Skip to content

V
vlc-gpu
Commit c3aa38ed authored by Pierre d'Herbemont's avatar Pierre d'Herbemont
Browse files
Options

ffmpeg-macosx-intel-mmx.patch: Update the patch to reflect latest ffmpeg... 

ffmpeg-macosx-intel-mmx.patch: Update the patch to reflect latest ffmpeg changes. (flac code moved to flacdsp_mmc.c and new sse2 optimization in flac that uses too much registers).
parent 120502a4
Hide whitespace changes
Inline Side-by-side
Showing with 99 additions and 55 deletions
extras/contrib/src/Patches/ffmpeg-macosx-intel-mmx.patch
View file @ c3aa38ed
Index: libavcodec/i386/motion_est_mmx.c
===================================================================
--- libavcodec/i386/motion_est_mmx.c (revision 10865)
--- libavcodec/i386/motion_est_mmx.c (revision 11270)
+++ libavcodec/i386/motion_est_mmx.c (working copy)
@@ -167,7 +167,7 @@
static inline void sad8_4_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
......@@ -40,7 +40,7 @@ Index: libavcodec/i386/motion_est_mmx.c
Index: libavcodec/i386/dsputil_h264_template_mmx.c
===================================================================
--- libavcodec/i386/dsputil_h264_template_mmx.c (revision 10865)
--- libavcodec/i386/dsputil_h264_template_mmx.c (revision 11270)
+++ libavcodec/i386/dsputil_h264_template_mmx.c (working copy)
@@ -188,8 +188,8 @@
"pxor %%mm7, %%mm7 \n\t"
......@@ -64,9 +64,9 @@ Index: libavcodec/i386/dsputil_h264_template_mmx.c
Index: libavcodec/i386/dsputil_mmx.c
===================================================================
--- libavcodec/i386/dsputil_mmx.c (revision 10865)
--- libavcodec/i386/dsputil_mmx.c (revision 11270)
+++ libavcodec/i386/dsputil_mmx.c (working copy)
@@ -1917,7 +1917,7 @@
@@ -1920,7 +1920,7 @@
#define QPEL_V_LOW(m3,m4,m5,m6, pw_20, pw_3, rnd, in0, in1, in2, in7, out, OP)\
"paddw " #m4 ", " #m3 " \n\t" /* x1 */\
......@@ -75,7 +75,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\
"movq "#in7", " #m3 " \n\t" /* d */\
"movq "#in0", %%mm5 \n\t" /* D */\
@@ -1929,7 +1929,7 @@
@@ -1932,7 +1932,7 @@
"paddw " #m5 ", %%mm6 \n\t" /* x2 */\
"paddw %%mm6, %%mm6 \n\t" /* 2x2 */\
"psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\
......@@ -84,7 +84,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"paddw " #rnd ", %%mm4 \n\t" /* x2 */\
"paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\
"psraw $5, %%mm5 \n\t"\
@@ -1963,10 +1963,10 @@
@@ -1966,10 +1966,10 @@
"paddw %%mm5, %%mm5 \n\t" /* 2b */\
"psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
"pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
......@@ -97,7 +97,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
"paddw %6, %%mm6 \n\t"\
"paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
@@ -1989,10 +1989,10 @@
@@ -1992,10 +1992,10 @@
"psrlq $24, %%mm6 \n\t" /* IJKLM000 */\
"punpcklbw %%mm7, %%mm2 \n\t" /* 0F0G0H0I */\
"punpcklbw %%mm7, %%mm6 \n\t" /* 0I0J0K0L */\
......@@ -110,7 +110,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"psubw %%mm4, %%mm3 \n\t" /* - 6b +3c - d */\
"paddw %6, %%mm1 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* 20a - 6b +3c - d */\
@@ -2015,7 +2015,7 @@
@@ -2018,7 +2018,7 @@
"psubw %%mm5, %%mm0 \n\t" /* c - 2b */\
