llvmpipe: Optimize blend swizzles by using bitmasks instead of shuffles for ubytes.
authorJosé Fonseca <jfonseca@vmware.com>
Tue, 4 Aug 2009 11:09:52 +0000 (12:09 +0100)
committerJosé Fonseca <jfonseca@vmware.com>
Sat, 29 Aug 2009 08:21:23 +0000 (09:21 +0100)
commit1dd7bb17c7331f9ecd0bc830b61ada235a56fe6d
tree31508bde5eead960ba644ef4ef50d4bc526fc43d
parenta77084ea4b5801dc3ba52f33533176da926aed16
llvmpipe: Optimize blend swizzles by using bitmasks instead of shuffles for ubytes.
src/gallium/drivers/llvmpipe/SConscript
src/gallium/drivers/llvmpipe/lp_bld_blend.c
src/gallium/drivers/llvmpipe/lp_bld_const.c
src/gallium/drivers/llvmpipe/lp_bld_const.h
src/gallium/drivers/llvmpipe/lp_bld_swizzle.c [new file with mode: 0644]
src/gallium/drivers/llvmpipe/lp_bld_swizzle.h [new file with mode: 0644]