Henrik Rydgård
9bbdd1907d
Kind-of optimized ARM software skinning (non-NEON)
2013-11-13 18:11:35 +01:00
Henrik Rydgård
821a2f10f8
Delete obsolete code
2013-11-13 18:10:58 +01:00
Henrik Rydgard
6221dbaf5d
Optimize software skinning for x86.
...
Can't seem to get a win on Windows vs hardware skinning though, even though
draw calls drop by 2/3rd...
2013-11-13 18:10:58 +01:00
Henrik Rydgard
9333d3ea76
Vtx dec jit: Combine the scale and offset registers to save 1 more xmm register.
2013-11-13 18:10:58 +01:00
Henrik Rydgard
f0cacf46d0
No reason to involve the FPU when loading matrices
2013-11-13 18:10:58 +01:00
Henrik Rydgard
6976d6a3a0
Enable the softskinning optimizations that let us merge drawcalls
2013-11-13 18:10:57 +01:00
Henrik Rydgard
179934ec9f
Decode step by step when sw skinning
2013-11-13 18:10:57 +01:00
Henrik Rydgard
46313ced55
Prepare transform pipeline for step by step decoding
2013-11-13 18:10:57 +01:00
Henrik Rydgard
4f78eda23b
Save a couple of registers in the x86 vertex decoder jit by SIMD-ing prescale UV
2013-11-13 18:10:57 +01:00
Henrik Rydgard
7e67476b00
Simple unoptimized software skinning.
...
Does not take advantage of the possible reduction in state changes yet.
2013-11-13 18:10:57 +01:00
Henrik Rydgård
ab3fe9ba86
Extract the software transform code into its own file.
2013-11-13 14:56:34 +01:00
Henrik Rydgård
67ca4419fe
Fix minor bug where we didn't clear dirtyUniforms if early-outing from UpdateUniform.
2013-11-13 10:02:58 +01:00
Sacha
04b338f39e
Buildfix for sse2 builds.
2013-11-13 12:08:46 +10:00
Henrik Rydgard
cf15ec8a53
Add BBOX support (very conservative test)
2013-11-12 17:06:03 +01:00
Henrik Rydgard
54217deb16
Speed up UpdateUniforms a little
2013-11-12 17:06:03 +01:00
Henrik Rydgard
4b98e0d6d6
Optimize LoadClut a little
2013-11-12 17:06:03 +01:00
Henrik Rydgard
84f20a1cad
Small optimizations
2013-11-12 14:05:50 +01:00
Sacha
aad70d8912
Symbian: Support software GPU. It seems to compile properly now. May allow Nokia 500 to run PPSSPP (albeit very slowly).
2013-11-12 16:20:52 +10:00
Henrik Rydgård
662ad32d4f
Fix and re-enable @unknownbracket's handcoded NEON tex hash
2013-11-11 17:12:38 +01:00
Henrik Rydgård
53800bd413
Disable the hand-coded NEON texture hash again. It's slow, for unknown reasons.
2013-11-11 14:39:29 +01:00
raven02
2bca62b26e
Don't reset texture width/height unless the size is different
2013-11-11 21:12:43 +08:00
raven02
d2546bed5b
Regression fix c69ac64
2013-11-11 08:53:47 +08:00
Unknown W. Brackets
1311d71455
softgpu: Fix reversed stencil test funcs.
...
LESS applies to the ref compared to the stencil, not the stencil compared
to the ref.
With this, stencil testing is finally being done properly in Star Ocean.
2013-11-10 03:21:01 -08:00
Unknown W. Brackets
2db98b8669
softgpu: Don't apply logic op to stencil value.
...
Also, they were not at all applying correctly.
2013-11-10 03:21:00 -08:00
Unknown W. Brackets
438928de40
softgpu: Apply stencil op even when depth disabled.
2013-11-10 03:21:00 -08:00
Unknown W. Brackets
5327ce80d5
softgpu: throughmode doesn't disable any tests.
2013-11-10 03:20:59 -08:00
Unknown W. Brackets
5751e0a0a4
softgpu: Simplify color clamping.
2013-11-10 03:20:58 -08:00
Unknown W. Brackets
21deaef35a
Invert gstate.getClearModeColorMask().
