corrects the color of some graphics on a game that has somewhat scrambled graphics either way, and only works with ZeroGS. Not even ZZOgl. :) + if (g_VUGameFixes & VUFIX_XGKICKDELAY2) + s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc - pc) / 8); + else + s_ScheduleXGKICK = 2; + } +} + +void recVU_UPPER_FD_00(VURegs* VU, s32 info); +void recVU_UPPER_FD_01(VURegs* VU, s32 info); +void recVU_UPPER_FD_10(VURegs* VU, s32 info); +void recVU_UPPER_FD_11(VURegs* VU, s32 info); +void recVULowerOP(VURegs* VU, s32 info); +void recVULowerOP_T3_00(VURegs* VU, s32 info); +void recVULowerOP_T3_01(VURegs* VU, s32 info); +void recVULowerOP_T3_10(VURegs* VU, s32 info); +void recVULowerOP_T3_11(VURegs* VU, s32 info); +void recVUunknown(VURegs* VU, s32 info); + +void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info) = +{ + recVUMI_LQ , recVUMI_SQ , recVUunknown , recVUunknown, + recVUMI_ILW , recVUMI_ISW , recVUunknown , recVUunknown, + recVUMI_IADDIU, recVUMI_ISUBIU, recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_FCEQ , recVUMI_FCSET , recVUMI_FCAND, recVUMI_FCOR, /* 0x10 */ + recVUMI_FSEQ , recVUMI_FSSET , recVUMI_FSAND, recVUMI_FSOR, + recVUMI_FMEQ , recVUunknown , recVUMI_FMAND, recVUMI_FMOR, + recVUMI_FCGET , recVUunknown , recVUunknown , recVUunknown, + recVUMI_B , recVUMI_BAL , recVUunknown , recVUunknown, /* 0x20 */ + recVUMI_JR , recVUMI_JALR , recVUunknown , recVUunknown, + recVUMI_IBEQ , recVUMI_IBNE , recVUunknown , recVUunknown, + recVUMI_IBLTZ , recVUMI_IBGTZ , recVUMI_IBLEZ, recVUMI_IBGEZ, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVULowerOP , recVUunknown , recVUunknown , recVUunknown, /* 0x40*/ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x50 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x60 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x70 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, +}; + +void (*recVULowerOP_T3_00_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_MOVE , recVUMI_LQI , recVUMI_DIV , recVUMI_MTIR, + recVUMI_RNEXT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_MFP , recVUMI_XTOP , recVUMI_XGKICK, + recVUMI_ESADD , recVUMI_EATANxy, recVUMI_ESQRT, recVUMI_ESIN, +}; + +void (*recVULowerOP_T3_01_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_MR32 , recVUMI_SQI , recVUMI_SQRT , recVUMI_MFIR, + recVUMI_RGET , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUMI_XITOP, recVUunknown, + recVUMI_ERSADD, recVUMI_EATANxz, recVUMI_ERSQRT, recVUMI_EATAN, +}; + +void (*recVULowerOP_T3_10_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_LQD , recVUMI_RSQRT, recVUMI_ILWR, + recVUMI_RINIT , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_ELENG , recVUMI_ESUM , recVUMI_ERCPR, recVUMI_EEXP, +}; + +void (*recVULowerOP_T3_11_OPCODE[32])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUMI_SQD , recVUMI_WAITQ, recVUMI_ISWR, + recVUMI_RXOR , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_ERLENG, recVUunknown , recVUMI_WAITP, recVUunknown, +}; + +void (*recVULowerOP_OPCODE[64])(VURegs* VU, s32 info) = +{ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x10 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x20 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUMI_IADD , recVUMI_ISUB , recVUMI_IADDI, recVUunknown, /* 0x30 */ + recVUMI_IAND , recVUMI_IOR , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVULowerOP_T3_00, recVULowerOP_T3_01, recVULowerOP_T3_10, recVULowerOP_T3_11, +}; + +void (*recVU_UPPER_OPCODE[64])(VURegs* VU, s32 info) = +{ + recVUMI_ADDx , recVUMI_ADDy , recVUMI_ADDz , recVUMI_ADDw, + recVUMI_SUBx , recVUMI_SUBy , recVUMI_SUBz , recVUMI_SUBw, + recVUMI_MADDx , recVUMI_MADDy , recVUMI_MADDz , recVUMI_MADDw, + recVUMI_MSUBx , recVUMI_MSUBy , recVUMI_MSUBz , recVUMI_MSUBw, + recVUMI_MAXx , recVUMI_MAXy , recVUMI_MAXz , recVUMI_MAXw, /* 0x10 */ + recVUMI_MINIx , recVUMI_MINIy , recVUMI_MINIz , recVUMI_MINIw, + recVUMI_MULx , recVUMI_MULy , recVUMI_MULz , recVUMI_MULw, + recVUMI_MULq , recVUMI_MAXi , recVUMI_MULi , recVUMI_MINIi, + recVUMI_ADDq , recVUMI_MADDq , recVUMI_ADDi , recVUMI_MADDi, /* 0x20 */ + recVUMI_SUBq , recVUMI_MSUBq , recVUMI_SUBi , recVUMI_MSUBi, + recVUMI_ADD , recVUMI_MADD , recVUMI_MUL , recVUMI_MAX, + recVUMI_SUB , recVUMI_MSUB , recVUMI_OPMSUB, recVUMI_MINI, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, /* 0x30 */ + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVUunknown , recVUunknown , recVUunknown , recVUunknown, + recVU_UPPER_FD_00, recVU_UPPER_FD_01, recVU_UPPER_FD_10, recVU_UPPER_FD_11, +}; + +void (*recVU_UPPER_FD_00_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAx, recVUMI_SUBAx , recVUMI_MADDAx, recVUMI_MSUBAx, + recVUMI_ITOF0, recVUMI_FTOI0, recVUMI_MULAx , recVUMI_MULAq , + recVUMI_ADDAq, recVUMI_SUBAq, recVUMI_ADDA , recVUMI_SUBA , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_01_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAy , recVUMI_SUBAy , recVUMI_MADDAy, recVUMI_MSUBAy, + recVUMI_ITOF4 , recVUMI_FTOI4 , recVUMI_MULAy , recVUMI_ABS , + recVUMI_MADDAq, recVUMI_MSUBAq, recVUMI_MADDA , recVUMI_MSUBA , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_10_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAz , recVUMI_SUBAz , recVUMI_MADDAz, recVUMI_MSUBAz, + recVUMI_ITOF12, recVUMI_FTOI12, recVUMI_MULAz , recVUMI_MULAi , + recVUMI_ADDAi, recVUMI_SUBAi , recVUMI_MULA , recVUMI_OPMULA, + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void (*recVU_UPPER_FD_11_TABLE[32])(VURegs* VU, s32 info) = +{ + recVUMI_ADDAw , recVUMI_SUBAw , recVUMI_MADDAw, recVUMI_MSUBAw, + recVUMI_ITOF15, recVUMI_FTOI15, recVUMI_MULAw , recVUMI_CLIP , + recVUMI_MADDAi, recVUMI_MSUBAi, recVUunknown , recVUMI_NOP , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , + recVUunknown , recVUunknown , recVUunknown , recVUunknown , +}; + +void recVU_UPPER_FD_00(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_00_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVU_UPPER_FD_01(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_01_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVU_UPPER_FD_10(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_10_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVU_UPPER_FD_11(VURegs* VU, s32 info) +{ + recVU_UPPER_FD_11_TABLE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP(VURegs* VU, s32 info) +{ + recVULowerOP_OPCODE[ VU->code & 0x3f ](VU, info); +} + +void recVULowerOP_T3_00(VURegs* VU, s32 info) +{ + recVULowerOP_T3_00_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_01(VURegs* VU, s32 info) +{ + recVULowerOP_T3_01_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_10(VURegs* VU, s32 info) +{ + recVULowerOP_T3_10_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVULowerOP_T3_11(VURegs* VU, s32 info) +{ + recVULowerOP_T3_11_OPCODE[(VU->code >> 6) & 0x1f ](VU, info); +} + +void recVUunknown(VURegs* VU, s32 info) +{ + Console::Notice("Unknown SVU micromode opcode called"); +} diff --git a/pcsx2/x86/iVUzerorec.h b/pcsx2/x86/sVU_zerorec.h similarity index 85% rename from pcsx2/x86/iVUzerorec.h rename to pcsx2/x86/sVU_zerorec.h index f81ec8d7b1..060ef7d9ce 100644 --- a/pcsx2/x86/iVUzerorec.h +++ b/pcsx2/x86/sVU_zerorec.h @@ -1,50 +1,47 @@ -/* Pcsx2 - Pc Ps2 Emulator - * Copyright (C) 2002-2009 Pcsx2 Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA - */ - -// Super VU recompiler - author: zerofrog(@gmail.com) - -#ifndef VU1_SUPER_RECOMPILER -#define VU1_SUPER_RECOMPILER - -#include "iVUmicro.h" - -extern void SuperVUAlloc(int vuindex); // global VU resources are automatically allocated if necessary. -extern void SuperVUDestroy(int vuindex); // if vuindex is -1, destroys everything -extern void SuperVUReset(int vuindex); // if vuindex is -1, resets everything - -//Using assembly code from an external file. -#ifdef __LINUX__ -extern "C" { -#endif -extern void SuperVUExecuteProgram(u32 startpc, int vuindex); -extern void SuperVUEndProgram(); -extern void svudispfntemp(); -#ifdef __LINUX__ -} -#endif -extern void __fastcall SuperVUClear(u32 startpc, u32 size, int vuindex); - -// read = 0, will write to reg -// read = 1, will read from reg -// read = 2, addr of previously written reg (used for status and clip flags) -extern u32 SuperVUGetVIAddr(int reg, int read); - -// if p == 0, flush q else flush p; if wait is != 0, waits for p/q -extern void SuperVUFlush(int p, int wait); - -#endif +/* Pcsx2 - Pc Ps2 Emulator + * Copyright (C) 2002-2009 Pcsx2 Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA + */ + +// Super VU recompiler - author: zerofrog(@gmail.com) + +#pragma once + +#include "sVU_Micro.h" + +extern void SuperVUAlloc(int vuindex); // global VU resources are automatically allocated if necessary. +extern void SuperVUDestroy(int vuindex); // if vuindex is -1, destroys everything +extern void SuperVUReset(int vuindex); // if vuindex is -1, resets everything + +//Using assembly code from an external file. +#ifdef __LINUX__ +extern "C" { +#endif +extern void SuperVUExecuteProgram(u32 startpc, int vuindex); +extern void SuperVUEndProgram(); +extern void svudispfntemp(); +#ifdef __LINUX__ +} +#endif +extern void __fastcall SuperVUClear(u32 startpc, u32 size, int vuindex); + +// read = 0, will write to reg +// read = 1, will read from reg +// read = 2, addr of previously written reg (used for status and clip flags) +extern u32 SuperVUGetVIAddr(int reg, int read); + +// if p == 0, flush q else flush p; if wait is != 0, waits for p/q +extern void SuperVUFlush(int p, int wait); From 6841f7fd9830021938a706c3fce1be02e3784973 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 11 Jun 2009 09:27:59 +0000 Subject: [PATCH 214/296] GSdx: that damned different pixel center of dx10... this one should fix the screen edge and other strange artifacts that appeared in non-native resolution. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1357 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSRendererHW10.cpp | 31 +++++++++++++++------------ plugins/GSdx/GSTextureFX10.h | 9 ++------ plugins/GSdx/res/tfx10.fx | 37 +++++++++++++++++---------------- 3 files changed, 39 insertions(+), 38 deletions(-) diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index f300ac44dd..09bec8dbe1 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -287,13 +287,13 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); - float ox = (float)((int)context->XYOFFSET.OFX - 8); - float oy = (float)((int)context->XYOFFSET.OFY - 8); - - // dx10 pixel center is different => offset - 8 (half pixel) + float ox = (float)(int)context->XYOFFSET.OFX; + float oy = (float)(int)context->XYOFFSET.OFY; + float ox2 = 1.0f / rt->GetWidth(); + float oy2 = 1.0f / rt->GetHeight(); vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); - vs_cb.VertexOffset = GSVector4(ox * sx + 1, -(oy * sy + 1), 0.0f, -1.0f); + vs_cb.VertexOffset = GSVector4(ox * sx - ox2 + 1, -(oy * sy - oy2 + 1), 0.0f, -1.0f); vs_cb.TextureScale = GSVector2(1.0f, 1.0f); if(PRIM->TME && PRIM->FST) @@ -337,8 +337,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache GSTextureFX10::PSConstantBuffer ps_cb; ps_cb.FogColorAREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - ps_cb.TA0 = (float)(int)env.TEXA.TA0 / 255; - ps_cb.TA1 = (float)(int)env.TEXA.TA1 / 255; + ps_cb.TA = GSVector4((int)env.TEXA.TA0, (int)env.TEXA.TA1) / 255; if(context->TEST.ATST == 2 || context->TEST.ATST == 5) { @@ -356,14 +355,16 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache int w = tex->m_texture->GetWidth(); int h = tex->m_texture->GetHeight(); - ps_cb.WH = GSVector2i(w, h); - switch(context->CLAMP.WMS) { case 0: + ps_cb.MinMax.x = w - 1; + ps_cb.MinMax.z = 0; ps_ssel.tau = 1; break; case 1: + ps_cb.MinMax.x = 0; + ps_cb.MinMax.z = w - 1; ps_ssel.tau = 0; break; case 2: @@ -374,8 +375,8 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tau = 0; break; case 3: - ps_cb.MskFix.x = context->CLAMP.MINU; - ps_cb.MskFix.z = context->CLAMP.MAXU; + ps_cb.MinMax.x = context->CLAMP.MINU; + ps_cb.MinMax.z = context->CLAMP.MAXU; ps_ssel.tau = 1; break; default: @@ -385,9 +386,13 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache switch(context->CLAMP.WMT) { case 0: + ps_cb.MinMax.y = h - 1; + ps_cb.MinMax.w = 0; ps_ssel.tav = 1; break; case 1: + ps_cb.MinMax.y = 0; + ps_cb.MinMax.w = h - 1; ps_ssel.tav = 0; break; case 2: @@ -398,8 +403,8 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_ssel.tav = 0; break; case 3: - ps_cb.MskFix.y = context->CLAMP.MINV; - ps_cb.MskFix.w = context->CLAMP.MAXV; + ps_cb.MinMax.y = context->CLAMP.MINV; + ps_cb.MinMax.w = context->CLAMP.MAXV; ps_ssel.tav = 1; break; default: diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 14e36c03e7..3775879d5f 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -78,12 +78,9 @@ public: __declspec(align(16)) struct PSConstantBuffer { GSVector4 FogColorAREF; - GSVector2i WH; - float TA0; - float TA1; + GSVector4 TA; GSVector4i MinMax; GSVector4 MinMaxF; - GSVector4i MskFix; struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} @@ -96,15 +93,13 @@ public: GSVector4i b1 = b[1]; GSVector4i b2 = b[2]; GSVector4i b3 = b[3]; - GSVector4i b4 = b[4]; - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4)).alltrue()) + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3)).alltrue()) { a[0] = b0; a[1] = b1; a[2] = b2; a[3] = b3; - a[4] = b4; return true; } diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index a16b2a8286..366b04188d 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -155,12 +155,9 @@ cbuffer cb1 { float3 FogColor; float AREF; - uint2 WH; - float TA0; - float TA1; + float4 TA; uint4 MinMax; float4 MinMaxF; - uint4 MskFix; }; struct PS_INPUT @@ -178,8 +175,8 @@ struct PS_OUTPUT #ifndef FST #define FST 0 -#define WMS 3 -#define WMT 3 +#define WMS 1 +#define WMT 1 #define BPP 0 #define AEM 0 #define TFX 0 @@ -216,28 +213,28 @@ int4 wrapuv(int4 uv) { switch(WMS) { - case 0: uv &= WH.xyxy - 1; break; - case 1: uv = clamp(uv, 0, WH.xyxy); break; + case 0: uv &= MinMax.xyxy; break; + case 1: uv = clamp(uv, 0, MinMax.zwzw); break; case 2: uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); break; - case 3: uv = (uv & MskFix.xyxy) | MskFix.zwzw; break; + case 3: uv = (uv & MinMax.xyxy) | MinMax.zwzw; break; } } else { switch(WMS) { - case 0: uv.xz &= WH.xx - 1; break; - case 1: uv.xz = clamp(uv.xz, 0, WH.xx); break; + case 0: uv.xz &= MinMax.xx; break; + case 1: uv.xz = clamp(uv.xz, 0, MinMax.zz); break; case 2: uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); break; - case 3: uv.xz = (uv.xz & MskFix.xx) | MskFix.zz; break; + case 3: uv.xz = (uv.xz & MinMax.xx) | MinMax.zz; break; } switch(WMT) { - case 0: uv.yw &= WH.yy - 1; break; - case 1: uv.yw = clamp(uv.yw, 0, WH.yy); break; + case 0: uv.yw &= MinMax.yy; break; + case 1: uv.yw = clamp(uv.yw, 0, MinMax.ww); break; case 2: uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); break; - case 3: uv.yw = (uv.yw & MskFix.yy) | MskFix.ww; break; + case 3: uv.yw = (uv.yw & MinMax.yy) | MinMax.ww; break; } } @@ -282,7 +279,11 @@ float4 sample(float2 tc, float w) } else { - float4 tc2 = (tc * WH).xyxy + float4(-0.499, -0.499, 0.501, 0.501); + float w, h; + + Texture.GetDimensions(w, h); + + float4 tc2 = (tc * float2(w, h)).xyxy + float4(-0.499, -0.499, 0.501, 0.501); float2 dd = frac(tc2.xy); int4 uv = wrapuv((int4)tc2); @@ -342,7 +343,7 @@ float4 sample(float2 tc, float w) if(BPP == 1) // 24 { - t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; + t.a = AEM == 0 || any(t.rgb) ? TA.x : 0; } else if(BPP == 2 || BPP == 5) // 16 || 16P { @@ -353,7 +354,7 @@ float4 sample(float2 tc, float w) // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; + t.a = t.a >= 0.5 ? TA.y : AEM == 0 || any(t.rgb) ? TA.x : 0; } return t; From 1c3a537d5cb9b665d9189bb0f2372812a72a2cd0 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 11 Jun 2009 12:56:55 +0000 Subject: [PATCH 215/296] Fix up Linux after r1356. Make a few changes that ought to make -fpermissive unneccessary, though it needs more testing before I actually remove said flag. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1358 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IopHw.cpp | 2 +- pcsx2/Linux/ConfigDlg.cpp | 6 ++-- pcsx2/Memory.cpp | 6 ++-- pcsx2/vtlb.cpp | 20 ++++++------ pcsx2/x86/Makefile.am | 18 +++++------ pcsx2/x86/aMicroVU.S | 62 -------------------------------------- pcsx2/x86/microVU_Misc.h | 4 +++ pcsx2/x86/microVU_Misc.inl | 2 ++ 8 files changed, 32 insertions(+), 88 deletions(-) delete mode 100644 pcsx2/x86/aMicroVU.S diff --git a/pcsx2/IopHw.cpp b/pcsx2/IopHw.cpp index 2b4635fb20..53ad416e20 100644 --- a/pcsx2/IopHw.cpp +++ b/pcsx2/IopHw.cpp @@ -618,7 +618,7 @@ u32 psxHwRead32(u32 add) { } // A buffer that stores messages until it gets a /n or the number of chars (g_pbufi) is more then 1023. -static s8 g_pbuf[1024]; +static char g_pbuf[1024]; static int g_pbufi; void psxHwWrite8(u32 add, u8 value) { if (add >= HW_USB_START && add < HW_USB_END) { diff --git a/pcsx2/Linux/ConfigDlg.cpp b/pcsx2/Linux/ConfigDlg.cpp index f37a141cc3..dd0f3f924d 100644 --- a/pcsx2/Linux/ConfigDlg.cpp +++ b/pcsx2/Linux/ConfigDlg.cpp @@ -134,7 +134,7 @@ void OnConfConf_Ok(GtkButton *button, gpointer user_data) plugin_types type; applychanges = TRUE; - for (type = GS; type <= BIOS; type = type + 1) + for (type = GS; type <= BIOS; type = (plugin_types)((int)type + 1)) { PluginConf *confs = ConfS(type); @@ -188,7 +188,7 @@ void UpdateConfDlg() plugin_types type; FindPlugins(); - for (type = GS; type <= BIOS; type = type + 1) + for (type = GS; type <= BIOS; type = (plugin_types)((int)type + 1)) { char tmp[50]; PluginConf *confs = ConfS(type); @@ -274,7 +274,7 @@ void FindPlugins() char plugin[g_MaxPath], name[g_MaxPath]; plugin_types type; - for (type = GS; type <= BIOS; type = type + 1) + for (type = GS; type <= BIOS; type = (plugin_types)((int)type + 1)) { PluginConf *confs = ConfS(type); diff --git a/pcsx2/Memory.cpp b/pcsx2/Memory.cpp index 050a5cc44d..8a7f22fc9d 100644 --- a/pcsx2/Memory.cpp +++ b/pcsx2/Memory.cpp @@ -927,9 +927,9 @@ void mmap_ClearCpuBlock( uint offset ) // manual protection. Indicates a logic error in the recompiler or protection code. jASSUME( m_PageProtectInfo[rampage].Mode != ProtMode_Manual ); - #ifndef __LINUX__ // this function is called from the signal handler - DbgCon::WriteLn( "Manual page @ 0x%05x", params m_PageProtectInfo[rampage].ReverseRamMap>>12 ); - #endif + //#ifndef __LINUX__ // this function is called from the signal handler + //DbgCon::WriteLn( "Manual page @ 0x%05x", params m_PageProtectInfo[rampage].ReverseRamMap>>12 ); + //#endif HostSys::MemProtect( &psM[rampage<<12], 1, Protect_ReadWrite ); m_PageProtectInfo[rampage].Mode = ProtMode_Manual; diff --git a/pcsx2/vtlb.cpp b/pcsx2/vtlb.cpp index 76edac5fb4..69692bd7e2 100644 --- a/pcsx2/vtlb.cpp +++ b/pcsx2/vtlb.cpp @@ -324,17 +324,17 @@ vtlbHandler vtlb_RegisterHandler( vtlbMemR8FP* r8,vtlbMemR16FP* r16,vtlbMemR32FP //write the code :p vtlbHandler rv=vtlbHandlerCount++; - vtlbdata.RWFT[0][0][rv] = (r8!=0) ? r8:vtlbDefaultPhyRead8; - vtlbdata.RWFT[1][0][rv] = (r16!=0) ? r16:vtlbDefaultPhyRead16; - vtlbdata.RWFT[2][0][rv] = (r32!=0) ? r32:vtlbDefaultPhyRead32; - vtlbdata.RWFT[3][0][rv] = (r64!=0) ? r64:vtlbDefaultPhyRead64; - vtlbdata.RWFT[4][0][rv] = (r128!=0) ? r128:vtlbDefaultPhyRead128; + vtlbdata.RWFT[0][0][rv] = (r8!=0) ? (void*)(r8): (void*)vtlbDefaultPhyRead8; + vtlbdata.RWFT[1][0][rv] = (r16!=0) ? (void*)r16: (void*)vtlbDefaultPhyRead16; + vtlbdata.RWFT[2][0][rv] = (r32!=0) ? (void*)r32: (void*)vtlbDefaultPhyRead32; + vtlbdata.RWFT[3][0][rv] = (r64!=0) ? (void*)r64: (void*)vtlbDefaultPhyRead64; + vtlbdata.RWFT[4][0][rv] = (r128!=0) ? (void*)r128: (void*)vtlbDefaultPhyRead128; - vtlbdata.RWFT[0][1][rv] = (w8!=0) ? w8:vtlbDefaultPhyWrite8; - vtlbdata.RWFT[1][1][rv] = (w16!=0) ? w16:vtlbDefaultPhyWrite16; - vtlbdata.RWFT[2][1][rv] = (w32!=0) ? w32:vtlbDefaultPhyWrite32; - vtlbdata.RWFT[3][1][rv] = (w64!=0) ? w64:vtlbDefaultPhyWrite64; - vtlbdata.RWFT[4][1][rv] = (w128!=0) ? w128:vtlbDefaultPhyWrite128; + vtlbdata.RWFT[0][1][rv] = (w8!=0) ? (void*)w8: (void*)vtlbDefaultPhyWrite8; + vtlbdata.RWFT[1][1][rv] = (w16!=0) ? (void*)w16: (void*)vtlbDefaultPhyWrite16; + vtlbdata.RWFT[2][1][rv] = (w32!=0) ? (void*)w32: (void*)vtlbDefaultPhyWrite32; + vtlbdata.RWFT[3][1][rv] = (w64!=0) ? (void*)w64: (void*)vtlbDefaultPhyWrite64; + vtlbdata.RWFT[4][1][rv] = (w128!=0) ? (void*)w128: (void*)vtlbDefaultPhyWrite128; return rv; } diff --git a/pcsx2/x86/Makefile.am b/pcsx2/x86/Makefile.am index cf43eabb46..24fa916809 100644 --- a/pcsx2/x86/Makefile.am +++ b/pcsx2/x86/Makefile.am @@ -18,19 +18,19 @@ ix86-32/iR5900Shift.cpp ix86-32/iR5900Arit.cpp ix86-32/iR5900Branch.cpp ix86-32 ix86-32/iR5900MultDiv.cpp ix86-32/iCore-32.cpp ix86-32/aR5900-32.S ix86-32/iR5900Templates.cpp ix86-32/recVTLB.cpp libx86recomp_a_SOURCES = \ -BaseblockEx.cpp iCOP0.cpp iCOP2.cpp iCore.cpp iFPU.cpp iFPUd.cpp iMMI.cpp iPsxMem.cpp iR3000A.cpp iR3000Atables.cpp \ -iR5900Misc.cpp iVU0micro.cpp iVU1micro.cpp iVUmicro.cpp iVUmicroLower.cpp iVUmicroUpper.cpp iVUzerorec.cpp iVif.cpp \ -ir5900tables.cpp fast_routines.S aR3000A.S aVUzerorec.S aVif.S $(archfiles) +BaseblockEx.cpp iCOP2.cpp iFPUd.cpp iR3000A.cpp iVU0micro.cpp ir5900tables.cpp sVU_Micro.cpp \ +iCore.cpp iMMI.cpp iR3000Atables.cpp iVU1micro.cpp microVU.cpp sVU_Upper.cpp \ +iCOP0.cpp iFPU.cpp iPsxMem.cpp iR5900Misc.cpp iVif.cpp sVU_Lower.cpp sVU_zerorec.cpp \ +aR3000A.S aVUzerorec.S aVif.S fast_routines.S $(archfiles) libx86recomp_a_SOURCES += \ -BaseblockEx.h iCOP0.h iCore.h iFPU.h iMMI.h iR3000A.h iR5900.h iR5900Arit.h iR5900AritImm.h iR5900Branch.h iR5900Jump.h \ -iR5900LoadStore.h iR5900Move.h iR5900MultDiv.h iR5900Shift.h iVUmicro.h iVUops.h iVUzerorec.h +microVU_Alloc.inl microVU_Compile.inl microVU_Flags.inl microVU_Lower.inl microVU_Tables.inl \ +microVU_Analyze.inl microVU_Execute.inl microVU_Log.inl microVU_Misc.inl microVU_Upper.inl -# Only active if PCSX2_MICROVU is defined. libx86recomp_a_SOURCES += \ -microVU.cpp microVU_Misc.inl microVU_Log.inl microVU_Analyze.inl microVU_Alloc.inl microVU_Upper.inl microVU_Lower.inl \ -microVU_Tables.inl microVU_Flags.inl microVU_Compile.inl microVU_Execute.inl \ -microVU.h microVU_Alloc.h microVU_Misc.h +BaseblockEx.h iFPU.h iR5900.h iR5900Branch.h iR5900Move.h microVU.h sVU_Debug.h \ +iCOP0.h iMMI.h iR5900Arit.h iR5900Jump.h iR5900MultDiv.h microVU_IR.h sVU_Micro.h \ +iCore.h iR3000A.h iR5900AritImm.h iR5900LoadStore.h iR5900Shift.h microVU_Misc.h sVU_zerorec.h libx86recomp_a_DEPENDENCIES = ix86/libix86.a diff --git a/pcsx2/x86/aMicroVU.S b/pcsx2/x86/aMicroVU.S deleted file mode 100644 index 04bce4fd11..0000000000 --- a/pcsx2/x86/aMicroVU.S +++ /dev/null @@ -1,62 +0,0 @@ -// microVU.cpp assembly routines -// arcum42(@gmail.com) -.intel_syntax noprefix - -.extern mVUexecuteVU0 -.extern mVUexecuteVU1 -.extern g_sseVUMXCSR -.extern g_sseMXCSR - -//------------------------------------------------------------------ -// Dispatcher Functions -//------------------------------------------------------------------ - -// Runs VU0 for number of cycles -// __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. -//void __fastcall startVU0(u32 startPC, u32 cycles) -.globl startVU0 -startVU0: - call mVUexecuteVU0 - - // backup cpu state - push ebx - push ebp - push esi - push edi - - ldmxcsr g_sseVUMXCSR - // Should set xmmZ? - jmp eax - -// Runs VU1 for number of cycles -// void __fastcall startVU1(u32 startPC, u32 cycles) -.globl startVU1 -startVU01: - call mVUexecuteVU1 - - // backup cpu state - push ebx - push ebp - push esi - push edi - - ldmxcsr g_sseVUMXCSR - - jmp eax - -// Exit point -// void __fastcall endVU0(u32 startPC, u32 cycles) -.globl endVU0 -endVU0: - //call mVUcleanUpVU0 - - /*restore cpu state*/ - pop edi; - pop esi; - pop ebp; - pop ebx; - - ldmxcsr g_sseMXCSR - - ret - \ No newline at end of file diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 685356beca..5538fcc26d 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -162,7 +162,10 @@ declareAllVariables #define pass4 if (recPass == 3) // Define mVUquickSearch +#ifndef __LINUX__ extern u8 mVUsearchXMM[0x1000]; +#endif + typedef u32 (__fastcall *mVUCall)(void*, void*); #ifndef __LINUX__ #define mVUquickSearch(dest, src, size) ((((mVUCall)((void*)mVUsearchXMM))(dest, src)) == 0xf) @@ -172,6 +175,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #else //#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) #define mVUquickSearch(dest, src, size) (!memcmp_mmx(dest, src, size)) +#define mVUemitSearch() #endif // Misc Macros... diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index ca60c20b33..9fbd866547 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -438,6 +438,7 @@ void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { // Micro VU - Custom Quick Search //------------------------------------------------------------------ +#ifndef __LINUX__ PCSX2_ALIGNED(0x1000, static u8 mVUsearchXMM[0x1000]); // Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) @@ -490,3 +491,4 @@ void mVUemitSearch() { xRET(); HostSys::MemProtect(mVUsearchXMM, 0x1000, Protect_ReadOnly, true ); } +#endif From 5736bbd449b2bdc5d4c02d3766aa00bbd08623e4 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 12 Jun 2009 19:09:17 +0000 Subject: [PATCH 216/296] GSdx: looking for an opengl guru... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1360 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPUDrawScanline.cpp | 4 +- plugins/GSdx/GPUSettingsDlg.cpp | 22 --- plugins/GSdx/GPUSettingsDlg.h | 3 +- plugins/GSdx/GS.cpp | 8 +- plugins/GSdx/GSDevice.h | 3 - plugins/GSdx/GSDevice9.cpp | 20 +-- plugins/GSdx/GSDevice9.h | 1 - plugins/GSdx/GSDeviceOGL.cpp | 211 ++++++++++++++++++++++++++++ plugins/GSdx/GSDeviceOGL.h | 70 +++++++++ plugins/GSdx/GSDrawScanline.cpp | 6 +- plugins/GSdx/GSFunctionMap.h | 2 +- plugins/GSdx/GSRenderer.cpp | 7 +- plugins/GSdx/GSRendererOGL.cpp | 80 +++++++++++ plugins/GSdx/GSRendererOGL.h | 40 ++++++ plugins/GSdx/GSSettingsDlg.cpp | 34 ++--- plugins/GSdx/GSSettingsDlg.h | 3 +- plugins/GSdx/GSTexture.h | 10 +- plugins/GSdx/GSTexture10.cpp | 11 +- plugins/GSdx/GSTexture10.h | 4 +- plugins/GSdx/GSTexture7.cpp | 11 +- plugins/GSdx/GSTexture7.h | 4 +- plugins/GSdx/GSTexture9.cpp | 11 +- plugins/GSdx/GSTexture9.h | 4 +- plugins/GSdx/GSTextureCache.cpp | 73 +++++++++- plugins/GSdx/GSTextureCache.h | 4 + plugins/GSdx/GSTextureCache10.cpp | 64 ++------- plugins/GSdx/GSTextureCache10.h | 3 - plugins/GSdx/GSTextureCache9.cpp | 65 ++------- plugins/GSdx/GSTextureCache9.h | 3 - plugins/GSdx/GSTextureCacheOGL.cpp | 62 ++++++++ plugins/GSdx/GSTextureCacheOGL.h | 60 ++++++++ plugins/GSdx/GSTextureFX9.cpp | 9 +- plugins/GSdx/GSTextureNull.h | 4 +- plugins/GSdx/GSTextureOGL.cpp | 84 +++++++++++ plugins/GSdx/GSTextureOGL.h | 47 +++++++ plugins/GSdx/GSVertexHW.h | 2 + plugins/GSdx/GSVertexSW.h | 2 +- plugins/GSdx/GSdx.rc | 83 +++++------ plugins/GSdx/GSdx_vs2008.vcproj | 32 +++++ plugins/GSdx/res/tfx10.fx | 2 + plugins/GSdx/stdafx.h | 10 ++ plugins/GSdx/vsprops/common.vsprops | 2 +- 42 files changed, 878 insertions(+), 302 deletions(-) create mode 100644 plugins/GSdx/GSDeviceOGL.cpp create mode 100644 plugins/GSdx/GSDeviceOGL.h create mode 100644 plugins/GSdx/GSRendererOGL.cpp create mode 100644 plugins/GSdx/GSRendererOGL.h create mode 100644 plugins/GSdx/GSTextureCacheOGL.cpp create mode 100644 plugins/GSdx/GSTextureCacheOGL.h create mode 100644 plugins/GSdx/GSTextureOGL.cpp create mode 100644 plugins/GSdx/GSTextureOGL.h diff --git a/plugins/GSdx/GPUDrawScanline.cpp b/plugins/GSdx/GPUDrawScanline.cpp index e067278dd0..a1e0e2cabb 100644 --- a/plugins/GSdx/GPUDrawScanline.cpp +++ b/plugins/GSdx/GPUDrawScanline.cpp @@ -69,7 +69,7 @@ void GPUDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) // - f->ssl = m_ds.Lookup(m_env.sel); + f->ssl = m_ds[m_env.sel]; f->sr = NULL; // TODO @@ -84,7 +84,7 @@ void GPUDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) sel.twin = m_env.sel.twin; sel.sprite = m_env.sel.sprite; - f->ssp = m_sp.Lookup(sel); + f->ssp = m_sp[sel]; } void GPUDrawScanline::EndDraw(const GSRasterizerStats& stats) diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 9b159e3130..56721daab8 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -33,15 +33,6 @@ GSSetting GPUSettingsDlg::g_renderers[] = // {3, "Null (Null)", NULL}, }; -GSSetting GPUSettingsDlg::g_psversion[] = -{ - {D3DPS_VERSION(3, 0), "Pixel Shader 3.0", NULL}, - {D3DPS_VERSION(2, 0), "Pixel Shader 2.0", NULL}, - //{D3DPS_VERSION(1, 4), "Pixel Shader 1.4", NULL}, - //{D3DPS_VERSION(1, 1), "Pixel Shader 1.1", NULL}, - //{D3DPS_VERSION(0, 0), "Fixed Pipeline (bogus)", NULL}, -}; - GSSetting GPUSettingsDlg::g_filter[] = { {0, "Nearest", NULL}, @@ -82,10 +73,6 @@ void GPUSettingsDlg::OnInit() { __super::OnInit(); - D3DCAPS9 caps; - memset(&caps, 0, sizeof(caps)); - caps.PixelShaderVersion = D3DPS_VERSION(0, 0); - m_modes.clear(); { @@ -116,8 +103,6 @@ void GPUSettingsDlg::OnInit() ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } - - d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); } } @@ -133,7 +118,6 @@ void GPUSettingsDlg::OnInit() } ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); ComboBoxInit(GetDlgItem(m_hWnd, IDC_FILTER), g_filter, countof(g_filter), theApp.GetConfig("filter", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_DITHERING), g_dithering, countof(g_dithering), theApp.GetConfig("dithering", 1)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); @@ -169,11 +153,6 @@ bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) theApp.SetConfig("Renderer", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SHADER), data)) - { - theApp.SetConfig("PixelShaderVersion2", (int)data); - } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_FILTER), data)) { theApp.SetConfig("filter", (int)data); @@ -214,7 +193,6 @@ void GPUSettingsDlg::UpdateControls() ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); - EnableWindow(GetDlgItem(m_hWnd, IDC_SHADER), dx9); EnableWindow(GetDlgItem(m_hWnd, IDC_SCALE), sw); EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS_EDIT), sw); EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS), sw); diff --git a/plugins/GSdx/GPUSettingsDlg.h b/plugins/GSdx/GPUSettingsDlg.h index f450ce647d..ef6a764ef2 100644 --- a/plugins/GSdx/GPUSettingsDlg.h +++ b/plugins/GSdx/GPUSettingsDlg.h @@ -37,8 +37,7 @@ protected: public: GPUSettingsDlg(); - static GSSetting g_renderers[]; - static GSSetting g_psversion[]; + static GSSetting g_renderers[]; static GSSetting g_filter[]; static GSSetting g_dithering[]; static GSSetting g_aspectratio[]; diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 3cc0cf8481..972fe0c5a0 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -23,6 +23,7 @@ #include "GSUtil.h" #include "GSRendererHW9.h" #include "GSRendererHW10.h" +#include "GSRendererOGL.h" #include "GSRendererSW.h" #include "GSRendererNull.h" #include "GSSettingsDlg.h" @@ -125,8 +126,11 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; - case 6: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; - case 7: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; + case 6: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; + case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; + case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; + case 9: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; + case 10: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; } if(!s_gs->Create(title)) diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index 41502653f8..d512fd37b5 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -103,7 +103,4 @@ public: void Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, const GSVector4& c); void Interlace(const GSVector2i& ds, int field, int mode, float yoffset); - - virtual void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) {} - virtual void OMSetRenderTargets(GSTexture* rt, GSTexture* ds) {} }; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index af728d7b42..3d7dd3a93e 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -114,23 +114,17 @@ bool GSDevice9::Create(GSWnd* wnd, bool vsync) // shaders - uint32 psver = (uint32)theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)); - - if(psver > m_d3dcaps.PixelShaderVersion) + if(m_d3dcaps.PixelShaderVersion < D3DPS_VERSION(2, 0)) { string s = format( - "Supported pixel shader version is too low!\n\nSupported: %d.%d\nSelected: %d.%d", - D3DSHADER_VERSION_MAJOR(m_d3dcaps.PixelShaderVersion), D3DSHADER_VERSION_MINOR(m_d3dcaps.PixelShaderVersion), - D3DSHADER_VERSION_MAJOR(psver), D3DSHADER_VERSION_MINOR(psver)); + "Supported pixel shader version is too low!\n\nSupported: %d.%d\nNeeded: 2.0 or higher", + D3DSHADER_VERSION_MAJOR(m_d3dcaps.PixelShaderVersion), D3DSHADER_VERSION_MINOR(m_d3dcaps.PixelShaderVersion)); MessageBox(NULL, s.c_str(), "GSdx", MB_OK); return false; } - m_d3dcaps.PixelShaderVersion = min(psver, m_d3dcaps.PixelShaderVersion); - m_d3dcaps.VertexShaderVersion = m_d3dcaps.PixelShaderVersion & ~0x10000; - // convert static const D3DVERTEXELEMENT9 il_convert[] = @@ -224,8 +218,6 @@ bool GSDevice9::Reset(int w, int h, bool fs) m_swapchain = NULL; - if(m_font) {m_font->OnLostDevice(); m_font = NULL;} - if(m_vs_cb) _aligned_free(m_vs_cb); if(m_ps_cb) _aligned_free(m_ps_cb); @@ -326,12 +318,6 @@ bool GSDevice9::Reset(int w, int h, bool fs) m_backbuffer = new GSTexture9(backbuffer); - D3DXFONT_DESC fd; - memset(&fd, 0, sizeof(fd)); - _tcscpy(fd.FaceName, _T("Arial")); - fd.Height = 20; - D3DXCreateFontIndirect(m_dev, &fd, &m_font); - m_dev->SetRenderState(D3DRS_CULLMODE, D3DCULL_NONE); m_dev->SetRenderState(D3DRS_LIGHTING, FALSE); m_dev->SetRenderState(D3DRS_ALPHATESTENABLE, FALSE); diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 13be056227..304c8ae502 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -104,7 +104,6 @@ private: public: // TODO D3DPRESENT_PARAMETERS m_pp; - CComPtr m_font; struct { diff --git a/plugins/GSdx/GSDeviceOGL.cpp b/plugins/GSdx/GSDeviceOGL.cpp new file mode 100644 index 0000000000..55ffa50648 --- /dev/null +++ b/plugins/GSdx/GSDeviceOGL.cpp @@ -0,0 +1,211 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSdx.h" +#include "GSDeviceOGL.h" +#include "resource.h" + +GSDeviceOGL::GSDeviceOGL() + : m_hDC(NULL) + , m_hGLRC(NULL) +{ +} + +GSDeviceOGL::~GSDeviceOGL() +{ + #ifdef _WINDOWS + + if(m_hGLRC) {wglMakeCurrent(NULL, NULL); wglDeleteContext(m_hGLRC);} + if(m_hDC) ReleaseDC((HWND)m_wnd->GetHandle(), m_hDC); + + #endif +} + +void GSDeviceOGL::OnCgError(CGcontext ctx, CGerror err) +{ + printf("%s\n", cgGetErrorString(err)); + printf("%s\n", cgGetLastListing(ctx)); // ? +} + +bool GSDeviceOGL::Create(GSWnd* wnd, bool vsync) +{ + if(!__super::Create(wnd, vsync)) + { + return false; + } + + cgSetErrorHandler(OnStaticCgError, this); + + #ifdef _WINDOWS + + PIXELFORMATDESCRIPTOR pfd; + + memset(&pfd, 0, sizeof(pfd)); + + pfd.nSize = sizeof(pfd); + pfd.nVersion = 1; + pfd.dwFlags = PFD_DRAW_TO_WINDOW | PFD_SUPPORT_OPENGL | PFD_DOUBLEBUFFER; + pfd.iPixelType = PFD_TYPE_RGBA; + pfd.cColorBits = 32; + pfd.cDepthBits = 32; // 24? + pfd.cStencilBits = 8; + pfd.iLayerType = PFD_MAIN_PLANE; + + m_hDC = GetDC((HWND)m_wnd->GetHandle()); + + if(!m_hDC) return false; + + if(!SetPixelFormat(m_hDC, ChoosePixelFormat(m_hDC, &pfd), &pfd)) + return false; + + m_hGLRC = wglCreateContext(m_hDC); + + if(!m_hGLRC) return false; + + if(!wglMakeCurrent(m_hDC, m_hGLRC)) + return false; + + #endif + + // const char* exts = (const char*)glGetString(GL_EXTENSIONS); + + // TODO + + Reset(1, 1, true); + + // + + return true; +} + +bool GSDeviceOGL::Reset(int w, int h, bool fs) +{ + if(!__super::Reset(w, h, fs)) + return false; + + m_backbuffer = new GSTextureOGL(0); // ??? + + glCullFace(GL_FRONT_AND_BACK); + glDisable(GL_LIGHTING); + glDisable(GL_ALPHA_TEST); + glDisable(GL_SCISSOR_TEST); + + glMatrixMode(GL_PROJECTION); + glLoadIdentity(); + glMatrixMode(GL_MODELVIEW); + glLoadIdentity(); + + glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); + + // glViewport(0, 0, ?, ?); + + return true; +} + +void GSDeviceOGL::Flip() +{ + // TODO +} + +void GSDeviceOGL::BeginScene() +{ + // TODO +} + +void GSDeviceOGL::DrawPrimitive() +{ + // TODO +} + +void GSDeviceOGL::EndScene() +{ + // TODO +} + +void GSDeviceOGL::ClearRenderTarget(GSTexture* t, const GSVector4& c) +{ + // TODO +} + +void GSDeviceOGL::ClearRenderTarget(GSTexture* t, uint32 c) +{ + // TODO +} + +void GSDeviceOGL::ClearDepth(GSTexture* t, float c) +{ + // TODO +} + +void GSDeviceOGL::ClearStencil(GSTexture* t, uint8 c) +{ + // TODO +} + +GSTexture* GSDeviceOGL::Create(int type, int w, int h, int format) +{ + // TODO + + return NULL; +} + +GSTexture* GSDeviceOGL::CreateRenderTarget(int w, int h, int format) +{ + return __super::CreateRenderTarget(w, h, format ? format : 0); // TODO +} + +GSTexture* GSDeviceOGL::CreateDepthStencil(int w, int h, int format) +{ + return __super::CreateDepthStencil(w, h, format ? format : 0); // TODO +} + +GSTexture* GSDeviceOGL::CreateTexture(int w, int h, int format) +{ + return __super::CreateTexture(w, h, format ? format : 0); // TODO +} + +GSTexture* GSDeviceOGL::CreateOffscreen(int w, int h, int format) +{ + return __super::CreateOffscreen(w, h, format ? format : 0); // TODO +} + +GSTexture* GSDeviceOGL::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format) +{ + // TODO + + return NULL; +} + +void GSDeviceOGL::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader, bool linear) +{ + // TODO +} + +void GSDeviceOGL::DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) +{ + // TODO +} + +void GSDeviceOGL::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) +{ + // TODO +} diff --git a/plugins/GSdx/GSDeviceOGL.h b/plugins/GSdx/GSDeviceOGL.h new file mode 100644 index 0000000000..0b7ab393f4 --- /dev/null +++ b/plugins/GSdx/GSDeviceOGL.h @@ -0,0 +1,70 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSDevice.h" +#include "GSTextureOGL.h" + +class GSDeviceOGL : public GSDevice +{ + #ifdef _WINDOWS + + HDC m_hDC; + HGLRC m_hGLRC; + + #endif + + static void OnStaticCgError(CGcontext ctx, CGerror err, void* p) {((GSDeviceOGL*)p)->OnCgError(ctx, err);} + void OnCgError(CGcontext ctx, CGerror err); + + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0); + +public: + GSDeviceOGL(); + virtual ~GSDeviceOGL(); + + bool Create(GSWnd* wnd, bool vsync); + bool Reset(int w, int h, bool fs); + + void Flip(); + + void BeginScene(); + void DrawPrimitive(); + void EndScene(); + + void ClearRenderTarget(GSTexture* t, const GSVector4& c); + void ClearRenderTarget(GSTexture* t, uint32 c); + void ClearDepth(GSTexture* t, float c); + void ClearStencil(GSTexture* t, uint8 c); + + GSTexture* CreateRenderTarget(int w, int h, int format = 0); + GSTexture* CreateDepthStencil(int w, int h, int format = 0); + GSTexture* CreateTexture(int w, int h, int format = 0); + GSTexture* CreateOffscreen(int w, int h, int format = 0); + + GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); + + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); +}; diff --git a/plugins/GSdx/GSDrawScanline.cpp b/plugins/GSdx/GSDrawScanline.cpp index c3452fe05c..8370c4ab04 100644 --- a/plugins/GSdx/GSDrawScanline.cpp +++ b/plugins/GSdx/GSDrawScanline.cpp @@ -163,7 +163,7 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) // - f->ssl = m_ds.Lookup(m_sel); + f->ssl = m_ds[m_sel]; if(m_sel.aa1)// && (m_state->m_perfmon.GetFrame() & 0x40)) { @@ -173,7 +173,7 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) sel.zwrite = 0; sel.edge = 1; - f->ssle = m_ds.Lookup(sel); + f->ssle = m_ds[sel]; } if(m_sel.IsSolidRect()) @@ -197,7 +197,7 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data, Functions* f) sel.zb = m_sel.zb; sel.zoverflow = m_sel.zoverflow; - f->ssp = m_sp.Lookup(sel); + f->ssp = m_sp[sel]; } void GSDrawScanline::EndDraw(const GSRasterizerStats& stats) diff --git a/plugins/GSdx/GSFunctionMap.h b/plugins/GSdx/GSFunctionMap.h index b0acb7daec..b825e7ea5a 100644 --- a/plugins/GSdx/GSFunctionMap.h +++ b/plugins/GSdx/GSFunctionMap.h @@ -72,7 +72,7 @@ public: } } - VALUE Lookup(KEY key) + VALUE operator [] (KEY key) { m_active = NULL; diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 0096c85607..288974c0f8 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -360,12 +360,11 @@ void GSRenderer::VSync(int field) if(GSTexture* offscreen = m_dev->CopyOffscreen(current, GSVector4(0, 0, 1, 1), size.x, size.y)) { - uint8* bits = NULL; - int pitch = 0; + GSTexture::GSMap m; - if(offscreen->Map(&bits, pitch)) + if(offscreen->Map(m)) { - m_capture.DeliverFrame(bits, pitch, m_dev->IsCurrentRGBA()); + m_capture.DeliverFrame(m.bits, m.pitch, m_dev->IsCurrentRGBA()); offscreen->Unmap(); } diff --git a/plugins/GSdx/GSRendererOGL.cpp b/plugins/GSdx/GSRendererOGL.cpp new file mode 100644 index 0000000000..942734e7d2 --- /dev/null +++ b/plugins/GSdx/GSRendererOGL.cpp @@ -0,0 +1,80 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererOGL.h" +#include "GSCrc.h" +#include "resource.h" + +GSRendererOGL::GSRendererOGL(uint8* base, bool mt, void (*irq)()) + : GSRendererHW(base, mt, irq, new GSDeviceOGL(), new GSTextureCacheOGL(this)) +{ + InitVertexKick(); +} + +bool GSRendererOGL::Create(const string& title) +{ + if(!__super::Create(title)) + return false; + + // TODO + + /* + + if(!m_tfx.Create((GSDeviceOGL*)m_dev)) + return false; + + */ + + return true; +} + +template +void GSRendererOGL::VertexKick(bool skip) +{ + GSVertexOGL& dst = m_vl.AddTail(); + + // TODO + + int count = 0; + + if(GSVertexOGL* v = DrawingKick(skip, count)) + { + // TODO + + m_count += count; + } +} + +void GSRendererOGL::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +{ + GSDrawingEnvironment& env = m_env; + GSDrawingContext* context = m_context; + + // m_perfmon.Put(GSPerfMon::Prim, prims); + m_perfmon.Put(GSPerfMon::Draw, 1); + + m_dev->BeginScene(); + + // TODO + + m_dev->EndScene(); +} diff --git a/plugins/GSdx/GSRendererOGL.h b/plugins/GSdx/GSRendererOGL.h new file mode 100644 index 0000000000..aa63b99183 --- /dev/null +++ b/plugins/GSdx/GSRendererOGL.h @@ -0,0 +1,40 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSRendererHW.h" +#include "GSVertexHW.h" +#include "GSTextureCacheOGL.h" +// TODO: #include "GSTextureFXOGL.h" + +class GSRendererOGL : public GSRendererHW +{ +protected: + void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + +public: + GSRendererOGL(uint8* base, bool mt, void (*irq)()); + + bool Create(const string& title); + + template void VertexKick(bool skip); +}; \ No newline at end of file diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index d4a2bdd627..abd06df0d3 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -33,17 +33,11 @@ GSSetting GSSettingsDlg::g_renderers[] = {3, "Direct3D10 (Hardware)", NULL}, {4, "Direct3D10 (Software)", NULL}, {5, "Direct3D10 (Null)", NULL}, - {6, "Null (Software)", NULL}, - {7, "Null (Null)", NULL}, -}; - -GSSetting GSSettingsDlg::g_psversion[] = -{ - {D3DPS_VERSION(3, 0), "Pixel Shader 3.0", NULL}, - {D3DPS_VERSION(2, 0), "Pixel Shader 2.0", NULL}, - //{D3DPS_VERSION(1, 4), "Pixel Shader 1.4", NULL}, - //{D3DPS_VERSION(1, 1), "Pixel Shader 1.1", NULL}, - //{D3DPS_VERSION(0, 0), "Fixed Pipeline (bogus)", NULL}, + {6, "OpenGL (Hardware)", NULL}, + {7, "OpenGL (Software)", NULL}, + {8, "OpenGL (Null)", NULL}, + {9, "Null (Software)", NULL}, + {10, "Null (Null)", NULL}, }; GSSetting GSSettingsDlg::g_interlace[] = @@ -73,10 +67,6 @@ void GSSettingsDlg::OnInit() { __super::OnInit(); - D3DCAPS9 caps; - memset(&caps, 0, sizeof(caps)); - caps.PixelShaderVersion = D3DPS_VERSION(0, 0); - m_modes.clear(); { @@ -107,8 +97,6 @@ void GSSettingsDlg::OnInit() ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } - - d3d->GetDeviceCaps(D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, &caps); } } @@ -124,7 +112,6 @@ void GSSettingsDlg::OnInit() } ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_SHADER), g_psversion, countof(g_psversion), theApp.GetConfig("PixelShaderVersion2", D3DPS_VERSION(2, 0)), caps.PixelShaderVersion); ComboBoxInit(GetDlgItem(m_hWnd, IDC_INTERLACE), g_interlace, countof(g_interlace), theApp.GetConfig("Interlace", 0)); ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); @@ -176,11 +163,6 @@ bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) theApp.SetConfig("Renderer", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SHADER), data)) - { - theApp.SetConfig("PixelShaderVersion2", (int)data); - } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_INTERLACE), data)) { theApp.SetConfig("Interlace", (int)data); @@ -215,15 +197,15 @@ void GSSettingsDlg::UpdateControls() { bool dx9 = i >= 0 && i <= 2; bool dx10 = i >= 3 && i <= 5; - bool hw = i == 0 || i == 3; - bool sw = i == 1 || i == 4 || i == 6; + bool ogl = i >= 6 && i <= 8; + bool hw = i == 0 || i == 3 || i == 6; + bool sw = i == 1 || i == 4 || i == 7; bool native = !!IsDlgButtonChecked(m_hWnd, IDC_NATIVERES); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); EnableWindow(GetDlgItem(m_hWnd, IDC_RESOLUTION), dx9); - EnableWindow(GetDlgItem(m_hWnd, IDC_SHADER), dx9); EnableWindow(GetDlgItem(m_hWnd, IDC_RESX), hw && !native); EnableWindow(GetDlgItem(m_hWnd, IDC_RESX_EDIT), hw && !native); EnableWindow(GetDlgItem(m_hWnd, IDC_RESY), hw && !native); diff --git a/plugins/GSdx/GSSettingsDlg.h b/plugins/GSdx/GSSettingsDlg.h index 2e3765a53a..82e947eabc 100644 --- a/plugins/GSdx/GSSettingsDlg.h +++ b/plugins/GSdx/GSSettingsDlg.h @@ -37,8 +37,7 @@ protected: public: GSSettingsDlg(); - static GSSetting g_renderers[]; - static GSSetting g_psversion[]; + static GSSetting g_renderers[]; static GSSetting g_interlace[]; static GSSetting g_aspectratio[]; }; \ No newline at end of file diff --git a/plugins/GSdx/GSTexture.h b/plugins/GSdx/GSTexture.h index a85199a8f9..41fa86f904 100644 --- a/plugins/GSdx/GSTexture.h +++ b/plugins/GSdx/GSTexture.h @@ -28,13 +28,15 @@ class GSTexture public: GSVector2 m_scale; + struct GSMap {uint8* bits; int pitch;}; + + enum {None, RenderTarget, DepthStencil, Texture, Offscreen}; + public: GSTexture() : m_scale(1, 1) {} virtual ~GSTexture() {} - enum {None, RenderTarget, DepthStencil, Texture, Offscreen}; - - virtual operator bool() = 0; + virtual operator bool() {ASSERT(0); return false;} virtual int GetType() const = 0; virtual int GetWidth() const = 0; @@ -42,7 +44,7 @@ public: virtual int GetFormat() const = 0; virtual bool Update(const GSVector4i& r, const void* data, int pitch) = 0; - virtual bool Map(uint8** bits, int& pitch, const GSVector4i* r = NULL) = 0; + virtual bool Map(GSMap& m, const GSVector4i* r = NULL) = 0; virtual void Unmap() = 0; virtual bool Save(const string& fn, bool dds = false) = 0; diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 8b4de5b233..421e650ec6 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -31,11 +31,6 @@ GSTexture10::GSTexture10(ID3D10Texture2D* texture) m_texture->GetDesc(&m_desc); } -GSTexture10::operator bool() -{ - return !!m_texture; -} - int GSTexture10::GetType() const { if(m_desc.BindFlags & D3D10_BIND_RENDER_TARGET) return GSTexture::RenderTarget; @@ -74,7 +69,7 @@ bool GSTexture10::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture10::Map(uint8** bits, int& pitch, const GSVector4i* r) +bool GSTexture10::Map(GSMap& m, const GSVector4i* r) { if(r != NULL) { @@ -89,8 +84,8 @@ bool GSTexture10::Map(uint8** bits, int& pitch, const GSVector4i* r) if(SUCCEEDED(m_texture->Map(0, D3D10_MAP_READ_WRITE, 0, &map))) { - *bits = (uint8*)map.pData; - pitch = (int)map.RowPitch; + m.bits = (uint8*)map.pData; + m.pitch = (int)map.RowPitch; return true; } diff --git a/plugins/GSdx/GSTexture10.h b/plugins/GSdx/GSTexture10.h index 5c8418ae6a..530ec3be3c 100644 --- a/plugins/GSdx/GSTexture10.h +++ b/plugins/GSdx/GSTexture10.h @@ -35,15 +35,13 @@ class GSTexture10 : public GSTexture public: explicit GSTexture10(ID3D10Texture2D* texture); - operator bool(); - int GetType() const; int GetWidth() const; int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch, const GSVector4i* r); + bool Map(GSMap& m, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture7.cpp b/plugins/GSdx/GSTexture7.cpp index 6c3d23b916..44dd81f7e9 100644 --- a/plugins/GSdx/GSTexture7.cpp +++ b/plugins/GSdx/GSTexture7.cpp @@ -45,11 +45,6 @@ GSTexture7::GSTexture7(int type, IDirectDrawSurface7* system, IDirectDrawSurface video->GetSurfaceDesc(&m_desc); } -GSTexture7::operator bool() -{ - return !!m_system; -} - int GSTexture7::GetType() const { return m_type; @@ -123,7 +118,7 @@ bool GSTexture7::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture7::Map(uint8** bits, int& pitch, const GSVector4i* r) +bool GSTexture7::Map(GSMap& m, const GSVector4i* r) { HRESULT hr; @@ -138,8 +133,8 @@ bool GSTexture7::Map(uint8** bits, int& pitch, const GSVector4i* r) if(SUCCEEDED(hr = m_system->Lock(NULL, &desc, DDLOCK_WAIT | DDLOCK_SURFACEMEMORYPTR, NULL))) { - *bits = (uint8*)desc.lpSurface; - pitch = (int)desc.lPitch; + m.bits = (uint8*)desc.lpSurface; + m.pitch = (int)desc.lPitch; return true; } diff --git a/plugins/GSdx/GSTexture7.h b/plugins/GSdx/GSTexture7.h index 8eb64d730c..c69b1a96b1 100644 --- a/plugins/GSdx/GSTexture7.h +++ b/plugins/GSdx/GSTexture7.h @@ -35,15 +35,13 @@ public: GSTexture7(int type, IDirectDrawSurface7* system); GSTexture7(int type, IDirectDrawSurface7* system, IDirectDrawSurface7* video); - operator bool(); - int GetType() const; int GetWidth() const; int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch, const GSVector4i* r); + bool Map(GSMap& m, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTexture9.cpp b/plugins/GSdx/GSTexture9.cpp index 50799a6767..daacf8e5dd 100644 --- a/plugins/GSdx/GSTexture9.cpp +++ b/plugins/GSdx/GSTexture9.cpp @@ -52,11 +52,6 @@ GSTexture9::~GSTexture9() { } -GSTexture9::operator bool() -{ - return !!m_surface; -} - int GSTexture9::GetType() const { if(m_desc.Usage & D3DUSAGE_RENDERTARGET) return GSTexture::RenderTarget; @@ -111,7 +106,7 @@ bool GSTexture9::Update(const GSVector4i& r, const void* data, int pitch) return false; } -bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) +bool GSTexture9::Map(GSMap& m, const GSVector4i* r) { HRESULT hr; @@ -121,8 +116,8 @@ bool GSTexture9::Map(uint8** bits, int& pitch, const GSVector4i* r) if(SUCCEEDED(hr = m_surface->LockRect(&lr, (LPRECT)r, 0))) { - *bits = (uint8*)lr.pBits; - pitch = (int)lr.Pitch; + m.bits = (uint8*)lr.pBits; + m.pitch = (int)lr.Pitch; return true; } diff --git a/plugins/GSdx/GSTexture9.h b/plugins/GSdx/GSTexture9.h index 3f42c46bb1..be8bac79fe 100644 --- a/plugins/GSdx/GSTexture9.h +++ b/plugins/GSdx/GSTexture9.h @@ -35,15 +35,13 @@ public: explicit GSTexture9(IDirect3DTexture9* texture); virtual ~GSTexture9(); - operator bool(); - int GetType() const; int GetWidth() const; int GetHeight() const; int GetFormat() const; bool Update(const GSVector4i& r, const void* data, int pitch); - bool Map(uint8** bits, int& pitch, const GSVector4i* r); + bool Map(GSMap& m, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 07d10931f8..609e9a5667 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -128,7 +128,12 @@ GSTextureCache::GSRenderTarget* GSTextureCache::GetRenderTarget(const GIFRegTEX0 { hh *= 2; } - +/* + if(hh < 512) + { + hh = 512; + } +*/ if(ww > 0 && hh > 0) { rt->m_texture->m_scale.x = (float)w / ww; @@ -647,6 +652,54 @@ bool GSTextureCache::GSRenderTarget::Create(int w, int h) return m_texture != NULL; } +void GSTextureCache::GSRenderTarget::Update() +{ + __super::Update(); + + // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) + + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); + + if(r.rempty()) return; + + int w = r.width(); + int h = r.height(); + + if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) + { + GIFRegTEXA TEXA; + + TEXA.AEM = 1; + TEXA.TA0 = 0; + TEXA.TA1 = 0x80; + + GSTexture::GSMap m; + + if(t->Map(m)) + { + m_renderer->m_mem.ReadTexture(r, m.bits, m.pitch, m_TEX0, TEXA); + + t->Unmap(); + } + else + { + static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); + + int pitch = ((w + 3) & ~3) * 4; + + m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA); + + t->Update(r.rsize(), buff, pitch); + } + + // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); + + m_renderer->m_dev->StretchRect(t, m_texture, GSVector4(r) * GSVector4(m_texture->m_scale).xyxy()); + + m_renderer->m_dev->Recycle(t); + } +} + // GSTextureCache::GSDepthStencil GSTextureCache::GSDepthStencil::GSDepthStencil(GSRenderer* r) @@ -664,6 +717,13 @@ bool GSTextureCache::GSDepthStencil::Create(int w, int h) return m_texture != NULL; } +void GSTextureCache::GSDepthStencil::Update() +{ + __super::Update(); + + // TODO +} + // GSTextureCache::GSCachedTexture GSTextureCache::GSCachedTexture::GSCachedTexture(GSRenderer* r) @@ -701,14 +761,13 @@ void GSTextureCache::GSCachedTexture::Update() m_valid = m_valid.runion(r); - uint8* bits = NULL; - int pitch = 0; - - if(m_texture->Map(&bits, pitch, &r)) + GSTexture::GSMap m; + + if(m_texture->Map(m, &r)) { // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource - m_renderer->m_mem.ReadTextureNP(r, bits, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); + m_renderer->m_mem.ReadTextureNP(r, m.bits, m.pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); m_texture->Unmap(); } @@ -716,7 +775,7 @@ void GSTextureCache::GSCachedTexture::Update() { static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - pitch = ((r.width() + 3) & ~3) * 4; + int pitch = ((r.width() + 3) & ~3) * 4; m_renderer->m_mem.ReadTextureNP(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 34bef246ca..a607a18315 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -52,6 +52,8 @@ public: explicit GSRenderTarget(GSRenderer* r); + void Update(); + virtual bool Create(int w, int h); virtual void Read(const GSVector4i& r) = 0; }; @@ -63,6 +65,8 @@ public: explicit GSDepthStencil(GSRenderer* renderer); + void Update(); + virtual bool Create(int w, int h); }; diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index e3c902d1e1..62c6c57b68 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -31,44 +31,6 @@ GSTextureCache10::GSTextureCache10(GSRenderer* r) // GSRenderTargetHW10 -void GSTextureCache10::GSRenderTargetHW10::Update() -{ - __super::Update(); - - // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - - GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); - - if(r.rempty()) return; - - int w = r.width(); - int h = r.height(); - - if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) - { - static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * 4, 16); - static int pitch = 1024 * 4; - - GIFRegTEXA TEXA; - - TEXA.AEM = 1; - TEXA.TA0 = 0; - TEXA.TA1 = 0x80; - - m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA); - - // s->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); - - t->Update(GSVector4i(0, 0, w, h), buff, pitch); - - GSVector4 dr = GSVector4(r) * GSVector4(m_texture->m_scale).xyxy(); - - m_renderer->m_dev->StretchRect(t, m_texture, dr); - - m_renderer->m_dev->Recycle(t); - } -} - void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) { if(m_TEX0.PSM != PSM_PSMCT32 @@ -98,10 +60,9 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) if(GSTexture* offscreen = m_renderer->m_dev->CopyOffscreen(m_texture, src, w, h, format)) { - uint8* bits; - int pitch; + GSTexture::GSMap m; - if(offscreen->Map(&bits, pitch)) + if(offscreen->Map(m)) { // TODO: block level write @@ -112,40 +73,40 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) if(m_TEX0.PSM == PSM_PSMCT32) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)m.bits)[i]); } } } else if(m_TEX0.PSM == PSM_PSMCT24) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)m.bits)[i]); } } } else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)bits)[i]); + m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)m.bits)[i]); } } } @@ -161,15 +122,6 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) } } -// GSDepthStencilHW10 - -void GSTextureCache10::GSDepthStencilHW10::Update() -{ - __super::Update(); - - // TODO -} - // GSTextureHW10 bool GSTextureCache10::GSCachedTextureHW10::Create() diff --git a/plugins/GSdx/GSTextureCache10.h b/plugins/GSdx/GSTextureCache10.h index c2c221cf32..bc480bea4c 100644 --- a/plugins/GSdx/GSTextureCache10.h +++ b/plugins/GSdx/GSTextureCache10.h @@ -31,7 +31,6 @@ class GSTextureCache10 : public GSTextureCache public: explicit GSRenderTargetHW10(GSRenderer* r) : GSRenderTarget(r) {} - void Update(); void Read(const GSVector4i& r); }; @@ -39,8 +38,6 @@ class GSTextureCache10 : public GSTextureCache { public: explicit GSDepthStencilHW10(GSRenderer* r) : GSDepthStencil(r) {} - - void Update(); }; class GSCachedTextureHW10 : public GSCachedTexture diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index 3c65349fdf..d288c5ebac 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -31,45 +31,6 @@ GSTextureCache9::GSTextureCache9(GSRenderer* r) // GSRenderTarget9 -void GSTextureCache9::GSRenderTarget9::Update() -{ - __super::Update(); - - // FIXME: the union of the rects may also update wrong parts of the render target (but a lot faster :) - - GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); - - if(r.rempty()) return; - - int w = r.width(); - int h = r.height(); - - if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) - { - uint8* bits; - int pitch; - - if(t->Map(&bits, pitch)) - { - GIFRegTEXA TEXA; - - TEXA.AEM = 1; - TEXA.TA0 = 0; - TEXA.TA1 = 0x80; - - m_renderer->m_mem.ReadTexture(r, bits, pitch, m_TEX0, TEXA); - - t->Unmap(); - - // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.Width() * r.Height() * 4); - - m_renderer->m_dev->StretchRect(t, m_texture, GSVector4(r) * GSVector4(m_texture->m_scale).xyxy()); - } - - m_renderer->m_dev->Recycle(t); - } -} - void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) { if(m_TEX0.PSM != PSM_PSMCT32 @@ -97,10 +58,9 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) if(GSTexture* offscreen = m_renderer->m_dev->CopyOffscreen(m_texture, src, w, h)) { - uint8* bits; - int pitch; + GSTexture::GSMap m; - if(offscreen->Map(&bits, pitch)) + if(offscreen->Map(m)) { // TODO: block level write @@ -111,40 +71,40 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) if(m_TEX0.PSM == PSM_PSMCT32) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)bits)[i]); + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)m.bits)[i]); } } } else if(m_TEX0.PSM == PSM_PSMCT24) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)bits)[i]); + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)m.bits)[i]); } } } else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) { - for(int y = r.top; y < r.bottom; y++, bits += pitch) + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) { uint32 addr = pa(0, y, bp, bw); int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; for(int x = r.left, i = 0; x < r.right; x++, i++) { - m_renderer->m_mem.WriteFrame16(addr + offset[x], ((uint32*)bits)[i]); + m_renderer->m_mem.WriteFrame16(addr + offset[x], ((uint32*)m.bits)[i]); } } } @@ -160,15 +120,6 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) } } -// GSDepthStencil9 - -void GSTextureCache9::GSDepthStencil9::Update() -{ - __super::Update(); - - // TODO -} - // GSTexture9 bool GSTextureCache9::GSCachedTexture9::Create() diff --git a/plugins/GSdx/GSTextureCache9.h b/plugins/GSdx/GSTextureCache9.h index 1b794db260..3f3afe0b2b 100644 --- a/plugins/GSdx/GSTextureCache9.h +++ b/plugins/GSdx/GSTextureCache9.h @@ -31,7 +31,6 @@ class GSTextureCache9 : public GSTextureCache public: explicit GSRenderTarget9(GSRenderer* r) : GSRenderTarget(r) {} - void Update(); void Read(const GSVector4i& r); }; @@ -39,8 +38,6 @@ class GSTextureCache9 : public GSTextureCache { public: explicit GSDepthStencil9(GSRenderer* r) : GSDepthStencil(r) {} - - void Update(); }; class GSCachedTexture9 : public GSCachedTexture diff --git a/plugins/GSdx/GSTextureCacheOGL.cpp b/plugins/GSdx/GSTextureCacheOGL.cpp new file mode 100644 index 0000000000..b470935181 --- /dev/null +++ b/plugins/GSdx/GSTextureCacheOGL.cpp @@ -0,0 +1,62 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "StdAfx.h" +#include "GSTextureCacheOGL.h" + +// GSTextureCacheOGL + +GSTextureCacheOGL::GSTextureCacheOGL(GSRenderer* r) + : GSTextureCache(r) +{ +} + +// GSRenderTargetOGL + +void GSTextureCacheOGL::GSRenderTargetOGL::Read(const GSVector4i& r) +{ + // TODO +} + +// GSTextureOGL + +bool GSTextureCacheOGL::GSCachedTextureOGL::Create() +{ + // TODO + + return true; +} + +bool GSTextureCacheOGL::GSCachedTextureOGL::Create(GSRenderTarget* rt) +{ + // TODO + + return true; +} + +bool GSTextureCacheOGL::GSCachedTextureOGL::Create(GSDepthStencil* ds) +{ + m_rendered = true; + + // TODO + + return false; +} diff --git a/plugins/GSdx/GSTextureCacheOGL.h b/plugins/GSdx/GSTextureCacheOGL.h new file mode 100644 index 0000000000..c553559bd9 --- /dev/null +++ b/plugins/GSdx/GSTextureCacheOGL.h @@ -0,0 +1,60 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSTextureCache.h" +#include "GSDeviceOGL.h" + +class GSTextureCacheOGL : public GSTextureCache +{ + class GSRenderTargetOGL : public GSRenderTarget + { + public: + explicit GSRenderTargetOGL(GSRenderer* r) : GSRenderTarget(r) {} + + void Read(const GSVector4i& r); + }; + + class GSDepthStencilOGL : public GSDepthStencil + { + public: + explicit GSDepthStencilOGL(GSRenderer* r) : GSDepthStencil(r) {} + }; + + class GSCachedTextureOGL : public GSCachedTexture + { + public: + explicit GSCachedTextureOGL(GSRenderer* r) : GSCachedTexture(r) {} + + bool Create(); + bool Create(GSRenderTarget* rt); + bool Create(GSDepthStencil* ds); + }; + +protected: + GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetOGL(m_renderer);} + GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilOGL(m_renderer);} + GSCachedTexture* CreateTexture() {return new GSCachedTextureOGL(m_renderer);} + +public: + GSTextureCacheOGL(GSRenderer* r); +}; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index f0ed275778..f59d70246f 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -64,14 +64,13 @@ GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) if(t) { - uint8* bits; - int pitch; - - if(t->Map(&bits, pitch)) + GSTexture::GSMap m; + + if(t->Map(m)) { for(uint32 i = 0; i < size; i++) { - ((float*)bits)[i] = (float)((i & msk) | fix) / size; + ((float*)m.bits)[i] = (float)((i & msk) | fix) / size; } t->Unmap(); diff --git a/plugins/GSdx/GSTextureNull.h b/plugins/GSdx/GSTextureNull.h index ec3ada8357..68cf9b2843 100644 --- a/plugins/GSdx/GSTextureNull.h +++ b/plugins/GSdx/GSTextureNull.h @@ -31,14 +31,12 @@ public: GSTextureNull(); GSTextureNull(int type, int w, int h, int format); - operator bool() {return m_desc.type != 0;} - int GetType() const {return m_desc.type;} int GetWidth() const {return m_desc.w;} int GetHeight() const {return m_desc.h;} int GetFormat() const {return m_desc.format;} bool Update(const GSVector4i& r, const void* data, int pitch) {return true;} - bool Map(uint8** bits, int& pitch, const GSVector4i* r) {return false;} + bool Map(GSMap& m, const GSVector4i* r) {return false;} void Unmap() {} bool Save(const string& fn, bool dds = false) {return false;} }; diff --git a/plugins/GSdx/GSTextureOGL.cpp b/plugins/GSdx/GSTextureOGL.cpp new file mode 100644 index 0000000000..35749fb43d --- /dev/null +++ b/plugins/GSdx/GSTextureOGL.cpp @@ -0,0 +1,84 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSTextureOGL.h" + +GSTextureOGL::GSTextureOGL(GLuint texture) + : m_texture(texture) + , m_type(None) + , m_width(0) + , m_height(0) + , m_format(0) +{ + // TODO: m_type, m_format, fb/ds? + + glBindTexture(GL_TEXTURE_2D, texture); + + glGetTexLevelParameteriv(GL_TEXTURE_2D, 0, GL_TEXTURE_WIDTH, &m_width); + glGetTexLevelParameteriv(GL_TEXTURE_2D, 0, GL_TEXTURE_HEIGHT, &m_height); +} + +int GSTextureOGL::GetType() const +{ + return m_type; +} + +int GSTextureOGL::GetWidth() const +{ + return m_width; +} + +int GSTextureOGL::GetHeight() const +{ + return m_height; +} + +int GSTextureOGL::GetFormat() const +{ + return m_format; +} + +bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch) +{ + // TODO + + return false; +} + +bool GSTextureOGL::Map(GSMap& m, const GSVector4i* r) +{ + // TODO + + return false; +} + +void GSTextureOGL::Unmap() +{ + // TODO +} + +bool GSTextureOGL::Save(const string& fn, bool dds) +{ + // TODO + + return false; +} diff --git a/plugins/GSdx/GSTextureOGL.h b/plugins/GSdx/GSTextureOGL.h new file mode 100644 index 0000000000..402aef62c6 --- /dev/null +++ b/plugins/GSdx/GSTextureOGL.h @@ -0,0 +1,47 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSTexture.h" + +class GSTextureOGL : public GSTexture +{ + GLuint m_texture; + + int m_type; + int m_width; + int m_height; + int m_format; + +public: + explicit GSTextureOGL(GLuint texture); + + int GetType() const; + int GetWidth() const; + int GetHeight() const; + int GetFormat() const; + + bool Update(const GSVector4i& r, const void* data, int pitch); + bool Map(GSMap& m, const GSVector4i* r); + void Unmap(); + bool Save(const string& fn, bool dds = false); +}; diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index f43c718692..2e49cd2d20 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -80,4 +80,6 @@ __declspec(align(16)) union GSVertexHW10 float GetQ() {return q;} }; +typedef GSVertexHW9 GSVertexOGL; // TODO + #pragma pack(pop) diff --git a/plugins/GSdx/GSVertexSW.h b/plugins/GSdx/GSVertexSW.h index 8766fc9249..3f51aa756a 100644 --- a/plugins/GSdx/GSVertexSW.h +++ b/plugins/GSdx/GSVertexSW.h @@ -255,7 +255,7 @@ public: key |= 1 << 4; } - m_map.Lookup(key)(v, count, m_min, m_max); + m_map[key](v, count, m_min, m_max); m_eq.value = (m_min.p == m_max.p).mask() | ((m_min.t == m_max.t).mask() << 4) | ((m_min.c == m_max.c).mask() << 8); diff --git a/plugins/GSdx/GSdx.rc b/plugins/GSdx/GSdx.rc index 0850a8481b..7b3a3e42fb 100644 --- a/plugins/GSdx/GSdx.rc +++ b/plugins/GSdx/GSdx.rc @@ -82,7 +82,7 @@ IDB_LOGO10 BITMAP "res\\logo10.bmp" // Dialog // -IDD_CONFIG DIALOGEX 0, 0, 189, 248 +IDD_CONFIG DIALOGEX 0, 0, 189, 233 STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_CAPTION | WS_SYSMENU CAPTION "Settings..." FONT 8, "MS Shell Dlg", 400, 0, 0x1 @@ -93,30 +93,27 @@ BEGIN COMBOBOX IDC_RESOLUTION,71,57,111,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Renderer:",IDC_STATIC,7,74,34,8 COMBOBOX IDC_RENDERER,71,72,111,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Shader:",IDC_STATIC,7,89,26,8 - COMBOBOX IDC_SHADER,71,87,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Interlacing (F5):",IDC_STATIC,7,105,53,8 - COMBOBOX IDC_INTERLACE,71,102,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Aspect Ratio (F6):",IDC_STATIC,7,120,60,8 - COMBOBOX IDC_ASPECTRATIO,71,117,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "D3D internal res:",IDC_STATIC,7,135,55,8 - EDITTEXT IDC_RESX_EDIT,71,132,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_RESX,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,135,11,14 - EDITTEXT IDC_RESY_EDIT,109,132,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_RESY,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,133,135,11,14 - CONTROL "Native",IDC_NATIVERES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,149,134,33,10 - LTEXT "SW rend. threads:",IDC_STATIC,7,149,60,8 - EDITTEXT IDC_SWTHREADS_EDIT,71,147,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,150,11,14 - CONTROL "Texture filtering",IDC_FILTER,"Button",BS_AUTO3STATE | WS_TABSTOP,7,167,67,10 - CONTROL "Logarithmic Z",IDC_LOGZ,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,168,58,10 - CONTROL "Wait vsync",IDC_VSYNC,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,180,51,10 - CONTROL "Alpha correction (FBA)",IDC_FBA,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,181,102,10 - CONTROL "Edge anti-aliasing (AA1, sw-mode only)",IDC_AA1, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,194,141,10 - CONTROL "Enable output merger blur effect",IDC_BLUR,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,208,121,10 - DEFPUSHBUTTON "OK",IDOK,43,227,50,14 - PUSHBUTTON "Cancel",IDCANCEL,96,227,50,14 + LTEXT "Interlacing (F5):",IDC_STATIC,7,90,53,8 + COMBOBOX IDC_INTERLACE,71,87,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "Aspect Ratio (F6):",IDC_STATIC,7,105,60,8 + COMBOBOX IDC_ASPECTRATIO,71,102,111,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "D3D internal res:",IDC_STATIC,7,120,55,8 + EDITTEXT IDC_RESX_EDIT,71,117,35,13,ES_AUTOHSCROLL | ES_NUMBER + CONTROL "",IDC_RESX,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,120,11,14 + EDITTEXT IDC_RESY_EDIT,109,117,35,13,ES_AUTOHSCROLL | ES_NUMBER + CONTROL "",IDC_RESY,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,133,120,11,14 + CONTROL "Native",IDC_NATIVERES,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,149,119,33,10 + LTEXT "SW rend. threads:",IDC_STATIC,7,134,60,8 + EDITTEXT IDC_SWTHREADS_EDIT,71,132,35,13,ES_AUTOHSCROLL | ES_NUMBER + CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,135,11,14 + CONTROL "Texture filtering",IDC_FILTER,"Button",BS_AUTO3STATE | WS_TABSTOP,7,152,67,10 + CONTROL "Logarithmic Z",IDC_LOGZ,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,153,58,10 + CONTROL "Wait vsync",IDC_VSYNC,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,165,51,10 + CONTROL "Alpha correction (FBA)",IDC_FBA,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,80,166,102,10 + CONTROL "Edge anti-aliasing (AA1, sw-mode only)",IDC_AA1,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,179,141,10 + CONTROL "Enable output merger blur effect",IDC_BLUR,"Button",BS_AUTOCHECKBOX | WS_TABSTOP,7,193,121,10 + DEFPUSHBUTTON "OK",IDOK,43,212,50,14 + PUSHBUTTON "Cancel",IDCANCEL,96,212,50,14 END IDD_CAPTURE DIALOGEX 0, 0, 279, 71 @@ -135,7 +132,7 @@ BEGIN DEFPUSHBUTTON "OK",IDOK,221,47,50,14 END -IDD_GPUCONFIG DIALOGEX 0, 0, 189, 235 +IDD_GPUCONFIG DIALOGEX 0, 0, 189, 199 STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_CAPTION | WS_SYSMENU CAPTION "Settings..." FONT 8, "MS Shell Dlg", 400, 0, 0x1 @@ -145,22 +142,20 @@ BEGIN COMBOBOX IDC_RESOLUTION,78,57,104,125,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP LTEXT "Renderer:",IDC_STATIC,7,74,34,8 COMBOBOX IDC_RENDERER,78,72,104,118,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Shader:",IDC_STATIC,7,89,26,8 - COMBOBOX IDC_SHADER,78,87,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Texture Filter (Del):",IDC_STATIC,7,105,64,8 - COMBOBOX IDC_FILTER,78,102,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Dithering (End):",IDC_STATIC,7,120,52,8 - COMBOBOX IDC_DITHERING,78,117,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Aspect Ratio (PgDn):",IDC_STATIC,7,135,68,8 - COMBOBOX IDC_ASPECTRATIO,78,132,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP - LTEXT "Rendering Threads:",IDC_STATIC,7,165,64,8 - EDITTEXT IDC_SWTHREADS_EDIT,78,163,35,13,ES_AUTOHSCROLL | ES_NUMBER - CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,169,11,14 - DEFPUSHBUTTON "OK",IDOK,43,214,50,14 - PUSHBUTTON "Cancel",IDCANCEL,96,214,50,14 - CONTROL 2022,IDC_LOGO10,"Static",SS_BITMAP,7,7,175,44 - LTEXT "Internal Resolution:",IDC_STATIC,7,150,64,8 - COMBOBOX IDC_SCALE,78,147,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "Texture Filter (Del):",IDC_STATIC,7,90,64,8 + COMBOBOX IDC_FILTER,78,87,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "Dithering (End):",IDC_STATIC,7,105,52,8 + COMBOBOX IDC_DITHERING,78,102,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "Aspect Ratio (PgDn):",IDC_STATIC,7,120,68,8 + COMBOBOX IDC_ASPECTRATIO,78,117,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP + LTEXT "Rendering Threads:",IDC_STATIC,7,150,64,8 + EDITTEXT IDC_SWTHREADS_EDIT,78,148,35,13,ES_AUTOHSCROLL | ES_NUMBER + CONTROL "",IDC_SWTHREADS,"msctls_updown32",UDS_SETBUDDYINT | UDS_ALIGNRIGHT | UDS_AUTOBUDDY | UDS_ARROWKEYS | UDS_NOTHOUSANDS,99,154,11,14 + DEFPUSHBUTTON "OK",IDOK,43,178,50,14 + PUSHBUTTON "Cancel",IDCANCEL,96,178,50,14 + CONTROL 2022,IDC_LOGO10,"Static",SS_BITMAP,7,7,173,42 + LTEXT "Internal Resolution:",IDC_STATIC,7,135,64,8 + COMBOBOX IDC_SCALE,78,132,104,98,CBS_DROPDOWNLIST | WS_VSCROLL | WS_TABSTOP END @@ -180,7 +175,7 @@ BEGIN VERTGUIDE, 80 VERTGUIDE, 182 TOPMARGIN, 7 - BOTTOMMARGIN, 241 + BOTTOMMARGIN, 226 END IDD_CAPTURE, DIALOG @@ -198,7 +193,7 @@ BEGIN VERTGUIDE, 78 VERTGUIDE, 182 TOPMARGIN, 7 - BOTTOMMARGIN, 228 + BOTTOMMARGIN, 192 END END #endif // APSTUDIO_INVOKED diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index ba86150a75..0a4cc487db 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1155,6 +1155,10 @@ RelativePath=".\GSDeviceNull.cpp" > + + @@ -1331,6 +1335,10 @@ RelativePath=".\GSRendererNull.cpp" > + + @@ -1415,6 +1423,10 @@ RelativePath=".\GSTextureCache9.cpp" > + + @@ -1447,6 +1459,10 @@ RelativePath=".\GSTextureNull.cpp" > + + @@ -1717,6 +1733,10 @@ RelativePath=".\GSDeviceNull.h" > + + @@ -1785,6 +1805,10 @@ RelativePath=".\GSRendererNull.h" > + + @@ -1841,6 +1865,10 @@ RelativePath=".\GSTextureCache9.h" > + + @@ -1857,6 +1885,10 @@ RelativePath=".\GSTextureNull.h" > + + diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 366b04188d..97f1893032 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -209,6 +209,8 @@ float4 Extract16(uint i) int4 wrapuv(int4 uv) { + // TODO: bitwise ops don't work with render target textures that were not rendered at the native resolution. + if(WMS == WMT) { switch(WMS) diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index e50189b796..eef0a46699 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -99,6 +99,8 @@ typedef signed long long int64; #define _M_AMD64 #endif +#ifdef _WINDOWS + // directx #include @@ -109,6 +111,14 @@ typedef signed long long int64; #define D3DCOLORWRITEENABLE_RGBA (D3DCOLORWRITEENABLE_RED | D3DCOLORWRITEENABLE_GREEN | D3DCOLORWRITEENABLE_BLUE | D3DCOLORWRITEENABLE_ALPHA) +#endif + +// opengl + +#include +#include +#include + // sse #if !defined(_M_SSE) && (defined(_M_AMD64) || defined(_M_IX86_FP) && _M_IX86_FP >= 2) diff --git a/plugins/GSdx/vsprops/common.vsprops b/plugins/GSdx/vsprops/common.vsprops index 29fe0375bb..e4990cacd0 100644 --- a/plugins/GSdx/vsprops/common.vsprops +++ b/plugins/GSdx/vsprops/common.vsprops @@ -18,7 +18,7 @@ /> Date: Fri, 12 Jun 2009 19:39:54 +0000 Subject: [PATCH 217/296] GSdx: added ogl/cg to the list of delay loaded dlls git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1361 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/vsprops/common.vsprops | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/GSdx/vsprops/common.vsprops b/plugins/GSdx/vsprops/common.vsprops index e4990cacd0..1d0bb33cad 100644 --- a/plugins/GSdx/vsprops/common.vsprops +++ b/plugins/GSdx/vsprops/common.vsprops @@ -20,7 +20,7 @@ Name="VCLinkerTool" AdditionalDependencies="JITProfiling.lib d3d10.lib d3dx10.lib d3d9.lib d3dx9.lib ddraw.lib dxguid.lib winmm.lib strmiids.lib xinput.lib cg.lib cgGL.lib glut32.lib" AdditionalLibraryDirectories="./vtune" - DelayLoadDLLs="d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll" + DelayLoadDLLs="d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll;cg.dll;cgGL.dll;glut32.dll" GenerateDebugInformation="true" SubSystem="2" RandomizedBaseAddress="1" From 690ef417b1b4db92327ef413997a0621a5cf112d Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 13 Jun 2009 19:41:34 +0000 Subject: [PATCH 218/296] GSdx: replaced switch with ifs in the shaders, kingdom hearts should be fixed git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1362 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GS.cpp | 2 + plugins/GSdx/GSDevice10.cpp | 2 +- plugins/GSdx/GSDevice10.h | 1 - plugins/GSdx/GSDevice7.cpp | 2 +- plugins/GSdx/GSDeviceNull.cpp | 2 +- plugins/GSdx/GSDeviceOGL.cpp | 171 ++++++++++++++++++++++++++++---- plugins/GSdx/GSDeviceOGL.h | 18 ++++ plugins/GSdx/GSRendererHW10.cpp | 34 +++---- plugins/GSdx/GSRendererHW9.cpp | 12 +-- plugins/GSdx/GSSettingsDlg.cpp | 2 + plugins/GSdx/GSTextureFX10.h | 14 ++- plugins/GSdx/GSTextureFX9.h | 6 +- plugins/GSdx/GSTextureOGL.cpp | 32 ++++-- plugins/GSdx/GSTextureOGL.h | 3 +- plugins/GSdx/res/tfx10.fx | 127 ++++++++++++++---------- plugins/GSdx/res/tfx9.fx | 79 +++++++++------ plugins/GSdx/stdafx.h | 5 + 17 files changed, 362 insertions(+), 150 deletions(-) diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 972fe0c5a0..cecc1019e7 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -126,9 +126,11 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; + #if 0 case 6: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; + #endif case 9: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; case 10: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; } diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 2a59eaa644..084b8d191a 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -221,7 +221,7 @@ bool GSDevice10::Create(GSWnd* wnd, bool vsync) // - Reset(1, 1, true); + Reset(1, 1, false); // diff --git a/plugins/GSdx/GSDevice10.h b/plugins/GSdx/GSDevice10.h index d12bdd9bad..5739532105 100644 --- a/plugins/GSdx/GSDevice10.h +++ b/plugins/GSdx/GSDevice10.h @@ -26,7 +26,6 @@ class GSDevice10 : public GSDevice { -private: ID3D10Buffer* m_vb; size_t m_vb_stride; ID3D10InputLayout* m_layout; diff --git a/plugins/GSdx/GSDevice7.cpp b/plugins/GSdx/GSDevice7.cpp index 6995fedd5a..2481fe5c0f 100644 --- a/plugins/GSdx/GSDevice7.cpp +++ b/plugins/GSdx/GSDevice7.cpp @@ -75,7 +75,7 @@ bool GSDevice7::Create(GSWnd* wnd, bool vsync) return false; } - Reset(1, 1, true); + Reset(1, 1, false); return true; } diff --git a/plugins/GSdx/GSDeviceNull.cpp b/plugins/GSdx/GSDeviceNull.cpp index 75112a544f..a2bca4dd4f 100644 --- a/plugins/GSdx/GSDeviceNull.cpp +++ b/plugins/GSdx/GSDeviceNull.cpp @@ -29,7 +29,7 @@ bool GSDeviceNull::Create(GSWnd* wnd, bool vsync) return false; } - Reset(1, 1, true); + Reset(1, 1, false); return true; } diff --git a/plugins/GSdx/GSDeviceOGL.cpp b/plugins/GSdx/GSDeviceOGL.cpp index 55ffa50648..c503b7c084 100644 --- a/plugins/GSdx/GSDeviceOGL.cpp +++ b/plugins/GSdx/GSDeviceOGL.cpp @@ -27,11 +27,21 @@ GSDeviceOGL::GSDeviceOGL() : m_hDC(NULL) , m_hGLRC(NULL) + , m_vbo(0) + , m_fbo(0) + , m_topology(-1) { + m_vertices.stride = 0; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = 0; } GSDeviceOGL::~GSDeviceOGL() { + if(m_vbo) glDeleteBuffers(1, &m_vbo); + if(m_fbo) glDeleteFramebuffers(1, &m_fbo); + #ifdef _WINDOWS if(m_hGLRC) {wglMakeCurrent(NULL, NULL); wglDeleteContext(m_hGLRC);} @@ -75,24 +85,40 @@ bool GSDeviceOGL::Create(GSWnd* wnd, bool vsync) if(!m_hDC) return false; if(!SetPixelFormat(m_hDC, ChoosePixelFormat(m_hDC, &pfd), &pfd)) + { return false; + } m_hGLRC = wglCreateContext(m_hDC); if(!m_hGLRC) return false; if(!wglMakeCurrent(m_hDC, m_hGLRC)) + { return false; + } + + if(WGLEW_EXT_swap_control) + { + wglSwapIntervalEXT(vsync ? 1 : 0); + } #endif + if(glewInit() != GLEW_OK) + { + return false; + } + // const char* exts = (const char*)glGetString(GL_EXTENSIONS); - // TODO + glGenBuffers(1, &m_vbo); + glBindBuffer(GL_ARRAY_BUFFER, m_vbo); + // TODO: setup layout? - Reset(1, 1, true); + GSVector4i r = wnd->GetClientRect(); - // + Reset(r.width(), r.height(), false); return true; } @@ -102,12 +128,12 @@ bool GSDeviceOGL::Reset(int w, int h, bool fs) if(!__super::Reset(w, h, fs)) return false; - m_backbuffer = new GSTextureOGL(0); // ??? + m_backbuffer = new GSTextureOGL(0, GSTexture::RenderTarget, w, h); glCullFace(GL_FRONT_AND_BACK); glDisable(GL_LIGHTING); glDisable(GL_ALPHA_TEST); - glDisable(GL_SCISSOR_TEST); + glEnable(GL_SCISSOR_TEST); glMatrixMode(GL_PROJECTION); glLoadIdentity(); @@ -116,76 +142,134 @@ bool GSDeviceOGL::Reset(int w, int h, bool fs) glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); - // glViewport(0, 0, ?, ?); + // glViewport(0, 0, w, h); + + if(m_fbo) {glDeleteFramebuffersEXT(1, &m_fbo); m_fbo = 0;} + + glGenFramebuffers(1, &m_fbo); + + if(m_fbo == 0) return false; + + glBindFramebuffer(GL_FRAMEBUFFER, m_fbo); return true; } void GSDeviceOGL::Flip() { - // TODO + #ifdef _WINDOWS + + SwapBuffers(m_hDC); + + #endif } void GSDeviceOGL::BeginScene() { - // TODO } void GSDeviceOGL::DrawPrimitive() { - // TODO + glDrawArrays(m_topology, m_vertices.count, m_vertices.start); } void GSDeviceOGL::EndScene() { - // TODO + m_vertices.start += m_vertices.count; + m_vertices.count = 0; } void GSDeviceOGL::ClearRenderTarget(GSTexture* t, const GSVector4& c) { - // TODO + // TODO: disable scissor, color mask + + glClearColor(c.r, c.g, c.b, c.a); + glClear(GL_COLOR_BUFFER_BIT); } void GSDeviceOGL::ClearRenderTarget(GSTexture* t, uint32 c) { - // TODO + ClearRenderTarget(t, GSVector4(c) * (1.0f / 255)); } void GSDeviceOGL::ClearDepth(GSTexture* t, float c) { - // TODO + // TODO: disable scissor, depth mask + + glClearDepth(c); + glClear(GL_DEPTH_BUFFER_BIT); } void GSDeviceOGL::ClearStencil(GSTexture* t, uint8 c) { - // TODO + // TODO: disable scissor, depth (?) mask + + glClearStencil((GLint)c); + glClear(GL_STENCIL_BUFFER_BIT); } GSTexture* GSDeviceOGL::Create(int type, int w, int h, int format) { - // TODO + GLuint texture = 0; - return NULL; + switch(type) + { + case GSTexture::RenderTarget: + glGenTextures(1, &texture); + glBindTexture(GL_TEXTURE_2D, texture); + glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); + break; + case GSTexture::DepthStencil: + glGenRenderbuffersEXT(1, &texture); + glBindRenderbufferEXT(GL_RENDERBUFFER_EXT, texture); + glRenderbufferStorageEXT(GL_RENDERBUFFER_EXT, GL_DEPTH32F_STENCIL8, w, h); + // TODO: depth textures? + break; + case GSTexture::Texture: + glGenTextures(1, &texture); + glBindTexture(GL_TEXTURE_2D, texture); + glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); + break; + case GSTexture::Offscreen: + // TODO: ??? + break; + } + + if(!texture) return false; + + GSTextureOGL* t = new GSTextureOGL(texture, type, w, h, format); + + switch(type) + { + case GSTexture::RenderTarget: + ClearRenderTarget(t, 0); + break; + case GSTexture::DepthStencil: + ClearDepth(t, 0); + break; + } + + return t; } GSTexture* GSDeviceOGL::CreateRenderTarget(int w, int h, int format) { - return __super::CreateRenderTarget(w, h, format ? format : 0); // TODO + return __super::CreateRenderTarget(w, h, format ? format : GL_RGBA8); } GSTexture* GSDeviceOGL::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(w, h, format ? format : 0); // TODO + return __super::CreateDepthStencil(w, h, format ? format : GL_DEPTH32F_STENCIL8); // TODO: GL_DEPTH24_STENCIL8_EXT, GL_DEPTH24_STENCIL8 } GSTexture* GSDeviceOGL::CreateTexture(int w, int h, int format) { - return __super::CreateTexture(w, h, format ? format : 0); // TODO + return __super::CreateTexture(w, h, format ? format : GL_RGBA8); } GSTexture* GSDeviceOGL::CreateOffscreen(int w, int h, int format) { - return __super::CreateOffscreen(w, h, format ? format : 0); // TODO + return __super::CreateOffscreen(w, h, format ? format : GL_RGBA8); } GSTexture* GSDeviceOGL::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format) @@ -209,3 +293,50 @@ void GSDeviceOGL::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool lin { // TODO } + +void GSDeviceOGL::IASetVertexBuffer(const void* vertices, size_t stride, size_t count) +{ + ASSERT(m_vertices.count == 0); + + glBindBuffer(GL_ARRAY_BUFFER, m_vbo); + + if(count * stride > m_vertices.limit * m_vertices.stride) + { + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = 0; + } + + if(m_vertices.limit == 0) + { + m_vertices.limit = max(count * 3 / 2, 10000); + + glBufferData(GL_ARRAY_BUFFER, m_vertices.limit * stride, NULL, GL_DYNAMIC_DRAW); // GL_STREAM_DRAW? + } + + if(m_vertices.start + count > m_vertices.limit || stride != m_vertices.stride) + { + m_vertices.start = 0; // TODO: how to discard and not overwrite previous data? + } + + if(GLvoid* v = glMapBuffer(GL_ARRAY_BUFFER, GL_WRITE_ONLY)) + { + GSVector4i::storent((uint8*)v + m_vertices.start * stride, vertices, count * stride); + + glUnmapBuffer(GL_ARRAY_BUFFER); + } + + m_vertices.count = count; + m_vertices.stride = stride; +} + +void GSDeviceOGL::IASetInputLayout() +{ + // TODO +} + +void GSDeviceOGL::IASetPrimitiveTopology(int topology) +{ + m_topology = topology; +} + diff --git a/plugins/GSdx/GSDeviceOGL.h b/plugins/GSdx/GSDeviceOGL.h index 0b7ab393f4..32c510b2d1 100644 --- a/plugins/GSdx/GSDeviceOGL.h +++ b/plugins/GSdx/GSDeviceOGL.h @@ -33,9 +33,23 @@ class GSDeviceOGL : public GSDevice #endif + GLuint m_vbo; + GLuint m_fbo; + + struct + { + size_t stride, start, count, limit; + } m_vertices; + + int m_topology; + + // + static void OnStaticCgError(CGcontext ctx, CGerror err, void* p) {((GSDeviceOGL*)p)->OnCgError(ctx, err);} void OnCgError(CGcontext ctx, CGerror err); + // + GSTexture* Create(int type, int w, int h, int format); void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); @@ -67,4 +81,8 @@ public: GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); + + void IASetVertexBuffer(const void* vertices, size_t stride, size_t count); + void IASetInputLayout(); // TODO + void IASetPrimitiveTopology(int topology); }; diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 09bec8dbe1..c8d4dfaa67 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -336,16 +336,15 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache GSTextureFX10::PSConstantBuffer ps_cb; - ps_cb.FogColorAREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - ps_cb.TA = GSVector4((int)env.TEXA.TA0, (int)env.TEXA.TA1) / 255; + ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; if(context->TEST.ATST == 2 || context->TEST.ATST == 5) { - ps_cb.FogColorAREF.a -= 0.9f / 255; + ps_cb.FogColor_AREF.a -= 0.9f / 255; } else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) { - ps_cb.FogColorAREF.a += 0.9f / 255; + ps_cb.FogColor_AREF.a += 0.9f / 255; } if(tex) @@ -355,28 +354,27 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache int w = tex->m_texture->GetWidth(); int h = tex->m_texture->GetHeight(); + ps_cb.WH_TA = GSVector4((int)(1 << context->TEX0.TW), (int)(1 << context->TEX0.TH), env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); + ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); + switch(context->CLAMP.WMS) { case 0: - ps_cb.MinMax.x = w - 1; - ps_cb.MinMax.z = 0; ps_ssel.tau = 1; break; case 1: - ps_cb.MinMax.x = 0; - ps_cb.MinMax.z = w - 1; ps_ssel.tau = 0; break; case 2: - ps_cb.MinMax.x = (int)context->CLAMP.MINU * w / (1 << context->TEX0.TW); - ps_cb.MinMax.z = (int)context->CLAMP.MAXU * w / (1 << context->TEX0.TW); + ps_cb.MinMax.x = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); + ps_cb.MinMax.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); ps_ssel.tau = 0; break; case 3: - ps_cb.MinMax.x = context->CLAMP.MINU; - ps_cb.MinMax.z = context->CLAMP.MAXU; + ps_cb.MskFix.x = context->CLAMP.MINU; + ps_cb.MskFix.z = context->CLAMP.MAXU; ps_ssel.tau = 1; break; default: @@ -386,25 +384,21 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache switch(context->CLAMP.WMT) { case 0: - ps_cb.MinMax.y = h - 1; - ps_cb.MinMax.w = 0; ps_ssel.tav = 1; break; case 1: - ps_cb.MinMax.y = 0; - ps_cb.MinMax.w = h - 1; ps_ssel.tav = 0; break; case 2: - ps_cb.MinMax.y = (int)context->CLAMP.MINV * h / (1 << context->TEX0.TH); - ps_cb.MinMax.w = (int)context->CLAMP.MAXV * h / (1 << context->TEX0.TH); + ps_cb.MinMax.y = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); + ps_cb.MinMax.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); ps_ssel.tav = 0; break; case 3: - ps_cb.MinMax.y = context->CLAMP.MINV; - ps_cb.MinMax.w = context->CLAMP.MAXV; + ps_cb.MskFix.y = context->CLAMP.MINV; + ps_cb.MskFix.w = context->CLAMP.MAXV; ps_ssel.tav = 1; break; default: diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index f88143c334..5a37df2530 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -308,17 +308,15 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: GSTextureFX9::PSConstantBuffer ps_cb; - ps_cb.FogColorAREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - ps_cb.TA0 = (float)(int)env.TEXA.TA0 / 255; - ps_cb.TA1 = (float)(int)env.TEXA.TA1 / 255; + ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; if(context->TEST.ATST == 2 || context->TEST.ATST == 5) { - ps_cb.FogColorAREF.a -= 0.9f / 255; + ps_cb.FogColor_AREF.a -= 0.9f / 255; } else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) { - ps_cb.FogColorAREF.a += 0.9f / 255; + ps_cb.FogColor_AREF.a += 0.9f / 255; } if(tex) @@ -328,8 +326,8 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: int w = tex->m_texture->GetWidth(); int h = tex->m_texture->GetHeight(); - ps_cb.WH = GSVector2(w, h); - ps_cb.HalfTexel = GSVector4(-0.5f / w, -0.5f / h, +0.5f / w, +0.5f / h); + ps_cb.WH_TA = GSVector4(w, h, env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); + ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); switch(context->CLAMP.WMS) { diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index abd06df0d3..3c84591fc9 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -33,9 +33,11 @@ GSSetting GSSettingsDlg::g_renderers[] = {3, "Direct3D10 (Hardware)", NULL}, {4, "Direct3D10 (Software)", NULL}, {5, "Direct3D10 (Null)", NULL}, + #if 0 {6, "OpenGL (Hardware)", NULL}, {7, "OpenGL (Software)", NULL}, {8, "OpenGL (Null)", NULL}, + #endif {9, "Null (Software)", NULL}, {10, "Null (Null)", NULL}, }; diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 3775879d5f..f50617052b 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -77,10 +77,12 @@ public: __declspec(align(16)) struct PSConstantBuffer { - GSVector4 FogColorAREF; - GSVector4 TA; - GSVector4i MinMax; + GSVector4 FogColor_AREF; + GSVector4 HalfTexel; + GSVector4 WH_TA; + GSVector4 MinMax; GSVector4 MinMaxF; + GSVector4i MskFix; struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} @@ -93,13 +95,17 @@ public: GSVector4i b1 = b[1]; GSVector4i b2 = b[2]; GSVector4i b3 = b[3]; + GSVector4i b4 = b[4]; + GSVector4i b5 = b[5]; - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3)).alltrue()) + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) { a[0] = b0; a[1] = b1; a[2] = b2; a[3] = b3; + a[4] = b4; + a[5] = b5; return true; } diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index c657a2b06d..b47c227a7b 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -53,11 +53,9 @@ public: struct PSConstantBuffer { - GSVector4 FogColorAREF; + GSVector4 FogColor_AREF; GSVector4 HalfTexel; - GSVector2 WH; - float TA0; - float TA1; + GSVector4 WH_TA; GSVector4 MinMax; GSVector4 MinMaxF; GSVector4i MskFix; diff --git a/plugins/GSdx/GSTextureOGL.cpp b/plugins/GSdx/GSTextureOGL.cpp index 35749fb43d..3e8f1f97eb 100644 --- a/plugins/GSdx/GSTextureOGL.cpp +++ b/plugins/GSdx/GSTextureOGL.cpp @@ -22,19 +22,29 @@ #include "stdafx.h" #include "GSTextureOGL.h" -GSTextureOGL::GSTextureOGL(GLuint texture) +GSTextureOGL::GSTextureOGL(GLuint texture, int type, int width, int height, int format) : m_texture(texture) - , m_type(None) - , m_width(0) - , m_height(0) - , m_format(0) + , m_type(type) + , m_width(width) + , m_height(height) + , m_format(format) { - // TODO: m_type, m_format, fb/ds? +} - glBindTexture(GL_TEXTURE_2D, texture); - - glGetTexLevelParameteriv(GL_TEXTURE_2D, 0, GL_TEXTURE_WIDTH, &m_width); - glGetTexLevelParameteriv(GL_TEXTURE_2D, 0, GL_TEXTURE_HEIGHT, &m_height); +GSTextureOGL::~GSTextureOGL() +{ + if(m_texture) + { + switch(m_type) + { + case DepthStencil: + glDeleteRenderbuffersEXT(1, &m_texture); + break; + default: + glDeleteTextures(1, &m_texture); + break; + } + } } int GSTextureOGL::GetType() const @@ -59,7 +69,7 @@ int GSTextureOGL::GetFormat() const bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch) { - // TODO + // TODO: glTexSubImage2D looks like UpdateSubresource but does not take a pitch return false; } diff --git a/plugins/GSdx/GSTextureOGL.h b/plugins/GSdx/GSTextureOGL.h index 402aef62c6..5e6e2a7926 100644 --- a/plugins/GSdx/GSTextureOGL.h +++ b/plugins/GSdx/GSTextureOGL.h @@ -33,7 +33,8 @@ class GSTextureOGL : public GSTexture int m_format; public: - explicit GSTextureOGL(GLuint texture); + GSTextureOGL(GLuint texture, int type, int width, int height, int format = 0); + virtual ~GSTextureOGL(); int GetType() const; int GetWidth() const; diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 97f1893032..a2b067a93b 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -155,9 +155,12 @@ cbuffer cb1 { float3 FogColor; float AREF; - float4 TA; - uint4 MinMax; + float4 HalfTexel; + float2 WH; + float2 TA; + float4 MinMax; float4 MinMaxF; + uint4 MskFix; }; struct PS_INPUT @@ -175,15 +178,15 @@ struct PS_OUTPUT #ifndef FST #define FST 0 -#define WMS 1 -#define WMT 1 +#define WMS 3 +#define WMT 3 #define BPP 0 #define AEM 0 #define TFX 0 #define TCC 1 -#define ATE 1 +#define ATE 0 #define ATST 4 -#define FOG 1 +#define FOG 0 #define CLR1 0 #define FBA 0 #define AOUT 0 @@ -207,58 +210,83 @@ float4 Extract16(uint i) return f; } -int4 wrapuv(int4 uv) +float4 wrapuv(float4 uv) { - // TODO: bitwise ops don't work with render target textures that were not rendered at the native resolution. - if(WMS == WMT) { - switch(WMS) + if(WMS == 0) { - case 0: uv &= MinMax.xyxy; break; - case 1: uv = clamp(uv, 0, MinMax.zwzw); break; - case 2: uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); break; - case 3: uv = (uv & MinMax.xyxy) | MinMax.zwzw; break; + uv = frac(uv); + } + else if(WMS == 1) + { + uv = saturate(uv); + } + else if(WMS == 2) + { + uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); + } + else if(WMS == 3) + { + uv = (float4)(((int4)(uv * WH.xyxy) & MskFix.xyxy) | MskFix.zwzw) / WH.xyxy; } } else { - switch(WMS) + if(WMS == 0) { - case 0: uv.xz &= MinMax.xx; break; - case 1: uv.xz = clamp(uv.xz, 0, MinMax.zz); break; - case 2: uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); break; - case 3: uv.xz = (uv.xz & MinMax.xx) | MinMax.zz; break; + uv.xz = frac(uv.xz); } - - switch(WMT) + else if(WMS == 1) { - case 0: uv.yw &= MinMax.yy; break; - case 1: uv.yw = clamp(uv.yw, 0, MinMax.ww); break; - case 2: uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); break; - case 3: uv.yw = (uv.yw & MinMax.yy) | MinMax.ww; break; + uv.xz = saturate(uv.xz); + } + else if(WMS == 2) + { + uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); + } + else if(WMS == 3) + { + uv.xz = (float2)(((int2)(uv * WH.xyxy).xz & MskFix.xx) | MskFix.zz) / WH; + } + + if(WMT == 0) + { + uv.yw = frac(uv.yw); + } + else if(WMT == 1) + { + uv.yw = saturate(uv.yw); + } + else if(WMT == 2) + { + uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); + } + else if(WMT == 3) + { + uv.yw = (float2)(((int2)(uv * WH.xyxy).yw & MskFix.yy) | MskFix.ww) / WH; } } return uv; } -float2 clampuv(float2 tc) +float2 clampuv(float2 uv) { if(WMS == 2 && WMT == 2) { - tc = clamp(tc, MinMaxF.xy, MinMaxF.zw); + uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); } else if(WMS == 2) { - tc.x = clamp(tc.x, MinMaxF.x, MinMaxF.z); + uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); } else if(WMT == 2) { - tc.y = clamp(tc.y, MinMaxF.y, MinMaxF.w); + uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); } - return tc; + return uv; } float4 sample(float2 tc, float w) @@ -282,12 +310,11 @@ float4 sample(float2 tc, float w) else { float w, h; - Texture.GetDimensions(w, h); - float4 tc2 = (tc * float2(w, h)).xyxy + float4(-0.499, -0.499, 0.501, 0.501); - float2 dd = frac(tc2.xy); - int4 uv = wrapuv((int4)tc2); + float4 uv2 = tc.xyxy + HalfTexel; + float2 dd = frac(uv2.xy * float2(w, h)); + float4 uv = wrapuv(uv2); float4 t00, t01, t10, t11; @@ -295,15 +322,15 @@ float4 sample(float2 tc, float w) { float4 a; - a.x = Texture.Load(int3(uv.xy, 0)).a; - a.y = Texture.Load(int3(uv.zy, 0)).a; - a.z = Texture.Load(int3(uv.xw, 0)).a; - a.w = Texture.Load(int3(uv.zw, 0)).a; + a.x = Texture.Sample(TextureSampler, uv.xy).a; + a.y = Texture.Sample(TextureSampler, uv.zy).a; + a.z = Texture.Sample(TextureSampler, uv.xw).a; + a.w = Texture.Sample(TextureSampler, uv.zw).a; - t00 = Palette.Load(a.x); - t01 = Palette.Load(a.y); - t10 = Palette.Load(a.z); - t11 = Palette.Load(a.w); + t00 = Palette.Sample(PaletteSampler, a.x); + t01 = Palette.Sample(PaletteSampler, a.y); + t10 = Palette.Sample(PaletteSampler, a.z); + t11 = Palette.Sample(PaletteSampler, a.w); } else if(BPP == 4) // 8HP + 16-bit palette { @@ -313,10 +340,10 @@ float4 sample(float2 tc, float w) { float4 r; - r.x = Texture.Load(int3(uv.xy, 0)).r; - r.y = Texture.Load(int3(uv.zy, 0)).r; - r.z = Texture.Load(int3(uv.xw, 0)).r; - r.w = Texture.Load(int3(uv.zw, 0)).r; + r.x = Texture.Sample(TextureSampler, uv.xy).r; + r.y = Texture.Sample(TextureSampler, uv.zy).r; + r.z = Texture.Sample(TextureSampler, uv.xw).r; + r.w = Texture.Sample(TextureSampler, uv.zw).r; uint4 i = r * 65535; @@ -327,10 +354,10 @@ float4 sample(float2 tc, float w) } else { - t00 = Texture.Load(int3(uv.xy, 0)); - t01 = Texture.Load(int3(uv.zy, 0)); - t10 = Texture.Load(int3(uv.xw, 0)); - t11 = Texture.Load(int3(uv.zw, 0)); + t00 = Texture.Sample(TextureSampler, uv.xy); + t01 = Texture.Sample(TextureSampler, uv.zy); + t10 = Texture.Sample(TextureSampler, uv.xw); + t11 = Texture.Sample(TextureSampler, uv.zw); } if(LTF) diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index 9c1011f735..1cbc591387 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -96,12 +96,12 @@ struct PS_INPUT #define AEM 0 #define TFX 0 #define TCC 1 -#define ATE 1 +#define ATE 0 #define ATST 4 -#define FOG 1 +#define FOG 0 #define CLR1 0 #define RT 0 -#define LTF 1 +#define LTF 0 #endif sampler Texture : register(s0); @@ -113,63 +113,84 @@ float4 wrapuv(float4 uv) { if(WMS == WMT) { - switch(WMS) + if(WMS == 0) + { + uv = frac(uv); + } + else if(WMS == 1) + { + uv = saturate(uv); + } + else if(WMS == 2) + { + uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); + } + else if(WMS == 3) { - case 0: uv = frac(uv); break; - case 1: uv = saturate(uv); break; - case 2: uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); break; - case 3: uv.x = tex1D(UMSKFIX, uv.x); uv.y = tex1D(VMSKFIX, uv.y); uv.z = tex1D(UMSKFIX, uv.z); uv.w = tex1D(VMSKFIX, uv.w); - break; } } else { - switch(WMS) + if(WMS == 0) + { + uv.xz = frac(uv.xz); + } + else if(WMS == 1) + { + uv.xz = saturate(uv.xz); + } + else if(WMS == 2) + { + uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); + } + else if(WMS == 3) { - case 0: uv.xz = frac(uv.xz); break; - case 1: uv.xz = saturate(uv.xz); break; - case 2: uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); break; - case 3: uv.x = tex1D(UMSKFIX, uv.x); uv.z = tex1D(UMSKFIX, uv.z); - break; } - switch(WMT) + if(WMT == 0) + { + uv.yw = frac(uv.yw); + } + else if(WMT == 1) + { + uv.yw = saturate(uv.yw); + } + else if(WMT == 2) + { + uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); + } + else if(WMT == 3) { - case 0: uv.yw = frac(uv.yw); break; - case 1: uv.yw = saturate(uv.yw); break; - case 2: uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); break; - case 3: uv.y = tex1D(VMSKFIX, uv.y); uv.w = tex1D(VMSKFIX, uv.w); - break; } } return uv; } -float2 clampuv(float2 tc) +float2 clampuv(float2 uv) { if(WMS == 2 && WMT == 2) { - tc = clamp(tc, MinMaxF.xy, MinMaxF.zw); + uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); } else if(WMS == 2) { - tc.x = clamp(tc.x, MinMaxF.x, MinMaxF.z); + uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); } else if(WMT == 2) { - tc.y = clamp(tc.y, MinMaxF.y, MinMaxF.w); + uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); } - return tc; + return uv; } float4 sample(float2 tc, float w) @@ -192,9 +213,9 @@ float4 sample(float2 tc, float w) } else { - float4 tc2 = tc.xyxy + HalfTexel; - float2 dd = frac(tc2.xy * WH); - float4 uv = wrapuv(tc2); + float4 uv2 = tc.xyxy + HalfTexel; + float2 dd = frac(uv2.xy * WH); + float4 uv = wrapuv(uv2); float4 t00, t01, t10, t11; diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index eef0a46699..ddc67531b0 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -115,10 +115,15 @@ typedef signed long long int64; // opengl +#include #include #include #include +#ifdef _WINDOWS +#include +#endif + // sse #if !defined(_M_SSE) && (defined(_M_AMD64) || defined(_M_IX86_FP) && _M_IX86_FP >= 2) From 359c596bd84df97e1c30fd8eec389ad00586fb08 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 13 Jun 2009 22:31:12 +0000 Subject: [PATCH 219/296] GSdx: added missing glew32.lib reference git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1363 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/vsprops/common.vsprops | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/GSdx/vsprops/common.vsprops b/plugins/GSdx/vsprops/common.vsprops index 1d0bb33cad..199b3bd8f4 100644 --- a/plugins/GSdx/vsprops/common.vsprops +++ b/plugins/GSdx/vsprops/common.vsprops @@ -18,7 +18,7 @@ /> Date: Sun, 14 Jun 2009 09:41:33 +0000 Subject: [PATCH 220/296] Linux: Experimental - Add a alternate, more flexible, build script written in Ruby. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1364 96395faa-99c1-11dd-bbfe-3dabce05a288 --- build.rb | 151 ++++++++++ plugins/zerogs/opengl/configure | 474 +++++++++++++++++++------------- 2 files changed, 437 insertions(+), 188 deletions(-) create mode 100755 build.rb diff --git a/build.rb b/build.rb new file mode 100755 index 0000000000..2198340f80 --- /dev/null +++ b/build.rb @@ -0,0 +1,151 @@ +#!/usr/bin/env ruby + +# Usage: ruby build.rb [option] [pcsx2, plugins, ] [dev,debug,release] [all, install, clean] +# If you don't specify pcsx2, plugins, or a plugin name, it will assume you want to rebuild everything. +# If you don't specify dev or debug, it assumes a release build. +# If it isn't all, install, or clean, it assumes install. + +# If you want other options, add them to $pcsx2_build_types. This is still a work in progress... +# --arcum42 + +require "fileutils.rb" +include FileUtils + +$main_dir = Dir.pwd +$pcsx2_install_dir = "#{$main_dir}/bin" +$plugin_install_dir = "#{$main_dir}/bin/plugins" + +$pcsx2_dir = "#{$main_dir}/pcsx2" +$plugins_dir = "#{$main_dir}/plugins" + +$pcsx2_prefix = " --prefix #{$main_dir}" +$plugins_prefix = " --prefix #{$plugin_install_dir}" + +$plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz"] + +$pcsx2_build_types = { + "dev" => " --enable-devbuild ", + "debug" => " --enable-debug ", + "release" => " " + } + +$pcsx2_release_params=["dev","debug","release"] +$make_params=["all", "clean","install"] + +def plugin_src_dir(plugin_name) + name = "#{$plugins_dir}/#{plugin_name}/" + case plugin_name + when "CDVDiso" then + name += "src" + when "CDVDisoEFP" then + name += "src/Linux" + when "CDVDlinuz" + name += "Src/Linux" + when "zerogs", "zzogl" + name += "opengl" + end + + return name +end + +def announce(my_program) + print "---------------\n" + print "Building #{my_program}\n" + print "---------------\n" +end + +def make(options) + system("make #{options}") +end + +def rebuild(options) + system("aclocal") + system("automake") + system("autoconf") + system("chmod +x configure") + system("./configure #{options}") + make "clean" +end + +def install(build_name) + make "install" + + case build_name + # If the package isn't inclined to obey simple instructions... + when "CDVDisoEFP" then + system("cp #{plugin_src_dir(build_name)}/cfgCDVDisoEFP #{$plugin_install_dir}") + system("cp #{plugin_src_dir(build_name)}/libCDVDisoEFP.so #{$plugin_install_dir}") + when "CDVDlinuz" then + system("cp #{plugin_src_dir(build_name)}/cfgCDVDlinuz #{$plugin_install_dir}") + system("cp #{plugin_src_dir(build_name)}/libCDVDlinuz.so #{$plugin_install_dir}") + when "PeopsSPU2" then + system("cp #{plugin_src_dir(build_name)}/libspu2Peops*.so* #{$plugin_install_dir}") + + # Copy the shaders over. Shouldn't the makefile do this? + when "zzogl","zerogs" then + system("cp #{plugin_src_dir(build_name)}/Win32/ps2hw.dat #{$plugin_install_dir}") + + #And while we have the opportunity... + when "pcsx2" then + svn_revision = `svn info | grep Revision:` + svn_revision = /[0-9]+/.match(svn_revision) + system("cp #{$pcsx2_install_dir}/pcsx2 #{$pcsx2_install_dir}/pcsx2-#{svn_revision}") + end +end + +def build(build_name, make_parameter) + announce "#{build_name.capitalize}" + + if build_name != "pcsx2" then + build_dir = plugin_src_dir(build_name) + else + build_dir = "#{$pcsx2_dir}" + end + + Dir.chdir build_dir + + case make_parameter + when "all" then + if build_name == "pcsx2" + rebuild($pcsx2_prefix) + else + rebuild($plugins_prefix) + end + install(build_name) + + when "clean" then + make "clean" + + else + install(build_name) + end + + Dir.chdir $main_dir +end + +build_parameter = "all" +make_parameter = "" +build_items = Array.new([]) + +ARGV.each do |x| + make_parameter = x if $make_params.include?(x) + + build_items.push(x) if $plugin_list.include?(x) or (x == "pcsx2") + $pcsx2_prefix = $pcsx2_build_types[x] + $pcsx2_prefix if $pcsx2_release_params.include?(x) + + if (x == "plugins") then + x = $plugin_list + build_items.push(x) + end +end + +if build_items.empty? then + build_items.push($plugin_list) + build_items.push("pcsx2") +end + +build_items.flatten! + +build_items.each do |x| + build(x,make_parameter) +end diff --git a/plugins/zerogs/opengl/configure b/plugins/zerogs/opengl/configure index e81eb7f7f7..9cd5f0cc3c 100644 --- a/plugins/zerogs/opengl/configure +++ b/plugins/zerogs/opengl/configure @@ -636,7 +636,9 @@ ac_includes_default="\ # include #endif" -ac_subst_vars='LTLIBOBJS +ac_subst_vars='am__EXEEXT_FALSE +am__EXEEXT_TRUE +LTLIBOBJS LIBOBJS EGREP GREP @@ -1859,7 +1861,7 @@ ac_compiler_gnu=$ac_cv_c_compiler_gnu -am__api_version='1.10' +am__api_version='1.11' ac_aux_dir= for ac_dir in "$srcdir" "$srcdir/.." "$srcdir/../.."; do @@ -1990,16 +1992,33 @@ $as_echo_n "checking whether build environment is sane... " >&6; } # Just in case sleep 1 echo timestamp > conftest.file +# Reject unsafe characters in $srcdir or the absolute working directory +# name. Accept space and tab only in the latter. +am_lf=' +' +case `pwd` in + *[\\\"\#\$\&\'\`$am_lf]*) + { { $as_echo "$as_me:$LINENO: error: unsafe absolute working directory name" >&5 +$as_echo "$as_me: error: unsafe absolute working directory name" >&2;} + { (exit 1); exit 1; }; };; +esac +case $srcdir in + *[\\\"\#\$\&\'\`$am_lf\ \ ]*) + { { $as_echo "$as_me:$LINENO: error: unsafe srcdir value: \`$srcdir'" >&5 +$as_echo "$as_me: error: unsafe srcdir value: \`$srcdir'" >&2;} + { (exit 1); exit 1; }; };; +esac + # Do `set' in a subshell so we don't clobber the current shell's # arguments. Must try -L first in case configure is actually a # symlink; some systems play weird games with the mod time of symlinks # (eg FreeBSD returns the mod time of the symlink's containing # directory). if ( - set X `ls -Lt $srcdir/configure conftest.file 2> /dev/null` + set X `ls -Lt "$srcdir/configure" conftest.file 2> /dev/null` if test "$*" = "X"; then # -L didn't work. - set X `ls -t $srcdir/configure conftest.file` + set X `ls -t "$srcdir/configure" conftest.file` fi rm -f conftest.file if test "$*" != "X $srcdir/configure conftest.file" \ @@ -2043,7 +2062,14 @@ program_transform_name=`$as_echo "$program_transform_name" | sed "$ac_script"` # expand $ac_aux_dir to an absolute path am_aux_dir=`cd $ac_aux_dir && pwd` -test x"${MISSING+set}" = xset || MISSING="\${SHELL} $am_aux_dir/missing" +if test x"${MISSING+set}" != xset; then + case $am_aux_dir in + *\ * | *\ *) + MISSING="\${SHELL} \"$am_aux_dir/missing\"" ;; + *) + MISSING="\${SHELL} $am_aux_dir/missing" ;; + esac +fi # Use eval to expand $SHELL if eval "$MISSING --run true"; then am_missing_run="$MISSING --run " @@ -2053,6 +2079,115 @@ else $as_echo "$as_me: WARNING: \`missing' script is too old or missing" >&2;} fi +if test x"${install_sh}" != xset; then + case $am_aux_dir in + *\ * | *\ *) + install_sh="\${SHELL} '$am_aux_dir/install-sh'" ;; + *) + install_sh="\${SHELL} $am_aux_dir/install-sh" + esac +fi + +# Installed binaries are usually stripped using `strip' when the user +# run `make install-strip'. However `strip' might not be the right +# tool to use in cross-compilation environments, therefore Automake +# will honor the `STRIP' environment variable to overrule this program. +if test "$cross_compiling" != no; then + if test -n "$ac_tool_prefix"; then + # Extract the first word of "${ac_tool_prefix}strip", so it can be a program name with args. +set dummy ${ac_tool_prefix}strip; ac_word=$2 +{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 +$as_echo_n "checking for $ac_word... " >&6; } +if test "${ac_cv_prog_STRIP+set}" = set; then + $as_echo_n "(cached) " >&6 +else + if test -n "$STRIP"; then + ac_cv_prog_STRIP="$STRIP" # Let the user override the test. +else +as_save_IFS=$IFS; IFS=$PATH_SEPARATOR +for as_dir in $PATH +do + IFS=$as_save_IFS + test -z "$as_dir" && as_dir=. + for ac_exec_ext in '' $ac_executable_extensions; do + if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then + ac_cv_prog_STRIP="${ac_tool_prefix}strip" + $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 + break 2 + fi +done +done +IFS=$as_save_IFS + +fi +fi +STRIP=$ac_cv_prog_STRIP +if test -n "$STRIP"; then + { $as_echo "$as_me:$LINENO: result: $STRIP" >&5 +$as_echo "$STRIP" >&6; } +else + { $as_echo "$as_me:$LINENO: result: no" >&5 +$as_echo "no" >&6; } +fi + + +fi +if test -z "$ac_cv_prog_STRIP"; then + ac_ct_STRIP=$STRIP + # Extract the first word of "strip", so it can be a program name with args. +set dummy strip; ac_word=$2 +{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 +$as_echo_n "checking for $ac_word... " >&6; } +if test "${ac_cv_prog_ac_ct_STRIP+set}" = set; then + $as_echo_n "(cached) " >&6 +else + if test -n "$ac_ct_STRIP"; then + ac_cv_prog_ac_ct_STRIP="$ac_ct_STRIP" # Let the user override the test. +else +as_save_IFS=$IFS; IFS=$PATH_SEPARATOR +for as_dir in $PATH +do + IFS=$as_save_IFS + test -z "$as_dir" && as_dir=. + for ac_exec_ext in '' $ac_executable_extensions; do + if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then + ac_cv_prog_ac_ct_STRIP="strip" + $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 + break 2 + fi +done +done +IFS=$as_save_IFS + +fi +fi +ac_ct_STRIP=$ac_cv_prog_ac_ct_STRIP +if test -n "$ac_ct_STRIP"; then + { $as_echo "$as_me:$LINENO: result: $ac_ct_STRIP" >&5 +$as_echo "$ac_ct_STRIP" >&6; } +else + { $as_echo "$as_me:$LINENO: result: no" >&5 +$as_echo "no" >&6; } +fi + + if test "x$ac_ct_STRIP" = x; then + STRIP=":" + else + case $cross_compiling:$ac_tool_warned in +yes:) +{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 +$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} +ac_tool_warned=yes ;; +esac + STRIP=$ac_ct_STRIP + fi +else + STRIP="$ac_cv_prog_STRIP" +fi + +fi +INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s" + { $as_echo "$as_me:$LINENO: checking for a thread-safe mkdir -p" >&5 $as_echo_n "checking for a thread-safe mkdir -p... " >&6; } if test -z "$MKDIR_P"; then @@ -2235,108 +2370,6 @@ AUTOHEADER=${AUTOHEADER-"${am_missing_run}autoheader"} MAKEINFO=${MAKEINFO-"${am_missing_run}makeinfo"} -install_sh=${install_sh-"\$(SHELL) $am_aux_dir/install-sh"} - -# Installed binaries are usually stripped using `strip' when the user -# run `make install-strip'. However `strip' might not be the right -# tool to use in cross-compilation environments, therefore Automake -# will honor the `STRIP' environment variable to overrule this program. -if test "$cross_compiling" != no; then - if test -n "$ac_tool_prefix"; then - # Extract the first word of "${ac_tool_prefix}strip", so it can be a program name with args. -set dummy ${ac_tool_prefix}strip; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_STRIP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$STRIP"; then - ac_cv_prog_STRIP="$STRIP" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_STRIP="${ac_tool_prefix}strip" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -STRIP=$ac_cv_prog_STRIP -if test -n "$STRIP"; then - { $as_echo "$as_me:$LINENO: result: $STRIP" >&5 -$as_echo "$STRIP" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - -fi -if test -z "$ac_cv_prog_STRIP"; then - ac_ct_STRIP=$STRIP - # Extract the first word of "strip", so it can be a program name with args. -set dummy strip; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_ac_ct_STRIP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$ac_ct_STRIP"; then - ac_cv_prog_ac_ct_STRIP="$ac_ct_STRIP" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_ac_ct_STRIP="strip" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -ac_ct_STRIP=$ac_cv_prog_ac_ct_STRIP -if test -n "$ac_ct_STRIP"; then - { $as_echo "$as_me:$LINENO: result: $ac_ct_STRIP" >&5 -$as_echo "$ac_ct_STRIP" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - if test "x$ac_ct_STRIP" = x; then - STRIP=":" - else - case $cross_compiling:$ac_tool_warned in -yes:) -{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 -$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} -ac_tool_warned=yes ;; -esac - STRIP=$ac_ct_STRIP - fi -else - STRIP="$ac_cv_prog_STRIP" -fi - -fi -INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s" - # We need awk for the "check" target. The system "awk" is bad on # some platforms. # Always define AMTAR for backward compatibility. @@ -3098,7 +3131,7 @@ ac_config_commands="$ac_config_commands depfiles" am_make=${MAKE-make} cat > confinc << 'END' am__doit: - @echo done + @echo this is the am__doit target .PHONY: am__doit END # If we don't find an include directive, just comment out the code. @@ -3109,24 +3142,24 @@ am__quote= _am_result=none # First try GNU make style include. echo "include confinc" > confmf -# We grep out `Entering directory' and `Leaving directory' -# messages which can occur if `w' ends up in MAKEFLAGS. -# In particular we don't look at `^make:' because GNU make might -# be invoked under some other name (usually "gmake"), in which -# case it prints its new name instead of `make'. -if test "`$am_make -s -f confmf 2> /dev/null | grep -v 'ing directory'`" = "done"; then - am__include=include - am__quote= - _am_result=GNU -fi +# Ignore all kinds of additional output from `make'. +case `$am_make -s -f confmf 2> /dev/null` in #( +*the\ am__doit\ target*) + am__include=include + am__quote= + _am_result=GNU + ;; +esac # Now try BSD make style include. if test "$am__include" = "#"; then echo '.include "confinc"' > confmf - if test "`$am_make -s -f confmf 2> /dev/null`" = "done"; then - am__include=.include - am__quote="\"" - _am_result=BSD - fi + case `$am_make -s -f confmf 2> /dev/null` in #( + *the\ am__doit\ target*) + am__include=.include + am__quote="\"" + _am_result=BSD + ;; + esac fi @@ -3183,6 +3216,11 @@ else if test "$am_compiler_list" = ""; then am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` fi + am__universal=false + case " $depcc " in #( + *\ -arch\ *\ -arch\ *) am__universal=true ;; + esac + for depmode in $am_compiler_list; do # Setup a source with many dependencies, because some compilers # like to wrap large dependency lists on column 80 (with \), and @@ -3200,7 +3238,17 @@ else done echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf + # We check with `-c' and `-o' for the sake of the "dashmstdout" + # mode. It turns out that the SunPro C++ compiler does not properly + # handle `-M -o', and we need to detect this. Also, some Intel + # versions had trouble with output in subdirs + am__obj=sub/conftest.${OBJEXT-o} + am__minus_obj="-o $am__obj" case $depmode in + gcc) + # This depmode causes a compiler race in universal mode. + test "$am__universal" = false || continue + ;; nosideeffect) # after this tag, mechanisms are not by side-effect, so they'll # only be used when explicitly requested @@ -3210,19 +3258,23 @@ else break fi ;; + msvisualcpp | msvcmsys) + # This compiler won't grok `-c -o', but also, the minuso test has + # not run yet. These depmodes are late enough in the game, and + # so weak that their functioning should not be impacted. + am__obj=conftest.${OBJEXT-o} + am__minus_obj= + ;; none) break ;; esac - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. if depmode=$depmode \ - source=sub/conftest.c object=sub/conftest.${OBJEXT-o} \ + source=sub/conftest.c object=$am__obj \ depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c -o sub/conftest.${OBJEXT-o} sub/conftest.c \ + $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ >/dev/null 2>conftest.err && grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftest.${OBJEXT-o} sub/conftest.Po > /dev/null 2>&1 && + grep $am__obj sub/conftest.Po > /dev/null 2>&1 && ${MAKE-make} -s -f confmf > /dev/null 2>&1; then # icc doesn't choke on unknown options, it will just issue warnings # or remarks (even with -Werror). So we grep stderr for any message @@ -3663,6 +3715,11 @@ else if test "$am_compiler_list" = ""; then am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` fi + am__universal=false + case " $depcc " in #( + *\ -arch\ *\ -arch\ *) am__universal=true ;; + esac + for depmode in $am_compiler_list; do # Setup a source with many dependencies, because some compilers # like to wrap large dependency lists on column 80 (with \), and @@ -3680,7 +3737,17 @@ else done echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf + # We check with `-c' and `-o' for the sake of the "dashmstdout" + # mode. It turns out that the SunPro C++ compiler does not properly + # handle `-M -o', and we need to detect this. Also, some Intel + # versions had trouble with output in subdirs + am__obj=sub/conftest.${OBJEXT-o} + am__minus_obj="-o $am__obj" case $depmode in + gcc) + # This depmode causes a compiler race in universal mode. + test "$am__universal" = false || continue + ;; nosideeffect) # after this tag, mechanisms are not by side-effect, so they'll # only be used when explicitly requested @@ -3690,19 +3757,23 @@ else break fi ;; + msvisualcpp | msvcmsys) + # This compiler won't grok `-c -o', but also, the minuso test has + # not run yet. These depmodes are late enough in the game, and + # so weak that their functioning should not be impacted. + am__obj=conftest.${OBJEXT-o} + am__minus_obj= + ;; none) break ;; esac - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. if depmode=$depmode \ - source=sub/conftest.c object=sub/conftest.${OBJEXT-o} \ + source=sub/conftest.c object=$am__obj \ depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c -o sub/conftest.${OBJEXT-o} sub/conftest.c \ + $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ >/dev/null 2>conftest.err && grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftest.${OBJEXT-o} sub/conftest.Po > /dev/null 2>&1 && + grep $am__obj sub/conftest.Po > /dev/null 2>&1 && ${MAKE-make} -s -f confmf > /dev/null 2>&1; then # icc doesn't choke on unknown options, it will just issue warnings # or remarks (even with -Werror). So we grep stderr for any message @@ -4333,6 +4404,9 @@ else if test "$am_compiler_list" = ""; then am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` fi + am__universal=false + + for depmode in $am_compiler_list; do # Setup a source with many dependencies, because some compilers # like to wrap large dependency lists on column 80 (with \), and @@ -4350,7 +4424,17 @@ else done echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf + # We check with `-c' and `-o' for the sake of the "dashmstdout" + # mode. It turns out that the SunPro C++ compiler does not properly + # handle `-M -o', and we need to detect this. Also, some Intel + # versions had trouble with output in subdirs + am__obj=sub/conftest.${OBJEXT-o} + am__minus_obj="-o $am__obj" case $depmode in + gcc) + # This depmode causes a compiler race in universal mode. + test "$am__universal" = false || continue + ;; nosideeffect) # after this tag, mechanisms are not by side-effect, so they'll # only be used when explicitly requested @@ -4360,19 +4444,23 @@ else break fi ;; + msvisualcpp | msvcmsys) + # This compiler won't grok `-c -o', but also, the minuso test has + # not run yet. These depmodes are late enough in the game, and + # so weak that their functioning should not be impacted. + am__obj=conftest.${OBJEXT-o} + am__minus_obj= + ;; none) break ;; esac - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. if depmode=$depmode \ - source=sub/conftest.c object=sub/conftest.${OBJEXT-o} \ + source=sub/conftest.c object=$am__obj \ depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c -o sub/conftest.${OBJEXT-o} sub/conftest.c \ + $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ >/dev/null 2>conftest.err && grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftest.${OBJEXT-o} sub/conftest.Po > /dev/null 2>&1 && + grep $am__obj sub/conftest.Po > /dev/null 2>&1 && ${MAKE-make} -s -f confmf > /dev/null 2>&1; then # icc doesn't choke on unknown options, it will just issue warnings # or remarks (even with -Werror). So we grep stderr for any message @@ -5997,6 +6085,14 @@ LIBOBJS=$ac_libobjs LTLIBOBJS=$ac_ltlibobjs + if test -n "$EXEEXT"; then + am__EXEEXT_TRUE= + am__EXEEXT_FALSE='#' +else + am__EXEEXT_TRUE='#' + am__EXEEXT_FALSE= +fi + if test -z "${AMDEP_TRUE}" && test -z "${AMDEP_FALSE}"; then { { $as_echo "$as_me:$LINENO: error: conditional \"AMDEP\" was never defined. Usually this means the macro was only invoked conditionally." >&5 @@ -7007,27 +7103,28 @@ $as_echo "$as_me: executing $ac_file commands" >&6;} case $ac_file$ac_mode in - "depfiles":C) test x"$AMDEP_TRUE" != x"" || # Autoconf 2.62 quotes --file arguments for eval, but not when files -# are listed without --file. Let's play safe and only enable the eval -# if we detect the quoting. -case $CONFIG_FILES in -*\'*) eval set x "$CONFIG_FILES" ;; -*) set x $CONFIG_FILES ;; -esac -shift -for mf -do - # Strip MF so we end up with the name of the file. - mf=`echo "$mf" | sed -e 's/:.*$//'` - # Check whether this is an Automake generated Makefile or not. - # We used to match only the files named `Makefile.in', but - # some people rename them; so instead we look at the file content. - # Grep'ing the first line is not enough: some people post-process - # each Makefile.in and add a new line on top of each file to say so. - # Grep'ing the whole file is not good either: AIX grep has a line - # limit of 2048, but all sed's we know have understand at least 4000. - if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then - dirpart=`$as_dirname -- "$mf" || + "depfiles":C) test x"$AMDEP_TRUE" != x"" || { + # Autoconf 2.62 quotes --file arguments for eval, but not when files + # are listed without --file. Let's play safe and only enable the eval + # if we detect the quoting. + case $CONFIG_FILES in + *\'*) eval set x "$CONFIG_FILES" ;; + *) set x $CONFIG_FILES ;; + esac + shift + for mf + do + # Strip MF so we end up with the name of the file. + mf=`echo "$mf" | sed -e 's/:.*$//'` + # Check whether this is an Automake generated Makefile or not. + # We used to match only the files named `Makefile.in', but + # some people rename them; so instead we look at the file content. + # Grep'ing the first line is not enough: some people post-process + # each Makefile.in and add a new line on top of each file to say so. + # Grep'ing the whole file is not good either: AIX grep has a line + # limit of 2048, but all sed's we know have understand at least 4000. + if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then + dirpart=`$as_dirname -- "$mf" || $as_expr X"$mf" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$mf" : 'X\(//\)[^/]' \| \ X"$mf" : 'X\(//\)$' \| \ @@ -7050,28 +7147,28 @@ $as_echo X"$mf" | q } s/.*/./; q'` - else - continue - fi - # Extract the definition of DEPDIR, am__include, and am__quote - # from the Makefile without running `make'. - DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"` - test -z "$DEPDIR" && continue - am__include=`sed -n 's/^am__include = //p' < "$mf"` - test -z "am__include" && continue - am__quote=`sed -n 's/^am__quote = //p' < "$mf"` - # When using ansi2knr, U may be empty or an underscore; expand it - U=`sed -n 's/^U = //p' < "$mf"` - # Find all dependency output files, they are included files with - # $(DEPDIR) in their names. We invoke sed twice because it is the - # simplest approach to changing $(DEPDIR) to its actual value in the - # expansion. - for file in `sed -n " - s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \ - sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g' -e 's/\$U/'"$U"'/g'`; do - # Make sure the directory exists. - test -f "$dirpart/$file" && continue - fdir=`$as_dirname -- "$file" || + else + continue + fi + # Extract the definition of DEPDIR, am__include, and am__quote + # from the Makefile without running `make'. + DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"` + test -z "$DEPDIR" && continue + am__include=`sed -n 's/^am__include = //p' < "$mf"` + test -z "am__include" && continue + am__quote=`sed -n 's/^am__quote = //p' < "$mf"` + # When using ansi2knr, U may be empty or an underscore; expand it + U=`sed -n 's/^U = //p' < "$mf"` + # Find all dependency output files, they are included files with + # $(DEPDIR) in their names. We invoke sed twice because it is the + # simplest approach to changing $(DEPDIR) to its actual value in the + # expansion. + for file in `sed -n " + s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \ + sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g' -e 's/\$U/'"$U"'/g'`; do + # Make sure the directory exists. + test -f "$dirpart/$file" && continue + fdir=`$as_dirname -- "$file" || $as_expr X"$file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$file" : 'X\(//\)[^/]' \| \ X"$file" : 'X\(//\)$' \| \ @@ -7094,7 +7191,7 @@ $as_echo X"$file" | q } s/.*/./; q'` - { as_dir=$dirpart/$fdir + { as_dir=$dirpart/$fdir case $as_dir in #( -*) as_dir=./$as_dir;; esac @@ -7135,10 +7232,11 @@ $as_echo X"$as_dir" | } || test -d "$as_dir" || { { $as_echo "$as_me:$LINENO: error: cannot create directory $as_dir" >&5 $as_echo "$as_me: error: cannot create directory $as_dir" >&2;} { (exit 1); exit 1; }; }; } - # echo "creating $dirpart/$file" - echo '# dummy' > "$dirpart/$file" + # echo "creating $dirpart/$file" + echo '# dummy' > "$dirpart/$file" + done done -done +} ;; esac From b9cb1d3a592fe862d141aa6ec07bcaf622b8ed55 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 14 Jun 2009 09:56:42 +0000 Subject: [PATCH 221/296] Delete a stray file that's automatically regenerated every time you build ZeroGS anyways... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1365 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/zerogs/opengl/configure | 7288 ------------------------------- 1 file changed, 7288 deletions(-) delete mode 100644 plugins/zerogs/opengl/configure diff --git a/plugins/zerogs/opengl/configure b/plugins/zerogs/opengl/configure deleted file mode 100644 index 9cd5f0cc3c..0000000000 --- a/plugins/zerogs/opengl/configure +++ /dev/null @@ -1,7288 +0,0 @@ -#! /bin/sh -# Guess values for system-dependent variables and create Makefiles. -# Generated by GNU Autoconf 2.63 for ZeroGSogl 0.96.2. -# -# Report bugs to . -# -# Copyright (C) 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000, 2001, -# 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc. -# This configure script is free software; the Free Software Foundation -# gives unlimited permission to copy, distribute and modify it. -## --------------------- ## -## M4sh Initialization. ## -## --------------------- ## - -# Be more Bourne compatible -DUALCASE=1; export DUALCASE # for MKS sh -if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then - emulate sh - NULLCMD=: - # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which - # is contrary to our usage. Disable this feature. - alias -g '${1+"$@"}'='"$@"' - setopt NO_GLOB_SUBST -else - case `(set -o) 2>/dev/null` in - *posix*) set -o posix ;; -esac - -fi - - - - -# PATH needs CR -# Avoid depending upon Character Ranges. -as_cr_letters='abcdefghijklmnopqrstuvwxyz' -as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ' -as_cr_Letters=$as_cr_letters$as_cr_LETTERS -as_cr_digits='0123456789' -as_cr_alnum=$as_cr_Letters$as_cr_digits - -as_nl=' -' -export as_nl -# Printing a long string crashes Solaris 7 /usr/bin/printf. -as_echo='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' -as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo -as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo$as_echo -if (test "X`printf %s $as_echo`" = "X$as_echo") 2>/dev/null; then - as_echo='printf %s\n' - as_echo_n='printf %s' -else - if test "X`(/usr/ucb/echo -n -n $as_echo) 2>/dev/null`" = "X-n $as_echo"; then - as_echo_body='eval /usr/ucb/echo -n "$1$as_nl"' - as_echo_n='/usr/ucb/echo -n' - else - as_echo_body='eval expr "X$1" : "X\\(.*\\)"' - as_echo_n_body='eval - arg=$1; - case $arg in - *"$as_nl"*) - expr "X$arg" : "X\\(.*\\)$as_nl"; - arg=`expr "X$arg" : ".*$as_nl\\(.*\\)"`;; - esac; - expr "X$arg" : "X\\(.*\\)" | tr -d "$as_nl" - ' - export as_echo_n_body - as_echo_n='sh -c $as_echo_n_body as_echo' - fi - export as_echo_body - as_echo='sh -c $as_echo_body as_echo' -fi - -# The user is always right. -if test "${PATH_SEPARATOR+set}" != set; then - PATH_SEPARATOR=: - (PATH='/bin;/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 && { - (PATH='/bin:/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 || - PATH_SEPARATOR=';' - } -fi - -# Support unset when possible. -if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then - as_unset=unset -else - as_unset=false -fi - - -# IFS -# We need space, tab and new line, in precisely that order. Quoting is -# there to prevent editors from complaining about space-tab. -# (If _AS_PATH_WALK were called with IFS unset, it would disable word -# splitting by setting IFS to empty value.) -IFS=" "" $as_nl" - -# Find who we are. Look in the path if we contain no directory separator. -case $0 in - *[\\/]* ) as_myself=$0 ;; - *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break -done -IFS=$as_save_IFS - - ;; -esac -# We did not find ourselves, most probably we were run as `sh COMMAND' -# in which case we are not to be found in the path. -if test "x$as_myself" = x; then - as_myself=$0 -fi -if test ! -f "$as_myself"; then - $as_echo "$as_myself: error: cannot find myself; rerun with an absolute file name" >&2 - { (exit 1); exit 1; } -fi - -# Work around bugs in pre-3.0 UWIN ksh. -for as_var in ENV MAIL MAILPATH -do ($as_unset $as_var) >/dev/null 2>&1 && $as_unset $as_var -done -PS1='$ ' -PS2='> ' -PS4='+ ' - -# NLS nuisances. -LC_ALL=C -export LC_ALL -LANGUAGE=C -export LANGUAGE - -# Required to use basename. -if expr a : '\(a\)' >/dev/null 2>&1 && - test "X`expr 00001 : '.*\(...\)'`" = X001; then - as_expr=expr -else - as_expr=false -fi - -if (basename -- /) >/dev/null 2>&1 && test "X`basename -- / 2>&1`" = "X/"; then - as_basename=basename -else - as_basename=false -fi - - -# Name of the executable. -as_me=`$as_basename -- "$0" || -$as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \ - X"$0" : 'X\(//\)$' \| \ - X"$0" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X/"$0" | - sed '/^.*\/\([^/][^/]*\)\/*$/{ - s//\1/ - q - } - /^X\/\(\/\/\)$/{ - s//\1/ - q - } - /^X\/\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - -# CDPATH. -$as_unset CDPATH - - -if test "x$CONFIG_SHELL" = x; then - if (eval ":") 2>/dev/null; then - as_have_required=yes -else - as_have_required=no -fi - - if test $as_have_required = yes && (eval ": -(as_func_return () { - (exit \$1) -} -as_func_success () { - as_func_return 0 -} -as_func_failure () { - as_func_return 1 -} -as_func_ret_success () { - return 0 -} -as_func_ret_failure () { - return 1 -} - -exitcode=0 -if as_func_success; then - : -else - exitcode=1 - echo as_func_success failed. -fi - -if as_func_failure; then - exitcode=1 - echo as_func_failure succeeded. -fi - -if as_func_ret_success; then - : -else - exitcode=1 - echo as_func_ret_success failed. -fi - -if as_func_ret_failure; then - exitcode=1 - echo as_func_ret_failure succeeded. -fi - -if ( set x; as_func_ret_success y && test x = \"\$1\" ); then - : -else - exitcode=1 - echo positional parameters were not saved. -fi - -test \$exitcode = 0) || { (exit 1); exit 1; } - -( - as_lineno_1=\$LINENO - as_lineno_2=\$LINENO - test \"x\$as_lineno_1\" != \"x\$as_lineno_2\" && - test \"x\`expr \$as_lineno_1 + 1\`\" = \"x\$as_lineno_2\") || { (exit 1); exit 1; } -") 2> /dev/null; then - : -else - as_candidate_shells= - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in /bin$PATH_SEPARATOR/usr/bin$PATH_SEPARATOR$PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - case $as_dir in - /*) - for as_base in sh bash ksh sh5; do - as_candidate_shells="$as_candidate_shells $as_dir/$as_base" - done;; - esac -done -IFS=$as_save_IFS - - - for as_shell in $as_candidate_shells $SHELL; do - # Try only shells that exist, to save several forks. - if { test -f "$as_shell" || test -f "$as_shell.exe"; } && - { ("$as_shell") 2> /dev/null <<\_ASEOF -if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then - emulate sh - NULLCMD=: - # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which - # is contrary to our usage. Disable this feature. - alias -g '${1+"$@"}'='"$@"' - setopt NO_GLOB_SUBST -else - case `(set -o) 2>/dev/null` in - *posix*) set -o posix ;; -esac - -fi - - -: -_ASEOF -}; then - CONFIG_SHELL=$as_shell - as_have_required=yes - if { "$as_shell" 2> /dev/null <<\_ASEOF -if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then - emulate sh - NULLCMD=: - # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which - # is contrary to our usage. Disable this feature. - alias -g '${1+"$@"}'='"$@"' - setopt NO_GLOB_SUBST -else - case `(set -o) 2>/dev/null` in - *posix*) set -o posix ;; -esac - -fi - - -: -(as_func_return () { - (exit $1) -} -as_func_success () { - as_func_return 0 -} -as_func_failure () { - as_func_return 1 -} -as_func_ret_success () { - return 0 -} -as_func_ret_failure () { - return 1 -} - -exitcode=0 -if as_func_success; then - : -else - exitcode=1 - echo as_func_success failed. -fi - -if as_func_failure; then - exitcode=1 - echo as_func_failure succeeded. -fi - -if as_func_ret_success; then - : -else - exitcode=1 - echo as_func_ret_success failed. -fi - -if as_func_ret_failure; then - exitcode=1 - echo as_func_ret_failure succeeded. -fi - -if ( set x; as_func_ret_success y && test x = "$1" ); then - : -else - exitcode=1 - echo positional parameters were not saved. -fi - -test $exitcode = 0) || { (exit 1); exit 1; } - -( - as_lineno_1=$LINENO - as_lineno_2=$LINENO - test "x$as_lineno_1" != "x$as_lineno_2" && - test "x`expr $as_lineno_1 + 1`" = "x$as_lineno_2") || { (exit 1); exit 1; } - -_ASEOF -}; then - break -fi - -fi - - done - - if test "x$CONFIG_SHELL" != x; then - for as_var in BASH_ENV ENV - do ($as_unset $as_var) >/dev/null 2>&1 && $as_unset $as_var - done - export CONFIG_SHELL - exec "$CONFIG_SHELL" "$as_myself" ${1+"$@"} -fi - - - if test $as_have_required = no; then - echo This script requires a shell more modern than all the - echo shells that I found on your system. Please install a - echo modern shell, or manually run the script under such a - echo shell if you do have one. - { (exit 1); exit 1; } -fi - - -fi - -fi - - - -(eval "as_func_return () { - (exit \$1) -} -as_func_success () { - as_func_return 0 -} -as_func_failure () { - as_func_return 1 -} -as_func_ret_success () { - return 0 -} -as_func_ret_failure () { - return 1 -} - -exitcode=0 -if as_func_success; then - : -else - exitcode=1 - echo as_func_success failed. -fi - -if as_func_failure; then - exitcode=1 - echo as_func_failure succeeded. -fi - -if as_func_ret_success; then - : -else - exitcode=1 - echo as_func_ret_success failed. -fi - -if as_func_ret_failure; then - exitcode=1 - echo as_func_ret_failure succeeded. -fi - -if ( set x; as_func_ret_success y && test x = \"\$1\" ); then - : -else - exitcode=1 - echo positional parameters were not saved. -fi - -test \$exitcode = 0") || { - echo No shell found that supports shell functions. - echo Please tell bug-autoconf@gnu.org about your system, - echo including any error possibly output before this message. - echo This can help us improve future autoconf versions. - echo Configuration will now proceed without shell functions. -} - - - - as_lineno_1=$LINENO - as_lineno_2=$LINENO - test "x$as_lineno_1" != "x$as_lineno_2" && - test "x`expr $as_lineno_1 + 1`" = "x$as_lineno_2" || { - - # Create $as_me.lineno as a copy of $as_myself, but with $LINENO - # uniformly replaced by the line number. The first 'sed' inserts a - # line-number line after each line using $LINENO; the second 'sed' - # does the real work. The second script uses 'N' to pair each - # line-number line with the line containing $LINENO, and appends - # trailing '-' during substitution so that $LINENO is not a special - # case at line end. - # (Raja R Harinath suggested sed '=', and Paul Eggert wrote the - # scripts with optimization help from Paolo Bonzini. Blame Lee - # E. McMahon (1931-1989) for sed's syntax. :-) - sed -n ' - p - /[$]LINENO/= - ' <$as_myself | - sed ' - s/[$]LINENO.*/&-/ - t lineno - b - :lineno - N - :loop - s/[$]LINENO\([^'$as_cr_alnum'_].*\n\)\(.*\)/\2\1\2/ - t loop - s/-\n.*// - ' >$as_me.lineno && - chmod +x "$as_me.lineno" || - { $as_echo "$as_me: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&2 - { (exit 1); exit 1; }; } - - # Don't try to exec as it changes $[0], causing all sort of problems - # (the dirname of $[0] is not the place where we might find the - # original and so on. Autoconf is especially sensitive to this). - . "./$as_me.lineno" - # Exit status is that of the last command. - exit -} - - -if (as_dir=`dirname -- /` && test "X$as_dir" = X/) >/dev/null 2>&1; then - as_dirname=dirname -else - as_dirname=false -fi - -ECHO_C= ECHO_N= ECHO_T= -case `echo -n x` in --n*) - case `echo 'x\c'` in - *c*) ECHO_T=' ';; # ECHO_T is single tab character. - *) ECHO_C='\c';; - esac;; -*) - ECHO_N='-n';; -esac -if expr a : '\(a\)' >/dev/null 2>&1 && - test "X`expr 00001 : '.*\(...\)'`" = X001; then - as_expr=expr -else - as_expr=false -fi - -rm -f conf$$ conf$$.exe conf$$.file -if test -d conf$$.dir; then - rm -f conf$$.dir/conf$$.file -else - rm -f conf$$.dir - mkdir conf$$.dir 2>/dev/null -fi -if (echo >conf$$.file) 2>/dev/null; then - if ln -s conf$$.file conf$$ 2>/dev/null; then - as_ln_s='ln -s' - # ... but there are two gotchas: - # 1) On MSYS, both `ln -s file dir' and `ln file dir' fail. - # 2) DJGPP < 2.04 has no symlinks; `ln -s' creates a wrapper executable. - # In both cases, we have to default to `cp -p'. - ln -s conf$$.file conf$$.dir 2>/dev/null && test ! -f conf$$.exe || - as_ln_s='cp -p' - elif ln conf$$.file conf$$ 2>/dev/null; then - as_ln_s=ln - else - as_ln_s='cp -p' - fi -else - as_ln_s='cp -p' -fi -rm -f conf$$ conf$$.exe conf$$.dir/conf$$.file conf$$.file -rmdir conf$$.dir 2>/dev/null - -if mkdir -p . 2>/dev/null; then - as_mkdir_p=: -else - test -d ./-p && rmdir ./-p - as_mkdir_p=false -fi - -if test -x / >/dev/null 2>&1; then - as_test_x='test -x' -else - if ls -dL / >/dev/null 2>&1; then - as_ls_L_option=L - else - as_ls_L_option= - fi - as_test_x=' - eval sh -c '\'' - if test -d "$1"; then - test -d "$1/."; - else - case $1 in - -*)set "./$1";; - esac; - case `ls -ld'$as_ls_L_option' "$1" 2>/dev/null` in - ???[sx]*):;;*)false;;esac;fi - '\'' sh - ' -fi -as_executable_p=$as_test_x - -# Sed expression to map a string onto a valid CPP name. -as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'" - -# Sed expression to map a string onto a valid variable name. -as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'" - - - -exec 7<&0 &1 - -# Name of the host. -# hostname on some systems (SVR3.2, Linux) returns a bogus exit status, -# so uname gets run too. -ac_hostname=`(hostname || uname -n) 2>/dev/null | sed 1q` - -# -# Initializations. -# -ac_default_prefix=/usr/local -ac_clean_files= -ac_config_libobj_dir=. -LIBOBJS= -cross_compiling=no -subdirs= -MFLAGS= -MAKEFLAGS= -SHELL=${CONFIG_SHELL-/bin/sh} - -# Identity of this package. -PACKAGE_NAME='ZeroGSogl' -PACKAGE_TARNAME='zerogsogl' -PACKAGE_VERSION='0.96.2' -PACKAGE_STRING='ZeroGSogl 0.96.2' -PACKAGE_BUGREPORT='zerofrog@gmail.com' - -# Factoring default headers for most tests. -ac_includes_default="\ -#include -#ifdef HAVE_SYS_TYPES_H -# include -#endif -#ifdef HAVE_SYS_STAT_H -# include -#endif -#ifdef STDC_HEADERS -# include -# include -#else -# ifdef HAVE_STDLIB_H -# include -# endif -#endif -#ifdef HAVE_STRING_H -# if !defined STDC_HEADERS && defined HAVE_MEMORY_H -# include -# endif -# include -#endif -#ifdef HAVE_STRINGS_H -# include -#endif -#ifdef HAVE_INTTYPES_H -# include -#endif -#ifdef HAVE_STDINT_H -# include -#endif -#ifdef HAVE_UNISTD_H -# include -#endif" - -ac_subst_vars='am__EXEEXT_FALSE -am__EXEEXT_TRUE -LTLIBOBJS -LIBOBJS -EGREP -GREP -SHARED_LDFLAGS -so_ext -GTK_CONFIG -SSE2_FALSE -SSE2_TRUE -RELEASE_TO_PUBLIC_FALSE -RELEASE_TO_PUBLIC_TRUE -ZEROGS_SONAME -ZEROGS_RELEASE -ZEROGS_AGE -ZEROGS_REVISION -ZEROGS_CURRENT -am__fastdepCCAS_FALSE -am__fastdepCCAS_TRUE -CCASDEPMODE -CCASFLAGS -CCAS -RANLIB -CPP -am__fastdepCXX_FALSE -am__fastdepCXX_TRUE -CXXDEPMODE -ac_ct_CXX -CXXFLAGS -CXX -am__fastdepCC_FALSE -am__fastdepCC_TRUE -CCDEPMODE -AMDEPBACKSLASH -AMDEP_FALSE -AMDEP_TRUE -am__quote -am__include -DEPDIR -OBJEXT -EXEEXT -ac_ct_CC -CPPFLAGS -LDFLAGS -CFLAGS -CC -am__untar -am__tar -AMTAR -am__leading_dot -SET_MAKE -AWK -mkdir_p -MKDIR_P -INSTALL_STRIP_PROGRAM -STRIP -install_sh -MAKEINFO -AUTOHEADER -AUTOMAKE -AUTOCONF -ACLOCAL -VERSION -PACKAGE -CYGPATH_W -am__isrc -INSTALL_DATA -INSTALL_SCRIPT -INSTALL_PROGRAM -target_alias -host_alias -build_alias -LIBS -ECHO_T -ECHO_N -ECHO_C -DEFS -mandir -localedir -libdir -psdir -pdfdir -dvidir -htmldir -infodir -docdir -oldincludedir -includedir -localstatedir -sharedstatedir -sysconfdir -datadir -datarootdir -libexecdir -sbindir -bindir -program_transform_name -prefix -exec_prefix -PACKAGE_BUGREPORT -PACKAGE_STRING -PACKAGE_VERSION -PACKAGE_TARNAME -PACKAGE_NAME -PATH_SEPARATOR -SHELL' -ac_subst_files='' -ac_user_opts=' -enable_option_checking -enable_dependency_tracking -enable_debug -enable_devbuild -enable_sse2 -' - ac_precious_vars='build_alias -host_alias -target_alias -CC -CFLAGS -LDFLAGS -LIBS -CPPFLAGS -CXX -CXXFLAGS -CCC -CPP -CCAS -CCASFLAGS' - - -# Initialize some variables set by options. -ac_init_help= -ac_init_version=false -ac_unrecognized_opts= -ac_unrecognized_sep= -# The variables have the same names as the options, with -# dashes changed to underlines. -cache_file=/dev/null -exec_prefix=NONE -no_create= -no_recursion= -prefix=NONE -program_prefix=NONE -program_suffix=NONE -program_transform_name=s,x,x, -silent= -site= -srcdir= -verbose= -x_includes=NONE -x_libraries=NONE - -# Installation directory options. -# These are left unexpanded so users can "make install exec_prefix=/foo" -# and all the variables that are supposed to be based on exec_prefix -# by default will actually change. -# Use braces instead of parens because sh, perl, etc. also accept them. -# (The list follows the same order as the GNU Coding Standards.) -bindir='${exec_prefix}/bin' -sbindir='${exec_prefix}/sbin' -libexecdir='${exec_prefix}/libexec' -datarootdir='${prefix}/share' -datadir='${datarootdir}' -sysconfdir='${prefix}/etc' -sharedstatedir='${prefix}/com' -localstatedir='${prefix}/var' -includedir='${prefix}/include' -oldincludedir='/usr/include' -docdir='${datarootdir}/doc/${PACKAGE_TARNAME}' -infodir='${datarootdir}/info' -htmldir='${docdir}' -dvidir='${docdir}' -pdfdir='${docdir}' -psdir='${docdir}' -libdir='${exec_prefix}/lib' -localedir='${datarootdir}/locale' -mandir='${datarootdir}/man' - -ac_prev= -ac_dashdash= -for ac_option -do - # If the previous option needs an argument, assign it. - if test -n "$ac_prev"; then - eval $ac_prev=\$ac_option - ac_prev= - continue - fi - - case $ac_option in - *=*) ac_optarg=`expr "X$ac_option" : '[^=]*=\(.*\)'` ;; - *) ac_optarg=yes ;; - esac - - # Accept the important Cygnus configure options, so we can diagnose typos. - - case $ac_dashdash$ac_option in - --) - ac_dashdash=yes ;; - - -bindir | --bindir | --bindi | --bind | --bin | --bi) - ac_prev=bindir ;; - -bindir=* | --bindir=* | --bindi=* | --bind=* | --bin=* | --bi=*) - bindir=$ac_optarg ;; - - -build | --build | --buil | --bui | --bu) - ac_prev=build_alias ;; - -build=* | --build=* | --buil=* | --bui=* | --bu=*) - build_alias=$ac_optarg ;; - - -cache-file | --cache-file | --cache-fil | --cache-fi \ - | --cache-f | --cache- | --cache | --cach | --cac | --ca | --c) - ac_prev=cache_file ;; - -cache-file=* | --cache-file=* | --cache-fil=* | --cache-fi=* \ - | --cache-f=* | --cache-=* | --cache=* | --cach=* | --cac=* | --ca=* | --c=*) - cache_file=$ac_optarg ;; - - --config-cache | -C) - cache_file=config.cache ;; - - -datadir | --datadir | --datadi | --datad) - ac_prev=datadir ;; - -datadir=* | --datadir=* | --datadi=* | --datad=*) - datadir=$ac_optarg ;; - - -datarootdir | --datarootdir | --datarootdi | --datarootd | --dataroot \ - | --dataroo | --dataro | --datar) - ac_prev=datarootdir ;; - -datarootdir=* | --datarootdir=* | --datarootdi=* | --datarootd=* \ - | --dataroot=* | --dataroo=* | --dataro=* | --datar=*) - datarootdir=$ac_optarg ;; - - -disable-* | --disable-*) - ac_useropt=`expr "x$ac_option" : 'x-*disable-\(.*\)'` - # Reject names that are not valid shell variable names. - expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && - { $as_echo "$as_me: error: invalid feature name: $ac_useropt" >&2 - { (exit 1); exit 1; }; } - ac_useropt_orig=$ac_useropt - ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` - case $ac_user_opts in - *" -"enable_$ac_useropt" -"*) ;; - *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--disable-$ac_useropt_orig" - ac_unrecognized_sep=', ';; - esac - eval enable_$ac_useropt=no ;; - - -docdir | --docdir | --docdi | --doc | --do) - ac_prev=docdir ;; - -docdir=* | --docdir=* | --docdi=* | --doc=* | --do=*) - docdir=$ac_optarg ;; - - -dvidir | --dvidir | --dvidi | --dvid | --dvi | --dv) - ac_prev=dvidir ;; - -dvidir=* | --dvidir=* | --dvidi=* | --dvid=* | --dvi=* | --dv=*) - dvidir=$ac_optarg ;; - - -enable-* | --enable-*) - ac_useropt=`expr "x$ac_option" : 'x-*enable-\([^=]*\)'` - # Reject names that are not valid shell variable names. - expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && - { $as_echo "$as_me: error: invalid feature name: $ac_useropt" >&2 - { (exit 1); exit 1; }; } - ac_useropt_orig=$ac_useropt - ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` - case $ac_user_opts in - *" -"enable_$ac_useropt" -"*) ;; - *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--enable-$ac_useropt_orig" - ac_unrecognized_sep=', ';; - esac - eval enable_$ac_useropt=\$ac_optarg ;; - - -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi \ - | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- \ - | --exec | --exe | --ex) - ac_prev=exec_prefix ;; - -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* \ - | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* \ - | --exec=* | --exe=* | --ex=*) - exec_prefix=$ac_optarg ;; - - -gas | --gas | --ga | --g) - # Obsolete; use --with-gas. - with_gas=yes ;; - - -help | --help | --hel | --he | -h) - ac_init_help=long ;; - -help=r* | --help=r* | --hel=r* | --he=r* | -hr*) - ac_init_help=recursive ;; - -help=s* | --help=s* | --hel=s* | --he=s* | -hs*) - ac_init_help=short ;; - - -host | --host | --hos | --ho) - ac_prev=host_alias ;; - -host=* | --host=* | --hos=* | --ho=*) - host_alias=$ac_optarg ;; - - -htmldir | --htmldir | --htmldi | --htmld | --html | --htm | --ht) - ac_prev=htmldir ;; - -htmldir=* | --htmldir=* | --htmldi=* | --htmld=* | --html=* | --htm=* \ - | --ht=*) - htmldir=$ac_optarg ;; - - -includedir | --includedir | --includedi | --included | --include \ - | --includ | --inclu | --incl | --inc) - ac_prev=includedir ;; - -includedir=* | --includedir=* | --includedi=* | --included=* | --include=* \ - | --includ=* | --inclu=* | --incl=* | --inc=*) - includedir=$ac_optarg ;; - - -infodir | --infodir | --infodi | --infod | --info | --inf) - ac_prev=infodir ;; - -infodir=* | --infodir=* | --infodi=* | --infod=* | --info=* | --inf=*) - infodir=$ac_optarg ;; - - -libdir | --libdir | --libdi | --libd) - ac_prev=libdir ;; - -libdir=* | --libdir=* | --libdi=* | --libd=*) - libdir=$ac_optarg ;; - - -libexecdir | --libexecdir | --libexecdi | --libexecd | --libexec \ - | --libexe | --libex | --libe) - ac_prev=libexecdir ;; - -libexecdir=* | --libexecdir=* | --libexecdi=* | --libexecd=* | --libexec=* \ - | --libexe=* | --libex=* | --libe=*) - libexecdir=$ac_optarg ;; - - -localedir | --localedir | --localedi | --localed | --locale) - ac_prev=localedir ;; - -localedir=* | --localedir=* | --localedi=* | --localed=* | --locale=*) - localedir=$ac_optarg ;; - - -localstatedir | --localstatedir | --localstatedi | --localstated \ - | --localstate | --localstat | --localsta | --localst | --locals) - ac_prev=localstatedir ;; - -localstatedir=* | --localstatedir=* | --localstatedi=* | --localstated=* \ - | --localstate=* | --localstat=* | --localsta=* | --localst=* | --locals=*) - localstatedir=$ac_optarg ;; - - -mandir | --mandir | --mandi | --mand | --man | --ma | --m) - ac_prev=mandir ;; - -mandir=* | --mandir=* | --mandi=* | --mand=* | --man=* | --ma=* | --m=*) - mandir=$ac_optarg ;; - - -nfp | --nfp | --nf) - # Obsolete; use --without-fp. - with_fp=no ;; - - -no-create | --no-create | --no-creat | --no-crea | --no-cre \ - | --no-cr | --no-c | -n) - no_create=yes ;; - - -no-recursion | --no-recursion | --no-recursio | --no-recursi \ - | --no-recurs | --no-recur | --no-recu | --no-rec | --no-re | --no-r) - no_recursion=yes ;; - - -oldincludedir | --oldincludedir | --oldincludedi | --oldincluded \ - | --oldinclude | --oldinclud | --oldinclu | --oldincl | --oldinc \ - | --oldin | --oldi | --old | --ol | --o) - ac_prev=oldincludedir ;; - -oldincludedir=* | --oldincludedir=* | --oldincludedi=* | --oldincluded=* \ - | --oldinclude=* | --oldinclud=* | --oldinclu=* | --oldincl=* | --oldinc=* \ - | --oldin=* | --oldi=* | --old=* | --ol=* | --o=*) - oldincludedir=$ac_optarg ;; - - -prefix | --prefix | --prefi | --pref | --pre | --pr | --p) - ac_prev=prefix ;; - -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*) - prefix=$ac_optarg ;; - - -program-prefix | --program-prefix | --program-prefi | --program-pref \ - | --program-pre | --program-pr | --program-p) - ac_prev=program_prefix ;; - -program-prefix=* | --program-prefix=* | --program-prefi=* \ - | --program-pref=* | --program-pre=* | --program-pr=* | --program-p=*) - program_prefix=$ac_optarg ;; - - -program-suffix | --program-suffix | --program-suffi | --program-suff \ - | --program-suf | --program-su | --program-s) - ac_prev=program_suffix ;; - -program-suffix=* | --program-suffix=* | --program-suffi=* \ - | --program-suff=* | --program-suf=* | --program-su=* | --program-s=*) - program_suffix=$ac_optarg ;; - - -program-transform-name | --program-transform-name \ - | --program-transform-nam | --program-transform-na \ - | --program-transform-n | --program-transform- \ - | --program-transform | --program-transfor \ - | --program-transfo | --program-transf \ - | --program-trans | --program-tran \ - | --progr-tra | --program-tr | --program-t) - ac_prev=program_transform_name ;; - -program-transform-name=* | --program-transform-name=* \ - | --program-transform-nam=* | --program-transform-na=* \ - | --program-transform-n=* | --program-transform-=* \ - | --program-transform=* | --program-transfor=* \ - | --program-transfo=* | --program-transf=* \ - | --program-trans=* | --program-tran=* \ - | --progr-tra=* | --program-tr=* | --program-t=*) - program_transform_name=$ac_optarg ;; - - -pdfdir | --pdfdir | --pdfdi | --pdfd | --pdf | --pd) - ac_prev=pdfdir ;; - -pdfdir=* | --pdfdir=* | --pdfdi=* | --pdfd=* | --pdf=* | --pd=*) - pdfdir=$ac_optarg ;; - - -psdir | --psdir | --psdi | --psd | --ps) - ac_prev=psdir ;; - -psdir=* | --psdir=* | --psdi=* | --psd=* | --ps=*) - psdir=$ac_optarg ;; - - -q | -quiet | --quiet | --quie | --qui | --qu | --q \ - | -silent | --silent | --silen | --sile | --sil) - silent=yes ;; - - -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb) - ac_prev=sbindir ;; - -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \ - | --sbi=* | --sb=*) - sbindir=$ac_optarg ;; - - -sharedstatedir | --sharedstatedir | --sharedstatedi \ - | --sharedstated | --sharedstate | --sharedstat | --sharedsta \ - | --sharedst | --shareds | --shared | --share | --shar \ - | --sha | --sh) - ac_prev=sharedstatedir ;; - -sharedstatedir=* | --sharedstatedir=* | --sharedstatedi=* \ - | --sharedstated=* | --sharedstate=* | --sharedstat=* | --sharedsta=* \ - | --sharedst=* | --shareds=* | --shared=* | --share=* | --shar=* \ - | --sha=* | --sh=*) - sharedstatedir=$ac_optarg ;; - - -site | --site | --sit) - ac_prev=site ;; - -site=* | --site=* | --sit=*) - site=$ac_optarg ;; - - -srcdir | --srcdir | --srcdi | --srcd | --src | --sr) - ac_prev=srcdir ;; - -srcdir=* | --srcdir=* | --srcdi=* | --srcd=* | --src=* | --sr=*) - srcdir=$ac_optarg ;; - - -sysconfdir | --sysconfdir | --sysconfdi | --sysconfd | --sysconf \ - | --syscon | --sysco | --sysc | --sys | --sy) - ac_prev=sysconfdir ;; - -sysconfdir=* | --sysconfdir=* | --sysconfdi=* | --sysconfd=* | --sysconf=* \ - | --syscon=* | --sysco=* | --sysc=* | --sys=* | --sy=*) - sysconfdir=$ac_optarg ;; - - -target | --target | --targe | --targ | --tar | --ta | --t) - ac_prev=target_alias ;; - -target=* | --target=* | --targe=* | --targ=* | --tar=* | --ta=* | --t=*) - target_alias=$ac_optarg ;; - - -v | -verbose | --verbose | --verbos | --verbo | --verb) - verbose=yes ;; - - -version | --version | --versio | --versi | --vers | -V) - ac_init_version=: ;; - - -with-* | --with-*) - ac_useropt=`expr "x$ac_option" : 'x-*with-\([^=]*\)'` - # Reject names that are not valid shell variable names. - expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && - { $as_echo "$as_me: error: invalid package name: $ac_useropt" >&2 - { (exit 1); exit 1; }; } - ac_useropt_orig=$ac_useropt - ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` - case $ac_user_opts in - *" -"with_$ac_useropt" -"*) ;; - *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--with-$ac_useropt_orig" - ac_unrecognized_sep=', ';; - esac - eval with_$ac_useropt=\$ac_optarg ;; - - -without-* | --without-*) - ac_useropt=`expr "x$ac_option" : 'x-*without-\(.*\)'` - # Reject names that are not valid shell variable names. - expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && - { $as_echo "$as_me: error: invalid package name: $ac_useropt" >&2 - { (exit 1); exit 1; }; } - ac_useropt_orig=$ac_useropt - ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` - case $ac_user_opts in - *" -"with_$ac_useropt" -"*) ;; - *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--without-$ac_useropt_orig" - ac_unrecognized_sep=', ';; - esac - eval with_$ac_useropt=no ;; - - --x) - # Obsolete; use --with-x. - with_x=yes ;; - - -x-includes | --x-includes | --x-include | --x-includ | --x-inclu \ - | --x-incl | --x-inc | --x-in | --x-i) - ac_prev=x_includes ;; - -x-includes=* | --x-includes=* | --x-include=* | --x-includ=* | --x-inclu=* \ - | --x-incl=* | --x-inc=* | --x-in=* | --x-i=*) - x_includes=$ac_optarg ;; - - -x-libraries | --x-libraries | --x-librarie | --x-librari \ - | --x-librar | --x-libra | --x-libr | --x-lib | --x-li | --x-l) - ac_prev=x_libraries ;; - -x-libraries=* | --x-libraries=* | --x-librarie=* | --x-librari=* \ - | --x-librar=* | --x-libra=* | --x-libr=* | --x-lib=* | --x-li=* | --x-l=*) - x_libraries=$ac_optarg ;; - - -*) { $as_echo "$as_me: error: unrecognized option: $ac_option -Try \`$0 --help' for more information." >&2 - { (exit 1); exit 1; }; } - ;; - - *=*) - ac_envvar=`expr "x$ac_option" : 'x\([^=]*\)='` - # Reject names that are not valid shell variable names. - expr "x$ac_envvar" : ".*[^_$as_cr_alnum]" >/dev/null && - { $as_echo "$as_me: error: invalid variable name: $ac_envvar" >&2 - { (exit 1); exit 1; }; } - eval $ac_envvar=\$ac_optarg - export $ac_envvar ;; - - *) - # FIXME: should be removed in autoconf 3.0. - $as_echo "$as_me: WARNING: you should use --build, --host, --target" >&2 - expr "x$ac_option" : ".*[^-._$as_cr_alnum]" >/dev/null && - $as_echo "$as_me: WARNING: invalid host type: $ac_option" >&2 - : ${build_alias=$ac_option} ${host_alias=$ac_option} ${target_alias=$ac_option} - ;; - - esac -done - -if test -n "$ac_prev"; then - ac_option=--`echo $ac_prev | sed 's/_/-/g'` - { $as_echo "$as_me: error: missing argument to $ac_option" >&2 - { (exit 1); exit 1; }; } -fi - -if test -n "$ac_unrecognized_opts"; then - case $enable_option_checking in - no) ;; - fatal) { $as_echo "$as_me: error: unrecognized options: $ac_unrecognized_opts" >&2 - { (exit 1); exit 1; }; } ;; - *) $as_echo "$as_me: WARNING: unrecognized options: $ac_unrecognized_opts" >&2 ;; - esac -fi - -# Check all directory arguments for consistency. -for ac_var in exec_prefix prefix bindir sbindir libexecdir datarootdir \ - datadir sysconfdir sharedstatedir localstatedir includedir \ - oldincludedir docdir infodir htmldir dvidir pdfdir psdir \ - libdir localedir mandir -do - eval ac_val=\$$ac_var - # Remove trailing slashes. - case $ac_val in - */ ) - ac_val=`expr "X$ac_val" : 'X\(.*[^/]\)' \| "X$ac_val" : 'X\(.*\)'` - eval $ac_var=\$ac_val;; - esac - # Be sure to have absolute directory names. - case $ac_val in - [\\/$]* | ?:[\\/]* ) continue;; - NONE | '' ) case $ac_var in *prefix ) continue;; esac;; - esac - { $as_echo "$as_me: error: expected an absolute directory name for --$ac_var: $ac_val" >&2 - { (exit 1); exit 1; }; } -done - -# There might be people who depend on the old broken behavior: `$host' -# used to hold the argument of --host etc. -# FIXME: To remove some day. -build=$build_alias -host=$host_alias -target=$target_alias - -# FIXME: To remove some day. -if test "x$host_alias" != x; then - if test "x$build_alias" = x; then - cross_compiling=maybe - $as_echo "$as_me: WARNING: If you wanted to set the --build type, don't use --host. - If a cross compiler is detected then cross compile mode will be used." >&2 - elif test "x$build_alias" != "x$host_alias"; then - cross_compiling=yes - fi -fi - -ac_tool_prefix= -test -n "$host_alias" && ac_tool_prefix=$host_alias- - -test "$silent" = yes && exec 6>/dev/null - - -ac_pwd=`pwd` && test -n "$ac_pwd" && -ac_ls_di=`ls -di .` && -ac_pwd_ls_di=`cd "$ac_pwd" && ls -di .` || - { $as_echo "$as_me: error: working directory cannot be determined" >&2 - { (exit 1); exit 1; }; } -test "X$ac_ls_di" = "X$ac_pwd_ls_di" || - { $as_echo "$as_me: error: pwd does not report name of working directory" >&2 - { (exit 1); exit 1; }; } - - -# Find the source files, if location was not specified. -if test -z "$srcdir"; then - ac_srcdir_defaulted=yes - # Try the directory containing this script, then the parent directory. - ac_confdir=`$as_dirname -- "$as_myself" || -$as_expr X"$as_myself" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$as_myself" : 'X\(//\)[^/]' \| \ - X"$as_myself" : 'X\(//\)$' \| \ - X"$as_myself" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$as_myself" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - srcdir=$ac_confdir - if test ! -r "$srcdir/$ac_unique_file"; then - srcdir=.. - fi -else - ac_srcdir_defaulted=no -fi -if test ! -r "$srcdir/$ac_unique_file"; then - test "$ac_srcdir_defaulted" = yes && srcdir="$ac_confdir or .." - { $as_echo "$as_me: error: cannot find sources ($ac_unique_file) in $srcdir" >&2 - { (exit 1); exit 1; }; } -fi -ac_msg="sources are in $srcdir, but \`cd $srcdir' does not work" -ac_abs_confdir=`( - cd "$srcdir" && test -r "./$ac_unique_file" || { $as_echo "$as_me: error: $ac_msg" >&2 - { (exit 1); exit 1; }; } - pwd)` -# When building in place, set srcdir=. -if test "$ac_abs_confdir" = "$ac_pwd"; then - srcdir=. -fi -# Remove unnecessary trailing slashes from srcdir. -# Double slashes in file names in object file debugging info -# mess up M-x gdb in Emacs. -case $srcdir in -*/) srcdir=`expr "X$srcdir" : 'X\(.*[^/]\)' \| "X$srcdir" : 'X\(.*\)'`;; -esac -for ac_var in $ac_precious_vars; do - eval ac_env_${ac_var}_set=\${${ac_var}+set} - eval ac_env_${ac_var}_value=\$${ac_var} - eval ac_cv_env_${ac_var}_set=\${${ac_var}+set} - eval ac_cv_env_${ac_var}_value=\$${ac_var} -done - -# -# Report the --help message. -# -if test "$ac_init_help" = "long"; then - # Omit some internal or obsolete options to make the list less imposing. - # This message is too long to be a string in the A/UX 3.1 sh. - cat <<_ACEOF -\`configure' configures ZeroGSogl 0.96.2 to adapt to many kinds of systems. - -Usage: $0 [OPTION]... [VAR=VALUE]... - -To assign environment variables (e.g., CC, CFLAGS...), specify them as -VAR=VALUE. See below for descriptions of some of the useful variables. - -Defaults for the options are specified in brackets. - -Configuration: - -h, --help display this help and exit - --help=short display options specific to this package - --help=recursive display the short help of all the included packages - -V, --version display version information and exit - -q, --quiet, --silent do not print \`checking...' messages - --cache-file=FILE cache test results in FILE [disabled] - -C, --config-cache alias for \`--cache-file=config.cache' - -n, --no-create do not create output files - --srcdir=DIR find the sources in DIR [configure dir or \`..'] - -Installation directories: - --prefix=PREFIX install architecture-independent files in PREFIX - [$ac_default_prefix] - --exec-prefix=EPREFIX install architecture-dependent files in EPREFIX - [PREFIX] - -By default, \`make install' will install all the files in -\`$ac_default_prefix/bin', \`$ac_default_prefix/lib' etc. You can specify -an installation prefix other than \`$ac_default_prefix' using \`--prefix', -for instance \`--prefix=\$HOME'. - -For better control, use the options below. - -Fine tuning of the installation directories: - --bindir=DIR user executables [EPREFIX/bin] - --sbindir=DIR system admin executables [EPREFIX/sbin] - --libexecdir=DIR program executables [EPREFIX/libexec] - --sysconfdir=DIR read-only single-machine data [PREFIX/etc] - --sharedstatedir=DIR modifiable architecture-independent data [PREFIX/com] - --localstatedir=DIR modifiable single-machine data [PREFIX/var] - --libdir=DIR object code libraries [EPREFIX/lib] - --includedir=DIR C header files [PREFIX/include] - --oldincludedir=DIR C header files for non-gcc [/usr/include] - --datarootdir=DIR read-only arch.-independent data root [PREFIX/share] - --datadir=DIR read-only architecture-independent data [DATAROOTDIR] - --infodir=DIR info documentation [DATAROOTDIR/info] - --localedir=DIR locale-dependent data [DATAROOTDIR/locale] - --mandir=DIR man documentation [DATAROOTDIR/man] - --docdir=DIR documentation root [DATAROOTDIR/doc/zerogsogl] - --htmldir=DIR html documentation [DOCDIR] - --dvidir=DIR dvi documentation [DOCDIR] - --pdfdir=DIR pdf documentation [DOCDIR] - --psdir=DIR ps documentation [DOCDIR] -_ACEOF - - cat <<\_ACEOF - -Program names: - --program-prefix=PREFIX prepend PREFIX to installed program names - --program-suffix=SUFFIX append SUFFIX to installed program names - --program-transform-name=PROGRAM run sed PROGRAM on installed program names -_ACEOF -fi - -if test -n "$ac_init_help"; then - case $ac_init_help in - short | recursive ) echo "Configuration of ZeroGSogl 0.96.2:";; - esac - cat <<\_ACEOF - -Optional Features: - --disable-option-checking ignore unrecognized --enable/--with options - --disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no) - --enable-FEATURE[=ARG] include FEATURE [ARG=yes] - --disable-dependency-tracking speeds up one-time build - --enable-dependency-tracking do not reject slow dependency extractors - --enable-debug debug build - --enable-devbuild Special Build for developers that simplifies testing - and adds extra checks - --enable-sse2 Enable sse2 instructions - -Some influential environment variables: - CC C compiler command - CFLAGS C compiler flags - LDFLAGS linker flags, e.g. -L if you have libraries in a - nonstandard directory - LIBS libraries to pass to the linker, e.g. -l - CPPFLAGS C/C++/Objective C preprocessor flags, e.g. -I if - you have headers in a nonstandard directory - CXX C++ compiler command - CXXFLAGS C++ compiler flags - CPP C preprocessor - CCAS assembler compiler command (defaults to CC) - CCASFLAGS assembler compiler flags (defaults to CFLAGS) - -Use these variables to override the choices made by `configure' or to help -it to find libraries and programs with nonstandard names/locations. - -Report bugs to . -_ACEOF -ac_status=$? -fi - -if test "$ac_init_help" = "recursive"; then - # If there are subdirs, report their specific --help. - for ac_dir in : $ac_subdirs_all; do test "x$ac_dir" = x: && continue - test -d "$ac_dir" || - { cd "$srcdir" && ac_pwd=`pwd` && srcdir=. && test -d "$ac_dir"; } || - continue - ac_builddir=. - -case "$ac_dir" in -.) ac_dir_suffix= ac_top_builddir_sub=. ac_top_build_prefix= ;; -*) - ac_dir_suffix=/`$as_echo "$ac_dir" | sed 's|^\.[\\/]||'` - # A ".." for each directory in $ac_dir_suffix. - ac_top_builddir_sub=`$as_echo "$ac_dir_suffix" | sed 's|/[^\\/]*|/..|g;s|/||'` - case $ac_top_builddir_sub in - "") ac_top_builddir_sub=. ac_top_build_prefix= ;; - *) ac_top_build_prefix=$ac_top_builddir_sub/ ;; - esac ;; -esac -ac_abs_top_builddir=$ac_pwd -ac_abs_builddir=$ac_pwd$ac_dir_suffix -# for backward compatibility: -ac_top_builddir=$ac_top_build_prefix - -case $srcdir in - .) # We are building in place. - ac_srcdir=. - ac_top_srcdir=$ac_top_builddir_sub - ac_abs_top_srcdir=$ac_pwd ;; - [\\/]* | ?:[\\/]* ) # Absolute name. - ac_srcdir=$srcdir$ac_dir_suffix; - ac_top_srcdir=$srcdir - ac_abs_top_srcdir=$srcdir ;; - *) # Relative name. - ac_srcdir=$ac_top_build_prefix$srcdir$ac_dir_suffix - ac_top_srcdir=$ac_top_build_prefix$srcdir - ac_abs_top_srcdir=$ac_pwd/$srcdir ;; -esac -ac_abs_srcdir=$ac_abs_top_srcdir$ac_dir_suffix - - cd "$ac_dir" || { ac_status=$?; continue; } - # Check for guested configure. - if test -f "$ac_srcdir/configure.gnu"; then - echo && - $SHELL "$ac_srcdir/configure.gnu" --help=recursive - elif test -f "$ac_srcdir/configure"; then - echo && - $SHELL "$ac_srcdir/configure" --help=recursive - else - $as_echo "$as_me: WARNING: no configuration information is in $ac_dir" >&2 - fi || ac_status=$? - cd "$ac_pwd" || { ac_status=$?; break; } - done -fi - -test -n "$ac_init_help" && exit $ac_status -if $ac_init_version; then - cat <<\_ACEOF -ZeroGSogl configure 0.96.2 -generated by GNU Autoconf 2.63 - -Copyright (C) 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000, 2001, -2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc. -This configure script is free software; the Free Software Foundation -gives unlimited permission to copy, distribute and modify it. -_ACEOF - exit -fi -cat >config.log <<_ACEOF -This file contains any messages produced by compilers while -running configure, to aid debugging if configure makes a mistake. - -It was created by ZeroGSogl $as_me 0.96.2, which was -generated by GNU Autoconf 2.63. Invocation command line was - - $ $0 $@ - -_ACEOF -exec 5>>config.log -{ -cat <<_ASUNAME -## --------- ## -## Platform. ## -## --------- ## - -hostname = `(hostname || uname -n) 2>/dev/null | sed 1q` -uname -m = `(uname -m) 2>/dev/null || echo unknown` -uname -r = `(uname -r) 2>/dev/null || echo unknown` -uname -s = `(uname -s) 2>/dev/null || echo unknown` -uname -v = `(uname -v) 2>/dev/null || echo unknown` - -/usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null || echo unknown` -/bin/uname -X = `(/bin/uname -X) 2>/dev/null || echo unknown` - -/bin/arch = `(/bin/arch) 2>/dev/null || echo unknown` -/usr/bin/arch -k = `(/usr/bin/arch -k) 2>/dev/null || echo unknown` -/usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null || echo unknown` -/usr/bin/hostinfo = `(/usr/bin/hostinfo) 2>/dev/null || echo unknown` -/bin/machine = `(/bin/machine) 2>/dev/null || echo unknown` -/usr/bin/oslevel = `(/usr/bin/oslevel) 2>/dev/null || echo unknown` -/bin/universe = `(/bin/universe) 2>/dev/null || echo unknown` - -_ASUNAME - -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - $as_echo "PATH: $as_dir" -done -IFS=$as_save_IFS - -} >&5 - -cat >&5 <<_ACEOF - - -## ----------- ## -## Core tests. ## -## ----------- ## - -_ACEOF - - -# Keep a trace of the command line. -# Strip out --no-create and --no-recursion so they do not pile up. -# Strip out --silent because we don't want to record it for future runs. -# Also quote any args containing shell meta-characters. -# Make two passes to allow for proper duplicate-argument suppression. -ac_configure_args= -ac_configure_args0= -ac_configure_args1= -ac_must_keep_next=false -for ac_pass in 1 2 -do - for ac_arg - do - case $ac_arg in - -no-create | --no-c* | -n | -no-recursion | --no-r*) continue ;; - -q | -quiet | --quiet | --quie | --qui | --qu | --q \ - | -silent | --silent | --silen | --sile | --sil) - continue ;; - *\'*) - ac_arg=`$as_echo "$ac_arg" | sed "s/'/'\\\\\\\\''/g"` ;; - esac - case $ac_pass in - 1) ac_configure_args0="$ac_configure_args0 '$ac_arg'" ;; - 2) - ac_configure_args1="$ac_configure_args1 '$ac_arg'" - if test $ac_must_keep_next = true; then - ac_must_keep_next=false # Got value, back to normal. - else - case $ac_arg in - *=* | --config-cache | -C | -disable-* | --disable-* \ - | -enable-* | --enable-* | -gas | --g* | -nfp | --nf* \ - | -q | -quiet | --q* | -silent | --sil* | -v | -verb* \ - | -with-* | --with-* | -without-* | --without-* | --x) - case "$ac_configure_args0 " in - "$ac_configure_args1"*" '$ac_arg' "* ) continue ;; - esac - ;; - -* ) ac_must_keep_next=true ;; - esac - fi - ac_configure_args="$ac_configure_args '$ac_arg'" - ;; - esac - done -done -$as_unset ac_configure_args0 || test "${ac_configure_args0+set}" != set || { ac_configure_args0=; export ac_configure_args0; } -$as_unset ac_configure_args1 || test "${ac_configure_args1+set}" != set || { ac_configure_args1=; export ac_configure_args1; } - -# When interrupted or exit'd, cleanup temporary files, and complete -# config.log. We remove comments because anyway the quotes in there -# would cause problems or look ugly. -# WARNING: Use '\'' to represent an apostrophe within the trap. -# WARNING: Do not start the trap code with a newline, due to a FreeBSD 4.0 bug. -trap 'exit_status=$? - # Save into config.log some information that might help in debugging. - { - echo - - cat <<\_ASBOX -## ---------------- ## -## Cache variables. ## -## ---------------- ## -_ASBOX - echo - # The following way of writing the cache mishandles newlines in values, -( - for ac_var in `(set) 2>&1 | sed -n '\''s/^\([a-zA-Z_][a-zA-Z0-9_]*\)=.*/\1/p'\''`; do - eval ac_val=\$$ac_var - case $ac_val in #( - *${as_nl}*) - case $ac_var in #( - *_cv_*) { $as_echo "$as_me:$LINENO: WARNING: cache variable $ac_var contains a newline" >&5 -$as_echo "$as_me: WARNING: cache variable $ac_var contains a newline" >&2;} ;; - esac - case $ac_var in #( - _ | IFS | as_nl) ;; #( - BASH_ARGV | BASH_SOURCE) eval $ac_var= ;; #( - *) $as_unset $ac_var ;; - esac ;; - esac - done - (set) 2>&1 | - case $as_nl`(ac_space='\'' '\''; set) 2>&1` in #( - *${as_nl}ac_space=\ *) - sed -n \ - "s/'\''/'\''\\\\'\'''\''/g; - s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='\''\\2'\''/p" - ;; #( - *) - sed -n "/^[_$as_cr_alnum]*_cv_[_$as_cr_alnum]*=/p" - ;; - esac | - sort -) - echo - - cat <<\_ASBOX -## ----------------- ## -## Output variables. ## -## ----------------- ## -_ASBOX - echo - for ac_var in $ac_subst_vars - do - eval ac_val=\$$ac_var - case $ac_val in - *\'\''*) ac_val=`$as_echo "$ac_val" | sed "s/'\''/'\''\\\\\\\\'\'''\''/g"`;; - esac - $as_echo "$ac_var='\''$ac_val'\''" - done | sort - echo - - if test -n "$ac_subst_files"; then - cat <<\_ASBOX -## ------------------- ## -## File substitutions. ## -## ------------------- ## -_ASBOX - echo - for ac_var in $ac_subst_files - do - eval ac_val=\$$ac_var - case $ac_val in - *\'\''*) ac_val=`$as_echo "$ac_val" | sed "s/'\''/'\''\\\\\\\\'\'''\''/g"`;; - esac - $as_echo "$ac_var='\''$ac_val'\''" - done | sort - echo - fi - - if test -s confdefs.h; then - cat <<\_ASBOX -## ----------- ## -## confdefs.h. ## -## ----------- ## -_ASBOX - echo - cat confdefs.h - echo - fi - test "$ac_signal" != 0 && - $as_echo "$as_me: caught signal $ac_signal" - $as_echo "$as_me: exit $exit_status" - } >&5 - rm -f core *.core core.conftest.