aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/i386
diff options
context:
space:
mode:
authorDiego Biurrun <diego@biurrun.de>2005-12-22 01:10:11 +0000
committerDiego Biurrun <diego@biurrun.de>2005-12-22 01:10:11 +0000
commitbb270c0896b39e1ae9277355e3c120ed3feb64a3 (patch)
treefc2fc2b1216d19acb3879abb6ea5a3b400f43fe4 /libavcodec/i386
parent50827fcf44f34521df4708cdb633809b56fb9df3 (diff)
downloadffmpeg-bb270c0896b39e1ae9277355e3c120ed3feb64a3.tar.gz
COSMETICS: tabs --> spaces, some prettyprinting
Originally committed as revision 4764 to svn://svn.ffmpeg.org/ffmpeg/trunk
Diffstat (limited to 'libavcodec/i386')
-rw-r--r--libavcodec/i386/cputest.c6
-rw-r--r--libavcodec/i386/dsputil_mmx.c1680
-rw-r--r--libavcodec/i386/dsputil_mmx_avg.h1280
-rw-r--r--libavcodec/i386/dsputil_mmx_rnd.h826
-rw-r--r--libavcodec/i386/fdct_mmx.c130
-rw-r--r--libavcodec/i386/fft_sse.c4
-rw-r--r--libavcodec/i386/h264dsp_mmx.c468
-rw-r--r--libavcodec/i386/idct_mmx.c514
-rw-r--r--libavcodec/i386/mmx.h424
-rw-r--r--libavcodec/i386/motion_est_mmx.c380
-rw-r--r--libavcodec/i386/mpegvideo_mmx.c900
-rw-r--r--libavcodec/i386/mpegvideo_mmx_template.c190
-rw-r--r--libavcodec/i386/simple_idct_mmx.c2054
13 files changed, 4428 insertions, 4428 deletions
diff --git a/libavcodec/i386/cputest.c b/libavcodec/i386/cputest.c
index f02c63d449..64656c65a5 100644
--- a/libavcodec/i386/cputest.c
+++ b/libavcodec/i386/cputest.c
@@ -15,7 +15,7 @@
/* ebx saving is necessary for PIC. gcc seems unable to see it alone */
#define cpuid(index,eax,ebx,ecx,edx)\
__asm __volatile\
- ("mov %%"REG_b", %%"REG_S"\n\t"\
+ ("mov %%"REG_b", %%"REG_S"\n\t"\
"cpuid\n\t"\
"xchg %%"REG_b", %%"REG_S\
: "=a" (eax), "=S" (ebx),\
@@ -89,8 +89,8 @@ int mm_support(void)
edx == 0x48727561 &&
ecx == 0x736c7561) { /* "CentaurHauls" */
/* VIA C3 */
- if(ext_caps & (1<<24))
- rval |= MM_MMXEXT;
+ if(ext_caps & (1<<24))
+ rval |= MM_MMXEXT;
} else if (ebx == 0x69727943 &&
edx == 0x736e4978 &&
ecx == 0x64616574) {
diff --git a/libavcodec/i386/dsputil_mmx.c b/libavcodec/i386/dsputil_mmx.c
index 7566b5d16a..ffdeafde72 100644
--- a/libavcodec/i386/dsputil_mmx.c
+++ b/libavcodec/i386/dsputil_mmx.c
@@ -89,56 +89,56 @@ static const uint64_t ff_pb_FC attribute_used __attribute__ ((aligned(8))) = 0xF
// first argument is unmodifed and second is trashed
// regfe is supposed to contain 0xfefefefefefefefe
#define PAVGB_MMX_NO_RND(rega, regb, regr, regfe) \
- "movq " #rega ", " #regr " \n\t"\
- "pand " #regb ", " #regr " \n\t"\
- "pxor " #rega ", " #regb " \n\t"\
- "pand " #regfe "," #regb " \n\t"\
- "psrlq $1, " #regb " \n\t"\
- "paddb " #regb ", " #regr " \n\t"
+ "movq " #rega ", " #regr " \n\t"\
+ "pand " #regb ", " #regr " \n\t"\
+ "pxor " #rega ", " #regb " \n\t"\
+ "pand " #regfe "," #regb " \n\t"\
+ "psrlq $1, " #regb " \n\t"\
+ "paddb " #regb ", " #regr " \n\t"
#define PAVGB_MMX(rega, regb, regr, regfe) \
- "movq " #rega ", " #regr " \n\t"\
- "por " #regb ", " #regr " \n\t"\
- "pxor " #rega ", " #regb " \n\t"\
- "pand " #regfe "," #regb " \n\t"\
- "psrlq $1, " #regb " \n\t"\
- "psubb " #regb ", " #regr " \n\t"
+ "movq " #rega ", " #regr " \n\t"\
+ "por " #regb ", " #regr " \n\t"\
+ "pxor " #rega ", " #regb " \n\t"\
+ "pand " #regfe "," #regb " \n\t"\
+ "psrlq $1, " #regb " \n\t"\
+ "psubb " #regb ", " #regr " \n\t"
// mm6 is supposed to contain 0xfefefefefefefefe
#define PAVGBP_MMX_NO_RND(rega, regb, regr, regc, regd, regp) \
- "movq " #rega ", " #regr " \n\t"\
- "movq " #regc ", " #regp " \n\t"\
- "pand " #regb ", " #regr " \n\t"\
- "pand " #regd ", " #regp " \n\t"\
- "pxor " #rega ", " #regb " \n\t"\
- "pxor " #regc ", " #regd " \n\t"\
- "pand %%mm6, " #regb " \n\t"\
- "pand %%mm6, " #regd " \n\t"\
- "psrlq $1, " #regb " \n\t"\
- "psrlq $1, " #regd " \n\t"\
- "paddb " #regb ", " #regr " \n\t"\
- "paddb " #regd ", " #regp " \n\t"
+ "movq " #rega ", " #regr " \n\t"\
+ "movq " #regc ", " #regp " \n\t"\
+ "pand " #regb ", " #regr " \n\t"\
+ "pand " #regd ", " #regp " \n\t"\
+ "pxor " #rega ", " #regb " \n\t"\
+ "pxor " #regc ", " #regd " \n\t"\
+ "pand %%mm6, " #regb " \n\t"\
+ "pand %%mm6, " #regd " \n\t"\
+ "psrlq $1, " #regb " \n\t"\
+ "psrlq $1, " #regd " \n\t"\
+ "paddb " #regb ", " #regr " \n\t"\
+ "paddb " #regd ", " #regp " \n\t"
#define PAVGBP_MMX(rega, regb, regr, regc, regd, regp) \
- "movq " #rega ", " #regr " \n\t"\
- "movq " #regc ", " #regp " \n\t"\
- "por " #regb ", " #regr " \n\t"\
- "por " #regd ", " #regp " \n\t"\
- "pxor " #rega ", " #regb " \n\t"\
- "pxor " #regc ", " #regd " \n\t"\
- "pand %%mm6, " #regb " \n\t"\
- "pand %%mm6, " #regd " \n\t"\
- "psrlq $1, " #regd " \n\t"\
- "psrlq $1, " #regb " \n\t"\
- "psubb " #regb ", " #regr " \n\t"\
- "psubb " #regd ", " #regp " \n\t"
+ "movq " #rega ", " #regr " \n\t"\
+ "movq " #regc ", " #regp " \n\t"\
+ "por " #regb ", " #regr " \n\t"\
+ "por " #regd ", " #regp " \n\t"\
+ "pxor " #rega ", " #regb " \n\t"\
+ "pxor " #regc ", " #regd " \n\t"\
+ "pand %%mm6, " #regb " \n\t"\
+ "pand %%mm6, " #regd " \n\t"\
+ "psrlq $1, " #regd " \n\t"\
+ "psrlq $1, " #regb " \n\t"\
+ "psubb " #regb ", " #regr " \n\t"\
+ "psubb " #regd ", " #regp " \n\t"
/***********************************/
/* MMX no rounding */
#define DEF(x, y) x ## _no_rnd_ ## y ##_mmx
#define SET_RND MOVQ_WONE
-#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX_NO_RND(a, b, c, d, e, f)
-#define PAVGB(a, b, c, e) PAVGB_MMX_NO_RND(a, b, c, e)
+#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX_NO_RND(a, b, c, d, e, f)
+#define PAVGB(a, b, c, e) PAVGB_MMX_NO_RND(a, b, c, e)
#include "dsputil_mmx_rnd.h"
@@ -151,8 +151,8 @@ static const uint64_t ff_pb_FC attribute_used __attribute__ ((aligned(8))) = 0xF
#define DEF(x, y) x ## _ ## y ##_mmx
#define SET_RND MOVQ_WTWO
-#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX(a, b, c, d, e, f)
-#define PAVGB(a, b, c, e) PAVGB_MMX(a, b, c, e)
+#define PAVGBP(a, b, c, d, e, f) PAVGBP_MMX(a, b, c, d, e, f)
+#define PAVGB(a, b, c, e) PAVGB_MMX(a, b, c, e)
#include "dsputil_mmx_rnd.h"
@@ -193,25 +193,25 @@ static const uint64_t ff_pb_FC attribute_used __attribute__ ((aligned(8))) = 0xF
static void get_pixels_mmx(DCTELEM *block, const uint8_t *pixels, int line_size)
{
asm volatile(
- "mov $-128, %%"REG_a" \n\t"
- "pxor %%mm7, %%mm7 \n\t"
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%0), %%mm0 \n\t"
- "movq (%0, %2), %%mm2 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "movq %%mm0, (%1, %%"REG_a")\n\t"
- "movq %%mm1, 8(%1, %%"REG_a")\n\t"
- "movq %%mm2, 16(%1, %%"REG_a")\n\t"
- "movq %%mm3, 24(%1, %%"REG_a")\n\t"
- "add %3, %0 \n\t"
- "add $32, %%"REG_a" \n\t"
- "js 1b \n\t"
+ "mov $-128, %%"REG_a" \n\t"
+ "pxor %%mm7, %%mm7 \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0), %%mm0 \n\t"
+ "movq (%0, %2), %%mm2 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "movq %%mm0, (%1, %%"REG_a") \n\t"
+ "movq %%mm1, 8(%1, %%"REG_a") \n\t"
+ "movq %%mm2, 16(%1, %%"REG_a") \n\t"
+ "movq %%mm3, 24(%1, %%"REG_a") \n\t"
+ "add %3, %0 \n\t"
+ "add $32, %%"REG_a" \n\t"
+ "js 1b \n\t"
: "+r" (pixels)
: "r" (block+64), "r" ((long)line_size), "r" ((long)line_size*2)
: "%"REG_a
@@ -221,26 +221,26 @@ static void get_pixels_mmx(DCTELEM *block, const uint8_t *pixels, int line_size)
static inline void diff_pixels_mmx(DCTELEM *block, const uint8_t *s1, const uint8_t *s2, int stride)
{
asm volatile(
- "pxor %%mm7, %%mm7 \n\t"
- "mov $-128, %%"REG_a" \n\t"
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%0), %%mm0 \n\t"
- "movq (%1), %%mm2 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "psubw %%mm2, %%mm0 \n\t"
- "psubw %%mm3, %%mm1 \n\t"
- "movq %%mm0, (%2, %%"REG_a")\n\t"
- "movq %%mm1, 8(%2, %%"REG_a")\n\t"
- "add %3, %0 \n\t"
- "add %3, %1 \n\t"
- "add $16, %%"REG_a" \n\t"
- "jnz 1b \n\t"
+ "pxor %%mm7, %%mm7 \n\t"
+ "mov $-128, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0), %%mm0 \n\t"
+ "movq (%1), %%mm2 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "psubw %%mm2, %%mm0 \n\t"
+ "psubw %%mm3, %%mm1 \n\t"
+ "movq %%mm0, (%2, %%"REG_a") \n\t"
+ "movq %%mm1, 8(%2, %%"REG_a") \n\t"
+ "add %3, %0 \n\t"
+ "add %3, %1 \n\t"
+ "add $16, %%"REG_a" \n\t"
+ "jnz 1b \n\t"
: "+r" (s1), "+r" (s2)
: "r" (block+64), "r" ((long)stride)
: "%"REG_a
@@ -257,25 +257,25 @@ void put_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size
p = block;
pix = pixels;
/* unrolled loop */
- __asm __volatile(
- "movq %3, %%mm0\n\t"
- "movq 8%3, %%mm1\n\t"
- "movq 16%3, %%mm2\n\t"
- "movq 24%3, %%mm3\n\t"
- "movq 32%3, %%mm4\n\t"
- "movq 40%3, %%mm5\n\t"
- "movq 48%3, %%mm6\n\t"
- "movq 56%3, %%mm7\n\t"
- "packuswb %%mm1, %%mm0\n\t"
- "packuswb %%mm3, %%mm2\n\t"
- "packuswb %%mm5, %%mm4\n\t"
- "packuswb %%mm7, %%mm6\n\t"
- "movq %%mm0, (%0)\n\t"
- "movq %%mm2, (%0, %1)\n\t"
- "movq %%mm4, (%0, %1, 2)\n\t"
- "movq %%mm6, (%0, %2)\n\t"
- ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "m"(*p)
- :"memory");
+ __asm __volatile(
+ "movq %3, %%mm0 \n\t"
+ "movq 8%3, %%mm1 \n\t"
+ "movq 16%3, %%mm2 \n\t"
+ "movq 24%3, %%mm3 \n\t"
+ "movq 32%3, %%mm4 \n\t"
+ "movq 40%3, %%mm5 \n\t"
+ "movq 48%3, %%mm6 \n\t"
+ "movq 56%3, %%mm7 \n\t"
+ "packuswb %%mm1, %%mm0 \n\t"
+ "packuswb %%mm3, %%mm2 \n\t"
+ "packuswb %%mm5, %%mm4 \n\t"
+ "packuswb %%mm7, %%mm6 \n\t"
+ "movq %%mm0, (%0) \n\t"
+ "movq %%mm2, (%0, %1) \n\t"
+ "movq %%mm4, (%0, %1, 2) \n\t"
+ "movq %%mm6, (%0, %2) \n\t"
+ ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "m"(*p)
+ :"memory");
pix += line_size*4;
p += 32;
@@ -283,24 +283,24 @@ void put_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size
// compiler would generate some very strange code
// thus using "r"
__asm __volatile(
- "movq (%3), %%mm0\n\t"
- "movq 8(%3), %%mm1\n\t"
- "movq 16(%3), %%mm2\n\t"
- "movq 24(%3), %%mm3\n\t"
- "movq 32(%3), %%mm4\n\t"
- "movq 40(%3), %%mm5\n\t"
- "movq 48(%3), %%mm6\n\t"
- "movq 56(%3), %%mm7\n\t"
- "packuswb %%mm1, %%mm0\n\t"
- "packuswb %%mm3, %%mm2\n\t"
- "packuswb %%mm5, %%mm4\n\t"
- "packuswb %%mm7, %%mm6\n\t"
- "movq %%mm0, (%0)\n\t"
- "movq %%mm2, (%0, %1)\n\t"
- "movq %%mm4, (%0, %1, 2)\n\t"
- "movq %%mm6, (%0, %2)\n\t"
- ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "r"(p)
- :"memory");
+ "movq (%3), %%mm0 \n\t"
+ "movq 8(%3), %%mm1 \n\t"
+ "movq 16(%3), %%mm2 \n\t"
+ "movq 24(%3), %%mm3 \n\t"
+ "movq 32(%3), %%mm4 \n\t"
+ "movq 40(%3), %%mm5 \n\t"
+ "movq 48(%3), %%mm6 \n\t"
+ "movq 56(%3), %%mm7 \n\t"
+ "packuswb %%mm1, %%mm0 \n\t"
+ "packuswb %%mm3, %%mm2 \n\t"
+ "packuswb %%mm5, %%mm4 \n\t"
+ "packuswb %%mm7, %%mm6 \n\t"
+ "movq %%mm0, (%0) \n\t"
+ "movq %%mm2, (%0, %1) \n\t"
+ "movq %%mm4, (%0, %1, 2) \n\t"
+ "movq %%mm6, (%0, %2) \n\t"
+ ::"r" (pix), "r" ((long)line_size), "r" ((long)line_size*3), "r"(p)
+ :"memory");
}
static const unsigned char __align8 vector128[8] =
@@ -333,30 +333,30 @@ void add_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size
MOVQ_ZERO(mm7);
i = 4;
do {
- __asm __volatile(
- "movq (%2), %%mm0\n\t"
- "movq 8(%2), %%mm1\n\t"
- "movq 16(%2), %%mm2\n\t"
- "movq 24(%2), %%mm3\n\t"
- "movq %0, %%mm4\n\t"
- "movq %1, %%mm6\n\t"
- "movq %%mm4, %%mm5\n\t"
- "punpcklbw %%mm7, %%mm4\n\t"
- "punpckhbw %%mm7, %%mm5\n\t"
- "paddsw %%mm4, %%mm0\n\t"
- "paddsw %%mm5, %%mm1\n\t"
- "movq %%mm6, %%mm5\n\t"
- "punpcklbw %%mm7, %%mm6\n\t"
- "punpckhbw %%mm7, %%mm5\n\t"
- "paddsw %%mm6, %%mm2\n\t"
- "paddsw %%mm5, %%mm3\n\t"
- "packuswb %%mm1, %%mm0\n\t"
- "packuswb %%mm3, %%mm2\n\t"
- "movq %%mm0, %0\n\t"
- "movq %%mm2, %1\n\t"
- :"+m"(*pix), "+m"(*(pix+line_size))
- :"r"(p)
- :"memory");
+ __asm __volatile(
+ "movq (%2), %%mm0 \n\t"
+ "movq 8(%2), %%mm1 \n\t"
+ "movq 16(%2), %%mm2 \n\t"
+ "movq 24(%2), %%mm3 \n\t"
+ "movq %0, %%mm4 \n\t"
+ "movq %1, %%mm6 \n\t"
+ "movq %%mm4, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddsw %%mm4, %%mm0 \n\t"
+ "paddsw %%mm5, %%mm1 \n\t"
+ "movq %%mm6, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm6 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddsw %%mm6, %%mm2 \n\t"
+ "paddsw %%mm5, %%mm3 \n\t"
+ "packuswb %%mm1, %%mm0 \n\t"
+ "packuswb %%mm3, %%mm2 \n\t"
+ "movq %%mm0, %0 \n\t"
+ "movq %%mm2, %1 \n\t"
+ :"+m"(*pix), "+m"(*(pix+line_size))
+ :"r"(p)
+ :"memory");
pix += line_size*2;
p += 16;
} while (--i);
@@ -365,101 +365,101 @@ void add_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size
static void put_pixels4_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movd (%1), %%mm0 \n\t"
- "movd (%1, %3), %%mm1 \n\t"
- "movd %%mm0, (%2) \n\t"
- "movd %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movd (%1), %%mm0 \n\t"
- "movd (%1, %3), %%mm1 \n\t"
- "movd %%mm0, (%2) \n\t"
- "movd %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- : "+g"(h), "+r" (pixels), "+r" (block)
- : "r"((long)line_size)
- : "%"REG_a, "memory"
- );
+ "lea (%3, %3), %%"REG_a" \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "movd (%1, %3), %%mm1 \n\t"
+ "movd %%mm0, (%2) \n\t"
+ "movd %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "movd (%1, %3), %%mm1 \n\t"
+ "movd %%mm0, (%2) \n\t"
+ "movd %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ : "+g"(h), "+r" (pixels), "+r" (block)
+ : "r"((long)line_size)
+ : "%"REG_a, "memory"
+ );
}
static void put_pixels8_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- : "+g"(h), "+r" (pixels), "+r" (block)
- : "r"((long)line_size)
- : "%"REG_a, "memory"
- );
+ "lea (%3, %3), %%"REG_a" \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ : "+g"(h), "+r" (pixels), "+r" (block)
+ : "r"((long)line_size)
+ : "%"REG_a, "memory"
+ );
}
static void put_pixels16_mmx(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm4 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq 8(%1, %3), %%mm5 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm4, 8(%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "movq %%mm5, 8(%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm4 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq 8(%1, %3), %%mm5 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm4, 8(%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "movq %%mm5, 8(%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- : "+g"(h), "+r" (pixels), "+r" (block)
- : "r"((long)line_size)
- : "%"REG_a, "memory"
- );
+ "lea (%3, %3), %%"REG_a" \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm4 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq 8(%1, %3), %%mm5 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm4, 8(%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "movq %%mm5, 8(%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm4 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq 8(%1, %3), %%mm5 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm4, 8(%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "movq %%mm5, 8(%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ : "+g"(h), "+r" (pixels), "+r" (block)
+ : "r"((long)line_size)
+ : "%"REG_a, "memory"
+ );
}
static void clear_blocks_mmx(DCTELEM *blocks)
{
__asm __volatile(
- "pxor %%mm7, %%mm7 \n\t"
- "mov $-128*6, %%"REG_a" \n\t"
- "1: \n\t"
- "movq %%mm7, (%0, %%"REG_a") \n\t"
- "movq %%mm7, 8(%0, %%"REG_a") \n\t"
- "movq %%mm7, 16(%0, %%"REG_a") \n\t"
- "movq %%mm7, 24(%0, %%"REG_a") \n\t"
- "add $32, %%"REG_a" \n\t"
- " js 1b \n\t"
+ "pxor %%mm7, %%mm7 \n\t"
+ "mov $-128*6, %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq %%mm7, (%0, %%"REG_a") \n\t"
+ "movq %%mm7, 8(%0, %%"REG_a") \n\t"
+ "movq %%mm7, 16(%0, %%"REG_a") \n\t"
+ "movq %%mm7, 24(%0, %%"REG_a") \n\t"
+ "add $32, %%"REG_a" \n\t"
+ " js 1b \n\t"
: : "r" (((uint8_t *)blocks)+128*6)
: "%"REG_a
);
@@ -472,31 +472,31 @@ static int pix_sum16_mmx(uint8_t * pix, int line_size){
long index= -line_size*h;
__asm __volatile(
- "pxor %%mm7, %%mm7 \n\t"
- "pxor %%mm6, %%mm6 \n\t"
- "1: \n\t"
- "movq (%2, %1), %%mm0 \n\t"
- "movq (%2, %1), %%mm1 \n\t"
- "movq 8(%2, %1), %%mm2 \n\t"
- "movq 8(%2, %1), %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "paddw %%mm0, %%mm1 \n\t"
- "paddw %%mm2, %%mm3 \n\t"
- "paddw %%mm1, %%mm3 \n\t"
- "paddw %%mm3, %%mm6 \n\t"
- "add %3, %1 \n\t"
- " js 1b \n\t"
- "movq %%mm6, %%mm5 \n\t"
- "psrlq $32, %%mm6 \n\t"
- "paddw %%mm5, %%mm6 \n\t"
- "movq %%mm6, %%mm5 \n\t"
- "psrlq $16, %%mm6 \n\t"
- "paddw %%mm5, %%mm6 \n\t"
- "movd %%mm6, %0 \n\t"
- "andl $0xFFFF, %0 \n\t"
+ "pxor %%mm7, %%mm7 \n\t"
+ "pxor %%mm6, %%mm6 \n\t"
+ "1: \n\t"
+ "movq (%2, %1), %%mm0 \n\t"
+ "movq (%2, %1), %%mm1 \n\t"
+ "movq 8(%2, %1), %%mm2 \n\t"
+ "movq 8(%2, %1), %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "paddw %%mm0, %%mm1 \n\t"
+ "paddw %%mm2, %%mm3 \n\t"
+ "paddw %%mm1, %%mm3 \n\t"
+ "paddw %%mm3, %%mm6 \n\t"
+ "add %3, %1 \n\t"
+ " js 1b \n\t"
+ "movq %%mm6, %%mm5 \n\t"
+ "psrlq $32, %%mm6 \n\t"
+ "paddw %%mm5, %%mm6 \n\t"
+ "movq %%mm6, %%mm5 \n\t"
+ "psrlq $16, %%mm6 \n\t"
+ "paddw %%mm5, %%mm6 \n\t"
+ "movd %%mm6, %0 \n\t"
+ "andl $0xFFFF, %0 \n\t"
: "=&r" (sum), "+r" (index)
: "r" (pix - index), "r" ((long)line_size)
);
@@ -508,18 +508,18 @@ static int pix_sum16_mmx(uint8_t * pix, int line_size){
static void add_bytes_mmx(uint8_t *dst, uint8_t *src, int w){
long i=0;
asm volatile(
- "1: \n\t"
- "movq (%1, %0), %%mm0 \n\t"
- "movq (%2, %0), %%mm1 \n\t"
- "paddb %%mm0, %%mm1 \n\t"
- "movq %%mm1, (%2, %0) \n\t"
- "movq 8(%1, %0), %%mm0 \n\t"
- "movq 8(%2, %0), %%mm1 \n\t"
- "paddb %%mm0, %%mm1 \n\t"
- "movq %%mm1, 8(%2, %0) \n\t"
- "add $16, %0 \n\t"
- "cmp %3, %0 \n\t"
- " jb 1b \n\t"
+ "1: \n\t"
+ "movq (%1, %0), %%mm0 \n\t"
+ "movq (%2, %0), %%mm1 \n\t"
+ "paddb %%mm0, %%mm1 \n\t"
+ "movq %%mm1, (%2, %0) \n\t"
+ "movq 8(%1, %0), %%mm0 \n\t"
+ "movq 8(%2, %0), %%mm1 \n\t"
+ "paddb %%mm0, %%mm1 \n\t"
+ "movq %%mm1, 8(%2, %0) \n\t"
+ "add $16, %0 \n\t"
+ "cmp %3, %0 \n\t"
+ " jb 1b \n\t"
: "+r" (i)
: "r"(src), "r"(dst), "r"((long)w-15)
);
@@ -528,75 +528,75 @@ static void add_bytes_mmx(uint8_t *dst, uint8_t *src, int w){
}
#define H263_LOOP_FILTER \
- "pxor %%mm7, %%mm7 \n\t"\
- "movq %0, %%mm0 \n\t"\
- "movq %0, %%mm1 \n\t"\
- "movq %3, %%mm2 \n\t"\
- "movq %3, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpckhbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpckhbw %%mm7, %%mm3 \n\t"\
- "psubw %%mm2, %%mm0 \n\t"\
- "psubw %%mm3, %%mm1 \n\t"\
- "movq %1, %%mm2 \n\t"\
- "movq %1, %%mm3 \n\t"\
- "movq %2, %%mm4 \n\t"\
- "movq %2, %%mm5 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpckhbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
- "punpckhbw %%mm7, %%mm5 \n\t"\
- "psubw %%mm2, %%mm4 \n\t"\
- "psubw %%mm3, %%mm5 \n\t"\
- "psllw $2, %%mm4 \n\t"\
- "psllw $2, %%mm5 \n\t"\
- "paddw %%mm0, %%mm4 \n\t"\
- "paddw %%mm1, %%mm5 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "pcmpgtw %%mm4, %%mm6 \n\t"\
- "pcmpgtw %%mm5, %%mm7 \n\t"\
- "pxor %%mm6, %%mm4 \n\t"\
- "pxor %%mm7, %%mm5 \n\t"\
- "psubw %%mm6, %%mm4 \n\t"\
- "psubw %%mm7, %%mm5 \n\t"\
- "psrlw $3, %%mm4 \n\t"\
- "psrlw $3, %%mm5 \n\t"\
- "packuswb %%mm5, %%mm4 \n\t"\
- "packsswb %%mm7, %%mm6 \n\t"\
- "pxor %%mm7, %%mm7 \n\t"\
- "movd %4, %%mm2 \n\t"\
- "punpcklbw %%mm2, %%mm2 \n\t"\
- "punpcklbw %%mm2, %%mm2 \n\t"\
- "punpcklbw %%mm2, %%mm2 \n\t"\
- "psubusb %%mm4, %%mm2 \n\t"\
- "movq %%mm2, %%mm3 \n\t"\
- "psubusb %%mm4, %%mm3 \n\t"\
- "psubb %%mm3, %%mm2 \n\t"\
- "movq %1, %%mm3 \n\t"\
- "movq %2, %%mm4 \n\t"\
- "pxor %%mm6, %%mm3 \n\t"\
- "pxor %%mm6, %%mm4 \n\t"\
- "paddusb %%mm2, %%mm3 \n\t"\
- "psubusb %%mm2, %%mm4 \n\t"\
- "pxor %%mm6, %%mm3 \n\t"\
- "pxor %%mm6, %%mm4 \n\t"\
- "paddusb %%mm2, %%mm2 \n\t"\
- "packsswb %%mm1, %%mm0 \n\t"\
- "pcmpgtb %%mm0, %%mm7 \n\t"\
- "pxor %%mm7, %%mm0 \n\t"\
- "psubb %%mm7, %%mm0 \n\t"\
- "movq %%mm0, %%mm1 \n\t"\
- "psubusb %%mm2, %%mm0 \n\t"\
- "psubb %%mm0, %%mm1 \n\t"\
- "pand %5, %%mm1 \n\t"\
- "psrlw $2, %%mm1 \n\t"\
- "pxor %%mm7, %%mm1 \n\t"\
- "psubb %%mm7, %%mm1 \n\t"\
- "movq %0, %%mm5 \n\t"\
- "movq %3, %%mm6 \n\t"\
- "psubb %%mm1, %%mm5 \n\t"\
- "paddb %%mm1, %%mm6 \n\t"
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movq %0, %%mm0 \n\t"\
+ "movq %0, %%mm1 \n\t"\
+ "movq %3, %%mm2 \n\t"\
+ "movq %3, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpckhbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpckhbw %%mm7, %%mm3 \n\t"\
+ "psubw %%mm2, %%mm0 \n\t"\
+ "psubw %%mm3, %%mm1 \n\t"\
+ "movq %1, %%mm2 \n\t"\
+ "movq %1, %%mm3 \n\t"\
+ "movq %2, %%mm4 \n\t"\
+ "movq %2, %%mm5 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpckhbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
+ "punpckhbw %%mm7, %%mm5 \n\t"\
+ "psubw %%mm2, %%mm4 \n\t"\
+ "psubw %%mm3, %%mm5 \n\t"\
+ "psllw $2, %%mm4 \n\t"\
+ "psllw $2, %%mm5 \n\t"\
+ "paddw %%mm0, %%mm4 \n\t"\
+ "paddw %%mm1, %%mm5 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "pcmpgtw %%mm4, %%mm6 \n\t"\
+ "pcmpgtw %%mm5, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm4 \n\t"\
+ "pxor %%mm7, %%mm5 \n\t"\
+ "psubw %%mm6, %%mm4 \n\t"\
+ "psubw %%mm7, %%mm5 \n\t"\
+ "psrlw $3, %%mm4 \n\t"\
+ "psrlw $3, %%mm5 \n\t"\
+ "packuswb %%mm5, %%mm4 \n\t"\
+ "packsswb %%mm7, %%mm6 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movd %4, %%mm2 \n\t"\
+ "punpcklbw %%mm2, %%mm2 \n\t"\
+ "punpcklbw %%mm2, %%mm2 \n\t"\
+ "punpcklbw %%mm2, %%mm2 \n\t"\
+ "psubusb %%mm4, %%mm2 \n\t"\
+ "movq %%mm2, %%mm3 \n\t"\
+ "psubusb %%mm4, %%mm3 \n\t"\
+ "psubb %%mm3, %%mm2 \n\t"\
+ "movq %1, %%mm3 \n\t"\
+ "movq %2, %%mm4 \n\t"\
+ "pxor %%mm6, %%mm3 \n\t"\
+ "pxor %%mm6, %%mm4 \n\t"\
+ "paddusb %%mm2, %%mm3 \n\t"\
+ "psubusb %%mm2, %%mm4 \n\t"\
+ "pxor %%mm6, %%mm3 \n\t"\
+ "pxor %%mm6, %%mm4 \n\t"\
+ "paddusb %%mm2, %%mm2 \n\t"\
+ "packsswb %%mm1, %%mm0 \n\t"\
+ "pcmpgtb %%mm0, %%mm7 \n\t"\
+ "pxor %%mm7, %%mm0 \n\t"\
+ "psubb %%mm7, %%mm0 \n\t"\
+ "movq %%mm0, %%mm1 \n\t"\
+ "psubusb %%mm2, %%mm0 \n\t"\
+ "psubb %%mm0, %%mm1 \n\t"\
+ "pand %5, %%mm1 \n\t"\
+ "psrlw $2, %%mm1 \n\t"\
+ "pxor %%mm7, %%mm1 \n\t"\
+ "psubb %%mm7, %%mm1 \n\t"\
+ "movq %0, %%mm5 \n\t"\
+ "movq %3, %%mm6 \n\t"\
+ "psubb %%mm1, %%mm5 \n\t"\
+ "paddb %%mm1, %%mm6 \n\t"
static void h263_v_loop_filter_mmx(uint8_t *src, int stride, int qscale){
const int strength= ff_h263_loop_filter_strength[qscale];
@@ -605,10 +605,10 @@ static void h263_v_loop_filter_mmx(uint8_t *src, int stride, int qscale){
H263_LOOP_FILTER
- "movq %%mm3, %1 \n\t"
- "movq %%mm4, %2 \n\t"
- "movq %%mm5, %0 \n\t"
- "movq %%mm6, %3 \n\t"
+ "movq %%mm3, %1 \n\t"
+ "movq %%mm4, %2 \n\t"
+ "movq %%mm5, %0 \n\t"
+ "movq %%mm6, %3 \n\t"
: "+m" (*(uint64_t*)(src - 2*stride)),
"+m" (*(uint64_t*)(src - 1*stride)),
"+m" (*(uint64_t*)(src + 0*stride)),
@@ -619,21 +619,21 @@ static void h263_v_loop_filter_mmx(uint8_t *src, int stride, int qscale){
static inline void transpose4x4(uint8_t *dst, uint8_t *src, int dst_stride, int src_stride){
asm volatile( //FIXME could save 1 instruction if done as 8x4 ...
