dsputil.c

This the source release kit for the following system configuration(s): - AMD Alchemy(TM) DBAu1200(

        src+=srcStride;
    }
}

STATIC_FUNC void put_h264_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride)
{
    const int w=8;
    uint8_t *cm = cropTbl + MAX_NEG_CROP;
    int i;
    for(i=0; i<w; i++)
    {
        const int srcB= src[-2*srcStride];
        const int srcA= src[-1*srcStride];
        const int src0= src[0 *srcStride];
        const int src1= src[1 *srcStride];
        const int src2= src[2 *srcStride];
        const int src3= src[3 *srcStride];
        const int src4= src[4 *srcStride];
        const int src5= src[5 *srcStride];
        const int src6= src[6 *srcStride];
        const int src7= src[7 *srcStride];
        const int src8= src[8 *srcStride];
        const int src9= src[9 *srcStride];
        const int src10=src[10*srcStride];
        dst[0*dstStride] = cm[(((src0+src1)*20 - (srcA+src2)*5 + (srcB+src3))+16)>>5];
        dst[1*dstStride] = cm[(((src1+src2)*20 - (src0+src3)*5 + (srcA+src4))+16)>>5];
        dst[2*dstStride] = cm[(((src2+src3)*20 - (src1+src4)*5 + (src0+src5))+16)>>5];
        dst[3*dstStride] = cm[(((src3+src4)*20 - (src2+src5)*5 + (src1+src6))+16)>>5];
        dst[4*dstStride] = cm[(((src4+src5)*20 - (src3+src6)*5 + (src2+src7))+16)>>5];
        dst[5*dstStride] = cm[(((src5+src6)*20 - (src4+src7)*5 + (src3+src8))+16)>>5];
        dst[6*dstStride] = cm[(((src6+src7)*20 - (src5+src8)*5 + (src4+src9))+16)>>5];
        dst[7*dstStride] = cm[(((src7+src8)*20 - (src6+src9)*5 + (src5+src10))+16)>>5];
        dst++;
        src++;
    }
}

#ifdef _MIPS_LINUX_  /* [ */
void put_h264_qpel8_hv_lowpass(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride)
{       
  const int hh = 8;
  const int ww = 8;
  uint8_t *cm = cropTbl + 1024;
  int ii;
  src -= 2 * srcStride;
  for (ii = 0; ii < hh + 5; ii++)
  {
    const int srcB = src[ - 2];
    const int srcA = src[ - 1];
    const int src0 = src[0];
    const int src1 = src[1];
    const int src2 = src[2];
    const int src3 = src[3];
    const int src4 = src[4];
    const int src5 = src[5];
    const int src6 = src[6];
    const int src7 = src[7];
    const int src8 = src[8];
    const int src9 = src[9];
    const int src10 = src[10];
    const int twenty = 20;
    const int five = 5;

    asm volatile ("mtlo %0": /* no outputs */ : "r"(srcB + src3));
    asm volatile ("madd %0,%1"::"r"(src0 + src1), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(srcA + src2), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[0]));
    //tmp[0] = (src0 + src1) * 20 - (srcA + src2) * 5 + (srcB + src3);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(srcA + src4));
    asm volatile ("madd %0,%1"::"r"(src1 + src2), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src0 + src3), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[1]));
    //tmp[1] = (src1 + src2) * 20 - (src0 + src3) * 5 + (srcA + src4);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src0 + src5));
    asm volatile ("madd %0,%1"::"r"(src2 + src3), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src1 + src4), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[2]));
    //tmp[2] = (src2 + src3) * 20 - (src1 + src4) * 5 + (src0 + src5);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src1 + src6));
    asm volatile ("madd %0,%1"::"r"(src3 + src4), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src2 + src5), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[3]));
    //tmp[3] = (src3 + src4) * 20 - (src2 + src5) * 5 + (src1 + src6);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src2 + src7));
    asm volatile ("madd %0,%1"::"r"(src4 + src5), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src3 + src6), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[4]));
    //tmp[4] = (src4 + src5) * 20 - (src3 + src6) * 5 + (src2 + src7);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src3 + src8));
    asm volatile ("madd %0,%1"::"r"(src5 + src6), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src4 + src7), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[5]));
    //tmp[5] = (src5 + src6) * 20 - (src4 + src7) * 5 + (src3 + src8);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src4 + src9));
    asm volatile ("madd %0,%1"::"r"(src6 + src7), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src5 + src8), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[6]));
    //tmp[6] = (src6 + src7) * 20 - (src5 + src8) * 5 + (src4 + src9);

    asm volatile ("mtlo %0": /* no outputs */ : "r"(src5 + src10));
    asm volatile ("madd %0,%1"::"r"(src7 + src8), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(src6 + src9), "r"(five));
    asm volatile ("mflo %0":"=r"(tmp[7]));
    //tmp[7] = (src7 + src8) * 20 - (src6 + src9) * 5 + (src5 + src10);