* && - rm -f -r conftest* confdefs* conf$$* $ac_clean_files && - exit $exit_status -' 0 -for ac_signal in 1 2 13 15; do - trap 'ac_signal='$ac_signal'; { (exit 1); exit 1; }' $ac_signal -done -ac_signal=0 - -# confdefs.h avoids OS command line length limits that DEFS can exceed. -rm -f -r conftest* confdefs.h - -# Predefined preprocessor variables. - -cat >>confdefs.h <<_ACEOF -#define PACKAGE_NAME "$PACKAGE_NAME" -_ACEOF - - -cat >>confdefs.h <<_ACEOF -#define PACKAGE_TARNAME "$PACKAGE_TARNAME" -_ACEOF - - -cat >>confdefs.h <<_ACEOF -#define PACKAGE_VERSION "$PACKAGE_VERSION" -_ACEOF - - -cat >>confdefs.h <<_ACEOF -#define PACKAGE_STRING "$PACKAGE_STRING" -_ACEOF - - -cat >>confdefs.h <<_ACEOF -#define PACKAGE_BUGREPORT "$PACKAGE_BUGREPORT" -_ACEOF - - -# Let the site file select an alternate cache file if it wants to. -# Prefer an explicitly selected file to automatically selected ones. -ac_site_file1=NONE -ac_site_file2=NONE -if test -n "$CONFIG_SITE"; then - ac_site_file1=$CONFIG_SITE -elif test "x$prefix" != xNONE; then - ac_site_file1=$prefix/share/config.site - ac_site_file2=$prefix/etc/config.site -else - ac_site_file1=$ac_default_prefix/share/config.site - ac_site_file2=$ac_default_prefix/etc/config.site -fi -for ac_site_file in "$ac_site_file1" "$ac_site_file2" -do - test "x$ac_site_file" = xNONE && continue - if test -r "$ac_site_file"; then - { $as_echo "$as_me:$LINENO: loading site script $ac_site_file" >&5 -$as_echo "$as_me: loading site script $ac_site_file" >&6;} - sed 's/^/| /' "$ac_site_file" >&5 - . "$ac_site_file" - fi -done - -if test -r "$cache_file"; then - # Some versions of bash will fail to source /dev/null (special - # files actually), so we avoid doing that. - if test -f "$cache_file"; then - { $as_echo "$as_me:$LINENO: loading cache $cache_file" >&5 -$as_echo "$as_me: loading cache $cache_file" >&6;} - case $cache_file in - [\\/]* | ?:[\\/]* ) . "$cache_file";; - *) . "./$cache_file";; - esac - fi -else - { $as_echo "$as_me:$LINENO: creating cache $cache_file" >&5 -$as_echo "$as_me: creating cache $cache_file" >&6;} - >$cache_file -fi - -# Check that the precious variables saved in the cache have kept the same -# value. -ac_cache_corrupted=false -for ac_var in $ac_precious_vars; do - eval ac_old_set=\$ac_cv_env_${ac_var}_set - eval ac_new_set=\$ac_env_${ac_var}_set - eval ac_old_val=\$ac_cv_env_${ac_var}_value - eval ac_new_val=\$ac_env_${ac_var}_value - case $ac_old_set,$ac_new_set in - set,) - { $as_echo "$as_me:$LINENO: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&5 -$as_echo "$as_me: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&2;} - ac_cache_corrupted=: ;; - ,set) - { $as_echo "$as_me:$LINENO: error: \`$ac_var' was not set in the previous run" >&5 -$as_echo "$as_me: error: \`$ac_var' was not set in the previous run" >&2;} - ac_cache_corrupted=: ;; - ,);; - *) - if test "x$ac_old_val" != "x$ac_new_val"; then - # differences in whitespace do not lead to failure. - ac_old_val_w=`echo x $ac_old_val` - ac_new_val_w=`echo x $ac_new_val` - if test "$ac_old_val_w" != "$ac_new_val_w"; then - { $as_echo "$as_me:$LINENO: error: \`$ac_var' has changed since the previous run:" >&5 -$as_echo "$as_me: error: \`$ac_var' has changed since the previous run:" >&2;} - ac_cache_corrupted=: - else - { $as_echo "$as_me:$LINENO: warning: ignoring whitespace changes in \`$ac_var' since the previous run:" >&5 -$as_echo "$as_me: warning: ignoring whitespace changes in \`$ac_var' since the previous run:" >&2;} - eval $ac_var=\$ac_old_val - fi - { $as_echo "$as_me:$LINENO: former value: \`$ac_old_val'" >&5 -$as_echo "$as_me: former value: \`$ac_old_val'" >&2;} - { $as_echo "$as_me:$LINENO: current value: \`$ac_new_val'" >&5 -$as_echo "$as_me: current value: \`$ac_new_val'" >&2;} - fi;; - esac - # Pass precious variables to config.status. - if test "$ac_new_set" = set; then - case $ac_new_val in - *\'*) ac_arg=$ac_var=`$as_echo "$ac_new_val" | sed "s/'/'\\\\\\\\''/g"` ;; - *) ac_arg=$ac_var=$ac_new_val ;; - esac - case " $ac_configure_args " in - *" '$ac_arg' "*) ;; # Avoid dups. Use of quotes ensures accuracy. - *) ac_configure_args="$ac_configure_args '$ac_arg'" ;; - esac - fi -done -if $ac_cache_corrupted; then - { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} - { $as_echo "$as_me:$LINENO: error: changes in the environment can compromise the build" >&5 -$as_echo "$as_me: error: changes in the environment can compromise the build" >&2;} - { { $as_echo "$as_me:$LINENO: error: run \`make distclean' and/or \`rm $cache_file' and start over" >&5 -$as_echo "$as_me: error: run \`make distclean' and/or \`rm $cache_file' and start over" >&2;} - { (exit 1); exit 1; }; } -fi - - - - - - - - - - - - - - - - - - - - - - - - - -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu - - - -am__api_version='1.11' - -ac_aux_dir= -for ac_dir in "$srcdir" "$srcdir/.." "$srcdir/../.."; do - if test -f "$ac_dir/install-sh"; then - ac_aux_dir=$ac_dir - ac_install_sh="$ac_aux_dir/install-sh -c" - break - elif test -f "$ac_dir/install.sh"; then - ac_aux_dir=$ac_dir - ac_install_sh="$ac_aux_dir/install.sh -c" - break - elif test -f "$ac_dir/shtool"; then - ac_aux_dir=$ac_dir - ac_install_sh="$ac_aux_dir/shtool install -c" - break - fi -done -if test -z "$ac_aux_dir"; then - { { $as_echo "$as_me:$LINENO: error: cannot find install-sh or install.sh in \"$srcdir\" \"$srcdir/..\" \"$srcdir/../..\"" >&5 -$as_echo "$as_me: error: cannot find install-sh or install.sh in \"$srcdir\" \"$srcdir/..\" \"$srcdir/../..\"" >&2;} - { (exit 1); exit 1; }; } -fi - -# These three variables are undocumented and unsupported, -# and are intended to be withdrawn in a future Autoconf release. -# They can cause serious problems if a builder's source tree is in a directory -# whose full name contains unusual characters. -ac_config_guess="$SHELL $ac_aux_dir/config.guess" # Please don't use this var. -ac_config_sub="$SHELL $ac_aux_dir/config.sub" # Please don't use this var. -ac_configure="$SHELL $ac_aux_dir/configure" # Please don't use this var. - - -# Find a good install program. We prefer a C program (faster), -# so one script is as good as another. But avoid the broken or -# incompatible versions: -# SysV /etc/install, /usr/sbin/install -# SunOS /usr/etc/install -# IRIX /sbin/install -# AIX /bin/install -# AmigaOS /C/install, which installs bootblocks on floppy discs -# AIX 4 /usr/bin/installbsd, which doesn't work without a -g flag -# AFS /usr/afsws/bin/install, which mishandles nonexistent args -# SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff" -# OS/2's system install, which has a completely different semantic -# ./install, which can be erroneously created by make from ./install.sh. -# Reject install programs that cannot install multiple files. -{ $as_echo "$as_me:$LINENO: checking for a BSD-compatible install" >&5 -$as_echo_n "checking for a BSD-compatible install... " >&6; } -if test -z "$INSTALL"; then -if test "${ac_cv_path_install+set}" = set; then - $as_echo_n "(cached) " >&6 -else - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - # Account for people who put trailing slashes in PATH elements. -case $as_dir/ in - ./ | .// | /cC/* | \ - /etc/* | /usr/sbin/* | /usr/etc/* | /sbin/* | /usr/afsws/bin/* | \ - ?:\\/os2\\/install\\/* | ?:\\/OS2\\/INSTALL\\/* | \ - /usr/ucb/* ) ;; - *) - # OSF1 and SCO ODT 3.0 have their own names for install. - # Don't use installbsd from OSF since it installs stuff as root - # by default. - for ac_prog in ginstall scoinst install; do - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_prog$ac_exec_ext" && $as_test_x "$as_dir/$ac_prog$ac_exec_ext"; }; then - if test $ac_prog = install && - grep dspmsg "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then - # AIX install. It has an incompatible calling convention. - : - elif test $ac_prog = install && - grep pwplus "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then - # program-specific install script used by HP pwplus--don't use. - : - else - rm -rf conftest.one conftest.two conftest.dir - echo one > conftest.one - echo two > conftest.two - mkdir conftest.dir - if "$as_dir/$ac_prog$ac_exec_ext" -c conftest.one conftest.two "`pwd`/conftest.dir" && - test -s conftest.one && test -s conftest.two && - test -s conftest.dir/conftest.one && - test -s conftest.dir/conftest.two - then - ac_cv_path_install="$as_dir/$ac_prog$ac_exec_ext -c" - break 3 - fi - fi - fi - done - done - ;; -esac - -done -IFS=$as_save_IFS - -rm -rf conftest.one conftest.two conftest.dir - -fi - if test "${ac_cv_path_install+set}" = set; then - INSTALL=$ac_cv_path_install - else - # As a last resort, use the slow shell script. Don't cache a - # value for INSTALL within a source directory, because that will - # break other packages using the cache if that directory is - # removed, or if the value is a relative name. - INSTALL=$ac_install_sh - fi -fi -{ $as_echo "$as_me:$LINENO: result: $INSTALL" >&5 -$as_echo "$INSTALL" >&6; } - -# Use test -z because SunOS4 sh mishandles braces in ${var-val}. -# It thinks the first close brace ends the variable substitution. -test -z "$INSTALL_PROGRAM" && INSTALL_PROGRAM='${INSTALL}' - -test -z "$INSTALL_SCRIPT" && INSTALL_SCRIPT='${INSTALL}' - -test -z "$INSTALL_DATA" && INSTALL_DATA='${INSTALL} -m 644' - -{ $as_echo "$as_me:$LINENO: checking whether build environment is sane" >&5 -$as_echo_n "checking whether build environment is sane... " >&6; } -# Just in case -sleep 1 -echo timestamp > conftest.file -# Reject unsafe characters in $srcdir or the absolute working directory -# name. Accept space and tab only in the latter. -am_lf=' -' -case `pwd` in - *[\\\"\#\$\&\'\`$am_lf]*) - { { $as_echo "$as_me:$LINENO: error: unsafe absolute working directory name" >&5 -$as_echo "$as_me: error: unsafe absolute working directory name" >&2;} - { (exit 1); exit 1; }; };; -esac -case $srcdir in - *[\\\"\#\$\&\'\`$am_lf\ \ ]*) - { { $as_echo "$as_me:$LINENO: error: unsafe srcdir value: \`$srcdir'" >&5 -$as_echo "$as_me: error: unsafe srcdir value: \`$srcdir'" >&2;} - { (exit 1); exit 1; }; };; -esac - -# Do `set' in a subshell so we don't clobber the current shell's -# arguments. Must try -L first in case configure is actually a -# symlink; some systems play weird games with the mod time of symlinks -# (eg FreeBSD returns the mod time of the symlink's containing -# directory). -if ( - set X `ls -Lt "$srcdir/configure" conftest.file 2> /dev/null` - if test "$*" = "X"; then - # -L didn't work. - set X `ls -t "$srcdir/configure" conftest.file` - fi - rm -f conftest.file - if test "$*" != "X $srcdir/configure conftest.file" \ - && test "$*" != "X conftest.file $srcdir/configure"; then - - # If neither matched, then we have a broken ls. This can happen - # if, for instance, CONFIG_SHELL is bash and it inherits a - # broken ls alias from the environment. This has actually - # happened. Such a system could not be considered "sane". - { { $as_echo "$as_me:$LINENO: error: ls -t appears to fail. Make sure there is not a broken -alias in your environment" >&5 -$as_echo "$as_me: error: ls -t appears to fail. Make sure there is not a broken -alias in your environment" >&2;} - { (exit 1); exit 1; }; } - fi - - test "$2" = conftest.file - ) -then - # Ok. - : -else - { { $as_echo "$as_me:$LINENO: error: newly created file is older than distributed files! -Check your system clock" >&5 -$as_echo "$as_me: error: newly created file is older than distributed files! -Check your system clock" >&2;} - { (exit 1); exit 1; }; } -fi -{ $as_echo "$as_me:$LINENO: result: yes" >&5 -$as_echo "yes" >&6; } -test "$program_prefix" != NONE && - program_transform_name="s&^&$program_prefix&;$program_transform_name" -# Use a double $ so make ignores it. -test "$program_suffix" != NONE && - program_transform_name="s&\$&$program_suffix&;$program_transform_name" -# Double any \ or $. -# By default was `s,x,x', remove it if useless. -ac_script='s/[\\$]/&&/g;s/;s,x,x,$//' -program_transform_name=`$as_echo "$program_transform_name" | sed "$ac_script"` - -# expand $ac_aux_dir to an absolute path -am_aux_dir=`cd $ac_aux_dir && pwd` - -if test x"${MISSING+set}" != xset; then - case $am_aux_dir in - *\ * | *\ *) - MISSING="\${SHELL} \"$am_aux_dir/missing\"" ;; - *) - MISSING="\${SHELL} $am_aux_dir/missing" ;; - esac -fi -# Use eval to expand $SHELL -if eval "$MISSING --run true"; then - am_missing_run="$MISSING --run " -else - am_missing_run= - { $as_echo "$as_me:$LINENO: WARNING: \`missing' script is too old or missing" >&5 -$as_echo "$as_me: WARNING: \`missing' script is too old or missing" >&2;} -fi - -if test x"${install_sh}" != xset; then - case $am_aux_dir in - *\ * | *\ *) - install_sh="\${SHELL} '$am_aux_dir/install-sh'" ;; - *) - install_sh="\${SHELL} $am_aux_dir/install-sh" - esac -fi - -# Installed binaries are usually stripped using `strip' when the user -# run `make install-strip'. However `strip' might not be the right -# tool to use in cross-compilation environments, therefore Automake -# will honor the `STRIP' environment variable to overrule this program. -if test "$cross_compiling" != no; then - if test -n "$ac_tool_prefix"; then - # Extract the first word of "${ac_tool_prefix}strip", so it can be a program name with args. -set dummy ${ac_tool_prefix}strip; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_STRIP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$STRIP"; then - ac_cv_prog_STRIP="$STRIP" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_STRIP="${ac_tool_prefix}strip" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -STRIP=$ac_cv_prog_STRIP -if test -n "$STRIP"; then - { $as_echo "$as_me:$LINENO: result: $STRIP" >&5 -$as_echo "$STRIP" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - -fi -if test -z "$ac_cv_prog_STRIP"; then - ac_ct_STRIP=$STRIP - # Extract the first word of "strip", so it can be a program name with args. -set dummy strip; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_ac_ct_STRIP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$ac_ct_STRIP"; then - ac_cv_prog_ac_ct_STRIP="$ac_ct_STRIP" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_ac_ct_STRIP="strip" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -ac_ct_STRIP=$ac_cv_prog_ac_ct_STRIP -if test -n "$ac_ct_STRIP"; then - { $as_echo "$as_me:$LINENO: result: $ac_ct_STRIP" >&5 -$as_echo "$ac_ct_STRIP" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - if test "x$ac_ct_STRIP" = x; then - STRIP=":" - else - case $cross_compiling:$ac_tool_warned in -yes:) -{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 -$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} -ac_tool_warned=yes ;; -esac - STRIP=$ac_ct_STRIP - fi -else - STRIP="$ac_cv_prog_STRIP" -fi - -fi -INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s" - -{ $as_echo "$as_me:$LINENO: checking for a thread-safe mkdir -p" >&5 -$as_echo_n "checking for a thread-safe mkdir -p... " >&6; } -if test -z "$MKDIR_P"; then - if test "${ac_cv_path_mkdir+set}" = set; then - $as_echo_n "(cached) " >&6 -else - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH$PATH_SEPARATOR/opt/sfw/bin -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_prog in mkdir gmkdir; do - for ac_exec_ext in '' $ac_executable_extensions; do - { test -f "$as_dir/$ac_prog$ac_exec_ext" && $as_test_x "$as_dir/$ac_prog$ac_exec_ext"; } || continue - case `"$as_dir/$ac_prog$ac_exec_ext" --version 2>&1` in #( - 'mkdir (GNU coreutils) '* | \ - 'mkdir (coreutils) '* | \ - 'mkdir (fileutils) '4.1*) - ac_cv_path_mkdir=$as_dir/$ac_prog$ac_exec_ext - break 3;; - esac - done - done -done -IFS=$as_save_IFS - -fi - - if test "${ac_cv_path_mkdir+set}" = set; then - MKDIR_P="$ac_cv_path_mkdir -p" - else - # As a last resort, use the slow shell script. Don't cache a - # value for MKDIR_P within a source directory, because that will - # break other packages using the cache if that directory is - # removed, or if the value is a relative name. - test -d ./--version && rmdir ./--version - MKDIR_P="$ac_install_sh -d" - fi -fi -{ $as_echo "$as_me:$LINENO: result: $MKDIR_P" >&5 -$as_echo "$MKDIR_P" >&6; } - -mkdir_p="$MKDIR_P" -case $mkdir_p in - [\\/$]* | ?:[\\/]*) ;; - */*) mkdir_p="\$(top_builddir)/$mkdir_p" ;; -esac - -for ac_prog in gawk mawk nawk awk -do - # Extract the first word of "$ac_prog", so it can be a program name with args. -set dummy $ac_prog; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_AWK+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$AWK"; then - ac_cv_prog_AWK="$AWK" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_AWK="$ac_prog" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -AWK=$ac_cv_prog_AWK -if test -n "$AWK"; then - { $as_echo "$as_me:$LINENO: result: $AWK" >&5 -$as_echo "$AWK" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - - test -n "$AWK" && break -done - -{ $as_echo "$as_me:$LINENO: checking whether ${MAKE-make} sets \$(MAKE)" >&5 -$as_echo_n "checking whether ${MAKE-make} sets \$(MAKE)... " >&6; } -set x ${MAKE-make} -ac_make=`$as_echo "$2" | sed 's/+/p/g; s/[^a-zA-Z0-9_]/_/g'` -if { as_var=ac_cv_prog_make_${ac_make}_set; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.make <<\_ACEOF -SHELL = /bin/sh -all: - @echo '@@@%%%=$(MAKE)=@@@%%%' -_ACEOF -# GNU make sometimes prints "make[1]: Entering...", which would confuse us. -case `${MAKE-make} -f conftest.make 2>/dev/null` in - *@@@%%%=?*=@@@%%%*) - eval ac_cv_prog_make_${ac_make}_set=yes;; - *) - eval ac_cv_prog_make_${ac_make}_set=no;; -esac -rm -f conftest.make -fi -if eval test \$ac_cv_prog_make_${ac_make}_set = yes; then - { $as_echo "$as_me:$LINENO: result: yes" >&5 -$as_echo "yes" >&6; } - SET_MAKE= -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } - SET_MAKE="MAKE=${MAKE-make}" -fi - -rm -rf .tst 2>/dev/null -mkdir .tst 2>/dev/null -if test -d .tst; then - am__leading_dot=. -else - am__leading_dot=_ -fi -rmdir .tst 2>/dev/null - -if test "`cd $srcdir && pwd`" != "`pwd`"; then - # Use -I$(srcdir) only when $(srcdir) != ., so that make's output - # is not polluted with repeated "-I." - am__isrc=' -I$(srcdir)' - # test to see if srcdir already configured - if test -f $srcdir/config.status; then - { { $as_echo "$as_me:$LINENO: error: source directory already configured; run \"make distclean\" there first" >&5 -$as_echo "$as_me: error: source directory already configured; run \"make distclean\" there first" >&2;} - { (exit 1); exit 1; }; } - fi -fi - -# test whether we have cygpath -if test -z "$CYGPATH_W"; then - if (cygpath --version) >/dev/null 2>/dev/null; then - CYGPATH_W='cygpath -w' - else - CYGPATH_W=echo - fi -fi - - -# Define the identity of the package. - PACKAGE=ZeroGSogl - VERSION=0.96.2 - - -cat >>confdefs.h <<_ACEOF -#define PACKAGE "$PACKAGE" -_ACEOF - - -cat >>confdefs.h <<_ACEOF -#define VERSION "$VERSION" -_ACEOF - -# Some tools Automake needs. - -ACLOCAL=${ACLOCAL-"${am_missing_run}aclocal-${am__api_version}"} - - -AUTOCONF=${AUTOCONF-"${am_missing_run}autoconf"} - - -AUTOMAKE=${AUTOMAKE-"${am_missing_run}automake-${am__api_version}"} - - -AUTOHEADER=${AUTOHEADER-"${am_missing_run}autoheader"} - - -MAKEINFO=${MAKEINFO-"${am_missing_run}makeinfo"} - -# We need awk for the "check" target. The system "awk" is bad on -# some platforms. -# Always define AMTAR for backward compatibility. - -AMTAR=${AMTAR-"${am_missing_run}tar"} - -am__tar='${AMTAR} chof - "$$tardir"'; am__untar='${AMTAR} xf -' - - - - - - -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu -if test -n "$ac_tool_prefix"; then - for ac_prog in gcc g++ cl KCC CC cxx cc++ xlC aCC c++ - do - # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. -set dummy $ac_tool_prefix$ac_prog; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_CC+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$CC"; then - ac_cv_prog_CC="$CC" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_CC="$ac_tool_prefix$ac_prog" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -CC=$ac_cv_prog_CC -if test -n "$CC"; then - { $as_echo "$as_me:$LINENO: result: $CC" >&5 -$as_echo "$CC" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - - test -n "$CC" && break - done -fi -if test -z "$CC"; then - ac_ct_CC=$CC - for ac_prog in gcc g++ cl KCC CC cxx cc++ xlC aCC c++ -do - # Extract the first word of "$ac_prog", so it can be a program name with args. -set dummy $ac_prog; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_ac_ct_CC+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$ac_ct_CC"; then - ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_ac_ct_CC="$ac_prog" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -ac_ct_CC=$ac_cv_prog_ac_ct_CC -if test -n "$ac_ct_CC"; then - { $as_echo "$as_me:$LINENO: result: $ac_ct_CC" >&5 -$as_echo "$ac_ct_CC" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - - test -n "$ac_ct_CC" && break -done - - if test "x$ac_ct_CC" = x; then - CC="" - else - case $cross_compiling:$ac_tool_warned in -yes:) -{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 -$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} -ac_tool_warned=yes ;; -esac - CC=$ac_ct_CC - fi -fi - - -test -z "$CC" && { { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: no acceptable C compiler found in \$PATH -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: no acceptable C compiler found in \$PATH -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } - -# Provide some information about the compiler. -$as_echo "$as_me:$LINENO: checking for C compiler version" >&5 -set X $ac_compile -ac_compiler=$2 -{ (ac_try="$ac_compiler --version >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler --version >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } -{ (ac_try="$ac_compiler -v >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler -v >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } -{ (ac_try="$ac_compiler -V >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler -V >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } - -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -ac_clean_files_save=$ac_clean_files -ac_clean_files="$ac_clean_files a.out a.out.dSYM a.exe b.out" -# Try to create an executable without -o first, disregard a.out. -# It will help us diagnose broken compilers, and finding out an intuition -# of exeext. -{ $as_echo "$as_me:$LINENO: checking for C compiler default output file name" >&5 -$as_echo_n "checking for C compiler default output file name... " >&6; } -ac_link_default=`$as_echo "$ac_link" | sed 's/ -o *conftest[^ ]*//'` - -# The possible output files: -ac_files="a.out conftest.exe conftest a.exe a_out.exe b.out conftest.*" - -ac_rmfiles= -for ac_file in $ac_files -do - case $ac_file in - *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) ;; - * ) ac_rmfiles="$ac_rmfiles $ac_file";; - esac -done -rm -f $ac_rmfiles - -if { (ac_try="$ac_link_default" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link_default") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; then - # Autoconf-2.13 could set the ac_cv_exeext variable to `no'. -# So ignore a value of `no', otherwise this would lead to `EXEEXT = no' -# in a Makefile. We should not override ac_cv_exeext if it was cached, -# so that the user can short-circuit this test for compilers unknown to -# Autoconf. -for ac_file in $ac_files '' -do - test -f "$ac_file" || continue - case $ac_file in - *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) - ;; - [ab].out ) - # We found the default executable, but exeext='' is most - # certainly right. - break;; - *.* ) - if test "${ac_cv_exeext+set}" = set && test "$ac_cv_exeext" != no; - then :; else - ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'` - fi - # We set ac_cv_exeext here because the later test for it is not - # safe: cross compilers may not add the suffix if given an `-o' - # argument, so we may need to know it at that point already. - # Even if this section looks crufty: it has the advantage of - # actually working. - break;; - * ) - break;; - esac -done -test "$ac_cv_exeext" = no && ac_cv_exeext= - -else - ac_file='' -fi - -{ $as_echo "$as_me:$LINENO: result: $ac_file" >&5 -$as_echo "$ac_file" >&6; } -if test -z "$ac_file"; then - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - -{ { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: C compiler cannot create executables -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: C compiler cannot create executables -See \`config.log' for more details." >&2;} - { (exit 77); exit 77; }; }; } -fi - -ac_exeext=$ac_cv_exeext - -# Check that the compiler produces executables we can run. If not, either -# the compiler is broken, or we cross compile. -{ $as_echo "$as_me:$LINENO: checking whether the C compiler works" >&5 -$as_echo_n "checking whether the C compiler works... " >&6; } -# FIXME: These cross compiler hacks should be removed for Autoconf 3.0 -# If not cross compiling, check that we can run a simple program. -if test "$cross_compiling" != yes; then - if { ac_try='./$ac_file' - { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; }; then - cross_compiling=no - else - if test "$cross_compiling" = maybe; then - cross_compiling=yes - else - { { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: cannot run C compiled programs. -If you meant to cross compile, use \`--host'. -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: cannot run C compiled programs. -If you meant to cross compile, use \`--host'. -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } - fi - fi -fi -{ $as_echo "$as_me:$LINENO: result: yes" >&5 -$as_echo "yes" >&6; } - -rm -f -r a.out a.out.dSYM a.exe conftest$ac_cv_exeext b.out -ac_clean_files=$ac_clean_files_save -# Check that the compiler produces executables we can run. If not, either -# the compiler is broken, or we cross compile. -{ $as_echo "$as_me:$LINENO: checking whether we are cross compiling" >&5 -$as_echo_n "checking whether we are cross compiling... " >&6; } -{ $as_echo "$as_me:$LINENO: result: $cross_compiling" >&5 -$as_echo "$cross_compiling" >&6; } - -{ $as_echo "$as_me:$LINENO: checking for suffix of executables" >&5 -$as_echo_n "checking for suffix of executables... " >&6; } -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; then - # If both `conftest.exe' and `conftest' are `present' (well, observable) -# catch `conftest.exe'. For instance with Cygwin, `ls conftest' will -# work properly (i.e., refer to `conftest.exe'), while it won't with -# `rm'. -for ac_file in conftest.exe conftest conftest.*; do - test -f "$ac_file" || continue - case $ac_file in - *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) ;; - *.* ) ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'` - break;; - * ) break;; - esac -done -else - { { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: cannot compute suffix of executables: cannot compile and link -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: cannot compute suffix of executables: cannot compile and link -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } -fi - -rm -f conftest$ac_cv_exeext -{ $as_echo "$as_me:$LINENO: result: $ac_cv_exeext" >&5 -$as_echo "$ac_cv_exeext" >&6; } - -rm -f conftest.$ac_ext -EXEEXT=$ac_cv_exeext -ac_exeext=$EXEEXT -{ $as_echo "$as_me:$LINENO: checking for suffix of object files" >&5 -$as_echo_n "checking for suffix of object files... " >&6; } -if test "${ac_cv_objext+set}" = set; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.o conftest.obj -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; then - for ac_file in conftest.o conftest.obj conftest.*; do - test -f "$ac_file" || continue; - case $ac_file in - *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM ) ;; - *) ac_cv_objext=`expr "$ac_file" : '.*\.\(.*\)'` - break;; - esac -done -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - -{ { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: cannot compute suffix of object files: cannot compile -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: cannot compute suffix of object files: cannot compile -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } -fi - -rm -f conftest.$ac_cv_objext conftest.$ac_ext -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_objext" >&5 -$as_echo "$ac_cv_objext" >&6; } -OBJEXT=$ac_cv_objext -ac_objext=$OBJEXT -{ $as_echo "$as_me:$LINENO: checking whether we are using the GNU C compiler" >&5 -$as_echo_n "checking whether we are using the GNU C compiler... " >&6; } -if test "${ac_cv_c_compiler_gnu+set}" = set; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ -#ifndef __GNUC__ - choke me -#endif - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_compiler_gnu=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_compiler_gnu=no -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -ac_cv_c_compiler_gnu=$ac_compiler_gnu - -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_c_compiler_gnu" >&5 -$as_echo "$ac_cv_c_compiler_gnu" >&6; } -if test $ac_compiler_gnu = yes; then - GCC=yes -else - GCC= -fi -ac_test_CFLAGS=${CFLAGS+set} -ac_save_CFLAGS=$CFLAGS -{ $as_echo "$as_me:$LINENO: checking whether $CC accepts -g" >&5 -$as_echo_n "checking whether $CC accepts -g... " >&6; } -if test "${ac_cv_prog_cc_g+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_save_c_werror_flag=$ac_c_werror_flag - ac_c_werror_flag=yes - ac_cv_prog_cc_g=no - CFLAGS="-g" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_prog_cc_g=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - CFLAGS="" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - : -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_c_werror_flag=$ac_save_c_werror_flag - CFLAGS="-g" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_prog_cc_g=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext - ac_c_werror_flag=$ac_save_c_werror_flag -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_prog_cc_g" >&5 -$as_echo "$ac_cv_prog_cc_g" >&6; } -if test "$ac_test_CFLAGS" = set; then - CFLAGS=$ac_save_CFLAGS -elif test $ac_cv_prog_cc_g = yes; then - if test "$GCC" = yes; then - CFLAGS="-g -O2" - else - CFLAGS="-g" - fi -else - if test "$GCC" = yes; then - CFLAGS="-O2" - else - CFLAGS= - fi -fi -{ $as_echo "$as_me:$LINENO: checking for $CC option to accept ISO C89" >&5 -$as_echo_n "checking for $CC option to accept ISO C89... " >&6; } -if test "${ac_cv_prog_cc_c89+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_cv_prog_cc_c89=no -ac_save_CC=$CC -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include -#include -#include -#include -/* Most of the following tests are stolen from RCS 5.7's src/conf.sh. */ -struct buf { int x; }; -FILE * (*rcsopen) (struct buf *, struct stat *, int); -static char *e (p, i) - char **p; - int i; -{ - return p[i]; -} -static char *f (char * (*g) (char **, int), char **p, ...) -{ - char *s; - va_list v; - va_start (v,p); - s = g (p, va_arg (v,int)); - va_end (v); - return s; -} - -/* OSF 4.0 Compaq cc is some sort of almost-ANSI by default. It has - function prototypes and stuff, but not '\xHH' hex character constants. - These don't provoke an error unfortunately, instead are silently treated - as 'x'. The following induces an error, until -std is added to get - proper ANSI mode. Curiously '\x00'!='x' always comes out true, for an - array size at least. It's necessary to write '\x00'==0 to get something - that's true only with -std. */ -int osf4_cc_array ['\x00' == 0 ? 1 : -1]; - -/* IBM C 6 for AIX is almost-ANSI by default, but it replaces macro parameters - inside strings and character constants. */ -#define FOO(x) 'x' -int xlc6_cc_array[FOO(a) == 'x' ? 1 : -1]; - -int test (int i, double x); -struct s1 {int (*f) (int a);}; -struct s2 {int (*f) (double a);}; -int pairnames (int, char **, FILE *(*)(struct buf *, struct stat *, int), int, int); -int argc; -char **argv; -int -main () -{ -return f (e, argv, 0) != argv[0] || f (e, argv, 1) != argv[1]; - ; - return 0; -} -_ACEOF -for ac_arg in '' -qlanglvl=extc89 -qlanglvl=ansi -std \ - -Ae "-Aa -D_HPUX_SOURCE" "-Xc -D__EXTENSIONS__" -do - CC="$ac_save_CC $ac_arg" - rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_prog_cc_c89=$ac_arg -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - -fi - -rm -f core conftest.err conftest.$ac_objext - test "x$ac_cv_prog_cc_c89" != "xno" && break -done -rm -f conftest.$ac_ext -CC=$ac_save_CC - -fi -# AC_CACHE_VAL -case "x$ac_cv_prog_cc_c89" in - x) - { $as_echo "$as_me:$LINENO: result: none needed" >&5 -$as_echo "none needed" >&6; } ;; - xno) - { $as_echo "$as_me:$LINENO: result: unsupported" >&5 -$as_echo "unsupported" >&6; } ;; - *) - CC="$CC $ac_cv_prog_cc_c89" - { $as_echo "$as_me:$LINENO: result: $ac_cv_prog_cc_c89" >&5 -$as_echo "$ac_cv_prog_cc_c89" >&6; } ;; -esac - - -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu -DEPDIR="${am__leading_dot}deps" - -ac_config_commands="$ac_config_commands depfiles" - - -am_make=${MAKE-make} -cat > confinc << 'END' -am__doit: - @echo this is the am__doit target -.PHONY: am__doit -END -# If we don't find an include directive, just comment out the code. -{ $as_echo "$as_me:$LINENO: checking for style of include used by $am_make" >&5 -$as_echo_n "checking for style of include used by $am_make... " >&6; } -am__include="#" -am__quote= -_am_result=none -# First try GNU make style include. -echo "include confinc" > confmf -# Ignore all kinds of additional output from `make'. -case `$am_make -s -f confmf 2> /dev/null` in #( -*the\ am__doit\ target*) - am__include=include - am__quote= - _am_result=GNU - ;; -esac -# Now try BSD make style include. -if test "$am__include" = "#"; then - echo '.include "confinc"' > confmf - case `$am_make -s -f confmf 2> /dev/null` in #( - *the\ am__doit\ target*) - am__include=.include - am__quote="\"" - _am_result=BSD - ;; - esac -fi - - -{ $as_echo "$as_me:$LINENO: result: $_am_result" >&5 -$as_echo "$_am_result" >&6; } -rm -f confinc confmf - -# Check whether --enable-dependency-tracking was given. -if test "${enable_dependency_tracking+set}" = set; then - enableval=$enable_dependency_tracking; -fi - -if test "x$enable_dependency_tracking" != xno; then - am_depcomp="$ac_aux_dir/depcomp" - AMDEPBACKSLASH='\' -fi - if test "x$enable_dependency_tracking" != xno; then - AMDEP_TRUE= - AMDEP_FALSE='#' -else - AMDEP_TRUE='#' - AMDEP_FALSE= -fi - - - -depcc="$CC" am_compiler_list= - -{ $as_echo "$as_me:$LINENO: checking dependency style of $depcc" >&5 -$as_echo_n "checking dependency style of $depcc... " >&6; } -if test "${am_cv_CC_dependencies_compiler_type+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -z "$AMDEP_TRUE" && test -f "$am_depcomp"; then - # We make a subdir and do the tests there. Otherwise we can end up - # making bogus files that we don't know about and never remove. For - # instance it was reported that on HP-UX the gcc test will end up - # making a dummy file named `D' -- because `-MD' means `put the output - # in D'. - mkdir conftest.dir - # Copy depcomp to subdir because otherwise we won't find it if we're - # using a relative directory. - cp "$am_depcomp" conftest.dir - cd conftest.dir - # We will build objects and dependencies in a subdirectory because - # it helps to detect inapplicable dependency modes. For instance - # both Tru64's cc and ICC support -MD to output dependencies as a - # side effect of compilation, but ICC will put the dependencies in - # the current directory while Tru64 will put them in the object - # directory. - mkdir sub - - am_cv_CC_dependencies_compiler_type=none - if test "$am_compiler_list" = ""; then - am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` - fi - am__universal=false - case " $depcc " in #( - *\ -arch\ *\ -arch\ *) am__universal=true ;; - esac - - for depmode in $am_compiler_list; do - # Setup a source with many dependencies, because some compilers - # like to wrap large dependency lists on column 80 (with \), and - # we should not choose a depcomp mode which is confused by this. - # - # We need to recreate these files for each test, as the compiler may - # overwrite some of them when testing with obscure command lines. - # This happens at least with the AIX C compiler. - : > sub/conftest.c - for i in 1 2 3 4 5 6; do - echo '#include "conftst'$i'.h"' >> sub/conftest.c - # Using `: > sub/conftst$i.h' creates only sub/conftst1.h with - # Solaris 8's {/usr,}/bin/sh. - touch sub/conftst$i.h - done - echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf - - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. Also, some Intel - # versions had trouble with output in subdirs - am__obj=sub/conftest.${OBJEXT-o} - am__minus_obj="-o $am__obj" - case $depmode in - gcc) - # This depmode causes a compiler race in universal mode. - test "$am__universal" = false || continue - ;; - nosideeffect) - # after this tag, mechanisms are not by side-effect, so they'll - # only be used when explicitly requested - if test "x$enable_dependency_tracking" = xyes; then - continue - else - break - fi - ;; - msvisualcpp | msvcmsys) - # This compiler won't grok `-c -o', but also, the minuso test has - # not run yet. These depmodes are late enough in the game, and - # so weak that their functioning should not be impacted. - am__obj=conftest.${OBJEXT-o} - am__minus_obj= - ;; - none) break ;; - esac - if depmode=$depmode \ - source=sub/conftest.c object=$am__obj \ - depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ - >/dev/null 2>conftest.err && - grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep $am__obj sub/conftest.Po > /dev/null 2>&1 && - ${MAKE-make} -s -f confmf > /dev/null 2>&1; then - # icc doesn't choke on unknown options, it will just issue warnings - # or remarks (even with -Werror). So we grep stderr for any message - # that says an option was ignored or not supported. - # When given -MP, icc 7.0 and 7.1 complain thusly: - # icc: Command line warning: ignoring option '-M'; no argument required - # The diagnosis changed in icc 8.0: - # icc: Command line remark: option '-MP' not supported - if (grep 'ignoring option' conftest.err || - grep 'not supported' conftest.err) >/dev/null 2>&1; then :; else - am_cv_CC_dependencies_compiler_type=$depmode - break - fi - fi - done - - cd .. - rm -rf conftest.dir -else - am_cv_CC_dependencies_compiler_type=none -fi - -fi -{ $as_echo "$as_me:$LINENO: result: $am_cv_CC_dependencies_compiler_type" >&5 -$as_echo "$am_cv_CC_dependencies_compiler_type" >&6; } -CCDEPMODE=depmode=$am_cv_CC_dependencies_compiler_type - - if - test "x$enable_dependency_tracking" != xno \ - && test "$am_cv_CC_dependencies_compiler_type" = gcc3; then - am__fastdepCC_TRUE= - am__fastdepCC_FALSE='#' -else - am__fastdepCC_TRUE='#' - am__fastdepCC_FALSE= -fi - - -ac_ext=cpp -ac_cpp='$CXXCPP $CPPFLAGS' -ac_compile='$CXX -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CXX -o conftest$ac_exeext $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_cxx_compiler_gnu -if test -z "$CXX"; then - if test -n "$CCC"; then - CXX=$CCC - else - if test -n "$ac_tool_prefix"; then - for ac_prog in gcc g++ cl KCC CC cxx cc++ xlC aCC c++ - do - # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. -set dummy $ac_tool_prefix$ac_prog; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_CXX+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$CXX"; then - ac_cv_prog_CXX="$CXX" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_CXX="$ac_tool_prefix$ac_prog" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -CXX=$ac_cv_prog_CXX -if test -n "$CXX"; then - { $as_echo "$as_me:$LINENO: result: $CXX" >&5 -$as_echo "$CXX" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - - test -n "$CXX" && break - done -fi -if test -z "$CXX"; then - ac_ct_CXX=$CXX - for ac_prog in gcc g++ cl KCC CC cxx cc++ xlC aCC c++ -do - # Extract the first word of "$ac_prog", so it can be a program name with args. -set dummy $ac_prog; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_ac_ct_CXX+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$ac_ct_CXX"; then - ac_cv_prog_ac_ct_CXX="$ac_ct_CXX" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_ac_ct_CXX="$ac_prog" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -ac_ct_CXX=$ac_cv_prog_ac_ct_CXX -if test -n "$ac_ct_CXX"; then - { $as_echo "$as_me:$LINENO: result: $ac_ct_CXX" >&5 -$as_echo "$ac_ct_CXX" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - - test -n "$ac_ct_CXX" && break -done - - if test "x$ac_ct_CXX" = x; then - CXX="g++" - else - case $cross_compiling:$ac_tool_warned in -yes:) -{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 -$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} -ac_tool_warned=yes ;; -esac - CXX=$ac_ct_CXX - fi -fi - - fi -fi -# Provide some information about the compiler. -$as_echo "$as_me:$LINENO: checking for C++ compiler version" >&5 -set X $ac_compile -ac_compiler=$2 -{ (ac_try="$ac_compiler --version >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler --version >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } -{ (ac_try="$ac_compiler -v >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler -v >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } -{ (ac_try="$ac_compiler -V >&5" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compiler -V >&5") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } - -{ $as_echo "$as_me:$LINENO: checking whether we are using the GNU C++ compiler" >&5 -$as_echo_n "checking whether we are using the GNU C++ compiler... " >&6; } -if test "${ac_cv_cxx_compiler_gnu+set}" = set; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ -#ifndef __GNUC__ - choke me -#endif - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_cxx_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_compiler_gnu=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_compiler_gnu=no -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -ac_cv_cxx_compiler_gnu=$ac_compiler_gnu - -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_cxx_compiler_gnu" >&5 -$as_echo "$ac_cv_cxx_compiler_gnu" >&6; } -if test $ac_compiler_gnu = yes; then - GXX=yes -else - GXX= -fi -ac_test_CXXFLAGS=${CXXFLAGS+set} -ac_save_CXXFLAGS=$CXXFLAGS -{ $as_echo "$as_me:$LINENO: checking whether $CXX accepts -g" >&5 -$as_echo_n "checking whether $CXX accepts -g... " >&6; } -if test "${ac_cv_prog_cxx_g+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_save_cxx_werror_flag=$ac_cxx_werror_flag - ac_cxx_werror_flag=yes - ac_cv_prog_cxx_g=no - CXXFLAGS="-g" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_cxx_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_prog_cxx_g=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - CXXFLAGS="" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_cxx_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - : -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cxx_werror_flag=$ac_save_cxx_werror_flag - CXXFLAGS="-g" - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_cxx_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_prog_cxx_g=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext - ac_cxx_werror_flag=$ac_save_cxx_werror_flag -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_prog_cxx_g" >&5 -$as_echo "$ac_cv_prog_cxx_g" >&6; } -if test "$ac_test_CXXFLAGS" = set; then - CXXFLAGS=$ac_save_CXXFLAGS -elif test $ac_cv_prog_cxx_g = yes; then - if test "$GXX" = yes; then - CXXFLAGS="-g -O2" - else - CXXFLAGS="-g" - fi -else - if test "$GXX" = yes; then - CXXFLAGS="-O2" - else - CXXFLAGS= - fi -fi -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu - -depcc="$CXX" am_compiler_list= - -{ $as_echo "$as_me:$LINENO: checking dependency style of $depcc" >&5 -$as_echo_n "checking dependency style of $depcc... " >&6; } -if test "${am_cv_CXX_dependencies_compiler_type+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -z "$AMDEP_TRUE" && test -f "$am_depcomp"; then - # We make a subdir and do the tests there. Otherwise we can end up - # making bogus files that we don't know about and never remove. For - # instance it was reported that on HP-UX the gcc test will end up - # making a dummy file named `D' -- because `-MD' means `put the output - # in D'. - mkdir conftest.dir - # Copy depcomp to subdir because otherwise we won't find it if we're - # using a relative directory. - cp "$am_depcomp" conftest.dir - cd conftest.dir - # We will build objects and dependencies in a subdirectory because - # it helps to detect inapplicable dependency modes. For instance - # both Tru64's cc and ICC support -MD to output dependencies as a - # side effect of compilation, but ICC will put the dependencies in - # the current directory while Tru64 will put them in the object - # directory. - mkdir sub - - am_cv_CXX_dependencies_compiler_type=none - if test "$am_compiler_list" = ""; then - am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` - fi - am__universal=false - case " $depcc " in #( - *\ -arch\ *\ -arch\ *) am__universal=true ;; - esac - - for depmode in $am_compiler_list; do - # Setup a source with many dependencies, because some compilers - # like to wrap large dependency lists on column 80 (with \), and - # we should not choose a depcomp mode which is confused by this. - # - # We need to recreate these files for each test, as the compiler may - # overwrite some of them when testing with obscure command lines. - # This happens at least with the AIX C compiler. - : > sub/conftest.c - for i in 1 2 3 4 5 6; do - echo '#include "conftst'$i'.h"' >> sub/conftest.c - # Using `: > sub/conftst$i.h' creates only sub/conftst1.h with - # Solaris 8's {/usr,}/bin/sh. - touch sub/conftst$i.h - done - echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf - - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. Also, some Intel - # versions had trouble with output in subdirs - am__obj=sub/conftest.${OBJEXT-o} - am__minus_obj="-o $am__obj" - case $depmode in - gcc) - # This depmode causes a compiler race in universal mode. - test "$am__universal" = false || continue - ;; - nosideeffect) - # after this tag, mechanisms are not by side-effect, so they'll - # only be used when explicitly requested - if test "x$enable_dependency_tracking" = xyes; then - continue - else - break - fi - ;; - msvisualcpp | msvcmsys) - # This compiler won't grok `-c -o', but also, the minuso test has - # not run yet. These depmodes are late enough in the game, and - # so weak that their functioning should not be impacted. - am__obj=conftest.${OBJEXT-o} - am__minus_obj= - ;; - none) break ;; - esac - if depmode=$depmode \ - source=sub/conftest.c object=$am__obj \ - depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ - >/dev/null 2>conftest.err && - grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep $am__obj sub/conftest.Po > /dev/null 2>&1 && - ${MAKE-make} -s -f confmf > /dev/null 2>&1; then - # icc doesn't choke on unknown options, it will just issue warnings - # or remarks (even with -Werror). So we grep stderr for any message - # that says an option was ignored or not supported. - # When given -MP, icc 7.0 and 7.1 complain thusly: - # icc: Command line warning: ignoring option '-M'; no argument required - # The diagnosis changed in icc 8.0: - # icc: Command line remark: option '-MP' not supported - if (grep 'ignoring option' conftest.err || - grep 'not supported' conftest.err) >/dev/null 2>&1; then :; else - am_cv_CXX_dependencies_compiler_type=$depmode - break - fi - fi - done - - cd .. - rm -rf conftest.dir -else - am_cv_CXX_dependencies_compiler_type=none -fi - -fi -{ $as_echo "$as_me:$LINENO: result: $am_cv_CXX_dependencies_compiler_type" >&5 -$as_echo "$am_cv_CXX_dependencies_compiler_type" >&6; } -CXXDEPMODE=depmode=$am_cv_CXX_dependencies_compiler_type - - if - test "x$enable_dependency_tracking" != xno \ - && test "$am_cv_CXX_dependencies_compiler_type" = gcc3; then - am__fastdepCXX_TRUE= - am__fastdepCXX_FALSE='#' -else - am__fastdepCXX_TRUE='#' - am__fastdepCXX_FALSE= -fi - - -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu -{ $as_echo "$as_me:$LINENO: checking how to run the C preprocessor" >&5 -$as_echo_n "checking how to run the C preprocessor... " >&6; } -# On Suns, sometimes $CPP names a directory. -if test -n "$CPP" && test -d "$CPP"; then - CPP= -fi -if test -z "$CPP"; then - if test "${ac_cv_prog_CPP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - # Double quotes because CPP needs to be expanded - for CPP in "$CC -E" "$CC -E -traditional-cpp" "/lib/cpp" - do - ac_preproc_ok=false -for ac_c_preproc_warn_flag in '' yes -do - # Use a header file that comes with gcc, so configuring glibc - # with a fresh cross-compiler works. - # Prefer to if __STDC__ is defined, since - # exists even on freestanding compilers. - # On the NeXT, cc -E runs the code through the compiler's parser, - # not just through cpp. "Syntax error" is here to catch this case. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#ifdef __STDC__ -# include -#else -# include -#endif - Syntax error -_ACEOF -if { (ac_try="$ac_cpp conftest.$ac_ext" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_cpp conftest.$ac_ext") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } >/dev/null && { - test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || - test ! -s conftest.err - }; then - : -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - # Broken: fails on valid input. -continue -fi - -rm -f conftest.err conftest.$ac_ext - - # OK, works on sane cases. Now check whether nonexistent headers - # can be detected and how. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include -_ACEOF -if { (ac_try="$ac_cpp conftest.$ac_ext" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_cpp conftest.$ac_ext") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } >/dev/null && { - test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || - test ! -s conftest.err - }; then - # Broken: success on invalid input. -continue -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - # Passes both tests. -ac_preproc_ok=: -break -fi - -rm -f conftest.err conftest.$ac_ext - -done -# Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped. -rm -f conftest.err conftest.$ac_ext -if $ac_preproc_ok; then - break -fi - - done - ac_cv_prog_CPP=$CPP - -fi - CPP=$ac_cv_prog_CPP -else - ac_cv_prog_CPP=$CPP -fi -{ $as_echo "$as_me:$LINENO: result: $CPP" >&5 -$as_echo "$CPP" >&6; } -ac_preproc_ok=false -for ac_c_preproc_warn_flag in '' yes -do - # Use a header file that comes with gcc, so configuring glibc - # with a fresh cross-compiler works. - # Prefer to if __STDC__ is defined, since - # exists even on freestanding compilers. - # On the NeXT, cc -E runs the code through the compiler's parser, - # not just through cpp. "Syntax error" is here to catch this case. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#ifdef __STDC__ -# include -#else -# include -#endif - Syntax error -_ACEOF -if { (ac_try="$ac_cpp conftest.$ac_ext" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_cpp conftest.$ac_ext") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } >/dev/null && { - test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || - test ! -s conftest.err - }; then - : -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - # Broken: fails on valid input. -continue -fi - -rm -f conftest.err conftest.$ac_ext - - # OK, works on sane cases. Now check whether nonexistent headers - # can be detected and how. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include -_ACEOF -if { (ac_try="$ac_cpp conftest.$ac_ext" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_cpp conftest.$ac_ext") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } >/dev/null && { - test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || - test ! -s conftest.err - }; then - # Broken: success on invalid input. -continue -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - # Passes both tests. -ac_preproc_ok=: -break -fi - -rm -f conftest.err conftest.$ac_ext - -done -# Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped. -rm -f conftest.err conftest.$ac_ext -if $ac_preproc_ok; then - : -else - { { $as_echo "$as_me:$LINENO: error: in \`$ac_pwd':" >&5 -$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} -{ { $as_echo "$as_me:$LINENO: error: C preprocessor \"$CPP\" fails sanity check -See \`config.log' for more details." >&5 -$as_echo "$as_me: error: C preprocessor \"$CPP\" fails sanity check -See \`config.log' for more details." >&2;} - { (exit 1); exit 1; }; }; } -fi - -ac_ext=c -ac_cpp='$CPP $CPPFLAGS' -ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' -ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' -ac_compiler_gnu=$ac_cv_c_compiler_gnu - -if test "x$CC" != xcc; then - { $as_echo "$as_me:$LINENO: checking whether $CC and cc understand -c and -o together" >&5 -$as_echo_n "checking whether $CC and cc understand -c and -o together... " >&6; } -else - { $as_echo "$as_me:$LINENO: checking whether cc understands -c and -o together" >&5 -$as_echo_n "checking whether cc understands -c and -o together... " >&6; } -fi -set dummy $CC; ac_cc=`$as_echo "$2" | - sed 's/[^a-zA-Z0-9_]/_/g;s/^[0-9]/_/'` -if { as_var=ac_cv_prog_cc_${ac_cc}_c_o; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - -int -main () -{ - - ; - return 0; -} -_ACEOF -# Make sure it works both with $CC and with simple cc. -# We do the test twice because some compilers refuse to overwrite an -# existing .o file with -o, though they will create one. -ac_try='$CC -c conftest.$ac_ext -o conftest2.$ac_objext >&5' -rm -f conftest2.* -if { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && - test -f conftest2.$ac_objext && { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; -then - eval ac_cv_prog_cc_${ac_cc}_c_o=yes - if test "x$CC" != xcc; then - # Test first that cc exists at all. - if { ac_try='cc -c conftest.$ac_ext >&5' - { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; }; then - ac_try='cc -c conftest.$ac_ext -o conftest2.$ac_objext >&5' - rm -f conftest2.* - if { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && - test -f conftest2.$ac_objext && { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; - then - # cc works too. - : - else - # cc exists but doesn't like -o. - eval ac_cv_prog_cc_${ac_cc}_c_o=no - fi - fi - fi -else - eval ac_cv_prog_cc_${ac_cc}_c_o=no -fi -rm -f core conftest* - -fi -if eval test \$ac_cv_prog_cc_${ac_cc}_c_o = yes; then - { $as_echo "$as_me:$LINENO: result: yes" >&5 -$as_echo "yes" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } - -cat >>confdefs.h <<\_ACEOF -#define NO_MINUS_C_MINUS_O 1 -_ACEOF - -fi - -# FIXME: we rely on the cache variable name because -# there is no other way. -set dummy $CC -am_cc=`echo $2 | sed 's/[^a-zA-Z0-9_]/_/g;s/^[0-9]/_/'` -eval am_t=\$ac_cv_prog_cc_${am_cc}_c_o -if test "$am_t" != yes; then - # Losing compiler, so override with the script. - # FIXME: It is wrong to rewrite CC. - # But if we don't then we get into trouble of one sort or another. - # A longer-term fix would be to have automake use am__CC in this case, - # and then we could set am__CC="\$(top_srcdir)/compile \$(CC)" - CC="$am_aux_dir/compile $CC" -fi - - - -# Find a good install program. We prefer a C program (faster), -# so one script is as good as another. But avoid the broken or -# incompatible versions: -# SysV /etc/install, /usr/sbin/install -# SunOS /usr/etc/install -# IRIX /sbin/install -# AIX /bin/install -# AmigaOS /C/install, which installs bootblocks on floppy discs -# AIX 4 /usr/bin/installbsd, which doesn't work without a -g flag -# AFS /usr/afsws/bin/install, which mishandles nonexistent args -# SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff" -# OS/2's system install, which has a completely different semantic -# ./install, which can be erroneously created by make from ./install.sh. -# Reject install programs that cannot install multiple files. -{ $as_echo "$as_me:$LINENO: checking for a BSD-compatible install" >&5 -$as_echo_n "checking for a BSD-compatible install... " >&6; } -if test -z "$INSTALL"; then -if test "${ac_cv_path_install+set}" = set; then - $as_echo_n "(cached) " >&6 -else - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - # Account for people who put trailing slashes in PATH elements. -case $as_dir/ in - ./ | .// | /cC/* | \ - /etc/* | /usr/sbin/* | /usr/etc/* | /sbin/* | /usr/afsws/bin/* | \ - ?:\\/os2\\/install\\/* | ?:\\/OS2\\/INSTALL\\/* | \ - /usr/ucb/* ) ;; - *) - # OSF1 and SCO ODT 3.0 have their own names for install. - # Don't use installbsd from OSF since it installs stuff as root - # by default. - for ac_prog in ginstall scoinst install; do - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_prog$ac_exec_ext" && $as_test_x "$as_dir/$ac_prog$ac_exec_ext"; }; then - if test $ac_prog = install && - grep dspmsg "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then - # AIX install. It has an incompatible calling convention. - : - elif test $ac_prog = install && - grep pwplus "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then - # program-specific install script used by HP pwplus--don't use. - : - else - rm -rf conftest.one conftest.two conftest.dir - echo one > conftest.one - echo two > conftest.two - mkdir conftest.dir - if "$as_dir/$ac_prog$ac_exec_ext" -c conftest.one conftest.two "`pwd`/conftest.dir" && - test -s conftest.one && test -s conftest.two && - test -s conftest.dir/conftest.one && - test -s conftest.dir/conftest.two - then - ac_cv_path_install="$as_dir/$ac_prog$ac_exec_ext -c" - break 3 - fi - fi - fi - done - done - ;; -esac - -done -IFS=$as_save_IFS - -rm -rf conftest.one conftest.two conftest.dir - -fi - if test "${ac_cv_path_install+set}" = set; then - INSTALL=$ac_cv_path_install - else - # As a last resort, use the slow shell script. Don't cache a - # value for INSTALL within a source directory, because that will - # break other packages using the cache if that directory is - # removed, or if the value is a relative name. - INSTALL=$ac_install_sh - fi -fi -{ $as_echo "$as_me:$LINENO: result: $INSTALL" >&5 -$as_echo "$INSTALL" >&6; } - -# Use test -z because SunOS4 sh mishandles braces in ${var-val}. -# It thinks the first close brace ends the variable substitution. -test -z "$INSTALL_PROGRAM" && INSTALL_PROGRAM='${INSTALL}' - -test -z "$INSTALL_SCRIPT" && INSTALL_SCRIPT='${INSTALL}' - -test -z "$INSTALL_DATA" && INSTALL_DATA='${INSTALL} -m 644' - -if test -n "$ac_tool_prefix"; then - # Extract the first word of "${ac_tool_prefix}ranlib", so it can be a program name with args. -set dummy ${ac_tool_prefix}ranlib; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_RANLIB+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$RANLIB"; then - ac_cv_prog_RANLIB="$RANLIB" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_RANLIB="${ac_tool_prefix}ranlib" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -RANLIB=$ac_cv_prog_RANLIB -if test -n "$RANLIB"; then - { $as_echo "$as_me:$LINENO: result: $RANLIB" >&5 -$as_echo "$RANLIB" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - -fi -if test -z "$ac_cv_prog_RANLIB"; then - ac_ct_RANLIB=$RANLIB - # Extract the first word of "ranlib", so it can be a program name with args. -set dummy ranlib; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_ac_ct_RANLIB+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$ac_ct_RANLIB"; then - ac_cv_prog_ac_ct_RANLIB="$ac_ct_RANLIB" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_ac_ct_RANLIB="ranlib" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -ac_ct_RANLIB=$ac_cv_prog_ac_ct_RANLIB -if test -n "$ac_ct_RANLIB"; then - { $as_echo "$as_me:$LINENO: result: $ac_ct_RANLIB" >&5 -$as_echo "$ac_ct_RANLIB" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - if test "x$ac_ct_RANLIB" = x; then - RANLIB=":" - else - case $cross_compiling:$ac_tool_warned in -yes:) -{ $as_echo "$as_me:$LINENO: WARNING: using cross tools not prefixed with host triplet" >&5 -$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} -ac_tool_warned=yes ;; -esac - RANLIB=$ac_ct_RANLIB - fi -else - RANLIB="$ac_cv_prog_RANLIB" -fi - - -# By default we simply use the C compiler to build assembly code. - -test "${CCAS+set}" = set || CCAS=$CC -test "${CCASFLAGS+set}" = set || CCASFLAGS=$CFLAGS - - - -depcc="$CCAS" am_compiler_list= - -{ $as_echo "$as_me:$LINENO: checking dependency style of $depcc" >&5 -$as_echo_n "checking dependency style of $depcc... " >&6; } -if test "${am_cv_CCAS_dependencies_compiler_type+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -z "$AMDEP_TRUE" && test -f "$am_depcomp"; then - # We make a subdir and do the tests there. Otherwise we can end up - # making bogus files that we don't know about and never remove. For - # instance it was reported that on HP-UX the gcc test will end up - # making a dummy file named `D' -- because `-MD' means `put the output - # in D'. - mkdir conftest.dir - # Copy depcomp to subdir because otherwise we won't find it if we're - # using a relative directory. - cp "$am_depcomp" conftest.dir - cd conftest.dir - # We will build objects and dependencies in a subdirectory because - # it helps to detect inapplicable dependency modes. For instance - # both Tru64's cc and ICC support -MD to output dependencies as a - # side effect of compilation, but ICC will put the dependencies in - # the current directory while Tru64 will put them in the object - # directory. - mkdir sub - - am_cv_CCAS_dependencies_compiler_type=none - if test "$am_compiler_list" = ""; then - am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` - fi - am__universal=false - - - for depmode in $am_compiler_list; do - # Setup a source with many dependencies, because some compilers - # like to wrap large dependency lists on column 80 (with \), and - # we should not choose a depcomp mode which is confused by this. - # - # We need to recreate these files for each test, as the compiler may - # overwrite some of them when testing with obscure command lines. - # This happens at least with the AIX C compiler. - : > sub/conftest.c - for i in 1 2 3 4 5 6; do - echo '#include "conftst'$i'.h"' >> sub/conftest.c - # Using `: > sub/conftst$i.h' creates only sub/conftst1.h with - # Solaris 8's {/usr,}/bin/sh. - touch sub/conftst$i.h - done - echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf - - # We check with `-c' and `-o' for the sake of the "dashmstdout" - # mode. It turns out that the SunPro C++ compiler does not properly - # handle `-M -o', and we need to detect this. Also, some Intel - # versions had trouble with output in subdirs - am__obj=sub/conftest.${OBJEXT-o} - am__minus_obj="-o $am__obj" - case $depmode in - gcc) - # This depmode causes a compiler race in universal mode. - test "$am__universal" = false || continue - ;; - nosideeffect) - # after this tag, mechanisms are not by side-effect, so they'll - # only be used when explicitly requested - if test "x$enable_dependency_tracking" = xyes; then - continue - else - break - fi - ;; - msvisualcpp | msvcmsys) - # This compiler won't grok `-c -o', but also, the minuso test has - # not run yet. These depmodes are late enough in the game, and - # so weak that their functioning should not be impacted. - am__obj=conftest.${OBJEXT-o} - am__minus_obj= - ;; - none) break ;; - esac - if depmode=$depmode \ - source=sub/conftest.c object=$am__obj \ - depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ - $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ - >/dev/null 2>conftest.err && - grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && - grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && - grep $am__obj sub/conftest.Po > /dev/null 2>&1 && - ${MAKE-make} -s -f confmf > /dev/null 2>&1; then - # icc doesn't choke on unknown options, it will just issue warnings - # or remarks (even with -Werror). So we grep stderr for any message - # that says an option was ignored or not supported. - # When given -MP, icc 7.0 and 7.1 complain thusly: - # icc: Command line warning: ignoring option '-M'; no argument required - # The diagnosis changed in icc 8.0: - # icc: Command line remark: option '-MP' not supported - if (grep 'ignoring option' conftest.err || - grep 'not supported' conftest.err) >/dev/null 2>&1; then :; else - am_cv_CCAS_dependencies_compiler_type=$depmode - break - fi - fi - done - - cd .. - rm -rf conftest.dir -else - am_cv_CCAS_dependencies_compiler_type=none -fi - -fi -{ $as_echo "$as_me:$LINENO: result: $am_cv_CCAS_dependencies_compiler_type" >&5 -$as_echo "$am_cv_CCAS_dependencies_compiler_type" >&6; } -CCASDEPMODE=depmode=$am_cv_CCAS_dependencies_compiler_type - - if - test "x$enable_dependency_tracking" != xno \ - && test "$am_cv_CCAS_dependencies_compiler_type" = gcc3; then - am__fastdepCCAS_TRUE= - am__fastdepCCAS_FALSE='#' -else - am__fastdepCCAS_TRUE='#' - am__fastdepCCAS_FALSE= -fi - - - -ZEROGS_CURRENT=0 - -ZEROGS_REVISION=96 - -ZEROGS_AGE=2 - -ZEROGS_RELEASE=$ZEROGS_CURRENT.$ZEROGS_REVISION.$ZEROGS_AGE - -ZEROGS_SONAME=libZeroGSogl.so.$ZEROGS_CURRENT.$ZEROGS_REVISION.$ZEROGS_AGE - - -CFLAGS=" -I/opt/cg/include -L/opt/cg/lib " -CPPFLAGS=" -I/opt/cg/include -L/opt/cg/lib " -CXXFLAGS=" -I/opt/cg/include -L/opt/cg/lib " -CCASFLAGS= - -{ $as_echo "$as_me:$LINENO: checking debug build" >&5 -$as_echo_n "checking debug build... " >&6; } -# Check whether --enable-debug was given. -if test "${enable_debug+set}" = set; then - enableval=$enable_debug; debug=$enableval -else - debug=no -fi - -if test "x$debug" == xyes -then - -cat >>confdefs.h <<\_ACEOF -#define _DEBUG 1 -_ACEOF - - CFLAGS+="-g -m32 " - CPPFLAGS+="-g -m32 " - CXXFLAGS+="-g -fpermissive -m32 " - CCASFLAGS+=" -m32 " -else - -cat >>confdefs.h <<\_ACEOF -#define NDEBUG 1 -_ACEOF - - CFLAGS+="-O3 -fomit-frame-pointer -m32 " - CPPFLAGS+="-O3 -fomit-frame-pointer -m32 " - CXXFLAGS+="-O3 -fomit-frame-pointer -fpermissive -m32 " - CCASFLAGS+=" -m32 " -fi -{ $as_echo "$as_me:$LINENO: result: $debug" >&5 -$as_echo "$debug" >&6; } - - - - -for ac_func in _aligned_malloc _aligned_free -do -as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh` -{ $as_echo "$as_me:$LINENO: checking for $ac_func" >&5 -$as_echo_n "checking for $ac_func... " >&6; } -if { as_var=$as_ac_var; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -/* Define $ac_func to an innocuous variant, in case declares $ac_func. - For example, HP-UX 11i declares gettimeofday. */ -#define $ac_func innocuous_$ac_func - -/* System header to define __stub macros and hopefully few prototypes, - which can conflict with char $ac_func (); below. - Prefer to if __STDC__ is defined, since - exists even on freestanding compilers. */ - -#ifdef __STDC__ -# include -#else -# include -#endif - -#undef $ac_func - -/* Override any GCC internal prototype to avoid an error. - Use char because int might match the return type of a GCC - builtin and then its argument prototype would still apply. */ -#ifdef __cplusplus -extern "C" -#endif -char $ac_func (); -/* The GNU C library defines this for functions which it implements - to always fail with ENOSYS. Some functions are actually named - something starting with __ and the normal name is an alias. */ -#if defined __stub_$ac_func || defined __stub___$ac_func -choke me -#endif - -int -main () -{ -return $ac_func (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - eval "$as_ac_var=yes" -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - eval "$as_ac_var=no" -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -fi -ac_res=`eval 'as_val=${'$as_ac_var'} - $as_echo "$as_val"'` - { $as_echo "$as_me:$LINENO: result: $ac_res" >&5 -$as_echo "$ac_res" >&6; } -as_val=`eval 'as_val=${'$as_ac_var'} - $as_echo "$as_val"'` - if test "x$as_val" = x""yes; then - cat >>confdefs.h <<_ACEOF -#define `$as_echo "HAVE_$ac_func" | $as_tr_cpp` 1 -_ACEOF - cat >>confdefs.h <<\_ACEOF -#define HAVE_ALIGNED_MALLOC 1 -_ACEOF - -fi -done - - -{ $as_echo "$as_me:$LINENO: checking for development build..." >&5 -$as_echo_n "checking for development build...... " >&6; } -# Check whether --enable-devbuild was given. -if test "${enable_devbuild+set}" = set; then - enableval=$enable_devbuild; devbuild=$enableval -else - devbuild=no -fi - -if test "x$devbuild" == xyes -then - -cat >>confdefs.h <<\_ACEOF -#define ZEROGS_DEVBUILD 1 -_ACEOF - -fi -{ $as_echo "$as_me:$LINENO: result: $devbuild" >&5 -$as_echo "$devbuild" >&6; } - if test x$devbuild = xno; then - RELEASE_TO_PUBLIC_TRUE= - RELEASE_TO_PUBLIC_FALSE='#' -else - RELEASE_TO_PUBLIC_TRUE='#' - RELEASE_TO_PUBLIC_FALSE= -fi - - -{ $as_echo "$as_me:$LINENO: checking check for sse2..." >&5 -$as_echo_n "checking check for sse2...... " >&6; } -# Check whether --enable-sse2 was given. -if test "${enable_sse2+set}" = set; then - enableval=$enable_sse2; sse2=$enableval -else - sse2=no -fi - -if test "x$sse2" == xyes -then - -cat >>confdefs.h <<\_ACEOF -#define ZEROGS_SSE2 1 -_ACEOF - -fi -{ $as_echo "$as_me:$LINENO: result: $sse2" >&5 -$as_echo "$sse2" >&6; } - if test x$sse2 = xyes; then - SSE2_TRUE= - SSE2_FALSE='#' -else - SSE2_TRUE='#' - SSE2_FALSE= -fi - - -{ $as_echo "$as_me:$LINENO: checking gtk2+" >&5 -$as_echo_n "checking gtk2+... " >&6; } -# Extract the first word of "pkg-config", so it can be a program name with args. -set dummy pkg-config; ac_word=$2 -{ $as_echo "$as_me:$LINENO: checking for $ac_word" >&5 -$as_echo_n "checking for $ac_word... " >&6; } -if test "${ac_cv_prog_GTK_CONFIG+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -n "$GTK_CONFIG"; then - ac_cv_prog_GTK_CONFIG="$GTK_CONFIG" # Let the user override the test. -else -as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_exec_ext in '' $ac_executable_extensions; do - if { test -f "$as_dir/$ac_word$ac_exec_ext" && $as_test_x "$as_dir/$ac_word$ac_exec_ext"; }; then - ac_cv_prog_GTK_CONFIG="pkg-config" - $as_echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5 - break 2 - fi -done -done -IFS=$as_save_IFS - -fi -fi -GTK_CONFIG=$ac_cv_prog_GTK_CONFIG -if test -n "$GTK_CONFIG"; then - { $as_echo "$as_me:$LINENO: result: $GTK_CONFIG" >&5 -$as_echo "$GTK_CONFIG" >&6; } -else - { $as_echo "$as_me:$LINENO: result: no" >&5 -$as_echo "no" >&6; } -fi - - -LIBS+=$(pkg-config --libs gtk+-2.0) - - -so_ext=".so.$ZEROGS_RELEASE" -SHARED_LDFLAGS="-shared" - - - -{ $as_echo "$as_me:$LINENO: checking OpenGL" >&5 -$as_echo_n "checking OpenGL... " >&6; } - - - -for ac_header in GL/gl.h GL/glu.h GL/glext.h -do -as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh` -{ $as_echo "$as_me:$LINENO: checking for $ac_header" >&5 -$as_echo_n "checking for $ac_header... " >&6; } -if { as_var=$as_ac_Header; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#if HAVE_GL_GL_H - #include - #endif - #if HAVE_GL_GLU_H - #include - #endif - - -#include <$ac_header> -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - eval "$as_ac_Header=yes" -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - eval "$as_ac_Header=no" -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi -ac_res=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - { $as_echo "$as_me:$LINENO: result: $ac_res" >&5 -$as_echo "$ac_res" >&6; } -as_val=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - if test "x$as_val" = x""yes; then - cat >>confdefs.h <<_ACEOF -#define `$as_echo "HAVE_$ac_header" | $as_tr_cpp` 1 -_ACEOF - -fi - -done - -{ $as_echo "$as_me:$LINENO: checking for main in -lGL" >&5 -$as_echo_n "checking for main in -lGL... " >&6; } -if test "${ac_cv_lib_GL_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lGL $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_GL_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_GL_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_GL_main" >&5 -$as_echo "$ac_cv_lib_GL_main" >&6; } -if test "x$ac_cv_lib_GL_main" = x""yes; then - LIBS="$LIBS -lGL" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lGLU" >&5 -$as_echo_n "checking for main in -lGLU... " >&6; } -if test "${ac_cv_lib_GLU_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lGLU $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_GLU_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_GLU_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_GLU_main" >&5 -$as_echo "$ac_cv_lib_GLU_main" >&6; } -if test "x$ac_cv_lib_GLU_main" = x""yes; then - LIBS="$LIBS -lGLU" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lGLEW" >&5 -$as_echo_n "checking for main in -lGLEW... " >&6; } -if test "${ac_cv_lib_GLEW_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lGLEW $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_GLEW_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_GLEW_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_GLEW_main" >&5 -$as_echo "$ac_cv_lib_GLEW_main" >&6; } -if test "x$ac_cv_lib_GLEW_main" = x""yes; then - LIBS="$LIBS -lGLEW" -fi - -{ $as_echo "$as_me:$LINENO: checking Cg" >&5 -$as_echo_n "checking Cg... " >&6; } -#AC_CHECK_HEADERS([Cg/cg.h Cg/cgGL.h]) - -{ $as_echo "$as_me:$LINENO: checking for grep that handles long lines and -e" >&5 -$as_echo_n "checking for grep that handles long lines and -e... " >&6; } -if test "${ac_cv_path_GREP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if test -z "$GREP"; then - ac_path_GREP_found=false - # Loop through the user's path and test for each of PROGNAME-LIST - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH$PATH_SEPARATOR/usr/xpg4/bin -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_prog in grep ggrep; do - for ac_exec_ext in '' $ac_executable_extensions; do - ac_path_GREP="$as_dir/$ac_prog$ac_exec_ext" - { test -f "$ac_path_GREP" && $as_test_x "$ac_path_GREP"; } || continue -# Check for GNU ac_path_GREP and select it if it is found. - # Check for GNU $ac_path_GREP -case `"$ac_path_GREP" --version 2>&1` in -*GNU*) - ac_cv_path_GREP="$ac_path_GREP" ac_path_GREP_found=:;; -*) - ac_count=0 - $as_echo_n 0123456789 >"conftest.in" - while : - do - cat "conftest.in" "conftest.in" >"conftest.tmp" - mv "conftest.tmp" "conftest.in" - cp "conftest.in" "conftest.nl" - $as_echo 'GREP' >> "conftest.nl" - "$ac_path_GREP" -e 'GREP$' -e '-(cannot match)-' < "conftest.nl" >"conftest.out" 2>/dev/null || break - diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break - ac_count=`expr $ac_count + 1` - if test $ac_count -gt ${ac_path_GREP_max-0}; then - # Best one so far, save it but keep looking for a better one - ac_cv_path_GREP="$ac_path_GREP" - ac_path_GREP_max=$ac_count - fi - # 10*(2^10) chars as input seems more than enough - test $ac_count -gt 10 && break - done - rm -f conftest.in conftest.tmp conftest.nl conftest.out;; -esac - - $ac_path_GREP_found && break 3 - done - done -done -IFS=$as_save_IFS - if test -z "$ac_cv_path_GREP"; then - { { $as_echo "$as_me:$LINENO: error: no acceptable grep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" >&5 -$as_echo "$as_me: error: no acceptable grep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" >&2;} - { (exit 1); exit 1; }; } - fi -else - ac_cv_path_GREP=$GREP -fi - -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_path_GREP" >&5 -$as_echo "$ac_cv_path_GREP" >&6; } - GREP="$ac_cv_path_GREP" - - -{ $as_echo "$as_me:$LINENO: checking for egrep" >&5 -$as_echo_n "checking for egrep... " >&6; } -if test "${ac_cv_path_EGREP+set}" = set; then - $as_echo_n "(cached) " >&6 -else - if echo a | $GREP -E '(a|b)' >/dev/null 2>&1 - then ac_cv_path_EGREP="$GREP -E" - else - if test -z "$EGREP"; then - ac_path_EGREP_found=false - # Loop through the user's path and test for each of PROGNAME-LIST - as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH$PATH_SEPARATOR/usr/xpg4/bin -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - for ac_prog in egrep; do - for ac_exec_ext in '' $ac_executable_extensions; do - ac_path_EGREP="$as_dir/$ac_prog$ac_exec_ext" - { test -f "$ac_path_EGREP" && $as_test_x "$ac_path_EGREP"; } || continue -# Check for GNU ac_path_EGREP and select it if it is found. - # Check for GNU $ac_path_EGREP -case `"$ac_path_EGREP" --version 2>&1` in -*GNU*) - ac_cv_path_EGREP="$ac_path_EGREP" ac_path_EGREP_found=:;; -*) - ac_count=0 - $as_echo_n 0123456789 >"conftest.in" - while : - do - cat "conftest.in" "conftest.in" >"conftest.tmp" - mv "conftest.tmp" "conftest.in" - cp "conftest.in" "conftest.nl" - $as_echo 'EGREP' >> "conftest.nl" - "$ac_path_EGREP" 'EGREP$' < "conftest.nl" >"conftest.out" 2>/dev/null || break - diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break - ac_count=`expr $ac_count + 1` - if test $ac_count -gt ${ac_path_EGREP_max-0}; then - # Best one so far, save it but keep looking for a better one - ac_cv_path_EGREP="$ac_path_EGREP" - ac_path_EGREP_max=$ac_count - fi - # 10*(2^10) chars as input seems more than enough - test $ac_count -gt 10 && break - done - rm -f conftest.in conftest.tmp conftest.nl conftest.out;; -esac - - $ac_path_EGREP_found && break 3 - done - done -done -IFS=$as_save_IFS - if test -z "$ac_cv_path_EGREP"; then - { { $as_echo "$as_me:$LINENO: error: no acceptable egrep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" >&5 -$as_echo "$as_me: error: no acceptable egrep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" >&2;} - { (exit 1); exit 1; }; } - fi -else - ac_cv_path_EGREP=$EGREP -fi - - fi -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_path_EGREP" >&5 -$as_echo "$ac_cv_path_EGREP" >&6; } - EGREP="$ac_cv_path_EGREP" - - -{ $as_echo "$as_me:$LINENO: checking for ANSI C header files" >&5 -$as_echo_n "checking for ANSI C header files... " >&6; } -if test "${ac_cv_header_stdc+set}" = set; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include -#include -#include -#include - -int -main () -{ - - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_cv_header_stdc=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_header_stdc=no -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext - -if test $ac_cv_header_stdc = yes; then - # SunOS 4.x string.h does not declare mem*, contrary to ANSI. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include - -_ACEOF -if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | - $EGREP "memchr" >/dev/null 2>&1; then - : -else - ac_cv_header_stdc=no -fi -rm -f conftest* - -fi - -if test $ac_cv_header_stdc = yes; then - # ISC 2.0.2 stdlib.h does not declare free, contrary to ANSI. - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include - -_ACEOF -if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | - $EGREP "free" >/dev/null 2>&1; then - : -else - ac_cv_header_stdc=no -fi -rm -f conftest* - -fi - -if test $ac_cv_header_stdc = yes; then - # /bin/cc in Irix-4.0.5 gets non-ANSI ctype macros unless using -ansi. - if test "$cross_compiling" = yes; then - : -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include -#include -#if ((' ' & 0x0FF) == 0x020) -# define ISLOWER(c) ('a' <= (c) && (c) <= 'z') -# define TOUPPER(c) (ISLOWER(c) ? 'A' + ((c) - 'a') : (c)) -#else -# define ISLOWER(c) \ - (('a' <= (c) && (c) <= 'i') \ - || ('j' <= (c) && (c) <= 'r') \ - || ('s' <= (c) && (c) <= 'z')) -# define TOUPPER(c) (ISLOWER(c) ? ((c) | 0x40) : (c)) -#endif - -#define XOR(e, f) (((e) && !(f)) || (!(e) && (f))) -int -main () -{ - int i; - for (i = 0; i < 256; i++) - if (XOR (islower (i), ISLOWER (i)) - || toupper (i) != TOUPPER (i)) - return 2; - return 0; -} -_ACEOF -rm -f conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { ac_try='./conftest$ac_exeext' - { (case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_try") 2>&5 - ac_status=$? - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); }; }; then - : -else - $as_echo "$as_me: program exited with status $ac_status" >&5 -$as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - -( exit $ac_status ) -ac_cv_header_stdc=no -fi -rm -rf conftest.dSYM -rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext conftest.$ac_objext conftest.$ac_ext -fi - - -fi -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_header_stdc" >&5 -$as_echo "$ac_cv_header_stdc" >&6; } -if test $ac_cv_header_stdc = yes; then - -cat >>confdefs.h <<\_ACEOF -#define STDC_HEADERS 1 -_ACEOF - -fi - -# On IRIX 5.3, sys/types and inttypes.h are conflicting. - - - - - - - - - -for ac_header in sys/types.h sys/stat.h stdlib.h string.h memory.h strings.h \ - inttypes.h stdint.h unistd.h -do -as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh` -{ $as_echo "$as_me:$LINENO: checking for $ac_header" >&5 -$as_echo_n "checking for $ac_header... " >&6; } -if { as_var=$as_ac_Header; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -$ac_includes_default - -#include <$ac_header> -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - eval "$as_ac_Header=yes" -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - eval "$as_ac_Header=no" -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -fi -ac_res=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - { $as_echo "$as_me:$LINENO: result: $ac_res" >&5 -$as_echo "$ac_res" >&6; } -as_val=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - if test "x$as_val" = x""yes; then - cat >>confdefs.h <<_ACEOF -#define `$as_echo "HAVE_$ac_header" | $as_tr_cpp` 1 -_ACEOF - -fi - -done - - - -for ac_header in jpeglib.h -do -as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh` -if { as_var=$as_ac_Header; eval "test \"\${$as_var+set}\" = set"; }; then - { $as_echo "$as_me:$LINENO: checking for $ac_header" >&5 -$as_echo_n "checking for $ac_header... " >&6; } -if { as_var=$as_ac_Header; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -fi -ac_res=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - { $as_echo "$as_me:$LINENO: result: $ac_res" >&5 -$as_echo "$ac_res" >&6; } -else - # Is the header compilable? -{ $as_echo "$as_me:$LINENO: checking $ac_header usability" >&5 -$as_echo_n "checking $ac_header usability... " >&6; } -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -$ac_includes_default -#include <$ac_header> -_ACEOF -rm -f conftest.$ac_objext -if { (ac_try="$ac_compile" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_compile") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest.$ac_objext; then - ac_header_compiler=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_header_compiler=no -fi - -rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext -{ $as_echo "$as_me:$LINENO: result: $ac_header_compiler" >&5 -$as_echo "$ac_header_compiler" >&6; } - -# Is the header present? -{ $as_echo "$as_me:$LINENO: checking $ac_header presence" >&5 -$as_echo_n "checking $ac_header presence... " >&6; } -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ -#include <$ac_header> -_ACEOF -if { (ac_try="$ac_cpp conftest.$ac_ext" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_cpp conftest.$ac_ext") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } >/dev/null && { - test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || - test ! -s conftest.err - }; then - ac_header_preproc=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_header_preproc=no -fi - -rm -f conftest.err conftest.$ac_ext -{ $as_echo "$as_me:$LINENO: result: $ac_header_preproc" >&5 -$as_echo "$ac_header_preproc" >&6; } - -# So? What about this header? -case $ac_header_compiler:$ac_header_preproc:$ac_c_preproc_warn_flag in - yes:no: ) - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: accepted by the compiler, rejected by the preprocessor!" >&5 -$as_echo "$as_me: WARNING: $ac_header: accepted by the compiler, rejected by the preprocessor!" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: proceeding with the compiler's result" >&5 -$as_echo "$as_me: WARNING: $ac_header: proceeding with the compiler's result" >&2;} - ac_header_preproc=yes - ;; - no:yes:* ) - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: present but cannot be compiled" >&5 -$as_echo "$as_me: WARNING: $ac_header: present but cannot be compiled" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: check for missing prerequisite headers?" >&5 -$as_echo "$as_me: WARNING: $ac_header: check for missing prerequisite headers?" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: see the Autoconf documentation" >&5 -$as_echo "$as_me: WARNING: $ac_header: see the Autoconf documentation" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: section \"Present But Cannot Be Compiled\"" >&5 -$as_echo "$as_me: WARNING: $ac_header: section \"Present But Cannot Be Compiled\"" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: proceeding with the preprocessor's result" >&5 -$as_echo "$as_me: WARNING: $ac_header: proceeding with the preprocessor's result" >&2;} - { $as_echo "$as_me:$LINENO: WARNING: $ac_header: in the future, the compiler will take precedence" >&5 -$as_echo "$as_me: WARNING: $ac_header: in the future, the compiler will take precedence" >&2;} - ( cat <<\_ASBOX -## --------------------------------- ## -## Report this to zerofrog@gmail.com ## -## --------------------------------- ## -_ASBOX - ) | sed "s/^/$as_me: WARNING: /" >&2 - ;; -esac -{ $as_echo "$as_me:$LINENO: checking for $ac_header" >&5 -$as_echo_n "checking for $ac_header... " >&6; } -if { as_var=$as_ac_Header; eval "test \"\${$as_var+set}\" = set"; }; then - $as_echo_n "(cached) " >&6 -else - eval "$as_ac_Header=\$ac_header_preproc" -fi -ac_res=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - { $as_echo "$as_me:$LINENO: result: $ac_res" >&5 -$as_echo "$ac_res" >&6; } - -fi -as_val=`eval 'as_val=${'$as_ac_Header'} - $as_echo "$as_val"'` - if test "x$as_val" = x""yes; then - cat >>confdefs.h <<_ACEOF -#define `$as_echo "HAVE_$ac_header" | $as_tr_cpp` 1 -_ACEOF - -else - { { $as_echo "$as_me:$LINENO: error: Install libjpeg-dev" >&5 -$as_echo "$as_me: error: Install libjpeg-dev" >&2;} - { (exit 1); exit 1; }; } -fi - -done - - -{ $as_echo "$as_me:$LINENO: checking for main in -ljpeg" >&5 -$as_echo_n "checking for main in -ljpeg... " >&6; } -if test "${ac_cv_lib_jpeg_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-ljpeg $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_jpeg_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_jpeg_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_jpeg_main" >&5 -$as_echo "$ac_cv_lib_jpeg_main" >&6; } -if test "x$ac_cv_lib_jpeg_main" = x""yes; then - LIBS="$LIBS -ljpeg" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lpthread" >&5 -$as_echo_n "checking for main in -lpthread... " >&6; } -if test "${ac_cv_lib_pthread_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lpthread $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_pthread_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_pthread_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_pthread_main" >&5 -$as_echo "$ac_cv_lib_pthread_main" >&6; } -if test "x$ac_cv_lib_pthread_main" = x""yes; then - LIBS="$LIBS -lpthread" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lstdc++" >&5 -$as_echo_n "checking for main in -lstdc++... " >&6; } -if test "${ac_cv_lib_stdcpp_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lstdc++ $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_stdcpp_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_stdcpp_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_stdcpp_main" >&5 -$as_echo "$ac_cv_lib_stdcpp_main" >&6; } -if test "x$ac_cv_lib_stdcpp_main" = x""yes; then - LIBS="$LIBS -lstdc++" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lz" >&5 -$as_echo_n "checking for main in -lz... " >&6; } -if test "${ac_cv_lib_z_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lz $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_z_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_z_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_z_main" >&5 -$as_echo "$ac_cv_lib_z_main" >&6; } -if test "x$ac_cv_lib_z_main" = x""yes; then - LIBS="$LIBS -lz" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -ldl" >&5 -$as_echo_n "checking for main in -ldl... " >&6; } -if test "${ac_cv_lib_dl_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-ldl $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_dl_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_dl_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_dl_main" >&5 -$as_echo "$ac_cv_lib_dl_main" >&6; } -if test "x$ac_cv_lib_dl_main" = x""yes; then - LIBS="$LIBS -ldl" -fi - -{ $as_echo "$as_me:$LINENO: checking for main in -lXxf86vm" >&5 -$as_echo_n "checking for main in -lXxf86vm... " >&6; } -if test "${ac_cv_lib_Xxf86vm_main+set}" = set; then - $as_echo_n "(cached) " >&6 -else - ac_check_lib_save_LIBS=$LIBS -LIBS="-lXxf86vm $LIBS" -cat >conftest.$ac_ext <<_ACEOF -/* confdefs.h. */ -_ACEOF -cat confdefs.h >>conftest.$ac_ext -cat >>conftest.$ac_ext <<_ACEOF -/* end confdefs.h. */ - - -int -main () -{ -return main (); - ; - return 0; -} -_ACEOF -rm -f conftest.$ac_objext conftest$ac_exeext -if { (ac_try="$ac_link" -case "(($ac_try" in - *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; - *) ac_try_echo=$ac_try;; -esac -eval ac_try_echo="\"\$as_me:$LINENO: $ac_try_echo\"" -$as_echo "$ac_try_echo") >&5 - (eval "$ac_link") 2>conftest.er1 - ac_status=$? - grep -v '^ *+' conftest.er1 >conftest.err - rm -f conftest.er1 - cat conftest.err >&5 - $as_echo "$as_me:$LINENO: \$? = $ac_status" >&5 - (exit $ac_status); } && { - test -z "$ac_c_werror_flag" || - test ! -s conftest.err - } && test -s conftest$ac_exeext && { - test "$cross_compiling" = yes || - $as_test_x conftest$ac_exeext - }; then - ac_cv_lib_Xxf86vm_main=yes -else - $as_echo "$as_me: failed program was:" >&5 -sed 's/^/| /' conftest.$ac_ext >&5 - - ac_cv_lib_Xxf86vm_main=no -fi - -rm -rf conftest.dSYM -rm -f core conftest.err conftest.$ac_objext conftest_ipa8_conftest.oo \ - conftest$ac_exeext conftest.$ac_ext -LIBS=$ac_check_lib_save_LIBS -fi -{ $as_echo "$as_me:$LINENO: result: $ac_cv_lib_Xxf86vm_main" >&5 -$as_echo "$ac_cv_lib_Xxf86vm_main" >&6; } -if test "x$ac_cv_lib_Xxf86vm_main" = x""yes; then - LIBS="$LIBS -lXxf86vm" -else - echo "Error! You need to have libXxf86vm-dev." ; exit -1 -fi - - - -LIBS+=" -lCg -lCgGL" - -ac_config_files="$ac_config_files Makefile Linux/Makefile" - -cat >confcache <<\_ACEOF -# This file is a shell script that caches the results of configure -# tests run on this system so they can be shared between configure -# scripts and configure runs, see configure's option --config-cache. -# It is not useful on other systems. If it contains results you don't -# want to keep, you may remove or edit it. -# -# config.status only pays attention to the cache file if you give it -# the --recheck option to rerun configure. -# -# `ac_cv_env_foo' variables (set or unset) will be overridden when -# loading this file, other *unset* `ac_cv_foo' will be assigned the -# following values. - -_ACEOF - -# The following way of writing the cache mishandles newlines in values, -# but we know of no workaround that is simple, portable, and efficient. -# So, we kill variables containing newlines. -# Ultrix sh set writes to stderr and can't be redirected directly, -# and sets the high bit in the cache file unless we assign to the vars. -( - for ac_var in `(set) 2>&1 | sed -n 's/^\([a-zA-Z_][a-zA-Z0-9_]*\)=.*/\1/p'`; do - eval ac_val=\$$ac_var - case $ac_val in #( - *${as_nl}*) - case $ac_var in #( - *_cv_*) { $as_echo "$as_me:$LINENO: WARNING: cache variable $ac_var contains a newline" >&5 -$as_echo "$as_me: WARNING: cache variable $ac_var contains a newline" >&2;} ;; - esac - case $ac_var in #( - _ | IFS | as_nl) ;; #( - BASH_ARGV | BASH_SOURCE) eval $ac_var= ;; #( - *) $as_unset $ac_var ;; - esac ;; - esac - done - - (set) 2>&1 | - case $as_nl`(ac_space=' '; set) 2>&1` in #( - *${as_nl}ac_space=\ *) - # `set' does not quote correctly, so add quotes (double-quote - # substitution turns \\\\ into \\, and sed turns \\ into \). - sed -n \ - "s/'/'\\\\''/g; - s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='\\2'/p" - ;; #( - *) - # `set' quotes correctly as required by POSIX, so do not add quotes. - sed -n "/^[_$as_cr_alnum]*_cv_[_$as_cr_alnum]*=/p" - ;; - esac | - sort -) | - sed ' - /^ac_cv_env_/b end - t clear - :clear - s/^\([^=]*\)=\(.*[{}].*\)$/test "${\1+set}" = set || &/ - t end - s/^\([^=]*\)=\(.*\)$/\1=${\1=\2}/ - :end' >>confcache -if diff "$cache_file" confcache >/dev/null 2>&1; then :; else - if test -w "$cache_file"; then - test "x$cache_file" != "x/dev/null" && - { $as_echo "$as_me:$LINENO: updating cache $cache_file" >&5 -$as_echo "$as_me: updating cache $cache_file" >&6;} - cat confcache >$cache_file - else - { $as_echo "$as_me:$LINENO: not updating unwritable cache $cache_file" >&5 -$as_echo "$as_me: not updating unwritable cache $cache_file" >&6;} - fi -fi -rm -f confcache - -test "x$prefix" = xNONE && prefix=$ac_default_prefix -# Let make expand exec_prefix. -test "x$exec_prefix" = xNONE && exec_prefix='${prefix}' - -# Transform confdefs.h into DEFS. -# Protect against shell expansion while executing Makefile rules. -# Protect against Makefile macro expansion. -# -# If the first sed substitution is executed (which looks for macros that -# take arguments), then branch to the quote section. Otherwise, -# look for a macro that doesn't take arguments. -ac_script=' -:mline -/\\$/{ - N - s,\\\n,, - b mline -} -t clear -:clear -s/^[ ]*#[ ]*define[ ][ ]*\([^ (][^ (]*([^)]*)\)[ ]*\(.*\)/-D\1=\2/g -t quote -s/^[ ]*#[ ]*define[ ][ ]*\([^ ][^ ]*\)[ ]*\(.*\)/-D\1=\2/g -t quote -b any -:quote -s/[ `~#$^&*(){}\\|;'\''"<>?]/\\&/g -s/\[/\\&/g -s/\]/\\&/g -s/\$/$$/g -H -:any -${ - g - s/^\n// - s/\n/ /g - p -} -' -DEFS=`sed -n "$ac_script" confdefs.h` - - -ac_libobjs= -ac_ltlibobjs= -for ac_i in : $LIBOBJS; do test "x$ac_i" = x: && continue - # 1. Remove the extension, and $U if already installed. - ac_script='s/\$U\././;s/\.o$//;s/\.obj$//' - ac_i=`$as_echo "$ac_i" | sed "$ac_script"` - # 2. Prepend LIBOBJDIR. When used with automake>=1.10 LIBOBJDIR - # will be set to the directory where LIBOBJS objects are built. - ac_libobjs="$ac_libobjs \${LIBOBJDIR}$ac_i\$U.$ac_objext" - ac_ltlibobjs="$ac_ltlibobjs \${LIBOBJDIR}$ac_i"'$U.lo' -done -LIBOBJS=$ac_libobjs - -LTLIBOBJS=$ac_ltlibobjs - - - if test -n "$EXEEXT"; then - am__EXEEXT_TRUE= - am__EXEEXT_FALSE='#' -else - am__EXEEXT_TRUE='#' - am__EXEEXT_FALSE= -fi - -if test -z "${AMDEP_TRUE}" && test -z "${AMDEP_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"AMDEP\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"AMDEP\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi -if test -z "${am__fastdepCC_TRUE}" && test -z "${am__fastdepCC_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"am__fastdepCC\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"am__fastdepCC\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi -if test -z "${am__fastdepCXX_TRUE}" && test -z "${am__fastdepCXX_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"am__fastdepCXX\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"am__fastdepCXX\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi -if test -z "${am__fastdepCCAS_TRUE}" && test -z "${am__fastdepCCAS_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"am__fastdepCCAS\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"am__fastdepCCAS\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi -if test -z "${RELEASE_TO_PUBLIC_TRUE}" && test -z "${RELEASE_TO_PUBLIC_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"RELEASE_TO_PUBLIC\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"RELEASE_TO_PUBLIC\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi -if test -z "${SSE2_TRUE}" && test -z "${SSE2_FALSE}"; then - { { $as_echo "$as_me:$LINENO: error: conditional \"SSE2\" was never defined. -Usually this means the macro was only invoked conditionally." >&5 -$as_echo "$as_me: error: conditional \"SSE2\" was never defined. -Usually this means the macro was only invoked conditionally." >&2;} - { (exit 1); exit 1; }; } -fi - -: ${CONFIG_STATUS=./config.status} -ac_write_fail=0 -ac_clean_files_save=$ac_clean_files -ac_clean_files="$ac_clean_files $CONFIG_STATUS" -{ $as_echo "$as_me:$LINENO: creating $CONFIG_STATUS" >&5 -$as_echo "$as_me: creating $CONFIG_STATUS" >&6;} -cat >$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -#! $SHELL -# Generated by $as_me. -# Run this file to recreate the current configuration. -# Compiler output produced by configure, useful for debugging -# configure, is in config.log if it exists. - -debug=false -ac_cs_recheck=false -ac_cs_silent=false -SHELL=\${CONFIG_SHELL-$SHELL} -_ACEOF - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -## --------------------- ## -## M4sh Initialization. ## -## --------------------- ## - -# Be more Bourne compatible -DUALCASE=1; export DUALCASE # for MKS sh -if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then - emulate sh - NULLCMD=: - # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which - # is contrary to our usage. Disable this feature. - alias -g '${1+"$@"}'='"$@"' - setopt NO_GLOB_SUBST -else - case `(set -o) 2>/dev/null` in - *posix*) set -o posix ;; -esac - -fi - - - - -# PATH needs CR -# Avoid depending upon Character Ranges. -as_cr_letters='abcdefghijklmnopqrstuvwxyz' -as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ' -as_cr_Letters=$as_cr_letters$as_cr_LETTERS -as_cr_digits='0123456789' -as_cr_alnum=$as_cr_Letters$as_cr_digits - -as_nl=' -' -export as_nl -# Printing a long string crashes Solaris 7 /usr/bin/printf. -as_echo='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' -as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo -as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo$as_echo -if (test "X`printf %s $as_echo`" = "X$as_echo") 2>/dev/null; then - as_echo='printf %s\n' - as_echo_n='printf %s' -else - if test "X`(/usr/ucb/echo -n -n $as_echo) 2>/dev/null`" = "X-n $as_echo"; then - as_echo_body='eval /usr/ucb/echo -n "$1$as_nl"' - as_echo_n='/usr/ucb/echo -n' - else - as_echo_body='eval expr "X$1" : "X\\(.*\\)"' - as_echo_n_body='eval - arg=$1; - case $arg in - *"$as_nl"*) - expr "X$arg" : "X\\(.*\\)$as_nl"; - arg=`expr "X$arg" : ".*$as_nl\\(.*\\)"`;; - esac; - expr "X$arg" : "X\\(.*\\)" | tr -d "$as_nl" - ' - export as_echo_n_body - as_echo_n='sh -c $as_echo_n_body as_echo' - fi - export as_echo_body - as_echo='sh -c $as_echo_body as_echo' -fi - -# The user is always right. -if test "${PATH_SEPARATOR+set}" != set; then - PATH_SEPARATOR=: - (PATH='/bin;/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 && { - (PATH='/bin:/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 || - PATH_SEPARATOR=';' - } -fi - -# Support unset when possible. -if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then - as_unset=unset -else - as_unset=false -fi - - -# IFS -# We need space, tab and new line, in precisely that order. Quoting is -# there to prevent editors from complaining about space-tab. -# (If _AS_PATH_WALK were called with IFS unset, it would disable word -# splitting by setting IFS to empty value.) -IFS=" "" $as_nl" - -# Find who we are. Look in the path if we contain no directory separator. -case $0 in - *[\\/]* ) as_myself=$0 ;; - *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR -for as_dir in $PATH -do - IFS=$as_save_IFS - test -z "$as_dir" && as_dir=. - test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break -done -IFS=$as_save_IFS - - ;; -esac -# We did not find ourselves, most probably we were run as `sh COMMAND' -# in which case we are not to be found in the path. -if test "x$as_myself" = x; then - as_myself=$0 -fi -if test ! -f "$as_myself"; then - $as_echo "$as_myself: error: cannot find myself; rerun with an absolute file name" >&2 - { (exit 1); exit 1; } -fi - -# Work around bugs in pre-3.0 UWIN ksh. -for as_var in ENV MAIL MAILPATH -do ($as_unset $as_var) >/dev/null 2>&1 && $as_unset $as_var -done -PS1='$ ' -PS2='> ' -PS4='+ ' - -# NLS nuisances. -LC_ALL=C -export LC_ALL -LANGUAGE=C -export LANGUAGE - -# Required to use basename. -if expr a : '\(a\)' >/dev/null 2>&1 && - test "X`expr 00001 : '.*\(...\)'`" = X001; then - as_expr=expr -else - as_expr=false -fi - -if (basename -- /) >/dev/null 2>&1 && test "X`basename -- / 2>&1`" = "X/"; then - as_basename=basename -else - as_basename=false -fi - - -# Name of the executable. -as_me=`$as_basename -- "$0" || -$as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \ - X"$0" : 'X\(//\)$' \| \ - X"$0" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X/"$0" | - sed '/^.*\/\([^/][^/]*\)\/*$/{ - s//\1/ - q - } - /^X\/\(\/\/\)$/{ - s//\1/ - q - } - /^X\/\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - -# CDPATH. -$as_unset CDPATH - - - - as_lineno_1=$LINENO - as_lineno_2=$LINENO - test "x$as_lineno_1" != "x$as_lineno_2" && - test "x`expr $as_lineno_1 + 1`" = "x$as_lineno_2" || { - - # Create $as_me.lineno as a copy of $as_myself, but with $LINENO - # uniformly replaced by the line number. The first 'sed' inserts a - # line-number line after each line using $LINENO; the second 'sed' - # does the real work. The second script uses 'N' to pair each - # line-number line with the line containing $LINENO, and appends - # trailing '-' during substitution so that $LINENO is not a special - # case at line end. - # (Raja R Harinath suggested sed '=', and Paul Eggert wrote the - # scripts with optimization help from Paolo Bonzini. Blame Lee - # E. McMahon (1931-1989) for sed's syntax. :-) - sed -n ' - p - /[$]LINENO/= - ' <$as_myself | - sed ' - s/[$]LINENO.*/&-/ - t lineno - b - :lineno - N - :loop - s/[$]LINENO\([^'$as_cr_alnum'_].*\n\)\(.*\)/\2\1\2/ - t loop - s/-\n.*// - ' >$as_me.lineno && - chmod +x "$as_me.lineno" || - { $as_echo "$as_me: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&2 - { (exit 1); exit 1; }; } - - # Don't try to exec as it changes $[0], causing all sort of problems - # (the dirname of $[0] is not the place where we might find the - # original and so on. Autoconf is especially sensitive to this). - . "./$as_me.lineno" - # Exit status is that of the last command. - exit -} - - -if (as_dir=`dirname -- /` && test "X$as_dir" = X/) >/dev/null 2>&1; then - as_dirname=dirname -else - as_dirname=false -fi - -ECHO_C= ECHO_N= ECHO_T= -case `echo -n x` in --n*) - case `echo 'x\c'` in - *c*) ECHO_T=' ';; # ECHO_T is single tab character. - *) ECHO_C='\c';; - esac;; -*) - ECHO_N='-n';; -esac -if expr a : '\(a\)' >/dev/null 2>&1 && - test "X`expr 00001 : '.*\(...\)'`" = X001; then - as_expr=expr -else - as_expr=false -fi - -rm -f conf$$ conf$$.exe conf$$.file -if test -d conf$$.dir; then - rm -f conf$$.dir/conf$$.file -else - rm -f conf$$.dir - mkdir conf$$.dir 2>/dev/null -fi -if (echo >conf$$.file) 2>/dev/null; then - if ln -s conf$$.file conf$$ 2>/dev/null; then - as_ln_s='ln -s' - # ... but there are two gotchas: - # 1) On MSYS, both `ln -s file dir' and `ln file dir' fail. - # 2) DJGPP < 2.04 has no symlinks; `ln -s' creates a wrapper executable. - # In both cases, we have to default to `cp -p'. - ln -s conf$$.file conf$$.dir 2>/dev/null && test ! -f conf$$.exe || - as_ln_s='cp -p' - elif ln conf$$.file conf$$ 2>/dev/null; then - as_ln_s=ln - else - as_ln_s='cp -p' - fi -else - as_ln_s='cp -p' -fi -rm -f conf$$ conf$$.exe conf$$.dir/conf$$.file conf$$.file -rmdir conf$$.dir 2>/dev/null - -if mkdir -p . 2>/dev/null; then - as_mkdir_p=: -else - test -d ./-p && rmdir ./-p - as_mkdir_p=false -fi - -if test -x / >/dev/null 2>&1; then - as_test_x='test -x' -else - if ls -dL / >/dev/null 2>&1; then - as_ls_L_option=L - else - as_ls_L_option= - fi - as_test_x=' - eval sh -c '\'' - if test -d "$1"; then - test -d "$1/."; - else - case $1 in - -*)set "./$1";; - esac; - case `ls -ld'$as_ls_L_option' "$1" 2>/dev/null` in - ???[sx]*):;;*)false;;esac;fi - '\'' sh - ' -fi -as_executable_p=$as_test_x - -# Sed expression to map a string onto a valid CPP name. -as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'" - -# Sed expression to map a string onto a valid variable name. -as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'" - - -exec 6>&1 - -# Save the log message, to keep $[0] and so on meaningful, and to -# report actual input values of CONFIG_FILES etc. instead of their -# values after options handling. -ac_log=" -This file was extended by ZeroGSogl $as_me 0.96.2, which was -generated by GNU Autoconf 2.63. Invocation command line was - - CONFIG_FILES = $CONFIG_FILES - CONFIG_HEADERS = $CONFIG_HEADERS - CONFIG_LINKS = $CONFIG_LINKS - CONFIG_COMMANDS = $CONFIG_COMMANDS - $ $0 $@ - -on `(hostname || uname -n) 2>/dev/null | sed 1q` -" - -_ACEOF - -case $ac_config_files in *" -"*) set x $ac_config_files; shift; ac_config_files=$*;; -esac - - - -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -# Files that config.status was made for. -config_files="$ac_config_files" -config_commands="$ac_config_commands" - -_ACEOF - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -ac_cs_usage="\ -\`$as_me' instantiates files from templates according to the -current configuration. - -Usage: $0 [OPTION]... [FILE]... - - -h, --help print this help, then exit - -V, --version print version number and configuration settings, then exit - -q, --quiet, --silent - do not print progress messages - -d, --debug don't remove temporary files - --recheck update $as_me by reconfiguring in the same conditions - --file=FILE[:TEMPLATE] - instantiate the configuration file FILE - -Configuration files: -$config_files - -Configuration commands: -$config_commands - -Report bugs to ." - -_ACEOF -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -ac_cs_version="\\ -ZeroGSogl config.status 0.96.2 -configured by $0, generated by GNU Autoconf 2.63, - with options \\"`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`\\" - -Copyright (C) 2008 Free Software Foundation, Inc. -This config.status script is free software; the Free Software Foundation -gives unlimited permission to copy, distribute and modify it." - -ac_pwd='$ac_pwd' -srcdir='$srcdir' -INSTALL='$INSTALL' -MKDIR_P='$MKDIR_P' -AWK='$AWK' -test -n "\$AWK" || AWK=awk -_ACEOF - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -# The default lists apply if the user does not specify any file. -ac_need_defaults=: -while test $# != 0 -do - case $1 in - --*=*) - ac_option=`expr "X$1" : 'X\([^=]*\)='` - ac_optarg=`expr "X$1" : 'X[^=]*=\(.*\)'` - ac_shift=: - ;; - *) - ac_option=$1 - ac_optarg=$2 - ac_shift=shift - ;; - esac - - case $ac_option in - # Handling of the options. - -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) - ac_cs_recheck=: ;; - --version | --versio | --versi | --vers | --ver | --ve | --v | -V ) - $as_echo "$ac_cs_version"; exit ;; - --debug | --debu | --deb | --de | --d | -d ) - debug=: ;; - --file | --fil | --fi | --f ) - $ac_shift - case $ac_optarg in - *\'*) ac_optarg=`$as_echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"` ;; - esac - CONFIG_FILES="$CONFIG_FILES '$ac_optarg'" - ac_need_defaults=false;; - --he | --h | --help | --hel | -h ) - $as_echo "$ac_cs_usage"; exit ;; - -q | -quiet | --quiet | --quie | --qui | --qu | --q \ - | -silent | --silent | --silen | --sile | --sil | --si | --s) - ac_cs_silent=: ;; - - # This is an error. - -*) { $as_echo "$as_me: error: unrecognized option: $1 -Try \`$0 --help' for more information." >&2 - { (exit 1); exit 1; }; } ;; - - *) ac_config_targets="$ac_config_targets $1" - ac_need_defaults=false ;; - - esac - shift -done - -ac_configure_extra_args= - -if $ac_cs_silent; then - exec 6>/dev/null - ac_configure_extra_args="$ac_configure_extra_args --silent" -fi - -_ACEOF -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -if \$ac_cs_recheck; then - set X '$SHELL' '$0' $ac_configure_args \$ac_configure_extra_args --no-create --no-recursion - shift - \$as_echo "running CONFIG_SHELL=$SHELL \$*" >&6 - CONFIG_SHELL='$SHELL' - export CONFIG_SHELL - exec "\$@" -fi - -_ACEOF -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -exec 5>>config.log -{ - echo - sed 'h;s/./-/g;s/^.../## /;s/...$/ ##/;p;x;p;x' <<_ASBOX -## Running $as_me. ## -_ASBOX - $as_echo "$ac_log" -} >&5 - -_ACEOF -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -# -# INIT-COMMANDS -# -AMDEP_TRUE="$AMDEP_TRUE" ac_aux_dir="$ac_aux_dir" - -_ACEOF - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 - -# Handling of arguments. -for ac_config_target in $ac_config_targets -do - case $ac_config_target in - "depfiles") CONFIG_COMMANDS="$CONFIG_COMMANDS depfiles" ;; - "Makefile") CONFIG_FILES="$CONFIG_FILES Makefile" ;; - "Linux/Makefile") CONFIG_FILES="$CONFIG_FILES Linux/Makefile" ;; - - *) { { $as_echo "$as_me:$LINENO: error: invalid argument: $ac_config_target" >&5 -$as_echo "$as_me: error: invalid argument: $ac_config_target" >&2;} - { (exit 1); exit 1; }; };; - esac -done - - -# If the user did not use the arguments to specify the items to instantiate, -# then the envvar interface is used. Set only those that are not. -# We use the long form for the default assignment because of an extremely -# bizarre bug on SunOS 4.1.3. -if $ac_need_defaults; then - test "${CONFIG_FILES+set}" = set || CONFIG_FILES=$config_files - test "${CONFIG_COMMANDS+set}" = set || CONFIG_COMMANDS=$config_commands -fi - -# Have a temporary directory for convenience. Make it in the build tree -# simply because there is no reason against having it here, and in addition, -# creating and moving files from /tmp can sometimes cause problems. -# Hook for its removal unless debugging. -# Note that there is a small window in which the directory will not be cleaned: -# after its creation but before its name has been assigned to `$tmp'. -$debug || -{ - tmp= - trap 'exit_status=$? - { test -z "$tmp" || test ! -d "$tmp" || rm -fr "$tmp"; } && exit $exit_status -' 0 - trap '{ (exit 1); exit 1; }' 1 2 13 15 -} -# Create a (secure) tmp directory for tmp files. - -{ - tmp=`(umask 077 && mktemp -d "./confXXXXXX") 2>/dev/null` && - test -n "$tmp" && test -d "$tmp" -} || -{ - tmp=./conf$$-$RANDOM - (umask 077 && mkdir "$tmp") -} || -{ - $as_echo "$as_me: cannot create a temporary directory in ." >&2 - { (exit 1); exit 1; } -} - -# Set up the scripts for CONFIG_FILES section. -# No need to generate them if there are no CONFIG_FILES. -# This happens for instance with `./config.status config.h'. -if test -n "$CONFIG_FILES"; then - - -ac_cr=' ' -ac_cs_awk_cr=`$AWK 'BEGIN { print "a\rb" }' /dev/null` -if test "$ac_cs_awk_cr" = "a${ac_cr}b"; then - ac_cs_awk_cr='\\r' -else - ac_cs_awk_cr=$ac_cr -fi - -echo 'BEGIN {' >"$tmp/subs1.awk" && -_ACEOF - - -{ - echo "cat >conf$$subs.awk <<_ACEOF" && - echo "$ac_subst_vars" | sed 's/.*/&!$&$ac_delim/' && - echo "_ACEOF" -} >conf$$subs.sh || - { { $as_echo "$as_me:$LINENO: error: could not make $CONFIG_STATUS" >&5 -$as_echo "$as_me: error: could not make $CONFIG_STATUS" >&2;} - { (exit 1); exit 1; }; } -ac_delim_num=`echo "$ac_subst_vars" | grep -c '$'` -ac_delim='%!_!# ' -for ac_last_try in false false false false false :; do - . ./conf$$subs.sh || - { { $as_echo "$as_me:$LINENO: error: could not make $CONFIG_STATUS" >&5 -$as_echo "$as_me: error: could not make $CONFIG_STATUS" >&2;} - { (exit 1); exit 1; }; } - - ac_delim_n=`sed -n "s/.*$ac_delim\$/X/p" conf$$subs.awk | grep -c X` - if test $ac_delim_n = $ac_delim_num; then - break - elif $ac_last_try; then - { { $as_echo "$as_me:$LINENO: error: could not make $CONFIG_STATUS" >&5 -$as_echo "$as_me: error: could not make $CONFIG_STATUS" >&2;} - { (exit 1); exit 1; }; } - else - ac_delim="$ac_delim!$ac_delim _$ac_delim!! " - fi -done -rm -f conf$$subs.sh - -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -cat >>"\$tmp/subs1.awk" <<\\_ACAWK && -_ACEOF -sed -n ' -h -s/^/S["/; s/!.*/"]=/ -p -g -s/^[^!]*!// -:repl -t repl -s/'"$ac_delim"'$// -t delim -:nl -h -s/\(.\{148\}\).*/\1/ -t more1 -s/["\\]/\\&/g; s/^/"/; s/$/\\n"\\/ -p -n -b repl -:more1 -s/["\\]/\\&/g; s/^/"/; s/$/"\\/ -p -g -s/.\{148\}// -t nl -:delim -h -s/\(.\{148\}\).*/\1/ -t more2 -s/["\\]/\\&/g; s/^/"/; s/$/"/ -p -b -:more2 -s/["\\]/\\&/g; s/^/"/; s/$/"\\/ -p -g -s/.\{148\}// -t delim -' >$CONFIG_STATUS || ac_write_fail=1 -rm -f conf$$subs.awk -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -_ACAWK -cat >>"\$tmp/subs1.awk" <<_ACAWK && - for (key in S) S_is_set[key] = 1 - FS = "" - -} -{ - line = $ 0 - nfields = split(line, field, "@") - substed = 0 - len = length(field[1]) - for (i = 2; i < nfields; i++) { - key = field[i] - keylen = length(key) - if (S_is_set[key]) { - value = S[key] - line = substr(line, 1, len) "" value "" substr(line, len + keylen + 3) - len += length(value) + length(field[++i]) - substed = 1 - } else - len += 1 + keylen - } - - print line -} - -_ACAWK -_ACEOF -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -if sed "s/$ac_cr//" < /dev/null > /dev/null 2>&1; then - sed "s/$ac_cr\$//; s/$ac_cr/$ac_cs_awk_cr/g" -else - cat -fi < "$tmp/subs1.awk" > "$tmp/subs.awk" \ - || { { $as_echo "$as_me:$LINENO: error: could not setup config files machinery" >&5 -$as_echo "$as_me: error: could not setup config files machinery" >&2;} - { (exit 1); exit 1; }; } -_ACEOF - -# VPATH may cause trouble with some makes, so we remove $(srcdir), -# ${srcdir} and @srcdir@ from VPATH if srcdir is ".", strip leading and -# trailing colons and then remove the whole line if VPATH becomes empty -# (actually we leave an empty line to preserve line numbers). -if test "x$srcdir" = x.; then - ac_vpsub='/^[ ]*VPATH[ ]*=/{ -s/:*\$(srcdir):*/:/ -s/:*\${srcdir}:*/:/ -s/:*@srcdir@:*/:/ -s/^\([^=]*=[ ]*\):*/\1/ -s/:*$// -s/^[^=]*=[ ]*$// -}' -fi - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -fi # test -n "$CONFIG_FILES" - - -eval set X " :F $CONFIG_FILES :C $CONFIG_COMMANDS" -shift -for ac_tag -do - case $ac_tag in - :[FHLC]) ac_mode=$ac_tag; continue;; - esac - case $ac_mode$ac_tag in - :[FHL]*:*);; - :L* | :C*:*) { { $as_echo "$as_me:$LINENO: error: invalid tag $ac_tag" >&5 -$as_echo "$as_me: error: invalid tag $ac_tag" >&2;} - { (exit 1); exit 1; }; };; - :[FH]-) ac_tag=-:-;; - :[FH]*) ac_tag=$ac_tag:$ac_tag.in;; - esac - ac_save_IFS=$IFS - IFS=: - set x $ac_tag - IFS=$ac_save_IFS - shift - ac_file=$1 - shift - - case $ac_mode in - :L) ac_source=$1;; - :[FH]) - ac_file_inputs= - for ac_f - do - case $ac_f in - -) ac_f="$tmp/stdin";; - *) # Look for the file first in the build tree, then in the source tree - # (if the path is not absolute). The absolute path cannot be DOS-style, - # because $ac_f cannot contain `:'. - test -f "$ac_f" || - case $ac_f in - [\\/$]*) false;; - *) test -f "$srcdir/$ac_f" && ac_f="$srcdir/$ac_f";; - esac || - { { $as_echo "$as_me:$LINENO: error: cannot find input file: $ac_f" >&5 -$as_echo "$as_me: error: cannot find input file: $ac_f" >&2;} - { (exit 1); exit 1; }; };; - esac - case $ac_f in *\'*) ac_f=`$as_echo "$ac_f" | sed "s/'/'\\\\\\\\''/g"`;; esac - ac_file_inputs="$ac_file_inputs '$ac_f'" - done - - # Let's still pretend it is `configure' which instantiates (i.e., don't - # use $as_me), people would be surprised to read: - # /* config.h. Generated by config.status. */ - configure_input='Generated from '` - $as_echo "$*" | sed 's|^[^:]*/||;s|:[^:]*/|, |g' - `' by configure.' - if test x"$ac_file" != x-; then - configure_input="$ac_file. $configure_input" - { $as_echo "$as_me:$LINENO: creating $ac_file" >&5 -$as_echo "$as_me: creating $ac_file" >&6;} - fi - # Neutralize special characters interpreted by sed in replacement strings. - case $configure_input in #( - *\&* | *\|* | *\\* ) - ac_sed_conf_input=`$as_echo "$configure_input" | - sed 's/[\\\\&|]/\\\\&/g'`;; #( - *) ac_sed_conf_input=$configure_input;; - esac - - case $ac_tag in - *:-:* | *:-) cat >"$tmp/stdin" \ - || { { $as_echo "$as_me:$LINENO: error: could not create $ac_file" >&5 -$as_echo "$as_me: error: could not create $ac_file" >&2;} - { (exit 1); exit 1; }; } ;; - esac - ;; - esac - - ac_dir=`$as_dirname -- "$ac_file" || -$as_expr X"$ac_file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$ac_file" : 'X\(//\)[^/]' \| \ - X"$ac_file" : 'X\(//\)$' \| \ - X"$ac_file" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$ac_file" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - { as_dir="$ac_dir" - case $as_dir in #( - -*) as_dir=./$as_dir;; - esac - test -d "$as_dir" || { $as_mkdir_p && mkdir -p "$as_dir"; } || { - as_dirs= - while :; do - case $as_dir in #( - *\'*) as_qdir=`$as_echo "$as_dir" | sed "s/'/'\\\\\\\\''/g"`;; #'( - *) as_qdir=$as_dir;; - esac - as_dirs="'$as_qdir' $as_dirs" - as_dir=`$as_dirname -- "$as_dir" || -$as_expr X"$as_dir" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$as_dir" : 'X\(//\)[^/]' \| \ - X"$as_dir" : 'X\(//\)$' \| \ - X"$as_dir" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$as_dir" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - test -d "$as_dir" && break - done - test -z "$as_dirs" || eval "mkdir $as_dirs" - } || test -d "$as_dir" || { { $as_echo "$as_me:$LINENO: error: cannot create directory $as_dir" >&5 -$as_echo "$as_me: error: cannot create directory $as_dir" >&2;} - { (exit 1); exit 1; }; }; } - ac_builddir=. - -case "$ac_dir" in -.) ac_dir_suffix= ac_top_builddir_sub=. ac_top_build_prefix= ;; -*) - ac_dir_suffix=/`$as_echo "$ac_dir" | sed 's|^\.[\\/]||'` - # A ".." for each directory in $ac_dir_suffix. - ac_top_builddir_sub=`$as_echo "$ac_dir_suffix" | sed 's|/[^\\/]*|/..|g;s|/||'` - case $ac_top_builddir_sub in - "") ac_top_builddir_sub=. ac_top_build_prefix= ;; - *) ac_top_build_prefix=$ac_top_builddir_sub/ ;; - esac ;; -esac -ac_abs_top_builddir=$ac_pwd -ac_abs_builddir=$ac_pwd$ac_dir_suffix -# for backward compatibility: -ac_top_builddir=$ac_top_build_prefix - -case $srcdir in - .) # We are building in place. - ac_srcdir=. - ac_top_srcdir=$ac_top_builddir_sub - ac_abs_top_srcdir=$ac_pwd ;; - [\\/]* | ?:[\\/]* ) # Absolute name. - ac_srcdir=$srcdir$ac_dir_suffix; - ac_top_srcdir=$srcdir - ac_abs_top_srcdir=$srcdir ;; - *) # Relative name. - ac_srcdir=$ac_top_build_prefix$srcdir$ac_dir_suffix - ac_top_srcdir=$ac_top_build_prefix$srcdir - ac_abs_top_srcdir=$ac_pwd/$srcdir ;; -esac -ac_abs_srcdir=$ac_abs_top_srcdir$ac_dir_suffix - - - case $ac_mode in - :F) - # - # CONFIG_FILE - # - - case $INSTALL in - [\\/$]* | ?:[\\/]* ) ac_INSTALL=$INSTALL ;; - *) ac_INSTALL=$ac_top_build_prefix$INSTALL ;; - esac - ac_MKDIR_P=$MKDIR_P - case $MKDIR_P in - [\\/$]* | ?:[\\/]* ) ;; - */*) ac_MKDIR_P=$ac_top_build_prefix$MKDIR_P ;; - esac -_ACEOF - -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -# If the template does not know about datarootdir, expand it. -# FIXME: This hack should be removed a few years after 2.60. -ac_datarootdir_hack=; ac_datarootdir_seen= - -ac_sed_dataroot=' -/datarootdir/ { - p - q -} -/@datadir@/p -/@docdir@/p -/@infodir@/p -/@localedir@/p -/@mandir@/p -' -case `eval "sed -n \"\$ac_sed_dataroot\" $ac_file_inputs"` in -*datarootdir*) ac_datarootdir_seen=yes;; -*@datadir@*|*@docdir@*|*@infodir@*|*@localedir@*|*@mandir@*) - { $as_echo "$as_me:$LINENO: WARNING: $ac_file_inputs seems to ignore the --datarootdir setting" >&5 -$as_echo "$as_me: WARNING: $ac_file_inputs seems to ignore the --datarootdir setting" >&2;} -_ACEOF -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 - ac_datarootdir_hack=' - s&@datadir@&$datadir&g - s&@docdir@&$docdir&g - s&@infodir@&$infodir&g - s&@localedir@&$localedir&g - s&@mandir@&$mandir&g - s&\\\${datarootdir}&$datarootdir&g' ;; -esac -_ACEOF - -# Neutralize VPATH when `$srcdir' = `.'. -# Shell code in configure.ac might set extrasub. -# FIXME: do we really want to maintain this feature? -cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 -ac_sed_extra="$ac_vpsub -$extrasub -_ACEOF -cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 -:t -/@[a-zA-Z_][a-zA-Z_0-9]*@/!b -s|@configure_input@|$ac_sed_conf_input|;t t -s&@top_builddir@&$ac_top_builddir_sub&;t t -s&@top_build_prefix@&$ac_top_build_prefix&;t t -s&@srcdir@&$ac_srcdir&;t t -s&@abs_srcdir@&$ac_abs_srcdir&;t t -s&@top_srcdir@&$ac_top_srcdir&;t t -s&@abs_top_srcdir@&$ac_abs_top_srcdir&;t t -s&@builddir@&$ac_builddir&;t t -s&@abs_builddir@&$ac_abs_builddir&;t t -s&@abs_top_builddir@&$ac_abs_top_builddir&;t t -s&@INSTALL@&$ac_INSTALL&;t t -s&@MKDIR_P@&$ac_MKDIR_P&;t t -$ac_datarootdir_hack -" -eval sed \"\$ac_sed_extra\" "$ac_file_inputs" | $AWK -f "$tmp/subs.awk" >$tmp/out \ - || { { $as_echo "$as_me:$LINENO: error: could not create $ac_file" >&5 -$as_echo "$as_me: error: could not create $ac_file" >&2;} - { (exit 1); exit 1; }; } - -test -z "$ac_datarootdir_hack$ac_datarootdir_seen" && - { ac_out=`sed -n '/\${datarootdir}/p' "$tmp/out"`; test -n "$ac_out"; } && - { ac_out=`sed -n '/^[ ]*datarootdir[ ]*:*=/p' "$tmp/out"`; test -z "$ac_out"; } && - { $as_echo "$as_me:$LINENO: WARNING: $ac_file contains a reference to the variable \`datarootdir' -which seems to be undefined. Please make sure it is defined." >&5 -$as_echo "$as_me: WARNING: $ac_file contains a reference to the variable \`datarootdir' -which seems to be undefined. Please make sure it is defined." >&2;} - - rm -f "$tmp/stdin" - case $ac_file in - -) cat "$tmp/out" && rm -f "$tmp/out";; - *) rm -f "$ac_file" && mv "$tmp/out" "$ac_file";; - esac \ - || { { $as_echo "$as_me:$LINENO: error: could not create $ac_file" >&5 -$as_echo "$as_me: error: could not create $ac_file" >&2;} - { (exit 1); exit 1; }; } - ;; - - - :C) { $as_echo "$as_me:$LINENO: executing $ac_file commands" >&5 -$as_echo "$as_me: executing $ac_file commands" >&6;} - ;; - esac - - - case $ac_file$ac_mode in - "depfiles":C) test x"$AMDEP_TRUE" != x"" || { - # Autoconf 2.62 quotes --file arguments for eval, but not when files - # are listed without --file. Let's play safe and only enable the eval - # if we detect the quoting. - case $CONFIG_FILES in - *\'*) eval set x "$CONFIG_FILES" ;; - *) set x $CONFIG_FILES ;; - esac - shift - for mf - do - # Strip MF so we end up with the name of the file. - mf=`echo "$mf" | sed -e 's/:.*$//'` - # Check whether this is an Automake generated Makefile or not. - # We used to match only the files named `Makefile.in', but - # some people rename them; so instead we look at the file content. - # Grep'ing the first line is not enough: some people post-process - # each Makefile.in and add a new line on top of each file to say so. - # Grep'ing the whole file is not good either: AIX grep has a line - # limit of 2048, but all sed's we know have understand at least 4000. - if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then - dirpart=`$as_dirname -- "$mf" || -$as_expr X"$mf" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$mf" : 'X\(//\)[^/]' \| \ - X"$mf" : 'X\(//\)$' \| \ - X"$mf" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$mf" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - else - continue - fi - # Extract the definition of DEPDIR, am__include, and am__quote - # from the Makefile without running `make'. - DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"` - test -z "$DEPDIR" && continue - am__include=`sed -n 's/^am__include = //p' < "$mf"` - test -z "am__include" && continue - am__quote=`sed -n 's/^am__quote = //p' < "$mf"` - # When using ansi2knr, U may be empty or an underscore; expand it - U=`sed -n 's/^U = //p' < "$mf"` - # Find all dependency output files, they are included files with - # $(DEPDIR) in their names. We invoke sed twice because it is the - # simplest approach to changing $(DEPDIR) to its actual value in the - # expansion. - for file in `sed -n " - s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \ - sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g' -e 's/\$U/'"$U"'/g'`; do - # Make sure the directory exists. - test -f "$dirpart/$file" && continue - fdir=`$as_dirname -- "$file" || -$as_expr X"$file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$file" : 'X\(//\)[^/]' \| \ - X"$file" : 'X\(//\)$' \| \ - X"$file" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$file" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - { as_dir=$dirpart/$fdir - case $as_dir in #( - -*) as_dir=./$as_dir;; - esac - test -d "$as_dir" || { $as_mkdir_p && mkdir -p "$as_dir"; } || { - as_dirs= - while :; do - case $as_dir in #( - *\'*) as_qdir=`$as_echo "$as_dir" | sed "s/'/'\\\\\\\\''/g"`;; #'( - *) as_qdir=$as_dir;; - esac - as_dirs="'$as_qdir' $as_dirs" - as_dir=`$as_dirname -- "$as_dir" || -$as_expr X"$as_dir" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ - X"$as_dir" : 'X\(//\)[^/]' \| \ - X"$as_dir" : 'X\(//\)$' \| \ - X"$as_dir" : 'X\(/\)' \| . 2>/dev/null || -$as_echo X"$as_dir" | - sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ - s//\1/ - q - } - /^X\(\/\/\)[^/].*/{ - s//\1/ - q - } - /^X\(\/\/\)$/{ - s//\1/ - q - } - /^X\(\/\).*/{ - s//\1/ - q - } - s/.*/./; q'` - test -d "$as_dir" && break - done - test -z "$as_dirs" || eval "mkdir $as_dirs" - } || test -d "$as_dir" || { { $as_echo "$as_me:$LINENO: error: cannot create directory $as_dir" >&5 -$as_echo "$as_me: error: cannot create directory $as_dir" >&2;} - { (exit 1); exit 1; }; }; } - # echo "creating $dirpart/$file" - echo '# dummy' > "$dirpart/$file" - done - done -} - ;; - - esac -done # for ac_tag - - -{ (exit 0); exit 0; } -_ACEOF -chmod +x $CONFIG_STATUS -ac_clean_files=$ac_clean_files_save - -test $ac_write_fail = 0 || - { { $as_echo "$as_me:$LINENO: error: write failure creating $CONFIG_STATUS" >&5 -$as_echo "$as_me: error: write failure creating $CONFIG_STATUS" >&2;} - { (exit 1); exit 1; }; } - - -# configure is writing to config.log, and then calls config.status. -# config.status does its own redirection, appending to config.log. -# Unfortunately, on DOS this fails, as config.log is still kept open -# by configure, so config.status won't be able to write to it; its -# output is simply discarded. So we exec the FD to /dev/null, -# effectively closing config.log, so it can be properly (re)opened and -# appended to by config.status. When coming back to configure, we -# need to make the FD available again. -if test "$no_create" != yes; then - ac_cs_success=: - ac_config_status_args= - test "$silent" = yes && - ac_config_status_args="$ac_config_status_args --quiet" - exec 5>/dev/null - $SHELL $CONFIG_STATUS $ac_config_status_args || ac_cs_success=false - exec 5>>config.log - # Use ||, not &&, to avoid exiting from the if with $? = 1, which - # would make configure fail if this is the last instruction. - $ac_cs_success || { (exit 1); exit 1; } -fi -if test -n "$ac_unrecognized_opts" && test "$enable_option_checking" != no; then - { $as_echo "$as_me:$LINENO: WARNING: unrecognized options: $ac_unrecognized_opts" >&5 -$as_echo "$as_me: WARNING: unrecognized options: $ac_unrecognized_opts" >&2;} -fi - - -echo "Configuration:" -echo " Target system type: $target" -echo " x86-64 build? $cpu64" -echo " SSE2 enabled? $sse2" -echo " Debug build? $debug" -echo " Dev build? $devbuild" From 136716a4f59bad75fac139830ca3e2ac0b69e1f3 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Mon, 15 Jun 2009 16:44:26 +0000 Subject: [PATCH 222/296] SPU2-X: * Fixed crash problems when using SPU2-X with old versions of Pcsx2 (0.9.4 and prior). * Fixed broken Device specification override (the device GUID wasn't being loaded from the INI). Pcsx2: * Added ThreadAffinity stuff to the CPUSpeed detection. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1368 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/ix86/ix86_cpudetect.cpp | 57 ++++++++++++++++---- plugins/spu2-x/src/DllInterface.cpp | 17 +++--- plugins/spu2-x/src/Dma.cpp | 4 +- plugins/spu2-x/src/Spu2.cpp | 15 +++--- plugins/spu2-x/src/Spu2.h | 5 +- plugins/spu2-x/src/Windows/CfgHelpers.cpp | 2 +- plugins/spu2-x/src/Windows/Dialogs.h | 2 +- plugins/spu2-x/src/Windows/SndOut_DSound.cpp | 4 -- 8 files changed, 71 insertions(+), 35 deletions(-) diff --git a/pcsx2/x86/ix86/ix86_cpudetect.cpp b/pcsx2/x86/ix86/ix86_cpudetect.cpp index 26a0dbb63c..34138f9ccd 100644 --- a/pcsx2/x86/ix86/ix86_cpudetect.cpp +++ b/pcsx2/x86/ix86/ix86_cpudetect.cpp @@ -180,20 +180,54 @@ static char* bool_to_char( bool testcond ) #endif +#ifdef _WINDOWS_ +static HANDLE s_threadId = NULL; +static DWORD s_oldmask = ERROR_INVALID_PARAMETER; +#endif + +static void SetSingleAffinity() +{ +#ifdef _WINDOWS_ + // Assign a single CPU thread affinity to ensure rdtsc() accuracy. + // (rdtsc for each CPU/core can differ, causing skewed results) + + DWORD_PTR availProcCpus, availSysCpus; + if( !GetProcessAffinityMask( GetCurrentProcess(), &availProcCpus, &availSysCpus ) ) return; + + int i; + for( i=0; i<32; ++i ) + { + if( availProcCpus & (1<>16 == 0x1f80) { @@ -430,7 +429,9 @@ EXPORT_C_(void) SPU2write(u32 rmem, u16 value) // If the SPU2 isn't in in sync with the IOP, samples can end up playing at rather // incorrect pitches and loop lengths. - TimeUpdate( *cPtr ); + if( cyclePtr != NULL ) + TimeUpdate( *cyclePtr ); + if (rmem>>16 == 0x1f80) SPU_ps1_write(rmem,value); else diff --git a/plugins/spu2-x/src/Dma.cpp b/plugins/spu2-x/src/Dma.cpp index b38bcb7149..31edd19ffc 100644 --- a/plugins/spu2-x/src/Dma.cpp +++ b/plugins/spu2-x/src/Dma.cpp @@ -313,7 +313,7 @@ void DoDMAWrite(int core,u16 *pMem,u32 size) void SPU2readDMA(int core, u16* pMem, u32 size) { - if(hasPtr) TimeUpdate(*cPtr); + if( cyclePtr != NULL ) TimeUpdate( *cyclePtr ); Cores[core].TSA &= 0xffff8; @@ -394,7 +394,7 @@ void SPU2readDMA(int core, u16* pMem, u32 size) void SPU2writeDMA(int core, u16* pMem, u32 size) { - if(hasPtr) TimeUpdate(*cPtr); + if(cyclePtr != NULL) TimeUpdate(*cyclePtr); Cores[core].DMAPtr=pMem; diff --git a/plugins/spu2-x/src/Spu2.cpp b/plugins/spu2-x/src/Spu2.cpp index cccd18f8ce..7c21143404 100644 --- a/plugins/spu2-x/src/Spu2.cpp +++ b/plugins/spu2-x/src/Spu2.cpp @@ -53,10 +53,8 @@ s16 OutPos; s16 InputPos; u32 Cycles; -u32* cPtr=NULL; -u32 lClocks=0; - -bool hasPtr=false; +u32* cyclePtr = NULL; +u32 lClocks = 0; int PlayMode; @@ -932,10 +930,13 @@ __forceinline void SPU2_FastWrite( u32 rmem, u16 value ) if( ((value>>15)&1) && (!thiscore.CoreEnabled) && (thiscore.InitDelay==0) ) // on init/reset { - if(hasPtr) + // When we have exact cycle update info from the Pcsx2 IOP unit, then use + // the more accurate delayed initialization system. + + if(cyclePtr != NULL) { - thiscore.InitDelay=1; - thiscore.Regs.STATX=0; + thiscore.InitDelay = 1; + thiscore.Regs.STATX = 0; } else { diff --git a/plugins/spu2-x/src/Spu2.h b/plugins/spu2-x/src/Spu2.h index 9d3bcaf70c..f04b797450 100644 --- a/plugins/spu2-x/src/Spu2.h +++ b/plugins/spu2-x/src/Spu2.h @@ -174,13 +174,12 @@ extern int PlayMode; extern int recording; extern u32 lClocks; -extern u32* cPtr; -extern bool hasPtr; +extern u32* cyclePtr; extern void SPU2writeLog( const char* action, u32 rmem, u16 value ); extern void TimeUpdate(u32 cClocks); -extern u16 SPU_ps1_read(u32 mem); +extern u16 SPU_ps1_read(u32 mem); extern void SPU_ps1_write(u32 mem, u16 value); extern void SPU2_FastWrite( u32 rmem, u16 value ); diff --git a/plugins/spu2-x/src/Windows/CfgHelpers.cpp b/plugins/spu2-x/src/Windows/CfgHelpers.cpp index 8b5a4f5194..ac4dad8159 100644 --- a/plugins/spu2-x/src/Windows/CfgHelpers.cpp +++ b/plugins/spu2-x/src/Windows/CfgHelpers.cpp @@ -110,7 +110,7 @@ void CfgReadStr(const TCHAR* Section, const TCHAR* Name, TCHAR* Data, int DataSi } } -void CfgReadStr(const TCHAR* Section, const TCHAR* Name, wstring Data, int DataSize, const TCHAR* Default) +void CfgReadStr(const TCHAR* Section, const TCHAR* Name, wstring& Data, int DataSize, const TCHAR* Default) { wchar_t workspace[512]; GetPrivateProfileString(Section,Name,L"",workspace,DataSize,CfgFile); diff --git a/plugins/spu2-x/src/Windows/Dialogs.h b/plugins/spu2-x/src/Windows/Dialogs.h index fca47d9e99..414a71db56 100644 --- a/plugins/spu2-x/src/Windows/Dialogs.h +++ b/plugins/spu2-x/src/Windows/Dialogs.h @@ -69,7 +69,7 @@ void CfgWriteInt(const TCHAR* Section, const TCHAR* Name, int Value); void CfgWriteStr(const TCHAR* Section, const TCHAR* Name, const wstring& Data); bool CfgReadBool(const TCHAR *Section,const TCHAR* Name, bool Default); -void CfgReadStr(const TCHAR* Section, const TCHAR* Name, wstring Data, int DataSize, const TCHAR* Default); +void CfgReadStr(const TCHAR* Section, const TCHAR* Name, wstring& Data, int DataSize, const TCHAR* Default); void CfgReadStr(const TCHAR* Section, const TCHAR* Name, TCHAR* Data, int DataSize, const TCHAR* Default); int CfgReadInt(const TCHAR* Section, const TCHAR* Name,int Default); diff --git a/plugins/spu2-x/src/Windows/SndOut_DSound.cpp b/plugins/spu2-x/src/Windows/SndOut_DSound.cpp index f30b6fb0a0..1715da1131 100644 --- a/plugins/spu2-x/src/Windows/SndOut_DSound.cpp +++ b/plugins/spu2-x/src/Windows/SndOut_DSound.cpp @@ -194,10 +194,6 @@ public: for(uint i=0;i Date: Mon, 15 Jun 2009 20:15:17 +0000 Subject: [PATCH 223/296] GSdx: Minor bugfix for DoMerge encountering a NULL pointer in rare cases. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1369 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDevice.cpp | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/plugins/GSdx/GSDevice.cpp b/plugins/GSdx/GSDevice.cpp index 7373028441..234c244c28 100644 --- a/plugins/GSdx/GSDevice.cpp +++ b/plugins/GSdx/GSDevice.cpp @@ -164,7 +164,11 @@ void GSDevice::Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVec // TODO: m_1x1 - DoMerge(st, sr, dr, m_merge, slbg, mmod, c); + // KH:COM crashes at startup when booting *through the bios* due to m_merge being NULL. + // (texture appears to be non-null, and is being re-created at a size around like 1700x340, + // dunno if that's relevant) -- air + if( m_merge ) + DoMerge(st, sr, dr, m_merge, slbg, mmod, c); m_current = m_merge; } From ce4b58ba61a8e956747512d7dc909202606a5de0 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 15 Jun 2009 22:28:05 +0000 Subject: [PATCH 224/296] microVU: - Minor changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1370 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 24 +++++++++++------------- pcsx2/x86/microVU_Misc.h | 13 +++++-------- pcsx2/x86/microVU_Misc.inl | 7 +++---- 3 files changed, 19 insertions(+), 25 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index 67169db9ed..abec8bcda8 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -277,21 +277,20 @@ microVUt(void) mVUanalyzeR2(mV, int Ft, bool canBeNOP) { iPC = curPC; \ } - microVUt(void) mVUanalyzeSflag(mV, int It) { + mVUlow.readFlags = 1; + analyzeVIreg2(It, mVUlow.VI_write, 1); if (!It) { mVUlow.isNOP = 1; } else { mVUinfo.swapOps = 1; mVUsFlagHack = 0; // Don't Optimize Out Status Flags for this block - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf /*<< mVUcount*/; } + flagSet(sFLAG.doNonSticky); + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf; } if (mVUcount >= 1) { incPC2(-2); mVUlow.useSflag = 1; incPC2(2); } // Note: useSflag is used for status flag optimizations when a FSSET instruction is called. // Do to stalls, it can only be set one instruction prior to the status flag read instruction // if we were guaranteed no-stalls were to happen, it could be set 4 instruction prior. } - mVUlow.readFlags = 1; - analyzeVIreg2(It, mVUlow.VI_write, 1); - flagSet(sFLAG.doNonSticky); } microVUt(void) mVUanalyzeFSSET(mV) { @@ -304,16 +303,15 @@ microVUt(void) mVUanalyzeFSSET(mV) { //------------------------------------------------------------------ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { - if (!It) { mVUlow.isNOP = 1; } - else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) - mVUinfo.swapOps = 1; - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 4); } - - } mVUlow.readFlags = 1; analyzeVIreg1(Is, mVUlow.VI_read[0]); analyzeVIreg2(It, mVUlow.VI_write, 1); - flagSet(mFLAG.doFlag); + if (!It) { mVUlow.isNOP = 1; } + else { // Need set _doMac for 4 previous Ops (need to do all 4 because stalls could change the result needed) + mVUinfo.swapOps = 1; + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << 4; } + flagSet(mFLAG.doFlag); + } } //------------------------------------------------------------------ @@ -323,7 +321,7 @@ microVUt(void) mVUanalyzeMflag(mV, int Is, int It) { microVUt(void) mVUanalyzeCflag(mV, int It) { mVUinfo.swapOps = 1; mVUlow.readFlags = 1; - if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << (/*mVUcount +*/ 8); } + if (mVUcount < 4) { mVUpBlock->pState.needExactMatch |= 0xf << 8; } analyzeVIreg2(It, mVUlow.VI_write, 1); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 5538fcc26d..4196d35e33 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -164,18 +164,15 @@ declareAllVariables // Define mVUquickSearch #ifndef __LINUX__ extern u8 mVUsearchXMM[0x1000]; -#endif - typedef u32 (__fastcall *mVUCall)(void*, void*); -#ifndef __LINUX__ #define mVUquickSearch(dest, src, size) ((((mVUCall)((void*)mVUsearchXMM))(dest, src)) == 0xf) -// Note: If GCC builds crash with above function, it means -// that they're not guaranteeing 16-byte alignment on the structs -// being compared. So use this function instead: +#define mVUemitSearch() { mVUcustomSearch(); } #else -//#define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) +// Note: GCC builds crash with custom search function, because +// they're not guaranteeing 16-byte alignment on the structs :( +// #define mVUquickSearch(dest, src, size) (!memcmp(dest, src, size)) #define mVUquickSearch(dest, src, size) (!memcmp_mmx(dest, src, size)) -#define mVUemitSearch() +#define mVUemitSearch() #endif // Misc Macros... diff --git a/pcsx2/x86/microVU_Misc.inl b/pcsx2/x86/microVU_Misc.inl index 9fbd866547..81534987f1 100644 --- a/pcsx2/x86/microVU_Misc.inl +++ b/pcsx2/x86/microVU_Misc.inl @@ -438,11 +438,11 @@ void SSE_ADD2PS_XMM_to_XMM(x86SSERegType to, x86SSERegType from) { // Micro VU - Custom Quick Search //------------------------------------------------------------------ -#ifndef __LINUX__ PCSX2_ALIGNED(0x1000, static u8 mVUsearchXMM[0x1000]); -// Generates a custom optimized block-search function (Note: Structs must be 16-byte aligned!) -void mVUemitSearch() { +// Generates a custom optimized block-search function +// Note: Structs must be 16-byte aligned! (GCC doesn't guarantee this) +void mVUcustomSearch() { using namespace x86Emitter; HostSys::MemProtect(mVUsearchXMM, 0x1000, Protect_ReadWrite, false); memset_8<0xcc,0x1000>(mVUsearchXMM); @@ -491,4 +491,3 @@ void mVUemitSearch() { xRET(); HostSys::MemProtect(mVUsearchXMM, 0x1000, Protect_ReadOnly, true ); } -#endif From c60095f6b0c7a128940171099d0e4b57a6b6d607 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Mon, 15 Jun 2009 23:21:17 +0000 Subject: [PATCH 225/296] GSdx: Fixed Software-mode crash on Escape/Exit, caused by a couple destructor race conditions. (C++ destructors and thread safety are fickle beasts that like to claw each others' eyes out) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1371 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPURendererSW.cpp | 1 + plugins/GSdx/GSRasterizer.cpp | 13 +++++++++++-- plugins/GSdx/GSRasterizer.h | 4 ++++ plugins/GSdx/GSRendererSW.cpp | 2 ++ 4 files changed, 18 insertions(+), 2 deletions(-) diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index c71f72fdcb..030741dcb2 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -32,6 +32,7 @@ GPURendererSW::GPURendererSW(GSDevice* dev) GPURendererSW::~GPURendererSW() { + m_rl.Terminate(); } void GPURendererSW::ResetDevice() diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index 7249849e03..1dcf723e75 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -823,9 +823,12 @@ GSRasterizerMT::GSRasterizerMT(IDrawScanline* ds, int id, int threads, long* syn } } -GSRasterizerMT::~GSRasterizerMT() +GSRasterizerMT::~GSRasterizerMT() {} + +void GSRasterizerMT::Terminate() { m_exit = true; + CloseThread(); } void GSRasterizerMT::Draw(const GSRasterizerData* data) @@ -877,8 +880,14 @@ GSRasterizerList::GSRasterizerList() GSRasterizerList::~GSRasterizerList() { _aligned_free(m_sync); +} - FreeRasterizers(); +void GSRasterizerList::Terminate() +{ + for(list::iterator i = begin(); i != end(); i++) + { + (*i)->Terminate(); + } } void GSRasterizerList::FreeRasterizers() diff --git a/plugins/GSdx/GSRasterizer.h b/plugins/GSdx/GSRasterizer.h index 00a2009baf..a35148571c 100644 --- a/plugins/GSdx/GSRasterizer.h +++ b/plugins/GSdx/GSRasterizer.h @@ -43,6 +43,7 @@ public: virtual void Draw(const GSRasterizerData* data) = 0; virtual void GetStats(GSRasterizerStats& stats) = 0; + virtual void Terminate() = 0; virtual void PrintStats() = 0; }; @@ -101,6 +102,7 @@ public: void Draw(const GSRasterizerData* data); void GetStats(GSRasterizerStats& stats); void PrintStats() {m_ds->PrintStats();} + void Terminate() {} }; class GSRasterizerMT : public GSRasterizer, private GSThread @@ -118,6 +120,7 @@ public: // IRasterizer void Draw(const GSRasterizerData* data); + void Terminate(); }; class GSRasterizerList : protected list, public IRasterizer @@ -148,4 +151,5 @@ public: void Draw(const GSRasterizerData* data); void GetStats(GSRasterizerStats& stats); void PrintStats(); + void Terminate(); }; diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index d9e0d732b6..40a7c93b7d 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -38,6 +38,8 @@ GSRendererSW::GSRendererSW(uint8* base, bool mt, void (*irq)(), GSDevice* dev) GSRendererSW::~GSRendererSW() { + m_rl.Terminate(); + delete m_tc; } From 48eae0c90790d0fe36ab9cf917cd4f7d9d4ed432 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 16 Jun 2009 01:42:08 +0000 Subject: [PATCH 226/296] GSdx: more opengl code, please review, I have no idea if it done the right way ^_^ git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1372 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPURendererSW.cpp | 1 - plugins/GSdx/GS.cpp | 2 +- plugins/GSdx/GSDevice.cpp | 9 +- plugins/GSdx/GSDeviceOGL.cpp | 153 ++++++++++++++++++++++++--------- plugins/GSdx/GSDeviceOGL.h | 29 +++++++ plugins/GSdx/GSRasterizer.cpp | 15 +--- plugins/GSdx/GSRasterizer.h | 4 - plugins/GSdx/GSRendererSW.cpp | 2 - plugins/GSdx/GSSettingsDlg.cpp | 2 +- plugins/GSdx/GSTextureOGL.cpp | 40 ++++++++- plugins/GSdx/GSTextureOGL.h | 3 + 11 files changed, 193 insertions(+), 67 deletions(-) diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index 030741dcb2..c71f72fdcb 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -32,7 +32,6 @@ GPURendererSW::GPURendererSW(GSDevice* dev) GPURendererSW::~GPURendererSW() { - m_rl.Terminate(); } void GPURendererSW::ResetDevice() diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index cecc1019e7..bf7ce343c9 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -602,4 +602,4 @@ EXPORT_C GSBenchmark(HWND hwnd, HINSTANCE hinst, LPSTR lpszCmdLine, int nCmdShow fclose(file); } -#endif \ No newline at end of file +#endif diff --git a/plugins/GSdx/GSDevice.cpp b/plugins/GSdx/GSDevice.cpp index 234c244c28..12d757278e 100644 --- a/plugins/GSdx/GSDevice.cpp +++ b/plugins/GSdx/GSDevice.cpp @@ -167,8 +167,15 @@ void GSDevice::Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVec // KH:COM crashes at startup when booting *through the bios* due to m_merge being NULL. // (texture appears to be non-null, and is being re-created at a size around like 1700x340, // dunno if that's relevant) -- air - if( m_merge ) + + if(m_merge) + { DoMerge(st, sr, dr, m_merge, slbg, mmod, c); + } + else + { + printf("GSdx: m_merge is NULL!\n"); + } m_current = m_merge; } diff --git a/plugins/GSdx/GSDeviceOGL.cpp b/plugins/GSdx/GSDeviceOGL.cpp index c503b7c084..8c8e929ed2 100644 --- a/plugins/GSdx/GSDeviceOGL.cpp +++ b/plugins/GSdx/GSDeviceOGL.cpp @@ -30,6 +30,8 @@ GSDeviceOGL::GSDeviceOGL() , m_vbo(0) , m_fbo(0) , m_topology(-1) + , m_rt((GLuint)-1) + , m_ds((GLuint)-1) { m_vertices.stride = 0; m_vertices.start = 0; @@ -110,10 +112,26 @@ bool GSDeviceOGL::Create(GSWnd* wnd, bool vsync) return false; } - // const char* exts = (const char*)glGetString(GL_EXTENSIONS); + #ifdef _WINDOWS - glGenBuffers(1, &m_vbo); - glBindBuffer(GL_ARRAY_BUFFER, m_vbo); + if(WGLEW_EXT_swap_control) + { + wglSwapIntervalEXT(vsync ? 1 : 0); + } + + #endif + + const char* vendor = (const char*)glGetString(GL_VENDOR); + const char* renderer = (const char*)glGetString(GL_RENDERER); + const char* version = (const char*)glGetString(GL_VERSION); + const char* exts = (const char*)glGetString(GL_EXTENSIONS); + + printf("%s, %s, OpenGL %s\n", vendor, renderer, version); + + const char* str = strstr(exts, "ARB_texture_non_power_of_two"); + + glGenBuffers(1, &m_vbo); CheckError(); + glBindBuffer(GL_ARRAY_BUFFER, m_vbo); CheckError(); // TODO: setup layout? GSVector4i r = wnd->GetClientRect(); @@ -128,33 +146,40 @@ bool GSDeviceOGL::Reset(int w, int h, bool fs) if(!__super::Reset(w, h, fs)) return false; - m_backbuffer = new GSTextureOGL(0, GSTexture::RenderTarget, w, h); + glCullFace(GL_FRONT_AND_BACK); CheckError(); + glDisable(GL_LIGHTING); CheckError(); + glDisable(GL_ALPHA_TEST); CheckError(); + glEnable(GL_SCISSOR_TEST); CheckError(); - glCullFace(GL_FRONT_AND_BACK); - glDisable(GL_LIGHTING); - glDisable(GL_ALPHA_TEST); - glEnable(GL_SCISSOR_TEST); + glMatrixMode(GL_PROJECTION); CheckError(); + glLoadIdentity(); CheckError(); + glMatrixMode(GL_MODELVIEW); CheckError(); + glLoadIdentity(); CheckError(); - glMatrixMode(GL_PROJECTION); - glLoadIdentity(); - glMatrixMode(GL_MODELVIEW); - glLoadIdentity(); - - glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); + glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); CheckError(); // glViewport(0, 0, w, h); if(m_fbo) {glDeleteFramebuffersEXT(1, &m_fbo); m_fbo = 0;} - glGenFramebuffers(1, &m_fbo); + glGenFramebuffers(1, &m_fbo); CheckError(); if(m_fbo == 0) return false; - glBindFramebuffer(GL_FRAMEBUFFER, m_fbo); + glBindFramebuffer(GL_FRAMEBUFFER, m_fbo); CheckError(); return true; } +void GSDeviceOGL::Present(const GSVector4i& r, int shader) +{ + glBindFramebuffer(GL_FRAMEBUFFER, 0); CheckError(); + + // TODO: m_current => backbuffer + + Flip(); +} + void GSDeviceOGL::Flip() { #ifdef _WINDOWS @@ -170,7 +195,7 @@ void GSDeviceOGL::BeginScene() void GSDeviceOGL::DrawPrimitive() { - glDrawArrays(m_topology, m_vertices.count, m_vertices.start); + glDrawArrays(m_topology, m_vertices.count, m_vertices.start); CheckError(); } void GSDeviceOGL::EndScene() @@ -181,10 +206,21 @@ void GSDeviceOGL::EndScene() void GSDeviceOGL::ClearRenderTarget(GSTexture* t, const GSVector4& c) { + GLuint texture = *(GSTextureOGL*)t; + + if(texture == 0) + { + // TODO: backbuffer + } + else + { + glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, GL_RENDERBUFFER, texture); CheckError(); + } + // TODO: disable scissor, color mask - glClearColor(c.r, c.g, c.b, c.a); - glClear(GL_COLOR_BUFFER_BIT); + glClearColor(c.r, c.g, c.b, c.a); CheckError(); + glClear(GL_COLOR_BUFFER_BIT); CheckError(); } void GSDeviceOGL::ClearRenderTarget(GSTexture* t, uint32 c) @@ -194,18 +230,22 @@ void GSDeviceOGL::ClearRenderTarget(GSTexture* t, uint32 c) void GSDeviceOGL::ClearDepth(GSTexture* t, float c) { + glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_DEPTH_COMPONENT, *(GSTextureOGL*)t); CheckError(); + // TODO: disable scissor, depth mask - glClearDepth(c); - glClear(GL_DEPTH_BUFFER_BIT); + glClearDepth(c); CheckError(); + glClear(GL_DEPTH_BUFFER_BIT); CheckError(); } void GSDeviceOGL::ClearStencil(GSTexture* t, uint8 c) { + glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_STENCIL_ATTACHMENT, GL_STENCIL_INDEX, *(GSTextureOGL*)t); CheckError(); + // TODO: disable scissor, depth (?) mask - glClearStencil((GLint)c); - glClear(GL_STENCIL_BUFFER_BIT); + glClearStencil((GLint)c); CheckError(); + glClear(GL_STENCIL_BUFFER_BIT); CheckError(); } GSTexture* GSDeviceOGL::Create(int type, int w, int h, int format) @@ -215,20 +255,20 @@ GSTexture* GSDeviceOGL::Create(int type, int w, int h, int format) switch(type) { case GSTexture::RenderTarget: - glGenTextures(1, &texture); - glBindTexture(GL_TEXTURE_2D, texture); - glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); + glGenTextures(1, &texture); CheckError(); + glBindTexture(GL_TEXTURE_2D, texture); CheckError(); + glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); CheckError(); break; case GSTexture::DepthStencil: - glGenRenderbuffersEXT(1, &texture); - glBindRenderbufferEXT(GL_RENDERBUFFER_EXT, texture); - glRenderbufferStorageEXT(GL_RENDERBUFFER_EXT, GL_DEPTH32F_STENCIL8, w, h); + glGenRenderbuffers(1, &texture); CheckError(); + glBindRenderbuffer(GL_RENDERBUFFER, texture); CheckError(); + glRenderbufferStorage(GL_RENDERBUFFER, GL_DEPTH32F_STENCIL8, w, h); CheckError(); // TODO: depth textures? break; case GSTexture::Texture: - glGenTextures(1, &texture); - glBindTexture(GL_TEXTURE_2D, texture); - glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); + glGenTextures(1, &texture); CheckError(); + glBindTexture(GL_TEXTURE_2D, texture); CheckError(); + glTexImage2D(GL_TEXTURE_2D, 0, format, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL); CheckError(); break; case GSTexture::Offscreen: // TODO: ??? @@ -298,34 +338,41 @@ void GSDeviceOGL::IASetVertexBuffer(const void* vertices, size_t stride, size_t { ASSERT(m_vertices.count == 0); - glBindBuffer(GL_ARRAY_BUFFER, m_vbo); + glBindBuffer(GL_ARRAY_BUFFER, m_vbo); CheckError(); + + bool growbuffer = false; + bool discard = false; // in opengl 3.x this should be a flag to glMapBuffer, as I read somewhere if(count * stride > m_vertices.limit * m_vertices.stride) { m_vertices.start = 0; m_vertices.count = 0; - m_vertices.limit = 0; - } - - if(m_vertices.limit == 0) - { m_vertices.limit = max(count * 3 / 2, 10000); - glBufferData(GL_ARRAY_BUFFER, m_vertices.limit * stride, NULL, GL_DYNAMIC_DRAW); // GL_STREAM_DRAW? + growbuffer = true; } if(m_vertices.start + count > m_vertices.limit || stride != m_vertices.stride) { - m_vertices.start = 0; // TODO: how to discard and not overwrite previous data? + m_vertices.start = 0; + + discard = true; } + if(growbuffer || discard) + { + glBufferData(GL_ARRAY_BUFFER, m_vertices.limit * stride, NULL, GL_DYNAMIC_DRAW); CheckError(); // GL_STREAM_DRAW? + } + + glBufferSubData(GL_ARRAY_BUFFER, m_vertices.start * stride, count * stride, vertices); CheckError(); +/* if(GLvoid* v = glMapBuffer(GL_ARRAY_BUFFER, GL_WRITE_ONLY)) { GSVector4i::storent((uint8*)v + m_vertices.start * stride, vertices, count * stride); - glUnmapBuffer(GL_ARRAY_BUFFER); + glUnmapBuffer(GL_ARRAY_BUFFER); CheckError(); } - +*/ m_vertices.count = count; m_vertices.stride = stride; } @@ -340,3 +387,25 @@ void GSDeviceOGL::IASetPrimitiveTopology(int topology) m_topology = topology; } +void GSDeviceOGL::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) +{ + GLuint rti = 0; + GLuint dsi = 0; + + if(rt) rti = *(GSTextureOGL*)rt; + if(ds) dsi = *(GSTextureOGL*)ds; + + // TODO: if(m_rt != rti) + { + glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, GL_RENDERBUFFER, rti); CheckError(); + + // TODO: m_rt = rti; + } + + // TODO: if(m_ds != dsi) + { + glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_DEPTH_COMPONENT, dsi); CheckError(); + + // TODO: m_ds = dsi; + } +} diff --git a/plugins/GSdx/GSDeviceOGL.h b/plugins/GSdx/GSDeviceOGL.h index 32c510b2d1..d95b8916b3 100644 --- a/plugins/GSdx/GSDeviceOGL.h +++ b/plugins/GSdx/GSDeviceOGL.h @@ -42,6 +42,8 @@ class GSDeviceOGL : public GSDevice } m_vertices; int m_topology; + GLuint m_rt; + GLuint m_ds; // @@ -62,6 +64,7 @@ public: bool Create(GSWnd* wnd, bool vsync); bool Reset(int w, int h, bool fs); + void Present(const GSVector4i& r, int shader); void Flip(); void BeginScene(); @@ -85,4 +88,30 @@ public: void IASetVertexBuffer(const void* vertices, size_t stride, size_t count); void IASetInputLayout(); // TODO void IASetPrimitiveTopology(int topology); + + void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); + + static void CheckError() + { + #ifdef _DEBUG + + GLenum error = glGetError(); + + if(error != GL_NO_ERROR) + { + printf("%s\n", gluErrorString(error)); + } + + #endif + } + + static void CheckFrameBuffer() + { + GLenum status = glCheckFramebufferStatus(GL_FRAMEBUFFER); + + if(status != GL_FRAMEBUFFER_COMPLETE) + { + printf("%d\n", status); + } + } }; diff --git a/plugins/GSdx/GSRasterizer.cpp b/plugins/GSdx/GSRasterizer.cpp index 1dcf723e75..abd730281a 100644 --- a/plugins/GSdx/GSRasterizer.cpp +++ b/plugins/GSdx/GSRasterizer.cpp @@ -823,12 +823,9 @@ GSRasterizerMT::GSRasterizerMT(IDrawScanline* ds, int id, int threads, long* syn } } -GSRasterizerMT::~GSRasterizerMT() {} - -void GSRasterizerMT::Terminate() +GSRasterizerMT::~GSRasterizerMT() { m_exit = true; - CloseThread(); } void GSRasterizerMT::Draw(const GSRasterizerData* data) @@ -879,15 +876,9 @@ GSRasterizerList::GSRasterizerList() GSRasterizerList::~GSRasterizerList() { - _aligned_free(m_sync); -} + FreeRasterizers(); -void GSRasterizerList::Terminate() -{ - for(list::iterator i = begin(); i != end(); i++) - { - (*i)->Terminate(); - } + _aligned_free(m_sync); } void GSRasterizerList::FreeRasterizers() diff --git a/plugins/GSdx/GSRasterizer.h b/plugins/GSdx/GSRasterizer.h index a35148571c..00a2009baf 100644 --- a/plugins/GSdx/GSRasterizer.h +++ b/plugins/GSdx/GSRasterizer.h @@ -43,7 +43,6 @@ public: virtual void Draw(const GSRasterizerData* data) = 0; virtual void GetStats(GSRasterizerStats& stats) = 0; - virtual void Terminate() = 0; virtual void PrintStats() = 0; }; @@ -102,7 +101,6 @@ public: void Draw(const GSRasterizerData* data); void GetStats(GSRasterizerStats& stats); void PrintStats() {m_ds->PrintStats();} - void Terminate() {} }; class GSRasterizerMT : public GSRasterizer, private GSThread @@ -120,7 +118,6 @@ public: // IRasterizer void Draw(const GSRasterizerData* data); - void Terminate(); }; class GSRasterizerList : protected list, public IRasterizer @@ -151,5 +148,4 @@ public: void Draw(const GSRasterizerData* data); void GetStats(GSRasterizerStats& stats); void PrintStats(); - void Terminate(); }; diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index 40a7c93b7d..d9e0d732b6 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -38,8 +38,6 @@ GSRendererSW::GSRendererSW(uint8* base, bool mt, void (*irq)(), GSDevice* dev) GSRendererSW::~GSRendererSW() { - m_rl.Terminate(); - delete m_tc; } diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 3c84591fc9..59e7a8ac7d 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -219,4 +219,4 @@ void GSSettingsDlg::UpdateControls() EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS_EDIT), sw); EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS), sw); } -} \ No newline at end of file +} diff --git a/plugins/GSdx/GSTextureOGL.cpp b/plugins/GSdx/GSTextureOGL.cpp index 3e8f1f97eb..ca59491c72 100644 --- a/plugins/GSdx/GSTextureOGL.cpp +++ b/plugins/GSdx/GSTextureOGL.cpp @@ -21,6 +21,7 @@ #include "stdafx.h" #include "GSTextureOGL.h" +#include "GSDeviceOGL.h" GSTextureOGL::GSTextureOGL(GLuint texture, int type, int width, int height, int format) : m_texture(texture) @@ -29,19 +30,27 @@ GSTextureOGL::GSTextureOGL(GLuint texture, int type, int width, int height, int , m_height(height) , m_format(format) { + // TODO: offscreen type should be just a memory array, also returned in Map + + glGenBuffers(1, &m_pbo); GSDeviceOGL::CheckError(); } GSTextureOGL::~GSTextureOGL() { + if(m_pbo) + { + glDeleteBuffers(1, &m_pbo); GSDeviceOGL::CheckError(); + } + if(m_texture) { switch(m_type) { case DepthStencil: - glDeleteRenderbuffersEXT(1, &m_texture); + glDeleteRenderbuffers(1, &m_texture); GSDeviceOGL::CheckError(); break; default: - glDeleteTextures(1, &m_texture); + glDeleteTextures(1, &m_texture); GSDeviceOGL::CheckError(); break; } } @@ -69,7 +78,32 @@ int GSTextureOGL::GetFormat() const bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch) { - // TODO: glTexSubImage2D looks like UpdateSubresource but does not take a pitch + glBindBuffer(GL_PIXEL_UNPACK_BUFFER, m_pbo); GSDeviceOGL::CheckError(); + + int w = r.width(); + int h = r.height(); + int bpp = 32; // TODO: should be in sync with m_format + int dstpitch = w * bpp >> 3; + + glBufferData(GL_PIXEL_UNPACK_BUFFER, h * dstpitch, NULL, GL_STREAM_DRAW); GSDeviceOGL::CheckError(); + + if(uint8* dst = (uint8*)glMapBuffer(GL_PIXEL_UNPACK_BUFFER, GL_WRITE_ONLY)) + { + uint8* src = (uint8*)data; + + for(int i = 0; i < h; i++, src += pitch, dst += dstpitch) + { + memcpy(dst, src, dstpitch); + } + + glUnmapBuffer(GL_PIXEL_UNPACK_BUFFER); GSDeviceOGL::CheckError(); + } + + glBindTexture(GL_TEXTURE_2D, m_texture); GSDeviceOGL::CheckError(); + + glTexSubImage2D(GL_TEXTURE_2D, 0, r.left, r.top, w, h, GL_RGBA, GL_UNSIGNED_BYTE, 0); GSDeviceOGL::CheckError(); + + glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0); GSDeviceOGL::CheckError(); return false; } diff --git a/plugins/GSdx/GSTextureOGL.h b/plugins/GSdx/GSTextureOGL.h index 5e6e2a7926..d69581a697 100644 --- a/plugins/GSdx/GSTextureOGL.h +++ b/plugins/GSdx/GSTextureOGL.h @@ -26,6 +26,7 @@ class GSTextureOGL : public GSTexture { GLuint m_texture; + GLuint m_pbo; int m_type; int m_width; @@ -45,4 +46,6 @@ public: bool Map(GSMap& m, const GSVector4i* r); void Unmap(); bool Save(const string& fn, bool dds = false); + + operator GLuint() {return m_texture;} }; From d1c7d2f1a6b66aa22414edc369dd69f60bb6a8cc Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 16 Jun 2009 02:28:43 +0000 Subject: [PATCH 227/296] xpad: fixed a crash that happened when subclassing an opengl window, also got rid of mfc. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1373 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/xpad/stdafx.h | 85 ++++++++++++++++++++++---------- plugins/xpad/xpad.cpp | 86 +++++++++++++-------------------- plugins/xpad/xpad.h | 15 ------ plugins/xpad/xpad_vs2008.vcproj | 16 +++--- 4 files changed, 101 insertions(+), 101 deletions(-) diff --git a/plugins/xpad/stdafx.h b/plugins/xpad/stdafx.h index 7d782ced5f..326e8d0bc1 100644 --- a/plugins/xpad/stdafx.h +++ b/plugins/xpad/stdafx.h @@ -4,58 +4,93 @@ #pragma once -#pragma warning(disable: 4996) +#pragma warning(disable: 4996 4995 4324 4100 4101 4201) -#ifndef VC_EXTRALEAN -#define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers -#endif +// The following macros define the minimum required platform. The minimum required platform +// is the earliest version of Windows, Internet Explorer etc. that has the necessary features to run +// your application. The macros work by enabling all features available on platform versions up to and +// including the version specified. // Modify the following defines if you have to target a platform prior to the ones specified below. // Refer to MSDN for the latest info on corresponding values for different platforms. -#ifndef WINVER // Allow use of features specific to Windows 95 and Windows NT 4 or later. -#define WINVER 0x0510 // Change this to the appropriate value to target Windows 98 and Windows 2000 or later. + +#ifndef WINVER // Specifies that the minimum required platform is Windows Vista. +#define WINVER 0x0600 // Change this to the appropriate value to target other versions of Windows. #endif -#ifndef _WIN32_WINNT // Allow use of features specific to Windows NT 4 or later. -#define _WIN32_WINNT 0x0400 // Change this to the appropriate value to target Windows 2000 or later. -#endif +#ifndef _WIN32_WINNT // Specifies that the minimum required platform is Windows Vista. +#define _WIN32_WINNT 0x0600 // Change this to the appropriate value to target other versions of Windows. +#endif -#ifndef _WIN32_WINDOWS // Allow use of features specific to Windows 98 or later. +#ifndef _WIN32_WINDOWS // Specifies that the minimum required platform is Windows 98. #define _WIN32_WINDOWS 0x0410 // Change this to the appropriate value to target Windows Me or later. #endif -#ifndef _WIN32_IE // Allow use of features specific to IE 4.0 or later. -#define _WIN32_IE 0x0400 // Change this to the appropriate value to target IE 5.0 or later. +#ifndef _WIN32_IE // Specifies that the minimum required platform is Internet Explorer 7.0. +#define _WIN32_IE 0x0700 // Change this to the appropriate value to target other versions of IE. #endif -#define _ATL_CSTRING_EXPLICIT_CONSTRUCTORS // some CString constructors will be explicit +#define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers -#include // MFC core and standard components -//#include // MFC extensions -#ifndef _AFX_NO_AFXCMN_SUPPORT -#include // MFC support for Windows Common Controls -#endif // _AFX_NO_AFXCMN_SUPPORT -//#include +#include +#include +#include +#include #include -#include -#include -#include #include -#define countof(a) (sizeof(a)/sizeof(a[0])) +// stdc + +#include +#include +#include + +#include +#include +#include +#include +#include + +using namespace std; +using namespace stdext; + +extern string format(const char* fmt, ...); + +// syntactic sugar + +// put these into vc9/common7/ide/usertype.dat to have them highlighted + +typedef unsigned char uint8; +typedef signed char int8; +typedef unsigned short uint16; +typedef signed short int16; +typedef unsigned int uint32; +typedef signed int int32; +typedef unsigned long long uint64; +typedef signed long long int64; + +#define countof(a) (sizeof(a) / sizeof(a[0])) #define EXPORT_C extern "C" __declspec(dllexport) void __stdcall #define EXPORT_C_(type) extern "C" __declspec(dllexport) type __stdcall +#define ALIGN_STACK(n) __declspec(align(n)) int __dummy; + #ifndef RESTRICT #ifdef __INTEL_COMPILER #define RESTRICT restrict - #elif _MSC_VER >= 1400 + #elif _MSC_VER >= 1400 // TODO: gcc #define RESTRICT __restrict #else #define RESTRICT #endif #endif -#pragma warning(disable : 4995 4324 4100) +#if defined(_DEBUG) && defined(_MSC_VER) + #define ASSERT assert +#else + #define ASSERT(exp) ((void)0) +#endif + + diff --git a/plugins/xpad/xpad.cpp b/plugins/xpad/xpad.cpp index 94aea468cd..26ca027d60 100644 --- a/plugins/xpad/xpad.cpp +++ b/plugins/xpad/xpad.cpp @@ -22,51 +22,19 @@ #include "stdafx.h" #include "xpad.h" -#ifdef _DEBUG -#define new DEBUG_NEW -#endif +static HMODULE s_hModule; -// -// Note! -// -// If this DLL is dynamically linked against the MFC -// DLLs, any functions exported from this DLL which -// call into MFC must have the AFX_MANAGE_STATE macro -// added at the very beginning of the function. -// -// For example: -// -// extern "C" BOOL PASCAL EXPORT ExportedFunction() -// { -// AFX_MANAGE_STATE(AfxGetStaticModuleState()); -// // normal function body here -// } -// -// It is very important that this macro appear in each -// function, prior to any calls into MFC. This means that -// it must appear as the first statement within the -// function, even before any object variable declarations -// as their constructors may generate calls into the MFC -// DLL. -// -// Please see MFC Technical Notes 33 and 58 for additional -// details. -// - -BEGIN_MESSAGE_MAP(xpadApp, CWinApp) -END_MESSAGE_MAP() - -xpadApp::xpadApp() +BOOL APIENTRY DllMain(HMODULE hModule, DWORD ul_reason_for_call, LPVOID lpReserved) { -} - -xpadApp theApp; - -BOOL xpadApp::InitInstance() -{ - __super::InitInstance(); - - SetRegistryKey(_T("Gabest")); + switch(ul_reason_for_call) + { + case DLL_PROCESS_ATTACH: + s_hModule = hModule; + case DLL_THREAD_ATTACH: + case DLL_THREAD_DETACH: + case DLL_PROCESS_DETACH: + break; + } return TRUE; } @@ -328,7 +296,7 @@ public: static class XPadPlugin { - CAtlArray m_pads; + vector m_pads; XPad* m_pad; int m_index; bool m_cfgreaddata; @@ -531,7 +499,7 @@ static class XPadPlugin BYTE UnknownCommand(int index, BYTE value) { - TRACE(_T("Unknown command %02x (%d, %02x)\n"), m_cmd, index, value); + // printf("Unknown command %02x (%d, %02x)\n", m_cmd, index, value); return 0; } @@ -544,8 +512,8 @@ public: , m_cfgreaddata(false) , m_handler(NULL) { - m_pads.Add(new XPad(0)); - m_pads.Add(new XPad(1)); + m_pads.push_back(new XPad(0)); + m_pads.push_back(new XPad(1)); for(int i = 0; i < countof(m_handlers); i++) { @@ -565,6 +533,16 @@ public: m_handlers['O'] = &XPadPlugin::SetDS2NativeMode; } + virtual ~XPadPlugin() + { + for(vector::iterator i = m_pads.begin(); i != m_pads.end(); i++) + { + delete *i; + } + + m_pads.clear(); + } + void StartPoll(int pad) { m_pad = m_pads[pad & 1]; @@ -602,7 +580,7 @@ static int s_nRefs = 0; static HWND s_hWnd = NULL; static WNDPROC s_GSWndProc = NULL; -static class CKeyEventList : protected CAtlList, protected CCritSec +static class CKeyEventList : protected list, protected CCritSec { public: void Push(UINT32 event, UINT32 key) @@ -614,16 +592,18 @@ public: e.event = event; e.key = key; - AddTail(e); + push_back(e); } bool Pop(KeyEvent& e) { CAutoLock cAutoLock(this); - if(IsEmpty()) return false; - - e = RemoveHead(); + if(empty()) return false; + + e = front(); + + pop_front(); return true; } @@ -647,7 +627,7 @@ LRESULT WINAPI PADwndProc(HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam) break; } - return s_GSWndProc(hWnd, msg, wParam, lParam); + return CallWindowProc(s_GSWndProc, hWnd, msg, wParam, lParam); } // diff --git a/plugins/xpad/xpad.h b/plugins/xpad/xpad.h index 9afa410498..1c4f48e165 100644 --- a/plugins/xpad/xpad.h +++ b/plugins/xpad/xpad.h @@ -21,21 +21,6 @@ #pragma once -#ifndef __AFXWIN_H__ - #error include 'stdafx.h' before including this file for PCH -#endif - -class xpadApp : public CWinApp -{ -public: - xpadApp(); - -public: - virtual BOOL InitInstance(); - - DECLARE_MESSAGE_MAP() -}; - // ps1 #define PSE_LT_PAD 8 diff --git a/plugins/xpad/xpad_vs2008.vcproj b/plugins/xpad/xpad_vs2008.vcproj index d44d4c5f32..105b923bb4 100644 --- a/plugins/xpad/xpad_vs2008.vcproj +++ b/plugins/xpad/xpad_vs2008.vcproj @@ -23,7 +23,7 @@ Name="Debug|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > @@ -281,7 +281,7 @@ Name="Debug SSE2|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops;.\vsprops\sse2.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > @@ -536,7 +536,7 @@ Name="Release SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\release.vsprops;.\vsprops\sse2.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" WholeProgramOptimization="1" > @@ -664,7 +664,7 @@ Name="Debug SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops;.\vsprops\sse2.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="0" CharacterSet="2" > From fb4fcc7d29486c3779b0b5be52ace6d2b43faa28 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Tue, 16 Jun 2009 22:49:06 +0000 Subject: [PATCH 228/296] Changed some code around in microVU that was causing 10+ minute link times in Release builds (LTCG mode), also added some new x86Emitter cleanups along the way. :) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1374 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/ix86/ix86_inlines.inl | 11 ++ pcsx2/x86/microVU.cpp | 196 +++++++++++++++++++++++++++++++ pcsx2/x86/microVU.h | 4 +- pcsx2/x86/microVU_Compile.inl | 201 ++------------------------------ pcsx2/x86/microVU_Execute.inl | 3 +- pcsx2/x86/microVU_Misc.h | 4 +- 6 files changed, 222 insertions(+), 197 deletions(-) diff --git a/pcsx2/x86/ix86/ix86_inlines.inl b/pcsx2/x86/ix86/ix86_inlines.inl index 4462aecd0b..45c9b75d75 100644 --- a/pcsx2/x86/ix86/ix86_inlines.inl +++ b/pcsx2/x86/ix86/ix86_inlines.inl @@ -231,4 +231,15 @@ namespace x86Emitter } } + // ------------------------------------------------------------------------ + // returns the inverted conditional type for this Jcc condition. Ie, JNS will become JS. + // + static __forceinline JccComparisonType xInvertCond( JccComparisonType src ) + { + jASSUME( src != Jcc_Unknown ); + if( Jcc_Unconditional == src ) return Jcc_Unconditional; + + // x86 conditionals are clever! To invert conditional types, just invert the lower bit: + return (JccComparisonType)((int)src ^ 1); + } } diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index a2f65ed38f..f3403d6455 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -252,6 +252,202 @@ microVUf(int) mVUsearchProg() { return 1; // If !cleared, then we're still on the same program as last-time ;) } +//------------------------------------------------------------------ +// JIT recompiler -- called from recompiled code when register jumps are made. +//------------------------------------------------------------------ + +microVUf(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState) { + + return mVUblockFetch( mVUx, startPC, pState ); +} + +//------------------------------------------------------------------ +// Recompiler +//------------------------------------------------------------------ + +static void* __fastcall mVUcompile( microVU* mVU, u32 startPC, uptr pState ) +{ + using namespace x86Emitter; + + // Setup Program Bounds/Range + mVUsetupRange(mVU, startPC); + + microBlock* pBlock = NULL; + u8* thisPtr = x86Ptr; + + const u32 microSizeDiv8 = (mVU->microSize-1) / 8; + + // First Pass + iPC = startPC / 4; + setCode(); + mVUbranch = 0; + mVUstartPC = iPC; + mVUcount = 0; + mVUcycles = 0; // Skips "M" phase, and starts counting cycles at "T" stage + mVU->p = 0; // All blocks start at p index #0 + mVU->q = 0; // All blocks start at q index #0 + memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info + mVUblock.x86ptrStart = thisPtr; + pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager + mVUpBlock = pBlock; + mVUregs.flags = 0; + mVUflagInfo = 0; + mVUsFlagHack = CHECK_VU_FLAGHACK; + + for (int branch = 0; mVUcount < microSizeDiv8; ) { + incPC(1); + startLoop(); + mVUincCycles(mVU, 1); + mVUopU(mVU, 0); + if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } + if (curI & _DTbit_) { branch = 4; } + if (curI & _Mbit_) { mVUup.mBit = 1; } + if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } + else { incPC(-1); mVUopL(mVU, 0); incPC(1); } + mVUsetCycles(mVU); + mVUinfo.readQ = mVU->q; + mVUinfo.writeQ = !mVU->q; + mVUinfo.readP = mVU->p; + mVUinfo.writeP = !mVU->p; + if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } + else if (branch == 1) { branch = 2; } + if (mVUbranch) { mVUsetFlagInfo(mVU); branch = 3; mVUbranch = 0; } + incPC(1); + mVUcount++; + } + + // Sets Up Flag instances + int xStatus[4], xMac[4], xClip[4]; + int xCycles = mVUsetFlags(mVU, xStatus, xMac, xClip); + mVUtestCycles(mVU); + + // Second Pass + iPC = mVUstartPC; + setCode(); + mVUbranch = 0; + uint x; + for (x = 0; x < microSizeDiv8; x++) { + if (mVUinfo.isEOB) { x = 0xffff; } + if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } + if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } + else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } + else { doSwapOp(); } + if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } + + if (!mVUinfo.isBdelay) { incPC(1); } + else { + microBlock* bBlock = NULL; + s32* ajmp = 0; + mVUsetupRange(mVU, xPC); + mVUdebugNOW(1); + + switch (mVUbranch) { + case 3: branchCase(Jcc_Equal); // IBEQ + case 4: branchCase(Jcc_GreaterOrEqual); // IBGEZ + case 5: branchCase(Jcc_Greater); // IBGTZ + case 6: branchCase(Jcc_LessOrEqual); // IBLEQ + case 7: branchCase(Jcc_Less); // IBLTZ + case 8: branchCase(Jcc_NotEqual); // IBNEQ + case 1: case 2: // B/BAL + + mVUprint("mVUcompile B/BAL"); + incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); + + if (mVUblocks[branchAddr/8] == NULL) + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); + + // Check if branch-block has already been compiled + pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); + if (pBlock) { xJMP(pBlock->x86ptrStart); } + else { mVUcompile(mVU, branchAddr, (uptr)&mVUregs); } + return thisPtr; + case 9: case 10: // JR/JALR + + mVUprint("mVUcompile JR/JALR"); + incPC(-3); // Go back to jump opcode + + if (mVUup.eBit) { // E-bit Jump + mVUendProgram(mVU, 2, xStatus, xMac, xClip); + MOV32MtoR(gprT1, (uptr)&mVU->branch); + MOV32RtoM((uptr)&mVU->regs->VI[REG_TPC].UL, gprT1); + xJMP(mVU->exitFunct); + return thisPtr; + } + + memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); + + mVUbackupRegs(mVU); + MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) + MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) + + if (!mVU->index) xCALL( mVUcompileJIT<0> ); //(u32 startPC, uptr pState) + else xCALL( mVUcompileJIT<1> ); + mVUrestoreRegs(mVU); + JMPR(gprT1); // Jump to rec-code address + return thisPtr; + } + // Conditional Branches + mVUprint("mVUcompile conditional branch"); + if (bBlock) { // Branch non-taken has already been compiled + incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if (mVUblocks[branchAddr/8] == NULL) + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); + + // Check if branch-block has already been compiled + pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); + if (pBlock) { xJMP( pBlock->x86ptrStart ); } + else if (!mVU->index) { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } + else { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } + } + else { + uptr jumpAddr; + u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs so back them up + memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); + + incPC2(1); // Get PC for branch not-taken + mVUcompile(mVU, xPC, (uptr)&mVUregs); + + iPC = bPC; + incPC(-3); // Go back to branch opcode (to get branch imm addr) + if (!mVU->index) jumpAddr = (uptr)mVUblockFetch(mVU, branchAddr, (uptr)&pBlock->pStateEnd); + else jumpAddr = (uptr)mVUblockFetch(mVU, branchAddr, (uptr)&pBlock->pStateEnd); + *ajmp = (jumpAddr - ((uptr)ajmp + 4)); + } + return thisPtr; + } + } + if (x == microSizeDiv8) { Console::Error("microVU%d: Possible infinite compiling loop!", params mVU->index); } + + // E-bit End + mVUendProgram(mVU, 1, xStatus, xMac, xClip); + return thisPtr; +} + + +microVUt(void*) mVUblockFetch( microVU* mVU, u32 startPC, uptr pState ) +{ + using namespace x86Emitter; + + if (startPC > mVU->microSize-1) { Console::Error("microVU%d: invalid startPC", params mVU->index); } + startPC &= ~7; + + if (mVUblocks[startPC/8] == NULL) { + mVUblocks[startPC/8] = microBlockManager::AlignedNew(); + } + + // Searches for Existing Compiled Block (if found, then returns; else, compile) + microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); + if (pBlock) { return pBlock->x86ptrStart; } + + return mVUcompile( mVU, startPC, pState ); +} + + //------------------------------------------------------------------ // Wrapper Functions - Called by other parts of the Emu //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 0dab141a4b..c722b0e5f8 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -155,17 +155,15 @@ microVUf(void) mVUinit(VURegs*); microVUx(void) mVUreset(); microVUf(void) mVUclose(); microVUf(void) mVUclear(u32, u32); +microVUt(void*) mVUblockFetch( microVU* mVU, u32 startPC, uptr pState ); // Prototypes for Linux void __fastcall mVUcleanUpVU0(); void __fastcall mVUcleanUpVU1(); -void* __fastcall mVUcompileVU0(u32 startPC, uptr pState); -void* __fastcall mVUcompileVU1(u32 startPC, uptr pState); mVUop(mVUopU); mVUop(mVUopL); // Private Functions -microVUf(void) mVUclearProg(microVU* mVU, int progIndex); microVUf(int) mVUfindLeastUsedProg(microVU* mVU); microVUf(int) mVUsearchProg(); microVUf(void) mVUcacheProg(int progIndex); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 30ac835a2b..6b713e24a9 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -22,20 +22,20 @@ // Helper Macros //------------------------------------------------------------------ -#define branchCase(JMPcc, nJMPcc, ebitJMP) \ +#define branchCase(JMPcc) \ mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ - CMP16ItoM((uptr)&mVU->branch, 0); \ + xCMP( ptr16[&mVU->branch], 0); \ if (mVUup.eBit) { /* Conditional Branch With E-Bit Set */ \ mVUendProgram(mVU, 2, xStatus, xMac, xClip); \ - u8* eJMP = ebitJMP(0); \ + xForwardJump8 eJMP( JMPcc ); \ incPC(1); /* Set PC to First instruction of Non-Taken Side */ \ - MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); \ - JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); \ - x86SetJ8(eJMP); \ + xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ + xJMP( mVU->exitFunct ); \ + eJMP.SetTarget(); \ incPC(-4); /* Go Back to Branch Opcode to get branchAddr */ \ iPC = branchAddr/4; \ - MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); \ - JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); \ + xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ + xJMP( mVU->exitFunct ); \ return thisPtr; \ } \ else { /* Normal Conditional Branch */ \ @@ -43,14 +43,14 @@ if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ - if (bBlock) { nJMPcc((uptr)bBlock->x86ptrStart - ((uptr)x86Ptr + 6)); } \ - else { ajmp = JMPcc((uptr)0); } \ + if (bBlock) { xJcc( xInvertCond( JMPcc ), bBlock->x86ptrStart ); } \ + else { ajmp = xJcc32( JMPcc ); } \ } \ break #define branchWarning() { \ if (mVUbranch) { \ - Console::Error("microVU%d Warning: Branch in E-bit/Branch delay slot! [%04x]", params vuIndex, xPC); \ + Console::Error("microVU%d Warning: Branch in E-bit/Branch delay slot! [%04x]", params mVU->index, xPC); \ mVUlow.isNOP = 1; \ } \ } @@ -282,182 +282,3 @@ microVUt(void) mVUtestCycles(mV) { mVUendProgram(mVU, 0, NULL, NULL, NULL); x86SetJ32(jmp32); } - -//------------------------------------------------------------------ -// Recompiler -//------------------------------------------------------------------ - -microVUf(void*) __fastcall mVUcompile(u32 startPC, uptr pState) { - microVU* mVU = mVUx; - u8* thisPtr = x86Ptr; - - if (startPC > ((vuIndex) ? 