- "movd %4, %%mm0 \n\t"
- "movd %5, %%mm1 \n\t"
- "movd %6, %%mm2 \n\t"
- "movd %7, %%mm3 \n\t"
- "punpcklbw %%mm1, %%mm0 \n\t"
- "punpcklbw %%mm3, %%mm2 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "punpcklwd %%mm2, %%mm0 \n\t"
- "punpckhwd %%mm2, %%mm1 \n\t"
- "movd %%mm0, %0 \n\t"
- "punpckhdq %%mm0, %%mm0 \n\t"
- "movd %%mm0, %1 \n\t"
- "movd %%mm1, %2 \n\t"
- "punpckhdq %%mm1, %%mm1 \n\t"
- "movd %%mm1, %3 \n\t"
+ "movd %4, %%mm0 \n\t"
+ "movd %5, %%mm1 \n\t"
+ "movd %6, %%mm2 \n\t"
+ "movd %7, %%mm3 \n\t"
+ "punpcklbw %%mm1, %%mm0 \n\t"
+ "punpcklbw %%mm3, %%mm2 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "punpcklwd %%mm2, %%mm0 \n\t"
+ "punpckhwd %%mm2, %%mm1 \n\t"
+ "movd %%mm0, %0 \n\t"
+ "punpckhdq %%mm0, %%mm0 \n\t"
+ "movd %%mm0, %1 \n\t"
+ "movd %%mm1, %2 \n\t"
+ "punpckhdq %%mm1, %%mm1 \n\t"
+ "movd %%mm1, %3 \n\t"
: "=m" (*(uint32_t*)(dst + 0*dst_stride)),
"=m" (*(uint32_t*)(dst + 1*dst_stride)),
@@ -666,30 +666,30 @@ static void h263_h_loop_filter_mmx(uint8_t *src, int stride, int qscale){
);
asm volatile(
- "movq %%mm5, %%mm1 \n\t"
- "movq %%mm4, %%mm0 \n\t"
- "punpcklbw %%mm3, %%mm5 \n\t"
- "punpcklbw %%mm6, %%mm4 \n\t"
- "punpckhbw %%mm3, %%mm1 \n\t"
- "punpckhbw %%mm6, %%mm0 \n\t"
- "movq %%mm5, %%mm3 \n\t"
- "movq %%mm1, %%mm6 \n\t"
- "punpcklwd %%mm4, %%mm5 \n\t"
- "punpcklwd %%mm0, %%mm1 \n\t"
- "punpckhwd %%mm4, %%mm3 \n\t"
- "punpckhwd %%mm0, %%mm6 \n\t"
- "movd %%mm5, (%0) \n\t"
- "punpckhdq %%mm5, %%mm5 \n\t"
- "movd %%mm5, (%0,%2) \n\t"
- "movd %%mm3, (%0,%2,2) \n\t"
- "punpckhdq %%mm3, %%mm3 \n\t"
- "movd %%mm3, (%0,%3) \n\t"
- "movd %%mm1, (%1) \n\t"
- "punpckhdq %%mm1, %%mm1 \n\t"
- "movd %%mm1, (%1,%2) \n\t"
- "movd %%mm6, (%1,%2,2) \n\t"
- "punpckhdq %%mm6, %%mm6 \n\t"
- "movd %%mm6, (%1,%3) \n\t"
+ "movq %%mm5, %%mm1 \n\t"
+ "movq %%mm4, %%mm0 \n\t"
+ "punpcklbw %%mm3, %%mm5 \n\t"
+ "punpcklbw %%mm6, %%mm4 \n\t"
+ "punpckhbw %%mm3, %%mm1 \n\t"
+ "punpckhbw %%mm6, %%mm0 \n\t"
+ "movq %%mm5, %%mm3 \n\t"
+ "movq %%mm1, %%mm6 \n\t"
+ "punpcklwd %%mm4, %%mm5 \n\t"
+ "punpcklwd %%mm0, %%mm1 \n\t"
+ "punpckhwd %%mm4, %%mm3 \n\t"
+ "punpckhwd %%mm0, %%mm6 \n\t"
+ "movd %%mm5, (%0) \n\t"
+ "punpckhdq %%mm5, %%mm5 \n\t"
+ "movd %%mm5, (%0,%2) \n\t"
+ "movd %%mm3, (%0,%2,2) \n\t"
+ "punpckhdq %%mm3, %%mm3 \n\t"
+ "movd %%mm3, (%0,%3) \n\t"
+ "movd %%mm1, (%1) \n\t"
+ "punpckhdq %%mm1, %%mm1 \n\t"
+ "movd %%mm1, (%1,%2) \n\t"
+ "movd %%mm6, (%1,%2,2) \n\t"
+ "punpckhdq %%mm6, %%mm6 \n\t"
+ "movd %%mm6, (%1,%3) \n\t"
:: "r" (src),
"r" (src + 4*stride),
"r" ((long) stride ),
@@ -705,26 +705,26 @@ static int pix_norm1_mmx(uint8_t *pix, int line_size) {
"pxor %%mm0,%%mm0\n"
"pxor %%mm7,%%mm7\n"
"1:\n"
- "movq (%0),%%mm2\n" /* mm2 = pix[0-7] */
- "movq 8(%0),%%mm3\n" /* mm3 = pix[8-15] */
+ "movq (%0),%%mm2\n" /* mm2 = pix[0-7] */
+ "movq 8(%0),%%mm3\n" /* mm3 = pix[8-15] */
- "movq %%mm2,%%mm1\n" /* mm1 = mm2 = pix[0-7] */
+ "movq %%mm2,%%mm1\n" /* mm1 = mm2 = pix[0-7] */
- "punpckhbw %%mm0,%%mm1\n" /* mm1 = [pix4-7] */
- "punpcklbw %%mm0,%%mm2\n" /* mm2 = [pix0-3] */
+ "punpckhbw %%mm0,%%mm1\n" /* mm1 = [pix4-7] */
+ "punpcklbw %%mm0,%%mm2\n" /* mm2 = [pix0-3] */
- "movq %%mm3,%%mm4\n" /* mm4 = mm3 = pix[8-15] */
- "punpckhbw %%mm0,%%mm3\n" /* mm3 = [pix12-15] */
- "punpcklbw %%mm0,%%mm4\n" /* mm4 = [pix8-11] */
+ "movq %%mm3,%%mm4\n" /* mm4 = mm3 = pix[8-15] */
+ "punpckhbw %%mm0,%%mm3\n" /* mm3 = [pix12-15] */
+ "punpcklbw %%mm0,%%mm4\n" /* mm4 = [pix8-11] */
- "pmaddwd %%mm1,%%mm1\n" /* mm1 = (pix0^2+pix1^2,pix2^2+pix3^2) */
- "pmaddwd %%mm2,%%mm2\n" /* mm2 = (pix4^2+pix5^2,pix6^2+pix7^2) */
+ "pmaddwd %%mm1,%%mm1\n" /* mm1 = (pix0^2+pix1^2,pix2^2+pix3^2) */
+ "pmaddwd %%mm2,%%mm2\n" /* mm2 = (pix4^2+pix5^2,pix6^2+pix7^2) */
"pmaddwd %%mm3,%%mm3\n"
"pmaddwd %%mm4,%%mm4\n"
- "paddd %%mm1,%%mm2\n" /* mm2 = (pix0^2+pix1^2+pix4^2+pix5^2,
- pix2^2+pix3^2+pix6^2+pix7^2) */
+ "paddd %%mm1,%%mm2\n" /* mm2 = (pix0^2+pix1^2+pix4^2+pix5^2,
+ pix2^2+pix3^2+pix6^2+pix7^2) */
"paddd %%mm3,%%mm4\n"
"paddd %%mm2,%%mm7\n"
@@ -734,7 +734,7 @@ static int pix_norm1_mmx(uint8_t *pix, int line_size) {
"jnz 1b\n"
"movq %%mm7,%%mm1\n"
- "psrlq $32, %%mm7\n" /* shift hi dword to lo */
+ "psrlq $32, %%mm7\n" /* shift hi dword to lo */
"paddd %%mm7,%%mm1\n"
"movd %%mm1,%1\n"
: "+r" (pix), "=r"(tmp) : "r" ((long)line_size) : "%ecx" );
@@ -746,13 +746,13 @@ static int sse8_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
asm volatile (
"movl %4,%%ecx\n"
"shr $1,%%ecx\n"
- "pxor %%mm0,%%mm0\n" /* mm0 = 0 */
- "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */
+ "pxor %%mm0,%%mm0\n" /* mm0 = 0 */
+ "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */
"1:\n"
- "movq (%0),%%mm1\n" /* mm1 = pix1[0][0-7] */
- "movq (%1),%%mm2\n" /* mm2 = pix2[0][0-7] */
- "movq (%0,%3),%%mm3\n" /* mm3 = pix1[1][0-7] */
- "movq (%1,%3),%%mm4\n" /* mm4 = pix2[1][0-7] */
+ "movq (%0),%%mm1\n" /* mm1 = pix1[0][0-7] */
+ "movq (%1),%%mm2\n" /* mm2 = pix2[0][0-7] */
+ "movq (%0,%3),%%mm3\n" /* mm3 = pix1[1][0-7] */
+ "movq (%1,%3),%%mm4\n" /* mm4 = pix2[1][0-7] */
/* todo: mm1-mm2, mm3-mm4 */
/* algo: substract mm1 from mm2 with saturation and vice versa */
@@ -773,16 +773,16 @@ static int sse8_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
"punpckhbw %%mm0,%%mm2\n"
"punpckhbw %%mm0,%%mm4\n"
- "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */
- "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */
+ "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */
+ "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */
"pmaddwd %%mm2,%%mm2\n"
"pmaddwd %%mm4,%%mm4\n"
"pmaddwd %%mm1,%%mm1\n"
"pmaddwd %%mm3,%%mm3\n"
- "lea (%0,%3,2), %0\n" /* pix1 += 2*line_size */
- "lea (%1,%3,2), %1\n" /* pix2 += 2*line_size */
+ "lea (%0,%3,2), %0\n" /* pix1 += 2*line_size */
+ "lea (%1,%3,2), %1\n" /* pix2 += 2*line_size */
"paddd %%mm2,%%mm1\n"
"paddd %%mm4,%%mm3\n"
@@ -793,7 +793,7 @@ static int sse8_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
"jnz 1b\n"
"movq %%mm7,%%mm1\n"
- "psrlq $32, %%mm7\n" /* shift hi dword to lo */
+ "psrlq $32, %%mm7\n" /* shift hi dword to lo */
"paddd %%mm7,%%mm1\n"
"movd %%mm1,%2\n"
: "+r" (pix1), "+r" (pix2), "=r"(tmp)
@@ -806,13 +806,13 @@ static int sse16_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
int tmp;
asm volatile (
"movl %4,%%ecx\n"
- "pxor %%mm0,%%mm0\n" /* mm0 = 0 */
- "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */
+ "pxor %%mm0,%%mm0\n" /* mm0 = 0 */
+ "pxor %%mm7,%%mm7\n" /* mm7 holds the sum */
"1:\n"
- "movq (%0),%%mm1\n" /* mm1 = pix1[0-7] */
- "movq (%1),%%mm2\n" /* mm2 = pix2[0-7] */
- "movq 8(%0),%%mm3\n" /* mm3 = pix1[8-15] */
- "movq 8(%1),%%mm4\n" /* mm4 = pix2[8-15] */
+ "movq (%0),%%mm1\n" /* mm1 = pix1[0-7] */
+ "movq (%1),%%mm2\n" /* mm2 = pix2[0-7] */
+ "movq 8(%0),%%mm3\n" /* mm3 = pix1[8-15] */
+ "movq 8(%1),%%mm4\n" /* mm4 = pix2[8-15] */
/* todo: mm1-mm2, mm3-mm4 */
/* algo: substract mm1 from mm2 with saturation and vice versa */
@@ -833,8 +833,8 @@ static int sse16_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
"punpckhbw %%mm0,%%mm2\n"
"punpckhbw %%mm0,%%mm4\n"
- "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */
- "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */
+ "punpcklbw %%mm0,%%mm1\n" /* mm1 now spread over (mm1,mm2) */
+ "punpcklbw %%mm0,%%mm3\n" /* mm4 now spread over (mm3,mm4) */
"pmaddwd %%mm2,%%mm2\n"
"pmaddwd %%mm4,%%mm4\n"
@@ -853,7 +853,7 @@ static int sse16_mmx(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int
"jnz 1b\n"
"movq %%mm7,%%mm1\n"
- "psrlq $32, %%mm7\n" /* shift hi dword to lo */
+ "psrlq $32, %%mm7\n" /* shift hi dword to lo */
"paddd %%mm7,%%mm1\n"
"movd %%mm1,%2\n"
: "+r" (pix1), "+r" (pix2), "=r"(tmp)
@@ -866,13 +866,13 @@ static int sse16_sse2(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in
int tmp;
asm volatile (
"shr $1,%2\n"
- "pxor %%xmm0,%%xmm0\n" /* mm0 = 0 */
- "pxor %%xmm7,%%xmm7\n" /* mm7 holds the sum */
+ "pxor %%xmm0,%%xmm0\n" /* mm0 = 0 */
+ "pxor %%xmm7,%%xmm7\n" /* mm7 holds the sum */
"1:\n"
- "movdqu (%0),%%xmm1\n" /* mm1 = pix1[0][0-15] */
- "movdqu (%1),%%xmm2\n" /* mm2 = pix2[0][0-15] */
- "movdqu (%0,%4),%%xmm3\n" /* mm3 = pix1[1][0-15] */
- "movdqu (%1,%4),%%xmm4\n" /* mm4 = pix2[1][0-15] */
+ "movdqu (%0),%%xmm1\n" /* mm1 = pix1[0][0-15] */
+ "movdqu (%1),%%xmm2\n" /* mm2 = pix2[0][0-15] */
+ "movdqu (%0,%4),%%xmm3\n" /* mm3 = pix1[1][0-15] */
+ "movdqu (%1,%4),%%xmm4\n" /* mm4 = pix2[1][0-15] */
/* todo: mm1-mm2, mm3-mm4 */
/* algo: substract mm1 from mm2 with saturation and vice versa */
@@ -893,16 +893,16 @@ static int sse16_sse2(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in
"punpckhbw %%xmm0,%%xmm2\n"
"punpckhbw %%xmm0,%%xmm4\n"
- "punpcklbw %%xmm0,%%xmm1\n" /* mm1 now spread over (mm1,mm2) */
- "punpcklbw %%xmm0,%%xmm3\n" /* mm4 now spread over (mm3,mm4) */
+ "punpcklbw %%xmm0,%%xmm1\n" /* mm1 now spread over (mm1,mm2) */
+ "punpcklbw %%xmm0,%%xmm3\n" /* mm4 now spread over (mm3,mm4) */
"pmaddwd %%xmm2,%%xmm2\n"
"pmaddwd %%xmm4,%%xmm4\n"
"pmaddwd %%xmm1,%%xmm1\n"
"pmaddwd %%xmm3,%%xmm3\n"
- "lea (%0,%4,2), %0\n" /* pix1 += 2*line_size */
- "lea (%1,%4,2), %1\n" /* pix2 += 2*line_size */
+ "lea (%0,%4,2), %0\n" /* pix1 += 2*line_size */
+ "lea (%1,%4,2), %1\n" /* pix2 += 2*line_size */
"paddd %%xmm2,%%xmm1\n"
"paddd %%xmm4,%%xmm3\n"
@@ -913,10 +913,10 @@ static int sse16_sse2(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in
"jnz 1b\n"
"movdqa %%xmm7,%%xmm1\n"
- "psrldq $8, %%xmm7\n" /* shift hi qword to lo */
+ "psrldq $8, %%xmm7\n" /* shift hi qword to lo */
"paddd %%xmm1,%%xmm7\n"
"movdqa %%xmm7,%%xmm1\n"
- "psrldq $4, %%xmm7\n" /* shift hi dword to lo */
+ "psrldq $4, %%xmm7\n" /* shift hi dword to lo */
"paddd %%xmm1,%%xmm7\n"
"movd %%xmm7,%3\n"
: "+r" (pix1), "+r" (pix2), "+r"(h), "=r"(tmp)
@@ -1427,18 +1427,18 @@ static int vsad16_mmx2(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, i
static void diff_bytes_mmx(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w){
long i=0;
asm volatile(
- "1: \n\t"
- "movq (%2, %0), %%mm0 \n\t"
- "movq (%1, %0), %%mm1 \n\t"
- "psubb %%mm0, %%mm1 \n\t"
- "movq %%mm1, (%3, %0) \n\t"
- "movq 8(%2, %0), %%mm0 \n\t"
- "movq 8(%1, %0), %%mm1 \n\t"
- "psubb %%mm0, %%mm1 \n\t"
- "movq %%mm1, 8(%3, %0) \n\t"
- "add $16, %0 \n\t"
- "cmp %4, %0 \n\t"
- " jb 1b \n\t"
+ "1: \n\t"
+ "movq (%2, %0), %%mm0 \n\t"
+ "movq (%1, %0), %%mm1 \n\t"
+ "psubb %%mm0, %%mm1 \n\t"
+ "movq %%mm1, (%3, %0) \n\t"
+ "movq 8(%2, %0), %%mm0 \n\t"
+ "movq 8(%1, %0), %%mm1 \n\t"
+ "psubb %%mm0, %%mm1 \n\t"
+ "movq %%mm1, 8(%3, %0) \n\t"
+ "add $16, %0 \n\t"
+ "cmp %4, %0 \n\t"
+ " jb 1b \n\t"
: "+r" (i)
: "r"(src1), "r"(src2), "r"(dst), "r"((long)w-15)
);
@@ -1451,24 +1451,24 @@ static void sub_hfyu_median_prediction_mmx2(uint8_t *dst, uint8_t *src1, uint8_t
uint8_t l, lt;
asm volatile(
- "1: \n\t"
- "movq -1(%1, %0), %%mm0 \n\t" // LT
- "movq (%1, %0), %%mm1 \n\t" // T
- "movq -1(%2, %0), %%mm2 \n\t" // L
- "movq (%2, %0), %%mm3 \n\t" // X
- "movq %%mm2, %%mm4 \n\t" // L
- "psubb %%mm0, %%mm2 \n\t"
- "paddb %%mm1, %%mm2 \n\t" // L + T - LT
- "movq %%mm4, %%mm5 \n\t" // L
- "pmaxub %%mm1, %%mm4 \n\t" // max(T, L)
- "pminub %%mm5, %%mm1 \n\t" // min(T, L)
- "pminub %%mm2, %%mm4 \n\t"
- "pmaxub %%mm1, %%mm4 \n\t"
- "psubb %%mm4, %%mm3 \n\t" // dst - pred
- "movq %%mm3, (%3, %0) \n\t"
- "add $8, %0 \n\t"
- "cmp %4, %0 \n\t"
- " jb 1b \n\t"
+ "1: \n\t"
+ "movq -1(%1, %0), %%mm0 \n\t" // LT
+ "movq (%1, %0), %%mm1 \n\t" // T
+ "movq -1(%2, %0), %%mm2 \n\t" // L
+ "movq (%2, %0), %%mm3 \n\t" // X
+ "movq %%mm2, %%mm4 \n\t" // L
+ "psubb %%mm0, %%mm2 \n\t"
+ "paddb %%mm1, %%mm2 \n\t" // L + T - LT
+ "movq %%mm4, %%mm5 \n\t" // L
+ "pmaxub %%mm1, %%mm4 \n\t" // max(T, L)
+ "pminub %%mm5, %%mm1 \n\t" // min(T, L)
+ "pminub %%mm2, %%mm4 \n\t"
+ "pmaxub %%mm1, %%mm4 \n\t"
+ "psubb %%mm4, %%mm3 \n\t" // dst - pred
+ "movq %%mm3, (%3, %0) \n\t"
+ "add $8, %0 \n\t"
+ "cmp %4, %0 \n\t"
+ " jb 1b \n\t"
: "+r" (i)
: "r"(src1), "r"(src2), "r"(dst), "r"((long)w)
);
@@ -1483,12 +1483,12 @@ static void sub_hfyu_median_prediction_mmx2(uint8_t *dst, uint8_t *src1, uint8_t
}
#define LBUTTERFLY2(a1,b1,a2,b2)\
- "paddw " #b1 ", " #a1 " \n\t"\
- "paddw " #b2 ", " #a2 " \n\t"\
- "paddw " #b1 ", " #b1 " \n\t"\
- "paddw " #b2 ", " #b2 " \n\t"\
- "psubw " #a1 ", " #b1 " \n\t"\
- "psubw " #a2 ", " #b2 " \n\t"
+ "paddw " #b1 ", " #a1 " \n\t"\
+ "paddw " #b2 ", " #a2 " \n\t"\
+ "paddw " #b1 ", " #b1 " \n\t"\
+ "paddw " #b2 ", " #b2 " \n\t"\
+ "psubw " #a1 ", " #b1 " \n\t"\
+ "psubw " #a2 ", " #b2 " \n\t"
#define HADAMARD48\
LBUTTERFLY2(%%mm0, %%mm1, %%mm2, %%mm3)\
@@ -1499,33 +1499,33 @@ static void sub_hfyu_median_prediction_mmx2(uint8_t *dst, uint8_t *src1, uint8_t
LBUTTERFLY2(%%mm2, %%mm6, %%mm3, %%mm7)\
#define MMABS(a,z)\
- "pxor " #z ", " #z " \n\t"\
- "pcmpgtw " #a ", " #z " \n\t"\
- "pxor " #z ", " #a " \n\t"\
- "psubw " #z ", " #a " \n\t"
+ "pxor " #z ", " #z " \n\t"\
+ "pcmpgtw " #a ", " #z " \n\t"\
+ "pxor " #z ", " #a " \n\t"\
+ "psubw " #z ", " #a " \n\t"
#define MMABS_SUM(a,z, sum)\
- "pxor " #z ", " #z " \n\t"\
- "pcmpgtw " #a ", " #z " \n\t"\
- "pxor " #z ", " #a " \n\t"\
- "psubw " #z ", " #a " \n\t"\
- "paddusw " #a ", " #sum " \n\t"
+ "pxor " #z ", " #z " \n\t"\
+ "pcmpgtw " #a ", " #z " \n\t"\
+ "pxor " #z ", " #a " \n\t"\
+ "psubw " #z ", " #a " \n\t"\
+ "paddusw " #a ", " #sum " \n\t"
#define MMABS_MMX2(a,z)\
- "pxor " #z ", " #z " \n\t"\
- "psubw " #a ", " #z " \n\t"\
- "pmaxsw " #z ", " #a " \n\t"
+ "pxor " #z ", " #z " \n\t"\
+ "psubw " #a ", " #z " \n\t"\
+ "pmaxsw " #z ", " #a " \n\t"
#define MMABS_SUM_MMX2(a,z, sum)\
- "pxor " #z ", " #z " \n\t"\
- "psubw " #a ", " #z " \n\t"\
- "pmaxsw " #z ", " #a " \n\t"\
- "paddusw " #a ", " #sum " \n\t"
+ "pxor " #z ", " #z " \n\t"\
+ "psubw " #a ", " #z " \n\t"\
+ "pmaxsw " #z ", " #a " \n\t"\
+ "paddusw " #a ", " #sum " \n\t"
#define SBUTTERFLY(a,b,t,n)\
- "movq " #a ", " #t " \n\t" /* abcd */\
- "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\
- "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\
+ "movq " #a ", " #t " \n\t" /* abcd */\
+ "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\
+ "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\
#define TRANSPOSE4(a,b,c,d,t)\
SBUTTERFLY(a,b,t,wd) /* a=aebf t=cgdh */\
@@ -1534,16 +1534,16 @@ static void sub_hfyu_median_prediction_mmx2(uint8_t *dst, uint8_t *src1, uint8_t
SBUTTERFLY(t,b,c,dq) /* t=cgko c=dhlp */
#define LOAD4(o, a, b, c, d)\
- "movq "#o"(%1), " #a " \n\t"\
- "movq "#o"+16(%1), " #b " \n\t"\
- "movq "#o"+32(%1), " #c " \n\t"\
- "movq "#o"+48(%1), " #d " \n\t"
+ "movq "#o"(%1), " #a " \n\t"\
+ "movq "#o"+16(%1), " #b " \n\t"\
+ "movq "#o"+32(%1), " #c " \n\t"\
+ "movq "#o"+48(%1), " #d " \n\t"
#define STORE4(o, a, b, c, d)\
- "movq "#a", "#o"(%1) \n\t"\
- "movq "#b", "#o"+16(%1) \n\t"\
- "movq "#c", "#o"+32(%1) \n\t"\
- "movq "#d", "#o"+48(%1) \n\t"\
+ "movq "#a", "#o"(%1) \n\t"\
+ "movq "#b", "#o"+16(%1) \n\t"\
+ "movq "#c", "#o"+32(%1) \n\t"\
+ "movq "#d", "#o"+48(%1) \n\t"\
static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride, int h){
uint64_t temp[16] __align8;
@@ -1559,12 +1559,12 @@ static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride,
HADAMARD48
- "movq %%mm7, 112(%1) \n\t"
+ "movq %%mm7, 112(%1) \n\t"
TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7)
STORE4(0 , %%mm0, %%mm3, %%mm7, %%mm2)
- "movq 112(%1), %%mm7 \n\t"
+ "movq 112(%1), %%mm7 \n\t"
TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0)
STORE4(64, %%mm4, %%mm7, %%mm0, %%mm6)
@@ -1573,23 +1573,23 @@ static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride,
HADAMARD48
- "movq %%mm7, 120(%1) \n\t"
+ "movq %%mm7, 120(%1) \n\t"
TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7)
STORE4(8 , %%mm0, %%mm3, %%mm7, %%mm2)
- "movq 120(%1), %%mm7 \n\t"
+ "movq 120(%1), %%mm7 \n\t"
TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0)
- "movq %%mm7, %%mm5 \n\t"//FIXME remove
- "movq %%mm6, %%mm7 \n\t"
- "movq %%mm0, %%mm6 \n\t"
+ "movq %%mm7, %%mm5 \n\t"//FIXME remove
+ "movq %%mm6, %%mm7 \n\t"
+ "movq %%mm0, %%mm6 \n\t"
// STORE4(72, %%mm4, %%mm7, %%mm0, %%mm6) //FIXME remove
LOAD4(64, %%mm0, %%mm1, %%mm2, %%mm3)
// LOAD4(72, %%mm4, %%mm5, %%mm6, %%mm7)
HADAMARD48
- "movq %%mm7, 64(%1) \n\t"
+ "movq %%mm7, 64(%1) \n\t"
MMABS(%%mm0, %%mm7)
MMABS_SUM(%%mm1, %%mm7, %%mm0)
MMABS_SUM(%%mm2, %%mm7, %%mm0)
@@ -1597,15 +1597,15 @@ static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride,
MMABS_SUM(%%mm4, %%mm7, %%mm0)
MMABS_SUM(%%mm5, %%mm7, %%mm0)
MMABS_SUM(%%mm6, %%mm7, %%mm0)
- "movq 64(%1), %%mm1 \n\t"
+ "movq 64(%1), %%mm1 \n\t"
MMABS_SUM(%%mm1, %%mm7, %%mm0)
- "movq %%mm0, 64(%1) \n\t"
+ "movq %%mm0, 64(%1) \n\t"
LOAD4(0 , %%mm0, %%mm1, %%mm2, %%mm3)
LOAD4(8 , %%mm4, %%mm5, %%mm6, %%mm7)
HADAMARD48
- "movq %%mm7, (%1) \n\t"
+ "movq %%mm7, (%1) \n\t"
MMABS(%%mm0, %%mm7)
MMABS_SUM(%%mm1, %%mm7, %%mm0)
MMABS_SUM(%%mm2, %%mm7, %%mm0)
@@ -1613,18 +1613,18 @@ static int hadamard8_diff_mmx(void *s, uint8_t *src1, uint8_t *src2, int stride,
MMABS_SUM(%%mm4, %%mm7, %%mm0)
MMABS_SUM(%%mm5, %%mm7, %%mm0)
MMABS_SUM(%%mm6, %%mm7, %%mm0)
- "movq (%1), %%mm1 \n\t"
+ "movq (%1), %%mm1 \n\t"
MMABS_SUM(%%mm1, %%mm7, %%mm0)
- "movq 64(%1), %%mm1 \n\t"
+ "movq 64(%1), %%mm1 \n\t"
MMABS_SUM(%%mm1, %%mm7, %%mm0)
- "movq %%mm0, %%mm1 \n\t"
- "psrlq $32, %%mm0 \n\t"
- "paddusw %%mm1, %%mm0 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "psrlq $16, %%mm0 \n\t"
- "paddusw %%mm1, %%mm0 \n\t"
- "movd %%mm0, %0 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "psrlq $32, %%mm0 \n\t"
+ "paddusw %%mm1, %%mm0 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "psrlq $16, %%mm0 \n\t"
+ "paddusw %%mm1, %%mm0 \n\t"
+ "movd %%mm0, %0 \n\t"
: "=r" (sum)
: "r"(temp)
@@ -1646,12 +1646,12 @@ static int hadamard8_diff_mmx2(void *s, uint8_t *src1, uint8_t *src2, int stride
HADAMARD48
- "movq %%mm7, 112(%1) \n\t"
+ "movq %%mm7, 112(%1) \n\t"
TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7)
STORE4(0 , %%mm0, %%mm3, %%mm7, %%mm2)
- "movq 112(%1), %%mm7 \n\t"
+ "movq 112(%1), %%mm7 \n\t"
TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0)
STORE4(64, %%mm4, %%mm7, %%mm0, %%mm6)
@@ -1660,23 +1660,23 @@ static int hadamard8_diff_mmx2(void *s, uint8_t *src1, uint8_t *src2, int stride
HADAMARD48
- "movq %%mm7, 120(%1) \n\t"
+ "movq %%mm7, 120(%1) \n\t"
TRANSPOSE4(%%mm0, %%mm1, %%mm2, %%mm3, %%mm7)
STORE4(8 , %%mm0, %%mm3, %%mm7, %%mm2)
- "movq 120(%1), %%mm7 \n\t"
+ "movq 120(%1), %%mm7 \n\t"
TRANSPOSE4(%%mm4, %%mm5, %%mm6, %%mm7, %%mm0)
- "movq %%mm7, %%mm5 \n\t"//FIXME remove
- "movq %%mm6, %%mm7 \n\t"
- "movq %%mm0, %%mm6 \n\t"
+ "movq %%mm7, %%mm5 \n\t"//FIXME remove
+ "movq %%mm6, %%mm7 \n\t"
+ "movq %%mm0, %%mm6 \n\t"
// STORE4(72, %%mm4, %%mm7, %%mm0, %%mm6) //FIXME remove
LOAD4(64, %%mm0, %%mm1, %%mm2, %%mm3)
// LOAD4(72, %%mm4, %%mm5, %%mm6, %%mm7)
HADAMARD48
- "movq %%mm7, 64(%1) \n\t"
+ "movq %%mm7, 64(%1) \n\t"
MMABS_MMX2(%%mm0, %%mm7)
MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm2, %%mm7, %%mm0)
@@ -1684,15 +1684,15 @@ static int hadamard8_diff_mmx2(void *s, uint8_t *src1, uint8_t *src2, int stride
MMABS_SUM_MMX2(%%mm4, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm5, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm6, %%mm7, %%mm0)
- "movq 64(%1), %%mm1 \n\t"
+ "movq 64(%1), %%mm1 \n\t"
MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0)
- "movq %%mm0, 64(%1) \n\t"
+ "movq %%mm0, 64(%1) \n\t"
LOAD4(0 , %%mm0, %%mm1, %%mm2, %%mm3)
LOAD4(8 , %%mm4, %%mm5, %%mm6, %%mm7)
HADAMARD48
- "movq %%mm7, (%1) \n\t"
+ "movq %%mm7, (%1) \n\t"
MMABS_MMX2(%%mm0, %%mm7)
MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm2, %%mm7, %%mm0)
@@ -1700,16 +1700,16 @@ static int hadamard8_diff_mmx2(void *s, uint8_t *src1, uint8_t *src2, int stride
MMABS_SUM_MMX2(%%mm4, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm5, %%mm7, %%mm0)
MMABS_SUM_MMX2(%%mm6, %%mm7, %%mm0)
- "movq (%1), %%mm1 \n\t"
+ "movq (%1), %%mm1 \n\t"
MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0)
- "movq 64(%1), %%mm1 \n\t"
+ "movq 64(%1), %%mm1 \n\t"
MMABS_SUM_MMX2(%%mm1, %%mm7, %%mm0)
"pshufw $0x0E, %%mm0, %%mm1 \n\t"
- "paddusw %%mm1, %%mm0 \n\t"
+ "paddusw %%mm1, %%mm0 \n\t"
"pshufw $0x01, %%mm0, %%mm1 \n\t"
- "paddusw %%mm1, %%mm0 \n\t"
- "movd %%mm0, %0 \n\t"
+ "paddusw %%mm1, %%mm0 \n\t"
+ "movd %%mm0, %0 \n\t"
: "=r" (sum)
: "r"(temp)
@@ -1726,24 +1726,24 @@ WARPER8_16_SQ(hadamard8_diff_mmx2, hadamard8_diff16_mmx2)
#define put_no_rnd_pixels16_mmx(a,b,c,d) put_pixels16_mmx(a,b,c,d)
#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 */\
- "movq "MANGLE(ff_pw_20)", %%mm4 \n\t" /* 20 */\
- "pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\
- "movq "#in7", " #m3 " \n\t" /* d */\
- "movq "#in0", %%mm5 \n\t" /* D */\
- "paddw " #m3 ", %%mm5 \n\t" /* x4 */\
- "psubw %%mm5, %%mm4 \n\t" /* 20x1 - x4 */\
- "movq "#in1", %%mm5 \n\t" /* C */\
- "movq "#in2", %%mm6 \n\t" /* B */\
- "paddw " #m6 ", %%mm5 \n\t" /* x3 */\
- "paddw " #m5 ", %%mm6 \n\t" /* x2 */\
- "paddw %%mm6, %%mm6 \n\t" /* 2x2 */\
- "psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\
- "pmullw "MANGLE(ff_pw_3)", %%mm5 \n\t" /* -6x2 + 3x3 */\
- "paddw " #rnd ", %%mm4 \n\t" /* x2 */\
- "paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\
- "psraw $5, %%mm5 \n\t"\
- "packuswb %%mm5, %%mm5 \n\t"\
+ "paddw " #m4 ", " #m3 " \n\t" /* x1 */\
+ "movq "MANGLE(ff_pw_20)", %%mm4 \n\t" /* 20 */\
+ "pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\
+ "movq "#in7", " #m3 " \n\t" /* d */\
+ "movq "#in0", %%mm5 \n\t" /* D */\
+ "paddw " #m3 ", %%mm5 \n\t" /* x4 */\
+ "psubw %%mm5, %%mm4 \n\t" /* 20x1 - x4 */\
+ "movq "#in1", %%mm5 \n\t" /* C */\
+ "movq "#in2", %%mm6 \n\t" /* B */\
+ "paddw " #m6 ", %%mm5 \n\t" /* x3 */\
+ "paddw " #m5 ", %%mm6 \n\t" /* x2 */\
+ "paddw %%mm6, %%mm6 \n\t" /* 2x2 */\
+ "psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm5 \n\t" /* -6x2 + 3x3 */\
+ "paddw " #rnd ", %%mm4 \n\t" /* x2 */\
+ "paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\
+ "psraw $5, %%mm5 \n\t"\
+ "packuswb %%mm5, %%mm5 \n\t"\
OP(%%mm5, out, %%mm7, d)
#define QPEL_BASE(OPNAME, ROUNDER, RND, OP_MMX2, OP_3DNOW)\
@@ -1751,116 +1751,116 @@ static void OPNAME ## mpeg4_qpel16_h_lowpass_mmx2(uint8_t *dst, uint8_t *src, in
uint64_t temp;\
\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\
- "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\
- "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\
- "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\
- "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\
- "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\
- "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\
- "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\
- "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\
- "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\
- "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\
- "psllq $24, %%mm4 \n\t" /* 000ABCDE */\
- "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\
- "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\
- "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\
- "paddw %%mm3, %%mm5 \n\t" /* b */\
- "paddw %%mm2, %%mm6 \n\t" /* c */\
- "paddw %%mm5, %%mm5 \n\t" /* 2b */\
- "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
- "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
- "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\
- "paddw %%mm4, %%mm0 \n\t" /* a */\
- "paddw %%mm1, %%mm5 \n\t" /* d */\
- "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\
- "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
- "paddw %6, %%mm6 \n\t"\
- "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
- "psraw $5, %%mm0 \n\t"\
- "movq %%mm0, %5 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\
+ "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\
+ "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\
+ "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\
+ "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\
+ "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\
+ "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\
+ "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\
+ "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\
+ "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\
+ "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\
+ "psllq $24, %%mm4 \n\t" /* 000ABCDE */\
+ "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\
+ "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\
+ "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\
+ "paddw %%mm3, %%mm5 \n\t" /* b */\
+ "paddw %%mm2, %%mm6 \n\t" /* c */\
+ "paddw %%mm5, %%mm5 \n\t" /* 2b */\
+ "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
+ "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\
+ "paddw %%mm4, %%mm0 \n\t" /* a */\
+ "paddw %%mm1, %%mm5 \n\t" /* d */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\
+ "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
+ "paddw %6, %%mm6 \n\t"\
+ "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
+ "psraw $5, %%mm0 \n\t"\
+ "movq %%mm0, %5 \n\t"\
/* mm1=EFGH, mm2=DEFG, mm3=CDEF, mm4=BCDE, mm7=0 */\
\
- "movq 5(%0), %%mm0 \n\t" /* FGHIJKLM */\
- "movq %%mm0, %%mm5 \n\t" /* FGHIJKLM */\
- "movq %%mm0, %%mm6 \n\t" /* FGHIJKLM */\
- "psrlq $8, %%mm0 \n\t" /* GHIJKLM0 */\
- "psrlq $16, %%mm5 \n\t" /* HIJKLM00 */\
- "punpcklbw %%mm7, %%mm0 \n\t" /* 0G0H0I0J */\
- "punpcklbw %%mm7, %%mm5 \n\t" /* 0H0I0J0K */\
- "paddw %%mm0, %%mm2 \n\t" /* b */\
- "paddw %%mm5, %%mm3 \n\t" /* c */\
- "paddw %%mm2, %%mm2 \n\t" /* 2b */\
- "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\
- "movq %%mm6, %%mm2 \n\t" /* FGHIJKLM */\
- "psrlq $24, %%mm6 \n\t" /* IJKLM000 */\
- "punpcklbw %%mm7, %%mm2 \n\t" /* 0F0G0H0I */\
- "punpcklbw %%mm7, %%mm6 \n\t" /* 0I0J0K0L */\
- "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\
- "paddw %%mm2, %%mm1 \n\t" /* a */\
- "paddw %%mm6, %%mm4 \n\t" /* d */\
- "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\
- "psubw %%mm4, %%mm3 \n\t" /* - 6b +3c - d */\
- "paddw %6, %%mm1 \n\t"\
- "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b +3c - d */\
- "psraw $5, %%mm3 \n\t"\
- "movq %5, %%mm1 \n\t"\
- "packuswb %%mm3, %%mm1 \n\t"\
+ "movq 5(%0), %%mm0 \n\t" /* FGHIJKLM */\
+ "movq %%mm0, %%mm5 \n\t" /* FGHIJKLM */\
+ "movq %%mm0, %%mm6 \n\t" /* FGHIJKLM */\
+ "psrlq $8, %%mm0 \n\t" /* GHIJKLM0 */\
+ "psrlq $16, %%mm5 \n\t" /* HIJKLM00 */\
+ "punpcklbw %%mm7, %%mm0 \n\t" /* 0G0H0I0J */\
+ "punpcklbw %%mm7, %%mm5 \n\t" /* 0H0I0J0K */\
+ "paddw %%mm0, %%mm2 \n\t" /* b */\
+ "paddw %%mm5, %%mm3 \n\t" /* c */\
+ "paddw %%mm2, %%mm2 \n\t" /* 2b */\
+ "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\
+ "movq %%mm6, %%mm2 \n\t" /* FGHIJKLM */\
+ "psrlq $24, %%mm6 \n\t" /* IJKLM000 */\
+ "punpcklbw %%mm7, %%mm2 \n\t" /* 0F0G0H0I */\
+ "punpcklbw %%mm7, %%mm6 \n\t" /* 0I0J0K0L */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\
+ "paddw %%mm2, %%mm1 \n\t" /* a */\
+ "paddw %%mm6, %%mm4 \n\t" /* d */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\
+ "psubw %%mm4, %%mm3 \n\t" /* - 6b +3c - d */\
+ "paddw %6, %%mm1 \n\t"\
+ "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b +3c - d */\
+ "psraw $5, %%mm3 \n\t"\
+ "movq %5, %%mm1 \n\t"\
+ "packuswb %%mm3, %%mm1 \n\t"\
OP_MMX2(%%mm1, (%1),%%mm4, q)\
/* mm0= GHIJ, mm2=FGHI, mm5=HIJK, mm6=IJKL, mm7=0 */\
\
- "movq 9(%0), %%mm1 \n\t" /* JKLMNOPQ */\
- "movq %%mm1, %%mm4 \n\t" /* JKLMNOPQ */\
- "movq %%mm1, %%mm3 \n\t" /* JKLMNOPQ */\
- "psrlq $8, %%mm1 \n\t" /* KLMNOPQ0 */\
- "psrlq $16, %%mm4 \n\t" /* LMNOPQ00 */\
- "punpcklbw %%mm7, %%mm1 \n\t" /* 0K0L0M0N */\
- "punpcklbw %%mm7, %%mm4 \n\t" /* 0L0M0N0O */\
- "paddw %%mm1, %%mm5 \n\t" /* b */\
- "paddw %%mm4, %%mm0 \n\t" /* c */\
- "paddw %%mm5, %%mm5 \n\t" /* 2b */\
- "psubw %%mm5, %%mm0 \n\t" /* c - 2b */\
- "movq %%mm3, %%mm5 \n\t" /* JKLMNOPQ */\
- "psrlq $24, %%mm3 \n\t" /* MNOPQ000 */\
- "pmullw "MANGLE(ff_pw_3)", %%mm0 \n\t" /* 3c - 6b */\
- "punpcklbw %%mm7, %%mm3 \n\t" /* 0M0N0O0P */\
- "paddw %%mm3, %%mm2 \n\t" /* d */\
- "psubw %%mm2, %%mm0 \n\t" /* -6b + 3c - d */\
- "movq %%mm5, %%mm2 \n\t" /* JKLMNOPQ */\
- "punpcklbw %%mm7, %%mm2 \n\t" /* 0J0K0L0M */\
- "punpckhbw %%mm7, %%mm5 \n\t" /* 0N0O0P0Q */\
- "paddw %%mm2, %%mm6 \n\t" /* a */\
- "pmullw "MANGLE(ff_pw_20)", %%mm6 \n\t" /* 20a */\
- "paddw %6, %%mm0 \n\t"\
- "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
- "psraw $5, %%mm0 \n\t"\
+ "movq 9(%0), %%mm1 \n\t" /* JKLMNOPQ */\
+ "movq %%mm1, %%mm4 \n\t" /* JKLMNOPQ */\
+ "movq %%mm1, %%mm3 \n\t" /* JKLMNOPQ */\
+ "psrlq $8, %%mm1 \n\t" /* KLMNOPQ0 */\
+ "psrlq $16, %%mm4 \n\t" /* LMNOPQ00 */\
+ "punpcklbw %%mm7, %%mm1 \n\t" /* 0K0L0M0N */\
+ "punpcklbw %%mm7, %%mm4 \n\t" /* 0L0M0N0O */\
+ "paddw %%mm1, %%mm5 \n\t" /* b */\
+ "paddw %%mm4, %%mm0 \n\t" /* c */\
+ "paddw %%mm5, %%mm5 \n\t" /* 2b */\
+ "psubw %%mm5, %%mm0 \n\t" /* c - 2b */\
+ "movq %%mm3, %%mm5 \n\t" /* JKLMNOPQ */\
+ "psrlq $24, %%mm3 \n\t" /* MNOPQ000 */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm0 \n\t" /* 3c - 6b */\
+ "punpcklbw %%mm7, %%mm3 \n\t" /* 0M0N0O0P */\
+ "paddw %%mm3, %%mm2 \n\t" /* d */\
+ "psubw %%mm2, %%mm0 \n\t" /* -6b + 3c - d */\
+ "movq %%mm5, %%mm2 \n\t" /* JKLMNOPQ */\
+ "punpcklbw %%mm7, %%mm2 \n\t" /* 0J0K0L0M */\
+ "punpckhbw %%mm7, %%mm5 \n\t" /* 0N0O0P0Q */\
+ "paddw %%mm2, %%mm6 \n\t" /* a */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm6 \n\t" /* 20a */\
+ "paddw %6, %%mm0 \n\t"\
+ "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
+ "psraw $5, %%mm0 \n\t"\
/* mm1=KLMN, mm2=JKLM, mm3=MNOP, mm4=LMNO, mm5=NOPQ mm7=0 */\
\
- "paddw %%mm5, %%mm3 \n\t" /* a */\
- "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0O0P0Q0Q */\
- "paddw %%mm4, %%mm6 \n\t" /* b */\
- "pshufw $0xBE, %%mm5, %%mm4 \n\t" /* 0P0Q0Q0P */\
- "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0Q0Q0P0O */\
- "paddw %%mm1, %%mm4 \n\t" /* c */\
- "paddw %%mm2, %%mm5 \n\t" /* d */\
- "paddw %%mm6, %%mm6 \n\t" /* 2b */\
- "psubw %%mm6, %%mm4 \n\t" /* c - 2b */\
- "pmullw "MANGLE(ff_pw_20)", %%mm3 \n\t" /* 20a */\
- "pmullw "MANGLE(ff_pw_3)", %%mm4 \n\t" /* 3c - 6b */\
- "psubw %%mm5, %%mm3 \n\t" /* -6b + 3c - d */\
- "paddw %6, %%mm4 \n\t"\
- "paddw %%mm3, %%mm4 \n\t" /* 20a - 6b + 3c - d */\
- "psraw $5, %%mm4 \n\t"\
- "packuswb %%mm4, %%mm0 \n\t"\
+ "paddw %%mm5, %%mm3 \n\t" /* a */\
+ "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0O0P0Q0Q */\
+ "paddw %%mm4, %%mm6 \n\t" /* b */\
+ "pshufw $0xBE, %%mm5, %%mm4 \n\t" /* 0P0Q0Q0P */\
+ "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0Q0Q0P0O */\
+ "paddw %%mm1, %%mm4 \n\t" /* c */\
+ "paddw %%mm2, %%mm5 \n\t" /* d */\
+ "paddw %%mm6, %%mm6 \n\t" /* 2b */\
+ "psubw %%mm6, %%mm4 \n\t" /* c - 2b */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm3 \n\t" /* 20a */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm4 \n\t" /* 3c - 6b */\
+ "psubw %%mm5, %%mm3 \n\t" /* -6b + 3c - d */\
+ "paddw %6, %%mm4 \n\t"\
+ "paddw %%mm3, %%mm4 \n\t" /* 20a - 6b + 3c - d */\
+ "psraw $5, %%mm4 \n\t"\
+ "packuswb %%mm4, %%mm0 \n\t"\
OP_MMX2(%%mm0, 8(%1), %%mm4, q)\
\
- "add %3, %0 \n\t"\
- "add %4, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add %3, %0 \n\t"\
+ "add %4, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
: "d"((long)srcStride), "S"((long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(temp), "m"(ROUNDER)\
: "memory"\
@@ -1890,21 +1890,21 @@ static void OPNAME ## mpeg4_qpel16_h_lowpass_3dnow(uint8_t *dst, uint8_t *src, i
temp[14]= (src[14]+src[15])*20 - (src[13]+src[16])*6 + (src[12]+src[16])*3 - (src[11]+src[15]);\
temp[15]= (src[15]+src[16])*20 - (src[14]+src[16])*6 + (src[13]+src[15])*3 - (src[12]+src[14]);\
asm volatile(\
- "movq (%0), %%mm0 \n\t"\
- "movq 8(%0), %%mm1 \n\t"\
- "paddw %2, %%mm0 \n\t"\
- "paddw %2, %%mm1 \n\t"\
- "psraw $5, %%mm0 \n\t"\
- "psraw $5, %%mm1 \n\t"\
- "packuswb %%mm1, %%mm0 \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 8(%0), %%mm1 \n\t"\
+ "paddw %2, %%mm0 \n\t"\
+ "paddw %2, %%mm1 \n\t"\
+ "psraw $5, %%mm0 \n\t"\
+ "psraw $5, %%mm1 \n\t"\
+ "packuswb %%mm1, %%mm0 \n\t"\
OP_3DNOW(%%mm0, (%1), %%mm1, q)\
- "movq 16(%0), %%mm0 \n\t"\
- "movq 24(%0), %%mm1 \n\t"\
- "paddw %2, %%mm0 \n\t"\
- "paddw %2, %%mm1 \n\t"\
- "psraw $5, %%mm0 \n\t"\
- "psraw $5, %%mm1 \n\t"\
- "packuswb %%mm1, %%mm0 \n\t"\
+ "movq 16(%0), %%mm0 \n\t"\
+ "movq 24(%0), %%mm1 \n\t"\
+ "paddw %2, %%mm0 \n\t"\
+ "paddw %2, %%mm1 \n\t"\
+ "psraw $5, %%mm0 \n\t"\
+ "psraw $5, %%mm1 \n\t"\
+ "packuswb %%mm1, %%mm0 \n\t"\
OP_3DNOW(%%mm0, 8(%1), %%mm1, q)\
:: "r"(temp), "r"(dst), "m"(ROUNDER)\
: "memory"\
@@ -1918,62 +1918,62 @@ static void OPNAME ## mpeg4_qpel8_h_lowpass_mmx2(uint8_t *dst, uint8_t *src, int
uint64_t temp;\
\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\
- "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\
- "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\
- "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\
- "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\
- "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\
- "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\
- "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\
- "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\
- "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\
- "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\
- "psllq $24, %%mm4 \n\t" /* 000ABCDE */\
- "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\
- "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\
- "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\
- "paddw %%mm3, %%mm5 \n\t" /* b */\
- "paddw %%mm2, %%mm6 \n\t" /* c */\
- "paddw %%mm5, %%mm5 \n\t" /* 2b */\
- "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
- "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
- "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\
- "paddw %%mm4, %%mm0 \n\t" /* a */\
- "paddw %%mm1, %%mm5 \n\t" /* d */\
- "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\
- "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
- "paddw %6, %%mm6 \n\t"\
- "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
