diff options
Diffstat (limited to 'libavcodec/dsputil.c')
| -rw-r--r-- | libavcodec/dsputil.c | 320 |
1 files changed, 160 insertions, 160 deletions
diff --git a/libavcodec/dsputil.c b/libavcodec/dsputil.c index bfab8e2341..933020a83a 100644 --- a/libavcodec/dsputil.c +++ b/libavcodec/dsputil.c @@ -19,12 +19,12 @@ * * gmc & q-pel & 32/64 bit based MC by Michael Niedermayer <michaelni@gmx.at> */ - + /** * @file dsputil.c * DSP utils */ - + #include "avcodec.h" #include "dsputil.h" #include "mpegvideo.h" @@ -65,72 +65,72 @@ const uint8_t ff_zigzag248_direct[64] = { uint16_t __align8 inv_zigzag_direct16[64] = {0, }; const uint8_t ff_alternate_horizontal_scan[64] = { - 0, 1, 2, 3, 8, 9, 16, 17, + 0, 1, 2, 3, 8, 9, 16, 17, 10, 11, 4, 5, 6, 7, 15, 14, - 13, 12, 19, 18, 24, 25, 32, 33, + 13, 12, 19, 18, 24, 25, 32, 33, 26, 27, 20, 21, 22, 23, 28, 29, - 30, 31, 34, 35, 40, 41, 48, 49, + 30, 31, 34, 35, 40, 41, 48, 49, 42, 43, 36, 37, 38, 39, 44, 45, - 46, 47, 50, 51, 56, 57, 58, 59, + 46, 47, 50, 51, 56, 57, 58, 59, 52, 53, 54, 55, 60, 61, 62, 63, }; const uint8_t ff_alternate_vertical_scan[64] = { - 0, 8, 16, 24, 1, 9, 2, 10, + 0, 8, 16, 24, 1, 9, 2, 10, 17, 25, 32, 40, 48, 56, 57, 49, - 41, 33, 26, 18, 3, 11, 4, 12, + 41, 33, 26, 18, 3, 11, 4, 12, 19, 27, 34, 42, 50, 58, 35, 43, - 51, 59, 20, 28, 5, 13, 6, 14, + 51, 59, 20, 28, 5, 13, 6, 14, 21, 29, 36, 44, 52, 60, 37, 45, - 53, 61, 22, 30, 7, 15, 23, 31, + 53, 61, 22, 30, 7, 15, 23, 31, 38, 46, 54, 62, 39, 47, 55, 63, }; /* a*inverse[b]>>32 == a/b for all 0<=a<=65536 && 2<=b<=255 */ const uint32_t inverse[256]={ - 0, 4294967295U,2147483648U,1431655766, 1073741824, 858993460, 715827883, 613566757, - 536870912, 477218589, 429496730, 390451573, 357913942, 330382100, 306783379, 286331154, - 268435456, 252645136, 238609295, 226050911, 214748365, 204522253, 195225787, 186737709, - 178956971, 171798692, 165191050, 159072863, 153391690, 148102321, 143165577, 138547333, - 134217728, 130150525, 126322568, 122713352, 119304648, 116080198, 113025456, 110127367, - 107374183, 104755300, 102261127, 99882961, 97612894, 95443718, 93368855, 91382283, - 89478486, 87652394, 85899346, 84215046, 82595525, 81037119, 79536432, 78090315, - 76695845, 75350304, 74051161, 72796056, 71582789, 70409300, 69273667, 68174085, - 67108864, 66076420, 65075263, 64103990, 63161284, 62245903, 61356676, 60492498, - 59652324, 58835169, 58040099, 57266231, 56512728, 55778797, 55063684, 54366675, - 53687092, 53024288, 52377650, 51746594, 51130564, 50529028, 49941481, 49367441, - 48806447, 48258060, 47721859, 47197443, 46684428, 46182445, 45691142, 45210183, - 44739243, 44278014, 43826197, 43383509, 42949673, 42524429, 42107523, 41698712, - 41297763, 40904451, 40518560, 40139882, 39768216, 39403370, 39045158, 38693400, - 38347923, 38008561, 37675152, 37347542, 37025581, 36709123, 36398028, 36092163, - 35791395, 35495598, 35204650, 34918434, 34636834, 34359739, 34087043, 33818641, - 33554432, 33294321, 33038210, 32786010, 32537632, 32292988, 32051995, 31814573, - 31580642, 31350127, 31122952, 30899046, 30678338, 30460761, 30246249, 30034737, - 29826162, 29620465, 29417585, 29217465, 29020050, 28825284, 28633116, 28443493, - 28256364, 28071682, 27889399, 27709467, 27531842, 27356480, 27183338, 27012373, - 26843546, 26676816, 26512144, 26349493, 