0x3fff : 0xfff)) { Console::Error("microVU%d: invalid startPC", params vuIndex); } - startPC &= (vuIndex ? 0x3ff8 : 0xff8); - - if (mVUblocks[startPC/8] == NULL) { - mVUblocks[startPC/8] = microBlockManager::AlignedNew(); - } - - // Searches for Existing Compiled Block (if found, then returns; else, compile) - microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); - if (pBlock) { return pBlock->x86ptrStart; } - - // Setup Program Bounds/Range - mVUsetupRange(mVU, startPC); - - // First Pass - iPC = startPC / 4; - setCode(); - mVUbranch = 0; - mVUstartPC = iPC; - mVUcount = 0; - mVUcycles = 0; // Skips "M" phase, and starts counting cycles at "T" stage - mVU->p = 0; // All blocks start at p index #0 - mVU->q = 0; // All blocks start at q index #0 - memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info - mVUblock.x86ptrStart = thisPtr; - pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager - mVUpBlock = pBlock; - mVUregs.flags = 0; - mVUflagInfo = 0; - mVUsFlagHack = CHECK_VU_FLAGHACK; - - for (int branch = 0; mVUcount < (vuIndex ? (0x3fff/8) : (0xfff/8)); ) { - incPC(1); - startLoop(); - mVUincCycles(mVU, 1); - mVUopU(mVU, 0); - if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } - if (curI & _DTbit_) { branch = 4; } - if (curI & _Mbit_) { mVUup.mBit = 1; } - if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } - else { incPC(-1); mVUopL(mVU, 0); incPC(1); } - mVUsetCycles(mVU); - mVUinfo.readQ = mVU->q; - mVUinfo.writeQ = !mVU->q; - mVUinfo.readP = mVU->p; - mVUinfo.writeP = !mVU->p; - if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } - else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(mVU); branch = 3; mVUbranch = 0; } - incPC(1); - mVUcount++; - } - - // Sets Up Flag instances - int xStatus[4], xMac[4], xClip[4]; - int xCycles = mVUsetFlags(mVU, xStatus, xMac, xClip); - mVUtestCycles(mVU); - - // Second Pass - iPC = mVUstartPC; - setCode(); - mVUbranch = 0; - int x; - for (x = 0; x < (vuIndex ? (0x3fff/8) : (0xfff/8)); x++) { - if (mVUinfo.isEOB) { x = 0xffff; } - if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } - if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } - else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { doSwapOp(); } - if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } - - if (!mVUinfo.isBdelay) { incPC(1); } - else { - microBlock* bBlock = NULL; - u32* ajmp = 0; - mVUsetupRange(mVU, xPC); - mVUdebugNOW(1); - - switch (mVUbranch) { - case 3: branchCase(JE32, JNE32, JE8); // IBEQ - case 4: branchCase(JGE32, JNGE32, JGE8); // IBGEZ - case 5: branchCase(JG32, JNG32, JG8); // IBGTZ - case 6: branchCase(JLE32, JNLE32, JLE8); // IBLEQ - case 7: branchCase(JL32, JNL32, JL8); // IBLTZ - case 8: branchCase(JNE32, JE32, JNE8); // IBNEQ - case 1: case 2: // B/BAL - - mVUprint("mVUcompile B/BAL"); - incPC(-3); // Go back to branch opcode (to get branch imm addr) - - if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - - // Check if branch-block has already been compiled - pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); - if (pBlock) { JMP32((uptr)pBlock->x86ptrStart - ((uptr)x86Ptr + 5)); } - else if (!vuIndex) { mVUcompileVU0(branchAddr, (uptr)&mVUregs); } - else { mVUcompileVU1(branchAddr, (uptr)&mVUregs); } - return thisPtr; - case 9: case 10: // JR/JALR - - mVUprint("mVUcompile JR/JALR"); - incPC(-3); // Go back to jump opcode - - if (mVUup.eBit) { // E-bit Jump - mVUendProgram(mVU, 2, xStatus, xMac, xClip); - MOV32MtoR(gprT1, (uptr)&mVU->branch); - MOV32RtoM((uptr)&mVU->regs->VI[REG_TPC].UL, gprT1); - JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); - return thisPtr; - } - - memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - - mVUbackupRegs(mVU); - MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) - MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) - - if (!vuIndex) CALLFunc((uptr)mVUcompileVU0); //(u32 startPC, uptr pState) - else CALLFunc((uptr)mVUcompileVU1); - mVUrestoreRegs(mVU); - JMPR(gprT1); // Jump to rec-code address - return thisPtr; - } - // Conditional Branches - mVUprint("mVUcompile conditional branch"); - if (bBlock) { // Branch non-taken has already been compiled - incPC(-3); // Go back to branch opcode (to get branch imm addr) - - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - - // Check if branch-block has already been compiled - pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); - if (pBlock) { JMP32((uptr)pBlock->x86ptrStart - ((uptr)x86Ptr + 5)); } - else if (!vuIndex) { mVUcompileVU0(branchAddr, (uptr)&mVUregs); } - else { mVUcompileVU1(branchAddr, (uptr)&mVUregs); } - } - else { - uptr jumpAddr; - u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs so back them up - memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - - incPC2(1); // Get PC for branch not-taken - if (!vuIndex) mVUcompileVU0(xPC, (uptr)&mVUregs); - else mVUcompileVU1(xPC, (uptr)&mVUregs); - - iPC = bPC; - incPC(-3); // Go back to branch opcode (to get branch imm addr) - if (!vuIndex) jumpAddr = (uptr)mVUcompileVU0(branchAddr, (uptr)&pBlock->pStateEnd); - else jumpAddr = (uptr)mVUcompileVU1(branchAddr, (uptr)&pBlock->pStateEnd); - *ajmp = (jumpAddr - ((uptr)ajmp + 4)); - } - return thisPtr; - } - } - if (x == (vuIndex?(0x3fff/8):(0xfff/8))) { Console::Error("microVU%d: Possible infinite compiling loop!", params vuIndex); } - - // E-bit End - mVUendProgram(mVU, 1, xStatus, xMac, xClip); - return thisPtr; -} - -void* __fastcall mVUcompileVU0(u32 startPC, uptr pState) { return mVUcompile<0>(startPC, pState); } -void* __fastcall mVUcompileVU1(u32 startPC, uptr pState) { return mVUcompile<1>(startPC, pState); } - diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 39d1c85e3a..1ca593aac8 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -128,8 +128,7 @@ microVUx(void*) __fastcall mVUexecute(u32 startPC, u32 cycles) { mVU->totalCycles = cycles; x86SetPtr(mVUcurProg.x86ptr); // Set x86ptr to where program left off - if (!vuIndex) return mVUcompileVU0(startPC, (uptr)&mVU->prog.lpState); - else return mVUcompileVU1(startPC, (uptr)&mVU->prog.lpState); + return mVUblockFetch(mVU, startPC, (uptr)&mVU->prog.lpState); } //------------------------------------------------------------------ diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 4196d35e33..66690240e7 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -94,7 +94,7 @@ declareAllVariables #define _Ftf_ ((mVU->code >> 23) & 0x03) #define _Imm5_ (s16)(((mVU->code & 0x400) ? 0xfff0 : 0) | ((mVU->code >> 6) & 0xf)) -#define _Imm11_ (s32)((mVU->code & 0x400) ? (0xfffffc00 | (mVU->code & 0x3ff)) : mVU->code & 0x3ff) +#define _Imm11_ (s32)((mVU->code & 0x400) ? (0xfffffc00 | (mVU->code & 0x3ff)) : (mVU->code & 0x3ff)) #define _Imm12_ (((mVU->code >> 21) & 0x1) << 11) | (mVU->code & 0x7ff) #define _Imm15_ (((mVU->code >> 10) & 0x7800) | (mVU->code & 0x7ff)) #define _Imm24_ (u32)(mVU->code & 0xffffff) @@ -204,7 +204,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } #define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) -#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & ((isVU1) ? 0x3ff8 : 0xff8)) +#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL From 8e979f1b96f5643ca038bb5f46260d522cb35bb1 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 17 Jun 2009 11:24:42 +0000 Subject: [PATCH 229/296] GSdx: moved 16 bit texture processing from gpu to cpu, with the pixel shader it was either incorrectly done or too slow, please check what got slower/faster or broken :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1375 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPURenderer.cpp | 2 +- plugins/GSdx/GPURendererSW.cpp | 2 +- plugins/GSdx/GS.cpp | 2 +- plugins/GSdx/GSDevice9.cpp | 12 ++- plugins/GSdx/GSDevice9.h | 4 +- plugins/GSdx/GSDeviceOGL.cpp | 137 ++++++++++++++++++++++++++++-- plugins/GSdx/GSDeviceOGL.h | 69 +++++++++++++++ plugins/GSdx/GSLocalMemory.cpp | 5 -- plugins/GSdx/GSLocalMemory.h | 2 - plugins/GSdx/GSRendererHW10.cpp | 8 +- plugins/GSdx/GSRendererHW9.cpp | 8 +- plugins/GSdx/GSTextureCache.cpp | 34 +++++--- plugins/GSdx/GSTextureCache.h | 2 +- plugins/GSdx/GSTextureCache10.cpp | 55 +++--------- plugins/GSdx/GSTextureCache9.cpp | 55 +++--------- plugins/GSdx/GSTextureCacheSW.cpp | 2 +- plugins/GSdx/GSTextureFX10.cpp | 20 ++--- plugins/GSdx/GSTextureFX10.h | 3 +- plugins/GSdx/GSTextureFX9.cpp | 3 +- plugins/GSdx/res/tfx10.fx | 49 +---------- plugins/GSdx/res/tfx9.fx | 2 +- 21 files changed, 271 insertions(+), 205 deletions(-) diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index 954732b2b1..944d5d7914 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -211,6 +211,6 @@ LRESULT GPURenderer::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) } } - return m_wndproc(m_hWnd, message, wParam, lParam); + return CallWindowProc(m_wndproc, m_hWnd, message, wParam, lParam); } diff --git a/plugins/GSdx/GPURendererSW.cpp b/plugins/GSdx/GPURendererSW.cpp index c71f72fdcb..c4b63199c2 100644 --- a/plugins/GSdx/GPURendererSW.cpp +++ b/plugins/GSdx/GPURendererSW.cpp @@ -78,7 +78,7 @@ GSTexture* GPURendererSW::GetOutput() } } - m_texture->Update(GSVector4i(0, 0, w, h), buff, m_mem.GetWidth() * sizeof(uint32)); + m_texture->Update(r.rsize(), buff, m_mem.GetWidth() * sizeof(uint32)); return m_texture; } diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index bf7ce343c9..f364ba8635 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -126,7 +126,7 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; - #if 0 + #if 1 case 6: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index 3d7dd3a93e..a88a24f97e 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -38,7 +38,6 @@ GSDevice9::GSDevice9() , m_ps_ss(NULL) , m_scissor(0, 0, 0, 0) , m_dss(NULL) - , m_sref(0) , m_bs(NULL) , m_bf(0xffffffff) , m_rtv(NULL) @@ -233,7 +232,6 @@ bool GSDevice9::Reset(int w, int h, bool fs) m_ps_ss = NULL; m_scissor = GSVector4i::zero(); m_dss = NULL; - m_sref = 0; m_bs = NULL; m_bf = 0xffffffff; m_rtv = NULL; @@ -570,7 +568,7 @@ void GSDevice9::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, c // om - OMSetDepthStencilState(&m_convert.dss, 0); + OMSetDepthStencilState(&m_convert.dss); OMSetBlendState(bs, 0); OMSetRenderTargets(dt, NULL); @@ -817,6 +815,7 @@ void GSDevice9::PSSetSamplerState(Direct3DSamplerState9* ss) { if(ss && m_ps_ss != ss) { + m_dev->SetSamplerState(0, D3DSAMP_ADDRESSU, ss->AddressU); m_dev->SetSamplerState(0, D3DSAMP_ADDRESSV, ss->AddressV); m_dev->SetSamplerState(1, D3DSAMP_ADDRESSU, ss->AddressU); @@ -850,9 +849,9 @@ void GSDevice9::RSSet(int width, int height, const GSVector4i* scissor) } } -void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref) +void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss) { - if(m_dss != dss || m_sref != sref) + if(m_dss != dss) { m_dev->SetRenderState(D3DRS_ZENABLE, dss->DepthEnable); m_dev->SetRenderState(D3DRS_ZWRITEENABLE, dss->DepthWriteMask); @@ -872,11 +871,10 @@ void GSDevice9::OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 s m_dev->SetRenderState(D3DRS_STENCILPASS, dss->StencilPassOp); m_dev->SetRenderState(D3DRS_STENCILFAIL, dss->StencilFailOp); m_dev->SetRenderState(D3DRS_STENCILZFAIL, dss->StencilDepthFailOp); - m_dev->SetRenderState(D3DRS_STENCILREF, sref); + m_dev->SetRenderState(D3DRS_STENCILREF, dss->StencilRef); } m_dss = dss; - m_sref = sref; } } diff --git a/plugins/GSdx/GSDevice9.h b/plugins/GSdx/GSDevice9.h index 304c8ae502..741c4c4296 100644 --- a/plugins/GSdx/GSDevice9.h +++ b/plugins/GSdx/GSDevice9.h @@ -44,6 +44,7 @@ struct Direct3DDepthStencilState9 D3DSTENCILOP StencilDepthFailOp; D3DSTENCILOP StencilPassOp; D3DCMPFUNC StencilFunc; + uint32 StencilRef; }; struct Direct3DBlendState9 @@ -75,7 +76,6 @@ private: Direct3DSamplerState9* m_ps_ss; GSVector4i m_scissor; Direct3DDepthStencilState9* m_dss; - uint32 m_sref; Direct3DBlendState9* m_bs; uint32 m_bf; IDirect3DSurface9* m_rtv; @@ -168,7 +168,7 @@ public: void PSSetShader(IDirect3DPixelShader9* ps, const float* ps_cb, int ps_cb_len); void PSSetSamplerState(Direct3DSamplerState9* ss); void RSSet(int width, int height, const GSVector4i* scissor = NULL); - void OMSetDepthStencilState(Direct3DDepthStencilState9* dss, uint32 sref); + void OMSetDepthStencilState(Direct3DDepthStencilState9* dss); void OMSetBlendState(Direct3DBlendState9* bs, uint32 bf); void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); diff --git a/plugins/GSdx/GSDeviceOGL.cpp b/plugins/GSdx/GSDeviceOGL.cpp index 8c8e929ed2..fd36ded466 100644 --- a/plugins/GSdx/GSDeviceOGL.cpp +++ b/plugins/GSdx/GSDeviceOGL.cpp @@ -29,7 +29,14 @@ GSDeviceOGL::GSDeviceOGL() , m_hGLRC(NULL) , m_vbo(0) , m_fbo(0) + , m_context(0) , m_topology(-1) + , m_ps_ss(NULL) + , m_scissor(0, 0, 0, 0) + , m_viewport(0, 0) + , m_dss(NULL) + , m_bs(NULL) + , m_bf(-1) , m_rt((GLuint)-1) , m_ds((GLuint)-1) { @@ -41,6 +48,7 @@ GSDeviceOGL::GSDeviceOGL() GSDeviceOGL::~GSDeviceOGL() { + if(m_context) cgDestroyContext(m_context); if(m_vbo) glDeleteBuffers(1, &m_vbo); if(m_fbo) glDeleteFramebuffers(1, &m_fbo); @@ -100,11 +108,6 @@ bool GSDeviceOGL::Create(GSWnd* wnd, bool vsync) return false; } - if(WGLEW_EXT_swap_control) - { - wglSwapIntervalEXT(vsync ? 1 : 0); - } - #endif if(glewInit() != GLEW_OK) @@ -134,6 +137,17 @@ bool GSDeviceOGL::Create(GSWnd* wnd, bool vsync) glBindBuffer(GL_ARRAY_BUFFER, m_vbo); CheckError(); // TODO: setup layout? + m_context = cgCreateContext(); + + // cgGLSetDebugMode(CG_FALSE); CheckCgError(); + + cgSetParameterSettingMode(m_context, CG_DEFERRED_PARAMETER_SETTING); CheckCgError(); +/* + struct {CGprofile vs, gs, ps;} m_profile; + m_profile.vs = cgGLGetLatestProfile(CG_GL_VERTEX); CheckCgError(); + m_profile.gs = cgGLGetLatestProfile(CG_GL_GEOMETRY); CheckCgError(); + m_profile.ps = cgGLGetLatestProfile(CG_GL_FRAGMENT); CheckCgError(); +*/ GSVector4i r = wnd->GetClientRect(); Reset(r.width(), r.height(), false); @@ -387,6 +401,111 @@ void GSDeviceOGL::IASetPrimitiveTopology(int topology) m_topology = topology; } +void GSDeviceOGL::PSSetSamplerState(SamplerStateOGL* ss) +{ + if(ss && m_ps_ss != ss) + { + glActiveTexture(GL_TEXTURE0); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, ss->wrap.s); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, ss->wrap.t); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, ss->filter); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, ss->filter); + + glActiveTexture(GL_TEXTURE1); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, ss->wrap.s); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, ss->wrap.t); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, ss->filter); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, ss->filter); + + glActiveTexture(GL_TEXTURE2); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_POINT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_POINT); + + glActiveTexture(GL_TEXTURE3); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_POINT); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_POINT); + + m_ps_ss = ss; + } +} + +void GSDeviceOGL::RSSet(int width, int height, const GSVector4i* scissor) +{ + if(m_viewport.x != width || m_viewport.y != height) + { + glViewport(0, 0, width, height); CheckError(); + + m_viewport = GSVector2i(width, height); + } + + GSVector4i r = scissor ? *scissor : GSVector4i(0, 0, width, height); + + if(!m_scissor.eq(r)) + { + glScissor(r.left, r.top, r.width(), r.height()); CheckError(); + + m_scissor = r; + } +} + +void GSDeviceOGL::OMSetDepthStencilState(DepthStencilStateOGL* dss) +{ + if(m_dss != dss) + { + if(dss->depth.enable) + { + glEnable(GL_DEPTH_TEST); CheckError(); + glDepthFunc(dss->depth.func); CheckError(); + glDepthMask(dss->depth.write); CheckError(); + } + else + { + glDisable(GL_DEPTH_TEST); CheckError(); + } + + if(dss->stencil.enable) + { + glEnable(GL_STENCIL_TEST); CheckError(); + glStencilFunc(dss->stencil.func, dss->stencil.ref, dss->stencil.mask); CheckError(); + glStencilOp(dss->stencil.sfail, dss->stencil.dpfail, dss->stencil.dppass); CheckError(); + glStencilMask(dss->stencil.wmask); CheckError(); + } + else + { + glDisable(GL_STENCIL_TEST); CheckError(); + } + + m_dss = dss; + } +} + +void GSDeviceOGL::OMSetBlendState(BlendStateOGL* bs, float bf) +{ + if(m_bs != bs || m_bf != bf) + { + if(bs->enable) + { + glEnable(GL_BLEND); CheckError(); + glBlendEquationSeparate(bs->modeRGB, bs->modeAlpha); CheckError(); + glBlendFuncSeparate(bs->srcRGB, bs->dstRGB, bs->srcAlpha, bs->dstAlpha); CheckError(); + glBlendColor(bf, bf, bf, 0); CheckError(); + } + else + { + glDisable(GL_BLEND); CheckError(); + } + + glColorMask(bs->mask.r, bs->mask.g, bs->mask.b, bs->mask.a); CheckError(); + + m_bs = bs; + m_bf = bf; + } +} + void GSDeviceOGL::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) { GLuint rti = 0; @@ -395,17 +514,17 @@ void GSDeviceOGL::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) if(rt) rti = *(GSTextureOGL*)rt; if(ds) dsi = *(GSTextureOGL*)ds; - // TODO: if(m_rt != rti) + if(m_rt != rti) { glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, GL_RENDERBUFFER, rti); CheckError(); - // TODO: m_rt = rti; + m_rt = rti; } - // TODO: if(m_ds != dsi) + if(m_ds != dsi) { glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_DEPTH_COMPONENT, dsi); CheckError(); - // TODO: m_ds = dsi; + m_ds = dsi; } } diff --git a/plugins/GSdx/GSDeviceOGL.h b/plugins/GSdx/GSDeviceOGL.h index d95b8916b3..850056120c 100644 --- a/plugins/GSdx/GSDeviceOGL.h +++ b/plugins/GSdx/GSDeviceOGL.h @@ -24,6 +24,46 @@ #include "GSDevice.h" #include "GSTextureOGL.h" +struct SamplerStateOGL +{ + struct {GLenum s, t;} wrap; + GLenum filter; +}; + +struct DepthStencilStateOGL +{ + struct + { + bool enable; + bool write; + GLenum func; + } depth; + + struct + { + bool enable; + GLenum func; + GLint ref; + GLuint mask; + GLenum sfail; + GLenum dpfail; + GLenum dppass; + GLuint wmask; + } stencil; +}; + +struct BlendStateOGL +{ + bool enable; + GLenum srcRGB; + GLenum dstRGB; + GLenum srcAlpha; + GLenum dstAlpha; + GLenum modeRGB; + GLenum modeAlpha; + union {uint8 r:1, g:1, b:1, a:1;} mask; +}; + class GSDeviceOGL : public GSDevice { #ifdef _WINDOWS @@ -42,11 +82,19 @@ class GSDeviceOGL : public GSDevice } m_vertices; int m_topology; + SamplerStateOGL* m_ps_ss; + GSVector4i m_scissor; + GSVector2i m_viewport; + DepthStencilStateOGL* m_dss; + BlendStateOGL* m_bs; + float m_bf; GLuint m_rt; GLuint m_ds; // + CGcontext m_context; + static void OnStaticCgError(CGcontext ctx, CGerror err, void* p) {((GSDeviceOGL*)p)->OnCgError(ctx, err);} void OnCgError(CGcontext ctx, CGerror err); @@ -89,6 +137,10 @@ public: void IASetInputLayout(); // TODO void IASetPrimitiveTopology(int topology); + void PSSetSamplerState(SamplerStateOGL* ss); + void RSSet(int width, int height, const GSVector4i* scissor); + void OMSetDepthStencilState(DepthStencilStateOGL* dss); + void OMSetBlendState(BlendStateOGL* bs, float bf); void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); static void CheckError() @@ -114,4 +166,21 @@ public: printf("%d\n", status); } } + + void CheckCgError() + { + CGerror error; + + const char* s = cgGetLastErrorString(&error); + + if(error != CG_NO_ERROR) + { + printf("%s\n", s); + + if(error == CG_COMPILER_ERROR) + { + printf("%s\n", cgGetLastListing(m_context)); + } + } + } }; diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index a13bdef939..8f3313dfe3 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -72,11 +72,6 @@ int GSLocalMemory::blockOffset4[256]; // -uint32 GSLocalMemory::m_xtbl[1024]; -uint32 GSLocalMemory::m_ytbl[1024]; - -// - GSLocalMemory::psm_t GSLocalMemory::m_psm[64]; // diff --git a/plugins/GSdx/GSLocalMemory.h b/plugins/GSdx/GSLocalMemory.h index f690fafa2a..a09e972c8d 100644 --- a/plugins/GSdx/GSLocalMemory.h +++ b/plugins/GSdx/GSLocalMemory.h @@ -841,8 +841,6 @@ public: // - static uint32 m_xtbl[1024], m_ytbl[1024]; - template void ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, readTexel rt, readTexture rtx); // diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index c8d4dfaa67..39a66640e3 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -252,17 +252,11 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache GSTextureFX10::VSSelector vs_sel; - vs_sel.bpp = 0; vs_sel.bppz = 0; vs_sel.tme = PRIM->TME; vs_sel.fst = PRIM->FST; vs_sel.prim = prim; - if(tex) - { - vs_sel.bpp = tex->m_bpp2; - } - if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) { if(context->ZBUF.PSM == PSM_PSMZ24) @@ -349,7 +343,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache if(tex) { - ps_sel.bpp = tex->m_bpp2; + ps_sel.bpp = tex->m_bpp; int w = tex->m_texture->GetWidth(); int h = tex->m_texture->GetHeight(); diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 5a37df2530..778bb9ef12 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -50,6 +50,7 @@ bool GSRendererHW9::Create(const string& title) m_date.dss.StencilWriteMask = 1; m_date.dss.StencilFunc = D3DCMP_ALWAYS; m_date.dss.StencilPassOp = D3DSTENCILOP_REPLACE; + m_date.dss.StencilRef = 1; memset(&m_date.bs, 0, sizeof(m_date.bs)); @@ -64,6 +65,7 @@ bool GSRendererHW9::Create(const string& title) m_fba.dss.StencilPassOp = D3DSTENCILOP_ZERO; m_fba.dss.StencilFailOp = D3DSTENCILOP_ZERO; m_fba.dss.StencilDepthFailOp = D3DSTENCILOP_ZERO; + m_fba.dss.StencilRef = 2; memset(&m_fba.bs, 0, sizeof(m_fba.bs)); @@ -321,7 +323,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: if(tex) { - ps_sel.bpp = tex->m_bpp2; + ps_sel.bpp = tex->m_bpp; int w = tex->m_texture->GetWidth(); int h = tex->m_texture->GetHeight(); @@ -523,7 +525,7 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) // om - dev->OMSetDepthStencilState(&m_date.dss, 1); + dev->OMSetDepthStencilState(&m_date.dss); dev->OMSetBlendState(&m_date.bs, 0); dev->OMSetRenderTargets(t, ds); @@ -575,7 +577,7 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) // om - dev->OMSetDepthStencilState(&m_fba.dss, 2); + dev->OMSetDepthStencilState(&m_fba.dss); dev->OMSetBlendState(&m_fba.bs, 0); // ia diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 609e9a5667..6bca8d5265 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -202,13 +202,14 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() { const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; + const GIFRegTEXA& TEXA = m_renderer->m_env.TEXA; + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; const uint32* clut = m_renderer->m_mem.m_clut; - const int pal = GSLocalMemory::m_psm[TEX0.PSM].pal; - if(pal > 0) + if(psm.pal > 0) { - m_renderer->m_mem.m_clut.Read(TEX0); + m_renderer->m_mem.m_clut.Read32(TEX0, TEXA); /* POSITION pos = m_tex.GetHeadPosition(); @@ -267,12 +268,17 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() { GSCachedTexture* t2 = *i; - if((((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) & 0xffefffff) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW (PSM & ~1) TW TH + if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH { continue; } - if(!(pal == 0 || t2->m_TEX0.CPSM == TEX0.CPSM && GSVector4i::compare(t2->m_clut, clut, pal * sizeof(clut[0])))) + if((psm.trbpp == 16 || psm.trbpp == 24) && TEX0.TCC && TEXA != t2->m_TEXA) + { + continue; + } + + if(psm.pal > 0 && !(t2->m_TEX0.CPSM == TEX0.CPSM && GSVector4i::compare(t2->m_clut, clut, psm.pal * sizeof(clut[0])))) { continue; } @@ -344,23 +350,23 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() m_tex.push_front(t); } - if(pal > 0) + if(psm.pal > 0) { - int size = pal * sizeof(clut[0]); + int size = psm.pal * sizeof(clut[0]); if(t->m_palette) { if(t->m_initpalette) { memcpy(t->m_clut, clut, size); - t->m_palette->Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); + t->m_palette->Update(GSVector4i(0, 0, psm.pal, 1), t->m_clut, size); t->m_initpalette = false; } else { if(GSVector4i::update(t->m_clut, clut, size)) { - t->m_palette->Update(GSVector4i(0, 0, pal, 1), t->m_clut, size); + t->m_palette->Update(GSVector4i(0, 0, psm.pal, 1), t->m_clut, size); } } } @@ -728,11 +734,11 @@ void GSTextureCache::GSDepthStencil::Update() GSTextureCache::GSCachedTexture::GSCachedTexture(GSRenderer* r) : GSSurface(r) - , m_valid(0, 0, 0, 0) , m_bpp(0) - , m_bpp2(0) , m_rendered(false) { + m_valid = GSVector4i::zero(); + m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); memset(m_clut, 0, sizeof(m_clut)); @@ -767,7 +773,7 @@ void GSTextureCache::GSCachedTexture::Update() { // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource - m_renderer->m_mem.ReadTextureNP(r, m.bits, m.pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); + m_renderer->m_mem.ReadTexture(r, m.bits, m.pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); m_texture->Unmap(); } @@ -777,12 +783,12 @@ void GSTextureCache::GSCachedTexture::Update() int pitch = ((r.width() + 3) & ~3) * 4; - m_renderer->m_mem.ReadTextureNP(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); + m_renderer->m_mem.ReadTexture(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); m_texture->Update(r, buff, pitch); } - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * m_bpp >> 3); + m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * 4); } bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& rr) diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index a607a18315..ad17e46eb3 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -38,6 +38,7 @@ public: int m_age; GSDirtyRectList m_dirty; GIFRegTEX0 m_TEX0; + GIFRegTEXA m_TEXA; explicit GSSurface(GSRenderer* r); virtual ~GSSurface(); @@ -79,7 +80,6 @@ public: uint32* m_clut; // * GSVector4i m_valid; int m_bpp; - int m_bpp2; bool m_rendered; explicit GSCachedTexture(GSRenderer* renderer); diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 62c6c57b68..2ec9f37db2 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -129,50 +129,13 @@ bool GSTextureCache10::GSCachedTextureHW10::Create() // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; - uint32 psm = m_TEX0.PSM; - - switch(psm) - { - case PSM_PSMT8: - case PSM_PSMT8H: - case PSM_PSMT4: - case PSM_PSMT4HL: - case PSM_PSMT4HH: - psm = m_TEX0.CPSM; - break; - } - - DXGI_FORMAT format; - - switch(psm) - { - default: - // printf("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n", m_TEX0.PSM, m_TEX0.CPSM); - case PSM_PSMCT32: - m_bpp = 32; - m_bpp2 = 0; - format = DXGI_FORMAT_R8G8B8A8_UNORM; - break; - case PSM_PSMCT24: - m_bpp = 32; - m_bpp2 = 1; - format = DXGI_FORMAT_R8G8B8A8_UNORM; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 16; - m_bpp2 = 5; - format = DXGI_FORMAT_R16_UNORM; - break; - } - - int w = 1 << m_TEX0.TW; - int h = 1 << m_TEX0.TH; + m_bpp = 0; ASSERT(m_texture == NULL); - m_texture = m_renderer->m_dev->CreateTexture(w, h, format); + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); return m_texture != NULL; } @@ -186,6 +149,8 @@ bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + m_rendered = true; int tw = 1 << m_TEX0.TW; @@ -309,18 +274,18 @@ bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) switch(m_TEX0.PSM) { case PSM_PSMCT32: - m_bpp2 = 0; + m_bpp = 0; break; case PSM_PSMCT24: - m_bpp2 = 1; + m_bpp = 1; break; case PSM_PSMCT16: case PSM_PSMCT16S: - m_bpp2 = 2; + m_bpp = 2; break; case PSM_PSMT8H: - m_bpp2 = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? DXGI_FORMAT_R8G8B8A8_UNORM : DXGI_FORMAT_R16_UNORM); // + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); m_initpalette = true; break; case PSM_PSMT4HL: diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index d288c5ebac..e9b685ae17 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -127,50 +127,13 @@ bool GSTextureCache9::GSCachedTexture9::Create() // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; - uint32 psm = m_TEX0.PSM; - - switch(psm) - { - case PSM_PSMT8: - case PSM_PSMT8H: - case PSM_PSMT4: - case PSM_PSMT4HL: - case PSM_PSMT4HH: - psm = m_TEX0.CPSM; - break; - } - - D3DFORMAT format; - - switch(psm) - { - default: - // printf("Invalid TEX0.PSM/CPSM (%I64d, %I64d)\n", m_TEX0.PSM, m_TEX0.CPSM); - case PSM_PSMCT32: - m_bpp = 32; - m_bpp2 = 0; - format = D3DFMT_A8R8G8B8; - break; - case PSM_PSMCT24: - m_bpp = 32; - m_bpp2 = 1; - format = D3DFMT_A8R8G8B8; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 16; - m_bpp2 = 2; - format = D3DFMT_A1R5G5B5; - break; - } - - int w = 1 << m_TEX0.TW; - int h = 1 << m_TEX0.TH; + m_bpp = 0; ASSERT(m_texture == NULL); - m_texture = m_renderer->m_dev->CreateTexture(w, h, format); + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); return m_texture != NULL; } @@ -184,6 +147,8 @@ bool GSTextureCache9::GSCachedTexture9::Create(GSRenderTarget* rt) // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + m_rendered = true; int tw = 1 << m_TEX0.TW; @@ -307,18 +272,18 @@ bool GSTextureCache9::GSCachedTexture9::Create(GSRenderTarget* rt) switch(m_TEX0.PSM) { case PSM_PSMCT32: - m_bpp2 = 0; + m_bpp = 0; break; case PSM_PSMCT24: - m_bpp2 = 1; + m_bpp = 1; break; case PSM_PSMCT16: case PSM_PSMCT16S: - m_bpp2 = 2; + m_bpp = 2; break; case PSM_PSMT8H: - m_bpp2 = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1, m_TEX0.CPSM == PSM_PSMCT32 ? D3DFMT_A8R8G8B8 : D3DFMT_A1R5G5B5); + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); m_initpalette = true; break; case PSM_PSMT4HL: diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index ca7144a438..1be3e4b6a6 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -51,7 +51,7 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE continue; } - if((psm.trbpp == 16 || psm.trbpp == 24) && (t2->m_TEX0.TCC != TEX0.TCC || TEX0.TCC && TEXA != t2->m_TEXA)) + if((psm.trbpp == 16 || psm.trbpp == 24) && TEX0.TCC && TEXA != t2->m_TEXA) { continue; } diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index b00de4e1d0..a3b04e1cc9 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -104,21 +104,19 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) if(i == m_vs.end()) { - string str[5]; + string str[4]; - str[0] = format("%d", sel.bpp); - str[1] = format("%d", sel.bppz); - str[2] = format("%d", sel.tme); - str[3] = format("%d", sel.fst); - str[4] = format("%d", sel.prim); + str[0] = format("%d", sel.bppz); + str[1] = format("%d", sel.tme); + str[2] = format("%d", sel.fst); + str[3] = format("%d", sel.prim); D3D10_SHADER_MACRO macro[] = { - {"VS_BPP", str[0].c_str()}, - {"VS_BPPZ", str[1].c_str()}, - {"VS_TME", str[2].c_str()}, - {"VS_FST", str[3].c_str()}, - {"VS_PRIM", str[4].c_str()}, + {"VS_BPPZ", str[0].c_str()}, + {"VS_TME", str[1].c_str()}, + {"VS_FST", str[2].c_str()}, + {"VS_PRIM", str[3].c_str()}, {NULL, NULL}, }; diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index f50617052b..285d930538 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -63,7 +63,6 @@ public: { struct { - uint32 bpp:3; uint32 bppz:2; uint32 tme:1; uint32 fst:1; @@ -72,7 +71,7 @@ public: uint32 key; - operator uint32() {return key & 0x3ff;} + operator uint32() {return key & 0x7f;} }; __declspec(align(16)) struct PSConstantBuffer diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index f59d70246f..c303db9652 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -293,6 +293,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, dss->StencilPassOp = dssel.fba ? D3DSTENCILOP_REPLACE : D3DSTENCILOP_KEEP; dss->StencilFailOp = dssel.fba ? D3DSTENCILOP_ZERO : D3DSTENCILOP_KEEP; dss->StencilDepthFailOp = dssel.fba ? D3DSTENCILOP_ZERO : D3DSTENCILOP_KEEP; + dss->StencilRef = 3; } if(!(dssel.zte && dssel.ztst == 1 && !dssel.zwe)) @@ -315,7 +316,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState((*i).second, 3); + m_dev->OMSetDepthStencilState((*i).second); hash_map::const_iterator j = m_om_bs.find(bsel); diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index a2b067a93b..529e528ab7 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -22,8 +22,7 @@ struct VS_OUTPUT float4 c : COLOR0; }; -#ifndef VS_BPP -#define VS_BPP 0 +#ifndef VS_BPPZ #define VS_BPPZ 0 #define VS_TME 1 #define VS_FST 1 @@ -193,23 +192,6 @@ struct PS_OUTPUT #define LTF 1 #endif -float4 Normalize16(float4 f) -{ - return f / float4(0x001f, 0x03e0, 0x7c00, 0x8000); -} - -float4 Extract16(uint i) -{ - float4 f; - - f.r = i & 0x001f; - f.g = i & 0x03e0; - f.b = i & 0x7c00; - f.a = i & 0x8000; - - return f; -} - float4 wrapuv(float4 uv) { if(WMS == WMT) @@ -318,7 +300,7 @@ float4 sample(float2 tc, float w) float4 t00, t01, t10, t11; - if(BPP == 3) // 8HP + 32-bit palette + if(BPP == 3) // 8HP { float4 a; @@ -332,26 +314,6 @@ float4 sample(float2 tc, float w) t10 = Palette.Sample(PaletteSampler, a.z); t11 = Palette.Sample(PaletteSampler, a.w); } - else if(BPP == 4) // 8HP + 16-bit palette - { - // TODO: yuck, just pre-convert the palette to 32-bit - } - else if(BPP == 5) // 16P - { - float4 r; - - r.x = Texture.Sample(TextureSampler, uv.xy).r; - r.y = Texture.Sample(TextureSampler, uv.zy).r; - r.z = Texture.Sample(TextureSampler, uv.xw).r; - r.w = Texture.Sample(TextureSampler, uv.zw).r; - - uint4 i = r * 65535; - - t00 = Extract16(i.x); - t01 = Extract16(i.y); - t10 = Extract16(i.z); - t11 = Extract16(i.w); - } else { t00 = Texture.Sample(TextureSampler, uv.xy); @@ -374,13 +336,8 @@ float4 sample(float2 tc, float w) { t.a = AEM == 0 || any(t.rgb) ? TA.x : 0; } - else if(BPP == 2 || BPP == 5) // 16 || 16P + else if(BPP == 2) // 16 { - if(BPP == 5) - { - t = Normalize16(t); - } - // a bit incompatible with up-scaling because the 1 bit alpha is interpolated t.a = t.a >= 0.5 ? TA.y : AEM == 0 || any(t.rgb) ? TA.x : 0; diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index 1cbc591387..c7a2c57896 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -219,7 +219,7 @@ float4 sample(float2 tc, float w) float4 t00, t01, t10, t11; - if(BPP == 3) // 8HP ln + if(BPP == 3) // 8HP { float4 a; From caa7a5dffd57a19585eb48e4c6236aef9d3c2075 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 17 Jun 2009 11:36:24 +0000 Subject: [PATCH 230/296] GSdx: forgot to disable opengl code, glew32.dll isn't needed yet but cannot be delay loaded git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1376 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GS.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index f364ba8635..bf7ce343c9 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -126,7 +126,7 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; - #if 1 + #if 0 case 6: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; From 3ddcf45325db366d9cf4ce6341717d7a0c892748 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 17 Jun 2009 14:34:01 +0000 Subject: [PATCH 231/296] Minor cleanups to my mVU commit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1377 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 20 ++++++++++---------- pcsx2/x86/microVU.h | 10 +++++----- pcsx2/x86/microVU_Compile.inl | 20 ++++++++++---------- pcsx2/x86/microVU_Log.inl | 2 +- pcsx2/x86/microVU_Misc.h | 8 ++++---- 5 files changed, 30 insertions(+), 30 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index f3403d6455..b56e266f1b 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -40,8 +40,8 @@ microVUf(void) mVUinit(VURegs* vuRegsPtr) { microVU* mVU = mVUx; mVU->regs = vuRegsPtr; mVU->index = vuIndex; - mVU->microSize = (vuIndex ? 0x4000 : 0x1000); - mVU->progSize = (vuIndex ? 0x4000 : 0x1000) / 4; + mVU->microMemSize = (vuIndex ? 0x4000 : 0x1000); + mVU->progMemSize = (vuIndex ? 0x4000 : 0x1000) / 4; mVU->cache = NULL; mVU->cacheSize = mVUcacheSize; memset(&mVU->prog, 0, sizeof(mVU->prog)); @@ -62,7 +62,7 @@ microVUx(void) mVUreset() { // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progSize / 2); j++) { + for (u32 j = 0; j < (mVU->progMemSize / 2); j++) { microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } } @@ -108,7 +108,7 @@ microVUf(void) mVUclose() { // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progSize / 2); j++) { + for (u32 j = 0; j < (mVU->progMemSize / 2); j++) { if (mVU->prog.prog[i].block[j]) { microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } @@ -137,7 +137,7 @@ microVUf(void) mVUclearProg(int progIndex) { mVU->prog.prog[progIndex].range[0] = -1; mVU->prog.prog[progIndex].range[1] = -1; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; - for (u32 i = 0; i < (mVU->progSize / 2); i++) { + for (u32 i = 0; i < (mVU->progMemSize / 2); i++) { if (mVU->prog.prog[progIndex].block[i]) mVU->prog.prog[progIndex].block[i]->reset(); } @@ -146,7 +146,7 @@ microVUf(void) mVUclearProg(int progIndex) { // Caches Micro Program microVUf(void) mVUcacheProg(int progIndex) { microVU* mVU = mVUx; - memcpy_fast(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microSize); + memcpy_fast(mVU->prog.prog[progIndex].data, mVU->regs->Micro, mVU->microMemSize); mVUdumpProg(progIndex); } @@ -214,7 +214,7 @@ microVUf(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, b microVU* mVU = mVUx; if (progUsed) { - if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microSize)) || + if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) || (!cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data + mVUprogI.range[0], (u8*)mVU->regs->Micro + mVUprogI.range[0], ((mVUprogI.range[1] + 8) - mVUprogI.range[0]))))) { mVU->prog.cur = progIndex; mVU->prog.cleared = 0; @@ -275,7 +275,7 @@ static void* __fastcall mVUcompile( microVU* mVU, u32 startPC, uptr pState ) microBlock* pBlock = NULL; u8* thisPtr = x86Ptr; - const u32 microSizeDiv8 = (mVU->microSize-1) / 8; + const u32 microSizeDiv8 = (mVU->microMemSize-1) / 8; // First Pass iPC = startPC / 4; @@ -433,8 +433,8 @@ microVUt(void*) mVUblockFetch( microVU* mVU, u32 startPC, uptr pState ) { using namespace x86Emitter; - if (startPC > mVU->microSize-1) { Console::Error("microVU%d: invalid startPC", params mVU->index); } - startPC &= ~7; + if (startPC > mVU->microMemSize-8) { Console::Error("microVU%d: invalid startPC", params mVU->index); } + startPC &= mVU->microMemSize-8; if (mVUblocks[startPC/8] == NULL) { mVUblocks[startPC/8] = microBlockManager::AlignedNew(); diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index c722b0e5f8..0d0cdad2f3 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -64,7 +64,7 @@ public: } return thisBlock; } - microBlock* search(microRegInfo* pState) { + __releaseinline microBlock* search(microRegInfo* pState) { if (listSize < 0) return NULL; if (pState->needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) for (int i = 0; i <= listSize; i++) { @@ -116,10 +116,10 @@ struct microVU { PCSX2_ALIGNED16(u32 xmmPQb[4]); // Backup for xmmPQ PCSX2_ALIGNED16(u32 xmmVFb[4]); // Backup for VF regs - u32 index; // VU Index (VU0 or VU1) - u32 microSize; // VU Micro Memory Size - u32 progSize; // VU Micro Program Size (microSize/4) - u32 cacheSize; // VU Cache Size + u32 index; // VU Index (VU0 or VU1) + u32 microMemSize; // VU Micro Memory Size (in bytes) + u32 progMemSize; // VU Micro Program Size (microSize/4, because each instruction of a program is 4 bytes) + u32 cacheSize; // VU Cache Size microProgManager<0x4000/4> prog; // Micro Program Data diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 6b713e24a9..01a2095f7b 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -22,20 +22,20 @@ // Helper Macros //------------------------------------------------------------------ -#define branchCase(JMPcc) \ +#define branchCase(JMPcc) \ mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ - xCMP( ptr16[&mVU->branch], 0); \ + xCMP( ptr16[&mVU->branch], 0); \ if (mVUup.eBit) { /* Conditional Branch With E-Bit Set */ \ mVUendProgram(mVU, 2, xStatus, xMac, xClip); \ - xForwardJump8 eJMP( JMPcc ); \ + xForwardJump8 eJMP( JMPcc ); \ incPC(1); /* Set PC to First instruction of Non-Taken Side */ \ - xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ - xJMP( mVU->exitFunct ); \ - eJMP.SetTarget(); \ + xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ + xJMP( mVU->exitFunct ); \ + eJMP.SetTarget(); \ incPC(-4); /* Go Back to Branch Opcode to get branchAddr */ \ iPC = branchAddr/4; \ - xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ - xJMP( mVU->exitFunct ); \ + xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ + xJMP( mVU->exitFunct ); \ return thisPtr; \ } \ else { /* Normal Conditional Branch */ \ @@ -43,7 +43,7 @@ if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ incPC2(-1); \ - if (bBlock) { xJcc( xInvertCond( JMPcc ), bBlock->x86ptrStart ); } \ + if (bBlock) { xJcc( xInvertCond( JMPcc ), bBlock->x86ptrStart ); } \ else { ajmp = xJcc32( JMPcc ); } \ } \ break @@ -102,7 +102,7 @@ microVUt(void) mVUcheckIsSame(mV) { if (mVU->prog.isSame == -1) { - mVU->prog.isSame = !memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microSize); + mVU->prog.isSame = !memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microMemSize); } if (mVU->prog.isSame == 0) { if (!isVU1) mVUcacheProg<0>(mVU->prog.cur); diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index 4b84ab61cc..a721642b01 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -62,7 +62,7 @@ microVUx(void) __mVUdumpProgram(int progIndex) { mVUlog("*********************\n\n

", progIndex); mVUlog(""); - for (u32 i = 0; i < mVU->progSize; i+=2) { + for (u32 i = 0; i < mVU->progMemSize; i+=2) { if (delay) { delay--; mVUlog(""); if (!delay) mVUlog("
"); } if (mVUbranch) { delay = 1; mVUbranch = 0; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 66690240e7..32089b006b 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -163,7 +163,7 @@ declareAllVariables // Define mVUquickSearch #ifndef __LINUX__ -extern u8 mVUsearchXMM[0x1000]; +PCSX2_ALIGNED16_EXTERN( u8 mVUsearchXMM[0x1000] ); typedef u32 (__fastcall *mVUCall)(void*, void*); #define mVUquickSearch(dest, src, size) ((((mVUCall)((void*)mVUsearchXMM))(dest, src)) == 0xf) #define mVUemitSearch() { mVUcustomSearch(); } @@ -201,10 +201,10 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define xPC ((iPC / 2) * 8) #define curI ((u32*)mVU->regs->Micro)[iPC] //mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } -#define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } -#define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } +#define incPC(x) { iPC = ((iPC + x) & (mVU->progMemSize-1)); setCode(); } +#define incPC2(x) { iPC = ((iPC + x) & (mVU->progMemSize-1)); } #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) -#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microSize-8)) +#define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microMemSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL From 1845c3cc381dfded5e22b042acf43d54d8b06377 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Wed, 17 Jun 2009 15:18:12 +0000 Subject: [PATCH 232/296] GSdx: just updating the vs2010 project files... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1378 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSdx_vs2010.vcxproj | 8 ++++++++ plugins/GSdx/GSdx_vs2010.vcxproj.filters | 24 ++++++++++++++++++++++++ plugins/GSdx/vsprops/common.props | 2 +- 3 files changed, 33 insertions(+), 1 deletion(-) diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj b/plugins/GSdx/GSdx_vs2010.vcxproj index fb4053ff4d..d4e7a8e2b2 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj +++ b/plugins/GSdx/GSdx_vs2010.vcxproj @@ -1306,6 +1306,7 @@ + @@ -1321,6 +1322,7 @@ + @@ -1334,10 +1336,12 @@ + + @@ -1418,6 +1422,7 @@ + @@ -1435,6 +1440,7 @@ + @@ -1449,10 +1455,12 @@ + + diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj.filters b/plugins/GSdx/GSdx_vs2010.vcxproj.filters index ee1b5c34d3..5cfd89b885 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj.filters +++ b/plugins/GSdx/GSdx_vs2010.vcxproj.filters @@ -159,6 +159,9 @@ Source Files + + Source Files + Source Files @@ -204,6 +207,9 @@ Source Files + + Source Files + Source Files @@ -243,6 +249,9 @@ Source Files + + Source Files + Source Files @@ -255,6 +264,9 @@ Source Files + + Source Files + Source Files @@ -440,6 +452,9 @@ Header Files + + Header Files + Header Files @@ -491,6 +506,9 @@ Header Files + + Header Files + Header Files @@ -533,6 +551,9 @@ Header Files + + Header Files + Header Files @@ -545,6 +566,9 @@ Header Files + + Header Files + Header Files diff --git a/plugins/GSdx/vsprops/common.props b/plugins/GSdx/vsprops/common.props index 488eba3520..357fe0d1b6 100644 --- a/plugins/GSdx/vsprops/common.props +++ b/plugins/GSdx/vsprops/common.props @@ -15,7 +15,7 @@ 4995;4324;%(DisableSpecificWarnings) - JITProfiling.lib;d3d10.lib;d3dx10.lib;d3d9.lib;d3dx9.lib;ddraw.lib;dxguid.lib;winmm.lib;strmiids.lib;xinput.lib;%(AdditionalDependencies) + JITProfiling.lib;d3d10.lib;d3dx10.lib;d3d9.lib;d3dx9.lib;ddraw.lib;dxguid.lib;winmm.lib;strmiids.lib;xinput.lib;cg.lib;cgGL.lib;glut32.lib;glew32.lib;%(AdditionalDependencies) ./vtune;%(AdditionalLibraryDirectories) d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll;%(DelayLoadDLLs) true From 43dbbb050f97992bb2f6f5c5a9a2be71987c87b7 Mon Sep 17 00:00:00 2001 From: refraction Date: Wed, 17 Jun 2009 17:49:01 +0000 Subject: [PATCH 233/296] Fix for Tekken Tag Issue 271 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1379 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index ac18745ca9..cb884ed0a1 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -535,6 +535,12 @@ void gifMFIFOInterrupt() mfifocycles = 0; if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + if((spr0->chcr & 0x100) && spr0->qwc == 0) + { + spr0->chcr &= ~0x100; + hwDmacIrq(DMAC_FROM_SPR); + } + if (!(gif->chcr & 0x100)) { Console::WriteLn("WTF GIFMFIFO"); cpuRegs.interrupt &= ~(1 << 11); @@ -546,13 +552,7 @@ void gifMFIFOInterrupt() //GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); CPU_INT(11,mfifocycles); return; - } - - if((spr0->chcr & 0x100) && spr0->qwc == 0) - { - spr0->chcr &= ~0x100; - hwDmacIrq(DMAC_FROM_SPR); - } + } if(gifstate != GIF_STATE_STALL) { if(gifqwc <= 0) { From 78746f0d3acf8654bc9b7bbeb03f8bf48754b3fe Mon Sep 17 00:00:00 2001 From: refraction Date: Wed, 17 Jun 2009 20:58:03 +0000 Subject: [PATCH 234/296] Modified how the GS interrupts work, solves the issue of the picture flashing on and off in some games (Street Fighter EX3). Needs some testing possibly, but should be better. Developer note: The reason for this is the event can only generate an interrupt if it is enabled (via CSRw). the CSRw has to also reflect the IMR status else it could give false readings. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1380 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Counters.cpp | 16 ++++++++++++---- pcsx2/GS.cpp | 3 ++- pcsx2/MTGS.cpp | 17 ++++++++++------- 3 files changed, 24 insertions(+), 12 deletions(-) diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index d8af6b3f80..62398ea880 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -336,8 +336,12 @@ static __forceinline void VSyncStart(u32 sCycle) EECNT_LOG( "///////// EE COUNTER VSYNC START \\\\\\\\\\\\\\\\\\\\ (frame: %d)", iFrame ); vSyncDebugStuff( iFrame ); // EE Profiling and Debug code - if ((CSRw & 0x8)) GSCSRr|= 0x8; - if (!(GSIMR&0x800)) gsIrq(); + if ((CSRw & 0x8)) + { + CSRw &= ~0x8; + GSCSRr|= 0x8; + if (!(GSIMR&0x800)) gsIrq(); + } hwIntcIrq(INTC_VBLANK_S); psxVBlankStart(); @@ -404,8 +408,12 @@ __forceinline void rcntUpdate_hScanline() hsyncCounter.Mode = MODE_HRENDER; } else { //HBLANK END / HRENDER Begin - if (CSRw & 0x4) GSCSRr |= 4; // signal - if (!(GSIMR&0x400)) gsIrq(); + if (CSRw & 0x4) + { + CSRw &= ~0x4; + GSCSRr |= 4; // signal + if (!(GSIMR&0x400)) gsIrq(); + } if (gates) rcntEndGate(false, hsyncCounter.sCycle); if (psxhblankgate) psxCheckEndGate16(0); diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index 8b8fe7e958..287a063dfc 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -243,6 +243,7 @@ void gsReset() memzero_obj(g_RealGSMem); + CSRw = 0; GSCSRr = 0x551B400F; // Set the FINISH bit to 1 for now GSIMR = 0x7f00; psHu32(GIF_STAT) = 0; @@ -323,7 +324,7 @@ void gsCSRwrite(u32 value) static void IMRwrite(u32 value) { GSIMR = (value & 0x1f00)|0x6000; - + CSRw |= (GSIMR >> 8) & 0x1f; // don't update mtgs mem } diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 55195caf5e..17416eb8e3 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -142,10 +142,12 @@ static void RegHandlerSIGNAL(const u32* data) GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); if ((CSRw & 0x1)) + { + CSRw &= ~0x1; GSCSRr |= 1; // signal - - if (!(GSIMR&0x100) ) - gsIrq(); + if (!(GSIMR&0x100) ) + gsIrq(); + } } static void RegHandlerFINISH(const u32* data) @@ -153,11 +155,12 @@ static void RegHandlerFINISH(const u32* data) MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); if ((CSRw & 0x2)) + { GSCSRr |= 2; // finish - - if (!(GSIMR&0x200) ) - gsIrq(); - + CSRw &= ~0x2; + if (!(GSIMR&0x200) ) + gsIrq(); + } } static void RegHandlerLABEL(const u32* data) From 4bd993e13573f121e1629a067e419257111fa55f Mon Sep 17 00:00:00 2001 From: refraction Date: Wed, 17 Jun 2009 21:04:13 +0000 Subject: [PATCH 235/296] Nothing to see here, small missed thing :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1381 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/GS.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index 287a063dfc..cb7e28edf6 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -324,7 +324,7 @@ void gsCSRwrite(u32 value) static void IMRwrite(u32 value) { GSIMR = (value & 0x1f00)|0x6000; - CSRw |= (GSIMR >> 8) & 0x1f; + CSRw |= ~(GSIMR >> 8) & 0x1f; // don't update mtgs mem } From 5278275dd442557485a6badc2e12aa64d761998b Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 00:13:50 +0000 Subject: [PATCH 236/296] Big fixes just for my friend Ferrarinews cos im sick of him GT4: Loading screen corruption + Freezing ingame fixed GT3: Freezing ingame fixed Soul Calibur 3: Now loads again Soul Calibur 2: Now loads again and should no longer have the Timeout issue. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1382 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Counters.cpp | 16 ++++------------ pcsx2/GS.cpp | 18 +++++++++++------- pcsx2/Gif.cpp | 10 ++++++++-- pcsx2/MTGS.cpp | 21 +++++++++------------ pcsx2/VifDma.cpp | 2 +- 5 files changed, 33 insertions(+), 34 deletions(-) diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index 62398ea880..d8af6b3f80 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -336,12 +336,8 @@ static __forceinline void VSyncStart(u32 sCycle) EECNT_LOG( "///////// EE COUNTER VSYNC START \\\\\\\\\\\\\\\\\\\\ (frame: %d)", iFrame ); vSyncDebugStuff( iFrame ); // EE Profiling and Debug code - if ((CSRw & 0x8)) - { - CSRw &= ~0x8; - GSCSRr|= 0x8; - if (!(GSIMR&0x800)) gsIrq(); - } + if ((CSRw & 0x8)) GSCSRr|= 0x8; + if (!(GSIMR&0x800)) gsIrq(); hwIntcIrq(INTC_VBLANK_S); psxVBlankStart(); @@ -408,12 +404,8 @@ __forceinline void rcntUpdate_hScanline() hsyncCounter.Mode = MODE_HRENDER; } else { //HBLANK END / HRENDER Begin - if (CSRw & 0x4) - { - CSRw &= ~0x4; - GSCSRr |= 4; // signal - if (!(GSIMR&0x400)) gsIrq(); - } + if (CSRw & 0x4) GSCSRr |= 4; // signal + if (!(GSIMR&0x400)) gsIrq(); if (gates) rcntEndGate(false, hsyncCounter.sCycle); if (psxhblankgate) psxCheckEndGate16(0); diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index cb7e28edf6..189327790f 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -243,8 +243,7 @@ void gsReset() memzero_obj(g_RealGSMem); - CSRw = 0; - GSCSRr = 0x551B400F; // Set the FINISH bit to 1 for now + GSCSRr = 0x551B4000; // Set the FINISH bit to 1 for now GSIMR = 0x7f00; psHu32(GIF_STAT) = 0; psHu32(GIF_CTRL) = 0; @@ -276,12 +275,14 @@ void gsGIFReset() { // fixme - should this be here? (air) //memzero_obj(g_RealGSMem); + // none of this should be here, its a GIF reset, not GS, only the dma side of it is reset. (Refraction) // perform a soft reset (but do not do a full reset if the soft reset API is unavailable) - gsGIFSoftReset( 7 ); + //gsGIFSoftReset( 7 ); - GSCSRr = 0x551B400F; // Set the FINISH bit to 1 for now - GSIMR = 0x7f00; + + //GSCSRr = 0x551B4000; // Set the FINISH bit to 1 for now + //GSIMR = 0x7f00; psHu32(GIF_STAT) = 0; psHu32(GIF_CTRL) = 0; psHu32(GIF_MODE) = 0; @@ -316,7 +317,7 @@ void gsCSRwrite(u32 value) GSreset(); } - GSCSRr = 0x551B400F; // Set the FINISH bit to 1 - GS is always at a finish state as we don't have a FIFO(saqib) + GSCSRr = 0x551B4000; // Set the FINISH bit to 1 - GS is always at a finish state as we don't have a FIFO(saqib) GSIMR = 0x7F00; //This is bits 14-8 thats all that should be 1 } } @@ -324,7 +325,8 @@ void gsCSRwrite(u32 value) static void IMRwrite(u32 value) { GSIMR = (value & 0x1f00)|0x6000; - CSRw |= ~(GSIMR >> 8) & 0x1f; + + if((GSCSRr & 0x1f) & (~(GSIMR >> 8) & 0x1f)) gsIrq(); // don't update mtgs mem } @@ -434,6 +436,8 @@ void __fastcall gsWrite64_page_00( u32 mem, const mem64_t* value ) void __fastcall gsWrite64_page_01( u32 mem, const mem64_t* value ) { + GIF_LOG("GS Write64 at %8.8lx with data %8.8x_%8.8x", mem, (u32*)value[1], (u32*)value[0]); + switch( mem ) { case GS_CSR: diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index cb884ed0a1..65f27fea9d 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -57,6 +57,12 @@ __forceinline void gsInterrupt() { //Console::WriteLn("Eh? why are you still interrupting! chcr %x, qwc %x, done = %x", params gif->chcr, gif->qwc, done); return; } + + if((vif1.cmd & 0x7f) == 0x51 && Path3progress != 0) + { + vif1Regs->stat &= ~VIF1_STAT_VGW; + } + if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 if (gif->qwc > 0 || gspath3done == 0) { @@ -193,7 +199,7 @@ void GIFdma() if(((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == 0) //Path2 gets priority in intermittent mode { - //GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); + GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); dmaGIFend(); return; } @@ -287,7 +293,7 @@ void GIFdma() return; } } - GIFchain(); //Transfers the data set by the switch + //GIFchain(); //Transfers the data set by the switch if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag GIF_LOG("dmaIrq Set"); diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 17416eb8e3..5ec6f831c1 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -137,17 +137,15 @@ void SaveState::mtgsFreeze() static void RegHandlerSIGNAL(const u32* data) { - MTGS_LOG("MTGS SIGNAL data %x_%x CSRw %x\n",data[0], data[1], CSRw); + GIF_LOG("MTGS SIGNAL data %x_%x CSRw %x IMR %x CSRr\n",data[0], data[1], CSRw, GSIMR, GSCSRr); GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); if ((CSRw & 0x1)) - { - CSRw &= ~0x1; GSCSRr |= 1; // signal - if (!(GSIMR&0x100) ) - gsIrq(); - } + + if (!(GSIMR&0x100) ) + gsIrq(); } static void RegHandlerFINISH(const u32* data) @@ -155,12 +153,11 @@ static void RegHandlerFINISH(const u32* data) MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); if ((CSRw & 0x2)) - { GSCSRr |= 2; // finish - CSRw &= ~0x2; - if (!(GSIMR&0x200) ) - gsIrq(); - } + + if (!(GSIMR&0x200) ) + gsIrq(); + } static void RegHandlerLABEL(const u32* data) @@ -529,7 +526,7 @@ int mtgsThreadObject::Callback() Console::WriteLn( "MTGS > GSopen Finished, return code: 0x%x", params m_returncode ); - GSCSRr = 0x551B400F; // 0x55190000 + GSCSRr = 0x551B4000; // 0x55190000 m_post_InitDone.Post(); if (m_returncode != 0) { return m_returncode; } // error msg will be issued to the user by Plugins.c diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index e6ca05a339..d7428c52ce 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2120,7 +2120,7 @@ static void Vif1CMDDirectHL() // DIRECT/HL //FIXME: This should have timing in both cases, see note below. if((vif1.cmd & 0x7f) == 0x51) { - if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 || Path3progress != 2)) //PATH3 is in image mode, so wait for end of transfer + if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 && Path3progress == 0)) //PATH3 is in image mode, so wait for end of transfer { //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); /*if(vif1Regs->mskpath3)*/vif1Regs->stat |= VIF1_STAT_VGW; From 378489fa1e3422e81ae11eda02b9282a20089c21 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 00:47:23 +0000 Subject: [PATCH 237/296] Seems all "Events" are enabled by default, which broke after my recent fix. Thanks for that Capcom git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1383 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/GS.cpp | 1 + 1 file changed, 1 insertion(+) diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index 189327790f..53789055bc 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -317,6 +317,7 @@ void gsCSRwrite(u32 value) GSreset(); } + CSRw = 0x1f; GSCSRr = 0x551B4000; // Set the FINISH bit to 1 - GS is always at a finish state as we don't have a FIFO(saqib) GSIMR = 0x7F00; //This is bits 14-8 thats all that should be 1 } From af0b33c7b5e20fdea45bb6e3732c0c8760a2e89e Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 18 Jun 2009 01:07:28 +0000 Subject: [PATCH 238/296] GSdx: minor fixes, ar tonelico 2 shadows in sw mode, lamune palette problems in hw mode git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1384 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSCrc.h | 1 + plugins/GSdx/GSDrawScanlineCodeGenerator.cpp | 15 +++++++-------- plugins/GSdx/GSRendererHW.h | 4 ++-- 4 files changed, 11 insertions(+), 10 deletions(-) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index b55978635e..7e27702525 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -28,6 +28,7 @@ CRC::Game CRC::m_games[] = {0x2113EA2E, MetalSlug6, Unknown}, {0x42E05BAF, TomoyoAfter, JP}, {0x7800DC84, Clannad, JP}, + {0xA6167B59, Lamune, JP}, {0xa39517ab, FFX, EU}, {0xa39517ae, FFX, FR}, {0x941bb7d9, FFX, DE}, diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index cc64037f8f..b47362e690 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -30,6 +30,7 @@ public: MetalSlug6, TomoyoAfter, Clannad, + Lamune, FFX, FFX2, FFXII, diff --git a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp index 6d61fb2b5b..cb1d003a72 100644 --- a/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp +++ b/plugins/GSdx/GSDrawScanlineCodeGenerator.cpp @@ -707,18 +707,17 @@ void GSDrawScanlineCodeGenerator::TestZ(const Xmm& temp1, const Xmm& temp2) switch(m_sel.ztst) { case ZTST_GEQUAL: - // test |= zso < zdo; + // test |= zso < zdo; // ~(zso >= zdo) pcmpgtd(xmm1, xmm0); por(xmm7, xmm1); break; - case ZTST_GREATER: - // test |= zso <= zdo; - movdqa(xmm4, xmm1); - pcmpgtd(xmm1, xmm0); - por(xmm7, xmm1); - pcmpeqd(xmm4, xmm0); - por(xmm7, xmm1); + case ZTST_GREATER: // TODO: tidus hair and chocobo wings only appear fully when this is tested as ZTST_GEQUAL + // test |= zso <= zdo; // ~(zso > zdo) + pcmpgtd(xmm0, xmm1); + pcmpeqd(xmm4, xmm4); + pxor(xmm0, xmm4); + por(xmm7, xmm0); break; } diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index aea77663f3..f58aa4e787 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -494,9 +494,9 @@ protected: #pragma endregion - #pragma region tomoyo after, clannad (palette uploaded in a point list, pure genius...) + #pragma region tomoyo after, clannad, lamune (palette uploaded in a point list, pure genius...) - if(m_game.title == CRC::TomoyoAfter || m_game.title == CRC::Clannad) + if(m_game.title == CRC::TomoyoAfter || m_game.title == CRC::Clannad || m_game.title == CRC::Lamune) { if(prim == GS_POINTLIST && !PRIM->TME) { From df2b0365addb63538be2ccd4cd12f0fa9c89242b Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 03:22:07 +0000 Subject: [PATCH 239/296] Jiggled some bits around to make VIF/GIF process slightly quicker (very insignificant speed change) Removed GIF Intermittent mode - While this maybe handy on the PS2, it causes us nothing but problems and extra load. Disabling this fixes NFS Most Wanted Path 3 Masking bugs (This was due to PATH2 doing a transfer half way through a PATH3 one) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1385 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 12 ++++---- pcsx2/VifDma.cpp | 74 +++++++++++++++++++++++++++--------------------- 2 files changed, 48 insertions(+), 38 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 65f27fea9d..c11762d302 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -58,9 +58,9 @@ __forceinline void gsInterrupt() { return; } - if((vif1.cmd & 0x7f) == 0x51 && Path3progress != 0) + if((vif1.cmd & 0x7f) == 0x51) { - vif1Regs->stat &= ~VIF1_STAT_VGW; + if(Path3progress != 0)vif1Regs->stat &= ~VIF1_STAT_VGW; } if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 @@ -128,7 +128,7 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) int _GIFchain() { - u32 qwc = ((psHu32(GIF_MODE) & 0x4) || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : min( gifsplit, (int)gif->qwc ); + u32 qwc = /*((psHu32(GIF_MODE) & 0x4) || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) :*/ min( gifsplit, (int)gif->qwc ); u32 *pMem; pMem = (u32*)dmaGetAddr(gif->madr); @@ -206,7 +206,7 @@ void GIFdma() if (vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1)) { if(gif->qwc == 0) { - if((gif->chcr & 0x10e) == 0x104) { + if((gif->chcr & 0x10c) == 0x104) { ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR if (ptag == NULL) { //Is ptag empty? @@ -229,10 +229,10 @@ void GIFdma() } } - if(Path3progress == 2/* && gif->qwc != 0*/) + if(Path3progress == 2 /*|| (vif1Regs->stat |= VIF1_STAT_VGW) == 0*/) { vif1Regs->stat &= ~VIF1_STAT_VGW; - dmaGIFend(); + if(gif->qwc == 0)dmaGIFend(); return; } diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index d7428c52ce..66adee139e 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -1857,6 +1857,17 @@ static int __fastcall Vif1TransDirectHL(u32 *data) { int ret = 0; + if((vif1.cmd & 0x7f) == 0x51) + { + if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 && Path3progress == 0)) //PATH3 is in image mode, so wait for end of transfer + { + vif1Regs->stat |= VIF1_STAT_VGW; + return 0; + } + } + + psHu32(GIF_STAT) |= (GIF_STAT_APATH2 | GIF_STAT_OPH); + if (splitptr > 0) //Leftover data from the last packet, filling the rest and sending to the GS { if ((splitptr < 4) && (vif1.vifpacketsize >= (4 - splitptr))) @@ -2032,12 +2043,14 @@ static void Vif1CMDSTMod() // STMOD vif1.cmd &= ~0x7f; } -static void Vif1CMDMskPath3() // MSKPATH3 +u8 schedulepath3msk = 0; + +static void Vif1MskPath3() // MSKPATH3 { - vif1Regs->mskpath3 = (vif1Regs->code >> 15) & 0x1; + vif1Regs->mskpath3 = schedulepath3msk & 0x1; //Console::WriteLn("VIF MSKPATH3 %x", params vif1Regs->mskpath3); - if ((vif1Regs->code >> 15) & 0x1) + if (vif1Regs->mskpath3) { psHu32(GIF_STAT) |= 0x2; } @@ -2045,10 +2058,27 @@ static void Vif1CMDMskPath3() // MSKPATH3 { Path3progress = 1; //Let the Gif know it can transfer again (making sure any vif stall isnt unset prematurely) psHu32(GIF_STAT) &= ~0x2; + CPU_INT(2, 4); + } + + schedulepath3msk = 0; +} +static void Vif1CMDMskPath3() // MSKPATH3 +{ + if(vif1ch->chcr & 0x100) + { + schedulepath3msk = 0x10 | ((vif1Regs->code >> 15) & 0x1); + vif1.vifstalled = true; + } + else + { + schedulepath3msk = (vif1Regs->code >> 15) & 0x1; + Vif1MskPath3(); } vif1.cmd &= ~0x7f; } + static void Vif1CMDMark() // MARK { vif1Regs->mark = (u16)vif1Regs->code; @@ -2064,7 +2094,8 @@ static void Vif1CMDFlush() // FLUSH/E/A { if((Path3progress != 2 || !vif1Regs->mskpath3) && gif->chcr & 0x100) // Gif is already transferring so wait for it. { - vif1Regs->stat |= VIF1_STAT_VGW; + vif1Regs->stat |= VIF1_STAT_VGW; + CPU_INT(2, 4); } } @@ -2116,19 +2147,6 @@ static void Vif1CMDDirectHL() // DIRECT/HL vif1.tag.size = 65536 << 2; else vif1.tag.size = vifImm << 2; - - //FIXME: This should have timing in both cases, see note below. - if((vif1.cmd & 0x7f) == 0x51) - { - if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 && Path3progress == 0)) //PATH3 is in image mode, so wait for end of transfer - { - //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); - /*if(vif1Regs->mskpath3)*/vif1Regs->stat |= VIF1_STAT_VGW; - //else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( - } - - } - psHu32(GIF_STAT) |= (GIF_STAT_APATH2 | GIF_STAT_OPH); } static void Vif1CMDNull() // invalid opcode @@ -2199,18 +2217,7 @@ int VIF1transfer(u32 *data, int size, int istag) vif1.vifpacketsize = size; while (vif1.vifpacketsize > 0) - { - if((vif1.cmd & 0x7f) == 0x51) - { - if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 || Path3progress != 2)) //PATH3 is in image mode, so wait for end of transfer - { - //DevCon::Notice("DirectHL gif chcr %x gif qwc %x mskpth3 %x", params gif->chcr, gif->qwc, vif1Regs->mskpath3); - /*if(vif1Regs->mskpath3)*/vif1Regs->stat |= VIF1_STAT_VGW; - //else while(gif->chcr & 0x100) gsInterrupt(); //Hacky as hell (no timing) but Soul Calibur 3 doesnt want timing :( - } - - } - + { if(vif1Regs->stat & VIF1_STAT_VGW) break; if (vif1.cmd) @@ -2271,9 +2278,10 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1.tag.size == 0) break; } } - if(!vif1.cmd) vif1Regs->stat &= ~VIF1_STAT_VPS_D; - if(vif1Regs->stat & VIF1_STAT_VGW) break; + if(!vif1.cmd) vif1Regs->stat &= ~VIF1_STAT_VPS_D; + + if((vif1Regs->stat & VIF1_STAT_VGW) || vif1.vifstalled == true) break; } // End of Transfer loop transferred += size - vif1.vifpacketsize; @@ -2501,11 +2509,13 @@ __forceinline void vif1Interrupt() g_vifCycles = 0; + if(schedulepath3msk) Vif1MskPath3(); + if((vif1Regs->stat & VIF1_STAT_VGW)) { if(gif->chcr & 0x100) { - CPU_INT(1, 16); + CPU_INT(1, gif->qwc * BIAS); return; } else vif1Regs->stat &= ~VIF1_STAT_VGW; From c7d2a3933c52b9cdf3d54c5d4a8c7fcb53c4624d Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 04:03:27 +0000 Subject: [PATCH 240/296] Fixed bug in IOP Counters stopping Dead to Rights booting. Ironicly we had already fixed this bug on the EE side. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1386 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/IopCounters.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pcsx2/IopCounters.cpp b/pcsx2/IopCounters.cpp index 5fa5212043..b92e55a082 100644 --- a/pcsx2/IopCounters.cpp +++ b/pcsx2/IopCounters.cpp @@ -402,11 +402,11 @@ void psxRcntUpdate() { s32 change = psxRegs.cycle - psxCounters[i].sCycleT; - // don't count disabled, gated, or hblank counters... + // don't count disabled or hblank counters... // We can't check the ALTSOURCE flag because the PSXCLOCK source *should* // be counted here. - if( psxCounters[i].mode & (IOPCNT_STOPPED | IOPCNT_ENABLE_GATE) ) continue; + if( psxCounters[i].mode & IOPCNT_STOPPED ) continue; if( psxCounters[i].rate == PSXHBLANK ) continue; if( change <= 0 ) continue; From 186403a789af4ef43b8c69f035d375a35e5aa798 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 04:51:09 +0000 Subject: [PATCH 241/296] Added some code to "skip" invalid memory addresses on VIF, the old PCSX2 use to ignore it, but it doesn't anymore. Fixes Ratatouille git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1387 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 66adee139e..0276185d61 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -1348,7 +1348,13 @@ int _VIF0chain() if ((vif0ch->qwc == 0) && !vif0.vifstalled) return 0; pMem = (u32*)dmaGetAddr(vif0ch->madr); - if (pMem == NULL) return -1; + if (pMem == NULL) + { + vif0.cmd = 0; + vif0.tag.size = 0; + vif0ch->qwc = 0; + return 0; + } if (vif0.vifstalled) ret = VIF0transfer(pMem + vif0.irqoffset, vif0ch->qwc * 4 - vif0.irqoffset, 0); @@ -2406,7 +2412,13 @@ int _VIF1chain() } pMem = (u32*)dmaGetAddr(vif1ch->madr); - if (pMem == NULL) return -1; + if (pMem == NULL) + { + vif1.cmd = 0; + vif1.tag.size = 0; + vif1ch->qwc = 0; + return 0; + } VIF_LOG("VIF1chain size=%d, madr=%lx, tadr=%lx", vif1ch->qwc, vif1ch->madr, vif1ch->tadr); From b974b815ff0fd9e4eff6970b2e24fbf00262849e Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 18 Jun 2009 08:20:19 +0000 Subject: [PATCH 242/296] Major Build System changes: * Changed w32pthreads library into a DLL so that it can be used from plugins correctly. (NOTE: you will need to make sure to build and copy w32pthreads.dll into your pcsx2 folder). * Switched pcsx2 from static CRT to shared CRT linking (needed to ensure correct exception handling behavior in multithreaded DLL environments). * Switched all standard plugins in the Suite to the shared CRT, to match pcsx2's new style. :) * Renamed _DEBUG (depreciated) to PCSX2_DEBUG (excluding Gabest projects since the ATL still uses it). * Added intrin.h to Pcsx2Defs.h (so that it is included universally), and added intrin_x86.h for GCC compatibility. Notes: * Current plugin version compatibility status should be unaffected. The new shared-CRT plugins work fine with older versions of Pcsx2, and the older plugins should work fine with the new shared-CRT version of pcsx2; so long as the necessary CRT DLLs are available on the user's system. * All future packagings of pcsx2 will include w32pthreads.dll and the Common Runtimes (CRTs). * Existing users who do not have MSVC installed can obtain the CRTs by downloading the Microsoft Visual C++ 2008 Redistributable Package (anyone with msvc installed should already have all they need). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1388 96395faa-99c1-11dd-bbfe-3dabce05a288 --- 3rdparty/SoundTouch/SoundTouch.vcproj | 19 +- 3rdparty/bzip2/bzip2.vcproj | 33 +- 3rdparty/w32pthreads/PCSX2.txt | 13 +- 3rdparty/w32pthreads/include/pthread.h | 2 + 3rdparty/w32pthreads/pthreads_2008.vcproj | 87 +- 3rdparty/zlib/zlib.vcproj | 28 +- common/include/PS2Edefs.h | 8 + common/include/Pcsx2Defs.h | 8 +- common/include/Pcsx2Types.h | 1 + common/include/intrin_x86.h | 948 +++++ .../vsprops/CodeGen_Debug.vsprops | 10 +- common/vsprops/CodeGen_Devel.vsprops | 18 + ...inking.vsprops => CodeGen_Release.vsprops} | 4 +- common/vsprops/pthreads.vsprops | 8 +- pcsx2/IPU/IPU.cpp | 9 +- pcsx2/MTGS.cpp | 35 +- pcsx2/PrecompiledHeader.h | 2 +- pcsx2/R3000AInterpreter.cpp | 5 - pcsx2/StringUtils.h | 2 +- pcsx2/System.h | 2 +- pcsx2/VU0microInterp.cpp | 7 - pcsx2/VU1micro.cpp | 2 +- pcsx2/VU1microInterp.cpp | 6 - pcsx2/VifDma.cpp | 12 +- pcsx2/tinyxml/tinyxml.h | 2 +- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 19 +- pcsx2/windows/WinMain.cpp | 4 +- pcsx2/windows/WinThreads.cpp | 11 +- pcsx2/x86/aVUzerorec.S | 4 +- pcsx2/x86/fast_routines.cpp | 2 +- pcsx2/x86/iCore.cpp | 4 - pcsx2/x86/iCore.h | 4 - pcsx2/x86/iR3000A.cpp | 90 +- pcsx2/x86/iVU1micro.cpp | 2 +- pcsx2/x86/ix86-32/iR5900-32.cpp | 18 +- pcsx2/x86/ix86-32/iR5900LoadStore.cpp | 6 +- pcsx2/x86/sVU_zerorec.cpp | 42 +- pcsx2_suite_2008.sln | 1 + plugins/CDVDiso/src/CDVDisop.cpp | 6 +- .../CDVDiso/src/Windows/CDVDiso_vs2008.vcproj | 59 +- plugins/CDVDnull/CDVD.h | 3 - .../CDVDnull/Windows/CDVDnull_vs2008.vcproj | 6 +- plugins/CDVDolio/cdvd_vs2008.vcproj | 16 +- plugins/CDVDolio/vsprops/debug.vsprops | 2 +- plugins/CDVDolio/vsprops/release.vsprops | 1 + plugins/FWnull/FW.h | 7 - plugins/FWnull/Windows/FWnull_vc2008.vcproj | 9 +- plugins/GSdx/GSdx_vs2008.vcproj | 3304 ++++++++--------- plugins/GSdx/vsprops/debug.vsprops | 2 +- plugins/GSdx/vsprops/release.vsprops | 1 + plugins/GSnull/Windows/GSnull_vc2008.vcproj | 2 +- plugins/LilyPad/LilyPad.cpp | 8 +- plugins/LilyPad/LilyPad_VC2008.vcproj | 25 +- plugins/PadNull/Windows/PadNull_vc2008.vcproj | 2 +- plugins/USBnull/USB.h | 3 - plugins/USBnull/Windows/USBnull_vc2008.vcproj | 9 +- plugins/dev9null/DEV9.h | 3 - .../dev9null/Windows/DEV9null_vc2008.vcproj | 6 +- plugins/spu2-x/src/BaseTypes.h | 2 +- plugins/spu2-x/src/DllInterface.cpp | 8 +- plugins/spu2-x/src/Mixer.cpp | 2 +- plugins/spu2-x/src/SaveStateSPU.cpp | 3 + plugins/spu2-x/src/Spu2.cpp | 2 +- plugins/spu2-x/src/Windows/Spu2-X.rc | 8 +- .../spu2-x/src/Windows/Spu2-X_vs2008.vcproj | 44 +- plugins/xpad/vsprops/debug.vsprops | 2 +- plugins/xpad/vsprops/release.vsprops | 1 + plugins/zerogs/dx/GS.h | 3 - plugins/zerogs/dx/GSmain.cpp | 10 +- plugins/zerogs/dx/Regs.cpp | 4 +- plugins/zerogs/dx/Windows/zerogs_2008.vcproj | 32 +- plugins/zerogs/dx/memcpy_amd.cpp | 4 +- plugins/zerogs/dx/targets.cpp | 32 +- plugins/zerogs/dx/zerogs.cpp | 14 +- plugins/zeropad/Windows/ZeroPAD_2008.vcproj | 68 +- plugins/zeropad/zeropad.cpp | 6 +- plugins/zeropad/zeropad.h | 4 +- plugins/zerospu2/Windows/ZeroSPU2.rc | 5 - plugins/zerospu2/Windows/ZeroSPU2_2008.vcproj | 26 +- plugins/zerospu2/misc.h | 64 +- plugins/zerospu2/zerospu2.cpp | 6 +- plugins/zerospu2/zerospu2.h | 2 - 82 files changed, 2946 insertions(+), 2358 deletions(-) create mode 100644 common/include/intrin_x86.h rename pcsx2/windows/VCprojects/vsprops/debug.vsprops => common/vsprops/CodeGen_Debug.vsprops (57%) create mode 100644 common/vsprops/CodeGen_Devel.vsprops rename common/vsprops/{GlobalLinking.vsprops => CodeGen_Release.vsprops} (83%) diff --git a/3rdparty/SoundTouch/SoundTouch.vcproj b/3rdparty/SoundTouch/SoundTouch.vcproj index b1cd6fa470..b7682bb958 100644 --- a/3rdparty/SoundTouch/SoundTouch.vcproj +++ b/3rdparty/SoundTouch/SoundTouch.vcproj @@ -18,7 +18,7 @@ @@ -95,12 +92,7 @@ /> @@ -154,12 +146,7 @@ /> @@ -96,17 +91,6 @@ /> @@ -160,17 +144,6 @@ /> _< Important: pthread_cond is an inherently flawed concept, and is quite incredibly slow on top of it. It's primarily meant for inter-process thread regulation, of which we do none. Therefore it has been excluded from the library builds. Don't use it! +I've also disabled rwlocks, since they use pthread_cond internally. Again, there should +be little or no reason to have to use those within the context of a high performance app +like Pcsx2. + diff --git a/3rdparty/w32pthreads/include/pthread.h b/3rdparty/w32pthreads/include/pthread.h index d33a9b14a8..12b4a169b5 100644 --- a/3rdparty/w32pthreads/include/pthread.h +++ b/3rdparty/w32pthreads/include/pthread.h @@ -1077,6 +1077,7 @@ PTW32_DLLPORT int PTW32_CDECL pthread_condattr_setpshared (pthread_condattr_t * /* * Condition Variable Functions */ +#ifdef PCSX2_ALLOW_COND PTW32_DLLPORT int PTW32_CDECL pthread_cond_init (pthread_cond_t * cond, const pthread_condattr_t * attr); @@ -1092,6 +1093,7 @@ PTW32_DLLPORT int PTW32_CDECL pthread_cond_timedwait (pthread_cond_t * cond, PTW32_DLLPORT int PTW32_CDECL pthread_cond_signal (pthread_cond_t * cond); PTW32_DLLPORT int PTW32_CDECL pthread_cond_broadcast (pthread_cond_t * cond); +#endif /* * Scheduling diff --git a/3rdparty/w32pthreads/pthreads_2008.vcproj b/3rdparty/w32pthreads/pthreads_2008.vcproj index 5459698111..26f91c289e 100644 --- a/3rdparty/w32pthreads/pthreads_2008.vcproj +++ b/3rdparty/w32pthreads/pthreads_2008.vcproj @@ -2,7 +2,7 @@ + @@ -74,14 +76,17 @@ + @@ -102,23 +107,11 @@ /> + @@ -145,14 +144,17 @@ + @@ -173,22 +175,11 @@ /> + @@ -215,6 +212,9 @@ + @@ -439,6 +439,10 @@ /> + + @@ -754,6 +758,7 @@ > @@ -96,15 +92,7 @@ /> @@ -158,17 +146,7 @@ /> +#else +# include +#endif + #include "Pcsx2Types.h" // Renamed ARRAYSIZE to ArraySize -- looks nice and gets rid of Windows.h conflicts (air) @@ -51,7 +57,7 @@ # endif # else # ifdef _MSC_VER -# define jBREAKPOINT() do { __asm int 3 } while(0); +# define jBREAKPOINT() __debugbreak(); # else # define jBREAKPOINT() ((void) *(volatile char *) 0) # endif diff --git a/common/include/Pcsx2Types.h b/common/include/Pcsx2Types.h index 41e2d893a1..236a5eb339 100644 --- a/common/include/Pcsx2Types.h +++ b/common/include/Pcsx2Types.h @@ -101,6 +101,7 @@ typedef s32 sptr; ////////////////////////////////////////////////////////////////////////////////////////// // A rough-and-ready cross platform 128-bit datatype, Non-SSE style. +// #ifdef __cplusplus struct u128 { diff --git a/common/include/intrin_x86.h b/common/include/intrin_x86.h new file mode 100644 index 0000000000..a47b68ae35 --- /dev/null +++ b/common/include/intrin_x86.h @@ -0,0 +1,948 @@ +/* + Compatibility header for GCC -- GCC equivalents of intrinsic + Microsoft Visual C++ functions. Originally developed for the ReactOS + () and TinyKrnl () + projects. + + Copyright (c) 2006 KJK::Hyperion + + Permission is hereby granted, free of charge, to any person obtaining a + copy of this software and associated documentation files (the "Software"), + to deal in the Software without restriction, including without limitation + the rights to use, copy, modify, merge, publish, distribute, sublicense, + and/or sell copies of the Software, and to permit persons to whom the + Software is furnished to do so, subject to the following conditions: + + The above copyright notice and this permission notice shall be included in + all copies or substantial portions of the Software. + + THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + DEALINGS IN THE SOFTWARE. +*/ + +#ifndef KJK_INTRIN_X86_H_ +#define KJK_INTRIN_X86_H_ + +/* + FIXME: review all "memory" clobbers, add/remove to match Visual C++ + behavior: some "obvious" memory barriers are not present in the Visual C++ + implementation - e.g. __stosX; on the other hand, some memory barriers that + *are* present could have been missed +*/ + +/* + NOTE: this is a *compatibility* header. Some functions may look wrong at + first, but they're only "as wrong" as they would be on Visual C++. Our + priority is compatibility + + NOTE: unlike most people who write inline asm for GCC, I didn't pull the + constraints and the uses of __volatile__ out of my... hat. Do not touch + them. I hate cargo cult programming + + NOTE: be very careful with declaring "memory" clobbers. Some "obvious" + barriers aren't there in Visual C++ (e.g. __stosX) + + NOTE: review all intrinsics with a return value, add/remove __volatile__ + where necessary. If an intrinsic whose value is ignored generates a no-op + under Visual C++, __volatile__ must be omitted; if it always generates code + (for example, if it has side effects), __volatile__ must be specified. GCC + will only optimize out non-volatile asm blocks with outputs, so input-only + blocks are safe. Oddities such as the non-volatile 'rdmsr' are intentional + and follow Visual C++ behavior + + NOTE: on GCC 4.1.0, please use the __sync_* built-ins for barriers and + atomic operations. Test the version like this: + + #if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100 + ... + + Pay attention to the type of barrier. Make it match with what Visual C++ + would use in the same case +*/ + +/*** Stack frame juggling ***/ +#define _ReturnAddress() (__builtin_return_address(0)) +#define _AddressOfReturnAddress() (&(((void **)(__builtin_frame_address(0)))[1])) +/* TODO: __getcallerseflags but how??? */ + + +/*** Atomic operations ***/ + +#if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100 +#define _ReadWriteBarrier() __sync_synchronize() +#else +static void __inline__ __attribute__((always_inline)) _MemoryBarrier(void) +{ + __asm__ __volatile__("" : : : "memory"); +} +#define _ReadWriteBarrier() _MemoryBarrier() +#endif + +/* BUGBUG: GCC only supports full barriers */ +#define _ReadBarrier _ReadWriteBarrier +#define _WriteBarrier _ReadWriteBarrier + +#if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100 + +static __inline__ __attribute__((always_inline)) char _InterlockedCompareExchange8(volatile char * const Destination, const char Exchange, const char Comperand) +{ + return __sync_val_compare_and_swap(Destination, Comperand, Exchange); +} + +static __inline__ __attribute__((always_inline)) short _InterlockedCompareExchange16(volatile short * const Destination, const short Exchange, const short Comperand) +{ + return __sync_val_compare_and_swap(Destination, Comperand, Exchange); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedCompareExchange(volatile long * const Destination, const long Exchange, const long Comperand) +{ + return __sync_val_compare_and_swap(Destination, Comperand, Exchange); +} + +static __inline__ __attribute__((always_inline)) long long _InterlockedCompareExchange64(volatile long long * const Destination, const long long Exchange, const long long Comperand) +{ + return __sync_val_compare_and_swap(Destination, Comperand, Exchange); +} + +static __inline__ __attribute__((always_inline)) void * _InterlockedCompareExchangePointer(void * volatile * const Destination, void * const Exchange, void * const Comperand) +{ + return __sync_val_compare_and_swap(Destination, Comperand, Exchange); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedExchange(volatile long * const Target, const long Value) +{ + /* NOTE: __sync_lock_test_and_set would be an acquire barrier, so we force a full barrier */ + __sync_synchronize(); + return __sync_lock_test_and_set(Target, Value); +} + +static __inline__ __attribute__((always_inline)) void * _InterlockedExchangePointer(void * volatile * const Target, void * const Value) +{ + /* NOTE: ditto */ + __sync_synchronize(); + return __sync_lock_test_and_set(Target, Value); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedExchangeAdd(volatile long * const Addend, const long Value) +{ + return __sync_fetch_and_add(Addend, Value); +} + +static __inline__ __attribute__((always_inline)) char _InterlockedAnd8(volatile char * const value, const char mask) +{ + return __sync_fetch_and_and(value, mask); +} + +static __inline__ __attribute__((always_inline)) short _InterlockedAnd16(volatile short * const value, const short mask) +{ + return __sync_fetch_and_and(value, mask); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedAnd(volatile long * const value, const long mask) +{ + return __sync_fetch_and_and(value, mask); +} + +static __inline__ __attribute__((always_inline)) char _InterlockedOr8(volatile char * const value, const char mask) +{ + return __sync_fetch_and_or(value, mask); +} + +static __inline__ __attribute__((always_inline)) short _InterlockedOr16(volatile short * const value, const short mask) +{ + return __sync_fetch_and_or(value, mask); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedOr(volatile long * const value, const long mask) +{ + return __sync_fetch_and_or(value, mask); +} + +static __inline__ __attribute__((always_inline)) char _InterlockedXor8(volatile char * const value, const char mask) +{ + return __sync_fetch_and_xor(value, mask); +} + +static __inline__ __attribute__((always_inline)) short _InterlockedXor16(volatile short * const value, const short mask) +{ + return __sync_fetch_and_xor(value, mask); +} + +static __inline__ __attribute__((always_inline)) long _InterlockedXor(volatile long * const value, const long mask) +{ + return __sync_fetch_and_xor(value, mask); +} + +#else + +static __inline__ __attribute__((always_inline)) char _InterlockedCompareExchange8(volatile char * const Destination, const char Exchange, const char Comperand) +{ + char retval = Comperand; + __asm__("lock; cmpxchgb %b[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) short _InterlockedCompareExchange16(volatile short * const Destination, const short Exchange, const short Comperand) +{ + short retval = Comperand; + __asm__("lock; cmpxchgw %w[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange): "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedCompareExchange(volatile long * const Destination, const long Exchange, const long Comperand) +{ + long retval = Comperand; + __asm__("lock; cmpxchgl %k[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange): "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) long long _InterlockedCompareExchange64(volatile long long * const Destination, const long long Exchange, const long long Comperand) +{ + long long retval = Comperand; + + __asm__ + ( + "cmpxchg8b %[Destination]" : + [retval] "+A" (retval) : + [Destination] "m" (*Destination), + "b" ((unsigned long)((Exchange >> 0) & 0xFFFFFFFF)), + "c" ((unsigned long)((Exchange >> 32) & 0xFFFFFFFF)) : + "memory" + ); + + return retval; +} + +static __inline__ __attribute__((always_inline)) void * _InterlockedCompareExchangePointer(void * volatile * const Destination, void * const Exchange, void * const Comperand) +{ + void * retval = (void *)Comperand; + __asm__("lock; cmpxchgl %k[Exchange], %[Destination]" : [retval] "=a" (retval) : "[retval]" (retval), [Destination] "m" (*Destination), [Exchange] "q" (Exchange) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedExchange(volatile long * const Target, const long Value) +{ + long retval = Value; + __asm__("xchgl %[retval], %[Target]" : [retval] "+r" (retval) : [Target] "m" (*Target) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) void * _InterlockedExchangePointer(void * volatile * const Target, void * const Value) +{ + void * retval = Value; + __asm__("xchgl %[retval], %[Target]" : [retval] "+r" (retval) : [Target] "m" (*Target) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedExchangeAdd(volatile long * const Addend, const long Value) +{ + long retval = Value; + __asm__("lock; xaddl %[retval], %[Addend]" : [retval] "+r" (retval) : [Addend] "m" (*Addend) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) char _InterlockedAnd8(volatile char * const value, const char mask) +{ + char x; + char y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange8(value, x & mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) short _InterlockedAnd16(volatile short * const value, const short mask) +{ + short x; + short y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange16(value, x & mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedAnd(volatile long * const value, const long mask) +{ + long x; + long y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange(value, x & mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) char _InterlockedOr8(volatile char * const value, const char mask) +{ + char x; + char y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange8(value, x | mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) short _InterlockedOr16(volatile short * const value, const short mask) +{ + short x; + short y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange16(value, x | mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedOr(volatile long * const value, const long mask) +{ + long x; + long y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange(value, x | mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) char _InterlockedXor8(volatile char * const value, const char mask) +{ + char x; + char y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange8(value, x ^ mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) short _InterlockedXor16(volatile short * const value, const short mask) +{ + short x; + short y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange16(value, x ^ mask, x); + } + while(y != x); + + return y; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedXor(volatile long * const value, const long mask) +{ + long x; + long y; + + y = *value; + + do + { + x = y; + y = _InterlockedCompareExchange(value, x ^ mask, x); + } + while(y != x); + + return y; +} + +#endif + +static __inline__ __attribute__((always_inline)) long _InterlockedAddLargeStatistic(volatile long long * const Addend, const long Value) +{ + __asm__ + ( + "lock; add %[Value], %[Lo32];" + "jae LABEL%=;" + "lock; adc $0, %[Hi32];" + "LABEL%=:;" : + [Lo32] "=m" (*((volatile long *)(Addend) + 0)), [Hi32] "=m" (*((volatile long *)(Addend) + 1)) : + [Value] "ir" (Value) + ); + + return Value; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedDecrement(volatile long * const lpAddend) +{ + return _InterlockedExchangeAdd(lpAddend, -1) - 1; +} + +static __inline__ __attribute__((always_inline)) long _InterlockedIncrement(volatile long * const lpAddend) +{ + return _InterlockedExchangeAdd(lpAddend, 1) + 1; +} + +static __inline__ __attribute__((always_inline)) unsigned char _interlockedbittestandreset(volatile long * a, const long b) +{ + unsigned char retval; + __asm__("lock; btrl %[b], %[a]; setb %b[retval]" : [retval] "=r" (retval), [a] "=m" (a) : [b] "Ir" (b) : "memory"); + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _interlockedbittestandset(volatile long * a, const long b) +{ + unsigned char retval; + __asm__("lock; btsl %[b], %[a]; setc %b[retval]" : [retval] "=r" (retval), [a] "=m" (a) : [b] "Ir" (b) : "memory"); + return retval; +} + + +/*** String operations ***/ +/* NOTE: we don't set a memory clobber in the __stosX functions because Visual C++ doesn't */ +static __inline__ __attribute__((always_inline)) void __stosb(unsigned char * Dest, const unsigned char Data, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; stosb" : + [Dest] "=D" (Dest), [Count] "=c" (Count) : + "[Dest]" (Dest), "a" (Data), "[Count]" (Count) + ); +} + +static __inline__ __attribute__((always_inline)) void __stosw(unsigned short * Dest, const unsigned short Data, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; stosw" : + [Dest] "=D" (Dest), [Count] "=c" (Count) : + "[Dest]" (Dest), "a" (Data), "[Count]" (Count) + ); +} + +static __inline__ __attribute__((always_inline)) void __stosd(unsigned long * Dest, const unsigned long Data, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; stosl" : + [Dest] "=D" (Dest), [Count] "=c" (Count) : + "[Dest]" (Dest), "a" (Data), "[Count]" (Count) + ); +} + +static __inline__ __attribute__((always_inline)) void __movsb(unsigned char * Destination, const unsigned char * Source, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; movsb" : + [Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) : + "[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count) + ); +} + +static __inline__ __attribute__((always_inline)) void __movsw(unsigned short * Destination, const unsigned short * Source, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; movsw" : + [Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) : + "[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count) + ); +} + +static __inline__ __attribute__((always_inline)) void __movsd(unsigned long * Destination, const unsigned long * Source, size_t Count) +{ + __asm__ __volatile__ + ( + "rep; movsd" : + [Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) : + "[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count) + ); +} + + +/*** FS segment addressing ***/ +static __inline__ __attribute__((always_inline)) void __writefsbyte(const unsigned long Offset, const unsigned char Data) +{ + __asm__("movb %b[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __writefsword(const unsigned long Offset, const unsigned short Data) +{ + __asm__("movw %w[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __writefsdword(const unsigned long Offset, const unsigned long Data) +{ + __asm__("movl %k[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + +static __inline__ __attribute__((always_inline)) unsigned char __readfsbyte(const unsigned long Offset) +{ + unsigned char value; + __asm__("movb %%fs:%a[Offset], %b[value]" : [value] "=q" (value) : [Offset] "irm" (Offset)); + return value; +} + +static __inline__ __attribute__((always_inline)) unsigned short __readfsword(const unsigned long Offset) +{ + unsigned short value; + __asm__("movw %%fs:%a[Offset], %w[value]" : [value] "=q" (value) : [Offset] "irm" (Offset)); + return value; +} + +static __inline__ __attribute__((always_inline)) unsigned long __readfsdword(const unsigned long Offset) +{ + unsigned long value; + __asm__("movl %%fs:%a[Offset], %k[value]" : [value] "=q" (value) : [Offset] "irm" (Offset)); + return value; +} + +static __inline__ __attribute__((always_inline)) void __incfsbyte(const unsigned long Offset) +{ + __asm__("incb %%fs:%a[Offset]" : : [Offset] "ir" (Offset)); +} + +static __inline__ __attribute__((always_inline)) void __incfsword(const unsigned long Offset) +{ + __asm__("incw %%fs:%a[Offset]" : : [Offset] "ir" (Offset)); +} + +static __inline__ __attribute__((always_inline)) void __incfsdword(const unsigned long Offset) +{ + __asm__("incl %%fs:%a[Offset]" : : [Offset] "ir" (Offset)); +} + +/* NOTE: the bizarre implementation of __addfsxxx mimics the broken Visual C++ behavior */ +static __inline__ __attribute__((always_inline)) void __addfsbyte(const unsigned long Offset, const unsigned char Data) +{ + if(!__builtin_constant_p(Offset)) + __asm__("addb %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset)); + else + __asm__("addb %b[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __addfsword(const unsigned long Offset, const unsigned short Data) +{ + if(!__builtin_constant_p(Offset)) + __asm__("addw %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset)); + else + __asm__("addw %w[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __addfsdword(const unsigned long Offset, const unsigned int Data) +{ + if(!__builtin_constant_p(Offset)) + __asm__("addl %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset)); + else + __asm__("addl %k[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data)); +} + + +/*** Bit manipulation ***/ +static __inline__ __attribute__((always_inline)) unsigned char _BitScanForward(unsigned long * const Index, const unsigned long Mask) +{ + __asm__("bsfl %[Mask], %[Index]" : [Index] "=r" (*Index) : [Mask] "mr" (Mask)); + return Mask ? 1 : 0; +} + +static __inline__ __attribute__((always_inline)) unsigned char _BitScanReverse(unsigned long * const Index, const unsigned long Mask) +{ + __asm__("bsrl %[Mask], %[Index]" : [Index] "=r" (*Index) : [Mask] "mr" (Mask)); + return Mask ? 1 : 0; +} + +/* NOTE: again, the bizarre implementation follows Visual C++ */ +static __inline__ __attribute__((always_inline)) unsigned char _bittest(const long * const a, const long b) +{ + unsigned char retval; + + if(__builtin_constant_p(b)) + __asm__("bt %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32)); + else + __asm__("bt %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b)); + + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _bittestandcomplement(long * const a, const long b) +{ + unsigned char retval; + + if(__builtin_constant_p(b)) + __asm__("btc %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32)); + else + __asm__("btc %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b)); + + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _bittestandreset(long * const a, const long b) +{ + unsigned char retval; + + if(__builtin_constant_p(b)) + __asm__("btr %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32)); + else + __asm__("btr %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b)); + + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _bittestandset(long * const a, const long b) +{ + unsigned char retval; + + if(__builtin_constant_p(b)) + __asm__("bts %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32)); + else + __asm__("bts %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b)); + + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _rotl8(const unsigned char value, const unsigned char shift) +{ + unsigned char retval; + __asm__("rolb %b[shift], %b[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift)); + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned short _rotl16(const unsigned short value, const unsigned char shift) +{ + unsigned short retval; + __asm__("rolw %b[shift], %w[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift)); + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned char _rotr8(const unsigned char value, const unsigned char shift) +{ + unsigned char retval; + __asm__("rorb %b[shift], %b[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift)); + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned short _rotr16(const unsigned short value, const unsigned char shift) +{ + unsigned short retval; + __asm__("rorw %b[shift], %w[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift)); + return retval; +} + +/* + NOTE: in __ll_lshift, __ll_rshift and __ull_rshift we use the "A" + constraint (edx:eax) for the Mask argument, because it's the only way GCC + can pass 64-bit operands around - passing the two 32 bit parts separately + just confuses it. Also we declare Bit as an int and then truncate it to + match Visual C++ behavior +*/ +static __inline__ __attribute__((always_inline)) unsigned long long __ll_lshift(const unsigned long long Mask, const int Bit) +{ + unsigned long long retval = Mask; + + __asm__ + ( + "shldl %b[Bit], %%eax, %%edx; sall %b[Bit], %%eax" : + "+A" (retval) : + [Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF) + ); + + return retval; +} + +static __inline__ __attribute__((always_inline)) long long __ll_rshift(const long long Mask, const int Bit) +{ + unsigned long long retval = Mask; + + __asm__ + ( + "shldl %b[Bit], %%eax, %%edx; sarl %b[Bit], %%eax" : + "+A" (retval) : + [Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF) + ); + + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned long long __ull_rshift(const unsigned long long Mask, int Bit) +{ + unsigned long long retval = Mask; + + __asm__ + ( + "shrdl %b[Bit], %%eax, %%edx; shrl %b[Bit], %%eax" : + "+A" (retval) : + [Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF) + ); + + return retval; +} + + +/*** 64-bit math ***/ +static __inline__ __attribute__((always_inline)) long long __emul(const int a, const int b) +{ + long long retval; + __asm__("imull %[b]" : "=A" (retval) : [a] "a" (a), [b] "rm" (b)); + return retval; +} + +static __inline__ __attribute__((always_inline)) unsigned long long __emulu(const unsigned int a, const unsigned int b) +{ + unsigned long long retval; + __asm__("mull %[b]" : "=A" (retval) : [a] "a" (a), [b] "rm" (b)); + return retval; +} + + +/*** Port I/O ***/ +static __inline__ __attribute__((always_inline)) unsigned char __inbyte(const unsigned short Port) +{ + unsigned char byte; + __asm__ __volatile__("inb %w[Port], %b[byte]" : [byte] "=a" (byte) : [Port] "Nd" (Port)); + return byte; +} + +static __inline__ __attribute__((always_inline)) unsigned short __inword(const unsigned short Port) +{ + unsigned short word; + __asm__ __volatile__("inw %w[Port], %w[word]" : [word] "=a" (word) : [Port] "Nd" (Port)); + return word; +} + +static __inline__ __attribute__((always_inline)) unsigned long __indword(const unsigned short Port) +{ + unsigned long dword; + __asm__ __volatile__("inl %w[Port], %k[dword]" : [dword] "=a" (dword) : [Port] "Nd" (Port)); + return dword; +} + +static __inline__ __attribute__((always_inline)) void __inbytestring(unsigned short Port, unsigned char * Buffer, unsigned long Count) +{ + __asm__ __volatile__ + ( + "rep; insb" : + [Buffer] "=D" (Buffer), [Count] "=c" (Count) : + "d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) : + "memory" + ); +} + +static __inline__ __attribute__((always_inline)) void __inwordstring(unsigned short Port, unsigned short * Buffer, unsigned long Count) +{ + __asm__ __volatile__ + ( + "rep; insw" : + [Buffer] "=D" (Buffer), [Count] "=c" (Count) : + "d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) : + "memory" + ); +} + +static __inline__ __attribute__((always_inline)) void __indwordstring(unsigned short Port, unsigned long * Buffer, unsigned long Count) +{ + __asm__ __volatile__ + ( + "rep; insl" : + [Buffer] "=D" (Buffer), [Count] "=c" (Count) : + "d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) : + "memory" + ); +} + +static __inline__ __attribute__((always_inline)) void __outbyte(unsigned short const Port, const unsigned char Data) +{ + __asm__ __volatile__("outb %b[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __outword(unsigned short const Port, const unsigned short Data) +{ + __asm__ __volatile__("outw %w[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __outdword(unsigned short const Port, const unsigned long Data) +{ + __asm__ __volatile__("outl %k[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data)); +} + +static __inline__ __attribute__((always_inline)) void __outbytestring(unsigned short const Port, const unsigned char * const Buffer, const unsigned long Count) +{ + __asm__ __volatile__("rep; outsb" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count)); +} + +static __inline__ __attribute__((always_inline)) void __outwordstring(unsigned short const Port, const unsigned short * const Buffer, const unsigned long Count) +{ + __asm__ __volatile__("rep; outsw" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count)); +} + +static __inline__ __attribute__((always_inline)) void __outdwordstring(unsigned short const Port, const unsigned long * const Buffer, const unsigned long Count) +{ + __asm__ __volatile__("rep; outsl" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count)); +} + + +/*** System information ***/ +static __inline__ __attribute__((always_inline)) void __cpuid(int CPUInfo[], const int InfoType) +{ + __asm__ __volatile__("cpuid" : "=a" (CPUInfo[0]), "=b" (CPUInfo[1]), "=c" (CPUInfo[2]), "=d" (CPUInfo[3]) : "a" (InfoType)); +} + +static __inline__ __attribute__((always_inline)) unsigned long long __rdtsc(void) +{ + unsigned long long retval; + __asm__ __volatile__("rdtsc" : "=A"(retval)); + return retval; +} + + +/*** Interrupts ***/ +static __inline__ __attribute__((always_inline)) void __debugbreak(void) +{ + __asm__("int $3"); +} + +static __inline__ __attribute__((always_inline)) void __int2c(void) +{ + __asm__("int $0x2c"); +} + +static __inline__ __attribute__((always_inline)) void _disable(void) +{ + __asm__("cli"); +} + +static __inline__ __attribute__((always_inline)) void _enable(void) +{ + __asm__("sti"); +} + + +/*** Protected memory management ***/ +static __inline__ __attribute__((always_inline)) unsigned long __readcr0(void) +{ + unsigned long value; + __asm__ __volatile__("mov %%cr0, %[value]" : [value] "=q" (value)); + return value; +} + +static __inline__ __attribute__((always_inline)) unsigned long __readcr2(void) +{ + unsigned long value; + __asm__ __volatile__("mov %%cr2, %[value]" : [value] "=q" (value)); + return value; +} + +static __inline__ __attribute__((always_inline)) unsigned long __readcr3(void) +{ + unsigned long value; + __asm__ __volatile__("mov %%cr3, %[value]" : [value] "=q" (value)); + return value; +} + +static __inline__ __attribute__((always_inline)) unsigned long __readcr4(void) +{ + unsigned long value; + __asm__ __volatile__("mov %%cr4, %[value]" : [value] "=q" (value)); + return value; +} + +static __inline__ __attribute__((always_inline)) void __writecr0(const unsigned long long Data) +{ + __asm__("mov %[Data], %%cr0" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory"); +} + +static __inline__ __attribute__((always_inline)) void __writecr3(const unsigned long long Data) +{ + __asm__("mov %[Data], %%cr3" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory"); +} + +static __inline__ __attribute__((always_inline)) void __writecr4(const unsigned long long Data) +{ + __asm__("mov %[Data], %%cr4" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory"); +} + +static __inline__ __attribute__((always_inline)) void __invlpg(void * const Address) +{ + __asm__("invlpg %[Address]" : : [Address] "m" (*((unsigned char *)(Address)))); +} + + +/*** System operations ***/ +static __inline__ __attribute__((always_inline)) unsigned long long __readmsr(const int reg) +{ + unsigned long long retval; + __asm__ __volatile__("rdmsr" : "=A" (retval) : "c" (reg)); + return retval; +} + +static __inline__ __attribute__((always_inline)) void __writemsr(const unsigned long Register, const unsigned long long Value) +{ + __asm__ __volatile__("wrmsr" : : "A" (Value), "c" (Register)); +} + +static __inline__ __attribute__((always_inline)) unsigned long long __readpmc(const int counter) +{ + unsigned long long retval; + __asm__ __volatile__("rdpmc" : "=A" (retval) : "c" (counter)); + return retval; +} + +/* NOTE: an immediate value for 'a' will raise an ICE in Visual C++ */ +static __inline__ __attribute__((always_inline)) unsigned long __segmentlimit(const unsigned long a) +{ + unsigned long retval; + __asm__ __volatile__("lsl %[a], %[retval]" : [retval] "=r" (retval) : [a] "rm" (a)); + return retval; +} + +static __inline__ __attribute__((always_inline)) void __wbinvd(void) +{ + __asm__ __volatile__("wbinvd"); +} + +#endif /* KJK_INTRIN_X86_H_ */ + +/* EOF */ diff --git a/pcsx2/windows/VCprojects/vsprops/debug.vsprops b/common/vsprops/CodeGen_Debug.vsprops similarity index 57% rename from pcsx2/windows/VCprojects/vsprops/debug.vsprops rename to common/vsprops/CodeGen_Debug.vsprops index af6f81ca44..b4fe7b5204 100644 --- a/pcsx2/windows/VCprojects/vsprops/debug.vsprops +++ b/common/vsprops/CodeGen_Debug.vsprops @@ -2,17 +2,13 @@ - diff --git a/common/vsprops/CodeGen_Devel.vsprops b/common/vsprops/CodeGen_Devel.vsprops new file mode 100644 index 0000000000..13aef9f8d3 --- /dev/null +++ b/common/vsprops/CodeGen_Devel.vsprops @@ -0,0 +1,18 @@ + + + + diff --git a/common/vsprops/GlobalLinking.vsprops b/common/vsprops/CodeGen_Release.vsprops similarity index 83% rename from common/vsprops/GlobalLinking.vsprops rename to common/vsprops/CodeGen_Release.vsprops index 5bf8cd3a63..7cfa66fc48 100644 --- a/common/vsprops/GlobalLinking.vsprops +++ b/common/vsprops/CodeGen_Release.vsprops @@ -2,7 +2,7 @@ - diff --git a/pcsx2/IPU/IPU.cpp b/pcsx2/IPU/IPU.cpp index a47e1ad681..4b4cd3845c 100644 --- a/pcsx2/IPU/IPU.cpp +++ b/pcsx2/IPU/IPU.cpp @@ -467,11 +467,7 @@ static __forceinline BOOL ipuIDEC(u32 val) return s_RoutineDone; } -#ifdef _DEBUG static int s_bdec = 0; -#else -#define s_bdec 0 -#endif static __forceinline BOOL ipuBDEC(u32 val) { @@ -497,9 +493,8 @@ static __forceinline BOOL ipuBDEC(u32 val) IPU_LOG(" Quantizer step=0x%X", bdec.QSC); -#ifdef _DEBUG - s_bdec++; -#endif + if( IsDebugBuild ) + s_bdec++; g_BP.BP += bdec.FB;//skip FB bits g_decoder.coding_type = I_TYPE; diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 5ec6f831c1..8b8c949302 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -32,10 +32,6 @@ // Uncomment this to enable profiling of the GS RingBufferCopy function. //#define PCSX2_GSRING_SAMPLING_STATS -#ifdef PCSX2_GSRING_TX_STATS -#include -#endif - using namespace Threading; using namespace std; @@ -178,7 +174,7 @@ mtgsThreadObject* mtgsThread = NULL; std::list ringposStack; #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // debug variable used to check for bad code bits where copies are started // but never closed, or closed without having been started. (GSRingBufCopy calls // should always be followed by a call to GSRINGBUF_DONECOPY) @@ -756,24 +752,25 @@ void mtgsThreadObject::SendDataPacket() jASSUME( temp <= m_RingBufferSize ); temp &= m_RingBufferMask; -#ifdef _DEBUG - if( m_packet_ringpos + m_packet_size < m_RingBufferSize ) + if( IsDebugBuild ) { - uint readpos = volatize(m_RingPos); - if( readpos != m_WritePos ) + if( m_packet_ringpos + m_packet_size < m_RingBufferSize ) { - // The writepos should never leapfrog the readpos - // since that indicates a bad write. - if( m_packet_ringpos < readpos ) - assert( temp < readpos ); - } + uint readpos = volatize(m_RingPos); + if( readpos != m_WritePos ) + { + // The writepos should never leapfrog the readpos + // since that indicates a bad write. + if( m_packet_ringpos < readpos ) + assert( temp < readpos ); + } - // Updating the writepos should never make it equal the readpos, since - // that would stop the buffer prematurely (and indicates bad code in the - // ringbuffer manager) - assert( readpos != temp ); + // Updating the writepos should never make it equal the readpos, since + // that would stop the buffer prematurely (and indicates bad code in the + // ringbuffer manager) + assert( readpos != temp ); + } } -#endif AtomicExchange( m_WritePos, temp ); diff --git a/pcsx2/PrecompiledHeader.h b/pcsx2/PrecompiledHeader.h index c2de341fea..764dc565e6 100644 --- a/pcsx2/PrecompiledHeader.h +++ b/pcsx2/PrecompiledHeader.h @@ -172,7 +172,7 @@ static const bool IsDevBuild = true; static const bool IsDevBuild = false; #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static const bool IsDebugBuild = true; #else static const bool IsDebugBuild = false; diff --git a/pcsx2/R3000AInterpreter.cpp b/pcsx2/R3000AInterpreter.cpp index ec321c609c..4c64e2a682 100644 --- a/pcsx2/R3000AInterpreter.cpp +++ b/pcsx2/R3000AInterpreter.cpp @@ -465,11 +465,6 @@ static void intExecute() { for (;;) execI(); } -#ifdef _DEBUG -extern u32 psxdump; -extern void iDumpPsxRegisters(u32,u32); -#endif - static s32 intExecuteBlock( s32 eeCycles ) { psxBreak = 0; diff --git a/pcsx2/StringUtils.h b/pcsx2/StringUtils.h index 4d37887eab..4d80602f2d 100644 --- a/pcsx2/StringUtils.h +++ b/pcsx2/StringUtils.h @@ -43,7 +43,7 @@ struct _VARG_PARAM uint someval; }; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG #define params va_arg_dummy, #define varg_assert() // jASSUME( dummy == &va_arg_dummy ); diff --git a/pcsx2/System.h b/pcsx2/System.h index 8447b8499f..ef32557ebf 100644 --- a/pcsx2/System.h +++ b/pcsx2/System.h @@ -224,7 +224,7 @@ using Console::Color_White; # define DevMsg #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG # define DbgCon Console #else # define DbgCon 0&&Console diff --git a/pcsx2/VU0microInterp.cpp b/pcsx2/VU0microInterp.cpp index cf70894bd1..da5c9b8bcd 100644 --- a/pcsx2/VU0microInterp.cpp +++ b/pcsx2/VU0microInterp.cpp @@ -220,18 +220,11 @@ namespace VU0micro { int i; - #ifdef _DEBUG - int prevbranch; - #endif - for (i = 128; i--;) { if ((VU0.VI[REG_VPU_STAT].UL & 0x1) == 0) break; - #ifdef _DEBUG - prevbranch = vu0branch; - #endif vu0Exec(&VU0); } diff --git a/pcsx2/VU1micro.cpp b/pcsx2/VU1micro.cpp index 8ebb1b2272..0f8098947c 100644 --- a/pcsx2/VU1micro.cpp +++ b/pcsx2/VU1micro.cpp @@ -32,7 +32,7 @@ VURegs* g_pVU1; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG u32 vudump = 0; #endif diff --git a/pcsx2/VU1microInterp.cpp b/pcsx2/VU1microInterp.cpp index fc8f7f793a..8ec7bc3cca 100644 --- a/pcsx2/VU1microInterp.cpp +++ b/pcsx2/VU1microInterp.cpp @@ -205,17 +205,11 @@ namespace VU1micro static void intExecuteBlock() { int i; - #ifdef _DEBUG - int prevbranch; - #endif for (i = 128; i--;) { if ((VU0.VI[REG_VPU_STAT].UL & 0x100) == 0) break; - #ifdef _DEBUG - prevbranch = vu1branch; - #endif vu1Exec(&VU1); } diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 0276185d61..f79dab837e 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -340,7 +340,7 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int VURegs * VU; u8 *cdata = (u8*)data; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG u32 memsize = VIFdmanum ? 0x4000 : 0x1000; #endif @@ -375,7 +375,7 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int size <<= 2; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG memsize = size; #endif @@ -455,7 +455,7 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int if (vif->cl != 0 || (size & 0xf)) //Check alignment for SSE unpacks { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static int s_count = 0; #endif @@ -552,7 +552,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i u32 tempsize = 0; const u32 memlimit = (VIFdmanum ? 0x4000 : 0x1000); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG u32 memsize = memlimit; #endif @@ -595,14 +595,14 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i size <<= 2; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG memsize = size; #endif if (vifRegs->cycle.cl >= vifRegs->cycle.wl) // skipping write { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static int s_count = 0; #endif if (v->addr >= memlimit) diff --git a/pcsx2/tinyxml/tinyxml.h b/pcsx2/tinyxml/tinyxml.h index 5391236e68..cdec7f3cf5 100644 --- a/pcsx2/tinyxml/tinyxml.h +++ b/pcsx2/tinyxml/tinyxml.h @@ -39,7 +39,7 @@ distribution. #include // Help out windows: -#if defined( _DEBUG ) && !defined( DEBUG ) +#if defined( PCSX2_DEBUG ) && !defined( DEBUG ) #define DEBUG #endif diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 5d2ca8cf13..8e49d65547 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -22,7 +22,7 @@ Name="Debug|Win32" IntermediateDirectory="$(ConfigurationName)" ConfigurationType="1" - InheritedPropertySheets=".\vsprops\common.vsprops;..\..\..\common\vsprops\BaseProperties.vsprops;..\..\..\common\vsprops\3rdpartyDeps.vsprops;.\vsprops\debug.vsprops;.\vsprops\devbuild.vsprops;..\..\..\common\vsprops\pthreads.vsprops;..\..\..\common\vsprops\IncrementalLinking.vsprops" + InheritedPropertySheets=".\vsprops\common.vsprops;..\..\..\common\vsprops\BaseProperties.vsprops;..\..\..\common\vsprops\3rdpartyDeps.vsprops;..\..\..\common\vsprops\pthreads.vsprops;.\vsprops\devbuild.vsprops;..\..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\..\common\vsprops\IncrementalLinking.vsprops" UseOfMFC="0" ATLMinimizesCRunTimeLibraryUsage="false" CharacterSet="2" @@ -54,7 +54,6 @@ Name="VCCLCompilerTool" AdditionalIncludeDirectories="" ExceptionHandling="2" - SmallerTypeCheck="false" UsePrecompiledHeader="2" PrecompiledHeaderThrough="PrecompiledHeader.h" PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch" @@ -99,8 +98,9 @@ diff --git a/pcsx2/windows/WinMain.cpp b/pcsx2/windows/WinMain.cpp index c74e88157f..31cbc0add0 100644 --- a/pcsx2/windows/WinMain.cpp +++ b/pcsx2/windows/WinMain.cpp @@ -171,7 +171,7 @@ void WinClose() ReleasePlugins(); Console::Close(); - pthread_win32_process_detach_np(); + //pthread_win32_process_detach_np(); SysShutdownDynarecs(); SysShutdownMem(); @@ -254,7 +254,7 @@ int WINAPI WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, LPSTR lpCmdLine pInstance=hInstance; FirstShow=true; // this is used by cheats.cpp to search for stuff (broken?) - pthread_win32_process_attach_np(); + //pthread_win32_process_attach_np(); gApp.hInstance = hInstance; gApp.hMenu = NULL; diff --git a/pcsx2/windows/WinThreads.cpp b/pcsx2/windows/WinThreads.cpp index 658aab7a0a..c8b6bbbd43 100644 --- a/pcsx2/windows/WinThreads.cpp +++ b/pcsx2/windows/WinThreads.cpp @@ -52,7 +52,7 @@ namespace Threading LogicalCoresPerPhysicalCPU = CPUs; cpuinfo.PhysicalCores = ( CPUs / LogicalCoresPerPhysicalCPU ) * PhysicalCoresPerPhysicalCPU; - ptw32_smp_system = ( cpuinfo.LogicalCores > 1 ) ? TRUE : FALSE; + //ptw32_smp_system = ( cpuinfo.LogicalCores > 1 ) ? TRUE : FALSE; } __forceinline void Timeslice() @@ -76,13 +76,13 @@ namespace Threading { jASSUME( itsme != NULL ); - pthread_win32_thread_attach_np(); + //pthread_win32_thread_attach_np(); Thread& owner = *((Thread*)itsme); owner.m_returncode = owner.Callback(); owner.m_terminated = true; - pthread_win32_thread_detach_np(); + //pthread_win32_thread_detach_np(); return NULL; } @@ -97,14 +97,13 @@ namespace Threading __forceinline long pcsx2_InterlockedExchange( volatile long* target, long srcval ) { - // Use the pthreads-win32 implementation... - return ptw32_InterlockedExchange( target, srcval ); + return _InterlockedExchange( target, srcval ); } __forceinline long pcsx2_InterlockedCompareExchange( volatile long* target, long srcval, long comp ) { // Use the pthreads-win32 implementation... - return ptw32_InterlockedCompareExchange( target, srcval, comp ); + return _InterlockedCompareExchange( target, srcval, comp ); } __forceinline long pcsx2_InterlockedExchangeAdd( volatile long* target, long srcval ) diff --git a/pcsx2/x86/aVUzerorec.S b/pcsx2/x86/aVUzerorec.S index 251a3327f0..26ecdb375e 100644 --- a/pcsx2/x86/aVUzerorec.S +++ b/pcsx2/x86/aVUzerorec.S @@ -26,7 +26,7 @@ SuperVUExecuteProgram: mov s_vu1esi, esi mov s_vuedi, edi mov s_vuebx, ebx -#ifdef _DEBUG +#ifdef PCSX2_DEBUG mov s_vu1esp, esp #endif @@ -44,7 +44,7 @@ SuperVUEndProgram: mov edi, s_vuedi mov ebx, s_vuebx -#ifdef _DEBUG +#ifdef PCSX2_DEBUG sub s_vu1esp, esp #endif diff --git a/pcsx2/x86/fast_routines.cpp b/pcsx2/x86/fast_routines.cpp index d907cfaf03..76f8de5b41 100644 --- a/pcsx2/x86/fast_routines.cpp +++ b/pcsx2/x86/fast_routines.cpp @@ -78,7 +78,7 @@ MEMCPY_AMD.CPP #if defined(_MSC_VER) -#ifdef _DEBUG +#ifdef PCSX2_DEBUG extern u8 g_globalMMXSaved; void checkregs() diff --git a/pcsx2/x86/iCore.cpp b/pcsx2/x86/iCore.cpp index 97600e541f..48c94d210e 100644 --- a/pcsx2/x86/iCore.cpp +++ b/pcsx2/x86/iCore.cpp @@ -36,10 +36,6 @@ u32 g_cpuRegHasSignExt = 0, g_cpuPrevRegHasSignExt = 0; // set if upper 32 bits // use FreezeMMXRegs, FreezeXMMRegs u32 g_recWriteback = 0; -#ifdef _DEBUG -char g_globalXMMLocked = 0; -#endif - _xmmregs xmmregs[iREGCNT_XMM], s_saveXMMregs[iREGCNT_XMM]; // X86 caching diff --git a/pcsx2/x86/iCore.h b/pcsx2/x86/iCore.h index 0139444df5..6c947f47b8 100644 --- a/pcsx2/x86/iCore.h +++ b/pcsx2/x86/iCore.h @@ -292,10 +292,6 @@ extern _xmmregs xmmregs[iREGCNT_XMM], s_saveXMMregs[iREGCNT_XMM]; extern u16 g_x86AllocCounter; extern u16 g_xmmAllocCounter; -#ifdef _DEBUG -extern char g_globalXMMLocked; -#endif - // allocates only if later insts use XMM, otherwise checks int _allocCheckGPRtoXMM(EEINST* pinst, int gprreg, int mode); int _allocCheckFPUtoXMM(EEINST* pinst, int fpureg, int mode); diff --git a/pcsx2/x86/iR3000A.cpp b/pcsx2/x86/iR3000A.cpp index d2a20cdaff..c754e2f390 100644 --- a/pcsx2/x86/iR3000A.cpp +++ b/pcsx2/x86/iR3000A.cpp @@ -133,11 +133,7 @@ void rpsxpropBSC(EEINST* prev, EEINST* pinst); static void iopClearRecLUT(BASEBLOCK* base, int count); -#ifdef _DEBUG -u32 psxdump = 0; -#else -#define psxdump 0 -#endif +static u32 psxdump = 0; #define PSX_GETBLOCK(x) PC_GETBLOCK_(x, psxRecLUT) @@ -146,10 +142,11 @@ u32 psxdump = 0; psxRecClearMem(mem) : 4) //////////////////////////////////////////////////// -#ifdef _DEBUG using namespace R3000A; static void iIopDumpBlock( int startpc, u8 * ptr ) { +#ifdef PCSX2_DEBUG + FILE *f; #ifdef __LINUX__ char command[256]; @@ -222,8 +219,8 @@ static void iIopDumpBlock( int startpc, u8 * ptr ) system(command); f = fopen( filename.c_str(), "a+" ); #endif -} #endif +} u8 _psxLoadWritesRs(u32 tempcode) { @@ -307,19 +304,9 @@ void _psxFlushConstRegs() MOV32ItoM((uptr)&psxRegs.GPR.r[i], g_psxConstRegs[i]); g_psxFlushedConstReg |= 1<>26) == 9 ) { //ADDIU, call bios -#ifdef _DEBUG - MOV32ItoM( (uptr)&psxRegs.code, psxRegs.code ); - MOV32ItoM( (uptr)&psxRegs.pc, psxpc ); - _psxFlushCall(FLUSH_NODESTROY); - CALLFunc((uptr)zeroEx); -#endif + if( IsDebugBuild ) + { + MOV32ItoM( (uptr)&psxRegs.code, psxRegs.code ); + MOV32ItoM( (uptr)&psxRegs.pc, psxpc ); + _psxFlushCall(FLUSH_NODESTROY); + CALLFunc((uptr)zeroEx); + } // Bios Call: Force the IOP to do a Branch Test ASAP. // Important! This helps prevent game freeze-ups during boot-up and stage loads. // Note: Fixes to cdvd have removed the need for this code. @@ -892,23 +880,13 @@ void psxRecompileNextInstruction(int delayslot) // pblock isn't used elsewhere in this function. BASEBLOCK* pblock = PSX_GETBLOCK(psxpc); -#ifdef _DEBUG - MOV32ItoR(EAX, psxpc); -#endif + if( IsDebugBuild ) + MOV32ItoR(EAX, psxpc); psxRegs.code = iopMemRead32( psxpc ); s_psxBlockCycles++; psxpc += 4; -//#ifdef _DEBUG -// CMP32ItoM((uptr)s_pCode, psxRegs.code); -// j8Ptr[0] = JE8(0); -// MOV32ItoR(EAX, psxpc); -// CALLFunc((uptr)checkcodefn); -// x86SetJ8( j8Ptr[ 0 ] ); -//#endif - - g_pCurInstInfo++; g_iopCyclePenalty = 0; @@ -926,9 +904,9 @@ void psxRecompileNextInstruction(int delayslot) _clearNeededX86regs(); } -#ifdef _DEBUG static void printfn() { +#ifdef PCSX2_DEBUG extern void iDumpPsxRegisters(u32 startpc, u32 temp); static int lastrec = 0; @@ -948,8 +926,8 @@ static void printfn() lastrec = g_psxlastpc; } -} #endif +} void iopRecRecompile(u32 startpc) { @@ -957,12 +935,11 @@ void iopRecRecompile(u32 startpc) u32 branchTo; u32 willbranch3 = 0; -#ifdef _DEBUG - extern void iDumpPsxRegisters(u32 startpc, u32 temp); - - if( psxdump & 4 ) + if( IsDebugBuild && (psxdump & 4) ) + { + extern void iDumpPsxRegisters(u32 startpc, u32 temp); iDumpPsxRegisters(startpc, 0); -#endif + } assert( startpc ); @@ -994,11 +971,11 @@ void iopRecRecompile(u32 startpc) _initX86regs(); -#ifdef _DEBUG - // for debugging purposes - MOV32ItoM((uptr)&g_psxlastpc, psxpc); - CALLFunc((uptr)printfn); -#endif + if( IsDebugBuild ) + { + MOV32ItoM((uptr)&g_psxlastpc, psxpc); + CALLFunc((uptr)printfn); + } // go until the next branch i = startpc; @@ -1082,27 +1059,26 @@ StartRecomp: } } -#ifdef _DEBUG // dump code - for(i = 0; i < ArraySize(s_psxrecblocks); ++i) { + if( IsDebugBuild ) + { + for(i = 0; i < ArraySize(s_psxrecblocks); ++i) { if( startpc == s_psxrecblocks[i] ) { iIopDumpBlock(startpc, recPtr); + } } - } - if( (psxdump & 1) ) - iIopDumpBlock(startpc, recPtr); -#endif + if( (psxdump & 1) ) + iIopDumpBlock(startpc, recPtr); + } g_pCurInstInfo = s_pInstCache; while (!psxbranch && psxpc < s_nEndBlock) { psxRecompileNextInstruction(0); } -#ifdef _DEBUG - if( (psxdump & 1) ) + if( IsDebugBuild && (psxdump & 1) ) iIopDumpBlock(startpc, recPtr); -#endif assert( (psxpc-startpc)>>2 <= 0xffff ); s_pCurBlockEx->size = (psxpc-startpc)>>2; diff --git a/pcsx2/x86/iVU1micro.cpp b/pcsx2/x86/iVU1micro.cpp index 379eacb4c5..16f2bdf99d 100644 --- a/pcsx2/x86/iVU1micro.cpp +++ b/pcsx2/x86/iVU1micro.cpp @@ -80,7 +80,7 @@ void VUtestPause() { void VUtestPause() {} #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG extern u32 vudump; #endif diff --git a/pcsx2/x86/ix86-32/iR5900-32.cpp b/pcsx2/x86/ix86-32/iR5900-32.cpp index 48212413e8..88a06b57eb 100644 --- a/pcsx2/x86/ix86-32/iR5900-32.cpp +++ b/pcsx2/x86/ix86-32/iR5900-32.cpp @@ -103,7 +103,7 @@ static EEINST* s_psaveInstInfo = NULL; static u32 s_savenBlockCycles = 0; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static u32 dumplog = 0; #else #define dumplog 0 @@ -1024,10 +1024,6 @@ static u32 eeScaleBlockCycles() // setting "branch = 2"; static void iBranchTest(u32 newpc, bool noDispatch) { -#ifdef _DEBUG - //CALLFunc((uptr)testfpu); -#endif - if( bExecBIOS ) CheckForBIOSEnd(); // Check the Event scheduler if our "cycle target" has been reached. @@ -1089,7 +1085,7 @@ void recompileNextInstruction(int delayslot) pc += 4; #if 0 -#ifdef _DEBUG +#ifdef PCSX2_DEBUG CMP32ItoM((u32)s_pCode, cpuRegs.code); j8Ptr[0] = JE8(0); MOV32ItoR(EAX, pc); @@ -1188,8 +1184,6 @@ void recompileNextInstruction(int delayslot) s_nEndBlock = pc; } -extern u32 psxdump; - static void printfn() { static int lastrec = 0; @@ -1241,7 +1235,7 @@ void recRecompile( const u32 startpc ) u32 willbranch3 = 0; u32 usecop2; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG //dumplog |= 4; if( dumplog & 4 ) iDumpRegisters(startpc, 0); @@ -1293,7 +1287,7 @@ void recRecompile( const u32 startpc ) _initXMMregs(); _initMMXregs(); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // for debugging purposes MOV32ItoM((uptr)&g_lastpc, pc); CALLFunc((uptr)printfn); @@ -1478,7 +1472,7 @@ StartRecomp: } } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // dump code for(i = 0; i < ArraySize(s_recblocks); ++i) { if( startpc == s_recblocks[i] ) { @@ -1571,7 +1565,7 @@ StartRecomp: recompileNextInstruction(0); // For the love of recursion, batman! } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( (dumplog & 1) ) iDumpBlock(startpc, recPtr); #endif diff --git a/pcsx2/x86/ix86-32/iR5900LoadStore.cpp b/pcsx2/x86/ix86-32/iR5900LoadStore.cpp index ba09e7f748..6416de45da 100644 --- a/pcsx2/x86/ix86-32/iR5900LoadStore.cpp +++ b/pcsx2/x86/ix86-32/iR5900LoadStore.cpp @@ -130,7 +130,7 @@ void recTransferX86ToReg(int x86reg, int gprreg, int sign) } } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG void testaddrs() { register int tempaddr; @@ -231,7 +231,7 @@ int recSetMemLocation(int regs, int imm, int mmreg, int msize, int j32) if ( imm != 0 ) ADD32ItoR( ECX, imm ); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG //CALLFunc((uptr)testaddrs); #endif @@ -256,7 +256,7 @@ int recSetMemLocation(int regs, int imm, int mmreg, int msize, int j32) x86SetJ8(ptr); if( msize == 1 ) AND8ItoR(ECX, 0xf8); else if( msize == 2 ) AND8ItoR(ECX, 0xf0); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG MOV32RtoR(EAX, ECX); SHR32ItoR(EAX, 28); CMP32ItoR(EAX, 1); diff --git a/pcsx2/x86/sVU_zerorec.cpp b/pcsx2/x86/sVU_zerorec.cpp index c02a7366fd..95aedcc6f7 100644 --- a/pcsx2/x86/sVU_zerorec.cpp +++ b/pcsx2/x86/sVU_zerorec.cpp @@ -57,7 +57,7 @@ extern void iDumpVU1Registers(); #define SUPERVU_PROPAGATEFLAGS // the correct behavior of VUs, for some reason superman breaks gfx with it on... // registers won't be flushed at block boundaries (faster) (nothing noticable speed-wise, causes SPS in Ratchet and clank (Nneeve) ) -#ifndef _DEBUG +#ifndef PCSX2_DEBUG //#define SUPERVU_INTERCACHING #endif @@ -104,7 +104,7 @@ extern void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info); //#define FORIT(it, v) for(it = (v).begin(); it != (v).end(); ++(it)) -#ifdef _DEBUG +#ifdef PCSX2_DEBUG u32 s_vucount = 0; static u32 g_vu1lastrec = 0, skipparent = -1; @@ -429,7 +429,7 @@ void SuperVUDestroy(int vuindex) // reset VU void SuperVUReset(int vuindex) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG s_vucount = 0; #endif @@ -722,16 +722,6 @@ void* SuperVUGetProgram(u32 startpc, int vuindex) if (*pheader == NULL) { -#ifdef _DEBUG -// if( vuindex ) VU1.VI[REG_TPC].UL = startpc; -// else VU0.VI[REG_TPC].UL = startpc; -// __Log("VU: %x\n", startpc); -// iDumpVU1Registers(); -// vudump |= 2; -#endif - - // measure run time - //QueryPerformanceCounter(&svubase); #ifdef SUPERVU_CACHING void* pmem = (vuindex & 1) ? VU1.Micro : VU0.Micro; @@ -764,9 +754,6 @@ void* SuperVUGetProgram(u32 startpc, int vuindex) return (void*)SuperVUEndProgram; } - //QueryPerformanceCounter(&svufinal); - //svutime += (u32)(svufinal.QuadPart-svubase.QuadPart); - assert((*pheader)->pprogfunc != NULL); } //else assert( (*pheader)->IsSame((vuindex&1) ? VU1.Micro : VU0.Micro) ); @@ -906,7 +893,7 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) SuperVUEliminateDeadCode(); SuperVUAssignRegs(); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); #endif @@ -934,7 +921,7 @@ static VuFunctionHeader* SuperVURecompileProgram(u32 startpc, int vuindex) s_pFnHeader->ranges.push_back(r); } -#if defined(_DEBUG) && defined(__LINUX__) +#if defined(PCSX2_DEBUG) && defined(__LINUX__) // dump at the end to capture the actual code if ((s_vu && (vudump&1)) || (!s_vu && (vudump&16))) SuperVUDumpBlock(s_listBlocks, s_vu); #endif @@ -1855,7 +1842,7 @@ static void SuperVUEliminateDeadCode() for(itblock = s_listBlocks.begin(); itblock != s_listBlocks.end(); itblock++) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG u32 startpc = (*itblock)->startpc; u32 curpc = startpc; #endif @@ -1965,8 +1952,8 @@ static void SuperVUEliminateDeadCode() continue; } } -#ifdef _DEBUG - else +#ifdef PCSX2_DEBUG + else { curpc += 8; } @@ -2601,7 +2588,7 @@ __declspec(naked) void SuperVUExecuteProgram(u32 startpc, int vuindex) mov s_vuedi, edi // have to save even in Release mov s_vuebx, ebx } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG __asm { mov s_vu1esp, esp @@ -2635,7 +2622,7 @@ __declspec(naked) static void SuperVUEndProgram() mov ebx, s_vuebx } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG __asm { sub s_vu1esp, esp @@ -2820,7 +2807,7 @@ void svudispfntemp() { #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static u32 i; if (((vudump&8) && g_curdebugvu) || ((vudump&0x80) && !g_curdebugvu)) //&& g_vu1lastrec != g_vu1last ) { @@ -2938,7 +2925,7 @@ void VuBaseBlock::Recompile() x86Align(16); pcode = x86Ptr; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG MOV32ItoM((uptr)&s_vufnheader, s_pFnHeader->startpc); MOV32ItoM((uptr)&VU->VI[REG_TPC], startpc); MOV32ItoM((uptr)&s_svulast, startpc); @@ -3593,9 +3580,8 @@ void VuInstruction::Recompile(list::iterator& itinst, u32 vuxyz) s_pCurBlock->prevFlagsOutOfBlock = 0; -#ifdef _DEBUG - MOV32ItoR(EAX, pc); -#endif + if( IsDebugBuild ) + MOV32ItoR(EAX, pc); assert(!(type & (INST_CLIP_WRITE | INST_STATUS_WRITE | INST_MAC_WRITE))); pc += 8; diff --git a/pcsx2_suite_2008.sln b/pcsx2_suite_2008.sln index 57f5664f52..861c65f948 100644 --- a/pcsx2_suite_2008.sln +++ b/pcsx2_suite_2008.sln @@ -77,6 +77,7 @@ EndProject Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "Include", "Include", "{0FADC26C-0E9D-4DD7-84B1-BF4F7754E90C}" ProjectSection(SolutionItems) = preProject common\include\afxresmw.h = common\include\afxresmw.h + common\include\intrin_x86.h = common\include\intrin_x86.h common\include\PS2Edefs.h = common\include\PS2Edefs.h common\include\PS2Etypes.h = common\include\PS2Etypes.h EndProjectSection diff --git a/plugins/CDVDiso/src/CDVDisop.cpp b/plugins/CDVDiso/src/CDVDisop.cpp index b2260b9798..96c16b75ce 100755 --- a/plugins/CDVDiso/src/CDVDisop.cpp +++ b/plugins/CDVDiso/src/CDVDisop.cpp @@ -47,7 +47,7 @@ char *methods[] = NULL }; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG char *LibName = "Linuz Iso CDVD (Debug) "; #else char *LibName = "Linuz Iso CDVD "; @@ -95,7 +95,7 @@ EXPORT_C_(u32) PS2EgetLibVersion2(u32 type) return (version << 16) | (revision << 8) | build; } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG void __Log(char *fmt, ...) { va_list list; @@ -113,7 +113,7 @@ void __Log(char *fmt, ...) EXPORT_C_(s32) CDVDinit() { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG cdvdLog = fopen("logs/cdvdLog.txt", "w"); if (cdvdLog == NULL) { diff --git a/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj b/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj index e99a890458..b9a31ebf39 100644 --- a/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj +++ b/plugins/CDVDiso/src/Windows/CDVDiso_vs2008.vcproj @@ -4,9 +4,8 @@ Version="9.00" Name="CDVDiso" ProjectGUID="{5F78E90B-BD22-47B1-9CA5-7A80F4DF5EF3}" - RootNamespace="CDVDiso_vs2008" - Keyword="Win32Proj" - TargetFrameworkVersion="196613" + RootNamespace="CDVDiso" + TargetFrameworkVersion="131072" > @@ -115,21 +110,8 @@ /> @@ -197,20 +178,8 @@ /> diff --git a/plugins/CDVDnull/CDVD.h b/plugins/CDVDnull/CDVD.h index 05858f3c41..a2c4c771fb 100644 --- a/plugins/CDVDnull/CDVD.h +++ b/plugins/CDVDnull/CDVD.h @@ -23,11 +23,8 @@ #include #endif -extern "C" -{ #define CDVDdefs #include "PS2Edefs.h" -} #ifdef __LINUX__ #include diff --git a/plugins/CDVDnull/Windows/CDVDnull_vs2008.vcproj b/plugins/CDVDnull/Windows/CDVDnull_vs2008.vcproj index 26d4c17132..129cc2fa95 100644 --- a/plugins/CDVDnull/Windows/CDVDnull_vs2008.vcproj +++ b/plugins/CDVDnull/Windows/CDVDnull_vs2008.vcproj @@ -18,7 +18,7 @@ @@ -43,12 +43,8 @@ /> @@ -152,7 +152,7 @@ Name="Debug SSE2|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops;.\vsprops\sse2.vsprops" - UseOfMFC="1" + UseOfMFC="2" CharacterSet="2" > @@ -281,7 +281,7 @@ Name="Release SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\release.vsprops;.\vsprops\sse2.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="2" CharacterSet="2" WholeProgramOptimization="1" > @@ -346,7 +346,7 @@ Name="Debug SSSE3|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops;.\vsprops\sse2.vsprops;.\vsprops\ssse3.vsprops" - UseOfMFC="1" + UseOfMFC="2" CharacterSet="2" > diff --git a/plugins/CDVDolio/vsprops/debug.vsprops b/plugins/CDVDolio/vsprops/debug.vsprops index 9d5c60bf96..b7506beeec 100644 --- a/plugins/CDVDolio/vsprops/debug.vsprops +++ b/plugins/CDVDolio/vsprops/debug.vsprops @@ -10,7 +10,7 @@ PreprocessorDefinitions="_DEBUG" MinimalRebuild="true" BasicRuntimeChecks="1" - RuntimeLibrary="1" + RuntimeLibrary="3" /> -#ifdef __cplusplus -extern "C" -{ -#endif #define FWdefs #include "PS2Edefs.h" -#ifdef __cplusplus -} -#endif #ifdef _WIN32 diff --git a/plugins/FWnull/Windows/FWnull_vc2008.vcproj b/plugins/FWnull/Windows/FWnull_vc2008.vcproj index 7ca94e36c7..651030099f 100644 --- a/plugins/FWnull/Windows/FWnull_vc2008.vcproj +++ b/plugins/FWnull/Windows/FWnull_vc2008.vcproj @@ -39,9 +39,6 @@ /> @@ -105,10 +102,6 @@ /> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - @@ -1206,6 +1190,14 @@ AssemblerOutput="4" /> + + + - - - @@ -1522,14 +1506,6 @@ UsePrecompiledHeader="1" /> - - - @@ -1538,14 +1514,6 @@ UsePrecompiledHeader="1" /> - - - @@ -1554,14 +1522,6 @@ UsePrecompiledHeader="1" /> - - - @@ -1570,14 +1530,6 @@ UsePrecompiledHeader="1" /> - - - @@ -1586,14 +1538,6 @@ UsePrecompiledHeader="1" /> - - - @@ -1603,7 +1547,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2030,14 +2014,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2046,14 +2022,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2062,14 +2030,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2078,14 +2038,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2094,14 +2046,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2111,7 +2055,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2166,14 +2150,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2182,14 +2158,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2198,14 +2166,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2214,14 +2174,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2230,14 +2182,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2247,7 +2191,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2302,14 +2286,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2318,14 +2294,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2334,14 +2302,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2350,14 +2310,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2366,14 +2318,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2383,7 +2327,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2438,14 +2422,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2454,14 +2430,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2470,14 +2438,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2486,14 +2446,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2502,14 +2454,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2519,7 +2463,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2574,14 +2558,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2590,14 +2566,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2606,14 +2574,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2622,14 +2582,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2638,14 +2590,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2655,7 +2599,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2710,14 +2694,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2726,14 +2702,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2742,14 +2710,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2758,14 +2718,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2774,14 +2726,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2791,7 +2735,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2862,14 +2846,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2878,14 +2854,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2894,14 +2862,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2910,14 +2870,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2926,14 +2878,6 @@ UsePrecompiledHeader="0" /> - - - @@ -2943,7 +2887,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -2998,14 +2982,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3014,14 +2990,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3030,14 +2998,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3046,14 +3006,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3062,14 +3014,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3079,7 +3023,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3134,14 +3118,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3150,14 +3126,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3166,14 +3134,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3182,14 +3142,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3198,14 +3150,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3215,7 +3159,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3270,14 +3254,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3286,14 +3262,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3302,14 +3270,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3318,14 +3278,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3334,14 +3286,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3351,7 +3295,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3406,14 +3390,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3422,14 +3398,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3438,14 +3406,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3454,14 +3414,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3470,14 +3422,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3487,7 +3431,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3546,14 +3530,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3562,14 +3538,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3578,14 +3546,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3594,14 +3554,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3610,14 +3562,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3627,7 +3571,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3682,14 +3666,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3698,14 +3674,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3714,14 +3682,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3730,14 +3690,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3746,14 +3698,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3763,7 +3707,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3818,14 +3802,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3834,14 +3810,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3850,14 +3818,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3866,14 +3826,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3882,14 +3834,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3899,7 +3843,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -3954,14 +3938,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3970,14 +3946,6 @@ UsePrecompiledHeader="0" /> - - - @@ -3986,14 +3954,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4002,14 +3962,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4018,14 +3970,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4035,7 +3979,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4094,14 +4078,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4110,14 +4086,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4126,14 +4094,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4142,14 +4102,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4158,14 +4110,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4175,7 +4119,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4230,14 +4214,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4246,14 +4222,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4262,14 +4230,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4278,14 +4238,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4294,14 +4246,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4311,7 +4255,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4366,14 +4350,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4382,14 +4358,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4398,14 +4366,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4414,14 +4374,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4430,14 +4382,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4447,7 +4391,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4502,14 +4486,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4518,14 +4494,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4534,14 +4502,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4550,14 +4510,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4566,14 +4518,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4583,7 +4527,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4638,14 +4622,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4654,14 +4630,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4670,14 +4638,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4686,14 +4646,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4702,14 +4654,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4719,7 +4663,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4774,14 +4758,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4790,14 +4766,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4806,14 +4774,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4822,14 +4782,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4838,14 +4790,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4855,7 +4799,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -4910,14 +4894,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4926,14 +4902,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4942,14 +4910,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4958,14 +4918,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4974,14 +4926,6 @@ UsePrecompiledHeader="0" /> - - - @@ -4991,7 +4935,7 @@ /> + + + + + + + + + + + + + + + + + + - - - @@ -5046,14 +5030,6 @@ UsePrecompiledHeader="0" /> - - - @@ -5062,14 +5038,6 @@ UsePrecompiledHeader="0" /> - - - @@ -5078,14 +5046,6 @@ UsePrecompiledHeader="0" /> - - - @@ -5094,14 +5054,6 @@ UsePrecompiledHeader="0" /> - - - @@ -5110,14 +5062,6 @@ UsePrecompiledHeader="0" /> - - - @@ -5127,7 +5071,7 @@ /> + + + + + + + + + + + + + + + + + + diff --git a/plugins/LilyPad/LilyPad.cpp b/plugins/LilyPad/LilyPad.cpp index 6f90966ac7..c878eed07d 100644 --- a/plugins/LilyPad/LilyPad.cpp +++ b/plugins/LilyPad/LilyPad.cpp @@ -18,7 +18,7 @@ #include "svnrev.h" #include "resource.h" -#ifdef _DEBUG +#ifdef PCSX2_DEBUG #include "crtdbg.h" #endif @@ -577,7 +577,7 @@ u32 CALLBACK PS2EgetLibVersion2(u32 type) { // Used in about and config screens. void GetNameAndVersionString(wchar_t *out) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG wsprintfW(out, L"LilyPad Debug %i.%i.%i (r%i)", (VERSION>>8)&0xFF, VERSION&0xFF, (VERSION>>24)&0xFF, SVN_REV); #elif (_MSC_VER != 1400) wsprintfW(out, L"LilyPad svn %i.%i.%i (r%i)", (VERSION>>8)&0xFF, VERSION&0xFF, (VERSION>>24)&0xFF, SVN_REV); @@ -587,7 +587,7 @@ void GetNameAndVersionString(wchar_t *out) { } char* CALLBACK PSEgetLibName() { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG static char version[50]; sprintf(version, "LilyPad Debug (r%i)", SVN_REV); return version; @@ -680,7 +680,7 @@ s32 CALLBACK PADinit(u32 flags) { return PADinit(2); } - #ifdef _DEBUG + #ifdef PCSX2_DEBUG int tmpFlag = _CrtSetDbgFlag( _CRTDBG_REPORT_FLAG ); tmpFlag |= _CRTDBG_LEAK_CHECK_DF; _CrtSetDbgFlag( tmpFlag ); diff --git a/plugins/LilyPad/LilyPad_VC2008.vcproj b/plugins/LilyPad/LilyPad_VC2008.vcproj index 4c97c07f5e..8e7255761f 100644 --- a/plugins/LilyPad/LilyPad_VC2008.vcproj +++ b/plugins/LilyPad/LilyPad_VC2008.vcproj @@ -21,7 +21,7 @@ @@ -49,14 +49,7 @@ diff --git a/plugins/PadNull/Windows/PadNull_vc2008.vcproj b/plugins/PadNull/Windows/PadNull_vc2008.vcproj index 8d8eade5da..5d19f70dfb 100644 --- a/plugins/PadNull/Windows/PadNull_vc2008.vcproj +++ b/plugins/PadNull/Windows/PadNull_vc2008.vcproj @@ -84,7 +84,7 @@ diff --git a/plugins/USBnull/USB.h b/plugins/USBnull/USB.h index 699ba1a432..382753ba3a 100644 --- a/plugins/USBnull/USB.h +++ b/plugins/USBnull/USB.h @@ -21,11 +21,8 @@ #include -extern "C" -{ #define USBdefs #include "PS2Edefs.h" -} #ifdef _WIN32 diff --git a/plugins/USBnull/Windows/USBnull_vc2008.vcproj b/plugins/USBnull/Windows/USBnull_vc2008.vcproj index 6aa6dc011d..fd6980a38b 100644 --- a/plugins/USBnull/Windows/USBnull_vc2008.vcproj +++ b/plugins/USBnull/Windows/USBnull_vc2008.vcproj @@ -39,9 +39,6 @@ /> @@ -105,10 +102,6 @@ /> #endif -extern "C" -{ #define DEV9defs #include "PS2Edefs.h" -} typedef struct { int Log; diff --git a/plugins/dev9null/Windows/DEV9null_vc2008.vcproj b/plugins/dev9null/Windows/DEV9null_vc2008.vcproj index 84c15bd706..d572808963 100644 --- a/plugins/dev9null/Windows/DEV9null_vc2008.vcproj +++ b/plugins/dev9null/Windows/DEV9null_vc2008.vcproj @@ -86,7 +86,7 @@ @@ -107,10 +107,6 @@ /> @@ -193,7 +180,7 @@ #include -// need C definitions -extern "C" { #define GSdefs #include "PS2Edefs.h" -} #ifdef _WIN32 diff --git a/plugins/zerogs/dx/GSmain.cpp b/plugins/zerogs/dx/GSmain.cpp index 0ee9b5ce74..8ccb6c057c 100644 --- a/plugins/zerogs/dx/GSmain.cpp +++ b/plugins/zerogs/dx/GSmain.cpp @@ -46,7 +46,7 @@ using namespace std; #pragma warning(disable:4244) #endif -#ifdef _DEBUG +#ifdef PCSX2_DEBUG HANDLE g_hCurrentThread = NULL; #endif @@ -73,7 +73,7 @@ unsigned char revision = 0; // revision and build gives plugin version unsigned char build = VER; unsigned char minor = 1; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG char *libraryName = "ZeroGS (Debug) "; #elif defined(RELEASE_TO_PUBLIC) @@ -354,7 +354,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) { GS_LOG("GSopen\n"); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG g_hCurrentThread = GetCurrentThread(); #endif @@ -760,7 +760,7 @@ void CALLBACK GSgetLastTag(u64* ptag) void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG assert( g_hCurrentThread == GetCurrentThread() ); #endif @@ -964,7 +964,7 @@ void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr) addr &= 0x3fff; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG PRIM_LOG("count: %d\n", count); count++; #endif diff --git a/plugins/zerogs/dx/Regs.cpp b/plugins/zerogs/dx/Regs.cpp index 7f4673a5e6..5d094feba1 100644 --- a/plugins/zerogs/dx/Regs.cpp +++ b/plugins/zerogs/dx/Regs.cpp @@ -77,7 +77,7 @@ u32 s_uTex1Data[2][2] = {0}, s_uClampData[2] = {0}; void __fastcall GIFPackedRegHandlerNull(u32* data) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG DEBUG_LOG("Unexpected packed reg handler %8.8lx_%8.8lx %x\n", data[0], data[1], data[2]); #endif } @@ -367,7 +367,7 @@ __forceinline void clampWrite(int i, u32 *data) void __fastcall GIFRegHandlerNull(u32* data) { -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( (((uptr)&data[2])&0xffff) == 0 ) return; diff --git a/plugins/zerogs/dx/Windows/zerogs_2008.vcproj b/plugins/zerogs/dx/Windows/zerogs_2008.vcproj index 06357c714d..da22f2bdc5 100644 --- a/plugins/zerogs/dx/Windows/zerogs_2008.vcproj +++ b/plugins/zerogs/dx/Windows/zerogs_2008.vcproj @@ -18,7 +18,7 @@ #endif -extern "C" { - #include "PS2Etypes.h" +extern "C" { + #if defined(_MSC_VER) && !defined(__x86_64__) void * memcpy_amd(void *dest, const void *src, size_t n) diff --git a/plugins/zerogs/dx/targets.cpp b/plugins/zerogs/dx/targets.cpp index 84b867cda0..5ab52c8b70 100644 --- a/plugins/zerogs/dx/targets.cpp +++ b/plugins/zerogs/dx/targets.cpp @@ -268,7 +268,7 @@ void ZeroGS::CRenderTarget::Resolve() D3DLOCKED_RECT locksrc; pd3dDevice->GetRenderTargetData(psurf, psys); -#if !defined(RELEASE_TO_PUBLIC) && defined(_DEBUG) +#if !defined(RELEASE_TO_PUBLIC) && defined(PCSX2_DEBUG) if( g_bSaveResolved ) { D3DXSaveSurfaceToFile("resolved.tga", D3DXIFF_TGA, psys, NULL, NULL); g_bSaveResolved = 0; @@ -293,7 +293,7 @@ void ZeroGS::CRenderTarget::Resolve(int startrange, int endrange) if( vb[0].prndr == this || vb[0].pdepth == this ) Flush(0); if( vb[1].prndr == this || vb[1].pdepth == this ) Flush(1); -#if !defined(RELEASE_TO_PUBLIC) && defined(_DEBUG) +#if !defined(RELEASE_TO_PUBLIC) && defined(PCSX2_DEBUG) if( g_bSaveResolved ) { D3DXSaveSurfaceToFile("resolved.tga", D3DXIFF_TGA, psys, NULL, NULL); g_bSaveResolved = 0; @@ -589,7 +589,7 @@ void ZeroGS::CRenderTarget::ConvertTo32() pd3dDevice->SetPixelShader(ppsConvert16to32); pd3dDevice->DrawPrimitive(D3DPT_TRIANGLESTRIP, 0, 2); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG //g_bSaveZUpdate = 1; //D3DXSaveSurfaceToFile("tex3.tga", D3DXIFF_TGA, psurfConv, NULL, NULL); if( g_bSaveZUpdate ) { @@ -801,7 +801,7 @@ void ZeroGS::CRenderTarget::ConvertTo16() pd3dDevice->SetPixelShader(ppsConvert32to16); pd3dDevice->DrawPrimitive(D3DPT_TRIANGLESTRIP, 0, 2); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG //g_bSaveZUpdate = 1; if( g_bSaveZUpdate ) { pd3dDevice->GetRenderTargetData(psurf, psys); @@ -1186,7 +1186,7 @@ void ZeroGS::CDepthTarget::Update(int context, ZeroGS::CRenderTarget* prndr) if( conf.options & GSOPTION_WIREFRAME ) SETRS(D3DRS_FILLMODE, D3DFILL_WIREFRAME); SETRS(D3DRS_SCISSORTESTENABLE, TRUE); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( g_bSaveZUpdate ) { if( pusetarg != NULL ) D3DXSaveTextureToFile("tex.tga", D3DXIFF_TGA, pusetarg->ptex, NULL); @@ -1268,7 +1268,7 @@ CRenderTarget* ZeroGS::CRenderTargetMngr::GetTarg(const frameInfo& frame, DWORD // can be both 16bit and 32bit if( (frame.psm&2) != (it->second->psm&2) ) { // a lot of games do this actually... -#ifdef _DEBUG +#ifdef PCSX2_DEBUG WARN_LOG("Really bad formats! %d %d\n", frame.psm, it->second->psm); #endif if( !(opts&TO_StrictHeight) ) { @@ -1949,7 +1949,7 @@ ZeroGS::CMemoryTarget* ZeroGS::CMemoryTargetMngr::GetMemoryTarget(const tex0Info ++it; } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG PRIM_LOG("memtarget: tbp: %x tbw: %x th: %x psm: %x\n", tex0.tbp0, tex0.tbw, tex0.th, tex0.psm); #endif @@ -2383,7 +2383,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) { int imin = 0, imax = (int)ranges.size(), imid; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2433,7 +2433,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) if( startindex == 0 && end < ranges.front().start ) { ranges.insert(ranges.begin(), RANGE(start, end)); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2462,7 +2462,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) // create a new range ranges.insert(ranges.begin()+startindex, RANGE(start, end)); -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2480,7 +2480,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) if( lastend > ranges.back().end ) ranges.back().end = lastend; if( end > ranges.back().end ) ranges.back().end = end; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2493,7 +2493,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) if( start < ranges.front().start ) ranges.front().start = start; if( end > ranges.front().end ) ranges.front().end = end; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2509,7 +2509,7 @@ void ZeroGS::CRangeManager::Insert(int start, int end) if( start < ranges[startindex].start ) ranges[startindex].start = start; if( end > ranges[startindex].end ) ranges[startindex].end = end; -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // sanity check for(int i = 0; i < (int)ranges.size()-1; ++i) assert( ranges[i].end < ranges[i+1].start ); #endif @@ -2569,7 +2569,7 @@ void FlushTransferRanges(const tex0Info* ptex) // Flush(0); // Flush(1); // -//#ifdef _DEBUG +//#ifdef PCSX2_DEBUG // // make sure targets are still the same // list::iterator it; // FORIT(it, listTransmissionUpdateTargs) { @@ -2811,7 +2811,7 @@ void TransferHostLocal(const void* pbyMem, u32 nQWordSize) } else s_vTransferCache.resize(0); -#if !defined(RELEASE_TO_PUBLIC) && defined(_DEBUG) +#if !defined(RELEASE_TO_PUBLIC) && defined(PCSX2_DEBUG) if( g_bSaveTrans ) { tex0Info t; t.tbp0 = gs.dstbuf.bp; @@ -3244,7 +3244,7 @@ void TransferLocalLocal() g_MemTargs.ClearRange(dststart, dstend); -#if !defined(RELEASE_TO_PUBLIC) && defined(_DEBUG) +#if !defined(RELEASE_TO_PUBLIC) && defined(PCSX2_DEBUG) if( g_bSaveTrans ) { tex0Info t; t.tbp0 = gs.dstbuf.bp; diff --git a/plugins/zerogs/dx/zerogs.cpp b/plugins/zerogs/dx/zerogs.cpp index af7ffadfa6..989bff2caf 100644 --- a/plugins/zerogs/dx/zerogs.cpp +++ b/plugins/zerogs/dx/zerogs.cpp @@ -598,7 +598,7 @@ void ZeroGS::VB::CheckFrame(int tbp) //u32 key = zbuf.zbp|(frame.fbw<<16); //CRenderTargetMngr::MAPTARGETS::iterator it = s_RTs.mapTargets.find(key); // if( it != s_RTs.mapTargets.end() ) { -//#ifdef _DEBUG +//#ifdef PCSX2_DEBUG // DEBUG_LOG("zbuf resolve\n"); //#endif // if( it->second->status & CRenderTarget::TS_Resolved ) @@ -856,7 +856,7 @@ HRESULT ZeroGS::Create(LONG _width, LONG _height) UINT adapter = D3DADAPTER_DEFAULT; D3DDEVTYPE devtype = !DEBUG_PS2 ? D3DDEVTYPE_HAL : D3DDEVTYPE_REF; -#ifndef _DEBUG +#ifndef PCSX2_DEBUG DWORD hwoptions = D3DCREATE_HARDWARE_VERTEXPROCESSING|D3DCREATE_PUREDEVICE; #else DWORD hwoptions = D3DCREATE_HARDWARE_VERTEXPROCESSING; @@ -1906,7 +1906,7 @@ void ZeroGS::Flush(int context) return; } -#if defined(PRIM_LOG) && defined(_DEBUG) +#if defined(PRIM_LOG) && defined(PCSX2_DEBUG) static const char* patst[8] = { "NEVER", "ALWAYS", "LESS", "LEQUAL", "EQUAL", "GEQUAL", "GREATER", "NOTEQUAL"}; static const char* pztst[4] = { "NEVER", "ALWAYS", "GEQUAL", "GREATER" }; static const char* pafail[4] = { "KEEP", "FB_ONLY", "ZB_ONLY", "RGB_ONLY" }; @@ -2067,7 +2067,7 @@ void ZeroGS::Flush(int context) else ptextarg = NULL; } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( g_bSaveFlushedFrame & 0x80000000 ) { char str[255]; sprintf(str, "rndr.tga", g_SaveFrameNum); @@ -2393,7 +2393,7 @@ void ZeroGS::Flush(int context) } else { // save the texture -#ifdef _DEBUG +#ifdef PCSX2_DEBUG // CMemoryTarget* pmemtarg = g_MemTargs.GetMemoryTarget(curvb.tex0, 0); // assert( curvb.pmemtarg == pmemtarg ); // if( PSMT_ISCLUT(curvb.tex0.psm) ) @@ -2546,7 +2546,7 @@ void ZeroGS::Flush(int context) } } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( bDestAlphaColor == 1 ) { WARN_LOG("dest alpha blending! manipulate alpha here\n"); } @@ -2719,7 +2719,7 @@ void ZeroGS::Flush(int context) //WARN_LOG("Need to reset dest alpha color\n"); } -#ifdef _DEBUG +#ifdef PCSX2_DEBUG if( g_bSaveFlushedFrame & 0xf ) { char str[255]; sprintf(str, "frames\\frame%.4d.jpg", g_SaveFrameNum++); diff --git a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj index 48f0aaf2d6..1080fb0720 100644 --- a/plugins/zeropad/Windows/ZeroPAD_2008.vcproj +++ b/plugins/zeropad/Windows/ZeroPAD_2008.vcproj @@ -19,7 +19,7 @@ @@ -113,12 +108,7 @@ /> @@ -184,13 +174,7 @@ /> - - - - + + + + + + + + + + - - - - - - @@ -114,15 +110,7 @@ /> @@ -192,14 +180,8 @@ /> diff --git a/plugins/zerospu2/misc.h b/plugins/zerospu2/misc.h index b1bf3b8315..5000683800 100644 --- a/plugins/zerospu2/misc.h +++ b/plugins/zerospu2/misc.h @@ -15,8 +15,10 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ - - #ifdef __LINUX__ + +#pragma once + +#ifdef __LINUX__ #include #include #include // ftime(), struct timeb @@ -87,61 +89,3 @@ static __forceinline void pcsx2_aligned_free(void* pmem) #define _aligned_malloc pcsx2_aligned_malloc #define _aligned_free pcsx2_aligned_free #endif - -// Atomic Operations -#if defined (_WIN32) - -#ifndef __x86_64__ -extern "C" LONG __cdecl _InterlockedExchangeAdd(LPLONG volatile Addend, LONG Value); -#endif - -#pragma intrinsic (_InterlockedExchangeAdd) -#define InterlockedExchangeAdd _InterlockedExchangeAdd - -#else - -//typedef void* PVOID; - -static __forceinline long InterlockedExchange(volatile long* Target, long Value) -{ - long result; - /* - * The XCHG instruction always locks the bus with or without the - * LOCKED prefix. This makes it significantly slower than CMPXCHG on - * uni-processor machines. The Windows InterlockedExchange function - * is nearly 3 times faster than the XCHG instruction, so this routine - * is not yet very useful for speeding up pthreads. - */ - - __asm__ __volatile__ ( - "xchgl %2,%1" - :"=r" (result) - :"m" (*Target), "0" (Value)); - - return result; -} - -static __forceinline long InterlockedExchangeAdd(volatile long* Addend, long Value) -{ - __asm__ __volatile__( - ".intel_syntax\n" - "lock xadd [%0], %%eax\n" - ".att_syntax\n" : : "r"(Addend), "a"(Value) : "memory"); -} - -static __forceinline long InterlockedCompareExchange(volatile long *dest, long value, long comp) -{ - long result; - - __asm__ __volatile__ ( - "lock\n\t" - "cmpxchgl %2,%1" /* if (EAX == [location]) */ - /* [location] = value */ - /* else */ - /* EAX = [location] */ - :"=a" (result) - :"m" (*dest), "r" (value), "a" (comp)); - - return result; -} -#endif diff --git a/plugins/zerospu2/zerospu2.cpp b/plugins/zerospu2/zerospu2.cpp index a0b5351572..ead28339a7 100644 --- a/plugins/zerospu2/zerospu2.cpp +++ b/plugins/zerospu2/zerospu2.cpp @@ -126,7 +126,7 @@ static void InitLibraryName() // subversion revision tags: strcpy( libraryName, "ZeroSPU2" -# ifdef _DEBUG +# ifdef PCSX2_DEBUG "-Debug" # elseif ZEROSPU2_DEVBUILD "-Dev" @@ -138,7 +138,7 @@ static void InitLibraryName() // to label the specific revision: sprintf_s( libraryName, "ZeroSPU2 r%d%s" -# ifdef _DEBUG +# ifdef PCSX2_DEBUG "-Debug" # elif ZEROSPU2_DEVBUILD "-Dev" @@ -151,7 +151,7 @@ static void InitLibraryName() // I'll hook in svn version code later. --arcum42 strcpy( libraryName, "ZeroSPU2 Playground" -# ifdef _DEBUG +# ifdef PCSX2_DEBUG "-Debug" # elif ZEROSPU2_DEVBUILD "-Dev" diff --git a/plugins/zerospu2/zerospu2.h b/plugins/zerospu2/zerospu2.h index 37698ddf1c..d68142c779 100644 --- a/plugins/zerospu2/zerospu2.h +++ b/plugins/zerospu2/zerospu2.h @@ -23,10 +23,8 @@ #include #include -extern "C" { #define SPU2defs #include "PS2Edefs.h" -} #include "reg.h" #include "misc.h" From 816e39851a8369d2112f766b801c481d563d2265 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 18 Jun 2009 09:32:35 +0000 Subject: [PATCH 243/296] Linux: Quick fixes to get the Linux build working with the changes in r1388. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1389 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/intrin_x86.h | 6 ++++++ plugins/zerospu2/misc.h | 2 ++ 2 files changed, 8 insertions(+) diff --git a/common/include/intrin_x86.h b/common/include/intrin_x86.h index a47b68ae35..7c04b795e0 100644 --- a/common/include/intrin_x86.h +++ b/common/include/intrin_x86.h @@ -65,6 +65,12 @@ would use in the same case */ + +#ifndef __SIZE_T +#define __SIZE_T +typedef unsigned int size_t; +#endif + /*** Stack frame juggling ***/ #define _ReturnAddress() (__builtin_return_address(0)) #define _AddressOfReturnAddress() (&(((void **)(__builtin_frame_address(0)))[1])) diff --git a/plugins/zerospu2/misc.h b/plugins/zerospu2/misc.h index 5000683800..dab6876483 100644 --- a/plugins/zerospu2/misc.h +++ b/plugins/zerospu2/misc.h @@ -40,6 +40,8 @@ inline u32 timeGetTime() #include +#define InterlockedExchangeAdd _InterlockedExchangeAdd + #else #include #include From 91be7889e2ed5622f4b66c2fd499a8d92ee21f66 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Thu, 18 Jun 2009 09:58:56 +0000 Subject: [PATCH 244/296] Suppose I should do something about the configure.ac files and _DEBUG, as well. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1390 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/configure.ac | 6 +++--- plugins/CDVDiso/src/configure.ac | 2 +- plugins/CDVDnull/configure.ac | 2 +- plugins/GSnull/configure.ac | 2 +- plugins/PadNull/configure.ac | 2 +- plugins/SPU2null/SPU2.cpp | 2 +- plugins/SPU2null/configure.ac | 2 +- plugins/USBnull/configure.ac | 2 +- plugins/dev9null/configure.ac | 2 +- plugins/spu2-x/src/configure.ac | 2 +- plugins/zeropad/configure.ac | 2 +- plugins/zerospu2/configure.ac | 2 +- 12 files changed, 14 insertions(+), 14 deletions(-) diff --git a/pcsx2/configure.ac b/pcsx2/configure.ac index 7b7a745f92..7fdcf65c03 100644 --- a/pcsx2/configure.ac +++ b/pcsx2/configure.ac @@ -38,7 +38,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), if test "x$debug" == "xyes" then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) AC_DEFINE(PCSX2_DEVBUILD,1,[PCSX2_DEVBUILD]) MYOBJDIR="Debug" optimization="no" @@ -112,7 +112,7 @@ then fi WARNING_FLAGS+=" -Wno-format -Wno-unused-parameter -Wno-unused-value " -NORMAL_FLAGS+=" -msse -msse2 ${WARNING_FLAGS} " +NORMAL_FLAGS+=" -msse -msse2 ${WARNING_FLAGS} -fno-strict-aliasing " dnl These optimizations seem to cause issues with GCC 4.3.3, so we'll turn them off. dnl I really need to run some tests and see if they are still needed. @@ -146,7 +146,7 @@ dnl Anything higher then -O0 seem to cause issues with GCC 4.4.0. NORMAL_FLAGS+=" -fsched-spec " dnl NORMAL_FLAGS+=" -fschedule-insns " NORMAL_FLAGS+=" -fschedule-insns2 " - NORMAL_FLAGS+=" -fstrict-aliasing -fstrict-overflow " + NORMAL_FLAGS+=" -fstrict-overflow " NORMAL_FLAGS+=" -ftree-switch-conversion -ftree-pre -ftree-vrp " else NORMAL_FLAGS+=" -pipe -O2 " diff --git a/plugins/CDVDiso/src/configure.ac b/plugins/CDVDiso/src/configure.ac index d42c8b2865..dc83464151 100755 --- a/plugins/CDVDiso/src/configure.ac +++ b/plugins/CDVDiso/src/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+=" -g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+=" -g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+=" -g -fPIC -Wall -Wno-unused-value -fpermissive -m32 " diff --git a/plugins/CDVDnull/configure.ac b/plugins/CDVDnull/configure.ac index 9075e3be14..de015970df 100644 --- a/plugins/CDVDnull/configure.ac +++ b/plugins/CDVDnull/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/GSnull/configure.ac b/plugins/GSnull/configure.ac index 8973091c0e..bf15cfa1c2 100644 --- a/plugins/GSnull/configure.ac +++ b/plugins/GSnull/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/PadNull/configure.ac b/plugins/PadNull/configure.ac index e8617e98dd..1fe5f6375c 100644 --- a/plugins/PadNull/configure.ac +++ b/plugins/PadNull/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/SPU2null/SPU2.cpp b/plugins/SPU2null/SPU2.cpp index 720be35249..5b2033ca7c 100644 --- a/plugins/SPU2null/SPU2.cpp +++ b/plugins/SPU2null/SPU2.cpp @@ -35,7 +35,7 @@ const u32 minor = 0; // increase that with each version #define SUSTAIN_MS 441L #define RELEASE_MS 437L -#ifdef _DEBUG +#ifdef PCSX2_DEBUG char *libraryName = "SPU2null (Debug)"; #else char *libraryName = "SPU2null "; diff --git a/plugins/SPU2null/configure.ac b/plugins/SPU2null/configure.ac index cee047ebe0..67a5de0468 100644 --- a/plugins/SPU2null/configure.ac +++ b/plugins/SPU2null/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/USBnull/configure.ac b/plugins/USBnull/configure.ac index a2177049ab..f754b11bcf 100644 --- a/plugins/USBnull/configure.ac +++ b/plugins/USBnull/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/dev9null/configure.ac b/plugins/dev9null/configure.ac index 7b9d8dbaef..31d42bfe74 100644 --- a/plugins/dev9null/configure.ac +++ b/plugins/dev9null/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/spu2-x/src/configure.ac b/plugins/spu2-x/src/configure.ac index feacc310e1..edb2e34e3a 100644 --- a/plugins/spu2-x/src/configure.ac +++ b/plugins/spu2-x/src/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " diff --git a/plugins/zeropad/configure.ac b/plugins/zeropad/configure.ac index 89e5c48c0a..5b067f0532 100644 --- a/plugins/zeropad/configure.ac +++ b/plugins/zeropad/configure.ac @@ -28,7 +28,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -m32 -fpic " CXXFLAGS+="-g -m32 -fpic " CCASFLAGS+=" -m32 -fpic " diff --git a/plugins/zerospu2/configure.ac b/plugins/zerospu2/configure.ac index 613f9be657..119f173f57 100644 --- a/plugins/zerospu2/configure.ac +++ b/plugins/zerospu2/configure.ac @@ -29,7 +29,7 @@ AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [debug build]), debug=$enableval,debug=no) if test "x$debug" == xyes then - AC_DEFINE(_DEBUG,1,[_DEBUG]) + AC_DEFINE(PCSX2_DEBUG,1,[PCSX2_DEBUG]) CFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CPPFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " CXXFLAGS+="-g -fPIC -Wall -Wno-unused-value -m32 " From 81293ede5d8a0233d19358d2319335e91b157933 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 13:00:21 +0000 Subject: [PATCH 245/296] Fixed bug which stopped Zone of Enders booting, introduced in r1385 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1391 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index f79dab837e..21fa1fa719 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2328,14 +2328,13 @@ int VIF1transfer(u32 *data, int size, int istag) if(vif1Regs->stat & VIF1_STAT_VGW) { vif1.vifstalled = true; - return -1; } if (vif1ch->qwc == 0 && (vif1.irqoffset == 0 || istag == 1)) vif1.inprogress = 0; - return 0; + return vif1.vifstalled ? -2 : 0; } void vif1TransferFromMemory() From e984c31b7b7f305762349245cf0c6cf9b69d2e18 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 13:52:26 +0000 Subject: [PATCH 246/296] Fixed small unpacking issue with Shox, hopefully rectified the games which stopped working in r1289 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1392 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 21fa1fa719..8bf107fae2 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -391,14 +391,11 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int VIFUNPACK_LOG("Aligning packet size = %d offset %d addr %x", size, vifRegs->offset, vif->tag.addr); - if(((size / ft->dsize) + vifRegs->offset) < (u32)ft->qsize) - VIFUNPACK_LOG("Warning! Size needed to align %x size chunks available %x offset %x", ft->qsize - ((size / ft->dsize) + vifRegs->offset), vifRegs->offset); - if (((u32)size / (u32)ft->dsize) < ((u32)ft->qsize - vifRegs->offset)) { DevCon::Error("Wasn't enough left size/dsize = %x left to write %x", params(size / ft->dsize), (ft->qsize - vifRegs->offset)); } - unpacksize = min(((u32)size / (u32)ft->dsize), ((u32)ft->qsize - vifRegs->offset)); + unpacksize = min((size / ft->dsize), (ft->qsize - vifRegs->offset)); VIFUNPACK_LOG("Increasing dest by %x from offset %x", (4 - ft->qsize) + unpacksize, vifRegs->offset); @@ -406,8 +403,17 @@ static int VIFalign(u32 *data, vifCode *v, unsigned int size, const unsigned int func(dest, (u32*)cdata, unpacksize); size -= unpacksize * ft->dsize; - vifRegs->num--; - ++vif->cl; + if(vifRegs->offset == 0) + { + vifRegs->num--; + ++vif->cl; + } + else + { + DevCon::Notice("Offset = %x", params vifRegs->offset); + vif->tag.addr += unpacksize * 4; + return size>>2; + } if (vif->cl == vifRegs->cycle.wl) { @@ -2593,7 +2599,11 @@ __forceinline void vif1Interrupt() vif1ch->chcr &= ~0x100; g_vifCycles = 0; hwDmacIrq(DMAC_VIF1); - if(vif1ch->chcr & 0x1)vif1Regs->stat &= ~0x1F000000; // FQC=0 + + //Im not totally sure why Path3 Masking makes it want to see stuff in the fifo + //Games effected by setting, Fatal Frame, KH2, Shox, Crash N Burn, GT3/4 possibly + //Im guessing due to the full gs fifo before the reverse? (Refraction) + if(!vif1Regs->mskpath3)vif1Regs->stat &= ~0x1F000000; // FQC=0 } void dmaVIF1() From 57bdc0e5ec049f54b4a525bfbf08c527bfd09cbb Mon Sep 17 00:00:00 2001 From: gabest11 Date: Thu, 18 Jun 2009 16:23:54 +0000 Subject: [PATCH 247/296] Added a setup project, main advantage that it can install vc_redist and other dlls. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1393 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/windows/VCprojects/pcsx2_2008.vcproj | 4 + pcsx2/windows/WinSysExec.cpp | 7 +- pcsx2/windows/pcsx2.rc | 1 + pcsx2/windows/resource.h | 3 +- pcsx2/windows/uninstall.ico | Bin 0 -> 4710 bytes pcsx2_suite_2008.sln | 7 + setup/setup.vdproj | 2906 ++++++++++++++++++++ 7 files changed, 2926 insertions(+), 2 deletions(-) create mode 100644 pcsx2/windows/uninstall.ico create mode 100644 setup/setup.vdproj diff --git a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj index 8e49d65547..9b9f3a4457 100644 --- a/pcsx2/windows/VCprojects/pcsx2_2008.vcproj +++ b/pcsx2/windows/VCprojects/pcsx2_2008.vcproj @@ -899,6 +899,10 @@ RelativePath="..\ps2_silver.bmp" > + + WWF@zw3m`^mVbiLn21<)TNJf;ld5EF zdzz+zWA0J?c1?j&Ihq*ud>Zl5SQZ~I^4K~dUgT^zOyq)H(21Ps6D7|2^}`ZpN-n;@ zr3+JY)PL~VctOp@Ymz6nthBF)*AxUk`sDCyG~s85F@g#rPbDvM(k0F*wZz+Hg5Xl5WZk`l-LOtqqRW zRyaF4;N^MnHXDX--^OTvKiKFfgn~UJZxqY-OD+2 z=_@hD52B~K3a#f0X^tK^tj%aE$;Uu@3vL_A(SM^3eO2^49gP^j)eMjELwKt{0W)X8 zWz52a`BMzmp2ld!F1RlL6Rs~(F;=+;W4bg<==Z^G$iPt5M_~1bP+ydRiKep{xpofy zXV&9R@g@wGZ3BNk0YeuPAzXeR-m7_V-z)=Pm4@+#Z1jBb5xh4vm}t8IPxlw_jP!vw zron5a`)w}<>nH*1EC;(|!Ek>EhHEpydM<#u?ts}o2eVgz-MU22b{P%YLm2D12-a7H z@$OQvejSZ7fH{otI?QkncVfUwIR~zwzA_(P!45WP0voyt?{Et|P9w!N@Q(Gu)u)Ar zCvCV6V*@&jJF98!8{r+KVpQ2LWun8J*Tj;#a{YnlR|7 zp}s9JRGk9zbQ5k8+E7$@2+Z3`zcnj(-U7S58WziW;+-fj&m-SMu%V#f-{fx-da-~1 zZt{1qdv_xF0Rr6v8zPJlXl^!!4I84ce*HR%C$Mwp4)PxK_1#8X+-8csaI=1Bb%o@I zFd(#GjPeb*yD&oeZVp&*sk;PqUDX(HIdHc1Bre%5LS;(A;if}C7JYjT2O#)FFqRcB z!>aSE@aCmA5vPxX%B(_qV>+z76+>ggxasOeC*6~FoJX5?99KLp^wazr`mL1^w1g@` zJ>dr7OM;EyBy`a`-%FUFc#Pg_5BbX@40*Lp_;SdFV-AKrHilh|4%G1sc9$2&tpa)k z27ZEJ#LbY~Iu0}SYjS!}(K8Nvl=Wag~g~+q-U!XetwWo;nfcypFa=sD<+lP zWNMdZW;^9>KQHwWzq3JvWrV5oNJVGY{Y^z@5h--`NRF-)&Nz=q6FIKa`tCD4(>`5F$0y1)t1Mqh(ZwB7UmET&QZJOug<15)le2buLkOXSSp)?^{!PJ2 zxm*M{f!;v{Q_-Z&2)*|`_lIbz2PF5k3%!R@+qd<^xAnxg^~C>uJs}>!-?(`sqf+I* znLVDPY8AIrrQUzY?f&d*T|7rpxGnE&+!*o4=ic7F_do8Euk?sa*}3JNw?mhHcaHxH zZ^vT=@^a*I>~bEXLBT z^z4skWQhFhs~-df&Yw4T&J~lHDy5pL%|$V*;In? ze-xL>WOEfuA89zc{@~TR#zuoq_fEK2y@?jNdSB7Gb0H5XW(DNdJdrF`b|zECy46z; zurBUwVPR3wgZ{JSkJr@d_sOW#BYoYJ9=%I*>P+G6-!7OdpJOqZ8ydtV%GkKkS5Epx zY(03K{w?1`NyY6XK$4_W9rz%|iPd*{dUSrBgPMNF*BQmqI zvp>x{c3g9IWMpyh61~xQwWu~@UuODgDM3Vz&py{y*zNXThWwfqqx4cnY0l&hP3cV? zH##~x^czl{Ia5?#UV84Uy0x|C`sEK7Km6lhS?Q7EV$MUp{4rS+f09~xL`kbCo4;VO z%oVvZCNn1|cX9zEA|j&xG%-T6kJv|o7Tp?|_}hywt>mV4X_fgdpDpU?>dJlbrB_z2 zjGh|r(|LZhU9uMIHc<`>dtoxaq>C4xr~k68t)nkXa`<1qs{6X!9K9Gu7QW)-^uT)p zo+nBje2sMW-!-aP+*CfesFa|Zf@&8i3892Q5$*RvtYv0tXFx^qBoyTDLAzL+Ou Date: Thu, 18 Jun 2009 17:19:01 +0000 Subject: [PATCH 248/296] MFIFO fixes for Project Altered Beast - Issue 155, Tekken Tag should be better now too (second time lucky!) