Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | r300compiler: fix scons build | Joakim Sindholt | 2010-06-05 | 1 | -0/+2 |
| | |||||
* | r300/compiler: implement SIN+COS+SCS for vertex shaders | Marek Olšák | 2010-06-05 | 3 | -21/+76 |
| | |||||
* | r300/compiler: implement SNE unwound for r3xx VS, natively for r5xx VS | Marek Olšák | 2010-06-05 | 2 | -1/+37 |
| | |||||
* | r300/compiler: implement SEQ unwound for r3xx VS, natively for r5xx VS | Marek Olšák | 2010-06-05 | 2 | -0/+36 |
| | | | | Fixes piglit/glsl-vs-vec4-indexing-4. | ||||
* | r300/compiler: implement SFL for vertex shaders | Marek Olšák | 2010-06-05 | 1 | -2/+3 |
| | | | | And sort the "case" statements alphabetically. | ||||
* | r300/compiler: print opcode names instead of numbers | Marek Olšák | 2010-06-03 | 3 | -8/+8 |
| | |||||
* | r300: fix blits for textures of width/height greater than 2048 on r5xx | Marek Olšák | 2010-05-29 | 1 | -5/+9 |
| | | | | Yes I am fixing r300c ... who knew? | ||||
* | r300/compiler: implement SGT+SLE opcodes | Marek Olšák | 2010-05-26 | 1 | -0/+20 |
| | | | | Reported-by: Gianluca Anzolin <gianluca@sottospazio.it> | ||||
* | r300/compiler: fix dumping r5xx vertex shaders | Marek Olšák | 2010-05-26 | 1 | -0/+3 |
| | |||||
* | r300/compiler: move hardware caps to the radeon_compiler base struct | Marek Olšák | 2010-05-26 | 6 | -18/+19 |
| | | | | Needed for vertex shaders too. | ||||
* | r300/compiler: shorten swizzle expressions | Marek Olšák | 2010-05-26 | 1 | -44/+65 |
| | |||||
* | r300/compiler: Implement constant folding | Nicolai Hähnle | 2010-05-19 | 1 | -1/+215 |
| | | | | Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> | ||||
* | r300/compiler: Emit 0.5 swizzle when necessary. | Tom Stellard | 2010-05-19 | 1 | -1/+3 |
| | | | | Signed-off-by: Marek Olšák <maraeo@gmail.com> | ||||
* | r300/compiler: fix peephole optimizer | Tom Stellard | 2010-05-16 | 1 | -1/+14 |
| | | | | Tested-by: Marek Olšák <maraeo@gmail.com> | ||||
* | r300/compiler: Implement simple peephole optimizer | Nicolai Hähnle | 2010-05-16 | 8 | -47/+308 |
| | | | | Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> | ||||
* | r300/compiler: silence a warning | Marek Olšák | 2010-05-15 | 1 | -1/+1 |
| | |||||
* | r300g: respect compare mode regardless of sampler type | Marek Olšák | 2010-05-08 | 2 | -1/+8 |
| | |||||
* | r300/compiler: fake equal/notequal for shadow samplers | Marek Olšák | 2010-05-08 | 1 | -0/+7 |
| | | | | | This is actually better than a real implementation, which would be useless due to inaccuracy (I know because I've tried to implement it). | ||||
* | r300/compiler: use perspective division only for TXP in shadow samplers | Marek Olšák | 2010-05-08 | 1 | -23/+36 |
| | |||||
* | r300/compiler: make shadow sampler codegen more readable, add comments | Marek Olšák | 2010-05-08 | 1 | -16/+28 |
| | |||||
* | r300/compiler: shorten RC_*SWIZZLE* expressions | Marek Olšák | 2010-05-08 | 4 | -4/+5 |
| | |||||
* | r300/compiler: generalize depth texture mode to support arbitrary swizzles | Marek Olšák | 2010-05-08 | 3 | -11/+7 |
| | |||||
* | r300/compiler: fix LIT instruction case 0^0 = 1 | Marek Olšák | 2010-05-08 | 1 | -0/+24 |
| | |||||
* | Merge branch 'gles2-2' | Kristian Høgsberg | 2010-05-02 | 2 | -2/+4 |
|\ | | | | | | | | | Conflicts: src/mesa/drivers/dri/common/dri_util.h | ||||
| * | dri: Add DRI entrypoints to create a context for a given API | Kristian Høgsberg | 2010-04-28 | 2 | -2/+4 |
| | | |||||
* | | Merge branch '7.8' | Brian Paul | 2010-04-22 | 1 | -7/+8 |
|\ \ | |/ |/| | | | | | | | Conflicts: src/mesa/state_tracker/st_format.c | ||||
| * | r300: fix vertex unit setup | Alex Deucher | 2010-04-22 | 1 | -7/+8 |
