-#if 0
-const uint8_t __attribute__((aligned(8))) dither_8x8_64[8][8]={
-{ 0, 48, 12, 60, 3, 51, 15, 63, },
-{ 32, 16, 44, 28, 35, 19, 47, 31, },
-{ 8, 56, 4, 52, 11, 59, 7, 55, },
-{ 40, 24, 36, 20, 43, 27, 39, 23, },
-{ 2, 50, 14, 62, 1, 49, 13, 61, },
-{ 34, 18, 46, 30, 33, 17, 45, 29, },
-{ 10, 58, 6, 54, 9, 57, 5, 53, },
-{ 42, 26, 38, 22, 41, 25, 37, 21, },
-};
-#endif
-
-const uint8_t __attribute__((aligned(8))) dither_8x8_73[8][8]={
-{ 0, 55, 14, 68, 3, 58, 17, 72, },
-{ 37, 18, 50, 32, 40, 22, 54, 35, },
-{ 9, 64, 5, 59, 13, 67, 8, 63, },
-{ 46, 27, 41, 23, 49, 31, 44, 26, },
-{ 2, 57, 16, 71, 1, 56, 15, 70, },
-{ 39, 21, 52, 34, 38, 19, 51, 33, },
-{ 11, 66, 7, 62, 10, 65, 6, 60, },
-{ 48, 30, 43, 25, 47, 29, 42, 24, },
-};
-
-#if 0
-const uint8_t __attribute__((aligned(8))) dither_8x8_128[8][8]={
-{ 68, 36, 92, 60, 66, 34, 90, 58, },
-{ 20, 116, 12, 108, 18, 114, 10, 106, },
-{ 84, 52, 76, 44, 82, 50, 74, 42, },
-{ 0, 96, 24, 120, 6, 102, 30, 126, },
-{ 64, 32, 88, 56, 70, 38, 94, 62, },
-{ 16, 112, 8, 104, 22, 118, 14, 110, },
-{ 80, 48, 72, 40, 86, 54, 78, 46, },
-{ 4, 100, 28, 124, 2, 98, 26, 122, },
-};
-#endif
-
-#if 1
-const uint8_t __attribute__((aligned(8))) dither_8x8_220[8][8]={
-{117, 62, 158, 103, 113, 58, 155, 100, },
-{ 34, 199, 21, 186, 31, 196, 17, 182, },
-{144, 89, 131, 76, 141, 86, 127, 72, },
-{ 0, 165, 41, 206, 10, 175, 52, 217, },
-{110, 55, 151, 96, 120, 65, 162, 107, },
-{ 28, 193, 14, 179, 38, 203, 24, 189, },
-{138, 83, 124, 69, 148, 93, 134, 79, },
-{ 7, 172, 48, 213, 3, 168, 45, 210, },
-};
-#elif 1
-// tries to correct a gamma of 1.5
-const uint8_t __attribute__((aligned(8))) dither_8x8_220[8][8]={
-{ 0, 143, 18, 200, 2, 156, 25, 215, },
-{ 78, 28, 125, 64, 89, 36, 138, 74, },
-{ 10, 180, 3, 161, 16, 195, 8, 175, },
-{109, 51, 93, 38, 121, 60, 105, 47, },
-{ 1, 152, 23, 210, 0, 147, 20, 205, },
-{ 85, 33, 134, 71, 81, 30, 130, 67, },
-{ 14, 190, 6, 171, 12, 185, 5, 166, },
-{117, 57, 101, 44, 113, 54, 97, 41, },
-};
-#elif 1
-// tries to correct a gamma of 2.0
-const uint8_t __attribute__((aligned(8))) dither_8x8_220[8][8]={
-{ 0, 124, 8, 193, 0, 140, 12, 213, },
-{ 55, 14, 104, 42, 66, 19, 119, 52, },
-{ 3, 168, 1, 145, 6, 187, 3, 162, },
-{ 86, 31, 70, 21, 99, 39, 82, 28, },
-{ 0, 134, 11, 206, 0, 129, 9, 200, },
-{ 62, 17, 114, 48, 58, 16, 109, 45, },
-{ 5, 181, 2, 157, 4, 175, 1, 151, },
