aboutsummaryrefslogtreecommitdiffstats
path: root/src/gallium/drivers/r300/r300_emit.c
Commit message (Collapse)AuthorAgeFilesLines
* r300g: final fix for r3xx constant buffer emissionMarek Olšák2010-07-171-1/+1
|
* r300g: fix typo in r3xx constant buffer emissionMarek Olšák2010-07-171-1/+1
| | | | Ooops.
* r300g: fix constant buffer emission on r3xxMarek Olšák2010-07-171-1/+1
| | | | FDO bug #29128.
* r300g: do not make copies of constant buffers, emit them directlyMarek Olšák2010-07-161-8/+8
|
* r300g: rebuild winsys and command submission to support multiple contextsMarek Olšák2010-07-161-60/+37
|
* r300g: extend and clean up debug loggingMarek Olšák2010-07-131-3/+3
|
* r300g: Remove unnecessary header.Vinson Lee2010-07-121-1/+0
|
* r300g: implement fast color clearMarek Olšák2010-07-121-5/+39
| | | | | | | | | | | | | | | An initial implementation made by Dave Airlie. For it to be used, a color-only clear must be invoked and exactly one point-sampled render target must be set. The render target must be macrotiled (for us to overcome alignment issues) and bpp must be either 16 or 32. I can't see a difference in performance. :( Conflicts: src/gallium/drivers/r300/r300_blit.c
* r300g: allow the GTT domain for samplersMarek Olšák2010-07-091-3/+4
| | | | This fixes sluggishness in vdrift.
* r300g: separate the hyperz state and pipelined FB regs out of the FB stateMarek Olšák2010-06-271-3/+20
|
* r300g: group debug messagesMarek Olšák2010-06-271-21/+24
|
* r300g: turn rasterizer state into a command bufferMarek Olšák2010-06-271-37/+5
|
* r300g: move emission of the MSPOS regs into the framebuffer stateMarek Olšák2010-06-271-41/+36
| | | | | | | | Now the question is whether we are allowed to ignore gl_rasterization_rules and pipe_rasterizer_state::multisample. The former is invariant anyway and I think the latter would need re-emitting the AA state which is quite costly, considering that it implicitly flushes the whole pipeline (all AA regs in the AA state are *unpipelined*).
* r300g: turn invariant state into a command bufferMarek Olšák2010-06-271-0/+7
|
* r300g: introduce VAP invariant stateMarek Olšák2010-06-261-10/+7
| | | | Unlike other invariant states, this one must be emitted after VAP flush.
* r300g: remove an XXX commentMarek Olšák2010-06-241-7/+0
|
* r300g: move AA registers into the new AA stateMarek Olšák2010-06-241-10/+21
|
* r300g: separate the cache flush from the framebuffer stateMarek Olšák2010-06-241-16/+17
|
* r300g: reorder and cleanup register writes everywhereMarek Olšák2010-06-241-7/+43
|
* r300g: fix random pixels appearing / incomplete renderingMarek Olšák2010-06-191-12/+19
| | | | | | | | | | | | This should fix the FDO bug #28612. Also, these piglit tests have been fixed: - fbo-copypix - scissor-copypixels - copytexsubimage - texredefine Finally, 2 flushes in the transfer path are no longer needed.
* r300g: more reasonable MSPOS defaults in error pathMarek Olšák2010-06-171-2/+2
|
* r300g: Fix uninitialized variable warnings in error path.Vinson Lee2010-06-161-0/+2
|
* r300g: Moar MSAA setup.Corbin Simpson2010-06-161-8/+46
| | | | Need to just add the resolve, then go switch to new DRM and test.
