virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2025-02-04 06:08:26 +02:00
Code Issues Releases Activity

1 c-asm loop less and 1x unroll of float_to_int16_sse()

Browse Source 
25% faster

Originally committed as revision 14104 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Michael Niedermayer 2008-07-07 21:25:18 +00:00
parent 560fa9bf51
commit 35ee72b1d7
1 changed files with 19 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
libavcodec/i386/dsputil_mmx.c
View File

@ -2045,18 +2045,25 @@ static void float_to_int16_3dnow(int16_t *dst, const float *src, long len){
);
}
static void float_to_int16_sse(int16_t *dst, const float *src, long len){
int i;
for(i=0; i<len; i+=4) {
asm volatile(
"cvtps2pi %1, %%mm0 \n\t"
"cvtps2pi %2, %%mm1 \n\t"
"packssdw %%mm1, %%mm0 \n\t"
"movq %%mm0, %0 \n\t"
:"=m"(dst[i])
:"m"(src[i]), "m"(src[i+2])
);
}
asm volatile("emms");
asm volatile(
"add %0 , %0 \n\t"
"lea (%2,%0,2) , %2 \n\t"
"add %0 , %1 \n\t"
"neg %0 \n\t"
"1: \n\t"
"cvtps2pi (%2,%0,2) , %%mm0 \n\t"
"cvtps2pi 8(%2,%0,2) , %%mm1 \n\t"
"cvtps2pi 16(%2,%0,2) , %%mm2 \n\t"
"cvtps2pi 24(%2,%0,2) , %%mm3 \n\t"
"packssdw %%mm1 , %%mm0 \n\t"
"packssdw %%mm3 , %%mm2 \n\t"
"movq %%mm0 , (%1,%0) \n\t"
"movq %%mm2 , 8(%1,%0) \n\t"
"add $16 , %0 \n\t"
" js 1b \n\t"
"emms \n\t"
:"+r"(len), "+r"(dst), "+r"(src)
);
}
extern void ff_snow_horizontal_compose97i_sse2(IDWTELEM *b, int width);