[x264-devel] x264: svn commit r696 (pengvado)

[Subversion daemon]

r696 | pengvado | 2007-11-20 07:07:17 +0100 (Tue, 20 Nov 2007) | 4 lines
Changed paths:
   M /trunk/common/amd64/amd64inc.asm
   M /trunk/common/amd64/pixel-sse2.asm
   M /trunk/common/frame.c
   M /trunk/common/i386/i386inc.asm
   M /trunk/common/i386/pixel-sse2.asm
   M /trunk/common/i386/pixel.h
   M /trunk/common/pixel.c
   M /trunk/tools/checkasm.c

avoid memory loads that span the border between two cachelines.
on core2 this makes x264_pixel_sad an average of 2x faster. other intel cpus gain various amounts. amd are unaffected.
overall speedup: 1-10%, depending on how much time is spent in fullpel motion estimation.

 > http://trac.videolan.org/x264/changeset/696