| | | | | | | | | | | | | RV3xx is 2, RV560,RV570 is 8 Noticed by Tormod Volden. | ||||
* | | r300/compiler: lower CEIL | Marek Olšák | 2010-04-19 | 3 | -1/+33 |
| | | |||||
* | | r300/compiler: enable branch emulation for vertex shaders | Marek Olšák | 2010-04-19 | 1 | -21/+22 |
| | | |||||
* | | r300/compiler: optimize CMP for vertex shaders a bit | Marek Olšák | 2010-04-19 | 1 | -20/+9 |
| | | |||||
* | | r300/compiler: Use memory_pool_array_reserve in deadcode elimination | Nicolai Hähnle | 2010-04-19 | 1 | -13/+2 |
| | | | | | | | | Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> | ||||
* | | r300/compiler: Use memory_pool_array_reserve in r500-fragprog_emit | Nicolai Hähnle | 2010-04-19 | 1 | -17/+2 |
| | | | | | | | | Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> | ||||
* | | r300/compiler: enable branch emulation for R500 fragment programs | Marek Olšák | 2010-04-19 | 1 | -4/+13 |
| | | |||||
* | | r300/compiler: Implement branch emulation for R300 fragment programs | Nicolai Hähnle | 2010-04-19 | 8 | -104/+512 |
| | | | | | | | | Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com> | ||||
* | | r300/compiler: Remove unnecessary header. | Vinson Lee | 2010-04-17 | 1 | -2/+0 |
| | | |||||
* | | r300/compiler: Add radeon_program_tex.c to SCons build. | Vinson Lee | 2010-04-16 | 1 | -0/+1 |
| | | | | | | | | This was missed in commit f8a14186809356871ae74159c774e9e3959a22e5. | ||||
* | | r300/compiler: add emulation of all mirrored-clamp wrap modes for NPOT textures | Marek Olšák | 2010-04-17 | 2 | -2/+19 |
| | | |||||
* | | r300/compiler: replace mirrored-repeat emulation with a faster version | Marek Olšák | 2010-04-17 | 1 | -62/+65 |
| | | | | | | | | | | Also, the Negate bitfield was 1 instead of RC_MASK_XYZ in the previous version, causing incorrect rendering. | ||||
* | | r300/compiler: add handy definitions for XYZ0 and smeared half swizzling | Marek Olšák | 2010-04-17 | 2 | -3/+5 |
| | | |||||
* | | r300/compiler: make ARB_shadow_ambient optional | Marek Olšák | 2010-04-16 | 4 | -6/+18 |
| | | | | | | | | | | This saves constant register space for r300g, which doesn't need this feature. | ||||
* | | r300/compiler: restore r500-specific bits of TEX transformations | Marek Olšák | 2010-04-16 | 1 | -23/+40 |
| | | |||||
* | | r300/compiler: Fix texture instruction readmasks. | Corbin Simpson | 2010-04-16 | 3 | -7/+27 |
| | | | | | | | | No immediate benefit, it was just bugging me. | ||||
* | | r300/compiler: Fix magic numbers in readmasks. | Corbin Simpson | 2010-04-16 | 1 | -3/+3 |
| | | |||||
* | | r300/compiler: Oops, this slipped through. | Corbin Simpson | 2010-04-16 | 1 | -2/+4 |
| | | |||||
* | | r300/compiler: Save Q for mirror and repeat. | Corbin Simpson | 2010-04-16 | 1 | -22/+20 |
| | | |||||
* | | r300/compiler: Split off texture fixes to their own file. | Corbin Simpson | 2010-04-16 | 8 | -555/+369 |
| | | | | | | | | Yes, I'm fully aware this generates subpar code on r500. | ||||
* | | r300/compiler: fix repeat wrap mode for TXP and NPOTs | Marek Olšák | 2010-04-16 | 1 | -6/+16 |
| | | | | | | | | | | No idea why st/mesa unnecessarily inserts TXP where TEX is sufficient. Also re-enabling the NPOT fallback for repeat in r300g. | ||||
* | | r300/compiler: kill off RC_WRAP_CLAMP | Marek Olšák | 2010-04-15 | 3 | -3/+2 |
| | | | | | | | | A variant thereof might be later reintroduced for the mirrored-clamp modes. | ||||
* | | r300/compiler: Hax around instructions limits in mirroring code. | Corbin Simpson | 2010-04-12 | 1 | -0/+10 |
| | | | | | | | | Stuff's starting to show up in arbnpot. | ||||
* | | r300/compiler: Comment code, add much better mirror maths. | Corbin Simpson | 2010-04-12 | 2 | -29/+161 |
| | |