- "psraw $5, %%mm0 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t" /* ABCDEFGH */\
+ "movq %%mm0, %%mm1 \n\t" /* ABCDEFGH */\
+ "movq %%mm0, %%mm2 \n\t" /* ABCDEFGH */\
+ "punpcklbw %%mm7, %%mm0 \n\t" /* 0A0B0C0D */\
+ "punpckhbw %%mm7, %%mm1 \n\t" /* 0E0F0G0H */\
+ "pshufw $0x90, %%mm0, %%mm5 \n\t" /* 0A0A0B0C */\
+ "pshufw $0x41, %%mm0, %%mm6 \n\t" /* 0B0A0A0B */\
+ "movq %%mm2, %%mm3 \n\t" /* ABCDEFGH */\
+ "movq %%mm2, %%mm4 \n\t" /* ABCDEFGH */\
+ "psllq $8, %%mm2 \n\t" /* 0ABCDEFG */\
+ "psllq $16, %%mm3 \n\t" /* 00ABCDEF */\
+ "psllq $24, %%mm4 \n\t" /* 000ABCDE */\
+ "punpckhbw %%mm7, %%mm2 \n\t" /* 0D0E0F0G */\
+ "punpckhbw %%mm7, %%mm3 \n\t" /* 0C0D0E0F */\
+ "punpckhbw %%mm7, %%mm4 \n\t" /* 0B0C0D0E */\
+ "paddw %%mm3, %%mm5 \n\t" /* b */\
+ "paddw %%mm2, %%mm6 \n\t" /* c */\
+ "paddw %%mm5, %%mm5 \n\t" /* 2b */\
+ "psubw %%mm5, %%mm6 \n\t" /* c - 2b */\
+ "pshufw $0x06, %%mm0, %%mm5 \n\t" /* 0C0B0A0A */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm6 \n\t" /* 3c - 6b */\
+ "paddw %%mm4, %%mm0 \n\t" /* a */\
+ "paddw %%mm1, %%mm5 \n\t" /* d */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm0 \n\t" /* 20a */\
+ "psubw %%mm5, %%mm0 \n\t" /* 20a - d */\
+ "paddw %6, %%mm6 \n\t"\
+ "paddw %%mm6, %%mm0 \n\t" /* 20a - 6b + 3c - d */\
+ "psraw $5, %%mm0 \n\t"\
/* mm1=EFGH, mm2=DEFG, mm3=CDEF, mm4=BCDE, mm7=0 */\
\
- "movd 5(%0), %%mm5 \n\t" /* FGHI */\
- "punpcklbw %%mm7, %%mm5 \n\t" /* 0F0G0H0I */\
- "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0G0H0I0I */\
- "paddw %%mm5, %%mm1 \n\t" /* a */\
- "paddw %%mm6, %%mm2 \n\t" /* b */\
- "pshufw $0xBE, %%mm5, %%mm6 \n\t" /* 0H0I0I0H */\
- "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0I0I0H0G */\
- "paddw %%mm6, %%mm3 \n\t" /* c */\
- "paddw %%mm5, %%mm4 \n\t" /* d */\
- "paddw %%mm2, %%mm2 \n\t" /* 2b */\
- "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\
- "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\
- "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\
- "psubw %%mm4, %%mm3 \n\t" /* -6b + 3c - d */\
- "paddw %6, %%mm1 \n\t"\
- "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b + 3c - d */\
- "psraw $5, %%mm3 \n\t"\
- "packuswb %%mm3, %%mm0 \n\t"\
+ "movd 5(%0), %%mm5 \n\t" /* FGHI */\
+ "punpcklbw %%mm7, %%mm5 \n\t" /* 0F0G0H0I */\
+ "pshufw $0xF9, %%mm5, %%mm6 \n\t" /* 0G0H0I0I */\
+ "paddw %%mm5, %%mm1 \n\t" /* a */\
+ "paddw %%mm6, %%mm2 \n\t" /* b */\
+ "pshufw $0xBE, %%mm5, %%mm6 \n\t" /* 0H0I0I0H */\
+ "pshufw $0x6F, %%mm5, %%mm5 \n\t" /* 0I0I0H0G */\
+ "paddw %%mm6, %%mm3 \n\t" /* c */\
+ "paddw %%mm5, %%mm4 \n\t" /* d */\
+ "paddw %%mm2, %%mm2 \n\t" /* 2b */\
+ "psubw %%mm2, %%mm3 \n\t" /* c - 2b */\
+ "pmullw "MANGLE(ff_pw_20)", %%mm1 \n\t" /* 20a */\
+ "pmullw "MANGLE(ff_pw_3)", %%mm3 \n\t" /* 3c - 6b */\
+ "psubw %%mm4, %%mm3 \n\t" /* -6b + 3c - d */\
+ "paddw %6, %%mm1 \n\t"\
+ "paddw %%mm1, %%mm3 \n\t" /* 20a - 6b + 3c - d */\
+ "psraw $5, %%mm3 \n\t"\
+ "packuswb %%mm3, %%mm0 \n\t"\
OP_MMX2(%%mm0, (%1), %%mm4, q)\
\
- "add %3, %0 \n\t"\
- "add %4, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add %3, %0 \n\t"\
+ "add %4, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
: "S"((long)srcStride), "D"((long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(temp), "m"(ROUNDER)\
: "memory"\
@@ -1995,13 +1995,13 @@ static void OPNAME ## mpeg4_qpel8_h_lowpass_3dnow(uint8_t *dst, uint8_t *src, in
temp[ 6]= (src[ 6]+src[ 7])*20 - (src[ 5]+src[ 8])*6 + (src[ 4]+src[ 8])*3 - (src[ 3]+src[ 7]);\
temp[ 7]= (src[ 7]+src[ 8])*20 - (src[ 6]+src[ 8])*6 + (src[ 5]+src[ 7])*3 - (src[ 4]+src[ 6]);\
asm volatile(\
- "movq (%0), %%mm0 \n\t"\
- "movq 8(%0), %%mm1 \n\t"\
- "paddw %2, %%mm0 \n\t"\
- "paddw %2, %%mm1 \n\t"\
- "psraw $5, %%mm0 \n\t"\
- "psraw $5, %%mm1 \n\t"\
- "packuswb %%mm1, %%mm0 \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 8(%0), %%mm1 \n\t"\
+ "paddw %2, %%mm0 \n\t"\
+ "paddw %2, %%mm1 \n\t"\
+ "psraw $5, %%mm0 \n\t"\
+ "psraw $5, %%mm1 \n\t"\
+ "packuswb %%mm1, %%mm0 \n\t"\
OP_3DNOW(%%mm0, (%1), %%mm1, q)\
:: "r"(temp), "r"(dst), "m"(ROUNDER)\
:"memory"\
@@ -2020,24 +2020,24 @@ static void OPNAME ## mpeg4_qpel16_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src,
\
/*FIXME unroll */\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq (%0), %%mm1 \n\t"\
- "movq 8(%0), %%mm2 \n\t"\
- "movq 8(%0), %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpckhbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpckhbw %%mm7, %%mm3 \n\t"\
- "movq %%mm0, (%1) \n\t"\
- "movq %%mm1, 17*8(%1) \n\t"\
- "movq %%mm2, 2*17*8(%1) \n\t"\
- "movq %%mm3, 3*17*8(%1) \n\t"\
- "add $8, %1 \n\t"\
- "add %3, %0 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq (%0), %%mm1 \n\t"\
+ "movq 8(%0), %%mm2 \n\t"\
+ "movq 8(%0), %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpckhbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpckhbw %%mm7, %%mm3 \n\t"\
+ "movq %%mm0, (%1) \n\t"\
+ "movq %%mm1, 17*8(%1) \n\t"\
+ "movq %%mm2, 2*17*8(%1) \n\t"\
+ "movq %%mm3, 3*17*8(%1) \n\t"\
+ "add $8, %1 \n\t"\
+ "add %3, %0 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+r" (src), "+r" (temp_ptr), "+r"(count)\
: "r" ((long)srcStride)\
: "memory"\
@@ -2048,42 +2048,42 @@ static void OPNAME ## mpeg4_qpel16_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src,
\
/*FIXME reorder for speed */\
asm volatile(\
- /*"pxor %%mm7, %%mm7 \n\t"*/\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq 8(%0), %%mm1 \n\t"\
- "movq 16(%0), %%mm2 \n\t"\
- "movq 24(%0), %%mm3 \n\t"\
+ /*"pxor %%mm7, %%mm7 \n\t"*/\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 8(%0), %%mm1 \n\t"\
+ "movq 16(%0), %%mm2 \n\t"\
+ "movq 24(%0), %%mm3 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 16(%0), 8(%0), (%0), 32(%0), (%1), OP)\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 8(%0), (%0), (%0), 40(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, (%0), (%0), 8(%0), 48(%0), (%1), OP)\
\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, (%0), 8(%0), 16(%0), 56(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 8(%0), 16(%0), 24(%0), 64(%0), (%1), OP)\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 16(%0), 24(%0), 32(%0), 72(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 24(%0), 32(%0), 40(%0), 80(%0), (%1), OP)\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 32(%0), 40(%0), 48(%0), 88(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 40(%0), 48(%0), 56(%0), 96(%0), (%1), OP)\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 48(%0), 56(%0), 64(%0),104(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 56(%0), 64(%0), 72(%0),112(%0), (%1), OP)\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 64(%0), 72(%0), 80(%0),120(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 72(%0), 80(%0), 88(%0),128(%0), (%1), OP)\
\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 80(%0), 88(%0), 96(%0),128(%0), (%1, %3), OP)\
- "add %4, %1 \n\t" \
+ "add %4, %1 \n\t" \
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 88(%0), 96(%0),104(%0),120(%0), (%1), OP)\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 96(%0),104(%0),112(%0),112(%0), (%1, %3), OP)\
\
- "add $136, %0 \n\t"\
- "add %6, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add $136, %0 \n\t"\
+ "add %6, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
\
: "+r"(temp_ptr), "+r"(dst), "+g"(count)\
: "r"((long)dstStride), "r"(2*(long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(ROUNDER), "g"(4-14*(long)dstStride)\
@@ -2098,18 +2098,18 @@ static void OPNAME ## mpeg4_qpel8_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src,
\
/*FIXME unroll */\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq (%0), %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpckhbw %%mm7, %%mm1 \n\t"\
- "movq %%mm0, (%1) \n\t"\
- "movq %%mm1, 9*8(%1) \n\t"\
- "add $8, %1 \n\t"\
- "add %3, %0 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq (%0), %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpckhbw %%mm7, %%mm1 \n\t"\
+ "movq %%mm0, (%1) \n\t"\
+ "movq %%mm1, 9*8(%1) \n\t"\
+ "add $8, %1 \n\t"\
+ "add %3, %0 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+r" (src), "+r" (temp_ptr), "+r"(count)\
: "r" ((long)srcStride)\
: "memory"\
@@ -2120,30 +2120,30 @@ static void OPNAME ## mpeg4_qpel8_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src,
\
/*FIXME reorder for speed */\
asm volatile(\
- /*"pxor %%mm7, %%mm7 \n\t"*/\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq 8(%0), %%mm1 \n\t"\
- "movq 16(%0), %%mm2 \n\t"\
- "movq 24(%0), %%mm3 \n\t"\
+ /*"pxor %%mm7, %%mm7 \n\t"*/\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 8(%0), %%mm1 \n\t"\
+ "movq 16(%0), %%mm2 \n\t"\
+ "movq 24(%0), %%mm3 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 16(%0), 8(%0), (%0), 32(%0), (%1), OP)\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 8(%0), (%0), (%0), 40(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, (%0), (%0), 8(%0), 48(%0), (%1), OP)\
\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, (%0), 8(%0), 16(%0), 56(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm0, %%mm1, %%mm2, %%mm3, %5, %6, %5, 8(%0), 16(%0), 24(%0), 64(%0), (%1), OP)\
\
QPEL_V_LOW(%%mm1, %%mm2, %%mm3, %%mm0, %5, %6, %5, 16(%0), 24(%0), 32(%0), 64(%0), (%1, %3), OP)\
- "add %4, %1 \n\t"\
+ "add %4, %1 \n\t"\
QPEL_V_LOW(%%mm2, %%mm3, %%mm0, %%mm1, %5, %6, %5, 24(%0), 32(%0), 40(%0), 56(%0), (%1), OP)\
QPEL_V_LOW(%%mm3, %%mm0, %%mm1, %%mm2, %5, %6, %5, 32(%0), 40(%0), 48(%0), 48(%0), (%1, %3), OP)\
\
- "add $72, %0 \n\t"\
- "add %6, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add $72, %0 \n\t"\
+ "add %6, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
\
: "+r"(temp_ptr), "+r"(dst), "+g"(count)\
: "r"((long)dstStride), "r"(2*(long)dstStride), /*"m"(ff_pw_20), "m"(ff_pw_3),*/ "m"(ROUNDER), "g"(4-6*(long)dstStride)\
@@ -2374,15 +2374,15 @@ static void OPNAME ## qpel16_mc22_ ## MMX(uint8_t *dst, uint8_t *src, int stride
OPNAME ## mpeg4_qpel16_v_lowpass_ ## MMX(dst, halfH, stride, 16);\
}
-#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t"
+#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t"
#define AVG_3DNOW_OP(a,b,temp, size) \
-"mov" #size " " #b ", " #temp " \n\t"\
-"pavgusb " #temp ", " #a " \n\t"\
-"mov" #size " " #a ", " #b " \n\t"
+"mov" #size " " #b ", " #temp " \n\t"\
+"pavgusb " #temp ", " #a " \n\t"\
+"mov" #size " " #a ", " #b " \n\t"
#define AVG_MMX2_OP(a,b,temp, size) \
-"mov" #size " " #b ", " #temp " \n\t"\
-"pavgb " #temp ", " #a " \n\t"\
-"mov" #size " " #a ", " #b " \n\t"
+"mov" #size " " #b ", " #temp " \n\t"\
+"pavgb " #temp ", " #a " \n\t"\
+"mov" #size " " #a ", " #b " \n\t"
QPEL_BASE(put_ , ff_pw_16, _ , PUT_OP, PUT_OP)
QPEL_BASE(avg_ , ff_pw_16, _ , AVG_MMX2_OP, AVG_3DNOW_OP)
@@ -2410,40 +2410,40 @@ static int try_8x8basis_mmx(int16_t rem[64], int16_t weight[64], int16_t basis[6
scale<<= 16 + 1 - BASIS_SHIFT + RECON_SHIFT;
asm volatile(
- "pcmpeqw %%mm6, %%mm6 \n\t" // -1w
- "psrlw $15, %%mm6 \n\t" // 1w
- "pxor %%mm7, %%mm7 \n\t"
- "movd %4, %%mm5 \n\t"
- "punpcklwd %%mm5, %%mm5 \n\t"
- "punpcklwd %%mm5, %%mm5 \n\t"
- "1: \n\t"
- "movq (%1, %0), %%mm0 \n\t"
- "movq 8(%1, %0), %%mm1 \n\t"
- "pmulhw %%mm5, %%mm0 \n\t"
- "pmulhw %%mm5, %%mm1 \n\t"
- "paddw %%mm6, %%mm0 \n\t"
- "paddw %%mm6, %%mm1 \n\t"
- "psraw $1, %%mm0 \n\t"
- "psraw $1, %%mm1 \n\t"
- "paddw (%2, %0), %%mm0 \n\t"
- "paddw 8(%2, %0), %%mm1 \n\t"
- "psraw $6, %%mm0 \n\t"
- "psraw $6, %%mm1 \n\t"
- "pmullw (%3, %0), %%mm0 \n\t"
- "pmullw 8(%3, %0), %%mm1 \n\t"
- "pmaddwd %%mm0, %%mm0 \n\t"
- "pmaddwd %%mm1, %%mm1 \n\t"
- "paddd %%mm1, %%mm0 \n\t"
- "psrld $4, %%mm0 \n\t"
- "paddd %%mm0, %%mm7 \n\t"
- "add $16, %0 \n\t"
- "cmp $128, %0 \n\t" //FIXME optimize & bench
- " jb 1b \n\t"
- "movq %%mm7, %%mm6 \n\t"
- "psrlq $32, %%mm7 \n\t"
- "paddd %%mm6, %%mm7 \n\t"
- "psrld $2, %%mm7 \n\t"
- "movd %%mm7, %0 \n\t"
+ "pcmpeqw %%mm6, %%mm6 \n\t" // -1w
+ "psrlw $15, %%mm6 \n\t" // 1w
+ "pxor %%mm7, %%mm7 \n\t"
+ "movd %4, %%mm5 \n\t"
+ "punpcklwd %%mm5, %%mm5 \n\t"
+ "punpcklwd %%mm5, %%mm5 \n\t"
+ "1: \n\t"
+ "movq (%1, %0), %%mm0 \n\t"
+ "movq 8(%1, %0), %%mm1 \n\t"
+ "pmulhw %%mm5, %%mm0 \n\t"
+ "pmulhw %%mm5, %%mm1 \n\t"
+ "paddw %%mm6, %%mm0 \n\t"
+ "paddw %%mm6, %%mm1 \n\t"
+ "psraw $1, %%mm0 \n\t"
+ "psraw $1, %%mm1 \n\t"
+ "paddw (%2, %0), %%mm0 \n\t"
+ "paddw 8(%2, %0), %%mm1 \n\t"
+ "psraw $6, %%mm0 \n\t"
+ "psraw $6, %%mm1 \n\t"
+ "pmullw (%3, %0), %%mm0 \n\t"
+ "pmullw 8(%3, %0), %%mm1 \n\t"
+ "pmaddwd %%mm0, %%mm0 \n\t"
+ "pmaddwd %%mm1, %%mm1 \n\t"
+ "paddd %%mm1, %%mm0 \n\t"
+ "psrld $4, %%mm0 \n\t"
+ "paddd %%mm0, %%mm7 \n\t"
+ "add $16, %0 \n\t"
+ "cmp $128, %0 \n\t" //FIXME optimize & bench
+ " jb 1b \n\t"
+ "movq %%mm7, %%mm6 \n\t"
+ "psrlq $32, %%mm7 \n\t"
+ "paddd %%mm6, %%mm7 \n\t"
+ "psrld $2, %%mm7 \n\t"
+ "movd %%mm7, %0 \n\t"
: "+r" (i)
: "r"(basis), "r"(rem), "r"(weight), "g"(scale)
@@ -2457,27 +2457,27 @@ static void add_8x8basis_mmx(int16_t rem[64], int16_t basis[64], int scale){
if(ABS(scale) < 256){
scale<<= 16 + 1 - BASIS_SHIFT + RECON_SHIFT;
asm volatile(
- "pcmpeqw %%mm6, %%mm6 \n\t" // -1w
- "psrlw $15, %%mm6 \n\t" // 1w
- "movd %3, %%mm5 \n\t"
- "punpcklwd %%mm5, %%mm5 \n\t"
- "punpcklwd %%mm5, %%mm5 \n\t"
- "1: \n\t"
- "movq (%1, %0), %%mm0 \n\t"
- "movq 8(%1, %0), %%mm1 \n\t"
- "pmulhw %%mm5, %%mm0 \n\t"
- "pmulhw %%mm5, %%mm1 \n\t"
- "paddw %%mm6, %%mm0 \n\t"
- "paddw %%mm6, %%mm1 \n\t"
- "psraw $1, %%mm0 \n\t"
- "psraw $1, %%mm1 \n\t"
- "paddw (%2, %0), %%mm0 \n\t"
- "paddw 8(%2, %0), %%mm1 \n\t"
- "movq %%mm0, (%2, %0) \n\t"
- "movq %%mm1, 8(%2, %0) \n\t"
- "add $16, %0 \n\t"
- "cmp $128, %0 \n\t" //FIXME optimize & bench
- " jb 1b \n\t"
+ "pcmpeqw %%mm6, %%mm6 \n\t" // -1w
+ "psrlw $15, %%mm6 \n\t" // 1w
+ "movd %3, %%mm5 \n\t"
+ "punpcklwd %%mm5, %%mm5 \n\t"
+ "punpcklwd %%mm5, %%mm5 \n\t"
+ "1: \n\t"
+ "movq (%1, %0), %%mm0 \n\t"
+ "movq 8(%1, %0), %%mm1 \n\t"
+ "pmulhw %%mm5, %%mm0 \n\t"
+ "pmulhw %%mm5, %%mm1 \n\t"
+ "paddw %%mm6, %%mm0 \n\t"
+ "paddw %%mm6, %%mm1 \n\t"
+ "psraw $1, %%mm0 \n\t"
+ "psraw $1, %%mm1 \n\t"
+ "paddw (%2, %0), %%mm0 \n\t"
+ "paddw 8(%2, %0), %%mm1 \n\t"
+ "movq %%mm0, (%2, %0) \n\t"
+ "movq %%mm1, 8(%2, %0) \n\t"
+ "add $16, %0 \n\t"
+ "cmp $128, %0 \n\t" //FIXME optimize & bench
+ " jb 1b \n\t"
: "+r" (i)
: "r"(basis), "r"(rem), "g"(scale)
@@ -2569,10 +2569,10 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
mm_flags = mm_support();
if (avctx->dsp_mask) {
- if (avctx->dsp_mask & FF_MM_FORCE)
- mm_flags |= (avctx->dsp_mask & 0xffff);
- else
- mm_flags &= ~(avctx->dsp_mask & 0xffff);
+ if (avctx->dsp_mask & FF_MM_FORCE)
+ mm_flags |= (avctx->dsp_mask & 0xffff);
+ else
+ mm_flags &= ~(avctx->dsp_mask & 0xffff);
}
#if 0
@@ -2598,7 +2598,7 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
if(dct_algo==FF_DCT_AUTO || dct_algo==FF_DCT_MMX){
if(mm_flags & MM_SSE2){
c->fdct = ff_fdct_sse2;
- }else if(mm_flags & MM_MMXEXT){
+ }else if(mm_flags & MM_MMXEXT){
c->fdct = ff_fdct_mmx2;
}else{
c->fdct = ff_fdct_mmx;
@@ -2709,13 +2709,13 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
c->hadamard8_diff[0]= hadamard8_diff16_mmx;
c->hadamard8_diff[1]= hadamard8_diff_mmx;
- c->pix_norm1 = pix_norm1_mmx;
- c->sse[0] = (mm_flags & MM_SSE2) ? sse16_sse2 : sse16_mmx;
- c->sse[1] = sse8_mmx;
+ c->pix_norm1 = pix_norm1_mmx;
+ c->sse[0] = (mm_flags & MM_SSE2) ? sse16_sse2 : sse16_mmx;
+ c->sse[1] = sse8_mmx;
c->vsad[4]= vsad_intra16_mmx;
- c->nsse[0] = nsse16_mmx;
- c->nsse[1] = nsse8_mmx;
+ c->nsse[0] = nsse16_mmx;
+ c->nsse[1] = nsse8_mmx;
if(!(avctx->flags & CODEC_FLAG_BITEXACT)){
c->vsad[0] = vsad16_mmx;
}
@@ -2729,7 +2729,7 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
c->h263_v_loop_filter= h263_v_loop_filter_mmx;
c->h263_h_loop_filter= h263_h_loop_filter_mmx;
- c->put_h264_chroma_pixels_tab[0]= put_h264_chroma_mc8_mmx;
+ c->put_h264_chroma_pixels_tab[0]= put_h264_chroma_mc8_mmx;
c->put_h264_chroma_pixels_tab[1]= put_h264_chroma_mc4_mmx;
if (mm_flags & MM_MMXEXT) {
@@ -2829,7 +2829,7 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
dspfunc(avg_h264_qpel, 2, 4);
#undef dspfunc
- c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_mmx2;
+ c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_mmx2;
c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_mmx2;
c->h264_v_loop_filter_luma= h264_v_loop_filter_luma_mmx2;
c->h264_h_loop_filter_luma= h264_h_loop_filter_luma_mmx2;
@@ -2941,7 +2941,7 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
dspfunc(avg_h264_qpel, 1, 8);
dspfunc(avg_h264_qpel, 2, 4);
- c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_3dnow;
+ c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_3dnow;
c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_3dnow;
}
}
diff --git a/libavcodec/i386/dsputil_mmx_avg.h b/libavcodec/i386/dsputil_mmx_avg.h
index 434bc3a0e8..684687b9ed 100644
--- a/libavcodec/i386/dsputil_mmx_avg.h
+++ b/libavcodec/i386/dsputil_mmx_avg.h
@@ -28,522 +28,522 @@
static void DEF(put_pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- "add %%"REG_a", %1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(put_pixels4_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movd (%1), %%mm0 \n\t"
- "movd (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $4, %2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movd (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movd (%1), %%mm1 \n\t"
- "movd (%2), %%mm2 \n\t"
- "movd 4(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movd %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "movd (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movd (%1), %%mm1 \n\t"
- "movd 8(%2), %%mm2 \n\t"
- "movd 12(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movd %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "add $16, %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "movd (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $4, %2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movd (%1), %%mm1 \n\t"
+ "movd (%2), %%mm2 \n\t"
+ "movd 4(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movd %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movd (%1), %%mm1 \n\t"
+ "movd 8(%2), %%mm2 \n\t"
+ "movd 12(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movd %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $16, %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
}
static void DEF(put_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $8, %2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" 16(%2), %%mm0 \n\t"
- PAVGB" 24(%2), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $8, %2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" 16(%2), %%mm0 \n\t"
+ PAVGB" 24(%2), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
static void DEF(put_no_rnd_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "pcmpeqb %%mm6, %%mm6 \n\t"
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $8, %2 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq (%2), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "pxor %%mm6, %%mm2 \n\t"
- "pxor %%mm6, %%mm3 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq 16(%2), %%mm2 \n\t"
- "movq 24(%2), %%mm3 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "pxor %%mm6, %%mm2 \n\t"
- "pxor %%mm6, %%mm3 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "pcmpeqb %%mm6, %%mm6 \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $8, %2 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%2), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "pxor %%mm6, %%mm2 \n\t"
+ "pxor %%mm6, %%mm3 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq 16(%2), %%mm2 \n\t"
+ "movq 24(%2), %%mm3 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "pxor %%mm6, %%mm2 \n\t"
+ "pxor %%mm6, %%mm3 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
static void DEF(avg_pixels4_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movd (%1), %%mm0 \n\t"
- "movd (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $4, %2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movd (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movd (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 4(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- PAVGB" (%3), %%mm1 \n\t"
- "movd %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "movd (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movd (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" 8(%2), %%mm0 \n\t"
- PAVGB" 12(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movd %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- PAVGB" (%3), %%mm1 \n\t"
- "movd %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "add $16, %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "movd (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $4, %2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movd (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 4(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ PAVGB" (%3), %%mm1 \n\t"
+ "movd %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movd (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movd (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" 8(%2), %%mm0 \n\t"
+ PAVGB" 12(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movd %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ PAVGB" (%3), %%mm1 \n\t"
+ "movd %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $16, %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
}
static void DEF(avg_pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $8, %2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- PAVGB" (%3), %%mm1 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" 16(%2), %%mm0 \n\t"
- PAVGB" 24(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- "movq %%mm0, (%3) \n\t"
- "add %5, %3 \n\t"
- PAVGB" (%3), %%mm1 \n\t"
- "movq %%mm1, (%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $8, %2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ PAVGB" (%3), %%mm1 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" 16(%2), %%mm0 \n\t"
+ PAVGB" 24(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "add %5, %3 \n\t"
+ PAVGB" (%3), %%mm1 \n\t"
+ "movq %%mm1, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
static void DEF(put_pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq 8(%1), %%mm2 \n\t"
- "movq 8(%1, %3), %%mm3 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- PAVGB" 9(%1), %%mm2 \n\t"
- PAVGB" 9(%1, %3), %%mm3 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "movq %%mm2, 8(%2) \n\t"
- "movq %%mm3, 8(%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq 8(%1), %%mm2 \n\t"
- "movq 8(%1, %3), %%mm3 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- PAVGB" 9(%1), %%mm2 \n\t"
- PAVGB" 9(%1, %3), %%mm3 \n\t"
- "add %%"REG_a", %1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "movq %%mm2, 8(%2) \n\t"
- "movq %%mm3, 8(%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq 8(%1), %%mm2 \n\t"
+ "movq 8(%1, %3), %%mm3 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ PAVGB" 9(%1), %%mm2 \n\t"
+ PAVGB" 9(%1, %3), %%mm3 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "movq %%mm2, 8(%2) \n\t"
+ "movq %%mm3, 8(%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq 8(%1), %%mm2 \n\t"
+ "movq 8(%1, %3), %%mm3 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ PAVGB" 9(%1), %%mm2 \n\t"
+ PAVGB" 9(%1, %3), %%mm3 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "movq %%mm2, 8(%2) \n\t"
+ "movq %%mm3, 8(%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(put_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $16, %2 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" 16(%2), %%mm0 \n\t"
- PAVGB" 24(%2), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $16, %2 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" 16(%2), %%mm0 \n\t"
+ PAVGB" 24(%2), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
static void DEF(avg_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $16, %2 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- PAVGB" 8(%3), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" 8(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- PAVGB" 8(%3), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- PAVGB" 16(%2), %%mm0 \n\t"
- PAVGB" 24(%2), %%mm1 \n\t"
- PAVGB" (%3), %%mm0 \n\t"
- PAVGB" 8(%3), %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $16, %2 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ PAVGB" 8(%3), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" 8(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ PAVGB" 8(%3), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ PAVGB" 16(%2), %%mm0 \n\t"
+ PAVGB" 24(%2), %%mm1 \n\t"
+ PAVGB" (%3), %%mm0 \n\t"
+ PAVGB" 8(%3), %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
static void DEF(put_no_rnd_pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
__asm __volatile(
- "pcmpeqb %%mm6, %%mm6\n\t"
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "movq (%2), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "pxor %%mm6, %%mm2 \n\t"
- "pxor %%mm6, %%mm3 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $16, %2 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq (%2), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "pxor %%mm6, %%mm2 \n\t"
- "pxor %%mm6, %%mm3 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 8(%1), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq 16(%2), %%mm2 \n\t"
- "movq 24(%2), %%mm3 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "pxor %%mm6, %%mm2 \n\t"
- "pxor %%mm6, %%mm3 \n\t"
- PAVGB" %%mm2, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm1 \n\t"
- "pxor %%mm6, %%mm0 \n\t"
- "pxor %%mm6, %%mm1 \n\t"
- "movq %%mm0, (%3) \n\t"
- "movq %%mm1, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
+ "pcmpeqb %%mm6, %%mm6 \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "movq (%2), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "pxor %%mm6, %%mm2 \n\t"
+ "pxor %%mm6, %%mm3 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $16, %2 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%2), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "pxor %%mm6, %%mm2 \n\t"
+ "pxor %%mm6, %%mm3 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 8(%1), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq 16(%2), %%mm2 \n\t"
+ "movq 24(%2), %%mm3 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "pxor %%mm6, %%mm2 \n\t"
+ "pxor %%mm6, %%mm3 \n\t"
+ PAVGB" %%mm2, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm1 \n\t"
+ "pxor %%mm6, %%mm0 \n\t"
+ "pxor %%mm6, %%mm1 \n\t"
+ "movq %%mm0, (%3) \n\t"
+ "movq %%mm1, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
//the following should be used, though better not with gcc ...
-/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
- :"r"(src1Stride), "r"(dstStride)
- :"memory");*/
+/* :"+g"(h), "+r"(src1), "+r"(src2), "+r"(dst)
+ :"r"(src1Stride), "r"(dstStride)
+ :"memory");*/
}
/* GL: this function does incorrect rounding if overflow */
@@ -551,67 +551,67 @@ static void DEF(put_no_rnd_pixels8_x2)(uint8_t *block, const uint8_t *pixels, in
{
MOVQ_BONE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- "add %%"REG_a", %1 \n\t"
- "psubusb %%mm6, %%mm0 \n\t"
- "psubusb %%mm6, %%mm2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm2 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- "psubusb %%mm6, %%mm0 \n\t"
- "psubusb %%mm6, %%mm2 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm3, %%mm2 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "psubusb %%mm6, %%mm0 \n\t"
+ "psubusb %%mm6, %%mm2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm2 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "psubusb %%mm6, %%mm0 \n\t"
+ "psubusb %%mm6, %%mm2 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm3, %%mm2 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(put_pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- "sub %3, %2 \n\t"
- "1: \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm2, %%mm1 \n\t"
- "movq %%mm0, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" %%mm1, %%mm2 \n\t"
- PAVGB" %%mm0, %%mm1 \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D" (block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "sub %3, %2 \n\t"
+ "1: \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm2, %%mm1 \n\t"
+ "movq %%mm0, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" %%mm1, %%mm2 \n\t"
+ PAVGB" %%mm0, %%mm1 \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D" (block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
/* GL: this function does incorrect rounding if overflow */
@@ -619,131 +619,131 @@ static void DEF(put_no_rnd_pixels8_y2)(uint8_t *block, const uint8_t *pixels, in
{
MOVQ_BONE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- "sub %3, %2 \n\t"
- "1: \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- "add %%"REG_a", %1 \n\t"
- "psubusb %%mm6, %%mm1 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm2, %%mm1 \n\t"
- "movq %%mm0, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- "psubusb %%mm6, %%mm1 \n\t"
- PAVGB" %%mm1, %%mm2 \n\t"
- PAVGB" %%mm0, %%mm1 \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D" (block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "sub %3, %2 \n\t"
+ "1: \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "psubusb %%mm6, %%mm1 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm2, %%mm1 \n\t"
+ "movq %%mm0, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "psubusb %%mm6, %%mm1 \n\t"
+ PAVGB" %%mm1, %%mm2 \n\t"
+ PAVGB" %%mm0, %%mm1 \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D" (block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(avg_pixels8)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "1: \n\t"
- "movq (%2), %%mm0 \n\t"
- "movq (%2, %3), %%mm1 \n\t"
- PAVGB" (%1), %%mm0 \n\t"
- PAVGB" (%1, %3), %%mm1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%2), %%mm0 \n\t"
- "movq (%2, %3), %%mm1 \n\t"
- PAVGB" (%1), %%mm0 \n\t"
- PAVGB" (%1, %3), %%mm1 \n\t"
- "add %%"REG_a", %1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq (%2), %%mm0 \n\t"
+ "movq (%2, %3), %%mm1 \n\t"
+ PAVGB" (%1), %%mm0 \n\t"
+ PAVGB" (%1, %3), %%mm1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%2), %%mm0 \n\t"
+ "movq (%2, %3), %%mm1 \n\t"
+ PAVGB" (%1), %%mm0 \n\t"
+ PAVGB" (%1, %3), %%mm1 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(avg_pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm2 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" (%2, %3), %%mm2 \n\t"
- "add %%"REG_a", %1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm2 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" (%2, %3), %%mm2 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm2 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" (%2, %3), %%mm2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm2 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" (%2, %3), %%mm2 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
static void DEF(avg_pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- "sub %3, %2 \n\t"
- "1: \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm2, %%mm1 \n\t"
- "movq (%2, %3), %%mm3 \n\t"
- "movq (%2, %%"REG_a"), %%mm4 \n\t"
- PAVGB" %%mm3, %%mm0 \n\t"
- PAVGB" %%mm4, %%mm1 \n\t"
- "movq %%mm0, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- PAVGB" %%mm1, %%mm2 \n\t"
- PAVGB" %%mm0, %%mm1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- "movq (%2, %3), %%mm3 \n\t"
- "movq (%2, %%"REG_a"), %%mm4 \n\t"
- PAVGB" %%mm3, %%mm2 \n\t"
- PAVGB" %%mm4, %%mm1 \n\t"
- "movq %%mm2, (%2, %3) \n\t"
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "sub %3, %2 \n\t"
+ "1: \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm2, %%mm1 \n\t"
+ "movq (%2, %3), %%mm3 \n\t"
+ "movq (%2, %%"REG_a"), %%mm4 \n\t"
+ PAVGB" %%mm3, %%mm0 \n\t"
+ PAVGB" %%mm4, %%mm1 \n\t"
+ "movq %%mm0, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ PAVGB" %%mm1, %%mm2 \n\t"
+ PAVGB" %%mm0, %%mm1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "movq (%2, %3), %%mm3 \n\t"
+ "movq (%2, %%"REG_a"), %%mm4 \n\t"
+ PAVGB" %%mm3, %%mm2 \n\t"
+ PAVGB" %%mm4, %%mm1 \n\t"
+ "movq %%mm2, (%2, %3) \n\t"
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
// Note this is not correctly rounded, but this function is only used for b frames so it doesnt matter
@@ -751,41 +751,41 @@ static void DEF(avg_pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line
{
MOVQ_BONE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- PAVGB" 1(%1), %%mm0 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "psubusb %%mm6, %%mm2 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- PAVGB" 1(%1, %%"REG_a"), %%mm2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" %%mm1, %%mm0 \n\t"
- PAVGB" %%mm2, %%mm1 \n\t"
- PAVGB" (%2), %%mm0 \n\t"
- PAVGB" (%2, %3), %%mm1 \n\t"
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- PAVGB" 1(%1, %3), %%mm1 \n\t"
- PAVGB" 1(%1, %%"REG_a"), %%mm0 \n\t"
- "add %%"REG_a", %2 \n\t"
- "add %%"REG_a", %1 \n\t"
- PAVGB" %%mm1, %%mm2 \n\t"
- PAVGB" %%mm0, %%mm1 \n\t"
- PAVGB" (%2), %%mm2 \n\t"
- PAVGB" (%2, %3), %%mm1 \n\t"
- "movq %%mm2, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r" ((long)line_size)
- :"%"REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ PAVGB" 1(%1), %%mm0 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "psubusb %%mm6, %%mm2 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ PAVGB" 1(%1, %%"REG_a"), %%mm2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" %%mm1, %%mm0 \n\t"
+ PAVGB" %%mm2, %%mm1 \n\t"
+ PAVGB" (%2), %%mm0 \n\t"
+ PAVGB" (%2, %3), %%mm1 \n\t"
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ PAVGB" 1(%1, %3), %%mm1 \n\t"
+ PAVGB" 1(%1, %%"REG_a"), %%mm0 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "add %%"REG_a", %1 \n\t"
+ PAVGB" %%mm1, %%mm2 \n\t"
+ PAVGB" %%mm0, %%mm1 \n\t"
+ PAVGB" (%2), %%mm2 \n\t"
+ PAVGB" (%2, %3), %%mm1 \n\t"
+ "movq %%mm2, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r" ((long)line_size)
+ :"%"REG_a, "memory");
}
//FIXME the following could be optimized too ...