26188825, 26030105, 25873297, 25718368, - 25565282, 25414008, 25264514, 25116768, 24970741, 24826401, 24683721, 24542671, - 24403224, 24265352, 24129030, 23994231, 23860930, 23729102, 23598722, 23469767, - 23342214, 23216040, 23091223, 22967740, 22845571, 22724695, 22605092, 22486740, - 22369622, 22253717, 22139007, 22025474, 21913099, 21801865, 21691755, 21582751, - 21474837, 21367997, 21262215, 21157475, 21053762, 20951060, 20849356, 20748635, - 20648882, 20550083, 20452226, 20355296, 20259280, 20164166, 20069941, 19976593, - 19884108, 19792477, 19701685, 19611723, 19522579, 19434242, 19346700, 19259944, - 19173962, 19088744, 19004281, 18920561, 18837576, 18755316, 18673771, 18592933, - 18512791, 18433337, 18354562, 18276457, 18199014, 18122225, 18046082, 17970575, - 17895698, 17821442, 17747799, 17674763, 17602325, 17530479, 17459217, 17388532, + 0, 4294967295U,2147483648U,1431655766, 1073741824, 858993460, 715827883, 613566757, + 536870912, 477218589, 429496730, 390451573, 357913942, 330382100, 306783379, 286331154, + 268435456, 252645136, 238609295, 226050911, 214748365, 204522253, 195225787, 186737709, + 178956971, 171798692, 165191050, 159072863, 153391690, 148102321, 143165577, 138547333, + 134217728, 130150525, 126322568, 122713352, 119304648, 116080198, 113025456, 110127367, + 107374183, 104755300, 102261127, 99882961, 97612894, 95443718, 93368855, 91382283, + 89478486, 87652394, 85899346, 84215046, 82595525, 81037119, 79536432, 78090315, + 76695845, 75350304, 74051161, 72796056, 71582789, 70409300, 69273667, 68174085, + 67108864, 66076420, 65075263, 64103990, 63161284, 62245903, 61356676, 60492498, + 59652324, 58835169, 58040099, 57266231, 56512728, 55778797, 55063684, 54366675, + 53687092, 53024288, 52377650, 51746594, 51130564, 50529028, 49941481, 49367441, + 48806447, 48258060, 47721859, 47197443, 46684428, 46182445, 45691142, 45210183, + 44739243, 44278014, 43826197, 43383509, 42949673, 42524429, 42107523, 41698712, + 41297763, 40904451, 40518560, 40139882, 39768216, 39403370, 39045158, 38693400, + 38347923, 38008561, 37675152, 37347542, 37025581, 36709123, 36398028, 36092163, + 35791395, 35495598, 35204650, 34918434, 34636834, 34359739, 34087043, 33818641, + 33554432, 33294321, 33038210, 32786010, 32537632, 32292988, 32051995, 31814573, + 31580642, 31350127, 31122952, 30899046, 30678338, 30460761, 30246249, 30034737, + 29826162, 29620465, 29417585, 29217465, 29020050, 28825284, 28633116, 28443493, + 28256364, 28071682, 27889399, 27709467, 27531842, 27356480, 27183338, 27012373, + 26843546, 26676816, 26512144, 26349493, 26188825, 26030105, 25873297, 25718368, + 25565282, 25414008, 25264514, 25116768, 24970741, 24826401, 24683721, 24542671, + 24403224, 24265352, 24129030, 23994231, 23860930, 23729102, 23598722, 23469767, + 23342214, 23216040, 23091223, 22967740, 22845571, 22724695, 22605092, 22486740, + 22369622, 22253717, 22139007, 22025474, 21913099, 21801865, 21691755, 21582751, + 21474837, 21367997, 21262215, 21157475, 21053762, 20951060, 20849356, 20748635, + 20648882, 20550083, 20452226, 20355296, 20259280, 20164166, 20069941, 19976593, + 19884108, 19792477, 19701685, 19611723, 19522579, 19434242, 19346700, 19259944, + 19173962, 19088744, 19004281, 18920561, 18837576, 18755316, 18673771, 18592933, + 18512791, 18433337, 18354562, 