* r300g: Old MSAA code from before gallium-msaa.Corbin Simpson2010-06-161-0/+7
|
* r300g: fix uploading RC state shader constants on r3xxMarek Olšák2010-06-141-1/+1
| | | | | | I've messed this up in one of my previous commits. Reported-by: Igor Murzov
* r300g: rewrite occlusion queriesMarek Olšák2010-06-141-16/+25
| | | | | | | | The previous implementation had issues with queries spanning over several command streams as well as using a very large number of queries. This fixes flickering in Enemy Territory: Quake Wars. The driver now renders everything correctly in this game and the graphics is awesome.
* r300g: emit viewport state as a dword tableMarek Olšák2010-06-141-6/+1
|
* r300g: subclass pipe_surfaceMarek Olšák2010-06-141-17/+10
|
* r300g: optimize emission of fragment shader constantsMarek Olšák2010-06-131-23/+12
|
* r300g: turn fragment shader into a CBMarek Olšák2010-06-131-123/+3
|
* r300g: turn depth stencil state into a CBMarek Olšák2010-06-131-16/+2
|
* r300g: turn clip state into a CBMarek Olšák2010-06-131-16/+2
|
* r300g: turn blend color into a CBMarek Olšák2010-06-131-11/+1
|
* r300g: turn blend state into a CBMarek Olšák2010-06-131-12/+2
|
* r300g: fix multiple render targetsMarek Olšák2010-06-131-11/+6
| | | | This fixes tests/drawbuffers.
* r300g: skip rendering for unsupported vertex formats instead of abortingMarek Olšák2010-06-081-12/+4
| | | | The next step is to replace skipping by an actual fallback.
* r300g: workaround index bias on chipsets that do not support itMarek Olšák2010-06-061-1/+1
|
* r300g: use the stored GEM domain in add_buffer/write_relocMarek Olšák2010-06-021-26/+26
| | | | | I admit blitting to/from GTT is actually enforced in this commit, texture transfers are now damned fast!
* r300g: use r300_buffer_domain everywhereMarek Olšák2010-06-021-21/+21
|
* r300g/swtcl: fix WPOSMarek Olšák2010-05-281-7/+6
| | | | | | | | There you are! r300->viewport_state is undefined when using SW TCL. Piglit score (r300g.tests): HW TCL: 231/275 SW TCL: 233/275
* r300g/swtcl: force vertex prefetching for non-indexed primitivesMarek Olšák2010-05-261-2/+2
|
* r300g/swtcl: move emitting AOS to prepare_for_renderingMarek Olšák2010-05-261-7/+1
|
* r300g/swtcl: do not use u_upload_mgr and do not compute max_indexMarek Olšák2010-05-261-1/+1
|
* r300g: Add RS block dump.Corbin Simpson2010-05-171-0/+4
| | | | Shows what's actually going on in the RS block.
* r300g: fix BO space accounting for occlusion queriesMarek Olšák2010-05-151-1/+4
|
* r300g: rename a few emit_query functionsMarek Olšák2010-05-151-9/+9
|
* r300g: Fix indentation.Corbin Simpson2010-05-121-10/+10
|
* r300g: move GA_POINT_MINMAX back to rs_stateMarek Olšák2010-05-081-3/+3
| | | | This basically reverts commit 74f94e8fdfe035fa68acdc19e6b0afc2957a4264.
* r300g: consolidate common render code into one functionMarek Olšák2010-05-081-2/+0
| | | | | | | | | | This reduces redundant code by moving: - CS space reservation - buffer validation - dirty state emission - index bias emission - AOS emission into r300_prepare_for_rendering.
* r300g: merge VS output mapping state (VAP) to rasterizer block state (RS)Marek Olšák2010-05-021-19/+7
| | | | | | | | | | | | These two should be tied together because what's set in VAP or stuffed in GA should be rasterized in RS. Not doing so causes a hardlock. The reason for the merge is that if stuffed texture coordinates (e.g. point sprite texgen) happen to occupy the texcoord slot dedicated to fog or wpos, the two must be relocated to other free slots, which needs remapping the vertex shader outputs. The rasterizer code is now literally a sequence read-rasterize-write.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-17 20:55:13 +0000