Commit Graph
7530 Commits
Author SHA1 Message Date
Unknown W. Brackets 0f36b45f2e Turns out this difference began in mpeg.prx 0105. 2014-03-02 10:40:54 -08:00
Unknown W. Brackets f300f46b9c Speed up MIPSGetInstruction() by 35%.
It's not that hot, but gets called a lot when loading modules / running
tests.  Well, and also interpreter.

Some of the previous checks were just not necessary (we never have Rese in
the tables, for example, and let's just not.)
2014-03-02 10:20:41 -08:00
Unknown W. Brackets eca06c60c7 Spend a bit less time in Read_Instruction on load. 2014-03-02 10:20:40 -08:00
Unknown W. Brackets e3506ead69 Clear the symbol map between each test.
This way we could, for example, test function replacement properly.

Actually slows down tests because it spends more time correctly
determining the function boundaries.
2014-03-02 10:20:40 -08:00
Henrik Rydgård 8f90681994 Merge pull request #5589 from unknownbrackets/mpeg-version
Track the sceMpeg_library version for behavior differences
2014-03-02 23:14:36 +07:00
Unknown W. Brackets fc1ab072b3 Write the ringbuffer struct based on mpeg version.
This way it can easily match tests.
2014-03-02 01:31:30 -08:00
Unknown W. Brackets cda1bc2772 Keep track of the loaded sceMpeg_library version.
They have variations in behavior.
2014-03-02 01:31:29 -08:00
Unknown W. Brackets 87a07d7355 Workaround bug in Kurohyou 2 (uninitialized data.)
The game calls scePsmfGetEPWithTimestamp(), but without any chapters.
This means nothing is written to its pointer.  But it takes the value
there anyway.

It happens that the value there is 0, but only because something wrote it
to the stack before the call.  It seems like scePsmfVerifyPsmf() is the
one writing it, since it has a different value shortly before that call.