    tmp += tmpStride;
    src += srcStride;
  }
  tmp -= tmpStride * (hh + 5 - 2);
  for (ii = 0; ii < ww; ii++)
  {
    const int tmpB = tmp[ - 2 *tmpStride];
    const int tmpA = tmp[ - 1 *tmpStride];
    const int tmp0 = tmp[0 *tmpStride];
    const int tmp1 = tmp[1 *tmpStride];
    const int tmp2 = tmp[2 *tmpStride];
    const int tmp3 = tmp[3 *tmpStride];
    const int tmp4 = tmp[4 *tmpStride];
    const int tmp5 = tmp[5 *tmpStride];
    const int tmp6 = tmp[6 *tmpStride];
    const int tmp7 = tmp[7 *tmpStride];
    const int tmp8 = tmp[8 *tmpStride];
    const int tmp9 = tmp[9 *tmpStride];
    const int tmp10 = tmp[10 *tmpStride];
    const int twenty = 20;
    const int five = 5;
    int t1;

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmpB + tmp3) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp0 + tmp1), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmpA + tmp2), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[0 *dstStride] = cm[t1 >> 10];
    // dst[0 * dstStride] = cm[(((tmp0 + tmp1)*20 - (tmpA + tmp2) * 5 + (tmpB + tmp3))+512)>>10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmpA + tmp4) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp1 + tmp2), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp0 + tmp3), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[1 * dstStride] = cm[t1 >> 10];
    // dst[1 * dstStride] = cm[(((tmp1 + tmp2) * 20 - (tmp0 + tmp3) * 5 + (tmpA + tmp4)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp0 + tmp5) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp2 + tmp3), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp1 + tmp4), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[2 * dstStride] = cm[t1 >> 10];
    // dst[2 * dstStride] = cm[(((tmp2 + tmp3) * 20 - (tmp1 + tmp4) * 5 + (tmp0 + tmp5)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp1 + tmp6) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp3 + tmp4), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp2 + tmp5), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[3 * dstStride] = cm[t1 >> 10];
    // dst[3 * dstStride] = cm[(((tmp3 + tmp4) * 20 - (tmp2 + tmp5) * 5 + (tmp1 + tmp6)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp2 + tmp7) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp4 + tmp5), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp3 + tmp6), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[4 * dstStride] = cm[t1 >> 10];
    // dst[4 * dstStride] = cm[(((tmp4 + tmp5) * 20 - (tmp3 + tmp6) * 5 + (tmp2 + tmp7)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp3 + tmp8) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp5 + tmp6), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp4 + tmp7), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[5 *dstStride] = cm[t1 >> 10];
    // dst[5 * dstStride] = cm[(((tmp5 + tmp6) * 20 - (tmp4 + tmp7) * 5 + (tmp3 + tmp8)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp4 + tmp9) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp6 + tmp7), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp5 + tmp8), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[6 *dstStride] = cm[t1 >> 10];
    // dst[6 * dstStride] = cm[(((tmp6 + tmp7) * 20 - (tmp5 + tmp8) * 5 + (tmp4 + tmp9)) + 512) >> 10];

    asm volatile ("mtlo %0": /* no outputs */ : "r"((tmp5 + tmp10) + 512));
    asm volatile ("madd %0,%1"::"r"(tmp7 + tmp8), "r"(twenty));
    asm volatile ("msub %0,%1"::"r"(tmp6 + tmp9), "r"(five));
    asm volatile ("mflo %0":"=r"(t1));
    dst[7 *dstStride] = cm[t1 >> 10];
    // dst[7 * dstStride] = cm[(((tmp7 + tmp8) * 20 - (tmp6 + tmp9) * 5 + (tmp5 + tmp10)) + 512) >> 10];

    dst++;
    tmp++;
  }
}

#else /* ] [ */
void put_h264_qpel8_hv_lowpass(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride)
{
  const int hh = 8;
  const int ww = 8;
  uint8_t *cm = cropTbl + MAX_NEG_CROP;
  int ii;
  src -= 2*srcStride;
  for (ii = 0; ii <hh + 5; ii++)
  {
    const srcB = src[-2];
    const srcA = src[-1];
    const src0 = src[0 ];
    const src1 = src[1 ];
    const src2 = src[2 ];
    const src3 = src[3];
    const src4 = src[4 ];
    const src5 = src[5 ];
    const src6 = src[6 ];
    const src7 = src[7 ];
    const src8 = src[8 ];
    const src9 = src[9 ];
    const src10 = src[10];

