LLVMFuzzerInitialize: 12| 2|extern "C" int LLVMFuzzerInitialize(int *argc, char ***argv) { 13| 2| memset(&glb, 0, sizeof(glb)); 14| 2| glb.version = XVID_VERSION; ------------------ | | 71| 2|#define XVID_VERSION XVID_MAKE_VERSION(1,4,-127) | | ------------------ | | | | 62| 2|#define XVID_MAKE_VERSION(a,b,c) ((((a)&0xff)<<16) | (((b)&0xff)<<8) | ((c)&0xff)) | | ------------------ ------------------ 15| 2| if ( xvid_global(nullptr, XVID_GBL_INIT, &glb, nullptr) ) { ------------------ | | 235| 2|#define XVID_GBL_INIT 0 /* initialize xvidcore; must be called before using xvid_decore, or xvid_encore) */ ------------------ | Branch (15:10): [True: 0, False: 2] ------------------ 16| 0| abort(); 17| 0| } 18| | 19| 2| return 0; 20| 2|} LLVMFuzzerTestOneInput: 22| 10.2k|extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) { 23| 10.2k| xvid_dec_stats_t stats; 24| 10.2k| xvid_dec_create_t ctx; 25| 10.2k| xvid_dec_frame_t frame; 26| | 27| 10.2k| uint32_t width = 0; 28| 10.2k| uint32_t height = 0; 29| 10.2k| int remaining = size; 30| | 31| 10.2k| uint8_t* out = nullptr; 32| 10.2k| uint8_t* dataCopy = (uint8_t*)calloc(1, size + 10240); 33| 10.2k| memcpy(dataCopy, data, size); 34| | 35| 10.2k| uint8_t* inptr = dataCopy; 36| | 37| 10.2k| { 38| 10.2k| memset(&ctx, 0, sizeof(ctx)); 39| | 40| 10.2k| ctx.version = XVID_VERSION; ------------------ | | 71| 10.2k|#define XVID_VERSION XVID_MAKE_VERSION(1,4,-127) | | ------------------ | | | | 62| 10.2k|#define XVID_MAKE_VERSION(a,b,c) ((((a)&0xff)<<16) | (((b)&0xff)<<8) | ((c)&0xff)) | | ------------------ ------------------ 41| 10.2k| ctx.width = width; 42| 10.2k| ctx.height = height; 43| 10.2k| } 44| | 45| 10.2k| if ( xvid_decore(nullptr, XVID_DEC_CREATE, &ctx, nullptr) ) { ------------------ | | 246| 10.2k|#define XVID_DEC_CREATE 0 /* create decore instance; return 0 on success */ ------------------ | Branch (45:10): [True: 0, False: 10.2k] ------------------ 46| 0| abort(); 47| 0| } 48| | 49| | 50| 10.2k| int loops = 0; 51| 69.9k| do { 52| 69.9k| { 53| 69.9k| memset(&stats, 0, sizeof(xvid_dec_stats_t)); 54| 69.9k| stats.version = XVID_VERSION; ------------------ | | 71| 69.9k|#define XVID_VERSION XVID_MAKE_VERSION(1,4,-127) | | ------------------ | | | | 62| 69.9k|#define XVID_MAKE_VERSION(a,b,c) ((((a)&0xff)<<16) | (((b)&0xff)<<8) | ((c)&0xff)) | | ------------------ ------------------ 55| 69.9k| } 56| | 57| 69.9k| { 58| 69.9k| memset(&frame, 0, sizeof(xvid_dec_frame_t)); 59| | 60| 69.9k| frame.version = XVID_VERSION; ------------------ | | 71| 69.9k|#define XVID_VERSION XVID_MAKE_VERSION(1,4,-127) | | ------------------ | | | | 62| 69.9k|#define XVID_MAKE_VERSION(a,b,c) ((((a)&0xff)<<16) | (((b)&0xff)<<8) | ((c)&0xff)) | | ------------------ ------------------ 61| 69.9k| frame.general = 0; 62| | 63| 69.9k| frame.bitstream = inptr; 64| 69.9k| frame.length = remaining; 65| | 66| 69.9k| frame.output.plane[0] = out; 67| 69.9k| frame.output.stride[0] = width * 3; 68| | 69| 69.9k| frame.output.csp = XVID_CSP_BGR; ------------------ | | 121| 69.9k|#define XVID_CSP_BGR (1<< 9) /* 24-bit bgr packed */ ------------------ 70| 69.9k| } 71| | 72| 69.9k| const int used_bytes = xvid_decore(ctx.handle, XVID_DEC_DECODE, &frame, &stats); ------------------ | | 248| 69.9k|#define XVID_DEC_DECODE 2 /* decode a frame: returns number of bytes consumed >= 0 */ ------------------ 73| 69.9k| if ( stats.type == XVID_TYPE_VOL) { ------------------ | | 165| 69.9k|#define XVID_TYPE_VOL -1 /* decoder only: vol was decoded */ ------------------ | Branch (73:14): [True: 34.5k, False: 35.4k] ------------------ 74| | /* Resize buffer */ 75| | 76| 34.5k| if ( (width != stats.data.vol.width) || (height != stats.data.vol.height) ) { ------------------ | Branch (76:18): [True: 12.4k, False: 22.1k] | Branch (76:53): [True: 528, False: 21.6k] ------------------ 77| 12.9k| if ( width * height < stats.data.vol.width * stats.data.vol.height ) { ------------------ | Branch (77:22): [True: 10.6k, False: 2.29k] ------------------ 78| 10.6k| if (out) { ------------------ | Branch (78:25): [True: 2.66k, False: 7.98k] ------------------ 79| 2.66k| free(out); 80| 2.66k| } 81| 10.6k| out = (uint8_t*)malloc(stats.data.vol.width * stats.data.vol.height * 4); 82| 10.6k| } 83| 12.9k| width = stats.data.vol.width; 84| 12.9k| height = stats.data.vol.height; 85| 12.9k| } 86| 34.5k| } 87| | 88| 69.9k| if ( used_bytes > 0 ) { ------------------ | Branch (88:14): [True: 69.2k, False: 717] ------------------ 89| 69.2k| inptr += used_bytes; 90| 69.2k| remaining -= used_bytes; 91| 69.2k| } else { 92| 717| break; 93| 717| } 94| | 95| 69.2k| loops++; 96| 69.2k| } while (stats.type <= 0 && remaining > 1); ------------------ | Branch (96:14): [True: 63.1k, False: 6.04k] | Branch (96:33): [True: 59.6k, False: 3.53k] ------------------ 97| | 98| 10.2k|end: 99| 10.2k| free(dataCopy); 100| 10.2k| free(out); 101| | 102| 10.2k| xvid_decore(ctx.handle, XVID_DEC_DESTROY, nullptr, nullptr); ------------------ | | 247| 10.2k|#define XVID_DEC_DESTROY 1 /* destroy decore instance: return 0 on success */ ------------------ 103| 10.2k| return 0; 104| 10.2k|} read_video_packet_header: 105| 28.2k|{ 106| 28.2k| int startcode_bits = NUMBITS_VP_RESYNC_MARKER + addbits; ------------------ | | 111| 28.2k|#define NUMBITS_VP_RESYNC_MARKER 17 ------------------ 107| 28.2k| int mbnum_bits = log2bin(dec->mb_width * dec->mb_height - 1); 108| 28.2k| int mbnum; 109| 28.2k| int hec = 0; 110| | 111| 28.2k| BitstreamSkip(bs, BitstreamNumBitsToByteAlign(bs)); 112| 28.2k| BitstreamSkip(bs, startcode_bits); 113| | 114| 28.2k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 28.2k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 115| | 116| 28.2k| if (dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) ------------------ | | 87| 28.2k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (116:6): [True: 14.7k, False: 13.5k] ------------------ 117| 14.7k| { 118| 14.7k| hec = BitstreamGetBit(bs); /* header_extension_code */ 119| 14.7k| if (hec && !(dec->sprite_enable == SPRITE_STATIC /* && current_coding_type = I_VOP */)) ------------------ | | 94| 11.7k|#define SPRITE_STATIC 1 ------------------ | Branch (119:7): [True: 11.7k, False: 2.95k] | Branch (119:14): [True: 8.69k, False: 3.07k] ------------------ 120| 8.69k| { 121| 8.69k| BitstreamSkip(bs, 13); /* vop_width */ 122| 8.69k| READ_MARKER(); ------------------ | | 99| 8.69k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 123| 8.69k| BitstreamSkip(bs, 13); /* vop_height */ 124| 8.69k| READ_MARKER(); ------------------ | | 99| 8.69k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 125| 8.69k| BitstreamSkip(bs, 13); /* vop_horizontal_mc_spatial_ref */ 126| 8.69k| READ_MARKER(); ------------------ | | 99| 8.69k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 127| 8.69k| BitstreamSkip(bs, 13); /* vop_vertical_mc_spatial_ref */ 128| 8.69k| READ_MARKER(); ------------------ | | 99| 8.69k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 129| 8.69k| } 130| 14.7k| } 131| | 132| 28.2k| mbnum = (mbnum_bits == 0) ? 0 : BitstreamGetBits(bs, mbnum_bits); /* macroblock_number */ ------------------ | Branch (132:10): [True: 9, False: 28.2k] ------------------ 133| 28.2k| DPRINTF(XVID_DEBUG_HEADER, "mbnum %i\n", mbnum); ------------------ | | 198| 28.2k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 134| | 135| 28.2k| if (dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) ------------------ | | 89| 28.2k|#define VIDOBJLAY_SHAPE_BINARY_ONLY 2 ------------------ | Branch (135:6): [True: 14.6k, False: 13.5k] ------------------ 136| 14.6k| { 137| 14.6k| *quant = BitstreamGetBits(bs, dec->quant_bits); /* quant_scale */ 138| 14.6k| DPRINTF(XVID_DEBUG_HEADER, "quant %i\n", *quant); ------------------ | | 198| 14.6k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 139| 14.6k| } 140| | 141| 28.2k| if (dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR) ------------------ | | 87| 28.2k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (141:6): [True: 13.5k, False: 14.7k] ------------------ 142| 13.5k| hec = BitstreamGetBit(bs); /* header_extension_code */ 143| | 144| | 145| 28.2k| DPRINTF(XVID_DEBUG_HEADER, "header_extension_code %i\n", hec); ------------------ | | 198| 28.2k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 146| 28.2k| if (hec) ------------------ | Branch (146:6): [True: 18.9k, False: 9.30k] ------------------ 147| 18.9k| { 148| 18.9k| int time_base; 149| 18.9k| int time_increment; 150| 18.9k| int coding_type; 151| | 152| 39.4k| for (time_base=0; BitstreamGetBit(bs)!=0; time_base++); /* modulo_time_base */ ------------------ | Branch (152:21): [True: 20.4k, False: 18.9k] ------------------ 153| 18.9k| READ_MARKER(); ------------------ | | 99| 18.9k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 154| 18.9k| if (dec->time_inc_bits) ------------------ | Branch (154:7): [True: 18.9k, False: 0] ------------------ 155| 18.9k| time_increment = (BitstreamGetBits(bs, dec->time_inc_bits)); /* vop_time_increment */ 156| 0| else 157| 0| time_increment = 0; 158| 18.9k| READ_MARKER(); ------------------ | | 99| 18.9k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 159| 18.9k| DPRINTF(XVID_DEBUG_HEADER,"time %i:%i\n", time_base, time_increment); ------------------ | | 198| 18.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 160| | 161| 18.9k| coding_type = BitstreamGetBits(bs, 2); 162| 18.9k| DPRINTF(XVID_DEBUG_HEADER,"coding_type %i\n", coding_type); ------------------ | | 198| 18.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 163| | 164| 18.9k| if (dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) ------------------ | | 87| 18.9k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (164:7): [True: 11.7k, False: 7.18k] ------------------ 165| 11.7k| { 166| 11.7k| BitstreamSkip(bs, 1); /* change_conv_ratio_disable */ 167| 11.7k| if (coding_type != I_VOP) ------------------ | | 104| 11.7k|#define I_VOP 0 ------------------ | Branch (167:8): [True: 3.68k, False: 8.08k] ------------------ 168| 3.68k| BitstreamSkip(bs, 1); /* vop_shape_coding_type */ 169| 11.7k| } 170| | 171| 18.9k| if (dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) ------------------ | | 89| 18.9k|#define VIDOBJLAY_SHAPE_BINARY_ONLY 2 ------------------ | Branch (171:7): [True: 8.13k, False: 10.8k] ------------------ 172| 8.13k| { 173| 8.13k| *intra_dc_threshold = intra_dc_threshold_table[BitstreamGetBits(bs, 3)]; 174| | 175| 8.13k| if (dec->sprite_enable == SPRITE_GMC && coding_type == S_VOP && ------------------ | | 95| 16.2k|#define SPRITE_GMC 2 ------------------ if (dec->sprite_enable == SPRITE_GMC && coding_type == S_VOP && ------------------ | | 107| 9.21k|#define S_VOP 3 ------------------ | Branch (175:8): [True: 1.08k, False: 7.04k] | Branch (175:44): [True: 143, False: 941] ------------------ 176| 8.13k| dec->sprite_warping_points > 0) ------------------ | Branch (176:5): [True: 138, False: 5] ------------------ 177| 138| { 178| | /* TODO: sprite trajectory */ 179| 138| } 180| 8.13k| if (dec->reduced_resolution_enable && ------------------ | Branch (180:8): [True: 4.35k, False: 3.77k] ------------------ 181| 8.13k| dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR && ------------------ | | 87| 12.4k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (181:5): [True: 3.54k, False: 807] ------------------ 182| 8.13k| (coding_type == P_VOP || coding_type == I_VOP)) ------------------ | | 105| 7.09k|#define P_VOP 1 ------------------ (coding_type == P_VOP || coding_type == I_VOP)) ------------------ | | 104| 2.49k|#define I_VOP 0 ------------------ | Branch (182:6): [True: 1.05k, False: 2.49k] | Branch (182:30): [True: 1.04k, False: 1.44k] ------------------ 183| 2.09k| { 184| 2.09k| BitstreamSkip(bs, 1); /* XXX: vop_reduced_resolution */ 185| 2.09k| } 186| | 187| 8.13k| if (coding_type != I_VOP && fcode_forward) ------------------ | | 104| 16.2k|#define I_VOP 0 ------------------ | Branch (187:8): [True: 5.04k, False: 3.08k] | Branch (187:32): [True: 2.67k, False: 2.36k] ------------------ 188| 2.67k| { 189| 2.67k| *fcode_forward = BitstreamGetBits(bs, 3); 190| 2.67k| DPRINTF(XVID_DEBUG_HEADER,"fcode_forward %i\n", *fcode_forward); ------------------ | | 198| 2.67k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 191| 2.67k| } 192| | 193| 8.13k| if (coding_type == B_VOP && fcode_backward) ------------------ | | 106| 16.2k|#define B_VOP 2 ------------------ | Branch (193:8): [True: 2.88k, False: 5.25k] | Branch (193:32): [True: 1.60k, False: 1.27k] ------------------ 194| 1.60k| { 195| 1.60k| *fcode_backward = BitstreamGetBits(bs, 3); 196| 1.60k| DPRINTF(XVID_DEBUG_HEADER,"fcode_backward %i\n", *fcode_backward); ------------------ | | 198| 1.60k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 197| 1.60k| } 198| 8.13k| } 199| 18.9k| } 200| | 201| 28.2k| if (dec->newpred_enable) ------------------ | Branch (201:6): [True: 11.8k, False: 16.3k] ------------------ 202| 11.8k| { 203| 11.8k| int vop_id; 204| 11.8k| int vop_id_for_prediction; 205| | 206| 11.8k| vop_id = BitstreamGetBits(bs, MIN(dec->time_inc_bits + 3, 15)); ------------------ | | 255| 11.8k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 11.2k, False: 680] | | ------------------ ------------------ 207| 11.8k| DPRINTF(XVID_DEBUG_HEADER, "vop_id %i\n", vop_id); ------------------ | | 198| 11.8k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 208| 11.8k| if (BitstreamGetBit(bs)) /* vop_id_for_prediction_indication */ ------------------ | Branch (208:7): [True: 5.85k, False: 6.02k] ------------------ 209| 5.85k| { 210| 5.85k| vop_id_for_prediction = BitstreamGetBits(bs, MIN(dec->time_inc_bits + 3, 15)); ------------------ | | 255| 5.85k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 5.25k, False: 598] | | ------------------ ------------------ 211| 5.85k| DPRINTF(XVID_DEBUG_HEADER, "vop_id_for_prediction %i\n", vop_id_for_prediction); ------------------ | | 198| 5.85k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 212| 5.85k| } 213| 11.8k| READ_MARKER(); ------------------ | | 99| 11.8k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 214| 11.8k| } 215| | 216| 28.2k| return mbnum; 217| 28.2k|} BitstreamReadHeaders: 397| 1.86M|{ 398| 1.86M| uint32_t vol_ver_id; 399| 1.86M| uint32_t coding_type; 400| 1.86M| uint32_t start_code; 401| 1.86M| uint32_t time_incr = 0; 402| 1.86M| int32_t time_increment = 0; 403| | 404| 22.8M| while ((BitstreamPos(bs) >> 3) + 4 <= bs->length) { ------------------ | Branch (404:9): [True: 22.8M, False: 3.06k] ------------------ 405| | 406| 22.8M| BitstreamByteAlign(bs); 407| 22.8M| start_code = BitstreamShowBits(bs, 32); 408| | 409| 22.8M| if (start_code == VISOBJSEQ_START_CODE) { ------------------ | | 42| 22.8M|#define VISOBJSEQ_START_CODE 0x000001b0 ------------------ | Branch (409:7): [True: 2.98k, False: 22.8M] ------------------ 410| | 411| 2.98k| int profile; 412| | 413| 2.98k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 2.98k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 414| | 415| 2.98k| BitstreamSkip(bs, 32); /* visual_object_sequence_start_code */ 416| 2.98k| profile = BitstreamGetBits(bs, 8); /* profile_and_level_indication */ 417| | 418| 2.98k| DPRINTF(XVID_DEBUG_HEADER, "profile_and_level_indication %i\n", profile); ------------------ | | 198| 2.98k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 419| | 420| 22.8M| } else if (start_code == VISOBJSEQ_STOP_CODE) { ------------------ | | 43| 22.8M|#define VISOBJSEQ_STOP_CODE 0x000001b1 /* ??? */ ------------------ | Branch (420:14): [True: 1.01k, False: 22.8M] ------------------ 421| | 422| 1.01k| BitstreamSkip(bs, 32); /* visual_object_sequence_stop_code */ 423| | 424| 1.01k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 1.01k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 425| | 426| 22.8M| } else if (start_code == VISOBJ_START_CODE) { ------------------ | | 47| 22.8M|#define VISOBJ_START_CODE 0x000001b5 ------------------ | Branch (426:14): [True: 8.49k, False: 22.8M] ------------------ 427| | 428| 8.49k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 8.49k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 429| | 430| 8.49k| BitstreamSkip(bs, 32); /* visual_object_start_code */ 431| 8.49k| if (BitstreamGetBit(bs)) /* is_visual_object_identified */ ------------------ | Branch (431:8): [True: 2.83k, False: 5.65k] ------------------ 432| 2.83k| { 433| 2.83k| dec->ver_id = BitstreamGetBits(bs, 4); /* visual_object_ver_id */ 434| 2.83k| DPRINTF(XVID_DEBUG_HEADER,"visobj_ver_id %i\n", dec->ver_id); ------------------ | | 198| 2.83k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 435| 2.83k| BitstreamSkip(bs, 3); /* visual_object_priority */ 436| 5.65k| } else { 437| 5.65k| dec->ver_id = 1; 438| 5.65k| } 439| | 440| 8.49k| if (BitstreamShowBits(bs, 4) != VISOBJ_TYPE_VIDEO) /* visual_object_type */ ------------------ | | 52| 8.49k|#define VISOBJ_TYPE_VIDEO 1 ------------------ | Branch (440:8): [True: 4.39k, False: 4.10k] ------------------ 441| 4.39k| { 442| 4.39k| DPRINTF(XVID_DEBUG_ERROR, "visual_object_type != video\n"); ------------------ | | 196| 4.39k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 443| 4.39k| return -1; 444| 4.39k| } 445| 4.10k| BitstreamSkip(bs, 4); 446| | 447| | /* video_signal_type */ 448| | 449| 4.10k| if (BitstreamGetBit(bs)) /* video_signal_type */ ------------------ | Branch (449:8): [True: 3.62k, False: 476] ------------------ 450| 3.62k| { 451| 3.62k| DPRINTF(XVID_DEBUG_HEADER,"+ video_signal_type\n"); ------------------ | | 198| 3.62k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 452| 3.62k| BitstreamSkip(bs, 3); /* video_format */ 453| 3.62k| BitstreamSkip(bs, 1); /* video_range */ 454| 3.62k| if (BitstreamGetBit(bs)) /* color_description */ ------------------ | Branch (454:9): [True: 2.77k, False: 850] ------------------ 455| 2.77k| { 456| 2.77k| DPRINTF(XVID_DEBUG_HEADER,"+ color_description"); ------------------ | | 198| 2.77k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 457| 2.77k| BitstreamSkip(bs, 8); /* color_primaries */ 458| 2.77k| BitstreamSkip(bs, 8); /* transfer_characteristics */ 459| 2.77k| BitstreamSkip(bs, 8); /* matrix_coefficients */ 460| 2.77k| } 461| 3.62k| } 462| 22.8M| } else if ((start_code & ~VIDOBJ_START_CODE_MASK) == VIDOBJ_START_CODE) { ------------------ | | 385| 22.8M|#define VIDOBJ_START_CODE_MASK 0x0000001f ------------------ } else if ((start_code & ~VIDOBJ_START_CODE_MASK) == VIDOBJ_START_CODE) { ------------------ | | 40| 22.8M|#define VIDOBJ_START_CODE 0x00000100 /* ..0x0000011f */ ------------------ | Branch (462:14): [True: 9.04k, False: 22.8M] ------------------ 463| | 464| 9.04k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 9.04k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 465| 9.04k| DPRINTF(XVID_DEBUG_HEADER, "vo id %i\n", start_code & VIDOBJ_START_CODE_MASK); ------------------ | | 198| 9.04k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ DPRINTF(XVID_DEBUG_HEADER, "vo id %i\n", start_code & VIDOBJ_START_CODE_MASK); ------------------ | | 385| 9.04k|#define VIDOBJ_START_CODE_MASK 0x0000001f ------------------ 466| | 467| 9.04k| BitstreamSkip(bs, 32); /* video_object_start_code */ 468| | 469| 22.8M| } else if ((start_code & ~VIDOBJLAY_START_CODE_MASK) == VIDOBJLAY_START_CODE) { ------------------ | | 386| 22.8M|#define VIDOBJLAY_START_CODE_MASK 0x0000000f ------------------ } else if ((start_code & ~VIDOBJLAY_START_CODE_MASK) == VIDOBJLAY_START_CODE) { ------------------ | | 41| 22.8M|#define VIDOBJLAY_START_CODE 0x00000120 /* ..0x0000012f */ ------------------ | Branch (469:14): [True: 45.3k, False: 22.7M] ------------------ 470| 45.3k| uint32_t width = 0, height = 0; 471| | 472| 45.3k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 45.3k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 473| 45.3k| DPRINTF(XVID_DEBUG_HEADER, "vol id %i\n", start_code & VIDOBJLAY_START_CODE_MASK); ------------------ | | 198| 45.3k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ DPRINTF(XVID_DEBUG_HEADER, "vol id %i\n", start_code & VIDOBJLAY_START_CODE_MASK); ------------------ | | 386| 45.3k|#define VIDOBJLAY_START_CODE_MASK 0x0000000f ------------------ 474| | 475| 45.3k| BitstreamSkip(bs, 32); /* video_object_layer_start_code */ 476| 45.3k| BitstreamSkip(bs, 1); /* random_accessible_vol */ 477| | 478| 45.3k| BitstreamSkip(bs, 8); /* video_object_type_indication */ 479| | 480| 45.3k| if (BitstreamGetBit(bs)) /* is_object_layer_identifier */ ------------------ | Branch (480:8): [True: 14.9k, False: 30.4k] ------------------ 481| 14.9k| { 482| 14.9k| DPRINTF(XVID_DEBUG_HEADER, "+ is_object_layer_identifier\n"); ------------------ | | 198| 14.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 483| 14.9k| vol_ver_id = BitstreamGetBits(bs, 4); /* video_object_layer_verid */ 484| 14.9k| DPRINTF(XVID_DEBUG_HEADER,"ver_id %i\n", vol_ver_id); ------------------ | | 198| 14.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 485| 14.9k| BitstreamSkip(bs, 3); /* video_object_layer_priority */ 486| 30.4k| } else { 487| 30.4k| vol_ver_id = dec->ver_id; 488| 30.4k| } 489| | 490| 45.3k| dec->aspect_ratio = BitstreamGetBits(bs, 4); 491| | 492| 45.3k| if (dec->aspect_ratio == VIDOBJLAY_AR_EXTPAR) /* aspect_ratio_info */ ------------------ | | 84| 45.3k|#define VIDOBJLAY_AR_EXTPAR 15 ------------------ | Branch (492:8): [True: 1.92k, False: 43.4k] ------------------ 493| 1.92k| { 494| 1.92k| DPRINTF(XVID_DEBUG_HEADER, "+ aspect_ratio_info\n"); ------------------ | | 198| 1.92k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 495| 1.92k| dec->par_width = BitstreamGetBits(bs, 8); /* par_width */ 496| 1.92k| dec->par_height = BitstreamGetBits(bs, 8); /* par_height */ 497| 1.92k| } 498| | 499| 45.3k| if (BitstreamGetBit(bs)) /* vol_control_parameters */ ------------------ | Branch (499:8): [True: 12.6k, False: 32.7k] ------------------ 500| 12.6k| { 501| 12.6k| DPRINTF(XVID_DEBUG_HEADER, "+ vol_control_parameters\n"); ------------------ | | 198| 12.6k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 502| 12.6k| BitstreamSkip(bs, 2); /* chroma_format */ 503| 12.6k| dec->low_delay = BitstreamGetBit(bs); /* low_delay */ 504| 12.6k| DPRINTF(XVID_DEBUG_HEADER, "low_delay %i\n", dec->low_delay); ------------------ | | 198| 12.6k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 505| 12.6k| if (BitstreamGetBit(bs)) /* vbv_parameters */ ------------------ | Branch (505:9): [True: 1.11k, False: 11.5k] ------------------ 506| 1.11k| { 507| 1.11k| unsigned int bitrate; 508| 1.11k| unsigned int buffer_size; 509| 1.11k| unsigned int occupancy; 510| | 511| 1.11k| DPRINTF(XVID_DEBUG_HEADER,"+ vbv_parameters\n"); ------------------ | | 198| 1.11k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 512| | 513| 1.11k| bitrate = BitstreamGetBits(bs,15) << 15; /* first_half_bit_rate */ 514| 1.11k| READ_MARKER(); ------------------ | | 99| 1.11k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 515| 1.11k| bitrate |= BitstreamGetBits(bs,15); /* latter_half_bit_rate */ 516| 1.11k| READ_MARKER(); ------------------ | | 99| 1.11k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 517| | 518| 1.11k| buffer_size = BitstreamGetBits(bs, 15) << 3; /* first_half_vbv_buffer_size */ 519| 1.11k| READ_MARKER(); ------------------ | | 99| 1.11k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 520| 1.11k| buffer_size |= BitstreamGetBits(bs, 3); /* latter_half_vbv_buffer_size */ 521| | 522| 1.11k| occupancy = BitstreamGetBits(bs, 11) << 15; /* first_half_vbv_occupancy */ 523| 1.11k| READ_MARKER(); ------------------ | | 99| 1.11k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 524| 1.11k| occupancy |= BitstreamGetBits(bs, 15); /* latter_half_vbv_occupancy */ 525| 1.11k| READ_MARKER(); ------------------ | | 99| 1.11k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 526| | 527| 1.11k| DPRINTF(XVID_DEBUG_HEADER,"bitrate %d (unit=400 bps)\n", bitrate); ------------------ | | 198| 1.11k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 528| 1.11k| DPRINTF(XVID_DEBUG_HEADER,"buffer_size %d (unit=16384 bits)\n", buffer_size); ------------------ | | 198| 1.11k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 529| 1.11k| DPRINTF(XVID_DEBUG_HEADER,"occupancy %d (unit=64 bits)\n", occupancy); ------------------ | | 198| 1.11k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 530| 1.11k| } 531| 32.7k| }else{ 532| 32.7k| dec->low_delay = dec->low_delay_default; 533| 32.7k| } 534| | 535| 45.3k| dec->shape = BitstreamGetBits(bs, 2); /* video_object_layer_shape */ 536| | 537| 45.3k| DPRINTF(XVID_DEBUG_HEADER, "shape %i\n", dec->shape); ------------------ | | 198| 45.3k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 538| 45.3k| if (dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) ------------------ | | 87| 45.3k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (538:8): [True: 17.5k, False: 27.8k] ------------------ 539| 17.5k| { 540| 17.5k| DPRINTF(XVID_DEBUG_ERROR,"non-rectangular shapes are not supported\n"); ------------------ | | 196| 17.5k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 541| 17.5k| } 542| | 543| 45.3k| if (dec->shape == VIDOBJLAY_SHAPE_GRAYSCALE && vol_ver_id != 1) { ------------------ | | 90| 90.7k|#define VIDOBJLAY_SHAPE_GRAYSCALE 3 ------------------ | Branch (543:8): [True: 5.87k, False: 39.5k] | Branch (543:51): [True: 1.29k, False: 4.57k] ------------------ 544| 1.29k| BitstreamSkip(bs, 4); /* video_object_layer_shape_extension */ 545| 1.29k| } 546| | 547| 45.3k| READ_MARKER(); ------------------ | | 99| 45.3k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 548| | 549| | /********************** for decode B-frame time ***********************/ 550| 45.3k| dec->time_inc_resolution = BitstreamGetBits(bs, 16); /* vop_time_increment_resolution */ 551| 45.3k| DPRINTF(XVID_DEBUG_HEADER,"vop_time_increment_resolution %i\n", dec->time_inc_resolution); ------------------ | | 198| 45.3k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 552| | 553| 45.3k| if (dec->time_inc_resolution > 0) { ------------------ | Branch (553:8): [True: 32.1k, False: 13.2k] ------------------ 554| 32.1k| dec->time_inc_bits = MAX(log2bin(dec->time_inc_resolution-1), 1); ------------------ | | 258| 32.1k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (258:20): [True: 25.2k, False: 6.90k] | | ------------------ ------------------ 555| 32.1k| } else { 556| | /* for "old" xvid compatibility, set time_inc_bits = 1 */ 557| 13.2k| dec->time_inc_bits = 1; 558| 13.2k| } 559| | 560| 45.3k| READ_MARKER(); ------------------ | | 99| 45.3k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 561| | 562| 45.3k| if (BitstreamGetBit(bs)) /* fixed_vop_rate */ ------------------ | Branch (562:8): [True: 10.2k, False: 35.1k] ------------------ 563| 10.2k| { 564| 10.2k| DPRINTF(XVID_DEBUG_HEADER, "+ fixed_vop_rate\n"); ------------------ | | 198| 10.2k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 565| 10.2k| BitstreamSkip(bs, dec->time_inc_bits); /* fixed_vop_time_increment */ 566| 10.2k| } 567| | 568| 45.3k| if (dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) { ------------------ | | 89| 45.3k|#define VIDOBJLAY_SHAPE_BINARY_ONLY 2 ------------------ | Branch (568:8): [True: 41.4k, False: 3.95k] ------------------ 569| | 570| 41.4k| if (dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR) { ------------------ | | 87| 41.4k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (570:9): [True: 27.8k, False: 13.5k] ------------------ 571| 27.8k| READ_MARKER(); ------------------ | | 99| 27.8k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 572| 27.8k| width = BitstreamGetBits(bs, 13); /* video_object_layer_width */ 573| 27.8k| READ_MARKER(); ------------------ | | 99| 27.8k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 574| 27.8k| height = BitstreamGetBits(bs, 13); /* video_object_layer_height */ 575| 27.8k| READ_MARKER(); ------------------ | | 99| 27.8k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 576| | 577| 27.8k| DPRINTF(XVID_DEBUG_HEADER, "width %i\n", width); ------------------ | | 198| 27.8k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 578| 27.8k| DPRINTF(XVID_DEBUG_HEADER, "height %i\n", height); ------------------ | | 198| 27.8k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 579| 27.8k| } 580| | 581| 41.4k| dec->interlacing = BitstreamGetBit(bs); 582| 41.4k| DPRINTF(XVID_DEBUG_HEADER, "interlacing %i\n", dec->interlacing); ------------------ | | 198| 41.4k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 583| | 584| 41.4k| if (!BitstreamGetBit(bs)) /* obmc_disable */ ------------------ | Branch (584:9): [True: 28.1k, False: 13.2k] ------------------ 585| 28.1k| { 586| 28.1k| DPRINTF(XVID_DEBUG_ERROR, "obmc_disabled==false not supported\n"); ------------------ | | 196| 28.1k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 587| | /* TODO */ 588| | /* fucking divx4.02 has this enabled */ 589| 28.1k| } 590| | 591| 41.4k| dec->sprite_enable = BitstreamGetBits(bs, (vol_ver_id == 1 ? 1 : 2)); /* sprite_enable */ ------------------ | Branch (591:48): [True: 27.2k, False: 14.1k] ------------------ 592| | 593| 41.4k| if (dec->sprite_enable == SPRITE_STATIC || dec->sprite_enable == SPRITE_GMC) ------------------ | | 94| 82.8k|#define SPRITE_STATIC 1 ------------------ if (dec->sprite_enable == SPRITE_STATIC || dec->sprite_enable == SPRITE_GMC) ------------------ | | 95| 35.4k|#define SPRITE_GMC 2 ------------------ | Branch (593:9): [True: 6.00k, False: 35.4k] | Branch (593:48): [True: 4.76k, False: 30.6k] ------------------ 594| 10.7k| { 595| 10.7k| int low_latency_sprite_enable; 596| | 597| 10.7k| if (dec->sprite_enable != SPRITE_GMC) ------------------ | | 95| 10.7k|#define SPRITE_GMC 2 ------------------ | Branch (597:10): [True: 6.00k, False: 4.76k] ------------------ 598| 6.00k| { 599| 6.00k| int sprite_width; 600| 6.00k| int sprite_height; 601| 6.00k| int sprite_left_coord; 602| 6.00k| int sprite_top_coord; 603| 6.00k| sprite_width = BitstreamGetBits(bs, 13); /* sprite_width */ 604| 6.00k| READ_MARKER(); ------------------ | | 99| 6.00k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 605| 6.00k| sprite_height = BitstreamGetBits(bs, 13); /* sprite_height */ 606| 6.00k| READ_MARKER(); ------------------ | | 99| 6.00k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 607| 6.00k| sprite_left_coord = BitstreamGetBits(bs, 13); /* sprite_left_coordinate */ 608| 6.00k| READ_MARKER(); ------------------ | | 99| 6.00k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 609| 6.00k| sprite_top_coord = BitstreamGetBits(bs, 13); /* sprite_top_coordinate */ 610| 6.00k| READ_MARKER(); ------------------ | | 99| 6.00k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 611| 6.00k| } 612| 10.7k| dec->sprite_warping_points = BitstreamGetBits(bs, 6); /* no_of_sprite_warping_points */ 613| 10.7k| dec->sprite_warping_accuracy = BitstreamGetBits(bs, 2); /* sprite_warping_accuracy */ 614| 10.7k| dec->sprite_brightness_change = BitstreamGetBits(bs, 1); /* brightness_change */ 615| 10.7k| if (dec->sprite_enable != SPRITE_GMC) ------------------ | | 95| 10.7k|#define SPRITE_GMC 2 ------------------ | Branch (615:10): [True: 6.00k, False: 4.76k] ------------------ 616| 6.00k| { 617| 6.00k| low_latency_sprite_enable = BitstreamGetBits(bs, 1); /* low_latency_sprite_enable */ 618| 6.00k| } 619| 10.7k| } 620| | 621| 41.4k| if (vol_ver_id != 1 && ------------------ | Branch (621:9): [True: 14.1k, False: 27.2k] ------------------ 622| 41.4k| dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) { ------------------ | | 87| 14.1k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (622:6): [True: 2.83k, False: 11.3k] ------------------ 623| 2.83k| BitstreamSkip(bs, 1); /* sadct_disable */ 624| 2.83k| } 625| | 626| 41.4k| if (BitstreamGetBit(bs)) /* not_8_bit */ ------------------ | Branch (626:9): [True: 11.9k, False: 29.4k] ------------------ 627| 11.9k| { 628| 11.9k| DPRINTF(XVID_DEBUG_HEADER, "not_8_bit==true (ignored)\n"); ------------------ | | 198| 11.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 629| 11.9k| dec->quant_bits = BitstreamGetBits(bs, 4); /* quant_precision */ 630| 11.9k| BitstreamSkip(bs, 4); /* bits_per_pixel */ 631| 29.4k| } else { 632| 29.4k| dec->quant_bits = 5; 633| 29.4k| } 634| | 635| 41.4k| if (dec->shape == VIDOBJLAY_SHAPE_GRAYSCALE) { ------------------ | | 90| 41.4k|#define VIDOBJLAY_SHAPE_GRAYSCALE 3 ------------------ | Branch (635:9): [True: 5.87k, False: 35.5k] ------------------ 636| 5.87k| BitstreamSkip(bs, 1); /* no_gray_quant_update */ 637| 5.87k| BitstreamSkip(bs, 1); /* composition_method */ 638| 5.87k| BitstreamSkip(bs, 1); /* linear_composition */ 639| 5.87k| } 640| | 641| 41.4k| dec->quant_type = BitstreamGetBit(bs); /* quant_type */ 642| 41.4k| DPRINTF(XVID_DEBUG_HEADER, "quant_type %i\n", dec->quant_type); ------------------ | | 198| 41.4k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 643| | 644| 41.4k| if (dec->quant_type) { ------------------ | Branch (644:9): [True: 12.0k, False: 29.3k] ------------------ 645| 12.0k| if (BitstreamGetBit(bs)) /* load_intra_quant_mat */ ------------------ | Branch (645:10): [True: 3.11k, False: 8.96k] ------------------ 646| 3.11k| { 647| 3.11k| uint8_t matrix[64]; 648| | 649| 3.11k| DPRINTF(XVID_DEBUG_HEADER, "load_intra_quant_mat\n"); ------------------ | | 198| 3.11k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 650| | 651| 3.11k| bs_get_matrix(bs, matrix); 652| 3.11k| set_intra_matrix(dec->mpeg_quant_matrices, matrix); 653| 3.11k| } else 654| 8.96k| set_intra_matrix(dec->mpeg_quant_matrices, get_default_intra_matrix()); 655| | 656| 12.0k| if (BitstreamGetBit(bs)) /* load_inter_quant_mat */ ------------------ | Branch (656:10): [True: 1.30k, False: 10.7k] ------------------ 657| 1.30k| { 658| 1.30k| uint8_t matrix[64]; 659| | 660| 1.30k| DPRINTF(XVID_DEBUG_HEADER, "load_inter_quant_mat\n"); ------------------ | | 198| 1.30k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 661| | 662| 1.30k| bs_get_matrix(bs, matrix); 663| 1.30k| set_inter_matrix(dec->mpeg_quant_matrices, matrix); 664| 1.30k| } else 665| 10.7k| set_inter_matrix(dec->mpeg_quant_matrices, get_default_inter_matrix()); 666| | 667| 12.0k| if (dec->shape == VIDOBJLAY_SHAPE_GRAYSCALE) { ------------------ | | 90| 12.0k|#define VIDOBJLAY_SHAPE_GRAYSCALE 3 ------------------ | Branch (667:10): [True: 2.37k, False: 9.70k] ------------------ 668| 2.37k| DPRINTF(XVID_DEBUG_ERROR, "greyscale matrix not supported\n"); ------------------ | | 196| 2.37k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 669| 2.37k| return -1; 670| 2.37k| } 671| | 672| 12.0k| } 673| | 674| | 675| 39.0k| if (vol_ver_id != 1) { ------------------ | Branch (675:9): [True: 13.9k, False: 25.1k] ------------------ 676| 13.9k| dec->quarterpel = BitstreamGetBit(bs); /* quarter_sample */ 677| 13.9k| DPRINTF(XVID_DEBUG_HEADER,"quarterpel %i\n", dec->quarterpel); ------------------ | | 198| 13.