Emulator-Archive/ppsspp
1
0
Fork 0
mirror of https://github.com/hrydgard/ppsspp.git synced 2025-02-26 00:45:49 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity
12,502 Commits 80 Branches 122 Tags
Commit Graph

2221 Commits

Author SHA1 Message Date
Unknown W. Brackets
466a5418d2 Cut time in EstimatePerVertexCost(). 
From profling, this cuts 4% time while lighting is enabled, and 14% while
disabled (on ARM.)  But maybe should just inline this...
 2014-03-30 00:42:26 -07:00
Unknown W. Brackets
7cde2311e0 Include sceKernelThread.h in less headers. 2014-03-29 17:02:41 -07:00
Unknown W. Brackets
a4327702f1 Reduce some includes under GPU/. 2014-03-29 16:51:38 -07:00
Henrik Rydgård
4c437218e4 Merge pull request #5766 from unknownbrackets/gpu-minor 
Update some flags in GLES_GPU
 2014-03-29 17:25:19 -04:00
Unknown W. Brackets
27550a2336 Update some flags in GLES_GPU. 
It's useful to know if we need to have a correct PC value or downcount.
 2014-03-29 14:10:35 -07:00
Unknown W. Brackets
0950e7b86d Optimize ComputeVertexShaderID() a bit. 
This is only like 1%.  Along the way found and fixed an overlap of bits.
 2014-03-29 10:35:09 -07:00
Unknown W. Brackets
1209bc7ba4 Optimize ComputeFragmentShaderID() a bit. 
I've seen it decently high on some profiles (1-2%).  This cuts it down by
12.5% when used in a tight loop.  The temporary was 8%, the ifs were the
other 4.5%.
 2014-03-29 10:21:04 -07:00
Unknown W. Brackets
2b7d2fdc4b Optimize > and != 0 alpha tests. 
There's no need to do rounding in this case.  These are very common.