"movq %%mm3, %%mm5 \n\t" /* JKLMNOPQ */\
"psrlq $24, %%mm3 \n\t" /* MNOPQ000 */\
......@@ -119,7 +119,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"punpcklbw %%mm7, %%mm3 \n\t" /* 0M0N0O0P */\
"paddw %%mm3, %%mm2 \n\t" /* d */\
"psubw %%mm2, %%mm0 \n\t" /* -6b + 3c - d */\
@@ -2023,7 +2023,7 @@
@@ -2026,7 +2026,7 @@
"punpcklbw %%mm7, %%mm2 \n\t" /* 0J0K0L0M */\
"punpckhbw %%mm7, %%mm5 \n\t" /* 0N0O0P0Q */\
"paddw %%mm2, %%mm6 \n\t" /* a */\
......@@ -128,7 +128,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"paddw %6, %%mm0 \n\t"\
"paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
"psraw $5, %%mm0 \n\t"\
@@ -2038,8 +2038,8 @@
@@ -2041,8 +2041,8 @@
"paddw %%mm2, %%mm5 \n\t" /* d */\
"paddw %%mm6, %%mm6 \n\t" /* 2b */\
"psubw %%mm6, %%mm4 \n\t" /* c - 2b */\
......@@ -139,7 +139,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"psubw %%mm5, %%mm3 \n\t" /* -6b + 3c - d */\
"paddw %6, %%mm4 \n\t"\
"paddw %%mm3, %%mm4 \n\t" /* 20a - 6b + 3c - d */\
@@ -2052,7 +2052,9 @@
@@ -2055,7 +2055,9 @@
"decl %2 \n\t"\
" jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
......@@ -150,7 +150,7 @@ Index: libavcodec/i386/dsputil_mmx.c
: "memory"\
);\
}\
@@ -2130,10 +2132,10 @@
@@ -2133,10 +2135,10 @@
"paddw %%mm5, %%mm5 \n\t" /* 2b */\
"psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
"pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
......@@ -163,7 +163,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
"paddw %6, %%mm6 \n\t"\
"paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
@@ -2151,8 +2153,8 @@
@@ -2154,8 +2156,8 @@
"paddw %%mm5, %%mm4 \n\t" /* d */\
"paddw %%mm2, %%mm2 \n\t" /* 2b */\
"psubw %%mm2, %%mm3 \n\t" /* c - 2b */\
......@@ -174,7 +174,7 @@ Index: libavcodec/i386/dsputil_mmx.c
"psubw %%mm4, %%mm3 \n\t" /* -6b + 3c - d */\
"paddw %6, %%mm1 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* 20a - 6b + 3c - d */\
@@ -2165,7 +2167,9 @@
@@ -2168,7 +2170,9 @@
"decl %2 \n\t"\
" jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
......@@ -185,7 +185,7 @@ Index: libavcodec/i386/dsputil_mmx.c
: "memory"\
);\
}\
@@ -2244,31 +2248,31 @@
@@ -2247,31 +2251,31 @@
"movq 8(%0), %%mm1 \n\t"\
"movq 16(%0), %%mm2 \n\t"\
"movq 24(%0), %%mm3 \n\t"\
......@@ -233,7 +233,7 @@ Index: libavcodec/i386/dsputil_mmx.c
\
"add $136, %0 \n\t"\
"add %6, %1 \n\t"\
@@ -2276,7 +2280,9 @@
@@ -2279,7 +2283,9 @@
" jnz 1b \n\t"\
\
: "+r"(temp_ptr), "+r"(dst), "+g"(count)\
......@@ -244,7 +244,7 @@ Index: libavcodec/i386/dsputil_mmx.c
:"memory"\
);\
}\
@@ -2316,19 +2322,19 @@
@@ -2319,19 +2325,19 @@
"movq 8(%0), %%mm1 \n\t"\
"movq 16(%0), %%mm2 \n\t"\
"movq 24(%0), %%mm3 \n\t"\
......@@ -272,7 +272,7 @@ Index: libavcodec/i386/dsputil_mmx.c
\
"add $72, %0 \n\t"\
"add %6, %1 \n\t"\
@@ -2336,7 +2342,9 @@
@@ -2339,7 +2345,9 @@
" jnz 1b \n\t"\
\
: "+r"(temp_ptr), "+r"(dst), "+g"(count)\
......@@ -283,43 +283,9 @@ Index: libavcodec/i386/dsputil_mmx.c
: "memory"\
);\
}\
@@ -2967,7 +2975,6 @@
double c = 2.0 / (len-1.0);
int n2 = len>>1;
long i = -n2*sizeof(int32_t);
- long j = n2*sizeof(int32_t);
asm volatile(
"movsd %0, %%xmm7 \n\t"
"movapd %1, %%xmm6 \n\t"
@@ -2985,17 +2992,18 @@
"movapd %%xmm6, %%xmm0 \n\t"\
"subpd %%xmm1, %%xmm0 \n\t"\