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 678| 13.9k| } 679| 25.1k| else 680| 25.1k| dec->quarterpel = 0; 681| | 682| | 683| 39.0k| dec->complexity_estimation_disable = BitstreamGetBit(bs); /* complexity estimation disable */ 684| 39.0k| if (!dec->complexity_estimation_disable) ------------------ | Branch (684:9): [True: 25.6k, False: 13.3k] ------------------ 685| 25.6k| { 686| 25.6k| read_vol_complexity_estimation_header(bs, dec); 687| 25.6k| } 688| | 689| 39.0k| BitstreamSkip(bs, 1); /* resync_marker_disable */ 690| | 691| 39.0k| if (BitstreamGetBit(bs)) /* data_partitioned */ ------------------ | Branch (691:9): [True: 11.7k, False: 27.2k] ------------------ 692| 11.7k| { 693| 11.7k| DPRINTF(XVID_DEBUG_ERROR, "data_partitioned not supported\n"); ------------------ | | 196| 11.7k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 694| 11.7k| BitstreamSkip(bs, 1); /* reversible_vlc */ 695| 11.7k| } 696| | 697| 39.0k| if (vol_ver_id != 1) { ------------------ | Branch (697:9): [True: 13.9k, False: 25.1k] ------------------ 698| 13.9k| dec->newpred_enable = BitstreamGetBit(bs); 699| 13.9k| if (dec->newpred_enable) /* newpred_enable */ ------------------ | Branch (699:10): [True: 4.02k, False: 9.91k] ------------------ 700| 4.02k| { 701| 4.02k| DPRINTF(XVID_DEBUG_HEADER, "+ newpred_enable\n"); ------------------ | | 198| 4.02k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 702| 4.02k| BitstreamSkip(bs, 2); /* requested_upstream_message_type */ 703| 4.02k| BitstreamSkip(bs, 1); /* newpred_segment_type */ 704| 4.02k| } 705| 13.9k| dec->reduced_resolution_enable = BitstreamGetBit(bs); /* reduced_resolution_vop_enable */ 706| 13.9k| DPRINTF(XVID_DEBUG_HEADER, "reduced_resolution_enable %i\n", dec->reduced_resolution_enable); ------------------ | | 198| 13.9k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 707| 13.9k| } 708| 25.1k| else 709| 25.1k| { 710| 25.1k| dec->newpred_enable = 0; 711| 25.1k| dec->reduced_resolution_enable = 0; 712| 25.1k| } 713| | 714| 39.0k| dec->scalability = BitstreamGetBit(bs); /* scalability */ 715| 39.0k| if (dec->scalability) ------------------ | Branch (715:9): [True: 8.14k, False: 30.9k] ------------------ 716| 8.14k| { 717| 8.14k| DPRINTF(XVID_DEBUG_ERROR, "scalability not supported\n"); ------------------ | | 196| 8.14k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 718| 8.14k| BitstreamSkip(bs, 1); /* hierarchy_type */ 719| 8.14k| BitstreamSkip(bs, 4); /* ref_layer_id */ 720| 8.14k| BitstreamSkip(bs, 1); /* ref_layer_sampling_direc */ 721| 8.14k| BitstreamSkip(bs, 5); /* hor_sampling_factor_n */ 722| 8.14k| BitstreamSkip(bs, 5); /* hor_sampling_factor_m */ 723| 8.14k| BitstreamSkip(bs, 5); /* vert_sampling_factor_n */ 724| 8.14k| BitstreamSkip(bs, 5); /* vert_sampling_factor_m */ 725| 8.14k| BitstreamSkip(bs, 1); /* enhancement_type */ 726| 8.14k| if(dec->shape == VIDOBJLAY_SHAPE_BINARY /* && hierarchy_type==0 */) { ------------------ | | 88| 8.14k|#define VIDOBJLAY_SHAPE_BINARY 1 ------------------ | Branch (726:9): [True: 1.66k, False: 6.48k] ------------------ 727| 1.66k| BitstreamSkip(bs, 1); /* use_ref_shape */ 728| 1.66k| BitstreamSkip(bs, 1); /* use_ref_texture */ 729| 1.66k| BitstreamSkip(bs, 5); /* shape_hor_sampling_factor_n */ 730| 1.66k| BitstreamSkip(bs, 5); /* shape_hor_sampling_factor_m */ 731| 1.66k| BitstreamSkip(bs, 5); /* shape_vert_sampling_factor_n */ 732| 1.66k| BitstreamSkip(bs, 5); /* shape_vert_sampling_factor_m */ 733| 1.66k| } 734| 8.14k| return -1; 735| 8.14k| } 736| 39.0k| } else /* dec->shape == BINARY_ONLY */ 737| 3.95k| { 738| 3.95k| if (vol_ver_id != 1) { ------------------ | Branch (738:9): [True: 2.02k, False: 1.93k] ------------------ 739| 2.02k| dec->scalability = BitstreamGetBit(bs); /* scalability */ 740| 2.02k| if (dec->scalability) ------------------ | Branch (740:10): [True: 333, False: 1.69k] ------------------ 741| 333| { 742| 333| DPRINTF(XVID_DEBUG_ERROR, "scalability not supported\n"); ------------------ | | 196| 333|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 743| 333| BitstreamSkip(bs, 4); /* ref_layer_id */ 744| 333| BitstreamSkip(bs, 5); /* hor_sampling_factor_n */ 745| 333| BitstreamSkip(bs, 5); /* hor_sampling_factor_m */ 746| 333| BitstreamSkip(bs, 5); /* vert_sampling_factor_n */ 747| 333| BitstreamSkip(bs, 5); /* vert_sampling_factor_m */ 748| 333| return -1; 749| 333| } 750| 2.02k| } 751| 3.62k| BitstreamSkip(bs, 1); /* resync_marker_disable */ 752| | 753| 3.62k| } 754| | 755| 34.5k| if (((width > 0) && (height > 0)) && (dec->width != width || dec->height != height)) ------------------ | Branch (755:9): [True: 16.4k, False: 18.1k] | Branch (755:24): [True: 14.8k, False: 1.58k] | Branch (755:42): [True: 12.4k, False: 2.44k] | Branch (755:65): [True: 528, False: 1.91k] ------------------ 756| 12.9k| { 757| 12.9k| if (dec->fixed_dimensions) ------------------ | Branch (757:9): [True: 0, False: 12.9k] ------------------ 758| 0| { 759| 0| DPRINTF(XVID_DEBUG_ERROR, "decoder width/height does not match bitstream\n"); ------------------ | | 196| 0|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 760| 0| return -1; 761| 0| } 762| 12.9k| dec->width = width; 763| 12.9k| dec->height = height; 764| 12.9k| return -3; 765| 12.9k| } 766| | 767| 21.6k| return -2; /* VOL */ 768| | 769| 22.7M| } else if (start_code == GRPOFVOP_START_CODE) { ------------------ | | 45| 22.7M|#define GRPOFVOP_START_CODE 0x000001b3 ------------------ | Branch (769:14): [True: 12.7k, False: 22.7M] ------------------ 770| | 771| 12.7k| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 12.7k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 772| | 773| 12.7k| BitstreamSkip(bs, 32); 774| 12.7k| { 775| 12.7k| int hours, minutes, seconds; 776| | 777| 12.7k| hours = BitstreamGetBits(bs, 5); 778| 12.7k| minutes = BitstreamGetBits(bs, 6); 779| 12.7k| READ_MARKER(); ------------------ | | 99| 12.7k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 780| 12.7k| seconds = BitstreamGetBits(bs, 6); 781| | 782| 12.7k| DPRINTF(XVID_DEBUG_HEADER, "time %ih%im%is\n", hours,minutes,seconds); ------------------ | | 198| 12.7k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 783| 12.7k| } 784| 12.7k| BitstreamSkip(bs, 1); /* closed_gov */ 785| 12.7k| BitstreamSkip(bs, 1); /* broken_link */ 786| | 787| 22.7M| } else if (start_code == VOP_START_CODE) { ------------------ | | 48| 22.7M|#define VOP_START_CODE 0x000001b6 ------------------ | Branch (787:14): [True: 1.81M, False: 20.9M] ------------------ 788| | 789| 1.81M| DPRINTF(XVID_DEBUG_STARTCODE, "\n"); ------------------ | | 197| 1.81M|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 790| | 791| 1.81M| BitstreamSkip(bs, 32); /* vop_start_code */ 792| | 793| 1.81M| coding_type = BitstreamGetBits(bs, 2); /* vop_coding_type */ 794| 1.81M| DPRINTF(XVID_DEBUG_HEADER, "coding_type %i\n", coding_type); ------------------ | | 198| 1.81M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 795| | 796| | /*********************** for decode B-frame time ***********************/ 797| 11.8M| while (BitstreamGetBit(bs) != 0) /* time_base */ ------------------ | Branch (797:11): [True: 10.0M, False: 1.81M] ------------------ 798| 10.0M| time_incr++; 799| | 800| 1.81M| READ_MARKER(); ------------------ | | 99| 1.81M|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 801| | 802| 1.81M| if (dec->time_inc_bits) { ------------------ | Branch (802:8): [True: 1.81M, False: 1.56k] ------------------ 803| 1.81M| time_increment = (BitstreamGetBits(bs, dec->time_inc_bits)); /* vop_time_increment */ 804| 1.81M| } 805| | 806| 1.81M| DPRINTF(XVID_DEBUG_HEADER, "time_base %i\n", time_incr); ------------------ | | 198| 1.81M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 807| 1.81M| DPRINTF(XVID_DEBUG_HEADER, "time_increment %i\n", time_increment); ------------------ | | 198| 1.81M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 808| | 809| 1.81M| DPRINTF(XVID_DEBUG_TIMECODE, "%c %i:%i\n", ------------------ | | 199| 1.81M|#define XVID_DEBUG_TIMECODE (1<< 3) ------------------ 810| 1.81M| coding_type == I_VOP ? 'I' : coding_type == P_VOP ? 'P' : coding_type == B_VOP ? 'B' : 'S', ------------------ | | 104| 1.81M|#define I_VOP 0 ------------------ coding_type == I_VOP ? 'I' : coding_type == P_VOP ? 'P' : coding_type == B_VOP ? 'B' : 'S', ------------------ | | 105| 1.75M|#define P_VOP 1 ------------------ coding_type == I_VOP ? 'I' : coding_type == P_VOP ? 'P' : coding_type == B_VOP ? 'B' : 'S', ------------------ | | 106| 1.68M|#define B_VOP 2 ------------------ | Branch (810:5): [True: 63.6k, False: 1.75M] | Branch (810:34): [True: 69.3k, False: 1.68M] | Branch (810:63): [True: 172k, False: 1.51M] ------------------ 811| 1.81M| time_incr, time_increment); 812| | 813| 1.81M| if (coding_type != B_VOP) { ------------------ | | 106| 1.81M|#define B_VOP 2 ------------------ | Branch (813:8): [True: 1.64M, False: 172k] ------------------ 814| 1.64M| dec->last_time_base = dec->time_base; 815| 1.64M| dec->time_base += time_incr; 816| 1.64M| dec->time = dec->time_base*dec->time_inc_resolution + time_increment; 817| 1.64M| dec->time_pp = (int32_t)(dec->time - dec->last_non_b_time); 818| 1.64M| dec->last_non_b_time = dec->time; 819| 1.64M| } else { 820| 172k| dec->time = (dec->last_time_base + time_incr)*dec->time_inc_resolution + time_increment; 821| 172k| dec->time_bp = dec->time_pp - (int32_t)(dec->last_non_b_time - dec->time); 822| 172k| } 823| 1.81M| if (dec->time_pp <= 0) dec->time_pp = 1; ------------------ | Branch (823:17): [True: 614k, False: 1.20M] ------------------ 824| 1.81M| DPRINTF(XVID_DEBUG_HEADER,"time_pp=%i\n", dec->time_pp); ------------------ | | 198| 1.81M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 825| 1.81M| DPRINTF(XVID_DEBUG_HEADER,"time_bp=%i\n", dec->time_bp); ------------------ | | 198| 1.81M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 826| | 827| 1.81M| READ_MARKER(); ------------------ | | 99| 1.81M|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 828| | 829| 1.81M| if (!BitstreamGetBit(bs)) /* vop_coded */ ------------------ | Branch (829:8): [True: 218k, False: 1.59M] ------------------ 830| 218k| { 831| 218k| DPRINTF(XVID_DEBUG_HEADER, "vop_coded==false\n"); ------------------ | | 198| 218k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 832| 218k| return N_VOP; ------------------ | | 108| 218k|#define N_VOP 4 ------------------ 833| 218k| } 834| | 835| 1.59M| if (dec->newpred_enable) ------------------ | Branch (835:8): [True: 616k, False: 980k] ------------------ 836| 616k| { 837| 616k| int vop_id; 838| 616k| int vop_id_for_prediction; 839| | 840| 616k| vop_id = BitstreamGetBits(bs, MIN(dec->time_inc_bits + 3, 15)); ------------------ | | 255| 616k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 488k, False: 127k] | | ------------------ ------------------ 841| 616k| DPRINTF(XVID_DEBUG_HEADER, "vop_id %i\n", vop_id); ------------------ | | 198| 616k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 842| 616k| if (BitstreamGetBit(bs)) /* vop_id_for_prediction_indication */ ------------------ | Branch (842:9): [True: 67.5k, False: 549k] ------------------ 843| 67.5k| { 844| 67.5k| vop_id_for_prediction = BitstreamGetBits(bs, MIN(dec->time_inc_bits + 3, 15)); ------------------ | | 255| 67.5k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 65.5k, False: 1.95k] | | ------------------ ------------------ 845| 67.5k| DPRINTF(XVID_DEBUG_HEADER, "vop_id_for_prediction %i\n", vop_id_for_prediction); ------------------ | | 198| 67.5k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 846| 67.5k| } 847| 616k| READ_MARKER(); ------------------ | | 99| 616k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 848| 616k| } 849| | 850| | 851| | 852| | /* fix a little bug by MinChen */ 853| 1.59M| if ((dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) && ------------------ | | 89| 1.59M|#define VIDOBJLAY_SHAPE_BINARY_ONLY 2 ------------------ | Branch (853:8): [True: 1.54M, False: 49.5k] ------------------ 854| 1.59M| ( (coding_type == P_VOP) || (coding_type == S_VOP && dec->sprite_enable == SPRITE_GMC) ) ) { ------------------ | | 105| 1.54M|#define P_VOP 1 ------------------ ( (coding_type == P_VOP) || (coding_type == S_VOP && dec->sprite_enable == SPRITE_GMC) ) ) { ------------------ | | 107| 3.00M|#define S_VOP 3 ------------------ ( (coding_type == P_VOP) || (coding_type == S_VOP && dec->sprite_enable == SPRITE_GMC) ) ) { ------------------ | | 95| 1.36M|#define SPRITE_GMC 2 ------------------ | Branch (854:7): [True: 45.1k, False: 1.50M] | Branch (854:34): [True: 1.36M, False: 137k] | Branch (854:58): [True: 416k, False: 948k] ------------------ 855| 461k| *rounding = BitstreamGetBit(bs); /* rounding_type */ 856| 461k| DPRINTF(XVID_DEBUG_HEADER, "rounding %i\n", *rounding); ------------------ | | 198| 461k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 857| 461k| } 858| | 859| 1.59M| if (dec->reduced_resolution_enable && ------------------ | Branch (859:8): [True: 691k, False: 906k] ------------------ 860| 1.59M| dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR && ------------------ | | 87| 2.28M|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (860:5): [True: 134k, False: 557k] ------------------ 861| 1.59M| (coding_type == P_VOP || coding_type == I_VOP)) { ------------------ | | 105| 268k|#define P_VOP 1 ------------------ (coding_type == P_VOP || coding_type == I_VOP)) { ------------------ | | 104| 128k|#define I_VOP 0 ------------------ | Branch (861:6): [True: 5.05k, False: 128k] | Branch (861:30): [True: 2.08k, False: 126k] ------------------ 862| | 863| 7.14k| if (BitstreamGetBit(bs)) { ------------------ | Branch (863:9): [True: 1.55k, False: 5.58k] ------------------ 864| 1.55k| DPRINTF(XVID_DEBUG_ERROR, "RRV not supported (anymore)\n"); ------------------ | | 196| 1.55k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 865| 1.55k| } 866| 7.14k| } 867| | 868| 1.59M| if (dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) { ------------------ | | 87| 1.59M|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (868:8): [True: 869k, False: 728k] ------------------ 869| 869k| if(!(dec->sprite_enable == SPRITE_STATIC && coding_type == I_VOP)) { ------------------ | | 94| 1.73M|#define SPRITE_STATIC 1 ------------------ if(!(dec->sprite_enable == SPRITE_STATIC && coding_type == I_VOP)) { ------------------ | | 104| 262k|#define I_VOP 0 ------------------ | Branch (869:10): [True: 262k, False: 607k] | Branch (869:49): [True: 5.17k, False: 256k] ------------------ 870| | 871| 864k| uint32_t width, height; 872| 864k| uint32_t horiz_mc_ref, vert_mc_ref; 873| | 874| 864k| width = BitstreamGetBits(bs, 13); 875| 864k| READ_MARKER(); ------------------ | | 99| 864k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 876| 864k| height = BitstreamGetBits(bs, 13); 877| 864k| READ_MARKER(); ------------------ | | 99| 864k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 878| 864k| horiz_mc_ref = BitstreamGetBits(bs, 13); 879| 864k| READ_MARKER(); ------------------ | | 99| 864k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 880| 864k| vert_mc_ref = BitstreamGetBits(bs, 13); 881| 864k| READ_MARKER(); ------------------ | | 99| 864k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 882| | 883| 864k| DPRINTF(XVID_DEBUG_HEADER, "width %i\n", width); ------------------ | | 198| 864k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 884| 864k| DPRINTF(XVID_DEBUG_HEADER, "height %i\n", height); ------------------ | | 198| 864k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 885| 864k| DPRINTF(XVID_DEBUG_HEADER, "horiz_mc_ref %i\n", horiz_mc_ref); ------------------ | | 198| 864k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 886| 864k| DPRINTF(XVID_DEBUG_HEADER, "vert_mc_ref %i\n", vert_mc_ref); ------------------ | | 198| 864k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 887| 864k| } 888| | 889| 869k| BitstreamSkip(bs, 1); /* change_conv_ratio_disable */ 890| 869k| if (BitstreamGetBit(bs)) /* vop_constant_alpha */ ------------------ | Branch (890:9): [True: 550k, False: 319k] ------------------ 891| 550k| { 892| 550k| BitstreamSkip(bs, 8); /* vop_constant_alpha_value */ 893| 550k| } 894| 869k| } 895| | 896| 1.59M| if (dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) { ------------------ | | 89| 1.59M|#define VIDOBJLAY_SHAPE_BINARY_ONLY 2 ------------------ | Branch (896:8): [True: 1.54M, False: 49.5k] ------------------ 897| | 898| 1.54M| if (!dec->complexity_estimation_disable) ------------------ | Branch (898:9): [True: 1.36M, False: 184k] ------------------ 899| 1.36M| { 900| 1.36M| read_vop_complexity_estimation_header(bs, dec, coding_type); 901| 1.36M| } 902| | 903| | /* intra_dc_vlc_threshold */ 904| 1.54M| *intra_dc_threshold = 905| 1.54M| intra_dc_threshold_table[BitstreamGetBits(bs, 3)]; 906| | 907| 1.54M| dec->top_field_first = 0; 908| 1.54M| dec->alternate_vertical_scan = 0; 909| | 910| 1.54M| if (dec->interlacing) { ------------------ | Branch (910:9): [True: 769k, False: 778k] ------------------ 911| 769k| dec->top_field_first = BitstreamGetBit(bs); 912| 769k| DPRINTF(XVID_DEBUG_HEADER, "interlace top_field_first %i\n", dec->top_field_first); ------------------ | | 198| 769k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 913| 769k| dec->alternate_vertical_scan = BitstreamGetBit(bs); 914| 769k| DPRINTF(XVID_DEBUG_HEADER, "interlace alternate_vertical_scan %i\n", dec->alternate_vertical_scan); ------------------ | | 198| 769k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 915| | 916| 769k| } 917| 1.54M| } 918| | 919| 1.59M| if ((dec->sprite_enable == SPRITE_STATIC || dec->sprite_enable== SPRITE_GMC) && coding_type == S_VOP) { ------------------ | | 94| 3.19M|#define SPRITE_STATIC 1 ------------------ if ((dec->sprite_enable == SPRITE_STATIC || dec->sprite_enable== SPRITE_GMC) && coding_type == S_VOP) { ------------------ | | 95| 1.00M|#define SPRITE_GMC 2 ------------------ if ((dec->sprite_enable == SPRITE_STATIC || dec->sprite_enable== SPRITE_GMC) && coding_type == S_VOP) { ------------------ | | 107| 1.03M|#define S_VOP 3 ------------------ | Branch (919:9): [True: 597k, False: 1.00M] | Branch (919:48): [True: 438k, False: 561k] | Branch (919:84): [True: 932k, False: 102k] ------------------ 920| | 921| 932k| int i; 922| | 923| 2.74M| for (i = 0 ; i < MIN(4, dec->sprite_warping_points); i++) ------------------ | | 255| 2.74M|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 2.19M, False: 542k] | | ------------------ ------------------ | Branch (923:18): [True: 1.80M, False: 932k] ------------------ 924| 1.80M| { 925| 1.80M| int length; 926| 1.80M| int x = 0, y = 0; 927| | 928| | /* sprite code borowed from ffmpeg; thx Michael Niedermayer */ 929| 1.80M| length = bs_get_spritetrajectory(bs); 930| 1.80M| if(length){ ------------------ | Branch (930:9): [True: 1.17M, False: 628k] ------------------ 931| 1.17M| x= BitstreamGetBits(bs, length); 932| 1.17M| if ((x >> (length - 1)) == 0) /* if MSB not set it is negative*/ ------------------ | Branch (932:11): [True: 878k, False: 299k] ------------------ 933| 878k| x = - (x ^ ((1 << length) - 1)); 934| 1.17M| } 935| 1.80M| READ_MARKER(); ------------------ | | 99| 1.80M|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 936| | 937| 1.80M| length = bs_get_spritetrajectory(bs); 938| 1.80M| if(length){ ------------------ | Branch (938:9): [True: 1.15M, False: 653k] ------------------ 939| 1.15M| y = BitstreamGetBits(bs, length); 940| 1.15M| if ((y >> (length - 1)) == 0) /* if MSB not set it is negative*/ ------------------ | Branch (940:11): [True: 876k, False: 277k] ------------------ 941| 876k| y = - (y ^ ((1 << length) - 1)); 942| 1.15M| } 943| 1.80M| READ_MARKER(); ------------------ | | 99| 1.80M|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 944| | 945| 1.80M| gmc_warp->duv[i].x = x; 946| 1.80M| gmc_warp->duv[i].y = y; 947| | 948| 1.80M| DPRINTF(XVID_DEBUG_HEADER,"sprite_warping_point[%i] xy=(%i,%i)\n", i, x, y); ------------------ | | 198| 1.80M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 949| 1.80M| } 950| | 951| 932k| if (dec->sprite_brightness_change) ------------------ | Branch (951:9): [True: 155k, False: 777k] ------------------ 952| 155k| { 953| | /* XXX: brightness_change_factor() */ 954| 155k| } 955| 932k| if (dec->sprite_enable == SPRITE_STATIC) ------------------ | | 94| 932k|#define SPRITE_STATIC 1 ------------------ | Branch (955:9): [True: 515k, False: 417k] ------------------ 956| 515k| { 957| | /* XXX: todo */ 958| 515k| } 959| | 960| 932k| } 961| | 962| 1.59M| if ((*quant = BitstreamGetBits(bs, dec->quant_bits)) < 1) /* vop_quant */ ------------------ | Branch (962:8): [True: 694k, False: 903k] ------------------ 963| 694k| *quant = 1; 964| 1.59M| DPRINTF(XVID_DEBUG_HEADER, "quant %i\n", *quant); ------------------ | | 198| 1.59M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 965| | 966| 1.59M| if (coding_type != I_VOP) { ------------------ | | 104| 1.59M|#define I_VOP 0 ------------------ | Branch (966:8): [True: 1.56M, False: 27.7k] ------------------ 967| 1.56M| *fcode_forward = BitstreamGetBits(bs, 3); /* fcode_forward */ 968| 1.56M| DPRINTF(XVID_DEBUG_HEADER, "fcode_forward %i\n", *fcode_forward); ------------------ | | 198| 1.56M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 969| 1.56M| } 970| | 971| 1.59M| if (coding_type == B_VOP) { ------------------ | | 106| 1.59M|#define B_VOP 2 ------------------ | Branch (971:8): [True: 112k, False: 1.48M] ------------------ 972| 112k| *fcode_backward = BitstreamGetBits(bs, 3); /* fcode_backward */ 973| 112k| DPRINTF(XVID_DEBUG_HEADER, "fcode_backward %i\n", *fcode_backward); ------------------ | | 198| 112k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 974| 112k| } 975| 1.59M| if (!dec->scalability) { ------------------ | Branch (975:8): [True: 509k, False: 1.08M] ------------------ 976| 509k| if ((dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) && ------------------ | | 87| 509k|#define VIDOBJLAY_SHAPE_RECTANGULAR 0 ------------------ | Branch (976:9): [True: 62.2k, False: 446k] ------------------ 977| 509k| (coding_type != I_VOP)) { ------------------ | | 104| 62.2k|#define I_VOP 0 ------------------ | Branch (977:6): [True: 61.1k, False: 1.08k] ------------------ 978| 61.1k| BitstreamSkip(bs, 1); /* vop_shape_coding_type */ 979| 61.1k| } 980| 509k| } 981| 1.59M| return coding_type; 982| | 983| 20.9M| } else if (start_code == USERDATA_START_CODE) { ------------------ | | 44| 20.9M|#define USERDATA_START_CODE 0x000001b2 ------------------ | Branch (983:14): [True: 7.75k, False: 20.9M] ------------------ 984| 7.75k| char tmp[256]; 985| 7.75k| int i, version = 0, build = 0; 986| 7.75k| char packed; 987| | 988| 7.75k| BitstreamSkip(bs, 32); /* user_data_start_code */ 989| | 990| 7.75k| memset(tmp, 0, 256); 991| 7.75k| tmp[0] = BitstreamShowBits(bs, 8); 992| | 993| 71.5k| for(i = 1; i < 255; i++){ ------------------ | Branch (993:15): [True: 71.4k, False: 71] ------------------ 994| 71.4k| tmp[i] = (BitstreamShowBits(bs, 16) & 0xFF); 995| | 996| 71.4k| if(tmp[i] == 0) ------------------ | Branch (996:8): [True: 7.68k, False: 63.7k] ------------------ 997| 7.68k| break; 998| | 999| 63.7k| BitstreamSkip(bs, 8); 1000| 63.7k| } 1001| | 1002| 7.75k| DPRINTF(XVID_DEBUG_STARTCODE, ": %s\n", tmp); ------------------ | | 197| 7.75k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 1003| | 1004| | /* read xvid bitstream version */ 1005| 7.75k| if(strncmp(tmp, "XviD", 4) == 0) { ------------------ | Branch (1005:7): [True: 2.98k, False: 4.76k] ------------------ 1006| 2.98k| if (tmp[strlen(tmp)-1] == 'C') { ------------------ | Branch (1006:9): [True: 286, False: 2.70k] ------------------ 1007| 286| sscanf(tmp, "XviD%dC", &dec->bs_version); 1008| 286| dec->cartoon_mode = 1; 1009| 286| } 1010| 2.70k| else 1011| 2.70k| sscanf(tmp, "XviD%d", &dec->bs_version); 1012| | 1013| 2.98k| DPRINTF(XVID_DEBUG_HEADER, "xvid bitstream version=%i\n", dec->bs_version); ------------------ | | 198| 2.98k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 1014| 2.98k| } 1015| | 1016| | /* divx detection */ 1017| 7.75k| i = sscanf(tmp, "DivX%dBuild%d%c", &version, &build, &packed); 1018| 7.75k| if (i < 2) ------------------ | Branch (1018:8): [True: 7.75k, False: 0] ------------------ 1019| 7.75k| i = sscanf(tmp, "DivX%db%d%c", &version, &build, &packed); 1020| | 1021| 7.75k| if (i >= 2) ------------------ | Branch (1021:8): [True: 2.53k, False: 5.21k] ------------------ 1022| 2.53k| { 1023| 2.53k| dec->packed_mode = (i == 3 && packed == 'p'); ------------------ | Branch (1023:25): [True: 613, False: 1.92k] | Branch (1023:35): [True: 35, False: 578] ------------------ 1024| 2.53k| DPRINTF(XVID_DEBUG_HEADER, "divx version=%i, build=%i packed=%i\n", ------------------ | | 198| 2.53k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 1025| 2.53k| version, build, dec->packed_mode); 1026| 2.53k| } 1027| | 1028| 7.75k| if ((dec->bs_version == 0) && (build > 0) && ------------------ | Branch (1028:8): [True: 4.46k, False: 3.29k] | Branch (1028:34): [True: 1.80k, False: 2.65k] ------------------ 1029| 7.75k| (build != 1393)) { /* non-xvid stream with xvid fourcc */ ------------------ | Branch (1029:5): [True: 1.72k, False: 81] ------------------ 1030| 1.72k| dec->bs_version = 0xffff; 1031| 1.72k| } 1032| | 1033| 7.75k| } else /* start_code == ? */ 1034| 20.9M| { 1035| 20.9M| if (BitstreamShowBits(bs, 24) == 0x000001) { ------------------ | Branch (1035:8): [True: 367k, False: 20.5M] ------------------ 1036| 367k| DPRINTF(XVID_DEBUG_STARTCODE, "\n", BitstreamShowBits(bs, 32)); ------------------ | | 197| 367k|#define XVID_DEBUG_STARTCODE (1<< 1) ------------------ 1037| 367k| } 1038| 20.9M| BitstreamSkip(bs, 8); 1039| 20.9M| } 1040| 22.8M| } 1041| | 1042| |#if 0 1043| | DPRINTF("*** WARNING: no vop_start_code found"); 1044| |#endif 1045| 3.06k| return -1; /* ignore it */ 1046| 1.86M|} bitstream.c:log2bin: 39| 85.6k|{ 40| 85.6k| int n = 0; 41| 85.6k| if (value & 0xffff0000) { ------------------ | Branch (41:7): [True: 436, False: 85.2k] ------------------ 42| 436| value >>= 16; 43| 436| n += 16; 44| 436| } 45| 85.6k| if (value & 0xff00) { ------------------ | Branch (45:7): [True: 56.5k, False: 29.1k] ------------------ 46| 56.5k| value >>= 8; 47| 56.5k| n += 8; 48| 56.5k| } 49| 85.6k| if (value & 0xf0) { ------------------ | Branch (49:7): [True: 48.7k, False: 36.9k] ------------------ 50| 48.7k| value >>= 4; 51| 48.7k| n += 4; 52| 48.7k| } 53| 85.6k| return n + log2_tab_16[value]; 54| 85.6k|} bitstream.c:bs_get_matrix: 71| 4.41k|{ 72| 4.41k| int i = 0; 73| 4.41k| int last, value = 0; 74| | 75| 42.5k| do { 76| 42.5k| last = value; 77| 42.5k| value = BitstreamGetBits(bs, 8); 78| 42.5k| matrix[scan_tables[0][i++]] = value; 79| 42.5k| } 80| 42.5k| while (value != 0 && i < 64); ------------------ | Branch (80:9): [True: 38.3k, False: 4.14k] | Branch (80:23): [True: 38.1k, False: 267] ------------------ 81| | 82| 4.41k| if (value != 0) return; ------------------ | Branch (82:6): [True: 267, False: 4.14k] ------------------ 83| | 84| 4.14k| i--; 85| 248k| while (i < 64) { ------------------ | Branch (85:9): [True: 244k, False: 4.14k] ------------------ 86| 244k| matrix[scan_tables[0][i++]] = last; 87| 244k| } 88| 4.14k|} bitstream.c:read_vol_complexity_estimation_header: 225| 25.6k|{ 226| 25.6k| ESTIMATION * e = &dec->estimation; 227| | 228| 25.6k| e->method = BitstreamGetBits(bs, 2); /* estimation_method */ 229| 25.6k| DPRINTF(XVID_DEBUG_HEADER,"+ complexity_estimation_header; method=%i\n", e->method); ------------------ | | 198| 25.6k|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 230| | 231| 25.6k| if (e->method == 0 || e->method == 1) ------------------ | Branch (231:6): [True: 17.0k, False: 8.67k] | Branch (231:24): [True: 5.27k, False: 3.39k] ------------------ 232| 22.2k| { 233| 22.2k| if (!BitstreamGetBit(bs)) /* shape_complexity_estimation_disable */ ------------------ | Branch (233:7): [True: 19.2k, False: 3.01k] ------------------ 234| 19.2k| { 235| 19.2k| e->opaque = BitstreamGetBit(bs); /* opaque */ 236| 19.2k| e->transparent = BitstreamGetBit(bs); /* transparent */ 237| 19.2k| e->intra_cae = BitstreamGetBit(bs); /* intra_cae */ 238| 19.2k| e->inter_cae = BitstreamGetBit(bs); /* inter_cae */ 239| 19.2k| e->no_update = BitstreamGetBit(bs); /* no_update */ 240| 19.2k| e->upsampling = BitstreamGetBit(bs); /* upsampling */ 241| 19.2k| } 242| | 243| 22.2k| if (!BitstreamGetBit(bs)) /* texture_complexity_estimation_set_1_disable */ ------------------ | Branch (243:7): [True: 15.1k, False: 7.17k] ------------------ 244| 15.1k| { 245| 15.1k| e->intra_blocks = BitstreamGetBit(bs); /* intra_blocks */ 246| 15.1k| e->inter_blocks = BitstreamGetBit(bs); /* inter_blocks */ 247| 15.1k| e->inter4v_blocks = BitstreamGetBit(bs); /* inter4v_blocks */ 248| 15.1k| e->not_coded_blocks = BitstreamGetBit(bs); /* not_coded_blocks */ 249| 15.1k| } 250| 22.2k| } 251| | 252| 25.6k| READ_MARKER(); ------------------ | | 99| 25.6k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 253| | 254| 25.6k| if (!BitstreamGetBit(bs)) /* texture_complexity_estimation_set_2_disable */ ------------------ | Branch (254:6): [True: 15.1k, False: 10.4k] ------------------ 255| 15.1k| { 256| 15.1k| e->dct_coefs = BitstreamGetBit(bs); /* dct_coefs */ 257| 15.1k| e->dct_lines = BitstreamGetBit(bs); /* dct_lines */ 258| 15.1k| e->vlc_symbols = BitstreamGetBit(bs); /* vlc_symbols */ 259| 15.1k| e->vlc_bits = BitstreamGetBit(bs); /* vlc_bits */ 260| 15.1k| } 261| | 262| 25.6k| if (!BitstreamGetBit(bs)) /* motion_compensation_complexity_disable */ ------------------ | Branch (262:6): [True: 17.2k, False: 8.38k] ------------------ 263| 17.2k| { 264| 17.2k| e->apm = BitstreamGetBit(bs); /* apm */ 265| 17.2k| e->npm = BitstreamGetBit(bs); /* npm */ 266| 17.2k| e->interpolate_mc_q = BitstreamGetBit(bs); /* interpolate_mc_q */ 267| 17.2k| e->forw_back_mc_q = BitstreamGetBit(bs); /* forw_back_mc_q */ 268| 17.2k| e->halfpel2 = BitstreamGetBit(bs); /* halfpel2 */ 269| 17.2k| e->halfpel4 = BitstreamGetBit(bs); /* halfpel4 */ 270| 17.2k| } 271| | 272| 25.6k| READ_MARKER(); ------------------ | | 99| 25.6k|#define READ_MARKER() BitstreamSkip(bs, 1) ------------------ 273| | 274| 25.6k| if (e->method == 1) ------------------ | Branch (274:6): [True: 5.27k, False: 20.4k] ------------------ 275| 5.27k| { 276| 5.27k| if (!BitstreamGetBit(bs)) /* version2_complexity_estimation_disable */ ------------------ | Branch (276:7): [True: 4.17k, False: 1.10k] ------------------ 277| 4.17k| { 278| 4.17k| e->sadct = BitstreamGetBit(bs); /* sadct */ 279| 4.17k| e->quarterpel = BitstreamGetBit(bs); /* quarterpel */ 280| 4.17k| } 281| 5.27k| } 282| 25.6k|} bitstream.c:read_vop_complexity_estimation_header: 288| 1.36M|{ 289| 1.36M| ESTIMATION * e = &dec->estimation; 290| | 291| 1.36M| if (e->method == 0 || e->method == 1) ------------------ | Branch (291:6): [True: 1.02M, False: 335k] | Branch (291:24): [True: 263k, False: 72.1k] ------------------ 292| 1.29M| { 293| 1.29M| if (coding_type == I_VOP) { ------------------ | | 104| 1.29M|#define I_VOP 0 ------------------ | Branch (293:7): [True: 18.6k, False: 1.27M] ------------------ 294| 18.6k| if (e->opaque) BitstreamSkip(bs, 8); /* dcecs_opaque */ ------------------ | Branch (294:8): [True: 4.09k, False: 14.5k] ------------------ 295| 18.6k| if (e->transparent) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (295:8): [True: 8.66k, False: 9.98k] ------------------ 296| 18.6k| if (e->intra_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (296:8): [True: 4.43k, False: 14.2k] ------------------ 297| 18.6k| if (e->inter_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (297:8): [True: 7.62k, False: 11.0k] ------------------ 298| 18.6k| if (e->no_update) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (298:8): [True: 5.22k, False: 13.4k] ------------------ 299| 18.6k| if (e->upsampling) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (299:8): [True: 3.65k, False: 14.9k] ------------------ 300| 18.6k| if (e->intra_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (300:8): [True: 4.65k, False: 13.9k] ------------------ 301| 18.6k| if (e->not_coded_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (301:8): [True: 6.20k, False: 12.4k] ------------------ 302| 18.6k| if (e->dct_coefs) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (302:8): [True: 4.17k, False: 14.4k] ------------------ 303| 18.6k| if (e->dct_lines) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (303:8): [True: 4.97k, False: 13.6k] ------------------ 304| 18.6k| if (e->vlc_symbols) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (304:8): [True: 6.79k, False: 11.8k] ------------------ 305| 18.6k| if (e->vlc_bits) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (305:8): [True: 3.99k, False: 14.6k] ------------------ 306| 18.6k| if (e->sadct) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (306:8): [True: 4.55k, False: 14.1k] ------------------ 307| 18.6k| } 308| | 309| 1.29M| if (coding_type == P_VOP) { ------------------ | | 105| 1.29M|#define P_VOP 1 ------------------ | Branch (309:7): [True: 35.1k, False: 1.25M] ------------------ 310| 35.1k| if (e->opaque) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (310:8): [True: 10.3k, False: 24.7k] ------------------ 311| 35.1k| if (e->transparent) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (311:8): [True: 7.63k, False: 27.5k] ------------------ 312| 35.1k| if (e->intra_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (312:8): [True: 11.5k, False: 23.6k] ------------------ 313| 35.1k| if (e->inter_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (313:8): [True: 9.23k, False: 25.9k] ------------------ 314| 35.1k| if (e->no_update) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (314:8): [True: 8.14k, False: 27.0k] ------------------ 315| 35.1k| if (e->upsampling) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (315:8): [True: 9.21k, False: 25.9k] ------------------ 316| 35.1k| if (e->intra_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (316:8): [True: 5.55k, False: 29.6k] ------------------ 317| 35.1k| if (e->not_coded_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (317:8): [True: 6.88k, False: 28.2k] ------------------ 318| 35.1k| if (e->dct_coefs) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (318:8): [True: 15.0k, False: 20.1k] ------------------ 319| 35.1k| if (e->dct_lines) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (319:8): [True: 8.63k, False: 26.5k] ------------------ 320| 35.1k| if (e->vlc_symbols) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (320:8): [True: 11.8k, False: 23.2k] ------------------ 321| 35.1k| if (e->vlc_bits) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (321:8): [True: 11.4k, False: 23.6k] ------------------ 322| 35.1k| if (e->inter_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (322:8): [True: 6.22k, False: 28.9k] ------------------ 323| 35.1k| if (e->inter4v_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (323:8): [True: 8.80k, False: 26.3k] ------------------ 324| 35.1k| if (e->apm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (324:8): [True: 15.6k, False: 19.5k] ------------------ 325| 35.1k| if (e->npm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (325:8): [True: 14.8k, False: 20.3k] ------------------ 326| 35.1k| if (e->forw_back_mc_q) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (326:8): [True: 19.3k, False: 15.8k] ------------------ 327| 35.1k| if (e->halfpel2) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (327:8): [True: 14.1k, False: 21.0k] ------------------ 328| 35.1k| if (e->halfpel4) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (328:8): [True: 10.5k, False: 24.5k] ------------------ 329| 35.1k| if (e->sadct) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (329:8): [True: 6.68k, False: 28.4k] ------------------ 330| 35.1k| if (e->quarterpel) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (330:8): [True: 6.20k, False: 28.9k] ------------------ 331| 35.1k| } 332| 1.29M| if (coding_type == B_VOP) { ------------------ | | 106| 1.29M|#define B_VOP 2 ------------------ | Branch (332:7): [True: 93.3k, False: 1.19M] ------------------ 333| 93.3k| if (e->opaque) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (333:8): [True: 42.2k, False: 51.0k] ------------------ 334| 93.3k| if (e->transparent) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (334:8): [True: 47.3k, False: 45.9k] ------------------ 335| 93.3k| if (e->intra_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (335:8): [True: 15.1k, False: 78.1k] ------------------ 336| 93.3k| if (e->inter_cae) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (336:8): [True: 49.3k, False: 44.0k] ------------------ 337| 93.3k| if (e->no_update) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (337:8): [True: 24.0k, False: 69.2k] ------------------ 338| 93.3k| if (e->upsampling) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (338:8): [True: 12.0k, False: 81.2k] ------------------ 339| 93.3k| if (e->intra_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (339:8): [True: 20.1k, False: 73.2k] ------------------ 340| 93.3k| if (e->not_coded_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (340:8): [True: 10.1k, False: 83.2k] ------------------ 341| 93.3k| if (e->dct_coefs) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (341:8): [True: 13.4k, False: 79.8k] ------------------ 342| 93.3k| if (e->dct_lines) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (342:8): [True: 18.1k, False: 75.2k] ------------------ 343| 93.3k| if (e->vlc_symbols) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (343:8): [True: 33.8k, False: 59.5k] ------------------ 344| 93.3k| if (e->vlc_bits) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (344:8): [True: 14.1k, False: 79.1k] ------------------ 345| 93.3k| if (e->inter_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (345:8): [True: 16.3k, False: 77.0k] ------------------ 346| 93.3k| if (e->inter4v_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (346:8): [True: 20.0k, False: 73.3k] ------------------ 347| 93.3k| if (e->apm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (347:8): [True: 54.7k, False: 38.5k] ------------------ 348| 93.3k| if (e->npm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (348:8): [True: 60.1k, False: 33.1k] ------------------ 349| 93.3k| if (e->forw_back_mc_q) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (349:8): [True: 23.6k, False: 69.7k] ------------------ 350| 93.3k| if (e->halfpel2) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (350:8): [True: 26.7k, False: 66.6k] ------------------ 351| 93.3k| if (e->halfpel4) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (351:8): [True: 38.6k, False: 54.7k] ------------------ 352| 93.3k| if (e->interpolate_mc_q) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (352:8): [True: 55.6k, False: 37.7k] ------------------ 353| 93.3k| if (e->sadct) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (353:8): [True: 37.5k, False: 55.7k] ------------------ 354| 93.3k| if (e->quarterpel) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (354:8): [True: 29.6k, False: 63.6k] ------------------ 355| 93.3k| } 356| | 357| 1.29M| if (coding_type == S_VOP && dec->sprite_enable == SPRITE_STATIC) { ------------------ | | 107| 2.58M|#define S_VOP 3 ------------------ if (coding_type == S_VOP && dec->sprite_enable == SPRITE_STATIC) { ------------------ | | 94| 1.14M|#define SPRITE_STATIC 1 ------------------ | Branch (357:7): [True: 1.14M, False: 147k] | Branch (357:31): [True: 421k, False: 721k] ------------------ 358| 421k| if (e->intra_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (358:8): [True: 49.6k, False: 372k] ------------------ 359| 421k| if (e->not_coded_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (359:8): [True: 251k, False: 170k] ------------------ 360| 421k| if (e->dct_coefs) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (360:8): [True: 33.4k, False: 388k] ------------------ 361| 421k| if (e->dct_lines) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (361:8): [True: 34.3k, False: 387k] ------------------ 362| 421k| if (e->vlc_symbols) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (362:8): [True: 108k, False: 313k] ------------------ 363| 421k| if (e->vlc_bits) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (363:8): [True: 20.1k, False: 401k] ------------------ 364| 421k| if (e->inter_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (364:8): [True: 50.8k, False: 371k] ------------------ 365| 421k| if (e->inter4v_blocks) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (365:8): [True: 265k, False: 156k] ------------------ 366| 421k| if (e->apm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (366:8): [True: 119k, False: 302k] ------------------ 367| 421k| if (e->npm) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (367:8): [True: 347k, False: 74.9k] ------------------ 368| 421k| if (e->forw_back_mc_q) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (368:8): [True: 263k, False: 158k] ------------------ 369| 421k| if (e->halfpel2) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (369:8): [True: 289k, False: 132k] ------------------ 370| 421k| if (e->halfpel4) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (370:8): [True: 269k, False: 151k] ------------------ 371| 421k| if (e->interpolate_mc_q) BitstreamSkip(bs, 8); /* */ ------------------ | Branch (371:8): [True: 355k, False: 66.8k] ------------------ 372| 421k| } 373| 1.29M| } 374| 1.36M|} mbcoding.c:BitstreamShowBits: 254| 685M|{ 255| 685M| int nbit = (bits + bs->pos) - 32; 256| | 257| 685M| if (nbit > 0) { ------------------ | Branch (257:6): [True: 301M, False: 383M] ------------------ 258| 301M| return ((bs->bufa & (0xffffffff >> bs->pos)) << nbit) | (bs-> 259| 301M| bufb >> (32 - 260| 301M| nbit)); 261| 383M| } else { 262| 383M| return (bs->bufa & (0xffffffff >> bs->pos)) >> (32 - bs->pos - bits); 263| 383M| } 264| 685M|} mbcoding.c:BitstreamSkip: 272| 277M|{ 273| 277M| bs->pos += bits; 274| | 275| 277M| if (bs->pos >= 32) { ------------------ | Branch (275:6): [True: 42.7M, False: 234M] ------------------ 276| 42.7M| uint32_t tmp; 277| | 278| 42.7M| bs->bufa = bs->bufb; 279| 42.7M|#if defined(XVID_SAFE_BS_TAIL) 280| 42.7M| if (bs->tail<(bs->start+((bs->length+3)>>2))) ------------------ | Branch (280:7): [True: 5.85M, False: 36.9M] ------------------ 281| 5.85M|#endif 282| 5.85M| { 283| 5.85M| tmp = *((uint32_t *) bs->tail + 2); 284| 5.85M|#ifndef ARCH_IS_BIG_ENDIAN 285| 5.85M| BSWAP(tmp); ------------------ | | 365| 5.85M| ((a) = (((a) & 0xff) << 24) | (((a) & 0xff00) << 8) | \ | | 366| 5.85M| (((a) >> 8) & 0xff00) | (((a) >> 24) & 0xff)) ------------------ 286| 5.85M|#endif 287| 5.85M| bs->bufb = tmp; 288| 5.85M| bs->tail++; 289| 5.85M| } 290| 36.9M|#if defined(XVID_SAFE_BS_TAIL) 291| 36.9M| else { 292| 36.9M| bs->bufb = 0; 293| 36.9M| } 294| 42.7M|#endif 295| 42.7M| bs->pos -= 32; 296| 42.7M| } 297| 277M|} mbcoding.c:BitstreamNumBitsToByteAlign: 303| 56.1M|{ 304| 56.1M| uint32_t n = (32 - bs->pos) % 8; 305| 56.1M| return n == 0 ? 