Optimize (amvd>2)+(amvd>32), about 1 cpu cycles faster.
authorZhou Zongyi <zhouzy@os.pku.edu.cn>
Fri, 26 Feb 2010 22:45:35 +0000 (22:45 +0000)
committerMichael Niedermayer <michaelni@gmx.at>
Fri, 26 Feb 2010 22:45:35 +0000 (22:45 +0000)
commit821fe7f3e60a23fc472b529c31e60ed4e30b2b96
treeece098ee44ba1dc991da41543afff08167e57543
parent4e06acbde4a75cf9aa6a3e46dd3a7c6ed1ecfb8f
Optimize (amvd>2)+(amvd>32), about 1 cpu cycles faster.
patch by Zhou Zongyi @ zhouzy () os punkt pku dot edu speck cn

Originally committed as revision 22084 to svn://svn.ffmpeg.org/ffmpeg/trunk
libavcodec/h264_cabac.c