diff --git a/libavcodec/i386/dsputil_mmx_rnd.h b/libavcodec/i386/dsputil_mmx_rnd.h
index 6d93f9d55f..e2745c353c 100644
--- a/libavcodec/i386/dsputil_mmx_rnd.h
+++ b/libavcodec/i386/dsputil_mmx_rnd.h
@@ -27,206 +27,206 @@ static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line
{
MOVQ_BFE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r"((long)line_size)
- :REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r"((long)line_size)
+ :REG_a, "memory");
}
static void attribute_unused DEF(put, pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
MOVQ_BFE(mm6);
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "add $8, %2 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm4, %%mm6)
- "movq %%mm4, (%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm5, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 16(%2), %%mm1 \n\t"
- "add %4, %1 \n\t"
- "movq (%1), %%mm2 \n\t"
- "movq 24(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- "add $32, %2 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%3) \n\t"
- "add %5, %3 \n\t"
- "movq %%mm5, (%3) \n\t"
- "add %5, %3 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "add $8, %2 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm4, %%mm6)
+ "movq %%mm4, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm5, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 16(%2), %%mm1 \n\t"
+ "add %4, %1 \n\t"
+ "movq (%1), %%mm2 \n\t"
+ "movq 24(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ "add $32, %2 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq %%mm5, (%3) \n\t"
+ "add %5, %3 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
:"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
:"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
}
static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
MOVQ_BFE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "movq 8(%1), %%mm0 \n\t"
- "movq 9(%1), %%mm1 \n\t"
- "movq 8(%1, %3), %%mm2 \n\t"
- "movq 9(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, 8(%2) \n\t"
- "movq %%mm5, 8(%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm1 \n\t"
- "movq (%1, %3), %%mm2 \n\t"
- "movq 1(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "movq 8(%1), %%mm0 \n\t"
- "movq 9(%1), %%mm1 \n\t"
- "movq 8(%1, %3), %%mm2 \n\t"
- "movq 9(%1, %3), %%mm3 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, 8(%2) \n\t"
- "movq %%mm5, 8(%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r"((long)line_size)
- :REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "movq 8(%1), %%mm0 \n\t"
+ "movq 9(%1), %%mm1 \n\t"
+ "movq 8(%1, %3), %%mm2 \n\t"
+ "movq 9(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, 8(%2) \n\t"
+ "movq %%mm5, 8(%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm1 \n\t"
+ "movq (%1, %3), %%mm2 \n\t"
+ "movq 1(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "movq 8(%1), %%mm0 \n\t"
+ "movq 9(%1), %%mm1 \n\t"
+ "movq 8(%1, %3), %%mm2 \n\t"
+ "movq 9(%1, %3), %%mm3 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, 8(%2) \n\t"
+ "movq %%mm5, 8(%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r"((long)line_size)
+ :REG_a, "memory");
}
static void attribute_unused DEF(put, pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int src1Stride, int h)
{
MOVQ_BFE(mm6);
__asm __volatile(
- "testl $1, %0 \n\t"
- " jz 1f \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "movq 8(%1), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- "add $16, %2 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%3) \n\t"
- "movq %%mm5, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "decl %0 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq (%2), %%mm1 \n\t"
- "movq 8(%1), %%mm2 \n\t"
- "movq 8(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%3) \n\t"
- "movq %%mm5, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "movq (%1), %%mm0 \n\t"
- "movq 16(%2), %%mm1 \n\t"
- "movq 8(%1), %%mm2 \n\t"
- "movq 24(%2), %%mm3 \n\t"
- "add %4, %1 \n\t"
- PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
- "movq %%mm4, (%3) \n\t"
- "movq %%mm5, 8(%3) \n\t"
- "add %5, %3 \n\t"
- "add $32, %2 \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
+ "testl $1, %0 \n\t"
+ " jz 1f \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "movq 8(%1), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ "add $16, %2 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%3) \n\t"
+ "movq %%mm5, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "decl %0 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq (%2), %%mm1 \n\t"
+ "movq 8(%1), %%mm2 \n\t"
+ "movq 8(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%3) \n\t"
+ "movq %%mm5, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 16(%2), %%mm1 \n\t"
+ "movq 8(%1), %%mm2 \n\t"
+ "movq 24(%2), %%mm3 \n\t"
+ "add %4, %1 \n\t"
+ PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5)
+ "movq %%mm4, (%3) \n\t"
+ "movq %%mm5, 8(%3) \n\t"
+ "add %5, %3 \n\t"
+ "add $32, %2 \n\t"
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
#ifdef PIC //Note "+bm" and "+mb" are buggy too (with gcc 3.2.2 at least) and cant be used
- :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+m"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#else
- :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
+ :"+b"(h), "+a"(src1), "+c"(src2), "+d"(dst)
#endif
- :"S"((long)src1Stride), "D"((long)dstStride)
- :"memory");
+ :"S"((long)src1Stride), "D"((long)dstStride)
+ :"memory");
}
static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
{
MOVQ_BFE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"),%%mm2 \n\t"
- PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"),%%mm0 \n\t"
- PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5)
- "movq %%mm4, (%2) \n\t"
- "movq %%mm5, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r"((long)line_size)
- :REG_a, "memory");
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"),%%mm2 \n\t"
+ PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"),%%mm0 \n\t"
+ PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5)
+ "movq %%mm4, (%2) \n\t"
+ "movq %%mm5, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r"((long)line_size)
+ :REG_a, "memory");
}
static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@@ -234,65 +234,65 @@ static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
MOVQ_ZERO(mm7);
SET_RND(mm6); // =2 for rnd and =1 for no_rnd version
__asm __volatile(
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm4 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm4, %%mm5 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm4 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm5 \n\t"
- "paddusw %%mm0, %%mm4 \n\t"
- "paddusw %%mm1, %%mm5 \n\t"
- "xor %%"REG_a", %%"REG_a" \n\t"
- "add %3, %1 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "paddusw %%mm2, %%mm0 \n\t"
- "paddusw %%mm3, %%mm1 \n\t"
- "paddusw %%mm6, %%mm4 \n\t"
- "paddusw %%mm6, %%mm5 \n\t"
- "paddusw %%mm0, %%mm4 \n\t"
- "paddusw %%mm1, %%mm5 \n\t"
- "psrlw $2, %%mm4 \n\t"
- "psrlw $2, %%mm5 \n\t"
- "packuswb %%mm5, %%mm4 \n\t"
- "movq %%mm4, (%2, %%"REG_a") \n\t"
- "add %3, %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm4 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm4, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddusw %%mm0, %%mm4 \n\t"
+ "paddusw %%mm1, %%mm5 \n\t"
+ "xor %%"REG_a", %%"REG_a" \n\t"
+ "add %3, %1 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "paddusw %%mm2, %%mm0 \n\t"
+ "paddusw %%mm3, %%mm1 \n\t"
+ "paddusw %%mm6, %%mm4 \n\t"
+ "paddusw %%mm6, %%mm5 \n\t"
+ "paddusw %%mm0, %%mm4 \n\t"
+ "paddusw %%mm1, %%mm5 \n\t"
+ "psrlw $2, %%mm4 \n\t"
+ "psrlw $2, %%mm5 \n\t"
+ "packuswb %%mm5, %%mm4 \n\t"
+ "movq %%mm4, (%2, %%"REG_a") \n\t"
+ "add %3, %%"REG_a" \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3
- "movq 1(%1, %%"REG_a"), %%mm4 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "movq %%mm4, %%mm5 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpcklbw %%mm7, %%mm4 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "punpckhbw %%mm7, %%mm5 \n\t"
- "paddusw %%mm2, %%mm4 \n\t"
- "paddusw %%mm3, %%mm5 \n\t"
- "paddusw %%mm6, %%mm0 \n\t"
- "paddusw %%mm6, %%mm1 \n\t"
- "paddusw %%mm4, %%mm0 \n\t"
- "paddusw %%mm5, %%mm1 \n\t"
- "psrlw $2, %%mm0 \n\t"
- "psrlw $2, %%mm1 \n\t"
- "packuswb %%mm1, %%mm0 \n\t"
- "movq %%mm0, (%2, %%"REG_a") \n\t"
- "add %3, %%"REG_a" \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3
+ "movq 1(%1, %%"REG_a"), %%mm4 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "movq %%mm4, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpcklbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddusw %%mm2, %%mm4 \n\t"
+ "paddusw %%mm3, %%mm5 \n\t"
+ "paddusw %%mm6, %%mm0 \n\t"
+ "paddusw %%mm6, %%mm1 \n\t"
+ "paddusw %%mm4, %%mm0 \n\t"
+ "paddusw %%mm5, %%mm1 \n\t"
+ "psrlw $2, %%mm0 \n\t"
+ "psrlw $2, %%mm1 \n\t"
+ "packuswb %%mm1, %%mm0 \n\t"
+ "movq %%mm0, (%2, %%"REG_a") \n\t"
+ "add %3, %%"REG_a" \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels)
- :"D"(block), "r"((long)line_size)
- :REG_a, "memory");
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels)
+ :"D"(block), "r"((long)line_size)
+ :REG_a, "memory");
}
// avg_pixels
@@ -301,16 +301,16 @@ static void attribute_unused DEF(avg, pixels4)(uint8_t *block, const uint8_t *pi
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movd %0, %%mm0 \n\t"
- "movd %1, %%mm1 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- "movd %%mm2, %0 \n\t"
- :"+m"(*block)
- :"m"(*pixels)
- :"memory");
- pixels += line_size;
- block += line_size;
+ __asm __volatile(
+ "movd %0, %%mm0 \n\t"
+ "movd %1, %%mm1 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ "movd %%mm2, %0 \n\t"
+ :"+m"(*block)
+ :"m"(*pixels)
+ :"memory");
+ pixels += line_size;
+ block += line_size;
}
while (--h);
}
@@ -321,16 +321,16 @@ static void DEF(avg, pixels8)(uint8_t *block, const uint8_t *pixels, int line_si
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %0, %%mm0 \n\t"
- "movq %1, %%mm1 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- "movq %%mm2, %0 \n\t"
- :"+m"(*block)
- :"m"(*pixels)
- :"memory");
- pixels += line_size;
- block += line_size;
+ __asm __volatile(
+ "movq %0, %%mm0 \n\t"
+ "movq %1, %%mm1 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ "movq %%mm2, %0 \n\t"
+ :"+m"(*block)
+ :"m"(*pixels)
+ :"memory");
+ pixels += line_size;
+ block += line_size;
}
while (--h);
}
@@ -340,20 +340,20 @@ static void DEF(avg, pixels16)(uint8_t *block, const uint8_t *pixels, int line_s
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %0, %%mm0 \n\t"
- "movq %1, %%mm1 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- "movq %%mm2, %0 \n\t"
- "movq 8%0, %%mm0 \n\t"
- "movq 8%1, %%mm1 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- "movq %%mm2, 8%0 \n\t"
- :"+m"(*block)
- :"m"(*pixels)
- :"memory");
- pixels += line_size;
- block += line_size;
+ __asm __volatile(
+ "movq %0, %%mm0 \n\t"
+ "movq %1, %%mm1 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ "movq %%mm2, %0 \n\t"
+ "movq 8%0, %%mm0 \n\t"
+ "movq 8%1, %%mm1 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ "movq %%mm2, 8%0 \n\t"
+ :"+m"(*block)
+ :"m"(*pixels)
+ :"memory");
+ pixels += line_size;
+ block += line_size;
}
while (--h);
}
@@ -363,18 +363,18 @@ static void DEF(avg, pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %1, %%mm0 \n\t"
- "movq 1%1, %%mm1 \n\t"
- "movq %0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, %0 \n\t"
- :"+m"(*block)
- :"m"(*pixels)
- :"memory");
- pixels += line_size;
- block += line_size;
+ __asm __volatile(
+ "movq %1, %%mm0 \n\t"
+ "movq 1%1, %%mm1 \n\t"
+ "movq %0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, %0 \n\t"
+ :"+m"(*block)
+ :"m"(*pixels)
+ :"memory");
+ pixels += line_size;
+ block += line_size;
} while (--h);
}
@@ -383,17 +383,17 @@ static __attribute__((unused)) void DEF(avg, pixels8_l2)(uint8_t *dst, uint8_t *
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %1, %%mm0 \n\t"
- "movq %2, %%mm1 \n\t"
- "movq %0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, %0 \n\t"
- :"+m"(*dst)
- :"m"(*src1), "m"(*src2)
- :"memory");
- dst += dstStride;
+ __asm __volatile(
+ "movq %1, %%mm0 \n\t"
+ "movq %2, %%mm1 \n\t"
+ "movq %0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, %0 \n\t"
+ :"+m"(*dst)
+ :"m"(*src1), "m"(*src2)
+ :"memory");
+ dst += dstStride;
src1 += src1Stride;
src2 += 8;
} while (--h);
@@ -404,24 +404,24 @@ static void DEF(avg, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int lin
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %1, %%mm0 \n\t"
- "movq 1%1, %%mm1 \n\t"
- "movq %0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, %0 \n\t"
- "movq 8%1, %%mm0 \n\t"
- "movq 9%1, %%mm1 \n\t"
- "movq 8%0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, 8%0 \n\t"
- :"+m"(*block)
- :"m"(*pixels)
- :"memory");
- pixels += line_size;
- block += line_size;
+ __asm __volatile(
+ "movq %1, %%mm0 \n\t"
+ "movq 1%1, %%mm1 \n\t"
+ "movq %0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, %0 \n\t"
+ "movq 8%1, %%mm0 \n\t"
+ "movq 9%1, %%mm1 \n\t"
+ "movq 8%0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, 8%0 \n\t"
+ :"+m"(*block)
+ :"m"(*pixels)
+ :"memory");
+ pixels += line_size;
+ block += line_size;
} while (--h);
}
@@ -430,23 +430,23 @@ static __attribute__((unused)) void DEF(avg, pixels16_l2)(uint8_t *dst, uint8_t
MOVQ_BFE(mm6);
JUMPALIGN();
do {
- __asm __volatile(
- "movq %1, %%mm0 \n\t"
- "movq %2, %%mm1 \n\t"
- "movq %0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, %0 \n\t"
- "movq 8%1, %%mm0 \n\t"
- "movq 8%2, %%mm1 \n\t"
- "movq 8%0, %%mm3 \n\t"
- PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
- PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
- "movq %%mm0, 8%0 \n\t"
- :"+m"(*dst)
- :"m"(*src1), "m"(*src2)
- :"memory");
- dst += dstStride;
+ __asm __volatile(
+ "movq %1, %%mm0 \n\t"
+ "movq %2, %%mm1 \n\t"
+ "movq %0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, %0 \n\t"
+ "movq 8%1, %%mm0 \n\t"
+ "movq 8%2, %%mm1 \n\t"
+ "movq 8%0, %%mm3 \n\t"
+ PAVGB(%%mm0, %%mm1, %%mm2, %%mm6)
+ PAVGB(%%mm3, %%mm2, %%mm0, %%mm6)
+ "movq %%mm0, 8%0 \n\t"
+ :"+m"(*dst)
+ :"m"(*src1), "m"(*src2)
+ :"memory");
+ dst += dstStride;
src1 += src1Stride;
src2 += 16;
} while (--h);
@@ -456,39 +456,39 @@ static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line
{
MOVQ_BFE(mm6);
__asm __volatile(
- "lea (%3, %3), %%"REG_a" \n\t"
- "movq (%1), %%mm0 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5)
- "movq (%2), %%mm3 \n\t"
- PAVGB(%%mm3, %%mm4, %%mm0, %%mm6)
- "movq (%2, %3), %%mm3 \n\t"
- PAVGB(%%mm3, %%mm5, %%mm1, %%mm6)
- "movq %%mm0, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
+ "lea (%3, %3), %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5)
+ "movq (%2), %%mm3 \n\t"
+ PAVGB(%%mm3, %%mm4, %%mm0, %%mm6)
+ "movq (%2, %3), %%mm3 \n\t"
+ PAVGB(%%mm3, %%mm5, %%mm1, %%mm6)
+ "movq %%mm0, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
- "movq (%1, %3), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5)
- "movq (%2), %%mm3 \n\t"
- PAVGB(%%mm3, %%mm4, %%mm2, %%mm6)
- "movq (%2, %3), %%mm3 \n\t"
- PAVGB(%%mm3, %%mm5, %%mm1, %%mm6)
- "movq %%mm2, (%2) \n\t"
- "movq %%mm1, (%2, %3) \n\t"
- "add %%"REG_a", %1 \n\t"
- "add %%"REG_a", %2 \n\t"
+ "movq (%1, %3), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5)
+ "movq (%2), %%mm3 \n\t"
+ PAVGB(%%mm3, %%mm4, %%mm2, %%mm6)
+ "movq (%2, %3), %%mm3 \n\t"
+ PAVGB(%%mm3, %%mm5, %%mm1, %%mm6)
+ "movq %%mm2, (%2) \n\t"
+ "movq %%mm1, (%2, %3) \n\t"
+ "add %%"REG_a", %1 \n\t"
+ "add %%"REG_a", %2 \n\t"
- "subl $4, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels), "+D"(block)
- :"r"((long)line_size)
- :REG_a, "memory");
+ "subl $4, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels), "+D"(block)
+ :"r"((long)line_size)
+ :REG_a, "memory");
}
// this routine is 'slightly' suboptimal but mostly unused
@@ -497,73 +497,73 @@ static void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
MOVQ_ZERO(mm7);
SET_RND(mm6); // =2 for rnd and =1 for no_rnd version
__asm __volatile(
- "movq (%1), %%mm0 \n\t"
- "movq 1(%1), %%mm4 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm4, %%mm5 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm4 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm5 \n\t"
- "paddusw %%mm0, %%mm4 \n\t"
- "paddusw %%mm1, %%mm5 \n\t"
- "xor %%"REG_a", %%"REG_a" \n\t"
- "add %3, %1 \n\t"
- ".balign 8 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "paddusw %%mm2, %%mm0 \n\t"
- "paddusw %%mm3, %%mm1 \n\t"
- "paddusw %%mm6, %%mm4 \n\t"
- "paddusw %%mm6, %%mm5 \n\t"
- "paddusw %%mm0, %%mm4 \n\t"
- "paddusw %%mm1, %%mm5 \n\t"
- "psrlw $2, %%mm4 \n\t"
- "psrlw $2, %%mm5 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "packuswb %%mm5, %%mm4 \n\t"
- "pcmpeqd %%mm2, %%mm2 \n\t"
- "paddb %%mm2, %%mm2 \n\t"
- PAVGB(%%mm3, %%mm4, %%mm5, %%mm2)
- "movq %%mm5, (%2, %%"REG_a") \n\t"
- "add %3, %%"REG_a" \n\t"
+ "movq (%1), %%mm0 \n\t"
+ "movq 1(%1), %%mm4 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm4, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddusw %%mm0, %%mm4 \n\t"
+ "paddusw %%mm1, %%mm5 \n\t"
+ "xor %%"REG_a", %%"REG_a" \n\t"
+ "add %3, %1 \n\t"
+ ".balign 8 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "paddusw %%mm2, %%mm0 \n\t"
+ "paddusw %%mm3, %%mm1 \n\t"
+ "paddusw %%mm6, %%mm4 \n\t"
+ "paddusw %%mm6, %%mm5 \n\t"
+ "paddusw %%mm0, %%mm4 \n\t"
+ "paddusw %%mm1, %%mm5 \n\t"
+ "psrlw $2, %%mm4 \n\t"
+ "psrlw $2, %%mm5 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "packuswb %%mm5, %%mm4 \n\t"
+ "pcmpeqd %%mm2, %%mm2 \n\t"
+ "paddb %%mm2, %%mm2 \n\t"
+ PAVGB(%%mm3, %%mm4, %%mm5, %%mm2)
+ "movq %%mm5, (%2, %%"REG_a") \n\t"
+ "add %3, %%"REG_a" \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3
- "movq 1(%1, %%"REG_a"), %%mm4 \n\t"
- "movq %%mm2, %%mm3 \n\t"
- "movq %%mm4, %%mm5 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpcklbw %%mm7, %%mm4 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "punpckhbw %%mm7, %%mm5 \n\t"
- "paddusw %%mm2, %%mm4 \n\t"
- "paddusw %%mm3, %%mm5 \n\t"
- "paddusw %%mm6, %%mm0 \n\t"
- "paddusw %%mm6, %%mm1 \n\t"
- "paddusw %%mm4, %%mm0 \n\t"
- "paddusw %%mm5, %%mm1 \n\t"
- "psrlw $2, %%mm0 \n\t"
- "psrlw $2, %%mm1 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "packuswb %%mm1, %%mm0 \n\t"
- "pcmpeqd %%mm2, %%mm2 \n\t"
- "paddb %%mm2, %%mm2 \n\t"
- PAVGB(%%mm3, %%mm0, %%mm1, %%mm2)
- "movq %%mm1, (%2, %%"REG_a") \n\t"
- "add %3, %%"REG_a" \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t" // 0 <-> 2 1 <-> 3
+ "movq 1(%1, %%"REG_a"), %%mm4 \n\t"
+ "movq %%mm2, %%mm3 \n\t"
+ "movq %%mm4, %%mm5 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpcklbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "punpckhbw %%mm7, %%mm5 \n\t"
+ "paddusw %%mm2, %%mm4 \n\t"
+ "paddusw %%mm3, %%mm5 \n\t"
+ "paddusw %%mm6, %%mm0 \n\t"
+ "paddusw %%mm6, %%mm1 \n\t"
+ "paddusw %%mm4, %%mm0 \n\t"
+ "paddusw %%mm5, %%mm1 \n\t"
+ "psrlw $2, %%mm0 \n\t"
+ "psrlw $2, %%mm1 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "packuswb %%mm1, %%mm0 \n\t"
+ "pcmpeqd %%mm2, %%mm2 \n\t"
+ "paddb %%mm2, %%mm2 \n\t"
+ PAVGB(%%mm3, %%mm0, %%mm1, %%mm2)
+ "movq %%mm1, (%2, %%"REG_a") \n\t"
+ "add %3, %%"REG_a" \n\t"
- "subl $2, %0 \n\t"
- "jnz 1b \n\t"
- :"+g"(h), "+S"(pixels)
- :"D"(block), "r"((long)line_size)
- :REG_a, "memory");
+ "subl $2, %0 \n\t"
+ "jnz 1b \n\t"
+ :"+g"(h), "+S"(pixels)
+ :"D"(block), "r"((long)line_size)
+ :REG_a, "memory");
}
//FIXME optimize
diff --git a/libavcodec/i386/fdct_mmx.c b/libavcodec/i386/fdct_mmx.c
index f3023549a2..f6150c83c9 100644
--- a/libavcodec/i386/fdct_mmx.c
+++ b/libavcodec/i386/fdct_mmx.c
@@ -30,21 +30,21 @@
//
//////////////////////////////////////////////////////////////////////
-#define BITS_FRW_ACC 3 //; 2 or 3 for accuracy
-#define SHIFT_FRW_COL BITS_FRW_ACC
-#define SHIFT_FRW_ROW (BITS_FRW_ACC + 17 - 3)
-#define RND_FRW_ROW (1 << (SHIFT_FRW_ROW-1))
-//#define RND_FRW_COL (1 << (SHIFT_FRW_COL-1))
+#define BITS_FRW_ACC 3 //; 2 or 3 for accuracy
+#define SHIFT_FRW_COL BITS_FRW_ACC
+#define SHIFT_FRW_ROW (BITS_FRW_ACC + 17 - 3)
+#define RND_FRW_ROW (1 << (SHIFT_FRW_ROW-1))
+//#define RND_FRW_COL (1 << (SHIFT_FRW_COL-1))
//concatenated table, for forward DCT transformation
static const int16_t fdct_tg_all_16[] ATTR_ALIGN(8) = {
- 13036, 13036, 13036, 13036, // tg * (2<<16) + 0.5
- 27146, 27146, 27146, 27146, // tg * (2<<16) + 0.5
- -21746, -21746, -21746, -21746, // tg * (2<<16) + 0.5
+ 13036, 13036, 13036, 13036, // tg * (2<<16) + 0.5
+ 27146, 27146, 27146, 27146, // tg * (2<<16) + 0.5
+ -21746, -21746, -21746, -21746, // tg * (2<<16) + 0.5
};
static const int16_t ocos_4_16[4] ATTR_ALIGN(8) = {
- 23170, 23170, 23170, 23170, //cos * (2<<15) + 0.5
+ 23170, 23170, 23170, 23170, //cos * (2<<15) + 0.5
};
static const int64_t fdct_one_corr ATTR_ALIGN(8) = 0x0001000100010001LL;
@@ -351,62 +351,62 @@ static always_inline void fdct_col(const int16_t *in, int16_t *out, int offset)
static always_inline void fdct_row_sse2(const int16_t *in, int16_t *out)
{
asm volatile(
- ".macro FDCT_ROW_SSE2_H1 i t \n\t"
- "movq \\i(%0), %%xmm2 \n\t"
- "movq \\i+8(%0), %%xmm0 \n\t"
- "movdqa \\t+32(%1), %%xmm3 \n\t"
- "movdqa \\t+48(%1), %%xmm7 \n\t"
- "movdqa \\t(%1), %%xmm4 \n\t"
- "movdqa \\t+16(%1), %%xmm5 \n\t"
- ".endm \n\t"
- ".macro FDCT_ROW_SSE2_H2 i t \n\t"
- "movq \\i(%0), %%xmm2 \n\t"
- "movq \\i+8(%0), %%xmm0 \n\t"
- "movdqa \\t+32(%1), %%xmm3 \n\t"
- "movdqa \\t+48(%1), %%xmm7 \n\t"
- ".endm \n\t"
- ".macro FDCT_ROW_SSE2 i \n\t"
- "movq %%xmm2, %%xmm1 \n\t"
- "pshuflw $27, %%xmm0, %%xmm0 \n\t"
- "paddsw %%xmm0, %%xmm1 \n\t"
- "psubsw %%xmm0, %%xmm2 \n\t"
- "punpckldq %%xmm2, %%xmm1 \n\t"
- "pshufd $78, %%xmm1, %%xmm2 \n\t"
- "pmaddwd %%xmm2, %%xmm3 \n\t"
- "pmaddwd %%xmm1, %%xmm7 \n\t"
- "pmaddwd %%xmm5, %%xmm2 \n\t"
- "pmaddwd %%xmm4, %%xmm1 \n\t"
- "paddd %%xmm7, %%xmm3 \n\t"
- "paddd %%xmm2, %%xmm1 \n\t"
- "paddd %%xmm6, %%xmm3 \n\t"
- "paddd %%xmm6, %%xmm1 \n\t"
- "psrad %3, %%xmm3 \n\t"
- "psrad %3, %%xmm1 \n\t"
- "packssdw %%xmm3, %%xmm1 \n\t"
- "movdqa %%xmm1, \\i(%4) \n\t"
- ".endm \n\t"
- "movdqa (%2), %%xmm6 \n\t"
- "FDCT_ROW_SSE2_H1 0 0 \n\t"
- "FDCT_ROW_SSE2 0 \n\t"
- "FDCT_ROW_SSE2_H2 64 0 \n\t"
- "FDCT_ROW_SSE2 64 \n\t"
-
- "FDCT_ROW_SSE2_H1 16 64 \n\t"
- "FDCT_ROW_SSE2 16 \n\t"
- "FDCT_ROW_SSE2_H2 112 64 \n\t"
- "FDCT_ROW_SSE2 112 \n\t"
-
- "FDCT_ROW_SSE2_H1 32 128 \n\t"
- "FDCT_ROW_SSE2 32 \n\t"
- "FDCT_ROW_SSE2_H2 96 128 \n\t"
- "FDCT_ROW_SSE2 96 \n\t"
-
- "FDCT_ROW_SSE2_H1 48 192 \n\t"
- "FDCT_ROW_SSE2 48 \n\t"
- "FDCT_ROW_SSE2_H2 80 192 \n\t"
- "FDCT_ROW_SSE2 80 \n\t"
- :
- : "r" (in), "r" (tab_frw_01234567_sse2.tab_frw_01234567_sse2), "r" (fdct_r_row_sse2.fdct_r_row_sse2), "i" (SHIFT_FRW_ROW), "r" (out)
+ ".macro FDCT_ROW_SSE2_H1 i t \n\t"
+ "movq \\i(%0), %%xmm2 \n\t"
+ "movq \\i+8(%0), %%xmm0 \n\t"
+ "movdqa \\t+32(%1), %%xmm3 \n\t"
+ "movdqa \\t+48(%1), %%xmm7 \n\t"
+ "movdqa \\t(%1), %%xmm4 \n\t"
+ "movdqa \\t+16(%1), %%xmm5 \n\t"
+ ".endm \n\t"
+ ".macro FDCT_ROW_SSE2_H2 i t \n\t"
+ "movq \\i(%0), %%xmm2 \n\t"
+ "movq \\i+8(%0), %%xmm0 \n\t"
+ "movdqa \\t+32(%1), %%xmm3 \n\t"
+ "movdqa \\t+48(%1), %%xmm7 \n\t"
+ ".endm \n\t"
+ ".macro FDCT_ROW_SSE2 i \n\t"
+ "movq %%xmm2, %%xmm1 \n\t"
+ "pshuflw $27, %%xmm0, %%xmm0 \n\t"
+ "paddsw %%xmm0, %%xmm1 \n\t"
+ "psubsw %%xmm0, %%xmm2 \n\t"
+ "punpckldq %%xmm2, %%xmm1 \n\t"
+ "pshufd $78, %%xmm1, %%xmm2 \n\t"
+ "pmaddwd %%xmm2, %%xmm3 \n\t"
+ "pmaddwd %%xmm1, %%xmm7 \n\t"
+ "pmaddwd %%xmm5, %%xmm2 \n\t"
+ "pmaddwd %%xmm4, %%xmm1 \n\t"
+ "paddd %%xmm7, %%xmm3 \n\t"
+ "paddd %%xmm2, %%xmm1 \n\t"
+ "paddd %%xmm6, %%xmm3 \n\t"
+ "paddd %%xmm6, %%xmm1 \n\t"
+ "psrad %3, %%xmm3 \n\t"
+ "psrad %3, %%xmm1 \n\t"
+ "packssdw %%xmm3, %%xmm1 \n\t"
+ "movdqa %%xmm1, \\i(%4) \n\t"
+ ".endm \n\t"
+ "movdqa (%2), %%xmm6 \n\t"
+ "FDCT_ROW_SSE2_H1 0 0 \n\t"
+ "FDCT_ROW_SSE2 0 \n\t"
+ "FDCT_ROW_SSE2_H2 64 0 \n\t"
+ "FDCT_ROW_SSE2 64 \n\t"
+
+ "FDCT_ROW_SSE2_H1 16 64 \n\t"
+ "FDCT_ROW_SSE2 16 \n\t"
+ "FDCT_ROW_SSE2_H2 112 64 \n\t"
+ "FDCT_ROW_SSE2 112 \n\t"
+
+ "FDCT_ROW_SSE2_H1 32 128 \n\t"
+ "FDCT_ROW_SSE2 32 \n\t"
+ "FDCT_ROW_SSE2_H2 96 128 \n\t"
+ "FDCT_ROW_SSE2 96 \n\t"
+
+ "FDCT_ROW_SSE2_H1 48 192 \n\t"
+ "FDCT_ROW_SSE2 48 \n\t"
+ "FDCT_ROW_SSE2_H2 80 192 \n\t"
+ "FDCT_ROW_SSE2 80 \n\t"
+ :
+ : "r" (in), "r" (tab_frw_01234567_sse2.tab_frw_01234567_sse2), "r" (fdct_r_row_sse2.fdct_r_row_sse2), "i" (SHIFT_FRW_ROW), "r" (out)
);
}
diff --git a/libavcodec/i386/fft_sse.c b/libavcodec/i386/fft_sse.c
index f8be644a3b..ce48c1040b 100644
--- a/libavcodec/i386/fft_sse.c
+++ b/libavcodec/i386/fft_sse.c
@@ -45,8 +45,8 @@ static void print_v4sf(const char *str, __m128 a)
void ff_fft_calc_sse(FFTContext *s, FFTComplex *z)
{
int ln = s->nbits;
- int j, np, np2;
- int nblocks, nloops;
+ int j, np, np2;
+ int nblocks, nloops;
register FFTComplex *p, *q;
FFTComplex *cptr, *cptr1;
int k;
diff --git a/libavcodec/i386/h264dsp_mmx.c b/libavcodec/i386/h264dsp_mmx.c
index 45a3c02f35..5356b9c23f 100644
--- a/libavcodec/i386/h264dsp_mmx.c
+++ b/libavcodec/i386/h264dsp_mmx.c
@@ -47,9 +47,9 @@
SUMSUB_BADC( d13, s02, s13, d02 )
#define SBUTTERFLY(a,b,t,n)\
- "movq " #a ", " #t " \n\t" /* abcd */\
- "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\
- "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\
+ "movq " #a ", " #t " \n\t" /* abcd */\
+ "punpckl" #n " " #b ", " #a " \n\t" /* aebf */\
+ "punpckh" #n " " #b ", " #t " \n\t" /* cgdh */\
#define TRANSPOSE4(a,b,c,d,t)\
SBUTTERFLY(a,b,t,wd) /* a=aebf t=cgdh */\
@@ -369,73 +369,73 @@ static void h264_h_loop_filter_chroma_intra_mmx2(uint8_t *pix, int stride, int a
/* motion compensation */
#define QPEL_H264V(A,B,C,D,E,F,OP)\
- "movd (%0), "#F" \n\t"\
- "movq "#C", %%mm6 \n\t"\
- "paddw "#D", %%mm6 \n\t"\
- "psllw $2, %%mm6 \n\t"\
- "psubw "#B", %%mm6 \n\t"\
- "psubw "#E", %%mm6 \n\t"\
- "pmullw %4, %%mm6 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, "#F" \n\t"\
- "paddw %5, "#A" \n\t"\
- "paddw "#F", "#A" \n\t"\
- "paddw "#A", %%mm6 \n\t"\
- "psraw $5, %%mm6 \n\t"\
- "packuswb %%mm6, %%mm6 \n\t"\
+ "movd (%0), "#F" \n\t"\
+ "movq "#C", %%mm6 \n\t"\
+ "paddw "#D", %%mm6 \n\t"\
+ "psllw $2, %%mm6 \n\t"\
+ "psubw "#B", %%mm6 \n\t"\
+ "psubw "#E", %%mm6 \n\t"\
+ "pmullw %4, %%mm6 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, "#F" \n\t"\
+ "paddw %5, "#A" \n\t"\
+ "paddw "#F", "#A" \n\t"\
+ "paddw "#A", %%mm6 \n\t"\
+ "psraw $5, %%mm6 \n\t"\
+ "packuswb %%mm6, %%mm6 \n\t"\
OP(%%mm6, (%1), A, d)\
- "add %3, %1 \n\t"
+ "add %3, %1 \n\t"
#define QPEL_H264HV(A,B,C,D,E,F,OF)\
- "movd (%0), "#F" \n\t"\
- "movq "#C", %%mm6 \n\t"\
- "paddw "#D", %%mm6 \n\t"\
- "psllw $2, %%mm6 \n\t"\
- "psubw "#B", %%mm6 \n\t"\
- "psubw "#E", %%mm6 \n\t"\
- "pmullw %3, %%mm6 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, "#F" \n\t"\
- "paddw "#F", "#A" \n\t"\
- "paddw "#A", %%mm6 \n\t"\
- "movq %%mm6, "#OF"(%1) \n\t"
+ "movd (%0), "#F" \n\t"\
+ "movq "#C", %%mm6 \n\t"\
+ "paddw "#D", %%mm6 \n\t"\
+ "psllw $2, %%mm6 \n\t"\
+ "psubw "#B", %%mm6 \n\t"\
+ "psubw "#E", %%mm6 \n\t"\
+ "pmullw %3, %%mm6 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, "#F" \n\t"\
+ "paddw "#F", "#A" \n\t"\
+ "paddw "#A", %%mm6 \n\t"\
+ "movq %%mm6, "#OF"(%1) \n\t"
#define QPEL_H264(OPNAME, OP, MMX)\
static void OPNAME ## h264_qpel4_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
int h=4;\
\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movq %5, %%mm4 \n\t"\
- "movq %6, %%mm5 \n\t"\
- "1: \n\t"\
- "movd -1(%0), %%mm1 \n\t"\
- "movd (%0), %%mm2 \n\t"\
- "movd 1(%0), %%mm3 \n\t"\
- "movd 2(%0), %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "paddw %%mm0, %%mm1 \n\t"\
- "paddw %%mm3, %%mm2 \n\t"\
- "movd -2(%0), %%mm0 \n\t"\
- "movd 3(%0), %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "paddw %%mm3, %%mm0 \n\t"\
- "psllw $2, %%mm2 \n\t"\
- "psubw %%mm1, %%mm2 \n\t"\
- "pmullw %%mm4, %%mm2 \n\t"\
- "paddw %%mm5, %%mm0 \n\t"\
- "paddw %%mm2, %%mm0 \n\t"\
- "psraw $5, %%mm0 \n\t"\
- "packuswb %%mm0, %%mm0 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movq %5, %%mm4 \n\t"\
+ "movq %6, %%mm5 \n\t"\
+ "1: \n\t"\
+ "movd -1(%0), %%mm1 \n\t"\
+ "movd (%0), %%mm2 \n\t"\
+ "movd 1(%0), %%mm3 \n\t"\
+ "movd 2(%0), %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "paddw %%mm0, %%mm1 \n\t"\
+ "paddw %%mm3, %%mm2 \n\t"\
+ "movd -2(%0), %%mm0 \n\t"\
+ "movd 3(%0), %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "paddw %%mm3, %%mm0 \n\t"\
+ "psllw $2, %%mm2 \n\t"\
+ "psubw %%mm1, %%mm2 \n\t"\
+ "pmullw %%mm4, %%mm2 \n\t"\
+ "paddw %%mm5, %%mm0 \n\t"\
+ "paddw %%mm2, %%mm0 \n\t"\
+ "psraw $5, %%mm0 \n\t"\
+ "packuswb %%mm0, %%mm0 \n\t"\
OP(%%mm0, (%1),%%mm6, d)\
- "add %3, %0 \n\t"\
- "add %4, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add %3, %0 \n\t"\
+ "add %4, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
: "d"((long)srcStride), "S"((long)dstStride), "m"(ff_pw_5), "m"(ff_pw_16)\
: "memory"\
@@ -444,22 +444,22 @@ static void OPNAME ## h264_qpel4_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, i
static void OPNAME ## h264_qpel4_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
src -= 2*srcStride;\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movd (%0), %%mm0 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm1 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm2 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm3 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm4 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movd (%0), %%mm0 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm1 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm2 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm3 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm4 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
QPEL_H264V(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\
QPEL_H264V(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\
QPEL_H264V(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\
@@ -476,22 +476,22 @@ static void OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp,
src -= 2*srcStride+2;\
while(w--){\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movd (%0), %%mm0 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm1 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm2 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm3 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm4 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movd (%0), %%mm0 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm1 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm2 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm3 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm4 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
QPEL_H264HV(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, 0*8*3)\
QPEL_H264HV(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, 1*8*3)\
QPEL_H264HV(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, 2*8*3)\
@@ -506,28 +506,28 @@ static void OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp,
}\
tmp -= 3*4;\
asm volatile(\
- "movq %4, %%mm6 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "paddw 10(%0), %%mm0 \n\t"\
- "movq 2(%0), %%mm1 \n\t"\
- "paddw 8(%0), %%mm1 \n\t"\
- "movq 4(%0), %%mm2 \n\t"\
- "paddw 6(%0), %%mm2 \n\t"\
- "psubw %%mm1, %%mm0 \n\t"/*a-b (abccba)*/\
- "psraw $2, %%mm0 \n\t"/*(a-b)/4 */\
- "psubw %%mm1, %%mm0 \n\t"/*(a-b)/4-b */\
- "paddsw %%mm2, %%mm0 \n\t"\
- "psraw $2, %%mm0 \n\t"/*((a-b)/4-b)/4 */\
- "paddw %%mm6, %%mm2 \n\t"\
- "paddw %%mm2, %%mm0 \n\t"\
- "psraw $6, %%mm0 \n\t"\
- "packuswb %%mm0, %%mm0 \n\t"\
+ "movq %4, %%mm6 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "paddw 10(%0), %%mm0 \n\t"\
+ "movq 2(%0), %%mm1 \n\t"\
+ "paddw 8(%0), %%mm1 \n\t"\
+ "movq 4(%0), %%mm2 \n\t"\
+ "paddw 6(%0), %%mm2 \n\t"\
+ "psubw %%mm1, %%mm0 \n\t"/*a-b (abccba)*/\
+ "psraw $2, %%mm0 \n\t"/*(a-b)/4 */\
+ "psubw %%mm1, %%mm0 \n\t"/*(a-b)/4-b */\
+ "paddsw %%mm2, %%mm0 \n\t"\
+ "psraw $2, %%mm0 \n\t"/*((a-b)/4-b)/4 */\
+ "paddw %%mm6, %%mm2 \n\t"\
+ "paddw %%mm2, %%mm0 \n\t"\
+ "psraw $6, %%mm0 \n\t"\
+ "packuswb %%mm0, %%mm0 \n\t"\
OP(%%mm0, (%1),%%mm7, d)\
- "add $24, %0 \n\t"\
- "add %3, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add $24, %0 \n\t"\
+ "add %3, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(tmp), "+c"(dst), "+m"(h)\
: "S"((long)dstStride), "m"(ff_pw_32)\
: "memory"\
@@ -537,54 +537,54 @@ static void OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp,
static void OPNAME ## h264_qpel8_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
int h=8;\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movq %5, %%mm6 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq 1(%0), %%mm2 \n\t"\
- "movq %%mm0, %%mm1 \n\t"\
- "movq %%mm2, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpckhbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpckhbw %%mm7, %%mm3 \n\t"\
