swscale.c

ffmpeg的完整源代码和作者自己写的文档。不但有在Linux的工程哦

#if defined(ARCH_X86) && defined (CONFIG_GPL)
void in_asm_used_var_warning_killer()
{
    volatile int i= bF8+bFC+w10+
    bm00001111+bm00000111+bm11111000+b16Mask+g16Mask+r16Mask+b15Mask+g15Mask+r15Mask+
    M24A+M24B+M24C+w02 + b5Dither+g5Dither+r5Dither+g6Dither+dither4[0]+dither8[0]+bm01010101;
    if (i) i=0;
}
#endif

static inline void yuv2yuvXinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
                               int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
                               uint8_t *dest, uint8_t *uDest, uint8_t *vDest, int dstW, int chrDstW)
{
    //FIXME Optimize (just quickly writen not opti..)
    int i;
    for (i=0; i<dstW; i++)
    {
        int val=1<<18;
        int j;
        for (j=0; j<lumFilterSize; j++)
            val += lumSrc[j][i] * lumFilter[j];

        dest[i]= av_clip_uint8(val>>19);
    }

    if (uDest != NULL)
        for (i=0; i<chrDstW; i++)
        {
            int u=1<<18;
            int v=1<<18;
            int j;
            for (j=0; j<chrFilterSize; j++)
            {
                u += chrSrc[j][i] * chrFilter[j];
                v += chrSrc[j][i + 2048] * chrFilter[j];
            }

            uDest[i]= av_clip_uint8(u>>19);
            vDest[i]= av_clip_uint8(v>>19);
        }
}

static inline void yuv2nv12XinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
                                int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
                                uint8_t *dest, uint8_t *uDest, int dstW, int chrDstW, int dstFormat)
{
    //FIXME Optimize (just quickly writen not opti..)
    int i;
    for (i=0; i<dstW; i++)
    {
        int val=1<<18;
        int j;
        for (j=0; j<lumFilterSize; j++)
            val += lumSrc[j][i] * lumFilter[j];

        dest[i]= av_clip_uint8(val>>19);
    }

    if (uDest == NULL)
        return;

    if (dstFormat == PIX_FMT_NV12)
        for (i=0; i<chrDstW; i++)
        {
            int u=1<<18;
            int v=1<<18;
            int j;
            for (j=0; j<chrFilterSize; j++)
            {
                u += chrSrc[j][i] * chrFilter[j];
                v += chrSrc[j][i + 2048] * chrFilter[j];
            }

            uDest[2*i]= av_clip_uint8(u>>19);
            uDest[2*i+1]= av_clip_uint8(v>>19);
        }
    else
        for (i=0; i<chrDstW; i++)
        {
            int u=1<<18;
            int v=1<<18;
            int j;
            for (j=0; j<chrFilterSize; j++)
            {
                u += chrSrc[j][i] * chrFilter[j];
                v += chrSrc[j][i + 2048] * chrFilter[j];
            }

            uDest[2*i]= av_clip_uint8(v>>19);
            uDest[2*i+1]= av_clip_uint8(u>>19);
        }
}

#define YSCALE_YUV_2_PACKEDX_C(type) \
    for (i=0; i<(dstW>>1); i++){\
        int j;\
        int Y1 = 1<<18;\
        int Y2 = 1<<18;\
        int U  = 1<<18;\
        int V  = 1<<18;\
        type av_unused *r, *b, *g;\
        const int i2= 2*i;\
        \
        for (j=0; j<lumFilterSize; j++)\
        {\
            Y1 += lumSrc[j][i2] * lumFilter[j];\
            Y2 += lumSrc[j][i2+1] * lumFilter[j];\
        }\
        for (j=0; j<chrFilterSize; j++)\
        {\
            U += chrSrc[j][i] * chrFilter[j];\
            V += chrSrc[j][i+2048] * chrFilter[j];\
        }\
        Y1>>=19;\
        Y2>>=19;\
        U >>=19;\
        V >>=19;\
        if ((Y1|Y2|U|V)&256)\
        {\
            if (Y1>255)   Y1=255; \
            else if (Y1<0)Y1=0;   \
            if (Y2>255)   Y2=255; \
            else if (Y2<0)Y2=0;   \
            if (U>255)    U=255;  \
            else if (U<0) U=0;    \
            if (V>255)    V=255;  \
            else if (V<0) V=0;    \
        }

#define YSCALE_YUV_2_RGBX_C(type) \
    YSCALE_YUV_2_PACKEDX_C(type)  \
    r = (type *)c->table_rV[V];   \
    g = (type *)(c->table_gU[U] + c->table_gV[V]); \
    b = (type *)c->table_bU[U];   \