-{ 95, 36, 78, 26, 90, 34, 74, 24, },
-};
-#else
-// tries to correct a gamma of 2.5
-const uint8_t __attribute__((aligned(8))) dither_8x8_220[8][8]={
-{ 0, 107, 3, 187, 0, 125, 6, 212, },
-{ 39, 7, 86, 28, 49, 11, 102, 36, },
-{ 1, 158, 0, 131, 3, 180, 1, 151, },
-{ 68, 19, 52, 12, 81, 25, 64, 17, },
-{ 0, 119, 5, 203, 0, 113, 4, 195, },
-{ 45, 9, 96, 33, 42, 8, 91, 30, },
-{ 2, 172, 1, 144, 2, 165, 0, 137, },
-{ 77, 23, 60, 15, 72, 21, 56, 14, },
-};
-#endif
-
-#ifdef HAVE_MMX
-
-/* hope these constant values are cache line aligned */
-static uint64_t attribute_used __attribute__((aligned(8))) mmx_00ffw = 0x00ff00ff00ff00ffULL;
-static uint64_t attribute_used __attribute__((aligned(8))) mmx_redmask = 0xf8f8f8f8f8f8f8f8ULL;
-static uint64_t attribute_used __attribute__((aligned(8))) mmx_grnmask = 0xfcfcfcfcfcfcfcfcULL;
-
-static uint64_t attribute_used __attribute__((aligned(8))) M24A= 0x00FF0000FF0000FFULL;
-static uint64_t attribute_used __attribute__((aligned(8))) M24B= 0xFF0000FF0000FF00ULL;
-static uint64_t attribute_used __attribute__((aligned(8))) M24C= 0x0000FF0000FF0000ULL;
-
-// the volatile is required because gcc otherwise optimizes some writes away not knowing that these
-// are read in the asm block
-static volatile uint64_t attribute_used __attribute__((aligned(8))) b5Dither;
-static volatile uint64_t attribute_used __attribute__((aligned(8))) g5Dither;
-static volatile uint64_t attribute_used __attribute__((aligned(8))) g6Dither;
-static volatile uint64_t attribute_used __attribute__((aligned(8))) r5Dither;
-
-static uint64_t __attribute__((aligned(8))) dither4[2]={
- 0x0103010301030103LL,
- 0x0200020002000200LL,};
-
-static uint64_t __attribute__((aligned(8))) dither8[2]={
- 0x0602060206020602LL,
- 0x0004000400040004LL,};
-
-#undef HAVE_MMX
-
-//MMX versions
-#undef RENAME
-#define HAVE_MMX
-#undef HAVE_MMX2
-#undef HAVE_3DNOW
-#define RENAME(a) a ## _MMX
-#include "yuv2rgb_template.c"
-
-//MMX2 versions
-#undef RENAME
-#define HAVE_MMX
-#define HAVE_MMX2
-#undef HAVE_3DNOW
-#define RENAME(a) a ## _MMX2
-#include "yuv2rgb_template.c"
-
-#endif /* defined(ARCH_X86) */
-
-const int32_t Inverse_Table_6_9[8][4] = {
- {117504, 138453, 13954, 34903}, /* no sequence_display_extension */
- {117504, 138453, 13954, 34903}, /* ITU-R Rec. 709 (1990) */
- {104597, 132201, 25675, 53279}, /* unspecified */
- {104597, 132201, 25675, 53279}, /* reserved */