    tmp[0]= (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3);
    tmp[1]= (src1+src2)*20 - (src0+src3)*5 + (srcA+src4);
    tmp[2]= (src2+src3)*20 - (src1+src4)*5 + (src0+src5);
    tmp[3]= (src3+src4)*20 - (src2+src5)*5 + (src1+src6);
    tmp[4]= (src4+src5)*20 - (src3+src6)*5 + (src2+src7);
    tmp[5]= (src5+src6)*20 - (src4+src7)*5 + (src3+src8);
    tmp[6]= (src6+src7)*20 - (src5+src8)*5 + (src4+src9);
    tmp[7]= (src7+src8)*20 - (src6+src9)*5 + (src5+src10);
    tmp += tmpStride;
    src += srcStride;
  }
  tmp -= tmpStride * (hh + 5 - 2);
  for (ii = 0; ii < ww; ii++)
  {
    const int tmpB= tmp[-2*tmpStride];
    const int tmpA= tmp[-1*tmpStride];
    const int tmp0= tmp[0 *tmpStride];
    const int tmp1= tmp[1 *tmpStride];
    const int tmp2= tmp[2 *tmpStride];
    const int tmp3= tmp[3 *tmpStride];
    const int tmp4= tmp[4 *tmpStride];
    const int tmp5= tmp[5 *tmpStride];
    const int tmp6= tmp[6 *tmpStride];
    const int tmp7= tmp[7 *tmpStride];
    const int tmp8= tmp[8 *tmpStride];
    const int tmp9= tmp[9 *tmpStride];
    const int tmp10=tmp[10*tmpStride];
    dst[0*dstStride] = cm[(((tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3))+512)>>10];
    dst[1*dstStride] = cm[(((tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4))+512)>>10];
    dst[2*dstStride] = cm[(((tmp2+tmp3)*20 - (tmp1+tmp4)*5 + (tmp0+tmp5))+512)>>10];
    dst[3*dstStride] = cm[(((tmp3+tmp4)*20 - (tmp2+tmp5)*5 + (tmp1+tmp6))+512)>>10];
    dst[4*dstStride] = cm[(((tmp4+tmp5)*20 - (tmp3+tmp6)*5 + (tmp2+tmp7))+512)>>10];
    dst[5*dstStride] = cm[(((tmp5+tmp6)*20 - (tmp4+tmp7)*5 + (tmp3+tmp8))+512)>>10];
    dst[6*dstStride] = cm[(((tmp6+tmp7)*20 - (tmp5+tmp8)*5 + (tmp4+tmp9))+512)>>10];
    dst[7*dstStride] = cm[(((tmp7+tmp8)*20 - (tmp6+tmp9)*5 + (tmp5+tmp10))+512)>>10];
    dst++;
    tmp++;
  }
}
#endif /* ] */
#endif /* ] */

STATIC_FUNC void put_h264_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride)
{
    put_h264_qpel8_v_lowpass(dst  , src  , dstStride, srcStride);
    put_h264_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride);
    src += 8*srcStride;
    dst += 8*dstStride;
    put_h264_qpel8_v_lowpass(dst  , src  , dstStride, srcStride);
    put_h264_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride);
}

STATIC_FUNC void put_h264_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride)
{
    put_h264_qpel8_h_lowpass(dst  , src  , dstStride, srcStride);
    put_h264_qpel8_h_lowpass(dst+8, src+8, dstStride, srcStride);
    src += 8*srcStride;
    dst += 8*dstStride;
    put_h264_qpel8_h_lowpass(dst  , src  , dstStride, srcStride);
    put_h264_qpel8_h_lowpass(dst+8, src+8, dstStride, srcStride);
}

STATIC_FUNC void put_h264_qpel16_hv_lowpass(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride)
{
    put_h264_qpel8_hv_lowpass(dst  , tmp  , src  , dstStride, tmpStride, srcStride);
    put_h264_qpel8_hv_lowpass(dst+8, tmp+8, src+8, dstStride, tmpStride, srcStride);
    src += 8*srcStride;
    dst += 8*dstStride;
    put_h264_qpel8_hv_lowpass(dst  , tmp  , src  , dstStride, tmpStride, srcStride);
    put_h264_qpel8_hv_lowpass(dst+8, tmp+8, src+8, dstStride, tmpStride, srcStride);
}

STATIC_FUNC void avg_h264_qpel2_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride)
{
    const int h=2;
    uint8_t *cm = cropTbl + MAX_NEG_CROP;
    int i;
    for(i=0; i<h; i++)
    {
        const int srcB= src[-2];
        const int srcA= src[-1];
        const int src0= src[0 ];
        const int src1= src[1];
        const int src2= src[2 ];
        const int src3= src[3 ];
        const int src4= src[4 ];
        dst[0] = (dst[0] + cm[(((src0+src1)*20 - (srcA+src2)*5 + (srcB+src3))+16)>>5]+1)>>1;
        dst[1] = (dst[1] + cm[(((src1+src2)*20 - (src0+src3)*5 + (srcA+src4))+16)>>5]+1)>>1;
        dst+=dstStride;
        src+=srcStride;
    }
}

STATIC_FUNC void avg_h264_qpel2_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride)
{
    const int w=2;
    uint8_t *cm = cropTbl + MAX_NEG_CROP;
    int i;
    for(i=0; i<w; i++)
    {
        const int srcB= src[-2*srcStride];
        const int srcA= src[-1*srcStride];
        const int src0= src[0 *srcStride];