...
It makes more sense matching gstate.getColorMask().
2013-11-10 03:20:58 -08:00
Unknown W. Brackets
5817684e28
softgpu: Never write alpha, only stencil.
2013-11-10 03:20:57 -08:00
Unknown W. Brackets
be9fe9ad9a
softgpu: Fix stencil read/write in some formats.
2013-11-10 03:20:56 -08:00
Henrik Rydgard
c69ac64d83
Don't reconvert light colors if they don't change.
...
Also prepare for a possible further optimization in GLES_GPU::FastRunLoop
2013-11-10 11:18:58 +01:00
Henrik Rydgard
f4ad7c64e5
Fix issue with texcoord speed hack (bPrescaleUV) in software transform
...
(and also thus rectangles of course even when hw transform is enabled)
2013-11-10 11:18:26 +01:00
Henrik Rydgård
179068823c
Merge pull request #4491 from raven02/patch-15
...
Attempt for another matching framebuffer logic
2013-11-10 01:58:29 -08:00
Unknown W. Brackets
1633aa689c
Remove the extra process queues hack.
...
It seems like it's not helping anymore so it could be hurting.
2013-11-09 23:08:44 -08:00
Unknown W. Brackets
afafa5fcd6
Make sure the gpu thread has a chance to run.
...
If the scheduler puts it on the same core, it may not even do anything
before we check if the framebuffer is dirty, so SyncThread will quit since
it's not even running.
Instead, let's wait until it's at least done something.
2013-11-09 23:08:28 -08:00
raven02
7c6a4cf87e
Attempt for another matching framebuffer logic
2013-11-10 10:38:33 +08:00
Henrik Rydgård
345a4ccf0e
Merge pull request #4487 from unknownbrackets/perf
...
Texture cache hashing tweaks, disable second cache on mobile
2013-11-09 17:30:40 -08:00
sum2012
e4b68099e0
Attempt to fix Colin Mcrae Dirt 2 Crash correctly
...
#4482
2013-11-10 08:23:31 +08:00
Unknown W. Brackets
fa79d7f13c
Handcode the non-NEON texcache hash as well on ARM.
...
Well, it could be a better algorithm, but at least this uses LDMIA.
2013-11-09 14:51:23 -08:00
Unknown W. Brackets
2dfa2379f4
Rewrite QuickTexHashNEON directly in asm.
...
Seems like gcc couldn't optimize it to this.
2013-11-09 12:54:55 -08:00
Unknown W. Brackets
3f57f1f447
Disable the secondary texcache on mobile.
...
It helps in games like Final Fantasy 2 and Popolocrois, but it seems to
cause out of memory errors despite the checks.
2013-11-09 12:52:58 -08:00
Henrik Rydgard
497915141b
Revert to normal for loop in the index generator, the < 3 case gets taken care of properly. Just existing caused some issues.
2013-11-09 19:33:40 +01:00
sum2012
886fcdd0b1
Fix Colin Mcrae Dirt 2 Crash
...
#4482
2013-11-10 02:23:22 +08:00
Henrik Rydgard
63334698e1
Add temporary setting to disable the vertex decoder jit while we debug it
2013-11-09 18:16:26 +01:00
Henrik Rydgård
d04415ac5d
Merge pull request #4479 from unknownbrackets/perf
...
Small optimizations (armjit) for imms
2013-11-08 11:45:16 -08:00
Unknown W. Brackets
385f9a457e
Use GetPointerUnchecked in how, pre-checked areas.
2013-11-08 11:39:24 -08:00
Henrik Rydgard
0e542b6ecc
vtxdec: small omission
2013-11-08 20:38:49 +01:00
Henrik Rydgard
4daebe9b0f
When logging PRIM commands, print the prim number
2013-11-08 20:36:42 +01:00
Henrik Rydgard
22d6b36005
Vertexdecoder "double": fix for x86 and very minor optimization for arm.
2013-11-08 20:03:28 +01:00
Henrik Rydgard
6b45c321b6
vtxdecjit: turn off excessive logging
2013-11-08 18:49:17 +01:00
Henrik Rydgård