8 : n; ------------------ | Branch (305:9): [True: 6.85M, False: 49.2M] ------------------ 306| 56.1M|} mbcoding.c:BitstreamShowBitsFromByteAlign: 312| 5.33M|{ 313| 5.33M| int bspos = bs->pos + BitstreamNumBitsToByteAlign(bs); 314| 5.33M| int nbit = (bits + bspos) - 32; 315| | 316| 5.33M| if (bspos >= 32) { ------------------ | Branch (316:6): [True: 1.10M, False: 4.22M] ------------------ 317| 1.10M| return bs->bufb >> (32 - nbit); 318| 4.22M| } else if (nbit > 0) { ------------------ | Branch (318:13): [True: 2.13M, False: 2.09M] ------------------ 319| 2.13M| return ((bs->bufa & (0xffffffff >> bspos)) << nbit) | (bs-> 320| 2.13M| bufb >> (32 - 321| 2.13M| nbit)); 322| 2.13M| } else { 323| 2.09M| return (bs->bufa & (0xffffffff >> bspos)) >> (32 - bspos - bits); 324| 2.09M| } 325| | 326| 5.33M|} mbcoding.c:BitstreamGetBit: 418| 1.45M|{ 419| 1.45M| return BitstreamGetBits(bs, 1); 420| 1.45M|} mbcoding.c:BitstreamGetBits: 406| 91.0M|{ 407| 91.0M| uint32_t ret = BitstreamShowBits(bs, n); 408| | 409| 91.0M| BitstreamSkip(bs, n); 410| 91.0M| return ret; 411| 91.0M|} decoder.c:BitstreamInit: 184| 69.9k|{ 185| 69.9k| uint32_t tmp; 186| 69.9k| size_t bitpos; 187| 69.9k| ptr_t adjbitstream = (ptr_t)bitstream; ------------------ | | 132| 69.9k|# define ptr_t uint64_t ------------------ 188| | 189| | /* 190| | * Start the stream on a uint32_t boundary, by rounding down to the 191| | * previous uint32_t and skipping the intervening bytes. 192| | */ 193| 69.9k| bitpos = ((sizeof(uint32_t)-1) & (size_t)bitstream); 194| 69.9k| adjbitstream = adjbitstream - bitpos; 195| 69.9k| bs->start = bs->tail = (uint32_t *) adjbitstream; 196| | 197| 69.9k| tmp = *bs->start; 198| 69.9k|#ifndef ARCH_IS_BIG_ENDIAN 199| 69.9k| BSWAP(tmp); ------------------ | | 365| 69.9k| ((a) = (((a) & 0xff) << 24) | (((a) & 0xff00) << 8) | \ | | 366| 69.9k| (((a) >> 8) & 0xff00) | (((a) >> 24) & 0xff)) ------------------ 200| 69.9k|#endif 201| 69.9k| bs->bufa = tmp; 202| | 203| 69.9k| tmp = *(bs->start + 1); 204| 69.9k|#ifndef ARCH_IS_BIG_ENDIAN 205| 69.9k| BSWAP(tmp); ------------------ | | 365| 69.9k| ((a) = (((a) & 0xff) << 24) | (((a) & 0xff00) << 8) | \ | | 366| 69.9k| (((a) >> 8) & 0xff00) | (((a) >> 24) & 0xff)) ------------------ 206| 69.9k|#endif 207| 69.9k| bs->bufb = tmp; 208| | 209| 69.9k| bs->pos = bs->initpos = (uint32_t) bitpos*8; 210| | /* preserve the intervening bytes */ 211| 69.9k| if (bs->initpos > 0) ------------------ | Branch (211:6): [True: 43.7k, False: 26.2k] ------------------ 212| 43.7k| bs->buf = bs->bufa & (0xffffffff << (32 - bs->initpos)); 213| 26.2k| else 214| 26.2k| bs->buf = 0; 215| 69.9k| bs->length = length; 216| 69.9k|} decoder.c:BitstreamShowBits: 254| 205M|{ 255| 205M| int nbit = (bits + bs->pos) - 32; 256| | 257| 205M| if (nbit > 0) { ------------------ | Branch (257:6): [True: 51.0M, False: 154M] ------------------ 258| 51.0M| return ((bs->bufa & (0xffffffff >> bs->pos)) << nbit) | (bs-> 259| 51.0M| bufb >> (32 - 260| 51.0M| nbit)); 261| 154M| } else { 262| 154M| return (bs->bufa & (0xffffffff >> bs->pos)) >> (32 - bs->pos - bits); 263| 154M| } 264| 205M|} decoder.c:BitstreamPos: 347| 69.9k|{ 348| 69.9k| return((uint32_t)(8*((ptr_t)bs->tail - (ptr_t)bs->start) + bs->pos - bs->initpos)); 349| 69.9k|} decoder.c:BitstreamSkip: 272| 112M|{ 273| 112M| bs->pos += bits; 274| | 275| 112M| if (bs->pos >= 32) { ------------------ | Branch (275:6): [True: 27.2M, False: 85.6M] ------------------ 276| 27.2M| uint32_t tmp; 277| | 278| 27.2M| bs->bufa = bs->bufb; 279| 27.2M|#if defined(XVID_SAFE_BS_TAIL) 280| 27.2M| if (bs->tail<(bs->start+((bs->length+3)>>2))) ------------------ | Branch (280:7): [True: 3.03M, False: 24.1M] ------------------ 281| 3.03M|#endif 282| 3.03M| { 283| 3.03M| tmp = *((uint32_t *) bs->tail + 2); 284| 3.03M|#ifndef ARCH_IS_BIG_ENDIAN 285| 3.03M| BSWAP(tmp); ------------------ | | 365| 3.03M| ((a) = (((a) & 0xff) << 24) | (((a) & 0xff00) << 8) | \ | | 366| 3.03M| (((a) >> 8) & 0xff00) | (((a) >> 24) & 0xff)) ------------------ 286| 3.03M|#endif 287| 3.03M| bs->bufb = tmp; 288| 3.03M| bs->tail++; 289| 3.03M| } 290| 24.1M|#if defined(XVID_SAFE_BS_TAIL) 291| 24.1M| else { 292| 24.1M| bs->bufb = 0; 293| 24.1M| } 294| 27.2M|#endif 295| 27.2M| bs->pos -= 32; 296| 27.2M| } 297| 112M|} decoder.c:BitstreamGetBit: 418| 108M|{ 419| 108M| return BitstreamGetBits(bs, 1); 420| 108M|} decoder.c:BitstreamGetBits: 406| 112M|{ 407| 112M| uint32_t ret = BitstreamShowBits(bs, n); 408| | 409| 112M| BitstreamSkip(bs, n); 410| 112M| return ret; 411| 112M|} bitstream.c:BitstreamSkip: 272| 75.2M|{ 273| 75.2M| bs->pos += bits; 274| | 275| 75.2M| if (bs->pos >= 32) { ------------------ | Branch (275:6): [True: 36.2M, False: 39.0M] ------------------ 276| 36.2M| uint32_t tmp; 277| | 278| 36.2M| bs->bufa = bs->bufb; 279| 36.2M|#if defined(XVID_SAFE_BS_TAIL) 280| 36.2M| if (bs->tail<(bs->start+((bs->length+3)>>2))) ------------------ | Branch (280:7): [True: 34.2M, False: 1.92M] ------------------ 281| 34.2M|#endif 282| 34.2M| { 283| 34.2M| tmp = *((uint32_t *) bs->tail + 2); 284| 34.2M|#ifndef ARCH_IS_BIG_ENDIAN 285| 34.2M| BSWAP(tmp); ------------------ | | 365| 34.2M| ((a) = (((a) & 0xff) << 24) | (((a) & 0xff00) << 8) | \ | | 366| 34.2M| (((a) >> 8) & 0xff00) | (((a) >> 24) & 0xff)) ------------------ 286| 34.2M|#endif 287| 34.2M| bs->bufb = tmp; 288| 34.2M| bs->tail++; 289| 34.2M| } 290| 1.92M|#if defined(XVID_SAFE_BS_TAIL) 291| 1.92M| else { 292| 1.92M| bs->bufb = 0; 293| 1.92M| } 294| 36.2M|#endif 295| 36.2M| bs->pos -= 32; 296| 36.2M| } 297| 75.2M|} bitstream.c:BitstreamNumBitsToByteAlign: 303| 28.2k|{ 304| 28.2k| uint32_t n = (32 - bs->pos) % 8; 305| 28.2k| return n == 0 ? 8 : n; ------------------ | Branch (305:9): [True: 364, False: 27.8k] ------------------ 306| 28.2k|} bitstream.c:BitstreamGetBit: 418| 18.1M|{ 419| 18.1M| return BitstreamGetBits(bs, 1); 420| 18.1M|} bitstream.c:BitstreamGetBits: 406| 33.6M|{ 407| 33.6M| uint32_t ret = BitstreamShowBits(bs, n); 408| | 409| 33.6M| BitstreamSkip(bs, n); 410| 33.6M| return ret; 411| 33.6M|} bitstream.c:BitstreamPos: 347| 22.8M|{ 348| 22.8M| return((uint32_t)(8*((ptr_t)bs->tail - (ptr_t)bs->start) + bs->pos - bs->initpos)); 349| 22.8M|} bitstream.c:BitstreamByteAlign: 334| 22.8M|{ 335| 22.8M| uint32_t remainder = bs->pos % 8; 336| | 337| 22.8M| if (remainder) { ------------------ | Branch (337:6): [True: 1.66M, False: 21.1M] ------------------ 338| 1.66M| BitstreamSkip(bs, 8 - remainder); 339| 1.66M| } 340| 22.8M|} bitstream.c:BitstreamShowBits: 254| 77.8M|{ 255| 77.8M| int nbit = (bits + bs->pos) - 32; 256| | 257| 77.8M| if (nbit > 0) { ------------------ | Branch (257:6): [True: 13.5M, False: 64.3M] ------------------ 258| 13.5M| return ((bs->bufa & (0xffffffff >> bs->pos)) << nbit) | (bs-> 259| 13.5M| bufb >> (32 - 260| 13.5M| nbit)); 261| 64.3M| } else { 262| 64.3M| return (bs->bufa & (0xffffffff >> bs->pos)) >> (32 - bs->pos - bits); 263| 64.3M| } 264| 77.8M|} bs_get_spritetrajectory: 71| 3.61M|{ 72| 3.61M| int i; 73| 25.3M| for (i = 0; i < 12; i++) ------------------ | Branch (73:14): [True: 23.9M, False: 1.34M] ------------------ 74| 23.9M| { 75| 23.9M| if (BitstreamShowBits(bs, sprite_trajectory_len[i].len) == sprite_trajectory_len[i].code) ------------------ | Branch (75:7): [True: 2.26M, False: 21.7M] ------------------ 76| 2.26M| { 77| 2.26M| BitstreamSkip(bs, sprite_trajectory_len[i].len); 78| 2.26M| return i; 79| 2.26M| } 80| 23.9M| } 81| 1.34M| return -1; 82| 3.61M|} init_vlc_tables: 86| 2|{ 87| 2| uint32_t i, j, k, intra, last, run, run_esc, level, level_esc, escape, escape_len, offset; 88| 2| int32_t l; 89| | 90| 6| for (intra = 0; intra < 2; intra++) ------------------ | Branch (90:18): [True: 4, False: 2] ------------------ 91| 16.3k| for (i = 0; i < 4096; i++) ------------------ | Branch (91:15): [True: 16.3k, False: 4] ------------------ 92| 16.3k| DCT3D[intra][i].event.level = 0; 93| | 94| 6| for (intra = 0; intra < 2; intra++) { ------------------ | Branch (94:18): [True: 4, False: 2] ------------------ 95| 12| for (last = 0; last < 2; last++) { ------------------ | Branch (95:18): [True: 8, False: 4] ------------------ 96| 516| for (run = 0; run < 63 + last; run++) { ------------------ | Branch (96:18): [True: 508, False: 8] ------------------ 97| 24.8k| for (level = 0; level < (uint32_t)(32 << intra); level++) { ------------------ | Branch (97:21): [True: 24.3k, False: 508] ------------------ 98| 24.3k| offset = !intra * LEVELOFFSET; ------------------ | | 46| 24.3k|#define LEVELOFFSET 32 ------------------ 99| 24.3k| coeff_VLC[intra][last][level + offset][run].len = 128; 100| 24.3k| } 101| 508| } 102| 8| } 103| 4| } 104| | 105| 6| for (intra = 0; intra < 2; intra++) { ------------------ | Branch (105:18): [True: 4, False: 2] ------------------ 106| 412| for (i = 0; i < 102; i++) { ------------------ | Branch (106:15): [True: 408, False: 4] ------------------ 107| 408| offset = !intra * LEVELOFFSET; ------------------ | | 46| 408|#define LEVELOFFSET 32 ------------------ 108| | 109| 16.6k| for (j = 0; j < (uint32_t)(1 << (12 - coeff_tab[intra][i].vlc.len)); j++) { ------------------ | Branch (109:16): [True: 16.2k, False: 408] ------------------ 110| 16.2k| DCT3D[intra][(coeff_tab[intra][i].vlc.code << (12 - coeff_tab[intra][i].vlc.len)) | j].len = coeff_tab[intra][i].vlc.len; 111| 16.2k| DCT3D[intra][(coeff_tab[intra][i].vlc.code << (12 - coeff_tab[intra][i].vlc.len)) | j].event = coeff_tab[intra][i].event; 112| 16.2k| } 113| | 114| 408| coeff_VLC[intra][coeff_tab[intra][i].event.last][coeff_tab[intra][i].event.level + offset][coeff_tab[intra][i].event.run].code 115| 408| = coeff_tab[intra][i].vlc.code << 1; 116| 408| coeff_VLC[intra][coeff_tab[intra][i].event.last][coeff_tab[intra][i].event.level + offset][coeff_tab[intra][i].event.run].len 117| 408| = coeff_tab[intra][i].vlc.len + 1; 118| | 119| 408| if (!intra) { ------------------ | Branch (119:8): [True: 204, False: 204] ------------------ 120| 204| coeff_VLC[intra][coeff_tab[intra][i].event.last][offset - coeff_tab[intra][i].event.level][coeff_tab[intra][i].event.run].code 121| 204| = (coeff_tab[intra][i].vlc.code << 1) | 1; 122| 204| coeff_VLC[intra][coeff_tab[intra][i].event.last][offset - coeff_tab[intra][i].event.level][coeff_tab[intra][i].event.run].len 123| 204| = coeff_tab[intra][i].vlc.len + 1; 124| 204| } 125| 408| } 126| 4| } 127| | 128| 6| for (intra = 0; intra < 2; intra++) { ------------------ | Branch (128:18): [True: 4, False: 2] ------------------ 129| 12| for (last = 0; last < 2; last++) { ------------------ | Branch (129:18): [True: 8, False: 4] ------------------ 130| 516| for (run = 0; run < 63 + last; run++) { ------------------ | Branch (130:18): [True: 508, False: 8] ------------------ 131| 24.3k| for (level = 1; level < (uint32_t)(32 << intra); level++) { ------------------ | Branch (131:21): [True: 23.8k, False: 508] ------------------ 132| | 133| 23.8k| if (level <= max_level[intra][last][run] && run <= max_run[intra][last][level]) ------------------ | Branch (133:10): [True: 408, False: 23.4k] | Branch (133:50): [True: 408, False: 0] ------------------ 134| 408| continue; 135| | 136| 23.4k| offset = !intra * LEVELOFFSET; ------------------ | | 46| 23.4k|#define LEVELOFFSET 32 ------------------ 137| 23.4k| level_esc = level - max_level[intra][last][run]; 138| 23.4k| run_esc = run - 1 - max_run[intra][last][level]; 139| | 140| 23.4k| if (level_esc <= max_level[intra][last][run] && run <= max_run[intra][last][level_esc]) { ------------------ | Branch (140:10): [True: 408, False: 23.0k] | Branch (140:54): [True: 408, False: 0] ------------------ 141| 408| escape = ESCAPE1; ------------------ | | 33| 408|#define ESCAPE1 6 ------------------ 142| 408| escape_len = 7 + 1; 143| 408| run_esc = run; 144| 23.0k| } else { 145| 23.0k| if (run_esc <= max_run[intra][last][level] && level <= max_level[intra][last][run_esc]) { ------------------ | Branch (145:11): [True: 496, False: 22.5k] | Branch (145:53): [True: 222, False: 274] ------------------ 146| 222| escape = ESCAPE2; ------------------ | | 34| 222|#define ESCAPE2 14 ------------------ 147| 222| escape_len = 7 + 2; 148| 222| level_esc = level; 149| 22.8k| } else { 150| 22.8k| if (!intra) { ------------------ | Branch (150:12): [True: 7.36k, False: 15.4k] ------------------ 151| 7.36k| coeff_VLC[intra][last][level + offset][run].code 152| 7.36k| = (ESCAPE3 << 21) | (last << 20) | (run << 14) | (1 << 13) | ((level & 0xfff) << 1) | 1; ------------------ | | 35| 7.36k|#define ESCAPE3 15 ------------------ 153| 7.36k| coeff_VLC[intra][last][level + offset][run].len = 30; 154| 7.36k| coeff_VLC[intra][last][offset - level][run].code 155| 7.36k| = (ESCAPE3 << 21) | (last << 20) | (run << 14) | (1 << 13) | ((-(int32_t)level & 0xfff) << 1) | 1; ------------------ | | 35| 7.36k|#define ESCAPE3 15 ------------------ 156| 7.36k| coeff_VLC[intra][last][offset - level][run].len = 30; 157| 7.36k| } 158| 22.8k| continue; 159| 22.8k| } 160| 23.0k| } 161| | 162| 630| coeff_VLC[intra][last][level + offset][run].code 163| 630| = (escape << coeff_VLC[intra][last][level_esc + offset][run_esc].len) 164| 630| | coeff_VLC[intra][last][level_esc + offset][run_esc].code; 165| 630| coeff_VLC[intra][last][level + offset][run].len 166| 630| = coeff_VLC[intra][last][level_esc + offset][run_esc].len + escape_len; 167| | 168| 630| if (!intra) { ------------------ | Branch (168:10): [True: 310, False: 320] ------------------ 169| 310| coeff_VLC[intra][last][offset - level][run].code 170| 310| = (escape << coeff_VLC[intra][last][level_esc + offset][run_esc].len) 171| 310| | coeff_VLC[intra][last][level_esc + offset][run_esc].code | 1; 172| 310| coeff_VLC[intra][last][offset - level][run].len 173| 310| = coeff_VLC[intra][last][level_esc + offset][run_esc].len + escape_len; 174| 310| } 175| 630| } 176| | 177| 508| if (!intra) { ------------------ | Branch (177:9): [True: 254, False: 254] ------------------ 178| 254| coeff_VLC[intra][last][0][run].code 179| 254| = (ESCAPE3 << 21) | (last << 20) | (run << 14) | (1 << 13) | ((-32 & 0xfff) << 1) | 1; ------------------ | | 35| 254|#define ESCAPE3 15 ------------------ 180| 254| coeff_VLC[intra][last][0][run].len = 30; 181| 254| } 182| 508| } 183| 8| } 184| 4| } 185| | 186| | /* init sprite_trajectory tables 187| | * even if GMC is not specified (it might be used later...) */ 188| | 189| 2| sprite_trajectory_code[0+16384].code = 0; 190| 2| sprite_trajectory_code[0+16384].len = 0; 191| 30| for (k=0;k<14;k++) { ------------------ | Branch (191:11): [True: 28, False: 2] ------------------ 192| 28| int limit = (1<>= 3; 840| | 841| 27.4M| BitstreamSkip(bs, mcbpc_intra_table[index].len); 842| | 843| 27.4M| return mcbpc_intra_table[index].code; 844| | 845| 27.4M|} get_mcbpc_inter: 849| 18.1M|{ 850| | 851| 18.1M| uint32_t index; 852| | 853| 18.1M| index = MIN(BitstreamShowBits(bs, 9), 256); ------------------ | | 255| 18.1M|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 18.0M, False: 144k] | | ------------------ ------------------ 854| | 855| 18.1M| BitstreamSkip(bs, mcbpc_inter_table[index].len); 856| | 857| 18.1M| return mcbpc_inter_table[index].code; 858| | 859| 18.1M|} get_cbpy: 864| 45.6M|{ 865| | 866| 45.6M| int cbpy; 867| 45.6M| uint32_t index = BitstreamShowBits(bs, 6); 868| | 869| 45.6M| BitstreamSkip(bs, cbpy_table[index].len); 870| 45.6M| cbpy = cbpy_table[index].code; 871| | 872| 45.6M| if (!intra) ------------------ | Branch (872:6): [True: 18.1M, False: 27.5M] ------------------ 873| 18.1M| cbpy = 15 - cbpy; 874| | 875| 45.6M| return cbpy; 876| | 877| 45.6M|} get_mv: 912| 1.45M|{ 913| | 914| 1.45M| int data; 915| 1.45M| int res; 916| 1.45M| int mv; 917| 1.45M| int scale_fac = 1 << (fcode - 1); 918| | 919| 1.45M| data = get_mv_data(bs); 920| | 921| 1.45M| if (scale_fac == 1 || data == 0) ------------------ | Branch (921:6): [True: 327k, False: 1.12M] | Branch (921:24): [True: 803k, False: 326k] ------------------ 922| 1.13M| return data; 923| | 924| 326k| res = BitstreamGetBits(bs, fcode - 1); 925| 326k| mv = ((abs(data) - 1) * scale_fac) + res + 1; 926| | 927| 326k| return data < 0 ? -mv : mv; ------------------ | Branch (927:9): [True: 157k, False: 169k] ------------------ 928| | 929| 1.45M|} get_dc_dif: 934| 44.7M|{ 935| | 936| 44.7M| int code = BitstreamGetBits(bs, dc_size); 937| 44.7M| int msb = code >> (dc_size - 1); 938| | 939| 44.7M| if (msb == 0) ------------------ | Branch (939:6): [True: 44.4M, False: 276k] ------------------ 940| 44.4M| return (-1 * (code ^ ((1 << dc_size) - 1))); 941| | 942| 276k| return code; 943| | 944| 44.7M|} get_dc_size_lum: 948| 89.0M|{ 949| | 950| 89.0M| int code, i; 951| | 952| 89.0M| code = BitstreamShowBits(bs, 11); 953| | 954| 801M| for (i = 11; i > 3; i--) { ------------------ | Branch (954:15): [True: 712M, False: 89.0M] ------------------ 955| 712M| if (code == 1) { ------------------ | Branch (955:7): [True: 93.0k, False: 712M] ------------------ 956| 93.0k| BitstreamSkip(bs, i); 957| 93.0k| return i + 1; 958| 93.0k| } 959| 712M| code >>= 1; 960| 712M| } 961| | 962| 89.0M| BitstreamSkip(bs, dc_lum_tab[code].len); 963| 89.0M| return dc_lum_tab[code].code; 964| | 965| 89.0M|} get_dc_size_chrom: 970| 44.5M|{ 971| | 972| 44.5M| uint32_t code, i; 973| | 974| 44.5M| code = BitstreamShowBits(bs, 12); 975| | 976| 489M| for (i = 12; i > 2; i--) { ------------------ | Branch (976:15): [True: 445M, False: 44.4M] ------------------ 977| 445M| if (code == 1) { ------------------ | Branch (977:7): [True: 60.7k, False: 445M] ------------------ 978| 60.7k| BitstreamSkip(bs, i); 979| 60.7k| return i; 980| 60.7k| } 981| 445M| code >>= 1; 982| 445M| } 983| | 984| 44.4M| return 3 - BitstreamGetBits(bs, 2); 985| | 986| 44.5M|} get_intra_block: 1087| 272M|{ 1088| | 1089| 272M| const uint16_t *scan = scan_tables[direction]; 1090| 272M| int level, run, last = 0; 1091| | 1092| 274M| do { 1093| 274M| level = get_coeff(bs, &run, &last, 1, 0); 1094| 274M| coeff += run; 1095| 274M| if (coeff & ~63) { ------------------ | Branch (1095:7): [True: 272M, False: 2.53M] ------------------ 1096| 272M| DPRINTF(XVID_DEBUG_ERROR,"fatal: invalid run or index"); ------------------ | | 196| 272M|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1097| 272M| break; 1098| 272M| } 1099| | 1100| 2.53M| block[scan[coeff]] = level; 1101| | 1102| 2.53M| DPRINTF(XVID_DEBUG_COEFF,"block[%i] %i\n", scan[coeff], level); ------------------ | | 201| 2.53M|#define XVID_DEBUG_COEFF (1<< 5) ------------------ 1103| |#if 0 1104| | DPRINTF(XVID_DEBUG_COEFF,"block[%i] %i %08x\n", scan[coeff], level, BitstreamShowBits(bs, 32)); 1105| |#endif 1106| | 1107| 2.53M| if (level < -2047 || level > 2047) { ------------------ | Branch (1107:7): [True: 13, False: 2.53M] | Branch (1107:24): [True: 0, False: 2.53M] ------------------ 1108| 13| DPRINTF(XVID_DEBUG_ERROR,"warning: intra_overflow %i\n", level); ------------------ | | 196| 13|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1109| 13| } 1110| 2.53M| coeff++; 1111| 2.53M| } while (!last); ------------------ | Branch (1111:11): [True: 2.21M, False: 319k] ------------------ 1112| | 1113| 272M|} get_inter_block_h263: 1122| 216k|{ 1123| | 1124| 216k| const uint16_t *scan = scan_tables[direction]; 1125| 216k| const uint16_t quant_m_2 = quant << 1; 1126| 216k| const uint16_t quant_add = (quant & 1 ? quant : quant - 1); ------------------ | Branch (1126:30): [True: 155k, False: 60.2k] ------------------ 1127| 216k| int p; 1128| 216k| int level; 1129| 216k| int run; 1130| 216k| int last = 0; 1131| | 1132| 216k| p = 0; 1133| 346k| do { 1134| 346k| level = get_coeff(bs, &run, &last, 0, 0); 1135| 346k| p += run; 1136| 346k| if (p & ~63) { ------------------ | Branch (1136:7): [True: 179k, False: 167k] ------------------ 1137| 179k| DPRINTF(XVID_DEBUG_ERROR,"fatal: invalid run or index"); ------------------ | | 196| 179k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1138| 179k| break; 1139| 179k| } 1140| | 1141| 167k| if (level < 0) { ------------------ | Branch (1141:7): [True: 89.6k, False: 77.4k] ------------------ 1142| 89.6k| level = level*quant_m_2 - quant_add; 1143| 89.6k| block[scan[p]] = (level >= -2048 ? level : -2048); ------------------ | Branch (1143:22): [True: 88.6k, False: 960] ------------------ 1144| 89.6k| } else { 1145| 77.4k| level = level * quant_m_2 + quant_add; 1146| 77.4k| block[scan[p]] = (level <= 2047 ? level : 2047); ------------------ | Branch (1146:22): [True: 76.7k, False: 731] ------------------ 1147| 77.4k| } 1148| 167k| p++; 1149| 167k| } while (!last); ------------------ | Branch (1149:11): [True: 129k, False: 37.1k] ------------------ 1150| 216k|} get_inter_block_mpeg: 1159| 310k|{ 1160| 310k| const uint16_t *scan = scan_tables[direction]; 1161| 310k| uint32_t sum = 0; 1162| 310k| int p; 1163| 310k| int level; 1164| 310k| int run; 1165| 310k| int last = 0; 1166| | 1167| 310k| p = 0; 1168| 533k| do { 1169| 533k| level = get_coeff(bs, &run, &last, 0, 0); 1170| 533k| p += run; 1171| 533k| if (p & ~63) { ------------------ | Branch (1171:7): [True: 263k, False: 269k] ------------------ 1172| 263k| DPRINTF(XVID_DEBUG_ERROR,"fatal: invalid run or index"); ------------------ | | 196| 263k|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1173| 263k| break; 1174| 263k| } 1175| | 1176| 269k| if (level < 0) { ------------------ | Branch (1176:7): [True: 151k, False: 118k] ------------------ 1177| 151k| level = ((2 * -level + 1) * matrix[scan[p]] * quant) >> 4; 1178| 151k| block[scan[p]] = (level <= 2048 ? -level : -2048); ------------------ | Branch (1178:22): [True: 146k, False: 4.15k] ------------------ 1179| 151k| } else { 1180| 118k| level = ((2 * level + 1) * matrix[scan[p]] * quant) >> 4; 1181| 118k| block[scan[p]] = (level <= 2047 ? level : 2047); ------------------ | Branch (1181:22): [True: 116k, False: 2.01k] ------------------ 1182| 118k| } 1183| | 1184| 269k| sum ^= block[scan[p]]; 1185| | 1186| 269k| p++; 1187| 269k| } while (!last); ------------------ | Branch (1187:11): [True: 222k, False: 46.6k] ------------------ 1188| | 1189| | /* mismatch control */ 1190| 310k| if ((sum & 1) == 0) { ------------------ | Branch (1190:6): [True: 275k, False: 35.3k] ------------------ 1191| 275k| block[63] ^= 1; 1192| 275k| } 1193| 310k|} mbcoding.c:get_mv_data: 881| 1.45M|{ 882| | 883| 1.45M| uint32_t index; 884| | 885| 1.45M| if (BitstreamGetBit(bs)) ------------------ | Branch (885:6): [True: 655k, False: 801k] ------------------ 886| 655k| return 0; 887| | 888| 801k| index = BitstreamShowBits(bs, 12); 889| | 890| 801k| if (index >= 512) { ------------------ | Branch (890:6): [True: 355k, False: 445k] ------------------ 891| 355k| index = (index >> 8) - 2; 892| 355k| BitstreamSkip(bs, TMNMVtab0[index].len); 893| 355k| return TMNMVtab0[index].code; 894| 355k| } 895| | 896| 445k| if (index >= 128) { ------------------ | Branch (896:6): [True: 50.0k, False: 395k] ------------------ 897| 50.0k| index = (index >> 2) - 32; 898| 50.0k| BitstreamSkip(bs, TMNMVtab1[index].len); 899| 50.0k| return TMNMVtab1[index].code; 900| 50.0k| } 901| | 902| 395k| index -= 4; 903| | 904| 395k| BitstreamSkip(bs, TMNMVtab2[index&0x7f].len); 905| 395k| return TMNMVtab2[index&0x7f].code; 906| | 907| 445k|} mbcoding.c:get_coeff: 996| 275M|{ 997| | 998| 275M| uint32_t mode; 999| 275M| int32_t level; 1000| 275M| REVERSE_EVENT *reverse_event; 1001| | 1002| 275M| uint32_t cache = BitstreamShowBits(bs, 32); 1003| | 1004| 275M| if (short_video_header) /* inter-VLCs will be used for both intra and inter blocks */ ------------------ | Branch (1004:6): [True: 0, False: 275M] ------------------ 1005| 0| intra = 0; 1006| | 1007| 275M| if (GET_BITS(cache, 7) != ESCAPE) { ------------------ | | 988| 275M|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ if (GET_BITS(cache, 7) != ESCAPE) { ------------------ | | 32| 275M|#define ESCAPE 3 ------------------ | Branch (1007:6): [True: 275M, False: 17.2k] ------------------ 1008| 275M| reverse_event = &DCT3D[intra][GET_BITS(cache, 12)]; ------------------ | | 988| 275M|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1009| | 1010| 275M| if ((level = reverse_event->event.level) == 0) ------------------ | Branch (1010:7): [True: 272M, False: 2.96M] ------------------ 1011| 272M| goto error; 1012| | 1013| 2.96M| *last = reverse_event->event.last; 1014| 2.96M| *run = reverse_event->event.run; 1015| | 1016| | /* Don't forget to update the bitstream position */ 1017| 2.96M| BitstreamSkip(bs, reverse_event->len+1); 1018| | 1019| 2.96M| return (GET_BITS(cache, reverse_event->len+1)&0x01) ? -level : level; ------------------ | | 988| 2.96M|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ | Branch (1019:10): [True: 2.04M, False: 923k] ------------------ 1020| 275M| } 1021| | 1022| | /* flush 7bits of cache */ 1023| 17.2k| cache <<= 7; 1024| | 1025| 17.2k| if (short_video_header) { ------------------ | Branch (1025:6): [True: 0, False: 17.2k] ------------------ 1026| | /* escape mode 4 - H.263 type, only used if short_video_header = 1 */ 1027| 0| *last = GET_BITS(cache, 1); ------------------ | | 988| 0|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1028| 0| *run = (GET_BITS(cache, 7) &0x3f); ------------------ | | 988| 0|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1029| 0| level = (GET_BITS(cache, 15)&0xff); ------------------ | | 988| 0|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1030| | 1031| 0| if (level == 0 || level == 128) ------------------ | Branch (1031:7): [True: 0, False: 0] | Branch (1031:21): [True: 0, False: 0] ------------------ 1032| 0| DPRINTF(XVID_DEBUG_ERROR, "Illegal LEVEL for ESCAPE mode 4: %d\n", level); ------------------ | | 196| 0|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1033| | 1034| | /* We've "eaten" 22 bits */ 1035| 0| BitstreamSkip(bs, 22); 1036| | 1037| 0| return (level << 24) >> 24; 1038| 0| } 1039| | 1040| 17.2k| if ((mode = GET_BITS(cache, 2)) < 3) { ------------------ | | 988| 17.2k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ | Branch (1040:6): [True: 15.3k, False: 1.91k] ------------------ 1041| 15.3k| const int skip[3] = {1, 1, 2}; 1042| 15.3k| cache <<= skip[mode]; 1043| | 1044| 15.3k| reverse_event = &DCT3D[intra][GET_BITS(cache, 12)]; ------------------ | | 988| 15.3k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1045| | 1046| 15.3k| if ((level = reverse_event->event.level) == 0) ------------------ | Branch (1046:7): [True: 6.83k, False: 8.51k] ------------------ 1047| 6.83k| goto error; 1048| | 1049| 8.51k| *last = reverse_event->event.last; 1050| 8.51k| *run = reverse_event->event.run; 1051| | 1052| 8.51k| if (mode < 2) { ------------------ | Branch (1052:7): [True: 6.60k, False: 1.91k] ------------------ 1053| | /* first escape mode, level is offset */ 1054| 6.60k| level += max_level[intra][*last][*run]; 1055| 6.60k| } else { 1056| | /* second escape mode, run is offset */ 1057| 1.91k| *run += max_run[intra][*last][level] + 1; 1058| 1.91k| } 1059| | 1060| | /* Update bitstream position */ 1061| 8.51k| BitstreamSkip(bs, 7 + skip[mode] + reverse_event->len + 1); 1062| | 1063| 8.51k| return (GET_BITS(cache, reverse_event->len+1)&0x01) ? -level : level; ------------------ | | 988| 8.51k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ | Branch (1063:10): [True: 3.24k, False: 5.26k] ------------------ 1064| 15.3k| } 1065| | 1066| | /* third escape mode - fixed length codes */ 1067| 1.91k| cache <<= 2; 1068| 1.91k| *last = GET_BITS(cache, 1); ------------------ | | 988| 1.91k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1069| 1.91k| *run = (GET_BITS(cache, 7)&0x3f); ------------------ | | 988| 1.91k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1070| 1.91k| level = (GET_BITS(cache, 20)&0xfff); ------------------ | | 988| 1.91k|#define GET_BITS(cache, n) ((cache)>>(32-(n))) ------------------ 1071| | 1072| | /* Update bitstream position */ 1073| 1.91k| BitstreamSkip(bs, 30); 1074| | 1075| 1.91k| return (level << 20) >> 20; 1076| | 1077| 272M| error: 1078| 272M| *run = 64; 1079| 272M| return 0; 1080| 17.2k|} idct_int32: 311| 273M|{ 312| 273M| int i, Rows = 0x07; 313| | 314| 273M| Idct_Row(In + 0*8, Tab04, Rnd0); ------------------ | | 49| 273M|#define Rnd0 65536 // 1<<(COL_SHIFT+ROW_SHIFT-1); ------------------ 315| 273M| Idct_Row(In + 1*8, Tab17, Rnd1); ------------------ | | 50| 273M|#define Rnd1 3597 // FIX (1.75683487303); ------------------ 316| 273M| Idct_Row(In + 2*8, Tab26, Rnd2); ------------------ | | 51| 273M|#define Rnd2 2260 // FIX (1.10355339059); ------------------ 317| 273M| if (Idct_Row(In + 3*8, Tab35, Rnd3)) Rows |= 0x08; ------------------ | | 52| 273M|#define Rnd3 1203 // FIX (0.587788325588); ------------------ | Branch (317:7): [True: 209k, False: 272M] ------------------ 318| 273M| if (Idct_Row(In + 4*8, Tab04, Rnd4)) Rows |= 0x10; ------------------ | | 53| 273M|#define Rnd4 0 ------------------ | Branch (318:7): [True: 84.7k, False: 273M] ------------------ 319| 273M| if (Idct_Row(In + 5*8, Tab35, Rnd5)) Rows |= 0x20; ------------------ | | 54| 273M|#define Rnd5 120 // FIX (0.058658283817); ------------------ | Branch (319:7): [True: 53.8k, False: 273M] ------------------ 320| 273M| if (Idct_Row(In + 6*8, Tab26, Rnd6)) Rows |= 0x40; ------------------ | | 55| 273M|#define Rnd6 512 // FIX (0.25); ------------------ | Branch (320:7): [True: 43.7k, False: 273M] ------------------ 321| 273M| if (Idct_Row(In + 7*8, Tab17, Rnd7)) Rows |= 0x80; ------------------ | | 56| 273M|#define Rnd7 512 // FIX (0.25); ------------------ | Branch (321:7): [True: 300k, False: 272M] ------------------ 322| | 323| 273M| if (Rows&0xf0) { ------------------ | Branch (323:7): [True: 375k, False: 272M] ------------------ 324| 3.37M| for(i=0; i<8; i++) ------------------ | Branch (324:14): [True: 3.00M, False: 375k] ------------------ 325| 3.00M| Idct_Col_8(In + i); 326| 375k| } 327| 272M| else if (Rows&0x08) { ------------------ | Branch (327:12): [True: 131k, False: 272M] ------------------ 328| 1.17M| for(i=0; i<8; i++) ------------------ | Branch (328:14): [True: 1.04M, False: 131k] ------------------ 329| 1.04M| Idct_Col_4(In + i); 330| 131k| } 331| 272M| else { 332| 2.45G| for(i=0; i<8; i++) ------------------ | Branch (332:14): [True: 2.18G, False: 272M] ------------------ 333| 2.18G| Idct_Col_3(In + i); 334| 272M| } 335| 273M|} idct.c:Idct_Row: 65| 2.18G|{ 66| 2.18G| const int C1 = Tab[0]; 67| 2.18G| const int C2 = Tab[1]; 68| 2.18G| const int C3 = Tab[2]; 69| 2.18G| const int C4 = Tab[3]; 70| 2.18G| const int C5 = Tab[4]; 71| 2.18G| const int C6 = Tab[5]; 72| 2.18G| const int C7 = Tab[6]; 73| | 74| 2.18G| const int Right = In[5]|In[6]|In[7]; 75| 2.18G| const int Left = In[1]|In[2]|In[3]; 76| 2.18G| if (!(Right | In[4])) ------------------ | Branch (76:7): [True: 2.18G, False: 444k] ------------------ 77| 2.18G| { 78| 2.18G| const int K = C4*In[0] + Rnd; 79| 2.18G| if (Left) ------------------ | Branch (79:9): [True: 848k, False: 2.18G] ------------------ 80| 848k| { 81| 848k| const int a0 = K + C2*In[2]; 82| 848k| const int a1 = K + C6*In[2]; 83| 848k| const int a2 = K - C6*In[2]; 84| 848k| const int a3 = K - C2*In[2]; 85| | 86| 848k| const int b0 = C1*In[1] + C3*In[3]; 87| 848k| const int b1 = C3*In[1] - C7*In[3]; 88| 848k| const int b2 = C5*In[1] - C1*In[3]; 89| 848k| const int b3 = C7*In[1] - C5*In[3]; 90| | 91| 848k| In[0] = (a0 + b0) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 92| 848k| In[1] = (a1 + b1) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 93| 848k| In[2] = (a2 + b2) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 94| 848k| In[3] = (a3 + b3) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 95| 848k| In[4] = (a3 - b3) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 96| 848k| In[5] = (a2 - b2) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 97| 848k| In[6] = (a1 - b1) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 98| 848k| In[7] = (a0 - b0) >> ROW_SHIFT; ------------------ | | 45| 848k|#define ROW_SHIFT 11 ------------------ 99| 848k| } 100| 2.18G| else 101| 2.18G| { 102| 2.18G| const int a0 = K >> ROW_SHIFT; ------------------ | | 45| 2.18G|#define ROW_SHIFT 11 ------------------ 103| 2.18G| if (a0) { ------------------ | Branch (103:11): [True: 818M, False: 1.36G] ------------------ 104| 818M| In[0] = In[1] = In[2] = In[3] = 105| 818M| In[4] = In[5] = In[6] = In[7] = a0; 106| 818M| } 107| 1.36G| else return 0; 108| 2.18G| } 109| 2.18G| } 110| 444k| else if (!