comparison i386/h264dsp_mmx.c @ 3105:2d35fb3cb940 libavcodec

h264: special case dc-only idct. ~1% faster overall

3104:78d6bfc238f3

        "add %1, %0             \n\t"
    STORE_DIFF_4P( %%mm4, %%mm1, %%mm7)
        : "+r"(dst)
        : "r" ((long)stride)
    );
}


/***********************************/
/* deblocking */

3105:2d35fb3cb940

        "add %1, %0             \n\t"
    STORE_DIFF_4P( %%mm4, %%mm1, %%mm7)
        : "+r"(dst)
        : "r" ((long)stride)
    );
}

void ff_h264_idct_dc_add_mmx2(uint8_t *dst, int16_t *block, int stride)
{
    int dc = (block[0] + 32) >> 6;
    asm volatile(
        "movd          %0, %%mm0 \n\t"
        "pxor       %%mm7, %%mm7 \n\t"
        "pshufw $0, %%mm0, %%mm0 \n\t"
        "pxor       %%mm1, %%mm1 \n\t"
        "psubw      %%mm0, %%mm1 \n\t"
        "pmaxsw     %%mm7, %%mm0 \n\t"
        "pmaxsw     %%mm7, %%mm1 \n\t"
        "packuswb   %%mm0, %%mm0 \n\t"
        "packuswb   %%mm1, %%mm1 \n\t"
        ::"r"(dc)
    );
    asm volatile(
        "movd          %0, %%mm2 \n\t"
        "movd          %1, %%mm3 \n\t"
        "movd          %2, %%mm4 \n\t"
        "movd          %3, %%mm5 \n\t"
        "paddusb    %%mm0, %%mm2 \n\t"
        "paddusb    %%mm0, %%mm3 \n\t"
        "paddusb    %%mm0, %%mm4 \n\t"
        "paddusb    %%mm0, %%mm5 \n\t"
        "psubusb    %%mm1, %%mm2 \n\t"
        "psubusb    %%mm1, %%mm3 \n\t"
        "psubusb    %%mm1, %%mm4 \n\t"
        "psubusb    %%mm1, %%mm5 \n\t"
        "movd       %%mm2, %0    \n\t"
        "movd       %%mm3, %1    \n\t"
        "movd       %%mm4, %2    \n\t"
        "movd       %%mm5, %3    \n\t"
        :"+m"(*(uint32_t*)(dst+0*stride)),
         "+m"(*(uint32_t*)(dst+1*stride)),
         "+m"(*(uint32_t*)(dst+2*stride)),
         "+m"(*(uint32_t*)(dst+3*stride))
    );
}

void ff_h264_idct8_dc_add_mmx2(uint8_t *dst, int16_t *block, int stride)
{
    int dc = (block[0] + 32) >> 6;
    int y;
    asm volatile(
        "movd          %0, %%mm0 \n\t"
        "pxor       %%mm7, %%mm7 \n\t"
        "pshufw $0, %%mm0, %%mm0 \n\t"
        "pxor       %%mm1, %%mm1 \n\t"
        "psubw      %%mm0, %%mm1 \n\t"
        "pmaxsw     %%mm7, %%mm0 \n\t"
        "pmaxsw     %%mm7, %%mm1 \n\t"
        "packuswb   %%mm0, %%mm0 \n\t"
        "packuswb   %%mm1, %%mm1 \n\t"
        ::"r"(dc)
    );
    for(y=2; y--; dst += 4*stride){
    asm volatile(
        "movq          %0, %%mm2 \n\t"
        "movq          %1, %%mm3 \n\t"
        "movq          %2, %%mm4 \n\t"
        "movq          %3, %%mm5 \n\t"
        "paddusb    %%mm0, %%mm2 \n\t"
        "paddusb    %%mm0, %%mm3 \n\t"
        "paddusb    %%mm0, %%mm4 \n\t"
        "paddusb    %%mm0, %%mm5 \n\t"
        "psubusb    %%mm1, %%mm2 \n\t"
        "psubusb    %%mm1, %%mm3 \n\t"
        "psubusb    %%mm1, %%mm4 \n\t"
        "psubusb    %%mm1, %%mm5 \n\t"
        "movq       %%mm2, %0    \n\t"
        "movq       %%mm3, %1    \n\t"
        "movq       %%mm4, %2    \n\t"
        "movq       %%mm5, %3    \n\t"
        :"+m"(*(uint64_t*)(dst+0*stride)),
         "+m"(*(uint64_t*)(dst+1*stride)),
         "+m"(*(uint64_t*)(dst+2*stride)),
         "+m"(*(uint64_t*)(dst+3*stride))
    );
    }
}


/***********************************/
/* deblocking */