#define YSCALE_YUV_2_PACKED2_C   \
    for (i=0; i<(dstW>>1); i++){ \
        const int i2= 2*i;       \
        int Y1= (buf0[i2  ]*yalpha1+buf1[i2  ]*yalpha)>>19;           \
        int Y2= (buf0[i2+1]*yalpha1+buf1[i2+1]*yalpha)>>19;           \
        int U= (uvbuf0[i     ]*uvalpha1+uvbuf1[i     ]*uvalpha)>>19;  \
        int V= (uvbuf0[i+2048]*uvalpha1+uvbuf1[i+2048]*uvalpha)>>19;  \

#define YSCALE_YUV_2_RGB2_C(type) \
    YSCALE_YUV_2_PACKED2_C\
    type *r, *b, *g;\
    r = (type *)c->table_rV[V];\
    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
    b = (type *)c->table_bU[U];\

#define YSCALE_YUV_2_PACKED1_C \
    for (i=0; i<(dstW>>1); i++){\
        const int i2= 2*i;\
        int Y1= buf0[i2  ]>>7;\
        int Y2= buf0[i2+1]>>7;\
        int U= (uvbuf1[i     ])>>7;\
        int V= (uvbuf1[i+2048])>>7;\

#define YSCALE_YUV_2_RGB1_C(type) \
    YSCALE_YUV_2_PACKED1_C\
    type *r, *b, *g;\
    r = (type *)c->table_rV[V];\
    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
    b = (type *)c->table_bU[U];\

#define YSCALE_YUV_2_PACKED1B_C \
    for (i=0; i<(dstW>>1); i++){\
        const int i2= 2*i;\
        int Y1= buf0[i2  ]>>7;\
        int Y2= buf0[i2+1]>>7;\
        int U= (uvbuf0[i     ] + uvbuf1[i     ])>>8;\
        int V= (uvbuf0[i+2048] + uvbuf1[i+2048])>>8;\

#define YSCALE_YUV_2_RGB1B_C(type) \
    YSCALE_YUV_2_PACKED1B_C\
    type *r, *b, *g;\
    r = (type *)c->table_rV[V];\
    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
    b = (type *)c->table_bU[U];\