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1394 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 4 ++-- pcsx2/SPR.cpp | 4 ++-- pcsx2/Vif.cpp | 52 ++++++++++++++++++++++++++++++++++-------------- pcsx2/VifDma.cpp | 4 ++-- 4 files changed, 43 insertions(+), 21 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index c11762d302..8b078d89b9 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -464,8 +464,8 @@ void mfifoGIFtransfer(int qwc) { if (gif->qwc == 0) { if (gif->tadr == spr0->madr) { //if( gifqwc > 1 ) DevCon::WriteLn("gif mfifo tadr==madr but qwc = %d", params gifqwc); - //hwDmacIrq(14); - + hwDmacIrq(14); + gifstate |= GIF_STATE_EMPTY; return; } gif->tadr = psHu32(DMAC_RBOR) + (gif->tadr & psHu32(DMAC_RBSR)); diff --git a/pcsx2/SPR.cpp b/pcsx2/SPR.cpp index 702cdb3306..6b8a1973b9 100644 --- a/pcsx2/SPR.cpp +++ b/pcsx2/SPR.cpp @@ -227,7 +227,7 @@ void SPRFROMinterrupt() //Console::WriteLn("mfifoGIFtransfer %x madr %x, tadr %x", params gif->chcr, gif->madr, gif->tadr); mfifoGIFtransfer(mfifotransferred); mfifotransferred = 0; - return; + if(gif->chcr & 0x100)return; } else if ((psHu32(DMAC_CTRL) & 0xC) == 0x8) // VIF1 MFIFO { @@ -236,7 +236,7 @@ void SPRFROMinterrupt() //Console::WriteLn("mfifoVIF1transfer %x madr %x, tadr %x", params vif1ch->chcr, vif1ch->madr, vif1ch->tadr); mfifoVIF1transfer(mfifotransferred); mfifotransferred = 0; - return; + if(vif1ch->chcr & 0x100)return; } } if (spr0finished == 0) return; diff --git a/pcsx2/Vif.cpp b/pcsx2/Vif.cpp index a3c7cfc4fa..76f466303c 100644 --- a/pcsx2/Vif.cpp +++ b/pcsx2/Vif.cpp @@ -397,7 +397,7 @@ static __forceinline int mfifo_VIF1chain() /* Is QWC = 0? if so there is nothing to transfer */ if ((vif1ch->qwc == 0) && (!vif1.vifstalled)) { - vif1.inprogress = 0; + vif1.inprogress &= ~1; return 0; } @@ -451,7 +451,7 @@ void mfifoVIF1transfer(int qwc) mfifodmairq = false; //Clear any previous TIE interrupt - if (vif1ch->qwc == 0) + if (vif1ch->qwc == 0 && vifqwc > 0) { ptag = (u32*)dmaGetAddr(vif1ch->tadr); @@ -550,7 +550,9 @@ void vifMFIFOInterrupt() hwDmacIrq(DMAC_FROM_SPR); } - if (vif1.inprogress == 1) mfifo_VIF1chain(); + + + if (vif1.irq && vif1.tag.size == 0) { @@ -565,16 +567,36 @@ void vifMFIFOInterrupt() } } - if (!vif1.done || vif1.inprogress & 1) + if (vif1.done == false || vif1ch->qwc) { - if (vifqwc <= 0) + + + switch(vif1.inprogress & 1) { - //Console::WriteLn("Empty 1"); - vif1.inprogress |= 0x10; - vif1Regs->stat &= ~0x1F000000; // FQC=0 - hwDmacIrq(14); - return; + case 0: //Set up transfer + if (vif1ch->tadr == spr0->madr) + { + // Console::WriteLn("Empty 1"); + vifqwc = 0; + vif1.inprogress |= 0x10; + vif1Regs->stat &= ~0x1F000000; // FQC=0 + hwDmacIrq(14); + return; + } + + mfifoVIF1transfer(0); + if (vif1ch->madr >= psHu32(DMAC_RBOR) && vif1ch->madr <= (psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR))) + CPU_INT(10, 0); + else + CPU_INT(10, vif1ch->qwc * BIAS); + + return; + case 1: //Transfer data + mfifo_VIF1chain(); + CPU_INT(10, 0); + return; } + if (!(vif1.inprogress & 0x1)) mfifoVIF1transfer(0); if (vif1ch->madr >= psHu32(DMAC_RBOR) && vif1ch->madr <= (psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR))) @@ -582,16 +604,16 @@ void vifMFIFOInterrupt() else CPU_INT(10, vif1ch->qwc * BIAS); - return; - } - else if (vifqwc <= 0) + } + + /*if (vifqwc <= 0) { //Console::WriteLn("Empty 2"); - vif1.inprogress |= 0x10; + //vif1.inprogress |= 0x10; vif1Regs->stat &= ~0x1F000000; // FQC=0 hwDmacIrq(14); - } + }*/ //On a TIE break we do not clear the MFIFO (Art of Fighting) //If we dont clear it on MFIFO end, Tekken Tag breaks, understandably (Refraction) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 8bf107fae2..aecf054914 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2307,7 +2307,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (((vif1Regs->code >> 24) & 0x7f) != 0x7) vif1Regs->stat |= VIF1_STAT_VIS; // Note: commenting this out fixes WALL-E if (vif1ch->qwc == 0 && (vif1.irqoffset == 0 || istag == 1)) - vif1.inprogress = 0; + vif1.inprogress &= ~0x1; // spiderman doesn't break on qw boundaries if (istag) return -2; @@ -2338,7 +2338,7 @@ int VIF1transfer(u32 *data, int size, int istag) if (vif1ch->qwc == 0 && (vif1.irqoffset == 0 || istag == 1)) - vif1.inprogress = 0; + vif1.inprogress &= ~0x1; return vif1.vifstalled ? -2 : 0; } From a2dc609822d388713873cecdfb8b64ae1bbfd8d2 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 19:24:39 +0000 Subject: [PATCH 249/296] Removed intermittant mode from GIF MFIFO git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1395 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 8b078d89b9..15765e0fcb 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -373,8 +373,7 @@ void dmaGIF() { // called from only one location, so forceinline it: static __forceinline int mfifoGIFrbTransfer() { - u32 qwc = (psHu32(GIF_MODE) & 0x4 || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; - int mfifoqwc = min(gifqwc, qwc); + u32 mfifoqwc = min(gifqwc, (u32)gif->qwc); u32 *src; /* Check if the transfer should wrap around the ring buffer */ @@ -432,11 +431,11 @@ static __forceinline int mfifoGIFchain() { } else { - int mfifoqwc = (psHu32(GIF_MODE) & 0x4 || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : gif->qwc; + int mfifoqwc = gif->qwc; u32 *pMem = (u32*)dmaGetAddr(gif->madr); if (pMem == NULL) return -1; - WRITERING_DMA(pMem, mfifoqwc); + mfifoqwc = WRITERING_DMA(pMem, mfifoqwc); //gif->madr += mfifoqwc*16; //gif->qwc -= mfifoqwc; mfifocycles += (mfifoqwc) * 2; /* guessing */ From b50fcc90d13ef5215f0bf8e9dd0310960d7af2d1 Mon Sep 17 00:00:00 2001 From: refraction Date: Thu, 18 Jun 2009 19:50:53 +0000 Subject: [PATCH 250/296] Fix for SSX3 not booting. VIF FIFO now clears when not in Reverse mode, regardless of MSKPATH3 git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1396 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index aecf054914..de78c47275 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2603,7 +2603,8 @@ __forceinline void vif1Interrupt() //Im not totally sure why Path3 Masking makes it want to see stuff in the fifo //Games effected by setting, Fatal Frame, KH2, Shox, Crash N Burn, GT3/4 possibly //Im guessing due to the full gs fifo before the reverse? (Refraction) - if(!vif1Regs->mskpath3)vif1Regs->stat &= ~0x1F000000; // FQC=0 + //Note also this is only the condition for reverse fifo mode, normal direction clears it as normal + if(!vif1Regs->mskpath3 || (vif1ch->chcr & 0x1))vif1Regs->stat &= ~0x1F000000; // FQC=0 } void dmaVIF1() From bc1b84295fec9d511b1fc915a986e49c801faf65 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Thu, 18 Jun 2009 20:44:20 +0000 Subject: [PATCH 251/296] Improvements to the new installer project: * installer automatically includes all three SSE build targets for GSdx (each built target needs to be built manually prior to building the installer). * Fixed some typos and abbreviated install paths from "PCSX2 Team/PCSX2" to just "PCSX2". * Added an "SSEtype" user define to GSdx's property sheets, which is used to generate the SSE postfix on the DLLs automatically. (ie, GSdx-SSE2.dll, etc). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1397 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2_suite_2008.sln | 12 +- plugins/GSdx/GSdx_vs2008.vcproj | 515 +-- plugins/GSdx/vsprops/sse2.vsprops | 6 +- plugins/GSdx/vsprops/sse4.vsprops | 4 + plugins/GSdx/vsprops/ssse3.vsprops | 6 +- setup/setup.vdproj | 5532 ++++++++++++++++++++++++++-- 6 files changed, 5270 insertions(+), 805 deletions(-) diff --git a/pcsx2_suite_2008.sln b/pcsx2_suite_2008.sln index 94eb641202..5a307a8885 100644 --- a/pcsx2_suite_2008.sln +++ b/pcsx2_suite_2008.sln @@ -33,7 +33,7 @@ Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "bzip2", "3rdparty\bzip2\bzi EndProject Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "LilyPad", "plugins\LilyPad\LilyPad_VC2008.vcproj", "{E4081455-398C-4610-A87C-90A8A7D72DC3}" EndProject -Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "pthreads", "3rdparty\w32pthreads\pthreads_2008.vcproj", "{26511268-2902-4997-8421-ECD7055F9E28}" +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "w32pthreads", "3rdparty\w32pthreads\pthreads_2008.vcproj", "{26511268-2902-4997-8421-ECD7055F9E28}" EndProject Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "ZeroSPU2", "plugins\zerospu2\Windows\ZeroSPU2_2008.vcproj", "{7F059854-568D-4E08-9D00-1E78E203E4DC}" ProjectSection(ProjectDependencies) = postProject @@ -273,11 +273,11 @@ Global {CDD9DB83-3BD9-4ED8-BB83-399A2F65F022}.Release SSE4|Win32.Build.0 = Release|Win32 {CDD9DB83-3BD9-4ED8-BB83-399A2F65F022}.Release SSSE3|Win32.ActiveCfg = Release|Win32 {CDD9DB83-3BD9-4ED8-BB83-399A2F65F022}.Release SSSE3|Win32.Build.0 = Release|Win32 - {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Debug|Win32.ActiveCfg = Release SSE2 - {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Devel|Win32.ActiveCfg = Release SSE2 - {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSE2|Win32.ActiveCfg = Release SSE2 - {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSE4|Win32.ActiveCfg = Release SSE4 - {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSSE3|Win32.ActiveCfg = Release SSSE3 + {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Debug|Win32.ActiveCfg = Release + {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Devel|Win32.ActiveCfg = Release + {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSE2|Win32.ActiveCfg = Release + {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSE4|Win32.ActiveCfg = Release + {14259DC9-2A12-4B05-A7D7-23C1A9FFEA03}.Release SSSE3|Win32.ActiveCfg = Release EndGlobalSection GlobalSection(SolutionProperties) = preSolution HideSolutionNode = FALSE diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 946f78609d..8b5f6a7d37 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -19,131 +19,6 @@ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - @@ -244,6 +120,7 @@ /> @@ -307,6 +184,7 @@ /> @@ -369,6 +247,7 @@ /> @@ -431,6 +310,7 @@ /> @@ -494,6 +374,7 @@ /> @@ -1490,22 +1371,6 @@ - - - - - - @@ -1998,22 +1863,6 @@ - - - - - - @@ -2134,22 +1983,6 @@ - - - - - - @@ -2270,22 +2103,6 @@ - - - - - - @@ -2406,22 +2223,6 @@ - - - - - - @@ -2542,22 +2343,6 @@ - - - - - - @@ -2678,22 +2463,6 @@ - - - - - - @@ -2830,22 +2599,6 @@ - - - - - - @@ -2966,22 +2719,6 @@ - - - - - - @@ -3102,22 +2839,6 @@ - - - - - - @@ -3238,22 +2959,6 @@ - - - - - - @@ -3374,22 +3079,6 @@ - - - - - - @@ -3514,22 +3203,6 @@ - - - - - - @@ -3650,22 +3323,6 @@ - - - - - - @@ -3786,22 +3443,6 @@ - - - - - - @@ -3922,22 +3563,6 @@ - - - - - - @@ -4062,22 +3687,6 @@ - - - - - - @@ -4198,22 +3807,6 @@ - - - - - - @@ -4334,22 +3927,6 @@ - - - - - - @@ -4470,22 +4047,6 @@ - - - - - - @@ -4606,22 +4167,6 @@ - - - - - - @@ -4742,22 +4287,6 @@ - - - - - - @@ -4878,22 +4407,6 @@ - - - - - - @@ -5014,22 +4527,6 @@ - - - - - - diff --git a/plugins/GSdx/vsprops/sse2.vsprops b/plugins/GSdx/vsprops/sse2.vsprops index 1ab36cb277..9211140cc0 100644 --- a/plugins/GSdx/vsprops/sse2.vsprops +++ b/plugins/GSdx/vsprops/sse2.vsprops @@ -6,7 +6,11 @@ > + diff --git a/plugins/GSdx/vsprops/sse4.vsprops b/plugins/GSdx/vsprops/sse4.vsprops index 601b39f585..da3ee6c270 100644 --- a/plugins/GSdx/vsprops/sse4.vsprops +++ b/plugins/GSdx/vsprops/sse4.vsprops @@ -9,4 +9,8 @@ PreprocessorDefinitions="_M_SSE=0x401" EnableEnhancedInstructionSet="2" /> + diff --git a/plugins/GSdx/vsprops/ssse3.vsprops b/plugins/GSdx/vsprops/ssse3.vsprops index ac7fa1a064..fc76dac7f5 100644 --- a/plugins/GSdx/vsprops/ssse3.vsprops +++ b/plugins/GSdx/vsprops/ssse3.vsprops @@ -6,7 +6,11 @@ > + diff --git a/setup/setup.vdproj b/setup/setup.vdproj index a943a5d8ed..db6139c1d3 100644 --- a/setup/setup.vdproj +++ b/setup/setup.vdproj @@ -15,8 +15,86 @@ { "Entry" { - "MsmKey" = "8:_0950432B04CD787CC7B938905C6C4399" - "OwnerKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "MsmKey" = "8:_008B1BB47C0A46C0B025FDE5376533F4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_00ABD03A61124E9F8BEC5E4D2D7F2715" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_03E75E8822CE48BFB4D8378A4FD3A9E8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_0435A4202F22485295FC9B14FF9E0DEF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_048B9870FB3E4278ADCB974F47D3AA3D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_069D4ED441804E6F91F93A1BA01C21B8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_072AB1B5BAC84646894F39D3E1BF092F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_072B429763E648A2AFA3744F69B35FD8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_07B925C7C3DD4023870E4C350A933F01" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_086724BEF26C4B91964547A4ABCFC33D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_094A3551F77548E28331D27415415CD4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_0953A3DB125D4969AA9CF0EAEE0174F8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_0A01FDD1F6A54A43A900C9AC92D4B2D9" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_0AA27B6457FC4913924E4F391F4581DC" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -39,12 +117,78 @@ } "Entry" { + "MsmKey" = "8:_0D6C5B8003914BC8B1F4C3EC09C58AB2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_0E8BBD04D986469EA64F1A82EC4DB8FA" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_1037D220B29040FDA7B60A0A9DA4865E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_110E08BF1DA04EF69BF2D475585D0500" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_1320398166C4413FB4BF04DC2C1FD8EF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_1430E155AE5F4B92A78B32E05FEAF3EF" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_14A843D75DC34B599F3463074612476E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_164C7824F6134DF3A56AF21FB50AE125" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_1750E80E95EC4B128595AC231181C330" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_17EF52CE5F174F708988D3792C4EF1FE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_1969096EAD2A4292BA27BED98ACA17FB" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_1B2AF463338C40B0AA1AE5F3149D343D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_1B7AD03818962FE68D82D1B4D4642FB6" "OwnerKey" = "8:_6827AD9625DF4C39A8A0E39FB82E3CD0" "MsmSig" = "8:_UNDEFINED" @@ -57,14 +201,56 @@ } "Entry" { + "MsmKey" = "8:_1FA9731E400641839E29839952B10A67" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_1FEA719F022CEAF9A269D43AD22AF93F" "OwnerKey" = "8:_6827AD9625DF4C39A8A0E39FB82E3CD0" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_22027664F1B34E9EA39D3B330D460FEC" - "OwnerKey" = "8:_9AB559EE78054703AF56D8E7744F3806" + "MsmKey" = "8:_223105795E5D49B8BD6AD93E12847151" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2286937AE4E6436F95D07AB1A588794C" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_22917529BC854387B3A9A7C7DF33EDAE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_231B4CAAD95A4D5391BD2A27D1BEA77E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2393441DF8524232AD40E46DF2640189" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_24AD39A905574544A69A1F67C0A3CB16" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_24E2C7FDC2EC4816A9E3D2C1860E5A41" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -75,18 +261,78 @@ } "Entry" { + "MsmKey" = "8:_25A855DA124946CF878441F4673A48D0" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_2753983542094EB781E4AA2FC90CF0D4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2806AFD2B45345C09E96EA1E139BE213" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_285DF38D8D4043129B7F2E2227603DB5" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2A2B73C24B1541B6A6B380B019877693" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2ADB9B378F3441FAA48C077E968E5929" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2BDA0576A7894E6481660D46B5A01437" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_2E849B77EC5944D9A6C38F02C8D1125E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_3148F098C35140D2B38C9EAE570278C7" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_316BCBF5300B4931ACBC48DBC4935C7D" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_32CF1A99F5714908AFDEF8B5686EB63E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_33545BED81CC46E2B3D3406385AB4BAB" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -99,38 +345,194 @@ } "Entry" { + "MsmKey" = "8:_34FCB627708446FA998E61BC379C3274" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_35CF60C53FC643E8BD271C8F4F257F7F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_363B16784029465C9F2F27ED2544DFB2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_368F66344EEB4252939476AE57FD170A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_385F635F391A45E7AF3C9ECCB8580BBB" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_389BF52BB2004C779C0B34C431B8DF6C" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_399D3AEE4A37416D8C89F4C87D4B2186" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_39C6B9B06D72499680F6F1AC6CB5BC41" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_3C0AF639574546BF806A7C2967B84D64" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_3D004382B4E34EBDA27EC6CB3542C77B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_3D0E6540B75342C983BD1B81A7D0687F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_3DE636877DB640AFBF78E771B6339E7B" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_40AF38FA32E74704B11F107D51809D67" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_40D1B86DB6384BBC8642DCCF1C19EF44" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_44BA53E81F0FD93AF224C5D26706AA36" - "OwnerKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "MsmKey" = "8:_410B722D4B1A40448CBAB425C76BB942" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_4E1164243AC854F0A069C84D57AE7EC1" - "OwnerKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "MsmKey" = "8:_439BE780DFAF488CAB0628CA9EE105C1" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_44610C616D8541A7AADF2A907D62CBC3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_45B9B43F0312436EB21A3EE185196698" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_45C9A622FD5C48778968D4FB6D6DF5A3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_497B59E060004BDD8AA9C9117771CC9F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_4C3F6AED1BC84C90867D52A478F93263" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_4DA88C47D7F24AF98C9EA65B22E3963A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_4E78F6174A7940B1A434F71096AE6599" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5083DF2E71904EFD87DF275C566E4DA3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5108D04A106F4BF0925ECDB0238858A0" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_519BA9151D974A46AB5AB56B0E4AA51F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_52EBADEB628C495898A151D2C9DFCE84" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_55DD5447D0704C5BA1D4485FFCB12750" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_56740A2D926040FB9DAB02218AF22A12" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_56EC8A30479C4EC19744D5D61814D1EE" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" + "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" "MsmSig" = "8:_UNDEFINED" } "Entry" { "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -147,42 +549,222 @@ } "Entry" { + "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" + "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" + "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5795B3AD60144867AAF76625BF8D12E6" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5816864CB37148EAA1B4E3694087D957" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_583A5A54C626479C895F30381B84C0A0" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5902179EF3C847EB89E23FF72968009A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5ACDADD74B40457B83F1981EFAA7857D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5BB0EB6C5231446AAC5204710D7B80AF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5DA8B5994E0C45E389DCB5E6A78A606F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_5FD67D7865EF469681C9064A909D284B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_600402FDB5064D0B82810CB4E6DBEEA1" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_601A2EF5BC9E4E18847D198965955979" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_601C974367D84D838C6C6CC9A6B0EF12" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_6159CD85FA674476A8500E4D4DBDE249" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_6177B9D05A0A4128BE5E565074A64B8A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6330D1F1B93B4ED782294CC53AE58C88" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_647EE324793341ECAF24A8ADE8855B94" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_6827AD9625DF4C39A8A0E39FB82E3CD0" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_6C73F449C09179BCC4C28824D77F9A65" + "MsmKey" = "8:_6843982E6E314D768F6CC5F102688B0E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_69328962B15D4340A2F60B4AC76D1BE6" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6B65629ABA014116BCEFB83DAD553D16" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6CE23E7F518243C7BA141E1E65C168CC" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6E07916227C246EB9FBC73A4BE564BE6" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6F2EA43D6CB84A66ADD39614C2DB53ED" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6F5170ED41D34E2696B817A3C44F11D5" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_6FB75A46740D4F2B9472C15B4E58AD88" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7064BEB3EECD4A94AEFA90955DA88FA7" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_718A74839C0D4897B1C66283597425E8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_71BE9607A7510034EDED2B591E6F981E" "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_7274EADA979045C380198FDA59E37468" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_741FC4B6CB49412FAF2FD02903557840" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_76BB1C3A20244799AAEAC764389DB226" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_78792FEB009343479E361944DCAF9C36" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_78F8BB67ADA242B1A949C41F909666E2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_797C85AEE487426EA22D8B0253111FA3" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_7AB0994101C146EFA3493863882452B4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_7B327706377AE5F414299F6736D8EC00" "OwnerKey" = "8:_87A61D50A0B944458FE3B315A47FA221" "MsmSig" = "8:_UNDEFINED" @@ -195,6 +777,78 @@ } "Entry" { + "MsmKey" = "8:_7C0CEAF559A84A0C8723C9328DBADDAE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7C346BC46D0B44BAB4750A35B70A0B46" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7D493307EEB847228F6CCEEE90D42DAA" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7E94D0A6FD0E48C8A7B7D8421F4D2DE6" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7EA7534B81A9415D89174750F58D27A2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7EDF89CE07114545909E655CC5567735" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_7FF341438FB64A7A9A22CD9B9839C63C" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8196C46459C542B1A8EC49445C14472F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_81D9FBCB8E3D4952A47B9BC4B2567BB2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_83305974B07F42C18E0AE99DBBF15037" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_84F3D3D23A1447339492C40CC699DD44" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_8743DB62D7EF4D72A820BAEC30E97DEC" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -225,8 +879,122 @@ } "Entry" { - "MsmKey" = "8:_919458435B414CF3A23DEBA86D49CE7D" - "OwnerKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" + "MsmKey" = "8:_8835232945DD4F2C8333D0E2CACA9DF8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8897C560DD094D698EA4A9B88C98D0A6" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_88FCE1823EA04CF8B1FE816A1A0E1E3B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_89BA174439B14C8F8FA942C9E150B718" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8AB21318425441BA811D919F6DC5D6E8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8AC15EDDB6994E7794DBA7A0A3DCA60B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8C0EC3CEAEB04DE7B6729EBBCE8C0B36" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8CF26805EE71422CBC1C152A7462FE44" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8E3C8EAC49BD49EE886229B91DF16166" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8E909C0BE4C942A7804FF65AD2DDAFB5" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_8F39494E0D4C4FF091091DAF5183388D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_90ED9EDA4346456F8957D9E7EB0D8DD9" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_91C427E57A3B48F399BAA6AF4A588293" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_924E5E9E11C245D383CFFA77545B69FE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_9356233C1CB34B1E8F0781EA7188920C" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_94DA87D95C7E4D009C9FFF2F647405C2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_95AA28E96394481AB463DEF29FAA78F8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_9636D19CCBD441D993A061FDE5FF7807" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_96AB009303B44C2CB19C172EB5D51CE2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_98FAE27F8FDB404098DC9E605F753E0D" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -237,13 +1005,25 @@ } "Entry" { - "MsmKey" = "8:_99C9A286D959DE56EFE6671DE8C65FAE" - "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" + "MsmKey" = "8:_9AB559EE78054703AF56D8E7744F3806" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_9AB559EE78054703AF56D8E7744F3806" + "MsmKey" = "8:_9C7D1B3E93F0481E804E68339BBEB021" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_9D1F177B2FFA4D3D84F11A8A20D1F611" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_9E96BFE206504BFEBF955CA9CBA846DC" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } @@ -255,31 +1035,163 @@ } "Entry" { + "MsmKey" = "8:_A2652446B2604985A24F992F42F2761E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_A28EBA4DE8914C988801F829035FFEC8" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_A3D55F63E0484B52BD23FBF66915D43D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_A3F76D4DE92743659E861547FE69DE1F" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_A40F34AB634C48D7B953C2E420915F07" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" + "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" + "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" + "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_A89761474AC748AE931002EE5E581284" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_A8F09DC7C504415BA06801BA7C5D750B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_AAB10625A4634D85821952D2CD42642B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_AB28BFBD30FC423293D42423F112B6BD" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_ADAC04E54F894404A83B379617E7A97E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_ADE00C6618D141C49F057DA07B235947" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_ADE37DD61B5347DB890F2957492F3745" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_AEB84272F38444F5BB708EF3E0833776" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_AF4D152A202841D2BF769FBDF5CEB413" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_AF7C7BE64EA846468642088F15DA4A1F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B2A5C84577F64477896DD21845527BD0" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B2F1836277ED493C958A02184D327E49" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B4A1253749E245B088BD031A677CFB4A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B50F62EC90CD4A07912A63B90AAD3293" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B63084B7B255496B836164646F5D536D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B8FEC8F89AD14BE5AF941C226AAA31AF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_B90FC469B0A7444AA681C869AB6704D4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_BA8C69F1D2364A2E9268819F2A8085FB" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "MsmKey" = "8:_BC30361C69EF4CD8869553D0AF1CA80A" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } @@ -291,20 +1203,110 @@ } "Entry" { + "MsmKey" = "8:_BD602A7A1C334F74815A52DF4C98D0CD" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_BDA20A14A241440B8632ABEBC26E35DB" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_BFE971F656CA46C7A298C91757E9AFF7" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C047CC641D1E4EAF9CF3032B69B49C46" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C10D7A1FC3A44F749156F781C6DB3140" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C188D42682084A7D983722A92046C3EA" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_C28E286C209D4FBCBBF1FD46837E0854" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_C40BE5E85657415296E733A2F29C985A" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_C46C31F5523D72D416CDEF1804E2A64D" "OwnerKey" = "8:_6827AD9625DF4C39A8A0E39FB82E3CD0" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_CAED6729570B459EBDCE72B9E32AB03D" - "OwnerKey" = "8:_22027664F1B34E9EA39D3B330D460FEC" + "MsmKey" = "8:_C555162941C34AE1B51CE4CB5352CD4E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C77702CF400646678BD9CC9242B1B6BB" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C80AC3C16BED483D8228AFA835D7F15F" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C82D0E331C0F4C368F91EE2A7BD0565D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_C980F539DB294419AE3FC77C33A77AEF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_CCDA5668BC2E426E9E05F8875B752536" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_CDAA2F2C190A4DCB928E78564F76BCA3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_CE6A0F7079024AF4A0933BB7F9BFE195" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -315,104 +1317,116 @@ } "Entry" { + "MsmKey" = "8:_D21F33B9B9CA4F28996FF4CB02872772" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_D2474D58E6AC4CA0A5FCEF9B4BA6B3E4" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_D31ED5CC46194EA6A81F9F9BE3DE5CD1" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_D4219ACEADB94026AA93732F128494E3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_D738FF1390814B24A9021A19F95E4290" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_D81E0DB3F5674742AB46156B10CCC968" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_D92503794DA8471A8BFE082269932306" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_D980AD62DE974BABA15438FF6ACFA8BF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_DAC88B65E7EB4A8E92FBF82174072D28" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_DB95F4450EFC4AAE94E1A2EDE746B2B2" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_DBB4560AE2924E349948FA5BB09FE28E" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_DCA1A5B0C13D49259445C449B0E50774" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { + "MsmKey" = "8:_DD150D2502704E67A072908310DB66F3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_DFE2B52C9EDF48DA81BC3E67D6564B0D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_E02F836E90974BFC8E673705BE975E95" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_E09F4497BBF14351A2528D6E215DE28D" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_E0E1B281255A4DC5934647A82D674F99" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_BCBBD06CDE4641D2B093302D05CB9004" + "MsmKey" = "8:_E17B0DADA8FC4512A4888DE697BC75B3" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_E663A731DB91460B991DAA889BD480EE" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_C28E286C209D4FBCBBF1FD46837E0854" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_6827AD9625DF4C39A8A0E39FB82E3CD0" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_BA8C69F1D2364A2E9268819F2A8085FB" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_AF4D152A202841D2BF769FBDF5CEB413" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_A3F76D4DE92743659E861547FE69DE1F" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_A22B625A9A0E4580B7976E1C76B13D93" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_9AB559EE78054703AF56D8E7744F3806" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_882694E02F0649F29ADA794E68DA525C" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_87A61D50A0B944458FE3B315A47FA221" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_600402FDB5064D0B82810CB4E6DBEEA1" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_E619C79FD5E14F6B965446BB0371AFDE" - "OwnerKey" = "8:_1430E155AE5F4B92A78B32E05FEAF3EF" + "MsmKey" = "8:_E508524696DD452E91B739C7EC466374" + "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" } "Entry" @@ -429,93 +1443,115 @@ } "Entry" { + "MsmKey" = "8:_E71C997965964853BA87889843181711" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_ECC63C4CBBE44D568436E4312BA47908" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_ECD3190AB8F14933A35236EC20488C61" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_EEECC8B200772D4A9F89A5F65A3B97F5" "OwnerKey" = "8:_BA8C69F1D2364A2E9268819F2A8085FB" "MsmSig" = "8:_UNDEFINED" } + "Entry" + { + "MsmKey" = "8:_F00371D156F84024BD88562FE8C9DDBF" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F165D6BEADEC6B529C157DDC87667E80" + "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F1A181FCDB444332B63B8E51A4EBD448" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F29E17EBECD5437C9244B842F5038BB1" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F43E0A4CFF22404996751E7EE5CFC3DE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F58D989DCC40429EBC0C9D43AF5D8BA1" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F6AF071E4E7F4C6CBB28DCDAA6A711C3" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F72D772B69424F28A3E406E8BCD1878B" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F84D270966164BAAA768673FDACED3CC" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_F9276AD7A6E840C59B3B10CD18910D35" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FBBAFABEF4884CCA881660F376B5F6FE" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FCB2559A1E524CAFA362E24808358B68" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FF64089169014510AEEBB6F516BFE6D5" + "OwnerKey" = "8:_UNDEFINED" + "MsmSig" = "8:_UNDEFINED" + } } "Configurations" { - "Release SSE2" + "Release" { - "DisplayName" = "8:Release SSE2" + "DisplayName" = "8:Release" "IsDebugOnly" = "11:FALSE" - "IsReleaseOnly" = "11:TRUE" - "OutputFilename" = "8:Release SSE2\\setup.msi" - "PackageFilesAs" = "3:2" - "PackageFileSize" = "3:-2147483648" - "CabType" = "3:1" - "Compression" = "3:3" - "SignOutput" = "11:FALSE" - "CertificateFile" = "8:" - "PrivateKeyFile" = "8:" - "TimeStampServer" = "8:" - "InstallerBootstrapper" = "3:2" - "BootstrapperCfg:{63ACBE69-63AA-4F98-B2B6-99F9E24495F2}" - { - "Enabled" = "11:TRUE" - "PromptEnabled" = "11:TRUE" - "PrerequisitesLocation" = "2:1" - "Url" = "8:" - "ComponentsUrl" = "8:" - "Items" - { - "{EDC2488A-8267-493A-A98E-7D9C3B36CDF3}:Microsoft.Visual.C++.9.0.x86" - { - "Name" = "8:Visual C++ Runtime Libraries (x86)" - "ProductCode" = "8:Microsoft.Visual.C++.9.0.x86" - } - "{EDC2488A-8267-493A-A98E-7D9C3B36CDF3}:Microsoft.Windows.Installer.3.1" - { - "Name" = "8:Windows Installer 3.1" - "ProductCode" = "8:Microsoft.Windows.Installer.3.1" - } - } - } - } - "Release SSE4" - { - "DisplayName" = "8:Release SSE4" - "IsDebugOnly" = "11:FALSE" - "IsReleaseOnly" = "11:TRUE" - "OutputFilename" = "8:Release SSE4\\setup.msi" - "PackageFilesAs" = "3:2" - "PackageFileSize" = "3:-2147483648" - "CabType" = "3:1" - "Compression" = "3:3" - "SignOutput" = "11:FALSE" - "CertificateFile" = "8:" - "PrivateKeyFile" = "8:" - "TimeStampServer" = "8:" - "InstallerBootstrapper" = "3:2" - "BootstrapperCfg:{63ACBE69-63AA-4F98-B2B6-99F9E24495F2}" - { - "Enabled" = "11:TRUE" - "PromptEnabled" = "11:TRUE" - "PrerequisitesLocation" = "2:1" - "Url" = "8:" - "ComponentsUrl" = "8:" - "Items" - { - "{EDC2488A-8267-493A-A98E-7D9C3B36CDF3}:Microsoft.Visual.C++.9.0.x86" - { - "Name" = "8:Visual C++ Runtime Libraries (x86)" - "ProductCode" = "8:Microsoft.Visual.C++.9.0.x86" - } - "{EDC2488A-8267-493A-A98E-7D9C3B36CDF3}:Microsoft.Windows.Installer.3.1" - { - "Name" = "8:Windows Installer 3.1" - "ProductCode" = "8:Microsoft.Windows.Installer.3.1" - } - } - } - } - "Release SSSE3" - { - "DisplayName" = "8:Release SSSE3" - "IsDebugOnly" = "11:FALSE" - "IsReleaseOnly" = "11:TRUE" - "OutputFilename" = "8:Release SSSE3\\setup.msi" + "IsReleaseOnly" = "11:FALSE" + "OutputFilename" = "8:Release\\pcsx2inst.msi" "PackageFilesAs" = "3:2" "PackageFileSize" = "3:-2147483648" "CabType" = "3:1" @@ -567,12 +1603,12 @@ } "File" { - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0950432B04CD787CC7B938905C6C4399" + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_008B1BB47C0A46C0B025FDE5376533F4" { - "SourcePath" = "8:glew32.dll" - "TargetName" = "8:glew32.dll" + "SourcePath" = "8:..\\bin\\patches\\A08C4057.pnach" + "TargetName" = "8:A08C4057.pnach" "Tag" = "8:" - "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" "Vital" = "11:TRUE" @@ -584,7 +1620,267 @@ "PackageAs" = "3:1" "Register" = "3:1" "Exclude" = "11:FALSE" - "IsDependency" = "11:TRUE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_00ABD03A61124E9F8BEC5E4D2D7F2715" + { + "SourcePath" = "8:..\\bin\\patches\\5E115FB6.pnach" + "TargetName" = "8:5E115FB6.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_03E75E8822CE48BFB4D8378A4FD3A9E8" + { + "SourcePath" = "8:..\\bin\\patches\\D48A92E1.pnach" + "TargetName" = "8:D48A92E1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0435A4202F22485295FC9B14FF9E0DEF" + { + "SourcePath" = "8:..\\bin\\patches\\60013EBD.pnach" + "TargetName" = "8:60013EBD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_048B9870FB3E4278ADCB974F47D3AA3D" + { + "SourcePath" = "8:..\\bin\\patches\\71584BAC.pnach" + "TargetName" = "8:71584BAC.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_069D4ED441804E6F91F93A1BA01C21B8" + { + "SourcePath" = "8:..\\bin\\patches\\848C6247.pnach" + "TargetName" = "8:848C6247.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_072AB1B5BAC84646894F39D3E1BF092F" + { + "SourcePath" = "8:..\\bin\\patches\\3FB69323.pnach" + "TargetName" = "8:3FB69323.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_072B429763E648A2AFA3744F69B35FD8" + { + "SourcePath" = "8:..\\bin\\patches\\2251E14D.pnach" + "TargetName" = "8:2251E14D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_07B925C7C3DD4023870E4C350A933F01" + { + "SourcePath" = "8:..\\bin\\patches\\512B5046.pnach" + "TargetName" = "8:512B5046.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_086724BEF26C4B91964547A4ABCFC33D" + { + "SourcePath" = "8:..\\bin\\patches\\4043F228.pnach" + "TargetName" = "8:4043F228.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_094A3551F77548E28331D27415415CD4" + { + "SourcePath" = "8:..\\bin\\patches\\33F7D21A.pnach" + "TargetName" = "8:33F7D21A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0953A3DB125D4969AA9CF0EAEE0174F8" + { + "SourcePath" = "8:..\\bin\\patches\\722BBD62.pnach" + "TargetName" = "8:722BBD62.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0A01FDD1F6A54A43A900C9AC92D4B2D9" + { + "SourcePath" = "8:..\\bin\\patches\\96B2F56D.pnach" + "TargetName" = "8:96B2F56D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0AA27B6457FC4913924E4F391F4581DC" + { + "SourcePath" = "8:..\\bin\\patches\\54D6BEE3.pnach" + "TargetName" = "8:54D6BEE3.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0AB0A47469D44DC28C0973A8B295B919" @@ -647,6 +1943,226 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0D6C5B8003914BC8B1F4C3EC09C58AB2" + { + "SourcePath" = "8:..\\bin\\patches\\2B2E1535.pnach" + "TargetName" = "8:2B2E1535.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_0E8BBD04D986469EA64F1A82EC4DB8FA" + { + "SourcePath" = "8:..\\bin\\patches\\337B927C.pnach" + "TargetName" = "8:337B927C.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1037D220B29040FDA7B60A0A9DA4865E" + { + "SourcePath" = "8:..\\bin\\patches\\9AAC530D.pnach" + "TargetName" = "8:9AAC530D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_110E08BF1DA04EF69BF2D475585D0500" + { + "SourcePath" = "8:..\\bin\\patches\\7130C553.pnach" + "TargetName" = "8:7130C553.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1320398166C4413FB4BF04DC2C1FD8EF" + { + "SourcePath" = "8:..\\bin\\patches\\B82E4C4B.pnach" + "TargetName" = "8:B82E4C4B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_14A843D75DC34B599F3463074612476E" + { + "SourcePath" = "8:..\\bin\\patches\\7377BC6F.pnach" + "TargetName" = "8:7377BC6F.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_164C7824F6134DF3A56AF21FB50AE125" + { + "SourcePath" = "8:..\\bin\\patches\\AA31B5BF.pnach" + "TargetName" = "8:AA31B5BF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1750E80E95EC4B128595AC231181C330" + { + "SourcePath" = "8:..\\bin\\patches\\7F6319C7.pnach" + "TargetName" = "8:7F6319C7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_17EF52CE5F174F708988D3792C4EF1FE" + { + "SourcePath" = "8:..\\bin\\patches\\A74F99CD.pnach" + "TargetName" = "8:A74F99CD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1969096EAD2A4292BA27BED98ACA17FB" + { + "SourcePath" = "8:..\\bin\\patches\\1468C474.pnach" + "TargetName" = "8:1468C474.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1B2AF463338C40B0AA1AE5F3149D343D" + { + "SourcePath" = "8:..\\bin\\patches\\163F0461.pnach" + "TargetName" = "8:163F0461.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_1B7AD03818962FE68D82D1B4D4642FB6" { "SourcePath" = "8:MSVFW32.dll" @@ -707,6 +2223,146 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_223105795E5D49B8BD6AD93E12847151" + { + "SourcePath" = "8:..\\bin\\patches\\E906EA37.pnach" + "TargetName" = "8:E906EA37.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2286937AE4E6436F95D07AB1A588794C" + { + "SourcePath" = "8:..\\bin\\patches\\6ADBC24B.pnach" + "TargetName" = "8:6ADBC24B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_22917529BC854387B3A9A7C7DF33EDAE" + { + "SourcePath" = "8:..\\bin\\patches\\EE838B5C.pnach" + "TargetName" = "8:EE838B5C.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_231B4CAAD95A4D5391BD2A27D1BEA77E" + { + "SourcePath" = "8:..\\bin\\patches\\DC85FC8F.pnach" + "TargetName" = "8:DC85FC8F.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2393441DF8524232AD40E46DF2640189" + { + "SourcePath" = "8:..\\bin\\patches\\015314A2.pnach" + "TargetName" = "8:015314A2.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_24AD39A905574544A69A1F67C0A3CB16" + { + "SourcePath" = "8:..\\bin\\patches\\F6F9A91D.pnach" + "TargetName" = "8:F6F9A91D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_24E2C7FDC2EC4816A9E3D2C1860E5A41" + { + "SourcePath" = "8:..\\bin\\patches\\8B0725D5.pnach" + "TargetName" = "8:8B0725D5.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_252DE06AE3874074A9070423E08DE75A" { "SourcePath" = "8:..\\bin\\Langs\\hb_HB\\LC_MESSAGES\\pcsx2.mo" @@ -727,6 +2383,186 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_25A855DA124946CF878441F4673A48D0" + { + "SourcePath" = "8:..\\bin\\patches\\DDA2FA6A.pnach" + "TargetName" = "8:DDA2FA6A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2753983542094EB781E4AA2FC90CF0D4" + { + "SourcePath" = "8:..\\bin\\patches\\0518D275.pnach" + "TargetName" = "8:0518D275.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2806AFD2B45345C09E96EA1E139BE213" + { + "SourcePath" = "8:..\\bin\\patches\\E4A275B2.pnach" + "TargetName" = "8:E4A275B2.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_285DF38D8D4043129B7F2E2227603DB5" + { + "SourcePath" = "8:..\\bin\\patches\\6A4EFE60.pnach" + "TargetName" = "8:6A4EFE60.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2A2B73C24B1541B6A6B380B019877693" + { + "SourcePath" = "8:..\\bin\\patches\\D08648B6.pnach" + "TargetName" = "8:D08648B6.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2ADB9B378F3441FAA48C077E968E5929" + { + "SourcePath" = "8:..\\bin\\patches\\83261085.pnach" + "TargetName" = "8:83261085.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2BDA0576A7894E6481660D46B5A01437" + { + "SourcePath" = "8:..\\bin\\patches\\60FA8C69.pnach" + "TargetName" = "8:60FA8C69.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_2E849B77EC5944D9A6C38F02C8D1125E" + { + "SourcePath" = "8:..\\bin\\patches\\C9C145BF.pnach" + "TargetName" = "8:C9C145BF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_3148F098C35140D2B38C9EAE570278C7" + { + "SourcePath" = "8:..\\bin\\patches\\DCC4EEEA.pnach" + "TargetName" = "8:DCC4EEEA.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_316BCBF5300B4931ACBC48DBC4935C7D" { "SourcePath" = "8:..\\bin\\Langs\\du_DU\\LC_MESSAGES\\pcsx2.mo" @@ -747,6 +2583,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_32CF1A99F5714908AFDEF8B5686EB63E" + { + "SourcePath" = "8:..\\bin\\patches\\77ECAAA0.pnach" + "TargetName" = "8:77ECAAA0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_33545BED81CC46E2B3D3406385AB4BAB" { "SourcePath" = "8:..\\bin\\Langs\\el_EL\\LC_MESSAGES\\pcsx2.mo" @@ -787,6 +2643,226 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_34FCB627708446FA998E61BC379C3274" + { + "SourcePath" = "8:..\\bin\\patches\\934F9081.pnach" + "TargetName" = "8:934F9081.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_35CF60C53FC643E8BD271C8F4F257F7F" + { + "SourcePath" = "8:..\\bin\\patches\\F6DC728D.pnach" + "TargetName" = "8:F6DC728D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_363B16784029465C9F2F27ED2544DFB2" + { + "SourcePath" = "8:..\\bin\\patches\\4D6DBB75.pnach" + "TargetName" = "8:4D6DBB75.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_368F66344EEB4252939476AE57FD170A" + { + "SourcePath" = "8:..\\bin\\patches\\3E68955A.pnach" + "TargetName" = "8:3E68955A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_385F635F391A45E7AF3C9ECCB8580BBB" + { + "SourcePath" = "8:..\\bin\\patches\\5A7635C1.pnach" + "TargetName" = "8:5A7635C1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_389BF52BB2004C779C0B34C431B8DF6C" + { + "SourcePath" = "8:..\\bin\\patches\\96B76E56.pnach" + "TargetName" = "8:96B76E56.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_399D3AEE4A37416D8C89F4C87D4B2186" + { + "SourcePath" = "8:..\\bin\\patches\\333F1F59.pnach" + "TargetName" = "8:333F1F59.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_39C6B9B06D72499680F6F1AC6CB5BC41" + { + "SourcePath" = "8:..\\bin\\patches\\AFAC88EF.pnach" + "TargetName" = "8:AFAC88EF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_3C0AF639574546BF806A7C2967B84D64" + { + "SourcePath" = "8:..\\bin\\patches\\624F11F1.pnach" + "TargetName" = "8:624F11F1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_3D004382B4E34EBDA27EC6CB3542C77B" + { + "SourcePath" = "8:..\\bin\\patches\\951555A0.pnach" + "TargetName" = "8:951555A0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_3D0E6540B75342C983BD1B81A7D0687F" + { + "SourcePath" = "8:..\\bin\\patches\\692CBA8E.pnach" + "TargetName" = "8:692CBA8E.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_3DE636877DB640AFBF78E771B6339E7B" { "SourcePath" = "8:..\\bin\\Langs\\fr_FR\\LC_MESSAGES\\pcsx2.mo" @@ -807,6 +2883,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_40AF38FA32E74704B11F107D51809D67" + { + "SourcePath" = "8:..\\bin\\patches\\3BD85DA4.pnach" + "TargetName" = "8:3BD85DA4.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_40D1B86DB6384BBC8642DCCF1C19EF44" { "SourcePath" = "8:..\\bin\\Langs\\po_PO\\LC_MESSAGES\\pcsx2.mo" @@ -827,12 +2923,12 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_44BA53E81F0FD93AF224C5D26706AA36" + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_410B722D4B1A40448CBAB425C76BB942" { - "SourcePath" = "8:DDRAW.dll" - "TargetName" = "8:DDRAW.dll" + "SourcePath" = "8:..\\bin\\patches\\E0127F2D.pnach" + "TargetName" = "8:E0127F2D.pnach" "Tag" = "8:" - "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" "Vital" = "11:TRUE" @@ -843,16 +2939,16 @@ "SharedLegacy" = "11:FALSE" "PackageAs" = "3:1" "Register" = "3:1" - "Exclude" = "11:TRUE" - "IsDependency" = "11:TRUE" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_4E1164243AC854F0A069C84D57AE7EC1" + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_439BE780DFAF488CAB0628CA9EE105C1" { - "SourcePath" = "8:OPENGL32.dll" - "TargetName" = "8:OPENGL32.dll" + "SourcePath" = "8:..\\bin\\patches\\B2BDE9F3.pnach" + "TargetName" = "8:B2BDE9F3.pnach" "Tag" = "8:" - "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" "Vital" = "11:TRUE" @@ -863,8 +2959,288 @@ "SharedLegacy" = "11:FALSE" "PackageAs" = "3:1" "Register" = "3:1" - "Exclude" = "11:TRUE" - "IsDependency" = "11:TRUE" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_44610C616D8541A7AADF2A907D62CBC3" + { + "SourcePath" = "8:..\\bin\\patches\\4B6DDB6B.pnach" + "TargetName" = "8:4B6DDB6B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_45B9B43F0312436EB21A3EE185196698" + { + "SourcePath" = "8:..\\bin\\patches\\2088950A.pnach" + "TargetName" = "8:2088950A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_45C9A622FD5C48778968D4FB6D6DF5A3" + { + "SourcePath" = "8:..\\bin\\patches\\B130E5BA.pnach" + "TargetName" = "8:B130E5BA.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_497B59E060004BDD8AA9C9117771CC9F" + { + "SourcePath" = "8:..\\bin\\patches\\013E349D.pnach" + "TargetName" = "8:013E349D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_4C3F6AED1BC84C90867D52A478F93263" + { + "SourcePath" = "8:..\\bin\\patches\\A3ACF3C7.pnach" + "TargetName" = "8:A3ACF3C7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_4DA88C47D7F24AF98C9EA65B22E3963A" + { + "SourcePath" = "8:..\\bin\\patches\\5162BCCA.pnach" + "TargetName" = "8:5162BCCA.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_4E78F6174A7940B1A434F71096AE6599" + { + "SourcePath" = "8:..\\bin\\patches\\32629F36.pnach" + "TargetName" = "8:32629F36.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5083DF2E71904EFD87DF275C566E4DA3" + { + "SourcePath" = "8:..\\bin\\patches\\8BE3D7B2.pnach" + "TargetName" = "8:8BE3D7B2.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5108D04A106F4BF0925ECDB0238858A0" + { + "SourcePath" = "8:..\\bin\\patches\\6D70F0E0.pnach" + "TargetName" = "8:6D70F0E0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_519BA9151D974A46AB5AB56B0E4AA51F" + { + "SourcePath" = "8:..\\bin\\patches\\B234036E.pnach" + "TargetName" = "8:B234036E.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_52EBADEB628C495898A151D2C9DFCE84" + { + "SourcePath" = "8:..\\bin\\patches\\844B58EE.pnach" + "TargetName" = "8:844B58EE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_55DD5447D0704C5BA1D4485FFCB12750" + { + "SourcePath" = "8:..\\bin\\patches\\0B359BBF.pnach" + "TargetName" = "8:0B359BBF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_56740A2D926040FB9DAB02218AF22A12" + { + "SourcePath" = "8:..\\bin\\patches\\F56C7948.pnach" + "TargetName" = "8:F56C7948.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_56EC8A30479C4EC19744D5D61814D1EE" + { + "SourcePath" = "8:..\\bin\\patches\\C4467D30.pnach" + "TargetName" = "8:C4467D30.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_572363F211D9AF197AFDB57B760E8636" @@ -887,6 +3263,206 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5795B3AD60144867AAF76625BF8D12E6" + { + "SourcePath" = "8:..\\bin\\patches\\7FD7A1B9.pnach" + "TargetName" = "8:7FD7A1B9.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5816864CB37148EAA1B4E3694087D957" + { + "SourcePath" = "8:..\\bin\\patches\\09D35D3F.pnach" + "TargetName" = "8:09D35D3F.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_583A5A54C626479C895F30381B84C0A0" + { + "SourcePath" = "8:..\\bin\\patches\\93223BE4.pnach" + "TargetName" = "8:93223BE4.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5902179EF3C847EB89E23FF72968009A" + { + "SourcePath" = "8:..\\bin\\patches\\DA0535FD.pnach" + "TargetName" = "8:DA0535FD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5ACDADD74B40457B83F1981EFAA7857D" + { + "SourcePath" = "8:..\\bin\\patches\\C04FB5FD.pnach" + "TargetName" = "8:C04FB5FD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5BB0EB6C5231446AAC5204710D7B80AF" + { + "SourcePath" = "8:..\\bin\\patches\\1510E1D1.pnach" + "TargetName" = "8:1510E1D1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5DA8B5994E0C45E389DCB5E6A78A606F" + { + "SourcePath" = "8:..\\bin\\patches\\1025D50A.pnach" + "TargetName" = "8:1025D50A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5FD67D7865EF469681C9064A909D284B" + { + "SourcePath" = "8:..\\bin\\patches\\85E92C92.pnach" + "TargetName" = "8:85E92C92.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_601A2EF5BC9E4E18847D198965955979" + { + "SourcePath" = "8:..\\bin\\patches\\116154AD.pnach" + "TargetName" = "8:116154AD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_601C974367D84D838C6C6CC9A6B0EF12" + { + "SourcePath" = "8:..\\bin\\patches\\302797DF.pnach" + "TargetName" = "8:302797DF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6159CD85FA674476A8500E4D4DBDE249" { "SourcePath" = "8:..\\bin\\Langs\\es_ES\\LC_MESSAGES\\pcsx2.mo" @@ -907,7 +3483,267 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6C73F449C09179BCC4C28824D77F9A65" + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6177B9D05A0A4128BE5E565074A64B8A" + { + "SourcePath" = "8:..\\bin\\patches\\3CFE3B37.pnach" + "TargetName" = "8:3CFE3B37.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6330D1F1B93B4ED782294CC53AE58C88" + { + "SourcePath" = "8:..\\bin\\patches\\F9E575D0.pnach" + "TargetName" = "8:F9E575D0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_647EE324793341ECAF24A8ADE8855B94" + { + "SourcePath" = "8:..\\bin\\patches\\5EB127E7.pnach" + "TargetName" = "8:5EB127E7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6843982E6E314D768F6CC5F102688B0E" + { + "SourcePath" = "8:..\\bin\\patches\\7EBEEBBD.pnach" + "TargetName" = "8:7EBEEBBD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_69328962B15D4340A2F60B4AC76D1BE6" + { + "SourcePath" = "8:..\\bin\\patches\\7ABDBB5E.pnach" + "TargetName" = "8:7ABDBB5E.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6B65629ABA014116BCEFB83DAD553D16" + { + "SourcePath" = "8:..\\bin\\patches\\4C9EE7DF.pnach" + "TargetName" = "8:4C9EE7DF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6CE23E7F518243C7BA141E1E65C168CC" + { + "SourcePath" = "8:..\\bin\\patches\\EF5B6AAD.pnach" + "TargetName" = "8:EF5B6AAD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6E07916227C246EB9FBC73A4BE564BE6" + { + "SourcePath" = "8:..\\bin\\patches\\E1C5F607.pnach" + "TargetName" = "8:E1C5F607.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6F2EA43D6CB84A66ADD39614C2DB53ED" + { + "SourcePath" = "8:..\\bin\\patches\\41A3191C.pnach" + "TargetName" = "8:41A3191C.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6F5170ED41D34E2696B817A3C44F11D5" + { + "SourcePath" = "8:..\\bin\\patches\\9D6F46F0.pnach" + "TargetName" = "8:9D6F46F0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_6FB75A46740D4F2B9472C15B4E58AD88" + { + "SourcePath" = "8:..\\bin\\patches\\A88ACA28.pnach" + "TargetName" = "8:A88ACA28.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7064BEB3EECD4A94AEFA90955DA88FA7" + { + "SourcePath" = "8:..\\bin\\patches\\1D2818AF.pnach" + "TargetName" = "8:1D2818AF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_718A74839C0D4897B1C66283597425E8" + { + "SourcePath" = "8:..\\bin\\patches\\E1FD9A2D.pnach" + "TargetName" = "8:E1FD9A2D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_71BE9607A7510034EDED2B591E6F981E" { "SourcePath" = "8:gnu_gettext.dll" "TargetName" = "8:gnu_gettext.dll" @@ -927,6 +3763,66 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7274EADA979045C380198FDA59E37468" + { + "SourcePath" = "8:..\\bin\\patches\\EDD7E0FF.pnach" + "TargetName" = "8:EDD7E0FF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_741FC4B6CB49412FAF2FD02903557840" + { + "SourcePath" = "8:..\\bin\\patches\\08FB9DCF.pnach" + "TargetName" = "8:08FB9DCF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_76BB1C3A20244799AAEAC764389DB226" + { + "SourcePath" = "8:..\\bin\\patches\\0C370E94.pnach" + "TargetName" = "8:0C370E94.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_78792FEB009343479E361944DCAF9C36" { "SourcePath" = "8:..\\bin\\Langs\\sw_SW\\LC_MESSAGES\\pcsx2.mo" @@ -947,6 +3843,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_78F8BB67ADA242B1A949C41F909666E2" + { + "SourcePath" = "8:..\\bin\\patches\\200BC0E6.pnach" + "TargetName" = "8:200BC0E6.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_797C85AEE487426EA22D8B0253111FA3" { "SourcePath" = "8:..\\bin\\Langs\\pl_PL\\LC_MESSAGES\\pcsx2.mo" @@ -967,6 +3883,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7AB0994101C146EFA3493863882452B4" + { + "SourcePath" = "8:..\\bin\\patches\\625AF967.pnach" + "TargetName" = "8:625AF967.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7B327706377AE5F414299F6736D8EC00" { "SourcePath" = "8:DSOUND.dll" @@ -987,6 +3923,226 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7C0CEAF559A84A0C8723C9328DBADDAE" + { + "SourcePath" = "8:..\\bin\\patches\\0A342A88.pnach" + "TargetName" = "8:0A342A88.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7C346BC46D0B44BAB4750A35B70A0B46" + { + "SourcePath" = "8:..\\bin\\patches\\B0AE1898.pnach" + "TargetName" = "8:B0AE1898.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7D493307EEB847228F6CCEEE90D42DAA" + { + "SourcePath" = "8:..\\bin\\patches\\C1767D64.pnach" + "TargetName" = "8:C1767D64.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7E94D0A6FD0E48C8A7B7D8421F4D2DE6" + { + "SourcePath" = "8:..\\bin\\patches\\A36CFF6C.pnach" + "TargetName" = "8:A36CFF6C.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7EA7534B81A9415D89174750F58D27A2" + { + "SourcePath" = "8:..\\bin\\patches\\36FEEE3A.pnach" + "TargetName" = "8:36FEEE3A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7EDF89CE07114545909E655CC5567735" + { + "SourcePath" = "8:..\\bin\\patches\\7D8F539A.pnach" + "TargetName" = "8:7D8F539A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_7FF341438FB64A7A9A22CD9B9839C63C" + { + "SourcePath" = "8:..\\bin\\patches\\789D6B71.pnach" + "TargetName" = "8:789D6B71.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8196C46459C542B1A8EC49445C14472F" + { + "SourcePath" = "8:..\\bin\\patches\\B884B6E9.pnach" + "TargetName" = "8:B884B6E9.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_81D9FBCB8E3D4952A47B9BC4B2567BB2" + { + "SourcePath" = "8:..\\bin\\patches\\582EED0D.pnach" + "TargetName" = "8:582EED0D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_83305974B07F42C18E0AE99DBBF15037" + { + "SourcePath" = "8:..\\bin\\patches\\F22CDDAF.pnach" + "TargetName" = "8:F22CDDAF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_84F3D3D23A1447339492C40CC699DD44" + { + "SourcePath" = "8:..\\bin\\patches\\default.xml" + "TargetName" = "8:default.xml" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8743DB62D7EF4D72A820BAEC30E97DEC" { "SourcePath" = "8:..\\bin\\Langs\\sh_SH\\LC_MESSAGES\\pcsx2.mo" @@ -1047,6 +4203,406 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8835232945DD4F2C8333D0E2CACA9DF8" + { + "SourcePath" = "8:..\\bin\\patches\\E7A35274.pnach" + "TargetName" = "8:E7A35274.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8897C560DD094D698EA4A9B88C98D0A6" + { + "SourcePath" = "8:..\\bin\\patches\\62F6F886.pnach" + "TargetName" = "8:62F6F886.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_88FCE1823EA04CF8B1FE816A1A0E1E3B" + { + "SourcePath" = "8:..\\bin\\patches\\AD9D2B54.pnach" + "TargetName" = "8:AD9D2B54.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_89BA174439B14C8F8FA942C9E150B718" + { + "SourcePath" = "8:..\\bin\\patches\\F1370E83.pnach" + "TargetName" = "8:F1370E83.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8AB21318425441BA811D919F6DC5D6E8" + { + "SourcePath" = "8:..\\bin\\patches\\83D0CE43.pnach" + "TargetName" = "8:83D0CE43.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8AC15EDDB6994E7794DBA7A0A3DCA60B" + { + "SourcePath" = "8:..\\bin\\patches\\38F29E28.pnach" + "TargetName" = "8:38F29E28.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8C0EC3CEAEB04DE7B6729EBBCE8C0B36" + { + "SourcePath" = "8:..\\bin\\patches\\F266B00B.pnach" + "TargetName" = "8:F266B00B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8CF26805EE71422CBC1C152A7462FE44" + { + "SourcePath" = "8:..\\bin\\patches\\B1BE3E51.pnach" + "TargetName" = "8:B1BE3E51.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8E3C8EAC49BD49EE886229B91DF16166" + { + "SourcePath" = "8:..\\bin\\patches\\06DE61E0.pnach" + "TargetName" = "8:06DE61E0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8E909C0BE4C942A7804FF65AD2DDAFB5" + { + "SourcePath" = "8:..\\bin\\patches\\54AD76D7.pnach" + "TargetName" = "8:54AD76D7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_8F39494E0D4C4FF091091DAF5183388D" + { + "SourcePath" = "8:..\\bin\\patches\\49DA19CE.pnach" + "TargetName" = "8:49DA19CE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_90ED9EDA4346456F8957D9E7EB0D8DD9" + { + "SourcePath" = "8:..\\bin\\patches\\CB4EBD11.pnach" + "TargetName" = "8:CB4EBD11.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_91C427E57A3B48F399BAA6AF4A588293" + { + "SourcePath" = "8:..\\bin\\patches\\6926B199.pnach" + "TargetName" = "8:6926B199.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_924E5E9E11C245D383CFFA77545B69FE" + { + "SourcePath" = "8:..\\bin\\patches\\46A7ECA4.pnach" + "TargetName" = "8:46A7ECA4.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_9356233C1CB34B1E8F0781EA7188920C" + { + "SourcePath" = "8:..\\bin\\patches\\05177ECE.pnach" + "TargetName" = "8:05177ECE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_94DA87D95C7E4D009C9FFF2F647405C2" + { + "SourcePath" = "8:..\\bin\\patches\\D03BEF2A.pnach" + "TargetName" = "8:D03BEF2A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_95AA28E96394481AB463DEF29FAA78F8" + { + "SourcePath" = "8:..\\bin\\patches\\A8083AE6.pnach" + "TargetName" = "8:A8083AE6.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_9636D19CCBD441D993A061FDE5FF7807" + { + "SourcePath" = "8:..\\bin\\patches\\D0E17D26.pnach" + "TargetName" = "8:D0E17D26.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_96AB009303B44C2CB19C172EB5D51CE2" + { + "SourcePath" = "8:..\\bin\\patches\\C1625F14.pnach" + "TargetName" = "8:C1625F14.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_98FAE27F8FDB404098DC9E605F753E0D" + { + "SourcePath" = "8:..\\bin\\patches\\48FE0C71.pnach" + "TargetName" = "8:48FE0C71.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_99C23BC9B7C5EEDB5AD1079051D199C3" { "SourcePath" = "8:AVIFIL32.dll" @@ -1067,12 +4623,152 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_99C9A286D959DE56EFE6671DE8C65FAE" + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_9C7D1B3E93F0481E804E68339BBEB021" { - "SourcePath" = "8:w32pthreads.dll" - "TargetName" = "8:w32pthreads.dll" + "SourcePath" = "8:..\\bin\\patches\\506644B3.pnach" + "TargetName" = "8:506644B3.pnach" "Tag" = "8:" - "Folder" = "8:_A396E583ED3A43A5B9FA30BABB3624E2" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_9D1F177B2FFA4D3D84F11A8A20D1F611" + { + "SourcePath" = "8:..\\bin\\patches\\4C4D7072.pnach" + "TargetName" = "8:4C4D7072.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_9E96BFE206504BFEBF955CA9CBA846DC" + { + "SourcePath" = "8:..\\bin\\patches\\8BC95883.pnach" + "TargetName" = "8:8BC95883.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A2652446B2604985A24F992F42F2761E" + { + "SourcePath" = "8:..\\bin\\patches\\DEFA4763.pnach" + "TargetName" = "8:DEFA4763.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A28EBA4DE8914C988801F829035FFEC8" + { + "SourcePath" = "8:..\\bin\\patches\\3EAD47FE.pnach" + "TargetName" = "8:3EAD47FE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A3D55F63E0484B52BD23FBF66915D43D" + { + "SourcePath" = "8:..\\bin\\patches\\E138094A.pnach" + "TargetName" = "8:E138094A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A40F34AB634C48D7B953C2E420915F07" + { + "SourcePath" = "8:..\\bin\\patches\\95BB1901.pnach" + "TargetName" = "8:95BB1901.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A6C0DAAB1BFF83F078FD451B8C7A5594" + { + "SourcePath" = "8:DDRAW.dll" + "TargetName" = "8:DDRAW.dll" + "Tag" = "8:" + "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" "Condition" = "8:" "Transitive" = "11:FALSE" "Vital" = "11:TRUE" @@ -1107,6 +4803,326 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A8F09DC7C504415BA06801BA7C5D750B" + { + "SourcePath" = "8:..\\bin\\patches\\C5DEFEA0.pnach" + "TargetName" = "8:C5DEFEA0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_AAB10625A4634D85821952D2CD42642B" + { + "SourcePath" = "8:..\\bin\\patches\\561BE340.pnach" + "TargetName" = "8:561BE340.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_AB28BFBD30FC423293D42423F112B6BD" + { + "SourcePath" = "8:..\\bin\\patches\\A7A2F7C5.pnach" + "TargetName" = "8:A7A2F7C5.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_ADAC04E54F894404A83B379617E7A97E" + { + "SourcePath" = "8:..\\bin\\patches\\94C56923.pnach" + "TargetName" = "8:94C56923.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_ADE00C6618D141C49F057DA07B235947" + { + "SourcePath" = "8:..\\bin\\patches\\A1B752C7.pnach" + "TargetName" = "8:A1B752C7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_ADE37DD61B5347DB890F2957492F3745" + { + "SourcePath" = "8:..\\bin\\patches\\9BE3F92D.pnach" + "TargetName" = "8:9BE3F92D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_AEB84272F38444F5BB708EF3E0833776" + { + "SourcePath" = "8:..\\bin\\patches\\A8A76AAC.pnach" + "TargetName" = "8:A8A76AAC.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_AF7C7BE64EA846468642088F15DA4A1F" + { + "SourcePath" = "8:..\\bin\\patches\\BF5D9AEC.pnach" + "TargetName" = "8:BF5D9AEC.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B2A5C84577F64477896DD21845527BD0" + { + "SourcePath" = "8:..\\bin\\patches\\1248FE3A.pnach" + "TargetName" = "8:1248FE3A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B2F1836277ED493C958A02184D327E49" + { + "SourcePath" = "8:..\\bin\\patches\\1CAC8A56.pnach" + "TargetName" = "8:1CAC8A56.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B4A1253749E245B088BD031A677CFB4A" + { + "SourcePath" = "8:..\\bin\\patches\\5CCA0737.pnach" + "TargetName" = "8:5CCA0737.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B50F62EC90CD4A07912A63B90AAD3293" + { + "SourcePath" = "8:..\\bin\\patches\\51F91783.pnach" + "TargetName" = "8:51F91783.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B63084B7B255496B836164646F5D536D" + { + "SourcePath" = "8:..\\bin\\patches\\B338676A.pnach" + "TargetName" = "8:B338676A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B8FEC8F89AD14BE5AF941C226AAA31AF" + { + "SourcePath" = "8:..\\bin\\patches\\3F0452DE.pnach" + "TargetName" = "8:3F0452DE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_B90FC469B0A7444AA681C869AB6704D4" + { + "SourcePath" = "8:..\\bin\\patches\\901AAC09.pnach" + "TargetName" = "8:901AAC09.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_BC30361C69EF4CD8869553D0AF1CA80A" + { + "SourcePath" = "8:..\\bin\\patches\\CA6243B9.pnach" + "TargetName" = "8:CA6243B9.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_BCED2304BC124DC79A400E6EC3DF991A" { "SourcePath" = "8:..\\bin\\Langs\\de_DE\\LC_MESSAGES\\pcsx2.mo" @@ -1127,6 +5143,146 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_BD602A7A1C334F74815A52DF4C98D0CD" + { + "SourcePath" = "8:..\\bin\\patches\\4D228733.pnach" + "TargetName" = "8:4D228733.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_BDA20A14A241440B8632ABEBC26E35DB" + { + "SourcePath" = "8:..\\bin\\patches\\2F56CBC9.pnach" + "TargetName" = "8:2F56CBC9.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_BFE971F656CA46C7A298C91757E9AFF7" + { + "SourcePath" = "8:..\\bin\\patches\\086273D2.pnach" + "TargetName" = "8:086273D2.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C047CC641D1E4EAF9CF3032B69B49C46" + { + "SourcePath" = "8:..\\bin\\patches\\44A61C8F.pnach" + "TargetName" = "8:44A61C8F.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C10D7A1FC3A44F749156F781C6DB3140" + { + "SourcePath" = "8:..\\bin\\patches\\C502AD6E.pnach" + "TargetName" = "8:C502AD6E.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C188D42682084A7D983722A92046C3EA" + { + "SourcePath" = "8:..\\bin\\patches\\C398F477.pnach" + "TargetName" = "8:C398F477.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C40BE5E85657415296E733A2F29C985A" + { + "SourcePath" = "8:..\\bin\\patches\\A5BF36A8.pnach" + "TargetName" = "8:A5BF36A8.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C46C31F5523D72D416CDEF1804E2A64D" { "SourcePath" = "8:d3dx9_41.dll" @@ -1147,6 +5303,166 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C555162941C34AE1B51CE4CB5352CD4E" + { + "SourcePath" = "8:..\\bin\\patches\\93F8A60B.pnach" + "TargetName" = "8:93F8A60B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C77702CF400646678BD9CC9242B1B6BB" + { + "SourcePath" = "8:..\\bin\\patches\\78168525.pnach" + "TargetName" = "8:78168525.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C80AC3C16BED483D8228AFA835D7F15F" + { + "SourcePath" = "8:..\\bin\\patches\\4C0C821D.pnach" + "TargetName" = "8:4C0C821D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C82D0E331C0F4C368F91EE2A7BD0565D" + { + "SourcePath" = "8:..\\bin\\patches\\BB3D833A.pnach" + "TargetName" = "8:BB3D833A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_C980F539DB294419AE3FC77C33A77AEF" + { + "SourcePath" = "8:..\\bin\\patches\\D4781770.pnach" + "TargetName" = "8:D4781770.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_CCDA5668BC2E426E9E05F8875B752536" + { + "SourcePath" = "8:..\\bin\\patches\\6BA2F6B9.pnach" + "TargetName" = "8:6BA2F6B9.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_CDAA2F2C190A4DCB928E78564F76BCA3" + { + "SourcePath" = "8:..\\bin\\patches\\6175FE7D.pnach" + "TargetName" = "8:6175FE7D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_CE6A0F7079024AF4A0933BB7F9BFE195" + { + "SourcePath" = "8:..\\bin\\patches\\68EAF48F.pnach" + "TargetName" = "8:68EAF48F.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D11A044D40114C19A0FDA835B19D37AF" { "SourcePath" = "8:..\\bin\\Langs\\ar_AR\\LC_MESSAGES\\pcsx2.mo" @@ -1167,6 +5483,46 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D21F33B9B9CA4F28996FF4CB02872772" + { + "SourcePath" = "8:..\\bin\\patches\\22C36E63.pnach" + "TargetName" = "8:22C36E63.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D2474D58E6AC4CA0A5FCEF9B4BA6B3E4" + { + "SourcePath" = "8:..\\bin\\patches\\C220951A.pnach" + "TargetName" = "8:C220951A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D31ED5CC46194EA6A81F9F9BE3DE5CD1" { "SourcePath" = "8:..\\bin\\Langs\\sw_SW\\LC_MESSAGES\\other.po" @@ -1187,6 +5543,166 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D4219ACEADB94026AA93732F128494E3" + { + "SourcePath" = "8:..\\bin\\patches\\25433CBD.pnach" + "TargetName" = "8:25433CBD.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D738FF1390814B24A9021A19F95E4290" + { + "SourcePath" = "8:..\\bin\\patches\\304C115C.pnach" + "TargetName" = "8:304C115C.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D81E0DB3F5674742AB46156B10CCC968" + { + "SourcePath" = "8:..\\bin\\patches\\D4FB6049.pnach" + "TargetName" = "8:D4FB6049.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D92503794DA8471A8BFE082269932306" + { + "SourcePath" = "8:..\\bin\\patches\\C488EC04.pnach" + "TargetName" = "8:C488EC04.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_D980AD62DE974BABA15438FF6ACFA8BF" + { + "SourcePath" = "8:..\\bin\\patches\\B99379B7.pnach" + "TargetName" = "8:B99379B7.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DAC88B65E7EB4A8E92FBF82174072D28" + { + "SourcePath" = "8:..\\bin\\patches\\E0DADD1A.pnach" + "TargetName" = "8:E0DADD1A.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DB95F4450EFC4AAE94E1A2EDE746B2B2" + { + "SourcePath" = "8:..\\bin\\patches\\F52FB2BE.pnach" + "TargetName" = "8:F52FB2BE.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DBB4560AE2924E349948FA5BB09FE28E" + { + "SourcePath" = "8:..\\bin\\patches\\B0621C55.pnach" + "TargetName" = "8:B0621C55.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DCA1A5B0C13D49259445C449B0E50774" { "SourcePath" = "8:..\\bin\\Langs\\it_IT\\LC_MESSAGES\\pcsx2.mo" @@ -1207,6 +5723,86 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DD150D2502704E67A072908310DB66F3" + { + "SourcePath" = "8:..\\bin\\patches\\37BA81B1.pnach" + "TargetName" = "8:37BA81B1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_DFE2B52C9EDF48DA81BC3E67D6564B0D" + { + "SourcePath" = "8:..\\bin\\patches\\0518D274.pnach" + "TargetName" = "8:0518D274.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E02F836E90974BFC8E673705BE975E95" + { + "SourcePath" = "8:..\\bin\\patches\\08901101.pnach" + "TargetName" = "8:08901101.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E09F4497BBF14351A2528D6E215DE28D" + { + "SourcePath" = "8:..\\bin\\patches\\129C8600.pnach" + "TargetName" = "8:129C8600.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E0E1B281255A4DC5934647A82D674F99" { "SourcePath" = "8:..\\bin\\Langs\\tc_TC\\LC_MESSAGES\\Pcsx2.mo" @@ -1227,6 +5823,46 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E17B0DADA8FC4512A4888DE697BC75B3" + { + "SourcePath" = "8:..\\bin\\patches\\F4992CC1.pnach" + "TargetName" = "8:F4992CC1.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E508524696DD452E91B739C7EC466374" + { + "SourcePath" = "8:..\\bin\\patches\\FA7E3081.pnach" + "TargetName" = "8:FA7E3081.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E6DE12DF8D434183B93158C458D72A61" { "SourcePath" = "8:..\\bin\\Langs\\ro_RO\\LC_MESSAGES\\pcsx2.mo" @@ -1247,6 +5883,66 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_E71C997965964853BA87889843181711" + { + "SourcePath" = "8:..\\bin\\patches\\EEC3B310.pnach" + "TargetName" = "8:EEC3B310.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_ECC63C4CBBE44D568436E4312BA47908" + { + "SourcePath" = "8:..\\bin\\patches\\EEE2F6A3.pnach" + "TargetName" = "8:EEE2F6A3.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_ECD3190AB8F14933A35236EC20488C61" + { + "SourcePath" = "8:..\\bin\\patches\\0F6B6315.pnach" + "TargetName" = "8:0F6B6315.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_EEECC8B200772D4A9F89A5F65A3B97F5" { "SourcePath" = "8:XINPUT1_3.dll" @@ -1267,6 +5963,266 @@ "IsDependency" = "11:TRUE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F00371D156F84024BD88562FE8C9DDBF" + { + "SourcePath" = "8:..\\bin\\patches\\9B1EE9EB.pnach" + "TargetName" = "8:9B1EE9EB.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F165D6BEADEC6B529C157DDC87667E80" + { + "SourcePath" = "8:w32pthreads.dll" + "TargetName" = "8:w32pthreads.dll" + "Tag" = "8:" + "Folder" = "8:_A396E583ED3A43A5B9FA30BABB3624E2" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:TRUE" + "IsDependency" = "11:TRUE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F1A181FCDB444332B63B8E51A4EBD448" + { + "SourcePath" = "8:..\\bin\\patches\\4334E17D.pnach" + "TargetName" = "8:4334E17D.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F29E17EBECD5437C9244B842F5038BB1" + { + "SourcePath" = "8:..\\bin\\patches\\E2F1DB6B.pnach" + "TargetName" = "8:E2F1DB6B.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F43E0A4CFF22404996751E7EE5CFC3DE" + { + "SourcePath" = "8:..\\bin\\patches\\AE3EAA05.pnach" + "TargetName" = "8:AE3EAA05.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F58D989DCC40429EBC0C9D43AF5D8BA1" + { + "SourcePath" = "8:..\\bin\\patches\\2C728173.pnach" + "TargetName" = "8:2C728173.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F6AF071E4E7F4C6CBB28DCDAA6A711C3" + { + "SourcePath" = "8:..\\bin\\patches\\54A548B4.pnach" + "TargetName" = "8:54A548B4.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F72D772B69424F28A3E406E8BCD1878B" + { + "SourcePath" = "8:..\\bin\\patches\\37C07E96.pnach" + "TargetName" = "8:37C07E96.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F84D270966164BAAA768673FDACED3CC" + { + "SourcePath" = "8:..\\bin\\patches\\21068223.pnach" + "TargetName" = "8:21068223.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_F9276AD7A6E840C59B3B10CD18910D35" + { + "SourcePath" = "8:..\\bin\\patches\\4F3D3CF0.pnach" + "TargetName" = "8:4F3D3CF0.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_FBBAFABEF4884CCA881660F376B5F6FE" + { + "SourcePath" = "8:..\\bin\\patches\\A929A697.pnach" + "TargetName" = "8:A929A697.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_FCB2559A1E524CAFA362E24808358B68" + { + "SourcePath" = "8:..\\bin\\patches\\53A803AF.pnach" + "TargetName" = "8:53A803AF.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_FF64089169014510AEEBB6F516BFE6D5" + { + "SourcePath" = "8:..\\bin\\patches\\5D67AE48.pnach" + "TargetName" = "8:5D67AE48.pnach" + "Tag" = "8:" + "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + } } "FileType" { @@ -1284,7 +6240,7 @@ { "{9EF0B969-E518-4E46-987F-47570745A589}:_E300D7C3BDA3486FB06E51133B66D4E1" { - "Name" = "8:PCSX2" + "Name" = "8:[ProductName]" "AlwaysCreate" = "11:FALSE" "Condition" = "8:" "Transitive" = "11:FALSE" @@ -1297,7 +6253,7 @@ } "{3C67513D-01DD-4637-8A68-80971EB9504F}:_A396E583ED3A43A5B9FA30BABB3624E2" { - "DefaultLocation" = "8:[ProgramFilesFolder][Manufacturer]\\[ProductName]" + "DefaultLocation" = "8:[ProgramFilesFolder]\\[ProductName]" "Name" = "8:#1925" "AlwaysCreate" = "11:FALSE" "Condition" = "8:" @@ -1907,7 +6863,7 @@ "Name" = "8:Microsoft Visual Studio" "ProductName" = "8:PCSX2" "ProductCode" = "8:{69F747B1-6251-4490-A3B4-AFADB306C112}" - "PackageCode" = "8:{894BBDF0-E030-4FE4-B213-2A4821043EDE}" + "PackageCode" = "8:{912DBF5E-E3BA-4797-BB29-02754B5B5813}" "UpgradeCode" = "8:{5237A49B-FA2E-47C4-8C0E-92338FDC9F2A}" "RestartWWWService" = "11:FALSE" "RemovePreviousVersions" = "11:FALSE" @@ -1917,12 +6873,12 @@ "Manufacturer" = "8:PCSX2 Team" "ARPHELPTELEPHONE" = "8:" "ARPHELPLINK" = "8:" - "Title" = "8:PCSX2 Intaller" - "Subject" = "8:" + "Title" = "8:PCSX2 Installer" + "Subject" = "8:Playstation 2 Emulator" "ARPCONTACT" = "8:PCSX2 Team" "Keywords" = "8:" - "ARPCOMMENTS" = "8:" - "ARPURLINFOABOUT" = "8:http://pcsx2.net/" + "ARPCOMMENTS" = "8:Official Installer for PCSX2 Emulator and Standard Plugins " + "ARPURLINFOABOUT" = "8:http://www.pcsx2.net/" "ARPPRODUCTICON" = "8:" "ARPIconIndex" = "3:0" "SearchPath" = "8:" @@ -1949,7 +6905,7 @@ { "{60EA8692-D2D5-43EB-80DC-7906BF13D6EF}:_C54E3A3195AE4A7FBC022E0BC47E4D68" { - "Name" = "8:[Manufacturer]" + "Name" = "8:[ProductName]" "Condition" = "8:" "AlwaysCreate" = "11:FALSE" "DeleteAtUninstall" = "11:FALSE" @@ -1983,7 +6939,7 @@ { "{60EA8692-D2D5-43EB-80DC-7906BF13D6EF}:_54C3FD07DA564F13AEBFBAC3BD00C177" { - "Name" = "8:[Manufacturer]" + "Name" = "8:[ProductName]" "Condition" = "8:" "AlwaysCreate" = "11:FALSE" "DeleteAtUninstall" = "11:FALSE" @@ -2450,62 +7406,6 @@ } "MergeModule" { - "{CEE29DC0-9FBA-4B99-8D47-5BC643D9B626}:_22027664F1B34E9EA39D3B330D460FEC" - { - "UseDynamicProperties" = "11:TRUE" - "IsDependency" = "11:TRUE" - "SourcePath" = "8:policy_9_0_Microsoft_VC90_MFC_x86.msm" - "Properties" - { - } - "LanguageId" = "3:0" - "Exclude" = "11:FALSE" - "Folder" = "8:" - "Feature" = "8:" - "IsolateTo" = "8:" - } - "{CEE29DC0-9FBA-4B99-8D47-5BC643D9B626}:_919458435B414CF3A23DEBA86D49CE7D" - { - "UseDynamicProperties" = "11:TRUE" - "IsDependency" = "11:TRUE" - "SourcePath" = "8:microsoft_vc90_crt_x86.msm" - "Properties" - { - } - "LanguageId" = "3:0" - "Exclude" = "11:FALSE" - "Folder" = "8:" - "Feature" = "8:" - "IsolateTo" = "8:" - } - "{CEE29DC0-9FBA-4B99-8D47-5BC643D9B626}:_CAED6729570B459EBDCE72B9E32AB03D" - { - "UseDynamicProperties" = "11:TRUE" - "IsDependency" = "11:TRUE" - "SourcePath" = "8:microsoft_vc90_mfc_x86.msm" - "Properties" - { - } - "LanguageId" = "3:0" - "Exclude" = "11:FALSE" - "Folder" = "8:" - "Feature" = "8:" - "IsolateTo" = "8:" - } - "{CEE29DC0-9FBA-4B99-8D47-5BC643D9B626}:_E619C79FD5E14F6B965446BB0371AFDE" - { - "UseDynamicProperties" = "11:TRUE" - "IsDependency" = "11:TRUE" - "SourcePath" = "8:policy_9_0_Microsoft_VC90_CRT_x86.msm" - "Properties" - { - } - "LanguageId" = "3:0" - "Exclude" = "11:FALSE" - "Folder" = "8:" - "Feature" = "8:" - "IsolateTo" = "8:" - } } "ProjectOutput" { @@ -2537,6 +7437,34 @@ { } } + "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_1FA9731E400641839E29839952B10A67" + { + "SourcePath" = "8:..\\bin\\plugins\\GSdx-SSSE3.dll" + "TargetName" = "8:" + "Tag" = "8:" + "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + "ProjectOutputGroupRegister" = "3:1" + "OutputConfiguration" = "8:Release SSSE3|Win32" + "OutputGroupCanonicalName" = "8:Built" + "OutputProjectGuid" = "8:{18E42F6F-3A62-41EE-B42F-79366C4F1E95}" + "ShowKeyOutput" = "11:TRUE" + "ExcludeFilters" + { + } + } "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_25EDB5C49F2745F199D2CF1036A1207F" { "SourcePath" = "8:..\\bin\\pcsx2.exe" @@ -2621,6 +7549,34 @@ { } } + "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_8037556EAB9F4BF298A6AA2BDA834930" + { + "SourcePath" = "8:..\\bin\\plugins\\GSdx-SSE2.dll" + "TargetName" = "8:" + "Tag" = "8:" + "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + "ProjectOutputGroupRegister" = "3:1" + "OutputConfiguration" = "8:Release SSE2|Win32" + "OutputGroupCanonicalName" = "8:Built" + "OutputProjectGuid" = "8:{18E42F6F-3A62-41EE-B42F-79366C4F1E95}" + "ShowKeyOutput" = "11:TRUE" + "ExcludeFilters" + { + } + } "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_87A61D50A0B944458FE3B315A47FA221" { "SourcePath" = "8:..\\bin\\plugins\\SPU2-X.dll" @@ -2817,34 +7773,6 @@ { } } - "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_BCBBD06CDE4641D2B093302D05CB9004" - { - "SourcePath" = "8:..\\bin\\plugins\\GSdx.dll" - "TargetName" = "8:" - "Tag" = "8:" - "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" - "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" - "ReadOnly" = "11:FALSE" - "Hidden" = "11:FALSE" - "System" = "11:FALSE" - "Permanent" = "11:FALSE" - "SharedLegacy" = "11:FALSE" - "PackageAs" = "3:1" - "Register" = "3:1" - "Exclude" = "11:FALSE" - "IsDependency" = "11:FALSE" - "IsolateTo" = "8:" - "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" - "OutputGroupCanonicalName" = "8:Built" - "OutputProjectGuid" = "8:{18E42F6F-3A62-41EE-B42F-79366C4F1E95}" - "ShowKeyOutput" = "11:TRUE" - "ExcludeFilters" - { - } - } "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_C28E286C209D4FBCBBF1FD46837E0854" { "SourcePath" = "8:..\\bin\\plugins\\ZeroPAD.dll" @@ -2873,6 +7801,34 @@ { } } + "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_C5FEBD35C45E4EAE824FD02E9F023FD0" + { + "SourcePath" = "8:..\\bin\\plugins\\GSdx-SSE4.dll" + "TargetName" = "8:" + "Tag" = "8:" + "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:FALSE" + "IsDependency" = "11:FALSE" + "IsolateTo" = "8:" + "ProjectOutputGroupRegister" = "3:1" + "OutputConfiguration" = "8:Release SSE4|Win32" + "OutputGroupCanonicalName" = "8:Built" + "OutputProjectGuid" = "8:{18E42F6F-3A62-41EE-B42F-79366C4F1E95}" + "ShowKeyOutput" = "11:TRUE" + "ExcludeFilters" + { + } + } "{5259A561-127C-4D43-A0A1-72F10C7B3BF8}:_E663A731DB91460B991DAA889BD480EE" { "SourcePath" = "8:..\\bin\\plugins\\ZeroSPU2.dll" From 7ad2d919b88692877843c1115b011b6d860fa04d Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 19 Jun 2009 00:11:56 +0000 Subject: [PATCH 252/296] GSdx: removed the last bit of mfc dependency from cdvdolio git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1398 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/CDVDolio/CDVDDialog.cpp | 177 +++++ plugins/CDVDolio/CDVDDialog.h | 59 ++ plugins/CDVDolio/SettingsDlg.cpp | 255 +++---- plugins/CDVDolio/SettingsDlg.h | 33 +- plugins/CDVDolio/cdvd.cpp | 172 +++-- plugins/CDVDolio/cdvd.def | 2 +- plugins/CDVDolio/cdvd.h | 83 ++- plugins/CDVDolio/cdvd_vs2008.vcproj | 1044 ++++++++++++++------------- plugins/CDVDolio/stdafx.cpp | 31 + plugins/CDVDolio/stdafx.h | 90 ++- plugins/GSdx/GPUSettingsDlg.cpp | 30 +- plugins/GSdx/GSCaptureDlg.cpp | 8 +- plugins/GSdx/GSDialog.cpp | 16 +- plugins/GSdx/GSDialog.h | 6 +- plugins/GSdx/GSSettingsDlg.cpp | 22 +- 15 files changed, 1154 insertions(+), 874 deletions(-) create mode 100644 plugins/CDVDolio/CDVDDialog.cpp create mode 100644 plugins/CDVDolio/CDVDDialog.h diff --git a/plugins/CDVDolio/CDVDDialog.cpp b/plugins/CDVDolio/CDVDDialog.cpp new file mode 100644 index 0000000000..b83f02a28c --- /dev/null +++ b/plugins/CDVDolio/CDVDDialog.cpp @@ -0,0 +1,177 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "StdAfx.h" +#include "CDVD.h" +#include "CDVDDialog.h" + +CDVDDialog::CDVDDialog(UINT id) + : m_id(id) + , m_hWnd(NULL) +{ +} + +INT_PTR CDVDDialog::DoModal() +{ + return DialogBoxParam(theApp.GetModuleHandle(), MAKEINTRESOURCE(m_id), NULL, DialogProc, (LPARAM)this); +} + +INT_PTR CALLBACK CDVDDialog::DialogProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) +{ + CDVDDialog* dlg = NULL; + + if(message == WM_INITDIALOG) + { + dlg = (CDVDDialog*)lParam; + SetWindowLongPtr(hWnd, GWL_USERDATA, (LONG_PTR)dlg); + dlg->m_hWnd = hWnd; + + MONITORINFO mi; + mi.cbSize = sizeof(mi); + GetMonitorInfo(MonitorFromWindow(hWnd, MONITOR_DEFAULTTONEAREST), &mi); + + RECT r; + GetWindowRect(hWnd, &r); + + int x = (mi.rcWork.left + mi.rcWork.right - (r.right - r.left)) / 2; + int y = (mi.rcWork.top + mi.rcWork.bottom - (r.bottom - r.top)) / 2; + + SetWindowPos(hWnd, NULL, x, y, -1, -1, SWP_NOSIZE | SWP_NOZORDER | SWP_NOACTIVATE); + + dlg->OnInit(); + + return true; + } + + dlg = (CDVDDialog*)GetWindowLongPtr(hWnd, GWL_USERDATA); + + return dlg != NULL ? dlg->OnMessage(message, wParam, lParam) : FALSE; +} + +bool CDVDDialog::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) +{ + return message == WM_COMMAND ? OnCommand((HWND)lParam, LOWORD(wParam), HIWORD(wParam)) : false; +} + +bool CDVDDialog::OnCommand(HWND hWnd, UINT id, UINT code) +{ + if(id == IDOK || id == IDCANCEL) + { + EndDialog(m_hWnd, id); + + return true; + } + + return false; +} + +string CDVDDialog::GetText(UINT id) +{ + string s; + + char* buff = NULL; + + for(int size = 256, limit = 65536; size < limit; size <<= 1) + { + buff = new char[size]; + + if(GetDlgItemText(m_hWnd, id, buff, size)) + { + s = buff; + size = limit; + } + + delete [] buff; + } + + return s; +} + +int CDVDDialog::GetTextAsInt(UINT id) +{ + return atoi(GetText(id).c_str()); +} + +void CDVDDialog::SetText(UINT id, const char* str) +{ + SetDlgItemText(m_hWnd, id, str); +} + +void CDVDDialog::SetTextAsInt(UINT id, int i) +{ + char buff[32] = {0}; + itoa(i, buff, 10); + SetText(id, buff); +} + +void CDVDDialog::ComboBoxInit(UINT id, const CDVDSetting* settings, int count, uint32 selid, uint32 maxid) +{ + HWND hWnd = GetDlgItem(m_hWnd, id); + + SendMessage(hWnd, CB_RESETCONTENT, 0, 0); + + for(int i = 0; i < count; i++) + { + if(settings[i].id <= maxid) + { + string str = settings[i].name; + + if(!settings[i].note.empty()) + { + str = str + " (" + settings[i].note + ")"; + } + + ComboBoxAppend(id, str.c_str(), (LPARAM)settings[i].id, settings[i].id == selid); + } + } +} + +int CDVDDialog::ComboBoxAppend(UINT id, const char* str, LPARAM data, bool select) +{ + HWND hWnd = GetDlgItem(m_hWnd, id); + + int item = (int)SendMessage(hWnd, CB_ADDSTRING, 0, (LPARAM)str); + + SendMessage(hWnd, CB_SETITEMDATA, item, (LPARAM)data); + + if(select) + { + SendMessage(hWnd, CB_SETCURSEL, item, 0); + } + + return item; +} + +bool CDVDDialog::ComboBoxGetSelData(UINT id, INT_PTR& data) +{ + HWND hWnd = GetDlgItem(m_hWnd, id); + + int item = SendMessage(hWnd, CB_GETCURSEL, 0, 0); + + if(item >= 0) + { + data = SendMessage(hWnd, CB_GETITEMDATA, item, 0); + + return true; + } + + return false; +} diff --git a/plugins/CDVDolio/CDVDDialog.h b/plugins/CDVDolio/CDVDDialog.h new file mode 100644 index 0000000000..0abd02d4df --- /dev/null +++ b/plugins/CDVDolio/CDVDDialog.h @@ -0,0 +1,59 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +struct CDVDSetting +{ + uint32 id; + string name; + string note; +}; + +class CDVDDialog +{ + int m_id; + + static INT_PTR CALLBACK DialogProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam); + +protected: + HWND m_hWnd; + + virtual void OnInit() {} + virtual bool OnMessage(UINT message, WPARAM wParam, LPARAM lParam); + virtual bool OnCommand(HWND hWnd, UINT id, UINT code); + +public: + CDVDDialog (UINT id); + virtual ~CDVDDialog () {} + + INT_PTR DoModal(); + + string GetText(UINT id); + int GetTextAsInt(UINT id); + + void SetText(UINT id, const char* str); + void SetTextAsInt(UINT id, int i); + + void ComboBoxInit(UINT id, const CDVDSetting* settings, int count, uint32 selid, uint32 maxid = ~0); + int ComboBoxAppend(UINT id, const char* str, LPARAM data = 0, bool select = false); + bool ComboBoxGetSelData(UINT id, INT_PTR& data); +}; diff --git a/plugins/CDVDolio/SettingsDlg.cpp b/plugins/CDVDolio/SettingsDlg.cpp index 57f6281fc0..d1bb7c3377 100644 --- a/plugins/CDVDolio/SettingsDlg.cpp +++ b/plugins/CDVDolio/SettingsDlg.cpp @@ -2,193 +2,156 @@ // #include "stdafx.h" -#include "cdvd.h" +#include "CDVD.h" #include "SettingsDlg.h" #include -#include -// CSettingsDlg dialog - -IMPLEMENT_DYNAMIC(CSettingsDlg, CDialog) - -CSettingsDlg::CSettingsDlg(CWnd* pParent /*=NULL*/) - : CDialog(CSettingsDlg::IDD, pParent) - , m_iso(_T("")) -{ - -} - -CSettingsDlg::~CSettingsDlg() +CDVDSettingsDlg::CDVDSettingsDlg() + : CDVDDialog(IDD_CONFIG) { } -BOOL CSettingsDlg::OnInitDialog() +void CDVDSettingsDlg::OnInit() { - __super::OnInitDialog(); + __super::OnInit(); - InitDrive(); - InitISO(); + UpdateDrives(); - UpdateData(FALSE); - - return TRUE; // return TRUE unless you set the focus to a control - // EXCEPTION: OCX Property Pages should return FALSE + SetText(IDC_EDIT1, theApp.GetConfig("iso", "").c_str()); } -void CSettingsDlg::InitDrive() +bool CDVDSettingsDlg::OnMessage(UINT message, WPARAM wParam, LPARAM lParam) { - int drive = AfxGetApp()->GetProfileInt(_T("Settings"), _T("drive"), -1); - - int sel = m_drive.GetCurSel(); - - if(sel >= 0) + if(message == WM_DEVICECHANGE && (wParam == DBT_DEVICEARRIVAL || wParam == DBT_DEVICEREMOVECOMPLETE)) { - drive = m_drive.GetItemData(sel); + UpdateDrives(); + + DEV_BROADCAST_HDR* p = (DEV_BROADCAST_HDR*)lParam; + + if(p->dbch_devicetype == DBT_DEVTYP_VOLUME) + { + DEV_BROADCAST_VOLUME* v = (DEV_BROADCAST_VOLUME*)p; + + for(int i = 0; i < 32; i++) + { + if(v->dbcv_unitmask & (1 << i)) + { + // printf("%c:\n", 'A' + i); + + // TODO + } + } + } } - while(m_drive.GetCount() > 0) + return __super::OnMessage(message, wParam, lParam); +} + +bool CDVDSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) +{ + if(id == IDOK) { - m_drive.DeleteString(0); + INT_PTR data = 0; + + if(!ComboBoxGetSelData(IDC_COMBO1, data)) + { + data = -1; + } + + theApp.SetConfig("drive", (int)data); + + theApp.SetConfig("iso", GetText(IDC_EDIT1).c_str()); } + else if(id == IDC_BUTTON1 && code == BN_CLICKED) + { + char buff[MAX_PATH] = {0}; + + OPENFILENAME ofn; + + memset(&ofn, 0, sizeof(ofn)); + + ofn.lStructSize = sizeof(ofn); + ofn.hwndOwner = m_hWnd; + ofn.lpstrFile = buff; + ofn.nMaxFile = countof(buff); + ofn.lpstrFilter = "ISO file\0*.iso\0All files\0*.*\0"; + ofn.Flags = OFN_EXPLORER | OFN_ENABLESIZING | OFN_PATHMUSTEXIST | OFN_FILEMUSTEXIST | OFN_NOCHANGEDIR; + + strcpy(ofn.lpstrFile, GetText(IDC_EDIT1).c_str()); + + if(GetOpenFileName(&ofn)) + { + SetText(IDC_EDIT1, ofn.lpstrFile); + + HWND hWnd = GetDlgItem(m_hWnd, IDC_COMBO1); + + SendMessage(hWnd, CB_SETCURSEL, SendMessage(hWnd, CB_GETCOUNT, 0, 0) - 1, 0); + } + + return true; + } + + return __super::OnCommand(hWnd, id, code); +} + +void CDVDSettingsDlg::UpdateDrives() +{ + int drive = theApp.GetConfig("drive", -1); + + INT_PTR data = 0; + + if(ComboBoxGetSelData(IDC_COMBO1, data)) + { + drive = (int)data; + } + + vector drives; for(int i = 'A'; i <= 'Z'; i++) { - CString path; + string path = format("%c:", i); - path.Format(_T("%c:"), i); - - if(GetDriveType(path) == DRIVE_CDROM) + if(GetDriveType(path.c_str()) == DRIVE_CDROM) { - CString label = path; + string label = path; - path.Format(_T("\\\\.\\%c:"), i); + path = format("\\\\.\\%c:", i); CDVD cdvd; - if(cdvd.Open(path)) + if(cdvd.Open(path.c_str())) { - CString str = cdvd.GetLabel(); + string str = cdvd.GetLabel(); - if(str.IsEmpty()) + if(str.empty()) { - str = _T("(no label)"); + str = "(no label)"; } - label.Format(_T("[%s] %s"), CString(label), str); + label = "[" + label + "] " + str; } else { - label.Format(_T("[%s] (not detected)"), CString(label)); + label = "[" + label + "] (not detected)"; } - m_drive.SetItemData(m_drive.AddString(label), (DWORD_PTR)i); + CDVDSetting s; + + s.id = i; + s.name = label; + + drives.push_back(s); } } - m_drive.SetItemData(m_drive.AddString(_T("Other...")), (DWORD_PTR)-1); - - for(int i = 0, j = m_drive.GetCount(); i < j; i++) { - if((int)m_drive.GetItemData(i) == drive) - { - m_drive.SetCurSel(i); + CDVDSetting s; - return; - } + s.id = -1; + s.name = "Other..."; + + drives.push_back(s); } - m_drive.SetCurSel(-1); -} - -void CSettingsDlg::InitISO() -{ - m_iso = AfxGetApp()->GetProfileString(_T("Settings"), _T("iso"), _T("")); -} - -void CSettingsDlg::DoDataExchange(CDataExchange* pDX) -{ - CDialog::DoDataExchange(pDX); - DDX_Control(pDX, IDC_COMBO1, m_drive); - DDX_Text(pDX, IDC_EDIT1, m_iso); -} - -BEGIN_MESSAGE_MAP(CSettingsDlg, CDialog) - ON_BN_CLICKED(IDC_BUTTON1, &CSettingsDlg::OnBrowse) - ON_BN_CLICKED(IDOK, &CSettingsDlg::OnBnClickedOk) -END_MESSAGE_MAP() - -// CSettingsDlg message handlers - -LRESULT CSettingsDlg::WindowProc(UINT message, WPARAM wParam, LPARAM lParam) -{ - if(message == WM_DEVICECHANGE) - { - if(wParam == DBT_DEVICEARRIVAL || wParam == DBT_DEVICEREMOVECOMPLETE) - { - InitDrive(); - - DEV_BROADCAST_HDR* p = (DEV_BROADCAST_HDR*)lParam; - - if(p->dbch_devicetype == DBT_DEVTYP_VOLUME) - { - DEV_BROADCAST_VOLUME* v = (DEV_BROADCAST_VOLUME*)p; - - for(int i = 0; i < 32; i++) - { - if(v->dbcv_unitmask & (1 << i)) - { - TRACE(_T("%c:\n"), 'A' + i); - - // TODO - } - } - } - } - } - - return __super::WindowProc(message, wParam, lParam); -} - -void CSettingsDlg::OnBrowse() -{ - UpdateData(); - - CFileDialog fd(TRUE, NULL, m_iso, - OFN_EXPLORER | OFN_ENABLESIZING | OFN_HIDEREADONLY, - _T("ISO file|*.iso|All files|*.*|"), this); - - if(fd.DoModal() == IDOK) - { - m_iso = fd.GetPathName(); - - UpdateData(FALSE); - - for(int i = 0, j = m_drive.GetCount(); i < j; i++) - { - if((int)m_drive.GetItemData(i) < 0) - { - m_drive.SetCurSel(i); - - break; - } - } - } -} - -void CSettingsDlg::OnBnClickedOk() -{ - UpdateData(); - - int i = m_drive.GetCurSel(); - - if(i >= 0) - { - i = (int)m_drive.GetItemData(i); - } - - AfxGetApp()->WriteProfileInt(_T("Settings"), _T("drive"), i); - - AfxGetApp()->WriteProfileString(_T("Settings"), _T("iso"), m_iso); - - OnOK(); + ComboBoxInit(IDC_COMBO1, &drives[0], drives.size(), drive); } diff --git a/plugins/CDVDolio/SettingsDlg.h b/plugins/CDVDolio/SettingsDlg.h index af06248c42..7c0962547a 100644 --- a/plugins/CDVDolio/SettingsDlg.h +++ b/plugins/CDVDolio/SettingsDlg.h @@ -1,35 +1,16 @@ #pragma once -#include "afxwin.h" +#include "CDVDDialog.h" #include "resource.h" -// CSettingsDlg dialog - -class CSettingsDlg : public CDialog +class CDVDSettingsDlg : public CDVDDialog { - DECLARE_DYNAMIC(CSettingsDlg) + void UpdateDrives(); protected: - void InitDrive(); - void InitISO(); + void OnInit(); + bool OnMessage(UINT message, WPARAM wParam, LPARAM lParam); + bool OnCommand(HWND hWnd, UINT id, UINT code); public: - CSettingsDlg(CWnd* pParent = NULL); // standard constructor - virtual ~CSettingsDlg(); - - virtual BOOL OnInitDialog(); - - enum { IDD = IDD_CONFIG }; - - CComboBox m_drive; - CString m_iso; - -protected: - virtual void DoDataExchange(CDataExchange* pDX); // DDX/DDV support - - DECLARE_MESSAGE_MAP() - - LRESULT WindowProc(UINT message, WPARAM wParam, LPARAM lParam); - - afx_msg void OnBrowse(); - afx_msg void OnBnClickedOk(); + CDVDSettingsDlg(); }; diff --git a/plugins/CDVDolio/cdvd.cpp b/plugins/CDVDolio/cdvd.cpp index 1568bf3777..ee3d75f90f 100644 --- a/plugins/CDVDolio/cdvd.cpp +++ b/plugins/CDVDolio/cdvd.cpp @@ -22,55 +22,60 @@ #include "stdafx.h" #include "cdvd.h" #include "SettingsDlg.h" -#include -#ifdef _DEBUG -#define new DEBUG_NEW -#endif +static HMODULE s_hModule; -// -// Note! -// -// If this DLL is dynamically linked against the MFC -// DLLs, any functions exported from this DLL which -// call into MFC must have the AFX_MANAGE_STATE macro -// added at the very beginning of the function. -// -// For example: -// -// extern "C" BOOL PASCAL EXPORT ExportedFunction() -// { -// AFX_MANAGE_STATE(AfxGetStaticModuleState()); -// // normal function body here -// } -// -// It is very important that this macro appear in each -// function, prior to any calls into MFC. This means that -// it must appear as the first statement within the -// function, even before any object variable declarations -// as their constructors may generate calls into the MFC -// DLL. -// -// Please see MFC Technical Notes 33 and 58 for additional -// details. -// +BOOL APIENTRY DllMain(HMODULE hModule, DWORD ul_reason_for_call, LPVOID lpReserved) +{ + switch(ul_reason_for_call) + { + case DLL_PROCESS_ATTACH: + s_hModule = hModule; + case DLL_THREAD_ATTACH: + case DLL_THREAD_DETACH: + case DLL_PROCESS_DETACH: + break; + } -BEGIN_MESSAGE_MAP(cdvdApp, CWinApp) -END_MESSAGE_MAP() + return TRUE; +} -cdvdApp::cdvdApp() +CDVDolioApp theApp; + +const char* CDVDolioApp::m_ini = "inis/CDVDolio.ini"; +const char* CDVDolioApp::m_section = "Settings"; + +CDVDolioApp::CDVDolioApp() { } -cdvdApp theApp; - -BOOL cdvdApp::InitInstance() +HMODULE CDVDolioApp::GetModuleHandle() { - __super::InitInstance(); + return s_hModule; +} - SetRegistryKey(_T("Gabest")); +string CDVDolioApp::GetConfig(const char* entry, const char* value) +{ + char buff[4096] = {0}; + GetPrivateProfileString(m_section, entry, value, buff, countof(buff), m_ini); + return string(buff); +} - return TRUE; +void CDVDolioApp::SetConfig(const char* entry, const char* value) +{ + WritePrivateProfileString(m_section, entry, value, m_ini); +} + +int CDVDolioApp::GetConfig(const char* entry, int value) +{ + return GetPrivateProfileInt(m_section, entry, value, m_ini); +} + +void CDVDolioApp::SetConfig(const char* entry, int value) +{ + char buff[32] = {0}; + itoa(value, buff, 10); + SetConfig(entry, buff); } // @@ -78,7 +83,7 @@ BOOL cdvdApp::InitInstance() #define PS2E_LT_CDVD 0x08 #define PS2E_CDVD_VERSION 0x0005 -EXPORT_C_(UINT32) PS2EgetLibType() +EXPORT_C_(uint32) PS2EgetLibType() { return PS2E_LT_CDVD; } @@ -88,11 +93,11 @@ EXPORT_C_(char*) PS2EgetLibName() return "CDVDolio"; // olio = OverLapped I/O (duh) } -EXPORT_C_(UINT32) PS2EgetLibVersion2(UINT32 type) +EXPORT_C_(uint32) PS2EgetLibVersion2(UINT32 type) { - const UINT32 revision = 0; - const UINT32 build = 1; - const UINT32 minor = 0; + const uint32 revision = 0; + const uint32 build = 1; + const uint32 minor = 0; return (build << 0) | (revision << 8) | (PS2E_CDVD_VERSION << 16) | (minor << 24); } @@ -125,12 +130,12 @@ bool CDVD::SyncRead(int lsn) return Read(lsn) && GetBuffer(); } -bool CDVD::Open(CString path) +bool CDVD::Open(const char* path) { - m_label.Empty(); + m_label.clear(); - DWORD share = FILE_SHARE_READ; - DWORD flags = FILE_ATTRIBUTE_READONLY | FILE_FLAG_SEQUENTIAL_SCAN | FILE_FLAG_OVERLAPPED; + uint32 share = FILE_SHARE_READ; + uint32 flags = FILE_ATTRIBUTE_READONLY | FILE_FLAG_SEQUENTIAL_SCAN | FILE_FLAG_OVERLAPPED; m_hFile = CreateFile(path, GENERIC_READ, share, NULL, OPEN_EXISTING, flags, (HANDLE)NULL); @@ -164,8 +169,23 @@ bool CDVD::Open(CString path) return false; } - m_label = CString(CStringA((char*)&m_buff[24 + 40], 32)); - m_label.Trim(); + m_label = string((const char*)&m_buff[24 + 40], 32); + + // trim + + { + string::size_type i = m_label.find_first_not_of(' '); + string::size_type j = m_label.find_last_not_of(' '); + + if(i == string::npos) + { + m_label.clear(); + } + else + { + m_label = m_label.substr(i, (j != string::npos ? j + 1 : string::npos) - i); + } + } // m_block.count = *(DWORD*)&m_buff[24 + 80]; @@ -184,12 +204,12 @@ void CDVD::Close() m_cache.pending = false; m_cache.count = 0; - m_label.Empty(); + m_label.clear(); } -CString CDVD::GetLabel() +const char* CDVD::GetLabel() { - return m_label; + return m_label.c_str(); } bool CDVD::Read(int lsn, int mode) @@ -237,7 +257,7 @@ bool CDVD::Read(int lsn, int mode) return true; } -BYTE* CDVD::GetBuffer() +uint8* CDVD::GetBuffer() { DWORD size = 0; @@ -258,7 +278,7 @@ BYTE* CDVD::GetBuffer() return &m_buff[24]; } -UINT32 CDVD::GetTN(cdvdTN* buff) +uint32 CDVD::GetTN(cdvdTN* buff) { buff->strack = 1; buff->etrack = 1; @@ -266,7 +286,7 @@ UINT32 CDVD::GetTN(cdvdTN* buff) return 0; } -UINT32 CDVD::GetTD(BYTE track, cdvdTD* buff) +uint32 CDVD::GetTD(BYTE track, cdvdTD* buff) { if(track == 0) { @@ -285,7 +305,7 @@ static CDVD s_cdvd; // -EXPORT_C_(UINT32) CDVDinit() +EXPORT_C_(uint32) CDVDinit() { return 0; } @@ -294,22 +314,22 @@ EXPORT_C CDVDshutdown() { } -EXPORT_C_(UINT32) CDVDopen(const char* title) +EXPORT_C_(uint32) CDVDopen(const char* title) { - CString path; + string path; - int i = AfxGetApp()->GetProfileInt(_T("Settings"), _T("drive"), -1); + int i = theApp.GetConfig("drive", -1); if(i >= 'A' && i <= 'Z') { - path.Format(_T("\\\\.\\%c:"), i); + path = format("\\\\.\\%c:", i); } else { - path = AfxGetApp()->GetProfileString(_T("Settings"), _T("iso"), _T("")); + path = theApp.GetConfig("iso", ""); } - return s_cdvd.Open(path) ? 0 : -1; + return s_cdvd.Open(path.c_str()) ? 0 : -1; } EXPORT_C CDVDclose() @@ -317,61 +337,59 @@ EXPORT_C CDVDclose() s_cdvd.Close(); } -EXPORT_C_(UINT32) CDVDreadTrack(int lsn, int mode) +EXPORT_C_(uint32) CDVDreadTrack(int lsn, int mode) { return s_cdvd.Read(lsn, mode) ? 0 : -1; } -EXPORT_C_(BYTE*) CDVDgetBuffer() +EXPORT_C_(uint8*) CDVDgetBuffer() { return s_cdvd.GetBuffer(); } -EXPORT_C_(UINT32) CDVDreadSubQ(UINT32 lsn, cdvdSubQ* subq) +EXPORT_C_(uint32) CDVDreadSubQ(uint32 lsn, cdvdSubQ* subq) { return -1; } -EXPORT_C_(UINT32) CDVDgetTN(cdvdTN* buff) +EXPORT_C_(uint32) CDVDgetTN(cdvdTN* buff) { return s_cdvd.GetTN(buff); } -EXPORT_C_(UINT32) CDVDgetTD(BYTE track, cdvdTD* buff) +EXPORT_C_(uint32) CDVDgetTD(uint8 track, cdvdTD* buff) { return s_cdvd.GetTD(track, buff); } -EXPORT_C_(UINT32) CDVDgetTOC(void* toc) +EXPORT_C_(uint32) CDVDgetTOC(void* toc) { return -1; // TODO } -EXPORT_C_(UINT32) CDVDgetDiskType() +EXPORT_C_(uint32) CDVDgetDiskType() { return CDVD_TYPE_PS2DVD; // TODO } -EXPORT_C_(UINT32) CDVDgetTrayStatus() +EXPORT_C_(uint32) CDVDgetTrayStatus() { return CDVD_TRAY_CLOSE; } -EXPORT_C_(UINT32) CDVDctrlTrayOpen() +EXPORT_C_(uint32) CDVDctrlTrayOpen() { return 0; } -EXPORT_C_(UINT32) CDVDctrlTrayClose() +EXPORT_C_(uint32) CDVDctrlTrayClose() { return 0; } EXPORT_C CDVDconfigure() { - AFX_MANAGE_STATE(AfxGetStaticModuleState()); - - CSettingsDlg dlg; + CDVDSettingsDlg dlg; if(IDOK == dlg.DoModal()) { @@ -384,7 +402,7 @@ EXPORT_C CDVDabout() { } -EXPORT_C_(UINT32) CDVDtest() +EXPORT_C_(uint32) CDVDtest() { return 0; } diff --git a/plugins/CDVDolio/cdvd.def b/plugins/CDVDolio/cdvd.def index 4d845a37ae..c47fee74cb 100644 --- a/plugins/CDVDolio/cdvd.def +++ b/plugins/CDVDolio/cdvd.def @@ -1,6 +1,6 @@ ; cdvd.def : Declares the module parameters for the DLL. -LIBRARY "cdvd" +LIBRARY "CDVDolio" EXPORTS ; Explicit exports can go here diff --git a/plugins/CDVDolio/cdvd.h b/plugins/CDVDolio/cdvd.h index 26d6c9c97b..23fb70f985 100644 --- a/plugins/CDVDolio/cdvd.h +++ b/plugins/CDVDolio/cdvd.h @@ -21,48 +21,31 @@ #pragma once -#ifndef __AFXWIN_H__ - #error include 'stdafx.h' before including this file for PCH -#endif - -class cdvdApp : public CWinApp -{ -public: - cdvdApp(); - -public: - virtual BOOL InitInstance(); - - DECLARE_MESSAGE_MAP() -}; - -// - struct cdvdSubQ { - BYTE ctrl:4; // control and mode bits - BYTE mode:4; // control and mode bits - BYTE trackNum; // current track number (1 to 99) - BYTE trackIndex; // current index within track (0 to 99) - BYTE trackM; // current minute location on the disc (BCD encoded) - BYTE trackS; // current sector location on the disc (BCD encoded) - BYTE trackF; // current frame location on the disc (BCD encoded) - BYTE pad; // unused - BYTE discM; // current minute offset from first track (BCD encoded) - BYTE discS; // current sector offset from first track (BCD encoded) - BYTE discF; // current frame offset from first track (BCD encoded) + uint8 ctrl:4; // control and mode bits + uint8 mode:4; // control and mode bits + uint8 trackNum; // current track number (1 to 99) + uint8 trackIndex; // current index within track (0 to 99) + uint8 trackM; // current minute location on the disc (BCD encoded) + uint8 trackS; // current sector location on the disc (BCD encoded) + uint8 trackF; // current frame location on the disc (BCD encoded) + uint8 pad; // unused + uint8 discM; // current minute offset from first track (BCD encoded) + uint8 discS; // current sector offset from first track (BCD encoded) + uint8 discF; // current frame offset from first track (BCD encoded) }; struct cdvdTD // NOT bcd coded { - UINT32 lsn; - BYTE type; + uint32 lsn; + uint8 type; }; struct cdvdTN { - BYTE strack; // number of the first track (usually 1) - BYTE etrack; // number of the last track + uint8 strack; // number of the first track (usually 1) + uint8 etrack; // number of the last track }; // CDVDreadTrack mode values: @@ -109,14 +92,34 @@ struct cdvdTN #define CACHE_BLOCK_COUNT 16 +class CDVDolioApp +{ + static const char* m_ini; + static const char* m_section; + +public: + CDVDolioApp(); + + HMODULE GetModuleHandle(); + + string GetConfig(const char* entry, const char* value); + void SetConfig(const char* entry, const char* value); + int GetConfig(const char* entry, int value); + void SetConfig(const char* entry, int value); +}; + +extern CDVDolioApp theApp; + +// + class CDVD { HANDLE m_hFile; - CString m_label; + string m_label; OVERLAPPED m_overlapped; struct {int count, size, offset;} m_block; - struct {BYTE buff[2048 * CACHE_BLOCK_COUNT]; bool pending; int start, count;} m_cache; - BYTE m_buff[2352]; + struct {uint8 buff[2048 * CACHE_BLOCK_COUNT]; bool pending; int start, count;} m_cache; + uint8 m_buff[2352]; LARGE_INTEGER MakeOffset(int lsn); bool SyncRead(int lsn); @@ -125,12 +128,12 @@ public: CDVD(); virtual ~CDVD(); - bool Open(CString path); + bool Open(const char* path); void Close(); - CString GetLabel(); + const char* GetLabel(); bool Read(int lsn, int mode = CDVD_MODE_2048); - BYTE* GetBuffer(); - UINT32 GetTN(cdvdTN* buff); - UINT32 GetTD(BYTE track, cdvdTD* buff); + uint8* GetBuffer(); + uint32 GetTN(cdvdTN* buff); + uint32 GetTD(uint8 track, cdvdTD* buff); }; diff --git a/plugins/CDVDolio/cdvd_vs2008.vcproj b/plugins/CDVDolio/cdvd_vs2008.vcproj index 52e9cbbd85..8ef2a9ee4f 100644 --- a/plugins/CDVDolio/cdvd_vs2008.vcproj +++ b/plugins/CDVDolio/cdvd_vs2008.vcproj @@ -23,7 +23,7 @@ Name="Debug|Win32" ConfigurationType="2" InheritedPropertySheets=".\vsprops\common.vsprops;.\vsprops\ProjectRootDir.vsprops;..\..\common\vsprops\BaseProperties.vsprops;.\vsprops\debug.vsprops" - UseOfMFC="2" + UseOfMFC="0" CharacterSet="2" > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - @@ -600,10 +147,75 @@ Name="VCPostBuildEventTool" /> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + @@ -916,10 +786,75 @@ Name="VCPostBuildEventTool" /> + + + + + + + + + + + + + + + + + + + @@ -979,10 +914,75 @@ Name="VCPostBuildEventTool" /> + + + + + + + + + + + + + + + + + + + + + @@ -1075,62 +1079,6 @@ UsePrecompiledHeader="1" /> - - - - - - - - - - - - - - - - - - - - - @@ -1139,6 +1087,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1147,6 +1103,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1155,6 +1119,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1163,6 +1135,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1171,6 +1151,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1179,6 +1167,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1187,6 +1183,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1206,6 +1210,10 @@ RelativePath=".\cdvd.h" > + + diff --git a/plugins/CDVDolio/stdafx.cpp b/plugins/CDVDolio/stdafx.cpp index e48c42db5b..ad51702df7 100644 --- a/plugins/CDVDolio/stdafx.cpp +++ b/plugins/CDVDolio/stdafx.cpp @@ -6,3 +6,34 @@ // TODO: reference any additional headers you need in STDAFX.H // and not in this file + +string format(const char* fmt, ...) +{ + va_list args; + va_start(args, fmt); + + int result = -1, length = 256; + + char* buffer = NULL; + + while(result == -1) + { + if(buffer) delete [] buffer; + + buffer = new char[length + 1]; + + memset(buffer, 0, length + 1); + + result = _vsnprintf(buffer, length, fmt, args); + + length *= 2; + } + + va_end(args); + + string s(buffer); + + delete [] buffer; + + return s; +} diff --git a/plugins/CDVDolio/stdafx.h b/plugins/CDVDolio/stdafx.h index 3f39fe0d2a..75fa739cf5 100644 --- a/plugins/CDVDolio/stdafx.h +++ b/plugins/CDVDolio/stdafx.h @@ -4,60 +4,98 @@ #pragma once -#pragma warning(disable: 4996) +#pragma warning(disable: 4996 4995 4324 4100 4101 4201) -#ifndef VC_EXTRALEAN -#define VC_EXTRALEAN // Exclude rarely-used stuff from Windows headers -#endif +#ifdef _WINDOWS + +// The following macros define the minimum required platform. The minimum required platform +// is the earliest version of Windows, Internet Explorer etc. that has the necessary features to run +// your application. The macros work by enabling all features available on platform versions up to and +// including the version specified. // Modify the following defines if you have to target a platform prior to the ones specified below. // Refer to MSDN for the latest info on corresponding values for different platforms. -#ifndef WINVER // Allow use of features specific to Windows 95 and Windows NT 4 or later. -#define WINVER 0x0510 // Change this to the appropriate value to target Windows 98 and Windows 2000 or later. + +#ifndef WINVER // Specifies that the minimum required platform is Windows Vista. +#define WINVER 0x0600 // Change this to the appropriate value to target other versions of Windows. #endif -#ifndef _WIN32_WINNT // Allow use of features specific to Windows NT 4 or later. -#define _WIN32_WINNT 0x0400 // Change this to the appropriate value to target Windows 2000 or later. -#endif +#ifndef _WIN32_WINNT // Specifies that the minimum required platform is Windows Vista. +#define _WIN32_WINNT 0x0600 // Change this to the appropriate value to target other versions of Windows. +#endif -#ifndef _WIN32_WINDOWS // Allow use of features specific to Windows 98 or later. +#ifndef _WIN32_WINDOWS // Specifies that the minimum required platform is Windows 98. #define _WIN32_WINDOWS 0x0410 // Change this to the appropriate value to target Windows Me or later. #endif -#ifndef _WIN32_IE // Allow use of features specific to IE 4.0 or later. -#define _WIN32_IE 0x0400 // Change this to the appropriate value to target IE 5.0 or later. +#ifndef _WIN32_IE // Specifies that the minimum required platform is Internet Explorer 7.0. +#define _WIN32_IE 0x0700 // Change this to the appropriate value to target other versions of IE. #endif -#define _ATL_CSTRING_EXPLICIT_CONSTRUCTORS // some CString constructors will be explicit +#define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers -#include // MFC core and standard components -//#include // MFC extensions -#ifndef _AFX_NO_AFXCMN_SUPPORT -#include // MFC support for Windows Common Controls -#endif // _AFX_NO_AFXCMN_SUPPORT -//#include +#include +#include +#include +#include #include -#include -#include -#include +#include -#define countof(a) (sizeof(a)/sizeof(a[0])) +#endif + +// stdc + +#include +#include +#include + +#include +#include +#include +#include +#include + +using namespace std; +using namespace stdext; + +extern string format(const char* fmt, ...); + +// syntactic sugar + +// put these into vc9/common7/ide/usertype.dat to have them highlighted + +typedef unsigned char uint8; +typedef signed char int8; +typedef unsigned short uint16; +typedef signed short int16; +typedef unsigned int uint32; +typedef signed int int32; +typedef unsigned long long uint64; +typedef signed long long int64; + +#define countof(a) (sizeof(a) / sizeof(a[0])) #define EXPORT_C extern "C" __declspec(dllexport) void __stdcall #define EXPORT_C_(type) extern "C" __declspec(dllexport) type __stdcall +#define ALIGN_STACK(n) __declspec(align(n)) int __dummy; + #ifndef RESTRICT #ifdef __INTEL_COMPILER #define RESTRICT restrict - #elif _MSC_VER >= 1400 + #elif _MSC_VER >= 1400 // TODO: gcc #define RESTRICT __restrict #else #define RESTRICT #endif #endif -#pragma warning(disable : 4995 4324 4100) +#if defined(_DEBUG) && defined(_MSC_VER) + #define ASSERT assert +#else + #define ASSERT(exp) ((void)0) +#endif #ifdef _M_SSE -#error No SSE please! + #error No SSE please! #endif diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index 56721daab8..bfc1871367 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -80,9 +80,7 @@ void GPUSettingsDlg::OnInit() memset(&mode, 0, sizeof(mode)); m_modes.push_back(mode); - HWND hWnd = GetDlgItem(m_hWnd, IDC_RESOLUTION); - - ComboBoxAppend(hWnd, "Windowed", (LPARAM)&m_modes.back(), true); + ComboBoxAppend(IDC_RESOLUTION, "Windowed", (LPARAM)&m_modes.back(), true); if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { @@ -100,7 +98,7 @@ void GPUSettingsDlg::OnInit() string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); - ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); + ComboBoxAppend(IDC_RESOLUTION, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } } @@ -117,11 +115,11 @@ void GPUSettingsDlg::OnInit() renderers.push_back(g_renderers[i]); } - ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_FILTER), g_filter, countof(g_filter), theApp.GetConfig("filter", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_DITHERING), g_dithering, countof(g_dithering), theApp.GetConfig("dithering", 1)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_SCALE), g_scale, countof(g_scale), theApp.GetConfig("scale_x", 0) | (theApp.GetConfig("scale_y", 0) << 2)); + ComboBoxInit(IDC_RENDERER, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); + ComboBoxInit(IDC_FILTER, g_filter, countof(g_filter), theApp.GetConfig("filter", 0)); + ComboBoxInit(IDC_DITHERING, g_dithering, countof(g_dithering), theApp.GetConfig("dithering", 1)); + ComboBoxInit(IDC_ASPECTRATIO, g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); + ComboBoxInit(IDC_SCALE, g_scale, countof(g_scale), theApp.GetConfig("scale_x", 0) | (theApp.GetConfig("scale_y", 0) << 2)); SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETRANGE, 0, MAKELPARAM(16, 1)); SendMessage(GetDlgItem(m_hWnd, IDC_SWTHREADS), UDM_SETPOS, 0, MAKELPARAM(theApp.GetConfig("swthreads", 1), 0)); @@ -139,7 +137,7 @@ bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { INT_PTR data; - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RESOLUTION), data)) + if(ComboBoxGetSelData(IDC_RESOLUTION, data)) { const D3DDISPLAYMODE* mode = (D3DDISPLAYMODE*)data; @@ -148,27 +146,27 @@ bool GPUSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) theApp.SetConfig("ModeRefreshRate", (int)mode->RefreshRate); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), data)) + if(ComboBoxGetSelData(IDC_RENDERER, data)) { theApp.SetConfig("Renderer", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_FILTER), data)) + if(ComboBoxGetSelData(IDC_FILTER, data)) { theApp.SetConfig("filter", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_DITHERING), data)) + if(ComboBoxGetSelData(IDC_DITHERING, data)) { theApp.SetConfig("dithering", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), data)) + if(ComboBoxGetSelData(IDC_ASPECTRATIO, data)) { theApp.SetConfig("AspectRatio", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_SCALE), data)) + if(ComboBoxGetSelData(IDC_SCALE, data)) { theApp.SetConfig("scale_x", data & 3); theApp.SetConfig("scale_y", (data >> 2) & 3); @@ -184,7 +182,7 @@ void GPUSettingsDlg::UpdateControls() { INT_PTR i; - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), i)) + if(ComboBoxGetSelData(IDC_RENDERER, i)) { bool dx9 = i == 1; bool dx10 = i == 2; diff --git a/plugins/GSdx/GSCaptureDlg.cpp b/plugins/GSdx/GSCaptureDlg.cpp index 1d82fde7d8..d918cbb18c 100644 --- a/plugins/GSdx/GSCaptureDlg.cpp +++ b/plugins/GSdx/GSCaptureDlg.cpp @@ -35,7 +35,7 @@ int GSCaptureDlg::GetSelCodec(Codec& c) { INT_PTR data = 0; - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_CODECS), data)) + if(ComboBoxGetSelData(IDC_CODECS, data)) { if(data == 0) return 2; @@ -66,9 +66,7 @@ void GSCaptureDlg::OnInit() string selected = theApp.GetConfig("CaptureVideoCodecDisplayName", ""); - HWND hWnd = GetDlgItem(m_hWnd, IDC_CODECS); - - ComboBoxAppend(hWnd, "Uncompressed", 0, true); + ComboBoxAppend(IDC_CODECS, "Uncompressed", 0, true); BeginEnumSysDev(CLSID_VideoCompressorCategory, moniker) { @@ -107,7 +105,7 @@ void GSCaptureDlg::OnInit() string s(c.FriendlyName.begin(), c.FriendlyName.end()); - ComboBoxAppend(hWnd, s.c_str(), (LPARAM)&m_codecs.back(), s == selected); + ComboBoxAppend(IDC_CODECS, s.c_str(), (LPARAM)&m_codecs.back(), s == selected); } EndEnumSysDev } diff --git a/plugins/GSdx/GSDialog.cpp b/plugins/GSdx/GSDialog.cpp index 11c2101188..c47cc35973 100644 --- a/plugins/GSdx/GSDialog.cpp +++ b/plugins/GSdx/GSDialog.cpp @@ -123,8 +123,12 @@ void GSDialog::SetTextAsInt(UINT id, int i) SetText(id, buff); } -void GSDialog::ComboBoxInit(HWND hWnd, const GSSetting* settings, int count, uint32 selid, uint32 maxid) +void GSDialog::ComboBoxInit(UINT id, const GSSetting* settings, int count, uint32 selid, uint32 maxid) { + HWND hWnd = GetDlgItem(m_hWnd, id); + + SendMessage(hWnd, CB_RESETCONTENT, 0, 0); + for(int i = 0; i < count; i++) { if(settings[i].id <= maxid) @@ -136,13 +140,15 @@ void GSDialog::ComboBoxInit(HWND hWnd, const GSSetting* settings, int count, uin str = str + " (" + settings[i].note + ")"; } - ComboBoxAppend(hWnd, str.c_str(), (LPARAM)settings[i].id, settings[i].id == selid); + ComboBoxAppend(id, str.c_str(), (LPARAM)settings[i].id, settings[i].id == selid); } } } -int GSDialog::ComboBoxAppend(HWND hWnd, const char* str, LPARAM data, bool select) +int GSDialog::ComboBoxAppend(UINT id, const char* str, LPARAM data, bool select) { + HWND hWnd = GetDlgItem(m_hWnd, id); + int item = (int)SendMessage(hWnd, CB_ADDSTRING, 0, (LPARAM)str); SendMessage(hWnd, CB_SETITEMDATA, item, (LPARAM)data); @@ -155,8 +161,10 @@ int GSDialog::ComboBoxAppend(HWND hWnd, const char* str, LPARAM data, bool selec return item; } -bool GSDialog::ComboBoxGetSelData(HWND hWnd, INT_PTR& data) +bool GSDialog::ComboBoxGetSelData(UINT id, INT_PTR& data) { + HWND hWnd = GetDlgItem(m_hWnd, id); + int item = SendMessage(hWnd, CB_GETCURSEL, 0, 0); if(item >= 0) diff --git a/plugins/GSdx/GSDialog.h b/plugins/GSdx/GSDialog.h index 269b1f8bdb..0f1b913569 100644 --- a/plugins/GSdx/GSDialog.h +++ b/plugins/GSdx/GSDialog.h @@ -48,7 +48,7 @@ public: void SetText(UINT id, const char* str); void SetTextAsInt(UINT id, int i); - static void ComboBoxInit(HWND hWnd, const GSSetting* settings, int count, uint32 selid, uint32 maxid = ~0); - static int ComboBoxAppend(HWND hWnd, const char* str, LPARAM data = 0, bool select = false); - static bool ComboBoxGetSelData(HWND hWnd, INT_PTR& data); + void ComboBoxInit(UINT id, const GSSetting* settings, int count, uint32 selid, uint32 maxid = ~0); + int ComboBoxAppend(UINT id, const char* str, LPARAM data = 0, bool select = false); + bool ComboBoxGetSelData(UINT id, INT_PTR& data); }; diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 59e7a8ac7d..023c473456 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -76,9 +76,7 @@ void GSSettingsDlg::OnInit() memset(&mode, 0, sizeof(mode)); m_modes.push_back(mode); - HWND hWnd = GetDlgItem(m_hWnd, IDC_RESOLUTION); - - ComboBoxAppend(hWnd, "Windowed", (LPARAM)&m_modes.back(), true); + ComboBoxAppend(IDC_RESOLUTION, "Windowed", (LPARAM)&m_modes.back(), true); if(CComPtr d3d = Direct3DCreate9(D3D_SDK_VERSION)) { @@ -96,7 +94,7 @@ void GSSettingsDlg::OnInit() string str = format("%dx%d %dHz", mode.Width, mode.Height, mode.RefreshRate); - ComboBoxAppend(hWnd, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); + ComboBoxAppend(IDC_RESOLUTION, str.c_str(), (LPARAM)&m_modes.back(), w == mode.Width && h == mode.Height && hz == mode.RefreshRate); } } } @@ -113,9 +111,9 @@ void GSSettingsDlg::OnInit() renderers.push_back(g_renderers[i]); } - ComboBoxInit(GetDlgItem(m_hWnd, IDC_RENDERER), &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_INTERLACE), g_interlace, countof(g_interlace), theApp.GetConfig("Interlace", 0)); - ComboBoxInit(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); + ComboBoxInit(IDC_RENDERER, &renderers[0], renderers.size(), theApp.GetConfig("Renderer", 0)); + ComboBoxInit(IDC_INTERLACE, g_interlace, countof(g_interlace), theApp.GetConfig("Interlace", 0)); + ComboBoxInit(IDC_ASPECTRATIO, g_aspectratio, countof(g_aspectratio), theApp.GetConfig("AspectRatio", 1)); CheckDlgButton(m_hWnd, IDC_FILTER, theApp.GetConfig("filter", 1)); CheckDlgButton(m_hWnd, IDC_VSYNC, theApp.GetConfig("vsync", 0)); @@ -151,7 +149,7 @@ bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) { INT_PTR data; - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RESOLUTION), data)) + if(ComboBoxGetSelData(IDC_RESOLUTION, data)) { const D3DDISPLAYMODE* mode = (D3DDISPLAYMODE*)data; @@ -160,17 +158,17 @@ bool GSSettingsDlg::OnCommand(HWND hWnd, UINT id, UINT code) theApp.SetConfig("ModeRefreshRate", (int)mode->RefreshRate); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), data)) + if(ComboBoxGetSelData(IDC_RENDERER, data)) { theApp.SetConfig("Renderer", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_INTERLACE), data)) + if(ComboBoxGetSelData(IDC_INTERLACE, data)) { theApp.SetConfig("Interlace", (int)data); } - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_ASPECTRATIO), data)) + if(ComboBoxGetSelData(IDC_ASPECTRATIO, data)) { theApp.SetConfig("AspectRatio", (int)data); } @@ -195,7 +193,7 @@ void GSSettingsDlg::UpdateControls() { INT_PTR i; - if(ComboBoxGetSelData(GetDlgItem(m_hWnd, IDC_RENDERER), i)) + if(ComboBoxGetSelData(IDC_RENDERER, i)) { bool dx9 = i >= 0 && i <= 2; bool dx10 = i >= 3 && i <= 5; From d15db78baf6ffba8e1c690cd76249567e9670763 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 19 Jun 2009 01:02:10 +0000 Subject: [PATCH 253/296] Linux: A bit more work on the ruby version of the build script. It now summarises what built and what didn't at the end, and can be used for building GSnull and PadNull as well... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1399 96395faa-99c1-11dd-bbfe-3dabce05a288 --- build.rb | 31 +++++++++++++++++++++++++++---- 1 file changed, 27 insertions(+), 4 deletions(-) diff --git a/build.rb b/build.rb index 2198340f80..58d3d474c2 100755 --- a/build.rb +++ b/build.rb @@ -22,6 +22,7 @@ $pcsx2_prefix = " --prefix #{$main_dir}" $plugins_prefix = " --prefix #{$plugin_install_dir}" $plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz"] +$full_plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz","GSnull","PadNull"] $pcsx2_build_types = { "dev" => " --enable-devbuild ", @@ -32,6 +33,9 @@ $pcsx2_build_types = { $pcsx2_release_params=["dev","debug","release"] $make_params=["all", "clean","install"] +$build_report ="" +$build_counter = 0 + def plugin_src_dir(plugin_name) name = "#{$plugins_dir}/#{plugin_name}/" case plugin_name @@ -56,6 +60,7 @@ end def make(options) system("make #{options}") + ($? == 0) end def rebuild(options) @@ -68,7 +73,7 @@ def rebuild(options) end def install(build_name) - make "install" + ret = make "install" case build_name # If the package isn't inclined to obey simple instructions... @@ -91,9 +96,13 @@ def install(build_name) svn_revision = /[0-9]+/.match(svn_revision) system("cp #{$pcsx2_install_dir}/pcsx2 #{$pcsx2_install_dir}/pcsx2-#{svn_revision}") end + + ret end def build(build_name, make_parameter) + completed = true + announce "#{build_name.capitalize}" if build_name != "pcsx2" then @@ -111,16 +120,24 @@ def build(build_name, make_parameter) else rebuild($plugins_prefix) end - install(build_name) + completed = install(build_name) when "clean" then make "clean" else - install(build_name) + completed = install(build_name) end Dir.chdir $main_dir + + if completed then + $build_report += "#{build_name} was built successfully.\n" + $build_counter += 1 + else + $build_report += "#{build_name} was not built successfully.\n" + end + end build_parameter = "all" @@ -130,7 +147,7 @@ build_items = Array.new([]) ARGV.each do |x| make_parameter = x if $make_params.include?(x) - build_items.push(x) if $plugin_list.include?(x) or (x == "pcsx2") + build_items.push(x) if $full_plugin_list.include?(x) or (x == "pcsx2") $pcsx2_prefix = $pcsx2_build_types[x] + $pcsx2_prefix if $pcsx2_release_params.include?(x) if (x == "plugins") then @@ -149,3 +166,9 @@ build_items.flatten! build_items.each do |x| build(x,make_parameter) end + +print "\n--\n" +print "Build Summary:\n" +print $build_report +print "\n" +print "#{$build_counter}/#{build_items.count} Successful.\n" \ No newline at end of file From 0f30bf62b54881185112f1afe7a3460a4185253b Mon Sep 17 00:00:00 2001 From: refraction Date: Fri, 19 Jun 2009 03:13:03 +0000 Subject: [PATCH 254/296] COP0.cpp: Updated the commented out Perf counter code to work on the newer builds, also removed the line which stopped them updating at all if the interrupt wasnt on. Others: Fixed a couple of unpack bugs, tried to tackle an MFIFO bug with Tekken Tag. Also re-jiggled a few bits on my recent changed, please negative if it breaks anything. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1400 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/COP0.cpp | 56 ++++++++++++++++++++++++++++++++++------------ pcsx2/Counters.cpp | 24 ++++++++++++++++---- pcsx2/GS.cpp | 32 +++++++++++++++++--------- pcsx2/HwWrite.cpp | 2 +- pcsx2/MTGS.cpp | 23 ++++++++++++------- pcsx2/Vif.cpp | 10 ++------- pcsx2/Vif.h | 8 +++---- pcsx2/VifDma.cpp | 15 ++++++++++--- pcsx2/VifDma.h | 2 ++ 9 files changed, 120 insertions(+), 52 deletions(-) diff --git a/pcsx2/COP0.cpp b/pcsx2/COP0.cpp index 500f2eeaf9..ea94cf3f7e 100644 --- a/pcsx2/COP0.cpp +++ b/pcsx2/COP0.cpp @@ -217,15 +217,15 @@ void COP0_DiagnosticPCCR() if( cpuRegs.PERF.n.pccr.b.Event1 >= 7 && cpuRegs.PERF.n.pccr.b.Event1 <= 10 ) Console::Notice( "PERF/PCR1 Unsupported Update Event Mode = 0x%x", params cpuRegs.PERF.n.pccr.b.Event1 ); } - +extern int branch; __forceinline void COP0_UpdatePCCR() { - if( cpuRegs.CP0.n.Status.b.ERL || !cpuRegs.PERF.n.pccr.b.CTE ) return; + //if( cpuRegs.CP0.n.Status.b.ERL || !cpuRegs.PERF.n.pccr.b.CTE ) return; // TODO : Implement memory mode checks here (kernel/super/user) // For now we just assume user mode. - if( cpuRegs.PERF.n.pccr.b.U0 ) + if( cpuRegs.PERF.n.pccr.val & 0xf ) { // ---------------------------------- // Update Performance Counter 0 @@ -243,24 +243,25 @@ __forceinline void COP0_UpdatePCCR() //prev ^= (1UL<<31); // XOR is fun! //if( (prev & cpuRegs.PERF.n.pcr0) & (1UL<<31) ) - if( cpuRegs.PERF.n.pcr0 & 0x80000000 ) + if( (cpuRegs.PERF.n.pcr0 & 0x80000000) && (cpuRegs.CP0.n.Status.b.ERL == 1) && cpuRegs.PERF.n.pccr.b.CTE) { // TODO: Vector to the appropriate exception here. // This code *should* be correct, but is untested (and other parts of the emu are // not prepared to handle proper Level 2 exception vectors yet) - /*if( delay_slot ) + //branch == 1 is probably not the best way to check for the delay slot, but it beats nothing! (Refraction) + /* if( branch == 1 ) { - cpuRegs.CP0.ErrorEPC = cpuRegs.pc - 4; - cpuRegs.CP0.Cause.BD2 = 1; + cpuRegs.CP0.n.ErrorEPC = cpuRegs.pc - 4; + cpuRegs.CP0.n.Cause |= 0x40000000; } else { - cpuRegs.CP0.ErrorEPC = cpuRegs.pc; - cpuRegs.CP0.Cause.BD2 = 0; + cpuRegs.CP0.n.ErrorEPC = cpuRegs.pc; + cpuRegs.CP0.n.Cause &= ~0x40000000; } - if( cpuRegs.CP0.Status.DEV ) + if( cpuRegs.CP0.n.Status.b.DEV ) { // Bootstrap vector cpuRegs.pc = 0xbfc00280; @@ -269,8 +270,8 @@ __forceinline void COP0_UpdatePCCR() { cpuRegs.pc = 0x80000080; } - cpuRegs.CP0.Status.ERL = 1; - cpuRegs.CP0.Cause.EXC2 = 2;*/ + cpuRegs.CP0.n.Status.b.ERL = 1; + cpuRegs.CP0.n.Cause |= 0x20000;*/ } } } @@ -289,9 +290,36 @@ __forceinline void COP0_UpdatePCCR() cpuRegs.PERF.n.pcr1 += incr; s_iLastPERFCycle[1] = cpuRegs.cycle; - if( cpuRegs.PERF.n.pcr1 & 0x80000000 ) + if( (cpuRegs.PERF.n.pcr1 & 0x80000000) && (cpuRegs.CP0.n.Status.b.ERL == 1) && cpuRegs.PERF.n.pccr.b.CTE) { - // See PCR0 comments for notes on exceptions + // TODO: Vector to the appropriate exception here. + // This code *should* be correct, but is untested (and other parts of the emu are + // not prepared to handle proper Level 2 exception vectors yet) + + //branch == 1 is probably not the best way to check for the delay slot, but it beats nothing! (Refraction) + + /*if( branch == 1 ) + { + cpuRegs.CP0.n.ErrorEPC = cpuRegs.pc - 4; + cpuRegs.CP0.n.Cause |= 0x40000000; + } + else + { + cpuRegs.CP0.n.ErrorEPC = cpuRegs.pc; + cpuRegs.CP0.n.Cause &= ~0x40000000; + } + + if( cpuRegs.CP0.n.Status.b.DEV ) + { + // Bootstrap vector + cpuRegs.pc = 0xbfc00280; + } + else + { + cpuRegs.pc = 0x80000080; + } + cpuRegs.CP0.n.Status.b.ERL = 1; + cpuRegs.CP0.n.Cause |= 0x20000;*/ } } } diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index d8af6b3f80..3825b41d59 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -336,8 +336,16 @@ static __forceinline void VSyncStart(u32 sCycle) EECNT_LOG( "///////// EE COUNTER VSYNC START \\\\\\\\\\\\\\\\\\\\ (frame: %d)", iFrame ); vSyncDebugStuff( iFrame ); // EE Profiling and Debug code - if ((CSRw & 0x8)) GSCSRr|= 0x8; - if (!(GSIMR&0x800)) gsIrq(); + if ((CSRw & 0x8)) + { + GSCSRr|= 0x8; + + if (!(GSIMR&0x800)) + { + gsIrq(); + } + CSRw &= ~0x8; //Disable the interrupt from triggering twice + } hwIntcIrq(INTC_VBLANK_S); psxVBlankStart(); @@ -404,8 +412,16 @@ __forceinline void rcntUpdate_hScanline() hsyncCounter.Mode = MODE_HRENDER; } else { //HBLANK END / HRENDER Begin - if (CSRw & 0x4) GSCSRr |= 4; // signal - if (!(GSIMR&0x400)) gsIrq(); + if (CSRw & 0x4) + { + GSCSRr |= 4; // signal + + if (!(GSIMR&0x400)) + { + gsIrq(); + } + CSRw &= ~0x4; //Disable the interrupt from triggering twice + } if (gates) rcntEndGate(false, hsyncCounter.sCycle); if (psxhblankgate) psxCheckEndGate16(0); diff --git a/pcsx2/GS.cpp b/pcsx2/GS.cpp index 53789055bc..abb171d06e 100644 --- a/pcsx2/GS.cpp +++ b/pcsx2/GS.cpp @@ -290,14 +290,7 @@ void gsGIFReset() void gsCSRwrite(u32 value) { - CSRw |= value & ~0x60; - - if( mtgsThread != NULL ) - mtgsThread->SendSimplePacket( GS_RINGTYPE_WRITECSR, CSRw, 0, 0 ); - else - GSwriteCSR(CSRw); - - GSCSRr = ((GSCSRr&~value)&0x1f)|(GSCSRr&~0x1f); + // Our emulated GS has no FIFO... /*if( value & 0x100 ) { // FLUSH @@ -317,17 +310,36 @@ void gsCSRwrite(u32 value) GSreset(); } - CSRw = 0x1f; + CSRw |= 0x1f; GSCSRr = 0x551B4000; // Set the FINISH bit to 1 - GS is always at a finish state as we don't have a FIFO(saqib) GSIMR = 0x7F00; //This is bits 14-8 thats all that should be 1 + } + else if( value & 0x100 ) // FLUSH + { + //Console::WriteLn("GS_CSR FLUSH GS fifo: %x (CSRr=%x)", params value, GSCSRr); } + else + { + CSRw |= value & 0x1f; + + if( mtgsThread != NULL ) + mtgsThread->SendSimplePacket( GS_RINGTYPE_WRITECSR, CSRw, 0, 0 ); + else + GSwriteCSR(CSRw); + + GSCSRr = ((GSCSRr&~value)&0x1f)|(GSCSRr&~0x1f); + } + } static void IMRwrite(u32 value) { GSIMR = (value & 0x1f00)|0x6000; - if((GSCSRr & 0x1f) & (~(GSIMR >> 8) & 0x1f)) gsIrq(); + if((GSCSRr & 0x1f) & (~(GSIMR >> 8) & 0x1f)) + { + gsIrq(); + } // don't update mtgs mem } diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index 43659dc34a..16155c09e7 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -949,7 +949,7 @@ void __fastcall hwWrite32_generic( u32 mem, u32 value ) if (value & 0x100) { vif1.done = false; //This must be done here! some games (ala Crash of the Titans) pause the dma to start MFIFO - } + } else cpuRegs.interrupt &= ~(1<<10) | ~(1<<1); //Tekken tag seems to stop vif and start it again in normal, so we will cancel the mfifo loop DmaExec(dmaVIF1, mem, value); return; diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 8b8c949302..dfebe33b8f 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -137,23 +137,30 @@ static void RegHandlerSIGNAL(const u32* data) GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); - if ((CSRw & 0x1)) + if ((CSRw & 0x1)) + { GSCSRr |= 1; // signal - if (!(GSIMR&0x100) ) - gsIrq(); + if (!(GSIMR&0x100) ) + { + gsIrq(); + } + CSRw &= ~0x1; //Disable the interrupt from triggering twice + } } static void RegHandlerFINISH(const u32* data) { - MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n",data[0], data[1], CSRw); + DevCon::Notice("MTGS FINISH data %x_%x CSRw %x\n", params data[0], data[1], CSRw); - if ((CSRw & 0x2)) + if ((CSRw & 0x2)) + { GSCSRr |= 2; // finish - - if (!(GSIMR&0x200) ) - gsIrq(); + if (!(GSIMR&0x200) ) + gsIrq(); + CSRw &= ~0x2; //Disable the interrupt from triggering twice + } } static void RegHandlerLABEL(const u32* data) diff --git a/pcsx2/Vif.cpp b/pcsx2/Vif.cpp index 76f466303c..f61b024b75 100644 --- a/pcsx2/Vif.cpp +++ b/pcsx2/Vif.cpp @@ -533,6 +533,8 @@ void vifMFIFOInterrupt() { g_vifCycles = 0; + if(schedulepath3msk) Vif1MskPath3(); + if((vif1Regs->stat & VIF1_STAT_VGW)) { if(gif->chcr & 0x100) @@ -596,14 +598,6 @@ void vifMFIFOInterrupt() CPU_INT(10, 0); return; } - - if (!(vif1.inprogress & 0x1)) mfifoVIF1transfer(0); - - if (vif1ch->madr >= psHu32(DMAC_RBOR) && vif1ch->madr <= (psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR))) - CPU_INT(10, 0); - else - CPU_INT(10, vif1ch->qwc * BIAS); - return; } diff --git a/pcsx2/Vif.h b/pcsx2/Vif.h index 2f84369b5a..64335c4dde 100644 --- a/pcsx2/Vif.h +++ b/pcsx2/Vif.h @@ -92,16 +92,16 @@ static __forceinline u32 setVifRowRegs(u32 reg, u32 data) switch (reg) { case 0: - vifRegs->r0 += data; + vifRegs->r0 = data; break; case 1: - vifRegs->r1 += data; + vifRegs->r1 = data; break; case 2: - vifRegs->r2 += data; + vifRegs->r2 = data; break; case 3: - vifRegs->r3 += data; + vifRegs->r3 = data; break; jNO_DEFAULT; } diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index de78c47275..fcbe73b6bf 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -328,6 +328,10 @@ static void ProcessMemSkip(int size, unsigned int unpackType, const unsigned int VIFUNPACK_LOG("addr aligned to %x", vif->tag.addr); vif->tag.addr = (vif->tag.addr & ~0xf) + (vifRegs->offset * 4); } + if(vif->tag.addr >= (u32)(VIFdmanum ? 0x4000 : 0x1000)) + { + vif->tag.addr &= (u32)(VIFdmanum ? 0x3fff : 0xfff); + } } @@ -634,14 +638,16 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i } else { - //DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); + DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); tempsize = size; size = 0; } } else { - tempsize = 0; + tempsize = 0; //Commenting out this then + //tempsize = size; // -\_uncommenting these Two enables non-SSE unpacks + //size = 0; // -/ } if (size >= ft->gsize) @@ -745,6 +751,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i { int incdest = ((vifRegs->cycle.cl - vifRegs->cycle.wl) << 2) + 4; size = 0; + int addrstart = v->addr; if((tempsize >> 2) != vif->tag.size) DevCon::Notice("split when size != tagsize"); VIFUNPACK_LOG("sorting tempsize :p, size %d, vifnum %d, addr %x", tempsize, vifRegs->num, vif->tag.addr); @@ -753,6 +760,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i { if(v->addr >= memlimit) { + DevCon::Notice("Mem limit ovf"); v->addr &= (memlimit - 1); dest = (u32*)(VU->Mem + v->addr); } @@ -789,6 +797,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i v->addr &= (memlimit - 1); dest = (u32*)(VU->Mem + v->addr); } + v->addr = addrstart; if(tempsize > 0) size = tempsize; } @@ -2057,7 +2066,7 @@ static void Vif1CMDSTMod() // STMOD u8 schedulepath3msk = 0; -static void Vif1MskPath3() // MSKPATH3 +void Vif1MskPath3() // MSKPATH3 { vif1Regs->mskpath3 = schedulepath3msk & 0x1; //Console::WriteLn("VIF MSKPATH3 %x", params vif1Regs->mskpath3); diff --git a/pcsx2/VifDma.h b/pcsx2/VifDma.h index 56c08d7d6f..3a5d85000b 100644 --- a/pcsx2/VifDma.h +++ b/pcsx2/VifDma.h @@ -55,6 +55,7 @@ struct vifStruct { extern vifStruct vif0, vif1; extern int Path3progress; +extern u8 schedulepath3msk; void __fastcall UNPACK_S_32( u32 *dest, u32 *data, int size ); @@ -95,6 +96,7 @@ void vif0Init(); void vif1Init(); extern void vif0Interrupt(); extern void vif1Interrupt(); +extern void Vif1MskPath3(); void vif0Write32(u32 mem, u32 value); void vif1Write32(u32 mem, u32 value); From edd6b785c13459de700dd902c82e4ea73acc599d Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Fri, 19 Jun 2009 11:45:07 +0000 Subject: [PATCH 255/296] Minor cleanups to the new installer; made it default to packaging Release output only (will add a custom installer for Devel packaging soon). git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1401 96395faa-99c1-11dd-bbfe-3dabce05a288 --- setup/setup.vdproj | 730 ++++++++++++++++++++++----------------------- 1 file changed, 365 insertions(+), 365 deletions(-) diff --git a/setup/setup.vdproj b/setup/setup.vdproj index db6139c1d3..8e5398236f 100644 --- a/setup/setup.vdproj +++ b/setup/setup.vdproj @@ -447,6 +447,24 @@ } "Entry" { + "MsmKey" = "8:_44BA53E81F0FD93AF224C5D26706AA36" + "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_44BA53E81F0FD93AF224C5D26706AA36" + "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_44BA53E81F0FD93AF224C5D26706AA36" + "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_45B9B43F0312436EB21A3EE185196698" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -525,42 +543,6 @@ } "Entry" { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_AF4D152A202841D2BF769FBDF5CEB413" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_882694E02F0649F29ADA794E68DA525C" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_572363F211D9AF197AFDB57B760E8636" - "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { "MsmKey" = "8:_5795B3AD60144867AAF76625BF8D12E6" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -1065,24 +1047,6 @@ } "Entry" { - "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" - "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" - "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { - "MsmKey" = "8:_A6C0DAAB1BFF83F078FD451B8C7A5594" - "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" - "MsmSig" = "8:_UNDEFINED" - } - "Entry" - { "MsmKey" = "8:_A89761474AC748AE931002EE5E581284" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -1539,6 +1503,42 @@ } "Entry" { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_1FA9731E400641839E29839952B10A67" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_C5FEBD35C45E4EAE824FD02E9F023FD0" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_AF4D152A202841D2BF769FBDF5CEB413" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_882694E02F0649F29ADA794E68DA525C" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_8037556EAB9F4BF298A6AA2BDA834930" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { + "MsmKey" = "8:_FD481E6914D27724507EE8D2A4BD4317" + "OwnerKey" = "8:_25EDB5C49F2745F199D2CF1036A1207F" + "MsmSig" = "8:_UNDEFINED" + } + "Entry" + { "MsmKey" = "8:_FF64089169014510AEEBB6F516BFE6D5" "OwnerKey" = "8:_UNDEFINED" "MsmSig" = "8:_UNDEFINED" @@ -1611,7 +1611,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1631,7 +1631,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1651,7 +1651,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1671,7 +1671,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1691,7 +1691,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1711,7 +1711,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1731,7 +1731,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1751,7 +1751,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1771,7 +1771,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1791,7 +1791,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1811,7 +1811,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1831,7 +1831,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1851,7 +1851,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1871,7 +1871,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1890,8 +1890,8 @@ "Tag" = "8:" "Folder" = "8:_69E3FE8CD77E49E3A2EB811D7CF5CE5C" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1930,8 +1930,8 @@ "Tag" = "8:" "Folder" = "8:_9897A3C67AA44BBE97E0185B4254592E" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1951,7 +1951,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1971,7 +1971,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -1991,7 +1991,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2011,7 +2011,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2031,7 +2031,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2051,7 +2051,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2071,7 +2071,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2091,7 +2091,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2111,7 +2111,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2131,7 +2131,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2151,7 +2151,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2190,8 +2190,8 @@ "Tag" = "8:" "Folder" = "8:_9BD92FDD1824471A943B71FB9DCC1BFC" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2231,7 +2231,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2251,7 +2251,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2271,7 +2271,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2291,7 +2291,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2311,7 +2311,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2331,7 +2331,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2351,7 +2351,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2370,8 +2370,8 @@ "Tag" = "8:" "Folder" = "8:_A01F7887979C490594F917213A282F1F" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2391,7 +2391,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2411,7 +2411,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2431,7 +2431,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2451,7 +2451,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2471,7 +2471,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2491,7 +2491,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2511,7 +2511,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2531,7 +2531,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2551,7 +2551,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2570,8 +2570,8 @@ "Tag" = "8:" "Folder" = "8:_9CC70A81B0B14388B8D25978A5DD64F3" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2591,7 +2591,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2630,8 +2630,8 @@ "Tag" = "8:" "Folder" = "8:_505652862F964FA8A185D0D9D41A3230" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2651,7 +2651,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2671,7 +2671,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2691,7 +2691,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2711,7 +2711,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2731,7 +2731,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2751,7 +2751,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2771,7 +2771,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2791,7 +2791,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2811,7 +2811,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2831,7 +2831,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2851,7 +2851,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2870,8 +2870,8 @@ "Tag" = "8:" "Folder" = "8:_AD3D7C491E3D4BFBB8E75C1EADBC803B" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2891,7 +2891,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2910,8 +2910,8 @@ "Tag" = "8:" "Folder" = "8:_55ABDA310DAE41369F53B2CE8F57D600" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2931,7 +2931,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2951,7 +2951,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2971,7 +2971,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -2983,6 +2983,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_44BA53E81F0FD93AF224C5D26706AA36" + { + "SourcePath" = "8:DDRAW.dll" + "TargetName" = "8:DDRAW.dll" + "Tag" = "8:" + "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:TRUE" + "IsDependency" = "11:TRUE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_45B9B43F0312436EB21A3EE185196698" { "SourcePath" = "8:..\\bin\\patches\\2088950A.pnach" @@ -2991,7 +3011,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3011,7 +3031,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3031,7 +3051,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3051,7 +3071,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3071,7 +3091,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3091,7 +3111,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3111,7 +3131,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3131,7 +3151,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3151,7 +3171,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3171,7 +3191,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3191,7 +3211,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3211,7 +3231,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3231,7 +3251,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3243,26 +3263,6 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_572363F211D9AF197AFDB57B760E8636" - { - "SourcePath" = "8:COMDLG32.dll" - "TargetName" = "8:COMDLG32.dll" - "Tag" = "8:" - "Folder" = "8:_A396E583ED3A43A5B9FA30BABB3624E2" - "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" - "ReadOnly" = "11:FALSE" - "Hidden" = "11:FALSE" - "System" = "11:FALSE" - "Permanent" = "11:FALSE" - "SharedLegacy" = "11:FALSE" - "PackageAs" = "3:1" - "Register" = "3:1" - "Exclude" = "11:TRUE" - "IsDependency" = "11:TRUE" - "IsolateTo" = "8:" - } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_5795B3AD60144867AAF76625BF8D12E6" { "SourcePath" = "8:..\\bin\\patches\\7FD7A1B9.pnach" @@ -3271,7 +3271,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3291,7 +3291,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3311,7 +3311,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3331,7 +3331,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3351,7 +3351,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3371,7 +3371,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3391,7 +3391,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3411,7 +3411,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3431,7 +3431,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3451,7 +3451,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3470,8 +3470,8 @@ "Tag" = "8:" "Folder" = "8:_F5073354FAFA4B98A39DCBA0BF8E855C" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3491,7 +3491,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3511,7 +3511,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3531,7 +3531,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3551,7 +3551,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3571,7 +3571,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3591,7 +3591,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3611,7 +3611,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3631,7 +3631,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3651,7 +3651,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3671,7 +3671,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3691,7 +3691,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3711,7 +3711,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3731,7 +3731,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3771,7 +3771,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3791,7 +3791,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3811,7 +3811,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3830,8 +3830,8 @@ "Tag" = "8:" "Folder" = "8:_0ACD0ED1DD734B27B3D547E45959012F" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3851,7 +3851,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3870,8 +3870,8 @@ "Tag" = "8:" "Folder" = "8:_39E58661BEE543E8A9B1269F2772FCAC" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3891,7 +3891,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3931,7 +3931,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3951,7 +3951,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3971,7 +3971,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -3991,7 +3991,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4011,7 +4011,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4031,7 +4031,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4051,7 +4051,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4071,7 +4071,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4091,7 +4091,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4111,7 +4111,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4131,7 +4131,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4150,8 +4150,8 @@ "Tag" = "8:" "Folder" = "8:_786BE5E01902493590964C4D4F509C2D" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4170,8 +4170,8 @@ "Tag" = "8:" "Folder" = "8:_DA9C9F5B385A448D97EE2AD52304A4FD" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4190,8 +4190,8 @@ "Tag" = "8:" "Folder" = "8:_F8763ACD40584CDAA0688B4AEA98D07E" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4211,7 +4211,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4231,7 +4231,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4251,7 +4251,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4271,7 +4271,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4291,7 +4291,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4311,7 +4311,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4331,7 +4331,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4351,7 +4351,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4371,7 +4371,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4391,7 +4391,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4411,7 +4411,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4431,7 +4431,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4451,7 +4451,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4471,7 +4471,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4491,7 +4491,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4511,7 +4511,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4531,7 +4531,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4551,7 +4551,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4571,7 +4571,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4591,7 +4591,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4631,7 +4631,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4651,7 +4651,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4671,7 +4671,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4691,7 +4691,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4711,7 +4711,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4731,7 +4731,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4751,7 +4751,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4763,26 +4763,6 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } - "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A6C0DAAB1BFF83F078FD451B8C7A5594" - { - "SourcePath" = "8:DDRAW.dll" - "TargetName" = "8:DDRAW.dll" - "Tag" = "8:" - "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" - "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" - "ReadOnly" = "11:FALSE" - "Hidden" = "11:FALSE" - "System" = "11:FALSE" - "Permanent" = "11:FALSE" - "SharedLegacy" = "11:FALSE" - "PackageAs" = "3:1" - "Register" = "3:1" - "Exclude" = "11:TRUE" - "IsDependency" = "11:TRUE" - "IsolateTo" = "8:" - } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_A89761474AC748AE931002EE5E581284" { "SourcePath" = "8:..\\bin\\Langs\\pe_PE\\LC_MESSAGES\\pcsx2.mo" @@ -4790,8 +4770,8 @@ "Tag" = "8:" "Folder" = "8:_5FC81602C55A4975B87E5712B58A9781" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4811,7 +4791,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4831,7 +4811,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4851,7 +4831,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4871,7 +4851,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4891,7 +4871,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4911,7 +4891,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4931,7 +4911,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4951,7 +4931,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4971,7 +4951,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -4991,7 +4971,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5011,7 +4991,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5031,7 +5011,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5051,7 +5031,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5071,7 +5051,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5091,7 +5071,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5111,7 +5091,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5130,8 +5110,8 @@ "Tag" = "8:" "Folder" = "8:_E5D304DFDDC94899A964648DA33FFEF3" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5151,7 +5131,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5171,7 +5151,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5191,7 +5171,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5211,7 +5191,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5231,7 +5211,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5251,7 +5231,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5271,7 +5251,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5311,7 +5291,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5331,7 +5311,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5351,7 +5331,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5371,7 +5351,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5391,7 +5371,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5411,7 +5391,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5431,7 +5411,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5451,7 +5431,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5470,8 +5450,8 @@ "Tag" = "8:" "Folder" = "8:_B3FF9E06585E4D588B8A11F45580C9C3" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5491,7 +5471,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5511,7 +5491,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5530,8 +5510,8 @@ "Tag" = "8:" "Folder" = "8:_0ACD0ED1DD734B27B3D547E45959012F" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5551,7 +5531,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5571,7 +5551,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5591,7 +5571,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5611,7 +5591,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5631,7 +5611,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5651,7 +5631,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5671,7 +5651,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5691,7 +5671,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5710,8 +5690,8 @@ "Tag" = "8:" "Folder" = "8:_E70A733A717B4988B3240BB9AA27F6FF" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5731,7 +5711,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5751,7 +5731,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5771,7 +5751,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5791,7 +5771,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5810,8 +5790,8 @@ "Tag" = "8:" "Folder" = "8:_99D64C98017D4A65B91D7963D17E86FA" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5831,7 +5811,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5851,7 +5831,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5870,8 +5850,8 @@ "Tag" = "8:" "Folder" = "8:_E02BBD7427474F9694A1DF1FB22A8895" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5891,7 +5871,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5911,7 +5891,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5931,7 +5911,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -5971,7 +5951,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6011,7 +5991,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6031,7 +6011,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6051,7 +6031,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6071,7 +6051,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6091,7 +6071,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6111,7 +6091,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6131,7 +6111,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6151,7 +6131,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6171,7 +6151,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6191,7 +6171,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6203,6 +6183,26 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" } + "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_FD481E6914D27724507EE8D2A4BD4317" + { + "SourcePath" = "8:COMDLG32.dll" + "TargetName" = "8:COMDLG32.dll" + "Tag" = "8:" + "Folder" = "8:_A396E583ED3A43A5B9FA30BABB3624E2" + "Condition" = "8:" + "Transitive" = "11:FALSE" + "Vital" = "11:TRUE" + "ReadOnly" = "11:FALSE" + "Hidden" = "11:FALSE" + "System" = "11:FALSE" + "Permanent" = "11:FALSE" + "SharedLegacy" = "11:FALSE" + "PackageAs" = "3:1" + "Register" = "3:1" + "Exclude" = "11:TRUE" + "IsDependency" = "11:TRUE" + "IsolateTo" = "8:" + } "{1FB2D0AE-D3B9-43D4-B9DD-F88EC61E35DE}:_FF64089169014510AEEBB6F516BFE6D5" { "SourcePath" = "8:..\\bin\\patches\\5D67AE48.pnach" @@ -6211,7 +6211,7 @@ "Folder" = "8:_42F0D922DB0247BBA7D42CB69E6D5400" "Condition" = "8:" "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -6240,7 +6240,7 @@ { "{9EF0B969-E518-4E46-987F-47570745A589}:_E300D7C3BDA3486FB06E51133B66D4E1" { - "Name" = "8:[ProductName]" + "Name" = "8:PCSX2" "AlwaysCreate" = "11:FALSE" "Condition" = "8:" "Transitive" = "11:FALSE" @@ -6862,22 +6862,22 @@ { "Name" = "8:Microsoft Visual Studio" "ProductName" = "8:PCSX2" - "ProductCode" = "8:{69F747B1-6251-4490-A3B4-AFADB306C112}" - "PackageCode" = "8:{912DBF5E-E3BA-4797-BB29-02754B5B5813}" + "ProductCode" = "8:{06D2A88C-85BA-4D97-8B6E-551B9DC899E1}" + "PackageCode" = "8:{319F53C1-4C70-4547-8A0D-1750F488911F}" "UpgradeCode" = "8:{5237A49B-FA2E-47C4-8C0E-92338FDC9F2A}" "RestartWWWService" = "11:FALSE" "RemovePreviousVersions" = "11:FALSE" "DetectNewerInstalledVersion" = "11:TRUE" "InstallAllUsers" = "11:TRUE" - "ProductVersion" = "8:1.0.0" + "ProductVersion" = "8:0.9.7" "Manufacturer" = "8:PCSX2 Team" "ARPHELPTELEPHONE" = "8:" "ARPHELPLINK" = "8:" "Title" = "8:PCSX2 Installer" "Subject" = "8:Playstation 2 Emulator" "ARPCONTACT" = "8:PCSX2 Team" - "Keywords" = "8:" - "ARPCOMMENTS" = "8:Official Installer for PCSX2 Emulator and Standard Plugins " + "Keywords" = "8:PCSX2 PS2 Playstation Emulation Emulator" + "ARPCOMMENTS" = "8:PCSX2 and Standard Plugins" "ARPURLINFOABOUT" = "8:http://www.pcsx2.net/" "ARPPRODUCTICON" = "8:" "ARPIconIndex" = "3:0" @@ -7149,8 +7149,8 @@ "Type" = "3:3" "ContextData" = "8:" "Attributes" = "3:0" - "Setting" = "3:1" - "Value" = "8:#1202" + "Setting" = "3:2" + "Value" = "8:Notice: This program falls under some copyleft mumbo-jumbo, more commonly known as the GNU Public License." "DefaultValue" = "8:#1202" "UsePlugInResources" = "11:TRUE" } @@ -7162,8 +7162,8 @@ "Type" = "3:3" "ContextData" = "8:" "Attributes" = "3:0" - "Setting" = "3:1" - "Value" = "8:#1203" + "Setting" = "3:2" + "Value" = "8:This program will install [ProductName] [Version] and other plugins and components necessary to run [ProductName]." "DefaultValue" = "8:#1203" "UsePlugInResources" = "11:TRUE" } @@ -7429,7 +7429,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{04439C5F-05FB-4A9C-AAD1-5388C25377DB}" "ShowKeyOutput" = "11:TRUE" @@ -7444,8 +7444,8 @@ "Tag" = "8:" "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -7485,7 +7485,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{1CEFD830-2B76-4596-A4EE-BCD7280A60BD}" "ShowKeyOutput" = "11:TRUE" @@ -7513,7 +7513,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{BF7B81A5-E348-4F7C-A69F-F74C8EEEAD70}" "ShowKeyOutput" = "11:TRUE" @@ -7528,8 +7528,8 @@ "Tag" = "8:" "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -7541,7 +7541,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{5C6B7D28-E73D-4F71-8FC0-17ADA640EBD8}" "ShowKeyOutput" = "11:TRUE" @@ -7556,8 +7556,8 @@ "Tag" = "8:" "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -7597,7 +7597,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{5307BBB7-EBB9-4AA4-8CB6-A94EC473C8C4}" "ShowKeyOutput" = "11:TRUE" @@ -7625,7 +7625,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{E4081455-398C-4610-A87C-90A8A7D72DC3}" "ShowKeyOutput" = "11:TRUE" @@ -7653,7 +7653,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{FCDF5AE2-EA47-4CC6-9F20-23A0517FEBCB}" "ShowKeyOutput" = "11:TRUE" @@ -7681,7 +7681,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{3D0EB14D-32F3-4D82-9C6D-B806ADBB859C}" "ShowKeyOutput" = "11:TRUE" @@ -7709,7 +7709,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{26511268-2902-4997-8421-ECD7055F9E28}" "ShowKeyOutput" = "11:TRUE" @@ -7737,7 +7737,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{5F78E90B-BD22-47B1-9CA5-7A80F4DF5EF3}" "ShowKeyOutput" = "11:TRUE" @@ -7765,7 +7765,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{6F3C4136-5801-4EBC-AC6E-37DF6FAB150A}" "ShowKeyOutput" = "11:TRUE" @@ -7793,7 +7793,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{CDD9DB83-3BD9-4ED8-BB83-399A2F65F022}" "ShowKeyOutput" = "11:TRUE" @@ -7808,8 +7808,8 @@ "Tag" = "8:" "Folder" = "8:_F71F199B36BB443EA7E2B933AB80A864" "Condition" = "8:" - "Transitive" = "11:FALSE" - "Vital" = "11:TRUE" + "Transitive" = "11:TRUE" + "Vital" = "11:FALSE" "ReadOnly" = "11:FALSE" "Hidden" = "11:FALSE" "System" = "11:FALSE" @@ -7849,7 +7849,7 @@ "IsDependency" = "11:FALSE" "IsolateTo" = "8:" "ProjectOutputGroupRegister" = "3:1" - "OutputConfiguration" = "8:" + "OutputConfiguration" = "8:Release|Win32" "OutputGroupCanonicalName" = "8:Built" "OutputProjectGuid" = "8:{7F059854-568D-4E08-9D00-1E78E203E4DC}" "ShowKeyOutput" = "11:TRUE" From df8c6110849d699053f7a67079e11ccf3e504000 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 19 Jun 2009 13:03:59 +0000 Subject: [PATCH 256/296] GSdx: trying to fix gaps between sprites that may still appear with d3d in native res git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1402 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCapture.cpp | 11 ++++++++--- plugins/GSdx/GSCrc.cpp | 1 + plugins/GSdx/GSCrc.h | 1 + plugins/GSdx/GSRenderer.cpp | 2 ++ plugins/GSdx/GSRendererHW.h | 4 ++-- plugins/GSdx/GSSettingsDlg.cpp | 1 + plugins/GSdx/res/tfx10.fx | 9 ++++++++- plugins/GSdx/res/tfx9.fx | 9 ++++++++- 8 files changed, 31 insertions(+), 7 deletions(-) diff --git a/plugins/GSdx/GSCapture.cpp b/plugins/GSdx/GSCapture.cpp index cf42560cc0..685141bd93 100644 --- a/plugins/GSdx/GSCapture.cpp +++ b/plugins/GSdx/GSCapture.cpp @@ -167,7 +167,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource } }; - CAutoPtr m_output; + GSSourceOutputPin* m_output; public: @@ -178,12 +178,17 @@ public: , m_atpf(10000000i64 / fps) , m_now(0) { - m_output.Attach(new GSSourceOutputPin(m_size, m_atpf, this, this, hr)); + m_output = new GSSourceOutputPin(m_size, m_atpf, this, this, hr); // FIXME if(fps == 60) m_atpf = 166834; // = 10000000i64 / 59.94 } + virtual ~GSSource() + { + delete m_output; + } + DECLARE_IUNKNOWN; int GetPinCount() @@ -193,7 +198,7 @@ public: CBasePin* GetPin(int n) { - return n == 0 ? m_output.m_p : NULL; + return n == 0 ? m_output : NULL; } // IGSSource diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index 7e27702525..dce65c2aa9 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -29,6 +29,7 @@ CRC::Game CRC::m_games[] = {0x42E05BAF, TomoyoAfter, JP}, {0x7800DC84, Clannad, JP}, {0xA6167B59, Lamune, JP}, + {0xDDB59F46, KyuuketsuKitanMoonties, JP}, {0xa39517ab, FFX, EU}, {0xa39517ae, FFX, FR}, {0x941bb7d9, FFX, DE}, diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index b47362e690..c0695d191e 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -31,6 +31,7 @@ public: TomoyoAfter, Clannad, Lamune, + KyuuketsuKitanMoonties, FFX, FFX2, FFXII, diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 288974c0f8..03a96686c9 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -35,6 +35,8 @@ GSRenderer::GSRenderer(uint8* base, bool mt, void (*irq)(), GSDevice* dev) m_aa1 = !!theApp.GetConfig("aa1", 0); m_blur = !!theApp.GetConfig("blur", 0); + if(m_nativeres) m_filter = 2; + s_n = 0; s_dump = !!theApp.GetConfig("dump", 0); s_save = !!theApp.GetConfig("save", 0); diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index f58aa4e787..af3bfefa6c 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -494,9 +494,9 @@ protected: #pragma endregion - #pragma region tomoyo after, clannad, lamune (palette uploaded in a point list, pure genius...) + #pragma region palette uploaded in a point list, pure genius... - if(m_game.title == CRC::TomoyoAfter || m_game.title == CRC::Clannad || m_game.title == CRC::Lamune) + if(m_game.title == CRC::TomoyoAfter || m_game.title == CRC::Clannad || m_game.title == CRC::Lamune || m_game.title == CRC::KyuuketsuKitanMoonties) { if(prim == GS_POINTLIST && !PRIM->TME) { diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 023c473456..3142679119 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -211,6 +211,7 @@ void GSSettingsDlg::UpdateControls() EnableWindow(GetDlgItem(m_hWnd, IDC_RESY), hw && !native); EnableWindow(GetDlgItem(m_hWnd, IDC_RESY_EDIT), hw && !native); EnableWindow(GetDlgItem(m_hWnd, IDC_NATIVERES), hw); + EnableWindow(GetDlgItem(m_hWnd, IDC_FILTER), hw && !native); EnableWindow(GetDlgItem(m_hWnd, IDC_LOGZ), dx9 && hw); EnableWindow(GetDlgItem(m_hWnd, IDC_FBA), dx9 && hw); EnableWindow(GetDlgItem(m_hWnd, IDC_AA1), sw); diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx index 529e528ab7..c0a3a87906 100644 --- a/plugins/GSdx/res/tfx10.fx +++ b/plugins/GSdx/res/tfx10.fx @@ -46,8 +46,15 @@ VS_OUTPUT vs_main(VS_INPUT input) } VS_OUTPUT output; + + // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) + // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty + // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel + // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 + + float4 p = float4(input.p, input.z, 0) - float4(0.05f, 0.05f, 0, 0); - output.p = float4(input.p, input.z, 0) * VertexScale - VertexOffset; + output.p = p * VertexScale - VertexOffset; if(VS_TME == 1) { diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx index c7a2c57896..0c1062fec8 100644 --- a/plugins/GSdx/res/tfx9.fx +++ b/plugins/GSdx/res/tfx9.fx @@ -38,8 +38,15 @@ VS_OUTPUT vs_main(VS_INPUT input) } VS_OUTPUT output; + + // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) + // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty + // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel + // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 + + float4 p = input.p - float4(0.05f, 0.05f, 0, 0); - output.p = input.p * VertexScale - VertexOffset; + output.p = p * VertexScale - VertexOffset; if(VS_LOGZ == 1) { From 254b105f6f02d7858417fe37f12d562425ff6b03 Mon Sep 17 00:00:00 2001 From: refraction Date: Fri, 19 Jun 2009 13:47:05 +0000 Subject: [PATCH 257/296] Just changing a couple of logs back i accidently changed during development git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1403 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/MTGS.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index dfebe33b8f..858d2536b6 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -133,7 +133,7 @@ void SaveState::mtgsFreeze() static void RegHandlerSIGNAL(const u32* data) { - GIF_LOG("MTGS SIGNAL data %x_%x CSRw %x IMR %x CSRr\n",data[0], data[1], CSRw, GSIMR, GSCSRr); + MTGS_LOG("MTGS SIGNAL data %x_%x CSRw %x IMR %x CSRr\n",data[0], data[1], CSRw, GSIMR, GSCSRr); GSSIGLBLID->SIGID = (GSSIGLBLID->SIGID&~data[1])|(data[0]&data[1]); @@ -151,7 +151,7 @@ static void RegHandlerSIGNAL(const u32* data) static void RegHandlerFINISH(const u32* data) { - DevCon::Notice("MTGS FINISH data %x_%x CSRw %x\n", params data[0], data[1], CSRw); + MTGS_LOG("MTGS FINISH data %x_%x CSRw %x\n", params data[0], data[1], CSRw); if ((CSRw & 0x2)) { From 6ef94cc707010db828beb5172ebac941f37dec97 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 19 Jun 2009 20:24:18 +0000 Subject: [PATCH 258/296] GSdx: ignore this... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1404 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPUSettingsDlg.cpp | 9 +- plugins/GSdx/GS.cpp | 14 +- plugins/GSdx/GSDevice.h | 22 + plugins/GSdx/GSDevice10.cpp | 44 +- plugins/GSdx/GSDevice9.cpp | 30 +- plugins/GSdx/GSRenderer.cpp | 1 - plugins/GSdx/GSSettingsDlg.cpp | 24 +- plugins/GSdx/GSTextureFX10.cpp | 40 +- plugins/GSdx/GSTextureFX9.cpp | 4 +- plugins/GSdx/GSUtil.cpp | 19 + plugins/GSdx/GSUtil.h | 1 + plugins/GSdx/GSVertexHW.h | 2 + plugins/GSdx/GSdx_vs2010.vcxproj | 22 +- plugins/GSdx/GSdx_vs2010.vcxproj.filters | 50 +- plugins/GSdx/res/GSdx.rc2 | 12 +- plugins/GSdx/res/{convert10.fx => convert.fx} | 91 ++- plugins/GSdx/res/convert9.fx | 82 -- .../GSdx/res/{interlace10.fx => interlace.fx} | 40 + plugins/GSdx/res/interlace9.fx | 35 - plugins/GSdx/res/merge.fx | 58 ++ plugins/GSdx/res/merge10.fx | 27 - plugins/GSdx/res/merge9.fx | 25 - plugins/GSdx/res/tfx.fx | 717 ++++++++++++++++++ plugins/GSdx/res/tfx10.fx | 472 ------------ plugins/GSdx/res/tfx9.fx | 382 ---------- plugins/GSdx/resource.h | 14 +- plugins/GSdx/stdafx.h | 11 +- plugins/GSdx/vsprops/common.props | 4 +- 28 files changed, 1114 insertions(+), 1138 deletions(-) rename plugins/GSdx/res/{convert10.fx => convert.fx} (51%) delete mode 100644 plugins/GSdx/res/convert9.fx rename plugins/GSdx/res/{interlace10.fx => interlace.fx} (50%) delete mode 100644 plugins/GSdx/res/interlace9.fx create mode 100644 plugins/GSdx/res/merge.fx delete mode 100644 plugins/GSdx/res/merge10.fx delete mode 100644 plugins/GSdx/res/merge9.fx create mode 100644 plugins/GSdx/res/tfx.fx delete mode 100644 plugins/GSdx/res/tfx10.fx delete mode 100644 plugins/GSdx/res/tfx9.fx diff --git a/plugins/GSdx/GPUSettingsDlg.cpp b/plugins/GSdx/GPUSettingsDlg.cpp index bfc1871367..815a6d77f8 100644 --- a/plugins/GSdx/GPUSettingsDlg.cpp +++ b/plugins/GSdx/GPUSettingsDlg.cpp @@ -30,7 +30,8 @@ GSSetting GPUSettingsDlg::g_renderers[] = {0, "Direct3D7 (Software)", NULL}, {1, "Direct3D9 (Software)", NULL}, {2, "Direct3D10 (Software)", NULL}, -// {3, "Null (Null)", NULL}, + {3, "Direct3D11 (Software)", NULL}, +// {4, "Null (Null)", NULL}, }; GSSetting GPUSettingsDlg::g_filter[] = @@ -105,12 +106,14 @@ void GPUSettingsDlg::OnInit() } bool isdx10avail = GSUtil::IsDirect3D10Available(); + bool isdx11avail = GSUtil::IsDirect3D11Available(); vector renderers; for(size_t i = 0; i < countof(g_renderers); i++) { if(i >= 3 && i <= 5 && !isdx10avail) continue; + if(i >= 6 && i <= 8 && !isdx11avail) continue; renderers.push_back(g_renderers[i]); } @@ -186,10 +189,12 @@ void GPUSettingsDlg::UpdateControls() { bool dx9 = i == 1; bool dx10 = i == 2; - bool sw = i >= 0 && i <= 2; + bool dx11 = i == 3; + bool sw = i >= 0 && i <= 3; ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); + // TODO: ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO11), dx11 ? SW_SHOW : SW_HIDE); EnableWindow(GetDlgItem(m_hWnd, IDC_SCALE), sw); EnableWindow(GetDlgItem(m_hWnd, IDC_SWTHREADS_EDIT), sw); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index bf7ce343c9..430bb23407 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -23,6 +23,7 @@ #include "GSUtil.h" #include "GSRendererHW9.h" #include "GSRendererHW10.h" +#include "GSRendererHW11.h" #include "GSRendererOGL.h" #include "GSRendererSW.h" #include "GSRendererNull.h" @@ -126,13 +127,16 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; + case 6: s_gs = new GSRendererHW11(s_basemem, !!mt, s_irq); break; + case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice11()); break; + case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice11()); break; #if 0 - case 6: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; - case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; - case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; + case 9: s_gs = new GSRendererOGL(s_basemem, !!mt, s_irq); break; + case 10: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; + case 11: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceOGL()); break; #endif - case 9: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; - case 10: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; + case 12: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; + case 13: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDeviceNull()); break; } if(!s_gs->Create(title)) diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index d512fd37b5..59799e814d 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -103,4 +103,26 @@ public: void Merge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, const GSVector2i& fs, bool slbg, bool mmod, const GSVector4& c); void Interlace(const GSVector2i& ds, int field, int mode, float yoffset); + + template void PrepareShaderMacro(vector& dst, const T* src, const char* model) + { + dst.clear(); + + while(src && src->Definition && src->Name) + { + dst.push_back(*src++); + } + + T m; + + m.Name = "SHADER_MODEL"; + m.Definition = model; + + dst.push_back(m); + + m.Name = NULL; + m.Definition = NULL; + + dst.push_back(m); + } }; diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 084b8d191a..7f3fbf2791 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -96,16 +96,6 @@ bool GSDevice10::Create(GSWnd* wnd, bool vsync) if(FAILED(hr)) return false; - // font -/* - // TODO: the driver crashes on alt-enter when using a font... - - D3DX10_FONT_DESC fd; - memset(&fd, 0, sizeof(fd)); - _tcscpy(fd.FaceName, _T("Arial")); - fd.Height = 20; - D3DX10CreateFontIndirect(m_dev, &fd, &m_font); -*/ // convert D3D10_INPUT_ELEMENT_DESC il_convert[] = @@ -114,11 +104,11 @@ bool GSDevice10::Create(GSWnd* wnd, bool vsync) {"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 16, D3D10_INPUT_PER_VERTEX_DATA, 0}, }; - hr = CompileShader(IDR_CONVERT10_FX, "vs_main", NULL, &m_convert.vs, il_convert, countof(il_convert), &m_convert.il); + hr = CompileShader(IDR_CONVERT_FX, "vs_main", NULL, &m_convert.vs, il_convert, countof(il_convert), &m_convert.il); for(int i = 0; i < countof(m_convert.ps); i++) { - hr = CompileShader(IDR_CONVERT10_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); + hr = CompileShader(IDR_CONVERT_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } memset(&dsd, 0, sizeof(dsd)); @@ -142,14 +132,12 @@ bool GSDevice10::Create(GSWnd* wnd, bool vsync) bd.ByteWidth = sizeof(MergeConstantBuffer); bd.Usage = D3D10_USAGE_DEFAULT; bd.BindFlags = D3D10_BIND_CONSTANT_BUFFER; - bd.CPUAccessFlags = 0; - bd.MiscFlags = 0; hr = m_dev->CreateBuffer(&bd, NULL, &m_merge.cb); for(int i = 0; i < countof(m_merge.ps); i++) { - hr = CompileShader(IDR_MERGE10_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); + hr = CompileShader(IDR_MERGE_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); } memset(&bsd, 0, sizeof(bsd)); @@ -172,14 +160,12 @@ bool GSDevice10::Create(GSWnd* wnd, bool vsync) bd.ByteWidth = sizeof(InterlaceConstantBuffer); bd.Usage = D3D10_USAGE_DEFAULT; bd.BindFlags = D3D10_BIND_CONSTANT_BUFFER; - bd.CPUAccessFlags = 0; - bd.MiscFlags = 0; hr = m_dev->CreateBuffer(&bd, NULL, &m_interlace.cb); for(int i = 0; i < countof(m_interlace.ps); i++) { - hr = CompileShader(IDR_INTERLACE10_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); + hr = CompileShader(IDR_INTERLACE_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); } // @@ -743,13 +729,17 @@ void GSDevice10::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) } } -HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** ps, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) +HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_MACRO* macro, ID3D10VertexShader** vs, D3D10_INPUT_ELEMENT_DESC* layout, int count, ID3D10InputLayout** il) { HRESULT hr; + vector m; + + PrepareShaderMacro(m, macro, "0x400"); + CComPtr shader, error; - hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "vs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), "vs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -761,7 +751,7 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M return hr; } - hr = m_dev->CreateVertexShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), ps); + hr = m_dev->CreateVertexShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), vs); if(FAILED(hr)) { @@ -782,9 +772,13 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M { HRESULT hr; + vector m; + + PrepareShaderMacro(m, macro, "0x400"); + CComPtr shader, error; - hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "gs_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), "gs_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { @@ -810,9 +804,13 @@ HRESULT GSDevice10::CompileShader(uint32 id, const string& entry, D3D10_SHADER_M { HRESULT hr; + vector m; + + PrepareShaderMacro(m, macro, "0x400"); + CComPtr shader, error; - hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, macro, NULL, entry.c_str(), "ps_4_0", 0, 0, NULL, &shader, &error, NULL); + hr = D3DX10CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), "ps_4_0", 0, 0, NULL, &shader, &error, NULL); if(error) { diff --git a/plugins/GSdx/GSDevice9.cpp b/plugins/GSdx/GSDevice9.cpp index a88a24f97e..37ef14415c 100644 --- a/plugins/GSdx/GSDevice9.cpp +++ b/plugins/GSdx/GSDevice9.cpp @@ -133,11 +133,11 @@ bool GSDevice9::Create(GSWnd* wnd, bool vsync) D3DDECL_END() }; - CompileShader(IDR_CONVERT9_FX, "vs_main", NULL, &m_convert.vs, il_convert, countof(il_convert), &m_convert.il); + CompileShader(IDR_CONVERT_FX, "vs_main", NULL, &m_convert.vs, il_convert, countof(il_convert), &m_convert.il); for(int i = 0; i < countof(m_convert.ps); i++) { - CompileShader(IDR_CONVERT9_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); + CompileShader(IDR_CONVERT_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); } m_convert.dss.DepthEnable = false; @@ -164,7 +164,7 @@ bool GSDevice9::Create(GSWnd* wnd, bool vsync) for(int i = 0; i < countof(m_merge.ps); i++) { - CompileShader(IDR_MERGE9_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); + CompileShader(IDR_MERGE_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); } m_merge.bs.BlendEnable = true; @@ -180,7 +180,7 @@ bool GSDevice9::Create(GSWnd* wnd, bool vsync) for(int i = 0; i < countof(m_interlace.ps); i++) { - CompileShader(IDR_INTERLACE9_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); + CompileShader(IDR_INTERLACE_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); } // @@ -952,25 +952,32 @@ static HRESULT LoadShader(uint32 id, LPCSTR& data, uint32& size) HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DVertexShader9** vs, const D3DVERTEXELEMENT9* layout, int count, IDirect3DVertexDeclaration9** il) { const char* target; + const char* model; if(m_d3dcaps.VertexShaderVersion >= D3DVS_VERSION(3, 0)) { target = "vs_3_0"; + model = "0x300"; } else if(m_d3dcaps.VertexShaderVersion >= D3DVS_VERSION(2, 0)) { target = "vs_2_0"; + model = "0x200"; } else { return E_FAIL; } + vector m; + + PrepareShaderMacro(m, macro, model); + HRESULT hr; CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), &m[0], NULL, entry.c_str(), target, 0, &shader, &error, NULL); LPCSTR data; uint32 size; @@ -979,7 +986,7 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO if(FAILED(hr)) return E_FAIL; - hr = D3DXCompileShader(data, size, macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); + hr = D3DXCompileShader(data, size, &m[0], NULL, entry.c_str(), target, 0, &shader, &error, NULL); if(SUCCEEDED(hr)) { @@ -1010,27 +1017,34 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO* macro, IDirect3DPixelShader9** ps) { const char* target = NULL; + const char* model; uint32 flags = 0; if(m_d3dcaps.PixelShaderVersion >= D3DPS_VERSION(3, 0)) { target = "ps_3_0"; + model = "0x300"; flags |= D3DXSHADER_AVOID_FLOW_CONTROL; } else if(m_d3dcaps.PixelShaderVersion >= D3DPS_VERSION(2, 0)) { target = "ps_2_0"; + model = "0x200"; } else { return false; } + vector m; + + PrepareShaderMacro(m, macro, model); + HRESULT hr; CComPtr shader, error; - // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), macro, NULL, entry.c_str(), target, flags, &shader, &error, NULL); + // FIXME: hr = D3DXCompileShaderFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), &m[0], NULL, entry.c_str(), target, flags, &shader, &error, NULL); LPCSTR data; uint32 size; @@ -1039,7 +1053,7 @@ HRESULT GSDevice9::CompileShader(uint32 id, const string& entry, const D3DXMACRO if(FAILED(hr)) return E_FAIL; - hr = D3DXCompileShader(data, size, macro, NULL, entry.c_str(), target, 0, &shader, &error, NULL); + hr = D3DXCompileShader(data, size, &m[0], NULL, entry.c_str(), target, 0, &shader, &error, NULL); if(SUCCEEDED(hr)) { diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 03a96686c9..8af6720dfc 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -251,7 +251,6 @@ bool GSRenderer::Merge(int field) return true; } - void GSRenderer::VSync(int field) { GSPerfMonAutoTimer pmat(m_perfmon); diff --git a/plugins/GSdx/GSSettingsDlg.cpp b/plugins/GSdx/GSSettingsDlg.cpp index 3142679119..002c5ae1e7 100644 --- a/plugins/GSdx/GSSettingsDlg.cpp +++ b/plugins/GSdx/GSSettingsDlg.cpp @@ -33,13 +33,16 @@ GSSetting GSSettingsDlg::g_renderers[] = {3, "Direct3D10 (Hardware)", NULL}, {4, "Direct3D10 (Software)", NULL}, {5, "Direct3D10 (Null)", NULL}, + {6, "Direct3D11 (Hardware)", NULL}, + {7, "Direct3D11 (Software)", NULL}, + {8, "Direct3D11 (Null)", NULL}, #if 0 - {6, "OpenGL (Hardware)", NULL}, - {7, "OpenGL (Software)", NULL}, - {8, "OpenGL (Null)", NULL}, + {9, "OpenGL (Hardware)", NULL}, + {10, "OpenGL (Software)", NULL}, + {11, "OpenGL (Null)", NULL}, #endif - {9, "Null (Software)", NULL}, - {10, "Null (Null)", NULL}, + {12, "Null (Software)", NULL}, + {13, "Null (Null)", NULL}, }; GSSetting GSSettingsDlg::g_interlace[] = @@ -101,12 +104,14 @@ void GSSettingsDlg::OnInit() } bool isdx10avail = GSUtil::IsDirect3D10Available(); + bool isdx11avail = GSUtil::IsDirect3D11Available(); vector renderers; for(size_t i = 0; i < countof(g_renderers); i++) { if(i >= 3 && i <= 5 && !isdx10avail) continue; + if(i >= 6 && i <= 8 && !isdx11avail) continue; renderers.push_back(g_renderers[i]); } @@ -197,13 +202,16 @@ void GSSettingsDlg::UpdateControls() { bool dx9 = i >= 0 && i <= 2; bool dx10 = i >= 3 && i <= 5; - bool ogl = i >= 6 && i <= 8; - bool hw = i == 0 || i == 3 || i == 6; - bool sw = i == 1 || i == 4 || i == 7; + bool dx11 = i >= 6 && i <= 8; + bool ogl = i >= 9 && i <= 12; + bool hw = i == 0 || i == 3 || i == 6 || i == 9; + bool sw = i == 1 || i == 4 || i == 7 || i == 10; bool native = !!IsDlgButtonChecked(m_hWnd, IDC_NATIVERES); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO9), dx9 ? SW_SHOW : SW_HIDE); ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO10), dx10 ? SW_SHOW : SW_HIDE); + // TODO: ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO11), dx11 ? SW_SHOW : SW_HIDE); + // TODO: ShowWindow(GetDlgItem(m_hWnd, IDC_LOGO_OGL), ogl ? SW_SHOW : SW_HIDE); EnableWindow(GetDlgItem(m_hWnd, IDC_RESOLUTION), dx9); EnableWindow(GetDlgItem(m_hWnd, IDC_RESX), hw && !native); diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index a3b04e1cc9..3005ca5918 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -72,7 +72,7 @@ bool GSTextureFX10::Create(GSDevice10* dev) memset(&sd, 0, sizeof(sd)); - sd.Filter = D3D10_ENCODE_BASIC_FILTER(D3D10_FILTER_TYPE_POINT, D3D10_FILTER_TYPE_POINT, D3D10_FILTER_TYPE_POINT, false); + sd.Filter = D3D10_FILTER_MIN_MAG_MIP_POINT; sd.AddressU = D3D10_TEXTURE_ADDRESS_CLAMP; sd.AddressV = D3D10_TEXTURE_ADDRESS_CLAMP; sd.AddressW = D3D10_TEXTURE_ADDRESS_CLAMP; @@ -133,7 +133,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) CComPtr il; CComPtr vs; - m_dev->CompileShader(IDR_TFX10_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); if(m_il == NULL) { @@ -178,12 +178,12 @@ bool GSTextureFX10::SetupGS(GSSelector sel) D3D10_SHADER_MACRO macro[] = { - {"IIP", str[0].c_str()}, - {"PRIM", str[1].c_str()}, + {"GS_IIP", str[0].c_str()}, + {"GS_PRIM", str[1].c_str()}, {NULL, NULL}, }; - hr = m_dev->CompileShader(IDR_TFX10_FX, "gs_main", macro, &gs); + hr = m_dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); m_gs[sel] = gs; } @@ -230,26 +230,26 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl D3D10_SHADER_MACRO macro[] = { - {"FST", str[0].c_str()}, - {"WMS", str[1].c_str()}, - {"WMT", str[2].c_str()}, - {"BPP", str[3].c_str()}, - {"AEM", str[4].c_str()}, - {"TFX", str[5].c_str()}, - {"TCC", str[6].c_str()}, - {"ATE", str[7].c_str()}, - {"ATST", str[8].c_str()}, - {"FOG", str[9].c_str()}, - {"CLR1", str[10].c_str()}, - {"FBA", str[11].c_str()}, - {"AOUT", str[12].c_str()}, - {"LTF", str[13].c_str()}, + {"PS_FST", str[0].c_str()}, + {"PS_WMS", str[1].c_str()}, + {"PS_WMT", str[2].c_str()}, + {"PS_BPP", str[3].c_str()}, + {"PS_AEM", str[4].c_str()}, + {"PS_TFX", str[5].c_str()}, + {"PS_TCC", str[6].c_str()}, + {"PS_ATE", str[7].c_str()}, + {"PS_ATST", str[8].c_str()}, + {"PS_FOG", str[9].c_str()}, + {"PS_CLR1", str[10].c_str()}, + {"PS_FBA", str[11].c_str()}, + {"PS_AOUT", str[12].c_str()}, + {"PS_LTF", str[13].c_str()}, {NULL, NULL}, }; CComPtr ps; - hr = m_dev->CompileShader(IDR_TFX10_FX, "ps_main", macro, &ps); + hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); m_ps[sel] = ps; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index c303db9652..24694a33ca 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -126,7 +126,7 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) CComPtr il; CComPtr vs; - m_dev->CompileShader(IDR_TFX9_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); if(m_il == NULL) { @@ -215,7 +215,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample CComPtr ps; - hr = m_dev->CompileShader(IDR_TFX9_FX, "ps_main", macro, &ps); + hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); m_ps[sel] = ps; diff --git a/plugins/GSdx/GSUtil.cpp b/plugins/GSdx/GSUtil.cpp index 64ce305c5d..4b38666b5c 100644 --- a/plugins/GSdx/GSUtil.cpp +++ b/plugins/GSdx/GSUtil.cpp @@ -172,6 +172,25 @@ bool GSUtil::IsDirect3D10Available() return false; } +bool GSUtil::IsDirect3D11Available() +{ + if(HMODULE hModule = LoadLibrary(_T("d3d11.dll"))) + { + FreeLibrary(hModule); + + return true; + } + + if(HMODULE hModule = LoadLibrary(_T("d3d11_beta.dll"))) + { + FreeLibrary(hModule); + + return true; + } + + return false; +} + char* GSUtil::GetLibName() { static string str; diff --git a/plugins/GSdx/GSUtil.h b/plugins/GSdx/GSUtil.h index bd3f580c6b..f7d81539b4 100644 --- a/plugins/GSdx/GSUtil.h +++ b/plugins/GSdx/GSUtil.h @@ -60,6 +60,7 @@ public: static bool CheckSSE(); static bool IsDirect3D10Available(); + static bool IsDirect3D11Available(); static char* GetLibName(); }; diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index 2e49cd2d20..76a84ed74d 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -80,6 +80,8 @@ __declspec(align(16)) union GSVertexHW10 float GetQ() {return q;} }; +typedef GSVertexHW10 GSVertexHW11; // TODO + typedef GSVertexHW9 GSVertexOGL; // TODO #pragma pack(pop) diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj b/plugins/GSdx/GSdx_vs2010.vcxproj index d4e7a8e2b2..8a7fc45bba 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj +++ b/plugins/GSdx/GSdx_vs2010.vcxproj @@ -1303,6 +1303,7 @@ + @@ -1320,6 +1321,7 @@ + @@ -1331,14 +1333,17 @@ + + + @@ -1419,6 +1424,7 @@ + @@ -1438,6 +1444,7 @@ + @@ -1450,14 +1457,17 @@ + + + @@ -1481,17 +1491,13 @@ - - + - - + - - - - + + diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj.filters b/plugins/GSdx/GSdx_vs2010.vcxproj.filters index 5cfd89b885..32c080fcf0 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj.filters +++ b/plugins/GSdx/GSdx_vs2010.vcxproj.filters @@ -150,6 +150,9 @@ Source Files + + Source Files + Source Files @@ -201,6 +204,9 @@ Source Files + + Source Files + Source Files @@ -234,6 +240,9 @@ Source Files + + Source Files + Source Files @@ -246,6 +255,9 @@ Source Files + + Source Files + Source Files @@ -258,6 +270,9 @@ Source Files + + Source Files + Source Files @@ -443,6 +458,9 @@ Header Files + + Header Files + Header Files @@ -500,6 +518,9 @@ Header Files + + Header Files + Header Files @@ -536,6 +557,9 @@ Header Files + + Header Files + Header Files @@ -548,6 +572,9 @@ Header Files + + Header Files + Header Files @@ -560,6 +587,9 @@ Header Files + + Header Files + Header Files @@ -625,19 +655,13 @@ Source Files - - Shaders - - + Shaders Resource Files - - Shaders - - + Shaders @@ -646,16 +670,10 @@ Resource Files - + Shaders - - Shaders - - - Shaders - - + Shaders diff --git a/plugins/GSdx/res/GSdx.rc2 b/plugins/GSdx/res/GSdx.rc2 index 45a9aad126..793e0342dd 100644 --- a/plugins/GSdx/res/GSdx.rc2 +++ b/plugins/GSdx/res/GSdx.rc2 @@ -10,13 +10,9 @@ ///////////////////////////////////////////////////////////////////////////// // Add manually edited resources here... -IDR_CONVERT9_FX RCDATA "res\\convert9.fx" -IDR_TFX9_FX RCDATA "res\\tfx9.fx" -IDR_MERGE9_FX RCDATA "res\\merge9.fx" -IDR_INTERLACE9_FX RCDATA "res\\interlace9.fx" -IDR_CONVERT10_FX RCDATA "res\\convert10.fx" -IDR_TFX10_FX RCDATA "res\\tfx10.fx" -IDR_MERGE10_FX RCDATA "res\\merge10.fx" -IDR_INTERLACE10_FX RCDATA "res\\interlace10.fx" +IDR_CONVERT_FX RCDATA "res\\convert.fx" +IDR_TFX_FX RCDATA "res\\tfx.fx" +IDR_MERGE_FX RCDATA "res\\merge.fx" +IDR_INTERLACE_FX RCDATA "res\\interlace.fx" ///////////////////////////////////////////////////////////////////////////// diff --git a/plugins/GSdx/res/convert10.fx b/plugins/GSdx/res/convert.fx similarity index 51% rename from plugins/GSdx/res/convert10.fx rename to plugins/GSdx/res/convert.fx index bcdfb22d7a..9749476e6d 100644 --- a/plugins/GSdx/res/convert10.fx +++ b/plugins/GSdx/res/convert.fx @@ -1,3 +1,5 @@ +#if SHADER_MODEL >= 0x400 + struct VS_INPUT { float4 p : POSITION; @@ -92,4 +94,91 @@ float4 ps_main6(PS_INPUT input) : SV_Target0 // diagonal uint4 p = (uint4)input.p; return ps_crt(input, (p.x + (p.y % 3)) % 3); -} \ No newline at end of file +} + +#elif SHADER_MODEL <= 0x300 + +struct VS_INPUT +{ + float4 p : POSITION; + float2 t : TEXCOORD0; +}; + +struct VS_OUTPUT +{ + float4 p : POSITION; + float2 t : TEXCOORD0; +}; + +VS_OUTPUT vs_main(VS_INPUT input) +{ + VS_OUTPUT output; + + output.p = input.p; + output.t = input.t; + + return output; +} + +sampler Texture : register(s0); + +float4 ps_main0(float2 t : TEXCOORD0) : COLOR +{ + return tex2D(Texture, t); +} + +float4 ps_main1(float2 t : TEXCOORD0) : COLOR +{ + float4 c = tex2D(Texture, t); + c.a *= 128.0f / 255; // *= 0.5f is no good here, need to do this in order to get 0x80 for 1.0f (instead of 0x7f) + return c; +} + +float4 ps_main2(float2 t : TEXCOORD0) : COLOR +{ + clip(tex2D(Texture, t).a - (1.0f - 0.9f/256)); + + return 0; +} + +float4 ps_main3(float2 t : TEXCOORD0) : COLOR +{ + clip((1.0f - 0.9f/256) - tex2D(Texture, t).a); + + return 0; +} + +float4 ps_main4() : COLOR +{ + return 1; +} + +float4 ps_crt(float2 t, int i) +{ + float4 mask[4] = + { + float4(1, 0, 0, 0), + float4(0, 1, 0, 0), + float4(0, 0, 1, 0), + float4(1, 1, 1, 0) + }; + + return tex2D(Texture, t) * saturate(mask[i] + 0.5f); +} + +float4 ps_main5(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // triangular +{ + int4 p = (int4)vPos; + + // return ps_crt(t, ((p.x + (p.y % 2) * 3) / 2) % 3); + return ps_crt(t, ((p.x + ((p.y / 2) % 2) * 3) / 2) % 3); +} + +float4 ps_main6(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // diagonal +{ + int4 p = (int4)vPos; + + return ps_crt(t, (p.x + (p.y % 3)) % 3); +} + +#endif \ No newline at end of file diff --git a/plugins/GSdx/res/convert9.fx b/plugins/GSdx/res/convert9.fx deleted file mode 100644 index 1341091681..0000000000 --- a/plugins/GSdx/res/convert9.fx +++ /dev/null @@ -1,82 +0,0 @@ -struct VS_INPUT -{ - float4 p : POSITION; - float2 t : TEXCOORD0; -}; - -struct VS_OUTPUT -{ - float4 p : POSITION; - float2 t : TEXCOORD0; -}; - -VS_OUTPUT vs_main(VS_INPUT input) -{ - VS_OUTPUT output; - - output.p = input.p; - output.t = input.t; - - return output; -} - -sampler Texture : register(s0); - -float4 ps_main0(float2 t : TEXCOORD0) : COLOR -{ - return tex2D(Texture, t); -} - -float4 ps_main1(float2 t : TEXCOORD0) : COLOR -{ - float4 c = tex2D(Texture, t); - c.a *= 128.0f / 255; // *= 0.5f is no good here, need to do this in order to get 0x80 for 1.0f (instead of 0x7f) - return c; -} - -float4 ps_main2(float2 t : TEXCOORD0) : COLOR -{ - clip(tex2D(Texture, t).a - (1.0f - 0.9f/256)); - - return 0; -} - -float4 ps_main3(float2 t : TEXCOORD0) : COLOR -{ - clip((1.0f - 0.9f/256) - tex2D(Texture, t).a); - - return 0; -} - -float4 ps_main4() : COLOR -{ - return 1; -} - -float4 ps_crt(float2 t, int i) -{ - float4 mask[4] = - { - float4(1, 0, 0, 0), - float4(0, 1, 0, 0), - float4(0, 0, 1, 0), - float4(1, 1, 1, 0) - }; - - return tex2D(Texture, t) * saturate(mask[i] + 0.5f); -} - -float4 ps_main5(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // triangular -{ - int4 p = (int4)vPos; - - // return ps_crt(t, ((p.x + (p.y % 2) * 3) / 2) % 3); - return ps_crt(t, ((p.x + ((p.y / 2) % 2) * 3) / 2) % 3); -} - -float4 ps_main6(float2 t : TEXCOORD0, float4 vPos : VPOS) : COLOR // diagonal -{ - int4 p = (int4)vPos; - - return ps_crt(t, (p.x + (p.y % 3)) % 3); -} diff --git a/plugins/GSdx/res/interlace10.fx b/plugins/GSdx/res/interlace.fx similarity index 50% rename from plugins/GSdx/res/interlace10.fx rename to plugins/GSdx/res/interlace.fx index 86fa44dd75..f726051260 100644 --- a/plugins/GSdx/res/interlace10.fx +++ b/plugins/GSdx/res/interlace.fx @@ -1,3 +1,4 @@ +#if SHADER_MODEL >= 0x400 Texture2D Texture; SamplerState Sampler; @@ -41,3 +42,42 @@ float4 ps_main3(PS_INPUT input) : SV_Target0 { return Texture.Sample(Sampler, input.t); } + +#elif SHADER_MODEL <= 0x300 + +sampler s0 : register(s0); + +float4 Params1 : register(c0); + +#define ZrH (Params1.xy) +#define hH (Params1.z) + +float4 ps_main0(float2 tex : TEXCOORD0) : COLOR +{ + clip(frac(tex.y * hH) - 0.5); + + return tex2D(s0, tex); +} + +float4 ps_main1(float2 tex : TEXCOORD0) : COLOR +{ + clip(0.5 - frac(tex.y * hH)); + + return tex2D(s0, tex); +} + +float4 ps_main2(float2 tex : TEXCOORD0) : COLOR +{ + float4 c0 = tex2D(s0, tex - ZrH); + float4 c1 = tex2D(s0, tex); + float4 c2 = tex2D(s0, tex + ZrH); + + return (c0 + c1 * 2 + c2) / 4; +} + +float4 ps_main3(float2 tex : TEXCOORD0) : COLOR +{ + return tex2D(s0, tex); +} + +#endif diff --git a/plugins/GSdx/res/interlace9.fx b/plugins/GSdx/res/interlace9.fx deleted file mode 100644 index 5f5db8bf9c..0000000000 --- a/plugins/GSdx/res/interlace9.fx +++ /dev/null @@ -1,35 +0,0 @@ - -sampler s0 : register(s0); - -float4 Params1 : register(c0); - -#define ZrH (Params1.xy) -#define hH (Params1.z) - -float4 ps_main0(float2 tex : TEXCOORD0) : COLOR -{ - clip(frac(tex.y * hH) - 0.5); - - return tex2D(s0, tex); -} - -float4 ps_main1(float2 tex : TEXCOORD0) : COLOR -{ - clip(0.5 - frac(tex.y * hH)); - - return tex2D(s0, tex); -} - -float4 ps_main2(float2 tex : TEXCOORD0) : COLOR -{ - float4 c0 = tex2D(s0, tex - ZrH); - float4 c1 = tex2D(s0, tex); - float4 c2 = tex2D(s0, tex + ZrH); - - return (c0 + c1 * 2 + c2) / 4; -} - -float4 ps_main3(float2 tex : TEXCOORD0) : COLOR -{ - return tex2D(s0, tex); -} diff --git a/plugins/GSdx/res/merge.fx b/plugins/GSdx/res/merge.fx new file mode 100644 index 0000000000..fc014c4a4e --- /dev/null +++ b/plugins/GSdx/res/merge.fx @@ -0,0 +1,58 @@ +#if SHADER_MODEL >= 0x400 + +Texture2D Texture; +SamplerState Sampler; + +cbuffer cb0 +{ + float4 BGColor; +}; + +struct PS_INPUT +{ + float4 p : SV_Position; + float2 t : TEXCOORD0; +}; + +float4 ps_main0(PS_INPUT input) : SV_Target0 +{ + float4 c = Texture.Sample(Sampler, input.t); + c.a = min(c.a * 2, 1); + return c; +} + +float4 ps_main1(PS_INPUT input) : SV_Target0 +{ + float4 c = Texture.Sample(Sampler, input.t); + c.a = BGColor.a; + return c; +} + +#elif SHADER_MODEL <= 0x300 + +sampler Texture : register(s0); + +float4 g_params[1]; + +#define BGColor (g_params[0]) + +struct PS_INPUT +{ + float2 t : TEXCOORD0; +}; + +float4 ps_main0(PS_INPUT input) : COLOR +{ + float4 c = tex2D(Texture, input.t); + // a = ; + return c.bgra; +} + +float4 ps_main1(PS_INPUT input) : COLOR +{ + float4 c = tex2D(Texture, input.t); + c.a = BGColor.a; + return c.bgra; +} + +#endif \ No newline at end of file diff --git a/plugins/GSdx/res/merge10.fx b/plugins/GSdx/res/merge10.fx deleted file mode 100644 index 1f91fe3e8f..0000000000 --- a/plugins/GSdx/res/merge10.fx +++ /dev/null @@ -1,27 +0,0 @@ -Texture2D Texture; -SamplerState Sampler; - -cbuffer cb0 -{ - float4 BGColor; -}; - -struct PS_INPUT -{ - float4 p : SV_Position; - float2 t : TEXCOORD0; -}; - -float4 ps_main0(PS_INPUT input) : SV_Target0 -{ - float4 c = Texture.Sample(Sampler, input.t); - c.a = min(c.a * 2, 1); - return c; -} - -float4 ps_main1(PS_INPUT input) : SV_Target0 -{ - float4 c = Texture.Sample(Sampler, input.t); - c.a = BGColor.a; - return c; -} \ No newline at end of file diff --git a/plugins/GSdx/res/merge9.fx b/plugins/GSdx/res/merge9.fx deleted file mode 100644 index dbb210660f..0000000000 --- a/plugins/GSdx/res/merge9.fx +++ /dev/null @@ -1,25 +0,0 @@ -sampler Texture : register(s0); - -float4 g_params[1]; - -#define BGColor (g_params[0]) - -struct PS_INPUT -{ - float2 t : TEXCOORD0; -}; - -float4 ps_main0(PS_INPUT input) : COLOR -{ - float4 c = tex2D(Texture, input.t); - // a = ; - return c.bgra; -} - -float4 ps_main1(PS_INPUT input) : COLOR -{ - float4 c = tex2D(Texture, input.t); - c.a = BGColor.a; - return c.bgra; -} - diff --git a/plugins/GSdx/res/tfx.fx b/plugins/GSdx/res/tfx.fx new file mode 100644 index 0000000000..ebd24e290b --- /dev/null +++ b/plugins/GSdx/res/tfx.fx @@ -0,0 +1,717 @@ +#if SHADER_MODEL >= 0x400 + +#ifndef VS_BPPZ +#define VS_BPPZ 0 +#define VS_TME 1 +#define VS_FST 1 +#define VS_PRIM 0 +#endif + +#ifndef GS_IIP +#define GS_IIP 0 +#define GS_PRIM 3 +#endif + +#ifndef PS_FST +#define PS_FST 0 +#define PS_WMS 3 +#define PS_WMT 3 +#define PS_BPP 0 +#define PS_AEM 0 +#define PS_TFX 0 +#define PS_TCC 1 +#define PS_ATE 0 +#define PS_ATST 4 +#define PS_FOG 0 +#define PS_CLR1 0 +#define PS_FBA 0 +#define PS_AOUT 0 +#define PS_LTF 1 +#endif + +struct VS_INPUT +{ + uint2 p : POSITION0; + uint z : POSITION1; + float2 t : TEXCOORD0; + float q : TEXCOORD1; + float4 c : COLOR0; + float4 f : COLOR1; +}; + +struct VS_OUTPUT +{ + float4 p : SV_Position; + float4 t : TEXCOORD0; + float4 c : COLOR0; +}; + +struct PS_INPUT +{ + float4 p : SV_Position; + float4 t : TEXCOORD0; + float4 c : COLOR0; +}; + +struct PS_OUTPUT +{ + float4 c0 : SV_Target0; + float4 c1 : SV_Target1; +}; + +Texture2D Texture; +Texture2D Palette; +SamplerState TextureSampler; +SamplerState PaletteSampler; + +cbuffer cb0 +{ + float4 VertexScale; + float4 VertexOffset; + float2 TextureScale; +}; + +cbuffer cb1 +{ + float3 FogColor; + float AREF; + float4 HalfTexel; + float2 WH; + float2 TA; + float4 MinMax; + float4 MinMaxF; + uint4 MskFix; +}; + +#elif SHADER_MODEL <= 0x300 + +#ifndef VS_BPPZ +#define VS_BPPZ 0 +#define VS_TME 1 +#define VS_FST 1 +#define VS_LOGZ 1 +#endif + +#ifndef PS_FST +#define PS_FST 0 +#define PS_WMS 3 +#define PS_WMT 3 +#define PS_BPP 0 +#define PS_AEM 0 +#define PS_TFX 0 +#define PS_TCC 1 +#define PS_ATE 0 +#define PS_ATST 4 +#define PS_FOG 0 +#define PS_CLR1 0 +#define PS_RT 0 +#define PS_LTF 0 +#endif + +struct VS_INPUT +{ + float4 p : POSITION0; + float2 t : TEXCOORD0; + float4 c : COLOR0; + float4 f : COLOR1; +}; + +struct VS_OUTPUT +{ + float4 p : POSITION; + float4 t : TEXCOORD0; + float4 c : COLOR0; +}; + +struct PS_INPUT +{ + float4 t : TEXCOORD0; + float4 c : COLOR0; +}; + +sampler Texture : register(s0); +sampler1D Palette : register(s1); +sampler1D UMSKFIX : register(s2); +sampler1D VMSKFIX : register(s3); + +float4 vs_params[3]; + +#define VertexScale vs_params[0] +#define VertexOffset vs_params[1] +#define TextureScale vs_params[2].xy + +float4 ps_params[5]; + +#define FogColor ps_params[0].bgr +#define AREF ps_params[0].a +#define HalfTexel ps_params[1] +#define WH ps_params[2].xy +#define TA0 ps_params[2].z +#define TA1 ps_params[2].w +#define MinMax ps_params[3] +#define MinMaxF ps_params[4] + +#endif + +float4 wrapuv(float4 uv) +{ + if(PS_WMS == PS_WMT) + { + if(PS_WMS == 0) + { + uv = frac(uv); + } + else if(PS_WMS == 1) + { + uv = saturate(uv); + } + else if(PS_WMS == 2) + { + uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); + } + else if(PS_WMS == 3) + { + #if SHADER_MODEL >= 0x400 + uv = (float4)(((int4)(uv * WH.xyxy) & MskFix.xyxy) | MskFix.zwzw) / WH.xyxy; + #elif SHADER_MODEL <= 0x300 + uv.x = tex1D(UMSKFIX, uv.x); + uv.y = tex1D(VMSKFIX, uv.y); + uv.z = tex1D(UMSKFIX, uv.z); + uv.w = tex1D(VMSKFIX, uv.w); + #endif + } + } + else + { + if(PS_WMS == 0) + { + uv.xz = frac(uv.xz); + } + else if(PS_WMS == 1) + { + uv.xz = saturate(uv.xz); + } + else if(PS_WMS == 2) + { + uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); + } + else if(PS_WMS == 3) + { + #if SHADER_MODEL >= 0x400 + uv.xz = (float2)(((int2)(uv * WH.xyxy).xz & MskFix.xx) | MskFix.zz) / WH; + #elif SHADER_MODEL <= 0x300 + uv.x = tex1D(UMSKFIX, uv.x); + uv.z = tex1D(UMSKFIX, uv.z); + #endif + } + + if(PS_WMT == 0) + { + uv.yw = frac(uv.yw); + } + else if(PS_WMT == 1) + { + uv.yw = saturate(uv.yw); + } + else if(PS_WMT == 2) + { + uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); + } + else if(PS_WMT == 3) + { + #if SHADER_MODEL >= 0x400 + uv.yw = (float2)(((int2)(uv * WH.xyxy).yw & MskFix.yy) | MskFix.ww) / WH; + #elif SHADER_MODEL <= 0x300 + uv.y = tex1D(VMSKFIX, uv.y); + uv.w = tex1D(VMSKFIX, uv.w); + #endif + } + } + + return uv; +} + +float2 clampuv(float2 uv) +{ + if(PS_WMS == 2 && PS_WMT == 2) + { + uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); + } + else if(PS_WMS == 2) + { + uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); + } + else if(PS_WMT == 2) + { + uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); + } + + return uv; +} + +float4 tfx(float4 t, float4 c) +{ + if(PS_TFX == 0) + { + if(PS_TCC == 0) + { + c.rgb = c.rgb * t.rgb * 255.0f / 128; + } + else + { + c = c * t * 255.0f / 128; + } + } + else if(PS_TFX == 1) + { + if(PS_TCC == 0) + { + c.rgb = t.rgb; + } + else + { + c = t; + } + } + else if(PS_TFX == 2) + { + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; + + if(PS_TCC == 1) + { + c.a += t.a; + } + } + else if(PS_TFX == 3) + { + c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; + + if(PS_TCC == 1) + { + c.a = t.a; + } + } + + return saturate(c); +} + +void atst(float4 c) +{ + if(PS_ATE == 1) + { + if(PS_ATST == 0) + { + discard; + } + else if(PS_ATST == 2 || PS_ATST == 3) // l, le + { + clip(AREF - c.a); + } + else if(PS_ATST == 4) // e + { + clip(0.6f / 255 - abs(c.a - AREF)); // FIXME: 0.5f is too small + } + else if(PS_ATST == 5 || PS_ATST == 6) // ge, g + { + clip(c.a - AREF); + } + else if(PS_ATST == 7) // ne + { + clip(abs(c.a - AREF) - 0.4f / 255); // FIXME: 0.5f is too much + } + } +} + +float4 fog(float4 c, float f) +{ + if(PS_FOG == 1) + { + c.rgb = lerp(FogColor, c.rgb, f); + } + + return c; +} + +#if SHADER_MODEL >= 0x400 + +VS_OUTPUT vs_main(VS_INPUT input) +{ + if(VS_BPPZ == 1) // 24 + { + input.z = input.z & 0xffffff; + } + else if(VS_BPPZ == 2) // 16 + { + input.z = input.z & 0xffff; + } + + if(VS_PRIM == 6) // sprite + { + //input.p.xy = (input.p.xy + 15) & ~15; // HACK + } + + VS_OUTPUT output; + + // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) + // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty + // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel + // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 + + float4 p = float4(input.p, input.z, 0) - float4(0.05f, 0.05f, 0, 0); + + output.p = p * VertexScale - VertexOffset; + + if(VS_TME == 1) + { + if(VS_FST == 1) + { + output.t.xy = input.t * TextureScale; + output.t.w = 1.0f; + } + else + { + output.t.xy = input.t; + output.t.w = input.q; + } + } + else + { + output.t.xy = 0; + output.t.w = 1.0f; + } + + output.c = input.c; + output.t.z = input.f.a; + + return output; +} + +#if GS_PRIM == 0 + +[maxvertexcount(1)] +void gs_main(point VS_OUTPUT input[1], inout PointStream stream) +{ + stream.Append(input[0]); +} + +#elif GS_PRIM == 1 + +[maxvertexcount(2)] +void gs_main(line VS_OUTPUT input[2], inout LineStream stream) +{ + #if GS_IIP == 0 + input[0].c = input[1].c; + #endif + + stream.Append(input[0]); + stream.Append(input[1]); +} + +#elif GS_PRIM == 2 + +[maxvertexcount(3)] +void gs_main(triangle VS_OUTPUT input[3], inout TriangleStream stream) +{ + #if GS_IIP == 0 + input[0].c = input[2].c; + input[1].c = input[2].c; + #endif + + stream.Append(input[0]); + stream.Append(input[1]); + stream.Append(input[2]); +} + +#elif GS_PRIM == 3 + +[maxvertexcount(4)] +void gs_main(line VS_OUTPUT input[2], inout TriangleStream stream) +{ + input[0].p.z = input[1].p.z; + input[0].t.zw = input[1].t.zw; + + #if GS_IIP == 0 + input[0].c = input[1].c; + #endif + + VS_OUTPUT lb = input[1]; + + lb.p.x = input[0].p.x; + lb.t.x = input[0].t.x; + + VS_OUTPUT rt = input[1]; + + rt.p.y = input[0].p.y; + rt.t.y = input[0].t.y; + + stream.Append(input[0]); + stream.Append(lb); + stream.Append(rt); + stream.Append(input[1]); +} + +#endif + +float4 sample(float2 tc, float w) +{ + if(PS_FST == 0) + { + tc /= w; + } + + float4 t; +/* + if(PS_BPP < 3 && PS_WMS < 2 && PS_WMT < 2) + { + t = Texture.Sample(TextureSampler, tc); + } +*/ + if(PS_BPP < 3 && PS_WMS < 3 && PS_WMT < 3) + { + t = Texture.Sample(TextureSampler, clampuv(tc)); + } + else + { + float w, h; + Texture.GetDimensions(w, h); + + float4 uv2 = tc.xyxy + HalfTexel; + float2 dd = frac(uv2.xy * float2(w, h)); + float4 uv = wrapuv(uv2); + + float4 t00, t01, t10, t11; + + if(PS_BPP == 3) // 8HP + { + float4 a; + + a.x = Texture.Sample(TextureSampler, uv.xy).a; + a.y = Texture.Sample(TextureSampler, uv.zy).a; + a.z = Texture.Sample(TextureSampler, uv.xw).a; + a.w = Texture.Sample(TextureSampler, uv.zw).a; + + t00 = Palette.Sample(PaletteSampler, a.x); + t01 = Palette.Sample(PaletteSampler, a.y); + t10 = Palette.Sample(PaletteSampler, a.z); + t11 = Palette.Sample(PaletteSampler, a.w); + } + else + { + t00 = Texture.Sample(TextureSampler, uv.xy); + t01 = Texture.Sample(TextureSampler, uv.zy); + t10 = Texture.Sample(TextureSampler, uv.xw); + t11 = Texture.Sample(TextureSampler, uv.zw); + } + + if(PS_LTF) + { + t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); + } + else + { + t = t00; + } + } + + if(PS_BPP == 1) // 24 + { + t.a = PS_AEM == 0 || any(t.rgb) ? TA.x : 0; + } + else if(PS_BPP == 2) // 16 + { + // a bit incompatible with up-scaling because the 1 bit alpha is interpolated + + t.a = t.a >= 0.5 ? TA.y : PS_AEM == 0 || any(t.rgb) ? TA.x : 0; + } + + return t; +} + +PS_OUTPUT ps_main(PS_INPUT input) +{ + float4 t = sample(input.t.xy, input.t.w); + + float4 c = tfx(t, input.c); + + atst(c); + + c = fog(c, input.t.z); + + if(PS_CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes + { + c.rgb = 1; + } + + PS_OUTPUT output; + + output.c1 = c.a * 2; // used for alpha blending + + if(PS_AOUT == 1) // 16 bit output + { + float a = 128.0f / 255; // alpha output will be 0x80 + + c.a = PS_FBA == 1 ? a : step(0.5, c.a) * a; + } + else if(PS_FBA == 1) + { + if(c.a < 0.5) c.a += 0.5; + } + + output.c0 = c; + + return output; +} + +#elif SHADER_MODEL <= 0x300 + +VS_OUTPUT vs_main(VS_INPUT input) +{ + if(VS_BPPZ == 1) // 24 + { + input.p.z = fmod(input.p.z, 0x1000000); + } + else if(VS_BPPZ == 2) // 16 + { + input.p.z = fmod(input.p.z, 0x10000); + } + + VS_OUTPUT output; + + // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) + // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty + // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel + // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 + + float4 p = input.p - float4(0.05f, 0.05f, 0, 0); + + output.p = p * VertexScale - VertexOffset; + + if(VS_LOGZ == 1) + { + output.p.z = log2(1.0f + input.p.z) / 32; + } + + if(VS_TME == 1) + { + if(VS_FST == 1) + { + output.t.xy = input.t * TextureScale; + output.t.w = 1.0f; + } + else + { + output.t.xy = input.t; + output.t.w = input.p.w; + } + } + else + { + output.t.xy = 0; + output.t.w = 1.0f; + } + + output.c = input.c; + output.t.z = input.f.a; + + return output; +} + +float4 sample(float2 tc, float w) +{ + if(PS_FST == 0) + { + tc /= w; + } + + float4 t; +/* + if(PS_BPP < 3 && PS_WMS < 2 && PS_WMT < 2) + { + t = tex2D(Texture, tc); + } +*/ + if(PS_BPP < 3 && PS_WMS < 3 && PS_WMT < 3) + { + t = tex2D(Texture, clampuv(tc)); + } + else + { + float4 uv2 = tc.xyxy + HalfTexel; + float2 dd = frac(uv2.xy * WH); + float4 uv = wrapuv(uv2); + + float4 t00, t01, t10, t11; + + if(PS_BPP == 3) // 8HP + { + float4 a; + + a.x = tex2D(Texture, uv.xy).a; + a.y = tex2D(Texture, uv.zy).a; + a.z = tex2D(Texture, uv.xw).a; + a.w = tex2D(Texture, uv.zw).a; + + if(PS_RT == 1) a *= 0.5; + + t00 = tex1D(Palette, a.x); + t01 = tex1D(Palette, a.y); + t10 = tex1D(Palette, a.z); + t11 = tex1D(Palette, a.w); + } + else + { + t00 = tex2D(Texture, uv.xy); + t01 = tex2D(Texture, uv.zy); + t10 = tex2D(Texture, uv.xw); + t11 = tex2D(Texture, uv.zw); + } + + if(PS_LTF) + { + t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); + } + else + { + t = t00; + } + } + + if(PS_BPP == 0) // 32 + { + if(PS_RT == 1) t.a *= 0.5; + } + else if(PS_BPP == 1) // 24 + { + t.a = PS_AEM == 0 || any(t.rgb) ? TA0 : 0; + } + else if(PS_BPP == 2) // 16 + { + // a bit incompatible with up-scaling because the 1 bit alpha is interpolated + + t.a = t.a >= 0.5 ? TA1 : PS_AEM == 0 || any(t.rgb) ? TA0 : 0; + } + + return t; +} + +float4 ps_main(PS_INPUT input) : COLOR +{ + float4 t = sample(input.t.xy, input.t.w); + + float4 c = tfx(t, input.c); + + atst(c); + + c = fog(c, input.t.z); + + if(PS_CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes + { + c.rgb = 1; + } + + c.a *= 2; + + return c; +} + +#endif \ No newline at end of file diff --git a/plugins/GSdx/res/tfx10.fx b/plugins/GSdx/res/tfx10.fx deleted file mode 100644 index c0a3a87906..0000000000 --- a/plugins/GSdx/res/tfx10.fx +++ /dev/null @@ -1,472 +0,0 @@ -cbuffer cb0 -{ - float4 VertexScale; - float4 VertexOffset; - float2 TextureScale; -}; - -struct VS_INPUT -{ - uint2 p : POSITION0; - uint z : POSITION1; - float2 t : TEXCOORD0; - float q : TEXCOORD1; - float4 c : COLOR0; - float4 f : COLOR1; -}; - -struct VS_OUTPUT -{ - float4 p : SV_Position; - float4 t : TEXCOORD0; - float4 c : COLOR0; -}; - -#ifndef VS_BPPZ -#define VS_BPPZ 0 -#define VS_TME 1 -#define VS_FST 1 -#define VS_PRIM 0 -#endif - -VS_OUTPUT vs_main(VS_INPUT input) -{ - if(VS_BPPZ == 1) // 24 - { - input.z = input.z & 0xffffff; - } - else if(VS_BPPZ == 2) // 16 - { - input.z = input.z & 0xffff; - } - - if(VS_PRIM == 6) // sprite - { - //input.p.xy = (input.p.xy + 15) & ~15; // HACK - } - - VS_OUTPUT output; - - // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) - // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty - // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel - // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 - - float4 p = float4(input.p, input.z, 0) - float4(0.05f, 0.05f, 0, 0); - - output.p = p * VertexScale - VertexOffset; - - if(VS_TME == 1) - { - if(VS_FST == 1) - { - output.t.xy = input.t * TextureScale; - output.t.w = 1.0f; - } - else - { - output.t.xy = input.t; - output.t.w = input.q; - } - } - else - { - output.t.xy = 0; - output.t.w = 1.0f; - } - - output.c = input.c; - output.t.z = input.f.a; - - return output; -} - -#ifndef IIP -#define IIP 0 -#define PRIM 3 -#endif - -#if PRIM == 0 - -[maxvertexcount(1)] -void gs_main(point VS_OUTPUT input[1], inout PointStream stream) -{ - stream.Append(input[0]); -} - -#elif PRIM == 1 - -[maxvertexcount(2)] -void gs_main(line VS_OUTPUT input[2], inout LineStream stream) -{ - #if IIP == 0 - input[0].c = input[1].c; - #endif - - stream.Append(input[0]); - stream.Append(input[1]); -} - -#elif PRIM == 2 - -[maxvertexcount(3)] -void gs_main(triangle VS_OUTPUT input[3], inout TriangleStream stream) -{ - #if IIP == 0 - input[0].c = input[2].c; - input[1].c = input[2].c; - #endif - - stream.Append(input[0]); - stream.Append(input[1]); - stream.Append(input[2]); -} - -#elif PRIM == 3 - -[maxvertexcount(4)] -void gs_main(line VS_OUTPUT input[2], inout TriangleStream stream) -{ - input[0].p.z = input[1].p.z; - input[0].t.zw = input[1].t.zw; - - #if IIP == 0 - input[0].c = input[1].c; - #endif - - VS_OUTPUT lb = input[1]; - - lb.p.x = input[0].p.x; - lb.t.x = input[0].t.x; - - VS_OUTPUT rt = input[1]; - - rt.p.y = input[0].p.y; - rt.t.y = input[0].t.y; - - stream.Append(input[0]); - stream.Append(lb); - stream.Append(rt); - stream.Append(input[1]); -} - -#endif - -Texture2D Texture; -Texture2D Palette; -SamplerState TextureSampler; -SamplerState PaletteSampler; - -cbuffer cb1 -{ - float3 FogColor; - float AREF; - float4 HalfTexel; - float2 WH; - float2 TA; - float4 MinMax; - float4 MinMaxF; - uint4 MskFix; -}; - -struct PS_INPUT -{ - float4 p : SV_Position; - float4 t : TEXCOORD0; - float4 c : COLOR0; -}; - -struct PS_OUTPUT -{ - float4 c0 : SV_Target0; - float4 c1 : SV_Target1; -}; - -#ifndef FST -#define FST 0 -#define WMS 3 -#define WMT 3 -#define BPP 0 -#define AEM 0 -#define TFX 0 -#define TCC 1 -#define ATE 0 -#define ATST 4 -#define FOG 0 -#define CLR1 0 -#define FBA 0 -#define AOUT 0 -#define LTF 1 -#endif - -float4 wrapuv(float4 uv) -{ - if(WMS == WMT) - { - if(WMS == 0) - { - uv = frac(uv); - } - else if(WMS == 1) - { - uv = saturate(uv); - } - else if(WMS == 2) - { - uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); - } - else if(WMS == 3) - { - uv = (float4)(((int4)(uv * WH.xyxy) & MskFix.xyxy) | MskFix.zwzw) / WH.xyxy; - } - } - else - { - if(WMS == 0) - { - uv.xz = frac(uv.xz); - } - else if(WMS == 1) - { - uv.xz = saturate(uv.xz); - } - else if(WMS == 2) - { - uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); - } - else if(WMS == 3) - { - uv.xz = (float2)(((int2)(uv * WH.xyxy).xz & MskFix.xx) | MskFix.zz) / WH; - } - - if(WMT == 0) - { - uv.yw = frac(uv.yw); - } - else if(WMT == 1) - { - uv.yw = saturate(uv.yw); - } - else if(WMT == 2) - { - uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); - } - else if(WMT == 3) - { - uv.yw = (float2)(((int2)(uv * WH.xyxy).yw & MskFix.yy) | MskFix.ww) / WH; - } - } - - return uv; -} - -float2 clampuv(float2 uv) -{ - if(WMS == 2 && WMT == 2) - { - uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); - } - else if(WMS == 2) - { - uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); - } - else if(WMT == 2) - { - uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); - } - - return uv; -} - -float4 sample(float2 tc, float w) -{ - if(FST == 0) - { - tc /= w; - } - - float4 t; -/* - if(BPP < 3 && WMS < 2 && WMT < 2) - { - t = Texture.Sample(TextureSampler, tc); - } -*/ - if(BPP < 3 && WMS < 3 && WMT < 3) - { - t = Texture.Sample(TextureSampler, clampuv(tc)); - } - else - { - float w, h; - Texture.GetDimensions(w, h); - - float4 uv2 = tc.xyxy + HalfTexel; - float2 dd = frac(uv2.xy * float2(w, h)); - float4 uv = wrapuv(uv2); - - float4 t00, t01, t10, t11; - - if(BPP == 3) // 8HP - { - float4 a; - - a.x = Texture.Sample(TextureSampler, uv.xy).a; - a.y = Texture.Sample(TextureSampler, uv.zy).a; - a.z = Texture.Sample(TextureSampler, uv.xw).a; - a.w = Texture.Sample(TextureSampler, uv.zw).a; - - t00 = Palette.Sample(PaletteSampler, a.x); - t01 = Palette.Sample(PaletteSampler, a.y); - t10 = Palette.Sample(PaletteSampler, a.z); - t11 = Palette.Sample(PaletteSampler, a.w); - } - else - { - t00 = Texture.Sample(TextureSampler, uv.xy); - t01 = Texture.Sample(TextureSampler, uv.zy); - t10 = Texture.Sample(TextureSampler, uv.xw); - t11 = Texture.Sample(TextureSampler, uv.zw); - } - - if(LTF) - { - t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); - } - else - { - t = t00; - } - } - - if(BPP == 1) // 24 - { - t.a = AEM == 0 || any(t.rgb) ? TA.x : 0; - } - else if(BPP == 2) // 16 - { - // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - - t.a = t.a >= 0.5 ? TA.y : AEM == 0 || any(t.rgb) ? TA.x : 0; - } - - return t; -} - -float4 tfx(float4 t, float4 c) -{ - if(TFX == 0) - { - if(TCC == 0) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128; - } - else - { - c = c * t * 255.0f / 128; - } - } - else if(TFX == 1) - { - if(TCC == 0) - { - c.rgb = t.rgb; - } - else - { - c = t; - } - } - else if(TFX == 2) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - - if(TCC == 1) - { - c.a += t.a; - } - } - else if(TFX == 3) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - - if(TCC == 1) - { - c.a = t.a; - } - } - - return saturate(c); -} - -void atst(float4 c) -{ - if(ATE == 1) - { - if(ATST == 0) - { - discard; - } - else if(ATST == 2 || ATST == 3) // l, le - { - clip(AREF - c.a); - } - else if(ATST == 4) // e - { - clip(0.6f / 255 - abs(c.a - AREF)); // FIXME: 0.5f is too small - } - else if(ATST == 5 || ATST == 6) // ge, g - { - clip(c.a - AREF); - } - else if(ATST == 7) // ne - { - clip(abs(c.a - AREF) - 0.4f / 255); // FIXME: 0.5f is too much - } - } -} - -float4 fog(float4 c, float f) -{ - if(FOG == 1) - { - c.rgb = lerp(FogColor.rgb, c.rgb, f); - } - - return c; -} - -PS_OUTPUT ps_main(PS_INPUT input) -{ - float4 t = sample(input.t.xy, input.t.w); - - float4 c = tfx(t, input.c); - - atst(c); - - c = fog(c, input.t.z); - - if(CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes - { - c.rgb = 1; - } - - PS_OUTPUT output; - - output.c1 = c.a * 2; // used for alpha blending - - if(AOUT == 1) // 16 bit output - { - float a = 128.0f / 255; // alpha output will be 0x80 - - c.a = FBA == 1 ? a : step(0.5, c.a) * a; - } - else if(FBA == 1) - { - if(c.a < 0.5) c.a += 0.5; - } - - output.c0 = c; - - return output; -} diff --git a/plugins/GSdx/res/tfx9.fx b/plugins/GSdx/res/tfx9.fx deleted file mode 100644 index 0c1062fec8..0000000000 --- a/plugins/GSdx/res/tfx9.fx +++ /dev/null @@ -1,382 +0,0 @@ -struct VS_INPUT -{ - float4 p : POSITION0; - float2 t : TEXCOORD0; - float4 c : COLOR0; - float4 f : COLOR1; -}; - -struct VS_OUTPUT -{ - float4 p : POSITION; - float4 t : TEXCOORD0; - float4 c : COLOR0; -}; - -float4 vs_params[3]; - -#define VertexScale vs_params[0] -#define VertexOffset vs_params[1] -#define TextureScale vs_params[2].xy - -#ifndef VS_BPPZ -#define VS_BPPZ 0 -#define VS_TME 1 -#define VS_FST 1 -#define VS_LOGZ 1 -#endif - -VS_OUTPUT vs_main(VS_INPUT input) -{ - if(VS_BPPZ == 1) // 24 - { - input.p.z = fmod(input.p.z, 0x1000000); - } - else if(VS_BPPZ == 2) // 16 - { - input.p.z = fmod(input.p.z, 0x10000); - } - - VS_OUTPUT output; - - // pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go) - // example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty - // input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel - // example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133 - - float4 p = input.p - float4(0.05f, 0.05f, 0, 0); - - output.p = p * VertexScale - VertexOffset; - - if(VS_LOGZ == 1) - { - output.p.z = log2(1.0f + input.p.z) / 32; - } - - if(VS_TME == 1) - { - if(VS_FST == 1) - { - output.t.xy = input.t * TextureScale; - output.t.w = 1.0f; - } - else - { - output.t.xy = input.t; - output.t.w = input.p.w; - } - } - else - { - output.t.xy = 0; - output.t.w = 1.0f; - } - - output.c = input.c; - output.t.z = input.f.a; - - return output; -} - -float4 ps_params[5]; - -#define FogColor ps_params[0].bgr -#define AREF ps_params[0].a -#define HalfTexel ps_params[1] -#define WH ps_params[2].xy -#define TA0 ps_params[2].z -#define TA1 ps_params[2].w -#define MinMax ps_params[3] -#define MinMaxF ps_params[4] - -struct PS_INPUT -{ - float4 t : TEXCOORD0; - float4 c : COLOR0; -}; - -#ifndef FST -#define FST 0 -#define WMS 3 -#define WMT 3 -#define BPP 0 -#define AEM 0 -#define TFX 0 -#define TCC 1 -#define ATE 0 -#define ATST 4 -#define FOG 0 -#define CLR1 0 -#define RT 0 -#define LTF 0 -#endif - -sampler Texture : register(s0); -sampler1D Palette : register(s1); -sampler1D UMSKFIX : register(s2); -sampler1D VMSKFIX : register(s3); - -float4 wrapuv(float4 uv) -{ - if(WMS == WMT) - { - if(WMS == 0) - { - uv = frac(uv); - } - else if(WMS == 1) - { - uv = saturate(uv); - } - else if(WMS == 2) - { - uv = clamp(uv, MinMax.xyxy, MinMax.zwzw); - } - else if(WMS == 3) - { - uv.x = tex1D(UMSKFIX, uv.x); - uv.y = tex1D(VMSKFIX, uv.y); - uv.z = tex1D(UMSKFIX, uv.z); - uv.w = tex1D(VMSKFIX, uv.w); - } - } - else - { - if(WMS == 0) - { - uv.xz = frac(uv.xz); - } - else if(WMS == 1) - { - uv.xz = saturate(uv.xz); - } - else if(WMS == 2) - { - uv.xz = clamp(uv.xz, MinMax.xx, MinMax.zz); - } - else if(WMS == 3) - { - uv.x = tex1D(UMSKFIX, uv.x); - uv.z = tex1D(UMSKFIX, uv.z); - } - - if(WMT == 0) - { - uv.yw = frac(uv.yw); - } - else if(WMT == 1) - { - uv.yw = saturate(uv.yw); - } - else if(WMT == 2) - { - uv.yw = clamp(uv.yw, MinMax.yy, MinMax.ww); - } - else if(WMT == 3) - { - uv.y = tex1D(VMSKFIX, uv.y); - uv.w = tex1D(VMSKFIX, uv.w); - } - } - - return uv; -} - -float2 clampuv(float2 uv) -{ - if(WMS == 2 && WMT == 2) - { - uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); - } - else if(WMS == 2) - { - uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); - } - else if(WMT == 2) - { - uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); - } - - return uv; -} - -float4 sample(float2 tc, float w) -{ - if(FST == 0) - { - tc /= w; - } - - float4 t; -/* - if(BPP < 3 && WMS < 2 && WMT < 2) - { - t = tex2D(Texture, tc); - } -*/ - if(BPP < 3 && WMS < 3 && WMT < 3) - { - t = tex2D(Texture, clampuv(tc)); - } - else - { - float4 uv2 = tc.xyxy + HalfTexel; - float2 dd = frac(uv2.xy * WH); - float4 uv = wrapuv(uv2); - - float4 t00, t01, t10, t11; - - if(BPP == 3) // 8HP - { - float4 a; - - a.x = tex2D(Texture, uv.xy).a; - a.y = tex2D(Texture, uv.zy).a; - a.z = tex2D(Texture, uv.xw).a; - a.w = tex2D(Texture, uv.zw).a; - - if(RT == 1) a *= 0.5; - - t00 = tex1D(Palette, a.x); - t01 = tex1D(Palette, a.y); - t10 = tex1D(Palette, a.z); - t11 = tex1D(Palette, a.w); - } - else - { - t00 = tex2D(Texture, uv.xy); - t01 = tex2D(Texture, uv.zy); - t10 = tex2D(Texture, uv.xw); - t11 = tex2D(Texture, uv.zw); - } - - if(LTF) - { - t = lerp(lerp(t00, t01, dd.x), lerp(t10, t11, dd.x), dd.y); - } - else - { - t = t00; - } - } - - if(BPP == 0) // 32 - { - if(RT == 1) t.a *= 0.5; - } - else if(BPP == 1) // 24 - { - t.a = AEM == 0 || any(t.rgb) ? TA0 : 0; - } - else if(BPP == 2) // 16 - { - // a bit incompatible with up-scaling because the 1 bit alpha is interpolated - - t.a = t.a >= 0.5 ? TA1 : AEM == 0 || any(t.rgb) ? TA0 : 0; - } - - return t; -} - -float4 tfx(float4 t, float4 c) -{ - if(TFX == 0) - { - if(TCC == 0) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128; - } - else - { - c = c * t * 255.0f / 128; - } - } - else if(TFX == 1) - { - if(TCC == 0) - { - c.rgb = t.rgb; - } - else - { - c = t; - } - } - else if(TFX == 2) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - - if(TCC == 1) - { - c.a += t.a; - } - } - else if(TFX == 3) - { - c.rgb = c.rgb * t.rgb * 255.0f / 128 + c.a; - - if(TCC == 1) - { - c.a = t.a; - } - } - - return saturate(c); -} - -void atst(float4 c) -{ - if(ATE == 1) - { - if(ATST == 0) - { - discard; - } - else if(ATST == 2 || ATST == 3) // l, le - { - clip(AREF - c.a); - } - else if(ATST == 4) // e - { - clip(0.6f / 255 - abs(c.a - AREF)); // FIXME: 0.5f is too small - } - else if(ATST == 5 || ATST == 6) // ge, g - { - clip(c.a - AREF); - } - else if(ATST == 7) // ne - { - clip(abs(c.a - AREF) - 0.4f / 255); // FIXME: 0.5f is too much - } - } -} - -float4 fog(float4 c, float f) -{ - if(FOG == 1) - { - c.rgb = lerp(FogColor, c.rgb, f); - } - - return c; -} - -float4 ps_main(PS_INPUT input) : COLOR -{ - float4 t = sample(input.t.xy, input.t.w); - - float4 c = tfx(t, input.c); - - atst(c); - - c = fog(c, input.t.z); - - if(CLR1 == 1) // needed for Cd * (As/Ad/F + 1) blending modes - { - c.rgb = 1; - } - - c.a *= 2; - - return c; -} diff --git a/plugins/GSdx/resource.h b/plugins/GSdx/resource.h index 2083fa82ee..409840b67e 100644 --- a/plugins/GSdx/resource.h +++ b/plugins/GSdx/resource.h @@ -40,20 +40,16 @@ #define IDC_WIDTH 2036 #define IDC_HEIGHT 2037 #define IDC_CONFIGURE 2038 -#define IDR_CONVERT9_FX 10000 -#define IDR_TFX9_FX 10001 -#define IDR_MERGE9_FX 10002 -#define IDR_INTERLACE9_FX 10003 -#define IDR_CONVERT10_FX 10004 -#define IDR_TFX10_FX 10005 -#define IDR_MERGE10_FX 10006 -#define IDR_INTERLACE10_FX 10007 +#define IDR_CONVERT_FX 10000 +#define IDR_TFX_FX 10001 +#define IDR_MERGE_FX 10002 +#define IDR_INTERLACE_FX 10003 // Next default values for new objects // #ifdef APSTUDIO_INVOKED #ifndef APSTUDIO_READONLY_SYMBOLS -#define _APS_NEXT_RESOURCE_VALUE 10009 +#define _APS_NEXT_RESOURCE_VALUE 10004 #define _APS_NEXT_COMMAND_VALUE 32771 #define _APS_NEXT_CONTROL_VALUE 2039 #define _APS_NEXT_SYMED_VALUE 5000 diff --git a/plugins/GSdx/stdafx.h b/plugins/GSdx/stdafx.h index ddc67531b0..375fb8d206 100644 --- a/plugins/GSdx/stdafx.h +++ b/plugins/GSdx/stdafx.h @@ -90,6 +90,7 @@ typedef signed long long int64; #endif #if defined(_DEBUG) && defined(_MSC_VER) + #include #define ASSERT assert #else #define ASSERT(exp) ((void)0) @@ -104,13 +105,19 @@ typedef signed long long int64; // directx #include -#include -#include +#include +#include #include #include +#include +#include #define D3DCOLORWRITEENABLE_RGBA (D3DCOLORWRITEENABLE_RED | D3DCOLORWRITEENABLE_GREEN | D3DCOLORWRITEENABLE_BLUE | D3DCOLORWRITEENABLE_ALPHA) +// dxsdk beta missing these: +#define D3D11_SHADER_MACRO D3D10_SHADER_MACRO +#define ID3D11Blob ID3D10Blob + #endif // opengl diff --git a/plugins/GSdx/vsprops/common.props b/plugins/GSdx/vsprops/common.props index 357fe0d1b6..0edc67dee6 100644 --- a/plugins/GSdx/vsprops/common.props +++ b/plugins/GSdx/vsprops/common.props @@ -15,9 +15,9 @@ 4995;4324;%(DisableSpecificWarnings) - JITProfiling.lib;d3d10.lib;d3dx10.lib;d3d9.lib;d3dx9.lib;ddraw.lib;dxguid.lib;winmm.lib;strmiids.lib;xinput.lib;cg.lib;cgGL.lib;glut32.lib;glew32.lib;%(AdditionalDependencies) + JITProfiling.lib;d3d11_beta.lib;d3dx11.lib;d3d10.lib;d3dx10.lib;d3d9.lib;d3dx9.lib;ddraw.lib;dxguid.lib;winmm.lib;strmiids.lib;xinput.lib;cg.lib;cgGL.lib;glut32.lib;glew32.lib;%(AdditionalDependencies) ./vtune;%(AdditionalLibraryDirectories) - d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll;%(DelayLoadDLLs) + d3d9.dll;d3dx9_41.dll;d3d10.dll;d3dx10_41.dll;d3d11.dll;d3d11_beta.dll;d3dx11_41.dll;%(DelayLoadDLLs) true Windows false From 191a3d0c7e26fe808251f67e372a177e680f686a Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 19 Jun 2009 20:31:36 +0000 Subject: [PATCH 259/296] GSdx: another useless revision... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1405 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSDevice11.cpp | 870 ++++++++++++++++++++++++++++++ plugins/GSdx/GSDevice11.h | 147 +++++ plugins/GSdx/GSRendererHW11.cpp | 593 ++++++++++++++++++++ plugins/GSdx/GSRendererHW11.h | 52 ++ plugins/GSdx/GSTexture11.cpp | 200 +++++++ plugins/GSdx/GSTexture11.h | 53 ++ plugins/GSdx/GSTextureCache11.cpp | 309 +++++++++++ plugins/GSdx/GSTextureCache11.h | 60 +++ plugins/GSdx/GSTextureFX11.cpp | 526 ++++++++++++++++++ plugins/GSdx/GSTextureFX11.h | 240 +++++++++ 10 files changed, 3050 insertions(+) create mode 100644 plugins/GSdx/GSDevice11.cpp create mode 100644 plugins/GSdx/GSDevice11.h create mode 100644 plugins/GSdx/GSRendererHW11.cpp create mode 100644 plugins/GSdx/GSRendererHW11.h create mode 100644 plugins/GSdx/GSTexture11.cpp create mode 100644 plugins/GSdx/GSTexture11.h create mode 100644 plugins/GSdx/GSTextureCache11.cpp create mode 100644 plugins/GSdx/GSTextureCache11.h create mode 100644 plugins/GSdx/GSTextureFX11.cpp create mode 100644 plugins/GSdx/GSTextureFX11.h diff --git a/plugins/GSdx/GSDevice11.cpp b/plugins/GSdx/GSDevice11.cpp new file mode 100644 index 0000000000..c31c0a8d92 --- /dev/null +++ b/plugins/GSdx/GSDevice11.cpp @@ -0,0 +1,870 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSdx.h" +#include "GSDevice11.h" +#include "resource.h" + +GSDevice11::GSDevice11() + : m_vb(NULL) + , m_vb_stride(0) + , m_layout(NULL) + , m_topology(D3D11_PRIMITIVE_TOPOLOGY_UNDEFINED) + , m_vs(NULL) + , m_vs_cb(NULL) + , m_gs(NULL) + , m_ps(NULL) + , m_ps_cb(NULL) + , m_scissor(0, 0, 0, 0) + , m_viewport(0, 0) + , m_dss(NULL) + , m_sref(0) + , m_bs(NULL) + , m_bf(-1) + , m_rtv(NULL) + , m_dsv(NULL) +{ + memset(m_ps_srv, 0, sizeof(m_ps_srv)); + memset(m_ps_ss, 0, sizeof(m_ps_ss)); + + m_vertices.stride = 0; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = 0; +} + +GSDevice11::~GSDevice11() +{ +} + +bool GSDevice11::Create(GSWnd* wnd, bool vsync) +{ + if(!__super::Create(wnd, vsync)) + { + return false; + } + + HRESULT hr; + + DXGI_SWAP_CHAIN_DESC scd; + D3D11_BUFFER_DESC bd; + D3D11_SAMPLER_DESC sd; + D3D11_DEPTH_STENCIL_DESC dsd; + D3D11_RASTERIZER_DESC rd; + D3D11_BLEND_DESC bsd; + + memset(&scd, 0, sizeof(scd)); + + scd.BufferCount = 2; + scd.BufferDesc.Width = 1; + scd.BufferDesc.Height = 1; + scd.BufferDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM; + //scd.BufferDesc.RefreshRate.Numerator = 60; + //scd.BufferDesc.RefreshRate.Denominator = 1; + scd.BufferUsage = DXGI_USAGE_RENDER_TARGET_OUTPUT; + scd.OutputWindow = (HWND)m_wnd->GetHandle(); + scd.SampleDesc.Count = 1; + scd.SampleDesc.Quality = 0; + scd.Windowed = TRUE; + + uint32 flags = 0; + +#ifdef DEBUG + flags |= D3D11_CREATE_DEVICE_DEBUG; +#endif + + hr = D3D11CreateDeviceAndSwapChain(NULL, D3D_DRIVER_TYPE_HARDWARE, NULL, 0, NULL, 0, D3D11_SDK_VERSION, &scd, &m_swapchain, &m_dev, &m_level, &m_ctx); + // hr = D3D11CreateDeviceAndSwapChain(NULL, D3D_DRIVER_TYPE_REFERENCE, NULL, 0, NULL, 0, D3D11_SDK_VERSION, &scd, &m_swapchain, &m_dev, &m_level, &m_ctx); + + if(FAILED(hr)) return false; + + D3D11_FEATURE_DATA_D3D10_X_HARDWARE_OPTIONS options; + + hr = m_dev->CheckFeatureSupport(D3D11_FEATURE_D3D10_X_HARDWARE_OPTIONS, &options, sizeof(D3D11_FEATURE_D3D10_X_HARDWARE_OPTIONS)); + + switch(m_level) + { + case D3D_FEATURE_LEVEL_9_1: + case D3D_FEATURE_LEVEL_9_2: + m_shader.model = "0x200"; + m_shader.vs = "vs_4_0_level_9_1"; + m_shader.ps = "ps_4_0_level_9_1"; + break; + case D3D_FEATURE_LEVEL_9_3: + m_shader.model = "0x300"; + m_shader.vs = "vs_4_0_level_9_3"; + m_shader.ps = "ps_4_0_level_9_3"; + break; + case D3D_FEATURE_LEVEL_10_0: + case D3D_FEATURE_LEVEL_10_1: + m_shader.model = "0x400"; + m_shader.vs = "vs_4_0"; + m_shader.gs = "gs_4_0"; + m_shader.ps = "ps_4_0"; + break; + case D3D_FEATURE_LEVEL_11_0: + m_shader.model = "0x500"; + m_shader.vs = "vs_5_0"; + m_shader.gs = "gs_5_0"; + m_shader.ps = "ps_5_0"; + break; + } + + if(m_level < D3D_FEATURE_LEVEL_10_0) + { + return false; + } + + // convert + + D3D11_INPUT_ELEMENT_DESC il_convert[] = + { + {"POSITION", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 0, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 16, D3D11_INPUT_PER_VERTEX_DATA, 0}, + }; + + hr = CompileShader(IDR_CONVERT_FX, "vs_main", NULL, &m_convert.vs, il_convert, countof(il_convert), &m_convert.il); + + for(int i = 0; i < countof(m_convert.ps); i++) + { + hr = CompileShader(IDR_CONVERT_FX, format("ps_main%d", i), NULL, &m_convert.ps[i]); + } + + memset(&dsd, 0, sizeof(dsd)); + + dsd.DepthEnable = false; + dsd.StencilEnable = false; + + hr = m_dev->CreateDepthStencilState(&dsd, &m_convert.dss); + + memset(&bsd, 0, sizeof(bsd)); + + bsd.RenderTarget[0].RenderTargetWriteMask = D3D11_COLOR_WRITE_ENABLE_ALL; + + hr = m_dev->CreateBlendState(&bsd, &m_convert.bs); + + // merge + + memset(&bd, 0, sizeof(bd)); + + bd.ByteWidth = sizeof(MergeConstantBuffer); + bd.Usage = D3D11_USAGE_DEFAULT; + bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; + + hr = m_dev->CreateBuffer(&bd, NULL, &m_merge.cb); + + for(int i = 0; i < countof(m_merge.ps); i++) + { + hr = CompileShader(IDR_MERGE_FX, format("ps_main%d", i), NULL, &m_merge.ps[i]); + } + + memset(&bsd, 0, sizeof(bsd)); + + bsd.RenderTarget[0].BlendEnable = true; + bsd.RenderTarget[0].BlendOp = D3D11_BLEND_OP_ADD; + bsd.RenderTarget[0].SrcBlend = D3D11_BLEND_SRC_ALPHA; + bsd.RenderTarget[0].DestBlend = D3D11_BLEND_INV_SRC_ALPHA; + bsd.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD; + bsd.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_ONE; + bsd.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ZERO; + bsd.RenderTarget[0].RenderTargetWriteMask = D3D11_COLOR_WRITE_ENABLE_ALL; + + hr = m_dev->CreateBlendState(&bsd, &m_merge.bs); + + // interlace + + memset(&bd, 0, sizeof(bd)); + + bd.ByteWidth = sizeof(InterlaceConstantBuffer); + bd.Usage = D3D11_USAGE_DEFAULT; + bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; + + hr = m_dev->CreateBuffer(&bd, NULL, &m_interlace.cb); + + for(int i = 0; i < countof(m_interlace.ps); i++) + { + hr = CompileShader(IDR_INTERLACE_FX, format("ps_main%d", i), NULL, &m_interlace.ps[i]); + } + + // + + memset(&rd, 0, sizeof(rd)); + + rd.FillMode = D3D11_FILL_SOLID; + rd.CullMode = D3D11_CULL_NONE; + rd.FrontCounterClockwise = false; + rd.DepthBias = false; + rd.DepthBiasClamp = 0; + rd.SlopeScaledDepthBias = 0; + rd.DepthClipEnable = false; // ??? + rd.ScissorEnable = true; + rd.MultisampleEnable = false; + rd.AntialiasedLineEnable = false; + + hr = m_dev->CreateRasterizerState(&rd, &m_rs); + + m_ctx->RSSetState(m_rs); + + // + + memset(&sd, 0, sizeof(sd)); + + sd.Filter = D3D11_FILTER_MIN_MAG_MIP_LINEAR; + sd.AddressU = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressV = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressW = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.MaxLOD = FLT_MAX; + sd.MaxAnisotropy = 16; + sd.ComparisonFunc = D3D11_COMPARISON_NEVER; + + hr = m_dev->CreateSamplerState(&sd, &m_convert.ln); + + sd.Filter = D3D11_FILTER_MIN_MAG_MIP_POINT; + + hr = m_dev->CreateSamplerState(&sd, &m_convert.pt); + + // + + Reset(1, 1, false); + + // + + return true; +} + +bool GSDevice11::Reset(int w, int h, bool fs) +{ + if(!__super::Reset(w, h, fs)) + return false; + + DXGI_SWAP_CHAIN_DESC scd; + memset(&scd, 0, sizeof(scd)); + m_swapchain->GetDesc(&scd); + m_swapchain->ResizeBuffers(scd.BufferCount, w, h, scd.BufferDesc.Format, 0); + + CComPtr backbuffer; + m_swapchain->GetBuffer(0, __uuidof(ID3D11Texture2D), (void**)&backbuffer); + m_backbuffer = new GSTexture11(backbuffer); + + return true; +} + +void GSDevice11::Flip() +{ + m_swapchain->Present(m_vsync ? 1 : 0, 0); +} + +void GSDevice11::BeginScene() +{ +} + +void GSDevice11::DrawPrimitive() +{ + m_ctx->Draw(m_vertices.count, m_vertices.start); +} + +void GSDevice11::EndScene() +{ + PSSetShaderResources(NULL, NULL); + + // not clearing the rt/ds gives a little fps boost in complex games (5-10%) + + // OMSetRenderTargets(NULL, NULL); + + m_vertices.start += m_vertices.count; + m_vertices.count = 0; +} + +void GSDevice11::ClearRenderTarget(GSTexture* t, const GSVector4& c) +{ + m_ctx->ClearRenderTargetView(*(GSTexture11*)t, c.v); +} + +void GSDevice11::ClearRenderTarget(GSTexture* t, uint32 c) +{ + GSVector4 color = GSVector4(c) * (1.0f / 255); + + m_ctx->ClearRenderTargetView(*(GSTexture11*)t, color.v); +} + +void GSDevice11::ClearDepth(GSTexture* t, float c) +{ + m_ctx->ClearDepthStencilView(*(GSTexture11*)t, D3D11_CLEAR_DEPTH, c, 0); +} + +void GSDevice11::ClearStencil(GSTexture* t, uint8 c) +{ + m_ctx->ClearDepthStencilView(*(GSTexture11*)t, D3D11_CLEAR_STENCIL, 0, c); +} + +GSTexture* GSDevice11::Create(int type, int w, int h, int format) +{ + HRESULT hr; + + D3D11_TEXTURE2D_DESC desc; + + memset(&desc, 0, sizeof(desc)); + + desc.Width = w; + desc.Height = h; + desc.Format = (DXGI_FORMAT)format; + desc.MipLevels = 1; + desc.ArraySize = 1; + desc.SampleDesc.Count = 1; + desc.SampleDesc.Quality = 0; + desc.Usage = D3D11_USAGE_DEFAULT; + + switch(type) + { + case GSTexture::RenderTarget: + desc.BindFlags = D3D11_BIND_RENDER_TARGET | D3D11_BIND_SHADER_RESOURCE; + break; + case GSTexture::DepthStencil: + desc.BindFlags = D3D11_BIND_DEPTH_STENCIL; + break; + case GSTexture::Texture: + desc.BindFlags = D3D11_BIND_SHADER_RESOURCE; + break; + case GSTexture::Offscreen: + desc.Usage = D3D11_USAGE_STAGING; + desc.CPUAccessFlags |= D3D11_CPU_ACCESS_READ | D3D11_CPU_ACCESS_WRITE; + break; + } + + GSTexture11* t = NULL; + + CComPtr texture; + + hr = m_dev->CreateTexture2D(&desc, NULL, &texture); + + if(SUCCEEDED(hr)) + { + t = new GSTexture11(texture); + + switch(type) + { + case GSTexture::RenderTarget: + ClearRenderTarget(t, 0); + break; + case GSTexture::DepthStencil: + ClearDepth(t, 0); + break; + } + } + + return t; +} + +GSTexture* GSDevice11::CreateRenderTarget(int w, int h, int format) +{ + return __super::CreateRenderTarget(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); +} + +GSTexture* GSDevice11::CreateDepthStencil(int w, int h, int format) +{ + return __super::CreateDepthStencil(w, h, format ? format : DXGI_FORMAT_D32_FLOAT_S8X24_UINT); +} + +GSTexture* GSDevice11::CreateTexture(int w, int h, int format) +{ + return __super::CreateTexture(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); +} + +GSTexture* GSDevice11::CreateOffscreen(int w, int h, int format) +{ + return __super::CreateOffscreen(w, h, format ? format : DXGI_FORMAT_R8G8B8A8_UNORM); +} + +GSTexture* GSDevice11::CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format) +{ + GSTexture* dst = NULL; + + if(format == 0) + { + format = DXGI_FORMAT_R8G8B8A8_UNORM; + } + + if(format != DXGI_FORMAT_R8G8B8A8_UNORM && format != DXGI_FORMAT_R16_UINT) + { + ASSERT(0); + + return false; + } + + if(GSTexture* rt = CreateRenderTarget(w, h, format)) + { + GSVector4 dr(0, 0, w, h); + + StretchRect(src, sr, rt, dr, m_convert.ps[format == DXGI_FORMAT_R16_UINT ? 1 : 0], NULL); + + dst = CreateOffscreen(w, h, format); + + if(dst) + { + m_ctx->CopyResource(*(GSTexture11*)dst, *(GSTexture11*)rt); + } + + Recycle(rt); + } + + return dst; +} + +void GSDevice11::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader, bool linear) +{ + StretchRect(st, sr, dt, dr, m_convert.ps[shader], NULL, linear); +} + +void GSDevice11::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D11PixelShader* ps, ID3D11Buffer* ps_cb, bool linear) +{ + StretchRect(st, sr, dt, dr, ps, ps_cb, m_convert.bs, linear); +} + +void GSDevice11::StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D11PixelShader* ps, ID3D11Buffer* ps_cb, ID3D11BlendState* bs, bool linear) +{ + BeginScene(); + + GSVector2i ds = dt->GetSize(); + + // om + + OMSetDepthStencilState(m_convert.dss, 0); + OMSetBlendState(bs, 0); + OMSetRenderTargets(dt, NULL); + + // ia + + float left = dr.x * 2 / ds.x - 1.0f; + float top = 1.0f - dr.y * 2 / ds.y; + float right = dr.z * 2 / ds.x - 1.0f; + float bottom = 1.0f - dr.w * 2 / ds.y; + + GSVertexPT1 vertices[] = + { + {GSVector4(left, top, 0.5f, 1.0f), GSVector2(sr.x, sr.y)}, + {GSVector4(right, top, 0.5f, 1.0f), GSVector2(sr.z, sr.y)}, + {GSVector4(left, bottom, 0.5f, 1.0f), GSVector2(sr.x, sr.w)}, + {GSVector4(right, bottom, 0.5f, 1.0f), GSVector2(sr.z, sr.w)}, + }; + + IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + IASetInputLayout(m_convert.il); + IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + VSSetShader(m_convert.vs, NULL); + + // gs + + GSSetShader(NULL); + + // ps + + PSSetShader(ps, ps_cb); + PSSetSamplerState(linear ? m_convert.ln : m_convert.pt, NULL); + PSSetShaderResources(st, NULL); + + // rs + + RSSet(ds.x, ds.y); + + // + + DrawPrimitive(); + + // + + EndScene(); +} + +void GSDevice11::DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c) +{ + ClearRenderTarget(dt, c); + + if(st[1] && !slbg) + { + StretchRect(st[1], sr[1], dt, dr[1], m_merge.ps[0], NULL, true); + } + + if(st[0]) + { + m_ctx->UpdateSubresource(m_merge.cb, 0, NULL, &c, 0, 0); + + StretchRect(st[0], sr[0], dt, dr[0], m_merge.ps[mmod ? 1 : 0], m_merge.cb, m_merge.bs, true); + } +} + +void GSDevice11::DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset) +{ + GSVector4 s = GSVector4(dt->GetSize()); + + GSVector4 sr(0, 0, 1, 1); + GSVector4 dr(0.0f, yoffset, s.x, s.y + yoffset); + + InterlaceConstantBuffer cb; + + cb.ZrH = GSVector2(0, 1.0f / s.y); + cb.hH = s.y / 2; + + m_ctx->UpdateSubresource(m_interlace.cb, 0, NULL, &cb, 0, 0); + + StretchRect(st, sr, dt, dr, m_interlace.ps[shader], m_interlace.cb, linear); +} + +void GSDevice11::IASetVertexBuffer(const void* vertices, size_t stride, size_t count) +{ + ASSERT(m_vertices.count == 0); + + if(count * stride > m_vertices.limit * m_vertices.stride) + { + m_vertices.vb_old = m_vertices.vb; + m_vertices.vb = NULL; + m_vertices.start = 0; + m_vertices.count = 0; + m_vertices.limit = max(count * 3 / 2, 10000); + } + + if(m_vertices.vb == NULL) + { + D3D11_BUFFER_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + bd.Usage = D3D11_USAGE_DYNAMIC; + bd.ByteWidth = m_vertices.limit * stride; + bd.BindFlags = D3D11_BIND_VERTEX_BUFFER; + bd.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE; + + HRESULT hr; + + hr = m_dev->CreateBuffer(&bd, NULL, &m_vertices.vb); + + if(FAILED(hr)) return; + } + + D3D11_MAP type = D3D11_MAP_WRITE_NO_OVERWRITE; + + if(m_vertices.start + count > m_vertices.limit || stride != m_vertices.stride) + { + m_vertices.start = 0; + + type = D3D11_MAP_WRITE_DISCARD; + } + + D3D11_MAPPED_SUBRESOURCE m; + + if(SUCCEEDED(m_ctx->Map(m_vertices.vb, 0, type, 0, &m))) + { + GSVector4i::storent((uint8*)m.pData + m_vertices.start * stride, vertices, count * stride); + + m_ctx->Unmap(m_vertices.vb, 0); + } + + m_vertices.count = count; + m_vertices.stride = stride; + + IASetVertexBuffer(m_vertices.vb, stride); +} + +void GSDevice11::IASetVertexBuffer(ID3D11Buffer* vb, size_t stride) +{ + if(m_vb != vb || m_vb_stride != stride) + { + uint32 offset = 0; + + m_ctx->IASetVertexBuffers(0, 1, &vb, &stride, &offset); + + m_vb = vb; + m_vb_stride = stride; + } +} + +void GSDevice11::IASetInputLayout(ID3D11InputLayout* layout) +{ + if(m_layout != layout) + { + m_ctx->IASetInputLayout(layout); + + m_layout = layout; + } +} + +void GSDevice11::IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY topology) +{ + if(m_topology != topology) + { + m_ctx->IASetPrimitiveTopology(topology); + + m_topology = topology; + } +} + +void GSDevice11::VSSetShader(ID3D11VertexShader* vs, ID3D11Buffer* vs_cb) +{ + if(m_vs != vs) + { + m_ctx->VSSetShader(vs, NULL, 0); + + m_vs = vs; + } + + if(m_vs_cb != vs_cb) + { + m_ctx->VSSetConstantBuffers(0, 1, &vs_cb); + + m_vs_cb = vs_cb; + } +} + +void GSDevice11::GSSetShader(ID3D11GeometryShader* gs) +{ + if(m_gs != gs) + { + m_ctx->GSSetShader(gs, NULL, 0); + + m_gs = gs; + } +} + +void GSDevice11::PSSetShaderResources(GSTexture* sr0, GSTexture* sr1) +{ + ID3D11ShaderResourceView* srv0 = NULL; + ID3D11ShaderResourceView* srv1 = NULL; + + if(sr0) srv0 = *(GSTexture11*)sr0; + if(sr1) srv1 = *(GSTexture11*)sr1; + + if(m_ps_srv[0] != srv0 || m_ps_srv[1] != srv1) + { + ID3D11ShaderResourceView* srvs[] = {srv0, srv1}; + + m_ctx->PSSetShaderResources(0, 2, srvs); + + m_ps_srv[0] = srv0; + m_ps_srv[1] = srv1; + } +} + +void GSDevice11::PSSetShader(ID3D11PixelShader* ps, ID3D11Buffer* ps_cb) +{ + if(m_ps != ps) + { + m_ctx->PSSetShader(ps, NULL, 0); + + m_ps = ps; + } + + if(m_ps_cb != ps_cb) + { + m_ctx->PSSetConstantBuffers(0, 1, &ps_cb); + + m_ps_cb = ps_cb; + } +} + +void GSDevice11::PSSetSamplerState(ID3D11SamplerState* ss0, ID3D11SamplerState* ss1) +{ + if(m_ps_ss[0] != ss0 || m_ps_ss[1] != ss1) + { + ID3D11SamplerState* sss[] = {ss0, ss1}; + + m_ctx->PSSetSamplers(0, 2, sss); + + m_ps_ss[0] = ss0; + m_ps_ss[1] = ss1; + } +} + +void GSDevice11::RSSet(int width, int height, const GSVector4i* scissor) +{ + if(m_viewport.x != width || m_viewport.y != height) + { + D3D11_VIEWPORT vp; + + memset(&vp, 0, sizeof(vp)); + + vp.TopLeftX = 0; + vp.TopLeftY = 0; + vp.Width = width; + vp.Height = height; + vp.MinDepth = 0.0f; + vp.MaxDepth = 1.0f; + + m_ctx->RSSetViewports(1, &vp); + + m_viewport = GSVector2i(width, height); + } + + GSVector4i r = scissor ? *scissor : GSVector4i(0, 0, width, height); + + if(!m_scissor.eq(r)) + { + m_ctx->RSSetScissorRects(1, r); + + m_scissor = r; + } +} + +void GSDevice11::OMSetDepthStencilState(ID3D11DepthStencilState* dss, uint8 sref) +{ + if(m_dss != dss || m_sref != sref) + { + m_ctx->OMSetDepthStencilState(dss, sref); + + m_dss = dss; + m_sref = sref; + } +} + +void GSDevice11::OMSetBlendState(ID3D11BlendState* bs, float bf) +{ + if(m_bs != bs || m_bf != bf) + { + float BlendFactor[] = {bf, bf, bf, 0}; + + m_ctx->OMSetBlendState(bs, BlendFactor, 0xffffffff); + + m_bs = bs; + m_bf = bf; + } +} + +void GSDevice11::OMSetRenderTargets(GSTexture* rt, GSTexture* ds) +{ + ID3D11RenderTargetView* rtv = NULL; + ID3D11DepthStencilView* dsv = NULL; + + if(rt) rtv = *(GSTexture11*)rt; + if(ds) dsv = *(GSTexture11*)ds; + + if(m_rtv != rtv || m_dsv != dsv) + { + m_ctx->OMSetRenderTargets(1, &rtv, dsv); + + m_rtv = rtv; + m_dsv = dsv; + } +} + +HRESULT GSDevice11::CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11VertexShader** vs, D3D11_INPUT_ELEMENT_DESC* layout, int count, ID3D11InputLayout** il) +{ + HRESULT hr; + + vector m; + + PrepareShaderMacro(m, macro, m_shader.model.c_str()); + + CComPtr shader, error; + + hr = D3DX11CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), m_shader.vs.c_str(), 0, 0, NULL, &shader, &error, NULL); + + if(error) + { + printf("%s\n", (const char*)error->GetBufferPointer()); + } + + if(FAILED(hr)) + { + return hr; + } + + hr = m_dev->CreateVertexShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), NULL, vs); + + if(FAILED(hr)) + { + return hr; + } + + hr = m_dev->CreateInputLayout(layout, count, shader->GetBufferPointer(), shader->GetBufferSize(), il); + + if(FAILED(hr)) + { + return hr; + } + + return hr; +} + +HRESULT GSDevice11::CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11GeometryShader** gs) +{ + HRESULT hr; + + vector m; + + PrepareShaderMacro(m, macro, m_shader.model.c_str()); + + CComPtr shader, error; + + hr = D3DX11CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), m_shader.gs.c_str(), 0, 0, NULL, &shader, &error, NULL); + + if(error) + { + printf("%s\n", (const char*)error->GetBufferPointer()); + } + + if(FAILED(hr)) + { + return hr; + } + + hr = m_dev->CreateGeometryShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(), NULL, gs); + + if(FAILED(hr)) + { + return hr; + } + + return hr; +} + +HRESULT GSDevice11::CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11PixelShader** ps) +{ + HRESULT hr; + + vector m; + + PrepareShaderMacro(m, macro, m_shader.model.c_str()); + + CComPtr shader, error; + + hr = D3DX11CompileFromResource(theApp.GetModuleHandle(), MAKEINTRESOURCE(id), NULL, &m[0], NULL, entry.c_str(), m_shader.ps.c_str(), 0, 0, NULL, &shader, &error, NULL); + + if(error) + { + printf("%s\n", (const char*)error->GetBufferPointer()); + } + + if(FAILED(hr)) + { + return hr; + } + + hr = m_dev->CreatePixelShader((void*)shader->GetBufferPointer(), shader->GetBufferSize(),NULL, ps); + + if(FAILED(hr)) + { + return hr; + } + + return hr; +} diff --git a/plugins/GSdx/GSDevice11.h b/plugins/GSdx/GSDevice11.h new file mode 100644 index 0000000000..2fa406e941 --- /dev/null +++ b/plugins/GSdx/GSDevice11.h @@ -0,0 +1,147 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSDevice.h" +#include "GSTexture11.h" + +class GSDevice11 : public GSDevice +{ + ID3D11Buffer* m_vb; + size_t m_vb_stride; + ID3D11InputLayout* m_layout; + D3D11_PRIMITIVE_TOPOLOGY m_topology; + ID3D11VertexShader* m_vs; + ID3D11Buffer* m_vs_cb; + ID3D11GeometryShader* m_gs; + ID3D11ShaderResourceView* m_ps_srv[2]; + ID3D11PixelShader* m_ps; + ID3D11Buffer* m_ps_cb; + ID3D11SamplerState* m_ps_ss[2]; + GSVector2i m_viewport; + GSVector4i m_scissor; + ID3D11DepthStencilState* m_dss; + uint8 m_sref; + ID3D11BlendState* m_bs; + float m_bf; + ID3D11RenderTargetView* m_rtv; + ID3D11DepthStencilView* m_dsv; + + // + + GSTexture* Create(int type, int w, int h, int format); + + void DoMerge(GSTexture* st[2], GSVector4* sr, GSVector4* dr, GSTexture* dt, bool slbg, bool mmod, const GSVector4& c); + void DoInterlace(GSTexture* st, GSTexture* dt, int shader, bool linear, float yoffset = 0); + + // + + D3D_FEATURE_LEVEL m_level; + CComPtr m_dev; + CComPtr m_ctx; + CComPtr m_swapchain; + struct {string model, vs, gs, ps;} m_shader; + + struct + { + CComPtr vb, vb_old; + size_t stride, start, count, limit; + } m_vertices; + +public: // TODO + CComPtr m_rs; + + struct + { + CComPtr il; + CComPtr vs; + CComPtr ps[7]; + CComPtr ln; + CComPtr pt; + CComPtr dss; + CComPtr bs; + } m_convert; + + struct + { + CComPtr ps[2]; + CComPtr cb; + CComPtr bs; + } m_merge; + + struct + { + CComPtr ps[4]; + CComPtr cb; + } m_interlace; + +public: + GSDevice11(); + virtual ~GSDevice11(); + + bool Create(GSWnd* wnd, bool vsync); + bool Reset(int w, int h, bool fs); + + void Flip(); + + void BeginScene(); + void DrawPrimitive(); + void EndScene(); + + void ClearRenderTarget(GSTexture* t, const GSVector4& c); + void ClearRenderTarget(GSTexture* t, uint32 c); + void ClearDepth(GSTexture* t, float c); + void ClearStencil(GSTexture* t, uint8 c); + + GSTexture* CreateRenderTarget(int w, int h, int format = 0); + GSTexture* CreateDepthStencil(int w, int h, int format = 0); + GSTexture* CreateTexture(int w, int h, int format = 0); + GSTexture* CreateOffscreen(int w, int h, int format = 0); + + GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sr, int w, int h, int format = 0); + + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, int shader = 0, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D11PixelShader* ps, ID3D11Buffer* ps_cb, bool linear = true); + void StretchRect(GSTexture* st, const GSVector4& sr, GSTexture* dt, const GSVector4& dr, ID3D11PixelShader* ps, ID3D11Buffer* ps_cb, ID3D11BlendState* bs, bool linear = true); + + void IASetVertexBuffer(const void* vertices, size_t stride, size_t count); + void IASetVertexBuffer(ID3D11Buffer* vb, size_t stride); + void IASetInputLayout(ID3D11InputLayout* layout); + void IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY topology); + void VSSetShader(ID3D11VertexShader* vs, ID3D11Buffer* vs_cb); + void GSSetShader(ID3D11GeometryShader* gs); + void PSSetShaderResources(GSTexture* sr0, GSTexture* sr1); + void PSSetShader(ID3D11PixelShader* ps, ID3D11Buffer* ps_cb); + void PSSetSamplerState(ID3D11SamplerState* ss0, ID3D11SamplerState* ss1); + void RSSet(int width, int height, const GSVector4i* scissor = NULL); + void OMSetDepthStencilState(ID3D11DepthStencilState* dss, uint8 sref); + void OMSetBlendState(ID3D11BlendState* bs, float bf); + void OMSetRenderTargets(GSTexture* rt, GSTexture* ds); + + ID3D11Device* operator->() {return m_dev;} + operator ID3D11Device*() {return m_dev;} + operator ID3D11DeviceContext*() {return m_ctx;} + + HRESULT CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11VertexShader** vs, D3D11_INPUT_ELEMENT_DESC* layout, int count, ID3D11InputLayout** il); + HRESULT CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11GeometryShader** gs); + HRESULT CompileShader(uint32 id, const string& entry, D3D11_SHADER_MACRO* macro, ID3D11PixelShader** ps); +}; diff --git a/plugins/GSdx/GSRendererHW11.cpp b/plugins/GSdx/GSRendererHW11.cpp new file mode 100644 index 0000000000..ec7f11447b --- /dev/null +++ b/plugins/GSdx/GSRendererHW11.cpp @@ -0,0 +1,593 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererHW11.h" +#include "GSCrc.h" +#include "resource.h" + +GSRendererHW11::GSRendererHW11(uint8* base, bool mt, void (*irq)()) + : GSRendererHW(base, mt, irq, new GSDevice11(), new GSTextureCache11(this)) +{ + InitVertexKick(); +} + +bool GSRendererHW11::Create(const string& title) +{ + if(!__super::Create(title)) + return false; + + if(!m_tfx.Create((GSDevice11*)m_dev)) + return false; + + // + + D3D11_DEPTH_STENCIL_DESC dsd; + + memset(&dsd, 0, sizeof(dsd)); + + dsd.DepthEnable = false; + dsd.StencilEnable = true; + dsd.StencilReadMask = 1; + dsd.StencilWriteMask = 1; + dsd.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS; + dsd.FrontFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; + dsd.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilFunc = D3D11_COMPARISON_ALWAYS; + dsd.BackFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; + dsd.BackFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + + (*(GSDevice11*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); + + D3D11_BLEND_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + (*(GSDevice11*)m_dev)->CreateBlendState(&bd, &m_date.bs); + + // + + return true; +} + +template +void GSRendererHW11::VertexKick(bool skip) +{ + GSVertexHW11& dst = m_vl.AddTail(); + + dst.vi[0] = m_v.vi[0]; + dst.vi[1] = m_v.vi[1]; + + if(tme && fst) + { + GSVector4::storel(&dst.ST, m_v.GetUV()); + } + + int count = 0; + + if(GSVertexHW11* v = DrawingKick(skip, count)) + { + GSVector4i scissor = m_context->scissor.dx10; + + #if _M_SSE >= 0x401 + + GSVector4i pmin, pmax, v0, v1, v2; + + switch(prim) + { + case GS_POINTLIST: + v0 = GSVector4i::load((int)v[0].p.xy).upl16(); + pmin = v0; + pmax = v0; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + pmin = v0.min_u16(v1).upl16(); + pmax = v0.max_u16(v1).upl16(); + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + v2 = GSVector4i::load((int)v[2].p.xy); + pmin = v0.min_u16(v1).min_u16(v2).upl16(); + pmax = v0.max_u16(v1).max_u16(v2).upl16(); + break; + } + + GSVector4i test = (pmax < scissor) | (pmin > scissor.zwxy()); + + if(test.mask() & 0xff) + { + return; + } + + #else + + switch(prim) + { + case GS_POINTLIST: + if(v[0].p.x < scissor.x + || v[0].p.x > scissor.z + || v[0].p.y < scissor.y + || v[0].p.y > scissor.w) + { + return; + } + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w) + { + return; + } + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x && v[2].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z && v[2].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y && v[2].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w && v[2].p.y > scissor.w) + { + return; + } + break; + } + + #endif + + m_count += count; + } +} + +void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +{ + GSDrawingEnvironment& env = m_env; + GSDrawingContext* context = m_context; +/* + if(s_dump) + { + TRACE(_T("\n")); + + TRACE(_T("PRIM = %d, ZMSK = %d, ZTE = %d, ZTST = %d, ATE = %d, ATST = %d, AFAIL = %d, AREF = %02x\n"), + PRIM->PRIM, context->ZBUF.ZMSK, + context->TEST.ZTE, context->TEST.ZTST, + context->TEST.ATE, context->TEST.ATST, context->TEST.AFAIL, context->TEST.AREF); + + for(int i = 0; i < m_count; i++) + { + TRACE(_T("[%d] %3.0f %3.0f %3.0f %3.0f\n"), i, (float)m_vertices[i].p.x / 16, (float)m_vertices[i].p.y / 16, (float)m_vertices[i].p.z, (float)m_vertices[i].a); + } + } +*/ + D3D11_PRIMITIVE_TOPOLOGY topology; + int prims = 0; + + switch(prim) + { + case GS_POINTLIST: + topology = D3D11_PRIMITIVE_TOPOLOGY_POINTLIST; + prims = m_count; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + topology = D3D11_PRIMITIVE_TOPOLOGY_LINELIST; + prims = m_count / 2; + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + topology = D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST; + prims = m_count / 3; + break; + default: + __assume(0); + } + + m_perfmon.Put(GSPerfMon::Prim, prims); + m_perfmon.Put(GSPerfMon::Draw, 1); + + // date + + SetupDATE(rt, ds); + + // + + m_dev->BeginScene(); + + // om + + GSTextureFX11::OMDepthStencilSelector om_dssel; + + om_dssel.zte = context->TEST.ZTE; + om_dssel.ztst = context->TEST.ZTST; + om_dssel.zwe = !context->ZBUF.ZMSK; + om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; + + GSTextureFX11::OMBlendSelector om_bsel; + + om_bsel.abe = PRIM->ABE || (prim == 1 || prim == 2) && PRIM->AA1; + om_bsel.a = context->ALPHA.A; + om_bsel.b = context->ALPHA.B; + om_bsel.c = context->ALPHA.C; + om_bsel.d = context->ALPHA.D; + om_bsel.wr = (context->FRAME.FBMSK & 0x000000ff) != 0x000000ff; + om_bsel.wg = (context->FRAME.FBMSK & 0x0000ff00) != 0x0000ff00; + om_bsel.wb = (context->FRAME.FBMSK & 0x00ff0000) != 0x00ff0000; + om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; + + float bf = (float)(int)context->ALPHA.FIX / 0x80; + + // vs + + GSTextureFX11::VSSelector vs_sel; + + vs_sel.bppz = 0; + vs_sel.tme = PRIM->TME; + vs_sel.fst = PRIM->FST; + vs_sel.prim = prim; + + if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) + { + if(context->ZBUF.PSM == PSM_PSMZ24) + { + if(WrapZ(0xffffff)) + { + vs_sel.bppz = 1; + om_dssel.ztst = 1; + } + } + else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) + { + if(WrapZ(0xffff)) + { + vs_sel.bppz = 2; + om_dssel.ztst = 1; + } + } + } + + GSTextureFX11::VSConstantBuffer vs_cb; + + float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); + float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); + float ox = (float)(int)context->XYOFFSET.OFX; + float oy = (float)(int)context->XYOFFSET.OFY; + float ox2 = 1.0f / rt->GetWidth(); + float oy2 = 1.0f / rt->GetHeight(); + + vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); + vs_cb.VertexOffset = GSVector4(ox * sx - ox2 + 1, -(oy * sy - oy2 + 1), 0.0f, -1.0f); + vs_cb.TextureScale = GSVector2(1.0f, 1.0f); + + if(PRIM->TME && PRIM->FST) + { + vs_cb.TextureScale.x = 1.0f / (16 << context->TEX0.TW); + vs_cb.TextureScale.y = 1.0f / (16 << context->TEX0.TH); + } + + // gs + + GSTextureFX11::GSSelector gs_sel; + + gs_sel.iip = PRIM->IIP; + gs_sel.prim = GSUtil::GetPrimClass(prim); + + // ps + + GSTextureFX11::PSSelector ps_sel; + + ps_sel.fst = PRIM->FST; + ps_sel.wms = context->CLAMP.WMS; + ps_sel.wmt = context->CLAMP.WMT; + ps_sel.bpp = 0; + ps_sel.aem = env.TEXA.AEM; + ps_sel.tfx = context->TEX0.TFX; + ps_sel.tcc = context->TEX0.TCC; + ps_sel.ate = context->TEST.ATE; + ps_sel.atst = context->TEST.ATST; + ps_sel.fog = PRIM->FGE; + ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; + ps_sel.fba = context->FBA.FBA; + ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; + ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; + + GSTextureFX11::PSSamplerSelector ps_ssel; + + ps_ssel.tau = 0; + ps_ssel.tav = 0; + ps_ssel.ltf = ps_sel.ltf; + + GSTextureFX11::PSConstantBuffer ps_cb; + + ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; + + if(context->TEST.ATST == 2 || context->TEST.ATST == 5) + { + ps_cb.FogColor_AREF.a -= 0.9f / 255; + } + else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) + { + ps_cb.FogColor_AREF.a += 0.9f / 255; + } + + if(tex) + { + ps_sel.bpp = tex->m_bpp; + + int w = tex->m_texture->GetWidth(); + int h = tex->m_texture->GetHeight(); + + ps_cb.WH_TA = GSVector4((int)(1 << context->TEX0.TW), (int)(1 << context->TEX0.TH), env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); + ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); + + switch(context->CLAMP.WMS) + { + case 0: + ps_ssel.tau = 1; + break; + case 1: + ps_ssel.tau = 0; + break; + case 2: + ps_cb.MinMax.x = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); + ps_cb.MinMax.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); + ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); + ps_ssel.tau = 0; + break; + case 3: + ps_cb.MskFix.x = context->CLAMP.MINU; + ps_cb.MskFix.z = context->CLAMP.MAXU; + ps_ssel.tau = 1; + break; + default: + __assume(0); + } + + switch(context->CLAMP.WMT) + { + case 0: + ps_ssel.tav = 1; + break; + case 1: + ps_ssel.tav = 0; + break; + case 2: + ps_cb.MinMax.y = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); + ps_cb.MinMax.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); + ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); + ps_ssel.tav = 0; + break; + case 3: + ps_cb.MskFix.y = context->CLAMP.MINV; + ps_cb.MskFix.w = context->CLAMP.MAXV; + ps_ssel.tav = 1; + break; + default: + __assume(0); + } + } + else + { + ps_sel.tfx = 4; + } + + // rs + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); + + // + + m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); + m_tfx.SetupIA(m_vertices, m_count, topology); + m_tfx.SetupVS(vs_sel, &vs_cb); + m_tfx.SetupGS(gs_sel); + m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); + m_tfx.SetupRS(w, h, scissor); + + // draw + + if(context->TEST.DoFirstPass()) + { + m_tfx.Draw(); + } + + if(context->TEST.DoSecondPass()) + { + ASSERT(!env.PABE.PABE); + + static const uint32 iatst[] = {1, 0, 5, 6, 7, 2, 3, 4}; + + ps_sel.atst = iatst[ps_sel.atst]; + + m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel); + + bool z = om_dssel.zwe; + bool r = om_bsel.wr; + bool g = om_bsel.wg; + bool b = om_bsel.wb; + bool a = om_bsel.wa; + + switch(context->TEST.AFAIL) + { + case 0: z = r = g = b = a = false; break; // none + case 1: z = false; break; // rgba + case 2: r = g = b = a = false; break; // z + case 3: z = a = false; break; // rgb + default: __assume(0); + } + + if(z || r || g || b || a) + { + om_dssel.zwe = z; + om_bsel.wr = r; + om_bsel.wg = g; + om_bsel.wb = b; + om_bsel.wa = a; + + m_tfx.UpdateOM(om_dssel, om_bsel, bf); + + m_tfx.Draw(); + } + } + + m_dev->EndScene(); +} + +bool GSRendererHW11::WrapZ(uint32 maxz) +{ + // should only run once if z values are in the z buffer range + + for(int i = 0, j = m_count; i < j; i++) + { + if(m_vertices[i].p.z <= maxz) + { + return false; + } + } + + return true; +} + +void GSRendererHW11::SetupDATE(GSTexture* rt, GSTexture* ds) +{ + if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) + + GSDevice11* dev = (GSDevice11*)m_dev; + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + if(GSTexture* t = dev->CreateRenderTarget(w, h)) + { + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + + GSVector4 mm; + + // TODO + + mm = GSVector4(-1, -1, 1, 1); + + // if(m_count < 100) + { + GSVector4 pmin(65535, 65535, 0, 0); + GSVector4 pmax = GSVector4::zero(); + + for(int i = 0, j = m_count; i < j; i++) + { + GSVector4 p(m_vertices[i].vi[0].uph16()); + + pmin = p.minv(pmin); + pmax = p.maxv(pmax); + } + + mm += pmin.xyxy(pmax); + + float sx = 2.0f * rt->m_scale.x / (w * 16); + float sy = 2.0f * rt->m_scale.y / (h * 16); + float ox = (float)(int)m_context->XYOFFSET.OFX; + float oy = (float)(int)m_context->XYOFFSET.OFY; + + mm.x = (mm.x - ox) * sx - 1; + mm.y = (mm.y - oy) * sy - 1; + mm.z = (mm.z - ox) * sx - 1; + mm.w = (mm.w - oy) * sy - 1; + + if(mm.x < -1) mm.x = -1; + if(mm.y < -1) mm.y = -1; + if(mm.z > +1) mm.z = +1; + if(mm.w > +1) mm.w = +1; + } + + GSVector4 uv = (mm + 1.0f) / 2.0f; + + // + + dev->BeginScene(); + + dev->ClearStencil(ds, 0); + + // om + + dev->OMSetDepthStencilState(m_date.dss, 1); + dev->OMSetBlendState(m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVertexPT1 vertices[] = + { + {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, + {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, + {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, + {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + }; + + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL); + + // gs + + dev->GSSetShader(NULL); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); + dev->PSSetSamplerState(dev->m_convert.pt, NULL); + + // rs + + dev->RSSet(w, h); + + // set + + dev->DrawPrimitive(); + + // + + dev->EndScene(); + + dev->Recycle(t); + } +} diff --git a/plugins/GSdx/GSRendererHW11.h b/plugins/GSdx/GSRendererHW11.h new file mode 100644 index 0000000000..f3763a7dd1 --- /dev/null +++ b/plugins/GSdx/GSRendererHW11.h @@ -0,0 +1,52 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSRendererHW.h" +#include "GSVertexHW.h" +#include "GSTextureCache11.h" +#include "GSTextureFX11.h" + +class GSRendererHW11 : public GSRendererHW +{ + bool WrapZ(uint32 maxz); + +protected: + GSTextureFX11 m_tfx; + + void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + + struct + { + CComPtr dss; + CComPtr bs; + } m_date; + + void SetupDATE(GSTexture* rt, GSTexture* ds); + +public: + GSRendererHW11(uint8* base, bool mt, void (*irq)()); + + bool Create(const string& title); + + template void VertexKick(bool skip); +}; \ No newline at end of file diff --git a/plugins/GSdx/GSTexture11.cpp b/plugins/GSdx/GSTexture11.cpp new file mode 100644 index 0000000000..c80cc7aae6 --- /dev/null +++ b/plugins/GSdx/GSTexture11.cpp @@ -0,0 +1,200 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSTexture11.h" + +GSTexture11::GSTexture11(ID3D11Texture2D* texture) + : m_texture(texture) +{ + ASSERT(m_texture); + + m_texture->GetDevice(&m_dev); + m_texture->GetDesc(&m_desc); + + m_dev->GetImmediateContext(&m_ctx); +} + +int GSTexture11::GetType() const +{ + if(m_desc.BindFlags & D3D11_BIND_RENDER_TARGET) return GSTexture::RenderTarget; + if(m_desc.BindFlags & D3D11_BIND_DEPTH_STENCIL) return GSTexture::DepthStencil; + if(m_desc.BindFlags & D3D11_BIND_SHADER_RESOURCE) return GSTexture::Texture; + if(m_desc.Usage == D3D11_USAGE_STAGING) return GSTexture::Offscreen; + return GSTexture::None; +} + +int GSTexture11::GetWidth() const +{ + return m_desc.Width; +} + +int GSTexture11::GetHeight() const +{ + return m_desc.Height; +} + +int GSTexture11::GetFormat() const +{ + return m_desc.Format; +} + +bool GSTexture11::Update(const GSVector4i& r, const void* data, int pitch) +{ + if(m_dev && m_texture) + { + D3D11_BOX box = {r.left, r.top, 0, r.right, r.bottom, 1}; + + m_ctx->UpdateSubresource(m_texture, 0, &box, data, pitch, 0); + + return true; + } + + return false; +} + +bool GSTexture11::Map(GSMap& m, const GSVector4i* r) +{ + if(r != NULL) + { + // ASSERT(0); // not implemented + + return false; + } + + if(m_texture) + { + D3D11_MAPPED_SUBRESOURCE map; + + if(SUCCEEDED(m_ctx->Map(m_texture, 0, D3D11_MAP_READ_WRITE, 0, &map))) + { + m.bits = (uint8*)map.pData; + m.pitch = (int)map.RowPitch; + + return true; + } + } + + return false; +} + +void GSTexture11::Unmap() +{ + if(m_texture) + { + m_ctx->Unmap(m_texture, 0); + } +} + +bool GSTexture11::Save(const string& fn, bool dds) +{ + CComPtr res; + + if(m_desc.BindFlags & D3D11_BIND_DEPTH_STENCIL) + { + HRESULT hr; + + D3D11_TEXTURE2D_DESC desc; + + memset(&desc, 0, sizeof(desc)); + + m_texture->GetDesc(&desc); + + desc.Usage = D3D11_USAGE_STAGING; + desc.BindFlags = 0; + desc.CPUAccessFlags = D3D11_CPU_ACCESS_READ; + + CComPtr src, dst; + + hr = m_dev->CreateTexture2D(&desc, NULL, &src); + + m_ctx->CopyResource(src, m_texture); + + desc.Format = DXGI_FORMAT_R8G8B8A8_UNORM; + desc.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE; + + hr = m_dev->CreateTexture2D(&desc, NULL, &dst); + + D3D11_MAPPED_SUBRESOURCE sm, dm; + + hr = m_ctx->Map(src, 0, D3D11_MAP_READ, 0, &sm); + hr = m_ctx->Map(dst, 0, D3D11_MAP_WRITE, 0, &dm); + + uint8* s = (uint8*)sm.pData; + uint8* d = (uint8*)dm.pData; + + for(uint32 y = 0; y < desc.Height; y++, s += sm.RowPitch, d += dm.RowPitch) + { + for(uint32 x = 0; x < desc.Width; x++) + { + ((uint32*)d)[x] = (uint32)(((float*)s)[x*2] * UINT_MAX); + } + } + + m_ctx->Unmap(src, 0); + m_ctx->Unmap(dst, 0); + + res = dst; + } + else + { + res = m_texture; + } + + return SUCCEEDED(D3DX11SaveTextureToFile(m_ctx, res, dds ? D3DX11_IFF_DDS : D3DX11_IFF_BMP, fn.c_str())); +} + +GSTexture11::operator ID3D11Texture2D*() +{ + return m_texture; +} + +GSTexture11::operator ID3D11ShaderResourceView*() +{ + if(!m_srv && m_dev && m_texture) + { + m_dev->CreateShaderResourceView(m_texture, NULL, &m_srv); + } + + return m_srv; +} + +GSTexture11::operator ID3D11RenderTargetView*() +{ + ASSERT(m_dev); + + if(!m_rtv && m_dev && m_texture) + { + m_dev->CreateRenderTargetView(m_texture, NULL, &m_rtv); + } + + return m_rtv; +} + +GSTexture11::operator ID3D11DepthStencilView*() +{ + if(!m_dsv && m_dev && m_texture) + { + m_dev->CreateDepthStencilView(m_texture, NULL, &m_dsv); + } + + return m_dsv; +} diff --git a/plugins/GSdx/GSTexture11.h b/plugins/GSdx/GSTexture11.h new file mode 100644 index 0000000000..6b3b16fb6c --- /dev/null +++ b/plugins/GSdx/GSTexture11.h @@ -0,0 +1,53 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSTexture.h" + +class GSTexture11 : public GSTexture +{ + CComPtr m_dev; + CComPtr m_ctx; + CComPtr m_texture; + D3D11_TEXTURE2D_DESC m_desc; + CComPtr m_srv; + CComPtr m_rtv; + CComPtr m_dsv; + +public: + explicit GSTexture11(ID3D11Texture2D* texture); + + int GetType() const; + int GetWidth() const; + int GetHeight() const; + int GetFormat() const; + + bool Update(const GSVector4i& r, const void* data, int pitch); + bool Map(GSMap& m, const GSVector4i* r); + void Unmap(); + bool Save(const string& fn, bool dds = false); + + operator ID3D11Texture2D*(); + operator ID3D11ShaderResourceView*(); + operator ID3D11RenderTargetView*(); + operator ID3D11DepthStencilView*(); +}; diff --git a/plugins/GSdx/GSTextureCache11.cpp b/plugins/GSdx/GSTextureCache11.cpp new file mode 100644 index 0000000000..221bfb7cd3 --- /dev/null +++ b/plugins/GSdx/GSTextureCache11.cpp @@ -0,0 +1,309 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "StdAfx.h" +#include "GSTextureCache11.h" + +// GSTextureCache11 + +GSTextureCache11::GSTextureCache11(GSRenderer* r) + : GSTextureCache(r) +{ +} + +// GSRenderTargetHW10 + +void GSTextureCache11::GSRenderTargetHW11::Read(const GSVector4i& r) +{ + if(m_TEX0.PSM != PSM_PSMCT32 + && m_TEX0.PSM != PSM_PSMCT24 + && m_TEX0.PSM != PSM_PSMCT16 + && m_TEX0.PSM != PSM_PSMCT16S) + { + //ASSERT(0); + return; + } + + if(!m_dirty.empty()) + { + return; + } + + // printf("GSRenderTarget::Read %d,%d - %d,%d (%08x)\n", r.left, r.top, r.right, r.bottom, m_TEX0.TBP0); + + // m_renderer->m_perfmon.Put(GSPerfMon::ReadRT, 1); + + int w = r.width(); + int h = r.height(); + + GSVector4 src = GSVector4(r) * GSVector4(m_texture->m_scale).xyxy() / GSVector4(m_texture->GetSize()).xyxy(); + + DXGI_FORMAT format = m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S ? DXGI_FORMAT_R16_UINT : DXGI_FORMAT_R8G8B8A8_UNORM; + + if(GSTexture* offscreen = m_renderer->m_dev->CopyOffscreen(m_texture, src, w, h, format)) + { + GSTexture::GSMap m; + + if(offscreen->Map(m)) + { + // TODO: block level write + + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; + + GSLocalMemory::pixelAddress pa = GSLocalMemory::m_psm[m_TEX0.PSM].pa; + + if(m_TEX0.PSM == PSM_PSMCT32) + { + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) + { + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel32(addr + offset[x], ((uint32*)m.bits)[i]); + } + } + } + else if(m_TEX0.PSM == PSM_PSMCT24) + { + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) + { + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel24(addr + offset[x], ((uint32*)m.bits)[i]); + } + } + } + else if(m_TEX0.PSM == PSM_PSMCT16 || m_TEX0.PSM == PSM_PSMCT16S) + { + for(int y = r.top; y < r.bottom; y++, m.bits += m.pitch) + { + uint32 addr = pa(0, y, bp, bw); + int* offset = GSLocalMemory::m_psm[m_TEX0.PSM].rowOffset[y & 7]; + + for(int x = r.left, i = 0; x < r.right; x++, i++) + { + m_renderer->m_mem.WritePixel16(addr + offset[x], ((uint16*)m.bits)[i]); + } + } + } + else + { + ASSERT(0); + } + + offscreen->Unmap(); + } + + m_renderer->m_dev->Recycle(offscreen); + } +} + +// GSTextureHW10 + +bool GSTextureCache11::GSCachedTextureHW11::Create() +{ + // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + m_bpp = 0; + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); + + return m_texture != NULL; +} + +bool GSTextureCache11::GSCachedTextureHW11::Create(GSRenderTarget* rt) +{ + // TODO: clean up this mess + + rt->Update(); + + // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + m_rendered = true; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + int tp = (int)m_TEX0.TW << 6; + + // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) + + int w = (int)(rt->m_texture->m_scale.x * tw); + int h = (int)(rt->m_texture->m_scale.y * th); + + GSVector2i rtsize = rt->m_texture->GetSize(); + + // pitch conversion + + if(rt->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM + { + // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) + + // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); + + GSVector4 size = GSVector4(rtsize).xyxy(); + GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); + + int bw = 64; + int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + + GSVector4i br(0, 0, bw, bh); + + int sw = (int)rt->m_TEX0.TBW << 6; + + int dw = (int)m_TEX0.TBW << 6; + int dh = 1 << m_TEX0.TH; + + if(sw != 0) + for(int dy = 0; dy < dh; dy += bh) + { + for(int dx = 0; dx < dw; dx += bw) + { + int o = dy * dw / bh + dx; + + int sx = o % sw; + int sy = o / sw; + + GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; + + m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); + + // TODO: this is quite a lot of StretchRect, do it with one Draw + } + } + } + else if(tw < tp) + { + // FIXME: timesplitters blurs the render target by blending itself over a couple of times + + if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) + { + return false; + } + } + + // width/height conversion + + GSVector2 scale = rt->m_texture->m_scale; + + GSVector4 dst(0, 0, w, h); + + if(w > rtsize.x) + { + scale.x = (float)rtsize.x / tw; + dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; + w = rtsize.x; + } + + if(h > rtsize.y) + { + scale.y = (float)rtsize.y / th; + dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; + h = rtsize.y; + } + + GSVector4 src(0, 0, w, h); + + GSTexture* st = m_texture ? m_texture : rt->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); + + if(!m_texture) + { + m_texture = dt; + } + + if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) + { + D3D11_BOX box = {0, 0, 0, w, h, 1}; + + ID3D11DeviceContext* ctx = *(GSDevice11*)m_renderer->m_dev; + + ctx->CopySubresourceRegion(*(GSTexture11*)dt, 0, 0, 0, 0, *(GSTexture11*)st, 0, &box); + } + else + { + src.z /= st->GetWidth(); + src.w /= st->GetHeight(); + + m_renderer->m_dev->StretchRect(st, src, dt, dst); + } + + if(dt != m_texture) + { + m_renderer->m_dev->Recycle(m_texture); + + m_texture = dt; + } + + m_texture->m_scale = scale; + + switch(m_TEX0.PSM) + { + case PSM_PSMCT32: + m_bpp = 0; + break; + case PSM_PSMCT24: + m_bpp = 1; + break; + case PSM_PSMCT16: + case PSM_PSMCT16S: + m_bpp = 2; + break; + case PSM_PSMT8H: + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); + m_initpalette = true; + break; + case PSM_PSMT4HL: + case PSM_PSMT4HH: + ASSERT(0); // TODO + break; + } + + return true; +} + +bool GSTextureCache11::GSCachedTextureHW11::Create(GSDepthStencil* ds) +{ + m_rendered = true; + + // TODO + + return false; +} diff --git a/plugins/GSdx/GSTextureCache11.h b/plugins/GSdx/GSTextureCache11.h new file mode 100644 index 0000000000..ec3d883f81 --- /dev/null +++ b/plugins/GSdx/GSTextureCache11.h @@ -0,0 +1,60 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSTextureCache.h" +#include "GSDevice11.h" + +class GSTextureCache11 : public GSTextureCache +{ + class GSRenderTargetHW11 : public GSRenderTarget + { + public: + explicit GSRenderTargetHW11(GSRenderer* r) : GSRenderTarget(r) {} + + void Read(const GSVector4i& r); + }; + + class GSDepthStencilHW11 : public GSDepthStencil + { + public: + explicit GSDepthStencilHW11(GSRenderer* r) : GSDepthStencil(r) {} + }; + + class GSCachedTextureHW11 : public GSCachedTexture + { + public: + explicit GSCachedTextureHW11(GSRenderer* r) : GSCachedTexture(r) {} + + bool Create(); + bool Create(GSRenderTarget* rt); + bool Create(GSDepthStencil* ds); + }; + +protected: + GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetHW11(m_renderer);} + GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilHW11(m_renderer);} + GSCachedTexture* CreateTexture() {return new GSCachedTextureHW11(m_renderer);} + +public: + GSTextureCache11(GSRenderer* r); +}; diff --git a/plugins/GSdx/GSTextureFX11.cpp b/plugins/GSdx/GSTextureFX11.cpp new file mode 100644 index 0000000000..d0efbc7ee3 --- /dev/null +++ b/plugins/GSdx/GSTextureFX11.cpp @@ -0,0 +1,526 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSTextureFX11.h" +#include "resource.h" + +GSTextureFX11::GSTextureFX11() + : m_dev(NULL) +{ + memset(&m_vs_cb_cache, 0, sizeof(m_vs_cb_cache)); + memset(&m_ps_cb_cache, 0, sizeof(m_ps_cb_cache)); +} + +bool GSTextureFX11::Create(GSDevice11* dev) +{ + m_dev = dev; + + VSSelector sel; + + sel.bppz = 0; + sel.tme = 0; + sel.fst = 0; + + VSConstantBuffer cb; + + SetupVS(sel, &cb); // creates layout + + HRESULT hr; + + D3D11_BUFFER_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + bd.ByteWidth = sizeof(VSConstantBuffer); + bd.Usage = D3D11_USAGE_DEFAULT; + bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; + + hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_vs_cb); + + if(FAILED(hr)) return false; + + memset(&bd, 0, sizeof(bd)); + + bd.ByteWidth = sizeof(PSConstantBuffer); + bd.Usage = D3D11_USAGE_DEFAULT; + bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; + + hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_ps_cb); + + if(FAILED(hr)) return false; + + D3D11_SAMPLER_DESC sd; + + memset(&sd, 0, sizeof(sd)); + + sd.Filter = D3D11_FILTER_MIN_MAG_MIP_POINT; + sd.AddressU = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressV = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressW = D3D11_TEXTURE_ADDRESS_CLAMP; + sd.MaxLOD = FLT_MAX; + sd.MaxAnisotropy = 16; + sd.ComparisonFunc = D3D11_COMPARISON_NEVER; + + hr = (*m_dev)->CreateSamplerState(&sd, &m_palette_ss); + + if(FAILED(hr)) return false; + + // + + return true; +} + +bool GSTextureFX11::SetupIA(const GSVertexHW11* vertices, int count, D3D11_PRIMITIVE_TOPOLOGY prim) +{ + m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); + m_dev->IASetInputLayout(m_il); + m_dev->IASetPrimitiveTopology(prim); + + return true; +} + +bool GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) +{ + hash_map >::const_iterator i = m_vs.find(sel); + + if(i == m_vs.end()) + { + string str[4]; + + str[0] = format("%d", sel.bppz); + str[1] = format("%d", sel.tme); + str[2] = format("%d", sel.fst); + str[3] = format("%d", sel.prim); + + D3D11_SHADER_MACRO macro[] = + { + {"VS_BPPZ", str[0].c_str()}, + {"VS_TME", str[1].c_str()}, + {"VS_FST", str[2].c_str()}, + {"VS_PRIM", str[3].c_str()}, + {NULL, NULL}, + }; + + D3D11_INPUT_ELEMENT_DESC layout[] = + { + {"POSITION", 0, DXGI_FORMAT_R16G16_UINT, 0, 8, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"POSITION", 1, DXGI_FORMAT_R32_UINT, 0, 12, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 0, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"TEXCOORD", 1, DXGI_FORMAT_R32_FLOAT, 0, 20, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 16, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"COLOR", 1, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 28, D3D11_INPUT_PER_VERTEX_DATA, 0}, + }; + + CComPtr il; + CComPtr vs; + + m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + + if(m_il == NULL) + { + m_il = il; + } + + m_vs[sel] = vs; + + i = m_vs.find(sel); + } + + if(m_vs_cb_cache.Update(cb)) + { + ID3D11DeviceContext* ctx = *m_dev; + + ctx->UpdateSubresource(m_vs_cb, 0, NULL, cb, 0, 0); + } + + m_dev->VSSetShader((*i).second, m_vs_cb); + + return true; +} + +bool GSTextureFX11::SetupGS(GSSelector sel) +{ + HRESULT hr; + + ID3D11GeometryShader* gs = NULL; + + if(sel.prim > 0 && (sel.iip == 0 || sel.prim == 3)) // geometry shader works in every case, but not needed + { + hash_map >::const_iterator i = m_gs.find(sel); + + if(i != m_gs.end()) + { + gs = (*i).second; + } + else + { + string str[2]; + + str[0] = format("%d", sel.iip); + str[1] = format("%d", sel.prim); + + D3D11_SHADER_MACRO macro[] = + { + {"IIP", str[0].c_str()}, + {"PRIM", str[1].c_str()}, + {NULL, NULL}, + }; + + hr = m_dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); + + m_gs[sel] = gs; + } + } + + m_dev->GSSetShader(gs); + + return true; +} + +bool GSTextureFX11::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) +{ + m_dev->PSSetShaderResources(tex, pal); + + UpdatePS(sel, cb, ssel); + + return true; +} + +void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) +{ + HRESULT hr; + + hash_map >::const_iterator i = m_ps.find(sel); + + if(i == m_ps.end()) + { + string str[14]; + + str[0] = format("%d", sel.fst); + str[1] = format("%d", sel.wms); + str[2] = format("%d", sel.wmt); + str[3] = format("%d", sel.bpp); + str[4] = format("%d", sel.aem); + str[5] = format("%d", sel.tfx); + str[6] = format("%d", sel.tcc); + str[7] = format("%d", sel.ate); + str[8] = format("%d", sel.atst); + str[9] = format("%d", sel.fog); + str[10] = format("%d", sel.clr1); + str[11] = format("%d", sel.fba); + str[12] = format("%d", sel.aout); + str[13] = format("%d", sel.ltf); + + D3D11_SHADER_MACRO macro[] = + { + {"FST", str[0].c_str()}, + {"WMS", str[1].c_str()}, + {"WMT", str[2].c_str()}, + {"BPP", str[3].c_str()}, + {"AEM", str[4].c_str()}, + {"TFX", str[5].c_str()}, + {"TCC", str[6].c_str()}, + {"ATE", str[7].c_str()}, + {"ATST", str[8].c_str()}, + {"FOG", str[9].c_str()}, + {"CLR1", str[10].c_str()}, + {"FBA", str[11].c_str()}, + {"AOUT", str[12].c_str()}, + {"LTF", str[13].c_str()}, + {NULL, NULL}, + }; + + CComPtr ps; + + hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); + + m_ps[sel] = ps; + + i = m_ps.find(sel); + } + + if(m_ps_cb_cache.Update(cb)) + { + ID3D11DeviceContext* ctx = *m_dev; + + ctx->UpdateSubresource(m_ps_cb, 0, NULL, cb, 0, 0); + } + + m_dev->PSSetShader((*i).second, m_ps_cb); + + ID3D11SamplerState* ss0 = NULL; + ID3D11SamplerState* ss1 = NULL; + + if(sel.tfx != 4) + { + if(!(sel.bpp < 3 && sel.wms < 3 && sel.wmt < 3)) + { + ssel.ltf = 0; + } + + hash_map >::const_iterator i = m_ps_ss.find(ssel); + + if(i != m_ps_ss.end()) + { + ss0 = (*i).second; + } + else + { + D3D11_SAMPLER_DESC sd; + + memset(&sd, 0, sizeof(sd)); + + sd.Filter = ssel.ltf ? D3D11_FILTER_MIN_MAG_LINEAR_MIP_POINT : D3D11_FILTER_MIN_MAG_MIP_POINT; + + sd.AddressU = ssel.tau ? D3D11_TEXTURE_ADDRESS_WRAP : D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressV = ssel.tav ? D3D11_TEXTURE_ADDRESS_WRAP : D3D11_TEXTURE_ADDRESS_CLAMP; + sd.AddressW = D3D11_TEXTURE_ADDRESS_CLAMP; + + sd.MaxLOD = FLT_MAX; + sd.MaxAnisotropy = 16; + sd.ComparisonFunc = D3D11_COMPARISON_NEVER; + + hr = (*m_dev)->CreateSamplerState(&sd, &ss0); + + m_ps_ss[ssel] = ss0; + } + + if(sel.bpp == 3) + { + ss1 = m_palette_ss; + } + } + + m_dev->PSSetSamplerState(ss0, ss1); +} + +void GSTextureFX11::SetupRS(int w, int h, const GSVector4i& scissor) +{ + m_dev->RSSet(w, h, &scissor); +} + +void GSTextureFX11::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds) +{ + UpdateOM(dssel, bsel, bf); + + m_dev->OMSetRenderTargets(rt, ds); +} + +void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf) +{ + HRESULT hr; + + hash_map >::const_iterator i = m_om_dss.find(dssel); + + if(i == m_om_dss.end()) + { + D3D11_DEPTH_STENCIL_DESC dsd; + + memset(&dsd, 0, sizeof(dsd)); + + if(dssel.date) + { + dsd.StencilEnable = true; + dsd.StencilReadMask = 1; + dsd.StencilWriteMask = 1; + dsd.FrontFace.StencilFunc = D3D11_COMPARISON_EQUAL; + dsd.FrontFace.StencilPassOp = D3D11_STENCIL_OP_KEEP; + dsd.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilFunc = D3D11_COMPARISON_EQUAL; + dsd.BackFace.StencilPassOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + } + + if(!(dssel.zte && dssel.ztst == 1 && !dssel.zwe)) + { + static const D3D11_COMPARISON_FUNC ztst[] = + { + D3D11_COMPARISON_NEVER, + D3D11_COMPARISON_ALWAYS, + D3D11_COMPARISON_GREATER_EQUAL, + D3D11_COMPARISON_GREATER + }; + + dsd.DepthEnable = dssel.zte; + dsd.DepthWriteMask = dssel.zwe ? D3D11_DEPTH_WRITE_MASK_ALL : D3D11_DEPTH_WRITE_MASK_ZERO; + dsd.DepthFunc = ztst[dssel.ztst]; + } + + CComPtr dss; + + hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); + + m_om_dss[dssel] = dss; + + i = m_om_dss.find(dssel); + } + + m_dev->OMSetDepthStencilState((*i).second, 1); + + hash_map >::const_iterator j = m_om_bs.find(bsel); + + if(j == m_om_bs.end()) + { + D3D11_BLEND_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + bd.RenderTarget[0].BlendEnable = bsel.abe; + + if(bsel.abe) + { + // (A:Cs/Cd/0 - B:Cs/Cd/0) * C:As/Ad/FIX + D:Cs/Cd/0 + + static const struct {int bogus; D3D11_BLEND_OP op; D3D11_BLEND src, dst;} map[3*3*3*3] = + { + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 0000: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 0001: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 0002: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 0010: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 0011: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 0012: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 0020: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 0021: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 0022: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {1, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_SRC1_ALPHA}, // * 0100: (Cs - Cd)*As + Cs ==> Cs*(As + 1) - Cd*As + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_INV_SRC1_ALPHA}, // 0101: (Cs - Cd)*As + Cd ==> Cs*As + Cd*(1 - As) + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_SRC1_ALPHA}, // 0102: (Cs - Cd)*As + 0 ==> Cs*As - Cd*As + {1, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_DEST_ALPHA}, // * 0110: (Cs - Cd)*Ad + Cs ==> Cs*(Ad + 1) - Cd*Ad + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_INV_DEST_ALPHA}, // 0111: (Cs - Cd)*Ad + Cd ==> Cs*Ad + Cd*(1 - Ad) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_DEST_ALPHA}, // 0112: (Cs - Cd)*Ad + 0 ==> Cs*Ad - Cd*Ad + {1, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_BLEND_FACTOR}, // * 0120: (Cs - Cd)*F + Cs ==> Cs*(F + 1) - Cd*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_INV_BLEND_FACTOR}, // 0121: (Cs - Cd)*F + Cd ==> Cs*F + Cd*(1 - F) + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_BLEND_FACTOR}, // 0122: (Cs - Cd)*F + 0 ==> Cs*F - Cd*F + {1, D3D11_BLEND_OP_ADD, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ZERO}, // * 0200: (Cs - 0)*As + Cs ==> Cs*(As + 1) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ONE}, // 0201: (Cs - 0)*As + Cd ==> Cs*As + Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ZERO}, // 0202: (Cs - 0)*As + 0 ==> Cs*As + {1, D3D11_BLEND_OP_ADD, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ZERO}, // * 0210: (Cs - 0)*Ad + Cs ==> Cs*(Ad + 1) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_ONE}, // 0211: (Cs - 0)*Ad + Cd ==> Cs*Ad + Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_ZERO}, // 0212: (Cs - 0)*Ad + 0 ==> Cs*Ad + {1, D3D11_BLEND_OP_ADD, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_ZERO}, // * 0220: (Cs - 0)*F + Cs ==> Cs*(F + 1) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_ONE}, // 0221: (Cs - 0)*F + Cd ==> Cs*F + Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_ZERO}, // 0222: (Cs - 0)*F + 0 ==> Cs*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_SRC1_ALPHA, D3D11_BLEND_SRC1_ALPHA}, // 1000: (Cd - Cs)*As + Cs ==> Cd*As + Cs*(1 - As) + {1, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_SRC1_ALPHA}, // * 1001: (Cd - Cs)*As + Cd ==> Cd*(As + 1) - Cs*As + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_SRC1_ALPHA}, // 1002: (Cd - Cs)*As + 0 ==> Cd*As - Cs*As + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_DEST_ALPHA, D3D11_BLEND_DEST_ALPHA}, // 1010: (Cd - Cs)*Ad + Cs ==> Cd*Ad + Cs*(1 - Ad) + {1, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_DEST_ALPHA}, // * 1011: (Cd - Cs)*Ad + Cd ==> Cd*(Ad + 1) - Cs*Ad + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_DEST_ALPHA}, // 1012: (Cd - Cs)*Ad + 0 ==> Cd*Ad - Cs*Ad + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_BLEND_FACTOR, D3D11_BLEND_BLEND_FACTOR}, // 1020: (Cd - Cs)*F + Cs ==> Cd*F + Cs*(1 - F) + {1, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_BLEND_FACTOR},// * 1021: (Cd - Cs)*F + Cd ==> Cd*(F + 1) - Cs*F + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_BLEND_FACTOR},// 1022: (Cd - Cs)*F + 0 ==> Cd*F - Cs*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 1100: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 1101: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 1102: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 1110: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 1111: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 1112: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 1120: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 1121: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 1122: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_SRC1_ALPHA}, // 1200: (Cd - 0)*As + Cs ==> Cs + Cd*As + {2, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_COLOR, D3D11_BLEND_SRC1_ALPHA}, // ** 1201: (Cd - 0)*As + Cd ==> Cd*(1 + As) // ffxii main menu background glow effect + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_SRC1_ALPHA}, // 1202: (Cd - 0)*As + 0 ==> Cd*As + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_DEST_ALPHA}, // 1210: (Cd - 0)*Ad + Cs ==> Cs + Cd*Ad + {2, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_COLOR, D3D11_BLEND_DEST_ALPHA}, // ** 1211: (Cd - 0)*Ad + Cd ==> Cd*(1 + Ad) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_DEST_ALPHA}, // 1212: (Cd - 0)*Ad + 0 ==> Cd*Ad + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_BLEND_FACTOR}, // 1220: (Cd - 0)*F + Cs ==> Cs + Cd*F + {2, D3D11_BLEND_OP_ADD, D3D11_BLEND_DEST_COLOR, D3D11_BLEND_BLEND_FACTOR}, // ** 1221: (Cd - 0)*F + Cd ==> Cd*(1 + F) + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_BLEND_FACTOR}, // 1222: (Cd - 0)*F + 0 ==> Cd*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_SRC1_ALPHA, D3D11_BLEND_ZERO}, // 2000: (0 - Cs)*As + Cs ==> Cs*(1 - As) + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ONE}, // 2001: (0 - Cs)*As + Cd ==> Cd - Cs*As + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_SRC1_ALPHA, D3D11_BLEND_ZERO}, // 2002: (0 - Cs)*As + 0 ==> 0 - Cs*As + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_DEST_ALPHA, D3D11_BLEND_ZERO}, // 2010: (0 - Cs)*Ad + Cs ==> Cs*(1 - Ad) + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_ONE}, // 2011: (0 - Cs)*Ad + Cd ==> Cd - Cs*Ad + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_ZERO}, // 2012: (0 - Cs)*Ad + 0 ==> 0 - Cs*Ad + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_INV_BLEND_FACTOR, D3D11_BLEND_ZERO}, // 2020: (0 - Cs)*F + Cs ==> Cs*(1 - F) + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_ONE}, // 2021: (0 - Cs)*F + Cd ==> Cd - Cs*F + {0, D3D11_BLEND_OP_REV_SUBTRACT, D3D11_BLEND_BLEND_FACTOR, D3D11_BLEND_ZERO}, // 2022: (0 - Cs)*F + 0 ==> 0 - Cs*F + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ONE, D3D11_BLEND_SRC1_ALPHA}, // 2100: (0 - Cd)*As + Cs ==> Cs - Cd*As + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_INV_SRC1_ALPHA}, // 2101: (0 - Cd)*As + Cd ==> Cd*(1 - As) + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ZERO, D3D11_BLEND_SRC1_ALPHA}, // 2102: (0 - Cd)*As + 0 ==> 0 - Cd*As + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ONE, D3D11_BLEND_DEST_ALPHA}, // 2110: (0 - Cd)*Ad + Cs ==> Cs - Cd*Ad + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_INV_DEST_ALPHA}, // 2111: (0 - Cd)*Ad + Cd ==> Cd*(1 - Ad) + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ONE, D3D11_BLEND_DEST_ALPHA}, // 2112: (0 - Cd)*Ad + 0 ==> 0 - Cd*Ad + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ONE, D3D11_BLEND_BLEND_FACTOR}, // 2120: (0 - Cd)*F + Cs ==> Cs - Cd*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_INV_BLEND_FACTOR}, // 2121: (0 - Cd)*F + Cd ==> Cd*(1 - F) + {0, D3D11_BLEND_OP_SUBTRACT, D3D11_BLEND_ONE, D3D11_BLEND_BLEND_FACTOR}, // 2122: (0 - Cd)*F + 0 ==> 0 - Cd*F + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 2200: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 2201: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 2202: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 2210: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 2211: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 2212: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ONE, D3D11_BLEND_ZERO}, // 2220: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cs ==> Cs + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ONE}, // 2221: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + Cd ==> Cd + {0, D3D11_BLEND_OP_ADD, D3D11_BLEND_ZERO, D3D11_BLEND_ZERO}, // 2222: (Cs/Cd/0 - Cs/Cd/0)*As/Ad/F + 0 ==> 0 + }; + + // bogus: 0100, 0110, 0120, 0200, 0210, 0220, 1001, 1011, 1021 + + // tricky: 1201, 1211, 1221 + // + // Source.rgb = float3(1, 1, 1); + // 1201 Cd*(1 + As) => Source * Dest color + Dest * Source1 alpha + // 1211 Cd*(1 + Ad) => Source * Dest color + Dest * Dest alpha + // 1221 Cd*(1 + F) => Source * Dest color + Dest * Factor + + int i = ((bsel.a * 3 + bsel.b) * 3 + bsel.c) * 3 + bsel.d; + + bd.RenderTarget[0].BlendOp = map[i].op; + bd.RenderTarget[0].SrcBlend = map[i].src; + bd.RenderTarget[0].DestBlend = map[i].dst; + bd.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD; + bd.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_ONE; + bd.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ZERO; + + if(map[i].bogus == 1) + { + ASSERT(0); + + (bsel.a == 0 ? bd.RenderTarget[0].SrcBlend : bd.RenderTarget[0].DestBlend) = D3D11_BLEND_ONE; + } + } + + if(bsel.wr) bd.RenderTarget[0].RenderTargetWriteMask |= D3D11_COLOR_WRITE_ENABLE_RED; + if(bsel.wg) bd.RenderTarget[0].RenderTargetWriteMask |= D3D11_COLOR_WRITE_ENABLE_GREEN; + if(bsel.wb) bd.RenderTarget[0].RenderTargetWriteMask |= D3D11_COLOR_WRITE_ENABLE_BLUE; + if(bsel.wa) bd.RenderTarget[0].RenderTargetWriteMask |= D3D11_COLOR_WRITE_ENABLE_ALPHA; + + CComPtr bs; + + hr = (*m_dev)->CreateBlendState(&bd, &bs); + + m_om_bs[bsel] = bs; + + j = m_om_bs.find(bsel); + } + + m_dev->OMSetBlendState((*j).second, bf); +} + +void GSTextureFX11::Draw() +{ + m_dev->DrawPrimitive(); +} diff --git a/plugins/GSdx/GSTextureFX11.h b/plugins/GSdx/GSTextureFX11.h new file mode 100644 index 0000000000..e64b9349ab --- /dev/null +++ b/plugins/GSdx/GSTextureFX11.h @@ -0,0 +1,240 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSDevice11.h" + +class GSTextureFX11 +{ +public: + #pragma pack(push, 1) + + __declspec(align(16)) struct VSConstantBuffer + { + GSVector4 VertexScale; + GSVector4 VertexOffset; + GSVector2 TextureScale; + float _pad[2]; + + struct VSConstantBuffer() {memset(this, 0, sizeof(*this));} + + __forceinline bool Update(const VSConstantBuffer* cb) + { + GSVector4i* a = (GSVector4i*)this; + GSVector4i* b = (GSVector4i*)cb; + + GSVector4i b0 = b[0]; + GSVector4i b1 = b[1]; + GSVector4i b2 = b[2]; + + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2)).alltrue()) + { + a[0] = b0; + a[1] = b1; + a[2] = b2; + + return true; + } + + return false; + } + }; + + union VSSelector + { + struct + { + uint32 bppz:2; + uint32 tme:1; + uint32 fst:1; + uint32 prim:3; + }; + + uint32 key; + + operator uint32() {return key & 0x7f;} + }; + + __declspec(align(16)) struct PSConstantBuffer + { + GSVector4 FogColor_AREF; + GSVector4 HalfTexel; + GSVector4 WH_TA; + GSVector4 MinMax; + GSVector4 MinMaxF; + GSVector4i MskFix; + + struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} + + __forceinline bool Update(const PSConstantBuffer* cb) + { + GSVector4i* a = (GSVector4i*)this; + GSVector4i* b = (GSVector4i*)cb; + + GSVector4i b0 = b[0]; + GSVector4i b1 = b[1]; + GSVector4i b2 = b[2]; + GSVector4i b3 = b[3]; + GSVector4i b4 = b[4]; + GSVector4i b5 = b[5]; + + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) + { + a[0] = b0; + a[1] = b1; + a[2] = b2; + a[3] = b3; + a[4] = b4; + a[5] = b5; + + return true; + } + + return false; + } + }; + + union GSSelector + { + struct + { + uint32 iip:1; + uint32 prim:2; + }; + + uint32 key; + + operator uint32() {return key & 0x7;} + }; + + union PSSelector + { + struct + { + uint32 fst:1; + uint32 wms:2; + uint32 wmt:2; + uint32 bpp:3; + uint32 aem:1; + uint32 tfx:3; + uint32 tcc:1; + uint32 ate:1; + uint32 atst:3; + uint32 fog:1; + uint32 clr1:1; + uint32 fba:1; + uint32 aout:1; + uint32 ltf:1; + }; + + uint32 key; + + operator uint32() {return key & 0x3fffff;} + }; + + union PSSamplerSelector + { + struct + { + uint32 tau:1; + uint32 tav:1; + uint32 ltf:1; + }; + + uint32 key; + + operator uint32() {return key & 0x7;} + }; + + union OMDepthStencilSelector + { + struct + { + uint32 zte:1; + uint32 ztst:2; + uint32 zwe:1; + uint32 date:1; + }; + + uint32 key; + + operator uint32() {return key & 0x1f;} + }; + + union OMBlendSelector + { + struct + { + uint32 abe:1; + uint32 a:2; + uint32 b:2; + uint32 c:2; + uint32 d:2; + uint32 wr:1; + uint32 wg:1; + uint32 wb:1; + uint32 wa:1; + }; + + uint32 key; + + operator uint32() {return key & 0x1fff;} + }; + + #pragma pack(pop) + +private: + GSDevice11* m_dev; + CComPtr m_il; + hash_map > m_vs; + CComPtr m_vs_cb; + hash_map > m_gs; + hash_map > m_ps; + CComPtr m_ps_cb; + hash_map > m_ps_ss; + CComPtr m_palette_ss; + hash_map > m_om_dss; + hash_map > m_om_bs; + + CComPtr m_vb, m_vb_old; + int m_vb_max; + int m_vb_start; + int m_vb_count; + + VSConstantBuffer m_vs_cb_cache; + PSConstantBuffer m_ps_cb_cache; + +public: + GSTextureFX11(); + + bool Create(GSDevice11* dev); + + bool SetupIA(const GSVertexHW11* vertices, int count, D3D11_PRIMITIVE_TOPOLOGY prim); + bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); + bool SetupGS(GSSelector sel); + bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); + void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); + void SetupRS(int w, int h, const GSVector4i& scissor); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); + void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); + void Draw(); +}; From 69e3dcfa63fbd254d491f4d9513da97d9007a45a Mon Sep 17 00:00:00 2001 From: gabest11 Date: Fri, 19 Jun 2009 20:43:08 +0000 Subject: [PATCH 260/296] GSdx: everything should be back to normal now git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1406 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTextureFX11.cpp | 32 +- plugins/GSdx/GSTextureFX9.cpp | 26 +- plugins/GSdx/GSdx_vs2008.vcproj | 3240 ++++++++++++++------------- plugins/GSdx/vsprops/common.vsprops | 4 +- 4 files changed, 1719 insertions(+), 1583 deletions(-) diff --git a/plugins/GSdx/GSTextureFX11.cpp b/plugins/GSdx/GSTextureFX11.cpp index d0efbc7ee3..4fe3de6a28 100644 --- a/plugins/GSdx/GSTextureFX11.cpp +++ b/plugins/GSdx/GSTextureFX11.cpp @@ -180,8 +180,8 @@ bool GSTextureFX11::SetupGS(GSSelector sel) D3D11_SHADER_MACRO macro[] = { - {"IIP", str[0].c_str()}, - {"PRIM", str[1].c_str()}, + {"GS_IIP", str[0].c_str()}, + {"GS_PRIM", str[1].c_str()}, {NULL, NULL}, }; @@ -232,20 +232,20 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl D3D11_SHADER_MACRO macro[] = { - {"FST", str[0].c_str()}, - {"WMS", str[1].c_str()}, - {"WMT", str[2].c_str()}, - {"BPP", str[3].c_str()}, - {"AEM", str[4].c_str()}, - {"TFX", str[5].c_str()}, - {"TCC", str[6].c_str()}, - {"ATE", str[7].c_str()}, - {"ATST", str[8].c_str()}, - {"FOG", str[9].c_str()}, - {"CLR1", str[10].c_str()}, - {"FBA", str[11].c_str()}, - {"AOUT", str[12].c_str()}, - {"LTF", str[13].c_str()}, + {"PS_FST", str[0].c_str()}, + {"PS_WMS", str[1].c_str()}, + {"PS_WMT", str[2].c_str()}, + {"PS_BPP", str[3].c_str()}, + {"PS_AEM", str[4].c_str()}, + {"PS_TFX", str[5].c_str()}, + {"PS_TCC", str[6].c_str()}, + {"PS_ATE", str[7].c_str()}, + {"PS_ATST", str[8].c_str()}, + {"PS_FOG", str[9].c_str()}, + {"PS_CLR1", str[10].c_str()}, + {"PS_FBA", str[11].c_str()}, + {"PS_AOUT", str[12].c_str()}, + {"PS_LTF", str[13].c_str()}, {NULL, NULL}, }; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 24694a33ca..ac64eed885 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -197,19 +197,19 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample D3DXMACRO macro[] = { - {"FST", str[0].c_str()}, - {"WMS", str[1].c_str()}, - {"WMT", str[2].c_str()}, - {"BPP", str[3].c_str()}, - {"AEM", str[4].c_str()}, - {"TFX", str[5].c_str()}, - {"TCC", str[6].c_str()}, - {"ATE", str[7].c_str()}, - {"ATST", str[8].c_str()}, - {"FOG", str[9].c_str()}, - {"CLR1", str[10].c_str()}, - {"RT", str[11].c_str()}, - {"LTF", str[12].c_str()}, + {"PS_FST", str[0].c_str()}, + {"PS_WMS", str[1].c_str()}, + {"PS_WMT", str[2].c_str()}, + {"PS_BPP", str[3].c_str()}, + {"PS_AEM", str[4].c_str()}, + {"PS_TFX", str[5].c_str()}, + {"PS_TCC", str[6].c_str()}, + {"PS_ATE", str[7].c_str()}, + {"PS_ATST", str[8].c_str()}, + {"PS_FOG", str[9].c_str()}, + {"PS_CLR1", str[10].c_str()}, + {"PS_RT", str[11].c_str()}, + {"PS_LTF", str[12].c_str()}, {NULL, NULL}, }; diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 8b5f6a7d37..aa3a060e89 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -82,6 +82,69 @@ Name="VCPostBuildEventTool" /> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - @@ -1016,6 +1128,10 @@ RelativePath=".\GSDevice10.cpp" > + + @@ -1140,7 +1256,7 @@ /> + + @@ -1268,6 +1388,10 @@ RelativePath=".\GSTexture10.cpp" > + + @@ -1292,6 +1416,10 @@ RelativePath=".\GSTextureCache10.cpp" > + + @@ -1324,6 +1452,10 @@ /> + + @@ -1379,6 +1511,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1387,6 +1527,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1395,6 +1543,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1403,6 +1559,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1411,6 +1575,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1419,6 +1591,14 @@ UsePrecompiledHeader="1" /> + + + @@ -1435,54 +1615,6 @@ UsePrecompiledHeader="1" /> - - - - - - - - - - - - - - - - - - + + @@ -1654,6 +1790,10 @@ RelativePath=".\GSRendererHW10.h" > + + @@ -1702,6 +1842,10 @@ RelativePath=".\GSTexture10.h" > + + @@ -1718,6 +1862,10 @@ RelativePath=".\GSTextureCache10.h" > + + @@ -1734,6 +1882,10 @@ RelativePath=".\GSTextureFX10.h" > + + @@ -1817,35 +1969,19 @@ Name="Shaders" > - - - - - - - - @@ -1871,6 +2007,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1879,6 +2023,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1887,6 +2039,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1895,6 +2055,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1903,6 +2071,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1911,6 +2087,14 @@ UsePrecompiledHeader="0" /> + + + @@ -1927,54 +2111,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -1999,6 +2143,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2007,6 +2159,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2015,6 +2175,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2023,6 +2191,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2031,6 +2207,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2047,54 +2231,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2119,6 +2263,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2127,6 +2279,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2135,6 +2295,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2143,6 +2311,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2151,6 +2327,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2167,54 +2351,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2239,6 +2383,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2247,6 +2399,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2255,6 +2415,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2263,6 +2431,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2271,6 +2447,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2287,54 +2471,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2359,6 +2503,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2367,6 +2519,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2375,6 +2535,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2383,6 +2551,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2391,6 +2567,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2407,54 +2591,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2479,6 +2623,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2487,6 +2639,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2495,6 +2655,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2503,6 +2671,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2511,6 +2687,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2527,54 +2711,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2615,6 +2759,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2623,6 +2775,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2631,6 +2791,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2639,6 +2807,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2647,6 +2823,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2663,54 +2847,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2735,6 +2879,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2743,6 +2895,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2751,6 +2911,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2759,6 +2927,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2767,6 +2943,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2783,54 +2967,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2855,6 +2999,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2863,6 +3015,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2871,6 +3031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2879,6 +3047,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2887,6 +3063,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2903,54 +3087,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -2975,6 +3119,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2983,6 +3135,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2991,6 +3151,14 @@ UsePrecompiledHeader="0" /> + + + @@ -2999,6 +3167,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3007,6 +3183,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3023,54 +3207,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3095,6 +3239,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3103,6 +3255,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3111,6 +3271,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3119,6 +3287,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3127,6 +3303,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3143,54 +3327,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3219,6 +3363,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3227,6 +3379,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3235,6 +3395,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3243,6 +3411,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3251,6 +3427,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3267,54 +3451,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3339,6 +3483,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3347,6 +3499,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3355,6 +3515,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3363,6 +3531,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3371,6 +3547,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3387,54 +3571,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3459,6 +3603,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3467,6 +3619,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3475,6 +3635,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3483,6 +3651,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3491,6 +3667,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3507,54 +3691,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3579,6 +3723,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3587,6 +3739,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3595,6 +3755,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3603,6 +3771,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3611,6 +3787,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3627,54 +3811,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3703,6 +3847,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3711,6 +3863,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3719,6 +3879,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3727,6 +3895,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3735,6 +3911,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3751,54 +3935,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3823,6 +3967,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3831,6 +3983,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3839,6 +3999,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3847,6 +4015,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3855,6 +4031,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3871,54 +4055,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -3943,6 +4087,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3951,6 +4103,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3959,6 +4119,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3967,6 +4135,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3975,6 +4151,14 @@ UsePrecompiledHeader="0" /> + + + @@ -3991,54 +4175,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -4063,6 +4207,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4071,6 +4223,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4079,6 +4239,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4087,6 +4255,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4095,6 +4271,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4111,54 +4295,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -4183,6 +4327,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4191,6 +4343,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4199,6 +4359,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4207,6 +4375,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4215,6 +4391,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4231,54 +4415,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -4303,6 +4447,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4311,6 +4463,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4319,6 +4479,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4327,6 +4495,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4335,6 +4511,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4351,54 +4535,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -4423,6 +4567,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4431,6 +4583,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4439,6 +4599,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4447,6 +4615,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4455,6 +4631,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4471,54 +4655,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - + + + @@ -4543,6 +4687,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4551,6 +4703,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4559,6 +4719,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4567,6 +4735,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4575,6 +4751,14 @@ UsePrecompiledHeader="0" /> + + + @@ -4591,54 +4775,6 @@ UsePrecompiledHeader="0" /> - - - - - - - - - - - - - - - - - - Date: Sat, 20 Jun 2009 03:45:44 +0000 Subject: [PATCH 261/296] microVU: minor changes git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1407 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 268 ++++------------------------------ pcsx2/x86/microVU.h | 14 +- pcsx2/x86/microVU_Compile.inl | 186 +++++++++++++++++++++++ pcsx2/x86/microVU_Log.inl | 2 +- pcsx2/x86/microVU_Misc.h | 7 +- 5 files changed, 223 insertions(+), 254 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index b56e266f1b..d6198eebee 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -35,15 +35,16 @@ declareAllVariables // Declares All Global Variables :D //------------------------------------------------------------------ // Only run this once per VU! ;) -microVUf(void) mVUinit(VURegs* vuRegsPtr) { +microVUt(void) mVUinit(VURegs* vuRegsPtr, int vuIndex) { - microVU* mVU = mVUx; - mVU->regs = vuRegsPtr; - mVU->index = vuIndex; - mVU->microMemSize = (vuIndex ? 0x4000 : 0x1000); - mVU->progMemSize = (vuIndex ? 0x4000 : 0x1000) / 4; - mVU->cache = NULL; - mVU->cacheSize = mVUcacheSize; + microVU* mVU = mVUx; + mVU->regs = vuRegsPtr; + mVU->index = vuIndex; + mVU->vuMemSize = (vuIndex ? 0x4000 : 0x1000); + mVU->microMemSize = (vuIndex ? 0x4000 : 0x1000); + mVU->progSize = (vuIndex ? 0x4000 : 0x1000) / 4; + mVU->cache = NULL; + mVU->cacheSize = mVUcacheSize; memset(&mVU->prog, 0, sizeof(mVU->prog)); mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); @@ -51,18 +52,17 @@ microVUf(void) mVUinit(VURegs* vuRegsPtr) { if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", (u32)mVU->cache)); mVUemitSearch(); - mVUreset(); + mVUreset(mVU); } // Resets Rec Data -microVUx(void) mVUreset() { +microVUt(void) mVUreset(mV) { - microVU* mVU = mVUx; - mVUprint((vuIndex) ? "microVU1: reset" : "microVU0: reset"); + mVUprint((mVU->index) ? "microVU1: reset" : "microVU0: reset"); // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progMemSize / 2); j++) { + for (u32 j = 0; j < (mVU->progSize / 2); j++) { microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } } @@ -99,16 +99,15 @@ microVUx(void) mVUreset() { } // Free Allocated Resources -microVUf(void) mVUclose() { +microVUt(void) mVUclose(mV) { - microVU* mVU = mVUx; - mVUprint((vuIndex) ? "microVU1: close" : "microVU0: close"); + mVUprint((mVU->index) ? "microVU1: close" : "microVU0: close"); - if ( mVU->cache ) { HostSys::Munmap( mVU->cache, mVU->cacheSize ); mVU->cache = NULL; } + if (mVU->cache) { HostSys::Munmap(mVU->cache, mVU->cacheSize); mVU->cache = NULL; } // Delete Block Managers for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progMemSize / 2); j++) { + for (u32 j = 0; j < (mVU->progSize / 2); j++) { if (mVU->prog.prog[i].block[j]) { microBlockManager::Delete( mVU->prog.prog[i].block[j] ); } @@ -117,8 +116,7 @@ microVUf(void) mVUclose() { } // Clears Block Data in specified range -microVUf(void) mVUclear(u32 addr, u32 size) { - microVU* mVU = mVUx; +microVUt(void) mVUclear(mV, u32 addr, u32 size) { if (!mVU->prog.cleared) { memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); mVU->prog.cleared = 1; // Next execution searches/creates a new microprogram @@ -137,7 +135,7 @@ microVUf(void) mVUclearProg(int progIndex) { mVU->prog.prog[progIndex].range[0] = -1; mVU->prog.prog[progIndex].range[1] = -1; mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; - for (u32 i = 0; i < (mVU->progMemSize / 2); i++) { + for (u32 i = 0; i < (mVU->progSize / 2); i++) { if (mVU->prog.prog[progIndex].block[i]) mVU->prog.prog[progIndex].block[i]->reset(); } @@ -191,9 +189,8 @@ microVUf(int) mVUfindLeastUsedProg() { // frame-based decrementing system in combination with a program-execution-based incrementing // system. In english: if last_used >= 2 it means the program has been used for the current // or prev frame. if it's 0, the program hasn't been used for a while. -microVUf(void) mVUvsyncUpdate() { +microVUt(void) mVUvsyncUpdate(mV) { - microVU* mVU = mVUx; if (mVU->prog.total < mVU->prog.max) return; for (int i = 0; i <= mVU->prog.total; i++) { @@ -252,232 +249,17 @@ microVUf(int) mVUsearchProg() { return 1; // If !cleared, then we're still on the same program as last-time ;) } -//------------------------------------------------------------------ -// JIT recompiler -- called from recompiled code when register jumps are made. -//------------------------------------------------------------------ - -microVUf(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState) { - - return mVUblockFetch( mVUx, startPC, pState ); -} - -//------------------------------------------------------------------ -// Recompiler -//------------------------------------------------------------------ - -static void* __fastcall mVUcompile( microVU* mVU, u32 startPC, uptr pState ) -{ - using namespace x86Emitter; - - // Setup Program Bounds/Range - mVUsetupRange(mVU, startPC); - - microBlock* pBlock = NULL; - u8* thisPtr = x86Ptr; - - const u32 microSizeDiv8 = (mVU->microMemSize-1) / 8; - - // First Pass - iPC = startPC / 4; - setCode(); - mVUbranch = 0; - mVUstartPC = iPC; - mVUcount = 0; - mVUcycles = 0; // Skips "M" phase, and starts counting cycles at "T" stage - mVU->p = 0; // All blocks start at p index #0 - mVU->q = 0; // All blocks start at q index #0 - memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info - mVUblock.x86ptrStart = thisPtr; - pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager - mVUpBlock = pBlock; - mVUregs.flags = 0; - mVUflagInfo = 0; - mVUsFlagHack = CHECK_VU_FLAGHACK; - - for (int branch = 0; mVUcount < microSizeDiv8; ) { - incPC(1); - startLoop(); - mVUincCycles(mVU, 1); - mVUopU(mVU, 0); - if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } - if (curI & _DTbit_) { branch = 4; } - if (curI & _Mbit_) { mVUup.mBit = 1; } - if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } - else { incPC(-1); mVUopL(mVU, 0); incPC(1); } - mVUsetCycles(mVU); - mVUinfo.readQ = mVU->q; - mVUinfo.writeQ = !mVU->q; - mVUinfo.readP = mVU->p; - mVUinfo.writeP = !mVU->p; - if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } - else if (branch == 1) { branch = 2; } - if (mVUbranch) { mVUsetFlagInfo(mVU); branch = 3; mVUbranch = 0; } - incPC(1); - mVUcount++; - } - - // Sets Up Flag instances - int xStatus[4], xMac[4], xClip[4]; - int xCycles = mVUsetFlags(mVU, xStatus, xMac, xClip); - mVUtestCycles(mVU); - - // Second Pass - iPC = mVUstartPC; - setCode(); - mVUbranch = 0; - uint x; - for (x = 0; x < microSizeDiv8; x++) { - if (mVUinfo.isEOB) { x = 0xffff; } - if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } - if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } - else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } - else { doSwapOp(); } - if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } - - if (!mVUinfo.isBdelay) { incPC(1); } - else { - microBlock* bBlock = NULL; - s32* ajmp = 0; - mVUsetupRange(mVU, xPC); - mVUdebugNOW(1); - - switch (mVUbranch) { - case 3: branchCase(Jcc_Equal); // IBEQ - case 4: branchCase(Jcc_GreaterOrEqual); // IBGEZ - case 5: branchCase(Jcc_Greater); // IBGTZ - case 6: branchCase(Jcc_LessOrEqual); // IBLEQ - case 7: branchCase(Jcc_Less); // IBLTZ - case 8: branchCase(Jcc_NotEqual); // IBNEQ - case 1: case 2: // B/BAL - - mVUprint("mVUcompile B/BAL"); - incPC(-3); // Go back to branch opcode (to get branch imm addr) - - if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - - // Check if branch-block has already been compiled - pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); - if (pBlock) { xJMP(pBlock->x86ptrStart); } - else { mVUcompile(mVU, branchAddr, (uptr)&mVUregs); } - return thisPtr; - case 9: case 10: // JR/JALR - - mVUprint("mVUcompile JR/JALR"); - incPC(-3); // Go back to jump opcode - - if (mVUup.eBit) { // E-bit Jump - mVUendProgram(mVU, 2, xStatus, xMac, xClip); - MOV32MtoR(gprT1, (uptr)&mVU->branch); - MOV32RtoM((uptr)&mVU->regs->VI[REG_TPC].UL, gprT1); - xJMP(mVU->exitFunct); - return thisPtr; - } - - memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - - mVUbackupRegs(mVU); - MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) - MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) - - if (!mVU->index) xCALL( mVUcompileJIT<0> ); //(u32 startPC, uptr pState) - else xCALL( mVUcompileJIT<1> ); - mVUrestoreRegs(mVU); - JMPR(gprT1); // Jump to rec-code address - return thisPtr; - } - // Conditional Branches - mVUprint("mVUcompile conditional branch"); - if (bBlock) { // Branch non-taken has already been compiled - incPC(-3); // Go back to branch opcode (to get branch imm addr) - - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - - // Check if branch-block has already been compiled - pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); - if (pBlock) { xJMP( pBlock->x86ptrStart ); } - else if (!mVU->index) { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } - else { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } - } - else { - uptr jumpAddr; - u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs so back them up - memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); - - incPC2(1); // Get PC for branch not-taken - mVUcompile(mVU, xPC, (uptr)&mVUregs); - - iPC = bPC; - incPC(-3); // Go back to branch opcode (to get branch imm addr) - if (!mVU->index) jumpAddr = (uptr)mVUblockFetch(mVU, branchAddr, (uptr)&pBlock->pStateEnd); - else jumpAddr = (uptr)mVUblockFetch(mVU, branchAddr, (uptr)&pBlock->pStateEnd); - *ajmp = (jumpAddr - ((uptr)ajmp + 4)); - } - return thisPtr; - } - } - if (x == microSizeDiv8) { Console::Error("microVU%d: Possible infinite compiling loop!", params mVU->index); } - - // E-bit End - mVUendProgram(mVU, 1, xStatus, xMac, xClip); - return thisPtr; -} - - -microVUt(void*) mVUblockFetch( microVU* mVU, u32 startPC, uptr pState ) -{ - using namespace x86Emitter; - - if (startPC > mVU->microMemSize-8) { Console::Error("microVU%d: invalid startPC", params mVU->index); } - startPC &= mVU->microMemSize-8; - - if (mVUblocks[startPC/8] == NULL) { - mVUblocks[startPC/8] = microBlockManager::AlignedNew(); - } - - // Searches for Existing Compiled Block (if found, then returns; else, compile) - microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); - if (pBlock) { return pBlock->x86ptrStart; } - - return mVUcompile( mVU, startPC, pState ); -} - - //------------------------------------------------------------------ // Wrapper Functions - Called by other parts of the Emu //------------------------------------------------------------------ -void initVUrec(VURegs* vuRegs, const int vuIndex) { - if (!vuIndex) mVUinit<0>(vuRegs); - else mVUinit<1>(vuRegs); -} - -void closeVUrec(const int vuIndex) { - if (!vuIndex) mVUclose<0>(); - else mVUclose<1>(); -} - -void resetVUrec(const int vuIndex) { - if (!vuIndex) mVUreset<0>(); - else mVUreset<1>(); -} - -void clearVUrec(u32 addr, u32 size, const int vuIndex) { - if (!vuIndex) mVUclear<0>(addr, size); - else mVUclear<1>(addr, size); -} +void initVUrec (VURegs* vuRegs, const int vuIndex) { mVUinit(vuRegs, vuIndex); } +void closeVUrec(const int vuIndex) { mVUclose(mVUx); } +void resetVUrec(const int vuIndex) { mVUreset(mVUx); } +void vsyncVUrec(const int vuIndex) { mVUvsyncUpdate(mVUx); } +void clearVUrec(u32 addr, u32 size, const int vuIndex) { mVUclear(mVUx, addr, size); } void runVUrec(u32 startPC, u32 cycles, const int vuIndex) { if (!vuIndex) startVU0(startPC, cycles); else startVU1(startPC, cycles); } - -void vsyncVUrec(const int vuIndex) { - if (!vuIndex) mVUvsyncUpdate<0>(); - else mVUvsyncUpdate<1>(); -} diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 0d0cdad2f3..96fd3349f7 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -117,8 +117,9 @@ struct microVU { PCSX2_ALIGNED16(u32 xmmVFb[4]); // Backup for VF regs u32 index; // VU Index (VU0 or VU1) + u32 vuMemSize; // VU Main Memory Size (in bytes) u32 microMemSize; // VU Micro Memory Size (in bytes) - u32 progMemSize; // VU Micro Program Size (microSize/4, because each instruction of a program is 4 bytes) + u32 progSize; // VU Micro Memory Size (in u32's) u32 cacheSize; // VU Cache Size microProgManager<0x4000/4> prog; // Micro Program Data @@ -151,11 +152,12 @@ extern void (*mVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); extern int mVUdebugNow; // Main Functions -microVUf(void) mVUinit(VURegs*); -microVUx(void) mVUreset(); -microVUf(void) mVUclose(); -microVUf(void) mVUclear(u32, u32); -microVUt(void*) mVUblockFetch( microVU* mVU, u32 startPC, uptr pState ); +microVUt(void) mVUinit(VURegs*, int); +microVUt(void) mVUreset(mV); +microVUt(void) mVUclose(mV); +microVUt(void) mVUclear(mV, u32, u32); +microVUt(void*) mVUblockFetch(microVU* mVU, u32 startPC, uptr pState); +microVUx(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState); // Prototypes for Linux void __fastcall mVUcleanUpVU0(); diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 01a2095f7b..78e3a5ca0a 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -282,3 +282,189 @@ microVUt(void) mVUtestCycles(mV) { mVUendProgram(mVU, 0, NULL, NULL, NULL); x86SetJ32(jmp32); } + +//------------------------------------------------------------------ +// Recompiler +//------------------------------------------------------------------ + +static void* __fastcall mVUcompile(microVU* mVU, u32 startPC, uptr pState) { + + using namespace x86Emitter; + microBlock* pBlock = NULL; + u8* thisPtr = x86Ptr; + const u32 endCount = (mVU->microMemSize / 8) - 1; + + // Setup Program Bounds/Range + mVUsetupRange(mVU, startPC); + + // First Pass + iPC = startPC / 4; + setCode(); + mVUbranch = 0; + mVUstartPC = iPC; + mVUcount = 0; + mVUcycles = 0; // Skips "M" phase, and starts counting cycles at "T" stage + mVU->p = 0; // All blocks start at p index #0 + mVU->q = 0; // All blocks start at q index #0 + memcpy_fast(&mVUregs, (microRegInfo*)pState, sizeof(microRegInfo)); // Loads up Pipeline State Info + mVUblock.x86ptrStart = thisPtr; + pBlock = mVUblocks[startPC/8]->add(&mVUblock); // Add this block to block manager + mVUpBlock = pBlock; + mVUregs.flags = 0; + mVUflagInfo = 0; + mVUsFlagHack = CHECK_VU_FLAGHACK; + + for (int branch = 0; mVUcount < endCount; mVUcount++) { + incPC(1); + startLoop(); + mVUincCycles(mVU, 1); + mVUopU(mVU, 0); + if (curI & _Ebit_) { branch = 1; mVUup.eBit = 1; } + if (curI & _DTbit_) { branch = 4; } + if (curI & _Mbit_) { mVUup.mBit = 1; } + if (curI & _Ibit_) { mVUlow.isNOP = 1; mVUup.iBit = 1; } + else { incPC(-1); mVUopL(mVU, 0); incPC(1); } + mVUsetCycles(mVU); + mVUinfo.readQ = mVU->q; + mVUinfo.writeQ = !mVU->q; + mVUinfo.readP = mVU->p; + mVUinfo.writeP = !mVU->p; + if (branch >= 2) { mVUinfo.isEOB = 1; if (branch == 3) { mVUinfo.isBdelay = 1; } mVUcount++; branchWarning(); break; } + else if (branch == 1) { branch = 2; } + if (mVUbranch) { mVUsetFlagInfo(mVU); branch = 3; mVUbranch = 0; } + incPC(1); + } + + // Sets Up Flag instances + int xStatus[4], xMac[4], xClip[4]; + int xCycles = mVUsetFlags(mVU, xStatus, xMac, xClip); + mVUtestCycles(mVU); + + // Second Pass + iPC = mVUstartPC; + setCode(); + mVUbranch = 0; + uint x; + for (x = 0; x < endCount; x++) { + if (mVUinfo.isEOB) { x = 0xffff; } + if (mVUup.mBit) { OR32ItoM((uptr)&mVU->regs->flags, VUFLAG_MFLAGSET); } + if (mVUlow.isNOP) { incPC(1); doUpperOp(); doIbit(); } + else if (!mVUinfo.swapOps) { incPC(1); doUpperOp(); doLowerOp(); } + else { doSwapOp(); } + if (mVUinfo.doXGKICK) { mVU_XGKICK_DELAY(mVU, 1); } + + if (!mVUinfo.isBdelay) { incPC(1); } + else { + microBlock* bBlock = NULL; + s32* ajmp = 0; + mVUsetupRange(mVU, xPC); + mVUdebugNOW(1); + + switch (mVUbranch) { + case 3: branchCase(Jcc_Equal); // IBEQ + case 4: branchCase(Jcc_GreaterOrEqual); // IBGEZ + case 5: branchCase(Jcc_Greater); // IBGTZ + case 6: branchCase(Jcc_LessOrEqual); // IBLEQ + case 7: branchCase(Jcc_Less); // IBLTZ + case 8: branchCase(Jcc_NotEqual); // IBNEQ + case 1: case 2: // B/BAL + + mVUprint("mVUcompile B/BAL"); + incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); + + if (mVUblocks[branchAddr/8] == NULL) + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); + + // Check if branch-block has already been compiled + pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); + if (pBlock) { xJMP(pBlock->x86ptrStart); } + else { mVUcompile(mVU, branchAddr, (uptr)&mVUregs); } + return thisPtr; + case 9: case 10: // JR/JALR + + mVUprint("mVUcompile JR/JALR"); + incPC(-3); // Go back to jump opcode + + if (mVUup.eBit) { // E-bit Jump + mVUendProgram(mVU, 2, xStatus, xMac, xClip); + MOV32MtoR(gprT1, (uptr)&mVU->branch); + MOV32RtoM((uptr)&mVU->regs->VI[REG_TPC].UL, gprT1); + xJMP(mVU->exitFunct); + return thisPtr; + } + + memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); + + mVUbackupRegs(mVU); + MOV32MtoR(gprT2, (uptr)&mVU->branch); // Get startPC (ECX first argument for __fastcall) + MOV32ItoR(gprR, (u32)&pBlock->pStateEnd); // Get pState (EDX second argument for __fastcall) + + if (!mVU->index) xCALL(mVUcompileJIT<0>); //(u32 startPC, uptr pState) + else xCALL(mVUcompileJIT<1>); + mVUrestoreRegs(mVU); + JMPR(gprT1); // Jump to rec-code address + return thisPtr; + } + // Conditional Branches + mVUprint("mVUcompile conditional branch"); + if (bBlock) { // Branch non-taken has already been compiled + incPC(-3); // Go back to branch opcode (to get branch imm addr) + + if (mVUblocks[branchAddr/8] == NULL) + mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); + + // Check if branch-block has already been compiled + pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); + if (pBlock) { xJMP( pBlock->x86ptrStart ); } + else { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } + } + else { + uptr jumpAddr; + u32 bPC = iPC; // mVUcompile can modify iPC and mVUregs so back them up + memcpy_fast(&pBlock->pStateEnd, &mVUregs, sizeof(microRegInfo)); + + incPC2(1); // Get PC for branch not-taken + mVUcompile(mVU, xPC, (uptr)&mVUregs); + + iPC = bPC; + incPC(-3); // Go back to branch opcode (to get branch imm addr) + jumpAddr = (uptr)mVUblockFetch(mVU, branchAddr, (uptr)&pBlock->pStateEnd); + *ajmp = (jumpAddr - ((uptr)ajmp + 4)); + } + return thisPtr; + } + } + if (x == endCount) { Console::Error("microVU%d: Possible infinite compiling loop!", params mVU->index); } + + // E-bit End + mVUendProgram(mVU, 1, xStatus, xMac, xClip); + return thisPtr; +} + + +microVUt(void*) mVUblockFetch(microVU* mVU, u32 startPC, uptr pState) { + + using namespace x86Emitter; + + if (startPC > mVU->microMemSize-8) { DevCon::Error("microVU%d: invalid startPC", params mVU->index); } + startPC &= mVU->microMemSize-8; + + if (mVUblocks[startPC/8] == NULL) { + mVUblocks[startPC/8] = microBlockManager::AlignedNew(); + } + + // Searches for Existing Compiled Block (if found, then returns; else, compile) + microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); + if (pBlock) { return pBlock->x86ptrStart; } + else { return mVUcompile(mVU, startPC, pState); } +} + +// Called By JR/JALR +microVUx(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState) { + return mVUblockFetch(mVUx, startPC, pState); +} + diff --git a/pcsx2/x86/microVU_Log.inl b/pcsx2/x86/microVU_Log.inl index a721642b01..4b84ab61cc 100644 --- a/pcsx2/x86/microVU_Log.inl +++ b/pcsx2/x86/microVU_Log.inl @@ -62,7 +62,7 @@ microVUx(void) __mVUdumpProgram(int progIndex) { mVUlog("*********************\n\n

", progIndex); mVUlog(""); - for (u32 i = 0; i < mVU->progMemSize; i+=2) { + for (u32 i = 0; i < mVU->progSize; i+=2) { if (delay) { delay--; mVUlog(""); if (!delay) mVUlog("
