| Commit message (Expand) | Author | Age | Files | Lines |
* | st/xorg: Don't try to use option values before processing options | Thomas Hellstrom | 2010-10-12 | 1 | -13/+13 |
* | Revert "llvmpipe: try to keep plane c values small" | Keith Whitwell | 2010-10-12 | 2 | -24/+17 |
* | gallivm: don't branch on KILLs near end of shader | Keith Whitwell | 2010-10-12 | 1 | -10/+47 |
* | r600g: add missing file to sconscript | Keith Whitwell | 2010-10-12 | 1 | -0/+1 |
* | gallium: move sse intrinsics debug helpers to u_sse.h | Keith Whitwell | 2010-10-12 | 3 | -117/+79 |
* | llvmpipe: Fix MSVC build. | José Fonseca | 2010-10-12 | 1 | -18/+18 |
* | llvmpipe: fix typo in last commit | Keith Whitwell | 2010-10-12 | 1 | -2/+2 |
* | llvmpipe: try to keep plane c values small | Keith Whitwell | 2010-10-12 | 2 | -17/+24 |
* | llvmpipe: add debug helpers for epi32 etc | Keith Whitwell | 2010-10-12 | 1 | -0/+115 |
* | llvmpipe: try to do more of rast_tri_3_16 with intrinsics | Keith Whitwell | 2010-10-12 | 2 | -9/+271 |
* | llvmpipe: Do not dispose the execution engine. | José Fonseca | 2010-10-12 | 1 | -3/+0 |
* | nouveau: Get larger push buffers. | Francisco Jerez | 2010-10-12 | 3 | -3/+3 |
* | dri/nouveau: Initialize tile_flags when allocating a render target. | Francisco Jerez | 2010-10-12 | 2 | -6/+14 |
* | r600g: fix typo in vertex sampling on r600 | Dave Airlie | 2010-10-12 | 1 | -1/+1 |
* | i965: Always use the new FS backend on gen6. | Eric Anholt | 2010-10-11 | 1 | -2/+7 |
* | i965: Fix gen6 pixel_[xy] setup to avoid mixing int and float src operands. | Eric Anholt | 2010-10-11 | 1 | -6/+15 |
* | i965: Don't compute-to-MRF in gen6 VS math. | Eric Anholt | 2010-10-11 | 1 | -7/+15 |
* | i965: Expand uniform args to gen6 math to full registers to get hstride == 1. | Eric Anholt | 2010-10-11 | 1 | -0/+25 |
* | i965: Don't compute-to-MRF in gen6 math instructions. | Eric Anholt | 2010-10-11 | 1 | -0/+16 |
* | i965: Add a couple of checks for gen6 math instruction limits. | Eric Anholt | 2010-10-11 | 1 | -0/+26 |
* | i965: Don't consider gen6 math instructions to write to MRFs. | Eric Anholt | 2010-10-11 | 1 | -17/+38 |
* | glsl: Changes in generated file glsl_lexer.cpp | Chad Versace | 2010-10-11 | 1 | -691/+716 |
* | glsl: Add lexer rules for uint and uvecN (N=2..4) | Chad Versace | 2010-10-11 | 1 | -0/+4 |
* | glsl: Add glsl_type::uvecN_type for N=2,3 | Chad Versace | 2010-10-11 | 2 | -0/+4 |
* | intel_extensions: Add ability to set GLSL version via environment | Chad Versace | 2010-10-11 | 1 | -1/+18 |
* | r200: revalidate after radeon_update_renderbuffers | Daniel Vetter | 2010-10-11 | 3 | -3/+10 |
* | i965: Compute to MRF in the new FS backend. | Eric Anholt | 2010-10-11 | 2 | -0/+124 |
* | i965: Give the FB write and texture opcodes the info on base MRF, like math. | Eric Anholt | 2010-10-11 | 2 | -38/+48 |
* | i965: Give the math opcodes information on base mrf/mrf len. | Eric Anholt | 2010-10-11 | 2 | -12/+57 |
* | i965: Move FS backend structures to a header. | Eric Anholt | 2010-10-11 | 5 | -363/+407 |
* | i965: Reduce register interference checks for changed FS_OPCODE_DISCARD. | Eric Anholt | 2010-10-11 | 1 | -5/+2 |
* | i965: Split FS_OPCODE_DISCARD into two steps. | Eric Anholt | 2010-10-11 | 1 | -9/+23 |
* | llvmpipe: Use lp_tgsi_info. | José Fonseca | 2010-10-11 | 4 | -30/+31 |
* | gallivm: More detailed analysis of tgsi shaders. | José Fonseca | 2010-10-11 | 4 | -0/+559 |
* | tgsi: Export some names for some tgsi enums. | José Fonseca | 2010-10-11 | 2 | -23/+35 |
* | gallium: Define C99 restrict keyword where absent. | José Fonseca | 2010-10-11 | 1 | -0/+21 |
* | gallivm: Eliminate unsigned integer arithmetic from texture coordinates. | José Fonseca | 2010-10-11 | 3 | -42/+32 |
* | llvmpipe: Remove outdated comment about stencil testing. | José Fonseca | 2010-10-11 | 1 | -9/+2 |
* | r600g: don't run with scissors. | Dave Airlie | 2010-10-11 | 2 | -0/+92 |
* | r600g: add TXL opcode support. | Dave Airlie | 2010-10-11 | 1 | -2/+2 |
* | r600g: enable vertex samplers. | Dave Airlie | 2010-10-11 | 4 | -9/+21 |
* | r600g: evergreen has no request size bit in texture word4 | Dave Airlie | 2010-10-11 | 2 | -4/+0 |
* | r600g: fix input/output Z export mixup for evergreen. | Dave Airlie | 2010-10-11 | 1 | -1/+1 |
* | gallivm: Pass texture coords derivates as scalars. | José Fonseca | 2010-10-10 | 4 | -26/+38 |
* | gallivm: Use variables instead of Phis in loops. | José Fonseca | 2010-10-10 | 2 | -42/+23 |
* | gallivm: Allow to disable bri-linear filtering with GALLIVM_DEBUG=no_brilinea... | José Fonseca | 2010-10-10 | 3 | -9/+10 |
* | gallivm: Fix a long standing bug with nested if-then-else emission. | José Fonseca | 2010-10-10 | 1 | -17/+6 |
* | dri/nv10: Fake fast Z clears for pre-nv17 cards. | Francisco Jerez | 2010-10-10 | 4 | -20/+127 |
* | dri/nouveau: Minor cleanup. | Francisco Jerez | 2010-10-10 | 4 | -23/+22 |
* | gallivm: Cleanup the rest of the flow module. | José Fonseca | 2010-10-09 | 3 | -211/+39 |