Files
ppsspp/Core/MIPS/IR/IRCompLoadStore.cpp
T
Henrik RydgårdandClaude Opus 5.5 c5d2adb303 IR: Compile data cache ops to nothing
The interpreter does nothing for any cache op but icache invalidation, but
the IR only knew four of them and sent the rest to the interpreter one at
a time. Media Engine code flushes its data cache with a loop of 0x14s,
often every pass of its main loop: with the ME emulated (on a branch),
mcidclan's me-simple-count went from 15.9 to 1.0 seconds for 3 emulated
ones.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-20 21:18:21 +02:00

169 lines
4.7 KiB
C++

// Copyright (c) 2012- PPSSPP Project.
// 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, version 2.0 or later versions.
// 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 2.0 for more details.
// A copy of the GPL 2.0 should have been included with the program.
// If not, see http://www.gnu.org/licenses/
// Official git repository and contact information can be found at
// https://github.com/hrydgard/ppsspp and http://www.ppsspp.org/.
#include "Core/MemMap.h"
#include "Core/MIPS/MIPS.h"
#include "Core/MIPS/MIPSAnalyst.h"
#include "Core/MIPS/MIPSCodeUtils.h"
#include "Core/MIPS/IR/IRFrontend.h"
#include "Core/MIPS/IR/IRRegCache.h"
#define _RS MIPS_GET_RS(op)
#define _RT MIPS_GET_RT(op)
#define _RD MIPS_GET_RD(op)
#define _FS MIPS_GET_FS(op)
#define _FT MIPS_GET_FT(op)
#define _FD MIPS_GET_FD(op)
#define _SA MIPS_GET_SA(op)
#define _POS ((op>> 6) & 0x1F)
#define _SIZE ((op>>11) & 0x1F)
#define _IMM16 (signed short)(op & 0xFFFF)
#define _IMM26 (op & 0x03FFFFFF)
// All functions should have CONDITIONAL_DISABLE, so we can narrow things down to a file quickly.
// Currently known non working ones should have DISABLE.
// #define CONDITIONAL_DISABLE(flag) { Comp_Generic(op); return; }
#define CONDITIONAL_DISABLE(flag) if (opts.disableFlags & (uint32_t)JitDisable::flag) { Comp_Generic(op); return; }
#define DISABLE { Comp_Generic(op); return; }
#define INVALIDOP { Comp_Generic(op); return; }
namespace MIPSComp {
void IRFrontend::Comp_ITypeMem(MIPSOpcode op) {
CONDITIONAL_DISABLE(LSU);
int offset = _IMM16;
MIPSGPReg rt = _RT;
MIPSGPReg rs = _RS;
int o = op >> 26;
CheckMemoryBreakpoint(rs, offset);
if (((op >> 29) & 1) == 0 && rt == MIPS_REG_ZERO) {
// Don't load anything into $zr
return;
}
if (js.kernelMode) {
// Kernel code may reach hardware registers. ResolveKernelMemory turns the plain memory
// accesses back into ordinary ones where it can tell. The rest go to the interpreter.
switch (o) {
case 32: ir.Write(IROp::LoadKernel8Ext, rt, rs, 0, offset); return;
case 33: ir.Write(IROp::LoadKernel16Ext, rt, rs, 0, offset); return;
case 35: ir.Write(IROp::LoadKernel32, rt, rs, 0, offset); return;
case 36: ir.Write(IROp::LoadKernel8, rt, rs, 0, offset); return;
case 37: ir.Write(IROp::LoadKernel16, rt, rs, 0, offset); return;
case 40: ir.Write(IROp::StoreKernel8, rt, rs, 0, offset); return;
case 41: ir.Write(IROp::StoreKernel16, rt, rs, 0, offset); return;
case 43: ir.Write(IROp::StoreKernel32, rt, rs, 0, offset); return;
default:
DISABLE;
}
}
switch (o) {
// Load
case 35:
ir.Write(IROp::Load32, rt, rs, 0, offset);
break;
case 37:
ir.Write(IROp::Load16, rt, rs, 0, offset);
break;
case 33:
ir.Write(IROp::Load16Ext, rt, rs, 0, offset);
break;
case 36:
ir.Write(IROp::Load8, rt, rs, 0, offset);
break;
case 32:
ir.Write(IROp::Load8Ext, rt, rs, 0, offset);
break;
// Store
case 43:
ir.Write(IROp::Store32, rt, rs, 0, offset);
break;
case 41:
ir.Write(IROp::Store16, rt, rs, 0, offset);
break;
case 40:
ir.Write(IROp::Store8, rt, rs, 0, offset);
break;
case 34: //lwl
ir.Write(IROp::Load32Left, rt, rs, 0, offset);
break;
case 38: //lwr
ir.Write(IROp::Load32Right, rt, rs, 0, offset);
break;
case 42: //swl
ir.Write(IROp::Store32Left, rt, rs, 0, offset);
break;
case 46: //swr
ir.Write(IROp::Store32Right, rt, rs, 0, offset);
break;
default:
INVALIDOP;
return;
}
}
void IRFrontend::Comp_StoreSync(MIPSOpcode op) {
CONDITIONAL_DISABLE(LSU);
int offset = _IMM16;
MIPSGPReg rt = _RT;
MIPSGPReg rs = _RS;
// Note: still does something even if loading to zero.
CheckMemoryBreakpoint(rs, offset);
if (js.kernelMode) {
// Send all memory accesses to the interpreter in kernel mode.
// TODO: Do something faster - but it hardly matters, currently this is VSH-only.
DISABLE;
return;
}
switch (op >> 26) {
case 48: // ll
ir.Write(IROp::Load32Linked, rt, rs, 0, offset);
break;
case 56: // sc
ir.Write(IROp::Store32Conditional, rt, rs, 0, offset);
break;
default:
INVALIDOP;
}
}
void IRFrontend::Comp_Cache(MIPSOpcode op) {
CONDITIONAL_DISABLE(LSU);
int func = (op >> 16) & 0x1F;
// See Int_Cache: only icache invalidation does anything, so the data cache ops (which ME
// code loops over to flush its cache) compile to nothing.
if (func == 8) {
// Fall back to the interpreter.
DISABLE;
}
}
}