- "paddw %%mm2, %%mm0 \n\t"\
- "paddw %%mm3, %%mm1 \n\t"\
- "psllw $2, %%mm0 \n\t"\
- "psllw $2, %%mm1 \n\t"\
- "movq -1(%0), %%mm2 \n\t"\
- "movq 2(%0), %%mm4 \n\t"\
- "movq %%mm2, %%mm3 \n\t"\
- "movq %%mm4, %%mm5 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpckhbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
- "punpckhbw %%mm7, %%mm5 \n\t"\
- "paddw %%mm4, %%mm2 \n\t"\
- "paddw %%mm3, %%mm5 \n\t"\
- "psubw %%mm2, %%mm0 \n\t"\
- "psubw %%mm5, %%mm1 \n\t"\
- "pmullw %%mm6, %%mm0 \n\t"\
- "pmullw %%mm6, %%mm1 \n\t"\
- "movd -2(%0), %%mm2 \n\t"\
- "movd 7(%0), %%mm5 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm5 \n\t"\
- "paddw %%mm3, %%mm2 \n\t"\
- "paddw %%mm5, %%mm4 \n\t"\
- "movq %6, %%mm5 \n\t"\
- "paddw %%mm5, %%mm2 \n\t"\
- "paddw %%mm5, %%mm4 \n\t"\
- "paddw %%mm2, %%mm0 \n\t"\
- "paddw %%mm4, %%mm1 \n\t"\
- "psraw $5, %%mm0 \n\t"\
- "psraw $5, %%mm1 \n\t"\
- "packuswb %%mm1, %%mm0 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movq %5, %%mm6 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 1(%0), %%mm2 \n\t"\
+ "movq %%mm0, %%mm1 \n\t"\
+ "movq %%mm2, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpckhbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpckhbw %%mm7, %%mm3 \n\t"\
+ "paddw %%mm2, %%mm0 \n\t"\
+ "paddw %%mm3, %%mm1 \n\t"\
+ "psllw $2, %%mm0 \n\t"\
+ "psllw $2, %%mm1 \n\t"\
+ "movq -1(%0), %%mm2 \n\t"\
+ "movq 2(%0), %%mm4 \n\t"\
+ "movq %%mm2, %%mm3 \n\t"\
+ "movq %%mm4, %%mm5 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpckhbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
+ "punpckhbw %%mm7, %%mm5 \n\t"\
+ "paddw %%mm4, %%mm2 \n\t"\
+ "paddw %%mm3, %%mm5 \n\t"\
+ "psubw %%mm2, %%mm0 \n\t"\
+ "psubw %%mm5, %%mm1 \n\t"\
+ "pmullw %%mm6, %%mm0 \n\t"\
+ "pmullw %%mm6, %%mm1 \n\t"\
+ "movd -2(%0), %%mm2 \n\t"\
+ "movd 7(%0), %%mm5 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm5 \n\t"\
+ "paddw %%mm3, %%mm2 \n\t"\
+ "paddw %%mm5, %%mm4 \n\t"\
+ "movq %6, %%mm5 \n\t"\
+ "paddw %%mm5, %%mm2 \n\t"\
+ "paddw %%mm5, %%mm4 \n\t"\
+ "paddw %%mm2, %%mm0 \n\t"\
+ "paddw %%mm4, %%mm1 \n\t"\
+ "psraw $5, %%mm0 \n\t"\
+ "psraw $5, %%mm1 \n\t"\
+ "packuswb %%mm1, %%mm0 \n\t"\
OP(%%mm0, (%1),%%mm5, q)\
- "add %3, %0 \n\t"\
- "add %4, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add %3, %0 \n\t"\
+ "add %4, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(src), "+c"(dst), "+m"(h)\
: "d"((long)srcStride), "S"((long)dstStride), "m"(ff_pw_5), "m"(ff_pw_16)\
: "memory"\
@@ -597,22 +597,22 @@ static void OPNAME ## h264_qpel8_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, i
\
while(h--){\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movd (%0), %%mm0 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm1 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm2 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm3 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm4 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movd (%0), %%mm0 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm1 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm2 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm3 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm4 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
QPEL_H264V(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\
QPEL_H264V(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\
QPEL_H264V(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\
@@ -636,22 +636,22 @@ static void OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp,
src -= 2*srcStride+2;\
while(w--){\
asm volatile(\
- "pxor %%mm7, %%mm7 \n\t"\
- "movd (%0), %%mm0 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm1 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm2 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm3 \n\t"\
- "add %2, %0 \n\t"\
- "movd (%0), %%mm4 \n\t"\
- "add %2, %0 \n\t"\
- "punpcklbw %%mm7, %%mm0 \n\t"\
- "punpcklbw %%mm7, %%mm1 \n\t"\
- "punpcklbw %%mm7, %%mm2 \n\t"\
- "punpcklbw %%mm7, %%mm3 \n\t"\
- "punpcklbw %%mm7, %%mm4 \n\t"\
+ "pxor %%mm7, %%mm7 \n\t"\
+ "movd (%0), %%mm0 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm1 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm2 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm3 \n\t"\
+ "add %2, %0 \n\t"\
+ "movd (%0), %%mm4 \n\t"\
+ "add %2, %0 \n\t"\
+ "punpcklbw %%mm7, %%mm0 \n\t"\
+ "punpcklbw %%mm7, %%mm1 \n\t"\
+ "punpcklbw %%mm7, %%mm2 \n\t"\
+ "punpcklbw %%mm7, %%mm3 \n\t"\
+ "punpcklbw %%mm7, %%mm4 \n\t"\
QPEL_H264HV(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, 0*8*4)\
QPEL_H264HV(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, 1*8*4)\
QPEL_H264HV(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, 2*8*4)\
@@ -670,42 +670,42 @@ static void OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp,
}\
tmp -= 4*4;\
asm volatile(\
- "movq %4, %%mm6 \n\t"\
- "1: \n\t"\
- "movq (%0), %%mm0 \n\t"\
- "movq 8(%0), %%mm3 \n\t"\
- "movq 2(%0), %%mm1 \n\t"\
- "movq 10(%0), %%mm4 \n\t"\
- "paddw %%mm4, %%mm0 \n\t"\
- "paddw %%mm3, %%mm1 \n\t"\
- "paddw 18(%0), %%mm3 \n\t"\
- "paddw 16(%0), %%mm4 \n\t"\
- "movq 4(%0), %%mm2 \n\t"\
- "movq 12(%0), %%mm5 \n\t"\
- "paddw 6(%0), %%mm2 \n\t"\
- "paddw 14(%0), %%mm5 \n\t"\
- "psubw %%mm1, %%mm0 \n\t"\
- "psubw %%mm4, %%mm3 \n\t"\
- "psraw $2, %%mm0 \n\t"\
- "psraw $2, %%mm3 \n\t"\
- "psubw %%mm1, %%mm0 \n\t"\
- "psubw %%mm4, %%mm3 \n\t"\
- "paddsw %%mm2, %%mm0 \n\t"\
- "paddsw %%mm5, %%mm3 \n\t"\
- "psraw $2, %%mm0 \n\t"\
- "psraw $2, %%mm3 \n\t"\
- "paddw %%mm6, %%mm2 \n\t"\
- "paddw %%mm6, %%mm5 \n\t"\
- "paddw %%mm2, %%mm0 \n\t"\
- "paddw %%mm5, %%mm3 \n\t"\
- "psraw $6, %%mm0 \n\t"\
- "psraw $6, %%mm3 \n\t"\
- "packuswb %%mm3, %%mm0 \n\t"\
+ "movq %4, %%mm6 \n\t"\
+ "1: \n\t"\
+ "movq (%0), %%mm0 \n\t"\
+ "movq 8(%0), %%mm3 \n\t"\
+ "movq 2(%0), %%mm1 \n\t"\
+ "movq 10(%0), %%mm4 \n\t"\
+ "paddw %%mm4, %%mm0 \n\t"\
+ "paddw %%mm3, %%mm1 \n\t"\
+ "paddw 18(%0), %%mm3 \n\t"\
+ "paddw 16(%0), %%mm4 \n\t"\
+ "movq 4(%0), %%mm2 \n\t"\
+ "movq 12(%0), %%mm5 \n\t"\
+ "paddw 6(%0), %%mm2 \n\t"\
+ "paddw 14(%0), %%mm5 \n\t"\
+ "psubw %%mm1, %%mm0 \n\t"\
+ "psubw %%mm4, %%mm3 \n\t"\
+ "psraw $2, %%mm0 \n\t"\
+ "psraw $2, %%mm3 \n\t"\
+ "psubw %%mm1, %%mm0 \n\t"\
+ "psubw %%mm4, %%mm3 \n\t"\
+ "paddsw %%mm2, %%mm0 \n\t"\
+ "paddsw %%mm5, %%mm3 \n\t"\
+ "psraw $2, %%mm0 \n\t"\
+ "psraw $2, %%mm3 \n\t"\
+ "paddw %%mm6, %%mm2 \n\t"\
+ "paddw %%mm6, %%mm5 \n\t"\
+ "paddw %%mm2, %%mm0 \n\t"\
+ "paddw %%mm5, %%mm3 \n\t"\
+ "psraw $6, %%mm0 \n\t"\
+ "psraw $6, %%mm3 \n\t"\
+ "packuswb %%mm3, %%mm0 \n\t"\
OP(%%mm0, (%1),%%mm7, q)\
- "add $32, %0 \n\t"\
- "add %3, %1 \n\t"\
- "decl %2 \n\t"\
- " jnz 1b \n\t"\
+ "add $32, %0 \n\t"\
+ "add %3, %1 \n\t"\
+ "decl %2 \n\t"\
+ " jnz 1b \n\t"\
: "+a"(tmp), "+c"(dst), "+m"(h)\
: "S"((long)dstStride), "m"(ff_pw_32)\
: "memory"\
@@ -862,15 +862,15 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## MMX(uint8_t *dst, uint8_t *
}\
-#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t"
+#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b " \n\t"
#define AVG_3DNOW_OP(a,b,temp, size) \
-"mov" #size " " #b ", " #temp " \n\t"\
-"pavgusb " #temp ", " #a " \n\t"\
-"mov" #size " " #a ", " #b " \n\t"
+"mov" #size " " #b ", " #temp " \n\t"\
+"pavgusb " #temp ", " #a " \n\t"\
+"mov" #size " " #a ", " #b " \n\t"
#define AVG_MMX2_OP(a,b,temp, size) \
-"mov" #size " " #b ", " #temp " \n\t"\
-"pavgb " #temp ", " #a " \n\t"\
-"mov" #size " " #a ", " #b " \n\t"
+"mov" #size " " #b ", " #temp " \n\t"\
+"pavgb " #temp ", " #a " \n\t"\
+"mov" #size " " #a ", " #b " \n\t"
QPEL_H264(put_, PUT_OP, 3dnow)
QPEL_H264(avg_, AVG_3DNOW_OP, 3dnow)
diff --git a/libavcodec/i386/idct_mmx.c b/libavcodec/i386/idct_mmx.c
index d1a84549d0..2b884fd832 100644
--- a/libavcodec/i386/idct_mmx.c
+++ b/libavcodec/i386/idct_mmx.c
@@ -38,7 +38,7 @@
#if 0
/* C row IDCT - its just here to document the MMXEXT and MMX versions */
static inline void idct_row (int16_t * row, int offset,
- int16_t * table, int32_t * rounder)
+ int16_t * table, int32_t * rounder)
{
int C1, C2, C3, C4, C5, C6, C7;
int a0, a1, a2, a3, b0, b1, b2, b3;
@@ -77,241 +77,241 @@ static inline void idct_row (int16_t * row, int offset,
/* MMXEXT row IDCT */
-#define mmxext_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, -c4, -c2, \
- c4, c6, c4, c6, \
- c1, c3, -c1, -c5, \
- c5, c7, c3, -c7, \
- c4, -c6, c4, -c6, \
- -c4, c2, c4, -c2, \
- c5, -c1, c3, -c1, \
- c7, c3, c7, -c5 }
+#define mmxext_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, -c4, -c2, \
+ c4, c6, c4, c6, \
+ c1, c3, -c1, -c5, \
+ c5, c7, c3, -c7, \
+ c4, -c6, c4, -c6, \
+ -c4, c2, c4, -c2, \
+ c5, -c1, c3, -c1, \
+ c7, c3, c7, -c5 }
static inline void mmxext_row_head (int16_t * row, int offset, const int16_t * table)
{
- movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
+ movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
- movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
- movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
+ movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
+ movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
- movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4
- movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
+ movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4
+ movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
- movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4
- pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2
+ movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4
+ pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2
- pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4
+ pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4
}
static inline void mmxext_row (const int16_t * table, const int32_t * rounder)
{
- movq_m2r (*(table+8), mm1); // mm1 = -C5 -C1 C3 C1
- pmaddwd_r2r (mm2, mm4); // mm4 = C4*x0+C6*x2 C4*x4+C6*x6
+ movq_m2r (*(table+8), mm1); // mm1 = -C5 -C1 C3 C1
+ pmaddwd_r2r (mm2, mm4); // mm4 = C4*x0+C6*x2 C4*x4+C6*x6
- pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x4-C6*x6 C4*x0-C6*x2
- pshufw_r2r (mm6, mm6, 0x4e); // mm6 = x3 x1 x7 x5
+ pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x4-C6*x6 C4*x0-C6*x2
+ pshufw_r2r (mm6, mm6, 0x4e); // mm6 = x3 x1 x7 x5
- movq_m2r (*(table+12), mm7); // mm7 = -C7 C3 C7 C5
- pmaddwd_r2r (mm5, mm1); // mm1 = -C1*x5-C5*x7 C1*x1+C3*x3
+ movq_m2r (*(table+12), mm7); // mm7 = -C7 C3 C7 C5
+ pmaddwd_r2r (mm5, mm1); // mm1 = -C1*x5-C5*x7 C1*x1+C3*x3
- paddd_m2r (*rounder, mm3); // mm3 += rounder
- pmaddwd_r2r (mm6, mm7); // mm7 = C3*x1-C7*x3 C5*x5+C7*x7
+ paddd_m2r (*rounder, mm3); // mm3 += rounder
+ pmaddwd_r2r (mm6, mm7); // mm7 = C3*x1-C7*x3 C5*x5+C7*x7
- pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x0-C2*x2 -C4*x4+C2*x6
- paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder
+ pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x0-C2*x2 -C4*x4+C2*x6
+ paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder
- pmaddwd_m2r (*(table+24), mm5); // mm5 = C3*x5-C1*x7 C5*x1-C1*x3
- movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder
+ pmaddwd_m2r (*(table+24), mm5); // mm5 = C3*x5-C1*x7 C5*x1-C1*x3
+ movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder
- pmaddwd_m2r (*(table+28), mm6); // mm6 = C7*x1-C5*x3 C7*x5+C3*x7
- paddd_r2r (mm7, mm1); // mm1 = b1 b0
+ pmaddwd_m2r (*(table+28), mm6); // mm6 = C7*x1-C5*x3 C7*x5+C3*x7
+ paddd_r2r (mm7, mm1); // mm1 = b1 b0
- paddd_m2r (*rounder, mm0); // mm0 += rounder
- psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder
+ paddd_m2r (*rounder, mm0); // mm0 += rounder
+ psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder
- psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7
- paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder
+ psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7
+ paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder
- paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder
- psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0
+ paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder
+ psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0
- paddd_r2r (mm6, mm5); // mm5 = b3 b2
- movq_r2r (mm0, mm4); // mm4 = a3 a2 + rounder
+ paddd_r2r (mm6, mm5); // mm5 = b3 b2
+ movq_r2r (mm0, mm4); // mm4 = a3 a2 + rounder
- paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder
- psubd_r2r (mm5, mm4); // mm4 = a3-b3 a2-b2 + rounder
+ paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder
+ psubd_r2r (mm5, mm4); // mm4 = a3-b3 a2-b2 + rounder
}
static inline void mmxext_row_tail (int16_t * row, int store)
{
- psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
+ psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
- psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5
+ psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5
- packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
+ packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
- packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5
+ packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5
- movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
- pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4
+ movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
+ pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4
/* slot */
- movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4
+ movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4
}
static inline void mmxext_row_mid (int16_t * row, int store,
- int offset, const int16_t * table)
+ int offset, const int16_t * table)
{
- movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
- psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
+ movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
+ psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
- movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
- psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5
+ movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
+ psrad_i2r (ROW_SHIFT, mm4); // mm4 = y4 y5
- packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
- movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
+ packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
+ movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
- packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5
- movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
+ packssdw_r2r (mm3, mm4); // mm4 = y6 y7 y4 y5
+ movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
- movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
- pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4
+ movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
+ pshufw_r2r (mm4, mm4, 0xb1); // mm4 = y7 y6 y5 y4
- movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4
- movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4
+ movq_m2r (*table, mm3); // mm3 = -C2 -C4 C2 C4
+ movq_r2m (mm4, *(row+store+4)); // save y7 y6 y5 y4
- pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2
+ pmaddwd_r2r (mm0, mm3); // mm3 = -C4*x4-C2*x6 C4*x0+C2*x2
- movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4
- pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4
+ movq_m2r (*(table+4), mm4); // mm4 = C6 C4 C6 C4
+ pshufw_r2r (mm2, mm2, 0x4e); // mm2 = x2 x0 x6 x4
}
/* MMX row IDCT */
-#define mmx_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, c4, c6, \
- c4, c6, -c4, -c2, \
- c1, c3, c3, -c7, \
- c5, c7, -c1, -c5, \
- c4, -c6, c4, -c2, \
- -c4, c2, c4, -c6, \
- c5, -c1, c7, -c5, \
- c7, c3, c3, -c1 }
+#define mmx_table(c1,c2,c3,c4,c5,c6,c7) { c4, c2, c4, c6, \
+ c4, c6, -c4, -c2, \
+ c1, c3, c3, -c7, \
+ c5, c7, -c1, -c5, \
+ c4, -c6, c4, -c2, \
+ -c4, c2, c4, -c6, \
+ c5, -c1, c7, -c5, \
+ c7, c3, c3, -c1 }
static inline void mmx_row_head (int16_t * row, int offset, const int16_t * table)
{
- movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
+ movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
- movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
- movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
+ movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
+ movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
- movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4
- movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
+ movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4
+ movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
- punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0
+ punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0
- movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4
- pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2
+ movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4
+ pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2
- movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1
- punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4
+ movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1
+ punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4
}
static inline void mmx_row (const int16_t * table, const int32_t * rounder)
{
- pmaddwd_r2r (mm2, mm4); // mm4 = -C4*x4-C2*x6 C4*x4+C6*x6
- punpckldq_r2r (mm5, mm5); // mm5 = x3 x1 x3 x1
+ pmaddwd_r2r (mm2, mm4); // mm4 = -C4*x4-C2*x6 C4*x4+C6*x6
+ punpckldq_r2r (mm5, mm5); // mm5 = x3 x1 x3 x1
- pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x0-C2*x2 C4*x0-C6*x2
- punpckhdq_r2r (mm6, mm6); // mm6 = x7 x5 x7 x5
+ pmaddwd_m2r (*(table+16), mm0); // mm0 = C4*x0-C2*x2 C4*x0-C6*x2
+ punpckhdq_r2r (mm6, mm6); // mm6 = x7 x5 x7 x5
- movq_m2r (*(table+12), mm7); // mm7 = -C5 -C1 C7 C5
- pmaddwd_r2r (mm5, mm1); // mm1 = C3*x1-C7*x3 C1*x1+C3*x3
+ movq_m2r (*(table+12), mm7); // mm7 = -C5 -C1 C7 C5
+ pmaddwd_r2r (mm5, mm1); // mm1 = C3*x1-C7*x3 C1*x1+C3*x3
- paddd_m2r (*rounder, mm3); // mm3 += rounder
- pmaddwd_r2r (mm6, mm7); // mm7 = -C1*x5-C5*x7 C5*x5+C7*x7
+ paddd_m2r (*rounder, mm3); // mm3 += rounder
+ pmaddwd_r2r (mm6, mm7); // mm7 = -C1*x5-C5*x7 C5*x5+C7*x7
- pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x4-C6*x6 -C4*x4+C2*x6
- paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder
+ pmaddwd_m2r (*(table+20), mm2); // mm2 = C4*x4-C6*x6 -C4*x4+C2*x6
+ paddd_r2r (mm4, mm3); // mm3 = a1 a0 + rounder
- pmaddwd_m2r (*(table+24), mm5); // mm5 = C7*x1-C5*x3 C5*x1-C1*x3
- movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder
+ pmaddwd_m2r (*(table+24), mm5); // mm5 = C7*x1-C5*x3 C5*x1-C1*x3
+ movq_r2r (mm3, mm4); // mm4 = a1 a0 + rounder
- pmaddwd_m2r (*(table+28), mm6); // mm6 = C3*x5-C1*x7 C7*x5+C3*x7
- paddd_r2r (mm7, mm1); // mm1 = b1 b0
+ pmaddwd_m2r (*(table+28), mm6); // mm6 = C3*x5-C1*x7 C7*x5+C3*x7
+ paddd_r2r (mm7, mm1); // mm1 = b1 b0
- paddd_m2r (*rounder, mm0); // mm0 += rounder
- psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder
+ paddd_m2r (*rounder, mm0); // mm0 += rounder
+ psubd_r2r (mm1, mm3); // mm3 = a1-b1 a0-b0 + rounder
- psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7
- paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder
+ psrad_i2r (ROW_SHIFT, mm3); // mm3 = y6 y7
+ paddd_r2r (mm4, mm1); // mm1 = a1+b1 a0+b0 + rounder
- paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder
- psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0
+ paddd_r2r (mm2, mm0); // mm0 = a3 a2 + rounder
+ psrad_i2r (ROW_SHIFT, mm1); // mm1 = y1 y0
- paddd_r2r (mm6, mm5); // mm5 = b3 b2
- movq_r2r (mm0, mm7); // mm7 = a3 a2 + rounder
+ paddd_r2r (mm6, mm5); // mm5 = b3 b2
+ movq_r2r (mm0, mm7); // mm7 = a3 a2 + rounder
- paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder
- psubd_r2r (mm5, mm7); // mm7 = a3-b3 a2-b2 + rounder
+ paddd_r2r (mm5, mm0); // mm0 = a3+b3 a2+b2 + rounder
+ psubd_r2r (mm5, mm7); // mm7 = a3-b3 a2-b2 + rounder
}
static inline void mmx_row_tail (int16_t * row, int store)
{
- psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
+ psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
- psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5
+ psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5
- packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
+ packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
- packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5
+ packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5
- movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
- movq_r2r (mm7, mm4); // mm4 = y6 y7 y4 y5
+ movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
+ movq_r2r (mm7, mm4); // mm4 = y6 y7 y4 y5
- pslld_i2r (16, mm7); // mm7 = y7 0 y5 0
+ pslld_i2r (16, mm7); // mm7 = y7 0 y5 0
- psrld_i2r (16, mm4); // mm4 = 0 y6 0 y4
+ psrld_i2r (16, mm4); // mm4 = 0 y6 0 y4
- por_r2r (mm4, mm7); // mm7 = y7 y6 y5 y4
+ por_r2r (mm4, mm7); // mm7 = y7 y6 y5 y4
/* slot */
- movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4
+ movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4
}
static inline void mmx_row_mid (int16_t * row, int store,
- int offset, const int16_t * table)
+ int offset, const int16_t * table)
{
- movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
- psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
+ movq_m2r (*(row+offset), mm2); // mm2 = x6 x4 x2 x0
+ psrad_i2r (ROW_SHIFT, mm0); // mm0 = y3 y2
- movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
- psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5
+ movq_m2r (*(row+offset+4), mm5); // mm5 = x7 x5 x3 x1
+ psrad_i2r (ROW_SHIFT, mm7); // mm7 = y4 y5
- packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
- movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
+ packssdw_r2r (mm0, mm1); // mm1 = y3 y2 y1 y0
+ movq_r2r (mm5, mm6); // mm6 = x7 x5 x3 x1
- packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5
- movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
+ packssdw_r2r (mm3, mm7); // mm7 = y6 y7 y4 y5
+ movq_r2r (mm2, mm0); // mm0 = x6 x4 x2 x0
- movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
- movq_r2r (mm7, mm1); // mm1 = y6 y7 y4 y5
+ movq_r2m (mm1, *(row+store)); // save y3 y2 y1 y0
+ movq_r2r (mm7, mm1); // mm1 = y6 y7 y4 y5
- punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0
- psrld_i2r (16, mm7); // mm7 = 0 y6 0 y4
+ punpckldq_r2r (mm0, mm0); // mm0 = x2 x0 x2 x0
+ psrld_i2r (16, mm7); // mm7 = 0 y6 0 y4
- movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4
- pslld_i2r (16, mm1); // mm1 = y7 0 y5 0
+ movq_m2r (*table, mm3); // mm3 = C6 C4 C2 C4
+ pslld_i2r (16, mm1); // mm1 = y7 0 y5 0
- movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4
- por_r2r (mm1, mm7); // mm7 = y7 y6 y5 y4
+ movq_m2r (*(table+4), mm4); // mm4 = -C2 -C4 C6 C4
+ por_r2r (mm1, mm7); // mm7 = y7 y6 y5 y4
- movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1
- punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4
+ movq_m2r (*(table+8), mm1); // mm1 = -C7 C3 C3 C1
+ punpckhdq_r2r (mm2, mm2); // mm2 = x6 x4 x6 x4
- movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4
- pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2
+ movq_r2m (mm7, *(row+store+4)); // save y7 y6 y5 y4
+ pmaddwd_r2r (mm0, mm3); // mm3 = C4*x0+C6*x2 C4*x0+C2*x2
}
@@ -403,132 +403,132 @@ static inline void idct_col (int16_t * col, int offset)
/* column code adapted from peter gubanov */
/* http://www.elecard.com/peter/idct.shtml */
- movq_m2r (*_T1, mm0); // mm0 = T1
+ movq_m2r (*_T1, mm0); // mm0 = T1
- movq_m2r (*(col+offset+1*8), mm1); // mm1 = x1
- movq_r2r (mm0, mm2); // mm2 = T1
+ movq_m2r (*(col+offset+1*8), mm1); // mm1 = x1
+ movq_r2r (mm0, mm2); // mm2 = T1
- movq_m2r (*(col+offset+7*8), mm4); // mm4 = x7
- pmulhw_r2r (mm1, mm0); // mm0 = T1*x1
+ movq_m2r (*(col+offset+7*8), mm4); // mm4 = x7
+ pmulhw_r2r (mm1, mm0); // mm0 = T1*x1
- movq_m2r (*_T3, mm5); // mm5 = T3
- pmulhw_r2r (mm4, mm2); // mm2 = T1*x7
+ movq_m2r (*_T3, mm5); // mm5 = T3
+ pmulhw_r2r (mm4, mm2); // mm2 = T1*x7
- movq_m2r (*(col+offset+5*8), mm6); // mm6 = x5
- movq_r2r (mm5, mm7); // mm7 = T3-1
+ movq_m2r (*(col+offset+5*8), mm6); // mm6 = x5
+ movq_r2r (mm5, mm7); // mm7 = T3-1
- movq_m2r (*(col+offset+3*8), mm3); // mm3 = x3
- psubsw_r2r (mm4, mm0); // mm0 = v17
+ movq_m2r (*(col+offset+3*8), mm3); // mm3 = x3
+ psubsw_r2r (mm4, mm0); // mm0 = v17
- movq_m2r (*_T2, mm4); // mm4 = T2
- pmulhw_r2r (mm3, mm5); // mm5 = (T3-1)*x3
+ movq_m2r (*_T2, mm4); // mm4 = T2
+ pmulhw_r2r (mm3, mm5); // mm5 = (T3-1)*x3
- paddsw_r2r (mm2, mm1); // mm1 = u17
- pmulhw_r2r (mm6, mm7); // mm7 = (T3-1)*x5
+ paddsw_r2r (mm2, mm1); // mm1 = u17
+ pmulhw_r2r (mm6, mm7); // mm7 = (T3-1)*x5
/* slot */
- movq_r2r (mm4, mm2); // mm2 = T2
- paddsw_r2r (mm3, mm5); // mm5 = T3*x3
+ movq_r2r (mm4, mm2); // mm2 = T2
+ paddsw_r2r (mm3, mm5); // mm5 = T3*x3
pmulhw_m2r (*(col+offset+2*8), mm4);// mm4 = T2*x2
- paddsw_r2r (mm6, mm7); // mm7 = T3*x5
+ paddsw_r2r (mm6, mm7); // mm7 = T3*x5
- psubsw_r2r (mm6, mm5); // mm5 = v35
- paddsw_r2r (mm3, mm7); // mm7 = u35
+ psubsw_r2r (mm6, mm5); // mm5 = v35
+ paddsw_r2r (mm3, mm7); // mm7 = u35
- movq_m2r (*(col+offset+6*8), mm3); // mm3 = x6
- movq_r2r (mm0, mm6); // mm6 = v17
+ movq_m2r (*(col+offset+6*8), mm3); // mm3 = x6
+ movq_r2r (mm0, mm6); // mm6 = v17
- pmulhw_r2r (mm3, mm2); // mm2 = T2*x6
- psubsw_r2r (mm5, mm0); // mm0 = b3
+ pmulhw_r2r (mm3, mm2); // mm2 = T2*x6
+ psubsw_r2r (mm5, mm0); // mm0 = b3
- psubsw_r2r (mm3, mm4); // mm4 = v26
- paddsw_r2r (mm6, mm5); // mm5 = v12
+ psubsw_r2r (mm3, mm4); // mm4 = v26
+ paddsw_r2r (mm6, mm5); // mm5 = v12
- movq_r2m (mm0, *(col+offset+3*8)); // save b3 in scratch0
- movq_r2r (mm1, mm6); // mm6 = u17
+ movq_r2m (mm0, *(col+offset+3*8)); // save b3 in scratch0
+ movq_r2r (mm1, mm6); // mm6 = u17
paddsw_m2r (*(col+offset+2*8), mm2);// mm2 = u26
- paddsw_r2r (mm7, mm6); // mm6 = b0
+ paddsw_r2r (mm7, mm6); // mm6 = b0
- psubsw_r2r (mm7, mm1); // mm1 = u12
- movq_r2r (mm1, mm7); // mm7 = u12
+ psubsw_r2r (mm7, mm1); // mm1 = u12
+ movq_r2r (mm1, mm7); // mm7 = u12
- movq_m2r (*(col+offset+0*8), mm3); // mm3 = x0
- paddsw_r2r (mm5, mm1); // mm1 = u12+v12
+ movq_m2r (*(col+offset+0*8), mm3); // mm3 = x0
+ paddsw_r2r (mm5, mm1); // mm1 = u12+v12
- movq_m2r (*_C4, mm0); // mm0 = C4/2
- psubsw_r2r (mm5, mm7); // mm7 = u12-v12
+ movq_m2r (*_C4, mm0); // mm0 = C4/2
+ psubsw_r2r (mm5, mm7); // mm7 = u12-v12
- movq_r2m (mm6, *(col+offset+5*8)); // save b0 in scratch1
- pmulhw_r2r (mm0, mm1); // mm1 = b1/2
+ movq_r2m (mm6, *(col+offset+5*8)); // save b0 in scratch1
+ pmulhw_r2r (mm0, mm1); // mm1 = b1/2
- movq_r2r (mm4, mm6); // mm6 = v26
- pmulhw_r2r (mm0, mm7); // mm7 = b2/2
+ movq_r2r (mm4, mm6); // mm6 = v26
+ pmulhw_r2r (mm0, mm7); // mm7 = b2/2
- movq_m2r (*(col+offset+4*8), mm5); // mm5 = x4
- movq_r2r (mm3, mm0); // mm0 = x0
+ movq_m2r (*(col+offset+4*8), mm5); // mm5 = x4
+ movq_r2r (mm3, mm0); // mm0 = x0
- psubsw_r2r (mm5, mm3); // mm3 = v04
- paddsw_r2r (mm5, mm0); // mm0 = u04
+ psubsw_r2r (mm5, mm3); // mm3 = v04
+ paddsw_r2r (mm5, mm0); // mm0 = u04
- paddsw_r2r (mm3, mm4); // mm4 = a1
- movq_r2r (mm0, mm5); // mm5 = u04
+ paddsw_r2r (mm3, mm4); // mm4 = a1
+ movq_r2r (mm0, mm5); // mm5 = u04
- psubsw_r2r (mm6, mm3); // mm3 = a2
- paddsw_r2r (mm2, mm5); // mm5 = a0
+ psubsw_r2r (mm6, mm3); // mm3 = a2
+ paddsw_r2r (mm2, mm5); // mm5 = a0
- paddsw_r2r (mm1, mm1); // mm1 = b1
- psubsw_r2r (mm2, mm0); // mm0 = a3
+ paddsw_r2r (mm1, mm1); // mm1 = b1
+ psubsw_r2r (mm2, mm0); // mm0 = a3
- paddsw_r2r (mm7, mm7); // mm7 = b2
- movq_r2r (mm3, mm2); // mm2 = a2
+ paddsw_r2r (mm7, mm7); // mm7 = b2
+ movq_r2r (mm3, mm2); // mm2 = a2
- movq_r2r (mm4, mm6); // mm6 = a1
- paddsw_r2r (mm7, mm3); // mm3 = a2+b2
+ movq_r2r (mm4, mm6); // mm6 = a1
+ paddsw_r2r (mm7, mm3); // mm3 = a2+b2
- psraw_i2r (COL_SHIFT, mm3); // mm3 = y2
- paddsw_r2r (mm1, mm4); // mm4 = a1+b1
+ psraw_i2r (COL_SHIFT, mm3); // mm3 = y2
+ paddsw_r2r (mm1, mm4); // mm4 = a1+b1
- psraw_i2r (COL_SHIFT, mm4); // mm4 = y1
- psubsw_r2r (mm1, mm6); // mm6 = a1-b1
+ psraw_i2r (COL_SHIFT, mm4); // mm4 = y1
+ psubsw_r2r (mm1, mm6); // mm6 = a1-b1
- movq_m2r (*(col+offset+5*8), mm1); // mm1 = b0
- psubsw_r2r (mm7, mm2); // mm2 = a2-b2
+ movq_m2r (*(col+offset+5*8), mm1); // mm1 = b0
+ psubsw_r2r (mm7, mm2); // mm2 = a2-b2
- psraw_i2r (COL_SHIFT, mm6); // mm6 = y6
- movq_r2r (mm5, mm7); // mm7 = a0
+ psraw_i2r (COL_SHIFT, mm6); // mm6 = y6
+ movq_r2r (mm5, mm7); // mm7 = a0
- movq_r2m (mm4, *(col+offset+1*8)); // save y1
- psraw_i2r (COL_SHIFT, mm2); // mm2 = y5
+ movq_r2m (mm4, *(col+offset+1*8)); // save y1
+ psraw_i2r (COL_SHIFT, mm2); // mm2 = y5
- movq_r2m (mm3, *(col+offset+2*8)); // save y2
- paddsw_r2r (mm1, mm5); // mm5 = a0+b0
+ movq_r2m (mm3, *(col+offset+2*8)); // save y2
+ paddsw_r2r (mm1, mm5); // mm5 = a0+b0
- movq_m2r (*(col+offset+3*8), mm4); // mm4 = b3
- psubsw_r2r (mm1, mm7); // mm7 = a0-b0
+ movq_m2r (*(col+offset+3*8), mm4); // mm4 = b3
+ psubsw_r2r (mm1, mm7); // mm7 = a0-b0
- psraw_i2r (COL_SHIFT, mm5); // mm5 = y0
- movq_r2r (mm0, mm3); // mm3 = a3
+ psraw_i2r (COL_SHIFT, mm5); // mm5 = y0
+ movq_r2r (mm0, mm3); // mm3 = a3
- movq_r2m (mm2, *(col+offset+5*8)); // save y5
- psubsw_r2r (mm4, mm3); // mm3 = a3-b3
+ movq_r2m (mm2, *(col+offset+5*8)); // save y5
+ psubsw_r2r (mm4, mm3); // mm3 = a3-b3
- psraw_i2r (COL_SHIFT, mm7); // mm7 = y7
- paddsw_r2r (mm0, mm4); // mm4 = a3+b3
+ psraw_i2r (COL_SHIFT, mm7); // mm7 = y7
+ paddsw_r2r (mm0, mm4); // mm4 = a3+b3
- movq_r2m (mm5, *(col+offset+0*8)); // save y0
- psraw_i2r (COL_SHIFT, mm3); // mm3 = y4
+ movq_r2m (mm5, *(col+offset+0*8)); // save y0
+ psraw_i2r (COL_SHIFT, mm3); // mm3 = y4
- movq_r2m (mm6, *(col+offset+6*8)); // save y6
- psraw_i2r (COL_SHIFT, mm4); // mm4 = y3
+ movq_r2m (mm6, *(col+offset+6*8)); // save y6
+ psraw_i2r (COL_SHIFT, mm4); // mm4 = y3
- movq_r2m (mm7, *(col+offset+7*8)); // save y7
+ movq_r2m (mm7, *(col+offset+7*8)); // save y7
- movq_r2m (mm3, *(col+offset+4*8)); // save y4
+ movq_r2m (mm3, *(col+offset+4*8)); // save y4
- movq_r2m (mm4, *(col+offset+3*8)); // save y3
+ movq_r2m (mm4, *(col+offset+3*8)); // save y3
#undef T1
#undef T2
@@ -540,61 +540,61 @@ static const int32_t rounder0[] ATTR_ALIGN(8) =
rounder ((1 << (COL_SHIFT - 1)) - 0.5);
static const int32_t rounder4[] ATTR_ALIGN(8) = rounder (0);
static const int32_t rounder1[] ATTR_ALIGN(8) =
- rounder (1.25683487303); /* C1*(C1/C4+C1+C7)/2 */
+ rounder (1.25683487303); /* C1*(C1/C4+C1+C7)/2 */
static const int32_t rounder7[] ATTR_ALIGN(8) =
- rounder (-0.25); /* C1*(C7/C4+C7-C1)/2 */
+ rounder (-0.25); /* C1*(C7/C4+C7-C1)/2 */
static const int32_t rounder2[] ATTR_ALIGN(8) =
- rounder (0.60355339059); /* C2 * (C6+C2)/2 */
+ rounder (0.60355339059); /* C2 * (C6+C2)/2 */
static const int32_t rounder6[] ATTR_ALIGN(8) =
- rounder (-0.25); /* C2 * (C6-C2)/2 */
+ rounder (-0.25); /* C2 * (C6-C2)/2 */
static const int32_t rounder3[] ATTR_ALIGN(8) =
- rounder (0.087788325588); /* C3*(-C3/C4+C3+C5)/2 */
+ rounder (0.087788325588); /* C3*(-C3/C4+C3+C5)/2 */
static const int32_t rounder5[] ATTR_ALIGN(8) =
- rounder (-0.441341716183); /* C3*(-C5/C4+C5-C3)/2 */
+ rounder (-0.441341716183); /* C3*(-C5/C4+C5-C3)/2 */
#undef COL_SHIFT
#undef ROW_SHIFT
-#define declare_idct(idct,table,idct_row_head,idct_row,idct_row_tail,idct_row_mid) \
-void idct (int16_t * block) \
-{ \
- static const int16_t table04[] ATTR_ALIGN(16) = \
- table (22725, 21407, 19266, 16384, 12873, 8867, 4520); \
- static const int16_t table17[] ATTR_ALIGN(16) = \
- table (31521, 29692, 26722, 22725, 17855, 12299, 6270); \
- static const int16_t table26[] ATTR_ALIGN(16) = \
- table (29692, 27969, 25172, 21407, 16819, 11585, 5906); \
- static const int16_t table35[] ATTR_ALIGN(16) = \
- table (26722, 25172, 22654, 19266, 15137, 10426, 5315); \
- \
- idct_row_head (block, 0*8, table04); \
- idct_row (table04, rounder0); \
- idct_row_mid (block, 0*8, 4*8, table04); \
- idct_row (table04, rounder4); \
- idct_row_mid (block, 4*8, 1*8, table17); \
- idct_row (table17, rounder1); \
- idct_row_mid (block, 1*8, 7*8, table17); \
- idct_row (table17, rounder7); \
- idct_row_mid (block, 7*8, 2*8, table26); \
- idct_row (table26, rounder2); \
- idct_row_mid (block, 2*8, 6*8, table26); \
- idct_row (table26, rounder6); \
- idct_row_mid (block, 6*8, 3*8, table35); \
- idct_row (table35, rounder3); \
- idct_row_mid (block, 3*8, 5*8, table35); \
- idct_row (table35, rounder5); \
- idct_row_tail (block, 5*8); \
- \
- idct_col (block, 0); \
- idct_col (block, 4); \
+#define declare_idct(idct,table,idct_row_head,idct_row,idct_row_tail,idct_row_mid) \
+void idct (int16_t * block) \
+{ \
+ static const int16_t table04[] ATTR_ALIGN(16) = \
+ table (22725, 21407, 19266, 16384, 12873, 8867, 4520); \
+ static const int16_t table17[] ATTR_ALIGN(16) = \
+ table (31521, 29692, 26722, 22725, 17855, 12299, 6270); \
+ static const int16_t table26[] ATTR_ALIGN(16) = \
+ table (29692, 27969, 25172, 21407, 16819, 11585, 5906); \
