summaryrefslogtreecommitdiffstats
path: root/src/gallium
Commit message (Collapse)AuthorAgeFilesLines
* [g3dvl] rework pot buffer handling and flushingChristian König2010-12-073-87/+83
|
* [g3dvl] move mapping/unmapping of buffers one layer upChristian König2010-12-053-34/+49
|
* [g3dvl] move vertex handling into vl_vb objectChristian König2010-12-055-117/+94
|
* [g3dvl] cleanup empty block handlingChristian König2010-12-052-70/+26
| | | | doing empty block handling in the mc code is indeed faster
* move empty block handling back into mc for testingChristian König2010-12-035-111/+190
|
* add rasterizer stateChristian König2010-12-034-5/+34
|
* use CMP also for referenz frame fetchChristian König2010-12-031-43/+71
|
* cleanup and use CMP instead of IF ELSE ENDIFChristian König2010-12-031-20/+10
|
* move to four component calculation for idct codeChristian König2010-12-031-70/+52
|
* copy only mv really needed to vbChristian König2010-12-031-11/+14
|
* use vertex buffer also for mc codeChristian König2010-12-032-162/+162
|
* give each vertex element its own bufferChristian König2010-12-032-78/+85
|
* use vl_vb_upload_quads also for mcChristian König2010-12-031-37/+1
|
* move vertex elemt states into mb type handlersChristian König2010-12-032-133/+72
|
* move macroblock type handling into its own structureChristian König2010-12-032-290/+151
|
* use a shadow buffer for vertex data to optimize memory accessChristian König2010-12-036-129/+249
|
* make nr of render targets configureable for testingChristian König2010-12-031-8/+23
|
* [g3dvl] join empty blocks to get larger slicesChristian König2010-11-282-52/+91
|
* [g3dvl] use 8 zslices for idctChristian König2010-11-272-65/+60
|
* [g3dvl] some more bugfixingChristian König2010-11-271-4/+2
|
* [g3dvl] idividual vs for each stage and a bunch of bugsfixesChristian König2010-11-272-68/+104
|
* [g3dvl] give idct stage 1 & 2 its own sb_state and viewportChristian König2010-11-272-24/+28
|
* [g3dvl] use four component fetch also for idct sourceChristian König2010-11-262-15/+16
|
* r600g: disable hardware blit for stream textureChristian König2010-11-261-1/+2
|
* [g3dvl] split matrix mul into seperate functionsChristian König2010-11-261-64/+114
|
* [g3dvl] use inline constants instead of vs_const for idctChristian König2010-11-263-55/+18
|
* [g3dvl] no need to keep the idct matrix multiple timesChristian König2010-11-253-54/+67
|
* [g3dvl] add dump option to xvmcChristian König2010-11-251-0/+14
|
* [g3dvl] also use four elemets on right side multiplikationChristian König2010-11-251-8/+11
|
* [g3dvl] use four elements in matrix texture fetchChristian König2010-11-241-19/+31
|
* r600g: disable not working formatsChristian König2010-11-241-5/+5
|
* r600g: reenable texture uploads, but keep R16_SNORM disabledChristian König2010-11-241-2/+5
|
* r600g: disable R32 float also in r600_translate_colorformatChristian König2010-11-241-2/+2
|
* [g3dvl] no need for all samplers at all stagesChristian König2010-11-242-17/+16
|
* [g3dvl] remove flushing between stagesChristian König2010-11-241-4/+0
|
* [g3dvl] spread scaling between idct stagesChristian König2010-11-241-18/+8
|
* [g3dvl] remove invalid use of assertChristian König2010-11-232-17/+27
|
* [g3dvl] switch to r32 float for idct matrixChristian König2010-11-231-6/+6
|
* [g3dvl] add some error handlingChristian König2010-11-232-35/+55
|
* Merge remote branch 'origin/master' into pipe-videoChristian König2010-11-2181-796/+1684
|\ | | | | | | | | | | Conflicts: src/gallium/auxiliary/Makefile src/gallium/auxiliary/SConscript
| * i915g: kill idws->poolDaniel Vetter2010-11-214-14/+7
| | | | | | | | | | | | | | | | | | The drm winsys only ever handles one gem memory manager. Rip out the unnecessary complication. Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * i915g: kill buf->map_gttDaniel Vetter2010-11-212-15/+2
| | | | | | | | | | | | | | | | | | Not using the gtt is considered harmful for performance. And for partial uploads there's always drm_intel_bo_subdata. Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * i915g: kill RGBA/X formatsDaniel Vetter2010-11-211-4/+0
| | | | | | | | | | | | | | | | It's intel, so always little endian! Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * i915g: add pineview pci idsDaniel Vetter2010-11-212-0/+10
| | | | | | | | | | | | Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * i915g: s/hw_tiled/tilingDaniel Vetter2010-11-212-6/+7
| | | | | | | | | | | | | | | | | | | | More in line with other intel drivers. Change to use enum by Jakob Bornecrantz. Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * i915g: rip out ->sw_tiledDaniel Vetter2010-11-214-15/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It looks like this was meant to facilitate unfenced access to textures/ color/renderbuffers. It's totally incomplete and fundamentally broken on a few levels: - broken: The kernel needs to about every tiled bo to fix up bit17 swizzling on swap-in. - unflexible: fenced/unfenced relocs from execbuffer2 do the same, much simpler. - unneeded: with relaxed fencing tiled gem bos are as memory-efficient as this trick. Hence kill it. Reviewed-by: Jakob Bornecrantz <[email protected]> Signed-off-by: Daniel Vetter <[email protected]> Signed-off-by: Jakob Bornecrantz <[email protected]>
| * r300g: silence guard band cap errorsJoakim Sindholt2010-11-211-0/+7
| | | | | | | | | | | | Somebody should find out what these are. It can be found on Windows getting a D3DCAPS9 from IDirect3D9::GetCaps() and reading the GuardBand* values.
| * st/vega: Fix vgReadPixels with a subrectangle.Chia-I Wu2010-11-211-4/+12
| | | | | | | | | | | | | | Fix a crash when the subrectangle is not inside the fb. Fix wrong pipe transfer when sx > 0 or sy + height != fb->height. This fixes "readpixels" demo.
| * st/vega: Set wrap_r for mask and blend samplers.Chia-I Wu2010-11-211-0/+2
| | | | | | | | | | | | | | | | | | | | | | These two samplers use non-normalized texture coordinates. wrap_r cannot be PIPE_TEX_WRAP_REPEAT (the default). This fixes sp_tex_sample.c:1790:get_linear_unorm_wrap: Assertion `0' failed assertion failure.
| * st/vega: vegaLookupSingle should validate the state.Chia-I Wu2010-11-211-0/+2
| | | | | | | | Fix "lookup" demo crash.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-27 10:52:31 +0000