(Left|Right)) ------------------ | Branch (110:12): [True: 18.8k, False: 425k] ------------------ 111| 18.8k| { 112| 18.8k| const int a0 = (Rnd + C4*(In[0]+In[4])) >> ROW_SHIFT; ------------------ | | 45| 18.8k|#define ROW_SHIFT 11 ------------------ 113| 18.8k| const int a1 = (Rnd + C4*(In[0]-In[4])) >> ROW_SHIFT; ------------------ | | 45| 18.8k|#define ROW_SHIFT 11 ------------------ 114| | 115| 18.8k| In[0] = a0; 116| 18.8k| In[3] = a0; 117| 18.8k| In[4] = a0; 118| 18.8k| In[7] = a0; 119| 18.8k| In[1] = a1; 120| 18.8k| In[2] = a1; 121| 18.8k| In[5] = a1; 122| 18.8k| In[6] = a1; 123| 18.8k| } 124| 425k| else 125| 425k| { 126| 425k| const int K = C4*In[0] + Rnd; 127| 425k| const int a0 = K + C2*In[2] + C4*In[4] + C6*In[6]; 128| 425k| const int a1 = K + C6*In[2] - C4*In[4] - C2*In[6]; 129| 425k| const int a2 = K - C6*In[2] - C4*In[4] + C2*In[6]; 130| 425k| const int a3 = K - C2*In[2] + C4*In[4] - C6*In[6]; 131| | 132| 425k| const int b0 = C1*In[1] + C3*In[3] + C5*In[5] + C7*In[7]; 133| 425k| const int b1 = C3*In[1] - C7*In[3] - C1*In[5] - C5*In[7]; 134| 425k| const int b2 = C5*In[1] - C1*In[3] + C7*In[5] + C3*In[7]; 135| 425k| const int b3 = C7*In[1] - C5*In[3] + C3*In[5] - C1*In[7]; 136| | 137| 425k| In[0] = (a0 + b0) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 138| 425k| In[1] = (a1 + b1) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 139| 425k| In[2] = (a2 + b2) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 140| 425k| In[3] = (a3 + b3) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 141| 425k| In[4] = (a3 - b3) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 142| 425k| In[5] = (a2 - b2) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 143| 425k| In[6] = (a1 - b1) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 144| 425k| In[7] = (a0 - b0) >> ROW_SHIFT; ------------------ | | 45| 425k|#define ROW_SHIFT 11 ------------------ 145| 425k| } 146| 820M| return 1; 147| 2.18G|} idct.c:Idct_Col_8: 168| 3.00M|{ 169| 3.00M| int mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7, Spill; 170| | 171| | // odd 172| | 173| 3.00M| mm4 = (int)In[7*8]; 174| 3.00M| mm5 = (int)In[5*8]; 175| 3.00M| mm6 = (int)In[3*8]; 176| 3.00M| mm7 = (int)In[1*8]; 177| | 178| 3.00M| mm0 = MULT(Tan1, mm4, 16) + mm7; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 179| 3.00M| mm1 = MULT(Tan1, mm7, 16) - mm4; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 180| 3.00M| mm2 = MULT(Tan3, mm5, 16) + mm6; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 181| 3.00M| mm3 = MULT(Tan3, mm6, 16) - mm5; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 182| | 183| 3.00M| mm7 = mm0 + mm2; 184| 3.00M| mm4 = mm1 - mm3; 185| 3.00M| mm0 = mm0 - mm2; 186| 3.00M| mm1 = mm1 + mm3; 187| 3.00M| mm6 = mm0 + mm1; 188| 3.00M| mm5 = mm0 - mm1; 189| 3.00M| mm5 = 2*MULT(Sqrt2, mm5, 16); // 2*sqrt2 ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 190| 3.00M| mm6 = 2*MULT(Sqrt2, mm6, 16); // Watch out: precision loss but done to match ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 191| | // the pmulhw used in mmx/sse versions 192| | 193| | // even 194| | 195| 3.00M| mm1 = (int)In[2*8]; 196| 3.00M| mm2 = (int)In[6*8]; 197| 3.00M| mm3 = MULT(Tan2,mm2, 16) + mm1; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 198| 3.00M| mm2 = MULT(Tan2,mm1, 16) - mm2; ------------------ | | 154| 3.00M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 199| | 200| 3.00M| LOAD_BUTF(mm0, mm1, 0*8, 4*8, Spill, In); ------------------ | | 164| 3.00M| (m1) = (S)[(a)] + (S)[(b)]; \ | | 165| 3.00M| (m2) = (S)[(a)] - (S)[(b)] ------------------ 201| | 202| 3.00M| BUTF(mm0, mm3, Spill); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 203| 3.00M| BUTF(mm0, mm7, Spill); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 204| 3.00M| In[8*0] = (int16_t) (mm0 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 205| 3.00M| In[8*7] = (int16_t) (mm7 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 206| 3.00M| BUTF(mm3, mm4, mm0); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 207| 3.00M| In[8*3] = (int16_t) (mm3 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 208| 3.00M| In[8*4] = (int16_t) (mm4 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 209| | 210| 3.00M| BUTF(mm1, mm2, mm0); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 211| 3.00M| BUTF(mm1, mm6, mm0); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 212| 3.00M| In[8*1] = (int16_t) (mm1 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 213| 3.00M| In[8*6] = (int16_t) (mm6 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 214| 3.00M| BUTF(mm2, mm5, mm0); ------------------ | | 159| 3.00M| (tmp) = (a)+(b); \ | | 160| 3.00M| (b) = (a)-(b); \ | | 161| 3.00M| (a) = (tmp) ------------------ 215| 3.00M| In[8*2] = (int16_t) (mm2 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 216| 3.00M| In[8*5] = (int16_t) (mm5 >> COL_SHIFT); ------------------ | | 46| 3.00M|#define COL_SHIFT 6 ------------------ 217| 3.00M|} idct.c:Idct_Col_4: 220| 1.04M|{ 221| 1.04M| int mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7, Spill; 222| | 223| | // odd 224| | 225| 1.04M| mm0 = (int)In[1*8]; 226| 1.04M| mm2 = (int)In[3*8]; 227| | 228| 1.04M| mm1 = MULT(Tan1, mm0, 16); ------------------ | | 154| 1.04M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 229| 1.04M| mm3 = MULT(Tan3, mm2, 16); ------------------ | | 154| 1.04M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 230| | 231| 1.04M| mm7 = mm0 + mm2; 232| 1.04M| mm4 = mm1 - mm3; 233| 1.04M| mm0 = mm0 - mm2; 234| 1.04M| mm1 = mm1 + mm3; 235| 1.04M| mm6 = mm0 + mm1; 236| 1.04M| mm5 = mm0 - mm1; 237| 1.04M| mm6 = 2*MULT(Sqrt2, mm6, 16); // 2*sqrt2 ------------------ | | 154| 1.04M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 238| 1.04M| mm5 = 2*MULT(Sqrt2, mm5, 16); ------------------ | | 154| 1.04M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 239| | 240| | // even 241| | 242| 1.04M| mm0 = mm1 = (int)In[0*8]; 243| 1.04M| mm3 = (int)In[2*8]; 244| 1.04M| mm2 = MULT(Tan2,mm3, 16); ------------------ | | 154| 1.04M|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 245| | 246| 1.04M| BUTF(mm0, mm3, Spill); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 247| 1.04M| BUTF(mm0, mm7, Spill); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 248| 1.04M| In[8*0] = (int16_t) (mm0 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 249| 1.04M| In[8*7] = (int16_t) (mm7 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 250| 1.04M| BUTF(mm3, mm4, mm0); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 251| 1.04M| In[8*3] = (int16_t) (mm3 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 252| 1.04M| In[8*4] = (int16_t) (mm4 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 253| | 254| 1.04M| BUTF(mm1, mm2, mm0); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 255| 1.04M| BUTF(mm1, mm6, mm0); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 256| 1.04M| In[8*1] = (int16_t) (mm1 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 257| 1.04M| In[8*6] = (int16_t) (mm6 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 258| 1.04M| BUTF(mm2, mm5, mm0); ------------------ | | 159| 1.04M| (tmp) = (a)+(b); \ | | 160| 1.04M| (b) = (a)-(b); \ | | 161| 1.04M| (a) = (tmp) ------------------ 259| 1.04M| In[8*2] = (int16_t) (mm2 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 260| 1.04M| In[8*5] = (int16_t) (mm5 >> COL_SHIFT); ------------------ | | 46| 1.04M|#define COL_SHIFT 6 ------------------ 261| 1.04M|} idct.c:Idct_Col_3: 264| 2.18G|{ 265| 2.18G| int mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7, Spill; 266| | 267| | // odd 268| | 269| 2.18G| mm7 = (int)In[1*8]; 270| 2.18G| mm4 = MULT(Tan1, mm7, 16); ------------------ | | 154| 2.18G|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 271| | 272| 2.18G| mm6 = mm7 + mm4; 273| 2.18G| mm5 = mm7 - mm4; 274| 2.18G| mm6 = 2*MULT(Sqrt2, mm6, 16); // 2*sqrt2 ------------------ | | 154| 2.18G|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 275| 2.18G| mm5 = 2*MULT(Sqrt2, mm5, 16); ------------------ | | 154| 2.18G|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 276| | 277| | // even 278| | 279| 2.18G| mm0 = mm1 = (int)In[0*8]; 280| 2.18G| mm3 = (int)In[2*8]; 281| 2.18G| mm2 = MULT(Tan2,mm3, 16); ------------------ | | 154| 2.18G|#define MULT(c,x, n) ( ((c) * (x)) >> (n) ) ------------------ 282| | 283| 2.18G| BUTF(mm0, mm3, Spill); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 284| 2.18G| BUTF(mm0, mm7, Spill); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 285| 2.18G| In[8*0] = (int16_t) (mm0 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 286| 2.18G| In[8*7] = (int16_t) (mm7 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 287| 2.18G| BUTF(mm3, mm4, mm0); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 288| 2.18G| In[8*3] = (int16_t) (mm3 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 289| 2.18G| In[8*4] = (int16_t) (mm4 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 290| | 291| 2.18G| BUTF(mm1, mm2, mm0); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 292| 2.18G| BUTF(mm1, mm6, mm0); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 293| 2.18G| In[8*1] = (int16_t) (mm1 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 294| 2.18G| In[8*6] = (int16_t) (mm6 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 295| 2.18G| BUTF(mm2, mm5, mm0); ------------------ | | 159| 2.18G| (tmp) = (a)+(b); \ | | 160| 2.18G| (b) = (a)-(b); \ | | 161| 2.18G| (a) = (tmp) ------------------ 296| 2.18G| In[8*2] = (int16_t) (mm2 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 297| 2.18G| In[8*5] = (int16_t) (mm5 >> COL_SHIFT); ------------------ | | 46| 2.18G|#define COL_SHIFT 6 ------------------ 298| 2.18G|} decoder_create: 151| 10.2k|{ 152| 10.2k| int ret = 0; 153| 10.2k| DECODER *dec; 154| | 155| 10.2k| if (XVID_VERSION_MAJOR(create->version) != 1) /* v1.x.x */ ------------------ | | 63| 10.2k|#define XVID_VERSION_MAJOR(a) ((char)(((a)>>16) & 0xff)) ------------------ | Branch (155:7): [True: 0, False: 10.2k] ------------------ 156| 0| return XVID_ERR_VERSION; ------------------ | | 98| 0|#define XVID_ERR_VERSION -4 /* structure version not supported */ ------------------ 157| | 158| 10.2k| dec = xvid_malloc(sizeof(DECODER), CACHE_LINE); ------------------ | | 131| 10.2k|# define CACHE_LINE 64 ------------------ 159| 10.2k| if (dec == NULL) { ------------------ | Branch (159:7): [True: 0, False: 10.2k] ------------------ 160| 0| return XVID_ERR_MEMORY; ------------------ | | 96| 0|#define XVID_ERR_MEMORY -2 /* memory allocation error */ ------------------ 161| 0| } 162| | 163| 10.2k| memset(dec, 0, sizeof(DECODER)); 164| | 165| 10.2k| dec->mpeg_quant_matrices = xvid_malloc(sizeof(uint16_t) * 64 * 8, CACHE_LINE); ------------------ | | 131| 10.2k|# define CACHE_LINE 64 ------------------ 166| 10.2k| if (dec->mpeg_quant_matrices == NULL) { ------------------ | Branch (166:7): [True: 0, False: 10.2k] ------------------ 167| 0| xvid_free(dec); 168| 0| return XVID_ERR_MEMORY; ------------------ | | 96| 0|#define XVID_ERR_MEMORY -2 /* memory allocation error */ ------------------ 169| 0| } 170| | 171| 10.2k| create->handle = dec; 172| | 173| 10.2k| dec->width = MAX(0, create->width); ------------------ | | 35| 10.2k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 0, False: 10.2k] | | ------------------ ------------------ 174| 10.2k| dec->height = MAX(0, create->height); ------------------ | | 35| 10.2k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 0, False: 10.2k] | | ------------------ ------------------ 175| | 176| 10.2k| dec->num_threads = MAX(0, create->num_threads); ------------------ | | 35| 10.2k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 0, False: 10.2k] | | ------------------ ------------------ 177| | 178| 10.2k| image_null(&dec->cur); 179| 10.2k| image_null(&dec->refn[0]); 180| 10.2k| image_null(&dec->refn[1]); 181| 10.2k| image_null(&dec->tmp); 182| 10.2k| image_null(&dec->qtmp); 183| | 184| | /* image based GMC */ 185| 10.2k| image_null(&dec->gmc); 186| | 187| 10.2k| dec->mbs = NULL; 188| 10.2k| dec->last_mbs = NULL; 189| 10.2k| dec->qscale = NULL; 190| | 191| 10.2k| init_timer(); 192| 10.2k| init_postproc(&dec->postproc); 193| 10.2k| init_mpeg_matrix(dec->mpeg_quant_matrices); 194| | 195| | /* For B-frame support (used to save reference frame's time */ 196| 10.2k| dec->frames = 0; 197| 10.2k| dec->time = dec->time_base = dec->last_time_base = 0; 198| 10.2k| dec->low_delay = 0; 199| 10.2k| dec->packed_mode = 0; 200| 10.2k| dec->time_inc_resolution = 1; /* until VOL header says otherwise */ 201| 10.2k| dec->ver_id = 1; 202| | 203| 10.2k| if (create->fourcc == ((int)('X')|((int)('V')<<8)| ------------------ | Branch (203:7): [True: 0, False: 10.2k] ------------------ 204| 10.2k| ((int)('I')<<16)|((int)('D')<<24))) { /* XVID */ 205| 0| dec->bs_version = 0; /* Initially assume oldest xvid version */ 206| 0| } 207| 10.2k| else { 208| 10.2k| dec->bs_version = 0xffff; /* Initialize to very high value -> assume bugfree stream */ 209| 10.2k| } 210| | 211| 10.2k| dec->fixed_dimensions = (dec->width > 0 && dec->height > 0); ------------------ | Branch (211:28): [True: 0, False: 10.2k] | Branch (211:46): [True: 0, False: 0] ------------------ 212| | 213| 10.2k| ret = decoder_resize(dec); 214| 10.2k| if (ret == XVID_ERR_MEMORY) create->handle = NULL; ------------------ | | 96| 10.2k|#define XVID_ERR_MEMORY -2 /* memory allocation error */ ------------------ | Branch (214:7): [True: 0, False: 10.2k] ------------------ 215| | 216| 10.2k| return ret; 217| 10.2k|} decoder_destroy: 222| 10.2k|{ 223| 10.2k| xvid_free(dec->last_mbs); 224| 10.2k| xvid_free(dec->mbs); 225| 10.2k| xvid_free(dec->qscale); 226| | 227| | /* image based GMC */ 228| 10.2k| image_destroy(&dec->gmc, dec->edged_width, dec->edged_height); 229| | 230| 10.2k| image_destroy(&dec->refn[0], dec->edged_width, dec->edged_height); 231| 10.2k| image_destroy(&dec->refn[1], dec->edged_width, dec->edged_height); 232| 10.2k| image_destroy(&dec->tmp, dec->edged_width, dec->edged_height); 233| 10.2k| image_destroy(&dec->qtmp, dec->edged_width, dec->edged_height); 234| 10.2k| image_destroy(&dec->cur, dec->edged_width, dec->edged_height); 235| 10.2k| xvid_free(dec->mpeg_quant_matrices); 236| 10.2k| xvid_free(dec); 237| | 238| 10.2k| write_timer(); 239| 10.2k| return 0; 240| 10.2k|} decoder_decode: 1566| 69.9k|{ 1567| | 1568| 69.9k| Bitstream bs; 1569| 69.9k| uint32_t rounding = 0; 1570| 69.9k| uint32_t quant = 2; 1571| 69.9k| uint32_t fcode_forward = 0; 1572| 69.9k| uint32_t fcode_backward = 0; 1573| 69.9k| uint32_t intra_dc_threshold = 0; 1574| 69.9k| WARPPOINTS gmc_warp; 1575| 69.9k| int coding_type = -1; 1576| 69.9k| int success, output, seen_something; 1577| | 1578| 69.9k| if (XVID_VERSION_MAJOR(frame->version) != 1 || (stats && XVID_VERSION_MAJOR(stats->version) != 1)) /* v1.x.x */ ------------------ | | 63| 69.9k|#define XVID_VERSION_MAJOR(a) ((char)(((a)>>16) & 0xff)) ------------------ if (XVID_VERSION_MAJOR(frame->version) != 1 || (stats && XVID_VERSION_MAJOR(stats->version) != 1)) /* v1.x.x */ ------------------ | | 63| 69.9k|#define XVID_VERSION_MAJOR(a) ((char)(((a)>>16) & 0xff)) ------------------ | Branch (1578:7): [True: 0, False: 69.9k] | Branch (1578:51): [True: 69.9k, False: 0] | Branch (1578:60): [True: 0, False: 69.9k] ------------------ 1579| 0| return XVID_ERR_VERSION; ------------------ | | 98| 0|#define XVID_ERR_VERSION -4 /* structure version not supported */ ------------------ 1580| | 1581| 69.9k| start_global_timer(); 1582| 69.9k| memset((void *)&gmc_warp, 0, sizeof(WARPPOINTS)); 1583| | 1584| 69.9k| dec->low_delay_default = (frame->general & XVID_LOWDELAY); ------------------ | | 268| 69.9k|#define XVID_LOWDELAY (1<<0) /* lowdelay mode */ ------------------ 1585| 69.9k| if ((frame->general & XVID_DISCONTINUITY)) ------------------ | | 269| 69.9k|#define XVID_DISCONTINUITY (1<<1) /* indicates break in stream */ ------------------ | Branch (1585:7): [True: 0, False: 69.9k] ------------------ 1586| 0| dec->frames = 0; 1587| 69.9k| dec->out_frm = (frame->output.csp == XVID_CSP_SLICE) ? &frame->output : NULL; ------------------ | | 124| 69.9k|#define XVID_CSP_SLICE (1<<12) /* decoder only: 4:2:0 planar, per slice rendering */ ------------------ | Branch (1587:18): [True: 0, False: 69.9k] ------------------ 1588| | 1589| 69.9k| if(frame->length<0) { /* decoder flush */ ------------------ | Branch (1589:6): [True: 0, False: 69.9k] ------------------ 1590| 0| int ret; 1591| | /* if not decoding "low_delay/packed", and this isn't low_delay and 1592| | we have a reference frame, then outout the reference frame */ 1593| 0| if (!(dec->low_delay_default && dec->packed_mode) && !dec->low_delay && dec->frames>0) { ------------------ | Branch (1593:11): [True: 0, False: 0] | Branch (1593:37): [True: 0, False: 0] | Branch (1593:58): [True: 0, False: 0] | Branch (1593:77): [True: 0, False: 0] ------------------ 1594| 0| decoder_output(dec, &dec->refn[0], dec->last_mbs, frame, stats, dec->last_coding_type, quant); 1595| 0| dec->frames = 0; 1596| 0| ret = 0; 1597| 0| } else { 1598| 0| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 0|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1598:11): [True: 0, False: 0] ------------------ 1599| 0| ret = XVID_ERR_END; ------------------ | | 99| 0|#define XVID_ERR_END -5 /* encoder only; end of stream reached */ ------------------ 1600| 0| } 1601| | 1602| 0| emms(); 1603| 0| stop_global_timer(); 1604| 0| return ret; 1605| 0| } 1606| | 1607| 69.9k| BitstreamInit(&bs, frame->bitstream, frame->length); 1608| | 1609| | /* XXX: 0x7f is only valid whilst decoding vfw xvid/divx5 avi's */ 1610| 69.9k| if(dec->low_delay_default && frame->length == 1 && BitstreamShowBits(&bs, 8) == 0x7f) ------------------ | Branch (1610:6): [True: 0, False: 69.9k] | Branch (1610:32): [True: 0, False: 0] | Branch (1610:54): [True: 0, False: 0] ------------------ 1611| 0| { 1612| 0| image_output(&dec->refn[0], dec->width, dec->height, dec->edged_width, 1613| 0| (uint8_t**)frame->output.plane, frame->output.stride, frame->output.csp, dec->interlacing); 1614| 0| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 0|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1614:9): [True: 0, False: 0] ------------------ 1615| 0| emms(); 1616| 0| return 1; /* one byte consumed */ 1617| 0| } 1618| | 1619| 69.9k| success = 0; 1620| 69.9k| output = 0; 1621| 69.9k| seen_something = 0; 1622| | 1623| 1.86M|repeat: 1624| | 1625| 1.86M| coding_type = BitstreamReadHeaders(&bs, dec, &rounding, 1626| 1.86M| &quant, &fcode_forward, &fcode_backward, &intra_dc_threshold, &gmc_warp); 1627| | 1628| 1.86M| DPRINTF(XVID_DEBUG_HEADER, "coding_type=%i, packed=%i, time=%" ------------------ | | 198| 1.86M|#define XVID_DEBUG_HEADER (1<< 2) ------------------ 1629| |#if defined(_MSC_VER) 1630| | "I64" 1631| |#else 1632| 1.86M| "ll" 1633| 1.86M|#endif 1634| 1.86M| "i, time_pp=%i, time_bp=%i\n", 1635| 1.86M| coding_type, dec->packed_mode, dec->time, dec->time_pp, dec->time_bp); 1636| | 1637| 1.86M| if (coding_type == -1) { /* nothing */ ------------------ | Branch (1637:7): [True: 18.3k, False: 1.85M] ------------------ 1638| 18.3k| if (success) goto done; ------------------ | Branch (1638:9): [True: 0, False: 18.3k] ------------------ 1639| 18.3k| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 18.3k|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1639:9): [True: 18.3k, False: 0] ------------------ 1640| 18.3k| emms(); 1641| 18.3k| return BitstreamPos(&bs)/8; 1642| 18.3k| } 1643| | 1644| 1.85M| if (coding_type == -2 || coding_type == -3) { /* vol and/or resize */ ------------------ | Branch (1644:7): [True: 21.6k, False: 1.82M] | Branch (1644:28): [True: 12.9k, False: 1.81M] ------------------ 1645| | 1646| 34.5k| if (coding_type == -3) ------------------ | Branch (1646:9): [True: 12.9k, False: 21.6k] ------------------ 1647| 12.9k| if (decoder_resize(dec)) return XVID_ERR_MEMORY; ------------------ | | 96| 0|#define XVID_ERR_MEMORY -2 /* memory allocation error */ ------------------ | Branch (1647:11): [True: 0, False: 12.9k] ------------------ 1648| | 1649| 34.5k| if(stats) { ------------------ | Branch (1649:8): [True: 34.5k, False: 0] ------------------ 1650| 34.5k| stats->type = XVID_TYPE_VOL; ------------------ | | 165| 34.5k|#define XVID_TYPE_VOL -1 /* decoder only: vol was decoded */ ------------------ 1651| 34.5k| stats->data.vol.general = 0; 1652| 34.5k| stats->data.vop.general = 0; 1653| 34.5k| if (dec->interlacing) { ------------------ | Branch (1653:8): [True: 10.1k, False: 24.3k] ------------------ 1654| 10.1k| stats->data.vol.general |= XVID_VOL_INTERLACING; ------------------ | | 661| 10.1k|#define XVID_VOL_INTERLACING (1<<5) /* enable interlaced encoding */ ------------------ 1655| 10.1k| if (dec->top_field_first) { ------------------ | Branch (1655:7): [True: 929, False: 9.22k] ------------------ 1656| 929| stats->data.vop.general |= XVID_VOP_TOPFIELDFIRST; ------------------ | | 686| 929|#define XVID_VOP_TOPFIELDFIRST (1<< 9) /* set top-field-first flag */ ------------------ 1657| 929| } 1658| 10.1k| } 1659| 34.5k| stats->data.vol.width = dec->width; 1660| 34.5k| stats->data.vol.height = dec->height; 1661| 34.5k| stats->data.vol.par = dec->aspect_ratio; 1662| 34.5k| stats->data.vol.par_width = dec->par_width; 1663| 34.5k| stats->data.vol.par_height = dec->par_height; 1664| 34.5k| emms(); 1665| 34.5k| return BitstreamPos(&bs)/8; /* number of bytes consumed */ 1666| 34.5k| } 1667| 0| goto repeat; 1668| 34.5k| } 1669| | 1670| 1.81M| if((dec->frames == 0 && coding_type != I_VOP) || (!dec->width || !dec->height)) { ------------------ | | 104| 1.80M|#define I_VOP 0 ------------------ | Branch (1670:7): [True: 1.80M, False: 10.1k] | Branch (1670:27): [True: 1.77M, False: 27.7k] | Branch (1670:53): [True: 20.7k, False: 17.1k] | Branch (1670:68): [True: 0, False: 17.1k] ------------------ 1671| | /* 1st frame is not an i-vop */ 1672| 1.79M| goto repeat; 1673| 1.79M| } 1674| | 1675| 17.1k| dec->p_bmv.x = dec->p_bmv.y = dec->p_fmv.x = dec->p_fmv.y = 0; /* init pred vector to 0 */ 1676| | 1677| | /* packed_mode: special-N_VOP treament */ 1678| 17.1k| if (dec->packed_mode && coding_type == N_VOP) { ------------------ | | 108| 669|#define N_VOP 4 ------------------ | Branch (1678:7): [True: 669, False: 16.4k] | Branch (1678:27): [True: 254, False: 415] ------------------ 1679| 254| if (dec->low_delay_default && dec->frames > 0) { ------------------ | Branch (1679:9): [True: 0, False: 254] | Branch (1679:35): [True: 0, False: 0] ------------------ 1680| 0| decoder_output(dec, &dec->refn[0], dec->last_mbs, frame, stats, dec->last_coding_type, quant); 1681| 0| output = 1; 1682| 0| } 1683| | /* ignore otherwise */ 1684| 16.8k| } else if (coding_type != B_VOP) { ------------------ | | 106| 16.8k|#define B_VOP 2 ------------------ | Branch (1684:14): [True: 10.5k, False: 6.33k] ------------------ 1685| 10.5k| switch(coding_type) { ------------------ | Branch (1685:12): [True: 0, False: 10.5k] ------------------ 1686| 7.00k| case I_VOP : ------------------ | | 104| 7.00k|#define I_VOP 0 ------------------ | Branch (1686:5): [True: 7.00k, False: 3.50k] ------------------ 1687| 7.00k| decoder_iframe(dec, &bs, quant, intra_dc_threshold); 1688| 7.00k| break; 1689| 988| case P_VOP : ------------------ | | 105| 988|#define P_VOP 1 ------------------ | Branch (1689:5): [True: 988, False: 9.51k] ------------------ 1690| 988| decoder_pframe(dec, &bs, rounding, quant, 1691| 988| fcode_forward, intra_dc_threshold, NULL); 1692| 988| break; 1693| 2.47k| case S_VOP : ------------------ | | 107| 2.47k|#define S_VOP 3 ------------------ | Branch (1693:5): [True: 2.47k, False: 8.03k] ------------------ 1694| 2.47k| decoder_pframe(dec, &bs, rounding, quant, 1695| 2.47k| fcode_forward, intra_dc_threshold, &gmc_warp); 1696| 2.47k| break; 1697| 42| case N_VOP : ------------------ | | 108| 42|#define N_VOP 4 ------------------ | Branch (1697:5): [True: 42, False: 10.4k] ------------------ 1698| | /* XXX: not_coded vops are not used for forward prediction */ 1699| | /* we should not swap(last_mbs,mbs) */ 1700| 42| image_copy(&dec->cur, &dec->refn[0], dec->edged_width, dec->height); 1701| 42| SWAP(MACROBLOCK *, dec->mbs, dec->last_mbs); /* it will be swapped back */ ------------------ | | 264| 42|#define SWAP(_T_,A,B) { _T_ tmp = A; A = B; B = tmp; } ------------------ 1702| 42| break; 1703| 10.5k| } 1704| | 1705| | /* note: for packed_mode, output is performed when the special-N_VOP is decoded */ 1706| 10.5k| if (!(dec->low_delay_default && dec->packed_mode)) { ------------------ | Branch (1706:11): [True: 0, False: 10.5k] | Branch (1706:37): [True: 0, False: 0] ------------------ 1707| 10.5k| if(dec->low_delay) { ------------------ | Branch (1707:10): [True: 286, False: 10.2k] ------------------ 1708| 286| decoder_output(dec, &dec->cur, dec->mbs, frame, stats, coding_type, quant); 1709| 286| output = 1; 1710| 10.2k| } else if (dec->frames > 0) { /* is the reference frame valid? */ ------------------ | Branch (1710:18): [True: 3.52k, False: 6.69k] ------------------ 1711| | /* output the reference frame */ 1712| 3.52k| decoder_output(dec, &dec->refn[0], dec->last_mbs, frame, stats, dec->last_coding_type, quant); 1713| 3.52k| output = 1; 1714| 3.52k| } 1715| 10.5k| } 1716| | 1717| 10.5k| image_swap(&dec->refn[0], &dec->refn[1]); 1718| 10.5k| dec->is_edged[1] = dec->is_edged[0]; 1719| 10.5k| image_swap(&dec->cur, &dec->refn[0]); 1720| 10.5k| dec->is_edged[0] = 0; 1721| 10.5k| SWAP(MACROBLOCK *, dec->mbs, dec->last_mbs); ------------------ | | 264| 10.5k|#define SWAP(_T_,A,B) { _T_ tmp = A; A = B; B = tmp; } ------------------ 1722| 10.5k| dec->last_coding_type = coding_type; 1723| | 1724| 10.5k| dec->frames++; 1725| 10.5k| seen_something = 1; 1726| | 1727| 10.5k| } else { /* B_VOP */ 1728| | 1729| 6.33k| if (dec->low_delay) { ------------------ | Branch (1729:9): [True: 86, False: 6.25k] ------------------ 1730| 86| DPRINTF(XVID_DEBUG_ERROR, "warning: bvop found in low_delay==1 stream\n"); ------------------ | | 196| 86|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1731| 86| dec->low_delay = 0; 1732| 86| } 1733| | 1734| 6.33k| if (dec->frames < 2) { ------------------ | Branch (1734:9): [True: 2.38k, False: 3.95k] ------------------ 1735| | /* attemping to decode a bvop without atleast 2 reference frames */ 1736| 2.38k| image_printf(&dec->cur, dec->edged_width, dec->height, 16, 16, 1737| 2.38k| "broken b-frame, mising ref frames"); 1738| 2.38k| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 2.38k|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1738:11): [True: 2.38k, False: 0] ------------------ 1739| 3.95k| } else if (dec->time_pp <= dec->time_bp) { ------------------ | Branch (1739:16): [True: 1.71k, False: 2.24k] ------------------ 1740| | /* this occurs when dx50_bvop_compatibility==0 sequences are 1741| | decoded in vfw. */ 1742| 1.71k| image_printf(&dec->cur, dec->edged_width, dec->height, 16, 16, 1743| 1.71k| "broken b-frame, tpp=%i tbp=%i", dec->time_pp, dec->time_bp); 1744| 1.71k| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 1.71k|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1744:11): [True: 1.71k, False: 0] ------------------ 1745| 2.24k| } else { 1746| 2.24k| decoder_bframe(dec, &bs, quant, fcode_forward, fcode_backward); 1747| 2.24k| decoder_output(dec, &dec->cur, dec->mbs, frame, stats, coding_type, quant); 1748| 2.24k| } 1749| | 1750| 6.33k| output = 1; 1751| 6.33k| dec->frames++; 1752| 6.33k| } 1753| | 1754| |#if 0 /* Avoids to read to much data because of 32bit reads in our BS functions */ 1755| | BitstreamByteAlign(&bs); 1756| |#endif 1757| | 1758| | /* low_delay_default mode: repeat in packed_mode */ 1759| 17.1k| if (dec->low_delay_default && dec->packed_mode && output == 0 && success == 0) { ------------------ | Branch (1759:7): [True: 0, False: 17.1k] | Branch (1759:33): [True: 0, False: 0] | Branch (1759:53): [True: 0, False: 0] | Branch (1759:68): [True: 0, False: 0] ------------------ 1760| 0| success = 1; 1761| 0| goto repeat; 1762| 0| } 1763| | 1764| 17.1k|done : 1765| | 1766| | /* if we reach here without outputing anything _and_ 1767| | the calling application has specified low_delay_default, 1768| | we *must* output something. 1769| | this always occurs on the first call to decode() call 1770| | when bframes are present in the bitstream. it may also 1771| | occur if no vops were seen in the bitstream 1772| | 1773| | if packed_mode is enabled, then we output the recently 1774| | decoded frame (the very first ivop). otherwise we have 1775| | nothing to display, and therefore output a black screen. 1776| | */ 1777| 17.1k| if (dec->low_delay_default && output == 0) { ------------------ | Branch (1777:7): [True: 0, False: 17.1k] | Branch (1777:33): [True: 0, False: 0] ------------------ 1778| 0| if (dec->packed_mode && seen_something) { ------------------ | Branch (1778:9): [True: 0, False: 0] | Branch (1778:29): [True: 0, False: 0] ------------------ 1779| 0| decoder_output(dec, &dec->refn[0], dec->last_mbs, frame, stats, dec->last_coding_type, quant); 1780| 0| } else { 1781| 0| image_clear(&dec->cur, dec->width, dec->height, dec->edged_width, 0, 128, 128); 1782| 0| decoder_output(dec, &dec->cur, NULL, frame, stats, P_VOP, quant); ------------------ | | 105| 0|#define P_VOP 1 ------------------ 1783| 0| if (stats) stats->type = XVID_TYPE_NOTHING; ------------------ | | 166| 0|#define XVID_TYPE_NOTHING 0 /* decoder only (encoder stats): nothing was decoded/encoded */ ------------------ | Branch (1783:11): [True: 0, False: 0] ------------------ 1784| 0| } 1785| 0| } 1786| | 1787| 17.1k| emms(); 1788| 17.1k| stop_global_timer(); 1789| | 1790| 17.1k| return (BitstreamPos(&bs)+7)/8; /* number of bytes consumed */ 1791| 17.1k|} decoder.c:decoder_resize: 70| 23.2k|{ 71| | /* free existing */ 72| 23.2k| image_destroy(&dec->cur, dec->edged_width, dec->edged_height); 73| 23.2k| image_destroy(&dec->refn[0], dec->edged_width, dec->edged_height); 74| 23.2k| image_destroy(&dec->refn[1], dec->edged_width, dec->edged_height); 75| 23.2k| image_destroy(&dec->tmp, dec->edged_width, dec->edged_height); 76| 23.2k| image_destroy(&dec->qtmp, dec->edged_width, dec->edged_height); 77| | 78| 23.2k| image_destroy(&dec->gmc, dec->edged_width, dec->edged_height); 79| | 80| 23.2k| image_null(&dec->cur); 81| 23.2k| image_null(&dec->refn[0]); 82| 23.2k| image_null(&dec->refn[1]); 83| 23.2k| image_null(&dec->tmp); 84| 23.2k| image_null(&dec->qtmp); 85| 23.2k| image_null(&dec->gmc); 86| | 87| | 88| 23.2k| xvid_free(dec->last_mbs); 89| 23.2k| xvid_free(dec->mbs); 90| 23.2k| xvid_free(dec->qscale); 91| 23.2k| dec->last_mbs = NULL; 92| 23.2k| dec->mbs = NULL; 93| 23.2k| dec->qscale = NULL; 94| | 95| | /* realloc */ 96| 23.2k| dec->mb_width = (dec->width + 15) / 16; 97| 23.2k| dec->mb_height = (dec->height + 15) / 16; 98| | 99| 23.2k| dec->edged_width = 16 * dec->mb_width + 2 * EDGE_SIZE; ------------------ | | 36| 23.2k|#define EDGE_SIZE 64 ------------------ 100| 23.2k| dec->edged_height = 16 * dec->mb_height + 2 * EDGE_SIZE; ------------------ | | 36| 23.2k|#define EDGE_SIZE 64 ------------------ 101| | 102| 23.2k| if ( image_create(&dec->cur, dec->edged_width, dec->edged_height) ------------------ | Branch (102:9): [True: 0, False: 23.2k] ------------------ 103| 23.2k| || image_create(&dec->refn[0], dec->edged_width, dec->edged_height) ------------------ | Branch (103:9): [True: 0, False: 23.2k] ------------------ 104| 23.2k| || image_create(&dec->refn[1], dec->edged_width, dec->edged_height) /* Support B-frame to reference last 2 frame */ ------------------ | Branch (104:9): [True: 0, False: 23.2k] ------------------ 105| 23.2k| || image_create(&dec->tmp, dec->edged_width, dec->edged_height) ------------------ | Branch (105:9): [True: 0, False: 23.2k] ------------------ 106| 23.2k| || image_create(&dec->qtmp, dec->edged_width, dec->edged_height) ------------------ | Branch (106:9): [True: 0, False: 23.2k] ------------------ 107| 23.2k| || image_create(&dec->gmc, dec->edged_width, dec->edged_height) ) ------------------ | Branch (107:10): [True: 0, False: 23.2k] ------------------ 108| 0| goto memory_error; 109| | 110| 23.2k| dec->mbs = 111| 23.2k| xvid_malloc(sizeof(MACROBLOCK) * dec->mb_width * dec->mb_height, 112| 23.2k| CACHE_LINE); ------------------ | | 131| 23.2k|# define CACHE_LINE 64 ------------------ 113| 23.2k| if (dec->mbs == NULL) ------------------ | Branch (113:6): [True: 0, False: 23.2k] ------------------ 114| 0| goto memory_error; 115| 23.2k| memset(dec->mbs, 0, sizeof(MACROBLOCK) * dec->mb_width * dec->mb_height); 116| | 117| | /* For skip MB flag */ 118| 23.2k| dec->last_mbs = 119| 23.2k| xvid_malloc(sizeof(MACROBLOCK) * dec->mb_width * dec->mb_height, 120| 23.2k| CACHE_LINE); ------------------ | | 131| 23.2k|# define CACHE_LINE 64 ------------------ 121| 23.2k| if (dec->last_mbs == NULL) ------------------ | Branch (121:6): [True: 0, False: 23.2k] ------------------ 122| 0| goto memory_error; 123| 23.2k| memset(dec->last_mbs, 0, sizeof(MACROBLOCK) * dec->mb_width * dec->mb_height); 124| | 125| | /* nothing happens if that fails */ 126| 23.2k| dec->qscale = 127| 23.2k| xvid_malloc(sizeof(int) * dec->mb_width * dec->mb_height, CACHE_LINE); ------------------ | | 131| 23.2k|# define CACHE_LINE 64 ------------------ 128| | 129| 23.2k| if (dec->qscale) ------------------ | Branch (129:6): [True: 23.2k, False: 0] ------------------ 130| 23.2k| memset(dec->qscale, 0, sizeof(int) * dec->mb_width * dec->mb_height); 131| | 132| 23.2k| return 0; 133| | 134| 0|memory_error: 135| | /* Most structures were deallocated / nullifieded, so it should be safe */ 136| | /* decoder_destroy(dec) minus the write_timer */ 137| 0| xvid_free(dec->mbs); 138| 0| image_destroy(&dec->cur, dec->edged_width, dec->edged_height); 139| 0| image_destroy(&dec->refn[0], dec->edged_width, dec->edged_height); 140| 0| image_destroy(&dec->refn[1], dec->edged_width, dec->edged_height); 141| 0| image_destroy(&dec->tmp, dec->edged_width, dec->edged_height); 142| 0| image_destroy(&dec->qtmp, dec->edged_width, dec->edged_height); 143| | 144| 0| xvid_free(dec); 145| 0| return XVID_ERR_MEMORY; ------------------ | | 96| 0|#define XVID_ERR_MEMORY -2 /* memory allocation error */ ------------------ 146| 23.2k|} decoder.c:decoder_output: 1524| 6.05k|{ 1525| 6.05k| const int brightness = XVID_VERSION_MINOR(frame->version) >= 1 ? frame->brightness : 0; ------------------ | | 64| 6.05k|#define XVID_VERSION_MINOR(a) ((char)(((a)>> 8) & 0xff)) ------------------ | Branch (1525:26): [True: 6.05k, False: 0] ------------------ 1526| | 1527| 6.05k| if (dec->cartoon_mode) ------------------ | Branch (1527:7): [True: 12, False: 6.04k] ------------------ 1528| 12| frame->general &= ~XVID_FILMEFFECT; ------------------ | | 272| 12|#define XVID_FILMEFFECT (1<<4) /* adds film grain */ ------------------ 1529| | 1530| 6.05k| if ((frame->general & (XVID_DEBLOCKY|XVID_DEBLOCKUV|XVID_FILMEFFECT) || brightness!=0) ------------------ | | 270| 6.05k|#define XVID_DEBLOCKY (1<<2) /* perform luma deblocking */ ------------------ if ((frame->general & (XVID_DEBLOCKY|XVID_DEBLOCKUV|XVID_FILMEFFECT) || brightness!=0) ------------------ | | 271| 6.05k|#define XVID_DEBLOCKUV (1<<3) /* perform chroma deblocking */ ------------------ if ((frame->general & (XVID_DEBLOCKY|XVID_DEBLOCKUV|XVID_FILMEFFECT) || brightness!