#define YSCALE_YUV_2_ANYRGB_C(func, func2)\
    switch(c->dstFormat)\
    {\
    case PIX_FMT_RGB32:\
    case PIX_FMT_BGR32:\
        func(uint32_t)\
            ((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];\
            ((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];\
        }                \
        break;\
    case PIX_FMT_RGB24:\
        func(uint8_t)\
            ((uint8_t*)dest)[0]= r[Y1];\
            ((uint8_t*)dest)[1]= g[Y1];\
            ((uint8_t*)dest)[2]= b[Y1];\
            ((uint8_t*)dest)[3]= r[Y2];\
            ((uint8_t*)dest)[4]= g[Y2];\
            ((uint8_t*)dest)[5]= b[Y2];\
            dest+=6;\
        }\
        break;\
    case PIX_FMT_BGR24:\
        func(uint8_t)\
            ((uint8_t*)dest)[0]= b[Y1];\
            ((uint8_t*)dest)[1]= g[Y1];\
            ((uint8_t*)dest)[2]= r[Y1];\
            ((uint8_t*)dest)[3]= b[Y2];\
            ((uint8_t*)dest)[4]= g[Y2];\
            ((uint8_t*)dest)[5]= r[Y2];\
            dest+=6;\
        }\
        break;\
    case PIX_FMT_RGB565:\
    case PIX_FMT_BGR565:\
        {\
            const int dr1= dither_2x2_8[y&1    ][0];\
            const int dg1= dither_2x2_4[y&1    ][0];\
            const int db1= dither_2x2_8[(y&1)^1][0];\
            const int dr2= dither_2x2_8[y&1    ][1];\
            const int dg2= dither_2x2_4[y&1    ][1];\
            const int db2= dither_2x2_8[(y&1)^1][1];\
            func(uint16_t)\
                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
            }\
        }\
        break;\
    case PIX_FMT_RGB555:\
    case PIX_FMT_BGR555:\
        {\
            const int dr1= dither_2x2_8[y&1    ][0];\
            const int dg1= dither_2x2_8[y&1    ][1];\
            const int db1= dither_2x2_8[(y&1)^1][0];\
            const int dr2= dither_2x2_8[y&1    ][1];\
            const int dg2= dither_2x2_8[y&1    ][0];\
            const int db2= dither_2x2_8[(y&1)^1][1];\
            func(uint16_t)\
                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
            }\
        }\
        break;\
    case PIX_FMT_RGB8:\
    case PIX_FMT_BGR8:\
        {\
            const uint8_t * const d64= dither_8x8_73[y&7];\
            const uint8_t * const d32= dither_8x8_32[y&7];\
            func(uint8_t)\
                ((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];\
                ((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];\
            }\
        }\
        break;\
    case PIX_FMT_RGB4:\
    case PIX_FMT_BGR4:\
        {\
            const uint8_t * const d64= dither_8x8_73 [y&7];\
            const uint8_t * const d128=dither_8x8_220[y&7];\
            func(uint8_t)\
                ((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]\
                                 + ((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);\
            }\
        }\
        break;\
    case PIX_FMT_RGB4_BYTE:\
    case PIX_FMT_BGR4_BYTE:\
        {\
            const uint8_t * const d64= dither_8x8_73 [y&7];\
            const uint8_t * const d128=dither_8x8_220[y&7];\
            func(uint8_t)\
                ((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];\
                ((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];\
            }\
        }\
        break;\
    case PIX_FMT_MONOBLACK:\
        {\
            const uint8_t * const d128=dither_8x8_220[y&7];\
            uint8_t *g= c->table_gU[128] + c->table_gV[128];\
            for (i=0; i<dstW-7; i+=8){\
                int acc;\
                acc =       g[((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19) + d128[0]];\
                acc+= acc + g[((buf0[i+1]*yalpha1+buf1[i+1]*yalpha)>>19) + d128[1]];\
                acc+= acc + g[((buf0[i+2]*yalpha1+buf1[i+2]*yalpha)>>19) + d128[2]];\
                acc+= acc + g[((buf0[i+3]*yalpha1+buf1[i+3]*yalpha)>>19) + d128[3]];\
                acc+= acc + g[((buf0[i+4]*yalpha1+buf1[i+4]*yalpha)>>19) + d128[4]];\
                acc+= acc + g[((buf0[i+5]*yalpha1+buf1[i+5]*yalpha)>>19) + d128[5]];\
                acc+= acc + g[((buf0[i+6]*yalpha1+buf1[i+6]*yalpha)>>19) + d128[6]];\
                acc+= acc + g[((buf0[i+7]*yalpha1+buf1[i+7]*yalpha)>>19) + d128[7]];\
                ((uint8_t*)dest)[0]= acc;\
                dest++;\
            }\
\
/*\
((uint8_t*)dest)-= dstW>>4;\
{\
            int acc=0;\
            int left=0;\
            static int top[1024];\
            static int last_new[1024][1024];\
            static int last_in3[1024][1024];\
            static int drift[1024][1024];\
            int topLeft=0;\
            int shift=0;\
            int count=0;\
            const uint8_t * const d128=dither_8x8_220[y&7];\
            int error_new=0;\
            int error_in3=0;\
            int f=0;\
            \
            for (i=dstW>>1; i<dstW; i++){\
                int in= ((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19);\
                int in2 = (76309 * (in - 16) + 32768) >> 16;\
                int in3 = (in2 < 0) ? 0 : ((in2 > 255) ? 255 : in2);\
                int old= (left*7 + topLeft + top[i]*5 + top[i+1]*3)/20 + in3\
                         + (last_new[y][i] - in3)*f/256;\
                int new= old> 128 ? 255 : 0;\
\
                error_new+= FFABS(last_new[y][i] - new);\
                error_in3+= FFABS(last_in3[y][i] - in3);\
                f= error_new - error_in3*4;\
                if (f<0) f=0;\
                if (f>256) f=256;\
\
                topLeft= top[i];\
                left= top[i]= old - new;\
                last_new[y][i]= new;\
                last_in3[y][i]= in3;\
\
                acc+= acc + (new&1);\
                if ((i&7)==6){\
                    ((uint8_t*)dest)[0]= acc;\
                    ((uint8_t*)dest)++;\
                }\
            }\
}\
*/\
        }\
        break;\
    case PIX_FMT_YUYV422:\
        func2\
            ((uint8_t*)dest)[2*i2+0]= Y1;\
            ((uint8_t*)dest)[2*i2+1]= U;\
            ((uint8_t*)dest)[2*i2+2]= Y2;\
            ((uint8_t*)dest)[2*i2+3]= V;\
        }                \
        break;\
    case PIX_FMT_UYVY422:\
        func2\
            ((uint8_t*)dest)[2*i2+0]= U;\
            ((uint8_t*)dest)[2*i2+1]= Y1;\
            ((uint8_t*)dest)[2*i2+2]= V;\
            ((uint8_t*)dest)[2*i2+3]= Y2;\
        }                \
        break;\
    }\


static inline void yuv2packedXinC(SwsContext *c, int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
                                  int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
                                  uint8_t *dest, int dstW, int y)
{
    int i;