projects / mesa.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 4c69823) | patch
swr/rast: Permute work for simd16
authorGeorge Kyriazis <george.kyriazis@intel.com>
Wed, 14 Mar 2018 18:38:18 +0000 (13:38 -0500)
committerGeorge Kyriazis <george.kyriazis@intel.com>
Wed, 18 Apr 2018 15:51:38 +0000 (10:51 -0500)
commit9103119cb36fbcfd591df8c722fa9941aaa9a911
tree1b13dbc16efd88520235c489f95937dd68d4246atree
parent4c69823d150805c6f1d1ea212efa4e20558768bdcommit | diff
swr/rast: Permute work for simd16

Fix slow permutes in PA tri lists under SIMD16 emulation on AVX

Added missing permute (interlane, immediate) to SIMDLIB

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
src/gallium/drivers/swr/rasterizer/common/simd16intrin.h diff | blob | history
src/gallium/drivers/swr/rasterizer/common/simdintrin.h diff | blob | history
src/gallium/drivers/swr/rasterizer/common/simdlib_256_avx.inl diff | blob | history
src/gallium/drivers/swr/rasterizer/common/simdlib_256_avx2.inl diff | blob | history
src/gallium/drivers/swr/rasterizer/common/simdlib_512_avx512.inl diff | blob | history
src/gallium/drivers/swr/rasterizer/common/simdlib_512_emu.inl diff | blob | history
src/gallium/drivers/swr/rasterizer/core/pa_avx.cpp diff | blob | history