"); } if (mVUbranch) { delay = 1; mVUbranch = 0; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 32089b006b..b91f9e74a9 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -201,8 +201,8 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define xPC ((iPC / 2) * 8) #define curI ((u32*)mVU->regs->Micro)[iPC] //mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } -#define incPC(x) { iPC = ((iPC + x) & (mVU->progMemSize-1)); setCode(); } -#define incPC2(x) { iPC = ((iPC + x) & (mVU->progMemSize-1)); } +#define incPC(x) { iPC = ((iPC + x) & (mVU->progSize-1)); setCode(); } +#define incPC2(x) { iPC = ((iPC + x) & (mVU->progSize-1)); } #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microMemSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) @@ -274,8 +274,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); uptr diff = ptr - start; \ if (diff >= limit) { \ Console::Error("microVU Error: Program went over its cache limit. Size = 0x%x", params diff); \ - if (!isVU1) mVUreset<0>(); \ - else mVUreset<1>(); \ + mVUreset(mVU); \ } \ } From 40ae3c6ecd1db8ea74978c22ab9c868fd27a1b16 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sat, 20 Jun 2009 04:33:55 +0000 Subject: [PATCH 262/296] microVU: more minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1408 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 2 +- pcsx2/x86/microVU_Compile.inl | 75 ++++++++++++++++------------------- 2 files changed, 36 insertions(+), 41 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 96fd3349f7..d7256904c3 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -64,7 +64,7 @@ public: } return thisBlock; } - __releaseinline microBlock* search(microRegInfo* pState) { + __forceinline microBlock* search(microRegInfo* pState) { if (listSize < 0) return NULL; if (pState->needExactMatch) { // Needs Detailed Search (Exact Match of Pipeline State) for (int i = 0; i <= listSize; i++) { diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 78e3a5ca0a..aa0f9ab271 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -22,31 +22,7 @@ // Helper Macros //------------------------------------------------------------------ -#define branchCase(JMPcc) \ - mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); \ - xCMP( ptr16[&mVU->branch], 0); \ - if (mVUup.eBit) { /* Conditional Branch With E-Bit Set */ \ - mVUendProgram(mVU, 2, xStatus, xMac, xClip); \ - xForwardJump8 eJMP( JMPcc ); \ - incPC(1); /* Set PC to First instruction of Non-Taken Side */ \ - xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ - xJMP( mVU->exitFunct ); \ - eJMP.SetTarget(); \ - incPC(-4); /* Go Back to Branch Opcode to get branchAddr */ \ - iPC = branchAddr/4; \ - xMOV( ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); \ - xJMP( mVU->exitFunct ); \ - return thisPtr; \ - } \ - else { /* Normal Conditional Branch */ \ - incPC2(1); /* Check if Branch Non-Taken Side has already been recompiled */ \ - if (!mVUblocks[iPC/2]) { mVUblocks[iPC/2] = microBlockManager::AlignedNew(); } \ - bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); \ - incPC2(-1); \ - if (bBlock) { xJcc( xInvertCond( JMPcc ), bBlock->x86ptrStart ); } \ - else { ajmp = xJcc32( JMPcc ); } \ - } \ - break +#define branchCase(JMPcond) branchCaseFunct(mVU, bBlock, xStatus, xMac, xClip, xCycles, ajmp, JMPcond); break #define branchWarning() { \ if (mVUbranch) { \ @@ -93,6 +69,7 @@ #define doLowerOp() { incPC(-1); mVUopL(mVU, 1); incPC(1); } #define doSwapOp() { doBackupVF1(); mVUopL(mVU, 1); doBackupVF2(); incPC(1); doUpperOp(); doBackupVF3(); } #define doIbit() { if (mVUup.iBit) { incPC(-1); MOV32ItoM((uptr)&mVU->regs->VI[REG_I].UL, curI); incPC(1); } } +#define blockCreate(addr) { if (!mVUblocks[addr]) mVUblocks[addr] = microBlockManager::AlignedNew(); } //------------------------------------------------------------------ // Helper Functions @@ -265,6 +242,32 @@ microVUt(void) mVUendProgram(mV, int isEbit, int* xStatus, int* xMac, int* xClip } } +void branchCaseFunct(mV, microBlock* &bBlock, int* xStatus, int* xMac, int* xClip, int &xCycles, s32* &ajmp, int JMPcc) { + using namespace x86Emitter; + mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); + xCMP(ptr16[&mVU->branch], 0); + if (mVUup.eBit) { // Conditional Branch With E-Bit Set + mVUendProgram(mVU, 2, xStatus, xMac, xClip); + xForwardJump8 eJMP((JccComparisonType)JMPcc); + incPC(1); // Set PC to First instruction of Non-Taken Side + xMOV(ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); + xJMP(mVU->exitFunct); + eJMP.SetTarget(); + incPC(-4); // Go Back to Branch Opcode to get branchAddr + iPC = branchAddr/4; + xMOV(ptr32[&mVU->regs->VI[REG_TPC].UL], xPC); + xJMP(mVU->exitFunct); + } + else { // Normal Conditional Branch + incPC2(1); // Check if Branch Non-Taken Side has already been recompiled + blockCreate(iPC/2); + bBlock = mVUblocks[iPC/2]->search((microRegInfo*)&mVUregs); + incPC2(-1); + if (bBlock) { xJcc( xInvertCond((JccComparisonType)JMPcc), bBlock->x86ptrStart ); } + else { ajmp = xJcc32((JccComparisonType)JMPcc); } + } +} + void __fastcall mVUwarning0(u32 PC) { Console::Error("microVU0 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUwarning1(u32 PC) { Console::Error("microVU1 Warning: Exiting from Possible Infinite Loop [%04x]", params PC); } void __fastcall mVUprintPC1(u32 PC) { Console::Write("Block PC [%04x] ", params PC); } @@ -287,7 +290,7 @@ microVUt(void) mVUtestCycles(mV) { // Recompiler //------------------------------------------------------------------ -static void* __fastcall mVUcompile(microVU* mVU, u32 startPC, uptr pState) { +microVUt(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { using namespace x86Emitter; microBlock* pBlock = NULL; @@ -375,10 +378,8 @@ static void* __fastcall mVUcompile(microVU* mVU, u32 startPC, uptr pState) { if (mVUup.eBit) { iPC = branchAddr/4; mVUendProgram(mVU, 1, xStatus, xMac, xClip); } // E-bit Branch mVUsetupBranch(mVU, xStatus, xMac, xClip, xCycles); - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - // Check if branch-block has already been compiled + blockCreate(branchAddr/8); pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); if (pBlock) { xJMP(pBlock->x86ptrStart); } else { mVUcompile(mVU, branchAddr, (uptr)&mVUregs); } @@ -411,13 +412,12 @@ static void* __fastcall mVUcompile(microVU* mVU, u32 startPC, uptr pState) { } // Conditional Branches mVUprint("mVUcompile conditional branch"); + if (mVUup.eBit) return thisPtr; // Handled in Branch Case if (bBlock) { // Branch non-taken has already been compiled incPC(-3); // Go back to branch opcode (to get branch imm addr) - if (mVUblocks[branchAddr/8] == NULL) - mVUblocks[branchAddr/8] = microBlockManager::AlignedNew(); - // Check if branch-block has already been compiled + blockCreate(branchAddr/8); pBlock = mVUblocks[branchAddr/8]->search((microRegInfo*)&mVUregs); if (pBlock) { xJMP( pBlock->x86ptrStart ); } else { mVUblockFetch(mVU, branchAddr, (uptr)&mVUregs); } @@ -445,25 +445,20 @@ static void* __fastcall mVUcompile(microVU* mVU, u32 startPC, uptr pState) { return thisPtr; } - + // Search for Existing Compiled Block (if found, return x86ptr; else, compile and return x86ptr) microVUt(void*) mVUblockFetch(microVU* mVU, u32 startPC, uptr pState) { using namespace x86Emitter; - if (startPC > mVU->microMemSize-8) { DevCon::Error("microVU%d: invalid startPC", params mVU->index); } startPC &= mVU->microMemSize-8; - if (mVUblocks[startPC/8] == NULL) { - mVUblocks[startPC/8] = microBlockManager::AlignedNew(); - } - - // Searches for Existing Compiled Block (if found, then returns; else, compile) + blockCreate(startPC/8); microBlock* pBlock = mVUblocks[startPC/8]->search((microRegInfo*)pState); if (pBlock) { return pBlock->x86ptrStart; } else { return mVUcompile(mVU, startPC, pState); } } -// Called By JR/JALR +// mVUcompileJIT() - Called By JR/JALR during execution microVUx(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState) { return mVUblockFetch(mVUx, startPC, pState); } From 3ad38e99cb4e48852d7854437a5fcb36a6dd1e56 Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 20 Jun 2009 13:08:17 +0000 Subject: [PATCH 263/296] Fix for Izumo Complete Issue 277, due to lack of understanding of VIF reset. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1409 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Gif.cpp | 11 +++-------- pcsx2/VifDma.cpp | 6 ++++++ 2 files changed, 9 insertions(+), 8 deletions(-) diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 15765e0fcb..55a4e0070c 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -249,14 +249,9 @@ void GIFdma() } GIFchain(); //Transfers the data set by the switch - - if (((gif->qwc == 0) && (gif->chcr & 0xc) == 0)) - gspath3done = 1; - else if(gif->qwc > 0 || gscycles) - { - GIFdmaEnd(); - return; - } + + GIFdmaEnd(); + return; } if ((gif->chcr & 0xc) == 0x4 && gspath3done == 0) { diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index fcbe73b6bf..0d176d2562 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -2691,6 +2691,12 @@ void vif1Write32(u32 mem, u32 value) psHu64(VIF1_FIFO) = 0; psHu64(0x10005008) = 0; // VIF1_FIFO + 8 vif1.done = true; + if(vif1Regs->mskpath3) + { + vif1Regs->mskpath3 = 0; + psHu32(GIF_STAT) &= ~0x2; + if(gif->chcr & 0x100) CPU_INT(2, 4); + } vif1Regs->err = 0; vif1.inprogress = 0; vif1Regs->stat &= ~(0x1F800000 | VIF1_STAT_INT | VIF1_STAT_VSS | VIF1_STAT_VIS | VIF1_STAT_VFS | VIF1_STAT_VPS); // FQC=0 From 7266dd12d28015ad5cfae9804e7c350a3439fd23 Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 20 Jun 2009 13:26:52 +0000 Subject: [PATCH 264/296] Fixed up Soul Calibur 3 regression in r1400. It was a test bit of code, which is obviously wrong :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1410 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Counters.cpp | 8 ++++---- pcsx2/MTGS.cpp | 12 +++++------- 2 files changed, 9 insertions(+), 11 deletions(-) diff --git a/pcsx2/Counters.cpp b/pcsx2/Counters.cpp index 3825b41d59..b43cf890f7 100644 --- a/pcsx2/Counters.cpp +++ b/pcsx2/Counters.cpp @@ -338,13 +338,13 @@ static __forceinline void VSyncStart(u32 sCycle) if ((CSRw & 0x8)) { - GSCSRr|= 0x8; + if (!(GSIMR&0x800)) { gsIrq(); } - CSRw &= ~0x8; //Disable the interrupt from triggering twice + GSCSRr|= 0x8; } hwIntcIrq(INTC_VBLANK_S); @@ -414,13 +414,13 @@ __forceinline void rcntUpdate_hScanline() else { //HBLANK END / HRENDER Begin if (CSRw & 0x4) { - GSCSRr |= 4; // signal + if (!(GSIMR&0x400)) { gsIrq(); } - CSRw &= ~0x4; //Disable the interrupt from triggering twice + GSCSRr |= 4; // signal } if (gates) rcntEndGate(false, hsyncCounter.sCycle); if (psxhblankgate) psxCheckEndGate16(0); diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 858d2536b6..131d80b150 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -139,13 +139,12 @@ static void RegHandlerSIGNAL(const u32* data) if ((CSRw & 0x1)) { - GSCSRr |= 1; // signal - if (!(GSIMR&0x100) ) { gsIrq(); } - CSRw &= ~0x1; //Disable the interrupt from triggering twice + + GSCSRr |= 1; // signal } } @@ -155,11 +154,10 @@ static void RegHandlerFINISH(const u32* data) if ((CSRw & 0x2)) { - GSCSRr |= 2; // finish - - if (!(GSIMR&0x200) ) + if (!(GSIMR&0x200)) gsIrq(); - CSRw &= ~0x2; //Disable the interrupt from triggering twice + + GSCSRr |= 2; // finish } } From d3f01bc572c3ccb98c996f51515fa8467cd132db Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 20 Jun 2009 16:55:18 +0000 Subject: [PATCH 265/296] Just a couple of unpack changes, some games thought they were overrunning when they weren't (punisher for one) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1411 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 32 +++++++++++++++++++------------- 1 file changed, 19 insertions(+), 13 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 0d176d2562..09b99613dc 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -622,23 +622,28 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i dest = (u32*)(VU->Mem + v->addr); } - tempsize = min(vifRegs->num, (size / ft->gsize)); - tempsize = (vif->tag.addr + (size / (ft->gsize * vifRegs->cycle.wl)) * - ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) + (tempsize * 16); + size = min(size, (int)vifRegs->num * ft->gsize); //size will always be the same or smaller + tempsize = vif->tag.addr + ((((vifRegs->num-1) / vifRegs->cycle.wl) * + (vifRegs->cycle.cl - vifRegs->cycle.wl)) * 16) + (vifRegs->num * 16); + + /*tempsize = vif->tag.addr + (((size / (ft->gsize * vifRegs->cycle.wl)) * + (vifRegs->cycle.cl - vifRegs->cycle.wl)) * 16) + (vifRegs->num * 16);*/ + + //Sanity Check (memory overflow) if (tempsize > memlimit) { if (((vifRegs->cycle.cl != vifRegs->cycle.wl) && - (memlimit + ((vifRegs->cycle.cl - vifRegs->cycle.wl) * 16)) == tempsize) || - (tempsize == memlimit)) + ((memlimit + (vifRegs->cycle.cl - vifRegs->cycle.wl) * 16) == tempsize) || + (tempsize == memlimit))) { //Its a red herring! so ignore it! SSE unpacks will be much quicker tempsize = 0; } else { - DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); + //DevCon::Notice("VIF%x Unpack ending %x > %x", params VIFdmanum, tempsize, VIFdmanum ? 0x4000 : 0x1000); tempsize = size; size = 0; } @@ -691,8 +696,6 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i vifRegs->cycle.cl = vifRegs->cycle.wl = 1; } - size = min(size, (int)vifRegs->num * ft->gsize); //size will always be the same or smaller - pfn = vif->usn ? VIFfuncTableSSE[unpackType].funcU : VIFfuncTableSSE[unpackType].funcS; writemask = VIFdmanum ? g_vif1HasMask3[min(vifRegs->cycle.wl,(u8)3)] : g_vif0HasMask3[min(vifRegs->cycle.wl,(u8)3)]; writemask = pfn[(((vifRegs->code & 0x10000000)>>28)<mode](dest, (u32*)cdata, size); @@ -952,20 +955,20 @@ static __forceinline void vif0UNPACK(u32 *data) if (vif0Regs->cycle.wl <= vif0Regs->cycle.cl) { - len = (((32 >> vl) * (vn + 1)) * vifNum + 31) >> 5; + vif0.tag.size = ((vifNum * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; } else { int n = vif0Regs->cycle.cl * (vifNum / vif0Regs->cycle.wl) + _limit(vifNum % vif0Regs->cycle.wl, vif0Regs->cycle.cl); - len = ((((32 >> vl) * (vn + 1)) * n) + 31) >> 5; + + vif0.tag.size = ((n * VIFfuncTable[ vif0.cmd & 0xf ].gsize) + 3) >> 2; } vif0.cl = 0; vif0.tag.cmd = vif0.cmd; vif0.tag.addr &= 0xfff; - vif0.tag.size = len; vif0Regs->offset = 0; @@ -1732,14 +1735,17 @@ static __forceinline void vif1UNPACK(u32 *data) if (vif1Regs->cycle.wl <= vif1Regs->cycle.cl) { - vif1.tag.size = (((32 >> vl) * (vn + 1)) * vifNum + 31) >> 5; + vif1.tag.size = ((vifNum * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; } else { int n = vif1Regs->cycle.cl * (vifNum / vif1Regs->cycle.wl) + _limit(vifNum % vif1Regs->cycle.wl, vif1Regs->cycle.cl); - vif1.tag.size = ((((32 >> vl) * (vn + 1)) * n) + 31) >> 5; + vif1.tag.size = ((n * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; + } + + if ((vif1Regs->code >> 15) & 0x1) vif1.tag.addr = (vif1Regs->code + vif1Regs->tops) & 0x3ff; else From 5f45219eb9a528e4a86c432e50669bd1d68b92aa Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 20 Jun 2009 20:28:36 +0000 Subject: [PATCH 266/296] GSdx: fixes for ffx2 menu git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1412 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSClut.cpp | 2 +- plugins/GSdx/GSCrc.cpp | 3 ++- plugins/GSdx/GSDrawScanline.cpp | 26 ++++++++++++++++---------- plugins/GSdx/GSTextureCache.cpp | 8 ++++++++ 4 files changed, 27 insertions(+), 12 deletions(-) diff --git a/plugins/GSdx/GSClut.cpp b/plugins/GSdx/GSClut.cpp index f428b29db4..59a717e82a 100644 --- a/plugins/GSdx/GSClut.cpp +++ b/plugins/GSdx/GSClut.cpp @@ -106,7 +106,7 @@ bool GSClut::WriteTest(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) case 3: m_CBP[1] = TEX0.CBP; break; case 4: if(m_CBP[0] == TEX0.CBP) return false; m_CBP[0] = TEX0.CBP; break; case 5: if(m_CBP[1] == TEX0.CBP) return false; m_CBP[1] = TEX0.CBP; break; - case 6: ASSERT(0); return false; + case 6: ASSERT(0); return false; // ffx2 menu case 7: ASSERT(0); return false; default: __assume(0); } diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index dce65c2aa9..fa70de419d 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -76,7 +76,8 @@ CRC::Game CRC::m_games[] = {0x35AA84D1, DBZBT2, Unknown}, {0x428113C2, DBZBT3, US}, {0xA422BB13, DBZBT3, EU}, - {0x983c53d2, DBZBT3, Unknown}, + {0x983C53D2, DBZBT3, Unknown}, + {0x983C53D3, DBZBT3, Unknown}, {0x72B3802A, SFEX3, US}, {0x71521863, SFEX3, US}, {0x28703748, Bully, US}, diff --git a/plugins/GSdx/GSDrawScanline.cpp b/plugins/GSdx/GSDrawScanline.cpp index 8370c4ab04..05cf43b328 100644 --- a/plugins/GSdx/GSDrawScanline.cpp +++ b/plugins/GSdx/GSDrawScanline.cpp @@ -298,19 +298,25 @@ void GSDrawScanline::DrawSolidRectT(const GSVector4i* row, int* col, const GSVec color = color.andnot(mask); - GSVector4i bm(8 * 4 / sizeof(T) - 1, 8 - 1); - GSVector4i br = (r + bm).andnot(bm.xyxy()); + GSVector4i br = r.ralign(GSVector2i(8 * 4 / sizeof(T), 8)); - FillRect(row, col, GSVector4i(r.x, r.y, r.z, br.y), c, m); - FillRect(row, col, GSVector4i(r.x, br.w, r.z, r.w), c, m); - - if(r.x < br.x || br.z < r.z) + if(!br.rempty()) { - FillRect(row, col, GSVector4i(r.x, br.y, br.x, br.w), c, m); - FillRect(row, col, GSVector4i(br.z, br.y, r.z, br.w), c, m); - } + FillRect(row, col, GSVector4i(r.x, r.y, r.z, br.y), c, m); + FillRect(row, col, GSVector4i(r.x, br.w, r.z, r.w), c, m); - FillBlock(row, col, br, color, mask); + if(r.x < br.x || br.z < r.z) + { + FillRect(row, col, GSVector4i(r.x, br.y, br.x, br.w), c, m); + FillRect(row, col, GSVector4i(br.z, br.y, r.z, br.w), c, m); + } + + FillBlock(row, col, br, color, mask); + } + else + { + FillRect(row, col, r, c, m); + } } template diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 6bca8d5265..df2a17ac10 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -286,6 +286,8 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() t = t2; m_tex.splice(m_tex.begin(), m_tex, i); + + break; } if(t == NULL) @@ -727,7 +729,13 @@ void GSTextureCache::GSDepthStencil::Update() { __super::Update(); + GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); + + if(r.rempty()) return; + // TODO + + m_renderer->m_dev->ClearDepth(m_texture, 0); } // GSTextureCache::GSCachedTexture From fffe95bc4b32c825a4136253f2d3c11e885fd969 Mon Sep 17 00:00:00 2001 From: refraction Date: Sat, 20 Jun 2009 22:46:20 +0000 Subject: [PATCH 267/296] Fixed a really silly typo which possibly broke loads (Including Star Ocean 3, thanks rama :P) masked it by removing some redundant code too. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1413 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VifDma.cpp | 25 +++++++------------------ 1 file changed, 7 insertions(+), 18 deletions(-) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 09b99613dc..ec5c6525b4 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -724,7 +724,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i if (size >= ft->dsize && vifRegs->num > 0) { - //VIF_LOG("warning, end with size = %d", size); + VIF_LOG("warning, end with size = %d", size); /* unpack one qword */ //vif->tag.addr += (size / ft->dsize) * 4; @@ -806,7 +806,7 @@ static void VIFunpack(u32 *data, vifCode *v, unsigned int size, const unsigned i } if (size >= ft->dsize && vifRegs->num > 0) //Else write what we do have { - //VIF_LOG("warning, end with size = %d", size); + VIF_LOG("warning, end with size = %d", size); /* unpack one qword */ //vif->tag.addr += (size / ft->dsize) * 4; @@ -933,8 +933,6 @@ void vif0Init() static __forceinline void vif0UNPACK(u32 *data) { int vifNum; - int vl, vn; - int len; if (vif0Regs->cycle.wl == 0 && vif0Regs->cycle.wl < vif0Regs->cycle.cl) { @@ -945,9 +943,7 @@ static __forceinline void vif0UNPACK(u32 *data) vif0FLUSH(); - vl = (vif0.cmd) & 0x3; - vn = (vif0.cmd >> 2) & 0x3; - vif0.tag.addr = (vif0Regs->code & 0x3ff) << 4; + vif0.tag.addr = (vif0Regs->code & 0xff) << 4; vif0.usn = (vif0Regs->code >> 14) & 0x1; vifNum = (vif0Regs->code >> 16) & 0xff; if (vifNum == 0) vifNum = 256; @@ -955,20 +951,18 @@ static __forceinline void vif0UNPACK(u32 *data) if (vif0Regs->cycle.wl <= vif0Regs->cycle.cl) { - vif0.tag.size = ((vifNum * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; + vif0.tag.size = ((vifNum * VIFfuncTable[ vif0.cmd & 0xf ].gsize) + 3) >> 2; } else { int n = vif0Regs->cycle.cl * (vifNum / vif0Regs->cycle.wl) + _limit(vifNum % vif0Regs->cycle.wl, vif0Regs->cycle.cl); - vif0.tag.size = ((n * VIFfuncTable[ vif0.cmd & 0xf ].gsize) + 3) >> 2; } vif0.cl = 0; vif0.tag.cmd = vif0.cmd; - vif0.tag.addr &= 0xfff; vif0Regs->offset = 0; @@ -1711,7 +1705,6 @@ void vif1Init() static __forceinline void vif1UNPACK(u32 *data) { int vifNum; - int vl, vn; if (vif1Regs->cycle.wl == 0) { @@ -1724,9 +1717,6 @@ static __forceinline void vif1UNPACK(u32 *data) } //vif1FLUSH(); - vl = (vif1.cmd) & 0x3; - vn = (vif1.cmd >> 2) & 0x3; - vif1.usn = (vif1Regs->code >> 14) & 0x1; vifNum = (vif1Regs->code >> 16) & 0xff; @@ -1741,10 +1731,9 @@ static __forceinline void vif1UNPACK(u32 *data) { int n = vif1Regs->cycle.cl * (vifNum / vif1Regs->cycle.wl) + _limit(vifNum % vif1Regs->cycle.wl, vif1Regs->cycle.cl); - vif1.tag.size = ((n * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; - - } - + + vif1.tag.size = ((n * VIFfuncTable[ vif1.cmd & 0xf ].gsize) + 3) >> 2; + } if ((vif1Regs->code >> 15) & 0x1) vif1.tag.addr = (vif1Regs->code + vif1Regs->tops) & 0x3ff; From 9d091450b14c93d351609aa7e41e87931da7c428 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Sun, 21 Jun 2009 07:01:41 +0000 Subject: [PATCH 268/296] microVU: minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1414 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/VUmicro.h | 4 +-- pcsx2/x86/microVU.cpp | 56 ++++++++++++++++++----------------- pcsx2/x86/microVU.h | 8 ++--- pcsx2/x86/microVU_Execute.inl | 2 -- pcsx2/x86/microVU_Misc.h | 1 + 5 files changed, 36 insertions(+), 35 deletions(-) diff --git a/pcsx2/VUmicro.h b/pcsx2/VUmicro.h index 922d74328a..2a88d9ce4d 100644 --- a/pcsx2/VUmicro.h +++ b/pcsx2/VUmicro.h @@ -76,9 +76,9 @@ void vuMicroCpuReset(); extern void initVUrec(VURegs* vuRegs, const int vuIndex); extern void closeVUrec(const int vuIndex); extern void resetVUrec(const int vuIndex); -extern void clearVUrec(u32 addr, u32 size, const int vuIndex); -extern void runVUrec(u32 startPC, u32 cycles, const int vuIndex); extern void vsyncVUrec(const int vuIndex); +extern void __fastcall clearVUrec(u32 addr, u32 size, const int vuIndex); +extern void __fastcall runVUrec(u32 startPC, u32 cycles, const int vuIndex); ///////////////////////////////////////////////////////////////// // Everything else does stuff on a per-VU basis. diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index d6198eebee..829ddd0803 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -49,8 +49,8 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr, int vuIndex) { mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); mVU->cache = SysMmapEx((vuIndex ? 0x5f240000 : 0x5e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); - if ( mVU->cache == NULL ) throw Exception::OutOfMemory(fmt_string( "microVU Error: Failed to allocate recompiler memory! (addr: 0x%x)", (u32)mVU->cache)); - + if (!mVU->cache) throw Exception::OutOfMemory(fmt_string("microVU Error: Failed to allocate recompiler memory!")); + mVUemitSearch(); mVUreset(mVU); } @@ -117,10 +117,10 @@ microVUt(void) mVUclose(mV) { // Clears Block Data in specified range microVUt(void) mVUclear(mV, u32 addr, u32 size) { - if (!mVU->prog.cleared) { - memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); + //if (!mVU->prog.cleared) { + //memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); mVU->prog.cleared = 1; // Next execution searches/creates a new microprogram - } + //} } //------------------------------------------------------------------ @@ -161,22 +161,20 @@ microVUf(int) mVUfindLeastUsedProg() { } else { - int startidx = (mVU->prog.cur + 1) & mVU->prog.max; - int endidx = mVU->prog.cur; - int smallidx = startidx; - u32 smallval = mVU->prog.prog[startidx].used; + const int pMax = mVU->prog.max; + int smallidx = (mVU->prog.cur+1)&pMax; + u32 smallval = mVU->prog.prog[smallidx].used; - for (int i = startidx; i != endidx; i = (i+1)&mVU->prog.max) { - u32 used = mVU->prog.prog[i].used; - if (smallval > used) { - smallval = used; - smallidx = i; + for (int i = 1, j = (smallidx+1)&pMax; i <= pMax; i++, j=(j+1)&pMax) { + if (smallval > mVU->prog.prog[j].used) { + smallval = mVU->prog.prog[j].used; + smallidx = j; } } mVUclearProg(smallidx); // Clear old data if overwriting old program mVUcacheProg(smallidx); // Cache Micro Program - //Console::Notice("microVU%d: Overwriting existing program in slot %d [%d times used]", params vuIndex, smallidx, smallval ); + //Console::Notice("microVU%d: Overwriting existing program in slot %d [%d times used]", params vuIndex, smallidx, smallval); return smallidx; } } @@ -187,13 +185,13 @@ microVUf(int) mVUfindLeastUsedProg() { // // To fix the program cache to more efficiently dispose of "obsolete" programs, we need to use a // frame-based decrementing system in combination with a program-execution-based incrementing -// system. In english: if last_used >= 2 it means the program has been used for the current +// system. In English: if last_used >= 2 it means the program has been used for the current // or prev frame. if it's 0, the program hasn't been used for a while. microVUt(void) mVUvsyncUpdate(mV) { if (mVU->prog.total < mVU->prog.max) return; - for (int i = 0; i <= mVU->prog.total; i++) { + for (int i = 0; i <= mVU->prog.max; i++) { if (mVU->prog.prog[i].last_used != 0) { if (mVU->prog.prog[i].last_used >= 3) { @@ -212,7 +210,7 @@ microVUf(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, b if (progUsed) { if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) || - (!cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data + mVUprogI.range[0], (u8*)mVU->regs->Micro + mVUprogI.range[0], ((mVUprogI.range[1] + 8) - mVUprogI.range[0]))))) { + (!cmpWholeProg && (!memcmp_mmx(cmpOffset(mVUprogI.data), cmpOffset(mVU->regs->Micro), ((mVUprogI.range[1] + 8) - mVUprogI.range[0]))))) { mVU->prog.cur = progIndex; mVU->prog.cleared = 0; mVU->prog.isSame = cmpWholeProg ? 1 : -1; @@ -229,7 +227,7 @@ microVUf(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, b // Searches for Cached Micro Program and sets prog.cur to it (returns 1 if program found, else returns 0) microVUf(int) mVUsearchProg() { microVU* mVU = mVUx; - + if (mVU->prog.cleared) { // If cleared, we need to search for new program for (int i = 0; i <= mVU->prog.total; i++) { if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 1, 0)) @@ -253,13 +251,17 @@ microVUf(int) mVUsearchProg() { // Wrapper Functions - Called by other parts of the Emu //------------------------------------------------------------------ -void initVUrec (VURegs* vuRegs, const int vuIndex) { mVUinit(vuRegs, vuIndex); } -void closeVUrec(const int vuIndex) { mVUclose(mVUx); } -void resetVUrec(const int vuIndex) { mVUreset(mVUx); } -void vsyncVUrec(const int vuIndex) { mVUvsyncUpdate(mVUx); } -void clearVUrec(u32 addr, u32 size, const int vuIndex) { mVUclear(mVUx, addr, size); } +void initVUrec (VURegs* vuRegs, const int vuIndex) { mVUinit(vuRegs, vuIndex); } +void closeVUrec(const int vuIndex) { mVUclose(mVUx); } +void resetVUrec(const int vuIndex) { mVUreset(mVUx); } +void vsyncVUrec(const int vuIndex) { mVUvsyncUpdate(mVUx); } -void runVUrec(u32 startPC, u32 cycles, const int vuIndex) { - if (!vuIndex) startVU0(startPC, cycles); - else startVU1(startPC, cycles); +void __fastcall clearVUrec(u32 addr, u32 size, const int vuIndex) { + mVUclear(mVUx, addr, size); } + +void __fastcall runVUrec(u32 startPC, u32 cycles, const int vuIndex) { + if (!vuIndex) ((mVUrecCall)microVU0.startFunct)(startPC, cycles); + else ((mVUrecCall)microVU1.startFunct)(startPC, cycles); +} + diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index d7256904c3..3ebd1d94ef 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -85,15 +85,15 @@ public: template // progSize = VU program memory size / 4 struct microProgram { - u32 data[progSize]; + PCSX2_ALIGNED16(u32 data[progSize]); + microBlockManager* block[progSize/2]; + microIR allocInfo; u32 used; // Number of times its been used u32 last_used; // Counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vSync) s32 range[2]; // The range of microMemory that has already been recompiled for the current program u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache - microBlockManager* block[progSize/2]; - microIR allocInfo; }; #define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) @@ -128,7 +128,7 @@ struct microVU { VURegs* regs; // VU Regs Struct u8* cache; // Dynarec Cache Start (where we will start writing the recompiled code to) u8* startFunct; // Ptr Function to the Start code for recompiled programs - u8* exitFunct; // Ptr Function to the Exit code for recompiled programs + u8* exitFunct; // Ptr Function to the Exit code for recompiled programs u32 code; // Contains the current Instruction u32 divFlag; // 1 instance of I/D flags u32 VIbackup; // Holds a backup of a VI reg if modified before a branch diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 1ca593aac8..3017e0636a 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -151,8 +151,6 @@ microVUx(void) mVUcleanUp() { // Caller Functions //------------------------------------------------------------------ -void __fastcall startVU0(u32 startPC, u32 cycles) { ((mVUrecCall)microVU0.startFunct)(startPC, cycles); } -void __fastcall startVU1(u32 startPC, u32 cycles) { ((mVUrecCall)microVU1.startFunct)(startPC, cycles); } void* __fastcall mVUexecuteVU0(u32 startPC, u32 cycles) { return mVUexecute<0>(startPC, cycles); } void* __fastcall mVUexecuteVU1(u32 startPC, u32 cycles) { return mVUexecute<1>(startPC, cycles); } void __fastcall mVUcleanUpVU0() { mVUcleanUp<0>(); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index b91f9e74a9..cfb3a54258 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -206,6 +206,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microMemSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) +#define cmpOffset(x) (&(((u8*)x)[mVUprogI.range[0]])) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL From 101dbdd1322288fa38aa5cc41afd2d191b825290 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sun, 21 Jun 2009 08:24:17 +0000 Subject: [PATCH 269/296] SPU2-X: *Really* fix compat with older versions of pcsx2 this time (0.9.4 and prior) >_< git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1415 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/spu2-x/src/DllInterface.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/plugins/spu2-x/src/DllInterface.cpp b/plugins/spu2-x/src/DllInterface.cpp index cc9b0097ad..33527ad774 100644 --- a/plugins/spu2-x/src/DllInterface.cpp +++ b/plugins/spu2-x/src/DllInterface.cpp @@ -372,7 +372,8 @@ EXPORT_C_(u16) SPU2read(u32 rmem) } else { - TimeUpdate( *cyclePtr ); + if( cyclePtr != NULL ) + TimeUpdate( *cyclePtr ); if (rmem>>16 == 0x1f80) { From ed89876ba6be0f12d106aa6ff798c79a9e31b5d5 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sun, 21 Jun 2009 09:26:28 +0000 Subject: [PATCH 270/296] Added a new function "DevAssert", which is intended to provide assert-like functionality to Devel builds of Pcsx2. Developers: Check the code comments for more details. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1417 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Exceptions.h | 10 ++++++++++ pcsx2/Misc.cpp | 34 ++++++++++++++++++++++++++++++++++ pcsx2/PrecompiledHeader.h | 9 +++++---- 3 files changed, 49 insertions(+), 4 deletions(-) diff --git a/pcsx2/Exceptions.h b/pcsx2/Exceptions.h index 640e61c0e1..ce2266aaf0 100644 --- a/pcsx2/Exceptions.h +++ b/pcsx2/Exceptions.h @@ -18,6 +18,8 @@ #pragma once +extern void DevAssert( bool condition, const char* msg ); + // This class provides an easy and clean method for ensuring objects are not copyable. class NoncopyableObject { @@ -121,6 +123,14 @@ namespace Exception {} }; + class AssertionFailure : public LogicError + { + public: + explicit AssertionFailure( const std::string& msg="Assertion Failure" ) : + LogicError( msg ) {} + virtual ~AssertionFailure() throw() {} + }; + ////////////////////////////////////////////////////////////////////////////////// // class OutOfMemory : public RuntimeError diff --git a/pcsx2/Misc.cpp b/pcsx2/Misc.cpp index d3799e17e2..a1a975c6fc 100644 --- a/pcsx2/Misc.cpp +++ b/pcsx2/Misc.cpp @@ -139,6 +139,40 @@ const char *LabelGreets = { N_( ) }; +// ------------------------------------------------------------------------ +// Force DevAssert to *not* inline for devel/debug builds (allows using breakpoints to trap +// assertions), and force it to inline for release builds (optimizes it out completely since +// IsDevBuild is false). Since Devel builds typically aren't enabled with Global Optimization/ +// LTCG, this currently isn't even necessary. But might as well, in case we decide at a later +// date to re-enable LTCG for devel. +#ifdef PCSX2_DEVBUILD +# define DEVASSERT_INLINE __noinline +#else +# define DEVASSERT_INLINE __forceinline +#endif + +////////////////////////////////////////////////////////////////////////////////////////// +// Assertion tool for Devel builds, intended for sanity checking and/or bounds checking +// variables in areas which are not performance critical. +// +// How it works: This function throws an exception of type Exception::AssertionFailure if +// the assertion conditional is false. Typically for the end-user, this exception is handled +// by the general handler, which (should eventually) create some state dumps and other +// information for troubleshooting purposes. +// +// From a debugging environment, you can trap your DevAssert by either breakpointing the +// exception throw below, or by adding either Exception::AssertionFailure or +// Exception::LogicError to your First-Chance Exception catch list (Visual Studio, under +// the Debug->Exceptions menu/dialog). +// +DEVASSERT_INLINE void DevAssert( bool condition, const char* msg ) +{ + if( IsDevBuild && !condition ) + { + throw Exception::AssertionFailure( msg ); + } +} + u32 GetBiosVersion() { unsigned int fileOffset=0; s8 *ROMVER; diff --git a/pcsx2/PrecompiledHeader.h b/pcsx2/PrecompiledHeader.h index 764dc565e6..47f62e9b4b 100644 --- a/pcsx2/PrecompiledHeader.h +++ b/pcsx2/PrecompiledHeader.h @@ -150,12 +150,13 @@ typedef int BOOL; #endif // ENABLE_NLS ////////////////////////////////////////////////////////////////////////////////////////// -// Forceinline macro that is enabled for RELEASE/PUBLIC builds ONLY. (non-inline in devel) +// Forceinline macro that is enabled for RELEASE/PUBLIC builds ONLY. (compiler-picked in devel) // This is useful because forceinline can make certain types of debugging problematic since // functions that look like they should be called won't breakpoint since their code is inlined. -// Henceforth, use release_inline for things which we want inlined on public/release builds but -// *not* in devel builds. - +// Henceforth, use __releaseinline for things which are generally large functions where trace +// debugging from Devel builds is likely useful; but which should be inlined in an optimized +// Release environment. +// #ifdef PCSX2_DEVBUILD # define __releaseinline #else From bd093b41bd418ddb995d3e3147013b1048e9302f Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 21 Jun 2009 10:26:56 +0000 Subject: [PATCH 271/296] Recursive inlining appears to be driving gcc to drink, or at least, to use up all available system memory trying to compile microVU.cpp and then failing. :( git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1418 96395faa-99c1-11dd-bbfe-3dabce05a288 --- build.rb | 2 +- pcsx2/x86/iR5900Misc.cpp | 2 +- pcsx2/x86/microVU_Compile.inl | 2 +- pcsx2/x86/microVU_Execute.inl | 2 +- pcsx2/x86/sVU_zerorec.cpp | 4 ++-- plugins/PadNull/Pad.h | 7 ------- 6 files changed, 6 insertions(+), 13 deletions(-) diff --git a/build.rb b/build.rb index 58d3d474c2..b9538a2aa1 100755 --- a/build.rb +++ b/build.rb @@ -22,7 +22,7 @@ $pcsx2_prefix = " --prefix #{$main_dir}" $plugins_prefix = " --prefix #{$plugin_install_dir}" $plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz"] -$full_plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz","GSnull","PadNull"] +$full_plugin_list=["CDVDnull", "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP", "CDVDlinuz","GSnull","PadNull","wxpad"] $pcsx2_build_types = { "dev" => " --enable-devbuild ", diff --git a/pcsx2/x86/iR5900Misc.cpp b/pcsx2/x86/iR5900Misc.cpp index e496a1d311..28bdae4b03 100644 --- a/pcsx2/x86/iR5900Misc.cpp +++ b/pcsx2/x86/iR5900Misc.cpp @@ -8,7 +8,7 @@ namespace R5900 { namespace Dynarec { -// R5900 branch hepler! +// R5900 branch helper! // Recompiles code for a branch test and/or skip, complete with delay slot // handling. Note, for "likely" branches use iDoBranchImm_Likely instead, which // handles delay slots differently. diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index aa0f9ab271..439c42d26b 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -290,7 +290,7 @@ microVUt(void) mVUtestCycles(mV) { // Recompiler //------------------------------------------------------------------ -microVUt(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { +void* mVUcompile(microVU* mVU, u32 startPC, uptr pState) { using namespace x86Emitter; microBlock* pBlock = NULL; diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 3017e0636a..8d34b48000 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -84,7 +84,7 @@ microVUt(void) mVUdispatcherA(mV) { } // Generates the code to exit from recompiled blocks -microVUt(void) mVUdispatcherB(mV) { +void mVUdispatcherB(mV) { mVU->exitFunct = x86Ptr; // Load EE's MXCSR state diff --git a/pcsx2/x86/sVU_zerorec.cpp b/pcsx2/x86/sVU_zerorec.cpp index 95aedcc6f7..67af485fc2 100644 --- a/pcsx2/x86/sVU_zerorec.cpp +++ b/pcsx2/x86/sVU_zerorec.cpp @@ -98,7 +98,7 @@ extern void (*recVU_LOWER_OPCODE[128])(VURegs* VU, s32 info); // Let's tempt fate by defining two different constants with almost identical names #define INST_DUMMY_ 0x8000 -#define INST_DUMMY 0x83c0 +#define INST_DUMMY 0x83c0 #define VFFREE_INVALID0 0x80000000 // (vffree[i]&0xf) is invalid @@ -955,7 +955,7 @@ static int _recbranchAddr(u32 vucode) static VuInstruction SuperVUFlushInst() { VuInstruction inst; - // don't need to raed q/p + // don't need to read q/p inst.type = INST_DUMMY_;//|INST_Q_READ|INST_P_READ; return inst; } diff --git a/plugins/PadNull/Pad.h b/plugins/PadNull/Pad.h index 4f391f5c25..fb24ff72ca 100644 --- a/plugins/PadNull/Pad.h +++ b/plugins/PadNull/Pad.h @@ -21,15 +21,8 @@ #include -#ifdef __cplusplus -extern "C" -{ -#endif #define PADdefs #include "PS2Edefs.h" -#ifdef __cplusplus -} -#endif #ifdef _WIN32 #include "PadWin.h" From c1dbdd7b62ae3ade540039541aee21153e25b235 Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Sun, 21 Jun 2009 11:00:53 +0000 Subject: [PATCH 272/296] SPU2-X: Upgraded license to the LPGLv3; and changed some #ifdef mess to #pragma once. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1419 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/spu2-x/GPL.txt | 674 ++++++++++++++++++ plugins/spu2-x/LGPL.txt | 657 +++++------------ plugins/spu2-x/License.txt | 29 +- plugins/spu2-x/src/ADSR.cpp | 34 +- plugins/spu2-x/src/Debug.cpp | 24 +- plugins/spu2-x/src/DllInterface.cpp | 24 +- plugins/spu2-x/src/Dma.cpp | 24 +- plugins/spu2-x/src/Linux/Alsa.cpp | 353 +++++---- plugins/spu2-x/src/Linux/Config.cpp | 26 +- plugins/spu2-x/src/Linux/Config.h | 26 +- plugins/spu2-x/src/Linux/Dialogs.cpp | 26 +- plugins/spu2-x/src/Linux/Dialogs.h | 26 +- plugins/spu2-x/src/Linux/Linux.cpp | 26 +- plugins/spu2-x/src/Linux/Linux.h | 26 +- plugins/spu2-x/src/Lowpass.cpp | 24 +- plugins/spu2-x/src/Lowpass.h | 38 +- plugins/spu2-x/src/Mixer.cpp | 24 +- plugins/spu2-x/src/ReadInput.cpp | 34 +- plugins/spu2-x/src/RegLog.cpp | 24 +- plugins/spu2-x/src/RegTable.cpp | 34 +- plugins/spu2-x/src/Reverb.cpp | 24 +- plugins/spu2-x/src/SaveStateSPU.cpp | 24 +- plugins/spu2-x/src/SndOut.cpp | 24 +- plugins/spu2-x/src/Spu2.cpp | 25 +- plugins/spu2-x/src/Spu2.h | 29 +- plugins/spu2-x/src/Spu2replay.h | 5 +- plugins/spu2-x/src/Timestretcher.cpp | 24 +- plugins/spu2-x/src/Wavedump_wav.cpp | 24 +- plugins/spu2-x/src/Windows/AboutBox.cpp | 24 +- plugins/spu2-x/src/Windows/CfgHelpers.cpp | 24 +- plugins/spu2-x/src/Windows/Config.cpp | 24 +- plugins/spu2-x/src/Windows/Config.h | 29 +- plugins/spu2-x/src/Windows/ConfigDebug.cpp | 24 +- .../spu2-x/src/Windows/ConfigSoundtouch.cpp | 24 +- plugins/spu2-x/src/Windows/Dialogs.h | 24 +- .../spu2-x/src/Windows/RealtimeDebugger.cpp | 24 +- plugins/spu2-x/src/Windows/SndOut_DSound.cpp | 24 +- plugins/spu2-x/src/Windows/SndOut_XAudio2.cpp | 24 +- plugins/spu2-x/src/Windows/SndOut_waveOut.cpp | 24 +- plugins/spu2-x/src/Windows/Spu2-X.def | 23 +- plugins/spu2-x/src/Windows/UIHelpers.cpp | 36 +- plugins/spu2-x/src/Windows/dsp.h | 2 + plugins/spu2-x/src/defs.h | 29 +- plugins/spu2-x/src/regs.h | 38 +- plugins/spu2-x/src/spdif.h | 5 +- 45 files changed, 1452 insertions(+), 1282 deletions(-) create mode 100644 plugins/spu2-x/GPL.txt diff --git a/plugins/spu2-x/GPL.txt b/plugins/spu2-x/GPL.txt new file mode 100644 index 0000000000..3f2e0ee4ce --- /dev/null +++ b/plugins/spu2-x/GPL.txt @@ -0,0 +1,674 @@ + GNU GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU General Public License is a free, copyleft license for +software and other kinds of works. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +the GNU General Public License is intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. We, the Free Software Foundation, use the +GNU General Public License for most of our software; it applies also to +any other work released this way by its authors. You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + To protect your rights, we need to prevent others from denying you +these rights or asking you to surrender the rights. Therefore, you have +certain responsibilities if you distribute copies of the software, or if +you modify it: responsibilities to respect the freedom of others. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must pass on to the recipients the same +freedoms that you received. You must make sure that they, too, receive +or can get the source code. And you must show them these terms so they +know their rights. + + Developers that use the GNU GPL protect your rights with two steps: +(1) assert copyright on the software, and (2) offer you this License +giving you legal permission to copy, distribute and/or modify it. + + For the developers' and authors' protection, the GPL clearly explains +that there is no warranty for this free software. For both users' and +authors' sake, the GPL requires that modified versions be marked as +changed, so that their problems will not be attributed erroneously to +authors of previous versions. + + Some devices are designed to deny users access to install or run +modified versions of the software inside them, although the manufacturer +can do so. This is fundamentally incompatible with the aim of +protecting users' freedom to change the software. The systematic +pattern of such abuse occurs in the area of products for individuals to +use, which is precisely where it is most unacceptable. Therefore, we +have designed this version of the GPL to prohibit the practice for those +products. If such problems arise substantially in other domains, we +stand ready to extend this provision to those domains in future versions +of the GPL, as needed to protect the freedom of users. + + Finally, every program is threatened constantly by software patents. +States should not allow patents to restrict development and use of +software on general-purpose computers, but in those that do, we wish to +avoid the special danger that patents applied to a free program could +make it effectively proprietary. To prevent this, the GPL assures that +patents cannot be used to render the program non-free. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Use with the GNU Affero General Public License. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU Affero General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the special requirements of the GNU Affero General Public License, +section 13, concerning interaction through a network will apply to the +combination as such. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If the program does terminal interaction, make it output a short +notice like this when it starts in an interactive mode: + + Copyright (C) + This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, your program's commands +might be different; for a GUI interface, you would use an "about box". + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU GPL, see +. + + The GNU General Public License does not permit incorporating your program +into proprietary programs. If your program is a subroutine library, you +may consider it more useful to permit linking proprietary applications with +the library. If this is what you want to do, use the GNU Lesser General +Public License instead of this License. But first, please read +. \ No newline at end of file diff --git a/plugins/spu2-x/LGPL.txt b/plugins/spu2-x/LGPL.txt index ebb7aa9340..a5043111dc 100644 --- a/plugins/spu2-x/LGPL.txt +++ b/plugins/spu2-x/LGPL.txt @@ -1,503 +1,164 @@ - GNU LESSER GENERAL PUBLIC LICENSE - Version 2.1, February 1999 + GNU LESSER GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 - Copyright (C) 1991, 1999 Free Software Foundation, Inc. - 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + Copyright (C) 2007 Free Software Foundation, Inc. Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. -[This is the first released version of the Lesser GPL. It also counts - as the successor of the GNU Library Public License, version 2, hence - the version number 2.1.] - - Preamble - - The licenses for most software are designed to take away your -freedom to share and change it. By contrast, the GNU General Public -Licenses are intended to guarantee your freedom to share and change -free software--to make sure the software is free for all its users. - - This license, the Lesser General Public License, applies to some -specially designated software packages--typically libraries--of the -Free Software Foundation and other authors who decide to use it. You -can use it too, but we suggest you first think carefully about whether -this license or the ordinary General Public License is the better -strategy to use in any particular case, based on the explanations below. - - When we speak of free software, we are referring to freedom of use, -not price. Our General Public Licenses are designed to make sure that -you have the freedom to distribute copies of free software (and charge -for this service if you wish); that you receive source code or can get -it if you want it; that you can change the software and use pieces of -it in new free programs; and that you are informed that you can do -these things. - - To protect your rights, we need to make restrictions that forbid -distributors to deny you these rights or to ask you to surrender these -rights. These restrictions translate to certain responsibilities for -you if you distribute copies of the library or if you modify it. - - For example, if you distribute copies of the library, whether gratis -or for a fee, you must give the recipients all the rights that we gave -you. You must make sure that they, too, receive or can get the source -code. If you link other code with the library, you must provide -complete object files to the recipients, so that they can relink them -with the library after making changes to the library and recompiling -it. And you must show them these terms so they know their rights. - - We protect your rights with a two-step method: (1) we copyright the -library, and (2) we offer you this license, which gives you legal -permission to copy, distribute and/or modify the library. - - To protect each distributor, we want to make it very clear that -there is no warranty for the free library. Also, if the library is -modified by someone else and passed on, the recipients should know -that what they have is not the original version, so that the original -author's reputation will not be affected by problems that might be -introduced by others. - - Finally, software patents pose a constant threat to the existence of -any free program. We wish to make sure that a company cannot -effectively restrict the users of a free program by obtaining a -restrictive license from a patent holder. Therefore, we insist that -any patent license obtained for a version of the library must be -consistent with the full freedom of use specified in this license. - - Most GNU software, including some libraries, is covered by the -ordinary GNU General Public License. This license, the GNU Lesser -General Public License, applies to certain designated libraries, and -is quite different from the ordinary General Public License. We use -this license for certain libraries in order to permit linking those -libraries into non-free programs. - - When a program is linked with a library, whether statically or using -a shared library, the combination of the two is legally speaking a -combined work, a derivative of the original library. The ordinary -General Public License therefore permits such linking only if the -entire combination fits its criteria of freedom. The Lesser General -Public License permits more lax criteria for linking other code with -the library. - - We call this license the "Lesser" General Public License because it -does Less to protect the user's freedom than the ordinary General -Public License. It also provides other free software developers Less -of an advantage over competing non-free programs. These disadvantages -are the reason we use the ordinary General Public License for many -libraries. However, the Lesser license provides advantages in certain -special circumstances. - - For example, on rare occasions, there may be a special need to -encourage the widest possible use of a certain library, so that it becomes -a de-facto standard. To achieve this, non-free programs must be -allowed to use the library. A more frequent case is that a free -library does the same job as widely used non-free libraries. In this -case, there is little to gain by limiting the free library to free -software only, so we use the Lesser General Public License. - - In other cases, permission to use a particular library in non-free -programs enables a greater number of people to use a large body of -free software. For example, permission to use the GNU C Library in -non-free programs enables many more people to use the whole GNU -operating system, as well as its variant, the GNU/Linux operating -system. - - Although the Lesser General Public License is Less protective of the -users' freedom, it does ensure that the user of a program that is -linked with the Library has the freedom and the wherewithal to run -that program using a modified version of the Library. - - The precise terms and conditions for copying, distribution and -modification follow. Pay close attention to the difference between a -"work based on the library" and a "work that uses the library". The -former contains code derived from the library, whereas the latter must -be combined with the library in order to run. - - GNU LESSER GENERAL PUBLIC LICENSE - TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION - - 0. This License Agreement applies to any software library or other -program which contains a notice placed by the copyright holder or -other authorized party saying it may be distributed under the terms of -this Lesser General Public License (also called "this License"). -Each licensee is addressed as "you". - - A "library" means a collection of software functions and/or data -prepared so as to be conveniently linked with application programs -(which use some of those functions and data) to form executables. - - The "Library", below, refers to any such software library or work -which has been distributed under these terms. A "work based on the -Library" means either the Library or any derivative work under -copyright law: that is to say, a work containing the Library or a -portion of it, either verbatim or with modifications and/or translated -straightforwardly into another language. (Hereinafter, translation is -included without limitation in the term "modification".) - - "Source code" for a work means the preferred form of the work for -making modifications to it. For a library, complete source code means -all the source code for all modules it contains, plus any associated -interface definition files, plus the scripts used to control compilation -and installation of the library. - - Activities other than copying, distribution and modification are not -covered by this License; they are outside its scope. The act of -running a program using the Library is not restricted, and output from -such a program is covered only if its contents constitute a work based -on the Library (independent of the use of the Library in a tool for -writing it). Whether that is true depends on what the Library does -and what the program that uses the Library does. - - 1. You may copy and distribute verbatim copies of the Library's -complete source code as you receive it, in any medium, provided that -you conspicuously and appropriately publish on each copy an -appropriate copyright notice and disclaimer of warranty; keep intact -all the notices that refer to this License and to the absence of any -warranty; and distribute a copy of this License along with the -Library. - - You may charge a fee for the physical act of transferring a copy, -and you may at your option offer warranty protection in exchange for a -fee. - - 2. You may modify your copy or copies of the Library or any portion -of it, thus forming a work based on the Library, and copy and -distribute such modifications or work under the terms of Section 1 -above, provided that you also meet all of these conditions: - - a) The modified work must itself be a software library. - - b) You must cause the files modified to carry prominent notices - stating that you changed the files and the date of any change. - - c) You must cause the whole of the work to be licensed at no - charge to all third parties under the terms of this License. - - d) If a facility in the modified Library refers to a function or a - table of data to be supplied by an application program that uses - the facility, other than as an argument passed when the facility - is invoked, then you must make a good faith effort to ensure that, - in the event an application does not supply such function or - table, the facility still operates, and performs whatever part of - its purpose remains meaningful. - - (For example, a function in a library to compute square roots has - a purpose that is entirely well-defined independent of the - application. Therefore, Subsection 2d requires that any - application-supplied function or table used by this function must - be optional: if the application does not supply it, the square - root function must still compute square roots.) - -These requirements apply to the modified work as a whole. If -identifiable sections of that work are not derived from the Library, -and can be reasonably considered independent and separate works in -themselves, then this License, and its terms, do not apply to those -sections when you distribute them as separate works. But when you -distribute the same sections as part of a whole which is a work based -on the Library, the distribution of the whole must be on the terms of -this License, whose permissions for other licensees extend to the -entire whole, and thus to each and every part regardless of who wrote -it. - -Thus, it is not the intent of this section to claim rights or contest -your rights to work written entirely by you; rather, the intent is to -exercise the right to control the distribution of derivative or -collective works based on the Library. - -In addition, mere aggregation of another work not based on the Library -with the Library (or with a work based on the Library) on a volume of -a storage or distribution medium does not bring the other work under -the scope of this License. - - 3. You may opt to apply the terms of the ordinary GNU General Public -License instead of this License to a given copy of the Library. To do -this, you must alter all the notices that refer to this License, so -that they refer to the ordinary GNU General Public License, version 2, -instead of to this License. (If a newer version than version 2 of the -ordinary GNU General Public License has appeared, then you can specify -that version instead if you wish.) Do not make any other change in -these notices. - - Once this change is made in a given copy, it is irreversible for -that copy, so the ordinary GNU General Public License applies to all -subsequent copies and derivative works made from that copy. - - This option is useful when you wish to copy part of the code of -the Library into a program that is not a library. - - 4. You may copy and distribute the Library (or a portion or -derivative of it, under Section 2) in object code or executable form -under the terms of Sections 1 and 2 above provided that you accompany -it with the complete corresponding machine-readable source code, which -must be distributed under the terms of Sections 1 and 2 above on a -medium customarily used for software interchange. - - If distribution of object code is made by offering access to copy -from a designated place, then offering equivalent access to copy the -source code from the same place satisfies the requirement to -distribute the source code, even though third parties are not -compelled to copy the source along with the object code. - - 5. A program that contains no derivative of any portion of the -Library, but is designed to work with the Library by being compiled or -linked with it, is called a "work that uses the Library". Such a -work, in isolation, is not a derivative work of the Library, and -therefore falls outside the scope of this License. - - However, linking a "work that uses the Library" with the Library -creates an executable that is a derivative of the Library (because it -contains portions of the Library), rather than a "work that uses the -library". The executable is therefore covered by this License. -Section 6 states terms for distribution of such executables. - - When a "work that uses the Library" uses material from a header file -that is part of the Library, the object code for the work may be a -derivative work of the Library even though the source code is not. -Whether this is true is especially significant if the work can be -linked without the Library, or if the work is itself a library. The -threshold for this to be true is not precisely defined by law. - - If such an object file uses only numerical parameters, data -structure layouts and accessors, and small macros and small inline -functions (ten lines or less in length), then the use of the object -file is unrestricted, regardless of whether it is legally a derivative -work. (Executables containing this object code plus portions of the -Library will still fall under Section 6.) - - Otherwise, if the work is a derivative of the Library, you may -distribute the object code for the work under the terms of Section 6. -Any executables containing that work also fall under Section 6, -whether or not they are linked directly with the Library itself. - - 6. As an exception to the Sections above, you may also combine or -link a "work that uses the Library" with the Library to produce a -work containing portions of the Library, and distribute that work -under terms of your choice, provided that the terms permit -modification of the work for the customer's own use and reverse -engineering for debugging such modifications. - - You must give prominent notice with each copy of the work that the -Library is used in it and that the Library and its use are covered by -this License. You must supply a copy of this License. If the work -during execution displays copyright notices, you must include the -copyright notice for the Library among them, as well as a reference -directing the user to the copy of this License. Also, you must do one -of these things: - - a) Accompany the work with the complete corresponding - machine-readable source code for the Library including whatever - changes were used in the work (which must be distributed under - Sections 1 and 2 above); and, if the work is an executable linked - with the Library, with the complete machine-readable "work that - uses the Library", as object code and/or source code, so that the - user can modify the Library and then relink to produce a modified - executable containing the modified Library. (It is understood - that the user who changes the contents of definitions files in the - Library will not necessarily be able to recompile the application - to use the modified definitions.) - - b) Use a suitable shared library mechanism for linking with the - Library. A suitable mechanism is one that (1) uses at run time a - copy of the library already present on the user's computer system, - rather than copying library functions into the executable, and (2) - will operate properly with a modified version of the library, if - the user installs one, as long as the modified version is - interface-compatible with the version that the work was made with. - - c) Accompany the work with a written offer, valid for at - least three years, to give the same user the materials - specified in Subsection 6a, above, for a charge no more - than the cost of performing this distribution. - - d) If distribution of the work is made by offering access to copy - from a designated place, offer equivalent access to copy the above - specified materials from the same place. - - e) Verify that the user has already received a copy of these - materials or that you have already sent this user a copy. - - For an executable, the required form of the "work that uses the -Library" must include any data and utility programs needed for -reproducing the executable from it. However, as a special exception, -the materials to be distributed need not include anything that is -normally distributed (in either source or binary form) with the major -components (compiler, kernel, and so on) of the operating system on -which the executable runs, unless that component itself accompanies -the executable. - - It may happen that this requirement contradicts the license -restrictions of other proprietary libraries that do not normally -accompany the operating system. Such a contradiction means you cannot -use both them and the Library together in an executable that you -distribute. - - 7. You may place library facilities that are a work based on the -Library side-by-side in a single library together with other library -facilities not covered by this License, and distribute such a combined -library, provided that the separate distribution of the work based on -the Library and of the other library facilities is otherwise -permitted, and provided that you do these two things: - - a) Accompany the combined library with a copy of the same work - based on the Library, uncombined with any other library - facilities. This must be distributed under the terms of the - Sections above. - - b) Give prominent notice with the combined library of the fact - that part of it is a work based on the Library, and explaining - where to find the accompanying uncombined form of the same work. - - 8. You may not copy, modify, sublicense, link with, or distribute -the Library except as expressly provided under this License. Any -attempt otherwise to copy, modify, sublicense, link with, or -distribute the Library is void, and will automatically terminate your -rights under this License. However, parties who have received copies, -or rights, from you under this License will not have their licenses -terminated so long as such parties remain in full compliance. - - 9. You are not required to accept this License, since you have not -signed it. However, nothing else grants you permission to modify or -distribute the Library or its derivative works. These actions are -prohibited by law if you do not accept this License. Therefore, by -modifying or distributing the Library (or any work based on the -Library), you indicate your acceptance of this License to do so, and -all its terms and conditions for copying, distributing or modifying -the Library or works based on it. - - 10. Each time you redistribute the Library (or any work based on the -Library), the recipient automatically receives a license from the -original licensor to copy, distribute, link with or modify the Library -subject to these terms and conditions. You may not impose any further -restrictions on the recipients' exercise of the rights granted herein. -You are not responsible for enforcing compliance by third parties with -this License. - - 11. If, as a consequence of a court judgment or allegation of patent -infringement or for any other reason (not limited to patent issues), -conditions are imposed on you (whether by court order, agreement or -otherwise) that contradict the conditions of this License, they do not -excuse you from the conditions of this License. If you cannot -distribute so as to satisfy simultaneously your obligations under this -License and any other pertinent obligations, then as a consequence you -may not distribute the Library at all. For example, if a patent -license would not permit royalty-free redistribution of the Library by -all those who receive copies directly or indirectly through you, then -the only way you could satisfy both it and this License would be to -refrain entirely from distribution of the Library. - -If any portion of this section is held invalid or unenforceable under any -particular circumstance, the balance of the section is intended to apply, -and the section as a whole is intended to apply in other circumstances. - -It is not the purpose of this section to induce you to infringe any -patents or other property right claims or to contest validity of any -such claims; this section has the sole purpose of protecting the -integrity of the free software distribution system which is -implemented by public license practices. Many people have made -generous contributions to the wide range of software distributed -through that system in reliance on consistent application of that -system; it is up to the author/donor to decide if he or she is willing -to distribute software through any other system and a licensee cannot -impose that choice. - -This section is intended to make thoroughly clear what is believed to -be a consequence of the rest of this License. - - 12. If the distribution and/or use of the Library is restricted in -certain countries either by patents or by copyrighted interfaces, the -original copyright holder who places the Library under this License may add -an explicit geographical distribution limitation excluding those countries, -so that distribution is permitted only in or among countries not thus -excluded. In such case, this License incorporates the limitation as if -written in the body of this License. - - 13. The Free Software Foundation may publish revised and/or new -versions of the Lesser General Public License from time to time. -Such new versions will be similar in spirit to the present version, -but may differ in detail to address new problems or concerns. - -Each version is given a distinguishing version number. If the Library -specifies a version number of this License which applies to it and -"any later version", you have the option of following the terms and -conditions either of that version or of any later version published by -the Free Software Foundation. If the Library does not specify a -license version number, you may choose any version ever published by -the Free Software Foundation. - - 14. If you wish to incorporate parts of the Library into other free -programs whose distribution conditions are incompatible with these, -write to the author to ask for permission. For software which is -copyrighted by the Free Software Foundation, write to the Free -Software Foundation; we sometimes make exceptions for this. Our -decision will be guided by the two goals of preserving the free status -of all derivatives of our free software and of promoting the sharing -and reuse of software generally. - - NO WARRANTY - - 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO -WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW. -EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR -OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY -KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR -PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE -LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME -THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. - - 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN -WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY -AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU -FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR -CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE -LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING -RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A -FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF -SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH -DAMAGES. - - END OF TERMS AND CONDITIONS - - How to Apply These Terms to Your New Libraries - - If you develop a new library, and you want it to be of the greatest -possible use to the public, we recommend making it free software that -everyone can redistribute and change. You can do so by permitting -redistribution under these terms (or, alternatively, under the terms of the -ordinary General Public License). - - To apply these terms, attach the following notices to the library. It is -safest to attach them to the start of each source file to most effectively -convey the exclusion of warranty; and each file should have at least the -"copyright" line and a pointer to where the full notice is found. - - - Copyright (C) - - This library is free software; you can redistribute it and/or - modify it under the terms of the GNU Lesser General Public - License as published by the Free Software Foundation; either - version 2.1 of the License, or (at your option) any later version. - - This library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public - License along with this library; if not, write to the Free Software - Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - -Also add information on how to contact you by electronic and paper mail. - -You should also get your employer (if you work as a programmer) or your -school, if any, to sign a "copyright disclaimer" for the library, if -necessary. Here is a sample; alter the names: - - Yoyodyne, Inc., hereby disclaims all copyright interest in the - library `Frob' (a library for tweaking knobs) written by James Random Hacker. - - , 1 April 1990 - Ty Coon, President of Vice - -That's all there is to it! + This version of the GNU Lesser General Public License incorporates +the terms and conditions of version 3 of the GNU General Public +License, supplemented by the additional permissions listed below. + + 0. Additional Definitions. + + As used herein, "this License" refers to version 3 of the GNU Lesser +General Public License, and the "GNU GPL" refers to version 3 of the GNU +General Public License. + + "The Library" refers to a covered work governed by this License, +other than an Application or a Combined Work as defined below. + + An "Application" is any work that makes use of an interface provided +by the Library, but which is not otherwise based on the Library. +Defining a subclass of a class defined by the Library is deemed a mode +of using an interface provided by the Library. + + A "Combined Work" is a work produced by combining or linking an +Application with the Library. The particular version of the Library +with which the Combined Work was made is also called the "Linked +Version". + + The "Minimal Corresponding Source" for a Combined Work means the +Corresponding Source for the Combined Work, excluding any source code +for portions of the Combined Work that, considered in isolation, are +based on the Application, and not on the Linked Version. + + The "Corresponding Application Code" for a Combined Work means the +object code and/or source code for the Application, including any data +and utility programs needed for reproducing the Combined Work from the +Application, but excluding the System Libraries of the Combined Work. + + 1. Exception to Section 3 of the GNU GPL. + + You may convey a covered work under sections 3 and 4 of this License +without being bound by section 3 of the GNU GPL. + + 2. Conveying Modified Versions. + + If you modify a copy of the Library, and, in your modifications, a +facility refers to a function or data to be supplied by an Application +that uses the facility (other than as an argument passed when the +facility is invoked), then you may convey a copy of the modified +version: + + a) under this License, provided that you make a good faith effort to + ensure that, in the event an Application does not supply the + function or data, the facility still operates, and performs + whatever part of its purpose remains meaningful, or + + b) under the GNU GPL, with none of the additional permissions of + this License applicable to that copy. + + 3. Object Code Incorporating Material from Library Header Files. + + The object code form of an Application may incorporate material from +a header file that is part of the Library. You may convey such object +code under terms of your choice, provided that, if the incorporated +material is not limited to numerical parameters, data structure +layouts and accessors, or small macros, inline functions and templates +(ten or fewer lines in length), you do both of the following: + + a) Give prominent notice with each copy of the object code that the + Library is used in it and that the Library and its use are + covered by this License. + + b) Accompany the object code with a copy of the GNU GPL and this license + document. + + 4. Combined Works. + + You may convey a Combined Work under terms of your choice that, +taken together, effectively do not restrict modification of the +portions of the Library contained in the Combined Work and reverse +engineering for debugging such modifications, if you also do each of +the following: + + a) Give prominent notice with each copy of the Combined Work that + the Library is used in it and that the Library and its use are + covered by this License. + + b) Accompany the Combined Work with a copy of the GNU GPL and this license + document. + + c) For a Combined Work that displays copyright notices during + execution, include the copyright notice for the Library among + these notices, as well as a reference directing the user to the + copies of the GNU GPL and this license document. + + d) Do one of the following: + + 0) Convey the Minimal Corresponding Source under the terms of this + License, and the Corresponding Application Code in a form + suitable for, and under terms that permit, the user to + recombine or relink the Application with a modified version of + the Linked Version to produce a modified Combined Work, in the + manner specified by section 6 of the GNU GPL for conveying + Corresponding Source. + + 1) Use a suitable shared library mechanism for linking with the + Library. A suitable mechanism is one that (a) uses at run time + a copy of the Library already present on the user's computer + system, and (b) will operate properly with a modified version + of the Library that is interface-compatible with the Linked + Version. + + e) Provide Installation Information, but only if you would otherwise + be required to provide such information under section 6 of the + GNU GPL, and only to the extent that such information is + necessary to install and execute a modified version of the + Combined Work produced by recombining or relinking the + Application with a modified version of the Linked Version. (If + you use option 4d0, the Installation Information must accompany + the Minimal Corresponding Source and Corresponding Application + Code. If you use option 4d1, you must provide the Installation + Information in the manner specified by section 6 of the GNU GPL + for conveying Corresponding Source.) + + 5. Combined Libraries. + + You may place library facilities that are a work based on the +Library side by side in a single library together with other library +facilities that are not Applications and are not covered by this +License, and convey such a combined library under terms of your +choice, if you do both of the following: + + a) Accompany the combined library with a copy of the same work based + on the Library, uncombined with any other library facilities, + conveyed under the terms of this License. + + b) Give prominent notice with the combined library that part of it + is a work based on the Library, and explaining where to find the + accompanying uncombined form of the same work. + + 6. Revised Versions of the GNU Lesser General Public License. + + The Free Software Foundation may publish revised and/or new versions +of the GNU Lesser General Public License from time to time. Such new +versions will be similar in spirit to the present version, but may +differ in detail to address new problems or concerns. + + Each version is given a distinguishing version number. If the +Library as you received it specifies that a certain numbered version +of the GNU Lesser General Public License "or any later version" +applies to it, you have the option of following the terms and +conditions either of that published version or of any later version +published by the Free Software Foundation. If the Library as you +received it does not specify a version number of the GNU Lesser +General Public License, you may choose any version of the GNU Lesser +General Public License ever published by the Free Software Foundation. + + If the Library as you received it specifies that a proxy can decide +whether future versions of the GNU Lesser General Public License shall +apply, that proxy's public statement of acceptance of any version is +permanent authorization for you to choose that version for the Library. diff --git a/plugins/spu2-x/License.txt b/plugins/spu2-x/License.txt index 4f562f9be8..fc02b8d6df 100644 --- a/plugins/spu2-x/License.txt +++ b/plugins/spu2-x/License.txt @@ -1,6 +1,7 @@ [This file contains the template for the SPU2-X code rights license. For the full - rant-like preamble of the LGP, see LGPL.TXT.] + rant-like preamble of the LGPL, see GPL.txt and LGPL.txt -- both lisences apply, + with the LGPL as a subset extension of the GPL.] /* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 @@ -8,18 +9,14 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ \ No newline at end of file diff --git a/plugins/spu2-x/src/ADSR.cpp b/plugins/spu2-x/src/ADSR.cpp index 39fb3549f3..e11925b2d0 100644 --- a/plugins/spu2-x/src/ADSR.cpp +++ b/plugins/spu2-x/src/ADSR.cpp @@ -1,23 +1,19 @@ /* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 -* Developed and maintained by the Pcsx2 Development Team. -* -* Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] -* -* This library is free software; you can redistribute it and/or modify it under -* the terms of the GNU Lesser General Public License as published by the Free -* Software Foundation; either version 2.1 of the the License, or (at your -* option) any later version. -* -* This library is distributed in the hope that it will be useful, but WITHOUT -* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS -* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License -* for more details. -* -* You should have received a copy of the GNU Lesser General Public License along -* with this library; if not, write to the Free Software Foundation, Inc., 59 -* Temple Place, Suite 330, Boston, MA 02111-1307 USA -* -*/ + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/Debug.cpp b/plugins/spu2-x/src/Debug.cpp index dbf6293c91..404db45524 100644 --- a/plugins/spu2-x/src/Debug.cpp +++ b/plugins/spu2-x/src/Debug.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/DllInterface.cpp b/plugins/spu2-x/src/DllInterface.cpp index 33527ad774..e6d39bb691 100644 --- a/plugins/spu2-x/src/DllInterface.cpp +++ b/plugins/spu2-x/src/DllInterface.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/Dma.cpp b/plugins/spu2-x/src/Dma.cpp index 31edd19ffc..8b2d4b102c 100644 --- a/plugins/spu2-x/src/Dma.cpp +++ b/plugins/spu2-x/src/Dma.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/Linux/Alsa.cpp b/plugins/spu2-x/src/Linux/Alsa.cpp index 97df382e53..8f6cfa0484 100644 --- a/plugins/spu2-x/src/Linux/Alsa.cpp +++ b/plugins/spu2-x/src/Linux/Alsa.cpp @@ -1,48 +1,45 @@ -/* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 - * Developed and maintained by the Pcsx2 Development Team. - * - * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] - * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ - // Adapted from ZeroSPU2 code by Zerofrog. Heavily modified by Arcum42. - -#include - -#define ALSA_MEM_DEF -#include "Alsa.h" -#include "SndOut.h" - +/* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ + + // Adapted from ZeroSPU2 code by Zerofrog. Heavily modified by Arcum42. + +#include + +#define ALSA_MEM_DEF +#include "Alsa.h" +#include "SndOut.h" + class AlsaMod: public SndOutModule { protected: static const int PacketsPerBuffer = 1; // increase this if ALSA can't keep up with 512-sample packets static const int MAX_BUFFER_COUNT = 4; - static const int NumBuffers = 4; // TODO: this should be configurable someday -- lower values reduce latency. + static const int NumBuffers = 4; // TODO: this should be configurable someday -- lower values reduce latency. unsigned int pspeed; - - snd_pcm_t *handle; - snd_pcm_uframes_t buffer_size; + + snd_pcm_t *handle; + snd_pcm_uframes_t buffer_size; snd_async_handler_t *pcm_callback; - - uint period_time; - uint buffer_time; - -protected: - // Invoked by the static ExternalCallback method below. + + uint period_time; + uint buffer_time; + +protected: + // Invoked by the static ExternalCallback method below. void _InternalCallback() { snd_pcm_sframes_t avail; @@ -61,148 +58,148 @@ protected: avail = snd_pcm_avail_update(handle); } } - - // Preps and invokes the _InternalCallback above. This provides a cdecl-compliant - // entry point for our C++ified object state. :) - static void ExternalCallback( snd_async_handler_t *pcm_callback ) - { + + // Preps and invokes the _InternalCallback above. This provides a cdecl-compliant + // entry point for our C++ified object state. :) + static void ExternalCallback( snd_async_handler_t *pcm_callback ) + { AlsaMod *data = (AlsaMod*)snd_async_handler_get_callback_private( pcm_callback ); - - jASSUME( data != NULL ); - jASSUME( data->handle == snd_async_handler_get_pcm(pcm_callback) ); - - // Not sure if we just need an assert, or something like this: - //if( data->handle != snd_async_handler_get_pcm(pcm_callback) ) return; - - data->_InternalCallback(); - } - -public: - - s32 Init() - { - snd_pcm_hw_params_t *hwparams; - snd_pcm_sw_params_t *swparams; - snd_pcm_status_t *status; - int pchannels = 2; - snd_pcm_format_t format = SND_PCM_FORMAT_S16_LE; - - handle = NULL; - pcm_callback = NULL; - pspeed = SAMPLE_RATE; - - // buffer time and period time are in microseconds... - // (don't simplify the equation below -- it'll just cause integer rounding errors. - period_time = (SndOutPacketSize*1000) / (SampleRate / 1000); - buffer_time = period_time * NumBuffers; - - int err; - - err = snd_pcm_open(&handle, "default", SND_PCM_STREAM_PLAYBACK, SND_PCM_NONBLOCK); - if(err < 0) - { - ERROR_LOG("Audio open error: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_nonblock(handle, 0); - if(err < 0) - { - ERROR_LOG("Can't set blocking mode: %s\n", snd_strerror(err)); - return -1; - } - - snd_pcm_hw_params_alloca(&hwparams); - snd_pcm_sw_params_alloca(&swparams); - - err = snd_pcm_hw_params_any(handle, hwparams); - if (err < 0) - { - ERROR_LOG("Broken configuration for this PCM: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params_set_access(handle, hwparams, SND_PCM_ACCESS_RW_INTERLEAVED); - if (err < 0) - { - ERROR_LOG("Access type not available: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params_set_format(handle, hwparams, format); - if (err < 0) - { - ERROR_LOG("Sample format not available: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params_set_channels(handle, hwparams, pchannels); - if (err < 0) - { - ERROR_LOG("Channels count not available: %s\n", snd_strerror(err)); - return -1; - } - err = snd_pcm_hw_params_set_rate_near(handle, hwparams, &pspeed, 0); - if (err < 0) - { - ERROR_LOG("Rate not available: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params_set_buffer_time_near(handle, hwparams, &buffer_time, 0); - if(err < 0) { - ERROR_LOG("Buffer time error: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params_set_period_time_near(handle, hwparams, &period_time, 0); - if (err < 0) - { - ERROR_LOG("Period time error: %s\n", snd_strerror(err)); - return -1; - } - - err = snd_pcm_hw_params(handle, hwparams); - if (err < 0) - { - ERROR_LOG("Unable to install hw params: %s\n", snd_strerror(err)); - return -1; - } - - snd_pcm_status_alloca(&status); - err = snd_pcm_status(handle, status); - if(err < 0) - { - ERROR_LOG("Unable to get status: %s\n", snd_strerror(err)); - return -1; - } - + + jASSUME( data != NULL ); + jASSUME( data->handle == snd_async_handler_get_pcm(pcm_callback) ); + + // Not sure if we just need an assert, or something like this: + //if( data->handle != snd_async_handler_get_pcm(pcm_callback) ) return; + + data->_InternalCallback(); + } + +public: + + s32 Init() + { + snd_pcm_hw_params_t *hwparams; + snd_pcm_sw_params_t *swparams; + snd_pcm_status_t *status; + int pchannels = 2; + snd_pcm_format_t format = SND_PCM_FORMAT_S16_LE; + + handle = NULL; + pcm_callback = NULL; + pspeed = SAMPLE_RATE; + + // buffer time and period time are in microseconds... + // (don't simplify the equation below -- it'll just cause integer rounding errors. + period_time = (SndOutPacketSize*1000) / (SampleRate / 1000); + buffer_time = period_time * NumBuffers; + + int err; + + err = snd_pcm_open(&handle, "default", SND_PCM_STREAM_PLAYBACK, SND_PCM_NONBLOCK); + if(err < 0) + { + ERROR_LOG("Audio open error: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_nonblock(handle, 0); + if(err < 0) + { + ERROR_LOG("Can't set blocking mode: %s\n", snd_strerror(err)); + return -1; + } + + snd_pcm_hw_params_alloca(&hwparams); + snd_pcm_sw_params_alloca(&swparams); + + err = snd_pcm_hw_params_any(handle, hwparams); + if (err < 0) + { + ERROR_LOG("Broken configuration for this PCM: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params_set_access(handle, hwparams, SND_PCM_ACCESS_RW_INTERLEAVED); + if (err < 0) + { + ERROR_LOG("Access type not available: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params_set_format(handle, hwparams, format); + if (err < 0) + { + ERROR_LOG("Sample format not available: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params_set_channels(handle, hwparams, pchannels); + if (err < 0) + { + ERROR_LOG("Channels count not available: %s\n", snd_strerror(err)); + return -1; + } + err = snd_pcm_hw_params_set_rate_near(handle, hwparams, &pspeed, 0); + if (err < 0) + { + ERROR_LOG("Rate not available: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params_set_buffer_time_near(handle, hwparams, &buffer_time, 0); + if(err < 0) { + ERROR_LOG("Buffer time error: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params_set_period_time_near(handle, hwparams, &period_time, 0); + if (err < 0) + { + ERROR_LOG("Period time error: %s\n", snd_strerror(err)); + return -1; + } + + err = snd_pcm_hw_params(handle, hwparams); + if (err < 0) + { + ERROR_LOG("Unable to install hw params: %s\n", snd_strerror(err)); + return -1; + } + + snd_pcm_status_alloca(&status); + err = snd_pcm_status(handle, status); + if(err < 0) + { + ERROR_LOG("Unable to get status: %s\n", snd_strerror(err)); + return -1; + } + // Bind our asynchronous callback magic: snd_async_add_pcm_handler( &pcm_callback, handle, ExternalCallback, this ); - - snd_pcm_start( handle ); - // Diagnostic code: - //buffer_size = snd_pcm_status_get_avail(status); - - return 0; - } - - void Close() - { - if(handle == NULL) return; - - snd_pcm_drop(handle); - snd_pcm_close(handle); - handle = NULL; - } - + + snd_pcm_start( handle ); + // Diagnostic code: + //buffer_size = snd_pcm_status_get_avail(status); + + return 0; + } + + void Close() + { + if(handle == NULL) return; + + snd_pcm_drop(handle); + snd_pcm_close(handle); + handle = NULL; + } + virtual void Configure(uptr parent) { } - virtual bool Is51Out() const { return false; } - + virtual bool Is51Out() const { return false; } + s32 Test() const { return 0; @@ -210,12 +207,12 @@ public: int GetEmptySampleCount() const { - if(handle == NULL) return 0; - - // Returns the amount of free buffer space, in samples. - uint l = snd_pcm_avail_update(handle); - if( l < 0 ) return 0; - return (l / 1000) * (SampleRate / 1000); + if(handle == NULL) return 0; + + // Returns the amount of free buffer space, in samples. + uint l = snd_pcm_avail_update(handle); + if( l < 0 ) return 0; + return (l / 1000) * (SampleRate / 1000); } const wchar_t* GetIdent() const diff --git a/plugins/spu2-x/src/Linux/Config.cpp b/plugins/spu2-x/src/Linux/Config.cpp index 1be745c561..8853b4346f 100644 --- a/plugins/spu2-x/src/Linux/Config.cpp +++ b/plugins/spu2-x/src/Linux/Config.cpp @@ -3,21 +3,17 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #include "Spu2.h" #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Linux/Config.h b/plugins/spu2-x/src/Linux/Config.h index ae3a159d8e..c7c6afe6a0 100644 --- a/plugins/spu2-x/src/Linux/Config.h +++ b/plugins/spu2-x/src/Linux/Config.h @@ -3,21 +3,17 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #ifndef CONFIG_H_INCLUDED #define CONFIG_H_INCLUDED diff --git a/plugins/spu2-x/src/Linux/Dialogs.cpp b/plugins/spu2-x/src/Linux/Dialogs.cpp index 5746dbd0d2..dfe8ea4b08 100644 --- a/plugins/spu2-x/src/Linux/Dialogs.cpp +++ b/plugins/spu2-x/src/Linux/Dialogs.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ // To be continued... \ No newline at end of file diff --git a/plugins/spu2-x/src/Linux/Dialogs.h b/plugins/spu2-x/src/Linux/Dialogs.h index 13348bd283..d8266ff3bd 100644 --- a/plugins/spu2-x/src/Linux/Dialogs.h +++ b/plugins/spu2-x/src/Linux/Dialogs.h @@ -3,21 +3,17 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #ifndef DIALOG_H_INCLUDED #define DIALOG_H_INCLUDED diff --git a/plugins/spu2-x/src/Linux/Linux.cpp b/plugins/spu2-x/src/Linux/Linux.cpp index c926b448cb..443917f845 100644 --- a/plugins/spu2-x/src/Linux/Linux.cpp +++ b/plugins/spu2-x/src/Linux/Linux.cpp @@ -3,21 +3,17 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #include "Linux.h" diff --git a/plugins/spu2-x/src/Linux/Linux.h b/plugins/spu2-x/src/Linux/Linux.h index ac3438fb51..0ba3c0c0cb 100644 --- a/plugins/spu2-x/src/Linux/Linux.h +++ b/plugins/spu2-x/src/Linux/Linux.h @@ -3,21 +3,17 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * - */ + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #ifndef __LINUX_H__ #define __LINUX_H__ diff --git a/plugins/spu2-x/src/Lowpass.cpp b/plugins/spu2-x/src/Lowpass.cpp index f862f32da5..c6d6e84fb0 100644 --- a/plugins/spu2-x/src/Lowpass.cpp +++ b/plugins/spu2-x/src/Lowpass.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Lowpass.h" diff --git a/plugins/spu2-x/src/Lowpass.h b/plugins/spu2-x/src/Lowpass.h index 13f0d742ce..23dc5f8380 100644 --- a/plugins/spu2-x/src/Lowpass.h +++ b/plugins/spu2-x/src/Lowpass.h @@ -1,24 +1,21 @@ -//GiGaHeRz's SPU2 Driver -//Copyright (c) 2003-2008, David Quintana -// -//This library is free software; you can redistribute it and/or -//modify it under the terms of the GNU Lesser General Public -//License as published by the Free Software Foundation; either -//version 2.1 of the License, or (at your option) any later version. -// -//This library is distributed in the hope that it will be useful, -//but WITHOUT ANY WARRANTY; without even the implied warranty of -//MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -//Lesser General Public License for more details. -// -//You should have received a copy of the GNU Lesser General Public -//License along with this library; if not, write to the Free Software -//Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -// -#pragma once +/* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 +* Developed and maintained by the Pcsx2 Development Team. +* +* Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] +* +* SPU2-X is free software: you can redistribute it and/or modify it under the terms +* of the GNU Lesser General Public License as published by the Free Software Found- +* ation, either version 3 of the License, or (at your option) any later version. +* +* SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; +* without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR +* PURPOSE. See the GNU Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public License +* along with SPU2-X. If not, see . +*/ -#ifndef _LOWPASS_H_ -#define _LOWPASS_H_ +#pragma once #include "BaseTypes.h" @@ -31,4 +28,3 @@ struct LPF_data double sample( double inval ); }; -#endif diff --git a/plugins/spu2-x/src/Mixer.cpp b/plugins/spu2-x/src/Mixer.cpp index b8debff5ee..066444f7e4 100644 --- a/plugins/spu2-x/src/Mixer.cpp +++ b/plugins/spu2-x/src/Mixer.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/ReadInput.cpp b/plugins/spu2-x/src/ReadInput.cpp index eaa50d1f50..b6ab222c86 100644 --- a/plugins/spu2-x/src/ReadInput.cpp +++ b/plugins/spu2-x/src/ReadInput.cpp @@ -1,23 +1,19 @@ /* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 -* Developed and maintained by the Pcsx2 Development Team. -* -* Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] -* -* This library is free software; you can redistribute it and/or modify it under -* the terms of the GNU Lesser General Public License as published by the Free -* Software Foundation; either version 2.1 of the the License, or (at your -* option) any later version. -* -* This library is distributed in the hope that it will be useful, but WITHOUT -* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS -* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License -* for more details. -* -* You should have received a copy of the GNU Lesser General Public License along -* with this library; if not, write to the Free Software Foundation, Inc., 59 -* Temple Place, Suite 330, Boston, MA 02111-1307 USA -* -*/ + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/RegLog.cpp b/plugins/spu2-x/src/RegLog.cpp index c240141270..7c8d58808e 100644 --- a/plugins/spu2-x/src/RegLog.cpp +++ b/plugins/spu2-x/src/RegLog.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/RegTable.cpp b/plugins/spu2-x/src/RegTable.cpp index 9ecf2e7f82..423a9cdad8 100644 --- a/plugins/spu2-x/src/RegTable.cpp +++ b/plugins/spu2-x/src/RegTable.cpp @@ -1,20 +1,20 @@ -//GiGaHeRz's SPU2 Driver -//Copyright (c) 2003-2008, David Quintana -// -//This library is free software; you can redistribute it and/or -//modify it under the terms of the GNU Lesser General Public -//License as published by the Free Software Foundation; either -//version 2.1 of the License, or (at your option) any later version. -// -//This library is distributed in the hope that it will be useful, -//but WITHOUT ANY WARRANTY; without even the implied warranty of -//MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -//Lesser General Public License for more details. -// -//You should have received a copy of the GNU Lesser General Public -//License along with this library; if not, write to the Free Software -//Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -// +/* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ + #include "Spu2.h" #include "RegTable.h" diff --git a/plugins/spu2-x/src/Reverb.cpp b/plugins/spu2-x/src/Reverb.cpp index 7f70c3d7a8..a2d91e2727 100644 --- a/plugins/spu2-x/src/Reverb.cpp +++ b/plugins/spu2-x/src/Reverb.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/SaveStateSPU.cpp b/plugins/spu2-x/src/SaveStateSPU.cpp index 8dc61bd452..5b19f5b59b 100644 --- a/plugins/spu2-x/src/SaveStateSPU.cpp +++ b/plugins/spu2-x/src/SaveStateSPU.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/SndOut.cpp b/plugins/spu2-x/src/SndOut.cpp index 55f035e692..0fcb169080 100644 --- a/plugins/spu2-x/src/SndOut.cpp +++ b/plugins/spu2-x/src/SndOut.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/Spu2.cpp b/plugins/spu2-x/src/Spu2.cpp index 9e23e25578..7928a8f0f2 100644 --- a/plugins/spu2-x/src/Spu2.cpp +++ b/plugins/spu2-x/src/Spu2.cpp @@ -3,23 +3,18 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ - #include "Spu2.h" #include "RegTable.h" diff --git a/plugins/spu2-x/src/Spu2.h b/plugins/spu2-x/src/Spu2.h index f04b797450..98fa15a3e1 100644 --- a/plugins/spu2-x/src/Spu2.h +++ b/plugins/spu2-x/src/Spu2.h @@ -1,26 +1,19 @@ /* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 * Developed and maintained by the Pcsx2 Development Team. * - * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ -#ifndef SPU2_H_INCLUDED -#define SPU2_H_INCLUDED +#pragma once #include "BaseTypes.h" #include "PS2Edefs.h" @@ -215,5 +208,3 @@ extern StereoOut32 DoReverb( V_Core& thiscore, const StereoOut32& Input ); extern s32 MulShr32( s32 srcval, s32 mulval ); //#define PCM24_S1_INTERLEAVE - -#endif // SPU2_H_INCLUDED // diff --git a/plugins/spu2-x/src/Spu2replay.h b/plugins/spu2-x/src/Spu2replay.h index 46efa0abe3..747a8adce0 100644 --- a/plugins/spu2-x/src/Spu2replay.h +++ b/plugins/spu2-x/src/Spu2replay.h @@ -16,8 +16,7 @@ //Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA // -#ifndef SPU2REPLAY_H_INCLUDED -#define SPU2REPLAY_H_INCLUDED +#pragma once // s2r dumping int s2r_open(char *filename); @@ -35,5 +34,3 @@ EXPORT_C_(void) s2r_replay(HWND hwnd, HINSTANCE hinst, LPSTR filename, int nCmdS #endif extern bool replay_mode; - -#endif//SPU2REPLAY_H_INCLUDED \ No newline at end of file diff --git a/plugins/spu2-x/src/Timestretcher.cpp b/plugins/spu2-x/src/Timestretcher.cpp index 08ab039337..d347bbe350 100644 --- a/plugins/spu2-x/src/Timestretcher.cpp +++ b/plugins/spu2-x/src/Timestretcher.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * -* This library is free software; you can redistribute it and/or modify it under -* the terms of the GNU Lesser General Public License as published by the Free -* Software Foundation; either version 2.1 of the the License, or (at your -* option) any later version. -* -* This library is distributed in the hope that it will be useful, but WITHOUT -* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS -* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License -* for more details. -* -* You should have received a copy of the GNU Lesser General Public License along -* with this library; if not, write to the Free Software Foundation, Inc., 59 -* Temple Place, Suite 330, Boston, MA 02111-1307 USA -* +* SPU2-X is free software: you can redistribute it and/or modify it under the terms +* of the GNU Lesser General Public License as published by the Free Software Found- +* ation, either version 3 of the License, or (at your option) any later version. +* +* SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; +* without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR +* PURPOSE. See the GNU Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public License +* along with SPU2-X. If not, see . */ #include "Spu2.h" diff --git a/plugins/spu2-x/src/Wavedump_wav.cpp b/plugins/spu2-x/src/Wavedump_wav.cpp index cfddf7cabb..bc09a67b51 100644 --- a/plugins/spu2-x/src/Wavedump_wav.cpp +++ b/plugins/spu2-x/src/Wavedump_wav.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include diff --git a/plugins/spu2-x/src/Windows/AboutBox.cpp b/plugins/spu2-x/src/Windows/AboutBox.cpp index a485862f9c..3049172f1b 100644 --- a/plugins/spu2-x/src/Windows/AboutBox.cpp +++ b/plugins/spu2-x/src/Windows/AboutBox.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/CfgHelpers.cpp b/plugins/spu2-x/src/Windows/CfgHelpers.cpp index ac4dad8159..b0b48b7a91 100644 --- a/plugins/spu2-x/src/Windows/CfgHelpers.cpp +++ b/plugins/spu2-x/src/Windows/CfgHelpers.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/Config.cpp b/plugins/spu2-x/src/Windows/Config.cpp index b400380b9c..5c3e6d6f9a 100644 --- a/plugins/spu2-x/src/Windows/Config.cpp +++ b/plugins/spu2-x/src/Windows/Config.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/Config.h b/plugins/spu2-x/src/Windows/Config.h index bd1ff56079..24ae24c2ba 100644 --- a/plugins/spu2-x/src/Windows/Config.h +++ b/plugins/spu2-x/src/Windows/Config.h @@ -3,24 +3,19 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ -#ifndef CONFIG_H_INCLUDED -#define CONFIG_H_INCLUDED +#pragma once #include @@ -150,5 +145,3 @@ void ReadSettings(); void WriteSettings(); void configure(); void AboutBox(); - -#endif // CONFIG_H_INCLUDED diff --git a/plugins/spu2-x/src/Windows/ConfigDebug.cpp b/plugins/spu2-x/src/Windows/ConfigDebug.cpp index e724bd7f25..0373c225c4 100644 --- a/plugins/spu2-x/src/Windows/ConfigDebug.cpp +++ b/plugins/spu2-x/src/Windows/ConfigDebug.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/ConfigSoundtouch.cpp b/plugins/spu2-x/src/Windows/ConfigSoundtouch.cpp index 1e2c9d6fa0..781ef22002 100644 --- a/plugins/spu2-x/src/Windows/ConfigSoundtouch.cpp +++ b/plugins/spu2-x/src/Windows/ConfigSoundtouch.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/Dialogs.h b/plugins/spu2-x/src/Windows/Dialogs.h index 414a71db56..72065d4d03 100644 --- a/plugins/spu2-x/src/Windows/Dialogs.h +++ b/plugins/spu2-x/src/Windows/Dialogs.h @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #pragma once diff --git a/plugins/spu2-x/src/Windows/RealtimeDebugger.cpp b/plugins/spu2-x/src/Windows/RealtimeDebugger.cpp index 0aa82ffd83..5c4e0da69a 100644 --- a/plugins/spu2-x/src/Windows/RealtimeDebugger.cpp +++ b/plugins/spu2-x/src/Windows/RealtimeDebugger.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ diff --git a/plugins/spu2-x/src/Windows/SndOut_DSound.cpp b/plugins/spu2-x/src/Windows/SndOut_DSound.cpp index 1715da1131..9d89a78aa0 100644 --- a/plugins/spu2-x/src/Windows/SndOut_DSound.cpp +++ b/plugins/spu2-x/src/Windows/SndOut_DSound.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #define _WIN32_DCOM diff --git a/plugins/spu2-x/src/Windows/SndOut_XAudio2.cpp b/plugins/spu2-x/src/Windows/SndOut_XAudio2.cpp index d6a2e4089b..ecfee395c4 100644 --- a/plugins/spu2-x/src/Windows/SndOut_XAudio2.cpp +++ b/plugins/spu2-x/src/Windows/SndOut_XAudio2.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #define _WIN32_DCOM diff --git a/plugins/spu2-x/src/Windows/SndOut_waveOut.cpp b/plugins/spu2-x/src/Windows/SndOut_waveOut.cpp index ed93c09520..6ef8272f68 100644 --- a/plugins/spu2-x/src/Windows/SndOut_waveOut.cpp +++ b/plugins/spu2-x/src/Windows/SndOut_waveOut.cpp @@ -3,20 +3,16 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ #include "Dialogs.h" diff --git a/plugins/spu2-x/src/Windows/Spu2-X.def b/plugins/spu2-x/src/Windows/Spu2-X.def index e2a13e4a3c..73f00f60f2 100644 --- a/plugins/spu2-x/src/Windows/Spu2-X.def +++ b/plugins/spu2-x/src/Windows/Spu2-X.def @@ -3,19 +3,16 @@ ; * ; * Originally based on SPU2ghz v1.9 beta, by David Quintana. ; * -; * This library is free software; you can redistribute it and/or modify it under -; * the terms of the GNU Lesser General Public License as published by the Free -; * Software Foundation; either version 2.1 of the the License, or (at your -; * option) any later version. -; * -; * This library is distributed in the hope that it will be useful, but WITHOUT -; * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS -; * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License -; * for more details. -; * -; * You should have received a copy of the GNU Lesser General Public License along -; * with this library; if not, write to the Free Software Foundation, Inc., 59 -; * Temple Place, Suite 330, Boston, MA 02111-1307 USA +; * SPU2-X is free software: you can redistribute it and/or modify it under the terms +; * of the GNU Lesser General Public License as published by the Free Software Found- +; * ation, either version 3 of the License, or (at your option) any later version. +; * +; * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; +; * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR +; * PURPOSE. See the GNU Lesser General Public License for more details. +; * +; * You should have received a copy of the GNU Lesser General Public License +; * along with SPU2-X. If not, see . ; SPU2-X.def : Declares the module parameters for the DLL. diff --git a/plugins/spu2-x/src/Windows/UIHelpers.cpp b/plugins/spu2-x/src/Windows/UIHelpers.cpp index 8fadeb2fde..924c55f9e8 100644 --- a/plugins/spu2-x/src/Windows/UIHelpers.cpp +++ b/plugins/spu2-x/src/Windows/UIHelpers.cpp @@ -1,24 +1,20 @@ /* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 -* Developed and maintained by the Pcsx2 Development Team. -* -* Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] -* -* This library is free software; you can redistribute it and/or modify it under -* the terms of the GNU Lesser General Public License as published by the Free -* Software Foundation; either version 2.1 of the the License, or (at your -* option) any later version. -* -* This library is distributed in the hope that it will be useful, but WITHOUT -* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS -* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License -* for more details. -* -* You should have received a copy of the GNU Lesser General Public License along -* with this library; if not, write to the Free Software Foundation, Inc., 59 -* Temple Place, Suite 330, Boston, MA 02111-1307 USA -* -*/ - + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ + #include "Dialogs.h" int SendDialogMsg( HWND hwnd, int dlgId, UINT code, WPARAM wParam, LPARAM lParam) diff --git a/plugins/spu2-x/src/Windows/dsp.h b/plugins/spu2-x/src/Windows/dsp.h index ca1d187484..e3d3c2ce4a 100644 --- a/plugins/spu2-x/src/Windows/dsp.h +++ b/plugins/spu2-x/src/Windows/dsp.h @@ -25,6 +25,8 @@ // Configuration data should be stored in \plugin.ini // (look at the vis plugin for configuration code) +#pragma once + typedef struct winampDSPModule { char *description; // description HWND hwndParent; // parent window (filled in by calling app) diff --git a/plugins/spu2-x/src/defs.h b/plugins/spu2-x/src/defs.h index aaabe1683f..fdbbc8eaf5 100644 --- a/plugins/spu2-x/src/defs.h +++ b/plugins/spu2-x/src/defs.h @@ -3,24 +3,19 @@ * * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] * - * This library is free software; you can redistribute it and/or modify it under - * the terms of the GNU Lesser General Public License as published by the Free - * Software Foundation; either version 2.1 of the the License, or (at your - * option) any later version. - * - * This library is distributed in the hope that it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License - * for more details. - * - * You should have received a copy of the GNU Lesser General Public License along - * with this library; if not, write to the Free Software Foundation, Inc., 59 - * Temple Place, Suite 330, Boston, MA 02111-1307 USA - * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . */ -#ifndef DEFS_H_INCLUDED -#define DEFS_H_INCLUDED +#pragma once struct V_VolumeLR { @@ -441,5 +436,3 @@ static __forceinline u32 timeGetTime() return (u32)(t.time*1000+t.millitm); } #endif - -#endif // DEFS_H_INCLUDED // diff --git a/plugins/spu2-x/src/regs.h b/plugins/spu2-x/src/regs.h index 2effce816a..10a789118f 100644 --- a/plugins/spu2-x/src/regs.h +++ b/plugins/spu2-x/src/regs.h @@ -1,22 +1,21 @@ -//GiGaHeRz's SPU2 Driver -//Copyright (c) 2003-2008, David Quintana -// -//This library is free software; you can redistribute it and/or -//modify it under the terms of the GNU Lesser General Public -//License as published by the Free Software Foundation; either -//version 2.1 of the License, or (at your option) any later version. -// -//This library is distributed in the hope that it will be useful, -//but WITHOUT ANY WARRANTY; without even the implied warranty of -//MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -//Lesser General Public License for more details. -// -//You should have received a copy of the GNU Lesser General Public -//License along with this library; if not, write to the Free Software -//Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA -// -#ifndef REGS_H_INCLUDED -#define REGS_H_INCLUDED +/* SPU2-X, A plugin for Emulating the Sound Processing Unit of the Playstation 2 + * Developed and maintained by the Pcsx2 Development Team. + * + * Original portions from SPU2ghz are (c) 2008 by David Quintana [gigaherz] + * + * SPU2-X is free software: you can redistribute it and/or modify it under the terms + * of the GNU Lesser General Public License as published by the Free Software Found- + * ation, either version 3 of the License, or (at your option) any later version. + * + * SPU2-X is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; + * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR + * PURPOSE. See the GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with SPU2-X. If not, see . + */ + +#pragma once #define SPU2_CORE0 0x00000000 #define SPU2_CORE1 0x00000400 @@ -181,4 +180,3 @@ Core attributes (SD_C) #define VOICE_ADDR_LSAX 0x4 // Loop point address #define VOICE_ADDR_NAX 0x8 // Waveform data that should be read next -#endif // REGS_H_INCLUDED // \ No newline at end of file diff --git a/plugins/spu2-x/src/spdif.h b/plugins/spu2-x/src/spdif.h index 16cca3ef5f..52391706aa 100644 --- a/plugins/spu2-x/src/spdif.h +++ b/plugins/spu2-x/src/spdif.h @@ -15,8 +15,8 @@ //License along with this library; if not, write to the Free Software //Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA // -#ifndef SPDIF_H_INCLUDED -#define SPDIF_H_INCLUDED + +#pragma once #ifndef u32 typedef unsigned int u32; @@ -109,4 +109,3 @@ typedef struct _chstatus u8 reservd2[22]; } chstatus: -#endif//SPDIF_H_INCLUDED \ No newline at end of file From a0bc6d33bbeadd6c5d464ab477d263fc8031bb72 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 23 Jun 2009 01:54:33 +0000 Subject: [PATCH 273/296] microVU: fixed a problem introduced in r1304 (fixes bios memcard viewer hanging, tekken 5, and probably a lot more...) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1421 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Analyze.inl | 67 +++++++++++++++++------------------ 1 file changed, 33 insertions(+), 34 deletions(-) diff --git a/pcsx2/x86/microVU_Analyze.inl b/pcsx2/x86/microVU_Analyze.inl index abec8bcda8..cd9d47b659 100644 --- a/pcsx2/x86/microVU_Analyze.inl +++ b/pcsx2/x86/microVU_Analyze.inl @@ -345,45 +345,44 @@ microVUt(void) mVUanalyzeXGkick(mV, int Fs, int xCycles) { // Branches - Branch Opcodes //------------------------------------------------------------------ -#define analyzeBranchVI(xReg, infoVar) { \ - if (xReg) { \ - int i; \ - int iEnd = aMin(5, mVUcount); \ - int bPC = iPC; \ - for (i = 0; i < iEnd; i++) { \ - incPC2(-2) \ - if ((mVUlow.VI_write.reg == xReg) && mVUlow.VI_write.used) { \ - if (mVUlow.readFlags || i == 5) break; \ - if (i == 0) continue; \ - if (((mVUlow.VI_read[0].reg == xReg) && (mVUlow.VI_read[0].used)) \ - || ((mVUlow.VI_read[1].reg == xReg) && (mVUlow.VI_read[1].used))) \ - { continue; } \ - } \ - break; \ - } \ - if (i) { \ - DevCon::Status("microVU%d: Branch VI-Delay (%d)", params getIndex, i); \ - incPC2(2); \ - mVUlow.backupVI = 1; \ - iPC = bPC; \ - infoVar = 1; \ - } \ - iPC = bPC; \ - } \ +microVUt(void) analyzeBranchVI(mV, int xReg, bool &infoVar) { + if (!xReg) return; + int i; + int iEnd = aMin(5, mVUcount); + int bPC = iPC; + incPC2(-2); + for (i = 0; i < iEnd; i++) { + if ((mVUlow.VI_write.reg == xReg) && mVUlow.VI_write.used) { + if (mVUlow.readFlags || i == 5) break; + if (i == 0) { incPC2(-2); continue; } + if (((mVUlow.VI_read[0].reg == xReg) && (mVUlow.VI_read[0].used)) + || ((mVUlow.VI_read[1].reg == xReg) && (mVUlow.VI_read[1].used))) + { incPC2(-2); continue; } + } + break; + } + if (i) { + incPC2(2); + mVUlow.backupVI = 1; + iPC = bPC; + infoVar = 1; + DevCon::Status("microVU%d: Branch VI-Delay (%d) [%04x]", params getIndex, i, xPC); + } + iPC = bPC; } microVUt(void) mVUanalyzeBranch1(mV, int Is) { - if (mVUregs.VI[Is] || mVUstall) { analyzeVIreg1(Is, mVUlow.VI_read[0]); } - else { analyzeBranchVI(Is, mVUlow.memReadIs); } + analyzeVIreg1(Is, mVUlow.VI_read[0]); + if (!mVUstall) { + analyzeBranchVI(mVU, Is, mVUlow.memReadIs); + } } microVUt(void) mVUanalyzeBranch2(mV, int Is, int It) { - if (mVUregs.VI[Is] || mVUregs.VI[It] || mVUstall) { - analyzeVIreg1(Is, mVUlow.VI_read[0]); - analyzeVIreg1(It, mVUlow.VI_read[1]); - } - else { - analyzeBranchVI(Is, mVUlow.memReadIs); - analyzeBranchVI(It, mVUlow.memReadIt); + analyzeVIreg1(Is, mVUlow.VI_read[0]); + analyzeVIreg1(It, mVUlow.VI_read[1]); + if (!mVUstall) { + analyzeBranchVI(mVU, Is, mVUlow.memReadIs); + analyzeBranchVI(mVU, It, mVUlow.memReadIt); } } From 98ff4fd12dcda11038c4dfcd2679cce3de7bc858 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 23 Jun 2009 02:07:38 +0000 Subject: [PATCH 274/296] microVU: minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1422 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 2 +- pcsx2/x86/microVU_Execute.inl | 2 +- pcsx2/x86/microVU_Misc.h | 8 ++++++++ 3 files changed, 10 insertions(+), 2 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 439c42d26b..c836d8b6f6 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -290,7 +290,7 @@ microVUt(void) mVUtestCycles(mV) { // Recompiler //------------------------------------------------------------------ -void* mVUcompile(microVU* mVU, u32 startPC, uptr pState) { +microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { using namespace x86Emitter; microBlock* pBlock = NULL; diff --git a/pcsx2/x86/microVU_Execute.inl b/pcsx2/x86/microVU_Execute.inl index 8d34b48000..9a8e4b28b1 100644 --- a/pcsx2/x86/microVU_Execute.inl +++ b/pcsx2/x86/microVU_Execute.inl @@ -23,7 +23,7 @@ //------------------------------------------------------------------ // Generates the code for entering recompiled blocks -microVUt(void) mVUdispatcherA(mV) { +void mVUdispatcherA(mV) { mVU->startFunct = x86Ptr; // __fastcall = The first two DWORD or smaller arguments are passed in ECX and EDX registers; all other arguments are passed right to left. diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index cfb3a54258..c895c7eee2 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -148,9 +148,17 @@ declareAllVariables #define mF int recPass #define mX mVU, recPass +// Recursive Inline +#ifndef __LINUX__ +#define __recInline __forceinline +#else +#define __recInline inline +#endif + // Function/Template Stuff #define mVUx (vuIndex ? µVU1 : µVU0) #define mVUop(opName) void opName (mP) +#define microVUr(aType) __recInline aType #define microVUt(aType) __forceinline aType #define microVUx(aType) template aType #define microVUf(aType) template __forceinline aType From a4a1feae644887b2e17a27a6be819148a6407a56 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 23 Jun 2009 02:48:06 +0000 Subject: [PATCH 275/296] microVU: fixed a typo that broke Crash Twinsanity :D git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1423 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index c836d8b6f6..d90735dd95 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -230,7 +230,7 @@ microVUt(void) mVUendProgram(mV, int isEbit, int* xStatus, int* xMac, int* xClip MOV32RtoM((uptr)&mVU->regs->VI[REG_MAC_FLAG].UL, gprT1); MOV32RtoM((uptr)&mVU->regs->VI[REG_CLIP_FLAG].UL, gprT2); - if (isEbit || !isVU1) { // Clear 'is busy' Flags + if (isEbit || isVU1) { // Clear 'is busy' Flags AND32ItoM((uptr)&VU0.VI[REG_VPU_STAT].UL, (isVU1 ? ~0x100 : ~0x001)); // VBS0/VBS1 flag AND32ItoM((uptr)&mVU->regs->vifRegs->stat, ~0x4); // Clear VU 'is busy' signal for vif } From bc5c78f124b70b4579c9e22cc79595f9c448f444 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 23 Jun 2009 03:56:52 +0000 Subject: [PATCH 276/296] microVU: fixed a bug where xgkick delay info wasn't propagating through blocks correctly (fixes Sega Classics Collection) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1424 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 7 ++++--- pcsx2/x86/microVU_Compile.inl | 2 +- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 3ebd1d94ef..cdf0011baa 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -73,9 +73,10 @@ public: } else { // Can do Simple Search (Only Matches the Important Pipeline Stuff) for (int i = 0; i <= listSize; i++) { - if ((blockList[i].pState.q == pState->q) - && (blockList[i].pState.p == pState->p) - && (blockList[i].pState.flags == pState->flags) + if ((blockList[i].pState.q == pState->q) + && (blockList[i].pState.p == pState->p) + && (blockList[i].pState.xgkick == pState->xgkick) + && (blockList[i].pState.flags == pState->flags) && !(blockList[i].pState.needExactMatch & 0xf0f)) { return &blockList[i]; } } } diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index d90735dd95..a67fb55217 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -317,7 +317,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { mVUflagInfo = 0; mVUsFlagHack = CHECK_VU_FLAGHACK; - for (int branch = 0; mVUcount < endCount; mVUcount++) { + for (int branch = 0; mVUcount < endCount; mVUcount++) { incPC(1); startLoop(); mVUincCycles(mVU, 1); From b284fae2d5ef13d244491ece5c5ab17928ec75ca Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 23 Jun 2009 04:12:32 +0000 Subject: [PATCH 277/296] GSdx: moved around some code and optimized texture caching a bit, there may be a slight speed-up in hw mode for those games that use many textures. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1425 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GS.cpp | 10 +- plugins/GSdx/GSDevice.h | 1 + plugins/GSdx/GSDevice10.cpp | 4 +- plugins/GSdx/GSRenderer.cpp | 311 ++++++++++++++ plugins/GSdx/GSRenderer.h | 15 +- plugins/GSdx/GSRendererHW.h | 219 ++-------- plugins/GSdx/GSRendererHW10.cpp | 83 ++-- plugins/GSdx/GSRendererHW10.h | 4 +- plugins/GSdx/GSRendererHW11.cpp | 83 ++-- plugins/GSdx/GSRendererHW11.h | 4 +- plugins/GSdx/GSRendererHW9.cpp | 60 +-- plugins/GSdx/GSRendererHW9.h | 4 +- plugins/GSdx/GSRendererSW.cpp | 363 +--------------- plugins/GSdx/GSRendererSW.h | 4 - plugins/GSdx/GSTexture10.cpp | 30 +- plugins/GSdx/GSTextureCache.cpp | 73 +--- plugins/GSdx/GSTextureCache.h | 4 +- plugins/GSdx/GSTextureCache10.cpp | 93 ++++- plugins/GSdx/GSTextureFX.cpp | 23 + plugins/GSdx/GSTextureFX.h | 243 +++++++++++ plugins/GSdx/GSTextureFX10.cpp | 17 +- plugins/GSdx/GSTextureFX10.h | 187 +-------- plugins/GSdx/GSTextureFX11.cpp | 17 +- plugins/GSdx/GSTextureFX11.h | 187 +-------- plugins/GSdx/GSTextureFX9.cpp | 10 - plugins/GSdx/GSTextureFX9.h | 117 +----- plugins/GSdx/GSVertex.h | 2 +- plugins/GSdx/GSVertexHW.h | 13 +- plugins/GSdx/GSVertexSW.cpp | 142 ------- plugins/GSdx/GSVertexSW.h | 78 ---- plugins/GSdx/GSVertexTrace.cpp | 670 ++++++++++++++++++++++++++++++ plugins/GSdx/GSVertexTrace.h | 99 +++++ plugins/GSdx/GSdx_vs2008.vcproj | 16 + plugins/GSdx/res/tfx.fx | 2 +- 34 files changed, 1667 insertions(+), 1521 deletions(-) create mode 100644 plugins/GSdx/GSTextureFX.cpp create mode 100644 plugins/GSdx/GSTextureFX.h create mode 100644 plugins/GSdx/GSVertexTrace.cpp create mode 100644 plugins/GSdx/GSVertexTrace.h diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 430bb23407..67aeab825d 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -73,6 +73,11 @@ EXPORT_C GSsetBaseMem(uint8* mem) EXPORT_C_(INT32) GSinit() { + if(!GSUtil::CheckSSE()) + { + return -1; + } + return 0; } @@ -113,11 +118,6 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) #endif - if(!GSUtil::CheckSSE()) - { - return -1; - } - switch(renderer) { default: diff --git a/plugins/GSdx/GSDevice.h b/plugins/GSdx/GSDevice.h index 59799e814d..a5d36775a3 100644 --- a/plugins/GSdx/GSDevice.h +++ b/plugins/GSdx/GSDevice.h @@ -81,6 +81,7 @@ public: virtual void Flip() {} virtual void BeginScene() {} + virtual void DrawPrimitive() {}; virtual void EndScene() {} virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {} diff --git a/plugins/GSdx/GSDevice10.cpp b/plugins/GSdx/GSDevice10.cpp index 7f3fbf2791..6d7d2b73ec 100644 --- a/plugins/GSdx/GSDevice10.cpp +++ b/plugins/GSdx/GSDevice10.cpp @@ -302,7 +302,7 @@ GSTexture* GSDevice10::Create(int type, int w, int h, int format) desc.BindFlags = D3D10_BIND_RENDER_TARGET | D3D10_BIND_SHADER_RESOURCE; break; case GSTexture::DepthStencil: - desc.BindFlags = D3D10_BIND_DEPTH_STENCIL; + desc.BindFlags = D3D10_BIND_DEPTH_STENCIL;// | D3D10_BIND_SHADER_RESOURCE; break; case GSTexture::Texture: desc.BindFlags = D3D10_BIND_SHADER_RESOURCE; @@ -344,7 +344,7 @@ GSTexture* GSDevice10::CreateRenderTarget(int w, int h, int format) GSTexture* GSDevice10::CreateDepthStencil(int w, int h, int format) { - return __super::CreateDepthStencil(w, h, format ? format : DXGI_FORMAT_D32_FLOAT_S8X24_UINT); + return __super::CreateDepthStencil(w, h, format ? format : DXGI_FORMAT_D32_FLOAT_S8X24_UINT); // DXGI_FORMAT_R32G8X24_TYPELESS } GSTexture* GSDevice10::CreateTexture(int w, int h, int format) diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 8af6720dfc..046483fba3 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -426,3 +426,314 @@ void GSRenderer::KeyEvent(GSKeyEventData* e) } } +void GSRenderer::GetTextureMinMax(GSVector4i& r) +{ + const GSDrawingContext* context = m_context; + + int tw = context->TEX0.TW; + int th = context->TEX0.TH; + + int w = 1 << tw; + int h = 1 << th; + + GSVector4i tr(0, 0, w, h); + + int wms = context->CLAMP.WMS; + int wmt = context->CLAMP.WMT; + + int minu = (int)context->CLAMP.MINU; + int minv = (int)context->CLAMP.MINV; + int maxu = (int)context->CLAMP.MAXU; + int maxv = (int)context->CLAMP.MAXV; + + GSVector4i vr = tr; + + switch(wms) + { + case CLAMP_REPEAT: + break; + case CLAMP_CLAMP: + break; + case CLAMP_REGION_CLAMP: + if(vr.x < minu) vr.x = minu; + if(vr.z > maxu + 1) vr.z = maxu + 1; + break; + case CLAMP_REGION_REPEAT: + vr.x = maxu; + vr.z = vr.x + (minu + 1); + break; + default: + __assume(0); + } + + switch(wmt) + { + case CLAMP_REPEAT: + break; + case CLAMP_CLAMP: + break; + case CLAMP_REGION_CLAMP: + if(vr.y < minv) vr.y = minv; + if(vr.w > maxv + 1) vr.w = maxv + 1; + break; + case CLAMP_REGION_REPEAT: + vr.y = maxv; + vr.w = vr.y + (minv + 1); + break; + default: + __assume(0); + } + + if(wms + wmt < 6) + { + GSVector4 st = m_vt.m_min.t.xyxy(m_vt.m_max.t); + + if(context->TEX1.IsLinear()) + { + st += GSVector4(-0x8000, 0x8000).xxyy(); + } + + GSVector4i uv = GSVector4i(st).sra32(16); + + GSVector4i u, v; + + int mask = 0; + + if(wms == CLAMP_REPEAT || wmt == CLAMP_REPEAT) + { + u = uv & GSVector4i::xffffffff().srl32(32 - tw); + v = uv & GSVector4i::xffffffff().srl32(32 - th); + + GSVector4i uu = uv.sra32(tw); + GSVector4i vv = uv.sra32(th); + + mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); + } + + uv = uv.rintersect(tr); + + switch(wms) + { + case CLAMP_REPEAT: + if(mask & 0x000f) {if(vr.x < u.x) vr.x = u.x; if(vr.z > u.z + 1) vr.z = u.z + 1;} + break; + case CLAMP_CLAMP: + case CLAMP_REGION_CLAMP: + if(vr.x < uv.x) vr.x = uv.x; + if(vr.z > uv.z + 1) vr.z = uv.z + 1; + break; + case CLAMP_REGION_REPEAT: // TODO + break; + default: + __assume(0); + } + + switch(wmt) + { + case CLAMP_REPEAT: + if(mask & 0xf000) {if(vr.y < v.y) vr.y = v.y; if(vr.w > v.w + 1) vr.w = v.w + 1;} + break; + case CLAMP_CLAMP: + case CLAMP_REGION_CLAMP: + if(vr.y < uv.y) vr.y = uv.y; + if(vr.w > uv.w + 1) vr.w = uv.w + 1; + break; + case CLAMP_REGION_REPEAT: // TODO + break; + default: + __assume(0); + } + } + + GSVector2i bs = GSLocalMemory::m_psm[context->TEX0.PSM].bs; + + r = vr.ralign(bs).rintersect(tr); +} + +void GSRenderer::GetAlphaMinMax() +{ + if(m_vt.m_alpha.valid) + { + return; + } + + const GSDrawingEnvironment& env = m_env; + const GSDrawingContext* context = m_context; + + GSVector4i a = m_vt.m_min.c.uph32(m_vt.m_max.c).zzww(); + + if(PRIM->TME && context->TEX0.TCC) + { + uint32 bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; + uint32 cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; + uint32 pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; + + if(bpp == 32) + { + a.y = 0; + a.w = 0xff; + } + else if(bpp == 24) + { + a.y = env.TEXA.AEM ? 0 : env.TEXA.TA0; + a.w = env.TEXA.TA0; + } + else if(bpp == 16) + { + a.y = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); + a.w = max(env.TEXA.TA0, env.TEXA.TA1); + } + else + { + m_mem.m_clut.GetAlphaMinMax32(a.y, a.w); + } + + switch(context->TEX0.TFX) + { + case TFX_MODULATE: + a.x = (a.x * a.y) >> 7; + a.z = (a.z * a.w) >> 7; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_DECAL: + a.x = a.y; + a.z = a.w; + break; + case TFX_HIGHLIGHT: + a.x = a.x + a.y; + a.z = a.z + a.w; + if(a.x > 0xff) a.x = 0xff; + if(a.z > 0xff) a.z = 0xff; + break; + case TFX_HIGHLIGHT2: + a.x = a.y; + a.z = a.w; + break; + default: + __assume(0); + } + } + + m_vt.m_alpha.min = a.x; + m_vt.m_alpha.max = a.z; + m_vt.m_alpha.valid = true; +} + +bool GSRenderer::TryAlphaTest(uint32& fm, uint32& zm) +{ + const GSDrawingContext* context = m_context; + + bool pass = true; + + if(context->TEST.ATST == ATST_NEVER) + { + pass = false; + } + else if(context->TEST.ATST != ATST_ALWAYS) + { + GetAlphaMinMax(); + + int amin = m_vt.m_alpha.min; + int amax = m_vt.m_alpha.max; + + int aref = context->TEST.AREF; + + switch(context->TEST.ATST) + { + case ATST_NEVER: + pass = false; + break; + case ATST_ALWAYS: + pass = true; + break; + case ATST_LESS: + if(amax < aref) pass = true; + else if(amin >= aref) pass = false; + else return false; + break; + case ATST_LEQUAL: + if(amax <= aref) pass = true; + else if(amin > aref) pass = false; + else return false; + break; + case ATST_EQUAL: + if(amin == aref && amax == aref) pass = true; + else if(amin > aref || amax < aref) pass = false; + else return false; + break; + case ATST_GEQUAL: + if(amin >= aref) pass = true; + else if(amax < aref) pass = false; + else return false; + break; + case ATST_GREATER: + if(amin > aref) pass = true; + else if(amax <= aref) pass = false; + else return false; + break; + case ATST_NOTEQUAL: + if(amin == aref && amax == aref) pass = false; + else if(amin > aref || amax < aref) pass = true; + else return false; + break; + default: + __assume(0); + } + } + + if(!pass) + { + switch(context->TEST.AFAIL) + { + case AFAIL_KEEP: fm = zm = 0xffffffff; break; + case AFAIL_FB_ONLY: zm = 0xffffffff; break; + case AFAIL_ZB_ONLY: fm = 0xffffffff; break; + case AFAIL_RGB_ONLY: fm |= 0xff000000; zm = 0xffffffff; break; + default: __assume(0); + } + } + + return true; +} + +bool GSRenderer::IsOpaque() +{ + if(PRIM->AA1) + { + return false; + } + + if(!PRIM->ABE) + { + return true; + } + + const GSDrawingContext* context = m_context; + + int amin = 0, amax = 0xff; + + if(context->ALPHA.A != context->ALPHA.B) + { + if(context->ALPHA.C == 0) + { + GetAlphaMinMax(); + + amin = m_vt.m_alpha.min; + amax = m_vt.m_alpha.max; + } + else if(context->ALPHA.C == 1) + { + if(context->FRAME.PSM == PSM_PSMCT24 || context->FRAME.PSM == PSM_PSMZ24) + { + amin = amax = 0x80; + } + } + else if(context->ALPHA.C == 1) + { + amin = amax = context->ALPHA.FIX; + } + } + + return context->ALPHA.IsOpaque(amin, amax); +} diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 97ad72df2c..21f1651d4b 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -24,6 +24,7 @@ #include "GSdx.h" #include "GSWnd.h" #include "GSState.h" +#include "GSVertexTrace.h" #include "GSVertexList.h" #include "GSSettingsDlg.h" #include "GSCapture.h" @@ -48,6 +49,15 @@ protected: virtual void ResetDevice() {} virtual GSTexture* GetOutput(int i) = 0; + GSVertexTrace m_vt; + + // following functions need m_vt to be initialized + + void GetTextureMinMax(GSVector4i& r); + void GetAlphaMinMax(); + bool TryAlphaTest(uint32& fm, uint32& zm); + bool IsOpaque(); + public: GSWnd m_wnd; GSDevice* m_dev; @@ -66,11 +76,6 @@ public: virtual bool MakeSnapshot(const string& path); virtual void KeyEvent(GSKeyEventData* e); - virtual void MinMaxUV(int w, int h, GSVector4i& r) - { - r = GSVector4i(0, 0, w, h); - } - virtual bool CanUpscale() { return !m_nativeres; diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index af3bfefa6c..67a16e449b 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -46,176 +46,6 @@ protected: __super::Reset(); } - void MinMaxUV(int w, int h, GSVector4i& r) - { - int wms = m_context->CLAMP.WMS; - int wmt = m_context->CLAMP.WMT; - - int minu = (int)m_context->CLAMP.MINU; - int minv = (int)m_context->CLAMP.MINV; - int maxu = (int)m_context->CLAMP.MAXU; - int maxv = (int)m_context->CLAMP.MAXV; - - GSVector4i vr = GSVector4i(0, 0, w, h); - - GSVector4i wm[3]; - - if(wms + wmt < 6) - { - GSVector4 mm; - - if(m_count < 100) - { - Vertex* v = m_vertices; - - GSVector4 minv(+1e10f); - GSVector4 maxv(-1e10f); - - int i = 0; - - if(PRIM->FST) - { - for(int j = m_count - 3; i < j; i += 4) - { - GSVector4 v0 = v[i + 0].vf[0]; - GSVector4 v1 = v[i + 1].vf[0]; - GSVector4 v2 = v[i + 2].vf[0]; - GSVector4 v3 = v[i + 3].vf[0]; - - minv = minv.minv((v0.minv(v1)).minv(v2.minv(v3))); - maxv = maxv.maxv((v0.maxv(v1)).maxv(v2.maxv(v3))); - } - - for(int j = m_count; i < j; i++) - { - GSVector4 v0 = v[i + 0].vf[0]; - - minv = minv.minv(v0); - maxv = maxv.maxv(v0); - } - - mm = minv.xyxy(maxv) * GSVector4(16 << m_context->TEX0.TW, 16 << m_context->TEX0.TH).xyxy().rcpnr(); - } - else - { - /* - for(int j = m_count - 3; i < j; i += 4) - { - GSVector4 v0 = GSVector4(v[i + 0].m128[0]) / GSVector4(v[i + 0].GetQ()); - GSVector4 v1 = GSVector4(v[i + 1].m128[0]) / GSVector4(v[i + 1].GetQ()); - GSVector4 v2 = GSVector4(v[i + 2].m128[0]) / GSVector4(v[i + 2].GetQ()); - GSVector4 v3 = GSVector4(v[i + 3].m128[0]) / GSVector4(v[i + 3].GetQ()); - - minv = minv.minv((v0.minv(v1)).minv(v2.minv(v3))); - maxv = maxv.maxv((v0.maxv(v1)).maxv(v2.maxv(v3))); - } - - for(int j = m_count; i < j; i++) - { - GSVector4 v0 = GSVector4(v[i + 0].m128[0]) / GSVector4(v[i + 0].GetQ());; - - minv = minv.minv(v0); - maxv = maxv.maxv(v0); - } - - mm = minv.xyxy(maxv); - */ - - // just can't beat the compiler generated scalar sse code with packed div or rcp - - mm.x = mm.y = +1e10; - mm.z = mm.w = -1e10; - - for(int j = m_count; i < j; i++) - { - float w = 1.0f / v[i].GetQ(); - - float x = v[i].t.x * w; - - if(x < mm.x) mm.x = x; - if(x > mm.z) mm.z = x; - - float y = v[i].t.y * w; - - if(y < mm.y) mm.y = y; - if(y > mm.w) mm.w = y; - } - } - } - else - { - mm = GSVector4(0.0f, 0.0f, 1.0f, 1.0f); - } - - GSVector4 v0 = GSVector4(vr); - GSVector4 v1 = v0.zwzw(); - - GSVector4 mmf = mm.floor(); - GSVector4 mask = mmf.xyxy() == mmf.zwzw(); - - wm[0] = GSVector4i(v0.blend8((mm - mmf) * v1, mask)); - - mm *= v1; - - wm[1] = GSVector4i(mm.sat(GSVector4::zero(), v1)); - wm[2] = GSVector4i(mm.sat(GSVector4(minu, minv, maxu, maxv))); - } - - GSVector4i v; - - switch(wms) - { - case CLAMP_REPEAT: - v = wm[0]; - if(v.x == 0 && v.z != w) v.z = w; // FIXME - vr.x = v.x; - vr.z = v.z; - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - v = wm[wms]; - if(v.x > v.z) v.x = v.z; - vr.x = v.x; - vr.z = v.z; - break; - case CLAMP_REGION_REPEAT: - vr.x = maxu; - vr.z = vr.x + (minu + 1); - break; - default: - __assume(0); - } - - switch(wmt) - { - case CLAMP_REPEAT: - v = wm[0]; - if(v.y == 0 && v.w != h) v.w = h; // FIXME - vr.y = v.y; - vr.w = v.w; - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - v = wm[wmt]; - if(v.y > v.w) v.y = v.w; - vr.y = v.y; - vr.w = v.w; - break; - case CLAMP_REGION_REPEAT: - vr.y = maxv; - vr.w = vr.y + (minv + 1); - break; - default: - __assume(0); - } - - r = vr + GSVector4i(-1, -1, 1, 1); // one more pixel because of bilinear filtering - - GSVector2i bs = GSLocalMemory::m_psm[m_context->TEX0.PSM].bs; - - r = r.ralign(bs).rintersect(GSVector4i(0, 0, w, h)); - } - void VSync(int field) { __super::VSync(field); @@ -285,10 +115,9 @@ protected: void Draw() { - if(IsBadFrame(m_skip)) - { - return; - } + if(IsBadFrame(m_skip)) return; + + m_vt.Update(m_vertices, m_count, GSUtil::GetPrimClass(PRIM->PRIM), PRIM, m_context); GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; @@ -311,7 +140,13 @@ protected: if(PRIM->TME) { - tex = m_tc->GetTexture(); + m_mem.m_clut.Read32(context->TEX0, env.TEXA); + + GSVector4i r; + + GetTextureMinMax(r); + + tex = m_tc->GetTexture(r); if(!tex) return; } @@ -366,7 +201,37 @@ protected: return; } - Draw(prim, rt->m_texture, ds->m_texture, tex); + // skip alpha test if possible + + GIFRegTEST TEST = context->TEST; + GIFRegFRAME FRAME = context->FRAME; + GIFRegZBUF ZBUF = context->ZBUF; + + uint32 fm = context->FRAME.FBMSK; + uint32 zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0; + + if(context->TEST.ATE && context->TEST.ATST != ATST_ALWAYS) + { + if(TryAlphaTest(fm, zm)) + { + context->TEST.ATE = 0; + } + } + + context->FRAME.FBMSK = fm; + context->ZBUF.ZMSK = zm != 0; + + // + + Draw(GSUtil::GetPrimClass(prim), rt->m_texture, ds->m_texture, tex); + + // + + context->TEST = TEST; + context->FRAME = FRAME; + context->ZBUF = ZBUF; + + // OverrideOutput(); @@ -396,7 +261,7 @@ protected: m_tc->InvalidateTextures(context->FRAME, context->ZBUF); } - virtual void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) = 0; + virtual void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) = 0; virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* t) { diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index 39a66640e3..b39afae7c0 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -169,44 +169,26 @@ void GSRendererHW10::VertexKick(bool skip) } } -void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +void GSRendererHW10::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; -/* - if(s_dump) - { - TRACE(_T("\n")); - TRACE(_T("PRIM = %d, ZMSK = %d, ZTE = %d, ZTST = %d, ATE = %d, ATST = %d, AFAIL = %d, AREF = %02x\n"), - PRIM->PRIM, context->ZBUF.ZMSK, - context->TEST.ZTE, context->TEST.ZTST, - context->TEST.ATE, context->TEST.ATST, context->TEST.AFAIL, context->TEST.AREF); - - for(int i = 0; i < m_count; i++) - { - TRACE(_T("[%d] %3.0f %3.0f %3.0f %3.0f\n"), i, (float)m_vertices[i].p.x / 16, (float)m_vertices[i].p.y / 16, (float)m_vertices[i].p.z, (float)m_vertices[i].a); - } - } -*/ D3D10_PRIMITIVE_TOPOLOGY topology; int prims = 0; - switch(prim) + switch(primclass) { - case GS_POINTLIST: + case GS_POINT_CLASS: topology = D3D10_PRIMITIVE_TOPOLOGY_POINTLIST; prims = m_count; break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: topology = D3D10_PRIMITIVE_TOPOLOGY_LINELIST; prims = m_count / 2; break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: + case GS_TRIANGLE_CLASS: topology = D3D10_PRIMITIVE_TOPOLOGY_TRIANGLELIST; prims = m_count / 3; break; @@ -227,16 +209,16 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // om - GSTextureFX10::OMDepthStencilSelector om_dssel; + GSTextureFX::OMDepthStencilSelector om_dssel; om_dssel.zte = context->TEST.ZTE; om_dssel.ztst = context->TEST.ZTST; om_dssel.zwe = !context->ZBUF.ZMSK; om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; - GSTextureFX10::OMBlendSelector om_bsel; + GSTextureFX::OMBlendSelector om_bsel; - om_bsel.abe = PRIM->ABE || (prim == 1 || prim == 2) && PRIM->AA1; + om_bsel.abe = !IsOpaque(); om_bsel.a = context->ALPHA.A; om_bsel.b = context->ALPHA.B; om_bsel.c = context->ALPHA.C; @@ -250,34 +232,38 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // vs - GSTextureFX10::VSSelector vs_sel; + GSTextureFX::VSSelector vs_sel; vs_sel.bppz = 0; vs_sel.tme = PRIM->TME; vs_sel.fst = PRIM->FST; - vs_sel.prim = prim; + vs_sel.prim = primclass; if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) { if(context->ZBUF.PSM == PSM_PSMZ24) { - if(WrapZ(0xffffff)) + if(m_vt.m_max.p.z > 0xffffff) { + ASSERT(m_vt.m_min.p.z > 0xffffff); + vs_sel.bppz = 1; om_dssel.ztst = 1; } } else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) { - if(WrapZ(0xffff)) + if(m_vt.m_max.p.z > 0xffff) { + ASSERT(m_vt.m_min.p.z > 0xffff); + vs_sel.bppz = 2; om_dssel.ztst = 1; } } } - GSTextureFX10::VSConstantBuffer vs_cb; + GSTextureFX::VSConstantBuffer vs_cb; float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); @@ -298,14 +284,14 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // gs - GSTextureFX10::GSSelector gs_sel; + GSTextureFX::GSSelector gs_sel; gs_sel.iip = PRIM->IIP; - gs_sel.prim = GSUtil::GetPrimClass(prim); + gs_sel.prim = primclass; // ps - GSTextureFX10::PSSelector ps_sel; + GSTextureFX::PSSelector ps_sel; ps_sel.fst = PRIM->FST; ps_sel.wms = context->CLAMP.WMS; @@ -322,21 +308,21 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - GSTextureFX10::PSSamplerSelector ps_ssel; + GSTextureFX::PSSamplerSelector ps_ssel; ps_ssel.tau = 0; ps_ssel.tav = 0; ps_ssel.ltf = ps_sel.ltf; - GSTextureFX10::PSConstantBuffer ps_cb; + GSTextureFX::PSConstantBuffer ps_cb; ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - if(context->TEST.ATST == 2 || context->TEST.ATST == 5) + if(ps_sel.atst == 2 || ps_sel.atst == 5) { ps_cb.FogColor_AREF.a -= 0.9f / 255; } - else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) + else if(ps_sel.atst == 3 || ps_sel.atst == 6) { ps_cb.FogColor_AREF.a += 0.9f / 255; } @@ -424,7 +410,7 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache if(context->TEST.DoFirstPass()) { - m_tfx.Draw(); + m_dev->DrawPrimitive(); } if(context->TEST.DoSecondPass()) @@ -462,28 +448,13 @@ void GSRendererHW10::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache m_tfx.UpdateOM(om_dssel, om_bsel, bf); - m_tfx.Draw(); + m_dev->DrawPrimitive(); } } m_dev->EndScene(); } -bool GSRendererHW10::WrapZ(uint32 maxz) -{ - // should only run once if z values are in the z buffer range - - for(int i = 0, j = m_count; i < j; i++) - { - if(m_vertices[i].p.z <= maxz) - { - return false; - } - } - - return true; -} - void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) { if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) diff --git a/plugins/GSdx/GSRendererHW10.h b/plugins/GSdx/GSRendererHW10.h index 51cd9648b4..0568758624 100644 --- a/plugins/GSdx/GSRendererHW10.h +++ b/plugins/GSdx/GSRendererHW10.h @@ -28,12 +28,10 @@ class GSRendererHW10 : public GSRendererHW { - bool WrapZ(uint32 maxz); - protected: GSTextureFX10 m_tfx; - void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); struct { diff --git a/plugins/GSdx/GSRendererHW11.cpp b/plugins/GSdx/GSRendererHW11.cpp index ec7f11447b..ac47dad0e7 100644 --- a/plugins/GSdx/GSRendererHW11.cpp +++ b/plugins/GSdx/GSRendererHW11.cpp @@ -169,44 +169,26 @@ void GSRendererHW11::VertexKick(bool skip) } } -void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +void GSRendererHW11::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; -/* - if(s_dump) - { - TRACE(_T("\n")); - TRACE(_T("PRIM = %d, ZMSK = %d, ZTE = %d, ZTST = %d, ATE = %d, ATST = %d, AFAIL = %d, AREF = %02x\n"), - PRIM->PRIM, context->ZBUF.ZMSK, - context->TEST.ZTE, context->TEST.ZTST, - context->TEST.ATE, context->TEST.ATST, context->TEST.AFAIL, context->TEST.AREF); - - for(int i = 0; i < m_count; i++) - { - TRACE(_T("[%d] %3.0f %3.0f %3.0f %3.0f\n"), i, (float)m_vertices[i].p.x / 16, (float)m_vertices[i].p.y / 16, (float)m_vertices[i].p.z, (float)m_vertices[i].a); - } - } -*/ D3D11_PRIMITIVE_TOPOLOGY topology; int prims = 0; - switch(prim) + switch(primclass) { - case GS_POINTLIST: + case GS_POINT_CLASS: topology = D3D11_PRIMITIVE_TOPOLOGY_POINTLIST; prims = m_count; break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: topology = D3D11_PRIMITIVE_TOPOLOGY_LINELIST; prims = m_count / 2; break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: + case GS_TRIANGLE_CLASS: topology = D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST; prims = m_count / 3; break; @@ -227,16 +209,16 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // om - GSTextureFX11::OMDepthStencilSelector om_dssel; + GSTextureFX::OMDepthStencilSelector om_dssel; om_dssel.zte = context->TEST.ZTE; om_dssel.ztst = context->TEST.ZTST; om_dssel.zwe = !context->ZBUF.ZMSK; om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; - GSTextureFX11::OMBlendSelector om_bsel; + GSTextureFX::OMBlendSelector om_bsel; - om_bsel.abe = PRIM->ABE || (prim == 1 || prim == 2) && PRIM->AA1; + om_bsel.abe = !IsOpaque(); om_bsel.a = context->ALPHA.A; om_bsel.b = context->ALPHA.B; om_bsel.c = context->ALPHA.C; @@ -250,34 +232,38 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // vs - GSTextureFX11::VSSelector vs_sel; + GSTextureFX::VSSelector vs_sel; vs_sel.bppz = 0; vs_sel.tme = PRIM->TME; vs_sel.fst = PRIM->FST; - vs_sel.prim = prim; + vs_sel.prim = primclass; if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) { if(context->ZBUF.PSM == PSM_PSMZ24) { - if(WrapZ(0xffffff)) + if(m_vt.m_max.p.z > 0xffffff) { + ASSERT(m_vt.m_min.p.z > 0xffffff); + vs_sel.bppz = 1; om_dssel.ztst = 1; } } else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) { - if(WrapZ(0xffff)) + if(m_vt.m_max.p.z > 0xffff) { + ASSERT(m_vt.m_min.p.z > 0xffff); + vs_sel.bppz = 2; om_dssel.ztst = 1; } } } - GSTextureFX11::VSConstantBuffer vs_cb; + GSTextureFX::VSConstantBuffer vs_cb; float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); @@ -298,14 +284,14 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache // gs - GSTextureFX11::GSSelector gs_sel; + GSTextureFX::GSSelector gs_sel; gs_sel.iip = PRIM->IIP; - gs_sel.prim = GSUtil::GetPrimClass(prim); + gs_sel.prim = primclass; // ps - GSTextureFX11::PSSelector ps_sel; + GSTextureFX::PSSelector ps_sel; ps_sel.fst = PRIM->FST; ps_sel.wms = context->CLAMP.WMS; @@ -322,21 +308,21 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - GSTextureFX11::PSSamplerSelector ps_ssel; + GSTextureFX::PSSamplerSelector ps_ssel; ps_ssel.tau = 0; ps_ssel.tav = 0; ps_ssel.ltf = ps_sel.ltf; - GSTextureFX11::PSConstantBuffer ps_cb; + GSTextureFX::PSConstantBuffer ps_cb; ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - if(context->TEST.ATST == 2 || context->TEST.ATST == 5) + if(ps_sel.atst == 2 || ps_sel.atst == 5) { ps_cb.FogColor_AREF.a -= 0.9f / 255; } - else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) + else if(ps_sel.atst == 3 || ps_sel.atst == 6) { ps_cb.FogColor_AREF.a += 0.9f / 255; } @@ -424,7 +410,7 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache if(context->TEST.DoFirstPass()) { - m_tfx.Draw(); + m_dev->DrawPrimitive(); } if(context->TEST.DoSecondPass()) @@ -462,28 +448,13 @@ void GSRendererHW11::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache m_tfx.UpdateOM(om_dssel, om_bsel, bf); - m_tfx.Draw(); + m_dev->DrawPrimitive(); } } m_dev->EndScene(); } -bool GSRendererHW11::WrapZ(uint32 maxz) -{ - // should only run once if z values are in the z buffer range - - for(int i = 0, j = m_count; i < j; i++) - { - if(m_vertices[i].p.z <= maxz) - { - return false; - } - } - - return true; -} - void GSRendererHW11::SetupDATE(GSTexture* rt, GSTexture* ds) { if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) diff --git a/plugins/GSdx/GSRendererHW11.h b/plugins/GSdx/GSRendererHW11.h index f3763a7dd1..fab4c76911 100644 --- a/plugins/GSdx/GSRendererHW11.h +++ b/plugins/GSdx/GSRendererHW11.h @@ -28,12 +28,10 @@ class GSRendererHW11 : public GSRendererHW { - bool WrapZ(uint32 maxz); - protected: GSTextureFX11 m_tfx; - void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); struct { diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index 778bb9ef12..cfc0d99cdc 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -171,7 +171,7 @@ void GSRendererHW9::VertexKick(bool skip) } } -void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +void GSRendererHW9::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; @@ -179,21 +179,18 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: D3DPRIMITIVETYPE topology; int prims = 0; - switch(prim) + switch(primclass) { - case GS_POINTLIST: + case GS_POINT_CLASS: topology = D3DPT_POINTLIST; prims = m_count; break; - case GS_LINELIST: - case GS_LINESTRIP: + case GS_LINE_CLASS: topology = D3DPT_LINELIST; prims = m_count / 2; break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - case GS_SPRITE: + case GS_TRIANGLE_CLASS: + case GS_SPRITE_CLASS: topology = D3DPT_TRIANGLELIST; prims = m_count / 3; break; @@ -216,7 +213,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: // om - GSTextureFX9::OMDepthStencilSelector om_dssel; + GSTextureFX::OMDepthStencilSelector om_dssel; om_dssel.zte = context->TEST.ZTE; om_dssel.ztst = context->TEST.ZTST; @@ -224,9 +221,9 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; om_dssel.fba = m_fba.enabled ? context->FBA.FBA : 0; - GSTextureFX9::OMBlendSelector om_bsel; + GSTextureFX::OMBlendSelector om_bsel; - om_bsel.abe = PRIM->ABE || (prim == 1 || prim == 2) && PRIM->AA1; + om_bsel.abe = !IsOpaque(); om_bsel.a = context->ALPHA.A; om_bsel.b = context->ALPHA.B; om_bsel.c = context->ALPHA.C; @@ -240,7 +237,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: // vs - GSTextureFX9::VSSelector vs_sel; + GSTextureFX::VSSelector vs_sel; vs_sel.bppz = 0; vs_sel.tme = PRIM->TME; @@ -251,23 +248,27 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: { if(context->ZBUF.PSM == PSM_PSMZ24) { - if(WrapZ(0xffffff)) + if(m_vt.m_max.p.z > 0xffffff) { + ASSERT(m_vt.m_min.p.z > 0xffffff); + vs_sel.bppz = 1; om_dssel.ztst = 1; } } else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) { - if(WrapZ(0xffff)) + if(m_vt.m_max.p.z > 0xffff) { + ASSERT(m_vt.m_min.p.z > 0xffff); + vs_sel.bppz = 2; om_dssel.ztst = 1; } } } - GSTextureFX9::VSConstantBuffer vs_cb; + GSTextureFX::VSConstantBuffer vs_cb; float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); @@ -286,7 +287,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: // ps - GSTextureFX9::PSSelector ps_sel; + GSTextureFX::PSSelector ps_sel; ps_sel.fst = PRIM->FST; ps_sel.wms = context->CLAMP.WMS; @@ -302,7 +303,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_sel.rt = tex && tex->m_rendered; ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - GSTextureFX9::PSSamplerSelector ps_ssel; + GSTextureFX::PSSamplerSelector ps_ssel; ps_ssel.tau = 0; ps_ssel.tav = 0; @@ -312,11 +313,11 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - if(context->TEST.ATST == 2 || context->TEST.ATST == 5) + if(ps_sel.atst == 2 || ps_sel.atst == 5) { ps_cb.FogColor_AREF.a -= 0.9f / 255; } - else if(context->TEST.ATST == 3 || context->TEST.ATST == 6) + else if(ps_sel.atst == 3 || ps_sel.atst == 6) { ps_cb.FogColor_AREF.a += 0.9f / 255; } @@ -403,7 +404,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: if(context->TEST.DoFirstPass()) { - m_tfx.Draw(); + m_dev->DrawPrimitive(); } if(context->TEST.DoSecondPass()) @@ -441,7 +442,7 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: m_tfx.UpdateOM(om_dssel, om_bsel, bf); - m_tfx.Draw(); + m_dev->DrawPrimitive(); } } @@ -450,21 +451,6 @@ void GSRendererHW9::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache: if(om_dssel.fba) UpdateFBA(rt); } -bool GSRendererHW9::WrapZ(float maxz) -{ - // should only run once if z values are in the z buffer range - - for(int i = 0, j = m_count; i < j; i++) - { - if(m_vertices[i].p.z <= maxz) - { - return false; - } - } - - return true; -} - void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) { if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) diff --git a/plugins/GSdx/GSRendererHW9.h b/plugins/GSdx/GSRendererHW9.h index 9b8ff4221a..896fe892a5 100644 --- a/plugins/GSdx/GSRendererHW9.h +++ b/plugins/GSdx/GSRendererHW9.h @@ -28,13 +28,11 @@ class GSRendererHW9 : public GSRendererHW { - bool WrapZ(float maxz); - protected: GSTextureFX9 m_tfx; bool m_logz; - void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); struct { diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index d9e0d732b6..9828f8757e 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -139,7 +139,7 @@ void GSRendererSW::Draw() { GS_PRIM_CLASS primclass = GSUtil::GetPrimClass(PRIM->PRIM); - m_vtrace.Update(m_vertices, m_count, primclass, PRIM->IIP, PRIM->TME, m_context->TEX0.TFX, m_context->TEX0.TCC); + m_vt.Update(m_vertices, m_count, primclass, PRIM, m_context); if(m_dump) { @@ -206,7 +206,7 @@ void GSRendererSW::Draw() m_perfmon.Put(GSPerfMon::Prim, stats.prims); m_perfmon.Put(GSPerfMon::Fillrate, stats.pixels); - GSVector4i r = GSVector4i(m_vtrace.m_min.p.xyxy(m_vtrace.m_max.p)).rintersect(data.scissor); + GSVector4i r = GSVector4i(m_vt.m_min.p.xyxy(m_vt.m_max.p)).rintersect(data.scissor); GIFRegBITBLTBUF BITBLTBUF; @@ -266,261 +266,6 @@ void GSRendererSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GS m_tc->InvalidateVideoMem(BITBLTBUF, r); } -void GSRendererSW::GetTextureMinMax(int w, int h, GSVector4i& r, uint32 fst) -{ - const GSDrawingContext* context = m_context; - - int wms = context->CLAMP.WMS; - int wmt = context->CLAMP.WMT; - - int minu = (int)context->CLAMP.MINU; - int minv = (int)context->CLAMP.MINV; - int maxu = (int)context->CLAMP.MAXU; - int maxv = (int)context->CLAMP.MAXV; - - GSVector4i vr(0, 0, w, h); - - switch(wms) - { - case CLAMP_REPEAT: - break; - case CLAMP_CLAMP: - break; - case CLAMP_REGION_CLAMP: - if(vr.x < minu) vr.x = minu; - if(vr.z > maxu + 1) vr.z = maxu + 1; - break; - case CLAMP_REGION_REPEAT: - vr.x = maxu; - vr.z = vr.x + (minu + 1); - break; - default: - __assume(0); - } - - switch(wmt) - { - case CLAMP_REPEAT: - break; - case CLAMP_CLAMP: - break; - case CLAMP_REGION_CLAMP: - if(vr.y < minv) vr.y = minv; - if(vr.w > maxv + 1) vr.w = maxv + 1; - break; - case CLAMP_REGION_REPEAT: - vr.y = maxv; - vr.w = vr.y + (minv + 1); - break; - default: - __assume(0); - } - - if(fst) - { - GSVector4i uv = GSVector4i(m_vtrace.m_min.t.xyxy(m_vtrace.m_max.t)).sra32(16); - - GSVector4i u, v; - - int mask = 0; - - if(wms == CLAMP_REPEAT || wmt == CLAMP_REPEAT) - { - int tw = context->TEX0.TW; - int th = context->TEX0.TH; - - u = uv & GSVector4i::xffffffff().srl32(32 - tw); - v = uv & GSVector4i::xffffffff().srl32(32 - th); - - GSVector4i uu = uv.sra32(tw); - GSVector4i vv = uv.sra32(th); - - mask = (uu.upl32(vv) == uu.uph32(vv)).mask(); - } - - switch(wms) - { - case CLAMP_REPEAT: - if(mask & 0x000f) {if(vr.x < u.x) vr.x = u.x; if(vr.z > u.z + 1) vr.z = u.z + 1;} - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - if(vr.x < uv.x) vr.x = uv.x; - if(vr.z > uv.z + 1) vr.z = uv.z + 1; - break; - case CLAMP_REGION_REPEAT: // TODO - break; - default: - __assume(0); - } - - switch(wmt) - { - case CLAMP_REPEAT: - if(mask & 0xf000) {if(vr.y < v.y) vr.y = v.y; if(vr.w > v.w + 1) vr.w = v.w + 1;} - break; - case CLAMP_CLAMP: - case CLAMP_REGION_CLAMP: - if(vr.y < uv.y) vr.y = uv.y; - if(vr.w > uv.w + 1) vr.w = uv.w + 1; - break; - case CLAMP_REGION_REPEAT: // TODO - break; - default: - __assume(0); - } - } - - r = vr.rintersect(GSVector4i(0, 0, w, h)); -} - -void GSRendererSW::GetAlphaMinMax() -{ - if(m_vtrace.m_alpha.valid) - { - return; - } - - const GSDrawingEnvironment& env = m_env; - const GSDrawingContext* context = m_context; - - GSVector4i a = GSVector4i(m_vtrace.m_min.c.wwww(m_vtrace.m_max.c)) >> 7; - - if(PRIM->TME && context->TEX0.TCC) - { - uint32 bpp = GSLocalMemory::m_psm[context->TEX0.PSM].trbpp; - uint32 cbpp = GSLocalMemory::m_psm[context->TEX0.CPSM].trbpp; - uint32 pal = GSLocalMemory::m_psm[context->TEX0.PSM].pal; - - if(bpp == 32) - { - a.y = 0; - a.w = 0xff; - } - else if(bpp == 24) - { - a.y = env.TEXA.AEM ? 0 : env.TEXA.TA0; - a.w = env.TEXA.TA0; - } - else if(bpp == 16) - { - a.y = env.TEXA.AEM ? 0 : min(env.TEXA.TA0, env.TEXA.TA1); - a.w = max(env.TEXA.TA0, env.TEXA.TA1); - } - else - { - m_mem.m_clut.GetAlphaMinMax32(a.y, a.w); - } - - switch(context->TEX0.TFX) - { - case TFX_MODULATE: - a.x = (a.x * a.y) >> 7; - a.z = (a.z * a.w) >> 7; - if(a.x > 0xff) a.x = 0xff; - if(a.z > 0xff) a.z = 0xff; - break; - case TFX_DECAL: - a.x = a.y; - a.z = a.w; - break; - case TFX_HIGHLIGHT: - a.x = a.x + a.y; - a.z = a.z + a.w; - if(a.x > 0xff) a.x = 0xff; - if(a.z > 0xff) a.z = 0xff; - break; - case TFX_HIGHLIGHT2: - a.x = a.y; - a.z = a.w; - break; - default: - __assume(0); - } - } - - m_vtrace.m_alpha.min = a.x; - m_vtrace.m_alpha.max = a.z; - m_vtrace.m_alpha.valid = true; -} - -bool GSRendererSW::TryAlphaTest(uint32& fm, uint32& zm) -{ - const GSDrawingContext* context = m_context; - - bool pass = true; - - if(context->TEST.ATST == ATST_NEVER) - { - pass = false; - } - else if(context->TEST.ATST != ATST_ALWAYS) - { - GetAlphaMinMax(); - - int amin = m_vtrace.m_alpha.min; - int amax = m_vtrace.m_alpha.max; - - int aref = context->TEST.AREF; - - switch(context->TEST.ATST) - { - case ATST_NEVER: - pass = false; - break; - case ATST_ALWAYS: - pass = true; - break; - case ATST_LESS: - if(amax < aref) pass = true; - else if(amin >= aref) pass = false; - else return false; - break; - case ATST_LEQUAL: - if(amax <= aref) pass = true; - else if(amin > aref) pass = false; - else return false; - break; - case ATST_EQUAL: - if(amin == aref && amax == aref) pass = true; - else if(amin > aref || amax < aref) pass = false; - else return false; - break; - case ATST_GEQUAL: - if(amin >= aref) pass = true; - else if(amax < aref) pass = false; - else return false; - break; - case ATST_GREATER: - if(amin > aref) pass = true; - else if(amax <= aref) pass = false; - else return false; - break; - case ATST_NOTEQUAL: - if(amin == aref && amax == aref) pass = false; - else if(amin > aref || amax < aref) pass = true; - else return false; - break; - default: - __assume(0); - } - } - - if(!pass) - { - switch(context->TEST.AFAIL) - { - case AFAIL_KEEP: fm = zm = 0xffffffff; break; - case AFAIL_FB_ONLY: zm = 0xffffffff; break; - case AFAIL_ZB_ONLY: fm = 0xffffffff; break; - case AFAIL_RGB_ONLY: fm |= 0xff000000; zm = 0xffffffff; break; - default: __assume(0); - } - } - - return true; -} - void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) { const GSDrawingEnvironment& env = m_env; @@ -574,7 +319,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) { p.sel.fpsm = GSUtil::EncodePSM(context->FRAME.PSM); - if((primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS) && m_vtrace.m_eq.rgba != 15) + if((primclass == GS_LINE_CLASS || primclass == GS_TRIANGLE_CLASS) && m_vt.m_eq.rgba != 0xffff) { p.sel.iip = PRIM->IIP; } @@ -589,14 +334,11 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) p.sel.wms = context->CLAMP.WMS; p.sel.wmt = context->CLAMP.WMT; - if(p.sel.iip == 0 && p.sel.tfx == TFX_MODULATE && p.sel.tcc) + if(/*p.sel.iip == 0 &&*/ p.sel.tfx == TFX_MODULATE && p.sel.tcc && m_vt.m_eq.rgba == 0xffff && m_vt.m_min.c.eq(GSVector4i(128))) { - if(m_vtrace.m_eq.rgba == 15 && (m_vtrace.m_min.c == GSVector4(128.0f * 128.0f)).alltrue()) - { - // modulate does not do anything when vertex color is 0x80 + // modulate does not do anything when vertex color is 0x80 - p.sel.tfx = TFX_DECAL; - } + p.sel.tfx = TFX_DECAL; } if(p.sel.fst == 0) @@ -605,7 +347,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) GSVertexSW* v = m_vertices; - if(m_vtrace.m_eq.q) + if(m_vt.m_eq.q) { p.sel.fst = 1; @@ -617,44 +359,29 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) { v[i].t *= w; } - - m_vtrace.m_min.t *= w; - m_vtrace.m_max.t *= w; } } else if(primclass == GS_SPRITE_CLASS) { p.sel.fst = 1; - GSVector4 tmin = GSVector4(FLT_MAX); - GSVector4 tmax = GSVector4(-FLT_MAX); - for(int i = 0, j = m_count; i < j; i += 2) { GSVector4 w = v[i + 1].t.zzzz().rcpnr(); - GSVector4 v0 = v[i + 0].t * w; - GSVector4 v1 = v[i + 1].t * w; - - v[i + 0].t = v0; - v[i + 1].t = v1; - - tmin = tmin.minv(v0).minv(v1); - tmax = tmax.maxv(v0).maxv(v1); + v[i + 0].t *= w; + v[i + 1].t *= w; } - - m_vtrace.m_max.t = tmax; - m_vtrace.m_min.t = tmin; } } - if(p.sel.fst) + if(p.sel.ltf) { - // if q is constant we can do the half pel shift for bilinear sampling on the vertices + GSVector4 half(0x8000, 0x8000); - if(p.sel.ltf) + if(p.sel.fst) { - GSVector4 half(0x8000, 0x8000); + // if q is constant we can do the half pel shift for bilinear sampling on the vertices GSVertexSW* v = m_vertices; @@ -662,46 +389,12 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) { v[i].t -= half; } - - m_vtrace.m_min.t -= half; - m_vtrace.m_max.t += half; } } - /* - else - { - GSVector4 tmin = GSVector4(FLT_MAX); - GSVector4 tmax = GSVector4(-FLT_MAX); - - GSVertexSW* v = m_vertices; - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 v0 = v[i].t * v[i].t.zzzz().rcpnr(); - - tmin = tmin.minv(v0); - tmax = tmax.maxv(v0); - } - - if(p.sel.ltf) - { - GSVector4 half(0x8000, 0x8000); - - tmin -= half; - tmax += half; - } - - m_vtrace.min.t = tmin; - m_vtrace.max.t = tmax; - } - */ - - int w = 1 << context->TEX0.TW; - int h = 1 << context->TEX0.TH; GSVector4i r; - GetTextureMinMax(w, h, r, p.sel.fst); + GetTextureMinMax(r); const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, r); @@ -720,31 +413,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) p.sel.datm = context->TEST.DATM; } - int amin = 0, amax = 0xff; - - if(PRIM->ABE && context->ALPHA.A != context->ALPHA.B && !PRIM->AA1) - { - if(context->ALPHA.C == 0) - { - GetAlphaMinMax(); - - amin = m_vtrace.m_alpha.min; - amax = m_vtrace.m_alpha.max; - } - else if(context->ALPHA.C == 1) - { - if(p.sel.fpsm == 1) - { - amin = amax = 0x80; - } - } - else if(context->ALPHA.C == 1) - { - amin = amax = context->ALPHA.FIX; - } - } - - if(PRIM->ABE && !context->ALPHA.IsOpaque(amin, amax) || PRIM->AA1) + if(!IsOpaque()) { p.sel.abe = PRIM->ABE; p.sel.ababcd = context->ALPHA.u32[0]; @@ -785,7 +454,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) { p.sel.zpsm = GSUtil::EncodePSM(context->ZBUF.PSM); p.sel.ztst = ztest ? context->TEST.ZTST : 1; - p.sel.zoverflow = GSVector4i(m_vtrace.m_max.p).z == 0x80000000; + p.sel.zoverflow = GSVector4i(m_vt.m_max.p).z == 0x80000000; } } diff --git a/plugins/GSdx/GSRendererSW.h b/plugins/GSdx/GSRendererSW.h index 0be46daee9..542e53e6b3 100644 --- a/plugins/GSdx/GSRendererSW.h +++ b/plugins/GSdx/GSRendererSW.h @@ -30,7 +30,6 @@ class GSRendererSW : public GSRendererT protected: GSRasterizerList m_rl; GSTextureCacheSW* m_tc; - GSVertexTrace m_vtrace; GSTexture* m_texture[2]; bool m_reset; @@ -42,9 +41,6 @@ protected: void Draw(); void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); - void GetTextureMinMax(int w, int h, GSVector4i& r, uint32 fst); - void GetAlphaMinMax(); - bool TryAlphaTest(uint32& fm, uint32& zm); void GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass); public: diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 421e650ec6..1062861db8 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -169,7 +169,21 @@ GSTexture10::operator ID3D10ShaderResourceView*() { if(!m_srv && m_dev && m_texture) { - m_dev->CreateShaderResourceView(m_texture, NULL, &m_srv); + D3D10_SHADER_RESOURCE_VIEW_DESC* desc = NULL; + + if(m_desc.Format == DXGI_FORMAT_R32G8X24_TYPELESS) + { + desc = new D3D10_SHADER_RESOURCE_VIEW_DESC(); + memset(desc, 0, sizeof(*desc)); + desc->Format = DXGI_FORMAT_R32_FLOAT_X8X24_TYPELESS; + desc->ViewDimension = D3D10_SRV_DIMENSION_TEXTURE2D; + desc->Texture2D.MostDetailedMip = 0; + desc->Texture2D.MipLevels = 1; + } + + m_dev->CreateShaderResourceView(m_texture, desc, &m_srv); + + delete desc; } return m_srv; @@ -191,7 +205,19 @@ GSTexture10::operator ID3D10DepthStencilView*() { if(!m_dsv && m_dev && m_texture) { - m_dev->CreateDepthStencilView(m_texture, NULL, &m_dsv); + D3D10_DEPTH_STENCIL_VIEW_DESC* desc = NULL; + + if(m_desc.Format == DXGI_FORMAT_R32G8X24_TYPELESS) + { + desc = new D3D10_DEPTH_STENCIL_VIEW_DESC(); + memset(desc, 0, sizeof(*desc)); + desc->Format = DXGI_FORMAT_D32_FLOAT_S8X24_UINT; + desc->ViewDimension = D3D10_DSV_DIMENSION_TEXTURE2D; + } + + m_dev->CreateDepthStencilView(m_texture, desc, &m_dsv); + + delete desc; } return m_dsv; diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index df2a17ac10..28ae26542c 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -198,7 +198,7 @@ GSTextureCache::GSDepthStencil* GSTextureCache::GetDepthStencil(const GIFRegTEX0 return ds; } -GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() +GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture(const GSVector4i& r) { const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; @@ -207,61 +207,6 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; const uint32* clut = m_renderer->m_mem.m_clut; - if(psm.pal > 0) - { - m_renderer->m_mem.m_clut.Read32(TEX0, TEXA); - - /* - POSITION pos = m_tex.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_tex.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_tex.RemoveAt(cur); - - delete s; - } - } - - pos = m_rt.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_rt.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_rt.RemoveAt(cur); - - delete s; - } - } - - pos = m_ds.GetHeadPosition(); - - while(pos) - { - POSITION cur = pos; - - GSSurface* s = m_ds.GetNext(pos); - - if(s->m_TEX0.TBP0 == TEX0.CBP) - { - m_ds.RemoveAt(cur); - - delete s; - } - } - */ - } - GSCachedTexture* t = NULL; for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) @@ -378,7 +323,7 @@ GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture() } } - t->Update(); + t->Update(r); m_tex_used = true; @@ -757,7 +702,7 @@ GSTextureCache::GSCachedTexture::~GSCachedTexture() _aligned_free(m_clut); } -void GSTextureCache::GSCachedTexture::Update() +void GSTextureCache::GSCachedTexture::Update(const GSVector4i& rect) { __super::Update(); @@ -766,7 +711,7 @@ void GSTextureCache::GSCachedTexture::Update() return; } - GSVector4i r; + GSVector4i r = rect; if(!GetDirtyRect(r)) { @@ -799,16 +744,16 @@ void GSTextureCache::GSCachedTexture::Update() m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * 4); } -bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& rr) +bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& r) { int w = 1 << m_TEX0.TW; int h = 1 << m_TEX0.TH; - GSVector4i r(0, 0, w, h); + GSVector4i tr(0, 0, w, h); for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) { - const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(r); + const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(tr); if(!m_valid.rintersect(dirty).rempty()) { @@ -837,8 +782,6 @@ bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& rr) m_dirty.clear(); - m_renderer->MinMaxUV(w, h, r); - if(GSUtil::IsRectInRect(r, m_valid)) { return false; @@ -867,7 +810,5 @@ bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& rr) return false; } - rr = r; - return true; } \ No newline at end of file diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index ad17e46eb3..300d648bfd 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -85,7 +85,7 @@ public: explicit GSCachedTexture(GSRenderer* renderer); virtual ~GSCachedTexture(); - void Update(); + void Update(const GSVector4i& rect); virtual bool Create() = 0; virtual bool Create(GSRenderTarget* rt) = 0; @@ -130,7 +130,7 @@ public: GSRenderTarget* GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb = false); GSDepthStencil* GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h); - GSCachedTexture* GetTexture(); + GSCachedTexture* GetTexture(const GSVector4i& r); void InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF); void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 2ec9f37db2..0c20e24c24 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -142,6 +142,8 @@ bool GSTextureCache10::GSCachedTextureHW10::Create() bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) { + m_rendered = true; + // TODO: clean up this mess rt->Update(); @@ -151,8 +153,6 @@ bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) m_TEX0 = m_renderer->m_context->TEX0; m_TEXA = m_renderer->m_env.TEXA; - m_rendered = true; - int tw = 1 << m_TEX0.TW; int th = 1 << m_TEX0.TH; int tp = (int)m_TEX0.TW << 6; @@ -301,7 +301,92 @@ bool GSTextureCache10::GSCachedTextureHW10::Create(GSDepthStencil* ds) { m_rendered = true; - // TODO - return false; +/* + // TODO: clean up this mess + + ds->Update(); + + // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + int tp = (int)m_TEX0.TW << 6; + + // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) + + int w = (int)(ds->m_texture->m_scale.x * tw); + int h = (int)(ds->m_texture->m_scale.y * th); + + GSVector2i dssize = ds->m_texture->GetSize(); + + // pitch conversion + + if(ds->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM + { + ASSERT(0); + } + else if(tw < tp) + { + } + + // width/height conversion + + GSVector2 scale = ds->m_texture->m_scale; + + GSVector4 dst(0, 0, w, h); + + if(w > dssize.x) + { + scale.x = (float)dssize.x / tw; + dst.z = (float)dssize.x * scale.x / ds->m_texture->m_scale.x; + w = dssize.x; + } + + if(h > dssize.y) + { + scale.y = (float)dssize.y / th; + dst.w = (float)dssize.y * scale.y / ds->m_texture->m_scale.y; + h = dssize.y; + } + + m_texture = m_renderer->m_dev->CreateRenderTarget(w, h); + + GSVector4 src(0, 0, w, h); + + src.z /= ds->m_texture->GetWidth(); + src.w /= ds->m_texture->GetHeight(); + + m_renderer->m_dev->StretchRect(ds->m_texture, src, m_texture, dst, 7); + + m_texture->m_scale = scale; + + switch(m_TEX0.PSM) + { + case PSM_PSMCT32: + m_bpp = 0; + break; + case PSM_PSMCT24: + m_bpp = 1; + break; + case PSM_PSMCT16: + case PSM_PSMCT16S: + m_bpp = 2; + break; + case PSM_PSMT8H: + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); + m_initpalette = true; + break; + case PSM_PSMT4HL: + case PSM_PSMT4HH: + ASSERT(0); // TODO + break; + } + + return true; +*/ } diff --git a/plugins/GSdx/GSTextureFX.cpp b/plugins/GSdx/GSTextureFX.cpp new file mode 100644 index 0000000000..16a86f2475 --- /dev/null +++ b/plugins/GSdx/GSTextureFX.cpp @@ -0,0 +1,23 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSTextureFX.h" diff --git a/plugins/GSdx/GSTextureFX.h b/plugins/GSdx/GSTextureFX.h new file mode 100644 index 0000000000..ed338a59b7 --- /dev/null +++ b/plugins/GSdx/GSTextureFX.h @@ -0,0 +1,243 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSVector.h" + +class GSTextureFX +{ +public: + #pragma pack(push, 1) + + __declspec(align(16)) struct VSConstantBuffer + { + GSVector4 VertexScale; + GSVector4 VertexOffset; + GSVector2 TextureScale; + float _pad[2]; + + struct VSConstantBuffer() + { + memset(this, 0, sizeof(*this)); + } + + __forceinline bool Update(const VSConstantBuffer* cb) + { + GSVector4i* a = (GSVector4i*)this; + GSVector4i* b = (GSVector4i*)cb; + + GSVector4i b0 = b[0]; + GSVector4i b1 = b[1]; + GSVector4i b2 = b[2]; + + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2)).alltrue()) + { + a[0] = b0; + a[1] = b1; + a[2] = b2; + + return true; + } + + return false; + } + }; + + struct VSSelector + { + union + { + struct + { + uint32 bppz:2; + uint32 tme:1; + uint32 fst:1; + uint32 logz:1; + uint32 prim:2; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x7f;} + + VSSelector() : key(0) {} + }; + + __declspec(align(16)) struct PSConstantBuffer + { + GSVector4 FogColor_AREF; + GSVector4 HalfTexel; + GSVector4 WH_TA; + GSVector4 MinMax; + GSVector4 MinMaxF; + GSVector4i MskFix; + + struct PSConstantBuffer() + { + memset(this, 0, sizeof(*this)); + } + + __forceinline bool Update(const PSConstantBuffer* cb) + { + GSVector4i* a = (GSVector4i*)this; + GSVector4i* b = (GSVector4i*)cb; + + GSVector4i b0 = b[0]; + GSVector4i b1 = b[1]; + GSVector4i b2 = b[2]; + GSVector4i b3 = b[3]; + GSVector4i b4 = b[4]; + GSVector4i b5 = b[5]; + + if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) + { + a[0] = b0; + a[1] = b1; + a[2] = b2; + a[3] = b3; + a[4] = b4; + a[5] = b5; + + return true; + } + + return false; + } + }; + + struct GSSelector + { + union + { + struct + { + uint32 iip:1; + uint32 prim:2; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x7;} + + GSSelector() : key(0) {} + }; + + struct PSSelector + { + union + { + struct + { + uint32 fst:1; + uint32 wms:2; + uint32 wmt:2; + uint32 bpp:3; + uint32 aem:1; + uint32 tfx:3; + uint32 tcc:1; + uint32 ate:1; + uint32 atst:3; + uint32 fog:1; + uint32 clr1:1; + uint32 fba:1; + uint32 aout:1; + uint32 rt:1; + uint32 ltf:1; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x7fffff;} + + PSSelector() : key(0) {} + }; + + struct PSSamplerSelector + { + union + { + struct + { + uint32 tau:1; + uint32 tav:1; + uint32 ltf:1; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x7;} + + PSSamplerSelector() : key(0) {} + }; + + struct OMDepthStencilSelector + { + union + { + struct + { + uint32 zte:1; + uint32 ztst:2; + uint32 zwe:1; + uint32 date:1; + uint32 fba:1; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x3f;} + + OMDepthStencilSelector() : key(0) {} + }; + + struct OMBlendSelector + { + union + { + struct + { + uint32 abe:1; + uint32 a:2; + uint32 b:2; + uint32 c:2; + uint32 d:2; + uint32 wr:1; + uint32 wg:1; + uint32 wb:1; + uint32 wa:1; + }; + + uint32 key; + }; + + operator uint32() {return key & 0x1fff;} + + OMBlendSelector() : key(0) {} + }; + + #pragma pack(pop) +}; diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 3005ca5918..0a34979799 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -36,10 +36,6 @@ bool GSTextureFX10::Create(GSDevice10* dev) VSSelector sel; - sel.bppz = 0; - sel.tme = 0; - sel.fst = 0; - VSConstantBuffer cb; SetupVS(sel, &cb); // creates layout @@ -122,11 +118,11 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) D3D10_INPUT_ELEMENT_DESC layout[] = { - {"POSITION", 0, DXGI_FORMAT_R16G16_UINT, 0, 8, D3D10_INPUT_PER_VERTEX_DATA, 0}, - {"POSITION", 1, DXGI_FORMAT_R32_UINT, 0, 12, D3D10_INPUT_PER_VERTEX_DATA, 0}, {"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 0, D3D10_INPUT_PER_VERTEX_DATA, 0}, - {"TEXCOORD", 1, DXGI_FORMAT_R32_FLOAT, 0, 20, D3D10_INPUT_PER_VERTEX_DATA, 0}, - {"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 16, D3D10_INPUT_PER_VERTEX_DATA, 0}, + {"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 8, D3D10_INPUT_PER_VERTEX_DATA, 0}, + {"TEXCOORD", 1, DXGI_FORMAT_R32_FLOAT, 0, 12, D3D10_INPUT_PER_VERTEX_DATA, 0}, + {"POSITION", 0, DXGI_FORMAT_R16G16_UINT, 0, 16, D3D10_INPUT_PER_VERTEX_DATA, 0}, + {"POSITION", 1, DXGI_FORMAT_R32_UINT, 0, 20, D3D10_INPUT_PER_VERTEX_DATA, 0}, {"COLOR", 1, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 28, D3D10_INPUT_PER_VERTEX_DATA, 0}, }; @@ -515,8 +511,3 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, m_dev->OMSetBlendState((*j).second, bf); } - -void GSTextureFX10::Draw() -{ - m_dev->DrawPrimitive(); -} diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 285d930538..9e3c18a5ee 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -21,188 +21,11 @@ #pragma once +#include "GSTextureFX.h" #include "GSDevice10.h" -class GSTextureFX10 +class GSTextureFX10 : public GSTextureFX { -public: - #pragma pack(push, 1) - - __declspec(align(16)) struct VSConstantBuffer - { - GSVector4 VertexScale; - GSVector4 VertexOffset; - GSVector2 TextureScale; - float _pad[2]; - - struct VSConstantBuffer() {memset(this, 0, sizeof(*this));} - - __forceinline bool Update(const VSConstantBuffer* cb) - { - GSVector4i* a = (GSVector4i*)this; - GSVector4i* b = (GSVector4i*)cb; - - GSVector4i b0 = b[0]; - GSVector4i b1 = b[1]; - GSVector4i b2 = b[2]; - - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2)).alltrue()) - { - a[0] = b0; - a[1] = b1; - a[2] = b2; - - return true; - } - - return false; - } - }; - - union VSSelector - { - struct - { - uint32 bppz:2; - uint32 tme:1; - uint32 fst:1; - uint32 prim:3; - }; - - uint32 key; - - operator uint32() {return key & 0x7f;} - }; - - __declspec(align(16)) struct PSConstantBuffer - { - GSVector4 FogColor_AREF; - GSVector4 HalfTexel; - GSVector4 WH_TA; - GSVector4 MinMax; - GSVector4 MinMaxF; - GSVector4i MskFix; - - struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} - - __forceinline bool Update(const PSConstantBuffer* cb) - { - GSVector4i* a = (GSVector4i*)this; - GSVector4i* b = (GSVector4i*)cb; - - GSVector4i b0 = b[0]; - GSVector4i b1 = b[1]; - GSVector4i b2 = b[2]; - GSVector4i b3 = b[3]; - GSVector4i b4 = b[4]; - GSVector4i b5 = b[5]; - - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) - { - a[0] = b0; - a[1] = b1; - a[2] = b2; - a[3] = b3; - a[4] = b4; - a[5] = b5; - - return true; - } - - return false; - } - }; - - union GSSelector - { - struct - { - uint32 iip:1; - uint32 prim:2; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} - }; - - union PSSelector - { - struct - { - uint32 fst:1; - uint32 wms:2; - uint32 wmt:2; - uint32 bpp:3; - uint32 aem:1; - uint32 tfx:3; - uint32 tcc:1; - uint32 ate:1; - uint32 atst:3; - uint32 fog:1; - uint32 clr1:1; - uint32 fba:1; - uint32 aout:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x3fffff;} - }; - - union PSSamplerSelector - { - struct - { - uint32 tau:1; - uint32 tav:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} - }; - - union OMDepthStencilSelector - { - struct - { - uint32 zte:1; - uint32 ztst:2; - uint32 zwe:1; - uint32 date:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1f;} - }; - - union OMBlendSelector - { - struct - { - uint32 abe:1; - uint32 a:2; - uint32 b:2; - uint32 c:2; - uint32 d:2; - uint32 wr:1; - uint32 wg:1; - uint32 wb:1; - uint32 wa:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1fff;} - }; - - #pragma pack(pop) - -private: GSDevice10* m_dev; CComPtr m_il; hash_map > m_vs; @@ -215,11 +38,6 @@ private: hash_map > m_om_dss; hash_map > m_om_bs; - CComPtr m_vb, m_vb_old; - int m_vb_max; - int m_vb_start; - int m_vb_count; - VSConstantBuffer m_vs_cb_cache; PSConstantBuffer m_ps_cb_cache; @@ -236,5 +54,4 @@ public: void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); - void Draw(); }; diff --git a/plugins/GSdx/GSTextureFX11.cpp b/plugins/GSdx/GSTextureFX11.cpp index 4fe3de6a28..d3c1205888 100644 --- a/plugins/GSdx/GSTextureFX11.cpp +++ b/plugins/GSdx/GSTextureFX11.cpp @@ -36,10 +36,6 @@ bool GSTextureFX11::Create(GSDevice11* dev) VSSelector sel; - sel.bppz = 0; - sel.tme = 0; - sel.fst = 0; - VSConstantBuffer cb; SetupVS(sel, &cb); // creates layout @@ -122,11 +118,11 @@ bool GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) D3D11_INPUT_ELEMENT_DESC layout[] = { - {"POSITION", 0, DXGI_FORMAT_R16G16_UINT, 0, 8, D3D11_INPUT_PER_VERTEX_DATA, 0}, - {"POSITION", 1, DXGI_FORMAT_R32_UINT, 0, 12, D3D11_INPUT_PER_VERTEX_DATA, 0}, {"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 0, D3D11_INPUT_PER_VERTEX_DATA, 0}, - {"TEXCOORD", 1, DXGI_FORMAT_R32_FLOAT, 0, 20, D3D11_INPUT_PER_VERTEX_DATA, 0}, - {"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 16, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 8, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"TEXCOORD", 1, DXGI_FORMAT_R32_FLOAT, 0, 12, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"POSITION", 0, DXGI_FORMAT_R16G16_UINT, 0, 16, D3D11_INPUT_PER_VERTEX_DATA, 0}, + {"POSITION", 1, DXGI_FORMAT_R32_UINT, 0, 20, D3D11_INPUT_PER_VERTEX_DATA, 0}, {"COLOR", 1, DXGI_FORMAT_R8G8B8A8_UNORM, 0, 28, D3D11_INPUT_PER_VERTEX_DATA, 0}, }; @@ -519,8 +515,3 @@ void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, m_dev->OMSetBlendState((*j).second, bf); } - -void GSTextureFX11::Draw() -{ - m_dev->DrawPrimitive(); -} diff --git a/plugins/GSdx/GSTextureFX11.h b/plugins/GSdx/GSTextureFX11.h index e64b9349ab..53a4f2f337 100644 --- a/plugins/GSdx/GSTextureFX11.h +++ b/plugins/GSdx/GSTextureFX11.h @@ -21,188 +21,11 @@ #pragma once +#include "GSTextureFX.h" #include "GSDevice11.h" -class GSTextureFX11 +class GSTextureFX11 : public GSTextureFX { -public: - #pragma pack(push, 1) - - __declspec(align(16)) struct VSConstantBuffer - { - GSVector4 VertexScale; - GSVector4 VertexOffset; - GSVector2 TextureScale; - float _pad[2]; - - struct VSConstantBuffer() {memset(this, 0, sizeof(*this));} - - __forceinline bool Update(const VSConstantBuffer* cb) - { - GSVector4i* a = (GSVector4i*)this; - GSVector4i* b = (GSVector4i*)cb; - - GSVector4i b0 = b[0]; - GSVector4i b1 = b[1]; - GSVector4i b2 = b[2]; - - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2)).alltrue()) - { - a[0] = b0; - a[1] = b1; - a[2] = b2; - - return true; - } - - return false; - } - }; - - union VSSelector - { - struct - { - uint32 bppz:2; - uint32 tme:1; - uint32 fst:1; - uint32 prim:3; - }; - - uint32 key; - - operator uint32() {return key & 0x7f;} - }; - - __declspec(align(16)) struct PSConstantBuffer - { - GSVector4 FogColor_AREF; - GSVector4 HalfTexel; - GSVector4 WH_TA; - GSVector4 MinMax; - GSVector4 MinMaxF; - GSVector4i MskFix; - - struct PSConstantBuffer() {memset(this, 0, sizeof(*this));} - - __forceinline bool Update(const PSConstantBuffer* cb) - { - GSVector4i* a = (GSVector4i*)this; - GSVector4i* b = (GSVector4i*)cb; - - GSVector4i b0 = b[0]; - GSVector4i b1 = b[1]; - GSVector4i b2 = b[2]; - GSVector4i b3 = b[3]; - GSVector4i b4 = b[4]; - GSVector4i b5 = b[5]; - - if(!((a[0] == b0) & (a[1] == b1) & (a[2] == b2) & (a[3] == b3) & (a[4] == b4) & (a[5] == b5)).alltrue()) - { - a[0] = b0; - a[1] = b1; - a[2] = b2; - a[3] = b3; - a[4] = b4; - a[5] = b5; - - return true; - } - - return false; - } - }; - - union GSSelector - { - struct - { - uint32 iip:1; - uint32 prim:2; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} - }; - - union PSSelector - { - struct - { - uint32 fst:1; - uint32 wms:2; - uint32 wmt:2; - uint32 bpp:3; - uint32 aem:1; - uint32 tfx:3; - uint32 tcc:1; - uint32 ate:1; - uint32 atst:3; - uint32 fog:1; - uint32 clr1:1; - uint32 fba:1; - uint32 aout:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x3fffff;} - }; - - union PSSamplerSelector - { - struct - { - uint32 tau:1; - uint32 tav:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} - }; - - union OMDepthStencilSelector - { - struct - { - uint32 zte:1; - uint32 ztst:2; - uint32 zwe:1; - uint32 date:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1f;} - }; - - union OMBlendSelector - { - struct - { - uint32 abe:1; - uint32 a:2; - uint32 b:2; - uint32 c:2; - uint32 d:2; - uint32 wr:1; - uint32 wg:1; - uint32 wb:1; - uint32 wa:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1fff;} - }; - - #pragma pack(pop) - -private: GSDevice11* m_dev; CComPtr m_il; hash_map > m_vs; @@ -215,11 +38,6 @@ private: hash_map > m_om_dss; hash_map > m_om_bs; - CComPtr m_vb, m_vb_old; - int m_vb_max; - int m_vb_start; - int m_vb_count; - VSConstantBuffer m_vs_cb_cache; PSConstantBuffer m_ps_cb_cache; @@ -236,5 +54,4 @@ public: void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); - void Draw(); }; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index ac64eed885..5327f2d93f 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -34,11 +34,6 @@ bool GSTextureFX9::Create(GSDevice9* dev) VSSelector sel; - sel.bppz = 0; - sel.tme = 0; - sel.fst = 0; - sel.logz = 0; - VSConstantBuffer cb; SetupVS(sel, &cb); // creates layout @@ -455,8 +450,3 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, m_dev->OMSetBlendState((*j).second, 0x010101 * bf); } - -void GSTextureFX9::Draw() -{ - m_dev->DrawPrimitive(); -} diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index b47c227a7b..50f187a3bf 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -21,123 +21,11 @@ #pragma once +#include "GSTextureFX.h" #include "GSDevice9.h" -class GSTextureFX9 +class GSTextureFX9 : public GSTextureFX { -public: - #pragma pack(push, 1) - - struct VSConstantBuffer - { - GSVector4 VertexScale; - GSVector4 VertexOffset; - GSVector2 TextureScale; - float _pad[2]; - }; - - union VSSelector - { - struct - { - uint32 bppz:2; - uint32 tme:1; - uint32 fst:1; - uint32 logz:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1f;} - }; - - struct PSConstantBuffer - { - GSVector4 FogColor_AREF; - GSVector4 HalfTexel; - GSVector4 WH_TA; - GSVector4 MinMax; - GSVector4 MinMaxF; - GSVector4i MskFix; - }; - - union PSSelector - { - struct - { - uint32 fst:1; - uint32 wms:2; - uint32 wmt:2; - uint32 bpp:3; - uint32 aem:1; - uint32 tfx:3; - uint32 tcc:1; - uint32 ate:1; - uint32 atst:3; - uint32 fog:1; - uint32 clr1:1; - uint32 rt:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1fffff;} - }; - - union PSSamplerSelector - { - struct - { - uint32 tau:1; - uint32 tav:1; - uint32 ltf:1; - }; - - uint32 key; - - operator uint32() {return key & 0x7;} - }; - - union OMDepthStencilSelector - { - struct - { - uint32 zte:1; - uint32 ztst:2; - uint32 zwe:1; - uint32 date:1; - uint32 fba:1; - }; - - uint32 key; - - operator uint32() {return key & 0x3f;} - }; - - union OMBlendSelector - { - struct - { - uint32 abe:1; - uint32 a:2; - uint32 b:2; - uint32 c:2; - uint32 d:2; - uint32 wr:1; - uint32 wg:1; - uint32 wb:1; - uint32 wa:1; - }; - - uint32 key; - - operator uint32() {return key & 0x1fff;} - }; - - #pragma pack(pop) - -private: GSDevice9* m_dev; CComPtr m_il; hash_map > m_vs; @@ -162,5 +50,4 @@ public: void SetupRS(int w, int h, const GSVector4i& scissor); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds); void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); - void Draw(); }; diff --git a/plugins/GSdx/GSVertex.h b/plugins/GSdx/GSVertex.h index ee3bdc9874..9634a61aef 100644 --- a/plugins/GSdx/GSVertex.h +++ b/plugins/GSdx/GSVertex.h @@ -35,8 +35,8 @@ __declspec(align(16)) struct GSVertex struct { GIFRegST ST; - GIFRegXYZ XYZ; GIFRegRGBAQ RGBAQ; + GIFRegXYZ XYZ; GIFRegFOG FOG; }; diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index 76a84ed74d..9948a76b1d 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -21,6 +21,7 @@ #pragma once +#include "GS.h" #include "GSVector.h" #pragma pack(push, 1) @@ -53,17 +54,17 @@ __declspec(align(16)) union GSVertexHW10 GIFRegST ST; }; - union - { - struct {union {struct {uint16 x, y;}; uint32 xy;}; uint32 z;} p; - GIFRegXYZ XYZ; - }; - union { union {struct {uint8 r, g, b, a; float q;}; uint32 c0;}; GIFRegRGBAQ RGBAQ; }; + + union + { + struct {union {struct {uint16 x, y;}; uint32 xy;}; uint32 z;} p; + GIFRegXYZ XYZ; + }; union { diff --git a/plugins/GSdx/GSVertexSW.cpp b/plugins/GSdx/GSVertexSW.cpp index 43702ed972..0706fa00f8 100644 --- a/plugins/GSdx/GSVertexSW.cpp +++ b/plugins/GSdx/GSVertexSW.cpp @@ -23,145 +23,3 @@ #include "stdafx.h" #include "GSVertexSW.h" - -using namespace Xbyak; - -GSVertexTrace::GSVertexTraceCodeGenerator::GSVertexTraceCodeGenerator(uint32 key, void* ptr, size_t maxsize) - : CodeGenerator(maxsize, ptr) -{ - #if _M_AMD64 - #error TODO - #endif - - const int params = 0; - - uint32 primclass = (key >> 0) & 3; - uint32 iip = (key >> 2) & 1; - uint32 tme = (key >> 3) & 1; - uint32 color = (key >> 4) & 1; - - int n = 1; - - switch(primclass) - { - case GS_POINT_CLASS: - n = 1; - break; - case GS_LINE_CLASS: - case GS_SPRITE_CLASS: - n = 2; - break; - case GS_TRIANGLE_CLASS: - n = 3; - break; - } - - const int _v = params + 4; - const int _count = params + 8; - const int _min = params + 12; - const int _max = params + 16; - - // - - static const float fmin = -FLT_MAX; - static const float fmax = FLT_MAX; - - movss(xmm0, xmmword[&fmax]); - shufps(xmm0, xmm0, _MM_SHUFFLE(0, 0, 0, 0)); - - movss(xmm1, xmmword[&fmin]); - shufps(xmm1, xmm1, _MM_SHUFFLE(0, 0, 0, 0)); - - if(color) - { - // min.c = FLT_MAX; - // max.c = -FLT_MAX; - - movaps(xmm2, xmm0); - movaps(xmm3, xmm1); - } - - // min.p = FLT_MAX; - // max.p = -FLT_MAX; - - movaps(xmm4, xmm0); - movaps(xmm5, xmm1); - - if(tme) - { - // min.t = FLT_MAX; - // max.t = -FLT_MAX; - - movaps(xmm6, xmm0); - movaps(xmm7, xmm1); - } - - // for(int i = 0; i < count; i += step) { - - mov(edx, dword[esp + _v]); - mov(ecx, dword[esp + _count]); - - align(16); - -L("loop"); - - for(int j = 0; j < n; j++) - { - if(color && (iip || j == n - 1)) - { - // min.c = min.c.minv(v[i + j].c); - // max.c = max.c.maxv(v[i + j].c); - - movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW)]); - - minps(xmm2, xmm0); - maxps(xmm3, xmm0); - } - - // min.p = min.p.minv(v[i + j].p); - // max.p = max.p.maxv(v[i + j].p); - - movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW) + 16]); - - minps(xmm4, xmm0); - maxps(xmm5, xmm0); - - if(tme) - { - // min.t = min.t.minv(v[i + j].t); - // max.t = max.t.maxv(v[i + j].t); - - movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW) + 32]); - - minps(xmm6, xmm0); - maxps(xmm7, xmm0); - } - } - - add(edx, n * sizeof(GSVertexSW)); - sub(ecx, n); - - jg("loop"); - - // } - - mov(eax, dword[esp + _min]); - mov(edx, dword[esp + _max]); - - if(color) - { - movaps(xmmword[eax], xmm2); - movaps(xmmword[edx], xmm3); - } - - movaps(xmmword[eax + 16], xmm4); - movaps(xmmword[edx + 16], xmm5); - - if(tme) - { - movaps(xmmword[eax + 32], xmm6); - movaps(xmmword[edx + 32], xmm7); - } - - ret(); -} diff --git a/plugins/GSdx/GSVertexSW.h b/plugins/GSdx/GSVertexSW.h index 3f51aa756a..e3e181d0ef 100644 --- a/plugins/GSdx/GSVertexSW.h +++ b/plugins/GSdx/GSVertexSW.h @@ -214,81 +214,3 @@ __forceinline GSVertexSW operator / (const GSVertexSW& v, float f) return v0; } -#include "GSFunctionMap.h" -#include "xbyak/xbyak.h" - -__declspec(align(16)) class GSVertexTrace -{ - class GSVertexTraceCodeGenerator : public Xbyak::CodeGenerator - { - public: - GSVertexTraceCodeGenerator(uint32 key, void* ptr, size_t maxsize); - }; - - typedef void (*VertexTracePtr)(const GSVertexSW* v, int count, GSVertexSW& min, GSVertexSW& max); - - class GSVertexTraceMap : public GSCodeGeneratorFunctionMap - { - public: - GSVertexTraceMap() : GSCodeGeneratorFunctionMap("VertexTrace") {} - GSVertexTraceCodeGenerator* Create(uint32 key, void* ptr, size_t maxsize) {return new GSVertexTraceCodeGenerator(key, ptr, maxsize);} - } m_map; - -public: - GSVertexSW m_min, m_max; - - struct {int min, max; bool valid;} m_alpha; // source alpha range after tfx - - union - { - uint32 value; - struct {uint32 x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1, r:1, g:1, b:1, a:1;}; - struct {uint32 xyzf:4, stq:4, rgba:4;}; - } m_eq; - - void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, uint32 iip, uint32 tme, uint32 tfx, uint32 tcc) - { - uint32 key = primclass | (iip << 2) | (tme << 3); - - if(!(tme && tfx == TFX_DECAL && tcc)) - { - key |= 1 << 4; - } - - m_map[key](v, count, m_min, m_max); - - m_eq.value = (m_min.p == m_max.p).mask() | ((m_min.t == m_max.t).mask() << 4) | ((m_min.c == m_max.c).mask() << 8); - - m_alpha.valid = false; - } -/* - void Update(const GSVertexSW* v, int count) - { - GSVertexSW min, max; - - min.c = v[0].c; - max.c = v[0].c; - min.t = v[0].t; - max.t = v[0].t; - min.p = v[0].p; - max.p = v[0].p; - - for(int i = 1; i < count; i++) - { - min.c = min.c.minv(v[i].c); - max.c = max.c.maxv(v[i].c); - min.p = min.p.minv(v[i].p); - max.p = max.p.maxv(v[i].p); - min.t = min.t.minv(v[i].t); - max.t = max.t.maxv(v[i].t); - } - - m_min = min; - m_max = max; - - m_eq.value = (min.p == max.p).mask() | ((min.t == max.t).mask() << 4) | ((min.c == max.c).mask() << 8); - - m_alpha.valid = false; - } -*/ -}; diff --git a/plugins/GSdx/GSVertexTrace.cpp b/plugins/GSdx/GSVertexTrace.cpp new file mode 100644 index 0000000000..2f58e9a555 --- /dev/null +++ b/plugins/GSdx/GSVertexTrace.cpp @@ -0,0 +1,670 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "stdafx.h" +#include "GSVertexTrace.h" + +void GSVertexTrace::Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context) +{ + uint32 key = primclass | (PRIM->IIP << 2) | (PRIM->TME << 3) | (PRIM->FST << 4); + + if(!