18276457, 18199014, 18122225, 18046082, 17970575, + 17895698, 17821442, 17747799, 17674763, 17602325, 17530479, 17459217, 17388532, 17318417, 17248865, 17179870, 17111424, 17043522, 16976156, 16909321, 16843010, }; /* Input permutation for the simple_idct_mmx */ static const uint8_t simple_mmx_permutation[64]={ - 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, - 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, - 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, - 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, - 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, - 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, - 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, + 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, + 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, + 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, + 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, + 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, + 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, + 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, }; @@ -206,7 +206,7 @@ static int pix_norm1_c(uint8_t * pix, int line_size) static void bswap_buf(uint32_t *dst, uint32_t *src, int w){ int i; - + for(i=0; i+8<=w; i+=8){ dst[i+0]= bswap_32(src[i+0]); dst[i+1]= bswap_32(src[i+1]); @@ -298,7 +298,7 @@ static inline int w_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in int tmp[16*16]; #if 0 int level, ori; - static const int scale[2][2][4][4]={ + static const int scale[2][2][4][4]={ { { //8x8 dec=3 @@ -350,7 +350,7 @@ static inline int w_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in int stride= 16<<(dec_count-level); int sy= (ori&2) ? stride>>1 : 0; int size= 1<<level; - + for(i=0; i<size; i++){ for(j=0; j<size; j++){ int v= tmp[sx + sy + i*stride + j] * scale[type][dec_count-3][level][ori]; @@ -368,8 +368,8 @@ static inline int w_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in s+= ABS(tmp[16*i+j+3]); } } - assert(s>=0); - + assert(s>=0); + return s>>2; #endif } @@ -435,7 +435,7 @@ static void put_pixels_clamped_c(const DCTELEM *block, uint8_t *restrict pixels, { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<8;i++) { pixels[0] = cm[block[0]]; @@ -457,7 +457,7 @@ static void put_pixels_clamped4_c(const DCTELEM *block, uint8_t *restrict pixels { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<4;i++) { pixels[0] = cm[block[0]]; @@ -475,7 +475,7 @@ static void put_pixels_clamped2_c(const DCTELEM *block, uint8_t *restrict pixels { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<2;i++) { pixels[0] = cm[block[0]]; @@ -486,7 +486,7 @@ static void put_pixels_clamped2_c(const DCTELEM *block, uint8_t *restrict pixels } } -static void put_signed_pixels_clamped_c(const DCTELEM *block, +static void put_signed_pixels_clamped_c(const DCTELEM *block, uint8_t *restrict pixels, int line_size) { @@ -512,7 +512,7 @@ static void add_pixels_clamped_c(const DCTELEM *block, uint8_t *restrict pixels, { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<8;i++) { pixels[0] = cm[pixels[0] + block[0]]; @@ -533,7 +533,7 @@ static void add_pixels_clamped4_c(const DCTELEM *block, uint8_t *restrict pixels { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<4;i++) { pixels[0] = cm[pixels[0] + block[0]]; @@ -550,7 +550,7 @@ static void add_pixels_clamped2_c(const DCTELEM *block, uint8_t *restrict pixels { int i; uint8_t *cm = cropTbl + MAX_NEG_CROP; - + /* read the pixels */ for(i=0;i<2;i++) { pixels[0] = cm[pixels[0] + block[0]]; @@ -1143,12 +1143,12 @@ static void gmc1_c(uint8_t *dst, uint8_t *src, int stride, int h, int x16, int y } } -static void gmc_c(uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy, +static void gmc_c(uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy, 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--; @@ -1166,7 +1166,7 @@ static void gmc_c(uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy, 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; @@ -1176,23 +1176,23 @@ static void gmc_c(uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy, + 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) + 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) + 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; + index= clip(src_x, 0, width) + clip(src_y, 0, height)*stride; dst[y*stride + x]= src[index ]; } } - + vx+= dxx; vy+= dyx; } @@ -1231,7 +1231,7 @@ static inline void put_tpel_pixels_mc20_c(uint8_t *dst, const uint8_t *src, int 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++) { @@ -1242,7 +1242,7 @@ static inline void put_tpel_pixels_mc01_c(uint8_t *dst, const uint8_t *src, int 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++) { @@ -1328,7 +1328,7 @@ static inline void avg_tpel_pixels_mc20_c(uint8_t *dst, const uint8_t *src, int 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++) { @@ -1339,7 +1339,7 @@ static inline void avg_tpel_pixels_mc01_c(uint8_t *dst, const uint8_t *src, int 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++) { @@ -2496,7 +2496,7 @@ static void wmv2_mspel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int dst[6]= cm[(9*(src[6] + src[7]) - (src[ 5] + src[8]) + 8)>>4]; dst[7]= cm[(9*(src[7] + src[8]) - (src[ 6] + src[9]) + 8)>>4]; dst+=dstStride; - src+=srcStride; + src+=srcStride; } } @@ -2580,7 +2580,7 @@ static void put_mspel8_mc22_c(uint8_t *dst, uint8_t *src, int stride){ static void h263_v_loop_filter_c(uint8_t *src, int stride, int qscale){ int x; const int strength= ff_h263_loop_filter_strength[qscale]; - + for(x=0; x<8; x++){ int d1, d2, ad1; int p0= src[x-2*stride]; @@ -2594,19 +2594,19 @@ static void h263_v_loop_filter_c(uint8_t *src, int stride, int qscale){ else if(d< strength) d1= d; else if(d< 2*strength) d1= 2*strength - d; else d1= 0; - + p1 += d1; p2 -= d1; if(p1&256) p1= ~(p1>>31); if(p2&256) p2= ~(p2>>31); - + src[x-1*stride] = p1; src[x+0*stride] = p2; ad1= ABS(d1)>>1; - + d2= clip((p0-p3)/4, -ad1, ad1); - + src[x-2*stride] = p0 - d2; src[x+ stride] = p3 + d2; } @@ -2615,7 +2615,7 @@ static void h263_v_loop_filter_c(uint8_t *src, int stride, int qscale){ static void h263_h_loop_filter_c(uint8_t *src, int stride, int qscale){ int y; const int strength= ff_h263_loop_filter_strength[qscale]; - + for(y=0; y<8; y++){ int d1, d2, ad1; int p0= src[y*stride-2]; @@ -2629,19 +2629,19 @@ static void h263_h_loop_filter_c(uint8_t *src, int stride, int qscale){ else if(d< strength) d1= d; else if(d< 2*strength) d1= 2*strength - d; else d1= 0; - + p1 += d1; p2 -= d1; if(p1&256) p1= ~(p1>>31); if(p2&256) p2= ~(p2>>31); - + src[y*stride-1] = p1; src[y*stride+0] = p2; ad1= ABS(d1)>>1; - + d2= clip((p0-p3)/4, -ad1, ad1); - + src[y*stride-2] = p0 - d2; src[y*stride+1] = p3 + d2; } @@ -2662,7 +2662,7 @@ static void h261_loop_filter_c(uint8_t *src, int stride){ temp[yz] = src[xy - stride] + 2*src[xy] + src[xy + stride]; } } - + for(y=0; y<8; y++){ src[ y*stride] = (temp[ y*8] + 2)>>2; src[7+y*stride] = (temp[7+y*8] + 2)>>2; @@ -2689,14 +2689,14 @@ static inline void h264_loop_filter_luma_c(uint8_t *pix, int xstride, int ystrid const int q0 = pix[0]; const int q1 = pix[1*xstride]; const int q2 = pix[2*xstride]; - + if( ABS( p0 - q0 ) < alpha && ABS( p1 - p0 ) < beta && ABS( q1 - q0 ) < beta ) { - + int tc = tc0[i]; int i_delta; - + if( ABS( p2 - p0 ) < beta ) { pix[-2*xstride] = p1 + clip( (( p2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - p1, -tc0[i], tc0[i] ); tc++; @@ -2705,7 +2705,7 @@ static inline void h264_loop_filter_luma_c(uint8_t *pix, int xstride, int ystrid pix[ xstride] = q1 + clip( (( q2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - q1, -tc0[i], tc0[i] ); tc++; } - + i_delta = clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc ); pix[-xstride] = clip_uint8( p0 + i_delta ); /* p0' */ pix[0] = clip_uint8( q0 - i_delta ); /* q0' */ @@ -3019,7 +3019,7 @@ static int nsse8_c(void *v, uint8_t *s1, uint8_t *s2, int stride, int h){ int score1=0; int score2=0; int x,y; - + for(y=0; y<h; y++){ for(x=0; x<8; x++){ score1+= (s1[x ] - s2[x ])*(s1[x ] - s2[x ]); @@ -3035,7 +3035,7 @@ static int nsse8_c(void *v, uint8_t *s1, uint8_t *s2, int stride, int h){ s1+= stride; s2+= stride; } - + if(c) return score1 + ABS(score2)*c->avctx->nsse_weight; else return score1 + ABS(score2)*8; } @@ -3060,7 +3060,7 @@ static void add_8x8basis_c(int16_t rem[64], int16_t basis[64], int scale){ for(i=0; i<8*8; i++){ rem[i] += (basis[i]*scale + (1<<(BASIS_SHIFT - RECON_SHIFT-1)))>>(BASIS_SHIFT - RECON_SHIFT); - } + } } /** @@ -3068,14 +3068,14 @@ static void add_8x8basis_c(int16_t rem[64], int16_t basis[64], int scale){ * @param block the block which will be permuted according to the given permutation vector * @param permutation the permutation vector * @param last the last non zero coefficient in scantable order, used to speed the permutation up - * @param scantable the used scantable, this is only used to speed the permutation up, the block is not + * @param scantable the used scantable, this is only used to speed the permutation up, the block is not * (inverse) permutated to scantable order! */ void ff_block_permute(DCTELEM *block, uint8_t *permutation, const uint8_t *scantable, int last) { int i; DCTELEM temp[64]; - + if(last<=0) return; //if(permutation[1]==1) return; //FIXME its ok but not clean and might fail for some perms @@ -3084,7 +3084,7 @@ void ff_block_permute(DCTELEM *block, uint8_t *permutation, const uint8_t *scant temp[j]= block[j]; block[j]=0; } - + for(i=0; i<=last; i++){ const int j= scantable[i]; const int perm_j= permutation[j]; @@ -3098,9 +3098,9 @@ static int zero_cmp(void *s, uint8_t *a, uint8_t *b, int stride, int h){ void ff_set_cmp(DSPContext* c, me_cmp_func *cmp, int type){ int i; - + memset(cmp, 0, sizeof(void*)*5); - + for(i=0; i<5; i++){ switch(type&0xFF){ case FF_CMP_SAD: @@ -3203,7 +3203,7 @@ static void sub_hfyu_median_prediction_c(uint8_t *dst, uint8_t *src1, uint8_t *s lt= src1[i]; l= src2[i]; dst[i]= l - pred; - } + } *left= l; *left_top= lt; @@ -3228,7 +3228,7 @@ static int hadamard8_diff8x8_c(/*MpegEncContext*/ void *s, uint8_t *dst, uint8_t int i; int temp[64]; int sum=0; - + assert(h==8); for(i=0; i<8; i++){ @@ -3237,12 +3237,12 @@ static int hadamard8_diff8x8_c(/*MpegEncContext*/ void *s, uint8_t *dst, uint8_t BUTTERFLY2(temp[8*i+2], temp[8*i+3], src[stride*i+2]-dst[stride*i+2],src[stride*i+3]-dst[stride*i+3]); BUTTERFLY2(temp[8*i+4], temp[8*i+5], src[stride*i+4]-dst[stride*i+4],src[stride*i+5]-dst[stride*i+5]); BUTTERFLY2(temp[8*i+6], temp[8*i+7], src[stride*i+6]-dst[stride*i+6],src[stride*i+7]-dst[stride*i+7]); - + BUTTERFLY1(temp[8*i+0], temp[8*i+2]); BUTTERFLY1(temp[8*i+1], temp[8*i+3]); BUTTERFLY1(temp[8*i+4], temp[8*i+6]); BUTTERFLY1(temp[8*i+5], temp[8*i+7]); - + BUTTERFLY1(temp[8*i+0], temp[8*i+4]); BUTTERFLY1(temp[8*i+1], temp[8*i+5]); BUTTERFLY1(temp[8*i+2], temp[8*i+6]); @@ -3254,13 +3254,13 @@ static int hadamard8_diff8x8_c(/*MpegEncContext*/ void *s, uint8_t *dst, uint8_t BUTTERFLY1(temp[8*2+i], temp[8*3+i]); BUTTERFLY1(temp[8*4+i], temp[8*5+i]); BUTTERFLY1(temp[8*6+i], temp[8*7+i]); - + BUTTERFLY1(temp[8*0+i], temp[8*2+i]); BUTTERFLY1(temp[8*1+i], temp[8*3+i]); BUTTERFLY1(temp[8*4+i], temp[8*6+i]); BUTTERFLY1(temp[8*5+i], temp[8*7+i]); - sum += + sum += BUTTERFLYA(temp[8*0+i], temp[8*4+i]) +BUTTERFLYA(temp[8*1+i], temp[8*5+i]) +BUTTERFLYA(temp[8*2+i], temp[8*6+i]) @@ -3280,21 +3280,21 @@ static int hadamard8_intra8x8_c(/*MpegEncContext*/ void *s, uint8_t *src, uint8_ int i; int temp[64]; int sum=0; - + assert(h==8); - + for(i=0; i<8; i++){ //FIXME try pointer walks BUTTERFLY2(temp[8*i+0], temp[8*i+1], src[stride*i+0],src[stride*i+1]); BUTTERFLY2(temp[8*i+2], temp[8*i+3], src[stride*i+2],src[stride*i+3]); BUTTERFLY2(temp[8*i+4], temp[8*i+5], src[stride*i+4],src[stride*i+5]); BUTTERFLY2(temp[8*i+6], temp[8*i+7], src[stride*i+6],src[stride*i+7]); - + BUTTERFLY1(temp[8*i+0], temp[8*i+2]); BUTTERFLY1(temp[8*i+1], temp[8*i+3]); BUTTERFLY1(temp[8*i+4], temp[8*i+6]); BUTTERFLY1(temp[8*i+5], temp[8*i+7]); - + BUTTERFLY1(temp[8*i+0], temp[8*i+4]); BUTTERFLY1(temp[8*i+1], temp[8*i+5]); BUTTERFLY1(temp[8*i+2], temp[8*i+6]); @@ -3306,21 +3306,21 @@ static int hadamard8_intra8x8_c(/*MpegEncContext*/ void *s, uint8_t *src, uint8_ BUTTERFLY1(temp[8*2+i], temp[8*3+i]); BUTTERFLY1(temp[8*4+i], temp[8*5+i]); BUTTERFLY1(temp[8*6+i], temp[8*7+i]); - + BUTTERFLY1(temp[8*0+i], temp[8*2+i]); BUTTERFLY1(temp[8*1+i], temp[8*3+i]); BUTTERFLY1(temp[8*4+i], temp[8*6+i]); BUTTERFLY1(temp[8*5+i], temp[8*7+i]); - - sum += + + sum += BUTTERFLYA(temp[8*0+i], temp[8*4+i]) +BUTTERFLYA(temp[8*1+i], temp[8*5+i]) +BUTTERFLYA(temp[8*2+i], temp[8*6+i]) +BUTTERFLYA(temp[8*3+i], temp[8*7+i]); } - + sum -= ABS(temp[8*0] + temp[8*4]); // -mean - + return sum; } @@ -3329,7 +3329,7 @@ static int dct_sad8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 uint64_t __align8 aligned_temp[sizeof(DCTELEM)*64/8]; DCTELEM * const temp= (DCTELEM*)aligned_temp; int sum=0, i; - + assert(h==8); s->dsp.diff_pixels(temp, src1, src2, stride); @@ -3337,7 +3337,7 @@ static int dct_sad8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 for(i=0; i<64; i++) sum+= ABS(temp[i]); - + return sum; } @@ -3346,7 +3346,7 @@ static int