summaryrefslogtreecommitdiffstats
path: root/src/gallium/auxiliary
Commit message (Collapse)AuthorAgeFilesLines
* u_format: Reduce code duplication.José Fonseca2010-04-221-118/+97
|
* gallivm: Remove unused variable.Vinson Lee2010-04-221-3/+0
|
* gallium/util: Fix deviation calculation in u_cache.Vinson Lee2010-04-211-1/+1
| | | | | The approximate deviation calculation was using the entire cache's total instead of each cache entry's total.
* gallivm: added some assertions in loop-gen codeBrian Paul2010-04-211-1/+6
| | | | We're hitting these assertions with nested loops...
* gallivm: fix copy&paste error: s/cont_stack_size/break_stack_size/Brian Paul2010-04-211-1/+1
|
* gallivm: emit_instruction() is booleanBrian Paul2010-04-211-39/+44
|
* gallivm: implement TGSI KILPBrian Paul2010-04-211-2/+33
| | | | | | As in tgsi_exec.c we don't actually rely on condition codes; we do an unconditional kill. The only predication comes from the execution mask which applies inside loops/conditionals.
* draw: Make LLVM integration work without llvmpipe too (e.g. sofptipe).José Fonseca2010-04-211-0/+1
| | | | Prevents the assertion failure.
* gallium/draw: fix point sprite handlingBrian Paul2010-04-216-4/+25
| | | | | | | | New draw API function to indicate whether or not to convert points to quads for sprite rasterization. Fix point-to-quad conversion regression in the wide-point stage. We need to check the pipe_rasterizer_state::point_quad_rasterization flag.
* gallium/draw: use local var to simplify codeBrian Paul2010-04-211-17/+18
|
* targets: Don't check if we should use traceJakob Bornecrantz2010-04-211-3/+2
| | | | | Trace does its own checking, and it used the GALLIUM_TRACE variable as well, but expected a file and not a bool argument.
* gallivm: Remove unnecessary headers.Vinson Lee2010-04-202-3/+0
|
* gallium/util: remove bogus return statementBrian Paul2010-04-201-1/+1
| | | | util_surfaces_do_detach() is a void-valued function.
* gallium: replace pipe_resource::_usage with pipe_resource::usageBrian Paul2010-04-202-2/+2
|
* gallivm: Universal format support on lp_build_fetch_rgba_aos via ↵José Fonseca2010-04-204-57/+92
| | | | | | util_format_description::fetch_rgba_float This therefore adds support to half float vertex buffers.
* gallivm: Cleanups and bugfixes to aos format translation.José Fonseca2010-04-202-128/+36
|
* draw: Fallback to gallivm for translation of ↵José Fonseca2010-04-201-4/+20
| | | | | | A8R8G8B8/B8G8R8A8/R10G10B10X2/R10G10B10X2 and other formats. These need swizzles, and bitshifts.
* gallivm: New function to fetch a pixel into a 4xfloat AoS vector.José Fonseca2010-04-202-10/+50
|
* gallivm: Bring aos format back to life.José Fonseca2010-04-203-0/+382
| | | | | | Useful for fetching vertices for formats that are straight arrays. This reverts commit aa364d091e7e2ef2296fb25f92efc79a8c88f77d.
* draw: Remove draw_vs_llvm.c.José Fonseca2010-04-205-142/+3
| | | | | | To silence some warnings. Super-seeded by Zack's new llvm middle end.
* Merge branch 'gallium-index-bias'José Fonseca2010-04-208-43/+63
|\
| * draw: No need to rebase for elt_bias.José Fonseca2010-04-201-7/+5
| | | | | | | | | | | | | | As we are rebasing to min_index + elt_bias, and the vertex buffer has no elt_bias. I still don't know how to exercise this code. I hope this is now right.
| * draw: Fix cache elt_bias implementation.José Fonseca2010-04-201-5/+8
| |
| * draw: Fix typo resulting from bad regular expression in index bias addition.José Fonseca2010-04-201-4/+4
| |
| * draw: Implement index bias.José Fonseca2010-04-198-46/+65
| |
* | draw llvm: Move dereference of pointer after NULL check.Vinson Lee2010-04-201-1/+3
| |
* | draw llvm: Remove unnecessary header.Vinson Lee2010-04-191-1/+0
| |
* | gallivm: pass 3D texture stride as an arrayBrian Paul2010-04-192-9/+18
| | | | | | | | This should have been included with the previous commit.
