Commit e79da294 authored by Pierre d'Herbemont's avatar Pierre d'Herbemont

contrib/ffmpeg: Yet an other fix attempt for ffmpeg. (Previous fix is broken...

contrib/ffmpeg: Yet an other fix attempt for ffmpeg. (Previous fix is broken because of registers don't get reseted properly in the look).
parent 4f31c784
...@@ -2,39 +2,36 @@ Index: ffmpeg/libavcodec/i386/dsputil_mmx.c ...@@ -2,39 +2,36 @@ Index: ffmpeg/libavcodec/i386/dsputil_mmx.c
=================================================================== ===================================================================
--- ffmpeg/libavcodec/i386/dsputil_mmx.c (revision 10759) --- ffmpeg/libavcodec/i386/dsputil_mmx.c (revision 10759)
+++ ffmpeg/libavcodec/i386/dsputil_mmx.c (working copy) +++ ffmpeg/libavcodec/i386/dsputil_mmx.c (working copy)
@@ -2978,6 +2978,7 @@ @@ -2967,7 +2967,6 @@
::"m"(c), "m"(*ff_pd_1), "m"(*ff_pd_2) double c = 2.0 / (len-1.0);
); int n2 = len>>1;
#define WELCH(MOVPD)\ long i = -n2*sizeof(int32_t);
+ do {\ - long j = n2*sizeof(int32_t);
asm volatile(\ asm volatile(
"1: \n\t"\ "movsd %0, %%xmm7 \n\t"
"movapd %%xmm7, %%xmm1 \n\t"\ "movapd %1, %%xmm6 \n\t"
@@ -2985,8 +2986,12 @@ @@ -2985,17 +2984,18 @@
"movapd %%xmm6, %%xmm0 \n\t"\ "movapd %%xmm6, %%xmm0 \n\t"\
"subpd %%xmm1, %%xmm0 \n\t"\ "subpd %%xmm1, %%xmm0 \n\t"\
"pshufd $0x4e, %%xmm0, %%xmm1 \n\t"\ "pshufd $0x4e, %%xmm0, %%xmm1 \n\t"\
- "cvtpi2pd (%4,%0), %%xmm2 \n\t"\ - "cvtpi2pd (%4,%0), %%xmm2 \n\t"\
- "cvtpi2pd (%5,%1), %%xmm3 \n\t"\ - "cvtpi2pd (%5,%1), %%xmm3 \n\t"\
+ "cvtpi2pd (%2,%0), %%xmm2 \n\t"\ + "cvtpi2pd (%3,%0), %%xmm2 \n\t"\
+ "cvtpi2pd (%3,%1), %%xmm3 \n\t"\
+ :"+&r"(i), "+&r"(j)\
+ :"r"(data+n2), "r"(data+len-2-n2)\
+ );\
+ asm volatile(\
"mulpd %%xmm0, %%xmm2 \n\t"\ "mulpd %%xmm0, %%xmm2 \n\t"\
+ "movapd %%xmm2, (%1,%0,2) \n\t"\
+ "negl %0\n\t"\
+ "cvtpi2pd (%4,%0), %%xmm3 \n\t"\
"mulpd %%xmm1, %%xmm3 \n\t"\ "mulpd %%xmm1, %%xmm3 \n\t"\
"movapd %%xmm2, (%2,%0,2) \n\t"\ - "movapd %%xmm2, (%2,%0,2) \n\t"\
@@ -2996,9 +3001,9 @@ - MOVPD" %%xmm3, (%3,%1,2) \n\t"\
+ MOVPD" %%xmm3, (%2,%0,2) \n\t"\
"subpd %%xmm5, %%xmm7 \n\t"\
- "sub $8, %1 \n\t"\
+ "negl %0\n\t"\
"add $8, %0 \n\t"\ "add $8, %0 \n\t"\
"jl 1b \n\t"\ "jl 1b \n\t"\
:"+&r"(i), "+&r"(j)\ - :"+&r"(i), "+&r"(j)\
- :"r"(w_data+n2), "r"(w_data+len-2-n2),\ + :"+&r"(i)\
- "r"(data+n2), "r"(data+len-2-n2)\ :"r"(w_data+n2), "r"(w_data+len-2-n2),\
- ); "r"(data+n2), "r"(data+len-2-n2)\
+ :"r"(w_data+n2), "r"(w_data+len-2-n2)\ );
+ );\
+ } while(0);
if(len&1)
WELCH("movupd")
else
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment