summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2010-03-12Revert "st/mesa: Always recalculate invalid index bounds."Corbin Simpson
This reverts commit 50876ddaaff72a324ac45e255985e0f84e108594. Per ML discussion.
2010-03-12gallivm: cube map sampling works nowBrian Paul
2010-03-12gallivm: fix bugs in nested if/then/else codegenBrian Paul
Quite a bit a hair pulling was involved...
2010-03-12gallivm: support non-vector float in lp_build_sgn()Brian Paul
2010-03-12i965: Clarify the roles of emit_pixel_xy(), emit_delta_xy(), emit_wpos_xy().Eric Anholt
2010-03-12i965: Clarify that DELTAXY always occurs for both X and Y.Eric Anholt
2010-03-12i965: Do FS SLT, SGT, and friends using CMP, SEL instead of CMP, MOV, MOV.Eric Anholt
2010-03-12i965: Do VS SGT, SLT, and friends using CMP, SEL instead of CMP, MOV, MOV.Eric Anholt
2010-03-12i965: Fix up VS DP4 sequences to avoid dependency control.Eric Anholt
This is recommended by the B-Spec. I wasn't able to measure any difference in ETQW.
2010-03-12i965: When doing a swizzled kill pixel, don't do redundant channel compares.Eric Anholt
This was obvious when looking at the compiled output of ETQW's shaders.
2010-03-12i965: Use the SEL instruction to implement MIN and MAX.Eric Anholt
Saves an instruction over doing conditional moves.
2010-03-12svga: Fix up for context transfers.Michel Dänzer
2010-03-12st/xorg: Remove flushes no longer necessary thanks to context transfers.Michel Dänzer
2010-03-12glx/single2.c: Don't call __builtin_expect on non-gnu compilersAlan Coopersmith
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
2010-03-12Merge branch '7.8'Michel Dänzer
2010-03-12vmwgfx/dri: Fix SCons build.Michel Dänzer
2010-03-12st/dri: Always try to set up R5G6B5 configs.Michel Dänzer
Allows compiz to work in depth 16. The DRI2 getBuffersWithFormat hook is only required for 16/32 bit depth buffers, for colour buffers the only requirement is that the format matches the drawable depth, which we can't check here. (cherry picked from commit c50477c255a34444720fb944c54373462ef39fb9)
2010-03-12r100/r200/r300/r300: only enable accelerated pixel ops with kmsAlex Deucher
fixes fdo bug 27043
2010-03-12dri/r700: include shader/programopt.h instead of programopt.c.Luc Verhaegen
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-03-12Grammar and spelling fixesJeff Smith
Signed-off-by: Jeff Smith <whydoubt@yahoo.com> Signed-off-by: Brian Paul <brianp@vmware.com>
2010-03-12st/xorg: context transfersKeith Whitwell
2010-03-12gallium: fix BGRA vertex color swizzlesMarek Olšák
The mapping for vertex_array_bgra: (gl -> st -> translate) GL_RGBA -> PIPE_FORMAT_R8G8B8A8 (RGBA) -> no swizzle (XYZW) GL_BGRA -> PIPE_FORMAT_A8R8G8B8 (ARGB) -> ZYXW (BGRA again??) Iẗ́'s pretty clear that PIPE_FORMAT_A8R8G8B8 here is wrong. This commit fixes the pipe format and removes obvious workarounds in util/translate. Tested with: softpipe, llvmpipe, r300g. Signed-off-by: José Fonseca <jfonseca@vmware.com>
2010-03-12st/mesa: Always recalculate invalid index bounds.Corbin Simpson
These should always be sanitized before heading towards the pipe driver, and if the calling function explicitly marked them as invalid, we need to regenerate them. Allows r300g to properly pass a bit more of Wine's d3d9 testing without dropping stuff on the floor.
2010-03-12st/mesa: Update debug infos.Corbin Simpson
Still commented out, of course.
2010-03-12r300g: Properly clamp index limits.Corbin Simpson
Make the kernel even less likely to barf on us.
2010-03-12Merge commit 'origin/gallium-context-transfers-2'Keith Whitwell
2010-03-11radeon: Fix memory leaks from early return.Vinson Lee
2010-03-11llvmpipe: Remove unnecessary header.Vinson Lee
2010-03-11gallivm: checkpoint WIP cubemap codeBrian Paul
2010-03-11gallivm: added lp_build_sum_vector()Brian Paul
2010-03-11softpipe: further tighen up sample_cube()Brian Paul
The code can fairly easily be translated to llvm...
2010-03-11softpipe: tighten up the code in sample_cube()Brian Paul
2010-03-11st/mesa: s/BUFFER_FRONT_LEFT/surfIndex/Brian Paul
2010-03-11gallivm: enable 3D texture samplingBrian Paul
2010-03-11gallivm/llvmpipe: replace 'int stride' with 'int row_stride[MAX_LEVELS]'Brian Paul
The stride depends on the mipmap level. Rename to row_stride to distinguish from img_stride for 3D textures. Fixes incorrect texel addressing in small mipmap levels.
2010-03-11llvmpipe: fix comment typoBrian Paul
2010-03-11gallivm/llvmpipe: include os_llvm.h instead of llvm-c/Core.hBrian Paul
2010-03-11gallium/os: wrapper for llvm-c/Core.h and #define HAVE_LLVM if neededBrian Paul
2010-03-11st/mesa: fix st_set_framebuffer_surface() state validationBrian Paul
Set the _NEW_BUFFERS flag and remove the code which updated the parent framebuffer size. Normal Mesa state validation will do that. Fixes issues with Warsow on r300g and possibly other bugs.
2010-03-11r200: support additional blit formatsAlex Deucher
swizzle in the pixel shader
2010-03-11gallivm: fix some bugs on the 1D texture pathsBrian Paul
2010-03-11st/mesa: don't enable extensions which aren't actually supportedRoland Scheidegger
don't enable APPLE_client_storage, TDFX_texture_compression_FXT1, EXT_cull_vertex, NV_vertex_program, NV_vertex_program1_1 - the latter two might work somewhat with some luck. Also don't enable ARB_imaging.
2010-03-11cell: adapt to context texture transfers (untested)Roland Scheidegger
2010-03-11gallium: remove pipe_context member from pipe_transferKeith Whitwell
There was very little use for this beyond permitting the pipe_context::tex_transfer_destroy() function to omit the pipe_context argument. This change adds the pipe_context argument into tex_transfer_destroy() so that it looks like other pipe_context functions, and removes the pipe_context pointer from pipe_transfer.
2010-03-11ARB prog parser: added (float) casts and regenerate filesBrian Paul
2010-03-11gallivm: include tgsi_dump.h to silence warningBrian Paul
2010-03-11gallivm: move declarations to silence unused var warningsBrian Paul
2010-03-11gallium/util: use memset() to initialize vars to avoid warningsBrian Paul
2010-03-11svga: use memset() to initialize u to avoid warningsBrian Paul
2010-03-11gallivm: Use bitmasks for scalar masks.José Fonseca
We could use single 1 bit conditions for scalar masks, but a lot of code expects masks. The compiler easily optimzes away masks extensions/truncations so consistency is preferable. We can revisit this when LLVM backends have more support for vector conditions.
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-10 01:43:35 +0000