=0) ------------------ | | 272| 6.05k|#define XVID_FILMEFFECT (1<<4) /* adds film grain */ ------------------ | Branch (1530:8): [True: 0, False: 6.05k] | Branch (1530:75): [True: 0, False: 6.05k] ------------------ 1531| 6.05k| && mbs != NULL) /* post process */ ------------------ | Branch (1531:8): [True: 0, False: 0] ------------------ 1532| 0| { 1533| | /* note: image is stored to tmp */ 1534| 0| image_copy(&dec->tmp, img, dec->edged_width, dec->height); 1535| 0| image_postproc(&dec->postproc, &dec->tmp, dec->edged_width, 1536| 0| mbs, dec->mb_width, dec->mb_height, dec->mb_width, 1537| 0| frame->general, brightness, dec->frames, (coding_type == B_VOP), dec->num_threads); ------------------ | | 106| 0|#define B_VOP 2 ------------------ 1538| 0| img = &dec->tmp; 1539| 0| } 1540| | 1541| 6.05k| if ((frame->output.csp == XVID_CSP_INTERNAL) || ------------------ | | 125| 6.05k|#define XVID_CSP_INTERNAL (1<<13) /* decoder only: 4:2:0 planar, returns ptrs to internal buffers */ ------------------ | Branch (1541:7): [True: 0, False: 6.05k] ------------------ 1542| 6.05k| ((frame->output.plane[0] != NULL) && (frame->output.stride[0] >= dec->width))) { ------------------ | Branch (1542:8): [True: 6.05k, False: 0] | Branch (1542:44): [True: 6.05k, False: 0] ------------------ 1543| 6.05k| image_output(img, dec->width, dec->height, 1544| 6.05k| dec->edged_width, (uint8_t**)frame->output.plane, frame->output.stride, 1545| 6.05k| frame->output.csp, dec->interlacing); 1546| 6.05k| } 1547| | 1548| 6.05k| if (stats) { ------------------ | Branch (1548:7): [True: 6.05k, False: 0] ------------------ 1549| 6.05k| stats->type = coding2type(coding_type); 1550| 6.05k| stats->data.vop.time_base = (int)dec->time_base; 1551| 6.05k| stats->data.vop.time_increment = 0; /* XXX: todo */ 1552| 6.05k| stats->data.vop.qscale_stride = dec->mb_width; 1553| 6.05k| stats->data.vop.qscale = dec->qscale; 1554| 6.05k| if (stats->data.vop.qscale != NULL && mbs != NULL) { ------------------ | Branch (1554:9): [True: 6.05k, False: 0] | Branch (1554:43): [True: 6.05k, False: 0] ------------------ 1555| 6.05k| unsigned int i; 1556| 43.4M| for (i = 0; i < dec->mb_width*dec->mb_height; i++) ------------------ | Branch (1556:19): [True: 43.4M, False: 6.05k] ------------------ 1557| 43.4M| stats->data.vop.qscale[i] = mbs[i].quant; 1558| 6.05k| } else 1559| 0| stats->data.vop.qscale = NULL; 1560| 6.05k| } 1561| 6.05k|} decoder.c:decoder_iframe: 735| 7.00k|{ 736| 7.00k| uint32_t bound; 737| 7.00k| uint32_t x, y; 738| 7.00k| const uint32_t mb_width = dec->mb_width; 739| 7.00k| const uint32_t mb_height = dec->mb_height; 740| | 741| 7.00k| bound = 0; 742| | 743| 182k| for (y = 0; y < mb_height; y++) { ------------------ | Branch (743:15): [True: 175k, False: 7.00k] ------------------ 744| 27.6M| for (x = 0; x < mb_width; x++) { ------------------ | Branch (744:17): [True: 27.4M, False: 175k] ------------------ 745| 27.4M| MACROBLOCK *mb; 746| 27.4M| uint32_t mcbpc; 747| 27.4M| uint32_t cbpc; 748| 27.4M| uint32_t acpred_flag; 749| 27.4M| uint32_t cbpy; 750| 27.4M| uint32_t cbp; 751| | 752| 27.4M| while (BitstreamShowBits(bs, 9) == 1) ------------------ | Branch (752:14): [True: 1.49k, False: 27.4M] ------------------ 753| 1.49k| BitstreamSkip(bs, 9); 754| | 755| 27.4M| if (check_resync_marker(bs, 0)) ------------------ | Branch (755:11): [True: 12.6k, False: 27.4M] ------------------ 756| 12.6k| { 757| 12.6k| bound = read_video_packet_header(bs, dec, 0, 758| 12.6k| &quant, NULL, NULL, &intra_dc_threshold); 759| 12.6k| x = bound % mb_width; 760| 12.6k| y = MIN((bound / mb_width), (mb_height-1)); ------------------ | | 34| 12.6k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 8.90k, False: 3.71k] | | ------------------ ------------------ 761| 12.6k| } 762| 27.4M| mb = &dec->mbs[y * dec->mb_width + x]; 763| | 764| 27.4M| DPRINTF(XVID_DEBUG_MB, "macroblock (%i,%i) %08x\n", x, y, BitstreamShowBits(bs, 32)); ------------------ | | 200| 27.4M|#define XVID_DEBUG_MB (1<< 4) ------------------ 765| | 766| 27.4M| mcbpc = get_mcbpc_intra(bs); 767| 27.4M| mb->mode = mcbpc & 7; 768| 27.4M| cbpc = (mcbpc >> 4); 769| | 770| 27.4M| acpred_flag = BitstreamGetBit(bs); 771| | 772| 27.4M| cbpy = get_cbpy(bs, 1); 773| 27.4M| cbp = (cbpy << 2) | cbpc; 774| | 775| 27.4M| if (mb->mode == MODE_INTRA_Q) { ------------------ | | 38| 27.4M|#define MODE_INTRA_Q 4 ------------------ | Branch (775:11): [True: 9.80k, False: 27.4M] ------------------ 776| 9.80k| quant += dquant_table[BitstreamGetBits(bs, 2)]; 777| 9.80k| if (quant > 31) { ------------------ | Branch (777:13): [True: 2.15k, False: 7.65k] ------------------ 778| 2.15k| quant = 31; 779| 7.65k| } else if (quant < 1) { ------------------ | Branch (779:20): [True: 1.77k, False: 5.87k] ------------------ 780| 1.77k| quant = 1; 781| 1.77k| } 782| 9.80k| } 783| 27.4M| mb->quant = quant; 784| 27.4M| mb->mvs[0].x = mb->mvs[0].y = 785| 27.4M| mb->mvs[1].x = mb->mvs[1].y = 786| 27.4M| mb->mvs[2].x = mb->mvs[2].y = 787| 27.4M| mb->mvs[3].x = mb->mvs[3].y =0; 788| | 789| 27.4M| if (dec->interlacing) { ------------------ | Branch (789:11): [True: 19.7M, False: 7.69M] ------------------ 790| 19.7M| mb->field_dct = BitstreamGetBit(bs); 791| 19.7M| DPRINTF(XVID_DEBUG_MB,"deci: field_dct: %i\n", mb->field_dct); ------------------ | | 200| 19.7M|#define XVID_DEBUG_MB (1<< 4) ------------------ 792| 19.7M| } 793| | 794| 27.4M| decoder_mbintra(dec, mb, x, y, acpred_flag, cbp, bs, quant, 795| 27.4M| intra_dc_threshold, bound); 796| | 797| 27.4M| } 798| 175k| if(dec->out_frm) ------------------ | Branch (798:8): [True: 0, False: 175k] ------------------ 799| 0| output_slice(&dec->cur, dec->edged_width,dec->width,dec->out_frm,0,y,mb_width); 800| 175k| } 801| | 802| 7.00k|} decoder.c:decoder_mbintra: 258| 45.4M|{ 259| | 260| 45.4M| DECLARE_ALIGNED_MATRIX(block, 6, 64, int16_t, CACHE_LINE); ------------------ | | 287| 45.4M| type name##_storage[(sizex)*(sizey)+(alignment)-1]; \ | | 288| 45.4M|type * name = (type *) (((ptr_t) name##_storage+(alignment - 1)) & ~((ptr_t)(alignment)-1)) ------------------ 261| 45.4M| DECLARE_ALIGNED_MATRIX(data, 6, 64, int16_t, CACHE_LINE); ------------------ | | 287| 45.4M| type name##_storage[(sizex)*(sizey)+(alignment)-1]; \ | | 288| 45.4M|type * name = (type *) (((ptr_t) name##_storage+(alignment - 1)) & ~((ptr_t)(alignment)-1)) ------------------ 262| | 263| 45.4M| uint32_t stride = dec->edged_width; 264| 45.4M| uint32_t stride2 = stride / 2; 265| 45.4M| uint32_t next_block = stride * 8; 266| 45.4M| uint32_t i; 267| 45.4M| uint32_t iQuant = MAX(1, pMB->quant); ------------------ | | 35| 45.4M|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 1.15M, False: 44.2M] | | ------------------ ------------------ 268| 45.4M| uint8_t *pY_Cur, *pU_Cur, *pV_Cur; 269| | 270| 45.4M| pY_Cur = dec->cur.y + (y_pos << 4) * stride + (x_pos << 4); 271| 45.4M| pU_Cur = dec->cur.u + (y_pos << 3) * stride2 + (x_pos << 3); 272| 45.4M| pV_Cur = dec->cur.v + (y_pos << 3) * stride2 + (x_pos << 3); 273| | 274| 45.4M| memset(block, 0, 6 * 64 * sizeof(int16_t)); /* clear */ 275| | 276| 318M| for (i = 0; i < 6; i++) { ------------------ | Branch (276:15): [True: 272M, False: 45.4M] ------------------ 277| 272M| uint32_t iDcScaler = get_dc_scaler(iQuant, i < 4); 278| 272M| int16_t predictors[8]; 279| 272M| int start_coeff; 280| | 281| 272M| start_timer(); 282| 272M| predict_acdc(dec->mbs, x_pos, y_pos, dec->mb_width, i, &block[i * 64], 283| 272M| iQuant, iDcScaler, predictors, bound); 284| 272M| if (!acpred_flag) { ------------------ | Branch (284:9): [True: 272M, False: 368k] ------------------ 285| 272M| pMB->acpred_directions[i] = 0; 286| 272M| } 287| 272M| stop_prediction_timer(); 288| | 289| 272M| if (quant < intra_dc_threshold) { ------------------ | Branch (289:9): [True: 133M, False: 139M] ------------------ 290| 133M| int dc_size; 291| 133M| int dc_dif; 292| | 293| 133M| dc_size = i < 4 ? get_dc_size_lum(bs) : get_dc_size_chrom(bs); ------------------ | Branch (293:17): [True: 89.0M, False: 44.5M] ------------------ 294| 133M| dc_dif = dc_size ? get_dc_dif(bs, dc_size) : 0; ------------------ | Branch (294:16): [True: 44.7M, False: 88.8M] ------------------ 295| | 296| 133M| if (dc_size > 8) { ------------------ | Branch (296:11): [True: 117k, False: 133M] ------------------ 297| 117k| BitstreamSkip(bs, 1); /* marker */ 298| 117k| } 299| | 300| 133M| block[i * 64 + 0] = dc_dif; 301| 133M| start_coeff = 1; 302| | 303| 133M| DPRINTF(XVID_DEBUG_COEFF,"block[0] %i\n", dc_dif); ------------------ | | 201| 133M|#define XVID_DEBUG_COEFF (1<< 5) ------------------ 304| 139M| } else { 305| 139M| start_coeff = 0; 306| 139M| } 307| | 308| 272M| start_timer(); 309| 272M| if (cbp & (1 << (5 - i))) /* coded */ ------------------ | Branch (309:9): [True: 272M, False: 299k] ------------------ 310| 272M| { 311| 272M| int direction = dec->alternate_vertical_scan ? ------------------ | Branch (311:23): [True: 91.6M, False: 180M] ------------------ 312| 180M| 2 : pMB->acpred_directions[i]; 313| | 314| 272M| get_intra_block(bs, &block[i * 64], direction, start_coeff); 315| 272M| } 316| 272M| stop_coding_timer(); 317| | 318| 272M| start_timer(); 319| 272M| add_acdc(pMB, i, &block[i * 64], iDcScaler, predictors, dec->bs_version); 320| 272M| stop_prediction_timer(); 321| | 322| 272M| start_timer(); 323| 272M| if (dec->quant_type == 0) { ------------------ | Branch (323:9): [True: 158M, False: 114M] ------------------ 324| 158M| dequant_h263_intra(&data[i * 64], &block[i * 64], iQuant, iDcScaler, dec->mpeg_quant_matrices); 325| 158M| } else { 326| 114M| dequant_mpeg_intra(&data[i * 64], &block[i * 64], iQuant, iDcScaler, dec->mpeg_quant_matrices); 327| 114M| } 328| 272M| stop_iquant_timer(); 329| | 330| 272M| start_timer(); 331| 272M| idct((short * const)&data[i * 64]); 332| 272M| stop_idct_timer(); 333| | 334| 272M| } 335| | 336| 45.4M| if (dec->interlacing && pMB->field_dct) { ------------------ | Branch (336:7): [True: 33.5M, False: 11.8M] | Branch (336:27): [True: 68.3k, False: 33.4M] ------------------ 337| 68.3k| next_block = stride; 338| 68.3k| stride *= 2; 339| 68.3k| } 340| | 341| 45.4M| start_timer(); 342| 45.4M| transfer_16to8copy(pY_Cur, &data[0 * 64], stride); 343| 45.4M| transfer_16to8copy(pY_Cur + 8, &data[1 * 64], stride); 344| 45.4M| transfer_16to8copy(pY_Cur + next_block, &data[2 * 64], stride); 345| 45.4M| transfer_16to8copy(pY_Cur + 8 + next_block, &data[3 * 64], stride); 346| 45.4M| transfer_16to8copy(pU_Cur, &data[4 * 64], stride2); 347| 45.4M| transfer_16to8copy(pV_Cur, &data[5 * 64], stride2); 348| 45.4M| stop_transfer_timer(); 349| 45.4M|} decoder.c:decoder_pframe: 948| 3.45k|{ 949| 3.45k| uint32_t x, y; 950| 3.45k| uint32_t bound; 951| 3.45k| int cp_mb, st_mb; 952| 3.45k| const uint32_t mb_width = dec->mb_width; 953| 3.45k| const uint32_t mb_height = dec->mb_height; 954| | 955| 3.45k| if (!dec->is_edged[0]) { ------------------ | Branch (955:7): [True: 3.45k, False: 0] ------------------ 956| 3.45k| start_timer(); 957| 3.45k| image_setedges(&dec->refn[0], dec->edged_width, dec->edged_height, 958| 3.45k| dec->width, dec->height, dec->bs_version); 959| 3.45k| dec->is_edged[0] = 1; 960| 3.45k| stop_edges_timer(); 961| 3.45k| } 962| | 963| 3.45k| if (gmc_warp) { ------------------ | Branch (963:7): [True: 2.47k, False: 988] ------------------ 964| | /* accuracy: 0==1/2, 1=1/4, 2=1/8, 3=1/16 */ 965| 2.47k| generate_GMCparameters( dec->sprite_warping_points, 966| 2.47k| dec->sprite_warping_accuracy, gmc_warp, 967| 2.47k| dec->width, dec->height, &dec->new_gmc_data); 968| | 969| | /* image warping is done block-based in decoder_mbgmc(), now */ 970| 2.47k| } 971| | 972| 3.45k| bound = 0; 973| | 974| 448k| for (y = 0; y < mb_height; y++) { ------------------ | Branch (974:15): [True: 444k, False: 3.45k] ------------------ 975| 444k| cp_mb = st_mb = 0; 976| 19.1M| for (x = 0; x < mb_width; x++) { ------------------ | Branch (976:17): [True: 18.7M, False: 444k] ------------------ 977| 18.7M| MACROBLOCK *mb; 978| | 979| | /* skip stuffing */ 980| 18.7M| while (BitstreamShowBits(bs, 10) == 1) ------------------ | Branch (980:14): [True: 20.9k, False: 18.7M] ------------------ 981| 20.9k| BitstreamSkip(bs, 10); 982| | 983| 18.7M| if (check_resync_marker(bs, fcode - 1)) { ------------------ | Branch (983:11): [True: 11.7k, False: 18.7M] ------------------ 984| 11.7k| bound = read_video_packet_header(bs, dec, fcode - 1, 985| 11.7k| &quant, &fcode, NULL, &intra_dc_threshold); 986| 11.7k| x = bound % mb_width; 987| 11.7k| y = MIN((bound / mb_width), (mb_height-1)); ------------------ | | 34| 11.7k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 10.7k, False: 1.03k] | | ------------------ ------------------ 988| 11.7k| } 989| 18.7M| mb = &dec->mbs[y * dec->mb_width + x]; 990| | 991| 18.7M| DPRINTF(XVID_DEBUG_MB, "macroblock (%i,%i) %08x\n", x, y, BitstreamShowBits(bs, 32)); ------------------ | | 200| 18.7M|#define XVID_DEBUG_MB (1<< 4) ------------------ 992| | 993| 18.7M| if (!(BitstreamGetBit(bs))) { /* block _is_ coded */ ------------------ | Branch (993:11): [True: 18.1M, False: 539k] ------------------ 994| 18.1M| uint32_t mcbpc, cbpc, cbpy, cbp; 995| 18.1M| uint32_t intra, acpred_flag = 0; 996| 18.1M| int mcsel = 0; /* mcsel: '0'=local motion, '1'=GMC */ 997| | 998| 18.1M| cp_mb++; 999| 18.1M| mcbpc = get_mcbpc_inter(bs); 1000| 18.1M| mb->mode = mcbpc & 7; 1001| 18.1M| cbpc = (mcbpc >> 4); 1002| | 1003| 18.1M| DPRINTF(XVID_DEBUG_MB, "mode %i\n", mb->mode); ------------------ | | 200| 18.1M|#define XVID_DEBUG_MB (1<< 4) ------------------ 1004| 18.1M| DPRINTF(XVID_DEBUG_MB, "cbpc %i\n", cbpc); ------------------ | | 200| 18.1M|#define XVID_DEBUG_MB (1<< 4) ------------------ 1005| | 1006| 18.1M| intra = (mb->mode == MODE_INTRA || mb->mode == MODE_INTRA_Q); ------------------ | | 37| 36.3M|#define MODE_INTRA 3 ------------------ intra = (mb->mode == MODE_INTRA || mb->mode == MODE_INTRA_Q); ------------------ | | 38| 18.1M|#define MODE_INTRA_Q 4 ------------------ | Branch (1006:18): [True: 21.3k, False: 18.1M] | Branch (1006:44): [True: 15.7k, False: 18.1M] ------------------ 1007| | 1008| 18.1M| if (gmc_warp && (mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q)) ------------------ | | 34| 28.1M|#define MODE_INTER 0 ------------------ if (gmc_warp && (mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q)) ------------------ | | 35| 13.9M|#define MODE_INTER_Q 1 ------------------ | Branch (1008:13): [True: 14.0M, False: 4.14M] | Branch (1008:26): [True: 109k, False: 13.9M] | Branch (1008:52): [True: 23.4k, False: 13.9M] ------------------ 1009| 132k| mcsel = BitstreamGetBit(bs); 1010| 18.0M| else if (intra) ------------------ | Branch (1010:18): [True: 37.0k, False: 18.0M] ------------------ 1011| 37.0k| acpred_flag = BitstreamGetBit(bs); 1012| | 1013| 18.1M| cbpy = get_cbpy(bs, intra); 1014| 18.1M| DPRINTF(XVID_DEBUG_MB, "cbpy %i mcsel %i \n", cbpy,mcsel); ------------------ | | 200| 18.1M|#define XVID_DEBUG_MB (1<< 4) ------------------ 1015| | 1016| 18.1M| cbp = (cbpy << 2) | cbpc; 1017| | 1018| 18.1M| if (mb->mode == MODE_INTER_Q || mb->mode == MODE_INTRA_Q) { ------------------ | | 35| 36.3M|#define MODE_INTER_Q 1 ------------------ if (mb->mode == MODE_INTER_Q || mb->mode == MODE_INTRA_Q) { ------------------ | | 38| 18.1M|#define MODE_INTRA_Q 4 ------------------ | Branch (1018:13): [True: 32.8k, False: 18.1M] | Branch (1018:41): [True: 15.7k, False: 18.1M] ------------------ 1019| 48.6k| int dquant = dquant_table[BitstreamGetBits(bs, 2)]; 1020| 48.6k| DPRINTF(XVID_DEBUG_MB, "dquant %i\n", dquant); ------------------ | | 200| 48.6k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1021| 48.6k| quant += dquant; 1022| 48.6k| if (quant > 31) { ------------------ | Branch (1022:15): [True: 4.47k, False: 44.1k] ------------------ 1023| 4.47k| quant = 31; 1024| 44.1k| } else if (quant < 1) { ------------------ | Branch (1024:22): [True: 18.2k, False: 25.8k] ------------------ 1025| 18.2k| quant = 1; 1026| 18.2k| } 1027| 48.6k| DPRINTF(XVID_DEBUG_MB, "quant %i\n", quant); ------------------ | | 200| 48.6k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1028| 48.6k| } 1029| 18.1M| mb->quant = quant; 1030| | 1031| 18.1M| mb->field_pred=0; 1032| 18.1M| if (dec->interlacing) { ------------------ | Branch (1032:13): [True: 13.9M, False: 4.28M] ------------------ 1033| 13.9M| if (cbp || intra) { ------------------ | Branch (1033:15): [True: 13.8M, False: 47.6k] | Branch (1033:22): [True: 1.05k, False: 46.5k] ------------------ 1034| 13.8M| mb->field_dct = BitstreamGetBit(bs); 1035| 13.8M| DPRINTF(XVID_DEBUG_MB,"decp: field_dct: %i\n", mb->field_dct); ------------------ | | 200| 13.8M|#define XVID_DEBUG_MB (1<< 4) ------------------ 1036| 13.8M| } 1037| | 1038| 13.9M| if ((mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q) && !mcsel) { ------------------ | | 34| 27.8M|#define MODE_INTER 0 ------------------ if ((mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q) && !mcsel) { ------------------ | | 35| 13.8M|#define MODE_INTER_Q 1 ------------------ | Branch (1038:16): [True: 94.4k, False: 13.8M] | Branch (1038:42): [True: 20.8k, False: 13.7M] | Branch (1038:71): [True: 75.1k, False: 40.1k] ------------------ 1039| 75.1k| mb->field_pred = BitstreamGetBit(bs); 1040| 75.1k| DPRINTF(XVID_DEBUG_MB, "decp: field_pred: %i\n", mb->field_pred); ------------------ | | 200| 75.1k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1041| | 1042| 75.1k| if (mb->field_pred) { ------------------ | Branch (1042:17): [True: 28.0k, False: 47.1k] ------------------ 1043| 28.0k| mb->field_for_top = BitstreamGetBit(bs); 1044| 28.0k| DPRINTF(XVID_DEBUG_MB,"decp: field_for_top: %i\n", mb->field_for_top); ------------------ | | 200| 28.0k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1045| 28.0k| mb->field_for_bot = BitstreamGetBit(bs); 1046| 28.0k| DPRINTF(XVID_DEBUG_MB,"decp: field_for_bot: %i\n", mb->field_for_bot); ------------------ | | 200| 28.0k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1047| 28.0k| } 1048| 75.1k| } 1049| 13.9M| } 1050| | 1051| 18.1M| if (mcsel) { ------------------ | Branch (1051:13): [True: 52.4k, False: 18.1M] ------------------ 1052| 52.4k| decoder_mbgmc(dec, mb, x, y, fcode, cbp, bs, rounding); 1053| 52.4k| continue; 1054| | 1055| 18.1M| } else if (mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q) { ------------------ | | 34| 36.2M|#define MODE_INTER 0 ------------------ } else if (mb->mode == MODE_INTER || mb->mode == MODE_INTER_Q) { ------------------ | | 35| 18.0M|#define MODE_INTER_Q 1 ------------------ | Branch (1055:20): [True: 123k, False: 18.0M] | Branch (1055:46): [True: 24.2k, False: 17.9M] ------------------ 1056| | 1057| 147k| if(dec->interlacing) { ------------------ | Branch (1057:14): [True: 75.1k, False: 72.7k] ------------------ 1058| | /* Get motion vectors interlaced, field_pred is handled there */ 1059| 75.1k| get_motion_vector_interlaced(dec, bs, x, y, 0, mb, fcode, bound); 1060| 75.1k| } else { 1061| 72.7k| get_motion_vector(dec, bs, x, y, 0, &mb->mvs[0], fcode, bound); 1062| 72.7k| mb->mvs[1] = mb->mvs[2] = mb->mvs[3] = mb->mvs[0]; 1063| 72.7k| } 1064| 17.9M| } else if (mb->mode == MODE_INTER4V ) { ------------------ | | 36| 17.9M|#define MODE_INTER4V 2 ------------------ | Branch (1064:20): [True: 36.9k, False: 17.9M] ------------------ 1065| | /* interlaced missing here */ 1066| 36.9k| get_motion_vector(dec, bs, x, y, 0, &mb->mvs[0], fcode, bound); 1067| 36.9k| get_motion_vector(dec, bs, x, y, 1, &mb->mvs[1], fcode, bound); 1068| 36.9k| get_motion_vector(dec, bs, x, y, 2, &mb->mvs[2], fcode, bound); 1069| 36.9k| get_motion_vector(dec, bs, x, y, 3, &mb->mvs[3], fcode, bound); 1070| 17.9M| } else { /* MODE_INTRA, MODE_INTRA_Q */ 1071| 17.9M| mb->mvs[0].x = mb->mvs[1].x = mb->mvs[2].x = mb->mvs[3].x = 0; 1072| 17.9M| mb->mvs[0].y = mb->mvs[1].y = mb->mvs[2].y = mb->mvs[3].y = 0; 1073| 17.9M| decoder_mbintra(dec, mb, x, y, acpred_flag, cbp, bs, quant, 1074| 17.9M| intra_dc_threshold, bound); 1075| 17.9M| continue; 1076| 17.9M| } 1077| | 1078| | /* See how to decode */ 1079| 184k| if(!mb->field_pred) ------------------ | Branch (1079:12): [True: 156k, False: 28.0k] ------------------ 1080| 156k| decoder_mbinter(dec, mb, x, y, cbp, bs, rounding, 0, 0); 1081| 28.0k| else 1082| 28.0k| decoder_mbinter_field(dec, mb, x, y, cbp, bs, rounding, 0, 0); 1083| | 1084| 539k| } else if (gmc_warp) { /* a not coded S(GMC)-VOP macroblock */ ------------------ | Branch (1084:18): [True: 329k, False: 209k] ------------------ 1085| 329k| mb->mode = MODE_NOT_CODED_GMC; ------------------ | | 40| 329k|#define MODE_NOT_CODED_GMC 17 ------------------ 1086| 329k| mb->quant = quant; 1087| 329k| decoder_mbgmc(dec, mb, x, y, fcode, 0x00, bs, rounding); 1088| | 1089| 329k| if(dec->out_frm && cp_mb > 0) { ------------------ | Branch (1089:12): [True: 0, False: 329k] | Branch (1089:28): [True: 0, False: 0] ------------------ 1090| 0| output_slice(&dec->cur, dec->edged_width,dec->width,dec->out_frm,st_mb,y,cp_mb); 1091| 0| cp_mb = 0; 1092| 0| } 1093| 329k| st_mb = x+1; 1094| 329k| } else { /* not coded P_VOP macroblock */ 1095| 209k| mb->mode = MODE_NOT_CODED; ------------------ | | 39| 209k|#define MODE_NOT_CODED 16 ------------------ 1096| 209k| mb->quant = quant; 1097| | 1098| 209k| mb->mvs[0].x = mb->mvs[1].x = mb->mvs[2].x = mb->mvs[3].x = 0; 1099| 209k| mb->mvs[0].y = mb->mvs[1].y = mb->mvs[2].y = mb->mvs[3].y = 0; 1100| 209k| mb->field_pred=0; /* (!) */ 1101| | 1102| 209k| decoder_mbinter(dec, mb, x, y, 0, bs, 1103| 209k| rounding, 0, 0); 1104| | 1105| 209k| if(dec->out_frm && cp_mb > 0) { ------------------ | Branch (1105:12): [True: 0, False: 209k] | Branch (1105:28): [True: 0, False: 0] ------------------ 1106| 0| output_slice(&dec->cur, dec->edged_width,dec->width,dec->out_frm,st_mb,y,cp_mb); 1107| 0| cp_mb = 0; 1108| 0| } 1109| 209k| st_mb = x+1; 1110| 209k| } 1111| 18.7M| } 1112| | 1113| 444k| if(dec->out_frm && cp_mb > 0) ------------------ | Branch (1113:8): [True: 0, False: 444k] | Branch (1113:24): [True: 0, False: 0] ------------------ 1114| 0| output_slice(&dec->cur, dec->edged_width,dec->width,dec->out_frm,st_mb,y,cp_mb); 1115| 444k| } 1116| 3.45k|} decoder.c:decoder_mbgmc: 690| 381k|{ 691| 381k| const uint32_t stride = dec->edged_width; 692| 381k| const uint32_t stride2 = stride / 2; 693| | 694| 381k| uint8_t *const pY_Cur=dec->cur.y + (y_pos << 4) * stride + (x_pos << 4); 695| 381k| uint8_t *const pU_Cur=dec->cur.u + (y_pos << 3) * stride2 + (x_pos << 3); 696| 381k| uint8_t *const pV_Cur=dec->cur.v + (y_pos << 3) * stride2 + (x_pos << 3); 697| | 698| 381k| NEW_GMC_DATA * gmc_data = &dec->new_gmc_data; 699| | 700| 381k| pMB->mvs[0] = pMB->mvs[1] = pMB->mvs[2] = pMB->mvs[3] = pMB->amv; 701| | 702| 381k| start_timer(); 703| | 704| |/* this is where the calculations are done */ 705| | 706| 381k| gmc_data->predict_16x16(gmc_data, 707| 381k| dec->cur.y + y_pos*16*stride + x_pos*16, dec->refn[0].y, 708| 381k| stride, stride, x_pos, y_pos, rounding); 709| | 710| 381k| gmc_data->predict_8x8(gmc_data, 711| 381k| dec->cur.u + y_pos*8*stride2 + x_pos*8, dec->refn[0].u, 712| 381k| dec->cur.v + y_pos*8*stride2 + x_pos*8, dec->refn[0].v, 713| 381k| stride2, stride2, x_pos, y_pos, rounding); 714| | 715| 381k| gmc_data->get_average_mv(gmc_data, &pMB->amv, x_pos, y_pos, dec->quarterpel); 716| | 717| 381k| pMB->amv.x = gmc_sanitize(pMB->amv.x, dec->quarterpel, fcode); 718| 381k| pMB->amv.y = gmc_sanitize(pMB->amv.y, dec->quarterpel, fcode); 719| | 720| 381k| pMB->mvs[0] = pMB->mvs[1] = pMB->mvs[2] = pMB->mvs[3] = pMB->amv; 721| | 722| 381k| stop_transfer_timer(); 723| | 724| 381k| if (cbp) ------------------ | Branch (724:7): [True: 37.5k, False: 344k] ------------------ 725| 37.5k| decoder_mb_decode(dec, cbp, bs, pY_Cur, pU_Cur, pV_Cur, pMB); 726| | 727| 381k|} decoder.c:decoder_mb_decode: 359| 262k|{ 360| 262k| DECLARE_ALIGNED_MATRIX(data, 1, 64, int16_t, CACHE_LINE); ------------------ | | 287| 262k| type name##_storage[(sizex)*(sizey)+(alignment)-1]; \ | | 288| 262k|type * name = (type *) (((ptr_t) name##_storage+(alignment - 1)) & ~((ptr_t)(alignment)-1)) ------------------ 361| | 362| 262k| int stride = dec->edged_width; 363| 262k| int i; 364| 262k| const uint32_t iQuant = MAX(1, pMB->quant); ------------------ | | 35| 262k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 1.38k, False: 261k] | | ------------------ ------------------ 365| 262k| const int direction = dec->alternate_vertical_scan ? 2 : 0; ------------------ | Branch (365:25): [True: 26.6k, False: 235k] ------------------ 366| 262k| typedef void (*get_inter_block_function_t)( 367| 262k| Bitstream * bs, 368| 262k| int16_t * block, 369| 262k| int direction, 370| 262k| const int quant, 371| 262k| const uint16_t *matrix); 372| 262k| typedef void (*add_residual_function_t)( 373| 262k| uint8_t *predicted_block, 374| 262k| const int16_t *residual, 375| 262k| int stride); 376| | 377| 262k| const get_inter_block_function_t get_inter_block = (dec->quant_type == 0) ------------------ | Branch (377:54): [True: 132k, False: 130k] ------------------ 378| 262k| ? (get_inter_block_function_t)get_inter_block_h263 379| 262k| : (get_inter_block_function_t)get_inter_block_mpeg; 380| | 381| 262k| uint8_t *dst[6]; 382| 262k| int strides[6]; 383| | 384| | 385| 262k| if (dec->interlacing && pMB->field_dct) { ------------------ | Branch (385:7): [True: 159k, False: 103k] | Branch (385:27): [True: 60.9k, False: 98.3k] ------------------ 386| 60.9k| dst[0] = pY_Cur; 387| 60.9k| dst[1] = pY_Cur + 8; 388| 60.9k| dst[2] = pY_Cur + stride; 389| 60.9k| dst[3] = dst[2] + 8; 390| 60.9k| dst[4] = pU_Cur; 391| 60.9k| dst[5] = pV_Cur; 392| 60.9k| strides[0] = strides[1] = strides[2] = strides[3] = stride*2; 393| 60.9k| strides[4] = stride/2; 394| 60.9k| strides[5] = stride/2; 395| 201k| } else { 396| 201k| dst[0] = pY_Cur; 397| 201k| dst[1] = pY_Cur + 8; 398| 201k| dst[2] = pY_Cur + 8*stride; 399| 201k| dst[3] = dst[2] + 8; 400| 201k| dst[4] = pU_Cur; 401| 201k| dst[5] = pV_Cur; 402| 201k| strides[0] = strides[1] = strides[2] = strides[3] = stride; 403| 201k| strides[4] = stride/2; 404| 201k| strides[5] = stride/2; 405| 201k| } 406| | 407| 1.83M| for (i = 0; i < 6; i++) { ------------------ | Branch (407:15): [True: 1.57M, False: 262k] ------------------ 408| | /* Process only coded blocks */ 409| 1.57M| if (cbp & (1 << (5 - i))) { ------------------ | Branch (409:9): [True: 526k, False: 1.04M] ------------------ 410| | 411| | /* Clear the block */ 412| 526k| memset(&data[0], 0, 64*sizeof(int16_t)); 413| | 414| | /* Decode coeffs and dequantize on the fly */ 415| 526k| start_timer(); 416| 526k| get_inter_block(bs, &data[0], direction, iQuant, get_inter_matrix(dec->mpeg_quant_matrices)); 417| 526k| stop_coding_timer(); 418| | 419| | /* iDCT */ 420| 526k| start_timer(); 421| 526k| idct((short * const)&data[0]); 422| 526k| stop_idct_timer(); 423| | 424| | /* Add this residual to the predicted block */ 425| 526k| start_timer(); 426| 526k| transfer_16to8add(dst[i], &data[0], strides[i]); 427| 526k| stop_transfer_timer(); 428| 526k| } 429| 1.57M| } 430| 262k|} decoder.c:get_motion_vector_interlaced: 858| 75.1k|{ 859| 75.1k| const int scale_fac = 1 << (fcode - 1); 860| 75.1k| const int high = (32 * scale_fac) - 1; 861| 75.1k| const int low = ((-32) * scale_fac); 862| 75.1k| const int range = (64 * scale_fac); 863| | 864| | /* Get interlaced prediction */ 865| 75.1k| const VECTOR pmv=get_pmv2_interlaced(dec->mbs,dec->mb_width,bound,x,y,k); 866| 75.1k| VECTOR mv,mvf1,mvf2; 867| | 868| 75.1k| if(!pMB->field_pred) ------------------ | Branch (868:6): [True: 47.1k, False: 28.0k] ------------------ 869| 47.1k| { 870| 47.1k| mv.x = get_mv(bs,fcode); 871| 47.1k| mv.y = get_mv(bs,fcode); 872| | 873| 47.1k| mv.x += pmv.x; 874| 47.1k| mv.y += pmv.y; 875| | 876| 47.1k| if(mv.xhigh) { ------------------ | Branch (878:16): [True: 25.1k, False: 12.9k] ------------------ 879| 25.1k| mv.x-=range; 880| 25.1k| } 881| | 882| 47.1k| if (mv.y < low) { ------------------ | Branch (882:9): [True: 8.89k, False: 38.2k] ------------------ 883| 8.89k| mv.y += range; 884| 38.2k| } else if (mv.y > high) { ------------------ | Branch (884:16): [True: 25.3k, False: 12.8k] ------------------ 885| 25.3k| mv.y -= range; 886| 25.3k| } 887| | 888| 47.1k| pMB->mvs[0]=pMB->mvs[1]=pMB->mvs[2]=pMB->mvs[3]=mv; 889| 47.1k| } 890| 28.0k| else 891| 28.0k| { 892| 28.0k| mvf1.x = get_mv(bs, fcode); 893| 28.0k| mvf1.y = get_mv(bs, fcode); 894| | 895| 28.0k| mvf1.x += pmv.x; 896| 28.0k| mvf1.y = 2*(mvf1.y+pmv.y/2); /* It's multiple of 2 */ 897| | 898| 28.0k| if (mvf1.x < low) { ------------------ | Branch (898:9): [True: 4.62k, False: 23.3k] ------------------ 899| 4.62k| mvf1.x += range; 900| 23.3k| } else if (mvf1.x > high) { ------------------ | Branch (900:16): [True: 15.1k, False: 8.22k] ------------------ 901| 15.1k| mvf1.x -= range; 902| 15.1k| } 903| | 904| 28.0k| if (mvf1.y < low) { ------------------ | Branch (904:9): [True: 7.14k, False: 20.8k] ------------------ 905| 7.14k| mvf1.y += range; 906| 20.8k| } else if (mvf1.y > high) { ------------------ | Branch (906:16): [True: 12.6k, False: 8.23k] ------------------ 907| 12.6k| mvf1.y -= range; 908| 12.6k| } 909| | 910| 28.0k| mvf2.x = get_mv(bs, fcode); 911| 28.0k| mvf2.y = get_mv(bs, fcode); 912| | 913| 28.0k| mvf2.x += pmv.x; 914| 28.0k| mvf2.y = 2*(mvf2.y+pmv.y/2); /* It's multiple of 2 */ 915| | 916| 28.0k| if (mvf2.x < low) { ------------------ | Branch (916:9): [True: 3.97k, False: 24.0k] ------------------ 917| 3.97k| mvf2.x += range; 918| 24.0k| } else if (mvf2.x > high) { ------------------ | Branch (918:16): [True: 15.7k, False: 8.29k] ------------------ 919| 15.7k| mvf2.x -= range; 920| 15.7k| } 921| | 922| 28.0k| if (mvf2.y < low) { ------------------ | Branch (922:9): [True: 4.13k, False: 23.8k] ------------------ 923| 4.13k| mvf2.y += range; 924| 23.8k| } else if (mvf2.y > high) { ------------------ | Branch (924:16): [True: 15.6k, False: 8.21k] ------------------ 925| 15.6k| mvf2.y -= range; 926| 15.6k| } 927| | 928| 28.0k| pMB->mvs[0]=mvf1; 929| 28.0k| pMB->mvs[1]=mvf2; 930| 28.0k| pMB->mvs[2].x=pMB->mvs[3].x=0; 931| 28.0k| pMB->mvs[2].y=pMB->mvs[3].y=0; 932| | 933| | /* Calculate average for as it is field predicted */ 934| 28.0k| pMB->mvs_avg.x=DIV2ROUND(pMB->mvs[0].x+pMB->mvs[1].x); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 935| 28.0k| pMB->mvs_avg.y=DIV2ROUND(pMB->mvs[0].y+pMB->mvs[1].y); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 936| 28.0k| } 937| 75.1k|} decoder.c:get_motion_vector: 814| 220k|{ 815| | 816| 220k| const int scale_fac = 1 << (fcode - 1); 817| 220k| const int high = (32 * scale_fac) - 1; 818| 220k| const int low = ((-32) * scale_fac); 819| 220k| const int range = (64 * scale_fac); 820| | 821| 220k| const VECTOR pmv = get_pmv2(dec->mbs, dec->mb_width, bound, x, y, k); 822| 220k| VECTOR mv; 823| | 824| 220k| mv.x = get_mv(bs, fcode); 825| 220k| mv.y = get_mv(bs, fcode); 826| | 827| 220k| DPRINTF(XVID_DEBUG_MV,"mv_diff (%i,%i) pred (%i,%i) result (%i,%i)\n", mv.x, mv.y, pmv.x, pmv.y, mv.x+pmv.x, mv.y+pmv.y); ------------------ | | 202| 220k|#define XVID_DEBUG_MV (1<< 6) ------------------ 828| | 829| 220k| mv.x += pmv.x; 830| 220k| mv.y += pmv.y; 831| | 832| 220k| if (mv.x < low) { ------------------ | Branch (832:7): [True: 36.2k, False: 184k] ------------------ 833| 36.2k| mv.x += range; 834| 184k| } else if (mv.x > high) { ------------------ | Branch (834:14): [True: 130k, False: 53.6k] ------------------ 835| 130k| mv.x -= range; 836| 130k| } 837| | 838| 220k| if (mv.y < low) { ------------------ | Branch (838:7): [True: 54.4k, False: 166k] ------------------ 839| 54.4k| mv.y += range; 840| 166k| } else if (mv.y > high) { ------------------ | Branch (840:14): [True: 112k, False: 53.3k] ------------------ 841| 112k| mv.y -= range; 842| 112k| } 843| | 844| 220k| ret_mv->x = mv.x; 845| 220k| ret_mv->y = mv.y; 846| 220k|} decoder.c:decoder_mbinter: 483| 435k|{ 484| 435k| uint32_t stride = dec->edged_width; 485| 435k| uint32_t stride2 = stride / 2; 486| 435k| uint32_t i; 487| | 488| 435k| uint8_t *pY_Cur, *pU_Cur, *pV_Cur; 489| | 490| 435k| int uv_dx, uv_dy; 491| 435k| VECTOR mv[4]; /* local copy of mvs */ 492| | 493| 435k| pY_Cur = dec->cur.y + (y_pos << 4) * stride + (x_pos << 4); 494| 435k| pU_Cur = dec->cur.u + (y_pos << 3) * stride2 + (x_pos << 3); 495| 435k| pV_Cur = dec->cur.v + (y_pos << 3) * stride2 + (x_pos << 3); 496| 2.17M| for (i = 0; i < 4; i++) ------------------ | Branch (496:15): [True: 1.74M, False: 435k] ------------------ 497| 1.74M| mv[i] = pMB->mvs[i]; 498| | 499| 435k| validate_vector(mv, x_pos, y_pos, dec); 500| | 501| 435k| start_timer(); 502| | 503| 435k| if ((pMB->mode != MODE_INTER4V) || (bvop)) { /* INTER, INTER_Q, NOT_CODED, FORWARD, BACKWARD */ ------------------ | | 36| 435k|#define MODE_INTER4V 2 ------------------ | Branch (503:7): [True: 354k, False: 81.1k] | Branch (503:38): [True: 44.1k, False: 36.9k] ------------------ 504| | 505| 398k| uv_dx = mv[0].x; 506| 398k| uv_dy = mv[0].y; 507| 398k| if (dec->quarterpel) { ------------------ | Branch (507:9): [True: 96.2k, False: 302k] ------------------ 508| 96.2k| if (dec->bs_version <= BS_VERSION_BUGGY_CHROMA_ROUNDING) { ------------------ | | 470| 96.2k|#define BS_VERSION_BUGGY_CHROMA_ROUNDING 1 ------------------ | Branch (508:8): [True: 1.66k, False: 94.5k] ------------------ 509| 1.66k| uv_dx = (uv_dx>>1) | (uv_dx&1); 510| 1.66k| uv_dy = (uv_dy>>1) | (uv_dy&1); 511| 1.66k| } 512| 94.5k| else { 513| 94.5k| uv_dx /= 2; 514| 94.5k| uv_dy /= 2; 515| 94.5k| } 516| 96.2k| } 517| 398k| uv_dx = (uv_dx >> 1) + roundtab_79[uv_dx & 0x3]; 518| 398k| uv_dy = (uv_dy >> 1) + roundtab_79[uv_dy & 0x3]; 519| | 520| 398k| if (dec->quarterpel) ------------------ | Branch (520:9): [True: 96.2k, False: 302k] ------------------ 521| 96.2k| interpolate16x16_quarterpel(dec->cur.y, dec->refn[ref].y, dec->qtmp.y, dec->qtmp.y + 64, 522| 96.2k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 523| 96.2k| mv[0].x, mv[0].y, stride, rounding); 524| 302k| else 525| 302k| interpolate16x16_switch(dec->cur.y, dec->refn[ref].y, 16*x_pos, 16*y_pos, 526| 302k| mv[0].x, mv[0].y, stride, rounding); 527| | 528| 398k| } else { /* MODE_INTER4V */ 529| | 530| 36.9k| if(dec->quarterpel) { ------------------ | Branch (530:8): [True: 13.7k, False: 23.2k] ------------------ 531| 13.7k| if (dec->bs_version <= BS_VERSION_BUGGY_CHROMA_ROUNDING) { ------------------ | | 470| 13.7k|#define BS_VERSION_BUGGY_CHROMA_ROUNDING 1 ------------------ | Branch (531:8): [True: 1.59k, False: 12.1k] ------------------ 532| 1.59k| int z; 533| 1.59k| uv_dx = 0; uv_dy = 0; 534| 7.98k| for (z = 0; z < 4; z++) { ------------------ | Branch (534:17): [True: 6.38k, False: 1.59k] ------------------ 535| 6.38k| uv_dx += ((mv[z].x>>1) | (mv[z].x&1)); 536| 6.38k| uv_dy += ((mv[z].y>>1) | (mv[z].y&1)); 537| 6.38k| } 538| 1.59k| } 539| 12.1k| else { 540| 12.1k| uv_dx = (mv[0].x / 2) + (mv[1].x / 2) + (mv[2].x / 2) + (mv[3].x / 2); 541| 12.1k| uv_dy = (mv[0].y / 2) + (mv[1].y / 2) + (mv[2].y / 2) + (mv[3].y / 2); 542| 12.1k| } 543| 23.2k| } else { 544| 23.2k| uv_dx = mv[0].x + mv[1].x + mv[2].x + mv[3].x; 545| 23.2k| uv_dy = mv[0].y + mv[1].y + mv[2].y + mv[3].y; 546| 23.2k| } 547| | 548| 36.9k| uv_dx = (uv_dx >> 3) + roundtab_76[uv_dx & 0xf]; 549| 36.9k| uv_dy = (uv_dy >> 3) + roundtab_76[uv_dy & 0xf]; 550| | 551| 36.9k| if (dec->quarterpel) { ------------------ | Branch (551:9): [True: 13.7k, False: 23.2k] ------------------ 552| 13.7k| interpolate8x8_quarterpel(dec->cur.y, dec->refn[0].y , dec->qtmp.y, dec->qtmp.y + 64, 553| 13.7k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 554| 13.7k| mv[0].x, mv[0].y, stride, rounding); 555| 13.7k| interpolate8x8_quarterpel(dec->cur.y, dec->refn[0].y , dec->qtmp.y, dec->qtmp.y + 64, 556| 13.7k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos, 557| 13.7k| mv[1].x, mv[1].y, stride, rounding); 558| 13.7k| interpolate8x8_quarterpel(dec->cur.y, dec->refn[0].y , dec->qtmp.y, dec->qtmp.y + 64, 559| 13.7k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos + 8, 560| 13.7k| mv[2].x, mv[2].y, stride, rounding); 561| 13.7k| interpolate8x8_quarterpel(dec->cur.y, dec->refn[0].y , dec->qtmp.y, dec->qtmp.y + 64, 562| 13.7k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos + 8, 563| 13.7k| mv[3].x, mv[3].y, stride, rounding); 564| 23.2k| } else { 565| 23.2k| interpolate8x8_switch(dec->cur.y, dec->refn[0].y , 16*x_pos, 16*y_pos, 566| 23.2k| mv[0].x, mv[0].y, stride, rounding); 567| 23.2k| interpolate8x8_switch(dec->cur.y, dec->refn[0].y , 16*x_pos + 8, 16*y_pos, 568| 23.2k| mv[1].x, mv[1].y, stride, rounding); 569| 23.2k| interpolate8x8_switch(dec->cur.y, dec->refn[0].y , 16*x_pos, 16*y_pos + 8, 570| 23.2k| mv[2].x, mv[2].y, stride, rounding); 571| 23.2k| interpolate8x8_switch(dec->cur.y, dec->refn[0].y , 16*x_pos + 8, 16*y_pos + 8, 572| 23.2k| mv[3].x, mv[3].y, stride, rounding); 573| 23.2k| } 574| 36.9k| } 575| | 576| | /* chroma */ 577| 435k| interpolate8x8_switch(dec->cur.u, dec->refn[ref].u, 8 * x_pos, 8 * y_pos, 578| 435k| uv_dx, uv_dy, stride2, rounding); 579| 435k| interpolate8x8_switch(dec->cur.v, dec->refn[ref].v, 8 * x_pos, 8 * y_pos, 580| 435k| uv_dx, uv_dy, stride2, rounding); 581| | 582| 435k| stop_comp_timer(); 583| | 584| 435k| if (cbp) ------------------ | Branch (584:7): [True: 142k, False: 293k] ------------------ 585| 142k| decoder_mb_decode(dec, cbp, bs, pY_Cur, pU_Cur, pV_Cur, pMB); 586| 435k|} decoder.c:validate_vector: 434| 1.70M|{ 435| | /* clip a vector to valid range 436| | prevents crashes if bitstream is broken 437| | */ 438| 1.70M| int shift = 5 + dec->quarterpel; 439| 1.70M| int xborder_high = (int)(dec->mb_width - x_pos) << shift; 440| 1.70M| int xborder_low = (-(int)x_pos-1) << shift; 441| 1.70M| int yborder_high = (int)(dec->mb_height - y_pos) << shift; 442| 1.70M| int yborder_low = (-(int)y_pos-1) << shift; 443| | 444| 1.70M|#define CHECK_MV(mv) \ 445| 1.70M| do { \ 446| 1.70M| if ((mv).x > xborder_high) { \ 447| 1.70M| DPRINTF(XVID_DEBUG_MV, "mv.x > max -- %d > %d, MB %d, %d", (mv).x, xborder_high, x_pos, y_pos); \ 448| 1.70M| (mv).x = xborder_high; \ 449| 1.70M| } else if ((mv).x < xborder_low) { \ 450| 1.70M| DPRINTF(XVID_DEBUG_MV, "mv.x < min -- %d < %d, MB %d, %d", (mv).x, xborder_low, x_pos, y_pos); \ 451| 1.70M| (mv).x = xborder_low; \ 452| 1.70M| } \ 453| 1.70M| if ((mv).y > yborder_high) { \ 454| 1.70M| DPRINTF(XVID_DEBUG_MV, "mv.y > max -- %d > %d, MB %d, %d", (mv).y, yborder_high, x_pos, y_pos); \ 455| 1.70M| (mv).y = yborder_high; \ 456| 1.70M| } else if ((mv).y < yborder_low) { \ 457| 1.70M| DPRINTF(XVID_DEBUG_MV, "mv.y < min -- %d < %d, MB %d, %d", (mv).y, yborder_low, x_pos, y_pos); \ 458| 1.70M| (mv).y = yborder_low; \ 459| 1.70M| } \ 460| 1.70M| } while (0) 461| | 462| 1.70M| CHECK_MV(mv[0]); ------------------ | | 445| 1.70M| do { \ | | 446| 1.70M| if ((mv).x > xborder_high) { \ | | ------------------ | | | Branch (446:7): [True: 34.9k, False: 1.66M] | | ------------------ | | 447| 34.9k| DPRINTF(XVID_DEBUG_MV, "mv.x > max -- %d > %d, MB %d, %d", (mv).x, xborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 34.9k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 448| 34.9k| (mv).x = xborder_high; \ | | 449| 1.66M| } else if ((mv).x < xborder_low) { \ | | ------------------ | | | Branch (449:14): [True: 40.8k, False: 1.62M] | | ------------------ | | 450| 40.8k| DPRINTF(XVID_DEBUG_MV, "mv.x < min -- %d < %d, MB %d, %d", (mv).x, xborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 40.8k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 451| 40.8k| (mv).x = xborder_low; \ | | 452| 40.8k| } \ | | 453| 1.70M| if ((mv).y > yborder_high) { \ | | ------------------ | | | Branch (453:7): [True: 36.3k, False: 1.66M] | | ------------------ | | 454| 36.3k| DPRINTF(XVID_DEBUG_MV, "mv.y > max -- %d > %d, MB %d, %d", (mv).y, yborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 36.3k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 455| 36.3k| (mv).y = yborder_high; \ | | 456| 1.66M| } else if ((mv).y < yborder_low) { \ | | ------------------ | | | Branch (456:14): [True: 40.4k, False: 1.62M] | | ------------------ | | 457| 40.4k| DPRINTF(XVID_DEBUG_MV, "mv.y < min -- %d < %d, MB %d, %d", (mv).y, yborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 40.4k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 458| 40.4k| (mv).y = yborder_low; \ | | 459| 40.4k| } \ | | 460| 1.70M| } while (0) | | ------------------ | | | Branch (460:12): [Folded - Ignored] | | ------------------ ------------------ 463| 1.70M| CHECK_MV(mv[1]); ------------------ | | 445| 1.70M| do { \ | | 446| 1.70M| if ((mv).x > xborder_high) { \ | | ------------------ | | | Branch (446:7): [True: 33.0k, False: 1.67M] | | ------------------ | | 447| 33.0k| DPRINTF(XVID_DEBUG_MV, "mv.x > max -- %d > %d, MB %d, %d", (mv).x, xborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 33.0k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 448| 33.0k| (mv).x = xborder_high; \ | | 449| 1.67M| } else if ((mv).x < xborder_low) { \ | | ------------------ | | | Branch (449:14): [True: 41.6k, False: 1.62M] | | ------------------ | | 450| 41.6k| DPRINTF(XVID_DEBUG_MV, "mv.x < min -- %d < %d, MB %d, %d", (mv).x, xborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 41.