+ static const int16_t table35[] ATTR_ALIGN(16) = \
+ table (26722, 25172, 22654, 19266, 15137, 10426, 5315); \
+ \
+ idct_row_head (block, 0*8, table04); \
+ idct_row (table04, rounder0); \
+ idct_row_mid (block, 0*8, 4*8, table04); \
+ idct_row (table04, rounder4); \
+ idct_row_mid (block, 4*8, 1*8, table17); \
+ idct_row (table17, rounder1); \
+ idct_row_mid (block, 1*8, 7*8, table17); \
+ idct_row (table17, rounder7); \
+ idct_row_mid (block, 7*8, 2*8, table26); \
+ idct_row (table26, rounder2); \
+ idct_row_mid (block, 2*8, 6*8, table26); \
+ idct_row (table26, rounder6); \
+ idct_row_mid (block, 6*8, 3*8, table35); \
+ idct_row (table35, rounder3); \
+ idct_row_mid (block, 3*8, 5*8, table35); \
+ idct_row (table35, rounder5); \
+ idct_row_tail (block, 5*8); \
+ \
+ idct_col (block, 0); \
+ idct_col (block, 4); \
}
void ff_mmx_idct(DCTELEM *block);
void ff_mmxext_idct(DCTELEM *block);
declare_idct (ff_mmxext_idct, mmxext_table,
- mmxext_row_head, mmxext_row, mmxext_row_tail, mmxext_row_mid)
+ mmxext_row_head, mmxext_row, mmxext_row_tail, mmxext_row_mid)
declare_idct (ff_mmx_idct, mmx_table,
- mmx_row_head, mmx_row, mmx_row_tail, mmx_row_mid)
+ mmx_row_head, mmx_row, mmx_row_tail, mmx_row_mid)
diff --git a/libavcodec/i386/mmx.h b/libavcodec/i386/mmx.h
index 96360ab16a..df4620e0ab 100644
--- a/libavcodec/i386/mmx.h
+++ b/libavcodec/i386/mmx.h
@@ -27,257 +27,257 @@
* values by ULL, lest they be truncated by the compiler)
*/
-typedef union {
- long long q; /* Quadword (64-bit) value */
- unsigned long long uq; /* Unsigned Quadword */
- int d[2]; /* 2 Doubleword (32-bit) values */
- unsigned int ud[2]; /* 2 Unsigned Doubleword */
- short w[4]; /* 4 Word (16-bit) values */
- unsigned short uw[4]; /* 4 Unsigned Word */
- char b[8]; /* 8 Byte (8-bit) values */
- unsigned char ub[8]; /* 8 Unsigned Byte */
- float s[2]; /* Single-precision (32-bit) value */
-} mmx_t; /* On an 8-byte (64-bit) boundary */
-
-
-#define mmx_i2r(op,imm,reg) \
- __asm__ __volatile__ (#op " %0, %%" #reg \
- : /* nothing */ \
- : "i" (imm) )
-
-#define mmx_m2r(op,mem,reg) \
- __asm__ __volatile__ (#op " %0, %%" #reg \
- : /* nothing */ \
- : "m" (mem))
-
-#define mmx_r2m(op,reg,mem) \
- __asm__ __volatile__ (#op " %%" #reg ", %0" \
- : "=m" (mem) \
- : /* nothing */ )
-
-#define mmx_r2r(op,regs,regd) \
- __asm__ __volatile__ (#op " %" #regs ", %" #regd)
-
-
-#define emms() __asm__ __volatile__ ("emms")
-
-#define movd_m2r(var,reg) mmx_m2r (movd, var, reg)
-#define movd_r2m(reg,var) mmx_r2m (movd, reg, var)
-#define movd_r2r(regs,regd) mmx_r2r (movd, regs, regd)
-
-#define movq_m2r(var,reg) mmx_m2r (movq, var, reg)
-#define movq_r2m(reg,var) mmx_r2m (movq, reg, var)
-#define movq_r2r(regs,regd) mmx_r2r (movq, regs, regd)
-
-#define packssdw_m2r(var,reg) mmx_m2r (packssdw, var, reg)
-#define packssdw_r2r(regs,regd) mmx_r2r (packssdw, regs, regd)
-#define packsswb_m2r(var,reg) mmx_m2r (packsswb, var, reg)
-#define packsswb_r2r(regs,regd) mmx_r2r (packsswb, regs, regd)
-
-#define packuswb_m2r(var,reg) mmx_m2r (packuswb, var, reg)
-#define packuswb_r2r(regs,regd) mmx_r2r (packuswb, regs, regd)
-
-#define paddb_m2r(var,reg) mmx_m2r (paddb, var, reg)
-#define paddb_r2r(regs,regd) mmx_r2r (paddb, regs, regd)
-#define paddd_m2r(var,reg) mmx_m2r (paddd, var, reg)
-#define paddd_r2r(regs,regd) mmx_r2r (paddd, regs, regd)
-#define paddw_m2r(var,reg) mmx_m2r (paddw, var, reg)
-#define paddw_r2r(regs,regd) mmx_r2r (paddw, regs, regd)
-
-#define paddsb_m2r(var,reg) mmx_m2r (paddsb, var, reg)
-#define paddsb_r2r(regs,regd) mmx_r2r (paddsb, regs, regd)
-#define paddsw_m2r(var,reg) mmx_m2r (paddsw, var, reg)
-#define paddsw_r2r(regs,regd) mmx_r2r (paddsw, regs, regd)
-
-#define paddusb_m2r(var,reg) mmx_m2r (paddusb, var, reg)
-#define paddusb_r2r(regs,regd) mmx_r2r (paddusb, regs, regd)
-#define paddusw_m2r(var,reg) mmx_m2r (paddusw, var, reg)
-#define paddusw_r2r(regs,regd) mmx_r2r (paddusw, regs, regd)
-
-#define pand_m2r(var,reg) mmx_m2r (pand, var, reg)
-#define pand_r2r(regs,regd) mmx_r2r (pand, regs, regd)
-
-#define pandn_m2r(var,reg) mmx_m2r (pandn, var, reg)
-#define pandn_r2r(regs,regd) mmx_r2r (pandn, regs, regd)
-
-#define pcmpeqb_m2r(var,reg) mmx_m2r (pcmpeqb, var, reg)
-#define pcmpeqb_r2r(regs,regd) mmx_r2r (pcmpeqb, regs, regd)
-#define pcmpeqd_m2r(var,reg) mmx_m2r (pcmpeqd, var, reg)
-#define pcmpeqd_r2r(regs,regd) mmx_r2r (pcmpeqd, regs, regd)
-#define pcmpeqw_m2r(var,reg) mmx_m2r (pcmpeqw, var, reg)
-#define pcmpeqw_r2r(regs,regd) mmx_r2r (pcmpeqw, regs, regd)
-
-#define pcmpgtb_m2r(var,reg) mmx_m2r (pcmpgtb, var, reg)
-#define pcmpgtb_r2r(regs,regd) mmx_r2r (pcmpgtb, regs, regd)
-#define pcmpgtd_m2r(var,reg) mmx_m2r (pcmpgtd, var, reg)
-#define pcmpgtd_r2r(regs,regd) mmx_r2r (pcmpgtd, regs, regd)
-#define pcmpgtw_m2r(var,reg) mmx_m2r (pcmpgtw, var, reg)
-#define pcmpgtw_r2r(regs,regd) mmx_r2r (pcmpgtw, regs, regd)
-
-#define pmaddwd_m2r(var,reg) mmx_m2r (pmaddwd, var, reg)
-#define pmaddwd_r2r(regs,regd) mmx_r2r (pmaddwd, regs, regd)
-
-#define pmulhw_m2r(var,reg) mmx_m2r (pmulhw, var, reg)
-#define pmulhw_r2r(regs,regd) mmx_r2r (pmulhw, regs, regd)
-
-#define pmullw_m2r(var,reg) mmx_m2r (pmullw, var, reg)
-#define pmullw_r2r(regs,regd) mmx_r2r (pmullw, regs, regd)
-
-#define por_m2r(var,reg) mmx_m2r (por, var, reg)
-#define por_r2r(regs,regd) mmx_r2r (por, regs, regd)
-
-#define pslld_i2r(imm,reg) mmx_i2r (pslld, imm, reg)
-#define pslld_m2r(var,reg) mmx_m2r (pslld, var, reg)
-#define pslld_r2r(regs,regd) mmx_r2r (pslld, regs, regd)
-#define psllq_i2r(imm,reg) mmx_i2r (psllq, imm, reg)
-#define psllq_m2r(var,reg) mmx_m2r (psllq, var, reg)
-#define psllq_r2r(regs,regd) mmx_r2r (psllq, regs, regd)
-#define psllw_i2r(imm,reg) mmx_i2r (psllw, imm, reg)
-#define psllw_m2r(var,reg) mmx_m2r (psllw, var, reg)
-#define psllw_r2r(regs,regd) mmx_r2r (psllw, regs, regd)
-
-#define psrad_i2r(imm,reg) mmx_i2r (psrad, imm, reg)
-#define psrad_m2r(var,reg) mmx_m2r (psrad, var, reg)
-#define psrad_r2r(regs,regd) mmx_r2r (psrad, regs, regd)
-#define psraw_i2r(imm,reg) mmx_i2r (psraw, imm, reg)
-#define psraw_m2r(var,reg) mmx_m2r (psraw, var, reg)
-#define psraw_r2r(regs,regd) mmx_r2r (psraw, regs, regd)
-
-#define psrld_i2r(imm,reg) mmx_i2r (psrld, imm, reg)
-#define psrld_m2r(var,reg) mmx_m2r (psrld, var, reg)
-#define psrld_r2r(regs,regd) mmx_r2r (psrld, regs, regd)
-#define psrlq_i2r(imm,reg) mmx_i2r (psrlq, imm, reg)
-#define psrlq_m2r(var,reg) mmx_m2r (psrlq, var, reg)
-#define psrlq_r2r(regs,regd) mmx_r2r (psrlq, regs, regd)
-#define psrlw_i2r(imm,reg) mmx_i2r (psrlw, imm, reg)
-#define psrlw_m2r(var,reg) mmx_m2r (psrlw, var, reg)
-#define psrlw_r2r(regs,regd) mmx_r2r (psrlw, regs, regd)
-
-#define psubb_m2r(var,reg) mmx_m2r (psubb, var, reg)
-#define psubb_r2r(regs,regd) mmx_r2r (psubb, regs, regd)
-#define psubd_m2r(var,reg) mmx_m2r (psubd, var, reg)
-#define psubd_r2r(regs,regd) mmx_r2r (psubd, regs, regd)
-#define psubw_m2r(var,reg) mmx_m2r (psubw, var, reg)
-#define psubw_r2r(regs,regd) mmx_r2r (psubw, regs, regd)
-
-#define psubsb_m2r(var,reg) mmx_m2r (psubsb, var, reg)
-#define psubsb_r2r(regs,regd) mmx_r2r (psubsb, regs, regd)
-#define psubsw_m2r(var,reg) mmx_m2r (psubsw, var, reg)
-#define psubsw_r2r(regs,regd) mmx_r2r (psubsw, regs, regd)
-
-#define psubusb_m2r(var,reg) mmx_m2r (psubusb, var, reg)
-#define psubusb_r2r(regs,regd) mmx_r2r (psubusb, regs, regd)
-#define psubusw_m2r(var,reg) mmx_m2r (psubusw, var, reg)
-#define psubusw_r2r(regs,regd) mmx_r2r (psubusw, regs, regd)
-
-#define punpckhbw_m2r(var,reg) mmx_m2r (punpckhbw, var, reg)
-#define punpckhbw_r2r(regs,regd) mmx_r2r (punpckhbw, regs, regd)
-#define punpckhdq_m2r(var,reg) mmx_m2r (punpckhdq, var, reg)
-#define punpckhdq_r2r(regs,regd) mmx_r2r (punpckhdq, regs, regd)
-#define punpckhwd_m2r(var,reg) mmx_m2r (punpckhwd, var, reg)
-#define punpckhwd_r2r(regs,regd) mmx_r2r (punpckhwd, regs, regd)
-
-#define punpcklbw_m2r(var,reg) mmx_m2r (punpcklbw, var, reg)
-#define punpcklbw_r2r(regs,regd) mmx_r2r (punpcklbw, regs, regd)
-#define punpckldq_m2r(var,reg) mmx_m2r (punpckldq, var, reg)
-#define punpckldq_r2r(regs,regd) mmx_r2r (punpckldq, regs, regd)
-#define punpcklwd_m2r(var,reg) mmx_m2r (punpcklwd, var, reg)
-#define punpcklwd_r2r(regs,regd) mmx_r2r (punpcklwd, regs, regd)
-
-#define pxor_m2r(var,reg) mmx_m2r (pxor, var, reg)
-#define pxor_r2r(regs,regd) mmx_r2r (pxor, regs, regd)
+typedef union {
+ long long q; /* Quadword (64-bit) value */
+ unsigned long long uq; /* Unsigned Quadword */
+ int d[2]; /* 2 Doubleword (32-bit) values */
+ unsigned int ud[2]; /* 2 Unsigned Doubleword */
+ short w[4]; /* 4 Word (16-bit) values */
+ unsigned short uw[4]; /* 4 Unsigned Word */
+ char b[8]; /* 8 Byte (8-bit) values */
+ unsigned char ub[8]; /* 8 Unsigned Byte */
+ float s[2]; /* Single-precision (32-bit) value */
+} mmx_t; /* On an 8-byte (64-bit) boundary */
+
+
+#define mmx_i2r(op,imm,reg) \
+ __asm__ __volatile__ (#op " %0, %%" #reg \
+ : /* nothing */ \
+ : "i" (imm) )
+
+#define mmx_m2r(op,mem,reg) \
+ __asm__ __volatile__ (#op " %0, %%" #reg \
+ : /* nothing */ \
+ : "m" (mem))
+
+#define mmx_r2m(op,reg,mem) \
+ __asm__ __volatile__ (#op " %%" #reg ", %0" \
+ : "=m" (mem) \
+ : /* nothing */ )
+
+#define mmx_r2r(op,regs,regd) \
+ __asm__ __volatile__ (#op " %" #regs ", %" #regd)
+
+
+#define emms() __asm__ __volatile__ ("emms")
+
+#define movd_m2r(var,reg) mmx_m2r (movd, var, reg)
+#define movd_r2m(reg,var) mmx_r2m (movd, reg, var)
+#define movd_r2r(regs,regd) mmx_r2r (movd, regs, regd)
+
+#define movq_m2r(var,reg) mmx_m2r (movq, var, reg)
+#define movq_r2m(reg,var) mmx_r2m (movq, reg, var)
+#define movq_r2r(regs,regd) mmx_r2r (movq, regs, regd)
+
+#define packssdw_m2r(var,reg) mmx_m2r (packssdw, var, reg)
+#define packssdw_r2r(regs,regd) mmx_r2r (packssdw, regs, regd)
+#define packsswb_m2r(var,reg) mmx_m2r (packsswb, var, reg)
+#define packsswb_r2r(regs,regd) mmx_r2r (packsswb, regs, regd)
+
+#define packuswb_m2r(var,reg) mmx_m2r (packuswb, var, reg)
+#define packuswb_r2r(regs,regd) mmx_r2r (packuswb, regs, regd)
+
+#define paddb_m2r(var,reg) mmx_m2r (paddb, var, reg)
+#define paddb_r2r(regs,regd) mmx_r2r (paddb, regs, regd)
+#define paddd_m2r(var,reg) mmx_m2r (paddd, var, reg)
+#define paddd_r2r(regs,regd) mmx_r2r (paddd, regs, regd)
+#define paddw_m2r(var,reg) mmx_m2r (paddw, var, reg)
+#define paddw_r2r(regs,regd) mmx_r2r (paddw, regs, regd)
+
+#define paddsb_m2r(var,reg) mmx_m2r (paddsb, var, reg)
+#define paddsb_r2r(regs,regd) mmx_r2r (paddsb, regs, regd)
+#define paddsw_m2r(var,reg) mmx_m2r (paddsw, var, reg)
+#define paddsw_r2r(regs,regd) mmx_r2r (paddsw, regs, regd)
+
+#define paddusb_m2r(var,reg) mmx_m2r (paddusb, var, reg)
+#define paddusb_r2r(regs,regd) mmx_r2r (paddusb, regs, regd)
+#define paddusw_m2r(var,reg) mmx_m2r (paddusw, var, reg)
+#define paddusw_r2r(regs,regd) mmx_r2r (paddusw, regs, regd)
+
+#define pand_m2r(var,reg) mmx_m2r (pand, var, reg)
+#define pand_r2r(regs,regd) mmx_r2r (pand, regs, regd)
+
+#define pandn_m2r(var,reg) mmx_m2r (pandn, var, reg)
+#define pandn_r2r(regs,regd) mmx_r2r (pandn, regs, regd)
+
+#define pcmpeqb_m2r(var,reg) mmx_m2r (pcmpeqb, var, reg)
+#define pcmpeqb_r2r(regs,regd) mmx_r2r (pcmpeqb, regs, regd)
+#define pcmpeqd_m2r(var,reg) mmx_m2r (pcmpeqd, var, reg)
+#define pcmpeqd_r2r(regs,regd) mmx_r2r (pcmpeqd, regs, regd)
+#define pcmpeqw_m2r(var,reg) mmx_m2r (pcmpeqw, var, reg)
+#define pcmpeqw_r2r(regs,regd) mmx_r2r (pcmpeqw, regs, regd)
+
+#define pcmpgtb_m2r(var,reg) mmx_m2r (pcmpgtb, var, reg)
+#define pcmpgtb_r2r(regs,regd) mmx_r2r (pcmpgtb, regs, regd)
+#define pcmpgtd_m2r(var,reg) mmx_m2r (pcmpgtd, var, reg)
+#define pcmpgtd_r2r(regs,regd) mmx_r2r (pcmpgtd, regs, regd)
+#define pcmpgtw_m2r(var,reg) mmx_m2r (pcmpgtw, var, reg)
+#define pcmpgtw_r2r(regs,regd) mmx_r2r (pcmpgtw, regs, regd)
+
+#define pmaddwd_m2r(var,reg) mmx_m2r (pmaddwd, var, reg)
+#define pmaddwd_r2r(regs,regd) mmx_r2r (pmaddwd, regs, regd)
+
+#define pmulhw_m2r(var,reg) mmx_m2r (pmulhw, var, reg)
+#define pmulhw_r2r(regs,regd) mmx_r2r (pmulhw, regs, regd)
+
+#define pmullw_m2r(var,reg) mmx_m2r (pmullw, var, reg)
+#define pmullw_r2r(regs,regd) mmx_r2r (pmullw, regs, regd)
+
+#define por_m2r(var,reg) mmx_m2r (por, var, reg)
+#define por_r2r(regs,regd) mmx_r2r (por, regs, regd)
+
+#define pslld_i2r(imm,reg) mmx_i2r (pslld, imm, reg)
+#define pslld_m2r(var,reg) mmx_m2r (pslld, var, reg)
+#define pslld_r2r(regs,regd) mmx_r2r (pslld, regs, regd)
+#define psllq_i2r(imm,reg) mmx_i2r (psllq, imm, reg)
+#define psllq_m2r(var,reg) mmx_m2r (psllq, var, reg)
+#define psllq_r2r(regs,regd) mmx_r2r (psllq, regs, regd)
+#define psllw_i2r(imm,reg) mmx_i2r (psllw, imm, reg)
+#define psllw_m2r(var,reg) mmx_m2r (psllw, var, reg)
+#define psllw_r2r(regs,regd) mmx_r2r (psllw, regs, regd)
+
+#define psrad_i2r(imm,reg) mmx_i2r (psrad, imm, reg)
+#define psrad_m2r(var,reg) mmx_m2r (psrad, var, reg)
+#define psrad_r2r(regs,regd) mmx_r2r (psrad, regs, regd)
+#define psraw_i2r(imm,reg) mmx_i2r (psraw, imm, reg)
+#define psraw_m2r(var,reg) mmx_m2r (psraw, var, reg)
+#define psraw_r2r(regs,regd) mmx_r2r (psraw, regs, regd)
+
+#define psrld_i2r(imm,reg) mmx_i2r (psrld, imm, reg)
+#define psrld_m2r(var,reg) mmx_m2r (psrld, var, reg)
+#define psrld_r2r(regs,regd) mmx_r2r (psrld, regs, regd)
+#define psrlq_i2r(imm,reg) mmx_i2r (psrlq, imm, reg)
+#define psrlq_m2r(var,reg) mmx_m2r (psrlq, var, reg)
+#define psrlq_r2r(regs,regd) mmx_r2r (psrlq, regs, regd)
+#define psrlw_i2r(imm,reg) mmx_i2r (psrlw, imm, reg)
+#define psrlw_m2r(var,reg) mmx_m2r (psrlw, var, reg)
+#define psrlw_r2r(regs,regd) mmx_r2r (psrlw, regs, regd)
+
+#define psubb_m2r(var,reg) mmx_m2r (psubb, var, reg)
+#define psubb_r2r(regs,regd) mmx_r2r (psubb, regs, regd)
+#define psubd_m2r(var,reg) mmx_m2r (psubd, var, reg)
+#define psubd_r2r(regs,regd) mmx_r2r (psubd, regs, regd)
+#define psubw_m2r(var,reg) mmx_m2r (psubw, var, reg)
+#define psubw_r2r(regs,regd) mmx_r2r (psubw, regs, regd)
+
+#define psubsb_m2r(var,reg) mmx_m2r (psubsb, var, reg)
+#define psubsb_r2r(regs,regd) mmx_r2r (psubsb, regs, regd)
+#define psubsw_m2r(var,reg) mmx_m2r (psubsw, var, reg)
+#define psubsw_r2r(regs,regd) mmx_r2r (psubsw, regs, regd)
+
+#define psubusb_m2r(var,reg) mmx_m2r (psubusb, var, reg)
+#define psubusb_r2r(regs,regd) mmx_r2r (psubusb, regs, regd)
+#define psubusw_m2r(var,reg) mmx_m2r (psubusw, var, reg)
+#define psubusw_r2r(regs,regd) mmx_r2r (psubusw, regs, regd)
+
+#define punpckhbw_m2r(var,reg) mmx_m2r (punpckhbw, var, reg)
+#define punpckhbw_r2r(regs,regd) mmx_r2r (punpckhbw, regs, regd)
+#define punpckhdq_m2r(var,reg) mmx_m2r (punpckhdq, var, reg)
+#define punpckhdq_r2r(regs,regd) mmx_r2r (punpckhdq, regs, regd)
+#define punpckhwd_m2r(var,reg) mmx_m2r (punpckhwd, var, reg)
+#define punpckhwd_r2r(regs,regd) mmx_r2r (punpckhwd, regs, regd)
+
+#define punpcklbw_m2r(var,reg) mmx_m2r (punpcklbw, var, reg)
+#define punpcklbw_r2r(regs,regd) mmx_r2r (punpcklbw, regs, regd)
+#define punpckldq_m2r(var,reg) mmx_m2r (punpckldq, var, reg)
+#define punpckldq_r2r(regs,regd) mmx_r2r (punpckldq, regs, regd)
+#define punpcklwd_m2r(var,reg) mmx_m2r (punpcklwd, var, reg)
+#define punpcklwd_r2r(regs,regd) mmx_r2r (punpcklwd, regs, regd)
+
+#define pxor_m2r(var,reg) mmx_m2r (pxor, var, reg)
+#define pxor_r2r(regs,regd) mmx_r2r (pxor, regs, regd)
/* 3DNOW extensions */
-#define pavgusb_m2r(var,reg) mmx_m2r (pavgusb, var, reg)
-#define pavgusb_r2r(regs,regd) mmx_r2r (pavgusb, regs, regd)
+#define pavgusb_m2r(var,reg) mmx_m2r (pavgusb, var, reg)
+#define pavgusb_r2r(regs,regd) mmx_r2r (pavgusb, regs, regd)
/* AMD MMX extensions - also available in intel SSE */
-#define mmx_m2ri(op,mem,reg,imm) \
+#define mmx_m2ri(op,mem,reg,imm) \
__asm__ __volatile__ (#op " %1, %0, %%" #reg \
: /* nothing */ \
: "X" (mem), "X" (imm))
-#define mmx_r2ri(op,regs,regd,imm) \
+#define mmx_r2ri(op,regs,regd,imm) \
__asm__ __volatile__ (#op " %0, %%" #regs ", %%" #regd \
: /* nothing */ \
: "X" (imm) )
-#define mmx_fetch(mem,hint) \
- __asm__ __volatile__ ("prefetch" #hint " %0" \
- : /* nothing */ \
- : "X" (mem))
+#define mmx_fetch(mem,hint) \
+ __asm__ __volatile__ ("prefetch" #hint " %0" \
+ : /* nothing */ \
+ : "X" (mem))
-#define maskmovq(regs,maskreg) mmx_r2ri (maskmovq, regs, maskreg)
+#define maskmovq(regs,maskreg) mmx_r2ri (maskmovq, regs, maskreg)
-#define movntq_r2m(mmreg,var) mmx_r2m (movntq, mmreg, var)
+#define movntq_r2m(mmreg,var) mmx_r2m (movntq, mmreg, var)
-#define pavgb_m2r(var,reg) mmx_m2r (pavgb, var, reg)
-#define pavgb_r2r(regs,regd) mmx_r2r (pavgb, regs, regd)
-#define pavgw_m2r(var,reg) mmx_m2r (pavgw, var, reg)
-#define pavgw_r2r(regs,regd) mmx_r2r (pavgw, regs, regd)
+#define pavgb_m2r(var,reg) mmx_m2r (pavgb, var, reg)
+#define pavgb_r2r(regs,regd) mmx_r2r (pavgb, regs, regd)
+#define pavgw_m2r(var,reg) mmx_m2r (pavgw, var, reg)
+#define pavgw_r2r(regs,regd) mmx_r2r (pavgw, regs, regd)
-#define pextrw_r2r(mmreg,reg,imm) mmx_r2ri (pextrw, mmreg, reg, imm)
+#define pextrw_r2r(mmreg,reg,imm) mmx_r2ri (pextrw, mmreg, reg, imm)
-#define pinsrw_r2r(reg,mmreg,imm) mmx_r2ri (pinsrw, reg, mmreg, imm)
+#define pinsrw_r2r(reg,mmreg,imm) mmx_r2ri (pinsrw, reg, mmreg, imm)
-#define pmaxsw_m2r(var,reg) mmx_m2r (pmaxsw, var, reg)
-#define pmaxsw_r2r(regs,regd) mmx_r2r (pmaxsw, regs, regd)
+#define pmaxsw_m2r(var,reg) mmx_m2r (pmaxsw, var, reg)
+#define pmaxsw_r2r(regs,regd) mmx_r2r (pmaxsw, regs, regd)
-#define pmaxub_m2r(var,reg) mmx_m2r (pmaxub, var, reg)
-#define pmaxub_r2r(regs,regd) mmx_r2r (pmaxub, regs, regd)
+#define pmaxub_m2r(var,reg) mmx_m2r (pmaxub, var, reg)
+#define pmaxub_r2r(regs,regd) mmx_r2r (pmaxub, regs, regd)
-#define pminsw_m2r(var,reg) mmx_m2r (pminsw, var, reg)
-#define pminsw_r2r(regs,regd) mmx_r2r (pminsw, regs, regd)
+#define pminsw_m2r(var,reg) mmx_m2r (pminsw, var, reg)
+#define pminsw_r2r(regs,regd) mmx_r2r (pminsw, regs, regd)
-#define pminub_m2r(var,reg) mmx_m2r (pminub, var, reg)
-#define pminub_r2r(regs,regd) mmx_r2r (pminub, regs, regd)
+#define pminub_m2r(var,reg) mmx_m2r (pminub, var, reg)
+#define pminub_r2r(regs,regd) mmx_r2r (pminub, regs, regd)
-#define pmovmskb(mmreg,reg) \
- __asm__ __volatile__ ("movmskps %" #mmreg ", %" #reg)
+#define pmovmskb(mmreg,reg) \
+ __asm__ __volatile__ ("movmskps %" #mmreg ", %" #reg)
-#define pmulhuw_m2r(var,reg) mmx_m2r (pmulhuw, var, reg)
-#define pmulhuw_r2r(regs,regd) mmx_r2r (pmulhuw, regs, regd)
+#define pmulhuw_m2r(var,reg) mmx_m2r (pmulhuw, var, reg)
+#define pmulhuw_r2r(regs,regd) mmx_r2r (pmulhuw, regs, regd)
-#define prefetcht0(mem) mmx_fetch (mem, t0)
-#define prefetcht1(mem) mmx_fetch (mem, t1)
-#define prefetcht2(mem) mmx_fetch (mem, t2)
-#define prefetchnta(mem) mmx_fetch (mem, nta)
+#define prefetcht0(mem) mmx_fetch (mem, t0)
+#define prefetcht1(mem) mmx_fetch (mem, t1)
+#define prefetcht2(mem) mmx_fetch (mem, t2)
+#define prefetchnta(mem) mmx_fetch (mem, nta)
-#define psadbw_m2r(var,reg) mmx_m2r (psadbw, var, reg)
-#define psadbw_r2r(regs,regd) mmx_r2r (psadbw, regs, regd)
+#define psadbw_m2r(var,reg) mmx_m2r (psadbw, var, reg)
+#define psadbw_r2r(regs,regd) mmx_r2r (psadbw, regs, regd)
-#define pshufw_m2r(var,reg,imm) mmx_m2ri(pshufw, var, reg, imm)
-#define pshufw_r2r(regs,regd,imm) mmx_r2ri(pshufw, regs, regd, imm)
+#define pshufw_m2r(var,reg,imm) mmx_m2ri(pshufw, var, reg, imm)
+#define pshufw_r2r(regs,regd,imm) mmx_r2ri(pshufw, regs, regd, imm)
-#define sfence() __asm__ __volatile__ ("sfence\n\t")
+#define sfence() __asm__ __volatile__ ("sfence\n\t")
/* SSE2 */
-#define pshufhw_m2r(var,reg,imm) mmx_m2ri(pshufhw, var, reg, imm)
-#define pshufhw_r2r(regs,regd,imm) mmx_r2ri(pshufhw, regs, regd, imm)
-#define pshuflw_m2r(var,reg,imm) mmx_m2ri(pshuflw, var, reg, imm)
-#define pshuflw_r2r(regs,regd,imm) mmx_r2ri(pshuflw, regs, regd, imm)
+#define pshufhw_m2r(var,reg,imm) mmx_m2ri(pshufhw, var, reg, imm)
+#define pshufhw_r2r(regs,regd,imm) mmx_r2ri(pshufhw, regs, regd, imm)
+#define pshuflw_m2r(var,reg,imm) mmx_m2ri(pshuflw, var, reg, imm)
+#define pshuflw_r2r(regs,regd,imm) mmx_r2ri(pshuflw, regs, regd, imm)
-#define pshufd_r2r(regs,regd,imm) mmx_r2ri(pshufd, regs, regd, imm)
+#define pshufd_r2r(regs,regd,imm) mmx_r2ri(pshufd, regs, regd, imm)
-#define movdqa_m2r(var,reg) mmx_m2r (movdqa, var, reg)
-#define movdqa_r2m(reg,var) mmx_r2m (movdqa, reg, var)
-#define movdqa_r2r(regs,regd) mmx_r2r (movdqa, regs, regd)
-#define movdqu_m2r(var,reg) mmx_m2r (movdqu, var, reg)
-#define movdqu_r2m(reg,var) mmx_r2m (movdqu, reg, var)
-#define movdqu_r2r(regs,regd) mmx_r2r (movdqu, regs, regd)
+#define movdqa_m2r(var,reg) mmx_m2r (movdqa, var, reg)
+#define movdqa_r2m(reg,var) mmx_r2m (movdqa, reg, var)
+#define movdqa_r2r(regs,regd) mmx_r2r (movdqa, regs, regd)
+#define movdqu_m2r(var,reg) mmx_m2r (movdqu, var, reg)
+#define movdqu_r2m(reg,var) mmx_r2m (movdqu, reg, var)
+#define movdqu_r2r(regs,regd) mmx_r2r (movdqu, regs, regd)
-#define pmullw_r2m(reg,var) mmx_r2m (pmullw, reg, var)
+#define pmullw_r2m(reg,var) mmx_r2m (pmullw, reg, var)
-#define pslldq_i2r(imm,reg) mmx_i2r (pslldq, imm, reg)
-#define psrldq_i2r(imm,reg) mmx_i2r (psrldq, imm, reg)
+#define pslldq_i2r(imm,reg) mmx_i2r (pslldq, imm, reg)
+#define psrldq_i2r(imm,reg) mmx_i2r (psrldq, imm, reg)
-#define punpcklqdq_r2r(regs,regd) mmx_r2r (punpcklqdq, regs, regd)
-#define punpckhqdq_r2r(regs,regd) mmx_r2r (punpckhqdq, regs, regd)
+#define punpcklqdq_r2r(regs,regd) mmx_r2r (punpcklqdq, regs, regd)
+#define punpckhqdq_r2r(regs,regd) mmx_r2r (punpckhqdq, regs, regd)
#endif /* AVCODEC_I386MMX_H */
diff --git a/libavcodec/i386/motion_est_mmx.c b/libavcodec/i386/motion_est_mmx.c
index 69e10f628b..51fdd33d56 100644
--- a/libavcodec/i386/motion_est_mmx.c
+++ b/libavcodec/i386/motion_est_mmx.c
@@ -34,33 +34,33 @@ static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm2 \n\t"
- "movq (%2, %%"REG_a"), %%mm4 \n\t"
- "add %3, %%"REG_a" \n\t"
- "psubusb %%mm0, %%mm2 \n\t"
- "psubusb %%mm4, %%mm0 \n\t"
- "movq (%1, %%"REG_a"), %%mm1 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "movq (%2, %%"REG_a"), %%mm5 \n\t"
- "psubusb %%mm1, %%mm3 \n\t"
- "psubusb %%mm5, %%mm1 \n\t"
- "por %%mm2, %%mm0 \n\t"
- "por %%mm1, %%mm3 \n\t"
- "movq %%mm0, %%mm1 \n\t"
- "movq %%mm3, %%mm2 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "punpcklbw %%mm7, %%mm3 \n\t"
- "punpckhbw %%mm7, %%mm2 \n\t"
- "paddw %%mm1, %%mm0 \n\t"
- "paddw %%mm3, %%mm2 \n\t"
- "paddw %%mm2, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %3, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm2 \n\t"
+ "movq (%2, %%"REG_a"), %%mm4 \n\t"
+ "add %3, %%"REG_a" \n\t"
+ "psubusb %%mm0, %%mm2 \n\t"
+ "psubusb %%mm4, %%mm0 \n\t"
+ "movq (%1, %%"REG_a"), %%mm1 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "movq (%2, %%"REG_a"), %%mm5 \n\t"
+ "psubusb %%mm1, %%mm3 \n\t"
+ "psubusb %%mm5, %%mm1 \n\t"
+ "por %%mm2, %%mm0 \n\t"
+ "por %%mm1, %%mm3 \n\t"
+ "movq %%mm0, %%mm1 \n\t"
+ "movq %%mm3, %%mm2 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "punpcklbw %%mm7, %%mm3 \n\t"
+ "punpckhbw %%mm7, %%mm2 \n\t"
+ "paddw %%mm1, %%mm0 \n\t"
+ "paddw %%mm3, %%mm2 \n\t"
+ "paddw %%mm2, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %3, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1 - len), "r" (blk2 - len), "r" ((long)stride)
);
@@ -70,19 +70,19 @@ static inline void sad8_1_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm2 \n\t"
- "psadbw %%mm2, %%mm0 \n\t"
- "add %3, %%"REG_a" \n\t"
- "movq (%1, %%"REG_a"), %%mm1 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "psadbw %%mm1, %%mm3 \n\t"
- "paddw %%mm3, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %3, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm2 \n\t"
+ "psadbw %%mm2, %%mm0 \n\t"
+ "add %3, %%"REG_a" \n\t"
+ "movq (%1, %%"REG_a"), %%mm1 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "psadbw %%mm1, %%mm3 \n\t"
+ "paddw %%mm3, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %3, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1 - len), "r" (blk2 - len), "r" ((long)stride)
);
@@ -92,23 +92,23 @@ static inline void sad8_2_mmx2(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2, in
{
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm2 \n\t"
- "pavgb %%mm2, %%mm0 \n\t"
- "movq (%3, %%"REG_a"), %%mm2 \n\t"
- "psadbw %%mm2, %%mm0 \n\t"
- "add %4, %%"REG_a" \n\t"
- "movq (%1, %%"REG_a"), %%mm1 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "pavgb %%mm1, %%mm3 \n\t"
- "movq (%3, %%"REG_a"), %%mm1 \n\t"
- "psadbw %%mm1, %%mm3 \n\t"
- "paddw %%mm3, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %4, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm2 \n\t"
+ "pavgb %%mm2, %%mm0 \n\t"
+ "movq (%3, %%"REG_a"), %%mm2 \n\t"
+ "psadbw %%mm2, %%mm0 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ "movq (%1, %%"REG_a"), %%mm1 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "pavgb %%mm1, %%mm3 \n\t"
+ "movq (%3, %%"REG_a"), %%mm1 \n\t"
+ "psadbw %%mm1, %%mm3 \n\t"
+ "paddw %%mm3, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1a - len), "r" (blk1b -len), "r" (blk2 - len), "r" ((long)stride)
);
@@ -118,34 +118,34 @@ static inline void sad8_4_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{ //FIXME reuse src
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "movq "MANGLE(bone)", %%mm5 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm2 \n\t"
- "movq 1(%1, %%"REG_a"), %%mm1 \n\t"
- "movq 1(%2, %%"REG_a"), %%mm3 \n\t"
- "pavgb %%mm2, %%mm0 \n\t"
- "pavgb %%mm1, %%mm3 \n\t"
- "psubusb %%mm5, %%mm3 \n\t"
- "pavgb %%mm3, %%mm0 \n\t"
- "movq (%3, %%"REG_a"), %%mm2 \n\t"
- "psadbw %%mm2, %%mm0 \n\t"
- "add %4, %%"REG_a" \n\t"
- "movq (%1, %%"REG_a"), %%mm1 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
- "movq 1(%2, %%"REG_a"), %%mm4 \n\t"
- "pavgb %%mm3, %%mm1 \n\t"
- "pavgb %%mm4, %%mm2 \n\t"
- "psubusb %%mm5, %%mm2 \n\t"
- "pavgb %%mm1, %%mm2 \n\t"
- "movq (%3, %%"REG_a"), %%mm1 \n\t"
- "psadbw %%mm1, %%mm2 \n\t"
- "paddw %%mm2, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %4, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "movq "MANGLE(bone)", %%mm5 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm2 \n\t"
+ "movq 1(%1, %%"REG_a"), %%mm1 \n\t"
+ "movq 1(%2, %%"REG_a"), %%mm3 \n\t"
+ "pavgb %%mm2, %%mm0 \n\t"
+ "pavgb %%mm1, %%mm3 \n\t"
+ "psubusb %%mm5, %%mm3 \n\t"
+ "pavgb %%mm3, %%mm0 \n\t"
+ "movq (%3, %%"REG_a"), %%mm2 \n\t"
+ "psadbw %%mm2, %%mm0 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ "movq (%1, %%"REG_a"), %%mm1 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
+ "movq 1(%2, %%"REG_a"), %%mm4 \n\t"
+ "pavgb %%mm3, %%mm1 \n\t"
+ "pavgb %%mm4, %%mm2 \n\t"
+ "psubusb %%mm5, %%mm2 \n\t"
+ "pavgb %%mm1, %%mm2 \n\t"
+ "movq (%3, %%"REG_a"), %%mm1 \n\t"
+ "psadbw %%mm1, %%mm2 \n\t"
+ "paddw %%mm2, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1 - len), "r" (blk1 - len + stride), "r" (blk2 - len), "r" ((long)stride)
);
@@ -155,35 +155,35 @@ static inline void sad8_2_mmx(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2, int
{
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm2 \n\t"
- "movq (%2, %%"REG_a"), %%mm3 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm3 \n\t"
- "paddw %%mm0, %%mm1 \n\t"
- "paddw %%mm2, %%mm3 \n\t"
- "movq (%3, %%"REG_a"), %%mm4 \n\t"
- "movq (%3, %%"REG_a"), %%mm2 \n\t"
- "paddw %%mm5, %%mm1 \n\t"
- "paddw %%mm5, %%mm3 \n\t"
- "psrlw $1, %%mm1 \n\t"
- "psrlw $1, %%mm3 \n\t"
- "packuswb %%mm3, %%mm1 \n\t"
- "psubusb %%mm1, %%mm4 \n\t"
- "psubusb %%mm2, %%mm1 \n\t"
- "por %%mm4, %%mm1 \n\t"
- "movq %%mm1, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "paddw %%mm1, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %4, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm2 \n\t"
+ "movq (%2, %%"REG_a"), %%mm3 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm3 \n\t"
+ "paddw %%mm0, %%mm1 \n\t"
+ "paddw %%mm2, %%mm3 \n\t"
+ "movq (%3, %%"REG_a"), %%mm4 \n\t"
+ "movq (%3, %%"REG_a"), %%mm2 \n\t"
+ "paddw %%mm5, %%mm1 \n\t"
+ "paddw %%mm5, %%mm3 \n\t"
+ "psrlw $1, %%mm1 \n\t"
+ "psrlw $1, %%mm3 \n\t"
+ "packuswb %%mm3, %%mm1 \n\t"
+ "psubusb %%mm1, %%mm4 \n\t"
+ "psubusb %%mm2, %%mm1 \n\t"
+ "por %%mm4, %%mm1 \n\t"
+ "movq %%mm1, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "paddw %%mm1, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1a - len), "r" (blk1b -len), "r" (blk2 - len), "r" ((long)stride)
);
@@ -193,47 +193,47 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
long len= -(stride*h);
asm volatile(
- ".balign 16 \n\t"
- "1: \n\t"
- "movq (%1, %%"REG_a"), %%mm0 \n\t"
- "movq (%2, %%"REG_a"), %%mm1 \n\t"
- "movq %%mm0, %%mm4 \n\t"
- "movq %%mm1, %%mm2 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm1 \n\t"
- "punpckhbw %%mm7, %%mm4 \n\t"
- "punpckhbw %%mm7, %%mm2 \n\t"
- "paddw %%mm1, %%mm0 \n\t"
- "paddw %%mm2, %%mm4 \n\t"
- "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
- "movq 1(%2, %%"REG_a"), %%mm3 \n\t"
- "movq %%mm2, %%mm1 \n\t"
- "punpcklbw %%mm7, %%mm2 \n\t"
- "punpckhbw %%mm7, %%mm1 \n\t"
- "paddw %%mm0, %%mm2 \n\t"
- "paddw %%mm4, %%mm1 \n\t"
- "movq %%mm3, %%mm4 \n\t"
- "punpcklbw %%mm7, %%mm3 \n\t"
- "punpckhbw %%mm7, %%mm4 \n\t"
- "paddw %%mm3, %%mm2 \n\t"
- "paddw %%mm4, %%mm1 \n\t"
- "movq (%3, %%"REG_a"), %%mm3 \n\t"
- "movq (%3, %%"REG_a"), %%mm4 \n\t"
- "paddw %%mm5, %%mm2 \n\t"
- "paddw %%mm5, %%mm1 \n\t"
- "psrlw $2, %%mm2 \n\t"
- "psrlw $2, %%mm1 \n\t"
- "packuswb %%mm1, %%mm2 \n\t"
- "psubusb %%mm2, %%mm3 \n\t"
- "psubusb %%mm4, %%mm2 \n\t"
- "por %%mm3, %%mm2 \n\t"
- "movq %%mm2, %%mm0 \n\t"
- "punpcklbw %%mm7, %%mm0 \n\t"
- "punpckhbw %%mm7, %%mm2 \n\t"
- "paddw %%mm2, %%mm0 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "add %4, %%"REG_a" \n\t"
- " js 1b \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%1, %%"REG_a"), %%mm0 \n\t"
+ "movq (%2, %%"REG_a"), %%mm1 \n\t"
+ "movq %%mm0, %%mm4 \n\t"
+ "movq %%mm1, %%mm2 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm1 \n\t"
+ "punpckhbw %%mm7, %%mm4 \n\t"
+ "punpckhbw %%mm7, %%mm2 \n\t"
+ "paddw %%mm1, %%mm0 \n\t"
+ "paddw %%mm2, %%mm4 \n\t"
+ "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
+ "movq 1(%2, %%"REG_a"), %%mm3 \n\t"
+ "movq %%mm2, %%mm1 \n\t"
+ "punpcklbw %%mm7, %%mm2 \n\t"
+ "punpckhbw %%mm7, %%mm1 \n\t"
+ "paddw %%mm0, %%mm2 \n\t"
+ "paddw %%mm4, %%mm1 \n\t"
+ "movq %%mm3, %%mm4 \n\t"
+ "punpcklbw %%mm7, %%mm3 \n\t"
+ "punpckhbw %%mm7, %%mm4 \n\t"
+ "paddw %%mm3, %%mm2 \n\t"
+ "paddw %%mm4, %%mm1 \n\t"
+ "movq (%3, %%"REG_a"), %%mm3 \n\t"
+ "movq (%3, %%"REG_a"), %%mm4 \n\t"
+ "paddw %%mm5, %%mm2 \n\t"
+ "paddw %%mm5, %%mm1 \n\t"
+ "psrlw $2, %%mm2 \n\t"
+ "psrlw $2, %%mm1 \n\t"
+ "packuswb %%mm1, %%mm2 \n\t"
+ "psubusb %%mm2, %%mm3 \n\t"
+ "psubusb %%mm4, %%mm2 \n\t"
+ "por %%mm3, %%mm2 \n\t"
+ "movq %%mm2, %%mm0 \n\t"
+ "punpcklbw %%mm7, %%mm0 \n\t"
+ "punpckhbw %%mm7, %%mm2 \n\t"
+ "paddw %%mm2, %%mm0 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "add %4, %%"REG_a" \n\t"
+ " js 1b \n\t"
: "+a" (len)
: "r" (blk1 - len), "r" (blk1 -len + stride), "r" (blk2 - len), "r" ((long)stride)
);
@@ -243,13 +243,13 @@ static inline int sum_mmx(void)
{
int ret;
asm volatile(
- "movq %%mm6, %%mm0 \n\t"
- "psrlq $32, %%mm6 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "movq %%mm6, %%mm0 \n\t"
- "psrlq $16, %%mm6 \n\t"
- "paddw %%mm0, %%mm6 \n\t"
- "movd %%mm6, %0 \n\t"
+ "movq %%mm6, %%mm0 \n\t"
+ "psrlq $32, %%mm6 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "movq %%mm6, %%mm0 \n\t"
+ "psrlq $16, %%mm6 \n\t"
+ "paddw %%mm0, %%mm6 \n\t"
+ "movd %%mm6, %0 \n\t"
: "=r" (ret)
);
return ret&0xFFFF;
@@ -259,7 +259,7 @@ static inline int sum_mmx2(void)
{
int ret;
asm volatile(
- "movd %%mm6, %0 \n\t"
+ "movd %%mm6, %0 \n\t"
: "=r" (ret)
);
return ret;
@@ -270,8 +270,8 @@ static inline int sum_mmx2(void)
static int sad8_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
assert(h==8);\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t":);\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t":);\
\
sad8_1_ ## suf(blk1, blk2, stride, 8);\
\
@@ -280,9 +280,9 @@ static int sad8_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h
static int sad8_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
assert(h==8);\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[1]) \
);\
\
@@ -294,9 +294,9 @@ static int sad8_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, in
static int sad8_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
assert(h==8);\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[1]) \
);\
\
@@ -308,9 +308,9 @@ static int sad8_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, in
static int sad8_xy2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
assert(h==8);\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[2]) \
);\
\
@@ -321,8 +321,8 @@ static int sad8_xy2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, i
\
static int sad16_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t":);\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t":);\
\
sad8_1_ ## suf(blk1 , blk2 , stride, h);\