Also, if the test was !=, convert it to >, which may work better on
PowerVR.
 2014-03-29 10:16:06 -07:00
Henrik Rydgard
f99925261d Merge branch 'raven02-patch-42' 
Conflicts:
	GPU/GLES/FragmentShaderGenerator.cpp
 2014-03-29 17:26:33 +01:00
raven02
664efe6812 Return false when alpha blending disabled 2014-03-29 22:04:59 +08:00
raven02
2944856c42 FS : Add missing gstate.isTextureMapEnabled() 2014-03-29 21:03:28 +08:00
Henrik Rydgard
37e9fc59b3 2x alpha: Safer change than the last attempt. Still fixes P2. I think this can be simplified, anyway. 2014-03-29 12:10:44 +01:00
Henrik Rydgard
27767fa85c Revert "Rethink 2x src alpha->color doubling. Fixes #3379." 
This reverts commit 47efded29fbde6e5cee23ad413b1da6e7566a7c0.
 2014-03-29 12:01:47 +01:00
Henrik Rydgard
47efded29f Rethink 2x src alpha->color doubling. Fixes #3379. 2014-03-29 11:40:17 +01:00
Henrik Rydgard
94c1e56d9b Merge branch 'gpu-minor' of github.com:unknownbrackets/ppsspp into unknownbrackets-gpu-minor 
Conflicts:
	GPU/GLES/VertexDecoder.cpp
 2014-03-29 10:15:57 +01:00
Henrik Rydgard
74f593362b Non-jit vertex decoder, morph: Clamp colors to 0-255. 2014-03-29 10:07:51 +01:00
Unknown W. Brackets
4b1737b0cd vertexjit: Oops, don't use NEON when not supported. 2014-03-29 02:05:10 -07:00
Unknown W. Brackets
18f6407705 Clamp morph colors in non-vertexjit. 
Not sure if any platforms use this...
 2014-03-29 01:53:28 -07:00
Unknown W. Brackets
3adab534c7 vertexjit: Fix dumb mistakes in NEON color morphs. 2014-03-29 01:00:02 -07:00
Unknown W. Brackets
162bd98877 vertexjit: Fix Jit_WriteMorphColor() on ARM. 
It was totally wrong.
 2014-03-29 00:24:34 -07:00
Unknown W. Brackets
6b606aa124 vertexjit: Fix some stupid mistakes in x86 morph. 
Good to have somewhere to test it properly.
 2014-03-29 00:15:35 -07:00
raven02
e0fae728ad Update few comments for 2x src/dst blending 2014-03-28 20:11:37 +08:00
Henrik Rydgard
1d21306dcc Use the VLD1 solution for the skinning problem, instead of the old VFP path. 2014-03-25 20:40:29 +01:00
Henrik Rydgård
98b6f0db3a Merge pull request #5717 from thedax/win32NVidiaVSync 
Win32/NVidia: go fullscreen later than previous builds, to avoid vsync being ignored
 2014-03-25 18:10:49 +01:00
Unknown W. Brackets
69c2500dc6 Make sure vai->flags are set while hashing. 
Since we decode the verts in this case we have a fresh flag.  Might
be #5718?
 2014-03-25 08:19:38 -07:00
Henrik Rydgard
6d648b075f Disable an optimization that broke Daxter (#5723) 2014-03-25 11:09:15 +01:00
Henrik Rydgård
cf3d75975d Merge pull request #5720 from unknownbrackets/texcache 
Optimize xxHash for ARM/NEON devices
 2014-03-25 09:32:15 +01:00
Unknown W. Brackets
006ed53e73 Use defines on x86/SSE to avoid func pointers. 
Unfortunately no such luck for Android.
 2014-03-25 01:28:09 -07:00
Henrik Rydgård
9ba51adb89 Merge pull request #5719 from unknownbrackets/gpu-minor 
Don't calculate lighting when not enabled (software transform)
 2014-03-25 09:19:33 +01:00
Unknown W. Brackets
b800762ceb Add a NEON-optimized version of XXH32. 
This takes at least 40% less time to hash on NEON/ARM devices.
 2014-03-25 00:34:55 -07:00
Unknown W. Brackets
a414abfe4a Don't calculate lighting when not enabled. 
We just ignore it anyway.  2% improvement in Tales of Phantasia X.
 2014-03-24 22:57:58 -07:00
The Dax
1ca3d39a50 Disable the adaptive vsync checks by commenting them out. No need to have a disabled check run every frame. 2014-03-24 20:15:41 -04:00
Henrik Rydgård
ff498ed63b Merge pull request #5710 from hrydgard/avoid-alpha-test 
Avoid alpha test when vertexFullAlpha && textureFullAlpha
 2014-03-25 00:51:53 +01:00
Henrik Rydgard
dc07d3410a More checks for alpha test elimination 2014-03-24 17:33:20 +01:00
Henrik Rydgard
f30d0d810e ARM buildfix 2014-03-24 14:44:46 +01:00
raven02
ffcbb8703d Divide by 65536 instead of 65535 2014-03-24 21:33:29 +08:00
Henrik Rydgard
f33ddad364 Eliminate further alpha tests 2014-03-24 12:41:33 +01:00
Henrik Rydgard
97e7270ec9 Merge branch 'unknownbrackets-vertexalpha' into avoid-alpha-test 
Conflicts:
	GPU/GLES/VertexDecoderX86.cpp
 2014-03-24 11:27:45 +01:00
Henrik Rydgard
975c564893 Merge branch 'vertexalpha' of github.com:unknownbrackets/ppsspp into unknownbrackets-vertexalpha 2014-03-24 11:22:05 +01:00
Henrik Rydgard
2c76e6d023 Correctly keep track of "full alpha" in vertices (x86 jit only). 2014-03-24 11:19:44 +01:00
Henrik Rydgard
6260ac4bc7 Fix the latest fog issue for software transform as well 2014-03-24 10:57:53 +01:00
Henrik Rydgard
f2f0355d94 Split up ApplyShader into ApplyVertexShader and ApplyFragmentShader. 
Will allow the alphatest avoiding optimization later.
 2014-03-24 10:55:07 +01:00
Unknown W. Brackets
604160f60c Switch to fullAlphaReg not hasAlphaReg. 2014-03-24 01:27:13 -07:00
Unknown W. Brackets
61f5d3d360 Initial stab at tracking vertex alpha. 
Not sure what efficient method to use on x86...
 2014-03-24 01:26:55 -07:00
Unknown W. Brackets
b589d3b170 vertexjit: Fix a silly mistake in weights > 4. 
Darn switch, took me way too long to notice this.
 2014-03-23 19:02:40 -07:00
Unknown W. Brackets
7be619cad2 vertexjit: Fix typo in NEON weights. 2014-03-23 17:50:53 -07:00
Unknown W. Brackets
75ca544e73 vertexjit: Avoid the remaining VFP instrs. 
So now, unless there's skin+morph, it should be all-NEON, I think.
 2014-03-23 16:26:22 -07:00
Unknown W. Brackets
5ee4ba54b7 vertexjit: Use NEON loads for up to 8 weights. 2014-03-23 16:26:21 -07:00
Unknown W. Brackets
cb02d66d8f vertexjit: Use NEON for prescale as well. 
Just to keep the VFP out of the picture.
 2014-03-23 16:26:20 -07:00
Unknown W. Brackets
0dc881e9aa vertexjit: Avoid the VFP more in some weight funcs. 2014-03-23 16:26:19 -07:00