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 451| 41.6k| (mv).x = xborder_low; \ | | 452| 41.6k| } \ | | 453| 1.70M| if ((mv).y > yborder_high) { \ | | ------------------ | | | Branch (453:7): [True: 35.6k, False: 1.66M] | | ------------------ | | 454| 35.6k| DPRINTF(XVID_DEBUG_MV, "mv.y > max -- %d > %d, MB %d, %d", (mv).y, yborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 35.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 455| 35.6k| (mv).y = yborder_high; \ | | 456| 1.66M| } else if ((mv).y < yborder_low) { \ | | ------------------ | | | Branch (456:14): [True: 40.6k, False: 1.62M] | | ------------------ | | 457| 40.6k| DPRINTF(XVID_DEBUG_MV, "mv.y < min -- %d < %d, MB %d, %d", (mv).y, yborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 40.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 458| 40.6k| (mv).y = yborder_low; \ | | 459| 40.6k| } \ | | 460| 1.70M| } while (0) | | ------------------ | | | Branch (460:12): [Folded - Ignored] | | ------------------ ------------------ 464| 1.70M| CHECK_MV(mv[2]); ------------------ | | 445| 1.70M| do { \ | | 446| 1.70M| if ((mv).x > xborder_high) { \ | | ------------------ | | | Branch (446:7): [True: 31.3k, False: 1.67M] | | ------------------ | | 447| 31.3k| DPRINTF(XVID_DEBUG_MV, "mv.x > max -- %d > %d, MB %d, %d", (mv).x, xborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 31.3k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 448| 31.3k| (mv).x = xborder_high; \ | | 449| 1.67M| } else if ((mv).x < xborder_low) { \ | | ------------------ | | | Branch (449:14): [True: 37.6k, False: 1.63M] | | ------------------ | | 450| 37.6k| DPRINTF(XVID_DEBUG_MV, "mv.x < min -- %d < %d, MB %d, %d", (mv).x, xborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 37.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 451| 37.6k| (mv).x = xborder_low; \ | | 452| 37.6k| } \ | | 453| 1.70M| if ((mv).y > yborder_high) { \ | | ------------------ | | | Branch (453:7): [True: 34.9k, False: 1.66M] | | ------------------ | | 454| 34.9k| DPRINTF(XVID_DEBUG_MV, "mv.y > max -- %d > %d, MB %d, %d", (mv).y, yborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 34.9k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 455| 34.9k| (mv).y = yborder_high; \ | | 456| 1.66M| } else if ((mv).y < yborder_low) { \ | | ------------------ | | | Branch (456:14): [True: 39.8k, False: 1.62M] | | ------------------ | | 457| 39.8k| DPRINTF(XVID_DEBUG_MV, "mv.y < min -- %d < %d, MB %d, %d", (mv).y, yborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 39.8k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 458| 39.8k| (mv).y = yborder_low; \ | | 459| 39.8k| } \ | | 460| 1.70M| } while (0) | | ------------------ | | | Branch (460:12): [Folded - Ignored] | | ------------------ ------------------ 465| 1.70M| CHECK_MV(mv[3]); ------------------ | | 445| 1.70M| do { \ | | 446| 1.70M| if ((mv).x > xborder_high) { \ | | ------------------ | | | Branch (446:7): [True: 31.3k, False: 1.67M] | | ------------------ | | 447| 31.3k| DPRINTF(XVID_DEBUG_MV, "mv.x > max -- %d > %d, MB %d, %d", (mv).x, xborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 31.3k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 448| 31.3k| (mv).x = xborder_high; \ | | 449| 1.67M| } else if ((mv).x < xborder_low) { \ | | ------------------ | | | Branch (449:14): [True: 37.8k, False: 1.63M] | | ------------------ | | 450| 37.8k| DPRINTF(XVID_DEBUG_MV, "mv.x < min -- %d < %d, MB %d, %d", (mv).x, xborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 37.8k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 451| 37.8k| (mv).x = xborder_low; \ | | 452| 37.8k| } \ | | 453| 1.70M| if ((mv).y > yborder_high) { \ | | ------------------ | | | Branch (453:7): [True: 35.6k, False: 1.66M] | | ------------------ | | 454| 35.6k| DPRINTF(XVID_DEBUG_MV, "mv.y > max -- %d > %d, MB %d, %d", (mv).y, yborder_high, x_pos, y_pos); \ | | ------------------ | | | | 202| 35.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 455| 35.6k| (mv).y = yborder_high; \ | | 456| 1.66M| } else if ((mv).y < yborder_low) { \ | | ------------------ | | | Branch (456:14): [True: 39.6k, False: 1.62M] | | ------------------ | | 457| 39.6k| DPRINTF(XVID_DEBUG_MV, "mv.y < min -- %d < %d, MB %d, %d", (mv).y, yborder_low, x_pos, y_pos); \ | | ------------------ | | | | 202| 39.6k|#define XVID_DEBUG_MV (1<< 6) | | ------------------ | | 458| 39.6k| (mv).y = yborder_low; \ | | 459| 39.6k| } \ | | 460| 1.70M| } while (0) | | ------------------ | | | Branch (460:12): [Folded - Ignored] | | ------------------ ------------------ 466| 1.70M|} decoder.c:decoder_mbinter_field: 599| 28.0k|{ 600| 28.0k| uint32_t stride = dec->edged_width; 601| 28.0k| uint32_t stride2 = stride / 2; 602| | 603| 28.0k| uint8_t *pY_Cur, *pU_Cur, *pV_Cur; 604| | 605| 28.0k| int uvtop_dx, uvtop_dy; 606| 28.0k| int uvbot_dx, uvbot_dy; 607| 28.0k| VECTOR mv[4]; /* local copy of mvs */ 608| | 609| | /* Get pointer to memory areas */ 610| 28.0k| pY_Cur = dec->cur.y + (y_pos << 4) * stride + (x_pos << 4); 611| 28.0k| pU_Cur = dec->cur.u + (y_pos << 3) * stride2 + (x_pos << 3); 612| 28.0k| pV_Cur = dec->cur.v + (y_pos << 3) * stride2 + (x_pos << 3); 613| | 614| 28.0k| mv[0] = pMB->mvs[0]; 615| 28.0k| mv[1] = pMB->mvs[1]; 616| 28.0k| memset(&mv[2],0,2*sizeof(VECTOR)); 617| | 618| 28.0k| validate_vector(mv, x_pos, y_pos, dec); 619| | 620| 28.0k| start_timer(); 621| | 622| 28.0k| if((pMB->mode!=MODE_INTER4V) || (bvop)) /* INTER, INTER_Q, NOT_CODED, FORWARD, BACKWARD */ ------------------ | | 36| 28.0k|#define MODE_INTER4V 2 ------------------ | Branch (622:6): [True: 28.0k, False: 0] | Branch (622:35): [True: 0, False: 0] ------------------ 623| 28.0k| { 624| | /* Prepare top field vector */ 625| 28.0k| uvtop_dx = DIV2ROUND(mv[0].x); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 626| 28.0k| uvtop_dy = DIV2ROUND(mv[0].y); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 627| | 628| | /* Prepare bottom field vector */ 629| 28.0k| uvbot_dx = DIV2ROUND(mv[1].x); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 630| 28.0k| uvbot_dy = DIV2ROUND(mv[1].y); ------------------ | | 64| 28.0k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 631| | 632| 28.0k| if(dec->quarterpel) ------------------ | Branch (632:8): [True: 6.73k, False: 21.2k] ------------------ 633| 6.73k| { 634| | /* NOT supported */ 635| 6.73k| } 636| 21.2k| else 637| 21.2k| { 638| | /* Interpolate top field left part(we use double stride for every 2nd line) */ 639| 21.2k| interpolate8x8_switch(dec->cur.y,dec->refn[ref].y+pMB->field_for_top*stride, 640| 21.2k| 16*x_pos,8*y_pos,mv[0].x, mv[0].y>>1,2*stride, rounding); 641| | /* top field right part */ 642| 21.2k| interpolate8x8_switch(dec->cur.y,dec->refn[ref].y+pMB->field_for_top*stride, 643| 21.2k| 16*x_pos+8,8*y_pos,mv[0].x, mv[0].y>>1,2*stride, rounding); 644| | 645| | /* Interpolate bottom field left part(we use double stride for every 2nd line) */ 646| 21.2k| interpolate8x8_switch(dec->cur.y+stride,dec->refn[ref].y+pMB->field_for_bot*stride, 647| 21.2k| 16*x_pos,8*y_pos,mv[1].x, mv[1].y>>1,2*stride, rounding); 648| | /* Bottom field right part */ 649| 21.2k| interpolate8x8_switch(dec->cur.y+stride,dec->refn[ref].y+pMB->field_for_bot*stride, 650| 21.2k| 16*x_pos+8,8*y_pos,mv[1].x, mv[1].y>>1,2*stride, rounding); 651| | 652| | /* Interpolate field1 U */ 653| 21.2k| interpolate8x4_switch(dec->cur.u,dec->refn[ref].u+pMB->field_for_top*stride2, 654| 21.2k| 8*x_pos,4*y_pos,uvtop_dx,DIV2ROUND(uvtop_dy),stride,rounding); ------------------ | | 64| 21.2k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 655| | 656| | /* Interpolate field1 V */ 657| 21.2k| interpolate8x4_switch(dec->cur.v,dec->refn[ref].v+pMB->field_for_top*stride2, 658| 21.2k| 8*x_pos,4*y_pos,uvtop_dx,DIV2ROUND(uvtop_dy),stride,rounding); ------------------ | | 64| 21.2k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 659| | 660| | /* Interpolate field2 U */ 661| 21.2k| interpolate8x4_switch(dec->cur.u+stride2,dec->refn[ref].u+pMB->field_for_bot*stride2, 662| 21.2k| 8*x_pos,4*y_pos,uvbot_dx,DIV2ROUND(uvbot_dy),stride,rounding); ------------------ | | 64| 21.2k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 663| | 664| | /* Interpolate field2 V */ 665| 21.2k| interpolate8x4_switch(dec->cur.v+stride2,dec->refn[ref].v+pMB->field_for_bot*stride2, 666| 21.2k| 8*x_pos,4*y_pos,uvbot_dx,DIV2ROUND(uvbot_dy),stride,rounding); ------------------ | | 64| 21.2k|#define DIV2ROUND(n) (((n)>>1)|((n)&1)) ------------------ 667| 21.2k| } 668| 28.0k| } 669| 0| else 670| 0| { 671| | /* We don't expect 4 motion vectors in interlaced mode */ 672| 0| } 673| | 674| 28.0k| stop_comp_timer(); 675| | 676| | /* Must add error correction? */ 677| 28.0k| if(cbp) ------------------ | Branch (677:6): [True: 15.7k, False: 12.3k] ------------------ 678| 15.7k| decoder_mb_decode(dec, cbp, bs, pY_Cur, pU_Cur, pV_Cur, pMB); 679| 28.0k|} decoder.c:decoder_bframe: 1360| 2.24k|{ 1361| 2.24k| uint32_t x, y; 1362| 2.24k| VECTOR mv; 1363| 2.24k| const VECTOR zeromv = {0,0}; 1364| 2.24k| int i; 1365| 2.24k| int resync_len; 1366| | 1367| 2.24k| if (!dec->is_edged[0]) { ------------------ | Branch (1367:7): [True: 2.24k, False: 0] ------------------ 1368| 2.24k| start_timer(); 1369| 2.24k| image_setedges(&dec->refn[0], dec->edged_width, dec->edged_height, 1370| 2.24k| dec->width, dec->height, dec->bs_version); 1371| 2.24k| dec->is_edged[0] = 1; 1372| 2.24k| stop_edges_timer(); 1373| 2.24k| } 1374| | 1375| 2.24k| if (!dec->is_edged[1]) { ------------------ | Branch (1375:7): [True: 2.24k, False: 0] ------------------ 1376| 2.24k| start_timer(); 1377| 2.24k| image_setedges(&dec->refn[1], dec->edged_width, dec->edged_height, 1378| 2.24k| dec->width, dec->height, dec->bs_version); 1379| 2.24k| dec->is_edged[1] = 1; 1380| 2.24k| stop_edges_timer(); 1381| 2.24k| } 1382| | 1383| 2.24k| resync_len = get_resync_len_b(fcode_backward, fcode_forward); 1384| 121k| for (y = 0; y < dec->mb_height; y++) { ------------------ | Branch (1384:15): [True: 119k, False: 2.24k] ------------------ 1385| | /* Initialize Pred Motion Vector */ 1386| 119k| dec->p_fmv = dec->p_bmv = zeromv; 1387| 4.70M| for (x = 0; x < dec->mb_width; x++) { ------------------ | Branch (1387:17): [True: 4.58M, False: 119k] ------------------ 1388| 4.58M| MACROBLOCK *mb = &dec->mbs[y * dec->mb_width + x]; 1389| 4.58M| MACROBLOCK *last_mb = &dec->last_mbs[y * dec->mb_width + x]; 1390| 4.58M| int intra_dc_threshold; /* fake variable */ 1391| | 1392| 4.58M| mv = 1393| 4.58M| mb->b_mvs[0] = mb->b_mvs[1] = mb->b_mvs[2] = mb->b_mvs[3] = 1394| 4.58M| mb->mvs[0] = mb->mvs[1] = mb->mvs[2] = mb->mvs[3] = zeromv; 1395| 4.58M| mb->quant = quant; 1396| | 1397| | /* 1398| | * skip if the co-located P_VOP macroblock is not coded 1399| | * if not codec in co-located S_VOP macroblock is _not_ 1400| | * automatically skipped 1401| | */ 1402| | 1403| 4.58M| if (last_mb->mode == MODE_NOT_CODED) { ------------------ | | 39| 4.58M|#define MODE_NOT_CODED 16 ------------------ | Branch (1403:11): [True: 0, False: 4.58M] ------------------ 1404| 0| mb->cbp = 0; 1405| 0| mb->mode = MODE_FORWARD; ------------------ | | 47| 0|#define MODE_FORWARD 3 ------------------ 1406| 0| decoder_mbinter(dec, mb, x, y, mb->cbp, bs, 0, 1, 1); 1407| 0| continue; 1408| 0| } 1409| | 1410| 4.58M| if (check_resync_marker(bs, resync_len)) { ------------------ | Branch (1410:11): [True: 3.84k, False: 4.58M] ------------------ 1411| 3.84k| int bound = read_video_packet_header(bs, dec, resync_len, &quant, 1412| 3.84k| &fcode_forward, &fcode_backward, &intra_dc_threshold); 1413| | 1414| 3.84k| bound = MAX(0, bound-1); /* valid bound must always be >0 */ ------------------ | | 35| 3.84k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (35:20): [True: 1.04k, False: 2.80k] | | ------------------ ------------------ 1415| 3.84k| x = bound % dec->mb_width; 1416| 3.84k| y = MIN((bound / dec->mb_width), (dec->mb_height-1)); ------------------ | | 34| 3.84k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 3.33k, False: 517] | | ------------------ ------------------ 1417| | /* reset predicted macroblocks */ 1418| 3.84k| dec->p_fmv = dec->p_bmv = zeromv; 1419| | /* update resync len with new fcodes */ 1420| 3.84k| resync_len = get_resync_len_b(fcode_backward, fcode_forward); 1421| 3.84k| continue; /* re-init loop */ 1422| 3.84k| } 1423| | 1424| 4.58M| if (!BitstreamGetBit(bs)) { /* modb=='0' */ ------------------ | Branch (1424:11): [True: 4.19M, False: 387k] ------------------ 1425| 4.19M| const uint8_t modb2 = BitstreamGetBit(bs); 1426| | 1427| 4.19M| mb->mode = get_mbtype(bs); 1428| | 1429| 4.19M| if (!modb2) /* modb=='00' */ ------------------ | Branch (1429:13): [True: 4.02M, False: 166k] ------------------ 1430| 4.02M| mb->cbp = BitstreamGetBits(bs, 6); 1431| 166k| else 1432| 166k| mb->cbp = 0; 1433| | 1434| 4.19M| if (mb->mode && mb->cbp) { ------------------ | Branch (1434:13): [True: 4.06M, False: 129k] | Branch (1434:25): [True: 95.4k, False: 3.96M] ------------------ 1435| 95.4k| quant += get_dbquant(bs); 1436| 95.4k| if (quant > 31) ------------------ | Branch (1436:15): [True: 5.62k, False: 89.8k] ------------------ 1437| 5.62k| quant = 31; 1438| 89.8k| else if (quant < 1) ------------------ | Branch (1438:20): [True: 5.55k, False: 84.2k] ------------------ 1439| 5.55k| quant = 1; 1440| 95.4k| } 1441| 4.19M| mb->quant = quant; 1442| | 1443| 4.19M| if (dec->interlacing) { ------------------ | Branch (1443:13): [True: 3.31M, False: 879k] ------------------ 1444| 3.31M| if (mb->cbp) { ------------------ | Branch (1444:15): [True: 93.4k, False: 3.22M] ------------------ 1445| 93.4k| mb->field_dct = BitstreamGetBit(bs); 1446| 93.4k| DPRINTF(XVID_DEBUG_MB,"decp: field_dct: %i\n", mb->field_dct); ------------------ | | 200| 93.4k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1447| 93.4k| } 1448| | 1449| 3.31M| if (mb->mode) { ------------------ | Branch (1449:15): [True: 3.22M, False: 92.1k] ------------------ 1450| 3.22M| mb->field_pred = BitstreamGetBit(bs); 1451| 3.22M| DPRINTF(XVID_DEBUG_MB, "decp: field_pred: %i\n", mb->field_pred); ------------------ | | 200| 3.22M|#define XVID_DEBUG_MB (1<< 4) ------------------ 1452| | 1453| 3.22M| if (mb->field_pred) { ------------------ | Branch (1453:17): [True: 68.6k, False: 3.15M] ------------------ 1454| 68.6k| mb->field_for_top = BitstreamGetBit(bs); 1455| 68.6k| DPRINTF(XVID_DEBUG_MB,"decp: field_for_top: %i\n", mb->field_for_top); ------------------ | | 200| 68.6k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1456| 68.6k| mb->field_for_bot = BitstreamGetBit(bs); 1457| 68.6k| DPRINTF(XVID_DEBUG_MB,"decp: field_for_bot: %i\n", mb->field_for_bot); ------------------ | | 200| 68.6k|#define XVID_DEBUG_MB (1<< 4) ------------------ 1458| 68.6k| } 1459| 3.22M| } 1460| 3.31M| } 1461| | 1462| 4.19M| } else { 1463| 387k| mb->mode = MODE_DIRECT_NONE_MV; ------------------ | | 48| 387k|#define MODE_DIRECT_NONE_MV 4 ------------------ 1464| 387k| mb->cbp = 0; 1465| 387k| } 1466| | 1467| 4.58M| switch (mb->mode) { 1468| 129k| case MODE_DIRECT: ------------------ | | 44| 129k|#define MODE_DIRECT 0 ------------------ | Branch (1468:7): [True: 129k, False: 4.45M] ------------------ 1469| 129k| get_b_motion_vector(bs, &mv, 1, zeromv, dec, x, y); 1470| | 1471| 517k| case MODE_DIRECT_NONE_MV: ------------------ | | 48| 517k|#define MODE_DIRECT_NONE_MV 4 ------------------ | Branch (1471:7): [True: 387k, False: 4.19M] ------------------ 1472| 2.58M| for (i = 0; i < 4; i++) { ------------------ | Branch (1472:21): [True: 2.06M, False: 517k] ------------------ 1473| 2.06M| mb->mvs[i].x = last_mb->mvs[i].x*dec->time_bp/dec->time_pp + mv.x; 1474| 2.06M| mb->mvs[i].y = last_mb->mvs[i].y*dec->time_bp/dec->time_pp + mv.y; 1475| | 1476| 2.06M| mb->b_mvs[i].x = (mv.x) ------------------ | Branch (1476:28): [True: 216k, False: 1.85M] ------------------ 1477| 2.06M| ? mb->mvs[i].x - last_mb->mvs[i].x 1478| 2.06M| : last_mb->mvs[i].x*(dec->time_bp - dec->time_pp)/dec->time_pp; 1479| 2.06M| mb->b_mvs[i].y = (mv.y) ------------------ | Branch (1479:28): [True: 171k, False: 1.89M] ------------------ 1480| 2.06M| ? mb->mvs[i].y - last_mb->mvs[i].y 1481| 2.06M| : last_mb->mvs[i].y*(dec->time_bp - dec->time_pp)/dec->time_pp; 1482| 2.06M| } 1483| | 1484| 517k| decoder_bf_interpolate_mbinter(dec, dec->refn[1], dec->refn[0], 1485| 517k| mb, x, y, bs, 1); 1486| 517k| break; 1487| | 1488| 102k| case MODE_INTERPOLATE: ------------------ | | 45| 102k|#define MODE_INTERPOLATE 1 ------------------ | Branch (1488:7): [True: 102k, False: 4.47M] ------------------ 1489| 102k| get_b_motion_vector(bs, &mb->mvs[0], fcode_forward, dec->p_fmv, dec, x, y); 1490| 102k| dec->p_fmv = mb->mvs[1] = mb->mvs[2] = mb->mvs[3] = mb->mvs[0]; 1491| | 1492| 102k| get_b_motion_vector(bs, &mb->b_mvs[0], fcode_backward, dec->p_bmv, dec, x, y); 1493| 102k| dec->p_bmv = mb->b_mvs[1] = mb->b_mvs[2] = mb->b_mvs[3] = mb->b_mvs[0]; 1494| | 1495| 102k| decoder_bf_interpolate_mbinter(dec, dec->refn[1], dec->refn[0], 1496| 102k| mb, x, y, bs, 0); 1497| 102k| break; 1498| | 1499| 44.1k| case MODE_BACKWARD: ------------------ | | 46| 44.1k|#define MODE_BACKWARD 2 ------------------ | Branch (1499:7): [True: 44.1k, False: 4.53M] ------------------ 1500| 44.1k| get_b_motion_vector(bs, &mb->mvs[0], fcode_backward, dec->p_bmv, dec, x, y); 1501| 44.1k| dec->p_bmv = mb->mvs[1] = mb->mvs[2] = mb->mvs[3] = mb->mvs[0]; 1502| | 1503| 44.1k| decoder_mbinter(dec, mb, x, y, mb->cbp, bs, 0, 0, 1); 1504| 44.1k| break; 1505| | 1506| 24.9k| case MODE_FORWARD: ------------------ | | 47| 24.9k|#define MODE_FORWARD 3 ------------------ | Branch (1506:7): [True: 24.9k, False: 4.55M] ------------------ 1507| 24.9k| get_b_motion_vector(bs, &mb->mvs[0], fcode_forward, dec->p_fmv, dec, x, y); 1508| 24.9k| dec->p_fmv = mb->mvs[1] = mb->mvs[2] = mb->mvs[3] = mb->mvs[0]; 1509| | 1510| 24.9k| decoder_mbinter(dec, mb, x, y, mb->cbp, bs, 0, 1, 1); 1511| 24.9k| break; 1512| | 1513| 3.89M| default: ------------------ | Branch (1513:7): [True: 3.89M, False: 689k] ------------------ 1514| 3.89M| DPRINTF(XVID_DEBUG_ERROR,"Not supported B-frame mb_type = %i\n", mb->mode); ------------------ | | 196| 3.89M|#define XVID_DEBUG_ERROR (1<< 0) ------------------ 1515| 4.58M| } 1516| 4.58M| } /* End of for */ 1517| 119k| } 1518| 2.24k|} decoder.c:get_resync_len_b: 1348| 6.08k| const int fcode_forward) { 1349| 6.08k| int resync_len = ((fcode_forward>fcode_backward) ? fcode_forward : fcode_backward) - 1; ------------------ | Branch (1349:21): [True: 1.35k, False: 4.73k] ------------------ 1350| 6.08k| if (resync_len < 1) resync_len = 1; ------------------ | Branch (1350:7): [True: 2.71k, False: 3.37k] ------------------ 1351| 6.08k| return resync_len; 1352| 6.08k|} decoder.c:get_mbtype: 1337| 4.19M|{ 1338| 4.19M| int32_t mb_type; 1339| | 1340| 20.0M| for (mb_type = 0; mb_type <= 3; mb_type++) ------------------ | Branch (1340:21): [True: 16.1M, False: 3.89M] ------------------ 1341| 16.1M| if (BitstreamGetBit(bs)) ------------------ | Branch (1341:9): [True: 301k, False: 15.8M] ------------------ 1342| 301k| return (mb_type); 1343| | 1344| 3.89M| return -1; 1345| 4.19M|} decoder.c:get_dbquant: 1318| 95.4k|{ 1319| 95.4k| if (!BitstreamGetBit(bs)) /* '0' */ ------------------ | Branch (1319:7): [True: 60.1k, False: 35.2k] ------------------ 1320| 60.1k| return (0); 1321| 35.2k| else if (!BitstreamGetBit(bs)) /* '10' */ ------------------ | Branch (1321:12): [True: 18.2k, False: 17.0k] ------------------ 1322| 18.2k| return (-2); 1323| 17.0k| else /* '11' */ 1324| 17.0k| return (2); 1325| 95.4k|} decoder.c:get_b_motion_vector: 1127| 404k|{ 1128| 404k| const int scale_fac = 1 << (fcode - 1); 1129| 404k| const int high = (32 * scale_fac) - 1; 1130| 404k| const int low = ((-32) * scale_fac); 1131| 404k| const int range = (64 * scale_fac); 1132| | 1133| 404k| int mv_x = get_mv(bs, fcode); 1134| 404k| int mv_y = get_mv(bs, fcode); 1135| | 1136| 404k| mv_x += pmv.x; 1137| 404k| mv_y += pmv.y; 1138| | 1139| 404k| if (mv_x < low) ------------------ | Branch (1139:7): [True: 62.9k, False: 341k] ------------------ 1140| 62.9k| mv_x += range; 1141| 341k| else if (mv_x > high) ------------------ | Branch (1141:12): [True: 105k, False: 236k] ------------------ 1142| 105k| mv_x -= range; 1143| | 1144| 404k| if (mv_y < low) ------------------ | Branch (1144:7): [True: 59.0k, False: 345k] ------------------ 1145| 59.0k| mv_y += range; 1146| 345k| else if (mv_y > high) ------------------ | Branch (1146:12): [True: 108k, False: 237k] ------------------ 1147| 108k| mv_y -= range; 1148| | 1149| 404k| mv->x = mv_x; 1150| 404k| mv->y = mv_y; 1151| 404k|} decoder.c:decoder_bf_interpolate_mbinter: 1163| 620k|{ 1164| 620k| uint32_t stride = dec->edged_width; 1165| 620k| uint32_t stride2 = stride / 2; 1166| 620k| int uv_dx, uv_dy; 1167| 620k| int b_uv_dx, b_uv_dy; 1168| 620k| uint8_t *pY_Cur, *pU_Cur, *pV_Cur; 1169| 620k| const uint32_t cbp = pMB->cbp; 1170| | 1171| 620k| pY_Cur = dec->cur.y + (y_pos << 4) * stride + (x_pos << 4); 1172| 620k| pU_Cur = dec->cur.u + (y_pos << 3) * stride2 + (x_pos << 3); 1173| 620k| pV_Cur = dec->cur.v + (y_pos << 3) * stride2 + (x_pos << 3); 1174| | 1175| 620k| validate_vector(pMB->mvs, x_pos, y_pos, dec); 1176| 620k| validate_vector(pMB->b_mvs, x_pos, y_pos, dec); 1177| | 1178| 620k| if (!direct) { ------------------ | Branch (1178:7): [True: 102k, False: 517k] ------------------ 1179| 102k| uv_dx = pMB->mvs[0].x; 1180| 102k| uv_dy = pMB->mvs[0].y; 1181| 102k| b_uv_dx = pMB->b_mvs[0].x; 1182| 102k| b_uv_dy = pMB->b_mvs[0].y; 1183| | 1184| 102k| if (dec->quarterpel) { ------------------ | Branch (1184:9): [True: 95.6k, False: 7.34k] ------------------ 1185| 95.6k| if (dec->bs_version <= BS_VERSION_BUGGY_CHROMA_ROUNDING) { ------------------ | | 470| 95.6k|#define BS_VERSION_BUGGY_CHROMA_ROUNDING 1 ------------------ | Branch (1185:8): [True: 549, False: 95.0k] ------------------ 1186| 549| uv_dx = (uv_dx>>1) | (uv_dx&1); 1187| 549| uv_dy = (uv_dy>>1) | (uv_dy&1); 1188| 549| b_uv_dx = (b_uv_dx>>1) | (b_uv_dx&1); 1189| 549| b_uv_dy = (b_uv_dy>>1) | (b_uv_dy&1); 1190| 549| } 1191| 95.0k| else { 1192| 95.0k| uv_dx /= 2; 1193| 95.0k| uv_dy /= 2; 1194| 95.0k| b_uv_dx /= 2; 1195| 95.0k| b_uv_dy /= 2; 1196| 95.0k| } 1197| 95.6k| } 1198| | 1199| 102k| uv_dx = (uv_dx >> 1) + roundtab_79[uv_dx & 0x3]; 1200| 102k| uv_dy = (uv_dy >> 1) + roundtab_79[uv_dy & 0x3]; 1201| 102k| b_uv_dx = (b_uv_dx >> 1) + roundtab_79[b_uv_dx & 0x3]; 1202| 102k| b_uv_dy = (b_uv_dy >> 1) + roundtab_79[b_uv_dy & 0x3]; 1203| | 1204| 517k| } else { 1205| 517k| if (dec->quarterpel) { /* for qpel the /2 shall be done before summation. We've done it right in the encoder in the past. */ ------------------ | Branch (1205:8): [True: 387k, False: 130k] ------------------ 1206| | /* TODO: figure out if we ever did it wrong on the encoder side. If yes, add some workaround */ 1207| 387k| if (dec->bs_version <= BS_VERSION_BUGGY_CHROMA_ROUNDING) { ------------------ | | 470| 387k|#define BS_VERSION_BUGGY_CHROMA_ROUNDING 1 ------------------ | Branch (1207:7): [True: 1.77k, False: 385k] ------------------ 1208| 1.77k| int z; 1209| 1.77k| uv_dx = 0; uv_dy = 0; 1210| 1.77k| b_uv_dx = 0; b_uv_dy = 0; 1211| 8.86k| for (z = 0; z < 4; z++) { ------------------ | Branch (1211:16): [True: 7.08k, False: 1.77k] ------------------ 1212| 7.08k| uv_dx += ((pMB->mvs[z].x>>1) | (pMB->mvs[z].x&1)); 1213| 7.08k| uv_dy += ((pMB->mvs[z].y>>1) | (pMB->mvs[z].y&1)); 1214| 7.08k| b_uv_dx += ((pMB->b_mvs[z].x>>1) | (pMB->b_mvs[z].x&1)); 1215| 7.08k| b_uv_dy += ((pMB->b_mvs[z].y>>1) | (pMB->b_mvs[z].y&1)); 1216| 7.08k| } 1217| 1.77k| } 1218| 385k| else { 1219| 385k| uv_dx = (pMB->mvs[0].x / 2) + (pMB->mvs[1].x / 2) + (pMB->mvs[2].x / 2) + (pMB->mvs[3].x / 2); 1220| 385k| uv_dy = (pMB->mvs[0].y / 2) + (pMB->mvs[1].y / 2) + (pMB->mvs[2].y / 2) + (pMB->mvs[3].y / 2); 1221| 385k| b_uv_dx = (pMB->b_mvs[0].x / 2) + (pMB->b_mvs[1].x / 2) + (pMB->b_mvs[2].x / 2) + (pMB->b_mvs[3].x / 2); 1222| 385k| b_uv_dy = (pMB->b_mvs[0].y / 2) + (pMB->b_mvs[1].y / 2) + (pMB->b_mvs[2].y / 2) + (pMB->b_mvs[3].y / 2); 1223| 385k| } 1224| 387k| } else { 1225| 130k| uv_dx = pMB->mvs[0].x + pMB->mvs[1].x + pMB->mvs[2].x + pMB->mvs[3].x; 1226| 130k| uv_dy = pMB->mvs[0].y + pMB->mvs[1].y + pMB->mvs[2].y + pMB->mvs[3].y; 1227| 130k| b_uv_dx = pMB->b_mvs[0].x + pMB->b_mvs[1].x + pMB->b_mvs[2].x + pMB->b_mvs[3].x; 1228| 130k| b_uv_dy = pMB->b_mvs[0].y + pMB->b_mvs[1].y + pMB->b_mvs[2].y + pMB->b_mvs[3].y; 1229| 130k| } 1230| | 1231| 517k| uv_dx = (uv_dx >> 3) + roundtab_76[uv_dx & 0xf]; 1232| 517k| uv_dy = (uv_dy >> 3) + roundtab_76[uv_dy & 0xf]; 1233| 517k| b_uv_dx = (b_uv_dx >> 3) + roundtab_76[b_uv_dx & 0xf]; 1234| 517k| b_uv_dy = (b_uv_dy >> 3) + roundtab_76[b_uv_dy & 0xf]; 1235| 517k| } 1236| | 1237| 620k| start_timer(); 1238| 620k| if(dec->quarterpel) { ------------------ | Branch (1238:6): [True: 482k, False: 137k] ------------------ 1239| 482k| if(!direct) { ------------------ | Branch (1239:8): [True: 95.6k, False: 387k] ------------------ 1240| 95.6k| interpolate16x16_quarterpel(dec->cur.y, forward.y, dec->qtmp.y, dec->qtmp.y + 64, 1241| 95.6k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 1242| 95.6k| pMB->mvs[0].x, pMB->mvs[0].y, stride, 0); 1243| 387k| } else { 1244| 387k| interpolate8x8_quarterpel(dec->cur.y, forward.y, dec->qtmp.y, dec->qtmp.y + 64, 1245| 387k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 1246| 387k| pMB->mvs[0].x, pMB->mvs[0].y, stride, 0); 1247| 387k| interpolate8x8_quarterpel(dec->cur.y, forward.y, dec->qtmp.y, dec->qtmp.y + 64, 1248| 387k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos, 1249| 387k| pMB->mvs[1].x, pMB->mvs[1].y, stride, 0); 1250| 387k| interpolate8x8_quarterpel(dec->cur.y, forward.y, dec->qtmp.y, dec->qtmp.y + 64, 1251| 387k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos + 8, 1252| 387k| pMB->mvs[2].x, pMB->mvs[2].y, stride, 0); 1253| 387k| interpolate8x8_quarterpel(dec->cur.y, forward.y, dec->qtmp.y, dec->qtmp.y + 64, 1254| 387k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos + 8, 1255| 387k| pMB->mvs[3].x, pMB->mvs[3].y, stride, 0); 1256| 387k| } 1257| 482k| } else { 1258| 137k| interpolate8x8_switch(dec->cur.y, forward.y, 16 * x_pos, 16 * y_pos, 1259| 137k| pMB->mvs[0].x, pMB->mvs[0].y, stride, 0); 1260| 137k| interpolate8x8_switch(dec->cur.y, forward.y, 16 * x_pos + 8, 16 * y_pos, 1261| 137k| pMB->mvs[1].x, pMB->mvs[1].y, stride, 0); 1262| 137k| interpolate8x8_switch(dec->cur.y, forward.y, 16 * x_pos, 16 * y_pos + 8, 1263| 137k| pMB->mvs[2].x, pMB->mvs[2].y, stride, 0); 1264| 137k| interpolate8x8_switch(dec->cur.y, forward.y, 16 * x_pos + 8, 16 * y_pos + 8, 1265| 137k| pMB->mvs[3].x, pMB->mvs[3].y, stride, 0); 1266| 137k| } 1267| | 1268| 620k| interpolate8x8_switch(dec->cur.u, forward.u, 8 * x_pos, 8 * y_pos, uv_dx, 1269| 620k| uv_dy, stride2, 0); 1270| 620k| interpolate8x8_switch(dec->cur.v, forward.v, 8 * x_pos, 8 * y_pos, uv_dx, 1271| 620k| uv_dy, stride2, 0); 1272| | 1273| | 1274| 620k| if(dec->quarterpel) { ------------------ | Branch (1274:6): [True: 482k, False: 137k] ------------------ 1275| 482k| if(!direct) { ------------------ | Branch (1275:8): [True: 95.6k, False: 387k] ------------------ 1276| 95.6k| interpolate16x16_add_quarterpel(dec->cur.y, backward.y, dec->qtmp.y, dec->qtmp.y + 64, 1277| 95.6k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 1278| 95.6k| pMB->b_mvs[0].x, pMB->b_mvs[0].y, stride, 0); 1279| 387k| } else { 1280| 387k| interpolate8x8_add_quarterpel(dec->cur.y, backward.y, dec->qtmp.y, dec->qtmp.y + 64, 1281| 387k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos, 1282| 387k| pMB->b_mvs[0].x, pMB->b_mvs[0].y, stride, 0); 1283| 387k| interpolate8x8_add_quarterpel(dec->cur.y, backward.y, dec->qtmp.y, dec->qtmp.y + 64, 1284| 387k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos, 1285| 387k| pMB->b_mvs[1].x, pMB->b_mvs[1].y, stride, 0); 1286| 387k| interpolate8x8_add_quarterpel(dec->cur.y, backward.y, dec->qtmp.y, dec->qtmp.y + 64, 1287| 387k| dec->qtmp.y + 128, 16*x_pos, 16*y_pos + 8, 1288| 387k| pMB->b_mvs[2].x, pMB->b_mvs[2].y, stride, 0); 1289| 387k| interpolate8x8_add_quarterpel(dec->cur.y, backward.y, dec->qtmp.y, dec->qtmp.y + 64, 1290| 387k| dec->qtmp.y + 128, 16*x_pos + 8, 16*y_pos + 8, 1291| 387k| pMB->b_mvs[3].x, pMB->b_mvs[3].y, stride, 0); 1292| 387k| } 1293| 482k| } else { 1294| 137k| interpolate8x8_add_switch(dec->cur.y, backward.y, 16 * x_pos, 16 * y_pos, 1295| 137k| pMB->b_mvs[0].x, pMB->b_mvs[0].y, stride, 0); 1296| 137k| interpolate8x8_add_switch(dec->cur.y, backward.y, 16 * x_pos + 8, 1297| 137k| 16 * y_pos, pMB->b_mvs[1].x, pMB->b_mvs[1].y, stride, 0); 1298| 137k| interpolate8x8_add_switch(dec->cur.y, backward.y, 16 * x_pos, 1299| 137k| 16 * y_pos + 8, pMB->b_mvs[2].x, pMB->b_mvs[2].y, stride, 0); 1300| 137k| interpolate8x8_add_switch(dec->cur.y, backward.y, 16 * x_pos + 8, 1301| 137k| 16 * y_pos + 8, pMB->b_mvs[3].x, pMB->b_mvs[3].y, stride, 0); 1302| 137k| } 1303| | 1304| 620k| interpolate8x8_add_switch(dec->cur.u, backward.u, 8 * x_pos, 8 * y_pos, 1305| 620k| b_uv_dx, b_uv_dy, stride2, 0); 1306| 620k| interpolate8x8_add_switch(dec->cur.v, backward.v, 8 * x_pos, 8 * y_pos, 1307| 620k| b_uv_dx, b_uv_dy, stride2, 0); 1308| | 1309| 620k| stop_comp_timer(); 1310| | 1311| 620k| if (cbp) ------------------ | Branch (1311:7): [True: 66.8k, False: 553k] ------------------ 1312| 66.8k| decoder_mb_decode(dec, cbp, bs, pY_Cur, pU_Cur, pV_Cur, pMB); 1313| 620k|} decoder.c:coding2type: 64| 6.05k|{ 65| 6.05k| return coding_type + 1; 66| 6.05k|} decoder.c:get_dc_scaler: 233| 272M|{ 234| 272M| if (quant < 5) ------------------ | Branch (234:6): [True: 69.1M, False: 203M] ------------------ 235| 69.1M| return 8; 236| | 237| 203M| if (quant < 25 && !lum) ------------------ | Branch (237:6): [True: 110M, False: 93.3M] | Branch (237:20): [True: 36.7M, False: 73.4M] ------------------ 238| 36.7M| return (quant + 13) / 2; 239| | 240| 166M| if (quant < 9) ------------------ | Branch (240:6): [True: 27.3M, False: 139M] ------------------ 241| 27.3M| return 2 * quant; 242| | 243| 139M| if (quant < 25) ------------------ | Branch (243:6): [True: 46.0M, False: 93.3M] ------------------ 244| 46.0M| return quant + 8; 245| | 246| 93.3M| if (lum) ------------------ | Branch (246:6): [True: 62.2M, False: 31.1M] ------------------ 247| 62.2M| return 2 * quant - 16; 248| 31.1M| else 249| 31.1M| return quant - 6; 250| 93.3M|} yv12_to_bgr_c: 98| 3.85k| int width, int height, int vflip) \ 99| 3.85k|{ \ 100| 3.85k| int fixed_width = (width + 1) & ~1; \ 101| 3.85k| int x_dif = x_stride - (SIZE)*fixed_width; \ 102| 3.85k| int y_dif = y_stride - fixed_width; \ 103| 3.85k| int uv_dif = uv_stride - (fixed_width / 2); \ 104| 3.85k| int x, y; \ 105| 3.85k| if ((x_ptr == NULL) || (x_dif < 0)) return; \ ------------------ | Branch (105:6): [True: 0, False: 3.85k] | Branch (105:25): [True: 0, False: 3.85k] ------------------ 106| 3.85k| if (vflip) { \ ------------------ | Branch (106:6): [True: 0, False: 3.85k] ------------------ 107| 0| x_ptr += (height - 1) * x_stride; \ 108| 0| x_dif = -(SIZE)*fixed_width - x_stride; \ 109| 0| x_stride = -x_stride; \ 110| 0| } \ 111| 1.01M| for (y = 0; y < height; y+=(VPIXELS)) { \ ------------------ | Branch (111:14): [True: 1.01M, False: 3.85k] ------------------ 112| 1.01M| FUNC##_ROW(SIZE,C1,C2,C3,C4); \ 113| 583M| for (x = 0; x < fixed_width; x+=(PIXELS)) { \ ------------------ | Branch (113:15): [True: 582M, False: 1.01M] ------------------ 114| 582M| FUNC(SIZE,C1,C2,C3,C4); \ ------------------ | | 450| 582M|MAKE_COLORSPACE(yv12_to_bgr_c, 3,2,2, YV12_TO_RGB, 2,1,0,0) | | ------------------ | | | | 405| 582M| int rgb_y; \ | | | | 406| 582M| int b_u0 = B_U_tab[ u_ptr[0] ]; \ | | | | 407| 582M| int g_uv0 = G_U_tab[ u_ptr[0] ] + G_V_tab[ v_ptr[0] ]; \ | | | | 408| 582M| int r_v0 = R_V_tab[ v_ptr[0] ]; \ | | | | 409| 582M| WRITE_RGB(SIZE, 0, 0, C1,C2,C3,C4) \ | | | | ------------------ | | | | | | 391| 582M| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 582M| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.57M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.57M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 582M| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 29.0M, False: 553M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 548M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 582M| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 582M| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 582M| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.57M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.57M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 29.0M, False: 553M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 548M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 582M| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | | | 410| 582M| WRITE_RGB(SIZE, 1, 0, C1,C2,C3,C4) | | | | ------------------ | | | | | | 391| 582M| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 582M| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.57M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.57M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 582M| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 29.0M, False: 553M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 548M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 582M| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 582M| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 582M| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.57M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.57M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 29.0M, False: 553M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 548M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 582M| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 1.