* | draw llvm: make sure use_llvm isn't used when not definedZack Rusin2010-04-191-1/+1
| |
* | draw llvm: fix typo (boolean, not bool)Zack Rusin2010-04-191-1/+1
| |
* | draw llvm: allow runtime switching of pipelines (yes/no to llvm)Zack Rusin2010-04-192-1/+4
| | | | | | | | use DRAW_USE_LLVM to disable or enable (default) llvm
* | draw llvm: fix constructor messZack Rusin2010-04-193-28/+9
| | | | | | | | use just one constructor to figure out whether to use llvm.
* | gallium/util: specify binding type for blit src textureBrian Paul2010-04-191-0/+1
| | | | | | | | Fixes fd.o bug 27711.
* | gallium/draw: use a local var to simplify some codeBrian Paul2010-04-191-8/+8
| |
* | Merge branch '7.8'Brian Paul2010-04-197-41/+187
|\ \ | | | | | | | | | | | | | | | | | | | | | Conflicts: src/gallium/auxiliary/draw/draw_context.c src/gallium/auxiliary/draw/draw_pipe_aaline.c src/gallium/drivers/llvmpipe/lp_context.c
| * | gallium/draw: use correct rasterization state for wide/AA points/linesBrian Paul2010-04-197-41/+186
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When points or lines are decomposed into triangles, we need to be sure to disable polygon culling, stippling, "un-filled" modes, etc. This patch sets the rasterization state to disable those things prior to drawing points/lines with triangles, then restores the previous state afterward. The new piglit point-no-line-cull test checks this problem & solution.
* | | gallivm: Remove redundant initialization of dst_vec_type.Vinson Lee2010-04-191-2/+0
| | | | | | | | | | | | | | | dec_vec_type is already initialized to lp_build_vec_type(dst_type) at its declaration.
* | | auxiliary: Move loop variable declaration outside for loop.Vinson Lee2010-04-181-1/+2
| | | | | | | | | | | | Fixes MSVC build.
* | | auxiliary: Initialize variable.Vinson Lee2010-04-181-1/+1
| | |
* | | auxiliary: Add util/u_surfaces.c to SCons build.Vinson Lee2010-04-181-0/+1
| | | | | | | | | | | | This was missed in commit ac4b8db62d056fa5b85661f67ba5e1d1d29e3fcf.
* | | u_inlines: split pipe_surface_init in pipe_surface_reset and *_initLuca Barbieri2010-04-182-3/+10
| | |
* | | tgsi: add comment about potential code removalBrian Paul2010-04-181-0/+6
| | |
* | | tgsi: replace 0xf with TGSI_WRITEMASK_XYZWBrian Paul2010-04-181-2/+2
| | |
* | | u_surfaces: silence warningLuca Barbieri2010-04-181-1/+1
| | |
* | | auxiliary: add util_dirty_surfaces to track dirty render temporariesLuca Barbieri2010-04-181-0/+88
| | |
* | | auxiliary: add util_surfaces to track a resource's pipe_surface structsLuca Barbieri2010-04-183-0/+167
| | |
* | | u_inlines: add function to initialize pipe_surfaceLuca Barbieri2010-04-181-0/+15
| | |
* | | u_blitter: add support for saving vertex buffersLuca Barbieri2010-04-182-0/+24
| |/ |/| | | | | | | | | | | | | | | | | | | | | Currently r300g does not save vertex buffer on blitter calls. It gets away with it because the current Mesa state tracker usually resets vertex buffers on every draw calls. However, this is wrong. nvfx won't be lucky because it needs to use the blitter inside draw calls.
* | draw: Cosmetic cleanups and comments.José Fonseca2010-04-184-4/+10
| |
* | os: Implement pipe_barrier for POSIX platforms without pthread_barrier_t.Vinson Lee2010-04-171-5/+26
| | | | | | | | This patch was tested on Mac OS X.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-18 14:43:16 +0000