(PRIM->TME && context->TEX0.TFX == TFX_DECAL && context->TEX0.TCC)) + { + key |= 1 << 5; + } + + m_map_sw[key](v, count, m_min, m_max); + + m_eq.value = (m_min.c == m_max.c).mask() | ((m_min.p == m_max.p).mask() << 16) | ((m_min.t == m_max.t).mask() << 20); + + m_alpha.valid = false; +} + +void GSVertexTrace::Update(const GSVertexHW9* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context) +{ + uint32 key = primclass | (PRIM->IIP << 2) | (PRIM->TME << 3) | (PRIM->FST << 4); + + if(!(PRIM->TME && context->TEX0.TFX == TFX_DECAL && context->TEX0.TCC)) + { + key |= 1 << 5; + } + + m_map_hw9[key](v, count, m_min, m_max); + + GSVector4 o(context->XYOFFSET); + GSVector4 s(1.0f / 16, 1.0f / 16, 1.0f, 1.0f); + + m_min.p = (m_min.p - o) * s; + m_max.p = (m_max.p - o) * s; + + if(PRIM->TME) + { + if(PRIM->FST) + { + s = GSVector4(1 << (16 - 4), 1).xxyy(); + } + else + { + s = GSVector4(0x10000 << context->TEX0.TW, 0x10000 << context->TEX0.TH, 1, 1); + } + + m_min.t *= s; + m_max.t *= s; + } + + m_eq.value = (m_min.c == m_max.c).mask() | ((m_min.p == m_max.p).mask() << 16) | ((m_min.t == m_max.t).mask() << 20); + + m_alpha.valid = false; +} + +void GSVertexTrace::Update(const GSVertexHW10* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context) +{ + uint32 key = primclass | (PRIM->IIP << 2) | (PRIM->TME << 3) | (PRIM->FST << 4); + + if(!(PRIM->TME && context->TEX0.TFX == TFX_DECAL && context->TEX0.TCC)) + { + key |= 1 << 5; + } + + m_map_hw10[key](v, count, m_min, m_max); + + GSVector4 o(context->XYOFFSET); + GSVector4 s(1.0f / 16, 1.0f / 16, 2.0f, 1.0f); + + m_min.p = (m_min.p - o) * s; + m_max.p = (m_max.p - o) * s; + + if(PRIM->TME) + { + if(PRIM->FST) + { + s = GSVector4(1 << (16 - 4), 1).xxyy(); + } + else + { + s = GSVector4(0x10000 << context->TEX0.TW, 0x10000 << context->TEX0.TH, 1, 1); + } + + m_min.t *= s; + m_max.t *= s; + } + + m_eq.value = (m_min.c == m_max.c).mask() | ((m_min.p == m_max.p).mask() << 16) | ((m_min.t == m_max.t).mask() << 20); + + m_alpha.valid = false; +} + +using namespace Xbyak; + +GSVertexTrace::CGSW::CGSW(uint32 key, void* ptr, size_t maxsize) + : CodeGenerator(maxsize, ptr) +{ + #if _M_AMD64 + #error TODO + #endif + + const int params = 0; + + uint32 primclass = (key >> 0) & 3; + uint32 iip = (key >> 2) & 1; + uint32 tme = (key >> 3) & 1; + uint32 fst = (key >> 4) & 1; + uint32 color = (key >> 5) & 1; + + int n = 1; + + switch(primclass) + { + case GS_POINT_CLASS: + n = 1; + break; + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: + n = 2; + break; + case GS_TRIANGLE_CLASS: + n = 3; + break; + } + + const int _v = params + 4; + const int _count = params + 8; + const int _min = params + 12; + const int _max = params + 16; + + // + + static const float fmin = -FLT_MAX; + static const float fmax = FLT_MAX; + + movss(xmm0, xmmword[&fmax]); + shufps(xmm0, xmm0, _MM_SHUFFLE(0, 0, 0, 0)); + + movss(xmm1, xmmword[&fmin]); + shufps(xmm1, xmm1, _MM_SHUFFLE(0, 0, 0, 0)); + + if(color) + { + // min.c = FLT_MAX; + // max.c = -FLT_MAX; + + movaps(xmm2, xmm0); + movaps(xmm3, xmm1); + } + + // min.p = FLT_MAX; + // max.p = -FLT_MAX; + + movaps(xmm4, xmm0); + movaps(xmm5, xmm1); + + if(tme) + { + // min.t = FLT_MAX; + // max.t = -FLT_MAX; + + movaps(xmm6, xmm0); + movaps(xmm7, xmm1); + } + + // for(int i = 0; i < count; i += step) { + + mov(edx, dword[esp + _v]); + mov(ecx, dword[esp + _count]); + + align(16); + +L("loop"); + + if(tme && !fst && primclass == GS_SPRITE_CLASS) + { + movaps(xmm1, xmmword[edx + 1 * sizeof(GSVertexSW) + 32]); + shufps(xmm1, xmm1, _MM_SHUFFLE(2, 2, 2, 2)); + } + + for(int j = 0; j < n; j++) + { + if(color && (iip || j == n - 1)) + { + // min.c = min.c.minv(v[i + j].c); + // max.c = max.c.maxv(v[i + j].c); + + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW)]); + + minps(xmm2, xmm0); + maxps(xmm3, xmm0); + } + + // min.p = min.p.minv(v[i + j].p); + // max.p = max.p.maxv(v[i + j].p); + + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW) + 16]); + + minps(xmm4, xmm0); + maxps(xmm5, xmm0); + + if(tme) + { + // min.t = min.t.minv(v[i + j].t); + // max.t = max.t.maxv(v[i + j].t); + + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexSW) + 32]); + + if(!fst) + { + if(primclass != GS_SPRITE_CLASS) + { + movaps(xmm1, xmm0); + shufps(xmm1, xmm1, _MM_SHUFFLE(2, 2, 2, 2)); + } + + divps(xmm0, xmm1); + shufps(xmm0, xmm1, _MM_SHUFFLE(3, 2, 1, 0)); + } + + minps(xmm6, xmm0); + maxps(xmm7, xmm0); + } + } + + add(edx, n * sizeof(GSVertexSW)); + sub(ecx, n); + + jg("loop"); + + // } + + mov(eax, dword[esp + _min]); + mov(edx, dword[esp + _max]); + + if(color) + { + cvttps2dq(xmm2, xmm2); + psrld(xmm2, 7); + movaps(xmmword[eax], xmm2); + + cvttps2dq(xmm3, xmm3); + psrld(xmm3, 7); + movaps(xmmword[edx], xmm3); + } + + movaps(xmmword[eax + 16], xmm4); + movaps(xmmword[edx + 16], xmm5); + + if(tme) + { + movaps(xmmword[eax + 32], xmm6); + movaps(xmmword[edx + 32], xmm7); + } + + ret(); +} + +GSVertexTrace::CGHW9::CGHW9(uint32 key, void* ptr, size_t maxsize) + : CodeGenerator(maxsize, ptr) +{ + #if _M_AMD64 + #error TODO + #endif + + const int params = 0; + + uint32 primclass = (key >> 0) & 3; + uint32 iip = (key >> 2) & 1; + uint32 tme = (key >> 3) & 1; + uint32 fst = (key >> 4) & 1; + uint32 color = (key >> 5) & 1; + + int n = 1; + + switch(primclass) + { + case GS_POINT_CLASS: + n = 1; + break; + case GS_LINE_CLASS: + n = 2; + break; + case GS_TRIANGLE_CLASS: + n = 3; + break; + case GS_SPRITE_CLASS: + n = 6; + break; + } + + const int _v = params + 4; + const int _count = params + 8; + const int _min = params + 12; + const int _max = params + 16; + + // + + static const float fmin = -FLT_MAX; + static const float fmax = FLT_MAX; + + movss(xmm0, xmmword[&fmax]); + shufps(xmm0, xmm0, _MM_SHUFFLE(0, 0, 0, 0)); + + movss(xmm1, xmmword[&fmin]); + shufps(xmm1, xmm1, _MM_SHUFFLE(0, 0, 0, 0)); + + if(color) + { + // min.c = 0xffffffff; + // max.c = 0; + + pcmpeqd(xmm2, xmm2); + pxor(xmm3, xmm3); + } + + // min.p = FLT_MAX; + // max.p = -FLT_MAX; + + movaps(xmm4, xmm0); + movaps(xmm5, xmm1); + + if(tme) + { + // min.t = FLT_MAX; + // max.t = -FLT_MAX; + + movaps(xmm6, xmm0); + movaps(xmm7, xmm1); + } + + // for(int i = 0; i < count; i += step) { + + mov(edx, dword[esp + _v]); + mov(ecx, dword[esp + _count]); + + align(16); + +L("loop"); + + if(tme && !fst && primclass == GS_SPRITE_CLASS) + { + movaps(xmm1, xmmword[edx + 5 * sizeof(GSVertexHW9) + 16]); + shufps(xmm1, xmm1, _MM_SHUFFLE(3, 3, 3, 3)); + } + + for(int j = 0; j < n; j++) + { + // min.p = min.p.minv(v[i + j].p); + // max.p = max.p.maxv(v[i + j].p); + + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexHW9) + 16]); + + minps(xmm4, xmm0); + maxps(xmm5, xmm0); + + if(tme && !fst && primclass != GS_SPRITE_CLASS) + { + movaps(xmm1, xmm0); + shufps(xmm1, xmm1, _MM_SHUFFLE(3, 3, 3, 3)); + } + + if(color && (iip || j == n - 1) || tme) + { + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexHW9)]); + } + + if(color && (iip || j == n - 1)) + { + // min.c = min.c.min_u8(v[i + j].c); + // max.c = max.c.min_u8(v[i + j].c); + + pminub(xmm2, xmm0); + pmaxub(xmm3, xmm0); + } + + if(tme) + { + shufps(xmm0, xmm0, _MM_SHUFFLE(1, 0, 1, 0)); // avoid FP assist, high part is integral + + if(!fst) + { + // t /= p.wwww(); + + divps(xmm0, xmm1); + } + + // min.t = min.t.minv(v[i + j].t); + // max.t = max.t.maxv(v[i + j].t); + + minps(xmm6, xmm0); + maxps(xmm7, xmm0); + } + } + + add(edx, n * sizeof(GSVertexHW9)); + sub(ecx, n); + + jg("loop"); + + // } + + mov(eax, dword[esp + _min]); + mov(edx, dword[esp + _max]); + + if(color) + { + // m_min.c = cmin.zzzz().u8to32(); + // m_max.c = cmax.zzzz().u8to32(); + + if(m_cpu.has(util::Cpu::tSSE41)) + { + pshufd(xmm2, xmm2, _MM_SHUFFLE(2, 2, 2, 2)); + pmovzxbd(xmm2, xmm2); + + pshufd(xmm3, xmm3, _MM_SHUFFLE(2, 2, 2, 2)); + pmovzxbd(xmm3, xmm3); + } + else + { + pxor(xmm0, xmm0); + + punpckhbw(xmm2, xmm0); + punpcklwd(xmm2, xmm0); + + punpckhbw(xmm3, xmm0); + punpcklwd(xmm3, xmm0); + } + + movaps(xmmword[eax], xmm2); + movaps(xmmword[edx], xmm3); + } + + // m_min.p = pmin; + // m_max.p = pmax; + + movaps(xmmword[eax + 16], xmm4); + movaps(xmmword[edx + 16], xmm5); + + if(tme) + { + // m_min.t = tmin.xyww(pmin); + // m_max.t = tmax.xyww(pmax); + + shufps(xmm6, xmm4, _MM_SHUFFLE(3, 3, 1, 0)); + shufps(xmm7, xmm5, _MM_SHUFFLE(3, 3, 1, 0)); + + movaps(xmmword[eax + 32], xmm6); + movaps(xmmword[edx + 32], xmm7); + } + + ret(); +} + +GSVertexTrace::CGHW10::CGHW10(uint32 key, void* ptr, size_t maxsize) + : CodeGenerator(maxsize, ptr) +{ + #if _M_AMD64 + #error TODO + #endif + + const int params = 0; + + uint32 primclass = (key >> 0) & 3; + uint32 iip = (key >> 2) & 1; + uint32 tme = (key >> 3) & 1; + uint32 fst = (key >> 4) & 1; + uint32 color = (key >> 5) & 1; + + int n = 1; + + switch(primclass) + { + case GS_POINT_CLASS: + n = 1; + break; + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: + n = 2; + break; + case GS_TRIANGLE_CLASS: + n = 3; + break; + } + + const int _v = params + 4; + const int _count = params + 8; + const int _min = params + 12; + const int _max = params + 16; + + // + + static const float fmin = -FLT_MAX; + static const float fmax = FLT_MAX; + + movss(xmm0, xmmword[&fmax]); + shufps(xmm0, xmm0, _MM_SHUFFLE(0, 0, 0, 0)); + + movss(xmm1, xmmword[&fmin]); + shufps(xmm1, xmm1, _MM_SHUFFLE(0, 0, 0, 0)); + + if(color) + { + // min.c = 0xffffffff; + // max.c = 0; + + pcmpeqd(xmm2, xmm2); + pxor(xmm3, xmm3); + } + + // min.p = FLT_MAX; + // max.p = -FLT_MAX; + + movaps(xmm4, xmm0); + movaps(xmm5, xmm1); + + if(tme) + { + // min.t = FLT_MAX; + // max.t = -FLT_MAX; + + movaps(xmm6, xmm0); + movaps(xmm7, xmm1); + } + + // for(int i = 0; i < count; i += step) { + + mov(edx, dword[esp + _v]); + mov(ecx, dword[esp + _count]); + + align(16); + +L("loop"); + + for(int j = 0; j < n; j++) + { + if(color && (iip || j == n - 1) || tme) + { + movaps(xmm0, xmmword[edx + j * sizeof(GSVertexHW10)]); + } + + if(color && (iip || j == n - 1)) + { + pminub(xmm2, xmm0); + pmaxub(xmm3, xmm0); + } + + if(tme) + { + if(!fst) + { + movaps(xmm1, xmm0); + } + + shufps(xmm0, xmm0, _MM_SHUFFLE(3, 3, 1, 0)); // avoid FP assist, third dword is integral + + if(!fst) + { + shufps(xmm1, xmm1, _MM_SHUFFLE(3, 3, 3, 3)); + divps(xmm0, xmm1); + shufps(xmm0, xmm1, _MM_SHUFFLE(3, 3, 1, 0)); // restore q + } + + minps(xmm6, xmm0); + maxps(xmm7, xmm0); + } + + movdqa(xmm0, xmmword[edx + j * sizeof(GSVertexHW10) + 16]); + + if(m_cpu.has(util::Cpu::tSSE41)) + { + pmovzxwd(xmm1, xmm0); + } + else + { + movdqa(xmm1, xmm0); + punpcklwd(xmm1, xmm1); + psrld(xmm1, 16); + } + + psrld(xmm0, 1); + punpcklqdq(xmm1, xmm0); + cvtdq2ps(xmm1, xmm1); + + minps(xmm4, xmm1); + maxps(xmm5, xmm1); + } + + add(edx, n * sizeof(GSVertexHW10)); + sub(ecx, n); + + jg("loop"); + + // } + + mov(eax, dword[esp + _min]); + mov(edx, dword[esp + _max]); + + if(color) + { + // m_min.c = cmin.zzzz().u8to32(); + // m_max.c = cmax.zzzz().u8to32(); + + if(m_cpu.has(util::Cpu::tSSE41)) + { + pshufd(xmm2, xmm2, _MM_SHUFFLE(2, 2, 2, 2)); + pmovzxbd(xmm2, xmm2); + + pshufd(xmm3, xmm3, _MM_SHUFFLE(2, 2, 2, 2)); + pmovzxbd(xmm3, xmm3); + } + else + { + pxor(xmm0, xmm0); + + punpckhbw(xmm2, xmm0); + punpcklwd(xmm2, xmm0); + + punpckhbw(xmm3, xmm0); + punpcklwd(xmm3, xmm0); + } + + movaps(xmmword[eax], xmm2); + movaps(xmmword[edx], xmm3); + } + + // m_min.p = pmin.xyww(); + // m_max.p = pmax.xyww(); + + shufps(xmm4, xmm4, _MM_SHUFFLE(3, 3, 1, 0)); + shufps(xmm5, xmm5, _MM_SHUFFLE(3, 3, 1, 0)); + + movaps(xmmword[eax + 16], xmm4); + movaps(xmmword[edx + 16], xmm5); + + if(tme) + { + // m_min.t = tmin; + // m_max.t = tmax; + + movaps(xmmword[eax + 32], xmm6); + movaps(xmmword[edx + 32], xmm7); + } + + ret(); +} \ No newline at end of file diff --git a/plugins/GSdx/GSVertexTrace.h b/plugins/GSdx/GSVertexTrace.h new file mode 100644 index 0000000000..802513fe2a --- /dev/null +++ b/plugins/GSdx/GSVertexTrace.h @@ -0,0 +1,99 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSDrawingContext.h" +#include "GSVertexSW.h" +#include "GSVertexHW.h" +#include "GSFunctionMap.h" +#include "xbyak/xbyak.h" +#include "xbyak/xbyak_util.h" + +__declspec(align(16)) class GSVertexTrace +{ + struct Vertex {GSVector4i c; GSVector4 p, t;}; + struct VertexAlpha {int min, max; bool valid;}; + + typedef void (*VertexTracePtr)(const void* v, int count, Vertex& min, Vertex& max); + + class CGSW : public Xbyak::CodeGenerator + { + public: + CGSW(uint32 key, void* ptr, size_t maxsize); + }; + + class GSVertexTraceMapSW : public GSCodeGeneratorFunctionMap + { + public: + GSVertexTraceMapSW() : GSCodeGeneratorFunctionMap("VertexTraceSW") {} + CGSW* Create(uint32 key, void* ptr, size_t maxsize) {return new CGSW(key, ptr, maxsize);} + }; + + class CGHW9 : public Xbyak::CodeGenerator + { + Xbyak::util::Cpu m_cpu; + + public: + CGHW9(uint32 key, void* ptr, size_t maxsize); + }; + + class GSVertexTraceMapHW9 : public GSCodeGeneratorFunctionMap + { + public: + GSVertexTraceMapHW9() : GSCodeGeneratorFunctionMap("VertexTraceHW9") {} + CGHW9* Create(uint32 key, void* ptr, size_t maxsize) {return new CGHW9(key, ptr, maxsize);} + }; + + class CGHW10 : public Xbyak::CodeGenerator + { + Xbyak::util::Cpu m_cpu; + + public: + CGHW10(uint32 key, void* ptr, size_t maxsize); + }; + + class GSVertexTraceMapHW10 : public GSCodeGeneratorFunctionMap + { + public: + GSVertexTraceMapHW10() : GSCodeGeneratorFunctionMap("VertexTraceHW10") {} + CGHW10* Create(uint32 key, void* ptr, size_t maxsize) {return new CGHW10(key, ptr, maxsize);} + }; + + GSVertexTraceMapSW m_map_sw; + GSVertexTraceMapHW9 m_map_hw9; + GSVertexTraceMapHW10 m_map_hw10; + +public: + Vertex m_min, m_max; // t.xy * 0x10000 + VertexAlpha m_alpha; // source alpha range after tfx, GSRenderer::GetAlphaMinMax() updates it + + union + { + uint32 value; + struct {uint32 r:4, g:4, b:4, a:4, x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1;}; + struct {uint32 rgba:16, xyzf:4, stq:4;}; + } m_eq; + + void Update(const GSVertexSW* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context); + void Update(const GSVertexHW9* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context); + void Update(const GSVertexHW10* v, int count, GS_PRIM_CLASS primclass, const GIFRegPRIM* PRIM, const GSDrawingContext* context); +}; diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index aa3a060e89..7addc2fb08 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1440,6 +1440,10 @@ /> + + @@ -1496,6 +1500,10 @@ RelativePath=".\GSVertexSW.cpp" > + + @@ -1878,6 +1886,10 @@ RelativePath=".\GSTextureCacheSW.h" > + + @@ -1926,6 +1938,10 @@ RelativePath=".\GSVertexSW.h" > + + diff --git a/plugins/GSdx/res/tfx.fx b/plugins/GSdx/res/tfx.fx index ebd24e290b..84ebbe190e 100644 --- a/plugins/GSdx/res/tfx.fx +++ b/plugins/GSdx/res/tfx.fx @@ -345,7 +345,7 @@ VS_OUTPUT vs_main(VS_INPUT input) input.z = input.z & 0xffff; } - if(VS_PRIM == 6) // sprite + if(VS_PRIM == 3) // sprite { //input.p.xy = (input.p.xy + 15) & ~15; // HACK } From c1fc72a634f99b45a5a79e4a897fa688fca55c18 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Tue, 23 Jun 2009 12:33:30 +0000 Subject: [PATCH 278/296] GSdx: small fixes to the last commit git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1426 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSCrc.cpp | 227 ++++++++++++----------- plugins/GSdx/GSCrc.h | 7 + plugins/GSdx/GSRendererHW.h | 2 +- plugins/GSdx/GSRendererHW10.cpp | 55 +----- plugins/GSdx/GSRendererHW11.cpp | 55 +----- plugins/GSdx/GSRendererHW9.cpp | 69 ++----- plugins/GSdx/GSTexture10.cpp | 2 +- plugins/GSdx/GSTexture11.cpp | 2 +- plugins/GSdx/GSdx_vs2010.vcxproj | 4 + plugins/GSdx/GSdx_vs2010.vcxproj.filters | 12 ++ 10 files changed, 179 insertions(+), 256 deletions(-) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index fa70de419d..a5c6977c03 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -24,119 +24,120 @@ CRC::Game CRC::m_games[] = { - {0x00000000, None, Unknown}, - {0x2113EA2E, MetalSlug6, Unknown}, - {0x42E05BAF, TomoyoAfter, JP}, - {0x7800DC84, Clannad, JP}, - {0xA6167B59, Lamune, JP}, - {0xDDB59F46, KyuuketsuKitanMoonties, JP}, - {0xa39517ab, FFX, EU}, - {0xa39517ae, FFX, FR}, - {0x941bb7d9, FFX, DE}, - {0xa39517a9, FFX, IT}, - {0x941bb7de, FFX, ES}, - {0xb4414ea1, FFX, RU}, - {0xee97db5b, FFX, RU}, - {0xaec495cc, FFX, RU}, - {0xbb3d833a, FFX, US}, - {0x6a4efe60, FFX, JP}, - {0x3866ca7e, FFX, ASIA}, // int. - {0x658597e2, FFX, JP}, // int. - {0x9aac5309, FFX2, EU}, - {0x9aac530c, FFX2, FR}, - {0x9aac530a, FFX2, FR}, // ? - {0x9aac530d, FFX2, DE}, - {0x9aac530b, FFX2, IT}, - {0x48fe0c71, FFX2, US}, - {0xe1fd9a2d, FFX2, JP}, // int. - {0x78da0252, FFXII, EU}, - {0xc1274668, FFXII, EU}, - {0xdc2a467e, FFXII, EU}, - {0xca284668, FFXII, EU}, - {0x280AD120, FFXII, JP}, - {0x8BE3D7B2, ShadowHearts, Unknown}, - {0xDEFA4763, ShadowHearts, US}, - {0x21068223, Okami, US}, - {0x891f223f, Okami, FR}, - {0xC5DEFEA0, Okami, JP}, - {0x053D2239, MetalGearSolid3, US}, - {0x086273D2, MetalGearSolid3, FR}, - {0x26A6E286, MetalGearSolid3, EU}, - {0xAA31B5BF, MetalGearSolid3, Unknown}, - {0x9F185CE1, MetalGearSolid3, Unknown}, - {0x98D4BC93, MetalGearSolid3, EU}, - {0x86BC3040, MetalGearSolid3, US}, //Subsistance disc 1 - {0x0481AD8A, MetalGearSolid3, JP}, - {0x79ED26AD, MetalGearSolid3, EU}, - {0x5E31EA42, MetalGearSolid3, EU}, - {0x278722BF, DBZBT2, US}, - {0xFE961D28, DBZBT2, US}, - {0x0393B6BE, DBZBT2, EU}, - {0xE2F289ED, DBZBT2, JP}, // Sparking Neo! - {0x35AA84D1, DBZBT2, Unknown}, - {0x428113C2, DBZBT3, US}, - {0xA422BB13, DBZBT3, EU}, - {0x983C53D2, DBZBT3, Unknown}, - {0x983C53D3, DBZBT3, Unknown}, - {0x72B3802A, SFEX3, US}, - {0x71521863, SFEX3, US}, - {0x28703748, Bully, US}, - {0xC78A495D, BullyCC, US}, - {0xC19A374E, SoTC, US}, - {0x7D8F539A, SoTC, EU}, - {0x3122B508, OnePieceGrandAdventure, US}, - {0x8DF14A24, OnePieceGrandAdventure, Unknown}, - {0x6F8545DB, ICO, US}, - {0xB01A4C95, ICO, JP}, - {0x5C991F4E, ICO, Unknown}, - {0x7ACF7E03, ICO, Unknown}, - {0xAEAD1CA3, GT4, JP}, - {0x44A61C8F, GT4, Unknown}, - {0x0086E35B, GT4, Unknown}, - {0x77E61C8A, GT4, Unknown}, - {0xC164550A, WildArms5, JPUNDUB}, - {0xC1640D2C, WildArms5, US}, - {0x0FCF8FE4, WildArms5, EU}, - {0x2294D322, WildArms5, JP}, - {0x565B6170, WildArms5, JP}, - {0x8B029334, Manhunt2, Unknown}, - {0x09F49E37, CrashBandicootWoC, Unknown}, - {0x013E349D, ResidentEvil4, US}, - {0x6BA2F6B9, ResidentEvil4, Unknown}, - {0x60FA8C69, ResidentEvil4, JP}, - {0x72E1E60E, Spartan, Unknown}, - {0x5ED8FB53, AceCombat4, JP}, - {0x1B9B7563, AceCombat4, Unknown}, - {0xEC432B24, Drakengard2, Unknown}, - {0xFC46EA61, Tekken5, JP}, - {0x1F88EE37, Tekken5, Unknown}, - {0x652050D2, Tekken5, Unknown}, - {0x9E98B8AE, IkkiTousen, JP}, - {0xD6385328, GodOfWar, US}, - {0xFB0E6D72, GodOfWar, EU}, - {0xEB001875, GodOfWar, EU}, - {0xA61A4C6D, GodOfWar, Unknown}, - {0xE23D532B, GodOfWar, Unknown}, - {0xDF1AF973, GodOfWar, Unknown}, - {0xD6385328, GodOfWar, Unknown}, - {0x2F123FD8, GodOfWar2, RU}, - {0x2F123FD8, GodOfWar2, US}, - {0x44A8A22A, GodOfWar2, EU}, - {0x4340C7C6, GodOfWar2, Unknown}, - {0x5D482F18, JackieChanAdv, Unknown}, - {0xf0a6d880, HarvestMoon, US}, - {0x75c01a04, NamcoXCapcom, US}, - {0xBF6F101F, GiTS, US}, - {0xA5768F53, GiTS, JP}, - {0x6BF11378, Onimusha3, US}, - {0xF442260C, MajokkoALaMode2, JP}, - {0x14FE77F7, TalesOfAbyss, US}, - {0x045D77E9, TalesOfAbyss, US}, // undub - {0xAA5EC3A3, TalesOfAbyss, JP}, - {0xFB236A46, SonicUnleashed, US}, - {0x4C7BB3C8, SimpsonsGame, Unknown}, - {0x4C94B32C, SimpsonsGame, Unknown}, - {0xD71B57F4, Genji, Unknown}, + {0x00000000, None, Unknown, 0}, + {0x2113EA2E, MetalSlug6, Unknown, 0}, + {0x42E05BAF, TomoyoAfter, JP, PointListPalette}, + {0x7800DC84, Clannad, JP, PointListPalette}, + {0xA6167B59, Lamune, JP, PointListPalette}, + {0xDDB59F46, KyuuketsuKitanMoonties, JP, PointListPalette}, + {0xC8EE2562, PiaCarroteYoukosoGPGakuenPrincess, JP, PointListPalette}, + {0xa39517ab, FFX, EU, 0}, + {0xa39517ae, FFX, FR, 0}, + {0x941bb7d9, FFX, DE, 0}, + {0xa39517a9, FFX, IT, 0}, + {0x941bb7de, FFX, ES, 0}, + {0xb4414ea1, FFX, RU, 0}, + {0xee97db5b, FFX, RU, 0}, + {0xaec495cc, FFX, RU, 0}, + {0xbb3d833a, FFX, US, 0}, + {0x6a4efe60, FFX, JP, 0}, + {0x3866ca7e, FFX, ASIA, 0}, // int. + {0x658597e2, FFX, JP, 0}, // int. + {0x9aac5309, FFX2, EU, 0}, + {0x9aac530c, FFX2, FR, 0}, + {0x9aac530a, FFX2, FR, 0}, // ? + {0x9aac530d, FFX2, DE, 0}, + {0x9aac530b, FFX2, IT, 0}, + {0x48fe0c71, FFX2, US, 0}, + {0xe1fd9a2d, FFX2, JP, 0}, // int. + {0x78da0252, FFXII, EU, 0}, + {0xc1274668, FFXII, EU, 0}, + {0xdc2a467e, FFXII, EU, 0}, + {0xca284668, FFXII, EU, 0}, + {0x280AD120, FFXII, JP, 0}, + {0x8BE3D7B2, ShadowHearts, Unknown, 0}, + {0xDEFA4763, ShadowHearts, US, 0}, + {0x21068223, Okami, US, 0}, + {0x891f223f, Okami, FR, 0}, + {0xC5DEFEA0, Okami, JP, 0}, + {0x053D2239, MetalGearSolid3, US, 0}, + {0x086273D2, MetalGearSolid3, FR, 0}, + {0x26A6E286, MetalGearSolid3, EU, 0}, + {0xAA31B5BF, MetalGearSolid3, Unknown, 0}, + {0x9F185CE1, MetalGearSolid3, Unknown, 0}, + {0x98D4BC93, MetalGearSolid3, EU, 0}, + {0x86BC3040, MetalGearSolid3, US, 0}, //Subsistance disc 1 + {0x0481AD8A, MetalGearSolid3, JP, 0}, + {0x79ED26AD, MetalGearSolid3, EU, 0}, + {0x5E31EA42, MetalGearSolid3, EU, 0}, + {0x278722BF, DBZBT2, US, 0}, + {0xFE961D28, DBZBT2, US, 0}, + {0x0393B6BE, DBZBT2, EU, 0}, + {0xE2F289ED, DBZBT2, JP, 0}, // Sparking Neo! + {0x35AA84D1, DBZBT2, Unknown, 0}, + {0x428113C2, DBZBT3, US, 0}, + {0xA422BB13, DBZBT3, EU, 0}, + {0x983C53D2, DBZBT3, Unknown, 0}, + {0x983C53D3, DBZBT3, Unknown, 0}, + {0x72B3802A, SFEX3, US, 0}, + {0x71521863, SFEX3, US, 0}, + {0x28703748, Bully, US, 0}, + {0xC78A495D, BullyCC, US, 0}, + {0xC19A374E, SoTC, US, 0}, + {0x7D8F539A, SoTC, EU, 0}, + {0x3122B508, OnePieceGrandAdventure, US, 0}, + {0x8DF14A24, OnePieceGrandAdventure, Unknown, 0}, + {0x6F8545DB, ICO, US, 0}, + {0xB01A4C95, ICO, JP, 0}, + {0x5C991F4E, ICO, Unknown, 0}, + {0x7ACF7E03, ICO, Unknown, 0}, + {0xAEAD1CA3, GT4, JP, 0}, + {0x44A61C8F, GT4, Unknown, 0}, + {0x0086E35B, GT4, Unknown, 0}, + {0x77E61C8A, GT4, Unknown, 0}, + {0xC164550A, WildArms5, JPUNDUB, 0}, + {0xC1640D2C, WildArms5, US, 0}, + {0x0FCF8FE4, WildArms5, EU, 0}, + {0x2294D322, WildArms5, JP, 0}, + {0x565B6170, WildArms5, JP, 0}, + {0x8B029334, Manhunt2, Unknown, 0}, + {0x09F49E37, CrashBandicootWoC, Unknown, 0}, + {0x013E349D, ResidentEvil4, US, 0}, + {0x6BA2F6B9, ResidentEvil4, Unknown, 0}, + {0x60FA8C69, ResidentEvil4, JP, 0}, + {0x72E1E60E, Spartan, Unknown, 0}, + {0x5ED8FB53, AceCombat4, JP, 0}, + {0x1B9B7563, AceCombat4, Unknown, 0}, + {0xEC432B24, Drakengard2, Unknown, 0}, + {0xFC46EA61, Tekken5, JP, 0}, + {0x1F88EE37, Tekken5, Unknown, 0}, + {0x652050D2, Tekken5, Unknown, 0}, + {0x9E98B8AE, IkkiTousen, JP, 0}, + {0xD6385328, GodOfWar, US, 0}, + {0xFB0E6D72, GodOfWar, EU, 0}, + {0xEB001875, GodOfWar, EU, 0}, + {0xA61A4C6D, GodOfWar, Unknown, 0}, + {0xE23D532B, GodOfWar, Unknown, 0}, + {0xDF1AF973, GodOfWar, Unknown, 0}, + {0xD6385328, GodOfWar, Unknown, 0}, + {0x2F123FD8, GodOfWar2, RU, 0}, + {0x2F123FD8, GodOfWar2, US, 0}, + {0x44A8A22A, GodOfWar2, EU, 0}, + {0x4340C7C6, GodOfWar2, Unknown, 0}, + {0x5D482F18, JackieChanAdv, Unknown, 0}, + {0xf0a6d880, HarvestMoon, US, 0}, + {0x75c01a04, NamcoXCapcom, US, 0}, + {0xBF6F101F, GiTS, US, 0}, + {0xA5768F53, GiTS, JP, 0}, + {0x6BF11378, Onimusha3, US, 0}, + {0xF442260C, MajokkoALaMode2, JP, 0}, + {0x14FE77F7, TalesOfAbyss, US, 0}, + {0x045D77E9, TalesOfAbyss, US, 0}, // undub + {0xAA5EC3A3, TalesOfAbyss, JP, 0}, + {0xFB236A46, SonicUnleashed, US, 0}, + {0x4C7BB3C8, SimpsonsGame, Unknown, 0}, + {0x4C94B32C, SimpsonsGame, Unknown, 0}, + {0xD71B57F4, Genji, Unknown, 0}, }; hash_map CRC::m_map; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index c0695d191e..d5791c7b74 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -32,6 +32,7 @@ public: Clannad, Lamune, KyuuketsuKitanMoonties, + PiaCarroteYoukosoGPGakuenPrincess, FFX, FFX2, FFXII, @@ -86,11 +87,17 @@ public: ASIA }; + enum Flags + { + PointListPalette = 1, + }; + struct Game { uint32 crc; Title title; Region region; + uint32 flags; }; private: diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 67a16e449b..392ac4066e 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -361,7 +361,7 @@ protected: #pragma region palette uploaded in a point list, pure genius... - if(m_game.title == CRC::TomoyoAfter || m_game.title == CRC::Clannad || m_game.title == CRC::Lamune || m_game.title == CRC::KyuuketsuKitanMoonties) + if(m_game.flags & CRC::PointListPalette) { if(prim == GS_POINTLIST && !PRIM->TME) { diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp index b39afae7c0..35a2da0ac6 100644 --- a/plugins/GSdx/GSRendererHW10.cpp +++ b/plugins/GSdx/GSRendererHW10.cpp @@ -468,47 +468,6 @@ void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) { // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - GSVector4 mm; - - // TODO - - mm = GSVector4(-1, -1, 1, 1); - - // if(m_count < 100) - { - GSVector4 pmin(65535, 65535, 0, 0); - GSVector4 pmax = GSVector4::zero(); - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 p(m_vertices[i].vi[0].uph16()); - - pmin = p.minv(pmin); - pmax = p.maxv(pmax); - } - - mm += pmin.xyxy(pmax); - - float sx = 2.0f * rt->m_scale.x / (w * 16); - float sy = 2.0f * rt->m_scale.y / (h * 16); - float ox = (float)(int)m_context->XYOFFSET.OFX; - float oy = (float)(int)m_context->XYOFFSET.OFY; - - mm.x = (mm.x - ox) * sx - 1; - mm.y = (mm.y - oy) * sy - 1; - mm.z = (mm.z - ox) * sx - 1; - mm.w = (mm.w - oy) * sy - 1; - - if(mm.x < -1) mm.x = -1; - if(mm.y < -1) mm.y = -1; - if(mm.z > +1) mm.z = +1; - if(mm.w > +1) mm.w = +1; - } - - GSVector4 uv = (mm + 1.0f) / 2.0f; - - // - dev->BeginScene(); dev->ClearStencil(ds, 0); @@ -521,12 +480,18 @@ void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) // ia + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + GSVertexPT1 vertices[] = { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, }; dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); diff --git a/plugins/GSdx/GSRendererHW11.cpp b/plugins/GSdx/GSRendererHW11.cpp index ac47dad0e7..e68eaad4d3 100644 --- a/plugins/GSdx/GSRendererHW11.cpp +++ b/plugins/GSdx/GSRendererHW11.cpp @@ -468,47 +468,6 @@ void GSRendererHW11::SetupDATE(GSTexture* rt, GSTexture* ds) { // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - GSVector4 mm; - - // TODO - - mm = GSVector4(-1, -1, 1, 1); - - // if(m_count < 100) - { - GSVector4 pmin(65535, 65535, 0, 0); - GSVector4 pmax = GSVector4::zero(); - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 p(m_vertices[i].vi[0].uph16()); - - pmin = p.minv(pmin); - pmax = p.maxv(pmax); - } - - mm += pmin.xyxy(pmax); - - float sx = 2.0f * rt->m_scale.x / (w * 16); - float sy = 2.0f * rt->m_scale.y / (h * 16); - float ox = (float)(int)m_context->XYOFFSET.OFX; - float oy = (float)(int)m_context->XYOFFSET.OFY; - - mm.x = (mm.x - ox) * sx - 1; - mm.y = (mm.y - oy) * sy - 1; - mm.z = (mm.z - ox) * sx - 1; - mm.w = (mm.w - oy) * sy - 1; - - if(mm.x < -1) mm.x = -1; - if(mm.y < -1) mm.y = -1; - if(mm.z > +1) mm.z = +1; - if(mm.w > +1) mm.w = +1; - } - - GSVector4 uv = (mm + 1.0f) / 2.0f; - - // - dev->BeginScene(); dev->ClearStencil(ds, 0); @@ -521,12 +480,18 @@ void GSRendererHW11::SetupDATE(GSTexture* rt, GSTexture* ds) // ia + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + GSVertexPT1 vertices[] = { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, }; dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp index cfc0d99cdc..a4d635fc87 100644 --- a/plugins/GSdx/GSRendererHW9.cpp +++ b/plugins/GSdx/GSRendererHW9.cpp @@ -464,47 +464,6 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) { // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - GSVector4 mm; - - // TODO - - mm = GSVector4(-1, -1, 1, 1); - - // if(m_count < 100) - { - GSVector4 pmin(65535, 65535, 0, 0); - GSVector4 pmax = GSVector4::zero(); - - for(int i = 0, j = m_count; i < j; i++) - { - GSVector4 p(m_vertices[i].p); - - pmin = p.minv(pmin); - pmax = p.maxv(pmax); - } - - mm += pmin.xyxy(pmax); - - float sx = 2.0f * rt->m_scale.x / (w * 16); - float sy = 2.0f * rt->m_scale.y / (h * 16); - float ox = (float)(int)m_context->XYOFFSET.OFX; - float oy = (float)(int)m_context->XYOFFSET.OFY; - - mm.x = (mm.x - ox) * sx - 1; - mm.y = (mm.y - oy) * sy - 1; - mm.z = (mm.z - ox) * sx - 1; - mm.w = (mm.w - oy) * sy - 1; - - if(mm.x < -1) mm.x = -1; - if(mm.y < -1) mm.y = -1; - if(mm.z > +1) mm.z = +1; - if(mm.w > +1) mm.w = +1; - } - - GSVector4 uv = (mm + 1.0f) / 2.0f; - - // - dev->BeginScene(); dev->ClearStencil(ds, 0); @@ -517,12 +476,18 @@ void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) // ia + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + GSVertexPT1 vertices[] = { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(uv.x, uv.y)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(uv.z, uv.y)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(uv.x, uv.w)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(uv.z, uv.w)}, + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, }; dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); @@ -568,14 +533,18 @@ void GSRendererHW9::UpdateFBA(GSTexture* rt) // ia - GSVector4 mm = GSVector4(-1, -1, 1, 1); + GSVector4 s = GSVector4(rt->m_scale.x / rt->GetWidth(), rt->m_scale.y / rt->GetHeight()); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); GSVertexPT1 vertices[] = { - {GSVector4(mm.x, -mm.y, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(mm.z, -mm.y, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(mm.x, -mm.w, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(mm.z, -mm.w, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, }; dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); diff --git a/plugins/GSdx/GSTexture10.cpp b/plugins/GSdx/GSTexture10.cpp index 1062861db8..57f88bea4d 100644 --- a/plugins/GSdx/GSTexture10.cpp +++ b/plugins/GSdx/GSTexture10.cpp @@ -78,7 +78,7 @@ bool GSTexture10::Map(GSMap& m, const GSVector4i* r) return false; } - if(m_texture) + if(m_texture && m_desc.Usage == D3D10_USAGE_STAGING) { D3D10_MAPPED_TEXTURE2D map; diff --git a/plugins/GSdx/GSTexture11.cpp b/plugins/GSdx/GSTexture11.cpp index c80cc7aae6..991128f156 100644 --- a/plugins/GSdx/GSTexture11.cpp +++ b/plugins/GSdx/GSTexture11.cpp @@ -80,7 +80,7 @@ bool GSTexture11::Map(GSMap& m, const GSVector4i* r) return false; } - if(m_texture) + if(m_texture && m_desc.Usage == D3D11_USAGE_STAGING) { D3D11_MAPPED_SUBRESOURCE map; diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj b/plugins/GSdx/GSdx_vs2010.vcxproj index 8a7fc45bba..6048be5f44 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj +++ b/plugins/GSdx/GSdx_vs2010.vcxproj @@ -1342,6 +1342,7 @@ + @@ -1352,6 +1353,7 @@ + Create @@ -1466,6 +1468,7 @@ + @@ -1478,6 +1481,7 @@ + diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj.filters b/plugins/GSdx/GSdx_vs2010.vcxproj.filters index 32c080fcf0..a9652e38e3 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj.filters +++ b/plugins/GSdx/GSdx_vs2010.vcxproj.filters @@ -267,6 +267,9 @@ Source Files + + Source Files + Source Files @@ -297,6 +300,9 @@ Source Files + + Source Files + Source Files @@ -584,6 +590,9 @@ Header Files + + Header Files + Header Files @@ -620,6 +629,9 @@ Header Files + + Header Files + Header Files From d9c40a7ccd0e42bfded9a5bd94597047a7af8a3c Mon Sep 17 00:00:00 2001 From: "Jake.Stine" Date: Wed, 24 Jun 2009 13:13:59 +0000 Subject: [PATCH 279/296] Many bugfixes and improvements to my SafeArray / SafeList utility classes. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1427 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/SafeArray.h | 180 ++++++++++++++++++++++++++++++++-------------- 1 file changed, 128 insertions(+), 52 deletions(-) diff --git a/pcsx2/SafeArray.h b/pcsx2/SafeArray.h index 2ec2f0e8ed..645ee2560f 100644 --- a/pcsx2/SafeArray.h +++ b/pcsx2/SafeArray.h @@ -16,8 +16,7 @@ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ -#ifndef __SAFEARRAY_H__ -#define __SAFEARRAY_H__ +#pragma once extern void* __fastcall pcsx2_aligned_malloc(size_t size, size_t align); extern void* __fastcall pcsx2_aligned_realloc(void* handle, size_t size, size_t align); @@ -76,21 +75,19 @@ public: static const int DefaultChunkSize = 0x1000 * sizeof(T); public: - const std::string Name; // user-assigned block name + const char* Name; // user-assigned block name int ChunkSize; protected: T* m_ptr; int m_size; // size of the allocation of memory - const static std::string m_str_Unnamed; - protected: // Internal constructor for use by derived classes. This allows a derived class to // use its own memory allocation (with an aligned memory, for example). // Throws: // Exception::OutOfMemory if the allocated_mem pointer is NULL. - explicit SafeArray( const std::string& name, T* allocated_mem, int initSize ) : + explicit SafeArray( const char* name, T* allocated_mem, int initSize ) : Name( name ) , ChunkSize( DefaultChunkSize ) , m_ptr( allocated_mem ) @@ -102,7 +99,10 @@ protected: virtual T* _virtual_realloc( int newsize ) { - return (T*)realloc( m_ptr, newsize * sizeof(T) ); + return (T*)((m_ptr == NULL) ? + malloc( newsize * sizeof(T) ) : + realloc( m_ptr, newsize * sizeof(T) ) + ); } public: @@ -111,7 +111,7 @@ public: safe_free( m_ptr ); } - explicit SafeArray( const std::string& name="Unnamed" ) : + explicit SafeArray( const char* name="Unnamed" ) : Name( name ) , ChunkSize( DefaultChunkSize ) , m_ptr( NULL ) @@ -119,39 +119,61 @@ public: { } - explicit SafeArray( int initialSize, const std::string& name="Unnamed" ) : + explicit SafeArray( int initialSize, const char* name="Unnamed" ) : Name( name ) , ChunkSize( DefaultChunkSize ) - , m_ptr( (T*)malloc( initialSize * sizeof(T) ) ) + , m_ptr( (initialSize==0) ? NULL : (T*)malloc( initialSize * sizeof(T) ) ) , m_size( initialSize ) { - if( m_ptr == NULL ) + if( (initialSize != 0) && (m_ptr == NULL) ) throw Exception::OutOfMemory(); } + + // Clears the contents of the array to zero, and frees all memory allocations. + void Dispose() + { + m_size = 0; + safe_free( m_ptr ); + } + + bool IsDisposed() const { return (m_ptr==NULL); } // Returns the size of the memory allocation, as according to the array type. int GetLength() const { return m_size; } // Returns the size of the memory allocation in bytes. int GetSizeInBytes() const { return m_size * sizeof(T); } + // reallocates the array to the explicit size. Can be used to shrink or grow an + // array, and bypasses the internal threshold growth indicators. + void ExactAlloc( int newsize ) + { + if( newsize == m_size ) return; + + m_ptr = _virtual_realloc( newsize ); + if( m_ptr == NULL ) + { + throw Exception::OutOfMemory( + "Out-of-memory on block re-allocation. " + "Old size: " + to_string( m_size ) + " bytes, " + "New size: " + to_string( newsize ) + " bytes" + ); + } + m_size = newsize; + } + // Ensures that the allocation is large enough to fit data of the // amount requested. The memory allocation is not resized smaller. - void MakeRoomFor( int blockSize ) + void MakeRoomFor( int newsize ) { - if( blockSize > m_size ) - { - const uint newalloc = blockSize + ChunkSize; - m_ptr = _virtual_realloc( newalloc ); - if( m_ptr == NULL ) - { - throw Exception::OutOfMemory( - "Out-of-memory on block re-allocation. " - "Old size: " + to_string( m_size ) + " bytes, " - "New size: " + to_string( newalloc ) + " bytes" - ); - } - m_size = newalloc; - } + if( newsize > m_size ) + ExactAlloc( newsize ); + } + + // Extends the containment area of the array. Extensions are performed + // in chunks. + void GrowBy( int items ) + { + MakeRoomFor( m_size + ChunkSize + items + 1 ); } // Gets a pointer to the requested allocation index. @@ -185,7 +207,7 @@ protected: { assert( 0 ); // makes debugging easier sometimes. :) throw Exception::IndexBoundsFault( - "Index out of bounds on SafeArray: " + Name + + "Index out of bounds on SafeArray: " + std::string(Name) + " (index=" + to_string(i) + ", size=" + to_string(m_size) + ")" ); @@ -193,20 +215,26 @@ protected: #endif return &m_ptr[i]; } - }; ////////////////////////////////////////////////////////////////////////////////////////// +// SafeList - Simple growable container without all the mess or hassle of std containers. // +// This container is intended for reasonably simple class types only. Things which this +// container does not handle with desired robustness: +// +// * Classes with non-trivial constructors (such that construction creates much overhead) +// * Classes with copy constructors (copying is done using performance memcpy) +// * Classes with destructors (they're not called, sorry!) // template< typename T > -class SafeList +class SafeList : public NoncopyableObject { public: static const int DefaultChunkSize = 0x80 * sizeof(T); public: - const std::string Name; // user-assigned block name + const char* Name; // user-assigned block name int ChunkSize; // assigned DefaultChunkSize on init, reconfigurable at any time. protected: @@ -214,22 +242,20 @@ protected: int m_allocsize; // size of the allocation of memory uint m_length; // length of the array (active items, not buffer allocation) - const static std::string m_str_Unnamed; - protected: virtual T* _virtual_realloc( int newsize ) { return (T*)realloc( m_ptr, newsize * sizeof(T) ); } - void _boundsCheck( uint i ) + void _boundsCheck( uint i ) const { #ifdef PCSX2_DEVBUILD if( i >= (uint)m_length ) { assert( 0 ); // makes debugging easier sometimes. :) throw Exception::IndexBoundsFault( - "Index out of bounds on SafeArray: " + Name + + "Index out of bounds on SafeArray: " + std::string(Name) + " (index=" + to_string(i) + ", length=" + to_string(m_length) + ")" ); @@ -240,9 +266,10 @@ protected: public: virtual ~SafeList() { + safe_free( m_ptr ); } - explicit SafeList( const std::string& name="Unnamed" ) : + explicit SafeList( const char* name="Unnamed" ) : Name( name ) , ChunkSize( DefaultChunkSize ) , m_ptr( NULL ) @@ -251,7 +278,7 @@ public: { } - explicit SafeList( int initialSize, const std::string& name="Unnamed" ) : + explicit SafeList( int initialSize, const char* name="Unnamed" ) : Name( name ) , ChunkSize( DefaultChunkSize ) , m_ptr( (T*)malloc( initialSize * sizeof(T) ) ) @@ -260,6 +287,12 @@ public: { if( m_ptr == NULL ) throw Exception::OutOfMemory(); + + for( int i=0; i m_allocsize ) { - const uint newalloc = blockSize + ChunkSize; + const int newalloc = blockSize + ChunkSize; m_ptr = _virtual_realloc( newalloc ); if( m_ptr == NULL ) { @@ -286,21 +319,57 @@ public: "New size: " + to_string( newalloc ) + " bytes" ); } - m_allocsize = newalloc; + + for( ; m_allocsize* Clone() const { SafeList* retval = new SafeList( m_length ); - memcpy_fast( retval->GetPtr(), m_ptr, sizeof(T) * m_length ); + memcpy_fast( retval->m_ptr, m_ptr, sizeof(T) * m_length ); return retval; } protected: + + void _MakeRoomFor_threshold( int newsize ) + { + MakeRoomFor( newsize + ChunkSize ); + } + // A safe array index fetcher. Throws an exception if the array index // is outside the bounds of the array. // Performance Considerations: This function adds quite a bit of overhead @@ -345,18 +420,21 @@ class SafeAlignedArray : public SafeArray protected: T* _virtual_realloc( int newsize ) { - return (T*)_aligned_realloc( this->m_ptr, newsize * sizeof(T), Alignment ); + return (T*)( ( m_ptr == NULL ) ? + _aligned_malloc( newsize * sizeof(T), Alignment ) : + _aligned_realloc( m_ptr, newsize * sizeof(T), Alignment ) + ); } // Appends "(align: xx)" to the name of the allocation in devel builds. // Maybe useful,maybe not... no harm in attaching it. :D - string _getName( const string& src ) + /*string _getName( const string& src ) { #ifdef PCSX2_DEVBUILD - return src + "(align:" + to_string(Alignment) + ")"; + return src + ; #endif return src; - } + }*/ public: virtual ~SafeAlignedArray() @@ -365,14 +443,14 @@ public: // mptr is set to null, so the parent class's destructor won't re-free it. } - explicit SafeAlignedArray( const std::string& name="Unnamed" ) : + explicit SafeAlignedArray( const char* name="Unnamed" ) : SafeArray::SafeArray( name ) { } - explicit SafeAlignedArray( int initialSize, const std::string& name="Unnamed" ) : + explicit SafeAlignedArray( int initialSize, const char* name="Unnamed" ) : SafeArray::SafeArray( - _getName(name), + name, (T*)_aligned_malloc( initialSize * sizeof(T), Alignment ), initialSize ) @@ -386,5 +464,3 @@ public: return retval; } }; - -#endif From d4132141d5eddeb1849c0ed6033d257c6a38c1c4 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Wed, 24 Jun 2009 23:17:32 +0000 Subject: [PATCH 280/296] microVU: changed VU1's cached program amount from 32 to 64, and vu0's from 32 to 8. speeds up some games that use a lot of microprograms at once (like ffxii) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1429 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 44 ++++++++++++++++++------------------- pcsx2/x86/microVU.h | 35 +++++++++++++---------------- pcsx2/x86/microVU_Flags.inl | 6 ++--- pcsx2/x86/microVU_Upper.inl | 4 ++-- 4 files changed, 41 insertions(+), 48 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 829ddd0803..f46d5d3a6c 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -37,7 +37,9 @@ declareAllVariables // Declares All Global Variables :D // Only run this once per VU! ;) microVUt(void) mVUinit(VURegs* vuRegsPtr, int vuIndex) { - microVU* mVU = mVUx; + microVU* mVU = mVUx; + memset(&mVU->prog, 0, sizeof(mVU->prog)); + mVU->regs = vuRegsPtr; mVU->index = vuIndex; mVU->vuMemSize = (vuIndex ? 0x4000 : 0x1000); @@ -45,7 +47,7 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr, int vuIndex) { mVU->progSize = (vuIndex ? 0x4000 : 0x1000) / 4; mVU->cache = NULL; mVU->cacheSize = mVUcacheSize; - memset(&mVU->prog, 0, sizeof(mVU->prog)); + mVU->prog.max = mMaxProg - 1; mVUprint((vuIndex) ? "microVU1: init" : "microVU0: init"); mVU->cache = SysMmapEx((vuIndex ? 0x5f240000 : 0x5e240000), mVU->cacheSize + 0x1000, 0, (vuIndex ? "Micro VU1" : "Micro VU0")); @@ -59,13 +61,7 @@ microVUt(void) mVUinit(VURegs* vuRegsPtr, int vuIndex) { microVUt(void) mVUreset(mV) { mVUprint((mVU->index) ? "microVU1: reset" : "microVU0: reset"); - - // Delete Block Managers - for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progSize / 2); j++) { - microBlockManager::Delete( mVU->prog.prog[i].block[j] ); - } - } + mVUclose(mVU, 1); // Dynarec Cache memset(mVU->cache, 0xcc, mVU->cacheSize + 0x1000); @@ -77,14 +73,15 @@ microVUt(void) mVUreset(mV) { // Clear All Program Data memset(&mVU->prog, 0, sizeof(mVU->prog)); + memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); // Program Variables - mVU->prog.isSame = -1; - mVU->prog.cleared = 1; - mVU->prog.cur = -1; - mVU->prog.total = -1; - memset(&mVU->prog.lpState, 0, sizeof(mVU->prog.lpState)); - //mVU->prog.lpState = &mVU->prog.prog[15].allocInfo.block.pState; // Blank Pipeline State (ToDo: finish implementation) + mVU->prog.cleared = 1; + mVU->prog.isSame = -1; + mVU->prog.cur = -1; + mVU->prog.total = -1; + mVU->prog.max = mMaxProg - 1; + mVU->prog.prog = (microProgram*)_aligned_malloc(sizeof(microProgram)*(mVU->prog.max+1), 64); // Setup Dynarec Cache Limits for Each Program u8* z = (mVU->cache + 0x1000); // Dispatcher Code is in first page of cache @@ -99,19 +96,20 @@ microVUt(void) mVUreset(mV) { } // Free Allocated Resources -microVUt(void) mVUclose(mV) { +microVUt(void) mVUclose(mV, bool isReset) { mVUprint((mVU->index) ? "microVU1: close" : "microVU0: close"); - if (mVU->cache) { HostSys::Munmap(mVU->cache, mVU->cacheSize); mVU->cache = NULL; } + if (!isReset && mVU->cache) { HostSys::Munmap(mVU->cache, mVU->cacheSize); mVU->cache = NULL; } - // Delete Block Managers - for (int i = 0; i <= mVU->prog.max; i++) { - for (u32 j = 0; j < (mVU->progSize / 2); j++) { - if (mVU->prog.prog[i].block[j]) { - microBlockManager::Delete( mVU->prog.prog[i].block[j] ); + // Delete Programs and Block Managers + if (mVU->prog.prog) { + for (int i = 0; i <= mVU->prog.max; i++) { + for (u32 j = 0; j < (mVU->progSize / 2); j++) { + microBlockManager::Delete(mVU->prog.prog[i].block[j]); } } + if (!isReset) { _aligned_free(mVU->prog.prog); } } } @@ -252,7 +250,7 @@ microVUf(int) mVUsearchProg() { //------------------------------------------------------------------ void initVUrec (VURegs* vuRegs, const int vuIndex) { mVUinit(vuRegs, vuIndex); } -void closeVUrec(const int vuIndex) { mVUclose(mVUx); } +void closeVUrec(const int vuIndex) { mVUclose(mVUx, 0); } void resetVUrec(const int vuIndex) { mVUreset(mVUx); } void vsyncVUrec(const int vuIndex) { mVUvsyncUpdate(mVUx); } diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index cdf0011baa..7371dc93d4 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -47,7 +47,7 @@ public: static void Delete(microBlockManager* dead) { if (dead == NULL) return; dead->~microBlockManager(); - _aligned_free( dead ); + _aligned_free(dead); } microBlockManager() { reset(); } @@ -84,11 +84,11 @@ public: } }; -template // progSize = VU program memory size / 4 +#define mProgSize 0x4000/4 struct microProgram { - PCSX2_ALIGNED16(u32 data[progSize]); - microBlockManager* block[progSize/2]; - microIR allocInfo; + u32 data [mProgSize]; + microBlockManager* block[mProgSize/2]; + microIR allocInfo; u32 used; // Number of times its been used u32 last_used; // Counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vSync) s32 range[2]; // The range of microMemory that has already been recompiled for the current program @@ -97,16 +97,15 @@ struct microProgram { u8* x86end; // Limit of program's rec-cache }; -#define mMaxProg 32 // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) -template // pSize = VU program memory size / 4 +#define mMaxProg ((mVU->index)?64:8) // The amount of Micro Programs Recs will 'remember' (For n = 1, 2, 4, 8, 16, etc...) struct microProgManager { - microProgram prog[mMaxProg]; // Store MicroPrograms in memory - static const int max = mMaxProg - 1; - int cur; // Index to Current MicroProgram thats running (-1 = uncached) - int total; // Total Number of valid MicroPrograms minus 1 - int isSame; // Current cached microProgram is Exact Same program as mVU->regs->Micro (-1 = unknown, 0 = No, 1 = Yes) - int cleared; // Micro Program is Indeterminate so must be searched for (and if no matches are found then recompile a new one) - microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) + microProgram* prog; // Store MicroPrograms in memory + int max; // Max Number of MicroPrograms minus 1 + int total; // Total Number of valid MicroPrograms minus 1 + int cur; // Index to Current MicroProgram thats running (-1 = uncached) + int isSame; // Current cached microProgram is Exact Same program as mVU->regs->Micro (-1 = unknown, 0 = No, 1 = Yes) + int cleared; // Micro Program is Indeterminate so must be searched for (and if no matches are found then recompile a new one) + microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; #define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) @@ -123,7 +122,7 @@ struct microVU { u32 progSize; // VU Micro Memory Size (in u32's) u32 cacheSize; // VU Cache Size - microProgManager<0x4000/4> prog; // Micro Program Data + microProgManager prog; // Micro Program Data FILE* logFile; // Log File Pointer VURegs* regs; // VU Regs Struct @@ -145,17 +144,13 @@ struct microVU { extern PCSX2_ALIGNED16(microVU microVU0); extern PCSX2_ALIGNED16(microVU microVU1); -// Opcode Tables -extern void (*mVU_UPPER_OPCODE[64])( VURegs* VU, s32 info ); -extern void (*mVU_LOWER_OPCODE[128])( VURegs* VU, s32 info ); - // Debug Helper extern int mVUdebugNow; // Main Functions microVUt(void) mVUinit(VURegs*, int); microVUt(void) mVUreset(mV); -microVUt(void) mVUclose(mV); +microVUt(void) mVUclose(mV, bool isReset); microVUt(void) mVUclear(mV, u32, u32); microVUt(void*) mVUblockFetch(microVU* mVU, u32 startPC, uptr pState); microVUx(void*) __fastcall mVUcompileJIT(u32 startPC, uptr pState); diff --git a/pcsx2/x86/microVU_Flags.inl b/pcsx2/x86/microVU_Flags.inl index 121038b3ed..57f35cca54 100644 --- a/pcsx2/x86/microVU_Flags.inl +++ b/pcsx2/x86/microVU_Flags.inl @@ -143,9 +143,9 @@ microVUt(int) mVUsetFlags(mV, int* xStatus, int* xMac, int* xClip) { mFLAG.lastWrite = (xM-1) & 3; cFLAG.lastWrite = (xC-1) & 3; - if (sFlagCond) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } - if (mFLAG.doFlag) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } - if (cFLAG.doFlag) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } + if (sFlagCond) { xStatus[xS] = cycles + 4; xS = (xS+1) & 3; } + if (mFLAG.doFlag) { xMac [xM] = cycles + 4; xM = (xM+1) & 3; } + if (cFLAG.doFlag) { xClip [xC] = cycles + 4; xC = (xC+1) & 3; } cycles++; incPC2(2); diff --git a/pcsx2/x86/microVU_Upper.inl b/pcsx2/x86/microVU_Upper.inl index eee4fa4ab1..380a2a4727 100644 --- a/pcsx2/x86/microVU_Upper.inl +++ b/pcsx2/x86/microVU_Upper.inl @@ -52,12 +52,12 @@ microVUt(void) mVUupdateFlags(mV, int reg, int regT1, int regT2, int xyzw, bool SSE_MOVMSKPS_XMM_to_R32(mReg, regT2); // Move the sign bits of the t1reg - AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation + AND32ItoR(mReg, AND_XYZW); // Grab "Is Signed" bits from the previous calculation SHL32ItoR(mReg, 4 + ADD_XYZW); //-------------------------Check for Zero flags------------------------------ - AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation + AND32ItoR(gprT2, AND_XYZW); // Grab "Is Zero" bits from the previous calculation if (mFLAG.doFlag) { SHIFT_XYZW(gprT2); } OR32RtoR(mReg, gprT2); From 8158636490fb9285fcc04939e12ee66d358c01d3 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 25 Jun 2009 05:17:54 +0000 Subject: [PATCH 281/296] microVU: Evenly distributed rec-cache between programs. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1430 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 7371dc93d4..100dbfd7b6 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -108,7 +108,7 @@ struct microProgManager { microRegInfo lpState; // Pipeline state from where program left off (useful for continuing execution) }; -#define mVUcacheSize (0x2000000 / ((vuIndex) ? 1 : 4)) +#define mVUcacheSize (mMaxProg * 0xCCCCC) // 0.8mb per program struct microVU { PCSX2_ALIGNED16(u32 macFlag[4]); // 4 instances of mac flag (used in execution) From 95db132180789bfb321ff100364d5b93b90844d2 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Thu, 25 Jun 2009 05:41:40 +0000 Subject: [PATCH 282/296] microVU: minor changes... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1431 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Lower.inl | 1 + pcsx2/x86/microVU_Misc.h | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/pcsx2/x86/microVU_Lower.inl b/pcsx2/x86/microVU_Lower.inl index a91266f159..8c82889596 100644 --- a/pcsx2/x86/microVU_Lower.inl +++ b/pcsx2/x86/microVU_Lower.inl @@ -1056,6 +1056,7 @@ mVUop(mVU_XITOP) { pass1 { if (!_It_) { mVUlow.isNOP = 1; } analyzeVIreg2(_It_, mVUlow.VI_write, 1); } pass2 { MOVZX32M16toR(gprT1, (uptr)&mVU->regs->vifRegs->itop); + AND32ItoR(gprT1, isVU1 ? 0x3ff : 0xff); mVUallocVIb(mVU, gprT1, _It_); } pass3 { mVUlog("XITOP vi%02d", _Ft_); } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index c895c7eee2..8fbf0ecbd0 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -109,7 +109,7 @@ declareAllVariables #define divI 0x1040000 #define divD 0x2080000 -#define isVU1 (mVU == µVU1) +#define isVU1 (mVU->index != 0) #define getIndex (isVU1 ? 1 : 0) #define getVUmem(x) (((isVU1) ? (x & 0x3ff) : ((x >= 0x400) ? (x & 0x43f) : (x & 0xff))) * 16) #define offsetSS ((_X) ? (0) : ((_Y) ? (4) : ((_Z) ? 8: 12))) From 7f00193900d3db5c301b01388f7c82d6febe2d4b Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 26 Jun 2009 00:30:19 +0000 Subject: [PATCH 283/296] Various changes, mainly Gif related cleanup. Fixed SafeArray.h to work with Linux again while I'm at it... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1432 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/COP0.cpp | 10 -- pcsx2/GS.h | 2 +- pcsx2/Gif.cpp | 385 ++++++++++++++++++++++++---------------------- pcsx2/Hw.cpp | 14 +- pcsx2/Hw.h | 6 +- pcsx2/MTGS.cpp | 29 ++-- pcsx2/R3000A.cpp | 2 +- pcsx2/SafeArray.h | 4 +- pcsx2/Vif.cpp | 4 +- pcsx2/VifDma.cpp | 16 +- pcsx2/VifDma.h | 9 +- 11 files changed, 253 insertions(+), 228 deletions(-) diff --git a/pcsx2/COP0.cpp b/pcsx2/COP0.cpp index ea94cf3f7e..36ecc2993d 100644 --- a/pcsx2/COP0.cpp +++ b/pcsx2/COP0.cpp @@ -440,11 +440,6 @@ int CPCOND0() { //#define CPCOND0 1 -/*#define BC0(cond) \ - if (CPCOND0() cond) { \ - intDoBranch(_BranchTarget_); \ - }*/ - void BC0F() { if (CPCOND0() == 0) intDoBranch(_BranchTarget_); COP0_LOG( "COP0 > BC0F" ); @@ -454,11 +449,6 @@ void BC0T() { if (CPCOND0() == 1) intDoBranch(_BranchTarget_); COP0_LOG( "COP0 > BC0T" ); } - -/*#define BC0L(cond) \ - if (CPCOND0() cond) { \ - intDoBranch(_BranchTarget_); \ - } else cpuRegs.pc+= 4;*/ void BC0FL() { if (CPCOND0() == 0) diff --git a/pcsx2/GS.h b/pcsx2/GS.h index a90d937815..39c2221844 100644 --- a/pcsx2/GS.h +++ b/pcsx2/GS.h @@ -277,7 +277,7 @@ extern mtgsThreadObject* mtgsThread; void mtgsWaitGS(); bool mtgsOpen(); -void mtgsRingBufSimplePacket( s32 command, u32 data0, u32 data1, u32 data2 ); +//void mtgsRingBufSimplePacket( s32 command, u32 data0, u32 data1, u32 data2 ); ///////////////////////////////////////////////////////////////////////////// // Generalized GS Functions and Stuff diff --git a/pcsx2/Gif.cpp b/pcsx2/Gif.cpp index 55a4e0070c..fb29f830b5 100644 --- a/pcsx2/Gif.cpp +++ b/pcsx2/Gif.cpp @@ -38,6 +38,7 @@ enum gifstate_t }; // A three-way toggle used to determine if the GIF is stalling (transferring) or done (finished). +// Should be a gifstate_t rather then int, but I don't feel like possibly interfering with savestates right now. static int gifstate = GIF_STATE_READY; static u64 s_gstag = 0; // used for querying the last tag @@ -48,25 +49,39 @@ static int gspath3done = 0; static u32 gscycles = 0, prevcycles = 0, mfifocycles = 0; static u32 gifqwc = 0; -bool gifmfifoirq = FALSE; +bool gifmfifoirq = false; -__forceinline void gsInterrupt() { +static __forceinline void clearFIFOstuff(bool full) +{ + GSCSRr &= ~0xC000; //Clear FIFO stuff + + if (full) + GSCSRr |= 0x8000; //FIFO full + else + GSCSRr |= 0x4000; //FIFO empty +} + +__forceinline void gsInterrupt() +{ GIF_LOG("gsInterrupt: %8.8x", cpuRegs.cycle); - if((gif->chcr & 0x100) == 0){ + if ((gif->chcr & 0x100) == 0) + { //Console::WriteLn("Eh? why are you still interrupting! chcr %x, qwc %x, done = %x", params gif->chcr, gif->qwc, done); return; } - if((vif1.cmd & 0x7f) == 0x51) + if ((vif1.cmd & 0x7f) == 0x51) { - if(Path3progress != 0)vif1Regs->stat &= ~VIF1_STAT_VGW; + if (Path3progress != IMAGE_MODE) vif1Regs->stat &= ~VIF1_STAT_VGW; } - if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + if (Path3progress == STOPPED_MODE) psHu32(GIF_STAT) &= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 - if (gif->qwc > 0 || gspath3done == 0) { - if (!(psHu32(DMAC_CTRL) & 0x1)) { + if ((gif->qwc > 0) || (gspath3done == 0)) + { + if (!(psHu32(DMAC_CTRL) & 0x1)) + { Console::Notice("gs dma masked, re-scheduling..."); // re-raise the int shortly in the future CPU_INT( 2, 64 ); @@ -77,16 +92,16 @@ __forceinline void gsInterrupt() { return; } - vif1Regs->stat &= ~VIF1_STAT_VGW; gspath3done = 0; gscycles = 0; gif->chcr &= ~0x100; + vif1Regs->stat &= ~VIF1_STAT_VGW; + psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 - GSCSRr &= ~0xC000; //Clear FIFO stuff - GSCSRr |= 0x4000; //FIFO empty psHu32(GIF_STAT) &= ~GIF_STAT_P3Q; - //psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 - psHu32(GIF_STAT)&= ~0x1F000000; // QFC=0 + psHu32(GIF_STAT) &= ~0x1F000000; // QFC=0 + + clearFIFOstuff(false); hwDmacIrq(DMAC_GIF); GIF_LOG("GIF DMA end"); @@ -96,7 +111,6 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) { psHu32(GIF_STAT) |= GIF_STAT_APATH3 | GIF_STAT_OPH; - if( mtgsThread != NULL ) { int sizetoread = qwc; @@ -126,13 +140,15 @@ static u32 WRITERING_DMA(u32 *pMem, u32 qwc) } } -int _GIFchain() { - - u32 qwc = /*((psHu32(GIF_MODE) & 0x4) || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) :*/ min( gifsplit, (int)gif->qwc ); +int _GIFchain() +{ + //u32 qwc = ((psHu32(GIF_MODE) & 0x4) || vif1Regs->mskpath3) ? min(8, (int)gif->qwc) : min( gifsplit, (int)gif->qwc ); + u32 qwc = min( gifsplit, (int)gif->qwc ); u32 *pMem; pMem = (u32*)dmaGetAddr(gif->madr); - if (pMem == NULL) { + if (pMem == NULL) + { // reset path3, fixes dark cloud 2 gsGIFSoftReset(4); @@ -142,11 +158,8 @@ int _GIFchain() { Console::Notice( "Hackfix - NULL GIFchain" ); return -1; } - qwc = WRITERING_DMA(pMem, qwc); - /*gif->madr+= qwc*16; - gif->qwc -= qwc;*/ - return (qwc); + return (WRITERING_DMA(pMem, qwc)); } static __forceinline void GIFchain() @@ -156,17 +169,47 @@ static __forceinline void GIFchain() FreezeRegs(0); } -static __forceinline void dmaGIFend() +static __forceinline bool checkTieBit(u32* &ptag) { - CPU_INT(2, 16); + if ((gif->chcr & 0x80) && (ptag[0] >> 31)) //Check TIE bit of CHCR and IRQ bit of tag + { + GIF_LOG("dmaIrq Set"); + gspath3done = 1; + return true; + } + + return false; } -// These could probably be consolidated into one function, -// but I wasn't absolutely sure if there was a good reason -// not to do the gif->qwc != 0 check. --arcum42 -static __forceinline void GIFdmaEnd() +static __forceinline bool ReadTag(u32* &ptag, u32 &id) { - CPU_INT(2, gscycles * BIAS); + ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR + + if (ptag == NULL) //Is ptag empty? + { + //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT)|= DMAC_STAT_BEIS; + return false; + } + gscycles+=2; // Add 1 cycles from the QW read for the tag + gif->chcr = ( gif->chcr & 0xFFFF ) | ((*ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 + + id = (ptag[0] >> 28) & 0x7; //ID for DmaChain copied from bit 28 of the tag + gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag + gif->madr = ptag[1]; //MADR = ADDR field + + gspath3done = hwDmacSrcChainWithStack(gif, id); + return true; +} + +static __forceinline void ReadTag2(u32* &ptag) +{ + ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR + gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag + gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 + gif->madr = ptag[1]; + + gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); } void GIFdma() @@ -174,200 +217,167 @@ void GIFdma() u32 *ptag; u32 id; - gscycles= prevcycles ? prevcycles: 0; + gscycles = prevcycles; - if ((psHu32(GIF_CTRL) & 8)) { // temporarily stop + if ((psHu32(GIF_CTRL) & 8)) // temporarily stop + { Console::WriteLn("Gif dma temp paused?"); return; } - if ((psHu32(DMAC_CTRL) & 0xC0) == 0x80 && prevcycles != 0) { // STD == GIF + if (((psHu32(DMAC_CTRL) & 0xC0) == 0x80) && (prevcycles != 0)) // STD == GIF + { Console::WriteLn("GS Stall Control Source = %x, Drain = %x\n MADR = %x, STADR = %x", params (psHu32(0xe000) >> 4) & 0x3, (psHu32(0xe000) >> 6) & 0x3, gif->madr, psHu32(DMAC_STADR)); - if( gif->madr + (gif->qwc * 16) > psHu32(DMAC_STADR) ) { + if ((gif->madr + (gif->qwc * 16)) > psHu32(DMAC_STADR)) + { CPU_INT(2, gscycles); gscycles = 0; return; } + prevcycles = 0; gif->qwc = 0; } - GSCSRr &= ~0xC000; //Clear FIFO stuff - GSCSRr |= 0x8000; //FIFO full - psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [ used to be 0xE00; // OPH=1 | APATH=3] - - if(((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == 0) //Path2 gets priority in intermittent mode + clearFIFOstuff(true); + psHu32(GIF_STAT) |= 0x10000000; // FQC=31, hack ;) [ used to be 0xE00; // OPH=1 | APATH=3] + + //Path2 gets priority in intermittent mode + if (((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && (Path3progress == IMAGE_MODE)) { GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); - dmaGIFend(); + CPU_INT(2, 16); return; } - if (vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1)) { - if(gif->qwc == 0) { - if((gif->chcr & 0x10c) == 0x104) { - ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR - - if (ptag == NULL) { //Is ptag empty? - psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register - return; - } - gscycles += 2; - gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 - id = (ptag[0] >> 28) & 0x7; //ID for DmaChain copied from bit 28 of the tag - gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag - gif->madr = ptag[1]; //MADR = ADDR field - gspath3done = hwDmacSrcChainWithStack(gif, id); + if (vif1Regs->mskpath3 || (psHu32(GIF_MODE) & 0x1)) + { + if (gif->qwc == 0) + { + if ((gif->chcr & 0x10c) == 0x104) + { + if (!ReadTag(ptag, id)) return; GIF_LOG("PTH3 MASK gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, id, gif->madr); - if ((gif->chcr & 0x80) && ptag[0] >> 31) { //Check TIE bit of CHCR and IRQ bit of tag - GIF_LOG("PATH3 MSK dmaIrq Set"); - Console::WriteLn("GIF TIE"); - gspath3done |= 1; - } + //Check TIE bit of CHCR and IRQ bit of tag + if (checkTieBit(ptag)) GIF_LOG("PATH3 MSK dmaIrq Set"); } } - if(Path3progress == 2 /*|| (vif1Regs->stat |= VIF1_STAT_VGW) == 0*/) + if (Path3progress == STOPPED_MODE) /*|| (vif1Regs->stat |= VIF1_STAT_VGW) == 0*/ { vif1Regs->stat &= ~VIF1_STAT_VGW; - if(gif->qwc == 0)dmaGIFend(); + if (gif->qwc == 0) CPU_INT(2, 16); return; } GIFchain(); - GIFdmaEnd(); + CPU_INT(2, gscycles * BIAS); return; } // Transfer Dn_QWC from Dn_MADR to GIF - if ((gif->chcr & 0xc) == 0 || gif->qwc > 0) { // Normal Mode + if (((gif->chcr & 0xc) == 0) || (gif->qwc > 0)) // Normal Mode + { - if ((((psHu32(DMAC_CTRL) & 0xC0) == 0x80) && ((gif->chcr & 0xc) == 0))) { + if (((psHu32(DMAC_CTRL) & 0xC0) == 0x80) && ((gif->chcr & 0xc) == 0)) + { Console::WriteLn("DMA Stall Control on GIF normal"); } GIFchain(); //Transfers the data set by the switch - GIFdmaEnd(); + CPU_INT(2, gscycles * BIAS); return; } - if ((gif->chcr & 0xc) == 0x4 && gspath3done == 0) + if (((gif->chcr & 0xc) == 0x4) && (gspath3done == 0)) // Chain Mode { - // Chain Mode - //while ((gspath3done == 0) && (gif->qwc == 0)) { //Loop if the transfers aren't intermittent - ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR - if (ptag == NULL) { //Is ptag empty? - psHu32(DMAC_STAT)|= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register + if (!ReadTag(ptag, id)) return; + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, id, gif->madr); + + if ((psHu32(DMAC_CTRL) & 0xC0) == 0x80) // STD == GIF + { + // there are still bugs, need to also check if gif->madr +16*qwc >= stadr, if not, stall + if (!gspath3done && ((gif->madr + (gif->qwc * 16)) > psHu32(DMAC_STADR)) && (id == 4)) + { + // stalled + Console::WriteLn("GS Stall Control Source = %x, Drain = %x\n MADR = %x, STADR = %x", params (psHu32(0xe000) >> 4) & 0x3, (psHu32(0xe000) >> 6) & 0x3,gif->madr, psHu32(DMAC_STADR)); + prevcycles = gscycles; + gif->tadr -= 16; + hwDmacIrq(DMAC_13); + CPU_INT(2, gscycles); + gscycles = 0; return; } - gscycles+=2; // Add 1 cycles from the QW read for the tag - - // We used to transfer dma tags if tte is set here - - gif->chcr = ( gif->chcr & 0xFFFF ) | ((*ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 - - id = (ptag[0] >> 28) & 0x7; //ID for DmaChain copied from bit 28 of the tag - gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag - gif->madr = ptag[1]; //MADR = ADDR field + } - gspath3done = hwDmacSrcChainWithStack(gif, id); - GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, id, gif->madr); - - if ((psHu32(DMAC_CTRL) & 0xC0) == 0x80) { // STD == GIF - // there are still bugs, need to also check if gif->madr +16*qwc >= stadr, if not, stall - if(!gspath3done && ((gif->madr + (gif->qwc * 16)) > psHu32(DMAC_STADR)) && (id == 4)) { - // stalled - Console::WriteLn("GS Stall Control Source = %x, Drain = %x\n MADR = %x, STADR = %x", params (psHu32(0xe000) >> 4) & 0x3, (psHu32(0xe000) >> 6) & 0x3,gif->madr, psHu32(DMAC_STADR)); - prevcycles = gscycles; - gif->tadr -= 16; - hwDmacIrq(13); - CPU_INT(2, gscycles); - gscycles = 0; - return; - } - } - //GIFchain(); //Transfers the data set by the switch - - if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag - GIF_LOG("dmaIrq Set"); - gspath3done = 1; - } - //} + checkTieBit(ptag); } prevcycles = 0; - if (gspath3done == 0 && gif->qwc == 0) + + if ((gspath3done == 0) && (gif->qwc == 0)) { - ptag = (u32*)dmaGetAddr(gif->tadr); //Set memory pointer to TADR gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 gif->madr = ptag[1]; gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); - if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag - GIF_LOG("dmaIrq Set"); - gspath3done = 1; - } - GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28) & 0x7, gif->madr); - GIFdmaEnd(); - return; - } else GIFdmaEnd(); - gscycles = 0; - + checkTieBit(ptag); + + GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28) & 0x7, gif->madr); + CPU_INT(2, gscycles * BIAS); + } + else + { + CPU_INT(2, gscycles * BIAS); + gscycles = 0; + } } -void dmaGIF() { +void dmaGIF() +{ //We used to add wait time for the buffer to fill here, fixing some timing problems in path 3 masking //It takes the time of 24 QW for the BUS to become ready - The Punisher And Streetball GIF_LOG("dmaGIFstart chcr = %lx, madr = %lx, qwc = %lx\n tadr = %lx, asr0 = %lx, asr1 = %lx", gif->chcr, gif->madr, gif->qwc, gif->tadr, gif->asr0, gif->asr1); - Path3progress = 2; - gspath3done = 0; // For some reason this doesnt clear? So when the system starts the thread, we will clear it :) + Path3progress = STOPPED_MODE; + gspath3done = 0; // For some reason this doesn't clear? So when the system starts the thread, we will clear it :) psHu32(GIF_STAT) |= GIF_STAT_P3Q; - GSCSRr &= ~0xC000; //Clear FIFO stuff - GSCSRr |= 0x8000; //FIFO full - psHu32(GIF_STAT)|= 0x10000000; // FQC=31, hack ;) [used to be 0xE00; // OPH=1 | APATH=3] + psHu32(GIF_STAT) |= 0x10000000; // FQC=31, hack ;) [used to be 0xE00; // OPH=1 | APATH=3] + clearFIFOstuff(true); - if ((psHu32(DMAC_CTRL) & 0xC) == 0xC ) { // GIF MFIFO + if ((psHu32(DMAC_CTRL) & 0xC) == 0xC ) // GIF MFIFO + { //Console::WriteLn("GIF MFIFO"); gifMFIFOInterrupt(); return; } - - if ((gif->qwc == 0) && ((gif->chcr & 0xc) != 0)){ + if ((gif->qwc == 0) && ((gif->chcr & 0xc) != 0)) + { u32 *ptag; - ptag = (u32*)dmaGetAddr(gif->tadr); - gif->qwc = (u16)ptag[0]; //QWC set to lower 16bits of the tag - gif->chcr = ( gif->chcr & 0xFFFF ) | ( (*ptag) & 0xFFFF0000 ); //Transfer upper part of tag to CHCR bits 31-15 - gif->madr = ptag[1]; - + + ReadTag2(ptag); GIF_LOG("gifdmaChain %8.8x_%8.8x size=%d, id=%d, addr=%lx", ptag[1], ptag[0], gif->qwc, (ptag[0] >> 28), gif->madr); - - gspath3done = hwDmacSrcChainWithStack(gif, (ptag[0] >> 28) & 0x7); - if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { //Check TIE bit of CHCR and IRQ bit of tag - GIF_LOG("dmaIrq Set"); - gspath3done = 1; - } - + checkTieBit(ptag); GIFdma(); - //gif->qwc = 0; return; } //Halflife sets a QWC amount in chain mode, no tadr set. - if(gif->qwc > 0) gspath3done = 1; + if (gif->qwc > 0) gspath3done = 1; GIFdma(); } // called from only one location, so forceinline it: -static __forceinline int mfifoGIFrbTransfer() { +static __forceinline int mfifoGIFrbTransfer() +{ u32 mfifoqwc = min(gifqwc, (u32)gif->qwc); u32 *src; @@ -384,15 +394,19 @@ static __forceinline int mfifoGIFrbTransfer() { if (src == NULL) return -1; s1 = WRITERING_DMA(src, s1); - if(s1 == (mfifoqwc - s2)) + if (s1 == (mfifoqwc - s2)) { /* and second copy 's2' bytes from 'maddr' to '&data[s1]' */ src = (u32*)PSM(psHu32(DMAC_RBOR)); if (src == NULL) return -1; s2 = WRITERING_DMA(src, s2); - } else s2 = 0; - mfifoqwc = s1 + s2; + } + else + { + s2 = 0; + } + mfifoqwc = s1 + s2; } else { @@ -406,21 +420,18 @@ static __forceinline int mfifoGIFrbTransfer() { } gifqwc -= mfifoqwc; - //gif->qwc -= mfifoqwc; - //gif->madr += mfifoqwc*16; - //mfifocycles += (mfifoqwc) * 2; /* guessing */ return 0; } // called from only one location, so forceinline it: -static __forceinline int mfifoGIFchain() { +static __forceinline int mfifoGIFchain() +{ /* Is QWC = 0? if so there is nothing to transfer */ - if (gif->qwc == 0) return 0; if (gif->madr >= psHu32(DMAC_RBOR) && - gif->madr <= (psHu32(DMAC_RBOR)+psHu32(DMAC_RBSR))) + gif->madr <= (psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR))) { if (mfifoGIFrbTransfer() == -1) return -1; } @@ -431,23 +442,23 @@ static __forceinline int mfifoGIFchain() { if (pMem == NULL) return -1; mfifoqwc = WRITERING_DMA(pMem, mfifoqwc); - //gif->madr += mfifoqwc*16; - //gif->qwc -= mfifoqwc; mfifocycles += (mfifoqwc) * 2; /* guessing */ } return 0; } -void mfifoGIFtransfer(int qwc) { +void mfifoGIFtransfer(int qwc) +{ u32 *ptag; int id; u32 temp = 0; mfifocycles = 0; - gifmfifoirq = FALSE; + gifmfifoirq = false; - if(qwc > 0 ) { + if(qwc > 0 ) + { gifqwc += qwc; if (gifstate != GIF_STATE_EMPTY) return; gifstate &= ~GIF_STATE_EMPTY; @@ -455,13 +466,16 @@ void mfifoGIFtransfer(int qwc) { GIF_LOG("mfifoGIFtransfer %x madr %x, tadr %x", gif->chcr, gif->madr, gif->tadr); - if (gif->qwc == 0) { - if (gif->tadr == spr0->madr) { + if (gif->qwc == 0) + { + if (gif->tadr == spr0->madr) + { //if( gifqwc > 1 ) DevCon::WriteLn("gif mfifo tadr==madr but qwc = %d", params gifqwc); - hwDmacIrq(14); + hwDmacIrq(DMAC_14); gifstate |= GIF_STATE_EMPTY; return; } + gif->tadr = psHu32(DMAC_RBOR) + (gif->tadr & psHu32(DMAC_RBSR)); ptag = (u32*)dmaGetAddr(gif->tadr); @@ -475,7 +489,8 @@ void mfifoGIFtransfer(int qwc) { ptag[1], ptag[0], gif->qwc, id, gif->madr, gif->tadr, gifqwc, spr0->madr); gifqwc--; - switch (id) { + switch (id) + { case 0: // Refe - Transfer Packet According to ADDR field gif->tadr = psHu32(DMAC_RBOR) + ((gif->tadr + 16) & psHu32(DMAC_RBSR)); gifstate = GIF_STATE_DONE; //End Transfer @@ -507,7 +522,8 @@ void mfifoGIFtransfer(int qwc) { break; } - if ((gif->chcr & 0x80) && (ptag[0] >> 31)) { + if ((gif->chcr & 0x80) && (ptag[0] >> 31)) + { SPR_LOG("dmaIrq Set"); gifstate = GIF_STATE_DONE; gifmfifoirq = TRUE; @@ -515,16 +531,15 @@ void mfifoGIFtransfer(int qwc) { } FreezeRegs(1); - - if (mfifoGIFchain() == -1) { - Console::WriteLn("GIF dmaChain error size=%d, madr=%lx, tadr=%lx", params - gif->qwc, gif->madr, gif->tadr); - gifstate = GIF_STATE_STALL; - } - + if (mfifoGIFchain() == -1) + { + Console::WriteLn("GIF dmaChain error size=%d, madr=%lx, tadr=%lx", params + gif->qwc, gif->madr, gif->tadr); + gifstate = GIF_STATE_STALL; + } FreezeRegs(0); - if(gif->qwc == 0 && gifstate == GIF_STATE_DONE) gifstate = GIF_STATE_STALL; + if ((gif->qwc == 0) && (gifstate == GIF_STATE_DONE)) gifstate = GIF_STATE_STALL; CPU_INT(11,mfifocycles); SPR_LOG("mfifoGIFtransfer end %x madr %x, tadr %x", gif->chcr, gif->madr, gif->tadr); @@ -533,40 +548,45 @@ void mfifoGIFtransfer(int qwc) { void gifMFIFOInterrupt() { mfifocycles = 0; - if(Path3progress == 2) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + if (Path3progress == STOPPED_MODE) psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 - if((spr0->chcr & 0x100) && spr0->qwc == 0) + if ((spr0->chcr & 0x100) && (spr0->qwc == 0)) { spr0->chcr &= ~0x100; hwDmacIrq(DMAC_FROM_SPR); } - if (!(gif->chcr & 0x100)) { + if (!(gif->chcr & 0x100)) + { Console::WriteLn("WTF GIFMFIFO"); cpuRegs.interrupt &= ~(1 << 11); - return ; + return; } - if(((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == 0) //Path2 gets priority in intermittent mode + if (((psHu32(GIF_STAT) & 0x100) || (vif1.cmd & 0x7f) == 0x50) && (psHu32(GIF_MODE) & 0x4) && Path3progress == IMAGE_MODE) //Path2 gets priority in intermittent mode { //GIF_LOG("Waiting VU %x, PATH2 %x, GIFMODE %x Progress %x", psHu32(GIF_STAT) & 0x100, (vif1.cmd & 0x7f), psHu32(GIF_MODE), Path3progress); CPU_INT(11,mfifocycles); return; } - if(gifstate != GIF_STATE_STALL) { - if(gifqwc <= 0) { + if (gifstate != GIF_STATE_STALL) + { + if (gifqwc <= 0) + { //Console::WriteLn("Empty"); gifstate |= GIF_STATE_EMPTY; psHu32(GIF_STAT)&= ~0xE00; // OPH=0 | APATH=0 - hwDmacIrq(14); + hwDmacIrq(DMAC_14); return; } mfifoGIFtransfer(0); return; } + #ifdef PCSX2_DEVBUILD - if(gifstate == GIF_STATE_READY || gif->qwc > 0) { + if (gifstate == GIF_STATE_READY || gif->qwc > 0) + { Console::Error("gifMFIFO Panic > Shouldn't go here!"); return; } @@ -576,15 +596,16 @@ void gifMFIFOInterrupt() gspath3done = 0; gscycles = 0; - psHu32(GIF_STAT)&= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 + + psHu32(GIF_STAT) &= ~(GIF_STAT_APATH3 | GIF_STAT_OPH); // OPH=0 | APATH=0 psHu32(GIF_STAT) &= ~GIF_STAT_P3Q; - gifstate = GIF_STATE_READY; - gif->chcr &= ~0x100; - vif1Regs->stat &= ~VIF1_STAT_VGW; - hwDmacIrq(DMAC_GIF); - GSCSRr &= ~0xC000; //Clear FIFO stuff - GSCSRr |= 0x4000; //FIFO empty psHu32(GIF_STAT)&= ~0x1F000000; // QFC=0 + + vif1Regs->stat &= ~VIF1_STAT_VGW; + gif->chcr &= ~0x100; + gifstate = GIF_STATE_READY; + hwDmacIrq(DMAC_GIF); + clearFIFOstuff(false); } void SaveState::gifFreeze() diff --git a/pcsx2/Hw.cpp b/pcsx2/Hw.cpp index 3ff49efc15..7d3f740e8e 100644 --- a/pcsx2/Hw.cpp +++ b/pcsx2/Hw.cpp @@ -90,7 +90,7 @@ __forceinline void intcInterrupt() cpuException(0x400, cpuRegs.branch); } -__forceinline void dmacInterrupt() +__forceinline void dmacInterrupt() { if ((cpuRegs.CP0.n.Status.val & 0x10807) != 0x10801) return; @@ -116,7 +116,7 @@ void hwDmacIrq(int n) { } /* Write 'size' bytes to memory address 'addr' from 'data'. */ -int hwMFIFOWrite(u32 addr, u8 *data, u32 size) { +bool hwMFIFOWrite(u32 addr, u8 *data, u32 size) { u32 msize = psHu32(DMAC_RBOR) + psHu32(DMAC_RBSR)+16; u8 *dst; @@ -129,22 +129,22 @@ int hwMFIFOWrite(u32 addr, u8 *data, u32 size) { /* it does, so first copy 's1' bytes from 'data' to 'addr' */ dst = (u8*)PSM(addr); - if (dst == NULL) return -1; + if (dst == NULL) return false; memcpy_fast(dst, data, s1); /* and second copy 's2' bytes from '&data[s1]' to 'maddr' */ dst = (u8*)PSM(psHu32(DMAC_RBOR)); - if (dst == NULL) return -1; + if (dst == NULL) return false; memcpy_fast(dst, &data[s1], s2); } else { /* it doesn't, so just copy 'size' bytes from 'data' to 'addr' */ dst = (u8*)PSM(addr); - if (dst == NULL) return -1; + if (dst == NULL) return false; memcpy_fast(dst, data, size); } - return 0; + return true; } @@ -206,7 +206,7 @@ bool hwDmacSrcChainWithStack(DMACh *dma, int id) { dma->asr0 = 0; //Clear ASR0 } else { //Else if ASR1 and ASR0 are empty //dma->tadr += 16; //Clear tag address - Kills Klonoa 2 - return 1; //End Transfer + return true; //End Transfer } } return false; diff --git a/pcsx2/Hw.h b/pcsx2/Hw.h index 7253727b1f..4df602d39e 100644 --- a/pcsx2/Hw.h +++ b/pcsx2/Hw.h @@ -306,6 +306,8 @@ enum DMACIrqs DMAC_SIF2, DMAC_FROM_SPR, DMAC_TO_SPR, + DMAC_13 = 13, // Stall? + DMAC_14 = 14, // Transfer? DMAC_ERROR = 15, }; @@ -519,8 +521,8 @@ extern void __fastcall hwWrite128_generic(u32 mem, const mem128_t *srcval); void hwIntcIrq(int n); void hwDmacIrq(int n); -int hwMFIFORead(u32 addr, u8 *data, u32 size); -int hwMFIFOWrite(u32 addr, u8 *data, u32 size); +//bool hwMFIFORead(u32 addr, u8 *data, u32 size); +bool hwMFIFOWrite(u32 addr, u8 *data, u32 size); bool hwDmacSrcChainWithStack(DMACh *dma, int id); bool hwDmacSrcChain(DMACh *dma, int id); diff --git a/pcsx2/MTGS.cpp b/pcsx2/MTGS.cpp index 131d80b150..4b18c5d001 100644 --- a/pcsx2/MTGS.cpp +++ b/pcsx2/MTGS.cpp @@ -224,8 +224,8 @@ void mtgsThreadObject::Start() m_post_InitDone.Wait(); - if( m_returncode != 0 ) // means the thread failed to init the GS plugin - throw Exception::PluginFailure( "GS", "The GS plugin failed to open/initialize." ); + // means the thread failed to init the GS plugin + if ( m_returncode != 0 ) throw Exception::PluginFailure( "GS", "The GS plugin failed to open/initialize." ); } mtgsThreadObject::~mtgsThreadObject() @@ -282,21 +282,23 @@ __forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u while(size > 0) { - if(path.tag.nloop == 0) + if (path.tag.nloop == 0) { path.SetTag( pMem ); pMem += sizeof(GIFTAG); --size; - if(pathidx == 2) + if (pathidx == 2) { - if(path.tag.flg != GIF_FLG_IMAGE)Path3progress = 1; //Other mode - else Path3progress = 0; //IMAGE mode + if (path.tag.flg != GIF_FLG_IMAGE) + Path3progress = TRANSFER_MODE; //Other mode (but not stopped, I guess?) + else + Path3progress = IMAGE_MODE; //IMAGE mode //if(pathidx == 2) GIF_LOG("Set Giftag NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); } - if( pathidx == 0 ) + if (pathidx == 0) { // int transize = 0; // hack: if too much data for VU1, just ignore. @@ -324,7 +326,8 @@ __forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u return ++size; } } - }else + } + else { // NOTE: size > 0 => do {} while(size > 0); should be faster than while(size > 0) {} @@ -453,7 +456,7 @@ __forceinline int mtgsThreadObject::_gifTransferDummy( GIF_PATH pathidx, const u //params path.tag.nloop, path.tag.eop, path.tag.flg, path.tag.nreg, Path3progress, vif1Regs->stat & VIF1_STAT_VGW); if(path.tag.eop) { - Path3progress = 2; + Path3progress = STOPPED_MODE; //GIF_LOG("Set progress NLoop %d EOP %x Mode %d Path3msk %x Path3progress %x ", path.tag.nloop, path.tag.eop, path.tag.flg, vif1Regs->mskpath3, Path3progress); } @@ -1134,7 +1137,7 @@ void mtgsThreadObject::Freeze( SaveState& state ) // this function is needed because of recompiled calls from iGS.cpp // (currently used in GCC only) -void mtgsRingBufSimplePacket( s32 command, u32 data0, u32 data1, u32 data2 ) -{ - mtgsThread->SendSimplePacket( (GS_RINGTYPE)command, data0, data1, data2 ); -} +//void mtgsRingBufSimplePacket( s32 command, u32 data0, u32 data1, u32 data2 ) +//{ +// mtgsThread->SendSimplePacket( (GS_RINGTYPE)command, data0, data1, data2 ); +//} diff --git a/pcsx2/R3000A.cpp b/pcsx2/R3000A.cpp index 368d5ad703..015a259803 100644 --- a/pcsx2/R3000A.cpp +++ b/pcsx2/R3000A.cpp @@ -164,7 +164,7 @@ __forceinline int psxTestCycle( u32 startCycle, s32 delta ) __forceinline void PSX_INT( IopEventId n, s32 ecycle ) { // Generally speaking games shouldn't throw ints that haven't been cleared yet. - // It's usually indicative os something amiss in our emulation, so uncomment this + // It's usually indicative of something amiss in our emulation, so uncomment this // code to help trap those sort of things. // Exception: IRQ16 - SIO - it drops ints like crazy when handling PAD stuff. diff --git a/pcsx2/SafeArray.h b/pcsx2/SafeArray.h index 645ee2560f..664db6fd01 100644 --- a/pcsx2/SafeArray.h +++ b/pcsx2/SafeArray.h @@ -420,9 +420,9 @@ class SafeAlignedArray : public SafeArray protected: T* _virtual_realloc( int newsize ) { - return (T*)( ( m_ptr == NULL ) ? + return (T*)( ( this->m_ptr == NULL ) ? _aligned_malloc( newsize * sizeof(T), Alignment ) : - _aligned_realloc( m_ptr, newsize * sizeof(T), Alignment ) + _aligned_realloc( this->m_ptr, newsize * sizeof(T), Alignment ) ); } diff --git a/pcsx2/Vif.cpp b/pcsx2/Vif.cpp index f61b024b75..f16a8e2cda 100644 --- a/pcsx2/Vif.cpp +++ b/pcsx2/Vif.cpp @@ -582,7 +582,7 @@ void vifMFIFOInterrupt() vifqwc = 0; vif1.inprogress |= 0x10; vif1Regs->stat &= ~0x1F000000; // FQC=0 - hwDmacIrq(14); + hwDmacIrq(DMAC_14); return; } @@ -606,7 +606,7 @@ void vifMFIFOInterrupt() //Console::WriteLn("Empty 2"); //vif1.inprogress |= 0x10; vif1Regs->stat &= ~0x1F000000; // FQC=0 - hwDmacIrq(14); + hwDmacIrq(DMAC_14); }*/ //On a TIE break we do not clear the MFIFO (Art of Fighting) diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index ec5c6525b4..1dd2bd0e12 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -58,7 +58,7 @@ static const unsigned int VIF0dmanum = 0; static const unsigned int VIF1dmanum = 1; int g_vifCycles = 0; -int Path3progress = 2; //0 = Image Mode (DirectHL), 1 = transferring, 2 = Stopped at End of Packet +Path3Modes Path3progress = STOPPED_MODE; u32 splittransfer[4]; u32 splitptr = 0; @@ -1873,9 +1873,9 @@ static int __fastcall Vif1TransDirectHL(u32 *data) { int ret = 0; - if((vif1.cmd & 0x7f) == 0x51) + if ((vif1.cmd & 0x7f) == 0x51) { - if(gif->chcr & 0x100 && (!vif1Regs->mskpath3 && Path3progress == 0)) //PATH3 is in image mode, so wait for end of transfer + if (gif->chcr & 0x100 && (!vif1Regs->mskpath3 && (Path3progress == IMAGE_MODE))) //PATH3 is in image mode, so wait for end of transfer { vif1Regs->stat |= VIF1_STAT_VGW; return 0; @@ -2072,7 +2072,8 @@ void Vif1MskPath3() // MSKPATH3 } else { - Path3progress = 1; //Let the Gif know it can transfer again (making sure any vif stall isnt unset prematurely) + //Let the Gif know it can transfer again (making sure any vif stall isnt unset prematurely) + Path3progress = TRANSFER_MODE; psHu32(GIF_STAT) &= ~0x2; CPU_INT(2, 4); } @@ -2106,9 +2107,10 @@ static void Vif1CMDFlush() // FLUSH/E/A { vif1FLUSH(); - if((vif1.cmd & 0x7f) == 0x13) + if ((vif1.cmd & 0x7f) == 0x13) { - if((Path3progress != 2 || !vif1Regs->mskpath3) && gif->chcr & 0x100) // Gif is already transferring so wait for it. + // Gif is already transferring so wait for it. + if (((Path3progress != STOPPED_MODE) || !vif1Regs->mskpath3) && gif->chcr & 0x100) { vif1Regs->stat |= VIF1_STAT_VGW; CPU_INT(2, 4); @@ -2488,7 +2490,7 @@ __forceinline void vif1SetupTransfer() if ((vif1ch->madr + vif1ch->qwc * 16) >= psHu32(DMAC_STADR)) { // stalled - hwDmacIrq(13); + hwDmacIrq(DMAC_13); return; } } diff --git a/pcsx2/VifDma.h b/pcsx2/VifDma.h index 3a5d85000b..627a5386e5 100644 --- a/pcsx2/VifDma.h +++ b/pcsx2/VifDma.h @@ -25,6 +25,13 @@ enum VifModes VIF_CHAIN_MODE = 2 }; +enum Path3Modes //0 = Image Mode (DirectHL), 1 = transferring, 2 = Stopped at End of Packet +{ + IMAGE_MODE = 0, + TRANSFER_MODE = 1, + STOPPED_MODE = 2 +}; + struct vifCode { u32 addr; u32 size; @@ -54,7 +61,7 @@ struct vifStruct { }; extern vifStruct vif0, vif1; -extern int Path3progress; +extern Path3Modes Path3progress; extern u8 schedulepath3msk; void __fastcall UNPACK_S_32( u32 *dest, u32 *data, int size ); From 12ba6027099197afb5f46e356d3c979b25d916a8 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 26 Jun 2009 01:18:08 +0000 Subject: [PATCH 284/296] GUST fix. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1433 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/HwWrite.cpp | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index 16155c09e7..4ebf145413 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -100,7 +100,9 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) if ((value & 0x100) && ((psHu32(mem) & 0x100) == 0x100) && ((psHu32(DMAC_CTRL) & 0x1) == 1)) { DMA_LOG( "DMAExec32 Attempt to run DMA while one is already active mem = %x", mem ); - return; + + // Returning here breaks every single Gust game written. :( + //return; } // Upper 16bits of QWC should not be written since QWC is 16bits in size. From b75e504fba801fbf3ced6ee78ccc2a22b10401fa Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 26 Jun 2009 01:47:55 +0000 Subject: [PATCH 285/296] Revert r1433. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1434 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/HwWrite.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index 4ebf145413..85645df444 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -102,7 +102,8 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) DMA_LOG( "DMAExec32 Attempt to run DMA while one is already active mem = %x", mem ); // Returning here breaks every single Gust game written. :( - //return; + // Not returning here breaks a bunch of other games. I'll let someone else figure it out... + return; } // Upper 16bits of QWC should not be written since QWC is 16bits in size. From ac7f3bc19df902d1b7d55f79cf21e2154b9a9767 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 26 Jun 2009 02:44:42 +0000 Subject: [PATCH 286/296] Super VU: - Implemented setting of GIF status regs for PATH1 transfers at execution time instead of at recompile time on XGkick instructions. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1435 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/sVU_Lower.cpp | 2 +- pcsx2/x86/sVU_zerorec.cpp | 24 ++++++++---------------- 2 files changed, 9 insertions(+), 17 deletions(-) diff --git a/pcsx2/x86/sVU_Lower.cpp b/pcsx2/x86/sVU_Lower.cpp index 26d6e76ee2..fae6df31ee 100644 --- a/pcsx2/x86/sVU_Lower.cpp +++ b/pcsx2/x86/sVU_Lower.cpp @@ -1983,7 +1983,7 @@ void VU1XGKICK_MTGSTransfer(u32 *pMem, u32 addr) // Chances are this should be a "loops around memory" situation, and the packet // should be continued starting at addr zero (0). - size = mtgsThread->PrepDataPacket( GIF_PATH_1, data, (0x4000-(addr&0x3fff)) >> 4); + size = mtgsThread->PrepDataPacket(GIF_PATH_1, data, (0x4000-(addr&0x3fff)) >> 4); jASSUME( size > 0 ); //if( size > 0 ) diff --git a/pcsx2/x86/sVU_zerorec.cpp b/pcsx2/x86/sVU_zerorec.cpp index 67af485fc2..386fe52516 100644 --- a/pcsx2/x86/sVU_zerorec.cpp +++ b/pcsx2/x86/sVU_zerorec.cpp @@ -4376,18 +4376,15 @@ void recVUMI_XGKICK_(VURegs *VU) _freeX86regs(); _freeXMMregs(); + OR32ItoM((uptr)&psHu32(GIF_STAT), (GIF_STAT_APATH1 | GIF_STAT_OPH)); // Set PATH1 GIF Status Flags PUSH32R(s_XGKICKReg); PUSH32I((uptr)VU->Mem); - //CALLFunc((u32)countfn); - - if (mtgsThread != NULL) - { + if (mtgsThread) { CALLFunc((uptr)VU1XGKICK_MTGSTransfer); ADD32ItoR(ESP, 8); } - else - { + else { #ifdef PCSX2_DEVBUILD CALLFunc((uptr)vu1xgkick); ADD32ItoR(ESP, 8); @@ -4395,14 +4392,13 @@ void recVUMI_XGKICK_(VURegs *VU) CALLFunc((uptr)GSgifTransfer1); #endif } - psHu32(GIF_STAT) &= ~(GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) + AND32ItoM((uptr)&psHu32(GIF_STAT), ~(GIF_STAT_APATH1 | GIF_STAT_OPH)); // Clear PATH1 GIF Status Flags s_ScheduleXGKICK = 0; } void recVUMI_XGKICK(VURegs *VU, int info) { - if (s_ScheduleXGKICK) - { + if (s_ScheduleXGKICK) { // second xgkick, so launch the first recVUMI_XGKICK_(VU); } @@ -4416,15 +4412,11 @@ void recVUMI_XGKICK(VURegs *VU, int info) SHL32ItoR(isreg, 4); AND32ItoR(isreg, 0x3fff); s_XGKICKReg = isreg; - psHu32(GIF_STAT) |= (GIF_STAT_APATH1 | GIF_STAT_OPH); //Probably should be in the recompilation but im a rec nab :( (Refraction) - - - if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) - { + + if (!SUPERVU_XGKICKDELAY || pc == s_pCurBlock->endpc) { recVUMI_XGKICK_(VU); } - else - { + else { // Erementar Gerad hack. // Corrects the color of some graphics on a game that has somewhat scrambled graphics either way, and only works with ZeroGS. Not even ZZOgl. :) if (g_VUGameFixes & VUFIX_XGKICKDELAY2) From 89dc306861543a711db9cf0c3c4876c49c979273 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 26 Jun 2009 03:12:36 +0000 Subject: [PATCH 287/296] The Gust fix is back in. The previous behavior is in as a gamehack for Fatal Frame, and any other games adversely affected. Checkbox not yet implemented for Windows. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1436 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 1 + pcsx2/HwWrite.cpp | 4 ++-- pcsx2/Linux/HacksDlg.cpp | 2 ++ pcsx2/Linux/Linux.h | 2 ++ pcsx2/Linux/callbacks.h | 16 ++++++++-------- pcsx2/Linux/interface.c | 7 +++++++ pcsx2/Linux/pcsx2.glade | 20 ++++++++++++++++++++ pcsx2/windows/WinMain.cpp | 4 +++- 8 files changed, 45 insertions(+), 11 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index 05a693f28b..a539cea503 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -70,6 +70,7 @@ extern SessionOverrideFlags g_Session; #define CHECK_FPUCOMPAREHACK (Config.GameFixes & 0x4) // Special Fix for Digimon Rumble Arena 2, fixes spinning/hanging on intro-menu. #define CHECK_VUCLIPFLAGHACK (Config.GameFixes & 0x2) // Special Fix for Persona games, maybe others. It's to do with the VU clip flag (again). #define CHECK_FPUMULHACK (Config.GameFixes & 0x8) // Special Fix for Tales of Destiny hangs. +#define CHECK_DMAEXECHACK (Config.GameFixes & 0x10) // Special Fix for Fatal Frame; breaks Gust and Tri-Ace games. //------------ Advanced Options!!! --------------- #define CHECK_VU_OVERFLOW (Config.vuOptions & 0x1) diff --git a/pcsx2/HwWrite.cpp b/pcsx2/HwWrite.cpp index 85645df444..17924fdada 100644 --- a/pcsx2/HwWrite.cpp +++ b/pcsx2/HwWrite.cpp @@ -102,8 +102,8 @@ static void DmaExec( void (*func)(), u32 mem, u32 value ) DMA_LOG( "DMAExec32 Attempt to run DMA while one is already active mem = %x", mem ); // Returning here breaks every single Gust game written. :( - // Not returning here breaks a bunch of other games. I'll let someone else figure it out... - return; + // Not returning here breaks Fatal Frame. Gamefix time. + if (CHECK_DMAEXECHACK) return; } // Upper 16bits of QWC should not be written since QWC is 16bits in size. diff --git a/pcsx2/Linux/HacksDlg.cpp b/pcsx2/Linux/HacksDlg.cpp index e9ebc80936..e620ba0ccc 100644 --- a/pcsx2/Linux/HacksDlg.cpp +++ b/pcsx2/Linux/HacksDlg.cpp @@ -29,6 +29,7 @@ GtkWidget *GameFixDlg, *SpeedHacksDlg; set_checked(GameFixDlg, "check_FPU_Compare", (Config.GameFixes & FLAG_FPU_Compare)); set_checked(GameFixDlg, "check_FPU_Mul", (Config.GameFixes & FLAG_FPU_MUL)); + set_checked(GameFixDlg, "check_DMAExec", (Config.GameFixes & FLAG_DMAExec)); gtk_widget_show_all(GameFixDlg); gtk_widget_set_sensitive(MainWindow, FALSE); @@ -45,6 +46,7 @@ void on_Game_Fix_OK(GtkButton *button, gpointer user_data) Config.GameFixes |= is_checked(GameFixDlg, "check_FPU_Compare") ? FLAG_FPU_Compare : 0; Config.GameFixes |= is_checked(GameFixDlg, "check_FPU_Mul") ? FLAG_FPU_MUL : 0; + Config.GameFixes |= is_checked(GameFixDlg, "check_DMAExec") ? FLAG_DMAExec : 0; SaveConfig(); diff --git a/pcsx2/Linux/Linux.h b/pcsx2/Linux/Linux.h index e1cf6b6272..15252e76c5 100644 --- a/pcsx2/Linux/Linux.h +++ b/pcsx2/Linux/Linux.h @@ -125,6 +125,8 @@ char ee_cycle_labels[3][256] = #define FLAG_FPU_Compare 0x4 //Tales of Destiny - IDC_GAMEFIX5 #define FLAG_FPU_MUL 0x8 +//Fatal Frame +#define FLAG_DMAExec 0x10 #define FLAG_VU_NO_OVERFLOW 0x2 #define FLAG_VU_EXTRA_OVERFLOW 0x40 diff --git a/pcsx2/Linux/callbacks.h b/pcsx2/Linux/callbacks.h index 5cc7803634..5b02c1aa37 100644 --- a/pcsx2/Linux/callbacks.h +++ b/pcsx2/Linux/callbacks.h @@ -1,6 +1,10 @@ #include +void +On_Dialog_Close (GtkDialog *dialog, + gpointer user_data); + void on_Advanced_Defaults (GtkButton *button, gpointer user_data); @@ -246,6 +250,10 @@ void OnHelp_About (GtkMenuItem *menuitem, gpointer user_data); +void +OnCpuCheckToggled (GtkToggleButton *togglebutton, + gpointer user_data); + void OnCpu_Ok (GtkButton *button, gpointer user_data); @@ -257,11 +265,3 @@ OnLogging_Ok (GtkButton *button, void OnMemcards_Ok (GtkButton *button, gpointer user_data); - -void -On_Dialog_Close (GtkDialog *dialog, - gpointer user_data); - -void -OnCpuCheckToggled (GtkToggleButton *togglebutton, - gpointer user_data); diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index 75943ba474..a6d8daaf5f 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -939,6 +939,7 @@ create_GameFixDlg (void) GtkWidget *check_VU_Add_Sub; GtkWidget *check_FPU_Mul; GtkWidget *check_VU_Clip; + GtkWidget *check_DMAExec; GtkWidget *label42; GtkWidget *dialog_action_area1; GtkWidget *cancelbutton1; @@ -990,6 +991,11 @@ create_GameFixDlg (void) gtk_widget_show (check_VU_Clip); gtk_box_pack_start (GTK_BOX (vbox30), check_VU_Clip, FALSE, FALSE, 0); + check_DMAExec = gtk_check_button_new_with_mnemonic (_("DMAExec Hack - Fixes Fatal Frame, possibly others.\nCauses freezing in Gust games and crashing in some TriAce games")); + gtk_widget_set_name (check_DMAExec, "check_DMAExec"); + gtk_widget_show (check_DMAExec); + gtk_box_pack_start (GTK_BOX (vbox30), check_DMAExec, FALSE, FALSE, 0); + label42 = gtk_label_new (_("Some games need special settings.\nConfigure them here.")); gtk_widget_set_name (label42, "label42"); gtk_widget_show (label42); @@ -1033,6 +1039,7 @@ create_GameFixDlg (void) GLADE_HOOKUP_OBJECT (GameFixDlg, check_VU_Add_Sub, "check_VU_Add_Sub"); GLADE_HOOKUP_OBJECT (GameFixDlg, check_FPU_Mul, "check_FPU_Mul"); GLADE_HOOKUP_OBJECT (GameFixDlg, check_VU_Clip, "check_VU_Clip"); + GLADE_HOOKUP_OBJECT (GameFixDlg, check_DMAExec, "check_DMAExec"); GLADE_HOOKUP_OBJECT (GameFixDlg, label42, "label42"); GLADE_HOOKUP_OBJECT_NO_REF (GameFixDlg, dialog_action_area1, "dialog_action_area1"); GLADE_HOOKUP_OBJECT (GameFixDlg, cancelbutton1, "cancelbutton1"); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index b71dc51c68..54f489309d 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -1964,6 +1964,26 @@ If you have problems, Disable all of these and try again. False + + + + True + True + DMAExec Hack - Fixes Fatal Frame, possibly others. +Causes freezing in Gust games and crashing in some TriAce games + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + diff --git a/pcsx2/windows/WinMain.cpp b/pcsx2/windows/WinMain.cpp index 31cbc0add0..c361fcfe37 100644 --- a/pcsx2/windows/WinMain.cpp +++ b/pcsx2/windows/WinMain.cpp @@ -525,6 +525,7 @@ BOOL APIENTRY GameFixes(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) if(Config.GameFixes & 0x4) CheckDlgButton(hDlg, IDC_GAMEFIX3, TRUE);//Digimon FPU compare fix if(Config.GameFixes & 0x2) CheckDlgButton(hDlg, IDC_GAMEFIX4, TRUE);//Persona3/4 fix if(Config.GameFixes & 0x8) CheckDlgButton(hDlg, IDC_GAMEFIX5, TRUE);//Tales of Destiny fix + //if(Config.GameFixes & 0x10) CheckDlgButton(hDlg, IDC_GAMEFIX?, TRUE);//Fatal Frame fix return TRUE; case WM_COMMAND: @@ -535,6 +536,7 @@ BOOL APIENTRY GameFixes(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX3) ? 0x4 : 0; newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX4) ? 0x2 : 0; newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX5) ? 0x8 : 0; + //newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX?) ? 0x10 : 0; EndDialog(hDlg, TRUE); @@ -1056,7 +1058,7 @@ void CreateMainWindow() if(Config.PsxOut) CheckMenuItem(gApp.hMenu,ID_CONSOLE,MF_CHECKED); if(Config.Patch) CheckMenuItem(gApp.hMenu,ID_PATCHES,MF_CHECKED); if(Config.Profiler) CheckMenuItem(gApp.hMenu,ID_PROFILER,MF_CHECKED); - if(Config.cdvdPrint)CheckMenuItem(gApp.hMenu,ID_CDVDPRINT,MF_CHECKED); + if(Config.cdvdPrint) CheckMenuItem(gApp.hMenu,ID_CDVDPRINT,MF_CHECKED); if(Config.closeGSonEsc)CheckMenuItem(gApp.hMenu,ID_CLOSEGS,MF_CHECKED); hStatusWnd = CreateStatusWindow(WS_CHILD | WS_VISIBLE, "", hWnd, 100); From 6b2c26a617bb6fa7162a9ee508c62cdb43e8819f Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Fri, 26 Jun 2009 04:27:18 +0000 Subject: [PATCH 288/296] Added 2 gamefixes to the Windows Side (untested since I don't have the games xD): - DMA Execution hack - Fixes Fatal frame problems by ignoring dma transfers while another one is being executed. - VU XGkick hack - Fixes Erementar Gerad by delaying XGkick. Similar to what SO3 needs, except this game needs more delay. Emulating this correctly is impossible with the current DMAC system, and will most-likely never be fixed correctly. The best (and fastest) way to simulate proper behavior is with a gamefix. (Super VU was doing this by a CRC hack, but I changed it to use this gamefix instead along with microVU.) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1437 96395faa-99c1-11dd-bbfe-3dabce05a288 --- common/include/Pcsx2Config.h | 11 ++++++----- pcsx2/windows/WinMain.cpp | 23 +++++++++++++---------- pcsx2/windows/pcsx2.rc | 26 ++++++++++++++------------ pcsx2/windows/resource.h | 2 ++ pcsx2/x86/microVU_Misc.h | 6 ++++-- pcsx2/x86/sVU_zerorec.cpp | 7 +------ 6 files changed, 40 insertions(+), 35 deletions(-) diff --git a/common/include/Pcsx2Config.h b/common/include/Pcsx2Config.h index a539cea503..7b467f8093 100644 --- a/common/include/Pcsx2Config.h +++ b/common/include/Pcsx2Config.h @@ -66,11 +66,12 @@ extern SessionOverrideFlags g_Session; #define CHECK_VU1REC (!g_Session.ForceDisableVU1rec && (Config.Options&PCSX2_VU1REC)) //------------ SPECIAL GAME FIXES!!! --------------- -#define CHECK_VUADDSUBHACK (Config.GameFixes & 0x1) // Special Fix for Tri-ace games, they use an encryption algorithm that requires VU addi opcode to be bit-accurate. -#define CHECK_FPUCOMPAREHACK (Config.GameFixes & 0x4) // Special Fix for Digimon Rumble Arena 2, fixes spinning/hanging on intro-menu. -#define CHECK_VUCLIPFLAGHACK (Config.GameFixes & 0x2) // Special Fix for Persona games, maybe others. It's to do with the VU clip flag (again). -#define CHECK_FPUMULHACK (Config.GameFixes & 0x8) // Special Fix for Tales of Destiny hangs. -#define CHECK_DMAEXECHACK (Config.GameFixes & 0x10) // Special Fix for Fatal Frame; breaks Gust and Tri-Ace games. +#define CHECK_VUADDSUBHACK (Config.GameFixes & 0x01) // Special Fix for Tri-ace games, they use an encryption algorithm that requires VU addi opcode to be bit-accurate. +#define CHECK_FPUCOMPAREHACK (Config.GameFixes & 0x04) // Special Fix for Digimon Rumble Arena 2, fixes spinning/hanging on intro-menu. +#define CHECK_VUCLIPFLAGHACK (Config.GameFixes & 0x02) // Special Fix for Persona games, maybe others. It's to do with the VU clip flag (again). +#define CHECK_FPUMULHACK (Config.GameFixes & 0x08) // Special Fix for Tales of Destiny hangs. +#define CHECK_DMAEXECHACK (Config.GameFixes & 0x10) // Special Fix for Fatal Frame; breaks Gust and Tri-Ace games. +#define CHECK_XGKICKHACK (Config.GameFixes & 0x20) // Special Fix for Erementar Gerad, adds more delay to VU XGkick instructions. Corrects the color of some graphics. //------------ Advanced Options!!! --------------- #define CHECK_VU_OVERFLOW (Config.vuOptions & 0x1) diff --git a/pcsx2/windows/WinMain.cpp b/pcsx2/windows/WinMain.cpp index c361fcfe37..5769719103 100644 --- a/pcsx2/windows/WinMain.cpp +++ b/pcsx2/windows/WinMain.cpp @@ -521,22 +521,25 @@ BOOL APIENTRY GameFixes(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) { switch (message) { case WM_INITDIALOG: - if(Config.GameFixes & 0x1) CheckDlgButton(hDlg, IDC_GAMEFIX2, TRUE);//Tri-Ace fix - if(Config.GameFixes & 0x4) CheckDlgButton(hDlg, IDC_GAMEFIX3, TRUE);//Digimon FPU compare fix - if(Config.GameFixes & 0x2) CheckDlgButton(hDlg, IDC_GAMEFIX4, TRUE);//Persona3/4 fix - if(Config.GameFixes & 0x8) CheckDlgButton(hDlg, IDC_GAMEFIX5, TRUE);//Tales of Destiny fix - //if(Config.GameFixes & 0x10) CheckDlgButton(hDlg, IDC_GAMEFIX?, TRUE);//Fatal Frame fix + if (Config.GameFixes & 0x01) CheckDlgButton(hDlg, IDC_GAMEFIX2, TRUE); // Tri-Ace fix + if (Config.GameFixes & 0x04) CheckDlgButton(hDlg, IDC_GAMEFIX3, TRUE); // Digimon FPU compare fix + if (Config.GameFixes & 0x02) CheckDlgButton(hDlg, IDC_GAMEFIX4, TRUE); // Persona3/4 fix + if (Config.GameFixes & 0x08) CheckDlgButton(hDlg, IDC_GAMEFIX5, TRUE); // Tales of Destiny fix + if (Config.GameFixes & 0x10) CheckDlgButton(hDlg, IDC_GAMEFIX6, TRUE); // Fatal Frame fix + if (Config.GameFixes & 0x20) CheckDlgButton(hDlg, IDC_GAMEFIX8, TRUE); // Erementar Gerad fix + return TRUE; case WM_COMMAND: if (LOWORD(wParam) == IDOK) { uint newfixes = 0; - newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX2) ? 0x1 : 0; - newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX3) ? 0x4 : 0; - newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX4) ? 0x2 : 0; - newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX5) ? 0x8 : 0; - //newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX?) ? 0x10 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX2) ? 0x01 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX3) ? 0x04 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX4) ? 0x02 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX5) ? 0x08 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX6) ? 0x10 : 0; + newfixes |= IsDlgButtonChecked(hDlg, IDC_GAMEFIX8) ? 0x20 : 0; EndDialog(hDlg, TRUE); diff --git a/pcsx2/windows/pcsx2.rc b/pcsx2/windows/pcsx2.rc index 4de07e7143..9f6d38bfbc 100644 --- a/pcsx2/windows/pcsx2.rc +++ b/pcsx2/windows/pcsx2.rc @@ -74,23 +74,26 @@ LANGUAGE LANG_GERMAN, SUBLANG_GERMAN // Dialog // -IDD_GAMEFIXES DIALOGEX 0, 0, 279, 123 +IDD_GAMEFIXES DIALOGEX 0, 0, 279, 154 STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_CAPTION | WS_SYSMENU -CAPTION "Game Special Fixes" FONT 8, "MS Shell Dlg", 400, 0, 0x1 BEGIN - DEFPUSHBUTTON "OK",IDOK,87,96,50,14 - PUSHBUTTON "Cancel",IDCANCEL,142,96,50,14 + DEFPUSHBUTTON "OK",IDOK,86,133,50,14 + PUSHBUTTON "Cancel",IDCANCEL,142,133,50,14 CTEXT "Some games need special settings.\nConfigure them here.",IDC_STATIC,7,7,265,17 - GROUPBOX "PCSX2 Gamefixes",IDC_STATIC,7,22,265,94 + GROUPBOX "PCSX2 Gamefixes",IDC_STATIC,7,22,265,104 CONTROL "FPU Compare Hack - Special fix for Digimon Rumble Arena 2.",IDC_GAMEFIX3, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,36,249,10 + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,51,249,10 CONTROL "VU Add Hack - Special fix for Tri-Ace games!",IDC_GAMEFIX2, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,65,252,10 - CONTROL "VU Clip Hack - Fixes missing ground geometry in Persona.",IDC_GAMEFIX4, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,79,238,10 + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,80,252,10 + CONTROL "VU Clip Hack - Fixes missing ground geometry in Persona. (Super VU)",IDC_GAMEFIX4, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,94,238,10 CONTROL "FPU Mul Hack - Special fix for Tales of Destiny.",IDC_GAMEFIX5, - "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,50,249,10 + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,65,249,10 + CONTROL "DMA Execution Hack - Special fix for Fatal Frame.",IDC_GAMEFIX6, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,37,238,10 + CONTROL "VU XGkick Hack - Special fix for Erementar Gerad.",IDC_GAMEFIX8, + "Button",BS_AUTOCHECKBOX | WS_TABSTOP,14,108,238,10 END @@ -108,8 +111,7 @@ BEGIN RIGHTMARGIN, 272 VERTGUIDE, 14 TOPMARGIN, 7 - BOTTOMMARGIN, 116 - HORZGUIDE, 103 + BOTTOMMARGIN, 147 END END #endif // APSTUDIO_INVOKED diff --git a/pcsx2/windows/resource.h b/pcsx2/windows/resource.h index 50cc46e57d..4aa25813f4 100644 --- a/pcsx2/windows/resource.h +++ b/pcsx2/windows/resource.h @@ -237,7 +237,9 @@ #define IDC_EE_ROUNDMODE0 1305 #define IDC_FCOR 1305 #define IDC_GAMEFIX7 1305 +#define IDC_GAMEFIX6 1305 #define IDC_EE_ROUNDMODE1 1306 +#define IDC_GAMEFIX8 1306 #define IDC_EE_ROUNDMODE2 1307 #define IDC_EE_ROUNDMODE3 1308 #define IDC_EESYNC_DEFAULT 1308 diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index 8fbf0ecbd0..aa9c991eb6 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -275,8 +275,10 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define CHECK_VU_MINMAXHACK (u32)Config.Hacks.vuMinMax // (Can cause SPS, Black Screens, etc...) // Unknown Data -#define mVU_XGKICK_CYCLES 1 // Its unknown at recompile time how long the xgkick transfer will take - // so give it a value that makes games happy :) (SO3 is fine at 1 cycle delay) +#define mVU_XGKICK_CYCLES ((CHECK_XGKICKHACK) ? 3 : 1) +// Its unknown at recompile time how long the xgkick transfer will take +// so give it a value that makes games happy :) (SO3 is fine at 1 cycle delay) + // Cache Limit Check #define mVUcacheCheck(ptr, start, limit) { \ diff --git a/pcsx2/x86/sVU_zerorec.cpp b/pcsx2/x86/sVU_zerorec.cpp index 386fe52516..1b8e3c582f 100644 --- a/pcsx2/x86/sVU_zerorec.cpp +++ b/pcsx2/x86/sVU_zerorec.cpp @@ -4417,12 +4417,7 @@ void recVUMI_XGKICK(VURegs *VU, int info) recVUMI_XGKICK_(VU); } else { - // Erementar Gerad hack. - // Corrects the color of some graphics on a game that has somewhat scrambled graphics either way, and only works with ZeroGS. Not even ZZOgl. :) - if (g_VUGameFixes & VUFIX_XGKICKDELAY2) - s_ScheduleXGKICK = min((u32)4, (s_pCurBlock->endpc - pc) / 8); - else - s_ScheduleXGKICK = 2; + s_ScheduleXGKICK = (CHECK_XGKICKHACK) ? (min((u32)4, (s_pCurBlock->endpc-pc)/8)) : 2; } } From 6e5d4a5d1539e3869d3bcb4d11e693e513aeaedc Mon Sep 17 00:00:00 2001 From: arcum42 Date: Fri, 26 Jun 2009 05:14:58 +0000 Subject: [PATCH 289/296] Brought the Erementar Gerad gamefix over to Linux, and got rid of some junk that isn't neccesary now that the crc version of the hack is gone. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1438 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/CDVD/CDVD.cpp | 2 -- pcsx2/Elfheader.cpp | 18 +----------------- pcsx2/Elfheader.h | 1 - pcsx2/Linux/HacksDlg.cpp | 2 ++ pcsx2/Linux/Linux.h | 2 ++ pcsx2/Linux/interface.c | 7 +++++++ pcsx2/Linux/pcsx2.glade | 20 ++++++++++++++++++++ pcsx2/Patch.cpp | 3 --- pcsx2/VU.h | 7 ------- 9 files changed, 32 insertions(+), 30 deletions(-) diff --git a/pcsx2/CDVD/CDVD.cpp b/pcsx2/CDVD/CDVD.cpp index 3bd4b07bc4..486c09af24 100644 --- a/pcsx2/CDVD/CDVD.cpp +++ b/pcsx2/CDVD/CDVD.cpp @@ -394,7 +394,6 @@ void cdvdReadKey(u8 arg0, u16 arg1, u32 arg2, u8* key) { { ElfCRC = loadElfCRC( str ); ElfApplyPatches(); - LoadGameSpecificSettings(); GSsetGameCRC( ElfCRC, 0 ); } } @@ -584,7 +583,6 @@ void cdvdNewDiskCB() { ElfCRC = loadElfCRC( str ); ElfApplyPatches(); - LoadGameSpecificSettings(); GSsetGameCRC( ElfCRC, 0 ); } diff --git a/pcsx2/Elfheader.cpp b/pcsx2/Elfheader.cpp index 5bf94204b3..d77d4b015f 100644 --- a/pcsx2/Elfheader.cpp +++ b/pcsx2/Elfheader.cpp @@ -565,23 +565,7 @@ int loadElfFile(const char *filename) Console::Status( "loadElfFile: %s; CRC = %8.8X", params filename, ElfCRC); ElfApplyPatches(); - LoadGameSpecificSettings(); - + return 0; } -#include "VU.h" -int g_VUGameFixes = 0; - -// fixme - this should be moved to patches or eliminated -void LoadGameSpecificSettings() -{ - // default - g_VUGameFixes = 0; - - switch(ElfCRC) { - case 0xb99379b7: // erementar gerad (discolored chars) - g_VUGameFixes |= VUFIX_XGKICKDELAY2; // Tested - still needed - arcum42 - break; - } -} diff --git a/pcsx2/Elfheader.h b/pcsx2/Elfheader.h index c6c93beb1d..4e352d020e 100644 --- a/pcsx2/Elfheader.h +++ b/pcsx2/Elfheader.h @@ -26,7 +26,6 @@ extern unsigned int args_ptr; //------------------- int loadElfFile(const char *filename); u32 loadElfCRC(const char *filename); -void LoadGameSpecificSettings(); void ElfApplyPatches(); extern u32 ElfCRC; diff --git a/pcsx2/Linux/HacksDlg.cpp b/pcsx2/Linux/HacksDlg.cpp index e620ba0ccc..c4e17fcab1 100644 --- a/pcsx2/Linux/HacksDlg.cpp +++ b/pcsx2/Linux/HacksDlg.cpp @@ -30,6 +30,7 @@ GtkWidget *GameFixDlg, *SpeedHacksDlg; set_checked(GameFixDlg, "check_FPU_Compare", (Config.GameFixes & FLAG_FPU_Compare)); set_checked(GameFixDlg, "check_FPU_Mul", (Config.GameFixes & FLAG_FPU_MUL)); set_checked(GameFixDlg, "check_DMAExec", (Config.GameFixes & FLAG_DMAExec)); + set_checked(GameFixDlg, "check_XGKick", (Config.GameFixes & FLAG_XGKick)); gtk_widget_show_all(GameFixDlg); gtk_widget_set_sensitive(MainWindow, FALSE); @@ -47,6 +48,7 @@ void on_Game_Fix_OK(GtkButton *button, gpointer user_data) Config.GameFixes |= is_checked(GameFixDlg, "check_FPU_Compare") ? FLAG_FPU_Compare : 0; Config.GameFixes |= is_checked(GameFixDlg, "check_FPU_Mul") ? FLAG_FPU_MUL : 0; Config.GameFixes |= is_checked(GameFixDlg, "check_DMAExec") ? FLAG_DMAExec : 0; + Config.GameFixes |= is_checked(GameFixDlg, "check_XGKick") ? FLAG_XGKick : 0; SaveConfig(); diff --git a/pcsx2/Linux/Linux.h b/pcsx2/Linux/Linux.h index 15252e76c5..49de616d43 100644 --- a/pcsx2/Linux/Linux.h +++ b/pcsx2/Linux/Linux.h @@ -127,6 +127,8 @@ char ee_cycle_labels[3][256] = #define FLAG_FPU_MUL 0x8 //Fatal Frame #define FLAG_DMAExec 0x10 +//Erementar Gerad +#define FLAG_XGKick 0x20 #define FLAG_VU_NO_OVERFLOW 0x2 #define FLAG_VU_EXTRA_OVERFLOW 0x40 diff --git a/pcsx2/Linux/interface.c b/pcsx2/Linux/interface.c index a6d8daaf5f..774b27f02c 100644 --- a/pcsx2/Linux/interface.c +++ b/pcsx2/Linux/interface.c @@ -940,6 +940,7 @@ create_GameFixDlg (void) GtkWidget *check_FPU_Mul; GtkWidget *check_VU_Clip; GtkWidget *check_DMAExec; + GtkWidget *check_XGKick; GtkWidget *label42; GtkWidget *dialog_action_area1; GtkWidget *cancelbutton1; @@ -996,6 +997,11 @@ create_GameFixDlg (void) gtk_widget_show (check_DMAExec); gtk_box_pack_start (GTK_BOX (vbox30), check_DMAExec, FALSE, FALSE, 0); + check_XGKick = gtk_check_button_new_with_mnemonic (_("VU XGkick Hack - Fix for Erementar Gerad. \nAdds more delay to VU XGkick instructions. Corrects the color of some graphics.")); + gtk_widget_set_name (check_XGKick, "check_XGKick"); + gtk_widget_show (check_XGKick); + gtk_box_pack_start (GTK_BOX (vbox30), check_XGKick, FALSE, FALSE, 0); + label42 = gtk_label_new (_("Some games need special settings.\nConfigure them here.")); gtk_widget_set_name (label42, "label42"); gtk_widget_show (label42); @@ -1040,6 +1046,7 @@ create_GameFixDlg (void) GLADE_HOOKUP_OBJECT (GameFixDlg, check_FPU_Mul, "check_FPU_Mul"); GLADE_HOOKUP_OBJECT (GameFixDlg, check_VU_Clip, "check_VU_Clip"); GLADE_HOOKUP_OBJECT (GameFixDlg, check_DMAExec, "check_DMAExec"); + GLADE_HOOKUP_OBJECT (GameFixDlg, check_XGKick, "check_XGKick"); GLADE_HOOKUP_OBJECT (GameFixDlg, label42, "label42"); GLADE_HOOKUP_OBJECT_NO_REF (GameFixDlg, dialog_action_area1, "dialog_action_area1"); GLADE_HOOKUP_OBJECT (GameFixDlg, cancelbutton1, "cancelbutton1"); diff --git a/pcsx2/Linux/pcsx2.glade b/pcsx2/Linux/pcsx2.glade index 54f489309d..841d42da45 100644 --- a/pcsx2/Linux/pcsx2.glade +++ b/pcsx2/Linux/pcsx2.glade @@ -1984,6 +1984,26 @@ Causes freezing in Gust games and crashing in some TriAce games False + + + + True + True + VU XGkick Hack - Fix for Erementar Gerad. +Adds more delay to VU XGkick instructions. Corrects the color of some graphics. + True + GTK_RELIEF_NORMAL + True + False + False + True + + + 0 + False + False + + diff --git a/pcsx2/Patch.cpp b/pcsx2/Patch.cpp index 618e6f5041..85d40b1277 100644 --- a/pcsx2/Patch.cpp +++ b/pcsx2/Patch.cpp @@ -653,13 +653,10 @@ int AddPatch(int Mode, int Place, int Address, int Size, u64 data) void patchFunc_ffxhack( char * cmd, char * param ) { - //Keeping this as a dummy a while :p - //g_FFXHack = 1; } void patchFunc_xkickdelay( char * cmd, char * param ) { - g_VUGameFixes |= VUFIX_XGKICKDELAY2; } void patchFunc_fastmemory( char * cmd, char * param ) diff --git a/pcsx2/VU.h b/pcsx2/VU.h index fb9602afe7..c339c4cf9b 100644 --- a/pcsx2/VU.h +++ b/pcsx2/VU.h @@ -196,10 +196,3 @@ static __forceinline u32* GET_VU_MEM(VURegs* VU, u32 addr) return (u32*)(VU0.Mem+(addr&0x0fff)); // for addr 0x0000 to 0x4000 just wrap around } - - -// various fixes to enable per game (all are off by default) -#define VUFIX_SIGNEDZERO 1 -#define VUFIX_EXTRAFLAGS 2 -#define VUFIX_XGKICKDELAY2 4 -extern int g_VUGameFixes; From 7bb9a3cc2555d3bcaf4120d2c249e1a0448db3f3 Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 27 Jun 2009 03:32:33 +0000 Subject: [PATCH 290/296] GSdx: changed a lot of things, expect new bugs :P git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1439 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GPURenderer.cpp | 2 +- plugins/GSdx/GS.cpp | 12 +- plugins/GSdx/GS.h | 33 +- plugins/GSdx/GSCrc.cpp | 3 +- plugins/GSdx/GSCrc.h | 1 + plugins/GSdx/GSDrawingContext.h | 2 +- plugins/GSdx/GSFunctionMap.h | 16 +- plugins/GSdx/GSLocalMemory.cpp | 149 +-- plugins/GSdx/GSRenderer.cpp | 19 +- plugins/GSdx/GSRenderer.h | 7 +- plugins/GSdx/GSRendererDX.cpp | 23 + plugins/GSdx/GSRendererDX.h | 338 +++++ plugins/GSdx/GSRendererDX10.cpp | 266 ++++ .../{GSRendererHW10.h => GSRendererDX10.h} | 11 +- plugins/GSdx/GSRendererDX11.cpp | 266 ++++ .../{GSRendererHW11.h => GSRendererDX11.h} | 10 +- plugins/GSdx/GSRendererDX9.cpp | 322 +++++ .../GSdx/{GSRendererHW9.h => GSRendererDX9.h} | 13 +- plugins/GSdx/GSRendererHW.h | 46 +- plugins/GSdx/GSRendererHW10.cpp | 529 -------- plugins/GSdx/GSRendererHW11.cpp | 529 -------- plugins/GSdx/GSRendererHW9.cpp | 573 --------- plugins/GSdx/GSRendererOGL.cpp | 3 +- plugins/GSdx/GSRendererOGL.h | 2 +- plugins/GSdx/GSRendererSW.cpp | 4 +- plugins/GSdx/GSTextureCache.cpp | 1133 +++++++++-------- plugins/GSdx/GSTextureCache.h | 125 +- plugins/GSdx/GSTextureCache10.cpp | 464 +++---- plugins/GSdx/GSTextureCache10.h | 34 +- plugins/GSdx/GSTextureCache11.cpp | 381 +++--- plugins/GSdx/GSTextureCache11.h | 34 +- plugins/GSdx/GSTextureCache9.cpp | 378 +++--- plugins/GSdx/GSTextureCache9.h | 34 +- plugins/GSdx/GSTextureCacheOGL.cpp | 23 +- plugins/GSdx/GSTextureCacheOGL.h | 34 +- plugins/GSdx/GSTextureCacheSW.cpp | 47 +- plugins/GSdx/GSTextureCacheSW.h | 3 - plugins/GSdx/GSTextureFX.cpp | 13 + plugins/GSdx/GSTextureFX.h | 25 +- plugins/GSdx/GSTextureFX10.cpp | 92 +- plugins/GSdx/GSTextureFX10.h | 15 +- plugins/GSdx/GSTextureFX11.cpp | 94 +- plugins/GSdx/GSTextureFX11.h | 15 +- plugins/GSdx/GSTextureFX9.cpp | 75 +- plugins/GSdx/GSTextureFX9.h | 14 +- plugins/GSdx/GSVector.cpp | 12 +- plugins/GSdx/GSVector.h | 27 +- plugins/GSdx/GSVertexHW.h | 18 +- plugins/GSdx/GSdx_vs2008.vcproj | 120 +- plugins/GSdx/GSdx_vs2010.vcxproj | 12 +- plugins/GSdx/GSdx_vs2010.vcxproj.filters | 12 +- plugins/GSdx/res/tfx.fx | 28 +- 52 files changed, 3004 insertions(+), 3437 deletions(-) create mode 100644 plugins/GSdx/GSRendererDX.cpp create mode 100644 plugins/GSdx/GSRendererDX.h create mode 100644 plugins/GSdx/GSRendererDX10.cpp rename plugins/GSdx/{GSRendererHW10.h => GSRendererDX10.h} (86%) create mode 100644 plugins/GSdx/GSRendererDX11.cpp rename plugins/GSdx/{GSRendererHW11.h => GSRendererDX11.h} (83%) create mode 100644 plugins/GSdx/GSRendererDX9.cpp rename plugins/GSdx/{GSRendererHW9.h => GSRendererDX9.h} (85%) delete mode 100644 plugins/GSdx/GSRendererHW10.cpp delete mode 100644 plugins/GSdx/GSRendererHW11.cpp delete mode 100644 plugins/GSdx/GSRendererHW9.cpp diff --git a/plugins/GSdx/GPURenderer.cpp b/plugins/GSdx/GPURenderer.cpp index 944d5d7914..4e2ba0d3a0 100644 --- a/plugins/GSdx/GPURenderer.cpp +++ b/plugins/GSdx/GPURenderer.cpp @@ -185,7 +185,7 @@ LRESULT CALLBACK GPURenderer::WndProc(HWND hWnd, UINT message, WPARAM wParam, LP if(i != m_wnd2gpu.end()) { - return (*i).second->OnMessage(message, wParam, lParam); + return i->second->OnMessage(message, wParam, lParam); } ASSERT(0); diff --git a/plugins/GSdx/GS.cpp b/plugins/GSdx/GS.cpp index 67aeab825d..f7751d5d33 100644 --- a/plugins/GSdx/GS.cpp +++ b/plugins/GSdx/GS.cpp @@ -21,9 +21,9 @@ #include "stdafx.h" #include "GSUtil.h" -#include "GSRendererHW9.h" -#include "GSRendererHW10.h" -#include "GSRendererHW11.h" +#include "GSRendererDX9.h" +#include "GSRendererDX10.h" +#include "GSRendererDX11.h" #include "GSRendererOGL.h" #include "GSRendererSW.h" #include "GSRendererNull.h" @@ -121,13 +121,13 @@ static INT32 GSopen(void* dsp, char* title, int mt, int renderer) switch(renderer) { default: - case 0: s_gs = new GSRendererHW9(s_basemem, !!mt, s_irq); break; + case 0: s_gs = new GSRendererDX9(s_basemem, !!mt, s_irq); break; case 1: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice9()); break; case 2: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice9()); break; - case 3: s_gs = new GSRendererHW10(s_basemem, !!mt, s_irq); break; + case 3: s_gs = new GSRendererDX10(s_basemem, !!mt, s_irq); break; case 4: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice10()); break; case 5: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice10()); break; - case 6: s_gs = new GSRendererHW11(s_basemem, !!mt, s_irq); break; + case 6: s_gs = new GSRendererDX11(s_basemem, !!mt, s_irq); break; case 7: s_gs = new GSRendererSW(s_basemem, !!mt, s_irq, new GSDevice11()); break; case 8: s_gs = new GSRendererNull(s_basemem, !!mt, s_irq, new GSDevice11()); break; #if 0 diff --git a/plugins/GSdx/GS.h b/plugins/GSdx/GS.h index 2e523d9003..3ef534c9bc 100644 --- a/plugins/GSdx/GS.h +++ b/plugins/GSdx/GS.h @@ -28,6 +28,9 @@ #define PLUGIN_VERSION 15 +#define MAX_PAGES 512 +#define MAX_BLOCKS 16384 + #include "GSVector.h" #pragma pack(push, 1) @@ -813,13 +816,39 @@ REG64_(GIFReg, TEX1) uint32 K:12; uint32 _PAD4:20; REG_END2 + + bool IsMinLinear() const {return (MMIN == 1) || (MMIN & 4);} + bool IsMagLinear() const {return MMAG;} + bool IsLinear() const { - bool mmag = (MMAG & 1); - bool mmin = (MMIN == 1) || (MMIN & 4); + bool mmin = IsMinLinear(); + bool mmag = IsMagLinear(); return !LCM ? mmag || mmin : K <= 0 ? mmag : mmin; } + + bool IsLinear(float qmin, float qmax) const + { + bool mmin = IsMinLinear(); + bool mmag = IsMagLinear(); + + if(mmag == mmin) return mmag; + + float LODmin = K; + float LODmax = K; + + if(!LCM) + { + float f = (float)(1 << L) / log(2.0f); + + LODmin += log(1.0f / abs(qmax)) * f; + LODmax += log(1.0f / abs(qmin)) * f; + } + + return LODmax <= 0 ? mmag : LODmin > 0 ? mmin : mmag || mmin; + } + REG_END2 REG64_(GIFReg, TEX2) diff --git a/plugins/GSdx/GSCrc.cpp b/plugins/GSdx/GSCrc.cpp index a5c6977c03..00fa942582 100644 --- a/plugins/GSdx/GSCrc.cpp +++ b/plugins/GSdx/GSCrc.cpp @@ -31,6 +31,7 @@ CRC::Game CRC::m_games[] = {0xA6167B59, Lamune, JP, PointListPalette}, {0xDDB59F46, KyuuketsuKitanMoonties, JP, PointListPalette}, {0xC8EE2562, PiaCarroteYoukosoGPGakuenPrincess, JP, PointListPalette}, + {0x6CF94A43, KazokuKeikakuKokoroNoKizuna, JP, PointListPalette}, {0xa39517ab, FFX, EU, 0}, {0xa39517ae, FFX, FR, 0}, {0x941bb7d9, FFX, DE, 0}, @@ -156,7 +157,7 @@ CRC::Game CRC::Lookup(uint32 crc) if(i != m_map.end()) { - return *(*i).second; + return *i->second; } return m_games[0]; diff --git a/plugins/GSdx/GSCrc.h b/plugins/GSdx/GSCrc.h index d5791c7b74..06f6698d55 100644 --- a/plugins/GSdx/GSCrc.h +++ b/plugins/GSdx/GSCrc.h @@ -33,6 +33,7 @@ public: Lamune, KyuuketsuKitanMoonties, PiaCarroteYoukosoGPGakuenPrincess, + KazokuKeikakuKokoroNoKizuna, FFX, FFX2, FFXII, diff --git a/plugins/GSdx/GSDrawingContext.h b/plugins/GSdx/GSDrawingContext.h index 0f72f2f930..75e4b73ce8 100644 --- a/plugins/GSdx/GSDrawingContext.h +++ b/plugins/GSdx/GSDrawingContext.h @@ -89,7 +89,7 @@ public: (int)SCISSOR.SCAX1 + 1, (int)SCISSOR.SCAY1 + 1); - scissor.ex = GSVector4i( + scissor.ex = GSVector4( (int)SCISSOR.SCAX0, (int)SCISSOR.SCAY0, (int)SCISSOR.SCAX1, diff --git a/plugins/GSdx/GSFunctionMap.h b/plugins/GSdx/GSFunctionMap.h index b825e7ea5a..050e49a92a 100644 --- a/plugins/GSdx/GSFunctionMap.h +++ b/plugins/GSdx/GSFunctionMap.h @@ -68,7 +68,7 @@ public: { for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - delete (*i).second; + delete i->second; } } @@ -80,7 +80,7 @@ public: if(i != m_map_active.end()) { - m_active = (*i).second; + m_active = i->second; } else { @@ -92,7 +92,7 @@ public: p->frame = (uint64)-1; - p->f = i != m_map.end() ? (*i).second : GetDefaultFunction(key); + p->f = i != m_map.end() ? i->second : GetDefaultFunction(key); m_map_active[key] = p; @@ -123,7 +123,7 @@ public: for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - ActivePtr* p = (*i).second; + ActivePtr* p = i->second; if(p->frames) { @@ -133,8 +133,8 @@ public: for(hash_map::iterator i = m_map_active.begin(); i != m_map_active.end(); i++) { - KEY key = (*i).first; - ActivePtr* p = (*i).second; + KEY key = i->first; + ActivePtr* p = i->second; if(p->frames > 0) { @@ -179,7 +179,7 @@ public: { for(hash_map::iterator i = m_cgmap.begin(); i != m_cgmap.end(); i++) { - delete (*i).second; + delete i->second; } } @@ -191,7 +191,7 @@ public: if(i != m_cgmap.end()) { - cg = (*i).second; + cg = i->second; } else { diff --git a/plugins/GSdx/GSLocalMemory.cpp b/plugins/GSdx/GSLocalMemory.cpp index 8f3313dfe3..7d5a78fbed 100644 --- a/plugins/GSdx/GSLocalMemory.cpp +++ b/plugins/GSdx/GSLocalMemory.cpp @@ -30,14 +30,17 @@ #define ASSERT_BLOCK(r, w, h) \ ASSERT((r).width() >= w && (r).height() >= h && !((r).left&(w-1)) && !((r).top&(h-1)) && !((r).right&(w-1)) && !((r).bottom&(h-1))); \ -#define FOREACH_BLOCK_START(w, h, bpp) \ +#define FOREACH_BLOCK_START(w, h, bpp, format) \ + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[format]; \ uint32 bp = TEX0.TBP0; \ uint32 bw = TEX0.TBW; \ int offset = dstpitch * h - r.width() * bpp / 8; \ for(int y = r.top, ye = r.bottom; y < ye; y += h, dst += offset) \ { ASSERT_BLOCK(r, w, h); \ + uint32 base = psm.bn(0, y, bp, bw); \ for(int x = r.left, xe = r.right; x < xe; x += w, dst += w * bpp / 8) \ { \ + const uint8* src = BlockPtr(base + psm.blockOffset[x >> 3]); \ #define FOREACH_BLOCK_END }} @@ -462,7 +465,7 @@ GSLocalMemory::~GSLocalMemory() for(hash_map::iterator i = m_omap.begin(); i != m_omap.end(); i++) { - Offset* o = (*i).second; + Offset* o = i->second; _aligned_free(o->col[0]); @@ -471,7 +474,7 @@ GSLocalMemory::~GSLocalMemory() for(hash_map::iterator i = m_o4map.begin(); i != m_o4map.end(); i++) { - _aligned_free((*i).second); + _aligned_free(i->second); } } @@ -487,7 +490,7 @@ GSLocalMemory::Offset* GSLocalMemory::GetOffset(uint32 bp, uint32 bw, uint32 psm if(i != m_omap.end()) { - return (*i).second; + return i->second; } Offset* o = (Offset*)_aligned_malloc(sizeof(Offset), 16); @@ -536,7 +539,7 @@ GSLocalMemory::Offset4* GSLocalMemory::GetOffset4(const GIFRegFRAME& FRAME, cons if(i != m_o4map.end()) { - return (*i).second; + return i->second; } Offset4* o = (Offset4*)_aligned_malloc(sizeof(Offset4), 16); @@ -1435,9 +1438,9 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB void GSLocalMemory::ReadTexture32(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMCT32) { - ReadBlock32(BlockPtr32(x, y, bp, bw), dst, dstpitch); + ReadBlock32(src, dst, dstpitch); } FOREACH_BLOCK_END } @@ -1446,17 +1449,17 @@ void GSLocalMemory::ReadTexture24(const GSVector4i& r, uint8* dst, int dstpitch, { if(TEXA.AEM) { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMCT24) { - ReadAndExpandBlock24(BlockPtr32(x, y, bp, bw), dst, dstpitch, TEXA); + ReadAndExpandBlock24(src, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } else { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMCT24) { - ReadAndExpandBlock24(BlockPtr32(x, y, bp, bw), dst, dstpitch, TEXA); + ReadAndExpandBlock24(src, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1466,9 +1469,9 @@ void GSLocalMemory::ReadTexture16(const GSVector4i& r, uint8* dst, int dstpitch, { __declspec(align(16)) uint16 block[16 * 8]; - FOREACH_BLOCK_START(16, 8, 32) + FOREACH_BLOCK_START(16, 8, 32, PSM_PSMCT16) { - ReadBlock16(BlockPtr16(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock16(src, (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1479,9 +1482,9 @@ void GSLocalMemory::ReadTexture16S(const GSVector4i& r, uint8* dst, int dstpitch { __declspec(align(16)) uint16 block[16 * 8]; - FOREACH_BLOCK_START(16, 8, 32) + FOREACH_BLOCK_START(16, 8, 32, PSM_PSMCT16S) { - ReadBlock16(BlockPtr16S(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock16(src, (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1492,9 +1495,9 @@ void GSLocalMemory::ReadTexture8(const GSVector4i& r, uint8* dst, int dstpitch, { const uint32* pal = m_clut; - FOREACH_BLOCK_START(16, 16, 32) + FOREACH_BLOCK_START(16, 16, 32, PSM_PSMT8) { - ReadAndExpandBlock8_32(BlockPtr8(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock8_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1503,9 +1506,9 @@ void GSLocalMemory::ReadTexture4(const GSVector4i& r, uint8* dst, int dstpitch, { const uint64* pal = m_clut; - FOREACH_BLOCK_START(32, 16, 32) + FOREACH_BLOCK_START(32, 16, 32, PSM_PSMT4) { - ReadAndExpandBlock4_32(BlockPtr4(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1514,9 +1517,9 @@ void GSLocalMemory::ReadTexture8H(const GSVector4i& r, uint8* dst, int dstpitch, { const uint32* pal = m_clut; - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT8H) { - ReadAndExpandBlock8H_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock8H_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1525,9 +1528,9 @@ void GSLocalMemory::ReadTexture4HL(const GSVector4i& r, uint8* dst, int dstpitch { const uint32* pal = m_clut; - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT4HL) { - ReadAndExpandBlock4HL_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4HL_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1536,18 +1539,18 @@ void GSLocalMemory::ReadTexture4HH(const GSVector4i& r, uint8* dst, int dstpitch { const uint32* pal = m_clut; - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT4HH) { - ReadAndExpandBlock4HH_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4HH_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture32Z(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMZ32) { - ReadBlock32(BlockPtr32Z(x, y, bp, bw), dst, dstpitch); + ReadBlock32(src, dst, dstpitch); } FOREACH_BLOCK_END } @@ -1556,17 +1559,17 @@ void GSLocalMemory::ReadTexture24Z(const GSVector4i& r, uint8* dst, int dstpitch { if(TEXA.AEM) { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMZ24) { - ReadAndExpandBlock24(BlockPtr32Z(x, y, bp, bw), dst, dstpitch, TEXA); + ReadAndExpandBlock24(src, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } else { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMZ24) { - ReadAndExpandBlock24(BlockPtr32Z(x, y, bp, bw), dst, dstpitch, TEXA); + ReadAndExpandBlock24(src, dst, dstpitch, TEXA); } FOREACH_BLOCK_END } @@ -1576,9 +1579,9 @@ void GSLocalMemory::ReadTexture16Z(const GSVector4i& r, uint8* dst, int dstpitch { __declspec(align(16)) uint16 block[16 * 8]; - FOREACH_BLOCK_START(16, 8, 32) + FOREACH_BLOCK_START(16, 8, 32, PSM_PSMZ16) { - ReadBlock16(BlockPtr16Z(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock16(src, (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1589,9 +1592,9 @@ void GSLocalMemory::ReadTexture16SZ(const GSVector4i& r, uint8* dst, int dstpitc { __declspec(align(16)) uint16 block[16 * 8]; - FOREACH_BLOCK_START(16, 8, 32) + FOREACH_BLOCK_START(16, 8, 32, PSM_PSMZ16S) { - ReadBlock16(BlockPtr16SZ(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock16(src, (uint8*)block, sizeof(block) / 8); ExpandBlock16(block, dst, dstpitch, TEXA); } @@ -1736,18 +1739,18 @@ void GSLocalMemory::ReadTexture(const GSVector4i& r, uint8* dst, int dstpitch, c void GSLocalMemory::ReadTexture16NP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(16, 8, 16) + FOREACH_BLOCK_START(16, 8, 16, PSM_PSMCT16) { - ReadBlock16(BlockPtr16(x, y, bp, bw), dst, dstpitch); + ReadBlock16(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture16SNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(16, 8, 16) + FOREACH_BLOCK_START(16, 8, 16, PSM_PSMCT16S) { - ReadBlock16(BlockPtr16S(x, y, bp, bw), dst, dstpitch); + ReadBlock16(src, dst, dstpitch); } FOREACH_BLOCK_END } @@ -1758,9 +1761,9 @@ void GSLocalMemory::ReadTexture8NP(const GSVector4i& r, uint8* dst, int dstpitch if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { - FOREACH_BLOCK_START(16, 16, 32) + FOREACH_BLOCK_START(16, 16, 32, PSM_PSMT8) { - ReadAndExpandBlock8_32(BlockPtr8(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock8_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1770,9 +1773,9 @@ void GSLocalMemory::ReadTexture8NP(const GSVector4i& r, uint8* dst, int dstpitch __declspec(align(16)) uint8 block[16 * 16]; - FOREACH_BLOCK_START(16, 16, 16) + FOREACH_BLOCK_START(16, 16, 16, PSM_PSMT8) { - ReadBlock8(BlockPtr8(x, y, bp, bw), (uint8*)block, sizeof(block) / 16); + ReadBlock8(src, (uint8*)block, sizeof(block) / 16); ExpandBlock8_16(block, dst, dstpitch, pal); } @@ -1786,9 +1789,9 @@ void GSLocalMemory::ReadTexture4NP(const GSVector4i& r, uint8* dst, int dstpitch if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { - FOREACH_BLOCK_START(32, 16, 32) + FOREACH_BLOCK_START(32, 16, 32, PSM_PSMT4) { - ReadAndExpandBlock4_32(BlockPtr4(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1798,9 +1801,9 @@ void GSLocalMemory::ReadTexture4NP(const GSVector4i& r, uint8* dst, int dstpitch __declspec(align(16)) uint8 block[(32 / 2) * 16]; - FOREACH_BLOCK_START(32, 16, 16) + FOREACH_BLOCK_START(32, 16, 16, PSM_PSMT4) { - ReadBlock4(BlockPtr4(x, y, bp, bw), (uint8*)block, sizeof(block) / 16); + ReadBlock4(src, (uint8*)block, sizeof(block) / 16); ExpandBlock4_16(block, dst, dstpitch, pal); } @@ -1814,9 +1817,9 @@ void GSLocalMemory::ReadTexture8HNP(const GSVector4i& r, uint8* dst, int dstpitc if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT8H) { - ReadAndExpandBlock8H_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock8H_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1826,9 +1829,9 @@ void GSLocalMemory::ReadTexture8HNP(const GSVector4i& r, uint8* dst, int dstpitc __declspec(align(16)) uint32 block[8 * 8]; - FOREACH_BLOCK_START(8, 8, 16) + FOREACH_BLOCK_START(8, 8, 16, PSM_PSMT8H) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); ExpandBlock8H_16(block, dst, dstpitch, pal); } @@ -1842,9 +1845,9 @@ void GSLocalMemory::ReadTexture4HLNP(const GSVector4i& r, uint8* dst, int dstpit if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT4HL) { - ReadAndExpandBlock4HL_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4HL_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1854,9 +1857,9 @@ void GSLocalMemory::ReadTexture4HLNP(const GSVector4i& r, uint8* dst, int dstpit __declspec(align(16)) uint32 block[8 * 8]; - FOREACH_BLOCK_START(8, 8, 16) + FOREACH_BLOCK_START(8, 8, 16, PSM_PSMT4HL) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); ExpandBlock4HL_16(block, dst, dstpitch, pal); } @@ -1870,9 +1873,9 @@ void GSLocalMemory::ReadTexture4HHNP(const GSVector4i& r, uint8* dst, int dstpit if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) { - FOREACH_BLOCK_START(8, 8, 32) + FOREACH_BLOCK_START(8, 8, 32, PSM_PSMT4HH) { - ReadAndExpandBlock4HH_32(BlockPtr32(x, y, bp, bw), dst, dstpitch, pal); + ReadAndExpandBlock4HH_32(src, dst, dstpitch, pal); } FOREACH_BLOCK_END } @@ -1882,9 +1885,9 @@ void GSLocalMemory::ReadTexture4HHNP(const GSVector4i& r, uint8* dst, int dstpit __declspec(align(16)) uint32 block[8 * 8]; - FOREACH_BLOCK_START(8, 8, 16) + FOREACH_BLOCK_START(8, 8, 16, PSM_PSMT4HH) { - ReadBlock32(BlockPtr32(x, y, bp, bw), (uint8*)block, sizeof(block) / 8); + ReadBlock32(src, (uint8*)block, sizeof(block) / 8); ExpandBlock4HH_16(block, dst, dstpitch, pal); } @@ -1894,18 +1897,18 @@ void GSLocalMemory::ReadTexture4HHNP(const GSVector4i& r, uint8* dst, int dstpit void GSLocalMemory::ReadTexture16ZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(16, 8, 16) + FOREACH_BLOCK_START(16, 8, 16, PSM_PSMZ16) { - ReadBlock16(BlockPtr16Z(x, y, bp, bw), dst, dstpitch); + ReadBlock16(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture16SZNP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(16, 8, 16) + FOREACH_BLOCK_START(16, 8, 16, PSM_PSMZ16S) { - ReadBlock16(BlockPtr16SZ(x, y, bp, bw), dst, dstpitch); + ReadBlock16(src, dst, dstpitch); } FOREACH_BLOCK_END } @@ -1958,45 +1961,45 @@ void GSLocalMemory::ReadTextureNP(const GSVector4i& r, uint8* dst, int dstpitch, void GSLocalMemory::ReadTexture8P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(16, 16, 8) + FOREACH_BLOCK_START(16, 16, 8, PSM_PSMT8) { - ReadBlock8(BlockPtr8(x, y, bp, bw), dst, dstpitch); + ReadBlock8(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture4P(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(32, 16, 8) + FOREACH_BLOCK_START(32, 16, 8, PSM_PSMT4) { - ReadBlock4P(BlockPtr4(x, y, bp, bw), dst, dstpitch); + ReadBlock4P(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture8HP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(8, 8, 8) + FOREACH_BLOCK_START(8, 8, 8, PSM_PSMT8H) { - ReadBlock8HP(BlockPtr32(x, y, bp, bw), dst, dstpitch); + ReadBlock8HP(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture4HLP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(8, 8, 8) + FOREACH_BLOCK_START(8, 8, 8, PSM_PSMT4HL) { - ReadBlock4HLP(BlockPtr32(x, y, bp, bw), dst, dstpitch); + ReadBlock4HLP(src, dst, dstpitch); } FOREACH_BLOCK_END } void GSLocalMemory::ReadTexture4HHP(const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) const { - FOREACH_BLOCK_START(8, 8, 8) + FOREACH_BLOCK_START(8, 8, 8, PSM_PSMT4HH) { - ReadBlock4HHP(BlockPtr32(x, y, bp, bw), dst, dstpitch); + ReadBlock4HHP(src, dst, dstpitch); } FOREACH_BLOCK_END } diff --git a/plugins/GSdx/GSRenderer.cpp b/plugins/GSdx/GSRenderer.cpp index 046483fba3..b83f9cdcf9 100644 --- a/plugins/GSdx/GSRenderer.cpp +++ b/plugins/GSdx/GSRenderer.cpp @@ -426,7 +426,7 @@ void GSRenderer::KeyEvent(GSKeyEventData* e) } } -void GSRenderer::GetTextureMinMax(GSVector4i& r) +void GSRenderer::GetTextureMinMax(GSVector4i& r, bool linear) { const GSDrawingContext* context = m_context; @@ -488,7 +488,7 @@ void GSRenderer::GetTextureMinMax(GSVector4i& r) { GSVector4 st = m_vt.m_min.t.xyxy(m_vt.m_max.t); - if(context->TEX1.IsLinear()) + if(linear) { st += GSVector4(-0x8000, 0x8000).xxyy(); } @@ -697,6 +697,21 @@ bool GSRenderer::TryAlphaTest(uint32& fm, uint32& zm) return true; } +bool GSRenderer::IsLinear() +{ + float qmin = m_vt.m_min.t.z; + float qmax = m_vt.m_max.t.z; + + if(PRIM->FST) + { + // assume Q = 1.0f => LOD > 0 (should not, but Q is very often bogus, 0 or DEN) + + qmin = qmax = 1.0f; + } + + return m_context->TEX1.IsLinear(qmin, qmax); +} + bool GSRenderer::IsOpaque() { if(PRIM->AA1) diff --git a/plugins/GSdx/GSRenderer.h b/plugins/GSdx/GSRenderer.h index 21f1651d4b..1817ce0786 100644 --- a/plugins/GSdx/GSRenderer.h +++ b/plugins/GSdx/GSRenderer.h @@ -53,9 +53,10 @@ protected: // following functions need m_vt to be initialized - void GetTextureMinMax(GSVector4i& r); + void GetTextureMinMax(GSVector4i& r, bool linear); void GetAlphaMinMax(); bool TryAlphaTest(uint32& fm, uint32& zm); + bool IsLinear(); bool IsOpaque(); public: @@ -127,6 +128,8 @@ protected: // FIXME: berserk fpsm = 27 (8H) Draw(); + + m_perfmon.Put(GSPerfMon::Draw, 1); } m_count = 0; @@ -222,9 +225,9 @@ protected: public: GSRendererT(uint8* base, bool mt, void (*irq)(), GSDevice* dev) : GSRenderer(base, mt, irq, dev) + , m_vertices(NULL) , m_count(0) , m_maxcount(0) - , m_vertices(NULL) { } diff --git a/plugins/GSdx/GSRendererDX.cpp b/plugins/GSdx/GSRendererDX.cpp new file mode 100644 index 0000000000..77f6bfd198 --- /dev/null +++ b/plugins/GSdx/GSRendererDX.cpp @@ -0,0 +1,23 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererDX.h" \ No newline at end of file diff --git a/plugins/GSdx/GSRendererDX.h b/plugins/GSdx/GSRendererDX.h new file mode 100644 index 0000000000..14a6fde4e0 --- /dev/null +++ b/plugins/GSdx/GSRendererDX.h @@ -0,0 +1,338 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#pragma once + +#include "GSRendererHW.h" +#include "GSTextureFX.h" + +template +class GSRendererDX : public GSRendererHW +{ + GSTextureFX* m_tfx; + bool m_logz; + bool m_fba; + +protected: + int m_topology; + GSVector2 m_pixelcenter; + + virtual void SetupDATE(GSTexture* rt, GSTexture* ds) {} + virtual void UpdateFBA(GSTexture* rt) {} + +public: + GSRendererDX(uint8* base, bool mt, void (*irq)(), GSDevice* dev, GSTextureCache* tc, GSTextureFX* tfx) + : GSRendererHW(base, mt, irq, dev, tc) + , m_tfx(tfx) + , m_topology(-1) + , m_pixelcenter(0, 0) + { + m_logz = !!theApp.GetConfig("logz", 0); + m_fba = !!theApp.GetConfig("fba", 1); + } + + virtual ~GSRendererDX() + { + delete m_tfx; + } + + bool Create(const string& title) + { + if(!__super::Create(title)) + return false; + + if(!m_tfx->Create(m_dev)) + return false; + + return true; + } + + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) + { + GSDrawingEnvironment& env = m_env; + GSDrawingContext* context = m_context; + + // + + SetupDATE(rt, ds); + + // + + m_dev->BeginScene(); + + // om + + GSTextureFX::OMDepthStencilSelector om_dssel; + + om_dssel.zte = context->TEST.ZTE; + om_dssel.ztst = context->TEST.ZTST; + om_dssel.zwe = !context->ZBUF.ZMSK; + om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; + om_dssel.fba = m_fba ? context->FBA.FBA : 0; + + GSTextureFX::OMBlendSelector om_bsel; + + om_bsel.abe = !IsOpaque(); + om_bsel.a = context->ALPHA.A; + om_bsel.b = context->ALPHA.B; + om_bsel.c = context->ALPHA.C; + om_bsel.d = context->ALPHA.D; + om_bsel.wr = (context->FRAME.FBMSK & 0x000000ff) != 0x000000ff; + om_bsel.wg = (context->FRAME.FBMSK & 0x0000ff00) != 0x0000ff00; + om_bsel.wb = (context->FRAME.FBMSK & 0x00ff0000) != 0x00ff0000; + om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; + + // vs + + GSTextureFX::VSSelector vs_sel; + + vs_sel.bppz = 0; + vs_sel.tme = PRIM->TME; + vs_sel.fst = PRIM->FST; + vs_sel.logz = m_logz ? 1 : 0; + vs_sel.prim = primclass; + + if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) + { + if(context->ZBUF.PSM == PSM_PSMZ24) + { + if(m_vt.m_max.p.z > 0xffffff) + { + ASSERT(m_vt.m_min.p.z > 0xffffff); + + vs_sel.bppz = 1; + om_dssel.ztst = 1; + } + } + else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) + { + if(m_vt.m_max.p.z > 0xffff) + { + ASSERT(m_vt.m_min.p.z > 0xffff); // sfex capcom logo + + vs_sel.bppz = 2; + om_dssel.ztst = 1; + } + } + } + + GSTextureFX::VSConstantBuffer vs_cb; + + float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); + float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); + float ox = (float)(int)context->XYOFFSET.OFX; + float oy = (float)(int)context->XYOFFSET.OFY; + float ox2 = 2.0f * m_pixelcenter.x / rt->GetWidth(); + float oy2 = 2.0f * m_pixelcenter.y / rt->GetHeight(); + + vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); + vs_cb.VertexOffset = GSVector4(ox * sx + ox2 + 1, -(oy * sy + oy2 + 1), 0.0f, -1.0f); + + if(PRIM->TME) + { + if(PRIM->FST) + { + vs_cb.TextureScale.x = 1.0f / (16 << context->TEX0.TW); + vs_cb.TextureScale.y = 1.0f / (16 << context->TEX0.TH); + } + else + { + vs_cb.TextureScale = GSVector2(1.0f, 1.0f); + } + } + + // gs + + GSTextureFX::GSSelector gs_sel; + + gs_sel.iip = PRIM->IIP; + gs_sel.prim = primclass; + + // ps + + GSTextureFX::PSSelector ps_sel; + + ps_sel.fst = PRIM->FST; + ps_sel.wms = context->CLAMP.WMS; + ps_sel.wmt = context->CLAMP.WMT; + ps_sel.bpp = 0; + ps_sel.aem = env.TEXA.AEM; + ps_sel.tfx = context->TEX0.TFX; + ps_sel.tcc = context->TEX0.TCC; + ps_sel.ate = context->TEST.ATE; + ps_sel.atst = context->TEST.ATST; + ps_sel.fog = PRIM->FGE; + ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; + ps_sel.fba = context->FBA.FBA; + ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; + ps_sel.ltf = m_filter == 2 ? IsLinear() : m_filter; + + GSTextureFX::PSSamplerSelector ps_ssel; + + ps_ssel.tau = 0; + ps_ssel.tav = 0; + ps_ssel.ltf = ps_sel.ltf; + + GSTextureFX::PSConstantBuffer ps_cb; + + ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; + + if(ps_sel.atst == 2 || ps_sel.atst == 5) + { + ps_cb.FogColor_AREF.a -= 0.9f / 255; + } + else if(ps_sel.atst == 3 || ps_sel.atst == 6) + { + ps_cb.FogColor_AREF.a += 0.9f / 255; + } + + if(tex) + { + ps_sel.bpp = tex->m_bpp; + + int w = tex->m_texture->GetWidth(); + int h = tex->m_texture->GetHeight(); + + ps_cb.WH = GSVector4((int)(1 << context->TEX0.TW), (int)(1 << context->TEX0.TH), w, h); + ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); + ps_cb.MinF_TA.z = (float)(int)env.TEXA.TA0 / 255; + ps_cb.MinF_TA.w = (float)(int)env.TEXA.TA1 / 255; + + switch(context->CLAMP.WMS) + { + case 0: + ps_ssel.tau = 1; + break; + case 1: + ps_ssel.tau = 0; + break; + case 2: + ps_cb.MinMax.x = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); + ps_cb.MinMax.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); + ps_cb.MinF_TA.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); + ps_ssel.tau = 0; + break; + case 3: + ps_cb.MskFix.x = context->CLAMP.MINU; + ps_cb.MskFix.z = context->CLAMP.MAXU; + ps_ssel.tau = 1; + break; + default: + __assume(0); + } + + switch(context->CLAMP.WMT) + { + case 0: + ps_ssel.tav = 1; + break; + case 1: + ps_ssel.tav = 0; + break; + case 2: + ps_cb.MinMax.y = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); + ps_cb.MinMax.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); + ps_cb.MinF_TA.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); + ps_ssel.tav = 0; + break; + case 3: + ps_cb.MskFix.y = context->CLAMP.MINV; + ps_cb.MskFix.w = context->CLAMP.MAXV; + ps_ssel.tav = 1; + break; + default: + __assume(0); + } + } + else + { + ps_sel.tfx = 4; + } + + // rs + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); + + // + + uint8 afix = context->ALPHA.FIX; + + m_tfx->SetupOM(om_dssel, om_bsel, afix, rt, ds); + m_tfx->SetupIA(m_vertices, m_count, m_topology); + m_tfx->SetupVS(vs_sel, &vs_cb); + m_tfx->SetupGS(gs_sel); + m_tfx->SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); + m_tfx->SetupRS(w, h, scissor); + + // draw + + if(context->TEST.DoFirstPass()) + { + m_dev->DrawPrimitive(); + } + + if(context->TEST.DoSecondPass()) + { + ASSERT(!env.PABE.PABE); + + static const uint32 iatst[] = {1, 0, 5, 6, 7, 2, 3, 4}; + + ps_sel.atst = iatst[ps_sel.atst]; + + m_tfx->UpdatePS(ps_sel, &ps_cb, ps_ssel); + + bool z = om_dssel.zwe; + bool r = om_bsel.wr; + bool g = om_bsel.wg; + bool b = om_bsel.wb; + bool a = om_bsel.wa; + + switch(context->TEST.AFAIL) + { + case 0: z = r = g = b = a = false; break; // none + case 1: z = false; break; // rgba + case 2: r = g = b = a = false; break; // z + case 3: z = a = false; break; // rgb + default: __assume(0); + } + + if(z || r || g || b || a) + { + om_dssel.zwe = z; + om_bsel.wr = r; + om_bsel.wg = g; + om_bsel.wb = b; + om_bsel.wa = a; + + m_tfx->UpdateOM(om_dssel, om_bsel, afix); + + m_dev->DrawPrimitive(); + } + } + + m_dev->EndScene(); + + if(om_dssel.fba) UpdateFBA(rt); + } +}; \ No newline at end of file diff --git a/plugins/GSdx/GSRendererDX10.cpp b/plugins/GSdx/GSRendererDX10.cpp new file mode 100644 index 0000000000..0bc018f78b --- /dev/null +++ b/plugins/GSdx/GSRendererDX10.cpp @@ -0,0 +1,266 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererDX10.h" +#include "GSCrc.h" +#include "resource.h" + +GSRendererDX10::GSRendererDX10(uint8* base, bool mt, void (*irq)()) + : GSRendererDX(base, mt, irq, new GSDevice10(), new GSTextureCache10(this), new GSTextureFX10()) +{ + InitVertexKick(); + + m_pixelcenter = GSVector2(-0.5f, -0.5f); +} + +bool GSRendererDX10::Create(const string& title) +{ + if(!__super::Create(title)) + return false; + + // + + D3D10_DEPTH_STENCIL_DESC dsd; + + memset(&dsd, 0, sizeof(dsd)); + + dsd.DepthEnable = false; + dsd.StencilEnable = true; + dsd.StencilReadMask = 1; + dsd.StencilWriteMask = 1; + dsd.FrontFace.StencilFunc = D3D10_COMPARISON_ALWAYS; + dsd.FrontFace.StencilPassOp = D3D10_STENCIL_OP_REPLACE; + dsd.FrontFace.StencilFailOp = D3D10_STENCIL_OP_KEEP; + dsd.FrontFace.StencilDepthFailOp = D3D10_STENCIL_OP_KEEP; + dsd.BackFace.StencilFunc = D3D10_COMPARISON_ALWAYS; + dsd.BackFace.StencilPassOp = D3D10_STENCIL_OP_REPLACE; + dsd.BackFace.StencilFailOp = D3D10_STENCIL_OP_KEEP; + dsd.BackFace.StencilDepthFailOp = D3D10_STENCIL_OP_KEEP; + + (*(GSDevice10*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); + + D3D10_BLEND_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + (*(GSDevice10*)m_dev)->CreateBlendState(&bd, &m_date.bs); + + // + + return true; +} + +template +void GSRendererDX10::VertexKick(bool skip) +{ + GSVertexHW10& dst = m_vl.AddTail(); + + dst.vi[0] = m_v.vi[0]; + dst.vi[1] = m_v.vi[1]; + + if(tme && fst) + { + GSVector4::storel(&dst.ST, m_v.GetUV()); + } + + int count = 0; + + if(GSVertexHW10* v = DrawingKick(skip, count)) + { + GSVector4i scissor = m_context->scissor.dx10; + + #if _M_SSE >= 0x401 + + GSVector4i pmin, pmax, v0, v1, v2; + + switch(prim) + { + case GS_POINTLIST: + v0 = GSVector4i::load((int)v[0].p.xy).upl16(); + pmin = v0; + pmax = v0; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + pmin = v0.min_u16(v1).upl16(); + pmax = v0.max_u16(v1).upl16(); + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + v2 = GSVector4i::load((int)v[2].p.xy); + pmin = v0.min_u16(v1).min_u16(v2).upl16(); + pmax = v0.max_u16(v1).max_u16(v2).upl16(); + break; + } + + GSVector4i test = (pmax < scissor) | (pmin > scissor.zwxy()); + + if(test.mask() & 0xff) + { + return; + } + + #else + + switch(prim) + { + case GS_POINTLIST: + if(v[0].p.x < scissor.x + || v[0].p.x > scissor.z + || v[0].p.y < scissor.y + || v[0].p.y > scissor.w) + { + return; + } + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w) + { + return; + } + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x && v[2].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z && v[2].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y && v[2].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w && v[2].p.y > scissor.w) + { + return; + } + break; + } + + #endif + + m_count += count; + } +} + +void GSRendererDX10::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) +{ + switch(primclass) + { + case GS_POINT_CLASS: + m_topology = D3D10_PRIMITIVE_TOPOLOGY_POINTLIST; + m_perfmon.Put(GSPerfMon::Prim, m_count); + break; + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: + m_topology = D3D10_PRIMITIVE_TOPOLOGY_LINELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 2); + break; + case GS_TRIANGLE_CLASS: + m_topology = D3D10_PRIMITIVE_TOPOLOGY_TRIANGLELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 3); + break; + default: + __assume(0); + } + + __super::Draw(primclass, rt, ds, tex); +} + +void GSRendererDX10::SetupDATE(GSTexture* rt, GSTexture* ds) +{ + if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) + + GSDevice10* dev = (GSDevice10*)m_dev; + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + if(GSTexture* t = dev->CreateRenderTarget(w, h)) + { + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + + dev->BeginScene(); + + dev->ClearStencil(ds, 0); + + // om + + dev->OMSetDepthStencilState(m_date.dss, 1); + dev->OMSetBlendState(m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + + GSVertexPT1 vertices[] = + { + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, + }; + + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL); + + // gs + + dev->GSSetShader(NULL); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); + dev->PSSetSamplerState(dev->m_convert.pt, NULL); + + // rs + + dev->RSSet(w, h); + + // set + + dev->DrawPrimitive(); + + // + + dev->EndScene(); + + dev->Recycle(t); + } +} diff --git a/plugins/GSdx/GSRendererHW10.h b/plugins/GSdx/GSRendererDX10.h similarity index 86% rename from plugins/GSdx/GSRendererHW10.h rename to plugins/GSdx/GSRendererDX10.h index 0568758624..1465fff56e 100644 --- a/plugins/GSdx/GSRendererHW10.h +++ b/plugins/GSdx/GSRendererDX10.h @@ -21,28 +21,25 @@ #pragma once -#include "GSRendererHW.h" +#include "GSRendererDX.h" #include "GSVertexHW.h" #include "GSTextureCache10.h" #include "GSTextureFX10.h" -class GSRendererHW10 : public GSRendererHW +class GSRendererDX10 : public GSRendererDX { protected: - GSTextureFX10 m_tfx; - - void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); - struct { CComPtr dss; CComPtr bs; } m_date; + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex); void SetupDATE(GSTexture* rt, GSTexture* ds); public: - GSRendererHW10(uint8* base, bool mt, void (*irq)()); + GSRendererDX10(uint8* base, bool mt, void (*irq)()); bool Create(const string& title); diff --git a/plugins/GSdx/GSRendererDX11.cpp b/plugins/GSdx/GSRendererDX11.cpp new file mode 100644 index 0000000000..063c513a8e --- /dev/null +++ b/plugins/GSdx/GSRendererDX11.cpp @@ -0,0 +1,266 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererDX11.h" +#include "GSCrc.h" +#include "resource.h" + +GSRendererDX11::GSRendererDX11(uint8* base, bool mt, void (*irq)()) + : GSRendererDX(base, mt, irq, new GSDevice11(), new GSTextureCache11(this), new GSTextureFX11()) +{ + InitVertexKick(); + + m_pixelcenter = GSVector2(-0.5f, -0.5f); +} + +bool GSRendererDX11::Create(const string& title) +{ + if(!__super::Create(title)) + return false; + + // + + D3D11_DEPTH_STENCIL_DESC dsd; + + memset(&dsd, 0, sizeof(dsd)); + + dsd.DepthEnable = false; + dsd.StencilEnable = true; + dsd.StencilReadMask = 1; + dsd.StencilWriteMask = 1; + dsd.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS; + dsd.FrontFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; + dsd.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilFunc = D3D11_COMPARISON_ALWAYS; + dsd.BackFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; + dsd.BackFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; + dsd.BackFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; + + (*(GSDevice11*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); + + D3D11_BLEND_DESC bd; + + memset(&bd, 0, sizeof(bd)); + + (*(GSDevice11*)m_dev)->CreateBlendState(&bd, &m_date.bs); + + // + + return true; +} + +template +void GSRendererDX11::VertexKick(bool skip) +{ + GSVertexHW11& dst = m_vl.AddTail(); + + dst.vi[0] = m_v.vi[0]; + dst.vi[1] = m_v.vi[1]; + + if(tme && fst) + { + GSVector4::storel(&dst.ST, m_v.GetUV()); + } + + int count = 0; + + if(GSVertexHW11* v = DrawingKick(skip, count)) + { + GSVector4i scissor = m_context->scissor.dx10; + + #if _M_SSE >= 0x401 + + GSVector4i pmin, pmax, v0, v1, v2; + + switch(prim) + { + case GS_POINTLIST: + v0 = GSVector4i::load((int)v[0].p.xy).upl16(); + pmin = v0; + pmax = v0; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + pmin = v0.min_u16(v1).upl16(); + pmax = v0.max_u16(v1).upl16(); + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + v0 = GSVector4i::load((int)v[0].p.xy); + v1 = GSVector4i::load((int)v[1].p.xy); + v2 = GSVector4i::load((int)v[2].p.xy); + pmin = v0.min_u16(v1).min_u16(v2).upl16(); + pmax = v0.max_u16(v1).max_u16(v2).upl16(); + break; + } + + GSVector4i test = (pmax < scissor) | (pmin > scissor.zwxy()); + + if(test.mask() & 0xff) + { + return; + } + + #else + + switch(prim) + { + case GS_POINTLIST: + if(v[0].p.x < scissor.x + || v[0].p.x > scissor.z + || v[0].p.y < scissor.y + || v[0].p.y > scissor.w) + { + return; + } + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w) + { + return; + } + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + if(v[0].p.x < scissor.x && v[1].p.x < scissor.x && v[2].p.x < scissor.x + || v[0].p.x > scissor.z && v[1].p.x > scissor.z && v[2].p.x > scissor.z + || v[0].p.y < scissor.y && v[1].p.y < scissor.y && v[2].p.y < scissor.y + || v[0].p.y > scissor.w && v[1].p.y > scissor.w && v[2].p.y > scissor.w) + { + return; + } + break; + } + + #endif + + m_count += count; + } +} + +void GSRendererDX11::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) +{ + switch(primclass) + { + case GS_POINT_CLASS: + m_topology = D3D11_PRIMITIVE_TOPOLOGY_POINTLIST; + m_perfmon.Put(GSPerfMon::Prim, m_count); + break; + case GS_LINE_CLASS: + case GS_SPRITE_CLASS: + m_topology = D3D11_PRIMITIVE_TOPOLOGY_LINELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 2); + break; + case GS_TRIANGLE_CLASS: + m_topology = D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 3); + break; + default: + __assume(0); + } + + __super::Draw(primclass, rt, ds, tex); +} + +void GSRendererDX11::SetupDATE(GSTexture* rt, GSTexture* ds) +{ + if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) + + GSDevice11* dev = (GSDevice11*)m_dev; + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + if(GSTexture* t = dev->CreateRenderTarget(w, h)) + { + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + + dev->BeginScene(); + + dev->ClearStencil(ds, 0); + + // om + + dev->OMSetDepthStencilState(m_date.dss, 1); + dev->OMSetBlendState(m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + + GSVertexPT1 vertices[] = + { + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, + }; + + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL); + + // gs + + dev->GSSetShader(NULL); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); + dev->PSSetSamplerState(dev->m_convert.pt, NULL); + + // rs + + dev->RSSet(w, h); + + // set + + dev->DrawPrimitive(); + + // + + dev->EndScene(); + + dev->Recycle(t); + } +} diff --git a/plugins/GSdx/GSRendererHW11.h b/plugins/GSdx/GSRendererDX11.h similarity index 83% rename from plugins/GSdx/GSRendererHW11.h rename to plugins/GSdx/GSRendererDX11.h index fab4c76911..e30f2e7c07 100644 --- a/plugins/GSdx/GSRendererHW11.h +++ b/plugins/GSdx/GSRendererDX11.h @@ -21,17 +21,15 @@ #pragma once -#include "GSRendererHW.h" +#include "GSRendererDX.h" #include "GSVertexHW.h" #include "GSTextureCache11.h" #include "GSTextureFX11.h" -class GSRendererHW11 : public GSRendererHW +class GSRendererDX11 : public GSRendererDX { protected: - GSTextureFX11 m_tfx; - - void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex); struct { @@ -42,7 +40,7 @@ protected: void SetupDATE(GSTexture* rt, GSTexture* ds); public: - GSRendererHW11(uint8* base, bool mt, void (*irq)()); + GSRendererDX11(uint8* base, bool mt, void (*irq)()); bool Create(const string& title); diff --git a/plugins/GSdx/GSRendererDX9.cpp b/plugins/GSdx/GSRendererDX9.cpp new file mode 100644 index 0000000000..c6e4be5a4f --- /dev/null +++ b/plugins/GSdx/GSRendererDX9.cpp @@ -0,0 +1,322 @@ +/* + * Copyright (C) 2007-2009 Gabest + * http://www.gabest.org + * + * This Program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This Program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * http://www.gnu.org/copyleft/gpl.html + * + */ + +#include "stdafx.h" +#include "GSRendererDX9.h" +#include "GSCrc.h" +#include "resource.h" + +GSRendererDX9::GSRendererDX9(uint8* base, bool mt, void (*irq)()) + : GSRendererDX(base, mt, irq, new GSDevice9(), new GSTextureCache9(this), new GSTextureFX9()) +{ + InitVertexKick(); +} + +bool GSRendererDX9::Create(const string& title) +{ + if(!__super::Create(title)) + return false; + + // + + memset(&m_date.dss, 0, sizeof(m_date.dss)); + + m_date.dss.StencilEnable = true; + m_date.dss.StencilReadMask = 1; + m_date.dss.StencilWriteMask = 1; + m_date.dss.StencilFunc = D3DCMP_ALWAYS; + m_date.dss.StencilPassOp = D3DSTENCILOP_REPLACE; + m_date.dss.StencilRef = 1; + + memset(&m_date.bs, 0, sizeof(m_date.bs)); + + // + + memset(&m_fba.dss, 0, sizeof(m_fba.dss)); + + m_fba.dss.StencilEnable = true; + m_fba.dss.StencilReadMask = 2; + m_fba.dss.StencilWriteMask = 2; + m_fba.dss.StencilFunc = D3DCMP_EQUAL; + m_fba.dss.StencilPassOp = D3DSTENCILOP_ZERO; + m_fba.dss.StencilFailOp = D3DSTENCILOP_ZERO; + m_fba.dss.StencilDepthFailOp = D3DSTENCILOP_ZERO; + m_fba.dss.StencilRef = 2; + + memset(&m_fba.bs, 0, sizeof(m_fba.bs)); + + m_fba.bs.RenderTargetWriteMask = D3DCOLORWRITEENABLE_ALPHA; + + // + + return true; +} + +template +void GSRendererDX9::VertexKick(bool skip) +{ + GSVertexHW9 v; + + v.p = GSVector4(((GSVector4i)m_v.XYZ).upl16()); + + if(tme && !fst) + { + v.p = v.p.xyxy(GSVector4((float)m_v.XYZ.Z, m_v.RGBAQ.Q)); + } + else + { + v.p = v.p.xyxy(GSVector4::load((float)m_v.XYZ.Z)); + } + + if(tme) + { + if(fst) + { + v.t = m_v.GetUV(); + } + else + { + v.t = GSVector4::loadl(&m_v.ST); + } + } + + GSVertexHW9& dst = m_vl.AddTail(); + + dst = v; + + dst.c0 = m_v.RGBAQ.u32[0]; + dst.c1 = m_v.FOG.u32[1]; + + int count = 0; + + if(GSVertexHW9* v = DrawingKick(skip, count)) + { + GSVector4 scissor = m_context->scissor.dx9; + + GSVector4 pmin, pmax; + + switch(prim) + { + case GS_POINTLIST: + pmin = v[0].p; + pmax = v[0].p; + break; + case GS_LINELIST: + case GS_LINESTRIP: + case GS_SPRITE: + pmin = v[0].p.minv(v[1].p); + pmax = v[0].p.maxv(v[1].p); + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + pmin = v[0].p.minv(v[1].p).minv(v[2].p); + pmax = v[0].p.maxv(v[1].p).maxv(v[2].p); + break; + } + + GSVector4 test = (pmax < scissor) | (pmin > scissor.zwxy()); + + if(test.mask() & 3) + { + return; + } + + switch(prim) + { + case GS_POINTLIST: + break; + case GS_LINELIST: + case GS_LINESTRIP: + if(PRIM->IIP == 0) {v[0].c0 = v[1].c0;} + break; + case GS_TRIANGLELIST: + case GS_TRIANGLESTRIP: + case GS_TRIANGLEFAN: + if(PRIM->IIP == 0) {v[0].c0 = v[1].c0 = v[2].c0;} + break; + case GS_SPRITE: + if(PRIM->IIP == 0) {v[0].c0 = v[1].c0;} + v[0].p.z = v[1].p.z; + v[0].p.w = v[1].p.w; + v[0].c1 = v[1].c1; + v[2] = v[1]; + v[3] = v[1]; + v[1].p.y = v[0].p.y; + v[1].t.y = v[0].t.y; + v[2].p.x = v[0].p.x; + v[2].t.x = v[0].t.x; + v[4] = v[1]; + v[5] = v[2]; + count += 4; + break; + } + + m_count += count; + } +} + +void GSRendererDX9::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) +{ + switch(primclass) + { + case GS_POINT_CLASS: + m_topology = D3DPT_POINTLIST; + m_perfmon.Put(GSPerfMon::Prim, m_count); + break; + case GS_LINE_CLASS: + m_topology = D3DPT_LINELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 2); + break; + case GS_TRIANGLE_CLASS: + case GS_SPRITE_CLASS: + m_topology = D3DPT_TRIANGLELIST; + m_perfmon.Put(GSPerfMon::Prim, m_count / 3); + break; + default: + __assume(0); + } + + (*(GSDevice9*)m_dev)->SetRenderState(D3DRS_SHADEMODE, PRIM->IIP ? D3DSHADE_GOURAUD : D3DSHADE_FLAT); // TODO + + __super::Draw(primclass, rt, ds, tex); +} + +void GSRendererDX9::SetupDATE(GSTexture* rt, GSTexture* ds) +{ + if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) + + GSDevice9* dev = (GSDevice9*)m_dev; + + int w = rt->GetWidth(); + int h = rt->GetHeight(); + + if(GSTexture* t = dev->CreateRenderTarget(w, h)) + { + // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows + + dev->BeginScene(); + + dev->ClearStencil(ds, 0); + + // om + + dev->OMSetDepthStencilState(&m_date.dss); + dev->OMSetBlendState(&m_date.bs, 0); + dev->OMSetRenderTargets(t, ds); + + // ia + + GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + + GSVertexPT1 vertices[] = + { + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, + }; + + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL, 0); + + // ps + + dev->PSSetShaderResources(rt, NULL); + dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL, 0); + dev->PSSetSamplerState(&dev->m_convert.pt); + + // rs + + dev->RSSet(w, h); + + // + + dev->DrawPrimitive(); + + // + + dev->EndScene(); + + dev->Recycle(t); + } +} + +void GSRendererDX9::UpdateFBA(GSTexture* rt) +{ + GSDevice9* dev = (GSDevice9*)m_dev; + + dev->BeginScene(); + + // om + + dev->OMSetDepthStencilState(&m_fba.dss); + dev->OMSetBlendState(&m_fba.bs, 0); + + // ia + + GSVector4 s = GSVector4(rt->m_scale.x / rt->GetWidth(), rt->m_scale.y / rt->GetHeight()); + GSVector4 o = GSVector4(-1.0f, 1.0f); + + GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); + GSVector4 dst = src * 2.0f + o.xxxx(); + + GSVertexPT1 vertices[] = + { + {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, + {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, + }; + + dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); + dev->IASetInputLayout(dev->m_convert.il); + dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); + + // vs + + dev->VSSetShader(dev->m_convert.vs, NULL, 0); + + // ps + + dev->PSSetShader(dev->m_convert.ps[4], NULL, 0); + + // rs + + dev->RSSet(rt->GetWidth(), rt->GetHeight()); + + // + + dev->DrawPrimitive(); + + // + + dev->EndScene(); +} diff --git a/plugins/GSdx/GSRendererHW9.h b/plugins/GSdx/GSRendererDX9.h similarity index 85% rename from plugins/GSdx/GSRendererHW9.h rename to plugins/GSdx/GSRendererDX9.h index 896fe892a5..1bebe4e490 100644 --- a/plugins/GSdx/GSRendererHW9.h +++ b/plugins/GSdx/GSRendererDX9.h @@ -21,19 +21,14 @@ #pragma once -#include "GSRendererHW.h" +#include "GSRendererDX.h" #include "GSVertexHW.h" #include "GSTextureCache9.h" #include "GSTextureFX9.h" -class GSRendererHW9 : public GSRendererHW +class GSRendererDX9 : public GSRendererDX { protected: - GSTextureFX9 m_tfx; - bool m_logz; - - void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); - struct { Direct3DDepthStencilState9 dss; @@ -42,16 +37,16 @@ protected: struct { - bool enabled; Direct3DDepthStencilState9 dss; Direct3DBlendState9 bs; } m_fba; + void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex); void SetupDATE(GSTexture* rt, GSTexture* ds); void UpdateFBA(GSTexture* rt); public: - GSRendererHW9(uint8* base, bool mt, void (*irq)()); + GSRendererDX9(uint8* base, bool mt, void (*irq)()); bool Create(const string& title); diff --git a/plugins/GSdx/GSRendererHW.h b/plugins/GSdx/GSRendererHW.h index 392ac4066e..536755108d 100644 --- a/plugins/GSdx/GSRendererHW.h +++ b/plugins/GSdx/GSRendererHW.h @@ -81,7 +81,7 @@ protected: GSTexture* t = NULL; - if(GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height, true)) + if(GSTextureCache::Target* rt = m_tc->LookupTarget(TEX0, m_width, m_height, GSTextureCache::RenderTarget, true, true)) { t = rt->m_texture; @@ -128,15 +128,15 @@ protected: TEX0.TBW = context->FRAME.FBW; TEX0.PSM = context->FRAME.PSM; - GSTextureCache::GSRenderTarget* rt = m_tc->GetRenderTarget(TEX0, m_width, m_height); + GSTextureCache::Target* rt = m_tc->LookupTarget(TEX0, m_width, m_height, GSTextureCache::RenderTarget, true); TEX0.TBP0 = context->ZBUF.Block(); TEX0.TBW = context->FRAME.FBW; TEX0.PSM = context->ZBUF.PSM; - GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height); + GSTextureCache::Target* ds = m_tc->LookupTarget(TEX0, m_width, m_height, GSTextureCache::DepthStencil, m_context->DepthWrite()); - GSTextureCache::GSCachedTexture* tex = NULL; + GSTextureCache::Source* tex = NULL; if(PRIM->TME) { @@ -144,9 +144,9 @@ protected: GSVector4i r; - GetTextureMinMax(r); + GetTextureMinMax(r, IsLinear()); - tex = m_tc->GetTexture(r); + tex = m_tc->LookupSource(context->TEX0, env.TEXA, r); if(!tex) return; } @@ -157,7 +157,7 @@ protected: string s; - if(s_save && PRIM->TME) + if(s_save && tex) { s = format("c:\\temp2\\_%05d_f%I64d_tex_%05x_%d_%d%d_%02x_%02x_%02x_%02x.dds", s_n, frame, (int)context->TEX0.TBP0, (int)context->TEX0.PSM, @@ -233,6 +233,30 @@ protected: // + GSVector4i r = GSVector4i(m_vt.m_min.p.xyxy(m_vt.m_max.p)).rintersect(GSVector4i(m_context->scissor.in)); + + GIFRegBITBLTBUF BITBLTBUF; + + BITBLTBUF.DBW = context->FRAME.FBW; + + if(fm != 0xffffffff) + { + BITBLTBUF.DBP = context->FRAME.Block(); + BITBLTBUF.DPSM = context->FRAME.PSM; + + m_tc->InvalidateVideoMem(BITBLTBUF, r, false); + } + + if(zm != 0xffffffff) + { + BITBLTBUF.DBP = context->ZBUF.Block(); + BITBLTBUF.DPSM = context->ZBUF.PSM; + + m_tc->InvalidateVideoMem(BITBLTBUF, r, false); + } + + // + OverrideOutput(); if(s_dump) @@ -257,13 +281,11 @@ protected: s_n++; } - - m_tc->InvalidateTextures(context->FRAME, context->ZBUF); } - virtual void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) = 0; + virtual void Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) = 0; - virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* t) + virtual bool OverrideInput(int& prim, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* t) { #pragma region ffxii pal video conversion @@ -424,7 +446,7 @@ protected: TEX0.TBW = FBW; TEX0.PSM = FPSM; - if(GSTextureCache::GSDepthStencil* ds = m_tc->GetDepthStencil(TEX0, m_width, m_height)) + if(GSTextureCache::Target* ds = m_tc->LookupTarget(TEX0, m_width, m_height, GSTextureCache::DepthStencil, true)) { m_dev->ClearDepth(ds->m_texture, 0); } diff --git a/plugins/GSdx/GSRendererHW10.cpp b/plugins/GSdx/GSRendererHW10.cpp deleted file mode 100644 index 35a2da0ac6..0000000000 --- a/plugins/GSdx/GSRendererHW10.cpp +++ /dev/null @@ -1,529 +0,0 @@ -/* - * Copyright (C) 2007-2009 Gabest - * http://www.gabest.org - * - * This Program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2, or (at your option) - * any later version. - * - * This Program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with GNU Make; see the file COPYING. If not, write to - * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. - * http://www.gnu.org/copyleft/gpl.html - * - */ - -#include "stdafx.h" -#include "GSRendererHW10.h" -#include "GSCrc.h" -#include "resource.h" - -GSRendererHW10::GSRendererHW10(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice10(), new GSTextureCache10(this)) -{ - InitVertexKick(); -} - -bool GSRendererHW10::Create(const string& title) -{ - if(!__super::Create(title)) - return false; - - if(!m_tfx.Create((GSDevice10*)m_dev)) - return false; - - // - - D3D10_DEPTH_STENCIL_DESC dsd; - - memset(&dsd, 0, sizeof(dsd)); - - dsd.DepthEnable = false; - dsd.StencilEnable = true; - dsd.StencilReadMask = 1; - dsd.StencilWriteMask = 1; - dsd.FrontFace.StencilFunc = D3D10_COMPARISON_ALWAYS; - dsd.FrontFace.StencilPassOp = D3D10_STENCIL_OP_REPLACE; - dsd.FrontFace.StencilFailOp = D3D10_STENCIL_OP_KEEP; - dsd.FrontFace.StencilDepthFailOp = D3D10_STENCIL_OP_KEEP; - dsd.BackFace.StencilFunc = D3D10_COMPARISON_ALWAYS; - dsd.BackFace.StencilPassOp = D3D10_STENCIL_OP_REPLACE; - dsd.BackFace.StencilFailOp = D3D10_STENCIL_OP_KEEP; - dsd.BackFace.StencilDepthFailOp = D3D10_STENCIL_OP_KEEP; - - (*(GSDevice10*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); - - D3D10_BLEND_DESC bd; - - memset(&bd, 0, sizeof(bd)); - - (*(GSDevice10*)m_dev)->CreateBlendState(&bd, &m_date.bs); - - // - - return true; -} - -template -void GSRendererHW10::VertexKick(bool skip) -{ - GSVertexHW10& dst = m_vl.AddTail(); - - dst.vi[0] = m_v.vi[0]; - dst.vi[1] = m_v.vi[1]; - - if(tme && fst) - { - GSVector4::storel(&dst.ST, m_v.GetUV()); - } - - int count = 0; - - if(GSVertexHW10* v = DrawingKick(skip, count)) - { - GSVector4i scissor = m_context->scissor.dx10; - - #if _M_SSE >= 0x401 - - GSVector4i pmin, pmax, v0, v1, v2; - - switch(prim) - { - case GS_POINTLIST: - v0 = GSVector4i::load((int)v[0].p.xy).upl16(); - pmin = v0; - pmax = v0; - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - v0 = GSVector4i::load((int)v[0].p.xy); - v1 = GSVector4i::load((int)v[1].p.xy); - pmin = v0.min_u16(v1).upl16(); - pmax = v0.max_u16(v1).upl16(); - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - v0 = GSVector4i::load((int)v[0].p.xy); - v1 = GSVector4i::load((int)v[1].p.xy); - v2 = GSVector4i::load((int)v[2].p.xy); - pmin = v0.min_u16(v1).min_u16(v2).upl16(); - pmax = v0.max_u16(v1).max_u16(v2).upl16(); - break; - } - - GSVector4i test = (pmax < scissor) | (pmin > scissor.zwxy()); - - if(test.mask() & 0xff) - { - return; - } - - #else - - switch(prim) - { - case GS_POINTLIST: - if(v[0].p.x < scissor.x - || v[0].p.x > scissor.z - || v[0].p.y < scissor.y - || v[0].p.y > scissor.w) - { - return; - } - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - if(v[0].p.x < scissor.x && v[1].p.x < scissor.x - || v[0].p.x > scissor.z && v[1].p.x > scissor.z - || v[0].p.y < scissor.y && v[1].p.y < scissor.y - || v[0].p.y > scissor.w && v[1].p.y > scissor.w) - { - return; - } - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - if(v[0].p.x < scissor.x && v[1].p.x < scissor.x && v[2].p.x < scissor.x - || v[0].p.x > scissor.z && v[1].p.x > scissor.z && v[2].p.x > scissor.z - || v[0].p.y < scissor.y && v[1].p.y < scissor.y && v[2].p.y < scissor.y - || v[0].p.y > scissor.w && v[1].p.y > scissor.w && v[2].p.y > scissor.w) - { - return; - } - break; - } - - #endif - - m_count += count; - } -} - -void GSRendererHW10::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) -{ - GSDrawingEnvironment& env = m_env; - GSDrawingContext* context = m_context; - - D3D10_PRIMITIVE_TOPOLOGY topology; - int prims = 0; - - switch(primclass) - { - case GS_POINT_CLASS: - topology = D3D10_PRIMITIVE_TOPOLOGY_POINTLIST; - prims = m_count; - break; - case GS_LINE_CLASS: - case GS_SPRITE_CLASS: - topology = D3D10_PRIMITIVE_TOPOLOGY_LINELIST; - prims = m_count / 2; - break; - case GS_TRIANGLE_CLASS: - topology = D3D10_PRIMITIVE_TOPOLOGY_TRIANGLELIST; - prims = m_count / 3; - break; - default: - __assume(0); - } - - m_perfmon.Put(GSPerfMon::Prim, prims); - m_perfmon.Put(GSPerfMon::Draw, 1); - - // date - - SetupDATE(rt, ds); - - // - - m_dev->BeginScene(); - - // om - - GSTextureFX::OMDepthStencilSelector om_dssel; - - om_dssel.zte = context->TEST.ZTE; - om_dssel.ztst = context->TEST.ZTST; - om_dssel.zwe = !context->ZBUF.ZMSK; - om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; - - GSTextureFX::OMBlendSelector om_bsel; - - om_bsel.abe = !IsOpaque(); - om_bsel.a = context->ALPHA.A; - om_bsel.b = context->ALPHA.B; - om_bsel.c = context->ALPHA.C; - om_bsel.d = context->ALPHA.D; - om_bsel.wr = (context->FRAME.FBMSK & 0x000000ff) != 0x000000ff; - om_bsel.wg = (context->FRAME.FBMSK & 0x0000ff00) != 0x0000ff00; - om_bsel.wb = (context->FRAME.FBMSK & 0x00ff0000) != 0x00ff0000; - om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; - - float bf = (float)(int)context->ALPHA.FIX / 0x80; - - // vs - - GSTextureFX::VSSelector vs_sel; - - vs_sel.bppz = 0; - vs_sel.tme = PRIM->TME; - vs_sel.fst = PRIM->FST; - vs_sel.prim = primclass; - - if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) - { - if(context->ZBUF.PSM == PSM_PSMZ24) - { - if(m_vt.m_max.p.z > 0xffffff) - { - ASSERT(m_vt.m_min.p.z > 0xffffff); - - vs_sel.bppz = 1; - om_dssel.ztst = 1; - } - } - else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) - { - if(m_vt.m_max.p.z > 0xffff) - { - ASSERT(m_vt.m_min.p.z > 0xffff); - - vs_sel.bppz = 2; - om_dssel.ztst = 1; - } - } - } - - GSTextureFX::VSConstantBuffer vs_cb; - - float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); - float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); - float ox = (float)(int)context->XYOFFSET.OFX; - float oy = (float)(int)context->XYOFFSET.OFY; - float ox2 = 1.0f / rt->GetWidth(); - float oy2 = 1.0f / rt->GetHeight(); - - vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); - vs_cb.VertexOffset = GSVector4(ox * sx - ox2 + 1, -(oy * sy - oy2 + 1), 0.0f, -1.0f); - vs_cb.TextureScale = GSVector2(1.0f, 1.0f); - - if(PRIM->TME && PRIM->FST) - { - vs_cb.TextureScale.x = 1.0f / (16 << context->TEX0.TW); - vs_cb.TextureScale.y = 1.0f / (16 << context->TEX0.TH); - } - - // gs - - GSTextureFX::GSSelector gs_sel; - - gs_sel.iip = PRIM->IIP; - gs_sel.prim = primclass; - - // ps - - GSTextureFX::PSSelector ps_sel; - - ps_sel.fst = PRIM->FST; - ps_sel.wms = context->CLAMP.WMS; - ps_sel.wmt = context->CLAMP.WMT; - ps_sel.bpp = 0; - ps_sel.aem = env.TEXA.AEM; - ps_sel.tfx = context->TEX0.TFX; - ps_sel.tcc = context->TEX0.TCC; - ps_sel.ate = context->TEST.ATE; - ps_sel.atst = context->TEST.ATST; - ps_sel.fog = PRIM->FGE; - ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; - ps_sel.fba = context->FBA.FBA; - ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; - ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - - GSTextureFX::PSSamplerSelector ps_ssel; - - ps_ssel.tau = 0; - ps_ssel.tav = 0; - ps_ssel.ltf = ps_sel.ltf; - - GSTextureFX::PSConstantBuffer ps_cb; - - ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - - if(ps_sel.atst == 2 || ps_sel.atst == 5) - { - ps_cb.FogColor_AREF.a -= 0.9f / 255; - } - else if(ps_sel.atst == 3 || ps_sel.atst == 6) - { - ps_cb.FogColor_AREF.a += 0.9f / 255; - } - - if(tex) - { - ps_sel.bpp = tex->m_bpp; - - int w = tex->m_texture->GetWidth(); - int h = tex->m_texture->GetHeight(); - - ps_cb.WH_TA = GSVector4((int)(1 << context->TEX0.TW), (int)(1 << context->TEX0.TH), env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); - ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); - - switch(context->CLAMP.WMS) - { - case 0: - ps_ssel.tau = 1; - break; - case 1: - ps_ssel.tau = 0; - break; - case 2: - ps_cb.MinMax.x = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); - ps_cb.MinMax.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); - ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); - ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); - ps_ssel.tau = 0; - break; - case 3: - ps_cb.MskFix.x = context->CLAMP.MINU; - ps_cb.MskFix.z = context->CLAMP.MAXU; - ps_ssel.tau = 1; - break; - default: - __assume(0); - } - - switch(context->CLAMP.WMT) - { - case 0: - ps_ssel.tav = 1; - break; - case 1: - ps_ssel.tav = 0; - break; - case 2: - ps_cb.MinMax.y = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); - ps_cb.MinMax.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); - ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); - ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); - ps_ssel.tav = 0; - break; - case 3: - ps_cb.MskFix.y = context->CLAMP.MINV; - ps_cb.MskFix.w = context->CLAMP.MAXV; - ps_ssel.tav = 1; - break; - default: - __assume(0); - } - } - else - { - ps_sel.tfx = 4; - } - - // rs - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); - - // - - m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); - m_tfx.SetupIA(m_vertices, m_count, topology); - m_tfx.SetupVS(vs_sel, &vs_cb); - m_tfx.SetupGS(gs_sel); - m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); - m_tfx.SetupRS(w, h, scissor); - - // draw - - if(context->TEST.DoFirstPass()) - { - m_dev->DrawPrimitive(); - } - - if(context->TEST.DoSecondPass()) - { - ASSERT(!env.PABE.PABE); - - static const uint32 iatst[] = {1, 0, 5, 6, 7, 2, 3, 4}; - - ps_sel.atst = iatst[ps_sel.atst]; - - m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel); - - bool z = om_dssel.zwe; - bool r = om_bsel.wr; - bool g = om_bsel.wg; - bool b = om_bsel.wb; - bool a = om_bsel.wa; - - switch(context->TEST.AFAIL) - { - case 0: z = r = g = b = a = false; break; // none - case 1: z = false; break; // rgba - case 2: r = g = b = a = false; break; // z - case 3: z = a = false; break; // rgb - default: __assume(0); - } - - if(z || r || g || b || a) - { - om_dssel.zwe = z; - om_bsel.wr = r; - om_bsel.wg = g; - om_bsel.wb = b; - om_bsel.wa = a; - - m_tfx.UpdateOM(om_dssel, om_bsel, bf); - - m_dev->DrawPrimitive(); - } - } - - m_dev->EndScene(); -} - -void GSRendererHW10::SetupDATE(GSTexture* rt, GSTexture* ds) -{ - if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) - - GSDevice10* dev = (GSDevice10*)m_dev; - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - if(GSTexture* t = dev->CreateRenderTarget(w, h)) - { - // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - - dev->BeginScene(); - - dev->ClearStencil(ds, 0); - - // om - - dev->OMSetDepthStencilState(m_date.dss, 1); - dev->OMSetBlendState(m_date.bs, 0); - dev->OMSetRenderTargets(t, ds); - - // ia - - GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); - GSVector4 o = GSVector4(-1.0f, 1.0f); - - GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); - GSVector4 dst = src * 2.0f + o.xxxx(); - - GSVertexPT1 vertices[] = - { - {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, - {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, - {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, - {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, - }; - - dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); - dev->IASetInputLayout(dev->m_convert.il); - dev->IASetPrimitiveTopology(D3D10_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); - - // vs - - dev->VSSetShader(dev->m_convert.vs, NULL); - - // gs - - dev->GSSetShader(NULL); - - // ps - - dev->PSSetShaderResources(rt, NULL); - dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); - dev->PSSetSamplerState(dev->m_convert.pt, NULL); - - // rs - - dev->RSSet(w, h); - - // set - - dev->DrawPrimitive(); - - // - - dev->EndScene(); - - dev->Recycle(t); - } -} diff --git a/plugins/GSdx/GSRendererHW11.cpp b/plugins/GSdx/GSRendererHW11.cpp deleted file mode 100644 index e68eaad4d3..0000000000 --- a/plugins/GSdx/GSRendererHW11.cpp +++ /dev/null @@ -1,529 +0,0 @@ -/* - * Copyright (C) 2007-2009 Gabest - * http://www.gabest.org - * - * This Program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2, or (at your option) - * any later version. - * - * This Program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with GNU Make; see the file COPYING. If not, write to - * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. - * http://www.gnu.org/copyleft/gpl.html - * - */ - -#include "stdafx.h" -#include "GSRendererHW11.h" -#include "GSCrc.h" -#include "resource.h" - -GSRendererHW11::GSRendererHW11(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice11(), new GSTextureCache11(this)) -{ - InitVertexKick(); -} - -bool GSRendererHW11::Create(const string& title) -{ - if(!