comparison x86/h264dsp_mmx.c @ 8510:cea216e44ee3 libavcodec

Add x264 SSE2 iDCT functions to H.264 decoder.

8509:5dffd656f606

            ff_h264_idct_add_mmx    (dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
        else if(block[i*16])
            ff_h264_idct_dc_add_mmx2(dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
    }
}

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

// out: o = |x-y|>a

8510:cea216e44ee3

            ff_h264_idct_add_mmx    (dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
        else if(block[i*16])
            ff_h264_idct_dc_add_mmx2(dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
    }
}

#if defined(CONFIG_GPL) && defined(HAVE_YASM)
static void ff_h264_idct_dc_add8_mmx2(uint8_t *dst, int16_t *block, int stride)
{
    __asm__ volatile(
        "movd             %0, %%mm0 \n\t"   //  0 0 X D
        "punpcklwd        %1, %%mm0 \n\t"   //  x X d D
        "paddsw           %2, %%mm0 \n\t"
        "psraw            $6, %%mm0 \n\t"
        "punpcklwd     %%mm0, %%mm0 \n\t"   //  d d D D
        "pxor          %%mm1, %%mm1 \n\t"   //  0 0 0 0
        "psubw         %%mm0, %%mm1 \n\t"   // -d-d-D-D
        "packuswb      %%mm1, %%mm0 \n\t"   // -d-d-D-D d d D D
        "pshufw $0xFA, %%mm0, %%mm1 \n\t"   // -d-d-d-d-D-D-D-D
        "punpcklwd     %%mm0, %%mm0 \n\t"   //  d d d d D D D D
        ::"m"(block[ 0]),
          "m"(block[16]),
          "m"(ff_pw_32)
    );
    __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))
    );
}

extern void ff_x264_add8x4_idct_sse2(uint8_t *dst, int16_t *block, int stride);

static void ff_h264_idct_add16_sse2(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
    int i;
    for(i=0; i<16; i+=2)
        if(nnzc[ scan8[i+0] ]|nnzc[ scan8[i+1] ])
            ff_x264_add8x4_idct_sse2 (dst + block_offset[i], block + i*16, stride);
}

static void ff_h264_idct_add16intra_sse2(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
    int i;
    for(i=0; i<16; i+=2){
        if(nnzc[ scan8[i+0] ]|nnzc[ scan8[i+1] ])
            ff_x264_add8x4_idct_sse2 (dst + block_offset[i], block + i*16, stride);
        else if(block[i*16]|block[i*16+16])
            ff_h264_idct_dc_add8_mmx2(dst + block_offset[i], block + i*16, stride);
    }
}

static void ff_h264_idct_add8_sse2(uint8_t **dest, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
    int i;
    for(i=16; i<16+8; i++){
        if(nnzc[ scan8[i+0] ]|nnzc[ scan8[i+1] ])
            ff_x264_add8x4_idct_sse2 (dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
        else if(block[i*16]|block[i*16+16])
            ff_h264_idct_dc_add8_mmx2(dest[(i&4)>>2] + block_offset[i], block + i*16, stride);
    }
}
#endif

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

// out: o = |x-y|>a