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 344M, False: 238M] | | | | | | | | | Branch (258:25): [True: 4.58M, False: 577M] | | | | | | | | | Branch (258:33): [True: 4.58M, False: 233M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 582M| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | ------------------ ------------------ 115| 582M| x_ptr += (PIXELS)*(SIZE); \ 116| 582M| y_ptr += (PIXELS); \ 117| 582M| u_ptr += (PIXELS)/2; \ 118| 582M| v_ptr += (PIXELS)/2; \ 119| 582M| } \ 120| 1.01M| x_ptr += x_dif + (VPIXELS-1)*x_stride; \ 121| 1.01M| y_ptr += y_dif + (VPIXELS-1)*y_stride; \ 122| 1.01M| u_ptr += uv_dif + ((VPIXELS/2)-1)*uv_stride; \ 123| 1.01M| v_ptr += uv_dif + ((VPIXELS/2)-1)*uv_stride; \ 124| 1.01M| } \ 125| 3.85k|} yv12_to_bgri_c: 98| 1.81k| int width, int height, int vflip) \ 99| 1.81k|{ \ 100| 1.81k| int fixed_width = (width + 1) & ~1; \ 101| 1.81k| int x_dif = x_stride - (SIZE)*fixed_width; \ 102| 1.81k| int y_dif = y_stride - fixed_width; \ 103| 1.81k| int uv_dif = uv_stride - (fixed_width / 2); \ 104| 1.81k| int x, y; \ 105| 1.81k| if ((x_ptr == NULL) || (x_dif < 0)) return; \ ------------------ | Branch (105:6): [True: 0, False: 1.81k] | Branch (105:25): [True: 0, False: 1.81k] ------------------ 106| 1.81k| if (vflip) { \ ------------------ | Branch (106:6): [True: 0, False: 1.81k] ------------------ 107| 0| x_ptr += (height - 1) * x_stride; \ 108| 0| x_dif = -(SIZE)*fixed_width - x_stride; \ 109| 0| x_stride = -x_stride; \ 110| 0| } \ 111| 1.88M| for (y = 0; y < height; y+=(VPIXELS)) { \ ------------------ | Branch (111:14): [True: 1.88M, False: 1.81k] ------------------ 112| 1.88M| FUNC##_ROW(SIZE,C1,C2,C3,C4); \ 113| 1.08G| for (x = 0; x < fixed_width; x+=(PIXELS)) { \ ------------------ | Branch (113:15): [True: 1.08G, False: 1.88M] ------------------ 114| 1.08G| FUNC(SIZE,C1,C2,C3,C4); \ ------------------ | | 461| 1.08G|MAKE_COLORSPACE(yv12_to_bgri_c, 3,2,4, YV12_TO_RGBI, 2,1,0, 0) | | ------------------ | | | | 414| 1.08G| int rgb_y; \ | | | | 415| 1.08G| int b_u0 = B_U_tab[ u_ptr[0] ]; \ | | | | 416| 1.08G| int g_uv0 = G_U_tab[ u_ptr[0] ] + G_V_tab[ v_ptr[0] ]; \ | | | | 417| 1.08G| int r_v0 = R_V_tab[ v_ptr[0] ]; \ | | | | 418| 1.08G| int b_u1 = B_U_tab[ u_ptr[uv_stride] ]; \ | | | | 419| 1.08G| int g_uv1 = G_U_tab[ u_ptr[uv_stride] ] + G_V_tab[ v_ptr[uv_stride] ]; \ | | | | 420| 1.08G| int r_v1 = R_V_tab[ v_ptr[uv_stride] ]; \ | | | | 421| 1.08G| WRITE_RGB(SIZE, 0, 0, C1,C2,C3,C4) \ | | | | ------------------ | | | | | | 391| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 1.08G| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 400k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 400k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 1.08G| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 837k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 837k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 1.08G| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 457k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 457k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 407k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 407k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 875k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 875k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 464k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 464k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | | | 422| 1.08G| WRITE_RGB(SIZE, 1, 1, C1,C2,C3,C4) \ | | | | ------------------ | | | | | | 391| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 1.08G| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 410k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 410k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 1.08G| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 830k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 830k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 1.08G| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 470k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 470k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 404k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 404k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 836k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 836k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 463k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 463k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | | | 423| 1.08G| WRITE_RGB(SIZE, 2, 0, C1,C2,C3,C4) \ | | | | ------------------ | | | | | | 391| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 1.08G| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 438k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 438k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 1.08G| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 857k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 857k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 1.08G| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 502k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 502k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 430k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 430k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 850k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 850k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 493k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 493k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | | | 424| 1.08G| WRITE_RGB(SIZE, 3, 1, C1,C2,C3,C4) | | | | ------------------ | | | | | | 391| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 0] ]; \ | | | | | | 392| 1.08G| x_ptr[(ROW)*x_stride+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 445k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 445k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 393| 1.08G| x_ptr[(ROW)*x_stride+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 863k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 863k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 394| 1.08G| x_ptr[(ROW)*x_stride+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 507k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 507k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 395| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(C4)] = 0; \ | | | | | | ------------------ | | | | | | | Branch (395:6): [Folded - Ignored] | | | | | | ------------------ | | | | | | 396| 1.08G| rgb_y = RGB_Y_tab[ y_ptr[(ROW)*y_stride + 1] ]; \ | | | | | | 397| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C3)] = MAX(0, MIN(255, (rgb_y + b_u##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 655M, False: 429M] | | | | | | | | | Branch (258:25): [True: 420k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 420k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 398| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C2)] = MAX(0, MIN(255, (rgb_y - g_uv##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 18.6M, False: 1.06G] | | | | | | | | | Branch (258:25): [True: 859k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 859k, False: 1.06G] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 399| 1.08G| x_ptr[(ROW)*x_stride+(SIZE)+(C1)] = MAX(0, MIN(255, (rgb_y + r_v##UV_ROW) >> SCALEBITS_OUT)); \ | | | | | | ------------------ | | | | | | | | 258| 2.16G|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | | | | | | | ------------------ | | | | | | | | | Branch (258:20): [True: 654M, False: 429M] | | | | | | | | | Branch (258:25): [True: 484k, False: 1.08G] | | | | | | | | | Branch (258:33): [True: 484k, False: 428M] | | | | | | | | ------------------ | | | | | | ------------------ | | | | | | 400| 1.08G| if ((SIZE)>3) x_ptr[(ROW)*x_stride+(SIZE)+(C4)] = 0; | | | | | | ------------------ | | | | | | | Branch (400:6): [Folded - Ignored] | | | | | | ------------------ | | | | ------------------ | | ------------------ ------------------ 115| 1.08G| x_ptr += (PIXELS)*(SIZE); \ 116| 1.08G| y_ptr += (PIXELS); \ 117| 1.08G| u_ptr += (PIXELS)/2; \ 118| 1.08G| v_ptr += (PIXELS)/2; \ 119| 1.08G| } \ 120| 1.88M| x_ptr += x_dif + (VPIXELS-1)*x_stride; \ 121| 1.88M| y_ptr += y_dif + (VPIXELS-1)*y_stride; \ 122| 1.88M| u_ptr += uv_dif + ((VPIXELS/2)-1)*uv_stride; \ 123| 1.88M| v_ptr += uv_dif + ((VPIXELS/2)-1)*uv_stride; \ 124| 1.88M| } \ 125| 1.81k|} colorspace_init: 528| 2|{ 529| 2| int32_t i; 530| | 531| 514| for (i = 0; i < 256; i++) { ------------------ | Branch (531:14): [True: 512, False: 2] ------------------ 532| 512| RGB_Y_tab[i] = FIX_OUT(RGB_Y_OUT) * (i - Y_ADD_OUT); ------------------ | | 331| 512|#define FIX_OUT(x) ((uint16_t) ((x) * (1L<= '!' && buf[i] <= '@') ------------------ | Branch (599:7): [True: 111k, False: 14.6k] | Branch (599:24): [True: 16.2k, False: 95.2k] ------------------ 600| 16.2k| font = ascii33[buf[i]-'!']; 601| 109k| else if (buf[i] >= 'A' && buf[i] <= 'Z') ------------------ | Branch (601:12): [True: 95.2k, False: 14.6k] | Branch (601:29): [True: 0, False: 95.2k] ------------------ 602| 0| font = ascii65[buf[i]-'A']; 603| 109k| else if (buf[i] >= '[' && buf[i] <= '`') ------------------ | Branch (603:12): [True: 95.2k, False: 14.6k] | Branch (603:29): [True: 0, False: 95.2k] ------------------ 604| 0| font = ascii91[buf[i]-'[']; 605| 109k| else if (buf[i] >= 'a' && buf[i] <= 'z') ------------------ | Branch (605:12): [True: 95.2k, False: 14.6k] | Branch (605:29): [True: 95.2k, False: 0] ------------------ 606| 95.2k| font = ascii65[buf[i]-'a']; 607| 14.6k| else 608| 14.6k| continue; 609| | 610| 111k| draw_num(img, edged_width, height, font, x + i*FONT_ZOOM*(FONT_WIDTH+1), y); ------------------ | | 562| 111k|#define FONT_ZOOM 4 ------------------ draw_num(img, edged_width, height, font, x + i*FONT_ZOOM*(FONT_WIDTH+1), y); ------------------ | | 32| 111k|#define FONT_WIDTH 4 ------------------ 611| 111k| } 612| 4.09k|} font.c:draw_num: 567| 111k|{ 568| 111k| int i, j; 569| | 570| 1.19M| for (j = 0; j < FONT_ZOOM * FONT_HEIGHT && y+j < height; j++) ------------------ | | 562| 1.19M|#define FONT_ZOOM 4 ------------------ for (j = 0; j < FONT_ZOOM * FONT_HEIGHT && y+j < height; j++) ------------------ | | 33| 2.39M|#define FONT_HEIGHT 6 ------------------ | Branch (570:14): [True: 1.15M, False: 41.1k] | Branch (570:45): [True: 1.08M, False: 70.2k] ------------------ 571| 8.40M| for (i = 0; i < FONT_ZOOM * FONT_WIDTH && x+i < stride; i++) ------------------ | | 562| 8.40M|#define FONT_ZOOM 4 ------------------ for (i = 0; i < FONT_ZOOM * FONT_WIDTH && x+i < stride; i++) ------------------ | | 32| 16.8M|#define FONT_WIDTH 4 ------------------ | Branch (571:15): [True: 7.95M, False: 451k] | Branch (571:45): [True: 7.31M, False: 636k] ------------------ 572| 7.31M| if (font[(j/FONT_ZOOM)*FONT_WIDTH + (i/FONT_ZOOM)]) ------------------ | | 562| 7.31M|#define FONT_ZOOM 4 ------------------ if (font[(j/FONT_ZOOM)*FONT_WIDTH + (i/FONT_ZOOM)]) ------------------ | | 32| 7.31M|#define FONT_WIDTH 4 ------------------ if (font[(j/FONT_ZOOM)*FONT_WIDTH + (i/FONT_ZOOM)]) ------------------ | | 562| 7.31M|#define FONT_ZOOM 4 ------------------ | Branch (572:8): [True: 3.99M, False: 3.32M] ------------------ 573| 3.99M| { 574| 3.99M| int offset = (y+j)*stride + (x+i); 575| 3.99M| int offset2 =((y+j)/2)*(stride/2) + ((x+i)/2); 576| 3.99M| img->y[offset] = 255; 577| 3.99M| img->u[offset2] = 127; 578| 3.99M| img->v[offset2] = 127; 579| 3.99M| } 580| 111k|} image_create: 49| 139k|{ 50| 139k| const uint32_t edged_width2 = edged_width / 2; 51| 139k| const uint32_t edged_height2 = edged_height / 2; 52| | 53| 139k| image->y = 54| 139k| xvid_malloc(edged_width * (edged_height + 1) + SAFETY, CACHE_LINE); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ xvid_malloc(edged_width * (edged_height + 1) + SAFETY, CACHE_LINE); ------------------ | | 131| 139k|# define CACHE_LINE 64 ------------------ 55| 139k| if (image->y == NULL) { ------------------ | Branch (55:6): [True: 0, False: 139k] ------------------ 56| 0| return -1; 57| 0| } 58| 139k| memset(image->y, 0, edged_width * (edged_height + 1) + SAFETY); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ 59| | 60| 139k| image->u = xvid_malloc(edged_width2 * edged_height2 + SAFETY, CACHE_LINE); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ image->u = xvid_malloc(edged_width2 * edged_height2 + SAFETY, CACHE_LINE); ------------------ | | 131| 139k|# define CACHE_LINE 64 ------------------ 61| 139k| if (image->u == NULL) { ------------------ | Branch (61:6): [True: 0, False: 139k] ------------------ 62| 0| xvid_free(image->y); 63| 0| image->y = NULL; 64| 0| return -1; 65| 0| } 66| 139k| memset(image->u, 0, edged_width2 * edged_height2 + SAFETY); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ 67| | 68| 139k| image->v = xvid_malloc(edged_width2 * edged_height2 + SAFETY, CACHE_LINE); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ image->v = xvid_malloc(edged_width2 * edged_height2 + SAFETY, CACHE_LINE); ------------------ | | 131| 139k|# define CACHE_LINE 64 ------------------ 69| 139k| if (image->v == NULL) { ------------------ | Branch (69:6): [True: 0, False: 139k] ------------------ 70| 0| xvid_free(image->u); 71| 0| image->u = NULL; 72| 0| xvid_free(image->y); 73| 0| image->y = NULL; 74| 0| return -1; 75| 0| } 76| 139k| memset(image->v, 0, edged_width2 * edged_height2 + SAFETY); ------------------ | | 41| 139k|#define SAFETY 64 ------------------ 77| | 78| 139k| image->y += EDGE_SIZE * edged_width + EDGE_SIZE; ------------------ | | 36| 139k|#define EDGE_SIZE 64 ------------------ image->y += EDGE_SIZE * edged_width + EDGE_SIZE; ------------------ | | 36| 139k|#define EDGE_SIZE 64 ------------------ 79| 139k| image->u += EDGE_SIZE2 * edged_width2 + EDGE_SIZE2; ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ image->u += EDGE_SIZE2 * edged_width2 + EDGE_SIZE2; ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ 80| 139k| image->v += EDGE_SIZE2 * edged_width2 + EDGE_SIZE2; ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ image->v += EDGE_SIZE2 * edged_width2 + EDGE_SIZE2; ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ 81| | 82| 139k| return 0; 83| 139k|} image_destroy: 91| 201k|{ 92| 201k| const uint32_t edged_width2 = edged_width / 2; 93| | 94| 201k| if (image->y) { ------------------ | Branch (94:6): [True: 139k, False: 61.7k] ------------------ 95| 139k| xvid_free(image->y - (EDGE_SIZE * edged_width + EDGE_SIZE)); ------------------ | | 36| 139k|#define EDGE_SIZE 64 ------------------ xvid_free(image->y - (EDGE_SIZE * edged_width + EDGE_SIZE)); ------------------ | | 36| 139k|#define EDGE_SIZE 64 ------------------ 96| 139k| image->y = NULL; 97| 139k| } 98| 201k| if (image->u) { ------------------ | Branch (98:6): [True: 139k, False: 61.7k] ------------------ 99| 139k| xvid_free(image->u - (EDGE_SIZE2 * edged_width2 + EDGE_SIZE2)); ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ xvid_free(image->u - (EDGE_SIZE2 * edged_width2 + EDGE_SIZE2)); ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ 100| 139k| image->u = NULL; 101| 139k| } 102| 201k| if (image->v) { ------------------ | Branch (102:6): [True: 139k, False: 61.7k] ------------------ 103| 139k| xvid_free(image->v - (EDGE_SIZE2 * edged_width2 + EDGE_SIZE2)); ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ xvid_free(image->v - (EDGE_SIZE2 * edged_width2 + EDGE_SIZE2)); ------------------ | | 42| 139k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 139k|#define EDGE_SIZE 64 | | ------------------ ------------------ 104| 139k| image->v = NULL; 105| 139k| } 106| 201k|} image_swap: 112| 21.0k|{ 113| 21.0k| SWAP(uint8_t*, image1->y, image2->y); ------------------ | | 264| 21.0k|#define SWAP(_T_,A,B) { _T_ tmp = A; A = B; B = tmp; } ------------------ 114| 21.0k| SWAP(uint8_t*, image1->u, image2->u); ------------------ | | 264| 21.0k|#define SWAP(_T_,A,B) { _T_ tmp = A; A = B; B = tmp; } ------------------ 115| 21.0k| SWAP(uint8_t*, image1->v, image2->v); ------------------ | | 264| 21.0k|#define SWAP(_T_,A,B) { _T_ tmp = A; A = B; B = tmp; } ------------------ 116| 21.0k|} image_copy: 124| 42|{ 125| 42| memcpy(image1->y, image2->y, edged_width * height); 126| 42| memcpy(image1->u, image2->u, edged_width * height / 4); 127| 42| memcpy(image1->v, image2->v, edged_width * height / 4); 128| 42|} image_setedges: 142| 7.93k|{ 143| 7.93k| const uint32_t edged_width2 = edged_width / 2; 144| 7.93k| uint32_t width2; 145| 7.93k| uint32_t i; 146| 7.93k| uint8_t *dst; 147| 7.93k| uint8_t *src; 148| | 149| 7.93k| dst = image->y - (EDGE_SIZE + EDGE_SIZE * edged_width); ------------------ | | 36| 7.93k|#define EDGE_SIZE 64 ------------------ dst = image->y - (EDGE_SIZE + EDGE_SIZE * edged_width); ------------------ | | 36| 7.93k|#define EDGE_SIZE 64 ------------------ 150| 7.93k| src = image->y; 151| | 152| | /* According to the Standard Clause 7.6.4, padding is done starting at 16 153| | * pixel width and height multiples. This was not respected in old xvids */ 154| 7.93k| if ((bs_version >= SETEDGES_BUG_BEFORE && ------------------ | | 131| 15.8k|#define SETEDGES_BUG_BEFORE 18 ------------------ | Branch (154:7): [True: 7.73k, False: 206] ------------------ 155| 7.93k| bs_version < SETEDGES_BUG_AFTER) || ------------------ | | 132| 7.73k|#define SETEDGES_BUG_AFTER 57 ------------------ | Branch (155:3): [True: 5, False: 7.72k] ------------------ 156| 7.93k| bs_version >= SETEDGES_BUG_REFIXED) { ------------------ | | 133| 7.93k|#define SETEDGES_BUG_REFIXED 63 ------------------ | Branch (156:3): [True: 7.72k, False: 207] ------------------ 157| 7.73k| width = (width+15)&~15; 158| 7.73k| height = (height+15)&~15; 159| 7.73k| } 160| | 161| 7.93k| width2 = MAX(1, width/2); ------------------ | | 258| 7.93k|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (258:20): [True: 36, False: 7.90k] | | ------------------ ------------------ 162| | 163| 515k| for (i = 0; i < EDGE_SIZE; i++) { ------------------ | | 36| 515k|#define EDGE_SIZE 64 ------------------ | Branch (163:14): [True: 508k, False: 7.93k] ------------------ 164| 508k| memset(dst, *src, EDGE_SIZE); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 165| 508k| memcpy(dst + EDGE_SIZE, src, width); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 166| 508k| memset(dst + edged_width - EDGE_SIZE, *(src + width - 1), ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 167| 508k| EDGE_SIZE); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 168| 508k| dst += edged_width; 169| 508k| } 170| | 171| 10.2M| for (i = 0; i < height; i++) { ------------------ | Branch (171:14): [True: 10.2M, False: 7.93k] ------------------ 172| 10.2M| memset(dst, *src, EDGE_SIZE); ------------------ | | 36| 10.2M|#define EDGE_SIZE 64 ------------------ 173| 10.2M| memset(dst + edged_width - EDGE_SIZE, src[width - 1], EDGE_SIZE); ------------------ | | 36| 10.2M|#define EDGE_SIZE 64 ------------------ memset(dst + edged_width - EDGE_SIZE, src[width - 1], EDGE_SIZE); ------------------ | | 36| 10.2M|#define EDGE_SIZE 64 ------------------ 174| 10.2M| dst += edged_width; 175| 10.2M| src += edged_width; 176| 10.2M| } 177| | 178| 7.93k| src -= edged_width; 179| 515k| for (i = 0; i < EDGE_SIZE; i++) { ------------------ | | 36| 515k|#define EDGE_SIZE 64 ------------------ | Branch (179:14): [True: 508k, False: 7.93k] ------------------ 180| 508k| memset(dst, *src, EDGE_SIZE); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 181| 508k| memcpy(dst + EDGE_SIZE, src, width); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 182| 508k| memset(dst + edged_width - EDGE_SIZE, *(src + width - 1), ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 183| 508k| EDGE_SIZE); ------------------ | | 36| 508k|#define EDGE_SIZE 64 ------------------ 184| 508k| dst += edged_width; 185| 508k| } 186| | 187| | 188| | /* U */ 189| 7.93k| dst = image->u - (EDGE_SIZE2 + EDGE_SIZE2 * edged_width2); ------------------ | | 42| 7.93k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 7.93k|#define EDGE_SIZE 64 | | ------------------ ------------------ dst = image->u - (EDGE_SIZE2 + EDGE_SIZE2 * edged_width2); ------------------ | | 42| 7.93k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 7.93k|#define EDGE_SIZE 64 | | ------------------ ------------------ 190| 7.93k| src = image->u; 191| | 192| 261k| for (i = 0; i < EDGE_SIZE2; i++) { ------------------ | | 42| 261k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 261k|#define EDGE_SIZE 64 | | ------------------ ------------------ | Branch (192:14): [True: 254k, False: 7.93k] ------------------ 193| 254k| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 194| 254k| memcpy(dst + EDGE_SIZE2, src, width2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 195| 254k| memset(dst + edged_width2 - EDGE_SIZE2, *(src + width2 - 1), ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 196| 254k| EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 197| 254k| dst += edged_width2; 198| 254k| } 199| | 200| 5.13M| for (i = 0; i < height / 2; i++) { ------------------ | Branch (200:14): [True: 5.12M, False: 7.93k] ------------------ 201| 5.12M| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ 202| 5.12M| memset(dst + edged_width2 - EDGE_SIZE2, src[width2 - 1], EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ memset(dst + edged_width2 - EDGE_SIZE2, src[width2 - 1], EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ 203| 5.12M| dst += edged_width2; 204| 5.12M| src += edged_width2; 205| 5.12M| } 206| 7.93k| src -= edged_width2; 207| 261k| for (i = 0; i < EDGE_SIZE2; i++) { ------------------ | | 42| 261k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 261k|#define EDGE_SIZE 64 | | ------------------ ------------------ | Branch (207:14): [True: 254k, False: 7.93k] ------------------ 208| 254k| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 209| 254k| memcpy(dst + EDGE_SIZE2, src, width2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 210| 254k| memset(dst + edged_width2 - EDGE_SIZE2, *(src + width2 - 1), ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 211| 254k| EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 212| 254k| dst += edged_width2; 213| 254k| } 214| | 215| | 216| | /* V */ 217| 7.93k| dst = image->v - (EDGE_SIZE2 + EDGE_SIZE2 * edged_width2); ------------------ | | 42| 7.93k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 7.93k|#define EDGE_SIZE 64 | | ------------------ ------------------ dst = image->v - (EDGE_SIZE2 + EDGE_SIZE2 * edged_width2); ------------------ | | 42| 7.93k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 7.93k|#define EDGE_SIZE 64 | | ------------------ ------------------ 218| 7.93k| src = image->v; 219| | 220| 261k| for (i = 0; i < EDGE_SIZE2; i++) { ------------------ | | 42| 261k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 261k|#define EDGE_SIZE 64 | | ------------------ ------------------ | Branch (220:14): [True: 254k, False: 7.93k] ------------------ 221| 254k| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 222| 254k| memcpy(dst + EDGE_SIZE2, src, width2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 223| 254k| memset(dst + edged_width2 - EDGE_SIZE2, *(src + width2 - 1), ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 224| 254k| EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 225| 254k| dst += edged_width2; 226| 254k| } 227| | 228| 5.13M| for (i = 0; i < height / 2; i++) { ------------------ | Branch (228:14): [True: 5.12M, False: 7.93k] ------------------ 229| 5.12M| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ 230| 5.12M| memset(dst + edged_width2 - EDGE_SIZE2, src[width2 - 1], EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ memset(dst + edged_width2 - EDGE_SIZE2, src[width2 - 1], EDGE_SIZE2); ------------------ | | 42| 5.12M|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 5.12M|#define EDGE_SIZE 64 | | ------------------ ------------------ 231| 5.12M| dst += edged_width2; 232| 5.12M| src += edged_width2; 233| 5.12M| } 234| 7.93k| src -= edged_width2; 235| 261k| for (i = 0; i < EDGE_SIZE2; i++) { ------------------ | | 42| 261k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 261k|#define EDGE_SIZE 64 | | ------------------ ------------------ | Branch (235:14): [True: 254k, False: 7.93k] ------------------ 236| 254k| memset(dst, *src, EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 237| 254k| memcpy(dst + EDGE_SIZE2, src, width2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 238| 254k| memset(dst + edged_width2 - EDGE_SIZE2, *(src + width2 - 1), ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 239| 254k| EDGE_SIZE2); ------------------ | | 42| 254k|#define EDGE_SIZE2 (EDGE_SIZE/2) | | ------------------ | | | | 36| 254k|#define EDGE_SIZE 64 | | ------------------ ------------------ 240| 254k| dst += edged_width2; 241| 254k| } 242| 7.93k|} image_output: 629| 6.05k|{ 630| 6.05k| const int edged_width2 = edged_width/2; 631| 6.05k| int height2 = height/2; 632| | 633| |/* 634| | if (interlacing) 635| | image_printf(image, edged_width, height, 5,100, "[i]=%i,%i",width,height); 636| | image_dump_yuvpgm(image, edged_width, width, height, "\\decode.pgm"); 637| |*/ 638| | 639| 6.05k| switch (csp & ~XVID_CSP_VFLIP) { ------------------ | | 127| 6.05k|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ | Branch (639:10): [True: 0, False: 6.05k] ------------------ 640| 0| case XVID_CSP_RGB555: ------------------ | | 122| 0|#define XVID_CSP_RGB555 (1<<10) /* 16-bit rgb555 packed */ ------------------ | Branch (640:2): [True: 0, False: 6.05k] ------------------ 641| 0| safe_packed_conv( 642| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 643| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 644| 0| interlacing?yv12_to_rgb555i :yv12_to_rgb555, ------------------ | Branch (644:4): [True: 0, False: 0] ------------------ 645| 0| interlacing?yv12_to_rgb555i_c:yv12_to_rgb555_c, 2, interlacing); ------------------ | Branch (645:4): [True: 0, False: 0] ------------------ 646| 0| return 0; 647| | 648| 0| case XVID_CSP_RGB565: ------------------ | | 123| 0|#define XVID_CSP_RGB565 (1<<11) /* 16-bit rgb565 packed */ ------------------ | Branch (648:2): [True: 0, False: 6.05k] ------------------ 649| 0| safe_packed_conv( 650| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 651| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 652| 0| interlacing?yv12_to_rgb565i :yv12_to_rgb565, ------------------ | Branch (652:4): [True: 0, False: 0] ------------------ 653| 0| interlacing?yv12_to_rgb565i_c:yv12_to_rgb565_c, 2, interlacing); ------------------ | Branch (653:4): [True: 0, False: 0] ------------------ 654| 0| return 0; 655| | 656| 6.05k| case XVID_CSP_BGR: ------------------ | | 121| 6.05k|#define XVID_CSP_BGR (1<< 9) /* 24-bit bgr packed */ ------------------ | Branch (656:5): [True: 6.05k, False: 0] ------------------ 657| 6.05k| safe_packed_conv( 658| 6.05k| dst[0], dst_stride[0], image->y, image->u, image->v, 659| 6.05k| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 6.05k|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 660| 6.05k| interlacing?yv12_to_bgri :yv12_to_bgr, ------------------ | Branch (660:4): [True: 1.94k, False: 4.10k] ------------------ 661| 6.05k| interlacing?yv12_to_bgri_c:yv12_to_bgr_c, 3, interlacing); ------------------ | Branch (661:4): [True: 1.94k, False: 4.10k] ------------------ 662| 6.05k| return 0; 663| | 664| 0| case XVID_CSP_BGRA: ------------------ | | 117| 0|#define XVID_CSP_BGRA (1<< 6) /* 32-bit bgra packed */ ------------------ | Branch (664:2): [True: 0, False: 6.05k] ------------------ 665| 0| safe_packed_conv( 666| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 667| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 668| 0| interlacing?yv12_to_bgrai :yv12_to_bgra, ------------------ | Branch (668:4): [True: 0, False: 0] ------------------ 669| 0| interlacing?yv12_to_bgrai_c:yv12_to_bgra_c, 4, interlacing); ------------------ | Branch (669:4): [True: 0, False: 0] ------------------ 670| 0| return 0; 671| | 672| 0| case XVID_CSP_ABGR: ------------------ | | 118| 0|#define XVID_CSP_ABGR (1<< 7) /* 32-bit abgr packed */ ------------------ | Branch (672:2): [True: 0, False: 6.05k] ------------------ 673| 0| safe_packed_conv( 674| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 675| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 676| 0| interlacing?yv12_to_abgri :yv12_to_abgr, ------------------ | Branch (676:4): [True: 0, False: 0] ------------------ 677| 0| interlacing?yv12_to_abgri_c:yv12_to_abgr_c, 4, interlacing); ------------------ | Branch (677:4): [True: 0, False: 0] ------------------ 678| 0| return 0; 679| | 680| 0| case XVID_CSP_RGB: ------------------ | | 116| 0|#define XVID_CSP_RGB (1<<16) /* 24-bit rgb packed */ ------------------ | Branch (680:2): [True: 0, False: 6.05k] ------------------ 681| 0| safe_packed_conv( 682| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 683| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 684| 0| interlacing?yv12_to_rgbi :yv12_to_rgb, ------------------ | Branch (684:4): [True: 0, False: 0] ------------------ 685| 0| interlacing?yv12_to_rgbi_c:yv12_to_rgb_c, 3, interlacing); ------------------ | Branch (685:4): [True: 0, False: 0] ------------------ 686| 0| return 0; 687| | 688| 0| case XVID_CSP_RGBA: ------------------ | | 119| 0|#define XVID_CSP_RGBA (1<< 8) /* 32-bit rgba packed */ ------------------ | Branch (688:2): [True: 0, False: 6.05k] ------------------ 689| 0| safe_packed_conv( 690| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 691| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 692| 0| interlacing?yv12_to_rgbai :yv12_to_rgba, ------------------ | Branch (692:4): [True: 0, False: 0] ------------------ 693| 0| interlacing?yv12_to_rgbai_c:yv12_to_rgba_c, 4, interlacing); ------------------ | Branch (693:4): [True: 0, False: 0] ------------------ 694| 0| return 0; 695| | 696| 0| case XVID_CSP_ARGB: ------------------ | | 120| 0|#define XVID_CSP_ARGB (1<<15) /* 32-bit argb packed */ ------------------ | Branch (696:2): [True: 0, False: 6.05k] ------------------ 697| 0| safe_packed_conv( 698| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 699| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 700| 0| interlacing?yv12_to_argbi :yv12_to_argb, ------------------ | Branch (700:4): [True: 0, False: 0] ------------------ 701| 0| interlacing?yv12_to_argbi_c:yv12_to_argb_c, 4, interlacing); ------------------ | Branch (701:4): [True: 0, False: 0] ------------------ 702| 0| return 0; 703| | 704| 0| case XVID_CSP_YUY2: ------------------ | | 113| 0|#define XVID_CSP_YUY2 (1<< 3) /* 4:2:2 packed */ ------------------ | Branch (704:2): [True: 0, False: 6.05k] ------------------ 705| 0| safe_packed_conv( 706| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 707| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 708| 0| interlacing?yv12_to_yuyvi :yv12_to_yuyv, ------------------ | Branch (708:4): [True: 0, False: 0] ------------------ 709| 0| interlacing?yv12_to_yuyvi_c:yv12_to_yuyv_c, 2, interlacing); ------------------ | Branch (709:4): [True: 0, False: 0] ------------------ 710| 0| return 0; 711| | 712| 0| case XVID_CSP_YVYU: /* u,v swapped */ ------------------ | | 115| 0|#define XVID_CSP_YVYU (1<< 5) /* 4:2:2 packed */ ------------------ | Branch (712:2): [True: 0, False: 6.05k] ------------------ 713| 0| safe_packed_conv( 714| 0| dst[0], dst_stride[0], image->y, image->v, image->u, 715| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 716| 0| interlacing?yv12_to_yuyvi :yv12_to_yuyv, ------------------ | Branch (716:4): [True: 0, False: 0] ------------------ 717| 0| interlacing?yv12_to_yuyvi_c:yv12_to_yuyv_c, 2, interlacing); ------------------ | Branch (717:4): [True: 0, False: 0] ------------------ 718| 0| return 0; 719| | 720| 0| case XVID_CSP_UYVY: ------------------ | | 114| 0|#define XVID_CSP_UYVY (1<< 4) /* 4:2:2 packed */ ------------------ | Branch (720:2): [True: 0, False: 6.05k] ------------------ 721| 0| safe_packed_conv( 722| 0| dst[0], dst_stride[0], image->y, image->u, image->v, 723| 0| edged_width, edged_width2, width, height, (csp & XVID_CSP_VFLIP), ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 724| 0| interlacing?yv12_to_uyvyi :yv12_to_uyvy, ------------------ | Branch (724:4): [True: 0, False: 0] ------------------ 725| 0| interlacing?yv12_to_uyvyi_c:yv12_to_uyvy_c, 2, interlacing); ------------------ | Branch (725:4): [True: 0, False: 0] ------------------ 726| 0| return 0; 727| | 728| 0| case XVID_CSP_I420: /* YCbCr == YUV == internal colorspace for MPEG */ ------------------ | | 111| 0|#define XVID_CSP_I420 (1<< 1) /* 4:2:0 planar */ ------------------ | Branch (728:2): [True: 0, False: 6.05k] ------------------ 729| 0| yv12_to_yv12(dst[0], dst[0] + dst_stride[0]*height, dst[0] + dst_stride[0]*height + (dst_stride[0]/2)*height2, 730| 0| dst_stride[0], dst_stride[0]/2, 731| 0| image->y, image->u, image->v, edged_width, edged_width2, 732| 0| width, height, (csp & XVID_CSP_VFLIP)); ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 733| 0| return 0; 734| | 735| 0| case XVID_CSP_YV12: /* YCrCb == YVU == U and V plane swapped */ ------------------ | | 112| 0|#define XVID_CSP_YV12 (1<< 2) /* 4:2:0 planar */ ------------------ | Branch (735:2): [True: 0, False: 6.05k] ------------------ 736| 0| yv12_to_yv12(dst[0], dst[0] + dst_stride[0]*height, dst[0] + dst_stride[0]*height + (dst_stride[0]/2)*height2, 737| 0| dst_stride[0], dst_stride[0]/2, 738| 0| image->y, image->v, image->u, edged_width, edged_width2, 739| 0| width, height, (csp & XVID_CSP_VFLIP)); ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 740| 0| return 0; 741| | 742| 0| case XVID_CSP_PLANAR: /* YCbCr with arbitrary pointers and different strides for Y and UV */ ------------------ | | 109| 0|#define XVID_CSP_PLANAR (1<< 0) /* 4:2:0 planar (==I420, except for pointers/strides) */ ------------------ | Branch (742:2): [True: 0, False: 6.05k] ------------------ 743| 0| yv12_to_yv12(dst[0], dst[1], dst[2], 744| 0| dst_stride[0], dst_stride[1], /* v: dst_stride[2] not yet supported */ 745| 0| image->y, image->u, image->v, edged_width, edged_width2, 746| 0| width, height, (csp & XVID_CSP_VFLIP)); ------------------ | | 127| 0|#define XVID_CSP_VFLIP (1<<31) /* vertical flip mask */ ------------------ 747| 0| return 0; 748| | 749| 0| case XVID_CSP_INTERNAL : ------------------ | | 125| 0|#define XVID_CSP_INTERNAL (1<<13) /* decoder only: 4:2:0 planar, returns ptrs to internal buffers */ ------------------ | Branch (749:2): [True: 0, False: 6.05k] ------------------ 750| 0| dst[0] = image->y; 751| 0| dst[1] = image->u; 752| 0| dst[2] = image->v; 753| 0| dst_stride[0] = edged_width; 754| 0| dst_stride[1] = edged_width/2; 755| 0| dst_stride[2] = edged_width/2; 756| 0| return 0; 757| | 758| 0| case XVID_CSP_NULL: ------------------ | | 126| 0|#define XVID_CSP_NULL (1<<14) /* decoder only: dont output anything */ ------------------ | Branch (758:2): [True: 0, False: 6.05k] ------------------ 759| 0| case XVID_CSP_SLICE: ------------------ | | 124| 0|#define XVID_CSP_SLICE (1<<12) /* decoder only: 4:2:0 planar, per slice rendering */ ------------------ | Branch (759:2): [True: 0, False: 6.05k] ------------------ 760| 0| return 0; 761| | 762| 6.05k| } 763| | 764| 0| return -1; 765| 6.05k|} image.c:safe_packed_conv: 394| 6.05k|{ 395| 6.05k| int width_opt, width_c, height_opt; 396| | 397| 6.05k| if (width<0 || width==1 || height==1) return; /* forget about it */ ------------------ | Branch (397:9): [True: 0, False: 6.05k] | Branch (397:20): [True: 231, False: 5.82k] | Branch (397:32): [True: 150, False: 5.67k] ------------------ 398| | 399| 5.67k| if (func_opt != func_c && x_stride < size*((width+15)/16)*16) ------------------ | Branch (399:6): [True: 0, False: 5.67k] | Branch (399:28): [True: 0, False: 0] ------------------ 400| 0| { 401| 0| width_opt = width & (~15); 402| 0| width_c = (width - width_opt) & (~1); 403| 0| } 404| 5.67k| else if (func_opt != func_c && !