dct_max8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 uint64_t __align8 aligned_temp[sizeof(DCTELEM)*64/8]; DCTELEM * const temp= (DCTELEM*)aligned_temp; int sum=0, i; - + assert(h==8); s->dsp.diff_pixels(temp, src1, src2, stride); @@ -3354,7 +3354,7 @@ static int dct_max8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 for(i=0; i<64; i++) sum= FFMAX(sum, ABS(temp[i])); - + return sum; } @@ -3369,18 +3369,18 @@ static int quant_psnr8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *s assert(h==8); s->mb_intra=0; - + s->dsp.diff_pixels(temp, src1, src2, stride); - + memcpy(bak, temp, 64*sizeof(DCTELEM)); - + s->block_last_index[0/*FIXME*/]= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i); s->dct_unquantize_inter(s, temp, 0, s->qscale); - simple_idct(temp); //FIXME - + simple_idct(temp); //FIXME + for(i=0; i<64; i++) sum+= (temp[i]-bak[i])*(temp[i]-bak[i]); - + return sum; } @@ -3395,7 +3395,7 @@ static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int const int esc_length= s->ac_esc_length; uint8_t * length; uint8_t * last_length; - + assert(h==8); for(i=0; i<8; i++){ @@ -3408,9 +3408,9 @@ static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int s->block_last_index[0/*FIXME*/]= last= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i); bits=0; - + if (s->mb_intra) { - start_i = 1; + start_i = 1; length = s->intra_ac_vlc_length; last_length= s->intra_ac_vlc_last_length; bits+= s->luma_dc_vlc_length[temp[0] + 256]; //FIXME chroma @@ -3419,13 +3419,13 @@ static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int length = s->inter_ac_vlc_length; last_length= s->inter_ac_vlc_last_length; } - + if(last>=start_i){ run=0; for(i=start_i; i<last; i++){ int j= scantable[i]; level= temp[j]; - + if(level){ level+=64; if((level&(~127)) == 0){ @@ -3437,16 +3437,16 @@ static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int run++; } i= scantable[last]; - + level= temp[i] + 64; assert(level - 64); - + if((level&(~127)) == 0){ bits+= last_length[UNI_AC_ENC_INDEX(run, level)]; }else bits+= esc_length; - + } if(last>=0){ @@ -3455,9 +3455,9 @@ static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int else s->dct_unquantize_inter(s, temp, 0, s->qscale); } - + s->dsp.idct_add(bak, stride, temp); - + distoration= s->dsp.sse[1](NULL, bak, src1, stride, 8); return distoration + ((bits*s->qscale*s->qscale*109 + 64)>>7); @@ -3474,15 +3474,15 @@ static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, in uint8_t * last_length; assert(h==8); - + s->dsp.diff_pixels(temp, src1, src2, stride); s->block_last_index[0/*FIXME*/]= last= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i); bits=0; - + if (s->mb_intra) { - start_i = 1; + start_i = 1; length = s->intra_ac_vlc_length; last_length= s->intra_ac_vlc_last_length; bits+= s->luma_dc_vlc_length[temp[0] + 256]; //FIXME chroma @@ -3491,13 +3491,13 @@ static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, in length = s->inter_ac_vlc_length; last_length= s->inter_ac_vlc_last_length; } - + if(last>=start_i){ run=0; for(i=start_i; i<last; i++){ int j= scantable[i]; level= temp[j]; - + if(level){ level+=64; if((level&(~127)) == 0){ @@ -3509,11 +3509,11 @@ static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, in run++; } i= scantable[last]; - + level= temp[i] + 