__super::Create(title)) - return false; - - if(!m_tfx.Create((GSDevice11*)m_dev)) - return false; - - // - - D3D11_DEPTH_STENCIL_DESC dsd; - - memset(&dsd, 0, sizeof(dsd)); - - dsd.DepthEnable = false; - dsd.StencilEnable = true; - dsd.StencilReadMask = 1; - dsd.StencilWriteMask = 1; - dsd.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS; - dsd.FrontFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; - dsd.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; - dsd.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; - dsd.BackFace.StencilFunc = D3D11_COMPARISON_ALWAYS; - dsd.BackFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE; - dsd.BackFace.StencilFailOp = D3D11_STENCIL_OP_KEEP; - dsd.BackFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP; - - (*(GSDevice11*)m_dev)->CreateDepthStencilState(&dsd, &m_date.dss); - - D3D11_BLEND_DESC bd; - - memset(&bd, 0, sizeof(bd)); - - (*(GSDevice11*)m_dev)->CreateBlendState(&bd, &m_date.bs); - - // - - return true; -} - -template -void GSRendererHW11::VertexKick(bool skip) -{ - GSVertexHW11& dst = m_vl.AddTail(); - - dst.vi[0] = m_v.vi[0]; - dst.vi[1] = m_v.vi[1]; - - if(tme && fst) - { - GSVector4::storel(&dst.ST, m_v.GetUV()); - } - - int count = 0; - - if(GSVertexHW11* v = DrawingKick(skip, count)) - { - GSVector4i scissor = m_context->scissor.dx10; - - #if _M_SSE >= 0x401 - - GSVector4i pmin, pmax, v0, v1, v2; - - switch(prim) - { - case GS_POINTLIST: - v0 = GSVector4i::load((int)v[0].p.xy).upl16(); - pmin = v0; - pmax = v0; - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - v0 = GSVector4i::load((int)v[0].p.xy); - v1 = GSVector4i::load((int)v[1].p.xy); - pmin = v0.min_u16(v1).upl16(); - pmax = v0.max_u16(v1).upl16(); - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - v0 = GSVector4i::load((int)v[0].p.xy); - v1 = GSVector4i::load((int)v[1].p.xy); - v2 = GSVector4i::load((int)v[2].p.xy); - pmin = v0.min_u16(v1).min_u16(v2).upl16(); - pmax = v0.max_u16(v1).max_u16(v2).upl16(); - break; - } - - GSVector4i test = (pmax < scissor) | (pmin > scissor.zwxy()); - - if(test.mask() & 0xff) - { - return; - } - - #else - - switch(prim) - { - case GS_POINTLIST: - if(v[0].p.x < scissor.x - || v[0].p.x > scissor.z - || v[0].p.y < scissor.y - || v[0].p.y > scissor.w) - { - return; - } - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - if(v[0].p.x < scissor.x && v[1].p.x < scissor.x - || v[0].p.x > scissor.z && v[1].p.x > scissor.z - || v[0].p.y < scissor.y && v[1].p.y < scissor.y - || v[0].p.y > scissor.w && v[1].p.y > scissor.w) - { - return; - } - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - if(v[0].p.x < scissor.x && v[1].p.x < scissor.x && v[2].p.x < scissor.x - || v[0].p.x > scissor.z && v[1].p.x > scissor.z && v[2].p.x > scissor.z - || v[0].p.y < scissor.y && v[1].p.y < scissor.y && v[2].p.y < scissor.y - || v[0].p.y > scissor.w && v[1].p.y > scissor.w && v[2].p.y > scissor.w) - { - return; - } - break; - } - - #endif - - m_count += count; - } -} - -void GSRendererHW11::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) -{ - GSDrawingEnvironment& env = m_env; - GSDrawingContext* context = m_context; - - D3D11_PRIMITIVE_TOPOLOGY topology; - int prims = 0; - - switch(primclass) - { - case GS_POINT_CLASS: - topology = D3D11_PRIMITIVE_TOPOLOGY_POINTLIST; - prims = m_count; - break; - case GS_LINE_CLASS: - case GS_SPRITE_CLASS: - topology = D3D11_PRIMITIVE_TOPOLOGY_LINELIST; - prims = m_count / 2; - break; - case GS_TRIANGLE_CLASS: - topology = D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST; - prims = m_count / 3; - break; - default: - __assume(0); - } - - m_perfmon.Put(GSPerfMon::Prim, prims); - m_perfmon.Put(GSPerfMon::Draw, 1); - - // date - - SetupDATE(rt, ds); - - // - - m_dev->BeginScene(); - - // om - - GSTextureFX::OMDepthStencilSelector om_dssel; - - om_dssel.zte = context->TEST.ZTE; - om_dssel.ztst = context->TEST.ZTST; - om_dssel.zwe = !context->ZBUF.ZMSK; - om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; - - GSTextureFX::OMBlendSelector om_bsel; - - om_bsel.abe = !IsOpaque(); - om_bsel.a = context->ALPHA.A; - om_bsel.b = context->ALPHA.B; - om_bsel.c = context->ALPHA.C; - om_bsel.d = context->ALPHA.D; - om_bsel.wr = (context->FRAME.FBMSK & 0x000000ff) != 0x000000ff; - om_bsel.wg = (context->FRAME.FBMSK & 0x0000ff00) != 0x0000ff00; - om_bsel.wb = (context->FRAME.FBMSK & 0x00ff0000) != 0x00ff0000; - om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; - - float bf = (float)(int)context->ALPHA.FIX / 0x80; - - // vs - - GSTextureFX::VSSelector vs_sel; - - vs_sel.bppz = 0; - vs_sel.tme = PRIM->TME; - vs_sel.fst = PRIM->FST; - vs_sel.prim = primclass; - - if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) - { - if(context->ZBUF.PSM == PSM_PSMZ24) - { - if(m_vt.m_max.p.z > 0xffffff) - { - ASSERT(m_vt.m_min.p.z > 0xffffff); - - vs_sel.bppz = 1; - om_dssel.ztst = 1; - } - } - else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) - { - if(m_vt.m_max.p.z > 0xffff) - { - ASSERT(m_vt.m_min.p.z > 0xffff); - - vs_sel.bppz = 2; - om_dssel.ztst = 1; - } - } - } - - GSTextureFX::VSConstantBuffer vs_cb; - - float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); - float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); - float ox = (float)(int)context->XYOFFSET.OFX; - float oy = (float)(int)context->XYOFFSET.OFY; - float ox2 = 1.0f / rt->GetWidth(); - float oy2 = 1.0f / rt->GetHeight(); - - vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); - vs_cb.VertexOffset = GSVector4(ox * sx - ox2 + 1, -(oy * sy - oy2 + 1), 0.0f, -1.0f); - vs_cb.TextureScale = GSVector2(1.0f, 1.0f); - - if(PRIM->TME && PRIM->FST) - { - vs_cb.TextureScale.x = 1.0f / (16 << context->TEX0.TW); - vs_cb.TextureScale.y = 1.0f / (16 << context->TEX0.TH); - } - - // gs - - GSTextureFX::GSSelector gs_sel; - - gs_sel.iip = PRIM->IIP; - gs_sel.prim = primclass; - - // ps - - GSTextureFX::PSSelector ps_sel; - - ps_sel.fst = PRIM->FST; - ps_sel.wms = context->CLAMP.WMS; - ps_sel.wmt = context->CLAMP.WMT; - ps_sel.bpp = 0; - ps_sel.aem = env.TEXA.AEM; - ps_sel.tfx = context->TEX0.TFX; - ps_sel.tcc = context->TEX0.TCC; - ps_sel.ate = context->TEST.ATE; - ps_sel.atst = context->TEST.ATST; - ps_sel.fog = PRIM->FGE; - ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; - ps_sel.fba = context->FBA.FBA; - ps_sel.aout = context->FRAME.PSM == PSM_PSMCT16 || context->FRAME.PSM == PSM_PSMCT16S || (context->FRAME.FBMSK & 0xff000000) == 0x7f000000 ? 1 : 0; - ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - - GSTextureFX::PSSamplerSelector ps_ssel; - - ps_ssel.tau = 0; - ps_ssel.tav = 0; - ps_ssel.ltf = ps_sel.ltf; - - GSTextureFX::PSConstantBuffer ps_cb; - - ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - - if(ps_sel.atst == 2 || ps_sel.atst == 5) - { - ps_cb.FogColor_AREF.a -= 0.9f / 255; - } - else if(ps_sel.atst == 3 || ps_sel.atst == 6) - { - ps_cb.FogColor_AREF.a += 0.9f / 255; - } - - if(tex) - { - ps_sel.bpp = tex->m_bpp; - - int w = tex->m_texture->GetWidth(); - int h = tex->m_texture->GetHeight(); - - ps_cb.WH_TA = GSVector4((int)(1 << context->TEX0.TW), (int)(1 << context->TEX0.TH), env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); - ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); - - switch(context->CLAMP.WMS) - { - case 0: - ps_ssel.tau = 1; - break; - case 1: - ps_ssel.tau = 0; - break; - case 2: - ps_cb.MinMax.x = ((float)(int)context->CLAMP.MINU) / (1 << context->TEX0.TW); - ps_cb.MinMax.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); - ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); - ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); - ps_ssel.tau = 0; - break; - case 3: - ps_cb.MskFix.x = context->CLAMP.MINU; - ps_cb.MskFix.z = context->CLAMP.MAXU; - ps_ssel.tau = 1; - break; - default: - __assume(0); - } - - switch(context->CLAMP.WMT) - { - case 0: - ps_ssel.tav = 1; - break; - case 1: - ps_ssel.tav = 0; - break; - case 2: - ps_cb.MinMax.y = ((float)(int)context->CLAMP.MINV) / (1 << context->TEX0.TH); - ps_cb.MinMax.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); - ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); - ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); - ps_ssel.tav = 0; - break; - case 3: - ps_cb.MskFix.y = context->CLAMP.MINV; - ps_cb.MskFix.w = context->CLAMP.MAXV; - ps_ssel.tav = 1; - break; - default: - __assume(0); - } - } - else - { - ps_sel.tfx = 4; - } - - // rs - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); - - // - - m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); - m_tfx.SetupIA(m_vertices, m_count, topology); - m_tfx.SetupVS(vs_sel, &vs_cb); - m_tfx.SetupGS(gs_sel); - m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); - m_tfx.SetupRS(w, h, scissor); - - // draw - - if(context->TEST.DoFirstPass()) - { - m_dev->DrawPrimitive(); - } - - if(context->TEST.DoSecondPass()) - { - ASSERT(!env.PABE.PABE); - - static const uint32 iatst[] = {1, 0, 5, 6, 7, 2, 3, 4}; - - ps_sel.atst = iatst[ps_sel.atst]; - - m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel); - - bool z = om_dssel.zwe; - bool r = om_bsel.wr; - bool g = om_bsel.wg; - bool b = om_bsel.wb; - bool a = om_bsel.wa; - - switch(context->TEST.AFAIL) - { - case 0: z = r = g = b = a = false; break; // none - case 1: z = false; break; // rgba - case 2: r = g = b = a = false; break; // z - case 3: z = a = false; break; // rgb - default: __assume(0); - } - - if(z || r || g || b || a) - { - om_dssel.zwe = z; - om_bsel.wr = r; - om_bsel.wg = g; - om_bsel.wb = b; - om_bsel.wa = a; - - m_tfx.UpdateOM(om_dssel, om_bsel, bf); - - m_dev->DrawPrimitive(); - } - } - - m_dev->EndScene(); -} - -void GSRendererHW11::SetupDATE(GSTexture* rt, GSTexture* ds) -{ - if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) - - GSDevice11* dev = (GSDevice11*)m_dev; - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - if(GSTexture* t = dev->CreateRenderTarget(w, h)) - { - // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - - dev->BeginScene(); - - dev->ClearStencil(ds, 0); - - // om - - dev->OMSetDepthStencilState(m_date.dss, 1); - dev->OMSetBlendState(m_date.bs, 0); - dev->OMSetRenderTargets(t, ds); - - // ia - - GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); - GSVector4 o = GSVector4(-1.0f, 1.0f); - - GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); - GSVector4 dst = src * 2.0f + o.xxxx(); - - GSVertexPT1 vertices[] = - { - {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, - {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, - {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, - {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, - }; - - dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); - dev->IASetInputLayout(dev->m_convert.il); - dev->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP); - - // vs - - dev->VSSetShader(dev->m_convert.vs, NULL); - - // gs - - dev->GSSetShader(NULL); - - // ps - - dev->PSSetShaderResources(rt, NULL); - dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL); - dev->PSSetSamplerState(dev->m_convert.pt, NULL); - - // rs - - dev->RSSet(w, h); - - // set - - dev->DrawPrimitive(); - - // - - dev->EndScene(); - - dev->Recycle(t); - } -} diff --git a/plugins/GSdx/GSRendererHW9.cpp b/plugins/GSdx/GSRendererHW9.cpp deleted file mode 100644 index a4d635fc87..0000000000 --- a/plugins/GSdx/GSRendererHW9.cpp +++ /dev/null @@ -1,573 +0,0 @@ -/* - * Copyright (C) 2007-2009 Gabest - * http://www.gabest.org - * - * This Program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2, or (at your option) - * any later version. - * - * This Program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with GNU Make; see the file COPYING. If not, write to - * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. - * http://www.gnu.org/copyleft/gpl.html - * - */ - -#include "stdafx.h" -#include "GSRendererHW9.h" -#include "GSCrc.h" -#include "resource.h" - -GSRendererHW9::GSRendererHW9(uint8* base, bool mt, void (*irq)()) - : GSRendererHW(base, mt, irq, new GSDevice9(), new GSTextureCache9(this)) -{ - m_fba.enabled = !!theApp.GetConfig("fba", 1); - m_logz = !!theApp.GetConfig("logz", 0); - - InitVertexKick(); -} - -bool GSRendererHW9::Create(const string& title) -{ - if(!__super::Create(title)) - return false; - - if(!m_tfx.Create((GSDevice9*)m_dev)) - return false; - - // - - memset(&m_date.dss, 0, sizeof(m_date.dss)); - - m_date.dss.StencilEnable = true; - m_date.dss.StencilReadMask = 1; - m_date.dss.StencilWriteMask = 1; - m_date.dss.StencilFunc = D3DCMP_ALWAYS; - m_date.dss.StencilPassOp = D3DSTENCILOP_REPLACE; - m_date.dss.StencilRef = 1; - - memset(&m_date.bs, 0, sizeof(m_date.bs)); - - // - - memset(&m_fba.dss, 0, sizeof(m_fba.dss)); - - m_fba.dss.StencilEnable = true; - m_fba.dss.StencilReadMask = 2; - m_fba.dss.StencilWriteMask = 2; - m_fba.dss.StencilFunc = D3DCMP_EQUAL; - m_fba.dss.StencilPassOp = D3DSTENCILOP_ZERO; - m_fba.dss.StencilFailOp = D3DSTENCILOP_ZERO; - m_fba.dss.StencilDepthFailOp = D3DSTENCILOP_ZERO; - m_fba.dss.StencilRef = 2; - - memset(&m_fba.bs, 0, sizeof(m_fba.bs)); - - m_fba.bs.RenderTargetWriteMask = D3DCOLORWRITEENABLE_ALPHA; - - // - - return true; -} - -template -void GSRendererHW9::VertexKick(bool skip) -{ - GSVertexHW9& dst = m_vl.AddTail(); - - dst.p.x = (float)(int)m_v.XYZ.X; - dst.p.y = (float)(int)m_v.XYZ.Y; - dst.p.z = (float)m_v.XYZ.Z; - - dst.c0 = m_v.RGBAQ.u32[0]; - dst.c1 = m_v.FOG.u32[1]; - - if(tme) - { - if(fst) - { - GSVector4::storel(&dst.t, m_v.GetUV()); - } - else - { - dst.t.x = m_v.ST.S; - dst.t.y = m_v.ST.T; - dst.p.w = m_v.RGBAQ.Q; - } - } - - int count = 0; - - if(GSVertexHW9* v = DrawingKick(skip, count)) - { - GSVector4 scissor = m_context->scissor.dx9; - - GSVector4 pmin, pmax; - - switch(prim) - { - case GS_POINTLIST: - pmin = v[0].p; - pmax = v[0].p; - break; - case GS_LINELIST: - case GS_LINESTRIP: - case GS_SPRITE: - pmin = v[0].p.minv(v[1].p); - pmax = v[0].p.maxv(v[1].p); - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - pmin = v[0].p.minv(v[1].p).minv(v[2].p); - pmax = v[0].p.maxv(v[1].p).maxv(v[2].p); - break; - } - - GSVector4 test = (pmax < scissor) | (pmin > scissor.zwxy()); - - if(test.mask() & 3) - { - return; - } - - switch(prim) - { - case GS_POINTLIST: - break; - case GS_LINELIST: - case GS_LINESTRIP: - if(PRIM->IIP == 0) {v[0].c0 = v[1].c0;} - break; - case GS_TRIANGLELIST: - case GS_TRIANGLESTRIP: - case GS_TRIANGLEFAN: - if(PRIM->IIP == 0) {v[0].c0 = v[1].c0 = v[2].c0;} - break; - case GS_SPRITE: - if(PRIM->IIP == 0) {v[0].c0 = v[1].c0;} - v[0].p.z = v[1].p.z; - v[0].p.w = v[1].p.w; - v[0].c1 = v[1].c1; - v[2] = v[1]; - v[3] = v[1]; - v[1].p.y = v[0].p.y; - v[1].t.y = v[0].t.y; - v[2].p.x = v[0].p.x; - v[2].t.x = v[0].t.x; - v[4] = v[1]; - v[5] = v[2]; - count += 4; - break; - } - - m_count += count; - } -} - -void GSRendererHW9::Draw(GS_PRIM_CLASS primclass, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) -{ - GSDrawingEnvironment& env = m_env; - GSDrawingContext* context = m_context; - - D3DPRIMITIVETYPE topology; - int prims = 0; - - switch(primclass) - { - case GS_POINT_CLASS: - topology = D3DPT_POINTLIST; - prims = m_count; - break; - case GS_LINE_CLASS: - topology = D3DPT_LINELIST; - prims = m_count / 2; - break; - case GS_TRIANGLE_CLASS: - case GS_SPRITE_CLASS: - topology = D3DPT_TRIANGLELIST; - prims = m_count / 3; - break; - default: - __assume(0); - } - - m_perfmon.Put(GSPerfMon::Prim, prims); - m_perfmon.Put(GSPerfMon::Draw, 1); - - // date - - SetupDATE(rt, ds); - - // - - m_dev->BeginScene(); - - (*(GSDevice9*)m_dev)->SetRenderState(D3DRS_SHADEMODE, PRIM->IIP ? D3DSHADE_GOURAUD : D3DSHADE_FLAT); // TODO - - // om - - GSTextureFX::OMDepthStencilSelector om_dssel; - - om_dssel.zte = context->TEST.ZTE; - om_dssel.ztst = context->TEST.ZTST; - om_dssel.zwe = !context->ZBUF.ZMSK; - om_dssel.date = context->FRAME.PSM != PSM_PSMCT24 ? context->TEST.DATE : 0; - om_dssel.fba = m_fba.enabled ? context->FBA.FBA : 0; - - GSTextureFX::OMBlendSelector om_bsel; - - om_bsel.abe = !IsOpaque(); - om_bsel.a = context->ALPHA.A; - om_bsel.b = context->ALPHA.B; - om_bsel.c = context->ALPHA.C; - om_bsel.d = context->ALPHA.D; - om_bsel.wr = (context->FRAME.FBMSK & 0x000000ff) != 0x000000ff; - om_bsel.wg = (context->FRAME.FBMSK & 0x0000ff00) != 0x0000ff00; - om_bsel.wb = (context->FRAME.FBMSK & 0x00ff0000) != 0x00ff0000; - om_bsel.wa = (context->FRAME.FBMSK & 0xff000000) != 0xff000000; - - uint8 bf = context->ALPHA.FIX >= 0x80 ? 0xff : (uint8)(context->ALPHA.FIX * 2); - - // vs - - GSTextureFX::VSSelector vs_sel; - - vs_sel.bppz = 0; - vs_sel.tme = PRIM->TME; - vs_sel.fst = PRIM->FST; - vs_sel.logz = m_logz ? 1 : 0; - - if(om_dssel.zte && om_dssel.ztst > 0 && om_dssel.zwe) - { - if(context->ZBUF.PSM == PSM_PSMZ24) - { - if(m_vt.m_max.p.z > 0xffffff) - { - ASSERT(m_vt.m_min.p.z > 0xffffff); - - vs_sel.bppz = 1; - om_dssel.ztst = 1; - } - } - else if(context->ZBUF.PSM == PSM_PSMZ16 || context->ZBUF.PSM == PSM_PSMZ16S) - { - if(m_vt.m_max.p.z > 0xffff) - { - ASSERT(m_vt.m_min.p.z > 0xffff); - - vs_sel.bppz = 2; - om_dssel.ztst = 1; - } - } - } - - GSTextureFX::VSConstantBuffer vs_cb; - - float sx = 2.0f * rt->m_scale.x / (rt->GetWidth() * 16); - float sy = 2.0f * rt->m_scale.y / (rt->GetHeight() * 16); - float ox = (float)(int)context->XYOFFSET.OFX; - float oy = (float)(int)context->XYOFFSET.OFY; - - vs_cb.VertexScale = GSVector4(sx, -sy, 1.0f / UINT_MAX, 0.0f); - vs_cb.VertexOffset = GSVector4(ox * sx + 1, -(oy * sy + 1), 0.0f, -1.0f); - vs_cb.TextureScale = GSVector2(1.0f, 1.0f); - - if(PRIM->TME && PRIM->FST) - { - vs_cb.TextureScale.x = 1.0f / (16 << context->TEX0.TW); - vs_cb.TextureScale.y = 1.0f / (16 << context->TEX0.TH); - } - - // ps - - GSTextureFX::PSSelector ps_sel; - - ps_sel.fst = PRIM->FST; - ps_sel.wms = context->CLAMP.WMS; - ps_sel.wmt = context->CLAMP.WMT; - ps_sel.bpp = 0; - ps_sel.aem = env.TEXA.AEM; - ps_sel.tfx = context->TEX0.TFX; - ps_sel.tcc = context->TEX0.TCC; - ps_sel.ate = context->TEST.ATE; - ps_sel.atst = context->TEST.ATST; - ps_sel.fog = PRIM->FGE; - ps_sel.clr1 = om_bsel.abe && om_bsel.a == 1 && om_bsel.b == 2 && om_bsel.d == 1; - ps_sel.rt = tex && tex->m_rendered; - ps_sel.ltf = m_filter == 2 ? context->TEX1.IsLinear() : m_filter; - - GSTextureFX::PSSamplerSelector ps_ssel; - - ps_ssel.tau = 0; - ps_ssel.tav = 0; - ps_ssel.ltf = ps_sel.ltf; - - GSTextureFX9::PSConstantBuffer ps_cb; - - ps_cb.FogColor_AREF = GSVector4((int)env.FOGCOL.FCR, (int)env.FOGCOL.FCG, (int)env.FOGCOL.FCB, (int)context->TEST.AREF) / 255; - - if(ps_sel.atst == 2 || ps_sel.atst == 5) - { - ps_cb.FogColor_AREF.a -= 0.9f / 255; - } - else if(ps_sel.atst == 3 || ps_sel.atst == 6) - { - ps_cb.FogColor_AREF.a += 0.9f / 255; - } - - if(tex) - { - ps_sel.bpp = tex->m_bpp; - - int w = tex->m_texture->GetWidth(); - int h = tex->m_texture->GetHeight(); - - ps_cb.WH_TA = GSVector4(w, h, env.TEXA.TA0, env.TEXA.TA1) / GSVector4(1, 255).xxyy(); - ps_cb.HalfTexel = GSVector4(-0.5f, 0.5f).xxyy() / GSVector4(w, h).xyxy(); - - switch(context->CLAMP.WMS) - { - case 0: - ps_ssel.tau = 1; - break; - case 1: - ps_ssel.tau = 0; - break; - case 2: - ps_cb.MinMax.x = (float)(int)context->CLAMP.MINU / (1 << context->TEX0.TW); - ps_cb.MinMax.z = (float)(int)context->CLAMP.MAXU / (1 << context->TEX0.TW); - ps_cb.MinMaxF.x = ((float)(int)context->CLAMP.MINU + 0.5f) / (1 << context->TEX0.TW); - ps_cb.MinMaxF.z = ((float)(int)context->CLAMP.MAXU) / (1 << context->TEX0.TW); - ps_ssel.tau = 0; - break; - case 3: - ps_cb.MskFix.x = context->CLAMP.MINU; - ps_cb.MskFix.z = context->CLAMP.MAXU; - ps_ssel.tau = 1; - break; - default: - __assume(0); - } - - switch(context->CLAMP.WMT) - { - case 0: - ps_ssel.tav = 1; - break; - case 1: - ps_ssel.tav = 0; - break; - case 2: - ps_cb.MinMax.y = (float)(int)context->CLAMP.MINV / (1 << context->TEX0.TH); - ps_cb.MinMax.w = (float)(int)context->CLAMP.MAXV / (1 << context->TEX0.TH); - ps_cb.MinMaxF.y = ((float)(int)context->CLAMP.MINV + 0.5f) / (1 << context->TEX0.TH); - ps_cb.MinMaxF.w = ((float)(int)context->CLAMP.MAXV) / (1 << context->TEX0.TH); - ps_ssel.tav = 0; - break; - case 3: - ps_cb.MskFix.y = context->CLAMP.MINV; - ps_cb.MskFix.w = context->CLAMP.MAXV; - ps_ssel.tav = 1; - break; - default: - __assume(0); - } - } - else - { - ps_sel.tfx = 4; - } - - // rs - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - GSVector4i scissor = GSVector4i(GSVector4(rt->m_scale).xyxy() * context->scissor.in).rintersect(GSVector4i(0, 0, w, h)); - - // - - m_tfx.SetupOM(om_dssel, om_bsel, bf, rt, ds); - m_tfx.SetupIA(m_vertices, m_count, topology); - m_tfx.SetupVS(vs_sel, &vs_cb); - m_tfx.SetupPS(ps_sel, &ps_cb, ps_ssel, tex ? tex->m_texture : NULL, tex ? tex->m_palette : NULL); - m_tfx.SetupRS(w, h, scissor); - - // draw - - if(context->TEST.DoFirstPass()) - { - m_dev->DrawPrimitive(); - } - - if(context->TEST.DoSecondPass()) - { - ASSERT(!env.PABE.PABE); - - static const uint32 iatst[] = {1, 0, 5, 6, 7, 2, 3, 4}; - - ps_sel.atst = iatst[ps_sel.atst]; - - m_tfx.UpdatePS(ps_sel, &ps_cb, ps_ssel); - - bool z = om_dssel.zwe; - bool r = om_bsel.wr; - bool g = om_bsel.wg; - bool b = om_bsel.wb; - bool a = om_bsel.wa; - - switch(context->TEST.AFAIL) - { - case 0: z = r = g = b = a = false; break; // none - case 1: z = false; break; // rgba - case 2: r = g = b = a = false; break; // z - case 3: z = a = false; break; // rgb - default: __assume(0); - } - - if(z || r || g || b || a) - { - om_dssel.zwe = z; - om_bsel.wr = r; - om_bsel.wg = g; - om_bsel.wb = b; - om_bsel.wa = a; - - m_tfx.UpdateOM(om_dssel, om_bsel, bf); - - m_dev->DrawPrimitive(); - } - } - - m_dev->EndScene(); - - if(om_dssel.fba) UpdateFBA(rt); -} - -void GSRendererHW9::SetupDATE(GSTexture* rt, GSTexture* ds) -{ - if(!m_context->TEST.DATE) return; // || (::GetAsyncKeyState(VK_CONTROL) & 0x8000) - - GSDevice9* dev = (GSDevice9*)m_dev; - - int w = rt->GetWidth(); - int h = rt->GetHeight(); - - if(GSTexture* t = dev->CreateRenderTarget(w, h)) - { - // sfex3 (after the capcom logo), vf4 (first menu fading in), ffxii shadows, rumble roses shadows, persona4 shadows - - dev->BeginScene(); - - dev->ClearStencil(ds, 0); - - // om - - dev->OMSetDepthStencilState(&m_date.dss); - dev->OMSetBlendState(&m_date.bs, 0); - dev->OMSetRenderTargets(t, ds); - - // ia - - GSVector4 s = GSVector4(rt->m_scale.x / w, rt->m_scale.y / h); - GSVector4 o = GSVector4(-1.0f, 1.0f); - - GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); - GSVector4 dst = src * 2.0f + o.xxxx(); - - GSVertexPT1 vertices[] = - { - {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(src.x, src.y)}, - {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(src.z, src.y)}, - {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(src.x, src.w)}, - {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(src.z, src.w)}, - }; - - dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); - dev->IASetInputLayout(dev->m_convert.il); - dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); - - // vs - - dev->VSSetShader(dev->m_convert.vs, NULL, 0); - - // ps - - dev->PSSetShaderResources(rt, NULL); - dev->PSSetShader(dev->m_convert.ps[m_context->TEST.DATM ? 2 : 3], NULL, 0); - dev->PSSetSamplerState(&dev->m_convert.pt); - - // rs - - dev->RSSet(w, h); - - // - - dev->DrawPrimitive(); - - // - - dev->EndScene(); - - dev->Recycle(t); - } -} - -void GSRendererHW9::UpdateFBA(GSTexture* rt) -{ - GSDevice9* dev = (GSDevice9*)m_dev; - - dev->BeginScene(); - - // om - - dev->OMSetDepthStencilState(&m_fba.dss); - dev->OMSetBlendState(&m_fba.bs, 0); - - // ia - - GSVector4 s = GSVector4(rt->m_scale.x / rt->GetWidth(), rt->m_scale.y / rt->GetHeight()); - GSVector4 o = GSVector4(-1.0f, 1.0f); - - GSVector4 src = ((m_vt.m_min.p.xyxy(m_vt.m_max.p) + o.xxyy()) * s.xyxy()).sat(o.zzyy()); - GSVector4 dst = src * 2.0f + o.xxxx(); - - GSVertexPT1 vertices[] = - { - {GSVector4(dst.x, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(dst.z, -dst.y, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(dst.x, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, - {GSVector4(dst.z, -dst.w, 0.5f, 1.0f), GSVector2(0, 0)}, - }; - - dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), countof(vertices)); - dev->IASetInputLayout(dev->m_convert.il); - dev->IASetPrimitiveTopology(D3DPT_TRIANGLESTRIP); - - // vs - - dev->VSSetShader(dev->m_convert.vs, NULL, 0); - - // ps - - dev->PSSetShader(dev->m_convert.ps[4], NULL, 0); - - // rs - - dev->RSSet(rt->GetWidth(), rt->GetHeight()); - - // - - dev->DrawPrimitive(); - - // - - dev->EndScene(); -} diff --git a/plugins/GSdx/GSRendererOGL.cpp b/plugins/GSdx/GSRendererOGL.cpp index 942734e7d2..107eea2bab 100644 --- a/plugins/GSdx/GSRendererOGL.cpp +++ b/plugins/GSdx/GSRendererOGL.cpp @@ -64,13 +64,12 @@ void GSRendererOGL::VertexKick(bool skip) } } -void GSRendererOGL::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex) +void GSRendererOGL::Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex) { GSDrawingEnvironment& env = m_env; GSDrawingContext* context = m_context; // m_perfmon.Put(GSPerfMon::Prim, prims); - m_perfmon.Put(GSPerfMon::Draw, 1); m_dev->BeginScene(); diff --git a/plugins/GSdx/GSRendererOGL.h b/plugins/GSdx/GSRendererOGL.h index aa63b99183..9dc3587b93 100644 --- a/plugins/GSdx/GSRendererOGL.h +++ b/plugins/GSdx/GSRendererOGL.h @@ -29,7 +29,7 @@ class GSRendererOGL : public GSRendererHW { protected: - void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::GSCachedTexture* tex); + void Draw(int prim, GSTexture* rt, GSTexture* ds, GSTextureCache::Source* tex); public: GSRendererOGL(uint8* base, bool mt, void (*irq)()); diff --git a/plugins/GSdx/GSRendererSW.cpp b/plugins/GSdx/GSRendererSW.cpp index 9828f8757e..63bee1db97 100644 --- a/plugins/GSdx/GSRendererSW.cpp +++ b/plugins/GSdx/GSRendererSW.cpp @@ -329,7 +329,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) p.sel.tfx = context->TEX0.TFX; p.sel.tcc = context->TEX0.TCC; p.sel.fst = PRIM->FST; - p.sel.ltf = context->TEX1.IsLinear(); + p.sel.ltf = IsLinear(); p.sel.tlu = GSLocalMemory::m_psm[context->TEX0.PSM].pal > 0; p.sel.wms = context->CLAMP.WMS; p.sel.wmt = context->CLAMP.WMT; @@ -394,7 +394,7 @@ void GSRendererSW::GetScanlineParam(GSScanlineParam& p, GS_PRIM_CLASS primclass) GSVector4i r; - GetTextureMinMax(r); + GetTextureMinMax(r, p.sel.ltf); const GSTextureCacheSW::GSTexture* t = m_tc->Lookup(context->TEX0, env.TEXA, r); diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index 28ae26542c..d8c71c3a06 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -34,94 +34,172 @@ GSTextureCache::~GSTextureCache() void GSTextureCache::RemoveAll() { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + m_src.RemoveAll(); + + for(int type = 0; type < 2; type++) { - delete *i; + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); i++) + { + delete *i; + } + + m_dst[type].clear(); } - - m_rt.clear(); - - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - delete *i; - } - - m_ds.clear(); - - for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) - { - delete *i; - } - - m_tex.clear(); } -GSTextureCache::GSRenderTarget* GSTextureCache::GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb) +GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r) { - GSRenderTarget* rt = NULL; + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; + const uint32* clut = m_renderer->m_mem.m_clut; + + Source* src = NULL; - if(rt == NULL) + const hash_map& map = m_src.m_map[TEX0.TBP0 >> 5]; + + for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + Source* s = i->first; + + if(((s->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((s->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH { - GSRenderTarget* rt2 = *i; + continue; + } - if(rt2->m_TEX0.TBP0 == TEX0.TBP0) + if((psm.trbpp == 16 || psm.trbpp == 24) && TEX0.TCC && TEXA != s->m_TEXA) + { + continue; + } + + if(psm.pal > 0 && !GSVector4i::compare(s->m_clut, clut, psm.pal * sizeof(clut[0]))) + { + continue; + } + + src = s; + + break; + } + + Target* dst = NULL; + + if(src == NULL) + { + for(int type = 0; type < 2 && dst == NULL; type++) + { + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); i++) { - m_rt.splice(m_rt.begin(), m_rt, i); + Target* t = *i; - rt = rt2; + if(t->m_used && t->m_dirty.empty() && GSUtil::HasSharedBits(t->m_TEX0.TBP0, t->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) + { + dst = t; - if(!fb) rt->m_TEX0 = TEX0; - - rt->Update(); - - break; + break; + } } } } - if(rt == NULL && fb) + if(src == NULL) { - // HACK: try to find something close to the base pointer + src = CreateSource(); - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) + if(!(dst ? src->Create(dst) : src->Create())) { - GSRenderTarget* rt2 = *i; - - if(rt2->m_TEX0.TBP0 <= TEX0.TBP0 && TEX0.TBP0 < rt2->m_TEX0.TBP0 + 0x700 && (!rt || rt2->m_TEX0.TBP0 >= rt->m_TEX0.TBP0)) - { - rt = rt2; - } - } - - if(rt) - { - rt->Update(); - } - } - - if(rt == NULL) - { - rt = CreateRenderTarget(); - - rt->m_TEX0 = TEX0; - - if(!rt->Create(w, h)) - { - delete rt; + delete src; return NULL; } - m_rt.push_front(rt); + if(psm.pal > 0) + { + memcpy(src->m_clut, clut, psm.pal * sizeof(clut[0])); + } + + m_src.Add(src, TEX0); + } + + if(psm.pal > 0) + { + int size = psm.pal * sizeof(clut[0]); + + if(src->m_palette) + { + if(src->m_initpalette || GSVector4i::update(src->m_clut, clut, size)) + { + src->m_palette->Update(GSVector4i(0, 0, psm.pal, 1), src->m_clut, size); + src->m_initpalette = false; + } + } + } + + src->Update(TEX0, TEXA, r); + + m_src.m_used = true; + + return src; +} + +GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, bool fb) +{ + Target* dst = NULL; + + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); i++) + { + Target* t = *i; + + if(t->m_TEX0.TBP0 == TEX0.TBP0) + { + m_dst[type].splice(m_dst[type].begin(), m_dst[type], i); + + dst = t; + + if(!fb) dst->m_TEX0 = TEX0; + + break; + } + } + + if(dst == NULL && fb) + { + // HACK: try to find something close to the base pointer + + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); i++) + { + Target* t = *i; + + if(t->m_TEX0.TBP0 <= TEX0.TBP0 && TEX0.TBP0 < t->m_TEX0.TBP0 + 0x700 && (!dst || t->m_TEX0.TBP0 >= dst->m_TEX0.TBP0)) + { + dst = t; + } + } + } + + if(dst == NULL) + { + dst = CreateTarget(); + + dst->m_TEX0 = TEX0; + + if(!dst->Create(w, h, type)) + { + delete dst; + + return NULL; + } + + m_dst[type].push_front(dst); + } + else + { + dst->Update(); } if(m_renderer->CanUpscale()) { GSVector4i fr = m_renderer->GetFrameRect(); - int ww = (int)(fr.left + rt->m_TEX0.TBW * 64); + int ww = (int)(fr.left + dst->m_TEX0.TBW * 64); int hh = (int)(fr.top + m_renderer->GetDisplayRect().height()); if(hh <= m_renderer->GetDeviceSize().y / 2) @@ -136,346 +214,129 @@ GSTextureCache::GSRenderTarget* GSTextureCache::GetRenderTarget(const GIFRegTEX0 */ if(ww > 0 && hh > 0) { - rt->m_texture->m_scale.x = (float)w / ww; - rt->m_texture->m_scale.y = (float)h / hh; + dst->m_texture->m_scale.x = (float)w / ww; + dst->m_texture->m_scale.y = (float)h / hh; } } - if(!fb) + if(used) { - rt->m_used = true; + dst->m_used = true; } - return rt; + return dst; } -GSTextureCache::GSDepthStencil* GSTextureCache::GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h) -{ - GSDepthStencil* ds = NULL; - - if(ds == NULL) - { - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - GSDepthStencil* ds2 = *i; - - if(ds2->m_TEX0.TBP0 == TEX0.TBP0) - { - m_ds.splice(m_ds.begin(), m_ds, i); - - ds = ds2; - - ds->m_TEX0 = TEX0; - - ds->Update(); - - break; - } - } - } - - if(ds == NULL) - { - ds = CreateDepthStencil(); - - ds->m_TEX0 = TEX0; - - if(!ds->Create(w, h)) - { - delete ds; - - return NULL; - } - - m_ds.push_front(ds); - } - - if(m_renderer->m_context->DepthWrite()) - { - ds->m_used = true; - } - - return ds; -} - -GSTextureCache::GSCachedTexture* GSTextureCache::GetTexture(const GSVector4i& r) -{ - const GIFRegTEX0& TEX0 = m_renderer->m_context->TEX0; - const GIFRegCLAMP& CLAMP = m_renderer->m_context->CLAMP; - const GIFRegTEXA& TEXA = m_renderer->m_env.TEXA; - - const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - const uint32* clut = m_renderer->m_mem.m_clut; - - GSCachedTexture* t = NULL; - - for(list::iterator i = m_tex.begin(); i != m_tex.end(); i++) - { - GSCachedTexture* t2 = *i; - - if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH - { - continue; - } - - if((psm.trbpp == 16 || psm.trbpp == 24) && TEX0.TCC && TEXA != t2->m_TEXA) - { - continue; - } - - if(psm.pal > 0 && !(t2->m_TEX0.CPSM == TEX0.CPSM && GSVector4i::compare(t2->m_clut, clut, psm.pal * sizeof(clut[0])))) - { - continue; - } - - t = t2; - - m_tex.splice(m_tex.begin(), m_tex, i); - - break; - } - - if(t == NULL) - { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); i++) - { - GSRenderTarget* rt = *i; - - if(rt->m_dirty.empty() && GSUtil::HasSharedBits(rt->m_TEX0.TBP0, rt->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) - { - t = CreateTexture(); - - if(!t->Create(rt)) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - - break; - } - } - } - - if(t == NULL) - { - for(list::iterator i = m_ds.begin(); i != m_ds.end(); i++) - { - GSDepthStencil* ds = *i; - - if(ds->m_dirty.empty() && ds->m_used && GSUtil::HasSharedBits(ds->m_TEX0.TBP0, ds->m_TEX0.PSM, TEX0.TBP0, TEX0.PSM)) - { - t = CreateTexture(); - - if(!t->Create(ds)) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - - break; - } - } - } - - if(t == NULL) - { - t = CreateTexture(); - - if(!t->Create()) - { - delete t; - - return NULL; - } - - m_tex.push_front(t); - } - - if(psm.pal > 0) - { - int size = psm.pal * sizeof(clut[0]); - - if(t->m_palette) - { - if(t->m_initpalette) - { - memcpy(t->m_clut, clut, size); - t->m_palette->Update(GSVector4i(0, 0, psm.pal, 1), t->m_clut, size); - t->m_initpalette = false; - } - else - { - if(GSVector4i::update(t->m_clut, clut, size)) - { - t->m_palette->Update(GSVector4i(0, 0, psm.pal, 1), t->m_clut, size); - } - } - } - else - { - memcpy(t->m_clut, clut, size); - } - } - - t->Update(r); - - m_tex_used = true; - - return t; -} - -void GSTextureCache::InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) -{ - for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) - { - list::iterator j = i++; - - GSCachedTexture* t = *j; - - if(GSUtil::HasSharedBits(FRAME.Block(), FRAME.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM) - || GSUtil::HasSharedBits(ZBUF.Block(), ZBUF.PSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) - { - m_tex.erase(j); - - delete t; - } - } -} - -void GSTextureCache::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) +void GSTextureCache::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& rect, bool target) { bool found = false; - for(list::iterator i = m_tex.begin(); i != m_tex.end(); ) + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[BITBLTBUF.DPSM]; + + uint32 bp = BITBLTBUF.DBP; + uint32 bw = BITBLTBUF.DBW; + + GSVector2i bs = (bp & 31) == 0 ? psm.pgs : psm.bs; + + GSVector4i r = rect.ralign(bs); + + if(!target) { - list::iterator j = i++; + const hash_map& map = m_src.m_map[bp >> 5]; - GSCachedTexture* t = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) + for(hash_map::const_iterator i = map.begin(); i != map.end(); ) { - if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) - { - t->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); + hash_map::const_iterator j = i++; - found = true; - } - else + Source* s = j->first; + + if(GSUtil::HasSharedBits(bp, BITBLTBUF.DPSM, s->m_TEX0.TBP0, s->m_TEX0.PSM)) { - m_tex.erase(j); - delete t; + m_src.RemoveAt(s); } } - else if(GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, t->m_TEX0.PSM)) + } + + for(int y = r.top; y < r.bottom; y += bs.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += bs.x) { - if(BITBLTBUF.DBW == t->m_TEX0.TBW && !t->m_rendered) + uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; + + if(page < MAX_PAGES) { - int rowsize = (int)BITBLTBUF.DBW * 8192; - int offset = ((int)BITBLTBUF.DBP - (int)t->m_TEX0.TBP0) * 256; + const hash_map& map = m_src.m_map[page]; - if(rowsize > 0 && offset % rowsize == 0) + for(hash_map::const_iterator i = map.begin(); i != map.end(); ) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; + hash_map::const_iterator j = i++; - GSVector4i r2(r.left, r.top + y, r.right, r.bottom + y); + Source* s = j->first; - int w = 1 << t->m_TEX0.TW; - int h = 1 << t->m_TEX0.TH; - - if(r2.bottom > 0 && r2.top < h && r2.right > 0 && r2.left < w) + if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, s->m_TEX0.PSM)) { - t->m_dirty.push_back(GSDirtyRect(r2, BITBLTBUF.DPSM)); + if(!s->m_target) + { + s->m_valid[page] = 0; + + found = true; + } + else + { + if(s->m_TEX0.TBP0 == bp) + { + m_src.RemoveAt(s); + } + } } } } } } - for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) + if(!target) return; + + for(int type = 0; type < 2; type++) { - list::iterator j = i++; - - GSRenderTarget* rt = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); ) { - if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM)) - { - rt->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); - rt->m_TEX0.TBW = BITBLTBUF.DBW; - } - else - { - m_rt.erase(j); - delete rt; - continue; - } - } + list::iterator j = i++; - if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, rt->m_TEX0.PSM) && BITBLTBUF.DBP < rt->m_TEX0.TBP0) - { - uint32 rowsize = BITBLTBUF.DBW * 8192; - uint32 offset = (uint32)((rt->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); + Target* t = *j; - if(rowsize > 0 && offset % rowsize == 0) + if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - - if(r.bottom > y) + if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, t->m_TEX0.PSM)) { - // TODO: do not add this rect above too - rt->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); - rt->m_TEX0.TBW = BITBLTBUF.DBW; + t->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); + t->m_TEX0.TBW = BITBLTBUF.DBW; + } + else + { + m_dst[type].erase(j); + delete t; continue; } } - } - } - // copypaste for ds - - for(list::iterator i = m_ds.begin(); i != m_ds.end(); ) - { - list::iterator j = i++; - - GSDepthStencil* ds = *j; - - if(GSUtil::HasSharedBits(BITBLTBUF.DBP, BITBLTBUF.DPSM, ds->m_TEX0.TBP0, ds->m_TEX0.PSM)) - { - if(!found && GSUtil::HasCompatibleBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM)) + if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, t->m_TEX0.PSM) && BITBLTBUF.DBP < t->m_TEX0.TBP0) { - ds->m_dirty.push_back(GSDirtyRect(r, BITBLTBUF.DPSM)); - ds->m_TEX0.TBW = BITBLTBUF.DBW; - } - else - { - m_ds.erase(j); - delete ds; - continue; - } - } + uint32 rowsize = BITBLTBUF.DBW * 8192; + uint32 offset = (uint32)((t->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); - if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, ds->m_TEX0.PSM) && BITBLTBUF.DBP < ds->m_TEX0.TBP0) - { - uint32 rowsize = BITBLTBUF.DBW * 8192; - uint32 offset = (uint32)((ds->m_TEX0.TBP0 - BITBLTBUF.DBP) * 256); - - if(rowsize > 0 && offset % rowsize == 0) - { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; - - if(r.bottom > y) + if(rowsize > 0 && offset % rowsize == 0) { - // TODO: do not add this rect above too - ds->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); - ds->m_TEX0.TBW = BITBLTBUF.DBW; - continue; + int y = GSLocalMemory::m_psm[BITBLTBUF.DPSM].pgs.y * offset / rowsize; + + if(r.bottom > y) + { + // TODO: do not add this rect above too + t->m_dirty.push_back(GSDirtyRect(GSVector4i(r.left, r.top - y, r.right, r.bottom - y), BITBLTBUF.DPSM)); + t->m_TEX0.TBW = BITBLTBUF.DBW; + continue; + } } } } @@ -484,31 +345,33 @@ void GSTextureCache::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const void GSTextureCache::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r) { - for(list::iterator i = m_rt.begin(); i != m_rt.end(); ) + for(list::iterator i = m_dst[RenderTarget].begin(); i != m_dst[RenderTarget].end(); ) { - list::iterator j = i++; + list::iterator j = i++; - GSRenderTarget* rt = *j; + Target* t = *j; - if(GSUtil::HasSharedBits(BITBLTBUF.SBP, BITBLTBUF.SPSM, rt->m_TEX0.TBP0, rt->m_TEX0.PSM)) + if(GSUtil::HasSharedBits(BITBLTBUF.SBP, BITBLTBUF.SPSM, t->m_TEX0.TBP0, t->m_TEX0.PSM)) { - if(GSUtil::HasCompatibleBits(BITBLTBUF.SPSM, rt->m_TEX0.PSM)) + if(GSUtil::HasCompatibleBits(BITBLTBUF.SPSM, t->m_TEX0.PSM)) { - rt->Read(r); + t->Read(r); + return; } - else if(BITBLTBUF.SPSM == PSM_PSMCT32 && (rt->m_TEX0.PSM == PSM_PSMCT16 || rt->m_TEX0.PSM == PSM_PSMCT16S)) + else if(BITBLTBUF.SPSM == PSM_PSMCT32 && (t->m_TEX0.PSM == PSM_PSMCT16 || t->m_TEX0.PSM == PSM_PSMCT16S)) { // ffx-2 riku changing to her default (shoots some reflecting glass at the end), 16-bit rt read as 32-bit - rt->Read(GSVector4i(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); + t->Read(GSVector4i(r.left, r.top, r.right, r.top + (r.bottom - r.top) * 2)); + return; } else { - m_rt.erase(j); - delete rt; - continue; + m_dst[RenderTarget].erase(j); + + delete t; } } } @@ -533,7 +396,7 @@ void GSTextureCache::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const if(rowsize > 0 && offset % rowsize == 0) { - int y = m_renderer->m_mem.m_psm[BITBLTBUF.SPSM].pgs.y * offset / rowsize; + int y = GSLocalMemory::m_psm[BITBLTBUF.SPSM].pgs.y * offset / rowsize; if(y < ymin && y < 512) { @@ -555,57 +418,291 @@ void GSTextureCache::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const void GSTextureCache::IncAge() { - RecycleByAge(m_tex, m_tex_used ? 2 : 30); - RecycleByAge(m_rt); - RecycleByAge(m_ds); + int maxage = m_src.m_used ? 3 : 30; - m_tex_used = false; + for(hash_map::iterator i = m_src.m_surfaces.begin(); i != m_src.m_surfaces.end(); ) + { + hash_map::iterator j = i++; + + Source* s = j->first; + + if(++s->m_age > maxage) + { + m_src.RemoveAt(s); + } + } + + m_src.m_used = false; + + maxage = 3; + + for(int type = 0; type < 2; type++) + { + for(list::iterator i = m_dst[type].begin(); i != m_dst[type].end(); ) + { + list::iterator j = i++; + + Target* t = *j; + + if(++t->m_age > maxage) + { + m_dst[type].erase(j); + + delete t; + } + } + } } -// GSTextureCache::GSSurface +// GSTextureCache::Surface -GSTextureCache::GSSurface::GSSurface(GSRenderer* r) +GSTextureCache::Surface::Surface(GSRenderer* r) : m_renderer(r) , m_texture(NULL) - , m_palette(NULL) - , m_initpalette(false) , m_age(0) { m_TEX0.TBP0 = (uint32)~0; } -GSTextureCache::GSSurface::~GSSurface() +GSTextureCache::Surface::~Surface() { m_renderer->m_dev->Recycle(m_texture); - m_renderer->m_dev->Recycle(m_palette); - - m_texture = NULL; - m_palette = NULL; } -void GSTextureCache::GSSurface::Update() +void GSTextureCache::Surface::Update() { m_age = 0; } -// GSTextureCache::GSRenderTarget +// GSTextureCache::Source -GSTextureCache::GSRenderTarget::GSRenderTarget(GSRenderer* r) - : GSSurface(r) - , m_used(true) +GSTextureCache::Source::Source(GSRenderer* r) + : Surface(r) + , m_palette(NULL) + , m_initpalette(false) + , m_bpp(0) + , m_target(false) +{ + memset(m_valid, 0, sizeof(m_valid)); + + m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); + + memset(m_clut, 0, sizeof(m_clut)); +} + +GSTextureCache::Source::~Source() +{ + m_renderer->m_dev->Recycle(m_palette); + + _aligned_free(m_clut); +} + +void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect) +{ + __super::Update(); + + if(m_target) + { + return; + } + + m_TEX0 = TEX0; + m_TEXA = TEXA; + + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_TEX0.PSM]; + + GSVector2i s = psm.bs; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + + GSVector4i tr(0, 0, tw, th); + + GSVector4i r = rect.ralign(s); + + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; + + uint32 blocks = 0; + + // TODO + static uint8* buff = (uint8*)_aligned_malloc(1024 * 16 * sizeof(uint32), 16); // max decompressed size for a row of blocks (1024 x 16, 4bpp) + + int pitch = max(tw, s.x) * sizeof(uint32); + + GSLocalMemory::readTexture rtx = psm.rtx; + + const GSLocalMemory& mem = m_renderer->m_mem; + + // TODO: bw == 0 (sfex) + + if(tw <= (bw << 6)) + { + // r.right = min(r.right, bw << 6); + + for(int y = r.top; y < r.bottom; y += s.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + int left = r.left; + int right = r.left; + + for(int x = r.left; x < r.right; x += s.x) + { + uint32 block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); + + if((m_valid[row] & col) == 0) + { + m_valid[row] |= col; + + if(right < x) + { + Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + + left = right = x; + } + + right += s.x; + + blocks++; + } + } + } + + if(left < right) + { + Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + } + } + } + else + { + // unfortunatelly a block may be part of the same texture multiple times at different places (tw 1024 > tbw 640, between 640 -> 1024 it is repeated from the next row), + // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side + + // TODO: still bogus if those repeated parts aren't fetched together + + for(int y = r.top; y < r.bottom; y += s.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + int left = r.left; + int right = r.left; + + for(int x = r.left; x < r.right; x += s.x) + { + uint32 block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); + + if((m_valid[row] & col) == 0) + { + if(right < x) + { + Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + + left = right = x; + } + + right += s.x; + + blocks++; + } + } + } + + if(left < right) + { + Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + } + } + + for(int y = r.top; y < r.bottom; y += s.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += s.x) + { + uint32 block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); + + m_valid[row] |= col; + } + } + } + } + + //_aligned_free(buff); + + m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, s.x * s.y * sizeof(uint32) * blocks); +} + +void GSTextureCache::Source::Write(const GSVector4i& r, uint8* buff, int pitch) +{ + if(r.rempty()) return; + + GSLocalMemory::readTexture rtx = GSLocalMemory::m_psm[m_TEX0.PSM].rtx; + + GSTexture::GSMap m; + + if(m_texture->Map(m, &r)) + { + (m_renderer->m_mem.*rtx)(r, m.bits, m.pitch, m_TEX0, m_TEXA); + + m_texture->Unmap(); + } + else + { + (m_renderer->m_mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); + + m_texture->Update(r, buff, pitch); + } +} + +// GSTextureCache::Target + +GSTextureCache::Target::Target(GSRenderer* r) + : Surface(r) + , m_type(-1) + , m_used(false) { } -bool GSTextureCache::GSRenderTarget::Create(int w, int h) +bool GSTextureCache::Target::Create(int w, int h, int type) { + ASSERT(m_texture == NULL); + // FIXME: initial data should be unswizzled from local mem in Update() if dirty - m_texture = m_renderer->m_dev->CreateRenderTarget(w, h); + m_type = type; + + if(type == RenderTarget) + { + m_texture = m_renderer->m_dev->CreateRenderTarget(w, h); + + m_used = true; + } + else if(type == DepthStencil) + { + m_texture = m_renderer->m_dev->CreateDepthStencil(w, h); + } return m_texture != NULL; } -void GSTextureCache::GSRenderTarget::Update() +void GSTextureCache::Target::Update() { __super::Update(); @@ -615,200 +712,112 @@ void GSTextureCache::GSRenderTarget::Update() if(r.rempty()) return; - int w = r.width(); - int h = r.height(); - - if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) + if(m_type == RenderTarget) { - GIFRegTEXA TEXA; + int w = r.width(); + int h = r.height(); - TEXA.AEM = 1; - TEXA.TA0 = 0; - TEXA.TA1 = 0x80; - - GSTexture::GSMap m; - - if(t->Map(m)) + if(GSTexture* t = m_renderer->m_dev->CreateTexture(w, h)) { - m_renderer->m_mem.ReadTexture(r, m.bits, m.pitch, m_TEX0, TEXA); + GIFRegTEXA TEXA; - t->Unmap(); - } - else - { - static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - - int pitch = ((w + 3) & ~3) * 4; + TEXA.AEM = 1; + TEXA.TA0 = 0; + TEXA.TA1 = 0x80; - m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA); - - t->Update(r.rsize(), buff, pitch); - } + GSTexture::GSMap m; - // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); + if(t->Map(m)) + { + m_renderer->m_mem.ReadTexture(r, m.bits, m.pitch, m_TEX0, TEXA); - m_renderer->m_dev->StretchRect(t, m_texture, GSVector4(r) * GSVector4(m_texture->m_scale).xyxy()); + t->Unmap(); + } + else + { + static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); + + int pitch = ((w + 3) & ~3) * 4; - m_renderer->m_dev->Recycle(t); - } -} + m_renderer->m_mem.ReadTexture(r, buff, pitch, m_TEX0, TEXA); + + t->Update(r.rsize(), buff, pitch); + } -// GSTextureCache::GSDepthStencil + // m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, w * h * 4); -GSTextureCache::GSDepthStencil::GSDepthStencil(GSRenderer* r) - : GSSurface(r) - , m_used(false) -{ -} + m_renderer->m_dev->StretchRect(t, m_texture, GSVector4(r) * GSVector4(m_texture->m_scale).xyxy()); -bool GSTextureCache::GSDepthStencil::Create(int w, int h) -{ - // FIXME: initial data should be unswizzled from local mem in Update() if dirty - - m_texture = m_renderer->m_dev->CreateDepthStencil(w, h); - - return m_texture != NULL; -} - -void GSTextureCache::GSDepthStencil::Update() -{ - __super::Update(); - - GSVector4i r = m_dirty.GetDirtyRectAndClear(m_TEX0, m_texture->GetSize()); - - if(r.rempty()) return; - - // TODO - - m_renderer->m_dev->ClearDepth(m_texture, 0); -} - -// GSTextureCache::GSCachedTexture - -GSTextureCache::GSCachedTexture::GSCachedTexture(GSRenderer* r) - : GSSurface(r) - , m_bpp(0) - , m_rendered(false) -{ - m_valid = GSVector4i::zero(); - - m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); - - memset(m_clut, 0, sizeof(m_clut)); -} - -GSTextureCache::GSCachedTexture::~GSCachedTexture() -{ - _aligned_free(m_clut); -} - -void GSTextureCache::GSCachedTexture::Update(const GSVector4i& rect) -{ - __super::Update(); - - if(m_rendered) - { - return; - } - - GSVector4i r = rect; - - if(!GetDirtyRect(r)) - { - return; - } - - m_valid = m_valid.runion(r); - - GSTexture::GSMap m; - - if(m_texture->Map(m, &r)) - { - // in dx9 managed textures can be written directly, less copying is faster, but still not as fast as dx10's UpdateResource - - m_renderer->m_mem.ReadTexture(r, m.bits, m.pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); - - m_texture->Unmap(); - } - else - { - static uint8* buff = (uint8*)::_aligned_malloc(1024 * 1024 * 4, 16); - - int pitch = ((r.width() + 3) & ~3) * 4; - - m_renderer->m_mem.ReadTexture(r, buff, pitch, m_renderer->m_context->TEX0, m_renderer->m_env.TEXA); - - m_texture->Update(r, buff, pitch); - } - - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, r.width() * r.height() * 4); -} - -bool GSTextureCache::GSCachedTexture::GetDirtyRect(GSVector4i& r) -{ - int w = 1 << m_TEX0.TW; - int h = 1 << m_TEX0.TH; - - GSVector4i tr(0, 0, w, h); - - for(list::iterator i = m_dirty.begin(); i != m_dirty.end(); i++) - { - const GSVector4i& dirty = i->GetDirtyRect(m_TEX0).rintersect(tr); - - if(!m_valid.rintersect(dirty).rempty()) - { - // find the rect having the largest area, outside dirty, inside m_valid - - GSVector4i left(m_valid.left, m_valid.top, min(m_valid.right, dirty.left), m_valid.bottom); - GSVector4i top(m_valid.left, m_valid.top, m_valid.right, min(m_valid.bottom, dirty.top)); - GSVector4i right(max(m_valid.left, dirty.right), m_valid.top, m_valid.right, m_valid.bottom); - GSVector4i bottom(m_valid.left, max(m_valid.top, dirty.bottom), m_valid.right, m_valid.bottom); - - int leftsize = !left.rempty() ? left.width() * left.height() : 0; - int topsize = !top.rempty() ? top.width() * top.height() : 0; - int rightsize = !right.rempty() ? right.width() * right.height() : 0; - int bottomsize = !bottom.rempty() ? bottom.width() * bottom.height() : 0; - - // TODO: sort - - m_valid = - leftsize > 0 ? left : - topsize > 0 ? top : - rightsize > 0 ? right : - bottomsize > 0 ? bottom : - GSVector4i::zero(); + m_renderer->m_dev->Recycle(t); } } + else if(m_type == DepthStencil) + { + // do the most likely thing a direct write would do, clear it - m_dirty.clear(); + m_renderer->m_dev->ClearDepth(m_texture, 0); + } +} - if(GSUtil::IsRectInRect(r, m_valid)) +// GSTextureCache::SourceMap + +void GSTextureCache::SourceMap::Add(Source* s, const GIFRegTEX0& TEX0) +{ + m_surfaces[s] = true; + + int tw = 1 << TEX0.TW; + int th = 1 << TEX0.TH; + + uint32 bp = TEX0.TBP0; + uint32 bw = TEX0.TBW; + + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; + + GSVector2i bs = (bp & 31) == 0 ? psm.pgs : psm.bs; + + int blocks = 0; + + for(int y = 0; y < th; y += bs.y) { - return false; + uint32 base = psm.bn(0, y, bp, bw); + + for(int x = 0; x < tw; x += bs.x) + { + uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; + + if(page < MAX_PAGES) + { + m_map[page][s] = true; + + s->m_pages.push_back(page); + } + } } - else if(GSUtil::IsRectInRectH(r, m_valid) && (r.left >= m_valid.left || r.right <= m_valid.right)) +} + +void GSTextureCache::SourceMap::RemoveAll() +{ + for(hash_map::iterator i = m_surfaces.begin(); i != m_surfaces.end(); i++) { - r.top = m_valid.top; - r.bottom = m_valid.bottom; - if(r.left < m_valid.left) r.right = m_valid.left; - else r.left = m_valid.right; // if(r.right > m_valid.right) - } - else if(GSUtil::IsRectInRectV(r, m_valid) && (r.top >= m_valid.top || r.bottom <= m_valid.bottom)) - { - r.left = m_valid.left; - r.right = m_valid.right; - if(r.top < m_valid.top) r.bottom = m_valid.top; - else r.top = m_valid.bottom; // if(r.bottom > m_valid.bottom) - } - else - { - r = r.runion(m_valid); + delete i->first; } - if(r.rempty()) + m_surfaces.clear(); + + for(int i = 0; i < MAX_PAGES; i++) { - return false; + m_map[i].clear(); + } +} + +void GSTextureCache::SourceMap::RemoveAt(Source* s) +{ + m_surfaces.erase(s); + + for(list::iterator i = s->m_pages.begin(); i != s->m_pages.end(); i++) + { + m_map[*i].erase(s); } - return true; -} \ No newline at end of file + delete s; +} diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 300d648bfd..5857ed5358 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -26,101 +26,86 @@ class GSTextureCache { public: - class GSSurface : public GSAlignedClass<16> + enum {RenderTarget, DepthStencil}; + + class Surface : public GSAlignedClass<16> { protected: GSRenderer* m_renderer; public: GSTexture* m_texture; - GSTexture* m_palette; - bool m_initpalette; - int m_age; - GSDirtyRectList m_dirty; GIFRegTEX0 m_TEX0; GIFRegTEXA m_TEXA; + int m_age; - explicit GSSurface(GSRenderer* r); - virtual ~GSSurface(); + public: + explicit Surface(GSRenderer* r); + virtual ~Surface(); virtual void Update(); }; - class GSRenderTarget : public GSSurface + class Target; + + class Source : public Surface { public: - bool m_used; - - explicit GSRenderTarget(GSRenderer* r); - - void Update(); - - virtual bool Create(int w, int h); - virtual void Read(const GSVector4i& r) = 0; - }; - - class GSDepthStencil : public GSSurface - { - public: - bool m_used; - - explicit GSDepthStencil(GSRenderer* renderer); - - void Update(); - - virtual bool Create(int w, int h); - }; - - class GSCachedTexture : public GSSurface - { - protected: - bool GetDirtyRect(GSVector4i& r); - - public: - uint32* m_clut; // * - GSVector4i m_valid; + GSTexture* m_palette; + bool m_initpalette; + list m_pages; + uint32 m_valid[MAX_PAGES]; // each uint32 bits map to the 32 blocks of that page + uint32* m_clut; int m_bpp; - bool m_rendered; + bool m_target; - explicit GSCachedTexture(GSRenderer* renderer); - virtual ~GSCachedTexture(); - - void Update(const GSVector4i& rect); + public: + explicit Source(GSRenderer* renderer); + virtual ~Source(); virtual bool Create() = 0; - virtual bool Create(GSRenderTarget* rt) = 0; - virtual bool Create(GSDepthStencil* ds) = 0; + virtual bool Create(Target* dst) = 0; + virtual void Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect); + + void Write(const GSVector4i& r, uint8* buff, int pitch); + }; + + class Target : public Surface + { + public: + int m_type; + bool m_used; + GSDirtyRectList m_dirty; + + public: + explicit Target(GSRenderer* r); + + virtual bool Create(int w, int h, int type); + virtual void Update(); + virtual void Read(const GSVector4i& r) = 0; }; protected: GSRenderer* m_renderer; - list m_rt; - list m_ds; - list m_tex; - - bool m_tex_used; - - template void RecycleByAge(list& l, int maxage = 60) + struct SourceMap { - for(list::iterator i = l.begin(); i != l.end(); ) - { - list::iterator j = i++; + hash_map m_surfaces; + hash_map m_map[MAX_PAGES]; + bool m_used; - T* t = *j; + SourceMap() : m_used(false) {} - if(++t->m_age > maxage) - { - l.erase(j); + void Add(Source* s, const GIFRegTEX0& TEX0); + void RemoveAll(); + void RemoveAt(Source* s); - delete t; - } - } - } + } m_src; - virtual GSRenderTarget* CreateRenderTarget() = 0; - virtual GSDepthStencil* CreateDepthStencil() = 0; - virtual GSCachedTexture* CreateTexture() = 0; + list m_dst[2]; + + virtual Source* CreateSource() = 0; + virtual Target* CreateTarget() = 0; public: GSTextureCache(GSRenderer* r); @@ -128,12 +113,10 @@ public: void RemoveAll(); - GSRenderTarget* GetRenderTarget(const GIFRegTEX0& TEX0, int w, int h, bool fb = false); - GSDepthStencil* GetDepthStencil(const GIFRegTEX0& TEX0, int w, int h); - GSCachedTexture* GetTexture(const GSVector4i& r); + Source* LookupSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r); + Target* LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, bool fb = false); - void InvalidateTextures(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF); - void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); + void InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r, bool target = true); void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r); void IncAge(); diff --git a/plugins/GSdx/GSTextureCache10.cpp b/plugins/GSdx/GSTextureCache10.cpp index 0c20e24c24..e4914cb09f 100644 --- a/plugins/GSdx/GSTextureCache10.cpp +++ b/plugins/GSdx/GSTextureCache10.cpp @@ -29,16 +29,207 @@ GSTextureCache10::GSTextureCache10(GSRenderer* r) { } -// GSRenderTargetHW10 +// Source10 -void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) +bool GSTextureCache10::Source10::Create() { + // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + m_bpp = 0; + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); + + return m_texture != NULL; +} + +bool GSTextureCache10::Source10::Create(Target* dst) +{ + m_target = true; + + if(dst->m_type != RenderTarget) + { + // TODO + + return false; + } + + // TODO: clean up this mess + + dst->Update(); + + // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + int tp = (int)m_TEX0.TW << 6; + + // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) + + int w = (int)(dst->m_texture->m_scale.x * tw); + int h = (int)(dst->m_texture->m_scale.y * th); + + GSVector2i dstsize = dst->m_texture->GetSize(); + + // pitch conversion + + if(dst->m_TEX0.TBW != m_TEX0.TBW) // && dst->m_TEX0.PSM == m_TEX0.PSM + { + // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) + + // ASSERT(dst->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateRenderTarget(dstsize.x, dstsize.y); + + GSVector4 size = GSVector4(dstsize).xyxy(); + GSVector4 scale = GSVector4(dst->m_texture->m_scale).xyxy(); + + int bw = 64; + int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + + GSVector4i br(0, 0, bw, bh); + + int sw = (int)dst->m_TEX0.TBW << 6; + + int dw = (int)m_TEX0.TBW << 6; + int dh = 1 << m_TEX0.TH; + + if(sw != 0) + for(int dy = 0; dy < dh; dy += bh) + { + for(int dx = 0; dx < dw; dx += bw) + { + int o = dy * dw / bh + dx; + + int sx = o % sw; + int sy = o / sw; + + GSVector4 sr = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dr = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; + + m_renderer->m_dev->StretchRect(dst->m_texture, sr, m_texture, dr); + + // TODO: this is quite a lot of StretchRect, do it with one Draw + } + } + } + else if(tw < tp) + { + // FIXME: timesplitters blurs the render target by blending itself over a couple of times + + if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) + { + return false; + } + } + + // width/height conversion + + GSVector2 scale = dst->m_texture->m_scale; + + GSVector4 dr(0, 0, w, h); + + if(w > dstsize.x) + { + scale.x = (float)dstsize.x / tw; + dr.z = (float)dstsize.x * scale.x / dst->m_texture->m_scale.x; + w = dstsize.x; + } + + if(h > dstsize.y) + { + scale.y = (float)dstsize.y / th; + dr.w = (float)dstsize.y * scale.y / dst->m_texture->m_scale.y; + h = dstsize.y; + } + + GSVector4 sr(0, 0, w, h); + + GSTexture* st = m_texture ? m_texture : dst->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); + + if(!m_texture) + { + m_texture = dt; + } + + if((sr == dr).alltrue()) + { + D3D10_BOX box = {0, 0, 0, w, h, 1}; + + (*(GSDevice10*)m_renderer->m_dev)->CopySubresourceRegion(*(GSTexture10*)dt, 0, 0, 0, 0, *(GSTexture10*)st, 0, &box); + } + else + { + sr.z /= st->GetWidth(); + sr.w /= st->GetHeight(); + + m_renderer->m_dev->StretchRect(st, sr, dt, dr); + } + + if(dt != m_texture) + { + m_renderer->m_dev->Recycle(m_texture); + + m_texture = dt; + } + + m_texture->m_scale = scale; + + switch(m_TEX0.PSM) + { + case PSM_PSMCT32: + m_bpp = 0; + break; + case PSM_PSMCT24: + m_bpp = 1; + break; + case PSM_PSMCT16: + case PSM_PSMCT16S: + m_bpp = 2; + break; + case PSM_PSMT8H: + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); + m_initpalette = true; + break; + case PSM_PSMT4HL: + case PSM_PSMT4HH: + ASSERT(0); // TODO + break; + } + + return true; +} + + +// Target10 + +void GSTextureCache10::Target10::Read(const GSVector4i& r) +{ + if(m_type != RenderTarget) + { + // TODO + + return; + } + if(m_TEX0.PSM != PSM_PSMCT32 && m_TEX0.PSM != PSM_PSMCT24 && m_TEX0.PSM != PSM_PSMCT16 && m_TEX0.PSM != PSM_PSMCT16S) { //ASSERT(0); + return; } @@ -121,272 +312,3 @@ void GSTextureCache10::GSRenderTargetHW10::Read(const GSVector4i& r) m_renderer->m_dev->Recycle(offscreen); } } - -// GSTextureHW10 - -bool GSTextureCache10::GSCachedTextureHW10::Create() -{ - // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - m_bpp = 0; - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); - - return m_texture != NULL; -} - -bool GSTextureCache10::GSCachedTextureHW10::Create(GSRenderTarget* rt) -{ - m_rendered = true; - - // TODO: clean up this mess - - rt->Update(); - - // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - int tw = 1 << m_TEX0.TW; - int th = 1 << m_TEX0.TH; - int tp = (int)m_TEX0.TW << 6; - - // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - - int w = (int)(rt->m_texture->m_scale.x * tw); - int h = (int)(rt->m_texture->m_scale.y * th); - - GSVector2i rtsize = rt->m_texture->GetSize(); - - // pitch conversion - - if(rt->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM - { - // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) - - // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); - - GSVector4 size = GSVector4(rtsize).xyxy(); - GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); - - int bw = 64; - int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; - - GSVector4i br(0, 0, bw, bh); - - int sw = (int)rt->m_TEX0.TBW << 6; - - int dw = (int)m_TEX0.TBW << 6; - int dh = 1 << m_TEX0.TH; - - if(sw != 0) - for(int dy = 0; dy < dh; dy += bh) - { - for(int dx = 0; dx < dw; dx += bw) - { - int o = dy * dw / bh + dx; - - int sx = o % sw; - int sy = o / sw; - - GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; - GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; - - m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); - - // TODO: this is quite a lot of StretchRect, do it with one Draw - } - } - } - else if(tw < tp) - { - // FIXME: timesplitters blurs the render target by blending itself over a couple of times - - if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) - { - return false; - } - } - - // width/height conversion - - GSVector2 scale = rt->m_texture->m_scale; - - GSVector4 dst(0, 0, w, h); - - if(w > rtsize.x) - { - scale.x = (float)rtsize.x / tw; - dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; - w = rtsize.x; - } - - if(h > rtsize.y) - { - scale.y = (float)rtsize.y / th; - dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; - h = rtsize.y; - } - - GSVector4 src(0, 0, w, h); - - GSTexture* st = m_texture ? m_texture : rt->m_texture; - GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); - - if(!m_texture) - { - m_texture = dt; - } - - if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) - { - D3D10_BOX box = {0, 0, 0, w, h, 1}; - - (*(GSDevice10*)m_renderer->m_dev)->CopySubresourceRegion(*(GSTexture10*)dt, 0, 0, 0, 0, *(GSTexture10*)st, 0, &box); - } - else - { - src.z /= st->GetWidth(); - src.w /= st->GetHeight(); - - m_renderer->m_dev->StretchRect(st, src, dt, dst); - } - - if(dt != m_texture) - { - m_renderer->m_dev->Recycle(m_texture); - - m_texture = dt; - } - - m_texture->m_scale = scale; - - switch(m_TEX0.PSM) - { - case PSM_PSMCT32: - m_bpp = 0; - break; - case PSM_PSMCT24: - m_bpp = 1; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 2; - break; - case PSM_PSMT8H: - m_bpp = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1); - m_initpalette = true; - break; - case PSM_PSMT4HL: - case PSM_PSMT4HH: - ASSERT(0); // TODO - break; - } - - return true; -} - -bool GSTextureCache10::GSCachedTextureHW10::Create(GSDepthStencil* ds) -{ - m_rendered = true; - - return false; -/* - // TODO: clean up this mess - - ds->Update(); - - // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - int tw = 1 << m_TEX0.TW; - int th = 1 << m_TEX0.TH; - int tp = (int)m_TEX0.TW << 6; - - // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - - int w = (int)(ds->m_texture->m_scale.x * tw); - int h = (int)(ds->m_texture->m_scale.y * th); - - GSVector2i dssize = ds->m_texture->GetSize(); - - // pitch conversion - - if(ds->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM - { - ASSERT(0); - } - else if(tw < tp) - { - } - - // width/height conversion - - GSVector2 scale = ds->m_texture->m_scale; - - GSVector4 dst(0, 0, w, h); - - if(w > dssize.x) - { - scale.x = (float)dssize.x / tw; - dst.z = (float)dssize.x * scale.x / ds->m_texture->m_scale.x; - w = dssize.x; - } - - if(h > dssize.y) - { - scale.y = (float)dssize.y / th; - dst.w = (float)dssize.y * scale.y / ds->m_texture->m_scale.y; - h = dssize.y; - } - - m_texture = m_renderer->m_dev->CreateRenderTarget(w, h); - - GSVector4 src(0, 0, w, h); - - src.z /= ds->m_texture->GetWidth(); - src.w /= ds->m_texture->GetHeight(); - - m_renderer->m_dev->StretchRect(ds->m_texture, src, m_texture, dst, 7); - - m_texture->m_scale = scale; - - switch(m_TEX0.PSM) - { - case PSM_PSMCT32: - m_bpp = 0; - break; - case PSM_PSMCT24: - m_bpp = 1; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 2; - break; - case PSM_PSMT8H: - m_bpp = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1); - m_initpalette = true; - break; - case PSM_PSMT4HL: - case PSM_PSMT4HH: - ASSERT(0); // TODO - break; - } - - return true; -*/ -} diff --git a/plugins/GSdx/GSTextureCache10.h b/plugins/GSdx/GSTextureCache10.h index bc480bea4c..857d37c4e7 100644 --- a/plugins/GSdx/GSTextureCache10.h +++ b/plugins/GSdx/GSTextureCache10.h @@ -26,34 +26,26 @@ class GSTextureCache10 : public GSTextureCache { - class GSRenderTargetHW10 : public GSRenderTarget + class Source10 : public Source { public: - explicit GSRenderTargetHW10(GSRenderer* r) : GSRenderTarget(r) {} + explicit Source10(GSRenderer* r) : Source(r) {} + + bool Create(); + bool Create(Target* dst); + }; + + class Target10 : public Target + { + public: + explicit Target10(GSRenderer* r) : Target(r) {} void Read(const GSVector4i& r); }; - class GSDepthStencilHW10 : public GSDepthStencil - { - public: - explicit GSDepthStencilHW10(GSRenderer* r) : GSDepthStencil(r) {} - }; - - class GSCachedTextureHW10 : public GSCachedTexture - { - public: - explicit GSCachedTextureHW10(GSRenderer* r) : GSCachedTexture(r) {} - - bool Create(); - bool Create(GSRenderTarget* rt); - bool Create(GSDepthStencil* ds); - }; - protected: - GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetHW10(m_renderer);} - GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilHW10(m_renderer);} - GSCachedTexture* CreateTexture() {return new GSCachedTextureHW10(m_renderer);} + Source* CreateSource() {return new Source10(m_renderer);} + Target* CreateTarget() {return new Target10(m_renderer);} public: GSTextureCache10(GSRenderer* r); diff --git a/plugins/GSdx/GSTextureCache11.cpp b/plugins/GSdx/GSTextureCache11.cpp index 221bfb7cd3..fc0e4ee0b4 100644 --- a/plugins/GSdx/GSTextureCache11.cpp +++ b/plugins/GSdx/GSTextureCache11.cpp @@ -29,16 +29,208 @@ GSTextureCache11::GSTextureCache11(GSRenderer* r) { } -// GSRenderTargetHW10 +// Source11 -void GSTextureCache11::GSRenderTargetHW11::Read(const GSVector4i& r) +bool GSTextureCache11::Source11::Create() { + // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + m_bpp = 0; + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); + + return m_texture != NULL; +} + +bool GSTextureCache11::Source11::Create(Target* dst) +{ + m_target = true; + + if(dst->m_type != RenderTarget) + { + // TODO + + return false; + } + + // TODO: clean up this mess + + dst->Update(); + + // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + int tp = (int)m_TEX0.TW << 6; + + // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) + + int w = (int)(dst->m_texture->m_scale.x * tw); + int h = (int)(dst->m_texture->m_scale.y * th); + + GSVector2i dstsize = dst->m_texture->GetSize(); + + // pitch conversion + + if(dst->m_TEX0.TBW != m_TEX0.TBW) // && dst->m_TEX0.PSM == m_TEX0.PSM + { + // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) + + // ASSERT(dst->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateRenderTarget(dstsize.x, dstsize.y); + + GSVector4 size = GSVector4(dstsize).xyxy(); + GSVector4 scale = GSVector4(dst->m_texture->m_scale).xyxy(); + + int bw = 64; + int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + + GSVector4i br(0, 0, bw, bh); + + int sw = (int)dst->m_TEX0.TBW << 6; + + int dw = (int)m_TEX0.TBW << 6; + int dh = 1 << m_TEX0.TH; + + if(sw != 0) + for(int dy = 0; dy < dh; dy += bh) + { + for(int dx = 0; dx < dw; dx += bw) + { + int o = dy * dw / bh + dx; + + int sx = o % sw; + int sy = o / sw; + + GSVector4 sr = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dr = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; + + m_renderer->m_dev->StretchRect(dst->m_texture, sr, m_texture, dr); + + // TODO: this is quite a lot of StretchRect, do it with one Draw + } + } + } + else if(tw < tp) + { + // FIXME: timesplitters blurs the render target by blending itself over a couple of times + + if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) + { + return false; + } + } + + // width/height conversion + + GSVector2 scale = dst->m_texture->m_scale; + + GSVector4 dr(0, 0, w, h); + + if(w > dstsize.x) + { + scale.x = (float)dstsize.x / tw; + dr.z = (float)dstsize.x * scale.x / dst->m_texture->m_scale.x; + w = dstsize.x; + } + + if(h > dstsize.y) + { + scale.y = (float)dstsize.y / th; + dr.w = (float)dstsize.y * scale.y / dst->m_texture->m_scale.y; + h = dstsize.y; + } + + GSVector4 sr(0, 0, w, h); + + GSTexture* st = m_texture ? m_texture : dst->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); + + if(!m_texture) + { + m_texture = dt; + } + + if((sr == dr).alltrue()) + { + D3D11_BOX box = {0, 0, 0, w, h, 1}; + + ID3D11DeviceContext* ctx = *(GSDevice11*)m_renderer->m_dev; + + ctx->CopySubresourceRegion(*(GSTexture11*)dt, 0, 0, 0, 0, *(GSTexture11*)st, 0, &box); + } + else + { + sr.z /= st->GetWidth(); + sr.w /= st->GetHeight(); + + m_renderer->m_dev->StretchRect(st, sr, dt, dr); + } + + if(dt != m_texture) + { + m_renderer->m_dev->Recycle(m_texture); + + m_texture = dt; + } + + m_texture->m_scale = scale; + + switch(m_TEX0.PSM) + { + case PSM_PSMCT32: + m_bpp = 0; + break; + case PSM_PSMCT24: + m_bpp = 1; + break; + case PSM_PSMCT16: + case PSM_PSMCT16S: + m_bpp = 2; + break; + case PSM_PSMT8H: + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); + m_initpalette = true; + break; + case PSM_PSMT4HL: + case PSM_PSMT4HH: + ASSERT(0); // TODO + break; + } + + return true; +} + +// Target11 + +void GSTextureCache11::Target11::Read(const GSVector4i& r) +{ + if(m_type != RenderTarget) + { + // TODO + + return; + } + if(m_TEX0.PSM != PSM_PSMCT32 && m_TEX0.PSM != PSM_PSMCT24 && m_TEX0.PSM != PSM_PSMCT16 && m_TEX0.PSM != PSM_PSMCT16S) { //ASSERT(0); + return; } @@ -122,188 +314,3 @@ void GSTextureCache11::GSRenderTargetHW11::Read(const GSVector4i& r) } } -// GSTextureHW10 - -bool GSTextureCache11::GSCachedTextureHW11::Create() -{ - // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - m_bpp = 0; - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); - - return m_texture != NULL; -} - -bool GSTextureCache11::GSCachedTextureHW11::Create(GSRenderTarget* rt) -{ - // TODO: clean up this mess - - rt->Update(); - - // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - m_rendered = true; - - int tw = 1 << m_TEX0.TW; - int th = 1 << m_TEX0.TH; - int tp = (int)m_TEX0.TW << 6; - - // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - - int w = (int)(rt->m_texture->m_scale.x * tw); - int h = (int)(rt->m_texture->m_scale.y * th); - - GSVector2i rtsize = rt->m_texture->GetSize(); - - // pitch conversion - - if(rt->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM - { - // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) - - // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); - - GSVector4 size = GSVector4(rtsize).xyxy(); - GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); - - int bw = 64; - int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; - - GSVector4i br(0, 0, bw, bh); - - int sw = (int)rt->m_TEX0.TBW << 6; - - int dw = (int)m_TEX0.TBW << 6; - int dh = 1 << m_TEX0.TH; - - if(sw != 0) - for(int dy = 0; dy < dh; dy += bh) - { - for(int dx = 0; dx < dw; dx += bw) - { - int o = dy * dw / bh + dx; - - int sx = o % sw; - int sy = o / sw; - - GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; - GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; - - m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); - - // TODO: this is quite a lot of StretchRect, do it with one Draw - } - } - } - else if(tw < tp) - { - // FIXME: timesplitters blurs the render target by blending itself over a couple of times - - if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) - { - return false; - } - } - - // width/height conversion - - GSVector2 scale = rt->m_texture->m_scale; - - GSVector4 dst(0, 0, w, h); - - if(w > rtsize.x) - { - scale.x = (float)rtsize.x / tw; - dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; - w = rtsize.x; - } - - if(h > rtsize.y) - { - scale.y = (float)rtsize.y / th; - dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; - h = rtsize.y; - } - - GSVector4 src(0, 0, w, h); - - GSTexture* st = m_texture ? m_texture : rt->m_texture; - GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); - - if(!m_texture) - { - m_texture = dt; - } - - if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) - { - D3D11_BOX box = {0, 0, 0, w, h, 1}; - - ID3D11DeviceContext* ctx = *(GSDevice11*)m_renderer->m_dev; - - ctx->CopySubresourceRegion(*(GSTexture11*)dt, 0, 0, 0, 0, *(GSTexture11*)st, 0, &box); - } - else - { - src.z /= st->GetWidth(); - src.w /= st->GetHeight(); - - m_renderer->m_dev->StretchRect(st, src, dt, dst); - } - - if(dt != m_texture) - { - m_renderer->m_dev->Recycle(m_texture); - - m_texture = dt; - } - - m_texture->m_scale = scale; - - switch(m_TEX0.PSM) - { - case PSM_PSMCT32: - m_bpp = 0; - break; - case PSM_PSMCT24: - m_bpp = 1; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 2; - break; - case PSM_PSMT8H: - m_bpp = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1); - m_initpalette = true; - break; - case PSM_PSMT4HL: - case PSM_PSMT4HH: - ASSERT(0); // TODO - break; - } - - return true; -} - -bool GSTextureCache11::GSCachedTextureHW11::Create(GSDepthStencil* ds) -{ - m_rendered = true; - - // TODO - - return false; -} diff --git a/plugins/GSdx/GSTextureCache11.h b/plugins/GSdx/GSTextureCache11.h index ec3d883f81..d02c9c5f71 100644 --- a/plugins/GSdx/GSTextureCache11.h +++ b/plugins/GSdx/GSTextureCache11.h @@ -26,34 +26,26 @@ class GSTextureCache11 : public GSTextureCache { - class GSRenderTargetHW11 : public GSRenderTarget + class Source11 : public Source { public: - explicit GSRenderTargetHW11(GSRenderer* r) : GSRenderTarget(r) {} + explicit Source11(GSRenderer* r) : Source(r) {} + + bool Create(); + bool Create(Target* dst); + }; + + class Target11 : public Target + { + public: + explicit Target11(GSRenderer* r) : Target(r) {} void Read(const GSVector4i& r); }; - class GSDepthStencilHW11 : public GSDepthStencil - { - public: - explicit GSDepthStencilHW11(GSRenderer* r) : GSDepthStencil(r) {} - }; - - class GSCachedTextureHW11 : public GSCachedTexture - { - public: - explicit GSCachedTextureHW11(GSRenderer* r) : GSCachedTexture(r) {} - - bool Create(); - bool Create(GSRenderTarget* rt); - bool Create(GSDepthStencil* ds); - }; - protected: - GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetHW11(m_renderer);} - GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilHW11(m_renderer);} - GSCachedTexture* CreateTexture() {return new GSCachedTextureHW11(m_renderer);} + Source* CreateSource() {return new Source11(m_renderer);} + Target* CreateTarget() {return new Target11(m_renderer);} public: GSTextureCache11(GSRenderer* r); diff --git a/plugins/GSdx/GSTextureCache9.cpp b/plugins/GSdx/GSTextureCache9.cpp index e9b685ae17..6b33249851 100644 --- a/plugins/GSdx/GSTextureCache9.cpp +++ b/plugins/GSdx/GSTextureCache9.cpp @@ -29,16 +29,207 @@ GSTextureCache9::GSTextureCache9(GSRenderer* r) { } -// GSRenderTarget9 +// Source9 -void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) +bool GSTextureCache9::Source9::Create() { + // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + m_bpp = 0; + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); + + return m_texture != NULL; +} + +bool GSTextureCache9::Source9::Create(Target* dst) +{ + m_target = true; + + if(dst->m_type != RenderTarget) + { + // TODO + + return false; + } + + // TODO: clean up this mess + + dst->Update(); + + // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); + + m_TEX0 = m_renderer->m_context->TEX0; + m_TEXA = m_renderer->m_env.TEXA; + + int tw = 1 << m_TEX0.TW; + int th = 1 << m_TEX0.TH; + int tp = (int)m_TEX0.TW << 6; + + // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) + + int w = (int)(dst->m_texture->m_scale.x * tw); + int h = (int)(dst->m_texture->m_scale.y * th); + + GSVector2i dstsize = dst->m_texture->GetSize(); + + // pitch conversion + + if(dst->m_TEX0.TBW != m_TEX0.TBW) // && dst->m_TEX0.PSM == m_TEX0.PSM + { + // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) + + // ASSERT(dst->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) + + ASSERT(m_texture == NULL); + + m_texture = m_renderer->m_dev->CreateRenderTarget(dstsize.x, dstsize.y); + + GSVector4 size = GSVector4(dstsize).xyxy(); + GSVector4 scale = GSVector4(dst->m_texture->m_scale).xyxy(); + + int bw = 64; + int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; + + GSVector4i br(0, 0, bw, bh); + + int sw = (int)dst->m_TEX0.TBW << 6; + + int dw = (int)m_TEX0.TBW << 6; + int dh = 1 << m_TEX0.TH; + + if(sw != 0) + for(int dy = 0; dy < dh; dy += bh) + { + for(int dx = 0; dx < dw; dx += bw) + { + int o = dy * dw / bh + dx; + + int sx = o % sw; + int sy = o / sw; + + GSVector4 sr = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; + GSVector4 dr = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; + + m_renderer->m_dev->StretchRect(dst->m_texture, sr, m_texture, dr); + + // TODO: this is quite a lot of StretchRect, do it with one Draw + } + } + } + else if(tw < tp) + { + // FIXME: timesplitters blurs the render target by blending itself over a couple of times + + if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) + { + return false; + } + } + + // width/height conversion + + GSVector2 scale = dst->m_texture->m_scale; + + GSVector4 dr(0, 0, w, h); + + if(w > dstsize.x) + { + scale.x = (float)dstsize.x / tw; + dr.z = (float)dstsize.x * scale.x / dst->m_texture->m_scale.x; + w = dstsize.x; + } + + if(h > dstsize.y) + { + scale.y = (float)dstsize.y / th; + dr.w = (float)dstsize.y * scale.y / dst->m_texture->m_scale.y; + h = dstsize.y; + } + + GSVector4 sr(0, 0, w, h); + + GSTexture* st = m_texture ? m_texture : dst->m_texture; + GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); + + if(!m_texture) + { + m_texture = dt; + } + + if((sr == dr).alltrue()) + { + GSVector4i r(0, 0, w, h); + + (*(GSDevice9*)m_renderer->m_dev)->StretchRect(*(GSTexture9*)st, r, *(GSTexture9*)dt, r, D3DTEXF_POINT); + } + else + { + sr.z /= st->GetWidth(); + sr.w /= st->GetHeight(); + + m_renderer->m_dev->StretchRect(st, sr, dt, dr); + } + + if(dt != m_texture) + { + m_renderer->m_dev->Recycle(m_texture); + + m_texture = dt; + } + + m_texture->m_scale = scale; + + switch(m_TEX0.PSM) + { + case PSM_PSMCT32: + m_bpp = 0; + break; + case PSM_PSMCT24: + m_bpp = 1; + break; + case PSM_PSMCT16: + case PSM_PSMCT16S: + m_bpp = 2; + break; + case PSM_PSMT8H: + m_bpp = 3; + m_palette = m_renderer->m_dev->CreateTexture(256, 1); + m_initpalette = true; + break; + case PSM_PSMT4HL: + case PSM_PSMT4HH: + ASSERT(0); // TODO + break; + } + + return true; +} + + +// Target9 + +void GSTextureCache9::Target9::Read(const GSVector4i& r) +{ + if(m_type != RenderTarget) + { + // TODO + + return; + } + if(m_TEX0.PSM != PSM_PSMCT32 && m_TEX0.PSM != PSM_PSMCT24 && m_TEX0.PSM != PSM_PSMCT16 && m_TEX0.PSM != PSM_PSMCT16S) { //ASSERT(0); + return; } @@ -120,186 +311,3 @@ void GSTextureCache9::GSRenderTarget9::Read(const GSVector4i& r) } } -// GSTexture9 - -bool GSTextureCache9::GSCachedTexture9::Create() -{ - // m_renderer->m_perfmon.Put(GSPerfMon::WriteTexture, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - m_bpp = 0; - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateTexture(1 << m_TEX0.TW, 1 << m_TEX0.TH); - - return m_texture != NULL; -} - -bool GSTextureCache9::GSCachedTexture9::Create(GSRenderTarget* rt) -{ - // TODO: clean up this mess - - rt->Update(); - - // m_renderer->m_perfmon.Put(GSPerfMon::ConvertRT2T, 1); - - m_TEX0 = m_renderer->m_context->TEX0; - m_TEXA = m_renderer->m_env.TEXA; - - m_rendered = true; - - int tw = 1 << m_TEX0.TW; - int th = 1 << m_TEX0.TH; - int tp = (int)m_TEX0.TW << 6; - - // do not round here!!! if edge becomes a black pixel and addressing mode is clamp => everything outside the clamped area turns into black (kh2 shadows) - - int w = (int)(rt->m_texture->m_scale.x * tw); - int h = (int)(rt->m_texture->m_scale.y * th); - - GSVector2i rtsize = rt->m_texture->GetSize(); - - // pitch conversion - - if(rt->m_TEX0.TBW != m_TEX0.TBW) // && rt->m_TEX0.PSM == m_TEX0.PSM - { - // sfex3 uses this trick (bw: 10 -> 5, wraps the right side below the left) - - // ASSERT(rt->m_TEX0.TBW > m_TEX0.TBW); // otherwise scale.x need to be reduced to make the larger texture fit (TODO) - - ASSERT(m_texture == NULL); - - m_texture = m_renderer->m_dev->CreateRenderTarget(rtsize.x, rtsize.y); - - GSVector4 size = GSVector4(rtsize).xyxy(); - GSVector4 scale = GSVector4(rt->m_texture->m_scale).xyxy(); - - int bw = 64; - int bh = m_TEX0.PSM == PSM_PSMCT32 || m_TEX0.PSM == PSM_PSMCT24 ? 32 : 64; - - GSVector4i br(0, 0, bw, bh); - - int sw = (int)rt->m_TEX0.TBW << 6; - - int dw = (int)m_TEX0.TBW << 6; - int dh = 1 << m_TEX0.TH; - - if(sw != 0) - for(int dy = 0; dy < dh; dy += bh) - { - for(int dx = 0; dx < dw; dx += bw) - { - int o = dy * dw / bh + dx; - - int sx = o % sw; - int sy = o / sw; - - GSVector4 src = GSVector4(GSVector4i(sx, sy).xyxy() + br) * scale / size; - GSVector4 dst = GSVector4(GSVector4i(dx, dy).xyxy() + br) * scale; - - m_renderer->m_dev->StretchRect(rt->m_texture, src, m_texture, dst); - - // TODO: this is quite a lot of StretchRect, do it with one Draw - } - } - } - else if(tw < tp) - { - // FIXME: timesplitters blurs the render target by blending itself over a couple of times - - if(tw == 256 && th == 128 && tp == 512 && (m_TEX0.TBP0 == 0 || m_TEX0.TBP0 == 0x00e00)) - { - return false; - } - } - - // width/height conversion - - GSVector2 scale = rt->m_texture->m_scale; - - GSVector4 dst(0, 0, w, h); - - if(w > rtsize.x) - { - scale.x = (float)rtsize.x / tw; - dst.z = (float)rtsize.x * scale.x / rt->m_texture->m_scale.x; - w = rtsize.x; - } - - if(h > rtsize.y) - { - scale.y = (float)rtsize.y / th; - dst.w = (float)rtsize.y * scale.y / rt->m_texture->m_scale.y; - h = rtsize.y; - } - - GSVector4 src(0, 0, w, h); - - GSTexture* st = m_texture ? m_texture : rt->m_texture; - GSTexture* dt = m_renderer->m_dev->CreateRenderTarget(w, h); - - if(!m_texture) - { - m_texture = dt; - } - - if(src.x == dst.x && src.y == dst.y && src.z == dst.z && src.w == dst.w) - { - GSVector4i r(0, 0, w, h); - - (*(GSDevice9*)m_renderer->m_dev)->StretchRect(*(GSTexture9*)st, r, *(GSTexture9*)dt, r, D3DTEXF_POINT); - } - else - { - src.z /= st->GetWidth(); - src.w /= st->GetHeight(); - - m_renderer->m_dev->StretchRect(st, src, dt, dst); - } - - if(dt != m_texture) - { - m_renderer->m_dev->Recycle(m_texture); - - m_texture = dt; - } - - m_texture->m_scale = scale; - - switch(m_TEX0.PSM) - { - case PSM_PSMCT32: - m_bpp = 0; - break; - case PSM_PSMCT24: - m_bpp = 1; - break; - case PSM_PSMCT16: - case PSM_PSMCT16S: - m_bpp = 2; - break; - case PSM_PSMT8H: - m_bpp = 3; - m_palette = m_renderer->m_dev->CreateTexture(256, 1); - m_initpalette = true; - break; - case PSM_PSMT4HL: - case PSM_PSMT4HH: - ASSERT(0); // TODO - break; - } - - return true; -} - -bool GSTextureCache9::GSCachedTexture9::Create(GSDepthStencil* ds) -{ - m_rendered = true; - - // TODO - - return false; -} diff --git a/plugins/GSdx/GSTextureCache9.h b/plugins/GSdx/GSTextureCache9.h index 3f3afe0b2b..0d0aedb2e9 100644 --- a/plugins/GSdx/GSTextureCache9.h +++ b/plugins/GSdx/GSTextureCache9.h @@ -26,34 +26,26 @@ class GSTextureCache9 : public GSTextureCache { - class GSRenderTarget9 : public GSRenderTarget + class Source9 : public Source { public: - explicit GSRenderTarget9(GSRenderer* r) : GSRenderTarget(r) {} + explicit Source9(GSRenderer* r) : Source(r) {} + + bool Create(); + bool Create(Target* dst); + }; + + class Target9 : public Target + { + public: + explicit Target9(GSRenderer* r) : Target(r) {} void Read(const GSVector4i& r); }; - class GSDepthStencil9 : public GSDepthStencil - { - public: - explicit GSDepthStencil9(GSRenderer* r) : GSDepthStencil(r) {} - }; - - class GSCachedTexture9 : public GSCachedTexture - { - public: - explicit GSCachedTexture9(GSRenderer* r) : GSCachedTexture(r) {} - - bool Create(); - bool Create(GSRenderTarget* rt); - bool Create(GSDepthStencil* ds); - }; - protected: - GSRenderTarget* CreateRenderTarget() {return new GSRenderTarget9(m_renderer);} - GSDepthStencil* CreateDepthStencil() {return new GSDepthStencil9(m_renderer);} - GSCachedTexture* CreateTexture() {return new GSCachedTexture9(m_renderer);} + Source* CreateSource() {return new Source9(m_renderer);} + Target* CreateTarget() {return new Target9(m_renderer);} public: GSTextureCache9(GSRenderer* r); diff --git a/plugins/GSdx/GSTextureCacheOGL.cpp b/plugins/GSdx/GSTextureCacheOGL.cpp index b470935181..99131040df 100644 --- a/plugins/GSdx/GSTextureCacheOGL.cpp +++ b/plugins/GSdx/GSTextureCacheOGL.cpp @@ -29,34 +29,27 @@ GSTextureCacheOGL::GSTextureCacheOGL(GSRenderer* r) { } -// GSRenderTargetOGL +// SourceOGL -void GSTextureCacheOGL::GSRenderTargetOGL::Read(const GSVector4i& r) -{ - // TODO -} - -// GSTextureOGL - -bool GSTextureCacheOGL::GSCachedTextureOGL::Create() +bool GSTextureCacheOGL::SourceOGL::Create() { // TODO return true; } -bool GSTextureCacheOGL::GSCachedTextureOGL::Create(GSRenderTarget* rt) +bool GSTextureCacheOGL::SourceOGL::Create(Target* dst) { + m_target = true; + // TODO return true; } -bool GSTextureCacheOGL::GSCachedTextureOGL::Create(GSDepthStencil* ds) +// TargetOGL + +void GSTextureCacheOGL::TargetOGL::Read(const GSVector4i& r) { - m_rendered = true; - // TODO - - return false; } diff --git a/plugins/GSdx/GSTextureCacheOGL.h b/plugins/GSdx/GSTextureCacheOGL.h index c553559bd9..41bb7d4c8b 100644 --- a/plugins/GSdx/GSTextureCacheOGL.h +++ b/plugins/GSdx/GSTextureCacheOGL.h @@ -26,34 +26,26 @@ class GSTextureCacheOGL : public GSTextureCache { - class GSRenderTargetOGL : public GSRenderTarget + class SourceOGL : public Source { public: - explicit GSRenderTargetOGL(GSRenderer* r) : GSRenderTarget(r) {} + explicit SourceOGL(GSRenderer* r) : Source(r) {} + + bool Create(); + bool Create(Target* dst); + }; + + class TargetOGL : public Target + { + public: + explicit TargetOGL(GSRenderer* r) : Target(r) {} void Read(const GSVector4i& r); }; - class GSDepthStencilOGL : public GSDepthStencil - { - public: - explicit GSDepthStencilOGL(GSRenderer* r) : GSDepthStencil(r) {} - }; - - class GSCachedTextureOGL : public GSCachedTexture - { - public: - explicit GSCachedTextureOGL(GSRenderer* r) : GSCachedTexture(r) {} - - bool Create(); - bool Create(GSRenderTarget* rt); - bool Create(GSDepthStencil* ds); - }; - protected: - GSRenderTarget* CreateRenderTarget() {return new GSRenderTargetOGL(m_renderer);} - GSDepthStencil* CreateDepthStencil() {return new GSDepthStencilOGL(m_renderer);} - GSCachedTexture* CreateTexture() {return new GSCachedTextureOGL(m_renderer);} + Source* CreateSource() {return new SourceOGL(m_renderer);} + Target* CreateTarget() {return new TargetOGL(m_renderer);} public: GSTextureCacheOGL(GSRenderer* r); diff --git a/plugins/GSdx/GSTextureCacheSW.cpp b/plugins/GSdx/GSTextureCacheSW.cpp index 1be3e4b6a6..17bd198170 100644 --- a/plugins/GSdx/GSTextureCacheSW.cpp +++ b/plugins/GSdx/GSTextureCacheSW.cpp @@ -34,17 +34,15 @@ GSTextureCacheSW::~GSTextureCacheSW() const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r) { - GSLocalMemory& mem = m_state->m_mem; - const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - const hash_map& map = m_map[TEX0.TBP0 >> 5]; - GSTexture* t = NULL; + const hash_map& map = m_map[TEX0.TBP0 >> 5]; + for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) { - GSTexture* t2 = (*i).first; + GSTexture* t2 = i->first; if(((t2->m_TEX0.u32[0] ^ TEX0.u32[0]) | ((t2->m_TEX0.u32[1] ^ TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH { @@ -85,12 +83,10 @@ const GSTextureCacheSW::GSTexture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TE { uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; - if(page >= MAX_PAGES) + if(page < MAX_PAGES) { - continue; + m_map[page][t] = true; } - - m_map[page][t] = true; } } } @@ -170,20 +166,21 @@ void GSTextureCacheSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, cons { uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; - if(page >= MAX_PAGES) + if(page < MAX_PAGES) { - continue; - } + const hash_map& map = m_map[page]; - const hash_map& map = m_map[page]; + for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) + { + GSTexture* t = i->first; - for(hash_map::const_iterator i = map.begin(); i != map.end(); i++) - { - GSTexture* t = (*i).first; + if(GSUtil::HasSharedBits(BITBLTBUF.DPSM, t->m_TEX0.PSM)) + { + t->m_valid[page] = 0; - t->m_valid[page] = 0; - - t->m_complete = false; + t->m_complete = false; + } + } } } } @@ -219,13 +216,8 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX m_TEX0 = TEX0; m_TEXA = TEXA; - GSLocalMemory& mem = m_state->m_mem; - const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - uint32 bp = TEX0.TBP0; - uint32 bw = TEX0.TBW; - GSVector2i s = psm.bs; int tw = max(1 << TEX0.TW, s.x); @@ -250,8 +242,13 @@ bool GSTextureCacheSW::GSTexture::Update(const GIFRegTEX0& TEX0, const GIFRegTEX m_complete = true; // lame, but better than nothing } + GSLocalMemory& mem = m_state->m_mem; + + uint32 bp = TEX0.TBP0; + uint32 bw = TEX0.TBW; + GSLocalMemory::readTextureBlock rtxb = psm.rtxbP; - + int bytes = psm.pal > 0 ? 1 : 4; uint32 pitch = (1 << m_tw) * bytes; diff --git a/plugins/GSdx/GSTextureCacheSW.h b/plugins/GSdx/GSTextureCacheSW.h index 4e99396ca9..ccb57495e4 100644 --- a/plugins/GSdx/GSTextureCacheSW.h +++ b/plugins/GSdx/GSTextureCacheSW.h @@ -23,9 +23,6 @@ #include "GSRenderer.h" -#define MAX_PAGES 512 -#define MAX_BLOCKS 16384 - class GSTextureCacheSW { public: diff --git a/plugins/GSdx/GSTextureFX.cpp b/plugins/GSdx/GSTextureFX.cpp index 16a86f2475..f243c41679 100644 --- a/plugins/GSdx/GSTextureFX.cpp +++ b/plugins/GSdx/GSTextureFX.cpp @@ -21,3 +21,16 @@ #include "stdafx.h" #include "GSTextureFX.h" + +GSTextureFX::GSTextureFX() + : m_dev(NULL) +{ +} + +bool GSTextureFX::Create(GSDevice* dev) +{ + m_dev = dev; + + return true; +} + diff --git a/plugins/GSdx/GSTextureFX.h b/plugins/GSdx/GSTextureFX.h index ed338a59b7..b3b3dcf692 100644 --- a/plugins/GSdx/GSTextureFX.h +++ b/plugins/GSdx/GSTextureFX.h @@ -22,8 +22,10 @@ #pragma once #include "GSVector.h" +#include "GSDevice.h" +#include "GSAlignedClass.h" -class GSTextureFX +class GSTextureFX : public GSAlignedClass<16> { public: #pragma pack(push, 1) @@ -87,9 +89,9 @@ public: { GSVector4 FogColor_AREF; GSVector4 HalfTexel; - GSVector4 WH_TA; + GSVector4 WH; GSVector4 MinMax; - GSVector4 MinMaxF; + GSVector4 MinF_TA; GSVector4i MskFix; struct PSConstantBuffer() @@ -240,4 +242,21 @@ public: }; #pragma pack(pop) + +protected: + GSDevice* m_dev; + +public: + GSTextureFX(); + + virtual bool Create(GSDevice* dev); + + virtual void SetupIA(const void* vertices, int count, int prim) = 0; + virtual void SetupVS(VSSelector sel, const VSConstantBuffer* cb) = 0; + virtual void SetupGS(GSSelector sel) = 0; + virtual void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) = 0; + virtual void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) = 0; + virtual void SetupRS(int w, int h, const GSVector4i& scissor) = 0; + virtual void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds) = 0; + virtual void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix) = 0; }; diff --git a/plugins/GSdx/GSTextureFX10.cpp b/plugins/GSdx/GSTextureFX10.cpp index 0a34979799..3750a5b9c8 100644 --- a/plugins/GSdx/GSTextureFX10.cpp +++ b/plugins/GSdx/GSTextureFX10.cpp @@ -24,21 +24,17 @@ #include "resource.h" GSTextureFX10::GSTextureFX10() - : m_dev(NULL) { memset(&m_vs_cb_cache, 0, sizeof(m_vs_cb_cache)); memset(&m_ps_cb_cache, 0, sizeof(m_ps_cb_cache)); } -bool GSTextureFX10::Create(GSDevice10* dev) +bool GSTextureFX10::Create(GSDevice* dev) { - m_dev = dev; - - VSSelector sel; - - VSConstantBuffer cb; - - SetupVS(sel, &cb); // creates layout + if(!__super::Create(dev)) + { + return false; + } HRESULT hr; @@ -50,7 +46,7 @@ bool GSTextureFX10::Create(GSDevice10* dev) bd.Usage = D3D10_USAGE_DEFAULT; bd.BindFlags = D3D10_BIND_CONSTANT_BUFFER; - hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_vs_cb); + hr = (*(GSDevice10*)dev)->CreateBuffer(&bd, NULL, &m_vs_cb); if(FAILED(hr)) return false; @@ -60,7 +56,7 @@ bool GSTextureFX10::Create(GSDevice10* dev) bd.Usage = D3D10_USAGE_DEFAULT; bd.BindFlags = D3D10_BIND_CONSTANT_BUFFER; - hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_ps_cb); + hr = (*(GSDevice10*)dev)->CreateBuffer(&bd, NULL, &m_ps_cb); if(FAILED(hr)) return false; @@ -76,7 +72,7 @@ bool GSTextureFX10::Create(GSDevice10* dev) sd.MaxAnisotropy = 16; sd.ComparisonFunc = D3D10_COMPARISON_NEVER; - hr = (*m_dev)->CreateSamplerState(&sd, &m_palette_ss); + hr = (*(GSDevice10*)dev)->CreateSamplerState(&sd, &m_palette_ss); if(FAILED(hr)) return false; @@ -85,17 +81,19 @@ bool GSTextureFX10::Create(GSDevice10* dev) return true; } -bool GSTextureFX10::SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMITIVE_TOPOLOGY prim) +void GSTextureFX10::SetupIA(const void* vertices, int count, int prim) { - m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); - m_dev->IASetInputLayout(m_il); - m_dev->IASetPrimitiveTopology(prim); + GSDevice10* dev = (GSDevice10*)m_dev; - return true; + dev->IASetVertexBuffer(vertices, sizeof(GSVertexHW10), count); + dev->IASetInputLayout(m_il); + dev->IASetPrimitiveTopology((D3D10_PRIMITIVE_TOPOLOGY)prim); } -bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) +void GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { + GSDevice10* dev = (GSDevice10*)m_dev; + hash_map >::const_iterator i = m_vs.find(sel); if(i == m_vs.end()) @@ -129,7 +127,7 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) CComPtr il; CComPtr vs; - m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); if(m_il == NULL) { @@ -143,17 +141,15 @@ bool GSTextureFX10::SetupVS(VSSelector sel, const VSConstantBuffer* cb) if(m_vs_cb_cache.Update(cb)) { - (*m_dev)->UpdateSubresource(m_vs_cb, 0, NULL, cb, 0, 0); + (*dev)->UpdateSubresource(m_vs_cb, 0, NULL, cb, 0, 0); } - m_dev->VSSetShader((*i).second, m_vs_cb); - - return true; + dev->VSSetShader(i->second, m_vs_cb); } -bool GSTextureFX10::SetupGS(GSSelector sel) +void GSTextureFX10::SetupGS(GSSelector sel) { - HRESULT hr; + GSDevice10* dev = (GSDevice10*)m_dev; ID3D10GeometryShader* gs = NULL; @@ -163,7 +159,7 @@ bool GSTextureFX10::SetupGS(GSSelector sel) if(i != m_gs.end()) { - gs = (*i).second; + gs = i->second; } else { @@ -179,29 +175,25 @@ bool GSTextureFX10::SetupGS(GSSelector sel) {NULL, NULL}, }; - hr = m_dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); + dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); m_gs[sel] = gs; } } - m_dev->GSSetShader(gs); - - return true; + dev->GSSetShader(gs); } -bool GSTextureFX10::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) +void GSTextureFX10::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) { - m_dev->PSSetShaderResources(tex, pal); + ((GSDevice10*)m_dev)->PSSetShaderResources(tex, pal); UpdatePS(sel, cb, ssel); - - return true; } void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) { - HRESULT hr; + GSDevice10* dev = (GSDevice10*)m_dev; hash_map >::const_iterator i = m_ps.find(sel); @@ -245,7 +237,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl CComPtr ps; - hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); + dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); m_ps[sel] = ps; @@ -254,10 +246,10 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(m_ps_cb_cache.Update(cb)) { - (*m_dev)->UpdateSubresource(m_ps_cb, 0, NULL, cb, 0, 0); + (*dev)->UpdateSubresource(m_ps_cb, 0, NULL, cb, 0, 0); } - m_dev->PSSetShader((*i).second, m_ps_cb); + dev->PSSetShader(i->second, m_ps_cb); ID3D10SamplerState* ss0 = NULL; ID3D10SamplerState* ss1 = NULL; @@ -291,7 +283,7 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl sd.MaxAnisotropy = 16; sd.ComparisonFunc = D3D10_COMPARISON_NEVER; - hr = (*m_dev)->CreateSamplerState(&sd, &ss0); + (*dev)->CreateSamplerState(&sd, &ss0); m_ps_ss[ssel] = ss0; } @@ -302,24 +294,24 @@ void GSTextureFX10::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl } } - m_dev->PSSetSamplerState(ss0, ss1); + dev->PSSetSamplerState(ss0, ss1); } void GSTextureFX10::SetupRS(int w, int h, const GSVector4i& scissor) { - m_dev->RSSet(w, h, &scissor); + ((GSDevice10*)m_dev)->RSSet(w, h, &scissor); } -void GSTextureFX10::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds) +void GSTextureFX10::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds) { - UpdateOM(dssel, bsel, bf); + UpdateOM(dssel, bsel, afix); - m_dev->OMSetRenderTargets(rt, ds); + ((GSDevice10*)m_dev)->OMSetRenderTargets(rt, ds); } -void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf) +void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix) { - HRESULT hr; + GSDevice10* dev = (GSDevice10*)m_dev; hash_map >::const_iterator i = m_om_dss.find(dssel); @@ -361,14 +353,14 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr dss; - hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); + (*dev)->CreateDepthStencilState(&dsd, &dss); m_om_dss[dssel] = dss; i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState((*i).second, 1); + dev->OMSetDepthStencilState((*i).second, 1); hash_map >::const_iterator j = m_om_bs.find(bsel); @@ -502,12 +494,12 @@ void GSTextureFX10::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr bs; - hr = (*m_dev)->CreateBlendState(&bd, &bs); + (*dev)->CreateBlendState(&bd, &bs); m_om_bs[bsel] = bs; j = m_om_bs.find(bsel); } - m_dev->OMSetBlendState((*j).second, bf); + dev->OMSetBlendState(j->second, (float)(int)afix / 0x80); } diff --git a/plugins/GSdx/GSTextureFX10.h b/plugins/GSdx/GSTextureFX10.h index 9e3c18a5ee..566809a360 100644 --- a/plugins/GSdx/GSTextureFX10.h +++ b/plugins/GSdx/GSTextureFX10.h @@ -26,7 +26,6 @@ class GSTextureFX10 : public GSTextureFX { - GSDevice10* m_dev; CComPtr m_il; hash_map > m_vs; CComPtr m_vs_cb; @@ -44,14 +43,14 @@ class GSTextureFX10 : public GSTextureFX public: GSTextureFX10(); - bool Create(GSDevice10* dev); + bool Create(GSDevice* dev); - bool SetupIA(const GSVertexHW10* vertices, int count, D3D10_PRIMITIVE_TOPOLOGY prim); - bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); - bool SetupGS(GSSelector sel); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); + void SetupIA(const void* vertices, int count, int prim); + void SetupVS(VSSelector sel, const VSConstantBuffer* cb); + void SetupGS(GSSelector sel); + void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupRS(int w, int h, const GSVector4i& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); - void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds); + void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix); }; diff --git a/plugins/GSdx/GSTextureFX11.cpp b/plugins/GSdx/GSTextureFX11.cpp index d3c1205888..24e2b25dbb 100644 --- a/plugins/GSdx/GSTextureFX11.cpp +++ b/plugins/GSdx/GSTextureFX11.cpp @@ -24,21 +24,17 @@ #include "resource.h" GSTextureFX11::GSTextureFX11() - : m_dev(NULL) { memset(&m_vs_cb_cache, 0, sizeof(m_vs_cb_cache)); memset(&m_ps_cb_cache, 0, sizeof(m_ps_cb_cache)); } -bool GSTextureFX11::Create(GSDevice11* dev) +bool GSTextureFX11::Create(GSDevice* dev) { - m_dev = dev; - - VSSelector sel; - - VSConstantBuffer cb; - - SetupVS(sel, &cb); // creates layout + if(!__super::Create(dev)) + { + return false; + } HRESULT hr; @@ -50,7 +46,7 @@ bool GSTextureFX11::Create(GSDevice11* dev) bd.Usage = D3D11_USAGE_DEFAULT; bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; - hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_vs_cb); + hr = (*(GSDevice11*)dev)->CreateBuffer(&bd, NULL, &m_vs_cb); if(FAILED(hr)) return false; @@ -60,7 +56,7 @@ bool GSTextureFX11::Create(GSDevice11* dev) bd.Usage = D3D11_USAGE_DEFAULT; bd.BindFlags = D3D11_BIND_CONSTANT_BUFFER; - hr = (*m_dev)->CreateBuffer(&bd, NULL, &m_ps_cb); + hr = (*(GSDevice11*)dev)->CreateBuffer(&bd, NULL, &m_ps_cb); if(FAILED(hr)) return false; @@ -76,7 +72,7 @@ bool GSTextureFX11::Create(GSDevice11* dev) sd.MaxAnisotropy = 16; sd.ComparisonFunc = D3D11_COMPARISON_NEVER; - hr = (*m_dev)->CreateSamplerState(&sd, &m_palette_ss); + hr = (*(GSDevice11*)dev)->CreateSamplerState(&sd, &m_palette_ss); if(FAILED(hr)) return false; @@ -85,17 +81,19 @@ bool GSTextureFX11::Create(GSDevice11* dev) return true; } -bool GSTextureFX11::SetupIA(const GSVertexHW11* vertices, int count, D3D11_PRIMITIVE_TOPOLOGY prim) +void GSTextureFX11::SetupIA(const void* vertices, int count, int prim) { - m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); - m_dev->IASetInputLayout(m_il); - m_dev->IASetPrimitiveTopology(prim); + GSDevice11* dev = (GSDevice11*)m_dev; - return true; + dev->IASetVertexBuffer(vertices, sizeof(GSVertexHW11), count); + dev->IASetInputLayout(m_il); + dev->IASetPrimitiveTopology((D3D11_PRIMITIVE_TOPOLOGY)prim); } -bool GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) +void GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { + GSDevice11* dev = (GSDevice11*)m_dev; + hash_map >::const_iterator i = m_vs.find(sel); if(i == m_vs.end()) @@ -129,7 +127,7 @@ bool GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) CComPtr il; CComPtr vs; - m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); if(m_il == NULL) { @@ -143,19 +141,17 @@ bool GSTextureFX11::SetupVS(VSSelector sel, const VSConstantBuffer* cb) if(m_vs_cb_cache.Update(cb)) { - ID3D11DeviceContext* ctx = *m_dev; + ID3D11DeviceContext* ctx = *dev; ctx->UpdateSubresource(m_vs_cb, 0, NULL, cb, 0, 0); } - m_dev->VSSetShader((*i).second, m_vs_cb); - - return true; + dev->VSSetShader(i->second, m_vs_cb); } -bool GSTextureFX11::SetupGS(GSSelector sel) +void GSTextureFX11::SetupGS(GSSelector sel) { - HRESULT hr; + GSDevice11* dev = (GSDevice11*)m_dev; ID3D11GeometryShader* gs = NULL; @@ -165,7 +161,7 @@ bool GSTextureFX11::SetupGS(GSSelector sel) if(i != m_gs.end()) { - gs = (*i).second; + gs = i->second; } else { @@ -181,29 +177,25 @@ bool GSTextureFX11::SetupGS(GSSelector sel) {NULL, NULL}, }; - hr = m_dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); + dev->CompileShader(IDR_TFX_FX, "gs_main", macro, &gs); m_gs[sel] = gs; } } - m_dev->GSSetShader(gs); - - return true; + dev->GSSetShader(gs); } -bool GSTextureFX11::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) +void GSTextureFX11::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) { - m_dev->PSSetShaderResources(tex, pal); + ((GSDevice11*)m_dev)->PSSetShaderResources(tex, pal); UpdatePS(sel, cb, ssel); - - return true; } void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) { - HRESULT hr; + GSDevice11* dev = (GSDevice11*)m_dev; hash_map >::const_iterator i = m_ps.find(sel); @@ -247,7 +239,7 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl CComPtr ps; - hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); + dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); m_ps[sel] = ps; @@ -256,12 +248,12 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(m_ps_cb_cache.Update(cb)) { - ID3D11DeviceContext* ctx = *m_dev; + ID3D11DeviceContext* ctx = *dev; ctx->UpdateSubresource(m_ps_cb, 0, NULL, cb, 0, 0); } - m_dev->PSSetShader((*i).second, m_ps_cb); + dev->PSSetShader(i->second, m_ps_cb); ID3D11SamplerState* ss0 = NULL; ID3D11SamplerState* ss1 = NULL; @@ -277,7 +269,7 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl if(i != m_ps_ss.end()) { - ss0 = (*i).second; + ss0 = i->second; } else { @@ -295,7 +287,7 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl sd.MaxAnisotropy = 16; sd.ComparisonFunc = D3D11_COMPARISON_NEVER; - hr = (*m_dev)->CreateSamplerState(&sd, &ss0); + (*dev)->CreateSamplerState(&sd, &ss0); m_ps_ss[ssel] = ss0; } @@ -306,24 +298,24 @@ void GSTextureFX11::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSampl } } - m_dev->PSSetSamplerState(ss0, ss1); + dev->PSSetSamplerState(ss0, ss1); } void GSTextureFX11::SetupRS(int w, int h, const GSVector4i& scissor) { - m_dev->RSSet(w, h, &scissor); + ((GSDevice11*)m_dev)->RSSet(w, h, &scissor); } -void GSTextureFX11::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds) +void GSTextureFX11::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds) { - UpdateOM(dssel, bsel, bf); + UpdateOM(dssel, bsel, afix); - m_dev->OMSetRenderTargets(rt, ds); + ((GSDevice11*)m_dev)->OMSetRenderTargets(rt, ds); } -void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf) +void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix) { - HRESULT hr; + GSDevice11* dev = (GSDevice11*)m_dev; hash_map >::const_iterator i = m_om_dss.find(dssel); @@ -365,14 +357,14 @@ void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr dss; - hr = (*m_dev)->CreateDepthStencilState(&dsd, &dss); + (*dev)->CreateDepthStencilState(&dsd, &dss); m_om_dss[dssel] = dss; i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState((*i).second, 1); + dev->OMSetDepthStencilState(i->second, 1); hash_map >::const_iterator j = m_om_bs.find(bsel); @@ -506,12 +498,12 @@ void GSTextureFX11::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, CComPtr bs; - hr = (*m_dev)->CreateBlendState(&bd, &bs); + (*dev)->CreateBlendState(&bd, &bs); m_om_bs[bsel] = bs; j = m_om_bs.find(bsel); } - m_dev->OMSetBlendState((*j).second, bf); + dev->OMSetBlendState(j->second, (float)(int)afix / 0x80); } diff --git a/plugins/GSdx/GSTextureFX11.h b/plugins/GSdx/GSTextureFX11.h index 53a4f2f337..5c2b67bcc1 100644 --- a/plugins/GSdx/GSTextureFX11.h +++ b/plugins/GSdx/GSTextureFX11.h @@ -26,7 +26,6 @@ class GSTextureFX11 : public GSTextureFX { - GSDevice11* m_dev; CComPtr m_il; hash_map > m_vs; CComPtr m_vs_cb; @@ -44,14 +43,14 @@ class GSTextureFX11 : public GSTextureFX public: GSTextureFX11(); - bool Create(GSDevice11* dev); + bool Create(GSDevice* dev); - bool SetupIA(const GSVertexHW11* vertices, int count, D3D11_PRIMITIVE_TOPOLOGY prim); - bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); - bool SetupGS(GSSelector sel); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); + void SetupIA(const void* vertices, int count, int prim); + void SetupVS(VSSelector sel, const VSConstantBuffer* cb); + void SetupGS(GSSelector sel); + void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupRS(int w, int h, const GSVector4i& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf, GSTexture* rt, GSTexture* ds); - void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, float bf); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds); + void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix); }; diff --git a/plugins/GSdx/GSTextureFX9.cpp b/plugins/GSdx/GSTextureFX9.cpp index 5327f2d93f..891085c949 100644 --- a/plugins/GSdx/GSTextureFX9.cpp +++ b/plugins/GSdx/GSTextureFX9.cpp @@ -24,19 +24,24 @@ #include "resource.h" GSTextureFX9::GSTextureFX9() - : m_dev(NULL) { } -bool GSTextureFX9::Create(GSDevice9* dev) +bool GSTextureFX9::Create(GSDevice* dev) { - m_dev = dev; + if(!__super::Create(dev)) + { + return false; + } + + // create layout VSSelector sel; - VSConstantBuffer cb; - SetupVS(sel, &cb); // creates layout + SetupVS(sel, &cb); + + // return true; } @@ -51,7 +56,7 @@ GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) if(i != m_mskfix.end()) { - t = (*i).second; + t = i->second; } else { @@ -78,17 +83,19 @@ GSTexture* GSTextureFX9::CreateMskFix(uint32 size, uint32 msk, uint32 fix) return t; } -bool GSTextureFX9::SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim) +void GSTextureFX9::SetupIA(const void* vertices, int count, int prim) { - m_dev->IASetVertexBuffer(vertices, sizeof(vertices[0]), count); - m_dev->IASetInputLayout(m_il); - m_dev->IASetPrimitiveTopology(prim); + GSDevice9* dev = (GSDevice9*)m_dev; - return true; + dev->IASetVertexBuffer(vertices, sizeof(GSVertexHW9), count); + dev->IASetInputLayout(m_il); + dev->IASetPrimitiveTopology((D3DPRIMITIVETYPE)prim); } -bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) +void GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) { + GSDevice9* dev = (GSDevice9*)m_dev; + hash_map >::const_iterator i = m_vs.find(sel); if(i == m_vs.end()) @@ -121,7 +128,7 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) CComPtr il; CComPtr vs; - m_dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); + dev->CompileShader(IDR_TFX_FX, "vs_main", macro, &vs, layout, countof(layout), &il); if(m_il == NULL) { @@ -133,14 +140,14 @@ bool GSTextureFX9::SetupVS(VSSelector sel, const VSConstantBuffer* cb) i = m_vs.find(sel); } - m_dev->VSSetShader((*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); - - return true; + dev->VSSetShader(i->second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); } -bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) +void GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal) { - m_dev->PSSetShaderResources(tex, pal); + GSDevice9* dev = (GSDevice9*)m_dev; + + dev->PSSetShaderResources(tex, pal); if(tex && (sel.wms == 3 || sel.wmt == 3)) { @@ -148,7 +155,7 @@ bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSampler { if(GSTexture* t = CreateMskFix(tex->GetWidth(), cb->MskFix.x, cb->MskFix.z)) { - (*m_dev)->SetTexture(2, *(GSTexture9*)t); + (*dev)->SetTexture(2, *(GSTexture9*)t); } } @@ -156,19 +163,17 @@ bool GSTextureFX9::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSampler { if(GSTexture* t = CreateMskFix(tex->GetHeight(), cb->MskFix.y, cb->MskFix.w)) { - (*m_dev)->SetTexture(3, *(GSTexture9*)t); + (*dev)->SetTexture(3, *(GSTexture9*)t); } } } UpdatePS(sel, cb, ssel); - - return true; } void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel) { - HRESULT hr; + GSDevice9* dev = (GSDevice9*)m_dev; hash_map >::const_iterator i = m_ps.find(sel); @@ -210,14 +215,14 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample CComPtr ps; - hr = m_dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); + dev->CompileShader(IDR_TFX_FX, "ps_main", macro, &ps); m_ps[sel] = ps; i = m_ps.find(sel); } - m_dev->PSSetShader((*i).second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); + dev->PSSetShader(i->second, (const float*)cb, sizeof(*cb) / sizeof(GSVector4)); Direct3DSamplerState9* ss = NULL; @@ -232,7 +237,7 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample if(i != m_ps_ss.end()) { - ss = (*i).second; + ss = i->second; } else { @@ -252,23 +257,25 @@ void GSTextureFX9::UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSample } } - m_dev->PSSetSamplerState(ss); + dev->PSSetSamplerState(ss); } void GSTextureFX9::SetupRS(int w, int h, const GSVector4i& scissor) { - m_dev->RSSet(w, h, &scissor); + ((GSDevice9*)m_dev)->RSSet(w, h, &scissor); } -void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds) +void GSTextureFX9::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds) { - UpdateOM(dssel, bsel, bf); + UpdateOM(dssel, bsel, afix); - m_dev->OMSetRenderTargets(rt, ds); + ((GSDevice9*)m_dev)->OMSetRenderTargets(rt, ds); } -void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf) +void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix) { + GSDevice9* dev = (GSDevice9*)m_dev; + Direct3DDepthStencilState9* dss = NULL; hash_map::const_iterator i = m_om_dss.find(dssel); @@ -311,7 +318,7 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, i = m_om_dss.find(dssel); } - m_dev->OMSetDepthStencilState((*i).second); + dev->OMSetDepthStencilState(i->second); hash_map::const_iterator j = m_om_bs.find(bsel); @@ -448,5 +455,5 @@ void GSTextureFX9::UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, j = m_om_bs.find(bsel); } - m_dev->OMSetBlendState((*j).second, 0x010101 * bf); + dev->OMSetBlendState(j->second, afix >= 0x80 ? 0xffffff : 0x020202 * afix); } diff --git a/plugins/GSdx/GSTextureFX9.h b/plugins/GSdx/GSTextureFX9.h index 50f187a3bf..317037f695 100644 --- a/plugins/GSdx/GSTextureFX9.h +++ b/plugins/GSdx/GSTextureFX9.h @@ -26,7 +26,6 @@ class GSTextureFX9 : public GSTextureFX { - GSDevice9* m_dev; CComPtr m_il; hash_map > m_vs; D3DXHANDLE m_vs_params; @@ -41,13 +40,14 @@ class GSTextureFX9 : public GSTextureFX public: GSTextureFX9(); - bool Create(GSDevice9* dev); + bool Create(GSDevice* dev); - bool SetupIA(const GSVertexHW9* vertices, int count, D3DPRIMITIVETYPE prim); - bool SetupVS(VSSelector sel, const VSConstantBuffer* cb); - bool SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); + void SetupIA(const void* vertices, int count, int prim); + void SetupVS(VSSelector sel, const VSConstantBuffer* cb); + void SetupGS(GSSelector sel) {} + void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel, GSTexture* tex, GSTexture* pal); void UpdatePS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupRS(int w, int h, const GSVector4i& scissor); - void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf, GSTexture* rt, GSTexture* ds); - void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 bf); + void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix, GSTexture* rt, GSTexture* ds); + void UpdateOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix); }; diff --git a/plugins/GSdx/GSVector.cpp b/plugins/GSdx/GSVector.cpp index a1741d5e0a..d7f6e7ec77 100644 --- a/plugins/GSdx/GSVector.cpp +++ b/plugins/GSdx/GSVector.cpp @@ -27,16 +27,16 @@ const GSVector4 GSVector4::m_ps4567(4.0f, 5.0f, 6.0f, 7.0f); const GSVector4 GSVector4::m_x3f800000(_mm_castsi128_ps(_mm_set1_epi32(0x3f800000))); const GSVector4 GSVector4::m_x4b000000(_mm_castsi128_ps(_mm_set1_epi32(0x4b000000))); -void GSVector4::operator = (const GSVector4i& v) -{ - m = _mm_cvtepi32_ps(v); -} - -void GSVector4i::operator = (const GSVector4& v) +GSVector4i::GSVector4i(const GSVector4& v) { m = _mm_cvttps_epi32(v); } +GSVector4::GSVector4(const GSVector4i& v) +{ + m = _mm_cvtepi32_ps(v); +} + GSVector4i GSVector4i::cast(const GSVector4& v) { return GSVector4i(_mm_castps_si128(v.m)); diff --git a/plugins/GSdx/GSVector.h b/plugins/GSdx/GSVector.h index cfd2bc0203..2b6455e273 100644 --- a/plugins/GSdx/GSVector.h +++ b/plugins/GSdx/GSVector.h @@ -121,18 +121,13 @@ public: this->m = m; } - explicit GSVector4i(const GSVector4& v) - { - *this = v; - } + explicit GSVector4i(const GSVector4& v); void operator = (const GSVector4i& v) { m = v.m; } - void operator = (const GSVector4& v); - void operator = (int i) { m = _mm_set1_epi32(i); @@ -1822,6 +1817,11 @@ public: GSVector4i* s = (GSVector4i*)src; GSVector4i* d = (GSVector4i*)dst; + if(!d[0].eq(s[0])) + { + return false; + } + GSVector4i v = GSVector4i::xffffffff(); for(int i = 0; i < size; i++) @@ -2312,18 +2312,13 @@ public: *this = GSVector4(GSVector4i::load((int)u32).u8to32()); } - explicit GSVector4(const GSVector4i& v) - { - *this = v; - } + explicit GSVector4(const GSVector4i& v); void operator = (const GSVector4& v) { m = v.m; } - void operator = (const GSVector4i& v); - void operator = (float f) { m = _mm_set1_ps(f); @@ -2625,10 +2620,10 @@ public: { GSVector4i mask = GSVector4i::x000000ff(); - a = v & mask; - b = (v >> 8) & mask; - c = (v >> 16) & mask; - d = (v >> 24); + a = GSVector4(v & mask); + b = GSVector4((v >> 8) & mask); + c = GSVector4((v >> 16) & mask); + d = GSVector4((v >> 24)); } __forceinline static void transpose(GSVector4& a, GSVector4& b, GSVector4& c, GSVector4& d) diff --git a/plugins/GSdx/GSVertexHW.h b/plugins/GSdx/GSVertexHW.h index 9948a76b1d..6314d31f39 100644 --- a/plugins/GSdx/GSVertexHW.h +++ b/plugins/GSdx/GSVertexHW.h @@ -30,9 +30,21 @@ __declspec(align(16)) union GSVertexHW9 { struct { - GSVector2 t; - union {struct {uint8 r, g, b, a;}; uint32 c0;}; - union {struct {uint8 ta0, ta1, res, f;}; uint32 c1;}; + union + { + struct + { + GSVector4 t; + }; + + struct + { + uint32 _pad[2]; + union {struct {uint8 r, g, b, a;}; uint32 c0;}; + union {struct {uint8 ta0, ta1, res, f;}; uint32 c1;}; + }; + }; + GSVector4 p; }; diff --git a/plugins/GSdx/GSdx_vs2008.vcproj b/plugins/GSdx/GSdx_vs2008.vcproj index 7addc2fb08..8de80b5287 100644 --- a/plugins/GSdx/GSdx_vs2008.vcproj +++ b/plugins/GSdx/GSdx_vs2008.vcproj @@ -1276,54 +1276,58 @@ RelativePath=".\GSRenderer.cpp" > + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - @@ -1790,22 +1794,26 @@ RelativePath=".\GSRenderer.h" > + + + + + + + + - - - - - - diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj b/plugins/GSdx/GSdx_vs2010.vcxproj index 6048be5f44..a926859248 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj +++ b/plugins/GSdx/GSdx_vs2010.vcxproj @@ -1320,9 +1320,9 @@ - - - + + + @@ -1445,9 +1445,9 @@ - - - + + + diff --git a/plugins/GSdx/GSdx_vs2010.vcxproj.filters b/plugins/GSdx/GSdx_vs2010.vcxproj.filters index a9652e38e3..4ca7e0481a 100644 --- a/plugins/GSdx/GSdx_vs2010.vcxproj.filters +++ b/plugins/GSdx/GSdx_vs2010.vcxproj.filters @@ -201,13 +201,13 @@ Source Files - + Source Files - + Source Files - + Source Files @@ -521,13 +521,13 @@ Header Files - + Header Files - + Header Files - + Header Files diff --git a/plugins/GSdx/res/tfx.fx b/plugins/GSdx/res/tfx.fx index 84ebbe190e..2bb2b3b968 100644 --- a/plugins/GSdx/res/tfx.fx +++ b/plugins/GSdx/res/tfx.fx @@ -76,10 +76,10 @@ cbuffer cb1 float3 FogColor; float AREF; float4 HalfTexel; - float2 WH; - float2 TA; + float4 WH; float4 MinMax; - float4 MinMaxF; + float2 MinF; + float2 TA; uint4 MskFix; }; @@ -145,11 +145,11 @@ float4 ps_params[5]; #define FogColor ps_params[0].bgr #define AREF ps_params[0].a #define HalfTexel ps_params[1] -#define WH ps_params[2].xy -#define TA0 ps_params[2].z -#define TA1 ps_params[2].w +#define WH ps_params[2] #define MinMax ps_params[3] -#define MinMaxF ps_params[4] +#define MinF ps_params[4].xy +#define TA0 ps_params[4].z +#define TA1 ps_params[4].w #endif @@ -198,7 +198,7 @@ float4 wrapuv(float4 uv) else if(PS_WMS == 3) { #if SHADER_MODEL >= 0x400 - uv.xz = (float2)(((int2)(uv * WH.xyxy).xz & MskFix.xx) | MskFix.zz) / WH; + uv.xz = (float2)(((int2)(uv * WH.xyxy).xz & MskFix.xx) | MskFix.zz) / WH.xy; #elif SHADER_MODEL <= 0x300 uv.x = tex1D(UMSKFIX, uv.x); uv.z = tex1D(UMSKFIX, uv.z); @@ -220,7 +220,7 @@ float4 wrapuv(float4 uv) else if(PS_WMT == 3) { #if SHADER_MODEL >= 0x400 - uv.yw = (float2)(((int2)(uv * WH.xyxy).yw & MskFix.yy) | MskFix.ww) / WH; + uv.yw = (float2)(((int2)(uv * WH.xyxy).yw & MskFix.yy) | MskFix.ww) / WH.xy; #elif SHADER_MODEL <= 0x300 uv.y = tex1D(VMSKFIX, uv.y); uv.w = tex1D(VMSKFIX, uv.w); @@ -235,15 +235,15 @@ float2 clampuv(float2 uv) { if(PS_WMS == 2 && PS_WMT == 2) { - uv = clamp(uv, MinMaxF.xy, MinMaxF.zw); + uv = clamp(uv, MinF, MinMax.zw); } else if(PS_WMS == 2) { - uv.x = clamp(uv.x, MinMaxF.x, MinMaxF.z); + uv.x = clamp(uv.x, MinF.x, MinMax.z); } else if(PS_WMT == 2) { - uv.y = clamp(uv.y, MinMaxF.y, MinMaxF.w); + uv.y = clamp(uv.y, MinF.y, MinMax.w); } return uv; @@ -476,7 +476,7 @@ float4 sample(float2 tc, float w) Texture.GetDimensions(w, h); float4 uv2 = tc.xyxy + HalfTexel; - float2 dd = frac(uv2.xy * float2(w, h)); + float2 dd = frac(uv2.xy * float2(w, h)); // * WH.zw float4 uv = wrapuv(uv2); float4 t00, t01, t10, t11; @@ -637,7 +637,7 @@ float4 sample(float2 tc, float w) else { float4 uv2 = tc.xyxy + HalfTexel; - float2 dd = frac(uv2.xy * WH); + float2 dd = frac(uv2.xy * WH.zw); float4 uv = wrapuv(uv2); float4 t00, t01, t10, t11; From 8abcc6dab20b14ed349d07da341e7116064404dc Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sat, 27 Jun 2009 19:05:36 +0000 Subject: [PATCH 291/296] GSdx: fixed a crashing in the previous revision, dx9 slowdown problems will be addressed later git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1440 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTextureCache.cpp | 39 ++++++++++++++++++++------------- plugins/GSdx/GSTextureCache.h | 2 +- 2 files changed, 25 insertions(+), 16 deletions(-) diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index d8c71c3a06..a62ec58c43 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -562,7 +562,7 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE if(right < x) { - Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); left = right = x; } @@ -576,7 +576,7 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE if(left < right) { - Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); } } } @@ -607,7 +607,7 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE { if(right < x) { - Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); left = right = x; } @@ -621,7 +621,7 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE if(left < right) { - Write(GSVector4i(left, y, right, y + s.y).rintersect(tr), buff, pitch); + Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); } } @@ -649,25 +649,34 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, s.x * s.y * sizeof(uint32) * blocks); } -void GSTextureCache::Source::Write(const GSVector4i& r, uint8* buff, int pitch) +void GSTextureCache::Source::Write(const GSVector4i& r, const GSVector4i& tr, uint8* buff, int pitch) { if(r.rempty()) return; GSLocalMemory::readTexture rtx = GSLocalMemory::m_psm[m_TEX0.PSM].rtx; - GSTexture::GSMap m; - - if(m_texture->Map(m, &r)) - { - (m_renderer->m_mem.*rtx)(r, m.bits, m.pitch, m_TEX0, m_TEXA); - - m_texture->Unmap(); - } - else + if((r > tr).mask() & 0xff00) { (m_renderer->m_mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); - m_texture->Update(r, buff, pitch); + m_texture->Update(r.rintersect(tr), buff, pitch); + } + else + { + GSTexture::GSMap m; + + if(m_texture->Map(m, &r)) + { + (m_renderer->m_mem.*rtx)(r, m.bits, m.pitch, m_TEX0, m_TEXA); + + m_texture->Unmap(); + } + else + { + (m_renderer->m_mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); + + m_texture->Update(r, buff, pitch); + } } } diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 5857ed5358..9e8af01dfb 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -67,7 +67,7 @@ public: virtual bool Create(Target* dst) = 0; virtual void Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect); - void Write(const GSVector4i& r, uint8* buff, int pitch); + void Write(const GSVector4i& r, const GSVector4i& tr, uint8* buff, int pitch); }; class Target : public Surface From 954744802041fd2e0a6d217df68adcf0431bae58 Mon Sep 17 00:00:00 2001 From: arcum42 Date: Sun, 28 Jun 2009 01:00:22 +0000 Subject: [PATCH 292/296] Worked on straightening up Plugins.cpp a bit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1441 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/Plugins.cpp | 453 ++++++++++++++++++++++++++-------------------- pcsx2/SPR.cpp | 2 +- pcsx2/VifDma.cpp | 6 +- 3 files changed, 257 insertions(+), 204 deletions(-) diff --git a/pcsx2/Plugins.cpp b/pcsx2/Plugins.cpp index 0c78f6cc41..70da26fc1e 100644 --- a/pcsx2/Plugins.cpp +++ b/pcsx2/Plugins.cpp @@ -214,10 +214,40 @@ _FWfreeze FWfreeze; _FWtest FWtest; _FWabout FWabout; - DEV9handler dev9Handler; USBhandler usbHandler; +enum PluginTypes +{ + GS = 0, + PAD, + PAD1, + PAD2, + SPU2, + CDVD, + DEV9, + USB, + FW +}; + +int PS2E_LT[9] = { +PS2E_LT_GS, +PS2E_LT_PAD,PS2E_LT_PAD, PS2E_LT_PAD, +PS2E_LT_SPU2, +PS2E_LT_CDVD, +PS2E_LT_DEV9, +PS2E_LT_USB, +PS2E_LT_FW}; + +int PS2E_VERSION[9] = { +PS2E_GS_VERSION, +PS2E_PAD_VERSION,PS2E_PAD_VERSION, PS2E_PAD_VERSION, +PS2E_SPU2_VERSION, +PS2E_CDVD_VERSION, +PS2E_DEV9_VERSION, +PS2E_USB_VERSION, +PS2E_FW_VERSION}; + #define Sfy(x) #x #define Strfy(x) Sfy(x) #define MapSymbolVarType(var,type,name) var = (type)SysLoadSym(drv,Strfy(name)) @@ -235,13 +265,13 @@ USBhandler usbHandler; #define MapSymbol_Error(name) MapSymbolVar_Error(name,name) // for pad1/2 -#define MapSymbolPAD(var,sym,name) MapSymbolVar(var##name,sym##name) -#define MapSymbolPAD_Fallback(var,sym,name) if((MapSymbolVarType(var##name,_##sym##name,sym##name))==NULL) var##name = var##_##name -#define MapSymbolPAD_Error(var,sym,name) MapSymbolVar_Error(var##name,sym##name) +#define MapSymbolPAD(var,name) MapSymbolVar(var##name,PAD##name) +#define MapSymbolPAD_Fallback(var,name) if((MapSymbolVarType(var##name,_PAD##name,PAD##name))==NULL) var##name = var##_##name +#define MapSymbolPAD_Error(var,name) MapSymbolVar_Error(var##name,PAD##name) -#define TestPS2Esyms(type) if(_TestPS2Esyms(drv,PS2E_LT_##type,PS2E_##type##_VERSION,filename) < 0) return -1; +void *GSplugin; -int _TestPS2Esyms(void* drv, int type, int expected_version, const string& filename) +static int _TestPS2Esyms(void* drv, int type, int expected_version, const string& filename) { _PS2EgetLibType PS2EgetLibType; _PS2EgetLibVersion2 PS2EgetLibVersion2; @@ -261,10 +291,11 @@ int _TestPS2Esyms(void* drv, int type, int expected_version, const string& filen return 0; } -//static const char *err; -//static int errval; - -void *GSplugin; +static __forceinline bool TestPS2Esyms(void* &drv, PluginTypes type, const string& filename) +{ + if (_TestPS2Esyms(drv, PS2E_LT[type],PS2E_VERSION[type],filename) < 0) return false; + return true; +} void CALLBACK GS_printf(int timeout, char *fmt, ...) { va_list list; @@ -292,7 +323,7 @@ int LoadGSplugin(const string& filename) GSplugin = SysLoadLibrary(filename.c_str()); if (GSplugin == NULL) { Msgbox::Alert ("Could Not Load GS Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = GSplugin; - TestPS2Esyms(GS); + if (!TestPS2Esyms(drv, GS, filename)) return -1; MapSymbol_Error(GSinit); MapSymbol_Error(GSshutdown); MapSymbol_Error(GSopen); @@ -347,24 +378,24 @@ int LoadPAD1plugin(const string& filename) { PAD1plugin = SysLoadLibrary(filename.c_str()); if (PAD1plugin == NULL) { Msgbox::Alert("Could Not Load PAD1 Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = PAD1plugin; - TestPS2Esyms(PAD); - MapSymbolPAD_Error(PAD1,PAD,init); - MapSymbolPAD_Error(PAD1,PAD,shutdown); - MapSymbolPAD_Error(PAD1,PAD,open); - MapSymbolPAD_Error(PAD1,PAD,close); - MapSymbolPAD_Error(PAD1,PAD,keyEvent); - MapSymbolPAD_Error(PAD1,PAD,startPoll); - MapSymbolPAD_Error(PAD1,PAD,poll); - MapSymbolPAD_Error(PAD1,PAD,query); - MapSymbolPAD(PAD1,PAD,update); + if (!TestPS2Esyms(drv, PAD, filename)) return -1; + MapSymbolPAD_Error(PAD1,init); + MapSymbolPAD_Error(PAD1,shutdown); + MapSymbolPAD_Error(PAD1,open); + MapSymbolPAD_Error(PAD1,close); + MapSymbolPAD_Error(PAD1,keyEvent); + MapSymbolPAD_Error(PAD1,startPoll); + MapSymbolPAD_Error(PAD1,poll); + MapSymbolPAD_Error(PAD1,query); + MapSymbolPAD(PAD1,update); - MapSymbolPAD(PAD1,PAD,gsDriverInfo); - MapSymbolPAD_Fallback(PAD1,PAD,configure); - MapSymbolPAD_Fallback(PAD1,PAD,about); - MapSymbolPAD_Fallback(PAD1,PAD,test); - MapSymbolPAD_Fallback(PAD1,PAD,freeze); - MapSymbolPAD_Fallback(PAD1,PAD,setSlot); - MapSymbolPAD_Fallback(PAD1,PAD,queryMtap); + MapSymbolPAD(PAD1,gsDriverInfo); + MapSymbolPAD_Fallback(PAD1,configure); + MapSymbolPAD_Fallback(PAD1,about); + MapSymbolPAD_Fallback(PAD1,test); + MapSymbolPAD_Fallback(PAD1,freeze); + MapSymbolPAD_Fallback(PAD1,setSlot); + MapSymbolPAD_Fallback(PAD1,queryMtap); return 0; } @@ -384,24 +415,24 @@ int LoadPAD2plugin(const string& filename) { PAD2plugin = SysLoadLibrary(filename.c_str()); if (PAD2plugin == NULL) { Msgbox::Alert("Could Not Load PAD2 Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = PAD2plugin; - TestPS2Esyms(PAD); - MapSymbolPAD_Error(PAD2,PAD,init); - MapSymbolPAD_Error(PAD2,PAD,shutdown); - MapSymbolPAD_Error(PAD2,PAD,open); - MapSymbolPAD_Error(PAD2,PAD,close); - MapSymbolPAD_Error(PAD2,PAD,keyEvent); - MapSymbolPAD_Error(PAD2,PAD,startPoll); - MapSymbolPAD_Error(PAD2,PAD,poll); - MapSymbolPAD_Error(PAD2,PAD,query); - MapSymbolPAD(PAD2,PAD,update); + if (!TestPS2Esyms(drv, PAD, filename)) return -1; + MapSymbolPAD_Error(PAD2,init); + MapSymbolPAD_Error(PAD2,shutdown); + MapSymbolPAD_Error(PAD2,open); + MapSymbolPAD_Error(PAD2,close); + MapSymbolPAD_Error(PAD2,keyEvent); + MapSymbolPAD_Error(PAD2,startPoll); + MapSymbolPAD_Error(PAD2,poll); + MapSymbolPAD_Error(PAD2,query); + MapSymbolPAD(PAD2,update); - MapSymbolPAD(PAD2,PAD,gsDriverInfo); - MapSymbolPAD_Fallback(PAD2,PAD,configure); - MapSymbolPAD_Fallback(PAD2,PAD,about); - MapSymbolPAD_Fallback(PAD2,PAD,test); - MapSymbolPAD_Fallback(PAD2,PAD,freeze); - MapSymbolPAD_Fallback(PAD2,PAD,setSlot); - MapSymbolPAD_Fallback(PAD2,PAD,queryMtap); + MapSymbolPAD(PAD2,gsDriverInfo); + MapSymbolPAD_Fallback(PAD2,configure); + MapSymbolPAD_Fallback(PAD2,about); + MapSymbolPAD_Fallback(PAD2,test); + MapSymbolPAD_Fallback(PAD2,freeze); + MapSymbolPAD_Fallback(PAD2,setSlot); + MapSymbolPAD_Fallback(PAD2,queryMtap); return 0; } @@ -419,7 +450,7 @@ int LoadSPU2plugin(const string& filename) { SPU2plugin = SysLoadLibrary(filename.c_str()); if (SPU2plugin == NULL) { Msgbox::Alert("Could Not Load SPU2 Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = SPU2plugin; - TestPS2Esyms(SPU2); + if (!TestPS2Esyms(drv, SPU2, filename)) return -1; MapSymbol_Error(SPU2init); MapSymbol_Error(SPU2shutdown); MapSymbol_Error(SPU2open); @@ -462,7 +493,7 @@ int LoadCDVDplugin(const string& filename) { CDVDplugin = SysLoadLibrary(filename.c_str()); if (CDVDplugin == NULL) { Msgbox::Alert("Could Not Load CDVD Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = CDVDplugin; - TestPS2Esyms(CDVD); + if (!TestPS2Esyms(drv, CDVD, filename)) return -1; MapSymbol_Error(CDVDinit); MapSymbol_Error(CDVDshutdown); MapSymbol_Error(CDVDopen); @@ -499,7 +530,7 @@ int LoadDEV9plugin(const string& filename) { DEV9plugin = SysLoadLibrary(filename.c_str()); if (DEV9plugin == NULL) { Msgbox::Alert("Could Not Load DEV9 Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = DEV9plugin; - TestPS2Esyms(DEV9); + if (!TestPS2Esyms(drv, DEV9, filename)) return -1; MapSymbol_Error(DEV9init); MapSymbol_Error(DEV9shutdown); MapSymbol_Error(DEV9open); @@ -536,7 +567,7 @@ int LoadUSBplugin(const string& filename) { USBplugin = SysLoadLibrary(filename.c_str()); if (USBplugin == NULL) { Msgbox::Alert("Could Not Load USB Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = USBplugin; - TestPS2Esyms(USB); + if (!TestPS2Esyms(drv, USB, filename)) return -1; MapSymbol_Error(USBinit); MapSymbol_Error(USBshutdown); MapSymbol_Error(USBopen); @@ -573,7 +604,7 @@ int LoadFWplugin(const string& filename) { FWplugin = SysLoadLibrary(filename.c_str()); if (FWplugin == NULL) { Msgbox::Alert("Could Not Load FW Plugin '%hs': %s", params &filename, SysLibError()); return -1; } drv = FWplugin; - TestPS2Esyms(FW); + if (!TestPS2Esyms(drv, FW, filename)) return -1; MapSymbol_Error(FWinit); MapSymbol_Error(FWshutdown); MapSymbol_Error(FWopen); @@ -605,13 +636,13 @@ struct PluginOpenStatusFlags static PluginOpenStatusFlags OpenStatus = {0}; -static bool loadp = false; -static bool initp = false; -static bool cdvdElf = false; +static bool plugins_loaded = false; +static bool plugins_initialized = false; +static bool only_loading_elf = false; int LoadPlugins() { - if( loadp ) return 0; + if (plugins_loaded) return 0; if (LoadGSplugin( Path::Combine( Config.PluginsDir, Config.GS )) == -1) return -1; if (LoadPAD1plugin( Path::Combine( Config.PluginsDir, Config.PAD1 )) == -1) return -1; @@ -622,80 +653,73 @@ int LoadPlugins() if (LoadUSBplugin( Path::Combine( Config.PluginsDir, Config.USB )) == -1) return -1; if (LoadFWplugin( Path::Combine( Config.PluginsDir, Config.FW )) == -1) return -1; - loadp = true; + plugins_loaded = true; return 0; } +bool ReportError(int err, const char *str) +{ + if (err != 0) + { + Msgbox::Alert("%s error: %d", params str, err); + return true; + } + return false; +} + +bool ReportError2(int err, const char *str) +{ + if (err != 0) + { + Msgbox::Alert("Error Opening %s Plugin", params str, err); + return true; + } + return false; +} int InitPlugins() { - if( initp ) return 0; + if (plugins_initialized) return 0; // Ensure plugins have been loaded.... - if( LoadPlugins() == -1 ) return -1; + if (LoadPlugins() == -1) return -1; - //if( !loadp ) - // throw Exception::InvalidOperation( "Bad coder mojo - InitPlugins called prior to plugins having been loaded." ); + //if( !plugins_loaded ) throw Exception::InvalidOperation( "Bad coder mojo - InitPlugins called prior to plugins having been loaded." ); -#ifndef _WIN32 - //chdir(MAIN_DIR); -#endif - int ret; + if (ReportError(GSinit(), "GSinit")) return -1; + if (ReportError(PAD1init(1), "PAD1init")) return -1; + if (ReportError(PAD2init(2), "PAD2init")) return -1; + if (ReportError(SPU2init(), "SPU2init")) return -1; + if (ReportError(CDVDinit(), "CDVDinit")) return -1; + if (ReportError(DEV9init(), "DEV9init")) return -1; + if (ReportError(USBinit(), "USBinit")) return -1; + if (ReportError(FWinit(), "FWinit")) return -1; - ret = GSinit(); - if (ret != 0) { Msgbox::Alert("GSinit error: %d", params ret); return -1; } - ret = PAD1init(1); - if (ret != 0) { Msgbox::Alert("PAD1init error: %d", params ret); return -1; } - ret = PAD2init(2); - if (ret != 0) { Msgbox::Alert("PAD2init error: %d", params ret); return -1; } - ret = SPU2init(); - if (ret != 0) { Msgbox::Alert("SPU2init error: %d", params ret); return -1; } - ret = CDVDinit(); - if (ret != 0) { Msgbox::Alert("CDVDinit error: %d", params ret); return -1; } - ret = DEV9init(); - if (ret != 0) { Msgbox::Alert("DEV9init error: %d", params ret); return -1; } - ret = USBinit(); - if (ret != 0) { Msgbox::Alert("USBinit error: %d", params ret); return -1; } - ret = FWinit(); - if (ret != 0) { Msgbox::Alert("FWinit error: %d", params ret); return -1; } - - cdvdElf = false; - initp = true; + only_loading_elf = false; + plugins_initialized = true; return 0; } void ShutdownPlugins() { - if( !initp ) return; + if (!plugins_initialized) return; mtgsWaitGS(); ClosePlugins( true ); - if( GSshutdown != NULL ) - GSshutdown(); + if (GSshutdown != NULL) GSshutdown(); - if( PAD1shutdown != NULL ) - PAD1shutdown(); - if( PAD2shutdown != NULL ) - PAD2shutdown(); + if (PAD1shutdown != NULL) PAD1shutdown(); + if (PAD2shutdown != NULL) PAD2shutdown(); - if( SPU2shutdown != NULL ) - SPU2shutdown(); + if (SPU2shutdown != NULL) SPU2shutdown(); + if (CDVDshutdown != NULL) CDVDshutdown(); + if (DEV9shutdown != NULL) DEV9shutdown(); + if (USBshutdown != NULL) USBshutdown(); + if (FWshutdown != NULL) FWshutdown(); - if( CDVDshutdown != NULL ) - CDVDshutdown(); - - if( DEV9shutdown != NULL ) - DEV9shutdown(); - - if( USBshutdown != NULL ) - USBshutdown(); - - if( FWshutdown != NULL ) - FWshutdown(); - - initp = false; + plugins_initialized = false; } uptr pDsp; @@ -703,131 +727,167 @@ extern void spu2DMA4Irq(); extern void spu2DMA7Irq(); extern void spu2Irq(); -int OpenPlugins(const char* pTitleFilename) +bool OpenGS() { GSdriverInfo info; - int ret; - - if ( !initp ) + + if (!OpenStatus.GS) { - if( InitPlugins() == -1 ) return -1; - } - -#ifndef _WIN32 - // change dir so that CDVD can find its config file - /*char file[255], pNewTitle[255]; - chdir(MAIN_DIR); - chdir(Config.PluginsDir); - - if( pTitleFilename != NULL && pTitleFilename[0] != '/' ) - { - // because we are changing the dir, we have to set a new title if it is a relative dir - sprintf(pNewTitle, "%s/%s", file, pTitleFilename); - pTitleFilename = pNewTitle; - }*/ -#endif - - // Don't Repetitively Open CDVD plugin if directly loading an elf file and Open failed once already - if( !OpenStatus.CDVD && !cdvdElf) - { - //first we need the data - if (CDVDnewDiskCB) CDVDnewDiskCB(cdvdNewDiskCB); - - ret = CDVDopen(pTitleFilename); - - if (ret != 0) { - if (g_Startup.BootMode != BootMode_Elf) { Msgbox::Alert("Error Opening CDVD Plugin"); goto OpenError; } - else { Console::Notice("Running ELF File Without CDVD Plugin Support!"); cdvdElf = 1; } + if (ReportError2(gsOpen(), "GS")) + { + ClosePlugins(true); + return false; } - OpenStatus.CDVD = true; - } - if( !OpenStatus.GS ) { - ret = gsOpen(); - if (ret != 0) { Msgbox::Alert("Error Opening GS Plugin"); goto OpenError; } - OpenStatus.GS = true; - - //then the user input - if (GSgetDriverInfo) { + //Get the user input. + if (GSgetDriverInfo) + { GSgetDriverInfo(&info); if (PAD1gsDriverInfo) PAD1gsDriverInfo(&info); if (PAD2gsDriverInfo) PAD2gsDriverInfo(&info); } + OpenStatus.GS = true; } + return true; +} - if( !OpenStatus.PAD1 ) +bool OpenCDVD(const char* pTitleFilename) +{ + // Don't repetitively open the CDVD plugin if directly loading an elf file and open failed once already. + if (!OpenStatus.CDVD && !only_loading_elf) { - ret = PAD1open((void *)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening PAD1 Plugin"); goto OpenError; } + //First, we need the data. + if (CDVDnewDiskCB) CDVDnewDiskCB(cdvdNewDiskCB); + + if (CDVDopen(pTitleFilename) != 0) + { + if (g_Startup.BootMode != BootMode_Elf) + { + Msgbox::Alert("Error Opening CDVD Plugin"); + ClosePlugins(true); + return false; + } + else + { + Console::Notice("Running ELF File Without CDVD Plugin Support!"); + only_loading_elf = true; + } + } + OpenStatus.CDVD = true; + } + return true; +} + +bool OpenPAD1() +{ + if (!OpenStatus.PAD1) + { + if (ReportError2(PAD1open((void *)&pDsp), "PAD1")) + { + ClosePlugins(true); + return false; + } OpenStatus.PAD1 = true; } + return true; +} - if( !OpenStatus.PAD2 ) +bool OpenPAD2() +{ + if (!OpenStatus.PAD2) { - ret = PAD2open((void *)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening PAD2 Plugin"); goto OpenError; } + if (ReportError2(PAD2open((void *)&pDsp), "PAD2")) + { + ClosePlugins(true); + return false; + } OpenStatus.PAD2 = true; } + return true; +} - //the sound - - if( !OpenStatus.SPU2 ) +bool OpenSPU2() +{ + if (!OpenStatus.SPU2) { SPU2irqCallback(spu2Irq,spu2DMA4Irq,spu2DMA7Irq); - if( SPU2setDMABaseAddr != NULL ) - SPU2setDMABaseAddr((uptr)psxM); + + if (SPU2setDMABaseAddr != NULL) SPU2setDMABaseAddr((uptr)psxM); + if (SPU2setClockPtr != NULL) SPU2setClockPtr(&psxRegs.cycle); - if(SPU2setClockPtr != NULL) - SPU2setClockPtr(&psxRegs.cycle); - - ret = SPU2open((void*)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening SPU2 Plugin"); goto OpenError; } + if (ReportError2(SPU2open((void*)&pDsp), "SPU2")) + { + ClosePlugins(true); + return false; + } OpenStatus.SPU2 = true; } + return true; +} - //and last the dev9 - if( !OpenStatus.DEV9 ) +bool OpenDEV9() +{ + if (!OpenStatus.DEV9) { DEV9irqCallback(dev9Irq); dev9Handler = DEV9irqHandler(); - ret = DEV9open(&psxRegs.pc); //((void *)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening DEV9 Plugin"); goto OpenError; } + + if (ReportError2(DEV9open(&psxRegs.pc)/*((void *)&pDsp)*/, "DEV9")) + { + ClosePlugins(true); + return false; + } OpenStatus.DEV9 = true; } - - if( !OpenStatus.USB ) + return true; +} +bool OpenUSB() +{ + if (!OpenStatus.USB) { USBirqCallback(usbIrq); usbHandler = USBirqHandler(); USBsetRAM(psxM); - ret = USBopen((void *)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening USB Plugin"); goto OpenError; } + + if (ReportError2(USBopen((void *)&pDsp), "USB")) + { + ClosePlugins(true); + return false; + } OpenStatus.USB = true; } + return true; +} - if( !OpenStatus.FW ) +bool OpenFW() +{ + if (!OpenStatus.FW) { FWirqCallback(fwIrq); - ret = FWopen((void *)&pDsp); - if (ret != 0) { Msgbox::Alert("Error Opening FW Plugin"); goto OpenError; } + + if (ReportError2(FWopen((void *)&pDsp), "FW")) + { + ClosePlugins(true); + return false; + } OpenStatus.FW = true; } + return true; +} - if( !cdvdElf ) - cdvdNewDiskCB(); +int OpenPlugins(const char* pTitleFilename) +{ + if (!plugins_initialized) + { + if( InitPlugins() == -1 ) return -1; + } -#ifndef _WIN32 - //chdir(MAIN_DIR); -#endif + if ((!OpenCDVD(pTitleFilename)) || (!OpenGS()) || (!OpenPAD1()) || (!OpenPAD2()) || + (!OpenSPU2()) || (!OpenDEV9()) || (!OpenUSB()) || (!OpenFW())) + return -1; + + if (!only_loading_elf) cdvdNewDiskCB(); return 0; - -OpenError: - ClosePlugins( true ); -#ifndef _WIN32 - //chdir(MAIN_DIR); -#endif - - return -1; } @@ -857,7 +917,9 @@ void ClosePlugins( bool closegs ) OpenStatus.GS = false; } else + { mtgsWaitGS(); + } } CLOSE_PLUGIN( CDVD ); @@ -865,16 +927,6 @@ void ClosePlugins( bool closegs ) CLOSE_PLUGIN( USB ); CLOSE_PLUGIN( FW ); CLOSE_PLUGIN( SPU2 ); - - // More special treatment for the GS. It needs a complete shutdown and re-init - // or else it will tend to error out when we try to use it again. - if( 0 ) //closegs ) - { - GSshutdown(); - - int ret = GSinit(); - if (ret != 0) { Msgbox::Alert("GSinit error: %d", params ret); } - } } //used to close the GS plugin window and pads, to switch gsdx renderer @@ -894,11 +946,11 @@ void CloseGS() void ReleasePlugins() { - if (!loadp) return; + if (!plugins_loaded) return; - if (GSplugin == NULL || PAD1plugin == NULL || PAD2plugin == NULL || - SPU2plugin == NULL || CDVDplugin == NULL || DEV9plugin == NULL || - USBplugin == NULL || FWplugin == NULL) return; + if ((GSplugin == NULL) || (PAD1plugin == NULL) || (PAD2plugin == NULL) || + (SPU2plugin == NULL) || (CDVDplugin == NULL) || (DEV9plugin == NULL) || + (USBplugin == NULL) || (FWplugin == NULL)) return; ShutdownPlugins(); @@ -910,7 +962,8 @@ void ReleasePlugins() SysCloseLibrary(DEV9plugin); DEV9plugin = NULL; SysCloseLibrary(USBplugin); USBplugin = NULL; SysCloseLibrary(FWplugin); FWplugin = NULL; - loadp = false; + + plugins_loaded = false; } void PluginsResetGS() diff --git a/pcsx2/SPR.cpp b/pcsx2/SPR.cpp index 6b8a1973b9..cb607ef85e 100644 --- a/pcsx2/SPR.cpp +++ b/pcsx2/SPR.cpp @@ -347,7 +347,7 @@ void _dmaSPR1() // toSPR work function { Console::WriteLn("SPR1 Tag BUSERR"); spr1->chcr = (spr1->chcr & 0xFFFF) | ((*ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 - psHu32(DMAC_STAT) |= 1 << 15; //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register done = TRUE; spr1finished = (done) ? 1: 0; return; diff --git a/pcsx2/VifDma.cpp b/pcsx2/VifDma.cpp index 1dd2bd0e12..7552899a5e 100644 --- a/pcsx2/VifDma.cpp +++ b/pcsx2/VifDma.cpp @@ -1385,7 +1385,7 @@ int _chainVIF0() { Console::Error("Vif0 Tag BUSERR"); vif0ch->chcr = (vif0ch->chcr & 0xFFFF) | ((*vif0ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 - psHu32(DMAC_STAT) |= 1 << 15; //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register return -1; //Return -1 as an error has occurred } @@ -2358,7 +2358,7 @@ void vif1TransferFromMemory() if (pMem == NULL) //Is vif0ptag empty? { Console::WriteLn("Vif1 Tag BUSERR"); - psHu32(DMAC_STAT) |= 1 << 15; //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register vif1.done = true; vif1Regs->stat &= ~0x1f000000; vif1ch->qwc = 0; @@ -2467,7 +2467,7 @@ __forceinline void vif1SetupTransfer() { Console::Error("Vif1 Tag BUSERR"); vif1ch->chcr = (vif1ch->chcr & 0xFFFF) | ((*vif1ptag) & 0xFFFF0000); //Transfer upper part of tag to CHCR bits 31-15 - psHu32(DMAC_STAT) |= 1 << 15; //If yes, set BEIS (BUSERR) in DMAC_STAT register + psHu32(DMAC_STAT) |= DMAC_STAT_BEIS; //If yes, set BEIS (BUSERR) in DMAC_STAT register return; //Return -1 as an error has occurred } From 7c2e51157e04752afcccfcd918cb28d4f332b64b Mon Sep 17 00:00:00 2001 From: gabest11 Date: Sun, 28 Jun 2009 02:02:14 +0000 Subject: [PATCH 293/296] GSdx: dx9 texture uploads should be at least as fast as before git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1442 96395faa-99c1-11dd-bbfe-3dabce05a288 --- plugins/GSdx/GSTextureCache.cpp | 312 ++++++++++++++++---------------- plugins/GSdx/GSTextureCache.h | 11 +- 2 files changed, 166 insertions(+), 157 deletions(-) diff --git a/plugins/GSdx/GSTextureCache.cpp b/plugins/GSdx/GSTextureCache.cpp index a62ec58c43..897f6be043 100644 --- a/plugins/GSdx/GSTextureCache.cpp +++ b/plugins/GSdx/GSTextureCache.cpp @@ -279,7 +279,10 @@ void GSTextureCache::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const { if(!s->m_target) { - s->m_valid[page] = 0; + s->m_blocks -= s->m_valid[page].count; + + s->m_valid[page].block = 0; + s->m_valid[page].count = 0; found = true; } @@ -488,6 +491,9 @@ GSTextureCache::Source::Source(GSRenderer* r) m_clut = (uint32*)_aligned_malloc(256 * sizeof(uint32), 16); memset(m_clut, 0, sizeof(m_clut)); + + m_write.rect = (GSVector4i*)_aligned_malloc(3 * sizeof(GSVector4i), 16); + m_write.count = 0; } GSTextureCache::Source::~Source() @@ -495,6 +501,8 @@ GSTextureCache::Source::~Source() m_renderer->m_dev->Recycle(m_palette); _aligned_free(m_clut); + + _aligned_free(m_write.rect); } void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect) @@ -509,175 +517,170 @@ void GSTextureCache::Source::Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TE m_TEX0 = TEX0; m_TEXA = TEXA; + if(m_blocks == m_total_blocks) + { + return; + } + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_TEX0.PSM]; - GSVector2i s = psm.bs; + GSVector2i bs = psm.bs; + + GSVector4i r = rect.ralign(bs); + + uint32 bp = m_TEX0.TBP0; + uint32 bw = m_TEX0.TBW; + + bool repeating = (1 << m_TEX0.TW) > (bw << 6); // TODO: bw == 0 + + uint32 blocks = 0; + + for(int y = r.top; y < r.bottom; y += bs.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += bs.x) + { + uint32 block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); + + if((m_valid[row].block & col) == 0) + { + if(!repeating) m_valid[row].block |= col; + + m_valid[row].count++; + + Write(GSVector4i(x, y, x + bs.x, y + bs.y)); + + blocks++; + } + } + } + } + + if(blocks > 0) + { + if(repeating) + { + for(int y = r.top; y < r.bottom; y += bs.y) + { + uint32 base = psm.bn(0, y, bp, bw); + + for(int x = r.left; x < r.right; x += bs.x) + { + uint32 block = base + psm.blockOffset[x >> 3]; + + if(block < MAX_BLOCKS) + { + uint32 row = block >> 5; + uint32 col = 1 << (block & 31); + + m_valid[row].block |= col; + } + } + } + } + + m_blocks += blocks; + + m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, bs.x * bs.y * sizeof(uint32) * blocks); + + Flush(m_write.count); + } +} + +void GSTextureCache::Source::Write(const GSVector4i& r) +{ + m_write.rect[m_write.count++] = r; + + while(m_write.count >= 2) + { + GSVector4i& a = m_write.rect[m_write.count - 2]; + GSVector4i& b = m_write.rect[m_write.count - 1]; + + if((a == b.zyxw()).mask() == 0xfff0) + { + a.right = b.right; // extend right + + m_write.count--; + } + else if((a == b.xwzy()).mask() == 0xff0f) + { + a.bottom = b.bottom; // extend down + + m_write.count--; + } + else + { + break; + } + } + + if(m_write.count > 2) + { + Flush(1); + } +} + +void GSTextureCache::Source::Flush(uint32 count) +{ + const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_TEX0.PSM]; int tw = 1 << m_TEX0.TW; int th = 1 << m_TEX0.TH; GSVector4i tr(0, 0, tw, th); - GSVector4i r = rect.ralign(s); - - uint32 bp = m_TEX0.TBP0; - uint32 bw = m_TEX0.TBW; - - uint32 blocks = 0; - // TODO - static uint8* buff = (uint8*)_aligned_malloc(1024 * 16 * sizeof(uint32), 16); // max decompressed size for a row of blocks (1024 x 16, 4bpp) + static uint8* buff = (uint8*)_aligned_malloc(1024 * 1024 * sizeof(uint32), 16); - int pitch = max(tw, s.x) * sizeof(uint32); - - GSLocalMemory::readTexture rtx = psm.rtx; + int pitch = max(tw, psm.bs.x) * sizeof(uint32); const GSLocalMemory& mem = m_renderer->m_mem; - // TODO: bw == 0 (sfex) + GSLocalMemory::readTexture rtx = psm.rtx; - if(tw <= (bw << 6)) + for(uint32 i = 0; i < count; i++) { - // r.right = min(r.right, bw << 6); + GSVector4i r = m_write.rect[i]; - for(int y = r.top; y < r.bottom; y += s.y) + if((r > tr).mask() & 0xff00) { - uint32 base = psm.bn(0, y, bp, bw); + (mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); - int left = r.left; - int right = r.left; - - for(int x = r.left; x < r.right; x += s.x) - { - uint32 block = base + psm.blockOffset[x >> 3]; - - if(block < MAX_BLOCKS) - { - uint32 row = block >> 5; - uint32 col = 1 << (block & 31); - - if((m_valid[row] & col) == 0) - { - m_valid[row] |= col; - - if(right < x) - { - Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); - - left = right = x; - } - - right += s.x; - - blocks++; - } - } - } - - if(left < right) - { - Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); - } - } - } - else - { - // unfortunatelly a block may be part of the same texture multiple times at different places (tw 1024 > tbw 640, between 640 -> 1024 it is repeated from the next row), - // so just can't set the block's bit to valid in one pass, even if 99.9% of the games don't address the repeated part at the right side - - // TODO: still bogus if those repeated parts aren't fetched together - - for(int y = r.top; y < r.bottom; y += s.y) - { - uint32 base = psm.bn(0, y, bp, bw); - - int left = r.left; - int right = r.left; - - for(int x = r.left; x < r.right; x += s.x) - { - uint32 block = base + psm.blockOffset[x >> 3]; - - if(block < MAX_BLOCKS) - { - uint32 row = block >> 5; - uint32 col = 1 << (block & 31); - - if((m_valid[row] & col) == 0) - { - if(right < x) - { - Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); - - left = right = x; - } - - right += s.x; - - blocks++; - } - } - } - - if(left < right) - { - Write(GSVector4i(left, y, right, y + s.y), tr, buff, pitch); - } - } - - for(int y = r.top; y < r.bottom; y += s.y) - { - uint32 base = psm.bn(0, y, bp, bw); - - for(int x = r.left; x < r.right; x += s.x) - { - uint32 block = base + psm.blockOffset[x >> 3]; - - if(block < MAX_BLOCKS) - { - uint32 row = block >> 5; - uint32 col = 1 << (block & 31); - - m_valid[row] |= col; - } - } - } - } - - //_aligned_free(buff); - - m_renderer->m_perfmon.Put(GSPerfMon::Unswizzle, s.x * s.y * sizeof(uint32) * blocks); -} - -void GSTextureCache::Source::Write(const GSVector4i& r, const GSVector4i& tr, uint8* buff, int pitch) -{ - if(r.rempty()) return; - - GSLocalMemory::readTexture rtx = GSLocalMemory::m_psm[m_TEX0.PSM].rtx; - - if((r > tr).mask() & 0xff00) - { - (m_renderer->m_mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); - - m_texture->Update(r.rintersect(tr), buff, pitch); - } - else - { - GSTexture::GSMap m; - - if(m_texture->Map(m, &r)) - { - (m_renderer->m_mem.*rtx)(r, m.bits, m.pitch, m_TEX0, m_TEXA); - - m_texture->Unmap(); + m_texture->Update(r.rintersect(tr), buff, pitch); } else { - (m_renderer->m_mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); + GSTexture::GSMap m; - m_texture->Update(r, buff, pitch); + if(m_texture->Map(m, &r)) + { + (mem.*rtx)(r, m.bits, m.pitch, m_TEX0, m_TEXA); + + m_texture->Unmap(); + } + else + { + (mem.*rtx)(r, buff, pitch, m_TEX0, m_TEXA); + + m_texture->Update(r, buff, pitch); + } } } + + if(count < m_write.count) + { + memcpy(m_write.rect[0], &m_write.rect[count], (m_write.count - count) * sizeof(m_write.rect[0])); + } + + m_write.count -= count; + + //_aligned_free(buff); } // GSTextureCache::Target @@ -782,7 +785,7 @@ void GSTextureCache::SourceMap::Add(Source* s, const GIFRegTEX0& TEX0) const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; - GSVector2i bs = (bp & 31) == 0 ? psm.pgs : psm.bs; + GSVector2i bs = psm.bs; int blocks = 0; @@ -792,16 +795,19 @@ void GSTextureCache::SourceMap::Add(Source* s, const GIFRegTEX0& TEX0) for(int x = 0; x < tw; x += bs.x) { - uint32 page = (base + psm.blockOffset[x >> 3]) >> 5; + uint32 block = base + psm.blockOffset[x >> 3]; - if(page < MAX_PAGES) + if(block < MAX_BLOCKS) { - m_map[page][s] = true; + m_map[block >> 5][s] = true; - s->m_pages.push_back(page); + blocks++; } } } + + s->m_blocks = 0; + s->m_total_blocks = blocks; } void GSTextureCache::SourceMap::RemoveAll() @@ -823,9 +829,9 @@ void GSTextureCache::SourceMap::RemoveAt(Source* s) { m_surfaces.erase(s); - for(list::iterator i = s->m_pages.begin(); i != s->m_pages.end(); i++) + for(int i = 0; i < countof(m_map); i++) { - m_map[*i].erase(s); + m_map[i].erase(s); } delete s; diff --git a/plugins/GSdx/GSTextureCache.h b/plugins/GSdx/GSTextureCache.h index 9e8af01dfb..277b5d046f 100644 --- a/plugins/GSdx/GSTextureCache.h +++ b/plugins/GSdx/GSTextureCache.h @@ -50,11 +50,16 @@ public: class Source : public Surface { + struct {GSVector4i* rect; uint32 count;} m_write; + + void Write(const GSVector4i& r); + void Flush(uint32 count); + public: GSTexture* m_palette; bool m_initpalette; - list m_pages; - uint32 m_valid[MAX_PAGES]; // each uint32 bits map to the 32 blocks of that page + uint32 m_blocks, m_total_blocks; + struct {uint32 block, count;} m_valid[MAX_PAGES]; // each uint32 bits map to the 32 blocks of that page uint32* m_clut; int m_bpp; bool m_target; @@ -66,8 +71,6 @@ public: virtual bool Create() = 0; virtual bool Create(Target* dst) = 0; virtual void Update(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& rect); - - void Write(const GSVector4i& r, const GSVector4i& tr, uint8* buff, int pitch); }; class Target : public Surface From 8f74247b4f49ea2d8162ece4b0fbacf227bc4c4e Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Mon, 29 Jun 2009 06:28:48 +0000 Subject: [PATCH 294/296] microVU: - Smarter microProgram comparison. Games will now use less microProgram 'slots'. Speedup in games using a lot of microPrograms at once... git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1443 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 54 +++++++++++++++++------------ pcsx2/x86/microVU.h | 17 ++++++--- pcsx2/x86/microVU_Compile.inl | 65 +++++++++++++++++++++++++++-------- pcsx2/x86/microVU_Misc.h | 3 +- 4 files changed, 97 insertions(+), 42 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index f46d5d3a6c..59a0ea433d 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -90,8 +90,11 @@ microVUt(void) mVUreset(mV) { mVU->prog.prog[i].x86ptr = z; z += (mVU->cacheSize / (mVU->prog.max + 1)); mVU->prog.prog[i].x86end = z; - mVU->prog.prog[i].range[0] = -1; // Set range to - mVU->prog.prog[i].range[1] = -1; // indeterminable status + for (int j = 0; j <= mVU->prog.prog[i].ranges.max; j++) { + mVU->prog.prog[i].ranges.range[j][0] = -1; // Set range to + mVU->prog.prog[i].ranges.range[j][1] = -1; // indeterminable status + mVU->prog.prog[i].ranges.total = -1; + } } } @@ -128,14 +131,16 @@ microVUt(void) mVUclear(mV, u32 addr, u32 size) { // Clears program data (Sets used to 1 because calling this function implies the program will be used at least once) microVUf(void) mVUclearProg(int progIndex) { microVU* mVU = mVUx; - mVU->prog.prog[progIndex].used = 1; - mVU->prog.prog[progIndex].last_used = 3; - mVU->prog.prog[progIndex].range[0] = -1; - mVU->prog.prog[progIndex].range[1] = -1; - mVU->prog.prog[progIndex].x86ptr = mVU->prog.prog[progIndex].x86start; + mVUprogI.used = 1; + mVUprogI.last_used = 3; + mVUprogI.x86ptr = mVUprogI.x86start; + for (int j = 0; j <= mVUprogI.ranges.max; j++) { + mVUprogI.ranges.range[j][0] = -1; // Set range to + mVUprogI.ranges.range[j][1] = -1; // indeterminable status + mVUprogI.ranges.total = -1; + } for (u32 i = 0; i < (mVU->progSize / 2); i++) { - if (mVU->prog.prog[progIndex].block[i]) - mVU->prog.prog[progIndex].block[i]->reset(); + if (mVUprogI.block[i]) mVUprogI.block[i]->reset(); } } @@ -161,7 +166,7 @@ microVUf(int) mVUfindLeastUsedProg() { const int pMax = mVU->prog.max; int smallidx = (mVU->prog.cur+1)&pMax; - u32 smallval = mVU->prog.prog[smallidx].used; + u64 smallval = mVU->prog.prog[smallidx].used; for (int i = 1, j = (smallidx+1)&pMax; i <= pMax; i++, j=(j+1)&pMax) { if (smallval > mVU->prog.prog[j].used) { @@ -192,9 +197,7 @@ microVUt(void) mVUvsyncUpdate(mV) { for (int i = 0; i <= mVU->prog.max; i++) { if (mVU->prog.prog[i].last_used != 0) { if (mVU->prog.prog[i].last_used >= 3) { - - if (mVU->prog.prog[i].used < 0x4fffffff) // program has been used recently. Give it a - mVU->prog.prog[i].used += 0x200; // 'weighted' bonus signifying it's importance + mVU->prog.prog[i].used += 0x200; // give 'weighted' bonus } mVU->prog.prog[i].last_used--; } @@ -202,20 +205,27 @@ microVUt(void) mVUvsyncUpdate(mV) { } } +microVUf(bool) mVUcmpPartial(int progIndex) { + microVU* mVU = mVUx; + for (int i = 0; i <= mVUprogI.ranges.total; i++) { + if (memcmp_mmx(cmpOffset(mVUprogI.data), cmpOffset(mVU->regs->Micro), ((mVUprogI.ranges.range[i][1] + 8) - mVUprogI.ranges.range[i][0]))) { + return 0; + } + } + return 1; +} + // Compare Cached microProgram to mVU->regs->Micro -microVUf(int) mVUcmpProg(int progIndex, bool progUsed, bool needOverflowCheck, bool cmpWholeProg) { +microVUf(bool) mVUcmpProg(int progIndex, bool progUsed, const bool cmpWholeProg) { microVU* mVU = mVUx; - if (progUsed) { - if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) || - (!cmpWholeProg && (!memcmp_mmx(cmpOffset(mVUprogI.data), cmpOffset(mVU->regs->Micro), ((mVUprogI.range[1] + 8) - mVUprogI.range[0]))))) { + if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) + ||(!cmpWholeProg && mVUcmpPartial(progIndex))) { mVU->prog.cur = progIndex; mVU->prog.cleared = 0; mVU->prog.isSame = cmpWholeProg ? 1 : -1; mVU->prog.prog[progIndex].last_used = 3; - if (needOverflowCheck && (mVU->prog.prog[progIndex].used < 0x7fffffff)) { - mVU->prog.prog[progIndex].used++; // increment 'used' (avoiding overflows if necessary) - } + mVU->prog.prog[progIndex].used++; // increment 'used' return 1; } } @@ -228,11 +238,11 @@ microVUf(int) mVUsearchProg() { if (mVU->prog.cleared) { // If cleared, we need to search for new program for (int i = 0; i <= mVU->prog.total; i++) { - if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 1, 0)) + if (mVUcmpProg(i, !!mVU->prog.prog[i].used, 0)) return 1; // Check Recently Used Programs } for (int i = 0; i <= mVU->prog.total; i++) { - if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0, 0)) + if (mVUcmpProg(i, !mVU->prog.prog[i].used, 0)) return 1; // Check Older Programs } mVU->prog.cur = mVUfindLeastUsedProg(); // If cleared and program not found, make a new program instance diff --git a/pcsx2/x86/microVU.h b/pcsx2/x86/microVU.h index 100dbfd7b6..3bd48c21b6 100644 --- a/pcsx2/x86/microVU.h +++ b/pcsx2/x86/microVU.h @@ -84,14 +84,21 @@ public: } }; +#define mMaxRanges 128 +struct microRange { + static const int max = mMaxRanges - 1; + int total; + s32 range[mMaxRanges][2]; +}; + #define mProgSize 0x4000/4 struct microProgram { - u32 data [mProgSize]; - microBlockManager* block[mProgSize/2]; - microIR allocInfo; - u32 used; // Number of times its been used + u32 data [mProgSize]; // Holds a copy of the VU microProgram + microBlockManager* block[mProgSize/2]; // Array of Block Managers + microIR allocInfo; // IR information + microRange ranges; // The ranges of the microProgram that have already been recompiled + u64 used; // Number of times its been used u32 last_used; // Counters # of frames since last use (starts at 3 and counts backwards to 0 for each 30fps vSync) - s32 range[2]; // The range of microMemory that has already been recompiled for the current program u8* x86ptr; // Pointer to program's recompilation code u8* x86start; // Start of program's rec-cache u8* x86end; // Limit of program's rec-cache diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index a67fb55217..b05ddbd5bf 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -79,7 +79,7 @@ microVUt(void) mVUcheckIsSame(mV) { if (mVU->prog.isSame == -1) { - mVU->prog.isSame = !memcmp_mmx(mVU->prog.prog[mVU->prog.cur].data, mVU->regs->Micro, mVU->microMemSize); + mVU->prog.isSame = !memcmp_mmx(mVUcurProg.data, mVU->regs->Micro, mVU->microMemSize); } if (mVU->prog.isSame == 0) { if (!isVU1) mVUcacheProg<0>(mVU->prog.cur); @@ -89,19 +89,56 @@ microVUt(void) mVUcheckIsSame(mV) { } // Sets up microProgram PC ranges based on whats been recompiled -microVUt(void) mVUsetupRange(mV, u32 pc) { +microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { - if (mVUcurProg.range[0] == -1) { - mVUcurProg.range[0] = (s32)pc; - mVUcurProg.range[1] = (s32)pc; + for (int i = 0; i <= mVUcurProg.ranges.total; i++) { + if ((pc >= mVUcurProg.ranges.range[i][0]) + && (pc <= mVUcurProg.ranges.range[i][1])) { return; } } - else if (mVUcurProg.range[0] > (s32)pc) { - mVUcurProg.range[0] = (s32)pc; - mVUcheckIsSame(mVU); + + mVUcheckIsSame(mVU); + + if (isStartPC) { + if (mVUcurProg.ranges.total < mVUcurProg.ranges.max) { + mVUcurProg.ranges.total++; + mVUrange[0] = pc; + } + else { + mVUcurProg.ranges.total = 0; + mVUrange[0] = 0; + mVUrange[1] = mVU->microMemSize - 8; + DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); + } } - else if (mVUcurProg.range[1] < (s32)pc) { - mVUcurProg.range[1] = (s32)pc; - mVUcheckIsSame(mVU); + else { + if (mVUrange[1] != -1) return; + if (mVUrange[0] <= pc) { + mVUrange[1] = pc; + for (int i = 0; i <= (mVUcurProg.ranges.total-1); i++) { + if ((mVUrange[0]-8) == mVUcurProg.ranges.range[i][1]) { + mVUcurProg.ranges.range[i][1] = pc; + mVUrange[0] = -1; + mVUrange[1] = -1; + mVUcurProg.ranges.total--; + //DevCon::Status("microVU%d: Prog Range Merging", params mVU->index); + } + } + } + else { + DevCon::Status("microVU%d: Prog Range Wrap", params mVU->index); + mVUrange[1] = mVU->microMemSize - 8; + if (mVUcurProg.ranges.total < mVUcurProg.ranges.max) { + mVUcurProg.ranges.total++; + mVUrange[0] = 0; + mVUrange[1] = pc; + } + else { + mVUcurProg.ranges.total = 0; + mVUrange[0] = 0; + mVUrange[1] = mVU->microMemSize - 8; + DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); + } + } } } @@ -236,7 +273,6 @@ microVUt(void) mVUendProgram(mV, int isEbit, int* xStatus, int* xMac, int* xClip } if (isEbit != 2) { // Save PC, and Jump to Exit Point - mVUsetupRange(mVU, xPC); MOV32ItoM((uptr)&mVU->regs->VI[REG_TPC].UL, xPC); JMP32((uptr)mVU->exitFunct - ((uptr)x86Ptr + 5)); } @@ -298,7 +334,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { const u32 endCount = (mVU->microMemSize / 8) - 1; // Setup Program Bounds/Range - mVUsetupRange(mVU, startPC); + mVUsetupRange(mVU, startPC, 1); // First Pass iPC = startPC / 4; @@ -360,7 +396,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { else { microBlock* bBlock = NULL; s32* ajmp = 0; - mVUsetupRange(mVU, xPC); + mVUsetupRange(mVU, xPC, 0); mVUdebugNOW(1); switch (mVUbranch) { @@ -441,6 +477,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { if (x == endCount) { Console::Error("microVU%d: Possible infinite compiling loop!", params mVU->index); } // E-bit End + mVUsetupRange(mVU, xPC, 0); mVUendProgram(mVU, 1, xStatus, xMac, xClip); return thisPtr; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index aa9c991eb6..cb252591a6 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -206,6 +206,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define cFLAG mVUinfo.cFlag #define mVUstartPC mVUallocInfo.startPC #define mVUflagInfo mVUregs.needExactMatch +#define mVUrange mVUcurProg.ranges.range[mVUcurProg.ranges.total] #define xPC ((iPC / 2) * 8) #define curI ((u32*)mVU->regs->Micro)[iPC] //mVUcurProg.data[iPC] #define setCode() { mVU->code = curI; } @@ -214,7 +215,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microMemSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) -#define cmpOffset(x) (&(((u8*)x)[mVUprogI.range[0]])) +#define cmpOffset(x) (&(((u8*)x)[mVUprogI.ranges.range[i][0]])) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL From 95579258dbd543d16ca1d6216b587bbe45e803e5 Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 30 Jun 2009 04:45:22 +0000 Subject: [PATCH 295/296] microVU: fixed some problems from my last revision. (fixes problems in DQ8) git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1444 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU.cpp | 4 ++-- pcsx2/x86/microVU_Compile.inl | 16 +++++++++------- pcsx2/x86/microVU_Misc.h | 2 +- 3 files changed, 12 insertions(+), 10 deletions(-) diff --git a/pcsx2/x86/microVU.cpp b/pcsx2/x86/microVU.cpp index 59a0ea433d..45b565b533 100644 --- a/pcsx2/x86/microVU.cpp +++ b/pcsx2/x86/microVU.cpp @@ -219,8 +219,8 @@ microVUf(bool) mVUcmpPartial(int progIndex) { microVUf(bool) mVUcmpProg(int progIndex, bool progUsed, const bool cmpWholeProg) { microVU* mVU = mVUx; if (progUsed) { - if (cmpWholeProg && (!memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) - ||(!cmpWholeProg && mVUcmpPartial(progIndex))) { + if ((cmpWholeProg && !memcmp_mmx((u8*)mVUprogI.data, mVU->regs->Micro, mVU->microMemSize)) + || (!cmpWholeProg && mVUcmpPartial(progIndex))) { mVU->prog.cur = progIndex; mVU->prog.cleared = 0; mVU->prog.isSame = cmpWholeProg ? 1 : -1; diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index b05ddbd5bf..6836109557 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -79,7 +79,7 @@ microVUt(void) mVUcheckIsSame(mV) { if (mVU->prog.isSame == -1) { - mVU->prog.isSame = !memcmp_mmx(mVUcurProg.data, mVU->regs->Micro, mVU->microMemSize); + mVU->prog.isSame = !memcmp_mmx((u8*)mVUcurProg.data, mVU->regs->Micro, mVU->microMemSize); } if (mVU->prog.isSame == 0) { if (!isVU1) mVUcacheProg<0>(mVU->prog.cur); @@ -96,6 +96,7 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { && (pc <= mVUcurProg.ranges.range[i][1])) { return; } } + const u32 endMem = mVU->microMemSize-8; mVUcheckIsSame(mVU); if (isStartPC) { @@ -106,16 +107,17 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { else { mVUcurProg.ranges.total = 0; mVUrange[0] = 0; - mVUrange[1] = mVU->microMemSize - 8; + mVUrange[1] = endMem; DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); } } else { - if (mVUrange[1] != -1) return; if (mVUrange[0] <= pc) { mVUrange[1] = pc; for (int i = 0; i <= (mVUcurProg.ranges.total-1); i++) { - if ((mVUrange[0]-8) == mVUcurProg.ranges.range[i][1]) { + if ((mVUrange[0] == mVUcurProg.ranges.range[i][1]) + ||(((mVUrange[0]-8)&endMem) == mVUcurProg.ranges.range[i][1])) { + if (mVUcurProg.ranges.range[i][1] > pc) { continue; } mVUcurProg.ranges.range[i][1] = pc; mVUrange[0] = -1; mVUrange[1] = -1; @@ -125,7 +127,7 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { } } else { - DevCon::Status("microVU%d: Prog Range Wrap", params mVU->index); + DevCon::Status("microVU%d: Prog Range Wrap [%04x] [%d]", params mVU->index, mVUrange[0], mVUrange[1]); mVUrange[1] = mVU->microMemSize - 8; if (mVUcurProg.ranges.total < mVUcurProg.ranges.max) { mVUcurProg.ranges.total++; @@ -135,7 +137,7 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { else { mVUcurProg.ranges.total = 0; mVUrange[0] = 0; - mVUrange[1] = mVU->microMemSize - 8; + mVUrange[1] = endMem; DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); } } @@ -477,7 +479,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { if (x == endCount) { Console::Error("microVU%d: Possible infinite compiling loop!", params mVU->index); } // E-bit End - mVUsetupRange(mVU, xPC, 0); + mVUsetupRange(mVU, xPC-8, 0); mVUendProgram(mVU, 1, xStatus, xMac, xClip); return thisPtr; } diff --git a/pcsx2/x86/microVU_Misc.h b/pcsx2/x86/microVU_Misc.h index cb252591a6..5f4d22bcf0 100644 --- a/pcsx2/x86/microVU_Misc.h +++ b/pcsx2/x86/microVU_Misc.h @@ -215,7 +215,7 @@ typedef u32 (__fastcall *mVUCall)(void*, void*); #define bSaveAddr (((xPC + (2 * 8)) & ((isVU1) ? 0x3ff8:0xff8)) / 8) #define branchAddr ((xPC + 8 + (_Imm11_ * 8)) & (mVU->microMemSize-8)) #define shufflePQ (((mVU->p) ? 0xb0 : 0xe0) | ((mVU->q) ? 0x01 : 0x04)) -#define cmpOffset(x) (&(((u8*)x)[mVUprogI.ranges.range[i][0]])) +#define cmpOffset(x) ((u8*)&(((u8*)x)[mVUprogI.ranges.range[i][0]])) #define Rmem (uptr)&mVU->regs->VI[REG_R].UL #define Roffset (uptr)&mVU->regs->VI[9].UL From 60e1fcffb7ec3da38704f1d2e0330c582bd9d6bb Mon Sep 17 00:00:00 2001 From: cottonvibes Date: Tue, 30 Jun 2009 20:42:25 +0000 Subject: [PATCH 296/296] microVU: Tweaked some stuff from my last commit. git-svn-id: http://pcsx2.googlecode.com/svn/trunk@1445 96395faa-99c1-11dd-bbfe-3dabce05a288 --- pcsx2/x86/microVU_Compile.inl | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/pcsx2/x86/microVU_Compile.inl b/pcsx2/x86/microVU_Compile.inl index 6836109557..4a01c6a31b 100644 --- a/pcsx2/x86/microVU_Compile.inl +++ b/pcsx2/x86/microVU_Compile.inl @@ -96,7 +96,6 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { && (pc <= mVUcurProg.ranges.range[i][1])) { return; } } - const u32 endMem = mVU->microMemSize-8; mVUcheckIsSame(mVU); if (isStartPC) { @@ -107,7 +106,7 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { else { mVUcurProg.ranges.total = 0; mVUrange[0] = 0; - mVUrange[1] = endMem; + mVUrange[1] = mVU->microMemSize - 8; DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); } } @@ -115,9 +114,10 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { if (mVUrange[0] <= pc) { mVUrange[1] = pc; for (int i = 0; i <= (mVUcurProg.ranges.total-1); i++) { - if ((mVUrange[0] == mVUcurProg.ranges.range[i][1]) - ||(((mVUrange[0]-8)&endMem) == mVUcurProg.ranges.range[i][1])) { - if (mVUcurProg.ranges.range[i][1] > pc) { continue; } + int rStart = (mVUrange[0] < 8) ? 0 : (mVUrange[0] - 8); + int rEnd = pc; + if((mVUcurProg.ranges.range[i][1] >= rStart) + && (mVUcurProg.ranges.range[i][1] <= rEnd)){ mVUcurProg.ranges.range[i][1] = pc; mVUrange[0] = -1; mVUrange[1] = -1; @@ -137,7 +137,7 @@ microVUt(void) mVUsetupRange(mV, s32 pc, bool isStartPC) { else { mVUcurProg.ranges.total = 0; mVUrange[0] = 0; - mVUrange[1] = endMem; + mVUrange[1] = mVU->microMemSize - 8; DevCon::Status("microVU%d: Prog Range List Full", params mVU->index); } } @@ -488,7 +488,7 @@ microVUr(void*) mVUcompile(microVU* mVU, u32 startPC, uptr pState) { microVUt(void*) mVUblockFetch(microVU* mVU, u32 startPC, uptr pState) { using namespace x86Emitter; - if (startPC > mVU->microMemSize-8) { DevCon::Error("microVU%d: invalid startPC", params mVU->index); } + if (startPC > mVU->microMemSize-8) { DevCon::Error("microVU%d: invalid startPC [%04x]", params mVU->index, startPC); } startPC &= mVU->microMemSize-8; blockCreate(startPC/8);