sad8_1_ ## suf(blk1+8, blk2+8, stride, h);\
@@ -331,9 +331,9 @@ static int sad16_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int
}\
static int sad16_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[1]) \
);\
\
@@ -344,9 +344,9 @@ static int sad16_x2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, i
}\
static int sad16_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[1]) \
);\
\
@@ -357,9 +357,9 @@ static int sad16_y2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, i
}\
static int sad16_xy2_ ## suf(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)\
{\
- asm volatile("pxor %%mm7, %%mm7 \n\t"\
- "pxor %%mm6, %%mm6 \n\t"\
- "movq %0, %%mm5 \n\t"\
+ asm volatile("pxor %%mm7, %%mm7 \n\t"\
+ "pxor %%mm6, %%mm6 \n\t"\
+ "movq %0, %%mm5 \n\t"\
:: "m"(round_tab[2]) \
);\
\
@@ -384,15 +384,15 @@ void dsputil_init_pix_mmx(DSPContext* c, AVCodecContext *avctx)
c->pix_abs[1][2] = sad8_y2_mmx;
c->pix_abs[1][3] = sad8_xy2_mmx;
- c->sad[0]= sad16_mmx;
+ c->sad[0]= sad16_mmx;
c->sad[1]= sad8_mmx;
}
if (mm_flags & MM_MMXEXT) {
- c->pix_abs[0][0] = sad16_mmx2;
- c->pix_abs[1][0] = sad8_mmx2;
+ c->pix_abs[0][0] = sad16_mmx2;
+ c->pix_abs[1][0] = sad8_mmx2;
- c->sad[0]= sad16_mmx2;
- c->sad[1]= sad8_mmx2;
+ c->sad[0]= sad16_mmx2;
+ c->sad[1]= sad8_mmx2;
if(!(avctx->flags & CODEC_FLAG_BITEXACT)){
c->pix_abs[0][1] = sad16_x2_mmx2;
diff --git a/libavcodec/i386/mpegvideo_mmx.c b/libavcodec/i386/mpegvideo_mmx.c
index af799b6b00..6899c3938c 100644
--- a/libavcodec/i386/mpegvideo_mmx.c
+++ b/libavcodec/i386/mpegvideo_mmx.c
@@ -57,52 +57,52 @@ static void dct_unquantize_h263_intra_mmx(MpegEncContext *s,
nCoeffs= s->inter_scantable.raster_end[ s->block_last_index[n] ];
//printf("%d %d ", qmul, qadd);
asm volatile(
- "movd %1, %%mm6 \n\t" //qmul
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "movd %2, %%mm5 \n\t" //qadd
- "pxor %%mm7, %%mm7 \n\t"
- "packssdw %%mm5, %%mm5 \n\t"
- "packssdw %%mm5, %%mm5 \n\t"
- "psubw %%mm5, %%mm7 \n\t"
- "pxor %%mm4, %%mm4 \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %3), %%mm0 \n\t"
- "movq 8(%0, %3), %%mm1 \n\t"
-
- "pmullw %%mm6, %%mm0 \n\t"
- "pmullw %%mm6, %%mm1 \n\t"
-
- "movq (%0, %3), %%mm2 \n\t"
- "movq 8(%0, %3), %%mm3 \n\t"
-
- "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
-
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
-
- "paddw %%mm7, %%mm0 \n\t"
- "paddw %%mm7, %%mm1 \n\t"
-
- "pxor %%mm0, %%mm2 \n\t"
- "pxor %%mm1, %%mm3 \n\t"
-
- "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0
-
- "pandn %%mm2, %%mm0 \n\t"
- "pandn %%mm3, %%mm1 \n\t"
-
- "movq %%mm0, (%0, %3) \n\t"
- "movq %%mm1, 8(%0, %3) \n\t"
-
- "add $16, %3 \n\t"
- "jng 1b \n\t"
- ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs))
- : "memory"
- );
+ "movd %1, %%mm6 \n\t" //qmul
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "movd %2, %%mm5 \n\t" //qadd
+ "pxor %%mm7, %%mm7 \n\t"
+ "packssdw %%mm5, %%mm5 \n\t"
+ "packssdw %%mm5, %%mm5 \n\t"
+ "psubw %%mm5, %%mm7 \n\t"
+ "pxor %%mm4, %%mm4 \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %3), %%mm0 \n\t"
+ "movq 8(%0, %3), %%mm1 \n\t"
+
+ "pmullw %%mm6, %%mm0 \n\t"
+ "pmullw %%mm6, %%mm1 \n\t"
+
+ "movq (%0, %3), %%mm2 \n\t"
+ "movq 8(%0, %3), %%mm3 \n\t"
+
+ "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+
+ "paddw %%mm7, %%mm0 \n\t"
+ "paddw %%mm7, %%mm1 \n\t"
+
+ "pxor %%mm0, %%mm2 \n\t"
+ "pxor %%mm1, %%mm3 \n\t"
+
+ "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0
+
+ "pandn %%mm2, %%mm0 \n\t"
+ "pandn %%mm3, %%mm1 \n\t"
+
+ "movq %%mm0, (%0, %3) \n\t"
+ "movq %%mm1, 8(%0, %3) \n\t"
+
+ "add $16, %3 \n\t"
+ "jng 1b \n\t"
+ ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs))
+ : "memory"
+ );
block[0]= level;
}
@@ -120,52 +120,52 @@ static void dct_unquantize_h263_inter_mmx(MpegEncContext *s,
nCoeffs= s->inter_scantable.raster_end[ s->block_last_index[n] ];
//printf("%d %d ", qmul, qadd);
asm volatile(
- "movd %1, %%mm6 \n\t" //qmul
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "movd %2, %%mm5 \n\t" //qadd
- "pxor %%mm7, %%mm7 \n\t"
- "packssdw %%mm5, %%mm5 \n\t"
- "packssdw %%mm5, %%mm5 \n\t"
- "psubw %%mm5, %%mm7 \n\t"
- "pxor %%mm4, %%mm4 \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %3), %%mm0 \n\t"
- "movq 8(%0, %3), %%mm1 \n\t"
-
- "pmullw %%mm6, %%mm0 \n\t"
- "pmullw %%mm6, %%mm1 \n\t"
-
- "movq (%0, %3), %%mm2 \n\t"
- "movq 8(%0, %3), %%mm3 \n\t"
-
- "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
-
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
-
- "paddw %%mm7, %%mm0 \n\t"
- "paddw %%mm7, %%mm1 \n\t"
-
- "pxor %%mm0, %%mm2 \n\t"
- "pxor %%mm1, %%mm3 \n\t"
-
- "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0
-
- "pandn %%mm2, %%mm0 \n\t"
- "pandn %%mm3, %%mm1 \n\t"
-
- "movq %%mm0, (%0, %3) \n\t"
- "movq %%mm1, 8(%0, %3) \n\t"
-
- "add $16, %3 \n\t"
- "jng 1b \n\t"
- ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs))
- : "memory"
- );
+ "movd %1, %%mm6 \n\t" //qmul
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "movd %2, %%mm5 \n\t" //qadd
+ "pxor %%mm7, %%mm7 \n\t"
+ "packssdw %%mm5, %%mm5 \n\t"
+ "packssdw %%mm5, %%mm5 \n\t"
+ "psubw %%mm5, %%mm7 \n\t"
+ "pxor %%mm4, %%mm4 \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %3), %%mm0 \n\t"
+ "movq 8(%0, %3), %%mm1 \n\t"
+
+ "pmullw %%mm6, %%mm0 \n\t"
+ "pmullw %%mm6, %%mm1 \n\t"
+
+ "movq (%0, %3), %%mm2 \n\t"
+ "movq 8(%0, %3), %%mm3 \n\t"
+
+ "pcmpgtw %%mm4, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm4, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+
+ "paddw %%mm7, %%mm0 \n\t"
+ "paddw %%mm7, %%mm1 \n\t"
+
+ "pxor %%mm0, %%mm2 \n\t"
+ "pxor %%mm1, %%mm3 \n\t"
+
+ "pcmpeqw %%mm7, %%mm0 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw %%mm7, %%mm1 \n\t" // block[i] == 0 ? -1 : 0
+
+ "pandn %%mm2, %%mm0 \n\t"
+ "pandn %%mm3, %%mm1 \n\t"
+
+ "movq %%mm0, (%0, %3) \n\t"
+ "movq %%mm1, 8(%0, %3) \n\t"
+
+ "add $16, %3 \n\t"
+ "jng 1b \n\t"
+ ::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(-nCoeffs))
+ : "memory"
+ );
}
@@ -216,54 +216,54 @@ static void dct_unquantize_mpeg1_intra_mmx(MpegEncContext *s,
/* XXX: only mpeg1 */
quant_matrix = s->intra_matrix;
asm volatile(
- "pcmpeqw %%mm7, %%mm7 \n\t"
- "psrlw $15, %%mm7 \n\t"
- "movd %2, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "mov %3, %%"REG_a" \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %%"REG_a"), %%mm0 \n\t"
- "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm4 \n\t"
- "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
- "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
- "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
- "pxor %%mm2, %%mm2 \n\t"
- "pxor %%mm3, %%mm3 \n\t"
- "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
- "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
- "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q
- "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q
- "pxor %%mm4, %%mm4 \n\t"
- "pxor %%mm5, %%mm5 \n\t" // FIXME slow
- "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
- "psraw $3, %%mm0 \n\t"
- "psraw $3, %%mm1 \n\t"
- "psubw %%mm7, %%mm0 \n\t"
- "psubw %%mm7, %%mm1 \n\t"
- "por %%mm7, %%mm0 \n\t"
- "por %%mm7, %%mm1 \n\t"
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t"
- "psubw %%mm3, %%mm1 \n\t"
- "pandn %%mm0, %%mm4 \n\t"
- "pandn %%mm1, %%mm5 \n\t"
- "movq %%mm4, (%0, %%"REG_a") \n\t"
- "movq %%mm5, 8(%0, %%"REG_a") \n\t"
-
- "add $16, %%"REG_a" \n\t"
- "js 1b \n\t"
- ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
- : "%"REG_a, "memory"
- );
+ "pcmpeqw %%mm7, %%mm7 \n\t"
+ "psrlw $15, %%mm7 \n\t"
+ "movd %2, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "mov %3, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %%"REG_a"), %%mm0 \n\t"
+ "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm4 \n\t"
+ "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
+ "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
+ "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
+ "pxor %%mm2, %%mm2 \n\t"
+ "pxor %%mm3, %%mm3 \n\t"
+ "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
+ "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
+ "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q
+ "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q
+ "pxor %%mm4, %%mm4 \n\t"
+ "pxor %%mm5, %%mm5 \n\t" // FIXME slow
+ "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+ "psraw $3, %%mm0 \n\t"
+ "psraw $3, %%mm1 \n\t"
+ "psubw %%mm7, %%mm0 \n\t"
+ "psubw %%mm7, %%mm1 \n\t"
+ "por %%mm7, %%mm0 \n\t"
+ "por %%mm7, %%mm1 \n\t"
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t"
+ "psubw %%mm3, %%mm1 \n\t"
+ "pandn %%mm0, %%mm4 \n\t"
+ "pandn %%mm1, %%mm5 \n\t"
+ "movq %%mm4, (%0, %%"REG_a") \n\t"
+ "movq %%mm5, 8(%0, %%"REG_a") \n\t"
+
+ "add $16, %%"REG_a" \n\t"
+ "js 1b \n\t"
+ ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+ : "%"REG_a, "memory"
+ );
block[0]= block0;
}
@@ -279,58 +279,58 @@ static void dct_unquantize_mpeg1_inter_mmx(MpegEncContext *s,
quant_matrix = s->inter_matrix;
asm volatile(
- "pcmpeqw %%mm7, %%mm7 \n\t"
- "psrlw $15, %%mm7 \n\t"
- "movd %2, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "mov %3, %%"REG_a" \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %%"REG_a"), %%mm0 \n\t"
- "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm4 \n\t"
- "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
- "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
- "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
- "pxor %%mm2, %%mm2 \n\t"
- "pxor %%mm3, %%mm3 \n\t"
- "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
- "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
- "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2
- "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2
- "paddw %%mm7, %%mm0 \n\t" // abs(block[i])*2 + 1
- "paddw %%mm7, %%mm1 \n\t" // abs(block[i])*2 + 1
- "pmullw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q
- "pmullw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q
- "pxor %%mm4, %%mm4 \n\t"
- "pxor %%mm5, %%mm5 \n\t" // FIXME slow
- "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
- "psraw $4, %%mm0 \n\t"
- "psraw $4, %%mm1 \n\t"
- "psubw %%mm7, %%mm0 \n\t"
- "psubw %%mm7, %%mm1 \n\t"
- "por %%mm7, %%mm0 \n\t"
- "por %%mm7, %%mm1 \n\t"
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t"
- "psubw %%mm3, %%mm1 \n\t"
- "pandn %%mm0, %%mm4 \n\t"
- "pandn %%mm1, %%mm5 \n\t"
- "movq %%mm4, (%0, %%"REG_a") \n\t"
- "movq %%mm5, 8(%0, %%"REG_a") \n\t"
-
- "add $16, %%"REG_a" \n\t"
- "js 1b \n\t"
- ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
- : "%"REG_a, "memory"
- );
+ "pcmpeqw %%mm7, %%mm7 \n\t"
+ "psrlw $15, %%mm7 \n\t"
+ "movd %2, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "mov %3, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %%"REG_a"), %%mm0 \n\t"
+ "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm4 \n\t"
+ "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
+ "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
+ "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
+ "pxor %%mm2, %%mm2 \n\t"
+ "pxor %%mm3, %%mm3 \n\t"
+ "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
+ "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
+ "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2
+ "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2
+ "paddw %%mm7, %%mm0 \n\t" // abs(block[i])*2 + 1
+ "paddw %%mm7, %%mm1 \n\t" // abs(block[i])*2 + 1
+ "pmullw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q
+ "pmullw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q
+ "pxor %%mm4, %%mm4 \n\t"
+ "pxor %%mm5, %%mm5 \n\t" // FIXME slow
+ "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+ "psraw $4, %%mm0 \n\t"
+ "psraw $4, %%mm1 \n\t"
+ "psubw %%mm7, %%mm0 \n\t"
+ "psubw %%mm7, %%mm1 \n\t"
+ "por %%mm7, %%mm0 \n\t"
+ "por %%mm7, %%mm1 \n\t"
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t"
+ "psubw %%mm3, %%mm1 \n\t"
+ "pandn %%mm0, %%mm4 \n\t"
+ "pandn %%mm1, %%mm5 \n\t"
+ "movq %%mm4, (%0, %%"REG_a") \n\t"
+ "movq %%mm5, 8(%0, %%"REG_a") \n\t"
+
+ "add $16, %%"REG_a" \n\t"
+ "js 1b \n\t"
+ ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+ : "%"REG_a, "memory"
+ );
}
static void dct_unquantize_mpeg2_intra_mmx(MpegEncContext *s,
@@ -351,50 +351,50 @@ static void dct_unquantize_mpeg2_intra_mmx(MpegEncContext *s,
block0 = block[0] * s->c_dc_scale;
quant_matrix = s->intra_matrix;
asm volatile(
- "pcmpeqw %%mm7, %%mm7 \n\t"
- "psrlw $15, %%mm7 \n\t"
- "movd %2, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "mov %3, %%"REG_a" \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %%"REG_a"), %%mm0 \n\t"
- "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm4 \n\t"
- "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
- "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
- "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
- "pxor %%mm2, %%mm2 \n\t"
- "pxor %%mm3, %%mm3 \n\t"
- "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
- "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
- "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q
- "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q
- "pxor %%mm4, %%mm4 \n\t"
- "pxor %%mm5, %%mm5 \n\t" // FIXME slow
- "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
- "psraw $3, %%mm0 \n\t"
- "psraw $3, %%mm1 \n\t"
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t"
- "psubw %%mm3, %%mm1 \n\t"
- "pandn %%mm0, %%mm4 \n\t"
- "pandn %%mm1, %%mm5 \n\t"
- "movq %%mm4, (%0, %%"REG_a") \n\t"
- "movq %%mm5, 8(%0, %%"REG_a") \n\t"
-
- "add $16, %%"REG_a" \n\t"
- "jng 1b \n\t"
- ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
- : "%"REG_a, "memory"
- );
+ "pcmpeqw %%mm7, %%mm7 \n\t"
+ "psrlw $15, %%mm7 \n\t"
+ "movd %2, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "mov %3, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %%"REG_a"), %%mm0 \n\t"
+ "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm4 \n\t"
+ "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
+ "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
+ "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
+ "pxor %%mm2, %%mm2 \n\t"
+ "pxor %%mm3, %%mm3 \n\t"
+ "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
+ "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
+ "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*q
+ "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*q
+ "pxor %%mm4, %%mm4 \n\t"
+ "pxor %%mm5, %%mm5 \n\t" // FIXME slow
+ "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+ "psraw $3, %%mm0 \n\t"
+ "psraw $3, %%mm1 \n\t"
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t"
+ "psubw %%mm3, %%mm1 \n\t"
+ "pandn %%mm0, %%mm4 \n\t"
+ "pandn %%mm1, %%mm5 \n\t"
+ "movq %%mm4, (%0, %%"REG_a") \n\t"
+ "movq %%mm5, 8(%0, %%"REG_a") \n\t"
+
+ "add $16, %%"REG_a" \n\t"
+ "jng 1b \n\t"
+ ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+ : "%"REG_a, "memory"
+ );
block[0]= block0;
//Note, we dont do mismatch control for intra as errors cannot accumulate
}
@@ -412,68 +412,68 @@ static void dct_unquantize_mpeg2_inter_mmx(MpegEncContext *s,
quant_matrix = s->inter_matrix;
asm volatile(
- "pcmpeqw %%mm7, %%mm7 \n\t"
- "psrlq $48, %%mm7 \n\t"
- "movd %2, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "packssdw %%mm6, %%mm6 \n\t"
- "mov %3, %%"REG_a" \n\t"
- ".balign 16\n\t"
- "1: \n\t"
- "movq (%0, %%"REG_a"), %%mm0 \n\t"
- "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
- "movq (%1, %%"REG_a"), %%mm4 \n\t"
- "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
- "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
- "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
- "pxor %%mm2, %%mm2 \n\t"
- "pxor %%mm3, %%mm3 \n\t"
- "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
- "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
- "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
- "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2
- "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2
- "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*2*q
- "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*2*q
- "paddw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q
- "paddw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q
- "pxor %%mm4, %%mm4 \n\t"
- "pxor %%mm5, %%mm5 \n\t" // FIXME slow
- "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
- "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
- "psrlw $4, %%mm0 \n\t"
- "psrlw $4, %%mm1 \n\t"
- "pxor %%mm2, %%mm0 \n\t"
- "pxor %%mm3, %%mm1 \n\t"
- "psubw %%mm2, %%mm0 \n\t"
- "psubw %%mm3, %%mm1 \n\t"
- "pandn %%mm0, %%mm4 \n\t"
- "pandn %%mm1, %%mm5 \n\t"
- "pxor %%mm4, %%mm7 \n\t"
- "pxor %%mm5, %%mm7 \n\t"
- "movq %%mm4, (%0, %%"REG_a") \n\t"
- "movq %%mm5, 8(%0, %%"REG_a") \n\t"
-
- "add $16, %%"REG_a" \n\t"
- "jng 1b \n\t"
- "movd 124(%0, %3), %%mm0 \n\t"
- "movq %%mm7, %%mm6 \n\t"
- "psrlq $32, %%mm7 \n\t"
- "pxor %%mm6, %%mm7 \n\t"
- "movq %%mm7, %%mm6 \n\t"
- "psrlq $16, %%mm7 \n\t"
- "pxor %%mm6, %%mm7 \n\t"
- "pslld $31, %%mm7 \n\t"
- "psrlq $15, %%mm7 \n\t"
- "pxor %%mm7, %%mm0 \n\t"
- "movd %%mm0, 124(%0, %3) \n\t"
-
- ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs)
- : "%"REG_a, "memory"
- );
+ "pcmpeqw %%mm7, %%mm7 \n\t"
+ "psrlq $48, %%mm7 \n\t"
+ "movd %2, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "packssdw %%mm6, %%mm6 \n\t"
+ "mov %3, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "movq (%0, %%"REG_a"), %%mm0 \n\t"
+ "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
+ "movq (%1, %%"REG_a"), %%mm4 \n\t"
+ "movq 8(%1, %%"REG_a"), %%mm5 \n\t"
+ "pmullw %%mm6, %%mm4 \n\t" // q=qscale*quant_matrix[i]
+ "pmullw %%mm6, %%mm5 \n\t" // q=qscale*quant_matrix[i]
+ "pxor %%mm2, %%mm2 \n\t"
+ "pxor %%mm3, %%mm3 \n\t"
+ "pcmpgtw %%mm0, %%mm2 \n\t" // block[i] < 0 ? -1 : 0
+ "pcmpgtw %%mm1, %%mm3 \n\t" // block[i] < 0 ? -1 : 0
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t" // abs(block[i])
+ "psubw %%mm3, %%mm1 \n\t" // abs(block[i])
+ "paddw %%mm0, %%mm0 \n\t" // abs(block[i])*2
+ "paddw %%mm1, %%mm1 \n\t" // abs(block[i])*2
+ "pmullw %%mm4, %%mm0 \n\t" // abs(block[i])*2*q
+ "pmullw %%mm5, %%mm1 \n\t" // abs(block[i])*2*q
+ "paddw %%mm4, %%mm0 \n\t" // (abs(block[i])*2 + 1)*q
+ "paddw %%mm5, %%mm1 \n\t" // (abs(block[i])*2 + 1)*q
+ "pxor %%mm4, %%mm4 \n\t"
+ "pxor %%mm5, %%mm5 \n\t" // FIXME slow
+ "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+ "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+ "psrlw $4, %%mm0 \n\t"
+ "psrlw $4, %%mm1 \n\t"
+ "pxor %%mm2, %%mm0 \n\t"
+ "pxor %%mm3, %%mm1 \n\t"
+ "psubw %%mm2, %%mm0 \n\t"
+ "psubw %%mm3, %%mm1 \n\t"
+ "pandn %%mm0, %%mm4 \n\t"
+ "pandn %%mm1, %%mm5 \n\t"
+ "pxor %%mm4, %%mm7 \n\t"
+ "pxor %%mm5, %%mm7 \n\t"
+ "movq %%mm4, (%0, %%"REG_a") \n\t"
+ "movq %%mm5, 8(%0, %%"REG_a") \n\t"
+
+ "add $16, %%"REG_a" \n\t"
+ "jng 1b \n\t"
+ "movd 124(%0, %3), %%mm0 \n\t"
+ "movq %%mm7, %%mm6 \n\t"
+ "psrlq $32, %%mm7 \n\t"
+ "pxor %%mm6, %%mm7 \n\t"
+ "movq %%mm7, %%mm6 \n\t"
+ "psrlq $16, %%mm7 \n\t"
+ "pxor %%mm6, %%mm7 \n\t"
+ "pslld $31, %%mm7 \n\t"
+ "psrlq $15, %%mm7 \n\t"
+ "pxor %%mm7, %%mm0 \n\t"
+ "movd %%mm0, 124(%0, %3) \n\t"
+
+ ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs)
+ : "%"REG_a, "memory"
+ );
}
/* draw the edges of width 'w' of an image of size width, height
@@ -488,79 +488,79 @@ static void draw_edges_mmx(uint8_t *buf, int wrap, int width, int height, int w)
ptr = buf;
if(w==8)
{
- asm volatile(
- "1: \n\t"
- "movd (%0), %%mm0 \n\t"
- "punpcklbw %%mm0, %%mm0 \n\t"
- "punpcklwd %%mm0, %%mm0 \n\t"
- "punpckldq %%mm0, %%mm0 \n\t"
- "movq %%mm0, -8(%0) \n\t"
- "movq -8(%0, %2), %%mm1 \n\t"
- "punpckhbw %%mm1, %%mm1 \n\t"
- "punpckhwd %%mm1, %%mm1 \n\t"
- "punpckhdq %%mm1, %%mm1 \n\t"
- "movq %%mm1, (%0, %2) \n\t"
- "add %1, %0 \n\t"
- "cmp %3, %0 \n\t"
- " jb 1b \n\t"
- : "+r" (ptr)
- : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height)
- );
+ asm volatile(
+ "1: \n\t"
+ "movd (%0), %%mm0 \n\t"
+ "punpcklbw %%mm0, %%mm0 \n\t"
+ "punpcklwd %%mm0, %%mm0 \n\t"
+ "punpckldq %%mm0, %%mm0 \n\t"
+ "movq %%mm0, -8(%0) \n\t"
+ "movq -8(%0, %2), %%mm1 \n\t"
+ "punpckhbw %%mm1, %%mm1 \n\t"
+ "punpckhwd %%mm1, %%mm1 \n\t"
+ "punpckhdq %%mm1, %%mm1 \n\t"
+ "movq %%mm1, (%0, %2) \n\t"
+ "add %1, %0 \n\t"
+ "cmp %3, %0 \n\t"
+ " jb 1b \n\t"
+ : "+r" (ptr)
+ : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height)
+ );
}
else
{
- asm volatile(
- "1: \n\t"
- "movd (%0), %%mm0 \n\t"
- "punpcklbw %%mm0, %%mm0 \n\t"
- "punpcklwd %%mm0, %%mm0 \n\t"
- "punpckldq %%mm0, %%mm0 \n\t"
- "movq %%mm0, -8(%0) \n\t"
- "movq %%mm0, -16(%0) \n\t"
- "movq -8(%0, %2), %%mm1 \n\t"
- "punpckhbw %%mm1, %%mm1 \n\t"
- "punpckhwd %%mm1, %%mm1 \n\t"
- "punpckhdq %%mm1, %%mm1 \n\t"
- "movq %%mm1, (%0, %2) \n\t"
- "movq %%mm1, 8(%0, %2) \n\t"
- "add %1, %0 \n\t"
- "cmp %3, %0 \n\t"
- " jb 1b \n\t"
- : "+r" (ptr)
- : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height)
- );
+ asm volatile(
+ "1: \n\t"
+ "movd (%0), %%mm0 \n\t"
+ "punpcklbw %%mm0, %%mm0 \n\t"
+ "punpcklwd %%mm0, %%mm0 \n\t"
+ "punpckldq %%mm0, %%mm0 \n\t"
+ "movq %%mm0, -8(%0) \n\t"
+ "movq %%mm0, -16(%0) \n\t"
+ "movq -8(%0, %2), %%mm1 \n\t"
+ "punpckhbw %%mm1, %%mm1 \n\t"
+ "punpckhwd %%mm1, %%mm1 \n\t"
+ "punpckhdq %%mm1, %%mm1 \n\t"
+ "movq %%mm1, (%0, %2) \n\t"
+ "movq %%mm1, 8(%0, %2) \n\t"
+ "add %1, %0 \n\t"
+ "cmp %3, %0 \n\t"
+ " jb 1b \n\t"
+ : "+r" (ptr)
+ : "r" ((long)wrap), "r" ((long)width), "r" (ptr + wrap*height)
+ );
}
for(i=0;i<w;i+=4) {
/* top and bottom (and hopefully also the corners) */
- ptr= buf - (i + 1) * wrap - w;
- asm volatile(
- "1: \n\t"
- "movq (%1, %0), %%mm0 \n\t"
- "movq %%mm0, (%0) \n\t"
- "movq %%mm0, (%0, %2) \n\t"
- "movq %%mm0, (%0, %2, 2) \n\t"
- "movq %%mm0, (%0, %3) \n\t"
- "add $8, %0 \n\t"
- "cmp %4, %0 \n\t"
- " jb 1b \n\t"
- : "+r" (ptr)
- : "r" ((long)buf - (long)ptr - w), "r" ((long)-wrap), "r" ((long)-wrap*3), "r" (ptr+width+2*w)
- );
- ptr= last_line + (i + 1) * wrap - w;
- asm volatile(
- "1: \n\t"
- "movq (%1, %0), %%mm0 \n\t"
- "movq %%mm0, (%0) \n\t"
- "movq %%mm0, (%0, %2) \n\t"
- "movq %%mm0, (%0, %2, 2) \n\t"
- "movq %%mm0, (%0, %3) \n\t"
- "add $8, %0 \n\t"
- "cmp %4, %0 \n\t"
- " jb 1b \n\t"
- : "+r" (ptr)
- : "r" ((long)last_line - (long)ptr - w), "r" ((long)wrap), "r" ((long)wrap*3), "r" (ptr+width+2*w)
- );
+ ptr= buf - (i + 1) * wrap - w;
+ asm volatile(
+ "1: \n\t"
+ "movq (%1, %0), %%mm0 \n\t"
+ "movq %%mm0, (%0) \n\t"
+ "movq %%mm0, (%0, %2) \n\t"
+ "movq %%mm0, (%0, %2, 2) \n\t"
+ "movq %%mm0, (%0, %3) \n\t"
+ "add $8, %0 \n\t"
+ "cmp %4, %0 \n\t"
+ " jb 1b \n\t"
+ : "+r" (ptr)
+ : "r" ((long)buf - (long)ptr - w), "r" ((long)-wrap), "r" ((long)-wrap*3), "r" (ptr+width+2*w)
+ );
+ ptr= last_line + (i + 1) * wrap - w;
+ asm volatile(
+ "1: \n\t"
+ "movq (%1, %0), %%mm0 \n\t"
+ "movq %%mm0, (%0) \n\t"
+ "movq %%mm0, (%0, %2) \n\t"
+ "movq %%mm0, (%0, %2, 2) \n\t"
+ "movq %%mm0, (%0, %3) \n\t"
+ "add $8, %0 \n\t"
+ "cmp %4, %0 \n\t"
+ " jb 1b \n\t"
+ : "+r" (ptr)
+ : "r" ((long)last_line - (long)ptr - w), "r" ((long)wrap), "r" ((long)wrap*3), "r" (ptr+width+2*w)
+ );
}
}
@@ -572,47 +572,47 @@ static void denoise_dct_mmx(MpegEncContext *s, DCTELEM *block){
s->dct_count[intra]++;
asm volatile(
- "pxor %%mm7, %%mm7 \n\t"
- "1: \n\t"
- "pxor %%mm0, %%mm0 \n\t"
- "pxor %%mm1, %%mm1 \n\t"
- "movq (%0), %%mm2 \n\t"
- "movq 8(%0), %%mm3 \n\t"
- "pcmpgtw %%mm2, %%mm0 \n\t"
- "pcmpgtw %%mm3, %%mm1 \n\t"
- "pxor %%mm0, %%mm2 \n\t"
- "pxor %%mm1, %%mm3 \n\t"
- "psubw %%mm0, %%mm2 \n\t"
- "psubw %%mm1, %%mm3 \n\t"
- "movq %%mm2, %%mm4 \n\t"
- "movq %%mm3, %%mm5 \n\t"
- "psubusw (%2), %%mm2 \n\t"
- "psubusw 8(%2), %%mm3 \n\t"
- "pxor %%mm0, %%mm2 \n\t"
- "pxor %%mm1, %%mm3 \n\t"
- "psubw %%mm0, %%mm2 \n\t"
- "psubw %%mm1, %%mm3 \n\t"
- "movq %%mm2, (%0) \n\t"
- "movq %%mm3, 8(%0) \n\t"
- "movq %%mm4, %%mm2 \n\t"
- "movq %%mm5, %%mm3 \n\t"
- "punpcklwd %%mm7, %%mm4 \n\t"
- "punpckhwd %%mm7, %%mm2 \n\t"
- "punpcklwd %%mm7, %%mm5 \n\t"
- "punpckhwd %%mm7, %%mm3 \n\t"
- "paddd (%1), %%mm4 \n\t"
- "paddd 8(%1), %%mm2 \n\t"
- "paddd 16(%1), %%mm5 \n\t"
- "paddd 24(%1), %%mm3 \n\t"
- "movq %%mm4, (%1) \n\t"
- "movq %%mm2, 8(%1) \n\t"
- "movq %%mm5, 16(%1) \n\t"
- "movq %%mm3, 24(%1) \n\t"
- "add $16, %0 \n\t"
- "add $32, %1 \n\t"
- "add $16, %2 \n\t"
- "cmp %3, %0 \n\t"
- " jb 1b \n\t"
+ "pxor %%mm7, %%mm7 \n\t"
+ "1: \n\t"
+ "pxor %%mm0, %%mm0 \n\t"
+ "pxor %%mm1, %%mm1 \n\t"
+ "movq (%0), %%mm2 \n\t"
+ "movq 8(%0), %%mm3 \n\t"
+ "pcmpgtw %%mm2, %%mm0 \n\t"
+ "pcmpgtw %%mm3, %%mm1 \n\t"
+ "pxor %%mm0, %%mm2 \n\t"
+ "pxor %%mm1, %%mm3 \n\t"
+ "psubw %%mm0, %%mm2 \n\t"
+ "psubw %%mm1, %%mm3 \n\t"
+ "movq %%mm2, %%mm4 \n\t"
+ "movq %%mm3, %%mm5 \n\t"
+ "psubusw (%2), %%mm2 \n\t"
+ "psubusw 8(%2), %%mm3 \n\t"
+ "pxor %%mm0, %%mm2 \n\t"
+ "pxor %%mm1, %%mm3 \n\t"
+ "psubw %%mm0, %%mm2 \n\t"
+ "psubw %%mm1, %%mm3 \n\t"
+ "movq %%mm2, (%0) \n\t"
+ "movq %%mm3, 8(%0) \n\t"
+ "movq %%mm4, %%mm2 \n\t"
+ "movq %%mm5, %%mm3 \n\t"
+ "punpcklwd %%mm7, %%mm4 \n\t"
+ "punpckhwd %%mm7, %%mm2 \n\t"
+ "punpcklwd %%mm7, %%mm5 \n\t"
+ "punpckhwd %%mm7, %%mm3 \n\t"
+ "paddd (%1), %%mm4 \n\t"
+ "paddd 8(%1), %%mm2 \n\t"
+ "paddd 16(%1), %%mm5 \n\t"
+ "paddd 24(%1), %%mm3 \n\t"
+ "movq %%mm4, (%1) \n\t"
+ "movq %%mm2, 8(%1) \n\t"
+ "movq %%mm5, 16(%1) \n\t"
+ "movq %%mm3, 24(%1) \n\t"
+ "add $16, %0 \n\t"
+ "add $32, %1 \n\t"
+ "add $16, %2 \n\t"
+ "cmp %3, %0 \n\t"
+ " jb 1b \n\t"
: "+r" (block), "+r" (sum), "+r" (offset)
: "r"(block+64)
);
@@ -626,47 +626,47 @@ static void denoise_dct_sse2(MpegEncContext *s, DCTELEM *block){
s->dct_count[intra]++;
asm volatile(
- "pxor %%xmm7, %%xmm7 \n\t"
- "1: \n\t"
- "pxor %%xmm0, %%xmm0 \n\t"
- "pxor %%xmm1, %%xmm1 \n\t"
- "movdqa (%0), %%xmm2 \n\t"
- "movdqa 16(%0), %%xmm3 \n\t"
- "pcmpgtw %%xmm2, %%xmm0 \n\t"
- "pcmpgtw %%xmm3, %%xmm1 \n\t"
- "pxor %%xmm0, %%xmm2 \n\t"
- "pxor %%xmm1, %%xmm3 \n\t"
- "psubw %%xmm0, %%xmm2 \n\t"
- "psubw %%xmm1, %%xmm3 \n\t"
- "movdqa %%xmm2, %%xmm4 \n\t"
- "movdqa %%xmm3, %%xmm5 \n\t"
- "psubusw (%2), %%xmm2 \n\t"
- "psubusw 16(%2), %%xmm3 \n\t"
- "pxor %%xmm0, %%xmm2 \n\t"
- "pxor %%xmm1, %%xmm3 \n\t"
- "psubw %%xmm0, %%xmm2 \n\t"
- "psubw %%xmm1, %%xmm3 \n\t"
- "movdqa %%xmm2, (%0) \n\t"
- "movdqa %%xmm3, 16(%0) \n\t"
- "movdqa %%xmm4, %%xmm6 \n\t"
- "movdqa %%xmm5, %%xmm0 \n\t"
- "punpcklwd %%xmm7, %%xmm4 \n\t"
- "punpckhwd %%xmm7, %%xmm6 \n\t"
- "punpcklwd %%xmm7, %%xmm5 \n\t"
- "punpckhwd %%xmm7, %%xmm0 \n\t"
- "paddd (%1), %%xmm4 \n\t"
- "paddd 16(%1), %%xmm6 \n\t"
- "paddd 32(%1), %%xmm5 \n\t"
- "paddd 48(%1), %%xmm0 \n\t"
- "movdqa %%xmm4, (%1) \n\t"
- "movdqa %%xmm6, 16(%1) \n\t"
- "movdqa %%xmm5, 32(%1) \n\t"
- "movdqa %%xmm0, 48(%1) \n\t"
- "add $32, %0 \n\t"
- "add $64, %1 \n\t"
- "add $32, %2 \n\t"
- "cmp %3, %0 \n\t"
- " jb 1b \n\t"
+ "pxor %%xmm7, %%xmm7 \n\t"
+ "1: \n\t"
+ "pxor %%xmm0, %%xmm0 \n\t"
+ "pxor %%xmm1, %%xmm1 \n\t"
+ "movdqa (%0), %%xmm2 \n\t"
+ "movdqa 16(%0), %%xmm3 \n\t"
+ "pcmpgtw %%xmm2, %%xmm0 \n\t"
+ "pcmpgtw %%xmm3, %%xmm1 \n\t"
+ "pxor %%xmm0, %%xmm2 \n\t"
+ "pxor %%xmm1, %%xmm3 \n\t"
+ "psubw %%xmm0, %%xmm2 \n\t"
+ "psubw %%xmm1, %%xmm3 \n\t"
+ "movdqa %%xmm2, %%xmm4 \n\t"
+ "movdqa %%xmm3, %%xmm5 \n\t"
+ "psubusw (%2), %%xmm2 \n\t"
+ "psubusw 16(%2), %%xmm3 \n\t"
+ "pxor %%xmm0, %%xmm2 \n\t"
+ "pxor %%xmm1, %%xmm3 \n\t"
+ "psubw %%xmm0, %%xmm2 \n\t"
+ "psubw %%xmm1, %%xmm3 \n\t"
+ "movdqa %%xmm2, (%0) \n\t"
+ "movdqa %%xmm3, 16(%0) \n\t"
+ "movdqa %%xmm4, %%xmm6 \n\t"
+ "movdqa %%xmm5, %%xmm0 \n\t"
+ "punpcklwd %%xmm7, %%xmm4 \n\t"
+ "punpckhwd %%xmm7, %%xmm6 \n\t"
+ "punpcklwd %%xmm7, %%xmm5 \n\t"
+ "punpckhwd %%xmm7, %%xmm0 \n\t"
+ "paddd (%1), %%xmm4 \n\t"
+ "paddd 16(%1), %%xmm6 \n\t"
+ "paddd 32(%1), %%xmm5 \n\t"
+ "paddd 48(%1), %%xmm0 \n\t"
+ "movdqa %%xmm4, (%1) \n\t"
+ "movdqa %%xmm6, 16(%1) \n\t"
+ "movdqa %%xmm5, 32(%1) \n\t"
+ "movdqa %%xmm0, 48(%1) \n\t"
+ "add $32, %0 \n\t"
+ "add $64, %1 \n\t"
+ "add $32, %2 \n\t"
+ "cmp %3, %0 \n\t"
+ " jb 1b \n\t"
: "+r" (block), "+r" (sum), "+r" (offset)
: "r"(block+64)
);
@@ -705,10 +705,10 @@ void MPV_common_init_mmx(MpegEncContext *s)
draw_edges = draw_edges_mmx;
if (mm_flags & MM_SSE2) {
- s->denoise_dct= denoise_dct_sse2;
- } else {
- s->denoise_dct= denoise_dct_mmx;
- }
+ s->denoise_dct= denoise_dct_sse2;
+ } else {
+ s->denoise_dct= denoise_dct_mmx;
+ }
if(dct_algo==FF_DCT_AUTO || dct_algo==FF_DCT_MMX){
if(mm_flags & MM_SSE2){
diff --git a/libavcodec/i386/mpegvideo_mmx_template.c b/libavcodec/i386/mpegvideo_mmx_template.c
index 28afdeef0c..01eb3b95e6 100644
--- a/libavcodec/i386/mpegvideo_mmx_template.c
+++ b/libavcodec/i386/mpegvideo_mmx_template.c
@@ -21,26 +21,26 @@
#undef PMAXW
#ifdef HAVE_MMX2
#define SPREADW(a) "pshufw $0, " #a ", " #a " \n\t"
-#define PMAXW(a,b) "pmaxsw " #a ", " #b " \n\t"
+#define PMAXW(a,b) "pmaxsw " #a ", " #b " \n\t"
#define PMAX(a,b) \
- "pshufw $0x0E," #a ", " #b " \n\t"\
- PMAXW(b, a)\
- "pshufw $0x01," #a ", " #b " \n\t"\
- PMAXW(b, a)
+ "pshufw $0x0E," #a ", " #b " \n\t"\
+ PMAXW(b, a)\
+ "pshufw $0x01," #a ", " #b " \n\t"\
+ PMAXW(b, a)
#else
#define SPREADW(a) \
- "punpcklwd " #a ", " #a " \n\t"\
- "punpcklwd " #a ", " #a " \n\t"
+ "punpcklwd " #a ", " #a " \n\t"\
+ "punpcklwd " #a ", " #a " \n\t"
#define PMAXW(a,b) \
- "psubusw " #a ", " #b " \n\t"\
- "paddw " #a ", " #b " \n\t"
+ "psubusw " #a ", " #b " \n\t"\
+ "paddw " #a ", " #b " \n\t"
#define PMAX(a,b) \
- "movq " #a ", " #b " \n\t"\
- "psrlq $32, " #a " \n\t"\
- PMAXW(b, a)\
- "movq " #a ", " #b " \n\t"\
- "psrlq $16, " #a " \n\t"\
- PMAXW(b, a)
+ "movq " #a ", " #b " \n\t"\
+ "psrlq $32, " #a " \n\t"\
+ PMAXW(b, a)\
+ "movq " #a ", " #b " \n\t"\
+ "psrlq $16, " #a " \n\t"\
+ PMAXW(b, a)
#endif
@@ -71,18 +71,18 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
if (!s->h263_aic) {
#if 1
asm volatile (
- "mul %%ecx \n\t"
- : "=d" (level), "=a"(dummy)
- : "a" ((block[0]>>2) + q), "c" (inverse[q<<1])
+ "mul %%ecx \n\t"
+ : "=d" (level), "=a"(dummy)
+ : "a" ((block[0]>>2) + q), "c" (inverse[q<<1])
);
#else
asm volatile (
- "xorl %%edx, %%edx \n\t"
- "divw %%cx \n\t"
- "movzwl %%ax, %%eax \n\t"
- : "=a" (level)
- : "a" ((block[0]>>2) + q), "c" (q<<1)
- : "%edx"
+ "xorl %%edx, %%edx \n\t"
+ "divw %%cx \n\t"
+ "movzwl %%ax, %%eax \n\t"
+ : "=a" (level)
+ : "a" ((block[0]>>2) + q), "c" (q<<1)
+ : "%edx"
);
#endif
} else
@@ -103,94 +103,94 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
if((s->out_format == FMT_H263 || s->out_format == FMT_H261) && s->mpeg_quant==0){
asm volatile(
- "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1
+ "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1
SPREADW(%%mm3)
- "pxor %%mm7, %%mm7 \n\t" // 0
- "pxor %%mm4, %%mm4 \n\t" // 0
- "movq (%2), %%mm5 \n\t" // qmat[0]
- "pxor %%mm6, %%mm6 \n\t"
- "psubw (%3), %%mm6 \n\t" // -bias[0]
- "mov $-128, %%"REG_a" \n\t"
- ".balign 16 \n\t"
- "1: \n\t"
- "pxor %%mm1, %%mm1 \n\t" // 0
- "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i]
- "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00
- "pxor %%mm1, %%mm0 \n\t"
- "psubw %%mm1, %%mm0 \n\t" // ABS(block[i])
- "psubusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0]
- "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16
- "por %%mm0, %%mm4 \n\t"
- "pxor %%mm1, %%mm0 \n\t"
- "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
- "movq %%mm0, (%5, %%"REG_a") \n\t"
- "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00
- "movq (%4, %%"REG_a"), %%mm1 \n\t"
- "movq %%mm7, (%1, %%"REG_a") \n\t" // 0
- "pandn %%mm1, %%mm0 \n\t"
- PMAXW(%%mm0, %%mm3)
- "add $8, %%"REG_a" \n\t"
- " js 1b \n\t"
- PMAX(%%mm3, %%mm0)
- "movd %%mm3, %%"REG_a" \n\t"
- "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1
- : "+a" (last_non_zero_p1)
+ "pxor %%mm7, %%mm7 \n\t" // 0
+ "pxor %%mm4, %%mm4 \n\t" // 0
+ "movq (%2), %%mm5 \n\t" // qmat[0]
+ "pxor %%mm6, %%mm6 \n\t"
+ "psubw (%3), %%mm6 \n\t" // -bias[0]
+ "mov $-128, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "pxor %%mm1, %%mm1 \n\t" // 0
+ "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i]
+ "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00
+ "pxor %%mm1, %%mm0 \n\t"
+ "psubw %%mm1, %%mm0 \n\t" // ABS(block[i])
+ "psubusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0]
+ "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16
+ "por %%mm0, %%mm4 \n\t"
+ "pxor %%mm1, %%mm0 \n\t"
+ "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
+ "movq %%mm0, (%5, %%"REG_a") \n\t"
+ "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00
+ "movq (%4, %%"REG_a"), %%mm1 \n\t"
+ "movq %%mm7, (%1, %%"REG_a") \n\t" // 0
+ "pandn %%mm1, %%mm0 \n\t"
+ PMAXW(%%mm0, %%mm3)
+ "add $8, %%"REG_a" \n\t"
+ " js 1b \n\t"
+ PMAX(%%mm3, %%mm0)
+ "movd %%mm3, %%"REG_a" \n\t"
+ "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1
+ : "+a" (last_non_zero_p1)
: "r" (block+64), "r" (qmat), "r" (bias),
"r" (inv_zigzag_direct16+64), "r" (temp_block+64)
);
// note the asm is split cuz gcc doesnt like that many operands ...
