projects / platform / upstream / libav.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2f41242) | patch
MMX, MMX2, SSE2 and SSSE3 optimizations for pred16x16/8x8_plane H264 intra
authorRonald S. Bultje <rsbultje@gmail.com>
Tue, 5 Oct 2010 22:06:18 +0000 (22:06 +0000)
committerRonald S. Bultje <rsbultje@gmail.com>
Tue, 5 Oct 2010 22:06:18 +0000 (22:06 +0000)
commitdd68d4db430089bff70f62a38cb84e07d183a435
tree80651f97cce9f9c3073a9fb8cf0695849c8abf7btree | snapshot
parent2f412421e8bb1927fcd866777dfedf223c9dfdfbcommit | diff
MMX, MMX2, SSE2 and SSSE3 optimizations for pred16x16/8x8_plane H264 intra
prediction (plus some with different rounding for svq3/rv40). Speedup (for
SSSE3) about ~6-fold, 3.6% faster overall with cathedral sample.

Originally committed as revision 25361 to svn://svn.ffmpeg.org/ffmpeg/trunk
libavcodec/x86/dsputil_mmx.c diff | blob | history
libavcodec/x86/h264_intrapred.asm diff | blob | history
libavcodec/x86/h264_intrapred_init.c diff | blob | history
Domain: Multimedia / Media Content;