RepoMirrors
/
ffmpeg
mirror of https://git.ffmpeg.org/ffmpeg.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

avcodec/x86/cavsdsp: remove incorrect LOCAL_ALIGN tmp 

This is faster and simpler as well

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This commit is contained in:
Michael Niedermayer 2015-03-16 14:28:11 +01:00
parent 12cf61c3b2
commit d79f7bf0d6
1 changed files with 11 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
libavcodec/x86/cavsdsp.c
View File

@ -142,8 +142,6 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
LOCAL_ALIGNED(16, int16_t, b2, [64]); LOCAL_ALIGNED(16, int16_t, b2, [64]);
for(i=0; i<2; i++){ for(i=0; i<2; i++){
LOCAL_ALIGNED(8, uint64_t, tmp);
cavs_idct8_1d(block+4*i, ff_pw_4.a); cavs_idct8_1d(block+4*i, ff_pw_4.a);
__asm__ volatile( __asm__ volatile(
@ -155,19 +153,19 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
"psraw $3, %%mm2 \n\t" "psraw $3, %%mm2 \n\t"
"psraw $3, %%mm1 \n\t" "psraw $3, %%mm1 \n\t"
"psraw $3, %%mm0 \n\t" "psraw $3, %%mm0 \n\t"
"movq %%mm7, %0 \n\t" "movq %%mm7, (%0) \n\t"
TRANSPOSE4( %%mm0, %%mm2, %%mm4, %%mm6, %%mm7 ) TRANSPOSE4( %%mm0, %%mm2, %%mm4, %%mm6, %%mm7 )
"movq %%mm0, 8(%1) \n\t" "movq %%mm0, 8(%0) \n\t"
"movq %%mm6, 24(%1) \n\t" "movq %%mm6, 24(%0) \n\t"
"movq %%mm7, 40(%1) \n\t" "movq %%mm7, 40(%0) \n\t"
"movq %%mm4, 56(%1) \n\t" "movq %%mm4, 56(%0) \n\t"
"movq %0, %%mm7 \n\t" "movq (%0), %%mm7 \n\t"
TRANSPOSE4( %%mm7, %%mm5, %%mm3, %%mm1, %%mm0 ) TRANSPOSE4( %%mm7, %%mm5, %%mm3, %%mm1, %%mm0 )
"movq %%mm7, (%1) \n\t" "movq %%mm7, (%0) \n\t"
"movq %%mm1, 16(%1) \n\t" "movq %%mm1, 16(%0) \n\t"
"movq %%mm0, 32(%1) \n\t" "movq %%mm0, 32(%0) \n\t"
"movq %%mm3, 48(%1) \n\t" "movq %%mm3, 48(%0) \n\t"
: "=m"(tmp) :
: "r"(b2+32*i) : "r"(b2+32*i)
: "memory" : "memory"
); );