"pshufd $0x4e, %%xmm0, %%xmm1 \n\t"\
- "cvtpi2pd (%4,%0), %%xmm2 \n\t"\
- "cvtpi2pd (%5,%1), %%xmm3 \n\t"\
+ "cvtpi2pd (%3,%0), %%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"\
- "movapd %%xmm2, (%2,%0,2) \n\t"\
- 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"\
"jl 1b \n\t"\
- :"+&r"(i), "+&r"(j)\
+ :"+&r"(i)\
:"r"(w_data+n2), "r"(w_data+len-2-n2),\
"r"(data+n2), "r"(data+len-2-n2)\
);
Index: libavcodec/i386/simple_idct_mmx.c
===================================================================
--- libavcodec/i386/simple_idct_mmx.c (revision 10865)
--- libavcodec/i386/simple_idct_mmx.c (revision 11270)
+++ libavcodec/i386/simple_idct_mmx.c (working copy)
@@ -363,7 +363,7 @@
"movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
......@@ -366,10 +332,9 @@ Index: libavcodec/i386/simple_idct_mmx.c
: "%eax"
);
}
Index: libavcodec/i386/cavsdsp_mmx.c
===================================================================
--- libavcodec/i386/cavsdsp_mmx.c (revision 11171)
--- libavcodec/i386/cavsdsp_mmx.c (revision 11270)
+++ libavcodec/i386/cavsdsp_mmx.c (working copy)
@@ -23,9 +23,15 @@
*/
......@@ -388,3 +353,82 @@ Index: libavcodec/i386/cavsdsp_mmx.c
/*****************************************************************************
*
* inverse transform
Index: libavcodec/i386/flacdsp_mmx.c
===================================================================
--- libavcodec/i386/flacdsp_mmx.c (revision 11270)
+++ libavcodec/i386/flacdsp_mmx.c (working copy)
@@ -26,7 +26,6 @@
double c = 2.0 / (len-1.0);
int n2 = len>>1;
long i = -n2*sizeof(int32_t);
- long j = n2*sizeof(int32_t);
asm volatile(
"movsd %0, %%xmm7 \n\t"
"movapd %1, %%xmm6 \n\t"
@@ -44,17 +43,18 @@
"movapd %%xmm6, %%xmm0 \n\t"\
"subpd %%xmm1, %%xmm0 \n\t"\
"pshufd $0x4e, %%xmm0, %%xmm1 \n\t"\
- "cvtpi2pd (%4,%0), %%xmm2 \n\t"\
- "cvtpi2pd (%5,%1), %%xmm3 \n\t"\
+ "cvtpi2pd (%3,%0), %%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"\
- "movapd %%xmm2, (%2,%0,2) \n\t"\
- 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"\
"jl 1b \n\t"\
- :"+&r"(i), "+&r"(j)\
+ :"+&r"(i)\
:"r"(w_data+n2), "r"(w_data+len-2-n2),\
"r"(data+n2), "r"(data+len-2-n2)\
);
@@ -85,9 +85,11 @@
long i = -len*sizeof(double);
if(j == lag-2) {
asm volatile(
- "movsd %6, %%xmm0 \n\t"
- "movsd %6, %%xmm1 \n\t"
- "movsd %6, %%xmm2 \n\t"
+ "movsd %0, %%xmm0 \n\t"
+ "movsd %0, %%xmm1 \n\t"
+ "movsd %0, %%xmm2 \n\t"
+ :: "m"(*ff_pd_1) );
+ asm volatile(
"1: \n\t"
"movapd (%4,%0), %%xmm3 \n\t"
"movupd -8(%5,%0), %%xmm4 \n\t"
@@ -110,12 +112,14 @@
"movsd %%xmm1, %2 \n\t"
"movsd %%xmm2, %3 \n\t"
:"+&r"(i), "=m"(autoc[j]), "=m"(autoc[j+1]), "=m"(autoc[j+2])
- :"r"(data1+len), "r"(data1+len-j), "m"(*ff_pd_1)
+ :"r"(data1+len), "r"(data1+len-j)
);
} else {
asm volatile(
- "movsd %5, %%xmm0 \n\t"
- "movsd %5, %%xmm1 \n\t"
+ "movsd %0, %%xmm0 \n\t"
+ "movsd %0, %%xmm1 \n\t"
+ :: "m"(*ff_pd_1) );
+ asm volatile(
"1: \n\t"
"movapd (%3,%0), %%xmm3 \n\t"
"movupd -8(%4,%0), %%xmm4 \n\t"
@@ -132,7 +136,7 @@
"movsd %%xmm0, %1 \n\t"
"movsd %%xmm1, %2 \n\t"
:"+&r"(i), "=m"(autoc[j]), "=m"(autoc[j+1])
- :"r"(data1+len), "r"(data1+len-j), "m"(*ff_pd_1)
+ :"r"(data1+len), "r"(data1+len-j)
);
}
}
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