提交 2b5269b5 编写于 作者: M Michael Niedermayer

moving lps state transition code a little up in the branched asm code (1% faster on P3)

Originally committed as revision 6658 to svn://svn.ffmpeg.org/ffmpeg/trunk
上级 b99f3cab
......@@ -405,11 +405,11 @@ static int get_cabac(CABACContext *c, uint8_t * const state){
"subl %%edx, %%ebx \n\t"
"movl %%esi, %%edx \n\t"
"shr $19, %%esi \n\t"
"movzbl "MANGLE(ff_h264_lps_state)"(%%eax), %%ecx \n\t"
"movb %%cl, (%1) \n\t"
"movzbl " MANGLE(ff_h264_norm_shift) "(%%esi), %%ecx \n\t"
"shll %%cl, %%ebx \n\t"
"shll %%cl, %%edx \n\t"
"movzbl "MANGLE(ff_h264_lps_state)"(%%eax), %%ecx \n\t"
"movb %%cl, (%1) \n\t"
"addl $1, %%eax \n\t"
"test %%bx, %%bx \n\t"
" jnz 2f \n\t"
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册