asm volatile(
- "movd %1, %%mm1 \n\t" // max_qcoeff
- SPREADW(%%mm1)
- "psubusw %%mm1, %%mm4 \n\t"
- "packuswb %%mm4, %%mm4 \n\t"
- "movd %%mm4, %0 \n\t" // *overflow
+ "movd %1, %%mm1 \n\t" // max_qcoeff
+ SPREADW(%%mm1)
+ "psubusw %%mm1, %%mm4 \n\t"
+ "packuswb %%mm4, %%mm4 \n\t"
+ "movd %%mm4, %0 \n\t" // *overflow
: "=g" (*overflow)
: "g" (s->max_qcoeff)
);
}else{ // FMT_H263
asm volatile(
- "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1
+ "movd %%"REG_a", %%mm3 \n\t" // last_non_zero_p1
SPREADW(%%mm3)
- "pxor %%mm7, %%mm7 \n\t" // 0
- "pxor %%mm4, %%mm4 \n\t" // 0
- "mov $-128, %%"REG_a" \n\t"
- ".balign 16 \n\t"
- "1: \n\t"
- "pxor %%mm1, %%mm1 \n\t" // 0
- "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i]
- "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00
- "pxor %%mm1, %%mm0 \n\t"
- "psubw %%mm1, %%mm0 \n\t" // ABS(block[i])
- "movq (%3, %%"REG_a"), %%mm6 \n\t" // bias[0]
- "paddusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0]
- "movq (%2, %%"REG_a"), %%mm5 \n\t" // qmat[i]
- "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] + bias[0]*qmat[0])>>16
- "por %%mm0, %%mm4 \n\t"
- "pxor %%mm1, %%mm0 \n\t"
- "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
- "movq %%mm0, (%5, %%"REG_a") \n\t"
- "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00
- "movq (%4, %%"REG_a"), %%mm1 \n\t"
- "movq %%mm7, (%1, %%"REG_a") \n\t" // 0
- "pandn %%mm1, %%mm0 \n\t"
- PMAXW(%%mm0, %%mm3)
- "add $8, %%"REG_a" \n\t"
- " js 1b \n\t"
- PMAX(%%mm3, %%mm0)
- "movd %%mm3, %%"REG_a" \n\t"
- "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1
- : "+a" (last_non_zero_p1)
+ "pxor %%mm7, %%mm7 \n\t" // 0
+ "pxor %%mm4, %%mm4 \n\t" // 0
+ "mov $-128, %%"REG_a" \n\t"
+ ".balign 16 \n\t"
+ "1: \n\t"
+ "pxor %%mm1, %%mm1 \n\t" // 0
+ "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i]
+ "pcmpgtw %%mm0, %%mm1 \n\t" // block[i] <= 0 ? 0xFF : 0x00
+ "pxor %%mm1, %%mm0 \n\t"
+ "psubw %%mm1, %%mm0 \n\t" // ABS(block[i])
+ "movq (%3, %%"REG_a"), %%mm6 \n\t" // bias[0]
+ "paddusw %%mm6, %%mm0 \n\t" // ABS(block[i]) + bias[0]
+ "movq (%2, %%"REG_a"), %%mm5 \n\t" // qmat[i]
+ "pmulhw %%mm5, %%mm0 \n\t" // (ABS(block[i])*qmat[0] + bias[0]*qmat[0])>>16
+ "por %%mm0, %%mm4 \n\t"
+ "pxor %%mm1, %%mm0 \n\t"
+ "psubw %%mm1, %%mm0 \n\t" // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
+ "movq %%mm0, (%5, %%"REG_a") \n\t"
+ "pcmpeqw %%mm7, %%mm0 \n\t" // out==0 ? 0xFF : 0x00
+ "movq (%4, %%"REG_a"), %%mm1 \n\t"
+ "movq %%mm7, (%1, %%"REG_a") \n\t" // 0
+ "pandn %%mm1, %%mm0 \n\t"
+ PMAXW(%%mm0, %%mm3)
+ "add $8, %%"REG_a" \n\t"
+ " js 1b \n\t"
+ PMAX(%%mm3, %%mm0)
+ "movd %%mm3, %%"REG_a" \n\t"
+ "movzb %%al, %%"REG_a" \n\t" // last_non_zero_p1
+ : "+a" (last_non_zero_p1)
: "r" (block+64), "r" (qmat+64), "r" (bias+64),
"r" (inv_zigzag_direct16+64), "r" (temp_block+64)
);
// note the asm is split cuz gcc doesnt like that many operands ...
asm volatile(
- "movd %1, %%mm1 \n\t" // max_qcoeff
- SPREADW(%%mm1)
- "psubusw %%mm1, %%mm4 \n\t"
- "packuswb %%mm4, %%mm4 \n\t"
- "movd %%mm4, %0 \n\t" // *overflow
+ "movd %1, %%mm1 \n\t" // max_qcoeff
+ SPREADW(%%mm1)
+ "psubusw %%mm1, %%mm4 \n\t"
+ "packuswb %%mm4, %%mm4 \n\t"
+ "movd %%mm4, %0 \n\t" // *overflow
: "=g" (*overflow)
: "g" (s->max_qcoeff)
);
diff --git a/libavcodec/i386/simple_idct_mmx.c b/libavcodec/i386/simple_idct_mmx.c
index 7b5084c7eb..b814fabb35 100644
--- a/libavcodec/i386/simple_idct_mmx.c
+++ b/libavcodec/i386/simple_idct_mmx.c
@@ -40,8 +40,8 @@
#define C4 16383 //cos(i*M_PI/16)*sqrt(2)*(1<<14) - 0.5
#endif
#define C5 12873 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
-#define C6 8867 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
-#define C7 4520 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define C6 8867 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define C7 4520 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
#define ROW_SHIFT 11
#define COL_SHIFT 20 // 6
@@ -50,13 +50,13 @@ static const uint64_t attribute_used __attribute__((aligned(8))) wm1010= 0xFFFF0
static const uint64_t attribute_used __attribute__((aligned(8))) d40000= 0x0000000000040000ULL;
static const int16_t __attribute__((aligned(8))) coeffs[]= {
- 1<<(ROW_SHIFT-1), 0, 1<<(ROW_SHIFT-1), 0,
-// 1<<(COL_SHIFT-1), 0, 1<<(COL_SHIFT-1), 0,
-// 0, 1<<(COL_SHIFT-1-16), 0, 1<<(COL_SHIFT-1-16),
- 1<<(ROW_SHIFT-1), 1, 1<<(ROW_SHIFT-1), 0,
- // the 1 = ((1<<(COL_SHIFT-1))/C4)<<ROW_SHIFT :)
-// 0, 0, 0, 0,
-// 0, 0, 0, 0,
+ 1<<(ROW_SHIFT-1), 0, 1<<(ROW_SHIFT-1), 0,
+// 1<<(COL_SHIFT-1), 0, 1<<(COL_SHIFT-1), 0,
+// 0, 1<<(COL_SHIFT-1-16), 0, 1<<(COL_SHIFT-1-16),
+ 1<<(ROW_SHIFT-1), 1, 1<<(ROW_SHIFT-1), 0,
+ // the 1 = ((1<<(COL_SHIFT-1))/C4)<<ROW_SHIFT :)
+// 0, 0, 0, 0,
+// 0, 0, 0, 0,
C4, C4, C4, C4,
C4, -C4, C4, -C4,
@@ -79,8 +79,8 @@ static const int16_t __attribute__((aligned(8))) coeffs[]= {
#if 0
static void unused_var_killer(){
- int a= wm1010 + d40000;
- temp[0]=a;
+ int a= wm1010 + d40000;
+ temp[0]=a;
}
static void inline idctCol (int16_t * col, int16_t *input)
@@ -93,21 +93,21 @@ static void inline idctCol (int16_t * col, int16_t *input)
#undef C5
#undef C6
#undef C7
- int a0, a1, a2, a3, b0, b1, b2, b3;
- const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ int a0, a1, a2, a3, b0, b1, b2, b3;
+ const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
/*
- if( !(col[8*1] | col[8*2] |col[8*3] |col[8*4] |col[8*5] |col[8*6] | col[8*7])) {
- col[8*0] = col[8*1] = col[8*2] = col[8*3] = col[8*4] =
- col[8*5] = col[8*6] = col[8*7] = col[8*0]<<3;
- return;
- }*/
+ if( !(col[8*1] | col[8*2] |col[8*3] |col[8*4] |col[8*5] |col[8*6] | col[8*7])) {
+ col[8*0] = col[8*1] = col[8*2] = col[8*3] = col[8*4] =
+ col[8*5] = col[8*6] = col[8*7] = col[8*0]<<3;
+ return;
+ }*/
col[8*0] = input[8*0 + 0];
col[8*1] = input[8*2 + 0];
@@ -118,39 +118,39 @@ col[8*5] = input[8*6 + 0];
col[8*6] = input[8*4 + 1];
col[8*7] = input[8*6 + 1];
- a0 = C4*col[8*0] + C2*col[8*2] + C4*col[8*4] + C6*col[8*6] + (1<<(COL_SHIFT-1));
- a1 = C4*col[8*0] + C6*col[8*2] - C4*col[8*4] - C2*col[8*6] + (1<<(COL_SHIFT-1));
- a2 = C4*col[8*0] - C6*col[8*2] - C4*col[8*4] + C2*col[8*6] + (1<<(COL_SHIFT-1));
- a3 = C4*col[8*0] - C2*col[8*2] + C4*col[8*4] - C6*col[8*6] + (1<<(COL_SHIFT-1));
-
- b0 = C1*col[8*1] + C3*col[8*3] + C5*col[8*5] + C7*col[8*7];
- b1 = C3*col[8*1] - C7*col[8*3] - C1*col[8*5] - C5*col[8*7];
- b2 = C5*col[8*1] - C1*col[8*3] + C7*col[8*5] + C3*col[8*7];
- b3 = C7*col[8*1] - C5*col[8*3] + C3*col[8*5] - C1*col[8*7];
-
- col[8*0] = (a0 + b0) >> COL_SHIFT;
- col[8*1] = (a1 + b1) >> COL_SHIFT;
- col[8*2] = (a2 + b2) >> COL_SHIFT;
- col[8*3] = (a3 + b3) >> COL_SHIFT;
- col[8*4] = (a3 - b3) >> COL_SHIFT;
- col[8*5] = (a2 - b2) >> COL_SHIFT;
- col[8*6] = (a1 - b1) >> COL_SHIFT;
- col[8*7] = (a0 - b0) >> COL_SHIFT;
+ a0 = C4*col[8*0] + C2*col[8*2] + C4*col[8*4] + C6*col[8*6] + (1<<(COL_SHIFT-1));
+ a1 = C4*col[8*0] + C6*col[8*2] - C4*col[8*4] - C2*col[8*6] + (1<<(COL_SHIFT-1));
+ a2 = C4*col[8*0] - C6*col[8*2] - C4*col[8*4] + C2*col[8*6] + (1<<(COL_SHIFT-1));
+ a3 = C4*col[8*0] - C2*col[8*2] + C4*col[8*4] - C6*col[8*6] + (1<<(COL_SHIFT-1));
+
+ b0 = C1*col[8*1] + C3*col[8*3] + C5*col[8*5] + C7*col[8*7];
+ b1 = C3*col[8*1] - C7*col[8*3] - C1*col[8*5] - C5*col[8*7];
+ b2 = C5*col[8*1] - C1*col[8*3] + C7*col[8*5] + C3*col[8*7];
+ b3 = C7*col[8*1] - C5*col[8*3] + C3*col[8*5] - C1*col[8*7];
+
+ col[8*0] = (a0 + b0) >> COL_SHIFT;
+ col[8*1] = (a1 + b1) >> COL_SHIFT;
+ col[8*2] = (a2 + b2) >> COL_SHIFT;
+ col[8*3] = (a3 + b3) >> COL_SHIFT;
+ col[8*4] = (a3 - b3) >> COL_SHIFT;
+ col[8*5] = (a2 - b2) >> COL_SHIFT;
+ col[8*6] = (a1 - b1) >> COL_SHIFT;
+ col[8*7] = (a0 - b0) >> COL_SHIFT;
}
static void inline idctRow (int16_t * output, int16_t * input)
{
- int16_t row[8];
-
- int a0, a1, a2, a3, b0, b1, b2, b3;
- const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
- const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ int16_t row[8];
+
+ int a0, a1, a2, a3, b0, b1, b2, b3;
+ const int C0 = 23170; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C1 = 22725; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C2 = 21407; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C3 = 19266; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C4 = 16383; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C5 = 12873; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C6 = 8867; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+ const int C7 = 4520; //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
row[0] = input[0];
row[2] = input[1];
@@ -161,290 +161,290 @@ row[3] = input[9];
row[5] = input[12];
row[7] = input[13];
- if( !(row[1] | row[2] |row[3] |row[4] |row[5] |row[6] | row[7]) ) {
- row[0] = row[1] = row[2] = row[3] = row[4] =
- row[5] = row[6] = row[7] = row[0]<<3;
- output[0] = row[0];
- output[2] = row[1];
- output[4] = row[2];
- output[6] = row[3];
- output[8] = row[4];
- output[10] = row[5];
- output[12] = row[6];
- output[14] = row[7];
- return;
- }
-
- a0 = C4*row[0] + C2*row[2] + C4*row[4] + C6*row[6] + (1<<(ROW_SHIFT-1));
- a1 = C4*row[0] + C6*row[2] - C4*row[4] - C2*row[6] + (1<<(ROW_SHIFT-1));
- a2 = C4*row[0] - C6*row[2] - C4*row[4] + C2*row[6] + (1<<(ROW_SHIFT-1));
- a3 = C4*row[0] - C2*row[2] + C4*row[4] - C6*row[6] + (1<<(ROW_SHIFT-1));
-
- b0 = C1*row[1] + C3*row[3] + C5*row[5] + C7*row[7];
- b1 = C3*row[1] - C7*row[3] - C1*row[5] - C5*row[7];
- b2 = C5*row[1] - C1*row[3] + C7*row[5] + C3*row[7];
- b3 = C7*row[1] - C5*row[3] + C3*row[5] - C1*row[7];
-
- row[0] = (a0 + b0) >> ROW_SHIFT;
- row[1] = (a1 + b1) >> ROW_SHIFT;
- row[2] = (a2 + b2) >> ROW_SHIFT;
- row[3] = (a3 + b3) >> ROW_SHIFT;
- row[4] = (a3 - b3) >> ROW_SHIFT;
- row[5] = (a2 - b2) >> ROW_SHIFT;
- row[6] = (a1 - b1) >> ROW_SHIFT;
- row[7] = (a0 - b0) >> ROW_SHIFT;
-
- output[0] = row[0];
- output[2] = row[1];
- output[4] = row[2];
- output[6] = row[3];
- output[8] = row[4];
- output[10] = row[5];
- output[12] = row[6];
- output[14] = row[7];
+ if( !(row[1] | row[2] |row[3] |row[4] |row[5] |row[6] | row[7]) ) {
+ row[0] = row[1] = row[2] = row[3] = row[4] =
+ row[5] = row[6] = row[7] = row[0]<<3;
+ output[0] = row[0];
+ output[2] = row[1];
+ output[4] = row[2];
+ output[6] = row[3];
+ output[8] = row[4];
+ output[10] = row[5];
+ output[12] = row[6];
+ output[14] = row[7];
+ return;
+ }
+
+ a0 = C4*row[0] + C2*row[2] + C4*row[4] + C6*row[6] + (1<<(ROW_SHIFT-1));
+ a1 = C4*row[0] + C6*row[2] - C4*row[4] - C2*row[6] + (1<<(ROW_SHIFT-1));
+ a2 = C4*row[0] - C6*row[2] - C4*row[4] + C2*row[6] + (1<<(ROW_SHIFT-1));
+ a3 = C4*row[0] - C2*row[2] + C4*row[4] - C6*row[6] + (1<<(ROW_SHIFT-1));
+
+ b0 = C1*row[1] + C3*row[3] + C5*row[5] + C7*row[7];
+ b1 = C3*row[1] - C7*row[3] - C1*row[5] - C5*row[7];
+ b2 = C5*row[1] - C1*row[3] + C7*row[5] + C3*row[7];
+ b3 = C7*row[1] - C5*row[3] + C3*row[5] - C1*row[7];
+
+ row[0] = (a0 + b0) >> ROW_SHIFT;
+ row[1] = (a1 + b1) >> ROW_SHIFT;
+ row[2] = (a2 + b2) >> ROW_SHIFT;
+ row[3] = (a3 + b3) >> ROW_SHIFT;
+ row[4] = (a3 - b3) >> ROW_SHIFT;
+ row[5] = (a2 - b2) >> ROW_SHIFT;
+ row[6] = (a1 - b1) >> ROW_SHIFT;
+ row[7] = (a0 - b0) >> ROW_SHIFT;
+
+ output[0] = row[0];
+ output[2] = row[1];
+ output[4] = row[2];
+ output[6] = row[3];
+ output[8] = row[4];
+ output[10] = row[5];
+ output[12] = row[6];
+ output[14] = row[7];
}
#endif
static inline void idct(int16_t *block)
{
- int64_t __attribute__((aligned(8))) align_tmp[16];
- int16_t * const temp= (int16_t*)align_tmp;
+ int64_t __attribute__((aligned(8))) align_tmp[16];
+ int16_t * const temp= (int16_t*)align_tmp;
- asm volatile(
+ asm volatile(
#if 0 //Alternative, simpler variant
#define ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm0, %%mm0 \n\t" \
- "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
- "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
- "movq %%mm7, " #dst " \n\t"\
- "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "movq %%mm2, 24+" #dst " \n\t"\
- "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
- "movq %%mm2, 8+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
- "movq %%mm4, 16+" #dst " \n\t"\
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm0, %%mm0 \n\t" \
+ "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
+ "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
+ "movq %%mm7, " #dst " \n\t"\
+ "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "movq %%mm2, 24+" #dst " \n\t"\
+ "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
+ "movq %%mm2, 8+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
+ "movq %%mm4, 16+" #dst " \n\t"\
#define COL_IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- #rounder ", %%mm0 \n\t"\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
- "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
- "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm7, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm2, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm2, 32+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm4, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"\
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ #rounder ", %%mm0 \n\t"\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
+ "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
+ "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm7, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm2, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm2, 32+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm4, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"\
#define DC_COND_ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq "MANGLE(wm1010)", %%mm4 \n\t"\
- "pand %%mm0, %%mm4 \n\t"\
- "por %%mm1, %%mm4 \n\t"\
- "por %%mm2, %%mm4 \n\t"\
- "por %%mm3, %%mm4 \n\t"\
- "packssdw %%mm4,%%mm4 \n\t"\
- "movd %%mm4, %%eax \n\t"\
- "orl %%eax, %%eax \n\t"\
- "jz 1f \n\t"\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm0, %%mm0 \n\t" \
- "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
- "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
- "movq %%mm7, " #dst " \n\t"\
- "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "movq %%mm2, 24+" #dst " \n\t"\
- "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
- "movq %%mm2, 8+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
- "movq %%mm4, 16+" #dst " \n\t"\
- "jmp 2f \n\t"\
- "1: \n\t"\
- "pslld $16, %%mm0 \n\t"\
- "#paddd "MANGLE(d40000)", %%mm0 \n\t"\
- "psrad $13, %%mm0 \n\t"\
- "packssdw %%mm0, %%mm0 \n\t"\
- "movq %%mm0, " #dst " \n\t"\
- "movq %%mm0, 8+" #dst " \n\t"\
- "movq %%mm0, 16+" #dst " \n\t"\
- "movq %%mm0, 24+" #dst " \n\t"\
- "2: \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq "MANGLE(wm1010)", %%mm4 \n\t"\
+ "pand %%mm0, %%mm4 \n\t"\
+ "por %%mm1, %%mm4 \n\t"\
+ "por %%mm2, %%mm4 \n\t"\
+ "por %%mm3, %%mm4 \n\t"\
+ "packssdw %%mm4,%%mm4 \n\t"\
+ "movd %%mm4, %%eax \n\t"\
+ "orl %%eax, %%eax \n\t"\
+ "jz 1f \n\t"\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm0, %%mm0 \n\t" \
+ "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
+ "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
+ "movq %%mm7, " #dst " \n\t"\
+ "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "movq %%mm2, 24+" #dst " \n\t"\
+ "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
+ "movq %%mm2, 8+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
+ "movq %%mm4, 16+" #dst " \n\t"\
+ "jmp 2f \n\t"\
+ "1: \n\t"\
+ "pslld $16, %%mm0 \n\t"\
+ "#paddd "MANGLE(d40000)", %%mm0 \n\t"\
+ "psrad $13, %%mm0 \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t"\
+ "movq %%mm0, " #dst " \n\t"\
+ "movq %%mm0, 8+" #dst " \n\t"\
+ "movq %%mm0, 16+" #dst " \n\t"\
+ "movq %%mm0, 24+" #dst " \n\t"\
+ "2: \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -467,236 +467,236 @@ COL_IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
#else
#define DC_COND_IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq "MANGLE(wm1010)", %%mm4 \n\t"\
- "pand %%mm0, %%mm4 \n\t"\
- "por %%mm1, %%mm4 \n\t"\
- "por %%mm2, %%mm4 \n\t"\
- "por %%mm3, %%mm4 \n\t"\
- "packssdw %%mm4,%%mm4 \n\t"\
- "movd %%mm4, %%eax \n\t"\
- "orl %%eax, %%eax \n\t"\
- "jz 1f \n\t"\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm0, %%mm0 \n\t" \
- "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
- "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
- "movq %%mm7, " #dst " \n\t"\
- "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "movq %%mm2, 24+" #dst " \n\t"\
- "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
- "movq %%mm2, 8+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
- "movq %%mm4, 16+" #dst " \n\t"\
- "jmp 2f \n\t"\
- "1: \n\t"\
- "pslld $16, %%mm0 \n\t"\
- "paddd "MANGLE(d40000)", %%mm0 \n\t"\
- "psrad $13, %%mm0 \n\t"\
- "packssdw %%mm0, %%mm0 \n\t"\
- "movq %%mm0, " #dst " \n\t"\
- "movq %%mm0, 8+" #dst " \n\t"\
- "movq %%mm0, 16+" #dst " \n\t"\
- "movq %%mm0, 24+" #dst " \n\t"\
- "2: \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq "MANGLE(wm1010)", %%mm4 \n\t"\
+ "pand %%mm0, %%mm4 \n\t"\
+ "por %%mm1, %%mm4 \n\t"\
+ "por %%mm2, %%mm4 \n\t"\
+ "por %%mm3, %%mm4 \n\t"\
+ "packssdw %%mm4,%%mm4 \n\t"\
+ "movd %%mm4, %%eax \n\t"\
+ "orl %%eax, %%eax \n\t"\
+ "jz 1f \n\t"\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm0, %%mm0 \n\t" \
+ "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
+ "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
+ "movq %%mm7, " #dst " \n\t"\
+ "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "movq %%mm2, 24+" #dst " \n\t"\
+ "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
+ "movq %%mm2, 8+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
+ "movq %%mm4, 16+" #dst " \n\t"\
+ "jmp 2f \n\t"\
+ "1: \n\t"\
+ "pslld $16, %%mm0 \n\t"\
+ "paddd "MANGLE(d40000)", %%mm0 \n\t"\
+ "psrad $13, %%mm0 \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t"\
+ "movq %%mm0, " #dst " \n\t"\
+ "movq %%mm0, 8+" #dst " \n\t"\
+ "movq %%mm0, 16+" #dst " \n\t"\
+ "movq %%mm0, 24+" #dst " \n\t"\
+ "2: \n\t"
#define Z_COND_IDCT(src0, src4, src1, src5, dst, rounder, shift, bt) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq %%mm0, %%mm4 \n\t"\
- "por %%mm1, %%mm4 \n\t"\
- "por %%mm2, %%mm4 \n\t"\
- "por %%mm3, %%mm4 \n\t"\
- "packssdw %%mm4,%%mm4 \n\t"\
- "movd %%mm4, %%eax \n\t"\
- "orl %%eax, %%eax \n\t"\
- "jz " #bt " \n\t"\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm0, %%mm0 \n\t" \
- "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
- "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
- "movq %%mm7, " #dst " \n\t"\
- "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "movq %%mm2, 24+" #dst " \n\t"\
- "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
- "movq %%mm2, 8+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
- "movq %%mm4, 16+" #dst " \n\t"\
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq %%mm0, %%mm4 \n\t"\
+ "por %%mm1, %%mm4 \n\t"\
+ "por %%mm2, %%mm4 \n\t"\
+ "por %%mm3, %%mm4 \n\t"\
+ "packssdw %%mm4,%%mm4 \n\t"\
+ "movd %%mm4, %%eax \n\t"\
+ "orl %%eax, %%eax \n\t"\
+ "jz " #bt " \n\t"\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm0, %%mm0 \n\t" \
+ "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
+ "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
+ "movq %%mm7, " #dst " \n\t"\
+ "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "movq %%mm2, 24+" #dst " \n\t"\
+ "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
+ "movq %%mm2, 8+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
+ "movq %%mm4, 16+" #dst " \n\t"\
#define ROW_IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm0, %%mm0 \n\t" \
- "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
- "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
- "movq %%mm7, " #dst " \n\t"\
- "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "movq %%mm2, 24+" #dst " \n\t"\
- "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
- "movq %%mm2, 8+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
- "movq %%mm4, 16+" #dst " \n\t"\
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq 56(%2), %%mm5 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm0, %%mm0 \n\t" \
+ "psubd %%mm1, %%mm0 \n\t" /* A2 a2 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm5, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm5 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm5 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm5 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm1, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm5, %%mm1 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm5, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm7 \n\t" /* A1+B1 a1+b1 A0+B0 a0+b0 */\
+ "packssdw %%mm4, %%mm2 \n\t" /* A0-B0 a0-b0 A1-B1 a1-b1 */\
+ "movq %%mm7, " #dst " \n\t"\
+ "movq " #src1 ", %%mm1 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "movq %%mm2, 24+" #dst " \n\t"\
+ "pmaddwd %%mm1, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm1 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm0, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm0 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm1, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm6, %%mm2 \n\t" /* A3+B3 a3+b3 A2+B2 a2+b2 */\
+ "movq %%mm2, 8+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm0, %%mm4 \n\t" /* A2-B2 a2-b2 A3-B3 a3-b3 */\
+ "movq %%mm4, 16+" #dst " \n\t"\
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
DC_COND_IDCT( 0(%0), 8(%0), 16(%0), 24(%0), 0(%1),paddd 8(%2), 11)
@@ -706,80 +706,80 @@ Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 1f)
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- #rounder ", %%mm0 \n\t"\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
- "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
- "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm7, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm2, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm2, 32+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm4, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ #rounder ", %%mm0 \n\t"\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
+ "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
+ "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm7, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm2, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm2, 32+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm4, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -787,144 +787,144 @@ IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "4: \n\t"
+ "#.balign 16 \n\t"\
+ "4: \n\t"
Z_COND_IDCT( 64(%0), 72(%0), 80(%0), 88(%0), 64(%1),paddd (%2), 11, 6f)
Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 5f)
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- #rounder ", %%mm0 \n\t"\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
- "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
- "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm1, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm2, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm1 \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm2, 32+" #dst " \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm1, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ #rounder ", %%mm0 \n\t"\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
+ "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
+ "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm1, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm2, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm2, 32+" #dst " \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm1, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "6: \n\t"
+ "#.balign 16 \n\t"\
+ "6: \n\t"
Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 7f)
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- #rounder ", %%mm0 \n\t"\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm1, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm2, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm1 \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm2, 32+" #dst " \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm1, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ #rounder ", %%mm0 \n\t"\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ "movq 72(%2), %%mm7 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm1 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm1, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm7, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm7, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm1, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm2, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq 88(%2), %%mm1 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm1, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm1, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm1 \n\t" /* A3 a3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm1 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm2, 32+" #dst " \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm1, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -932,144 +932,144 @@ IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "2: \n\t"
+ "#.balign 16 \n\t"\
+ "2: \n\t"
Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 3f)
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- #rounder ", %%mm0 \n\t"\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
- "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
- "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
- "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm7, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm2, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
- "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
- "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
- "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
- "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
- "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
- "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm2, 32+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm4, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq " #src5 ", %%mm3 \n\t" /* R7 R5 r7 r5 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ #rounder ", %%mm0 \n\t"\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 56(%2), %%mm1 \n\t" /* C7 C5 C7 C5 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* C7R7+C5R5 C7r7+C5r5 */\
+ "pmaddwd 64(%2), %%mm2 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm1, %%mm7 \n\t" /* B0 b0 */\
+ "movq 72(%2), %%mm1 \n\t" /* -C5 -C1 -C5 -C1 */\
+ "pmaddwd %%mm3, %%mm1 \n\t" /* -C5R7-C1R5 -C5r7-C1r5 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "paddd %%mm2, %%mm1 \n\t" /* B1 b1 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm2 \n\t" /* A1 a1 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm1, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm7, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm2, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq " #src1 ", %%mm0 \n\t" /* R3 R1 r3 r1 */\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "movq 88(%2), %%mm7 \n\t" /* C3 C7 C3 C7 */\
+ "pmaddwd 96(%2), %%mm0 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C3R7+C7R5 C3r7+C7r5 */\
+ "movq %%mm5, %%mm2 \n\t" /* A2 a2 */\
+ "pmaddwd 104(%2), %%mm3 \n\t" /* -C1R7+C3R5 -C1r7+C3r5 */\
+ "paddd %%mm7, %%mm4 \n\t" /* B2 b2 */\
+ "paddd %%mm4, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm0, %%mm3 \n\t" /* B3 b3 */\
+ "paddd %%mm3, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm3, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm2, %%mm2 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm2, 32+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm4, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "3: \n\t"
+ "#.balign 16 \n\t"\
+ "3: \n\t"
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- #rounder ", %%mm0 \n\t"\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 64(%2), %%mm3 \n\t"\
- "pmaddwd %%mm2, %%mm3 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm1 \n\t" /* A1 a1 */\
- "paddd %%mm3, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm3, %%mm1 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm1 \n\t"\
- "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm7, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm1, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "movq %%mm5, %%mm1 \n\t" /* A2 a2 */\
- "paddd %%mm4, %%mm1 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm1 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm1, %%mm1 \n\t" /* A2+B2 a2+b2 */\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm1, 32+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "movd %%mm4, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ #rounder ", %%mm0 \n\t"\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 64(%2), %%mm3 \n\t"\
+ "pmaddwd %%mm2, %%mm3 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm1 \n\t" /* A1 a1 */\
+ "paddd %%mm3, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm3, %%mm1 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm7, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm1, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "movq %%mm5, %%mm1 \n\t" /* A2 a2 */\
+ "paddd %%mm4, %%mm1 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm1, %%mm1 \n\t" /* A2+B2 a2+b2 */\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm1, 32+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "movd %%mm4, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -1077,67 +1077,67 @@ IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "5: \n\t"
+ "#.balign 16 \n\t"\
+ "5: \n\t"
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- #rounder ", %%mm0 \n\t"\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
- "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
- "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\
- "movq 8+" #src4 ", %%mm3 \n\t" /* R6 R2 r6 r2 */\
- "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm3, %%mm7 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "pmaddwd 40(%2), %%mm3 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm1 \n\t"\
- "paddd %%mm1, %%mm7 \n\t" /* A0 a0 */\
- "paddd %%mm1, %%mm1 \n\t" /* 2C0 2c0 */\
- #rounder ", %%mm2 \n\t"\
- "psubd %%mm7, %%mm1 \n\t" /* A3 a3 */\
- "paddd %%mm2, %%mm3 \n\t" /* A1 a1 */\
- "paddd %%mm2, %%mm2 \n\t" /* 2C1 2c1 */\
- "psubd %%mm3, %%mm2 \n\t" /* A2 a2 */\
- "psrad $" #shift ", %%mm4 \n\t"\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm3 \n\t"\
- "packssdw %%mm7, %%mm4 \n\t" /* A0 a0 */\
- "movq %%mm4, " #dst " \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "packssdw %%mm3, %%mm0 \n\t" /* A1 a1 */\
- "movq %%mm0, 16+" #dst " \n\t"\
- "movq %%mm0, 96+" #dst " \n\t"\
- "movq %%mm4, 112+" #dst " \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "psrad $" #shift ", %%mm6 \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm2, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movq %%mm5, 32+" #dst " \n\t"\
- "psrad $" #shift ", %%mm1 \n\t"\
- "packssdw %%mm1, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movq %%mm6, 48+" #dst " \n\t"\
- "movq %%mm6, 64+" #dst " \n\t"\
- "movq %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ #rounder ", %%mm0 \n\t"\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
+ "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
+ "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\
+ "movq 8+" #src4 ", %%mm3 \n\t" /* R6 R2 r6 r2 */\
+ "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm3, %%mm7 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "pmaddwd 40(%2), %%mm3 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm1 \n\t"\
+ "paddd %%mm1, %%mm7 \n\t" /* A0 a0 */\
+ "paddd %%mm1, %%mm1 \n\t" /* 2C0 2c0 */\
+ #rounder ", %%mm2 \n\t"\
+ "psubd %%mm7, %%mm1 \n\t" /* A3 a3 */\
+ "paddd %%mm2, %%mm3 \n\t" /* A1 a1 */\
+ "paddd %%mm2, %%mm2 \n\t" /* 2C1 2c1 */\
+ "psubd %%mm3, %%mm2 \n\t" /* A2 a2 */\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm3 \n\t"\
+ "packssdw %%mm7, %%mm4 \n\t" /* A0 a0 */\
+ "movq %%mm4, " #dst " \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "packssdw %%mm3, %%mm0 \n\t" /* A1 a1 */\
+ "movq %%mm0, 16+" #dst " \n\t"\
+ "movq %%mm0, 96+" #dst " \n\t"\
+ "movq %%mm4, 112+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm2, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movq %%mm5, 32+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "packssdw %%mm1, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movq %%mm6, 48+" #dst " \n\t"\
+ "movq %%mm6, 64+" #dst " \n\t"\
+ "movq %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -1145,75 +1145,75 @@ IDCT( 0(%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
//IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
//IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"\
- "1: \n\t"
+ "#.balign 16 \n\t"\
+ "1: \n\t"
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
- "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
- "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
- "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
- "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
- #rounder ", %%mm4 \n\t"\
- "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
- #rounder ", %%mm0 \n\t"\
- "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
- "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
- "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
- "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
- "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
- "movq 64(%2), %%mm1 \n\t"\
- "pmaddwd %%mm2, %%mm1 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
- "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
- "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "psrad $" #shift ", %%mm7 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "movq %%mm0, %%mm3 \n\t" /* A1 a1 */\
- "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "psubd %%mm1, %%mm3 \n\t" /* A1-B1 a1-b1 */\
- "psrad $" #shift ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm3 \n\t"\
- "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
- "movd %%mm7, " #dst " \n\t"\
- "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
- "movd %%mm0, 16+" #dst " \n\t"\
- "packssdw %%mm3, %%mm3 \n\t" /* A1-B1 a1-b1 */\
- "movd %%mm3, 96+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
- "movd %%mm4, 112+" #dst " \n\t"\
- "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
- "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
- "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
- "movq %%mm5, %%mm3 \n\t" /* A2 a2 */\
- "paddd %%mm4, %%mm3 \n\t" /* A2+B2 a2+b2 */\
- "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
- "psrad $" #shift ", %%mm3 \n\t"\
- "psrad $" #shift ", %%mm5 \n\t"\
- "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
- "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\
- "psrad $" #shift ", %%mm6 \n\t"\
- "packssdw %%mm3, %%mm3 \n\t" /* A2+B2 a2+b2 */\
- "movd %%mm3, 32+" #dst " \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
- "movd %%mm6, 48+" #dst " \n\t"\
- "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
- "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
- "movd %%mm4, 64+" #dst " \n\t"\
- "movd %%mm5, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq " #src4 ", %%mm1 \n\t" /* R6 R2 r6 r2 */\
+ "movq " #src1 ", %%mm2 \n\t" /* R3 R1 r3 r1 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm5 \n\t" /* C6 C2 C6 C2 */\
+ "pmaddwd %%mm1, %%mm5 \n\t" /* C6R6+C2R2 C6r6+C2r2 */\
+ "movq 40(%2), %%mm6 \n\t" /* -C2 C6 -C2 C6 */\
+ "pmaddwd %%mm6, %%mm1 \n\t" /* -C2R6+C6R2 -C2r6+C6r2 */\
+ #rounder ", %%mm4 \n\t"\
+ "movq %%mm4, %%mm6 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 48(%2), %%mm7 \n\t" /* C3 C1 C3 C1 */\
+ #rounder ", %%mm0 \n\t"\
+ "pmaddwd %%mm2, %%mm7 \n\t" /* C3R3+C1R1 C3r3+C1r1 */\
+ "paddd %%mm5, %%mm4 \n\t" /* A0 a0 */\
+ "psubd %%mm5, %%mm6 \n\t" /* A3 a3 */\
+ "movq %%mm0, %%mm5 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1 a1 */\
+ "psubd %%mm1, %%mm5 \n\t" /* A2 a2 */\
+ "movq 64(%2), %%mm1 \n\t"\
+ "pmaddwd %%mm2, %%mm1 \n\t" /* -C7R3+C3R1 -C7r3+C3r1 */\
+ "paddd %%mm4, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "paddd %%mm4, %%mm4 \n\t" /* 2A0 2a0 */\
+ "psubd %%mm7, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "psrad $" #shift ", %%mm7 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "movq %%mm0, %%mm3 \n\t" /* A1 a1 */\
+ "paddd %%mm1, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "psubd %%mm1, %%mm3 \n\t" /* A1-B1 a1-b1 */\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm3 \n\t"\
+ "packssdw %%mm7, %%mm7 \n\t" /* A0+B0 a0+b0 */\
+ "movd %%mm7, " #dst " \n\t"\
+ "packssdw %%mm0, %%mm0 \n\t" /* A1+B1 a1+b1 */\
+ "movd %%mm0, 16+" #dst " \n\t"\
+ "packssdw %%mm3, %%mm3 \n\t" /* A1-B1 a1-b1 */\
+ "movd %%mm3, 96+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A0-B0 a0-b0 */\
+ "movd %%mm4, 112+" #dst " \n\t"\
+ "movq 80(%2), %%mm4 \n\t" /* -C1 C5 -C1 C5 */\
+ "pmaddwd %%mm2, %%mm4 \n\t" /* -C1R3+C5R1 -C1r3+C5r1 */\
+ "pmaddwd 96(%2), %%mm2 \n\t" /* -C5R3+C7R1 -C5r3+C7r1 */\
+ "movq %%mm5, %%mm3 \n\t" /* A2 a2 */\
+ "paddd %%mm4, %%mm3 \n\t" /* A2+B2 a2+b2 */\
+ "psubd %%mm4, %%mm5 \n\t" /* a2-B2 a2-b2 */\
+ "psrad $" #shift ", %%mm3 \n\t"\
+ "psrad $" #shift ", %%mm5 \n\t"\
+ "movq %%mm6, %%mm4 \n\t" /* A3 a3 */\
+ "paddd %%mm2, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "psubd %%mm2, %%mm4 \n\t" /* a3-B3 a3-b3 */\
+ "psrad $" #shift ", %%mm6 \n\t"\
+ "packssdw %%mm3, %%mm3 \n\t" /* A2+B2 a2+b2 */\
+ "movd %%mm3, 32+" #dst " \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "packssdw %%mm6, %%mm6 \n\t" /* A3+B3 a3+b3 */\
+ "movd %%mm6, 48+" #dst " \n\t"\
+ "packssdw %%mm4, %%mm4 \n\t" /* A3-B3 a3-b3 */\
+ "packssdw %%mm5, %%mm5 \n\t" /* A2-B2 a2-b2 */\
+ "movd %%mm4, 64+" #dst " \n\t"\
+ "movd %%mm5, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
@@ -1221,42 +1221,42 @@ IDCT( (%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
IDCT( 8(%1), 72(%1), 40(%1), 104(%1), 4(%0),/nop, 20)
IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0),/nop, 20)
IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0),/nop, 20)
- "jmp 9f \n\t"
+ "jmp 9f \n\t"
- "#.balign 16 \n\t"
- "7: \n\t"
+ "#.balign 16 \n\t"
+ "7: \n\t"
#undef IDCT
#define IDCT(src0, src4, src1, src5, dst, rounder, shift) \
- "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
- "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- #rounder ", %%mm4 \n\t"\
- #rounder ", %%mm0 \n\t"\
- "psrad $" #shift ", %%mm4 \n\t"\
- "psrad $" #shift ", %%mm0 \n\t"\
- "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\
- "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\
- "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
- "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\
- "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
- "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\
- #rounder ", %%mm1 \n\t"\
- #rounder ", %%mm2 \n\t"\
- "psrad $" #shift ", %%mm1 \n\t"\
- "packssdw %%mm1, %%mm4 \n\t" /* A0 a0 */\
- "movq %%mm4, " #dst " \n\t"\
- "psrad $" #shift ", %%mm2 \n\t"\
- "packssdw %%mm2, %%mm0 \n\t" /* A1 a1 */\
- "movq %%mm0, 16+" #dst " \n\t"\
- "movq %%mm0, 96+" #dst " \n\t"\
- "movq %%mm4, 112+" #dst " \n\t"\
- "movq %%mm0, 32+" #dst " \n\t"\
- "movq %%mm4, 48+" #dst " \n\t"\
- "movq %%mm4, 64+" #dst " \n\t"\
- "movq %%mm0, 80+" #dst " \n\t"
+ "movq " #src0 ", %%mm0 \n\t" /* R4 R0 r4 r0 */\
+ "movq 16(%2), %%mm4 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm0, %%mm4 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm5 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm5, %%mm0 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ #rounder ", %%mm4 \n\t"\
+ #rounder ", %%mm0 \n\t"\
+ "psrad $" #shift ", %%mm4 \n\t"\
+ "psrad $" #shift ", %%mm0 \n\t"\
+ "movq 8+" #src0 ", %%mm2 \n\t" /* R4 R0 r4 r0 */\
+ "movq 16(%2), %%mm1 \n\t" /* C4 C4 C4 C4 */\
+ "pmaddwd %%mm2, %%mm1 \n\t" /* C4R4+C4R0 C4r4+C4r0 */\
+ "movq 24(%2), %%mm7 \n\t" /* -C4 C4 -C4 C4 */\
+ "pmaddwd %%mm7, %%mm2 \n\t" /* -C4R4+C4R0 -C4r4+C4r0 */\
+ "movq 32(%2), %%mm7 \n\t" /* C6 C2 C6 C2 */\
+ #rounder ", %%mm1 \n\t"\
+ #rounder ", %%mm2 \n\t"\
+ "psrad $" #shift ", %%mm1 \n\t"\
+ "packssdw %%mm1, %%mm4 \n\t" /* A0 a0 */\
+ "movq %%mm4, " #dst " \n\t"\
+ "psrad $" #shift ", %%mm2 \n\t"\
+ "packssdw %%mm2, %%mm0 \n\t" /* A1 a1 */\
+ "movq %%mm0, 16+" #dst " \n\t"\
+ "movq %%mm0, 96+" #dst " \n\t"\
+ "movq %%mm4, 112+" #dst " \n\t"\
+ "movq %%mm0, 32+" #dst " \n\t"\
+ "movq %%mm4, 48+" #dst " \n\t"\
+ "movq %%mm4, 64+" #dst " \n\t"\
+ "movq %%mm0, 80+" #dst " \n\t"
//IDCT( src0, src4, src1, src5, dst, rounder, shift)
IDCT( 0(%1), 64(%1), 32(%1), 96(%1), 0(%0),/nop, 20)
@@ -1290,9 +1290,9 @@ Temp
*/
"9: \n\t"
- :: "r" (block), "r" (temp), "r" (coeffs)
- : "%eax"
- );
+ :: "r" (block), "r" (temp), "r" (coeffs)
+ : "%eax"
+ );
}
void ff_simple_idct_mmx(int16_t *block)
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-04 23:26:31 +0000