- {104448, 132798, 24759, 53109}, /* FCC */
- {104597, 132201, 25675, 53279}, /* ITU-R Rec. 624-4 System B, G */
- {104597, 132201, 25675, 53279}, /* SMPTE 170M */
- {117579, 136230, 16907, 35559} /* SMPTE 240M (1987) */
-};
-
-#define RGB(i) \
- U = pu[i]; \
- V = pv[i]; \
- r = (void *)c->table_rV[V]; \
- g = (void *)(c->table_gU[U] + c->table_gV[V]); \
- b = (void *)c->table_bU[U];
-
-#define DST1(i) \
- Y = py_1[2*i]; \
- dst_1[2*i] = r[Y] + g[Y] + b[Y]; \
- Y = py_1[2*i+1]; \
- dst_1[2*i+1] = r[Y] + g[Y] + b[Y];
-
-#define DST2(i) \
- Y = py_2[2*i]; \
- dst_2[2*i] = r[Y] + g[Y] + b[Y]; \
- Y = py_2[2*i+1]; \
- dst_2[2*i+1] = r[Y] + g[Y] + b[Y];
-
-#define DST1RGB(i) \
- Y = py_1[2*i]; \
- dst_1[6*i] = r[Y]; dst_1[6*i+1] = g[Y]; dst_1[6*i+2] = b[Y]; \
- Y = py_1[2*i+1]; \
- dst_1[6*i+3] = r[Y]; dst_1[6*i+4] = g[Y]; dst_1[6*i+5] = b[Y];
-
-#define DST2RGB(i) \
- Y = py_2[2*i]; \
- dst_2[6*i] = r[Y]; dst_2[6*i+1] = g[Y]; dst_2[6*i+2] = b[Y]; \
- Y = py_2[2*i+1]; \
- dst_2[6*i+3] = r[Y]; dst_2[6*i+4] = g[Y]; dst_2[6*i+5] = b[Y];
-
-#define DST1BGR(i) \
- Y = py_1[2*i]; \
- dst_1[6*i] = b[Y]; dst_1[6*i+1] = g[Y]; dst_1[6*i+2] = r[Y]; \
- Y = py_1[2*i+1]; \
- dst_1[6*i+3] = b[Y]; dst_1[6*i+4] = g[Y]; dst_1[6*i+5] = r[Y];
-
-#define DST2BGR(i) \
- Y = py_2[2*i]; \
- dst_2[6*i] = b[Y]; dst_2[6*i+1] = g[Y]; dst_2[6*i+2] = r[Y]; \
- Y = py_2[2*i+1]; \
- dst_2[6*i+3] = b[Y]; dst_2[6*i+4] = g[Y]; dst_2[6*i+5] = r[Y];
-
-#define PROLOG(func_name, dst_type) \
-static int func_name(SwsContext *c, uint8_t* src[], int srcStride[], int srcSliceY, \
- int srcSliceH, uint8_t* dst[], int dstStride[]){\
- int y;\
-\
- if(c->srcFormat == PIX_FMT_YUV422P){\
- srcStride[1] *= 2;\
- srcStride[2] *= 2;\
- }\
- for(y=0; y<srcSliceH; y+=2){\
- dst_type *dst_1= (dst_type*)(dst[0] + (y+srcSliceY )*dstStride[0]);\
- dst_type *dst_2= (dst_type*)(dst[0] + (y+srcSliceY+1)*dstStride[0]);\
- dst_type attribute_unused *r, *b;\
- dst_type *g;\
- uint8_t *py_1= src[0] + y*srcStride[0];\
- uint8_t *py_2= py_1 + srcStride[0];\
- uint8_t *pu= src[1] + (y>>1)*srcStride[1];\
- uint8_t *pv= src[2] + (y>>1)*srcStride[2];\
- unsigned int h_size= c->dstW>>3;\
- while (h_size--) {\
- int attribute_unused U, V;\
- int Y;\
-
-#define EPILOG(dst_delta)\
- pu += 4;\
- pv += 4;\
- py_1 += 8;\
- py_2 += 8;\
- dst_1 += dst_delta;\
- dst_2 += dst_delta;\
- }\
- }\
- return srcSliceH;\