summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* nv50: occlusion queriesBen Skeggs2009-01-132-13/+73
| | | | Not quite working, but the general idea is right I think.
* nv50: implement KIL enough for progs/fp/kil to workBen Skeggs2009-01-131-0/+28
|
* cell: Add missing suffix to SHUFFLE macroJonathan Adamczewski2009-01-131-1/+1
|
* cell: allocate batch buffers w/ 16-byte alignmentJonathan Adamczewski2009-01-128-143/+111
| | | | | | | | | | | Replace cell_batch{align,alloc)*() with cell_batch_alloc16(), allocating multiples of 16 bytes that are 16 byte aligned. Opcodes are stored in preferred slot of SPU machine word. Various structures are explicitly padded to 16 byte multiples. Added STATIC_ASSERT().
* gallivm: Make sure the bitcode buffer is followed by a 0 byte.Michel Dänzer2009-01-123-4/+4
| | | | May fail to parse otherwise.
* gallivm: Print error message from ParseBitcodeFile() in case it fails.Michel Dänzer2009-01-121-2/+3
|
* gallivm: Adapt to header file move in LLVM 2.4.Michel Dänzer2009-01-125-5/+5
|
* nv50: make rtt work againBen Skeggs2009-01-121-1/+33
|
* nv50: fix handling of depth texturesBen Skeggs2009-01-122-7/+29
|
* nv50: another typo..Ben Skeggs2009-01-121-1/+1
|
* nouveau: fix warningBen Skeggs2009-01-121-1/+1
|
* nv50: fix assertion failureBen Skeggs2009-01-121-2/+9
|
* nv50: remove previous hack to manage tiled surfacesBen Skeggs2009-01-122-33/+1
|
* nv50: fix a typo and a thinkoBen Skeggs2009-01-121-3/+4
|
* nv50: enable npot texturesBen Skeggs2009-01-121-1/+1
|
* nv50: disable shader debugBen Skeggs2009-01-121-1/+7
|
* nv50: any cpu access to a texture is done on its backing imagesBen Skeggs2009-01-123-7/+72
| | | | | | Still a little dodgy: - RTT will hit an assertion (hopefully!) and fail - 3D textures with depth >= 32 will cause bad things to happen
* nv50: create buffers for each image that makes up a textureBen Skeggs2009-01-122-13/+38
|
* nouveau: return buffer map to something sane.Ben Skeggs2009-01-121-20/+0
| | | | | | Sorry, but no, we're not doing this.. Correctness always takes precedence over speed. Implement this higher up where you know it's safe to do so, and doesn't break other things in the process.
* nouveau: use usage, not uninitialised flags value...Ben Skeggs2009-01-121-1/+1
|
* cell: optimize unpack_colors() function, saving 12 cyclesBrian Paul2009-01-111-31/+34
|
* cell: move color unpacking code into separate functionBrian Paul2009-01-111-76/+89
|
* cell: re-order the z/stencil fetch/extract/convert instructions for better perfBrian Paul2009-01-111-55/+51
| | | | The new instruction order is 10 cycles faster.
* cell: datatype clean-ups in SPE rtasmBrian Paul2009-01-112-105/+99
|
* cell: simplify the 'optional register' codeBrian Paul2009-01-111-62/+50
|
* cell: asst datatype clean-upsBrian Paul2009-01-111-73/+75
|
* cell: move depth/stencil code into separate functionBrian Paul2009-01-111-179/+213
|
* cell: clean-up, re-indent, commentsBrian Paul2009-01-111-95/+139
|
* cell: use tgsi_dump_instruction() instead of spe_comment()Brian Paul2009-01-101-101/+25
|
* gallium: use tgsi_dump_instruction() instead of ppc_comment()Brian Paul2009-01-101-84/+7
|
* gallium: remove unused struct typeBrian Paul2009-01-101-8/+0
|
* gallium: fix register clobber bug in TGSI->PPC codegenBrian Paul2009-01-101-10/+41
| | | | | | | | When negating a src vector that's stored in a altivec register, need to put negated value into a new register so we don't upset the original value. This solves the dark colors in the mandelbrot GLSL demo. Also, use new predicate functions to check if a TGSI temp is stored in an altivec register.
* gallium: emit comments in TGSI->PPC codegenBrian Paul2009-01-101-2/+90
|
* gallium: code to dump/debug PPC code (disabled)Brian Paul2009-01-101-0/+5
|
* gallium: added comment/annotation support to PPC rtasmBrian Paul2009-01-102-62/+187
|
* cell: added rule to produce .s assembly filesBrian Paul2009-01-101-0/+3
|
* gallium: s/false/FALSE/Brian Paul2009-01-101-1/+1
|
* nouveau: Update nv30 swizzling.Younes Manton2009-01-102-9/+10
|
* nouveau: Factor out common winsys bits into libnouveaudrm.aYounes Manton2009-01-1044-922/+927
|
* g3dvl: Use Gallium thread wrappers.Younes Manton2009-01-101-4/+4
|
* g3dvl: Use Gallium MALLOC wrappers.Younes Manton2009-01-109-31/+34
|
* g3dvl: Get rid of old unbuffered motion compensation code.Younes Manton2009-01-104-2365/+1
|
* g3dvl: Map vertex bufs once per frame/flush.Younes Manton2009-01-101-36/+48
|
* nouveau: Catch some more leaks.Younes Manton2009-01-102-0/+3
|
* g3dvl: Fix some memory leaks.Younes Manton2009-01-102-0/+21
|
* g3dvl: Allocate one set of bufs, let winsys rename them as necessary.Younes Manton2009-01-102-158/+178
|
* nouveau: Add busy() query, determines if BOs can be mapped immediately.Younes Manton2009-01-102-0/+37
|
* nouveau: Use PIPE_BUFFER_USAGE_CPU_* instead of custom.Younes Manton2009-01-104-5/+4
|
* g3dvl: Expand YCbCr to full RGB range by default.Younes Manton2009-01-101-1/+1
|
* g3dvl: Work around mplayer weirdness in XvMCPutSurface().Younes Manton2009-01-101-2/+8
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-24 16:47:00 +0000