dsputil.c

arm平台下的H264编码和解码源代码

                  int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height)
{
    int y, vx, vy;
    const int s= 1<<shift;
    
    width--;
    height--;
    for(y=0; y<h; y++){
        int x;
        vx= ox;
        vy= oy;
        for(x=0; x<8; x++){ //XXX FIXME optimize
            int src_x, src_y, frac_x, frac_y, index;
            src_x= vx>>16;
            src_y= vy>>16;
            frac_x= src_x&(s-1);
            frac_y= src_y&(s-1);
            src_x>>=shift;
            src_y>>=shift;
  
            if((unsigned)src_x < width){
                if((unsigned)src_y < height){
                    index= src_x + src_y*stride;
                    dst[y*stride + x]= (  (  src[index         ]*(s-frac_x)
                                           + src[index       +1]*   frac_x )*(s-frac_y)
                                        + (  src[index+stride  ]*(s-frac_x)
                                           + src[index+stride+1]*   frac_x )*   frac_y
                                        + r)>>(shift*2);
                }else{
                    index= src_x + clip(src_y, 0, height)*stride;                    
                    dst[y*stride + x]= ( (  src[index         ]*(s-frac_x) 
                                          + src[index       +1]*   frac_x )*s
                                        + r)>>(shift*2);
                }
            }else{
                if((unsigned)src_y < height){
                    index= clip(src_x, 0, width) + src_y*stride;                    
                    dst[y*stride + x]= (  (  src[index         ]*(s-frac_y) 
                                           + src[index+stride  ]*   frac_y )*s
                                        + r)>>(shift*2);
                }else{
                    index= clip(src_x, 0, width) + clip(src_y, 0, height)*stride;                    
                    dst[y*stride + x]=    src[index         ];
                }
            }
            
            vx+= dxx;
            vy+= dyx;
        }
        ox += dxy;
        oy += dyy;
    }
}
static inline void put_tpel_pixels_mc00_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    switch(width){
    case 2: put_pixels2_c (dst, src, stride, height); break;
    case 4: put_pixels4_c (dst, src, stride, height); break;
    case 8: put_pixels8_c (dst, src, stride, height); break;
    case 16:put_pixels16_c(dst, src, stride, height); break;
    }
}
static inline void put_tpel_pixels_mc10_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (683*(2*src[j] + src[j+1] + 1)) >> 11;
      }
      src += stride;
      dst += stride;
    }
}
static inline void put_tpel_pixels_mc20_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (683*(src[j] + 2*src[j+1] + 1)) >> 11;
      }
      src += stride;
      dst += stride;
    }
}
    
static inline void put_tpel_pixels_mc01_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (683*(2*src[j] + src[j+stride] + 1)) >> 11;
      }
      src += stride;
      dst += stride;
    }
}
    
static inline void put_tpel_pixels_mc11_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15;
      }
      src += stride;
      dst += stride;
    }
}
static inline void put_tpel_pixels_mc12_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15;
      }
      src += stride;
      dst += stride;
    }
}
static inline void put_tpel_pixels_mc02_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (683*(src[j] + 2*src[j+stride] + 1)) >> 11;
      }
      src += stride;
      dst += stride;
    }
}
static inline void put_tpel_pixels_mc21_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15;
      }
      src += stride;
      dst += stride;
    }
}
static inline void put_tpel_pixels_mc22_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc00_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    switch(width){
    case 2: avg_pixels2_c (dst, src, stride, height); break;
    case 4: avg_pixels4_c (dst, src, stride, height); break;
    case 8: avg_pixels8_c (dst, src, stride, height); break;
    case 16:avg_pixels16_c(dst, src, stride, height); break;
    }
}
static inline void avg_tpel_pixels_mc10_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((683*(2*src[j] + src[j+1] + 1)) >> 11) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc20_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+1] + 1)) >> 11) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
    
static inline void avg_tpel_pixels_mc01_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((683*(2*src[j] + src[j+stride] + 1)) >> 11) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
    
static inline void avg_tpel_pixels_mc11_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((2731*(4*src[j] + 3*src[j+1] + 3*src[j+stride] + 2*src[j+stride+1] + 6)) >> 15) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc12_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((2731*(3*src[j] + 2*src[j+1] + 4*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc02_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((683*(src[j] + 2*src[j+stride] + 1)) >> 11) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc21_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((2731*(3*src[j] + 4*src[j+1] + 2*src[j+stride] + 3*src[j+stride+1] + 6)) >> 15) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
static inline void avg_tpel_pixels_mc22_c(uint8_t *dst, const uint8_t *src, int stride, int width, int height){
    int i,j;
    for (i=0; i < height; i++) {
      for (j=0; j < width; j++) {
	dst[j] = (dst[j] + ((2731*(2*src[j] + 3*src[j+1] + 3*src[j+stride] + 4*src[j+stride+1] + 6)) >> 15) + 1) >> 1;
      }
      src += stride;
      dst += stride;
    }
}
#if 0
#define TPEL_WIDTH(width)\
static void put_tpel_pixels ## width ## _mc00_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc00_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc10_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc10_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc20_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc20_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc01_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc01_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc11_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc11_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc21_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc21_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc02_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc02_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc12_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc12_c(dst, src, stride, width, height);}\
static void put_tpel_pixels ## width ## _mc22_c(uint8_t *dst, const uint8_t *src, int stride, int height){\
    void put_tpel_pixels_mc22_c(dst, src, stride, width, height);}
#endif
#define H264_CHROMA_MC(OPNAME, OP)\
static void OPNAME ## h264_chroma_mc2_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
    const int A=(8-x)*(8-y);\
    const int B=(  x)*(8-y);\
    const int C=(8-x)*(  y);\
    const int D=(  x)*(  y);\
    int i;\
    \
    assert(x<8 && y<8 && x>=0 && y>=0);\
\
    for(i=0; i<h; i++)\
    {\
        OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
        OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
        dst+= stride;\
        src+= stride;\
    }\
}\
\
static void OPNAME ## h264_chroma_mc4_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
    const int A=(8-x)*(8-y);\
    const int B=(  x)*(8-y);\
    const int C=(8-x)*(  y);\
    const int D=(  x)*(  y);\
    int i;\
    \
    assert(x<8 && y<8 && x>=0 && y>=0);\
\
    for(i=0; i<h; i++)\
    {\
        OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
        OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
        OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3]));\
        OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4]));\
        dst+= stride;\
        src+= stride;\
    }\
}\
\
static void OPNAME ## h264_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
    const int A=(8-x)*(8-y);\
    const int B=(  x)*(8-y);\
    const int C=(8-x)*(  y);\
    const int D=(  x)*(  y);\
    int i;\
    \
    assert(x<8 && y<8 && x>=0 && y>=0);\
\
    for(i=0; i<h; i++)\
    {\
        OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
        OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
        OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3]));\
        OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4]));\
        OP(dst[4], (A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5]));\
        OP(dst[5], (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6]));\
        OP(dst[6], (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7]));\
        OP(dst[7], (A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8]));\
        dst+= stride;\
        src+= stride;\
    }\
}
#define op_avg(a, b) a = (((a)+(((b) + 32)>>6)+1)>>1)
#define op_put(a, b) a = (((b) + 32)>>6)
H264_CHROMA_MC(put_       , op_put)
H264_CHROMA_MC(avg_       , op_avg)
#undef op_avg
#undef op_put
static inline void copy_block4(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
{
    int i;
    for(i=0; i<h; i++)
    {
        //ST32(dst   , LD32(src   ));
		memcpy(dst,src,sizeof(src));