So, as a workaround, just write some 0s into the stack.  Games without
bugs should not care.  Hard to predict what will happen to another buggy
game.
2014-03-02 01:22:25 -08:00
raven02 6888ea8d79 Typo fix 2014-03-02 15:12:50 +08:00
Unknown W. Brackets da1e4a80a6 Tone down sas raw output reporting, found a game.
Crimson Gem Saga apparently uses this mode.
2014-03-01 15:32:34 -08:00
Unknown W. Brackets c7f1197ab1 Cut down on a debug check in release mode.
In Jewel Summoner, for example, this gets called a lot, and eats a
measurable percentage of time.
2014-03-01 15:14:17 -08:00
Unknown W. Brackets 444dd15a25 Report module versions of HLE'd modules.
sceMpeg has slightly different functionality, so collecting the unique
versions will help.  So far I know of 4.
2014-03-01 14:32:42 -08:00
Unknown W. Brackets 1b5e54abb4 Fix nearest size check in optimum, cleanup. 2014-03-01 02:23:56 -08:00
Unknown W. Brackets 58e29062a5 Improve error checking in sceFontGetNumFontList(). 2014-03-01 01:54:15 -08:00
Unknown W. Brackets 160d5e3279 Fix sceFontGetFontList() error codes. 2014-03-01 01:47:35 -08:00
Unknown W. Brackets 140982223a Correct sceFontGetFontInfoByIndexNumber().
It just gets font style info, pretty simple.
2014-03-01 01:30:43 -08:00
Unknown W. Brackets 0e39b6fe36 Fix results of sceFontGetFontInfo() for user fonts. 2014-03-01 01:30:25 -08:00
Unknown W. Brackets 131cbc073c Implement sceFontGetShadowInfo(). 2014-03-01 00:18:19 -08:00
Unknown W. Brackets 29d2e81676 Cleanup errors in sceFontGet*ImageRect(). 2014-02-28 23:04:34 -08:00
Unknown W. Brackets 96a93d9bae Fix error handling in sceFontPointToPixelH()/etc. 2014-02-28 22:05:37 -08:00
Unknown W. Brackets 57e64547b3 Fix sceFontSetResolution() error codes. 2014-02-28 21:52:34 -08:00
Unknown W. Brackets 67fcdc0f4a Reflect font altCharCode in psp mem for tests.
Makes it easier to be sure it's working right.  Not completely impossible
a game will check it... but hopefully not.
2014-02-28 21:31:02 -08:00
Unknown W. Brackets bc3e123b61 Oops, fix negative clip regions.
Since they're summed, need to clamp the left/top side.  And the
bottom/right side is offset by x/y.
2014-02-28 21:25:45 -08:00
Unknown W. Brackets 692cffffd1 Use the correct default for font altCharCode. 2014-02-28 21:20:06 -08:00
Unknown W. Brackets 46370e2860 Implement sceFontGetShadowGlyphImage().
Not sure if any games actually use it, though...
2014-02-28 20:57:18 -08:00
Unknown W. Brackets 5dec831abc Reset initing if boot fails with threading off.
Fixes #5580.
2014-02-28 09:58:10 -08:00
Henrik Rydgård 909fc3d3f9 Merge pull request #5576 from thedax/decryptedEbootDump
Add a basic decrypted EBOOT.BIN dump feature.
2014-02-27 18:46:18 +07:00
Unknown W. Brackets 4a1c9645e8 Get rid of Memory::GetStruct(), use PSPPointer. 2014-02-27 00:28:57 -08:00
Unknown W. Brackets ab6b998710 Correct clipping in sceFontGetCharGlyphImage_Clip.
It was not clipping completely, and didn't handle negative clip w/h.
Also, reduce to debug log (seems to work fine now and has been working
fine in most games.)
2014-02-27 00:19:40 -08:00
The Dax 3fbf715d78 Fix wrong dump length. 2014-02-27 01:54:00 -05:00
The Dax 8c31e1ca21 Add a basic decrypted EBOOT.BIN dump function.
It won't dump already decrypted EBOOTs.
2014-02-27 01:35:17 -05:00
Henrik Rydgård 056968129c Merge pull request #5571 from unknownbrackets/font
Correct sceFindFont*() behavior
2014-02-25 23:28:37 +07:00
Unknown W. Brackets e0258bf347 Match sceFindFont*() behavior much more accurately.
Per tests.  Fixes bad font selection in Fieldrunners.
2014-02-25 08:16:10 -08:00
Henrik Rydgård 680283908c Merge pull request #5570 from thedax/warningFix
Fix GCC compiler warning.
2014-02-25 21:02:16 +07:00
The Dax 5bca8f1119 Fix GCC compiler warning. 2014-02-25 07:39:01 -05:00
Unknown W. Brackets db4e4c1c32 Improve sceFontFindFont() errors. 2014-02-25 00:43:24 -08:00
Henrik Rydgård 10a9d4967a Merge pull request #5565 from thedax/ps1EbootCheck
Add additional check to prevent PS1 EBOOTs from loading.
2014-02-25 15:16:27 +07:00
Henrik Rydgård 71f0f9d722 Merge pull request #5566 from chinhodado/patch-5
remove a check that is always false
2014-02-25 15:15:48 +07:00
Unknown W. Brackets 0e1dd5cb33 Improve error codes in sceFontFindOptimumFont(). 2014-02-24 23:42:22 -08:00
The Dax ffcae56209 Actually force the pause button on Symbian, iOS, and Meego Harmattan.
Instead of leaving a loophole open which could allow an ini edit, just force it on for these.
2014-02-25 01:57:32 -05:00
The Dax 57ed2149ef Let the touchscreen pause/menu button be available for everybody.
It's still forced ON on iOS, Symbian, and Meego Harmattan.
2014-02-25 01:39:10 -05:00
chinhodado 4f7a7d9434 remove a check that is always false 2014-02-25 00:22:52 -05:00
The Dax 7f89863319 Add additional check to prevent PS1 EBOOTs from loading. 2014-02-24 23:16:26 -05:00
Henrik Rydgård afdaaf3170 Merge pull request #5554 from hilesaz/master
Implemented sceKernelICacheInvalidateRange
2014-02-24 17:58:53 +07:00
Henrik Rydgård 40d867f74a Merge pull request #5551 from unknownbrackets/sas-minor
Revert resampling changes
2014-02-24 17:56:59 +07:00
Henrik Rydgård 1dbb925393 Merge pull request #5552 from unknownbrackets/font
Improve error codes in sceFont
2014-02-24 14:19:07 +07:00
Bovine 139c7ff5bb Implemented sceKernelICacheInvalidateRange
Fixes crash in mhp3 hot springs after some missions
2014-02-23 19:25:18 -07:00
Unknown W. Brackets 7298133c9a Allow open fonts in a few more calls.
It turns out if you open the same internal font twice, they both work
until you close both of them (it's reference counted, but the font handles
are different.)

Anyway, these funcs seem to crash on a PSP when used with a non-open font
so there's not a ton of risk here.
2014-02-23 11:48:31 -08:00
Unknown W. Brackets be79670efb Revert "Smooth SAS resampling when pitch != 0x1000."
This reverts commit 1efcebb121.
2014-02-23 08:48:49 -08:00
Unknown W. Brackets c806ae0982 Revert "Optimize mixing samples at common pitches."
This reverts commit dd434cd7c5.

Conflicts:
	Core/HW/SasAudio.cpp
2014-02-23 08:48:30 -08:00