motion_est_mmx.c

mplayer播放器的源码

/*
 * MMX optimized motion estimation
 * Copyright (c) 2001 Fabrice Bellard.
 * Copyright (c) 2002-2004 Michael Niedermayer
 *
 * mostly by Michael Niedermayer <[e-mail]>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */
#include "dsputil.h"
#include "x86_cpu.h"

static const __attribute__ ((aligned(8))) uint64_t round_tab[3]={
0x0000000000000000ULL,
0x0001000100010001ULL,
0x0002000200020002ULL,
};

static attribute_used __attribute__ ((aligned(8))) uint64_t bone= 0x0101010101010101LL;

static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
    long len= -(stride*h);
    asm volatile(
        ASMALIGN(4)
        "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)
    );
}

static inline void sad8_1_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
    asm volatile(
        ASMALIGN(4)
        "1:                             \n\t"
        "movq (%1), %%mm0               \n\t"
        "movq (%1, %3), %%mm1           \n\t"
        "psadbw (%2), %%mm0             \n\t"
        "psadbw (%2, %3), %%mm1         \n\t"
        "paddw %%mm0, %%mm6             \n\t"
        "paddw %%mm1, %%mm6             \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
}

static int sad16_sse2(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)
{
    int ret;
    asm volatile(
        "pxor %%xmm6, %%xmm6            \n\t"
        ASMALIGN(4)
        "1:                             \n\t"
        "movdqu (%1), %%xmm0            \n\t"
        "movdqu (%1, %3), %%xmm1        \n\t"
        "psadbw (%2), %%xmm0            \n\t"
        "psadbw (%2, %3), %%xmm1        \n\t"
        "paddw %%xmm0, %%xmm6           \n\t"
        "paddw %%xmm1, %%xmm6           \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
    asm volatile(
        "movhlps %%xmm6, %%xmm0         \n\t"
        "paddw   %%xmm0, %%xmm6         \n\t"
        "movd    %%xmm6, %0             \n\t"
        : "=r"(ret)
    );
    return ret;
}

static inline void sad8_x2a_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
    asm volatile(
        ASMALIGN(4)
        "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"
        "psadbw (%2), %%mm0             \n\t"
        "psadbw (%2, %3), %%mm1         \n\t"
        "paddw %%mm0, %%mm6             \n\t"
        "paddw %%mm1, %%mm6             \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
}

static inline void sad8_y2a_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
    asm volatile(
        "movq (%1), %%mm0               \n\t"
        "add %3, %1                     \n\t"
        ASMALIGN(4)
        "1:                             \n\t"
        "movq (%1), %%mm1               \n\t"
        "movq (%1, %3), %%mm2           \n\t"
        "pavgb %%mm1, %%mm0             \n\t"
        "pavgb %%mm2, %%mm1             \n\t"
        "psadbw (%2), %%mm0             \n\t"
        "psadbw (%2, %3), %%mm1         \n\t"
        "paddw %%mm0, %%mm6             \n\t"
        "paddw %%mm1, %%mm6             \n\t"
        "movq %%mm2, %%mm0              \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
}

static inline void sad8_4_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
{
    asm volatile(
        "movq "MANGLE(bone)", %%mm5     \n\t"
        "movq (%1), %%mm0               \n\t"
        "pavgb 1(%1), %%mm0             \n\t"
        "add %3, %1                     \n\t"
        ASMALIGN(4)
        "1:                             \n\t"
        "movq (%1), %%mm1               \n\t"
        "movq (%1,%3), %%mm2            \n\t"
        "pavgb 1(%1), %%mm1             \n\t"
        "pavgb 1(%1,%3), %%mm2          \n\t"
        "psubusb %%mm5, %%mm1           \n\t"
        "pavgb %%mm1, %%mm0             \n\t"
        "pavgb %%mm2, %%mm1             \n\t"
        "psadbw (%2), %%mm0             \n\t"
        "psadbw (%2,%3), %%mm1          \n\t"
        "paddw %%mm0, %%mm6             \n\t"
        "paddw %%mm1, %%mm6             \n\t"
        "movq %%mm2, %%mm0              \n\t"
        "lea (%1,%3,2), %1              \n\t"
        "lea (%2,%3,2), %2              \n\t"
        "sub $2, %0                     \n\t"
        " jg 1b                         \n\t"
        : "+r" (h), "+r" (blk1), "+r" (blk2)
        : "r" ((long)stride)
    );
}

static inline void sad8_2_mmx(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2, int stride, int h)
{
    long len= -(stride*h);
    asm volatile(
        ASMALIGN(4)
        "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)