64; - + assert(level - 64); - + if((level&(~127)) == 0){ bits+= last_length[UNI_AC_ENC_INDEX(run, level)]; }else @@ -3526,22 +3526,22 @@ static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, in static int vsad_intra16_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){ int score=0; int x,y; - + for(y=1; y<h; y++){ for(x=0; x<16; x+=4){ - score+= ABS(s[x ] - s[x +stride]) + ABS(s[x+1] - s[x+1+stride]) + score+= ABS(s[x ] - s[x +stride]) + ABS(s[x+1] - s[x+1+stride]) +ABS(s[x+2] - s[x+2+stride]) + ABS(s[x+3] - s[x+3+stride]); } s+= stride; } - + return score; } static int vsad16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int stride, int h){ int score=0; int x,y; - + for(y=1; y<h; y++){ for(x=0; x<16; x++){ score+= ABS(s1[x ] - s2[x ] - s1[x +stride] + s2[x +stride]); @@ -3549,7 +3549,7 @@ static int vsad16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int st s1+= stride; s2+= stride; } - + return score; } @@ -3557,22 +3557,22 @@ static int vsad16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int st static int vsse_intra16_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){ int score=0; int x,y; - + for(y=1; y<h; y++){ for(x=0; x<16; x+=4){ - score+= SQ(s[x ] - s[x +stride]) + SQ(s[x+1] - s[x+1+stride]) + score+= SQ(s[x ] - s[x +stride]) + SQ(s[x+1] - s[x+1+stride]) +SQ(s[x+2] - s[x+2+stride]) + SQ(s[x+3] - s[x+3+stride]); } s+= stride; } - + return score; } static int vsse16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int stride, int h){ int score=0; int x,y; - + for(y=1; y<h; y++){ for(x=0; x<16; x++){ score+= SQ(s1[x ] - s2[x ] - s1[x +stride] + s2[x +stride]); @@ -3580,7 +3580,7 @@ static int vsse16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int st s1+= stride; s2+= stride; } - + return score; } @@ -3650,11 +3650,11 @@ void dsputil_static_init(void) cropTbl[i] = 0; cropTbl[i + MAX_NEG_CROP + 256] = 255; } - + for(i=0;i<512;i++) { squareTbl[i] = (i - 256) * (i - 256); } - + for(i=0; i<64; i++) inv_zigzag_direct16[ff_zigzag_direct[i]]= i+1; } @@ -3667,11 +3667,11 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) if(avctx->dct_algo==FF_DCT_FASTINT) { c->fdct = fdct_ifast; c->fdct248 = fdct_ifast248; - } + } else if(avctx->dct_algo==FF_DCT_FAAN) { c->fdct = ff_faandct; - c->fdct248 = ff_faandct248; - } + c->fdct248 = ff_faandct248; + } else { c->fdct = ff_jpeg_fdct_islow; //slow/accurate/default c->fdct248 = ff_fdct248_islow; @@ -3861,11 +3861,11 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) c->put_mspel_pixels_tab[5]= put_mspel8_mc12_c; c->put_mspel_pixels_tab[6]= put_mspel8_mc22_c; c->put_mspel_pixels_tab[7]= put_mspel8_mc32_c; - + #define SET_CMP_FUNC(name) \ c->name[0]= name ## 16_c;\ c->name[1]= name ## 8x8_c; - + SET_CMP_FUNC(hadamard8_diff) c->hadamard8_diff[4]= hadamard8_intra16_c; SET_CMP_FUNC(dct_sad) @@ -3900,12 +3900,12 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) c->h264_h_loop_filter_chroma= h264_h_loop_filter_chroma_c; c->h264_v_loop_filter_chroma_intra= h264_v_loop_filter_chroma_intra_c; c->h264_h_loop_filter_chroma_intra= h264_h_loop_filter_chroma_intra_c; - + c->h263_h_loop_filter= h263_h_loop_filter_c; c->h263_v_loop_filter= h263_v_loop_filter_c; - + c->h261_loop_filter= h261_loop_filter_c; - + c->try_8x8basis= try_8x8basis_c; c->add_8x8basis= add_8x8basis_c; |