(width&1) && (size==3)) ------------------ | Branch (404:11): [True: 0, False: 5.67k] | Branch (404:33): [True: 0, False: 0] | Branch (404:47): [True: 0, False: 0] ------------------ 405| 0| { 406| | /* MMX reads 4 bytes per pixel for RGB/BGR */ 407| 0| width_opt = width - 2; 408| 0| width_c = 2; 409| 0| } 410| 5.67k| else { 411| | /* Enforce the width to be divisable by two. */ 412| 5.67k| width_opt = width & (~1); 413| 5.67k| width_c = 0; 414| 5.67k| } 415| | 416| | /* packed conversions require height to be divisable by 2 417| | (or even by 4 for interlaced conversion) */ 418| 5.67k| if (interlacing) ------------------ | Branch (418:12): [True: 1.81k, False: 3.85k] ------------------ 419| 1.81k| height_opt = height & (~3); 420| 3.85k| else 421| 3.85k| height_opt = height & (~1); 422| | 423| 5.67k| func_opt(x_ptr, x_stride, 424| 5.67k| y_ptr, u_ptr, v_ptr, y_stride, uv_stride, 425| 5.67k| width_opt, height_opt, vflip); 426| | 427| 5.67k| if (width_c) ------------------ | Branch (427:6): [True: 0, False: 5.67k] ------------------ 428| 0| { 429| 0| func_c(x_ptr + size*width_opt, x_stride, 430| 0| y_ptr + width_opt, u_ptr + width_opt/2, v_ptr + width_opt/2, 431| 0| y_stride, uv_stride, width_c, height_opt, vflip); 432| 0| } 433| 5.67k|} decoder.c:image_null: 43| 201k|{ 44| 201k| image->y = image->u = image->v = NULL; 45| 201k|} interpolate8x8_avg2_c: 61| 2.99M|{ 62| 2.99M| uint32_t i; 63| 2.99M| const int32_t round = 1 - rounding; 64| | 65| 26.9M| for(i = 0; i < height; i++) { ------------------ | Branch (65:16): [True: 23.9M, False: 2.99M] ------------------ 66| 23.9M| dst[0] = (src1[0] + src2[0] + round) >> 1; 67| 23.9M| dst[1] = (src1[1] + src2[1] + round) >> 1; 68| 23.9M| dst[2] = (src1[2] + src2[2] + round) >> 1; 69| 23.9M| dst[3] = (src1[3] + src2[3] + round) >> 1; 70| 23.9M| dst[4] = (src1[4] + src2[4] + round) >> 1; 71| 23.9M| dst[5] = (src1[5] + src2[5] + round) >> 1; 72| 23.9M| dst[6] = (src1[6] + src2[6] + round) >> 1; 73| 23.9M| dst[7] = (src1[7] + src2[7] + round) >> 1; 74| | 75| 23.9M| dst += stride; 76| 23.9M| src1 += stride; 77| 23.9M| src2 += stride; 78| 23.9M| } 79| 2.99M|} interpolate8x8_halfpel_add_c: 83| 2.99M|{ 84| 2.99M| interpolate8x8_avg2_c(dst, dst, src, stride, 0, 8); 85| 2.99M|} interpolate8x8_halfpel_h_c: 117| 284k|{ 118| 284k| uintptr_t j; ------------------ | | 138| 284k|# define uintptr_t uint64_t ------------------ 119| | 120| 284k| if (rounding) { ------------------ | Branch (120:6): [True: 29.8k, False: 254k] ------------------ 121| 268k| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (121:15): [True: 238k, False: 29.8k] ------------------ 122| 238k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + 1] )>>1); 123| 238k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + 2] )>>1); 124| 238k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + 3] )>>1); 125| 238k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + 4] )>>1); 126| 238k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + 5] )>>1); 127| 238k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + 6] )>>1); 128| 238k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + 7] )>>1); 129| 238k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + 8] )>>1); 130| 238k| } 131| 254k| } else { 132| 2.29M| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (132:15): [True: 2.03M, False: 254k] ------------------ 133| 2.03M| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + 1] + 1)>>1); 134| 2.03M| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + 2] + 1)>>1); 135| 2.03M| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + 3] + 1)>>1); 136| 2.03M| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + 4] + 1)>>1); 137| 2.03M| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + 5] + 1)>>1); 138| 2.03M| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + 6] + 1)>>1); 139| 2.03M| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + 7] + 1)>>1); 140| 2.03M| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + 8] + 1)>>1); 141| 2.03M| } 142| 254k| } 143| 284k|} interpolate8x4_halfpel_h_c: 152| 16.1k|{ 153| 16.1k| uintptr_t j; ------------------ | | 138| 16.1k|# define uintptr_t uint64_t ------------------ 154| | 155| 16.1k| if (rounding) { ------------------ | Branch (155:6): [True: 9.36k, False: 6.83k] ------------------ 156| 46.8k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (156:15): [True: 37.4k, False: 9.36k] ------------------ 157| 37.4k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + 1] )>>1); 158| 37.4k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + 2] )>>1); 159| 37.4k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + 3] )>>1); 160| 37.4k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + 4] )>>1); 161| 37.4k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + 5] )>>1); 162| 37.4k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + 6] )>>1); 163| 37.4k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + 7] )>>1); 164| 37.4k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + 8] )>>1); 165| 37.4k| } 166| 9.36k| } else { 167| 34.1k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (167:15): [True: 27.3k, False: 6.83k] ------------------ 168| 27.3k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + 1] + 1)>>1); 169| 27.3k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + 2] + 1)>>1); 170| 27.3k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + 3] + 1)>>1); 171| 27.3k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + 4] + 1)>>1); 172| 27.3k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + 5] + 1)>>1); 173| 27.3k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + 6] + 1)>>1); 174| 27.3k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + 7] + 1)>>1); 175| 27.3k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + 8] + 1)>>1); 176| 27.3k| } 177| 6.83k| } 178| 16.1k|} interpolate8x8_halfpel_h_add_c: 187| 117k|{ 188| 117k| uintptr_t j; ------------------ | | 138| 117k|# define uintptr_t uint64_t ------------------ 189| | 190| 117k| if (rounding) { ------------------ | Branch (190:6): [True: 0, False: 117k] ------------------ 191| 0| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (191:15): [True: 0, False: 0] ------------------ 192| 0| dst[j + 0] = (uint8_t)((((src[j + 0] + src[j + 1] )>>1) + dst[j+0] + 1)>>1); 193| 0| dst[j + 1] = (uint8_t)((((src[j + 1] + src[j + 2] )>>1) + dst[j+1] + 1)>>1); 194| 0| dst[j + 2] = (uint8_t)((((src[j + 2] + src[j + 3] )>>1) + dst[j+2] + 1)>>1); 195| 0| dst[j + 3] = (uint8_t)((((src[j + 3] + src[j + 4] )>>1) + dst[j+3] + 1)>>1); 196| 0| dst[j + 4] = (uint8_t)((((src[j + 4] + src[j + 5] )>>1) + dst[j+4] + 1)>>1); 197| 0| dst[j + 5] = (uint8_t)((((src[j + 5] + src[j + 6] )>>1) + dst[j+5] + 1)>>1); 198| 0| dst[j + 6] = (uint8_t)((((src[j + 6] + src[j + 7] )>>1) + dst[j+6] + 1)>>1); 199| 0| dst[j + 7] = (uint8_t)((((src[j + 7] + src[j + 8] )>>1) + dst[j+7] + 1)>>1); 200| 0| } 201| 117k| } else { 202| 1.05M| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (202:15): [True: 938k, False: 117k] ------------------ 203| 938k| dst[j + 0] = (uint8_t)((((src[j + 0] + src[j + 1] + 1)>>1) + dst[j+0] + 1)>>1); 204| 938k| dst[j + 1] = (uint8_t)((((src[j + 1] + src[j + 2] + 1)>>1) + dst[j+1] + 1)>>1); 205| 938k| dst[j + 2] = (uint8_t)((((src[j + 2] + src[j + 3] + 1)>>1) + dst[j+2] + 1)>>1); 206| 938k| dst[j + 3] = (uint8_t)((((src[j + 3] + src[j + 4] + 1)>>1) + dst[j+3] + 1)>>1); 207| 938k| dst[j + 4] = (uint8_t)((((src[j + 4] + src[j + 5] + 1)>>1) + dst[j+4] + 1)>>1); 208| 938k| dst[j + 5] = (uint8_t)((((src[j + 5] + src[j + 6] + 1)>>1) + dst[j+5] + 1)>>1); 209| 938k| dst[j + 6] = (uint8_t)((((src[j + 6] + src[j + 7] + 1)>>1) + dst[j+6] + 1)>>1); 210| 938k| dst[j + 7] = (uint8_t)((((src[j + 7] + src[j + 8] + 1)>>1) + dst[j+7] + 1)>>1); 211| 938k| } 212| 117k| } 213| 117k|} interpolate8x8_halfpel_v_c: 222| 241k|{ 223| 241k| uintptr_t j; ------------------ | | 138| 241k|# define uintptr_t uint64_t ------------------ 224| | 225| | 226| 241k| if (rounding) { ------------------ | Branch (226:6): [True: 23.3k, False: 217k] ------------------ 227| 210k| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (227:15): [True: 186k, False: 23.3k] ------------------ 228| 186k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + stride + 0] )>>1); 229| 186k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + stride + 1] )>>1); 230| 186k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + stride + 2] )>>1); 231| 186k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + stride + 3] )>>1); 232| 186k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + stride + 4] )>>1); 233| 186k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + stride + 5] )>>1); 234| 186k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + stride + 6] )>>1); 235| 186k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + stride + 7] )>>1); 236| 186k| } 237| 217k| } else { 238| 1.95M| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (238:15): [True: 1.74M, False: 217k] ------------------ 239| 1.74M| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + stride + 0] + 1)>>1); 240| 1.74M| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + stride + 1] + 1)>>1); 241| 1.74M| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + stride + 2] + 1)>>1); 242| 1.74M| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + stride + 3] + 1)>>1); 243| 1.74M| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + stride + 4] + 1)>>1); 244| 1.74M| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + stride + 5] + 1)>>1); 245| 1.74M| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + stride + 6] + 1)>>1); 246| 1.74M| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + stride + 7] + 1)>>1); 247| 1.74M| } 248| 217k| } 249| 241k|} interpolate8x4_halfpel_v_c: 258| 19.0k|{ 259| 19.0k| uintptr_t j; ------------------ | | 138| 19.0k|# define uintptr_t uint64_t ------------------ 260| | 261| | 262| 19.0k| if (rounding) { ------------------ | Branch (262:6): [True: 9.55k, False: 9.48k] ------------------ 263| 47.7k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (263:15): [True: 38.2k, False: 9.55k] ------------------ 264| 38.2k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + stride + 0] )>>1); 265| 38.2k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + stride + 1] )>>1); 266| 38.2k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + stride + 2] )>>1); 267| 38.2k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + stride + 3] )>>1); 268| 38.2k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + stride + 4] )>>1); 269| 38.2k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + stride + 5] )>>1); 270| 38.2k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + stride + 6] )>>1); 271| 38.2k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + stride + 7] )>>1); 272| 38.2k| } 273| 9.55k| } else { 274| 47.4k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (274:15): [True: 37.9k, False: 9.48k] ------------------ 275| 37.9k| dst[j + 0] = (uint8_t)((src[j + 0] + src[j + stride + 0] + 1)>>1); 276| 37.9k| dst[j + 1] = (uint8_t)((src[j + 1] + src[j + stride + 1] + 1)>>1); 277| 37.9k| dst[j + 2] = (uint8_t)((src[j + 2] + src[j + stride + 2] + 1)>>1); 278| 37.9k| dst[j + 3] = (uint8_t)((src[j + 3] + src[j + stride + 3] + 1)>>1); 279| 37.9k| dst[j + 4] = (uint8_t)((src[j + 4] + src[j + stride + 4] + 1)>>1); 280| 37.9k| dst[j + 5] = (uint8_t)((src[j + 5] + src[j + stride + 5] + 1)>>1); 281| 37.9k| dst[j + 6] = (uint8_t)((src[j + 6] + src[j + stride + 6] + 1)>>1); 282| 37.9k| dst[j + 7] = (uint8_t)((src[j + 7] + src[j + stride + 7] + 1)>>1); 283| 37.9k| } 284| 9.48k| } 285| 19.0k|} interpolate8x8_halfpel_v_add_c: 294| 66.8k|{ 295| 66.8k| uintptr_t j; ------------------ | | 138| 66.8k|# define uintptr_t uint64_t ------------------ 296| | 297| | 298| 66.8k| if (rounding) { ------------------ | Branch (298:6): [True: 0, False: 66.8k] ------------------ 299| 0| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (299:15): [True: 0, False: 0] ------------------ 300| 0| dst[j + 0] = (uint8_t)((((src[j + 0] + src[j + stride + 0] )>>1) + dst[j+0] + 1)>>1); 301| 0| dst[j + 1] = (uint8_t)((((src[j + 1] + src[j + stride + 1] )>>1) + dst[j+1] + 1)>>1); 302| 0| dst[j + 2] = (uint8_t)((((src[j + 2] + src[j + stride + 2] )>>1) + dst[j+2] + 1)>>1); 303| 0| dst[j + 3] = (uint8_t)((((src[j + 3] + src[j + stride + 3] )>>1) + dst[j+3] + 1)>>1); 304| 0| dst[j + 4] = (uint8_t)((((src[j + 4] + src[j + stride + 4] )>>1) + dst[j+4] + 1)>>1); 305| 0| dst[j + 5] = (uint8_t)((((src[j + 5] + src[j + stride + 5] )>>1) + dst[j+5] + 1)>>1); 306| 0| dst[j + 6] = (uint8_t)((((src[j + 6] + src[j + stride + 6] )>>1) + dst[j+6] + 1)>>1); 307| 0| dst[j + 7] = (uint8_t)((((src[j + 7] + src[j + stride + 7] )>>1) + dst[j+7] + 1)>>1); 308| 0| } 309| 66.8k| } else { 310| 601k| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (310:15): [True: 534k, False: 66.8k] ------------------ 311| 534k| dst[j + 0] = (uint8_t)((((src[j + 0] + src[j + stride + 0] + 1)>>1) + dst[j+0] + 1)>>1); 312| 534k| dst[j + 1] = (uint8_t)((((src[j + 1] + src[j + stride + 1] + 1)>>1) + dst[j+1] + 1)>>1); 313| 534k| dst[j + 2] = (uint8_t)((((src[j + 2] + src[j + stride + 2] + 1)>>1) + dst[j+2] + 1)>>1); 314| 534k| dst[j + 3] = (uint8_t)((((src[j + 3] + src[j + stride + 3] + 1)>>1) + dst[j+3] + 1)>>1); 315| 534k| dst[j + 4] = (uint8_t)((((src[j + 4] + src[j + stride + 4] + 1)>>1) + dst[j+4] + 1)>>1); 316| 534k| dst[j + 5] = (uint8_t)((((src[j + 5] + src[j + stride + 5] + 1)>>1) + dst[j+5] + 1)>>1); 317| 534k| dst[j + 6] = (uint8_t)((((src[j + 6] + src[j + stride + 6] + 1)>>1) + dst[j+6] + 1)>>1); 318| 534k| dst[j + 7] = (uint8_t)((((src[j + 7] + src[j + stride + 7] + 1)>>1) + dst[j+7] + 1)>>1); 319| 534k| } 320| 66.8k| } 321| 66.8k|} interpolate8x8_halfpel_hv_c: 330| 135k|{ 331| 135k| uintptr_t j; ------------------ | | 138| 135k|# define uintptr_t uint64_t ------------------ 332| | 333| 135k| if (rounding) { ------------------ | Branch (333:6): [True: 19.1k, False: 116k] ------------------ 334| 172k| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (334:15): [True: 153k, False: 19.1k] ------------------ 335| 153k| dst[j + 0] = (uint8_t)((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +1)>>2); 336| 153k| dst[j + 1] = (uint8_t)((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +1)>>2); 337| 153k| dst[j + 2] = (uint8_t)((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +1)>>2); 338| 153k| dst[j + 3] = (uint8_t)((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +1)>>2); 339| 153k| dst[j + 4] = (uint8_t)((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +1)>>2); 340| 153k| dst[j + 5] = (uint8_t)((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +1)>>2); 341| 153k| dst[j + 6] = (uint8_t)((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +1)>>2); 342| 153k| dst[j + 7] = (uint8_t)((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +1)>>2); 343| 153k| } 344| 116k| } else { 345| 1.05M| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (345:15): [True: 933k, False: 116k] ------------------ 346| 933k| dst[j + 0] = (uint8_t)((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +2)>>2); 347| 933k| dst[j + 1] = (uint8_t)((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +2)>>2); 348| 933k| dst[j + 2] = (uint8_t)((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +2)>>2); 349| 933k| dst[j + 3] = (uint8_t)((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +2)>>2); 350| 933k| dst[j + 4] = (uint8_t)((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +2)>>2); 351| 933k| dst[j + 5] = (uint8_t)((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +2)>>2); 352| 933k| dst[j + 6] = (uint8_t)((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +2)>>2); 353| 933k| dst[j + 7] = (uint8_t)((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +2)>>2); 354| 933k| } 355| 116k| } 356| 135k|} interpolate8x4_halfpel_hv_c: 365| 12.4k|{ 366| 12.4k| uintptr_t j; ------------------ | | 138| 12.4k|# define uintptr_t uint64_t ------------------ 367| | 368| 12.4k| if (rounding) { ------------------ | Branch (368:6): [True: 7.80k, False: 4.64k] ------------------ 369| 39.0k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (369:15): [True: 31.2k, False: 7.80k] ------------------ 370| 31.2k| dst[j + 0] = (uint8_t)((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +1)>>2); 371| 31.2k| dst[j + 1] = (uint8_t)((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +1)>>2); 372| 31.2k| dst[j + 2] = (uint8_t)((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +1)>>2); 373| 31.2k| dst[j + 3] = (uint8_t)((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +1)>>2); 374| 31.2k| dst[j + 4] = (uint8_t)((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +1)>>2); 375| 31.2k| dst[j + 5] = (uint8_t)((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +1)>>2); 376| 31.2k| dst[j + 6] = (uint8_t)((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +1)>>2); 377| 31.2k| dst[j + 7] = (uint8_t)((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +1)>>2); 378| 31.2k| } 379| 7.80k| } else { 380| 23.2k| for (j = 0; j < 4*stride; j+=stride) { ------------------ | Branch (380:15): [True: 18.5k, False: 4.64k] ------------------ 381| 18.5k| dst[j + 0] = (uint8_t)((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +2)>>2); 382| 18.5k| dst[j + 1] = (uint8_t)((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +2)>>2); 383| 18.5k| dst[j + 2] = (uint8_t)((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +2)>>2); 384| 18.5k| dst[j + 3] = (uint8_t)((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +2)>>2); 385| 18.5k| dst[j + 4] = (uint8_t)((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +2)>>2); 386| 18.5k| dst[j + 5] = (uint8_t)((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +2)>>2); 387| 18.5k| dst[j + 6] = (uint8_t)((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +2)>>2); 388| 18.5k| dst[j + 7] = (uint8_t)((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +2)>>2); 389| 18.5k| } 390| 4.64k| } 391| 12.4k|} interpolate8x8_halfpel_hv_add_c: 400| 44.0k|{ 401| 44.0k| uintptr_t j; ------------------ | | 138| 44.0k|# define uintptr_t uint64_t ------------------ 402| | 403| 44.0k| if (rounding) { ------------------ | Branch (403:6): [True: 0, False: 44.0k] ------------------ 404| 0| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (404:15): [True: 0, False: 0] ------------------ 405| 0| dst[j + 0] = (uint8_t)((((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +1)>>2) + dst[j+0])>>1); 406| 0| dst[j + 1] = (uint8_t)((((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +1)>>2) + dst[j+1])>>1); 407| 0| dst[j + 2] = (uint8_t)((((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +1)>>2) + dst[j+2])>>1); 408| 0| dst[j + 3] = (uint8_t)((((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +1)>>2) + dst[j+3])>>1); 409| 0| dst[j + 4] = (uint8_t)((((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +1)>>2) + dst[j+4])>>1); 410| 0| dst[j + 5] = (uint8_t)((((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +1)>>2) + dst[j+5])>>1); 411| 0| dst[j + 6] = (uint8_t)((((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +1)>>2) + dst[j+6])>>1); 412| 0| dst[j + 7] = (uint8_t)((((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +1)>>2) + dst[j+7])>>1); 413| 0| } 414| 44.0k| } else { 415| 396k| for (j = 0; j < 8*stride; j+=stride) { ------------------ | Branch (415:15): [True: 352k, False: 44.0k] ------------------ 416| 352k| dst[j + 0] = (uint8_t)((((src[j+0] + src[j+1] + src[j+stride+0] + src[j+stride+1] +2)>>2) + dst[j+0] + 1)>>1); 417| 352k| dst[j + 1] = (uint8_t)((((src[j+1] + src[j+2] + src[j+stride+1] + src[j+stride+2] +2)>>2) + dst[j+1] + 1)>>1); 418| 352k| dst[j + 2] = (uint8_t)((((src[j+2] + src[j+3] + src[j+stride+2] + src[j+stride+3] +2)>>2) + dst[j+2] + 1)>>1); 419| 352k| dst[j + 3] = (uint8_t)((((src[j+3] + src[j+4] + src[j+stride+3] + src[j+stride+4] +2)>>2) + dst[j+3] + 1)>>1); 420| 352k| dst[j + 4] = (uint8_t)((((src[j+4] + src[j+5] + src[j+stride+4] + src[j+stride+5] +2)>>2) + dst[j+4] + 1)>>1); 421| 352k| dst[j + 5] = (uint8_t)((((src[j+5] + src[j+6] + src[j+stride+5] + src[j+stride+6] +2)>>2) + dst[j+5] + 1)>>1); 422| 352k| dst[j + 6] = (uint8_t)((((src[j+6] + src[j+7] + src[j+stride+6] + src[j+stride+7] +2)>>2) + dst[j+6] + 1)>>1); 423| 352k| dst[j + 7] = (uint8_t)((((src[j+7] + src[j+8] + src[j+stride+7] + src[j+stride+8] +2)>>2) + dst[j+7] + 1)>>1); 424| 352k| } 425| 44.0k| } 426| 44.0k|} decoder.c:interpolate16x16_switch: 325| 302k|{ 326| 302k| interpolate8x8_switch(cur, refn, x, y, dx, dy, stride, rounding); 327| 302k| interpolate8x8_switch(cur, refn, x+8, y, dx, dy, stride, rounding); 328| 302k| interpolate8x8_switch(cur, refn, x, y+8, dx, dy, stride, rounding); 329| 302k| interpolate8x8_switch(cur, refn, x+8, y+8, dx, dy, stride, rounding); 330| 302k|} decoder.c:interpolate8x8_switch: 265| 4.05M|{ 266| | 267| 4.05M| const uint8_t * const src = refn + (int)((y + (dy>>1)) * stride + x + (dx>>1)); 268| 4.05M| uint8_t * const dst = cur + (int)(y * stride + x); 269| | 270| 4.05M| switch (((dx & 1) << 1) + (dy & 1)) { /* ((dx%2)?2:0)+((dy%2)?1:0) */ 271| 3.38M| case 0: ------------------ | Branch (271:2): [True: 3.38M, False: 661k] ------------------ 272| 3.38M| transfer8x8_copy(dst, src, stride); 273| 3.38M| break; 274| 241k| case 1: ------------------ | Branch (274:2): [True: 241k, False: 3.81M] ------------------ 275| 241k| interpolate8x8_halfpel_v(dst, src, stride, rounding); 276| 241k| break; 277| 284k| case 2: ------------------ | Branch (277:2): [True: 284k, False: 3.76M] ------------------ 278| 284k| interpolate8x8_halfpel_h(dst, src, stride, rounding); 279| 284k| break; 280| 135k| default: ------------------ | Branch (280:2): [True: 135k, False: 3.91M] ------------------ 281| 135k| interpolate8x8_halfpel_hv(dst, src, stride, rounding); 282| 135k| break; 283| 4.05M| } 284| 4.05M|} decoder.c:interpolate8x4_switch: 234| 85.1k|{ 235| | 236| 85.1k| const uint8_t * const src = refn + (int)((y + (dy>>1)) * stride + x + (dx>>1)); 237| 85.1k| uint8_t * const dst = cur + (int)(y * stride + x); 238| | 239| 85.1k| switch (((dx & 1) << 1) + (dy & 1)) 240| 85.1k| { /* ((dx%2)?2:0)+((dy%2)?1:0) */ 241| 37.4k| case 0: ------------------ | Branch (241:2): [True: 37.4k, False: 47.6k] ------------------ 242| 37.4k| transfer8x4_copy(dst, src, stride); 243| 37.4k| break; 244| 19.0k| case 1: ------------------ | Branch (244:2): [True: 19.0k, False: 66.0k] ------------------ 245| 19.0k| interpolate8x4_halfpel_v(dst, src, stride, rounding); 246| 19.0k| break; 247| 16.1k| case 2: ------------------ | Branch (247:2): [True: 16.1k, False: 68.9k] ------------------ 248| 16.1k| interpolate8x4_halfpel_h(dst, src, stride, rounding); 249| 16.1k| break; 250| 12.4k| default: ------------------ | Branch (250:2): [True: 12.4k, False: 72.6k] ------------------ 251| 12.4k| interpolate8x4_halfpel_hv(dst, src, stride, rounding); 252| 12.4k| break; 253| 85.1k| } 254| 85.1k|} decoder.c:interpolate8x8_add_switch: 295| 1.79M|{ 296| | 297| 1.79M| const uint8_t * const src = refn + (int)((y + (dy>>1)) * stride + x + (dx>>1)); 298| 1.79M| uint8_t * const dst = cur + (int)(y * stride + x); 299| | 300| 1.79M| switch (((dx & 1) << 1) + (dy & 1)) { /* ((dx%2)?2:0)+((dy%2)?1:0) */ 301| 1.56M| case 0: ------------------ | Branch (301:2): [True: 1.56M, False: 228k] ------------------ 302| 1.56M| interpolate8x8_halfpel_add(dst, src, stride, rounding); 303| 1.56M| break; 304| 66.8k| case 1: ------------------ | Branch (304:2): [True: 66.8k, False: 1.72M] ------------------ 305| 66.8k| interpolate8x8_halfpel_v_add(dst, src, stride, rounding); 306| 66.8k| break; 307| 117k| case 2: ------------------ | Branch (307:2): [True: 117k, False: 1.67M] ------------------ 308| 117k| interpolate8x8_halfpel_h_add(dst, src, stride, rounding); 309| 117k| break; 310| 44.0k| default: ------------------ | Branch (310:2): [True: 44.0k, False: 1.74M] ------------------ 311| 44.0k| interpolate8x8_halfpel_hv_add(dst, src, stride, rounding); 312| 44.0k| break; 313| 1.79M| } 314| 1.79M|} init_postproc: 51| 10.2k|{ 52| 10.2k| init_deblock(tbls); 53| 10.2k| init_noise(tbls); 54| 10.2k|} init_deblock: 214| 10.2k|{ 215| 10.2k| int i; 216| | 217| 5.27M| for(i = -255; i < 256; i++) { ------------------ | Branch (217:16): [True: 5.26M, False: 10.2k] ------------------ 218| 5.26M| tbls->xvid_thresh_tbl[i + 255] = 0; 219| 5.26M| if(ABS(i) < THR1) ------------------ | | 48| 5.26M|#define ABS(X) (((X)>0)?(X):-(X)) | | ------------------ | | | Branch (48:20): [True: 2.62M, False: 2.63M] | | ------------------ ------------------ if(ABS(i) < THR1) ------------------ | | 34| 5.26M|#define THR1 2 ------------------ | Branch (219:6): [True: 30.8k, False: 5.23M] ------------------ 220| 30.8k| tbls->xvid_thresh_tbl[i + 255] = 1; 221| 5.26M| tbls->xvid_abs_tbl[i + 255] = ABS(i); ------------------ | | 48| 5.26M|#define ABS(X) (((X)>0)?(X):-(X)) | | ------------------ | | | Branch (48:20): [True: 2.62M, False: 2.63M] | | ------------------ ------------------ 222| 5.26M| } 223| 10.2k|} init_noise: 427| 10.2k|{ 428| 10.2k| int i, j; 429| 10.2k| int patt[4] = { -1,0,1,0 }; 430| | 431| 10.2k| emms(); 432| | 433| 10.2k| srand(123457); 434| | 435| 42.1M| for(i = 0, j = 0; i < MAX_NOISE; i++, j++) ------------------ | | 37| 42.1M|#define MAX_NOISE 4096 ------------------ | Branch (435:20): [True: 42.1M, False: 10.2k] ------------------ 436| 42.1M| { 437| 42.1M| double x1, x2, w, y1, y2; 438| | 439| 54.0M| do { 440| 54.0M| x1 = 2.0 * rand() / (float) RAND_MAX - 1.0; 441| 54.0M| x2 = 2.0 * rand() / (float) RAND_MAX - 1.0; 442| 54.0M| w = x1 * x1 + x2 * x2; 443| 54.0M| } while (w >= 1.0); ------------------ | Branch (443:12): [True: 11.8M, False: 42.1M] ------------------ 444| | 445| 42.1M| w = sqrt((-2.0 * log(w)) / w); 446| 42.1M| y1 = x1 * w; 447| 42.1M| y2 = x1 * w; 448| | 449| 42.1M| y1 *= STRENGTH1 / sqrt(3.0); ------------------ | | 423| 42.1M|#define STRENGTH1 12 ------------------ 450| 42.1M| y2 *= STRENGTH2 / sqrt(3.0); ------------------ | | 424| 42.1M|#define STRENGTH2 8 ------------------ 451| | 452| 42.1M| y1 /= 2; 453| 42.1M| y2 /= 2; 454| 42.1M| y1 += patt[j%4] * STRENGTH1 * 0.35; ------------------ | | 423| 42.1M|#define STRENGTH1 12 ------------------ 455| 42.1M| y2 += patt[j%4] * STRENGTH2 * 0.35; ------------------ | | 424| 42.1M|#define STRENGTH2 8 ------------------ 456| | 457| 42.1M| if (y1 < -128) { ------------------ | Branch (457:7): [True: 0, False: 42.1M] ------------------ 458| 0| y1=-128; 459| 0| } 460| 42.1M| else if (y1 > 127) { ------------------ | Branch (460:12): [True: 0, False: 42.1M] ------------------ 461| 0| y1= 127; 462| 0| } 463| | 464| 42.1M| if (y2 < -128) { ------------------ | Branch (464:7): [True: 0, False: 42.1M] ------------------ 465| 0| y2=-128; 466| 0| } 467| 42.1M| else if (y2 > 127) { ------------------ | Branch (467:12): [True: 0, False: 42.1M] ------------------ 468| 0| y2= 127; 469| 0| } 470| | 471| 42.1M| y1 /= 3.0; 472| 42.1M| y2 /= 3.0; 473| 42.1M| tbls->xvid_noise1[i] = (int) y1; 474| 42.1M| tbls->xvid_noise2[i] = (int) y2; 475| | 476| 42.1M| if (RAND_N(6) == 0) { ------------------ | | 422| 42.1M|#define RAND_N(range) ((int) ((double)range * rand() / (RAND_MAX + 1.0))) ------------------ | Branch (476:7): [True: 7.09M, False: 35.0M] ------------------ 477| 7.09M| j--; 478| 7.09M| } 479| 42.1M| } 480| | 481| 31.6M| for (i = 0; i < MAX_RES; i++) ------------------ | | 39| 31.6M|#define MAX_RES (MAX_NOISE - MAX_SHIFT) | | ------------------ | | | | 37| 31.6M|#define MAX_NOISE 4096 | | ------------------ | | #define MAX_RES (MAX_NOISE - MAX_SHIFT) | | ------------------ | | | | 38| 31.6M|#define MAX_SHIFT 1024 | | ------------------ ------------------ | Branch (481:14): [True: 31.6M, False: 10.2k] ------------------ 482| 126M| for (j = 0; j < 3; j++) { ------------------ | Branch (482:15): [True: 94.9M, False: 31.6M] ------------------ 483| 94.9M| tbls->xvid_prev_shift[i][j] = tbls->xvid_noise1 + (rand() & (MAX_SHIFT - 1)); ------------------ | | 38| 94.9M|#define MAX_SHIFT 1024 ------------------ 484| 94.9M| tbls->xvid_prev_shift[i][3 + j] = tbls->xvid_noise2 + (rand() & (MAX_SHIFT - 1)); ------------------ | | 38| 94.9M|#define MAX_SHIFT 1024 ------------------ 485| 94.9M| } 486| 10.2k|} xvid_Init_QP: 412| 2|{ 413| |#if defined (ARCH_IS_IA32) || defined (ARCH_IS_X86_64) 414| | int i; 415| | 416| | for(i=0; i<256; ++i) { 417| | xvid_Expand_mmx[i][0] = i; 418| | xvid_Expand_mmx[i][1] = i; 419| | xvid_Expand_mmx[i][2] = i; 420| | xvid_Expand_mmx[i][3] = i; 421| | } 422| |#endif 423| | 424| | /* Alternate way of filtering (cf. USE_TABLES flag in qpel_mmx.asm) */ 425| | 426| 2| Init_FIR_Table(xvid_FIR_1_0_0_0, -1, 0, 0, 0); 427| 2| Init_FIR_Table(xvid_FIR_3_1_0_0, 3, -1, 0, 0); 428| 2| Init_FIR_Table(xvid_FIR_6_3_1_0, -6, 3, -1, 0); 429| 2| Init_FIR_Table(xvid_FIR_14_3_2_1, 14, -3, 2, -1); 430| 2| Init_FIR_Table(xvid_FIR_20_6_3_1, 20, -6, 3, -1); 431| 2| Init_FIR_Table(xvid_FIR_20_20_6_3, 20, 20, -6, 3); 432| 2| Init_FIR_Table(xvid_FIR_23_19_6_3, 23, 19, -6, 3); 433| 2| Init_FIR_Table(xvid_FIR_7_20_20_6, -7, 20, 20, -6); 434| 2| Init_FIR_Table(xvid_FIR_6_20_20_6, -6, 20, 20, -6); 435| 2| Init_FIR_Table(xvid_FIR_6_20_20_7, -6, 20, 20, -7); 436| 2| Init_FIR_Table(xvid_FIR_3_6_20_20, 3, -6, 20, 20); 437| 2| Init_FIR_Table(xvid_FIR_3_6_19_23, 3, -6, 19, 23); 438| 2| Init_FIR_Table(xvid_FIR_1_3_6_20, -1, 3, -6, 20); 439| 2| Init_FIR_Table(xvid_FIR_1_2_3_14, -1, 2, -3, 14); 440| 2| Init_FIR_Table(xvid_FIR_0_1_3_6, 0, -1, 3, -6); 441| 2| Init_FIR_Table(xvid_FIR_0_0_1_3, 0, 0, -1, 3); 442| 2| Init_FIR_Table(xvid_FIR_0_0_0_1, 0, 0, 0, -1); 443| | 444| 2|} qpel.c:H_Pass_16_C: 617| 27.8k|{ 618| 27.8k|#if (SIZE==16) 619| 489k| while(H-->0) { ------------------ | Branch (619:9): [True: 461k, False: 27.8k] ------------------ 620| 461k| int C; 621| 461k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 622| 461k| CLIP_STORE(Dst[ 0],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.64k, False: 459k] | | | Branch (612:28): [True: 5.62k, False: 454k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 623| 461k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 624| 461k| CLIP_STORE(Dst[ 1],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.82k, False: 458k] | | | Branch (612:28): [True: 6.42k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 625| 461k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 626| 461k| CLIP_STORE(Dst[ 2],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.15k, False: 459k] | | | Branch (612:28): [True: 5.29k, False: 454k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 627| 461k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 628| 461k| CLIP_STORE(Dst[ 3],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.06k, False: 458k] | | | Branch (612:28): [True: 5.97k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 629| 461k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 630| 461k| CLIP_STORE(Dst[ 4],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.72k, False: 458k] | | | Branch (612:28): [True: 6.07k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 631| 461k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 632| 461k| CLIP_STORE(Dst[ 5],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.69k, False: 458k] | | | Branch (612:28): [True: 5.60k, False: 453k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 633| 461k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 634| 461k| CLIP_STORE(Dst[ 6],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.49k, False: 459k] | | | Branch (612:28): [True: 5.30k, False: 453k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 635| 461k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 636| 461k| CLIP_STORE(Dst[ 7],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.18k, False: 458k] | | | Branch (612:28): [True: 5.97k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 637| 461k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 638| 461k| CLIP_STORE(Dst[ 8],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.17k, False: 458k] | | | Branch (612:28): [True: 5.17k, False: 453k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 639| 461k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 640| 461k| CLIP_STORE(Dst[ 9],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.51k, False: 458k] | | | Branch (612:28): [True: 5.86k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 641| 461k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 642| 461k| CLIP_STORE(Dst[10],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.09k, False: 458k] | | | Branch (612:28): [True: 5.60k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 643| 461k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 644| 461k| CLIP_STORE(Dst[11],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.77k, False: 458k] | | | Branch (612:28): [True: 5.88k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 645| 461k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 646| 461k| CLIP_STORE(Dst[12],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.79k, False: 458k] | | | Branch (612:28): [True: 6.04k, False: 452k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 647| 461k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 648| 461k| CLIP_STORE(Dst[13],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.67k, False: 459k] | | | Branch (612:28): [True: 5.54k, False: 454k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 649| 461k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 650| 461k| CLIP_STORE(Dst[14],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.66k, False: 458k] | | | Branch (612:28): [True: 5.77k, False: 453k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 651| 461k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 652| 461k| CLIP_STORE(Dst[15],C); ------------------ | | 612| 461k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.32k, False: 460k] | | | Branch (612:28): [True: 5.47k, False: 454k] | | ------------------ | | 613| 461k| STORE(D, C) | | ------------------ | | | | 145| 461k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 653| 461k| Src += BpS; 654| 461k| Dst += BpS; 655| 461k| } 656| |#else 657| | while(H-->0) { 658| | int C; 659| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 660| | CLIP_STORE(Dst[0],C); 661| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 662| | CLIP_STORE(Dst[1],C); 663| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 664| | CLIP_STORE(Dst[2],C); 665| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 666| | CLIP_STORE(Dst[3],C); 667| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 668| | CLIP_STORE(Dst[4],C); 669| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 670| | CLIP_STORE(Dst[5],C); 671| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 672| | CLIP_STORE(Dst[6],C); 673| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 674| | CLIP_STORE(Dst[7],C); 675| | Src += BpS; 676| | Dst += BpS; 677| | } 678| |#endif 679| 27.8k|} qpel.c:H_Pass_Avrg_16_C: 689| 32.3k|{ 690| 32.3k|#if (SIZE==16) 691| 569k| while(H-->0) { ------------------ | Branch (691:9): [True: 537k, False: 32.3k] ------------------ 692| 537k| int C; 693| 537k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 694| 537k| CLIP_STORE(0,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.51k, False: 536k] | | | Branch (683:28): [True: 5.89k, False: 530k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 695| 537k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 696| 537k| CLIP_STORE( 1,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.37k, False: 535k] | | | Branch (683:28): [True: 5.49k, False: 529k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 697| 537k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 698| 537k| CLIP_STORE( 2,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.96k, False: 534k] | | | Branch (683:28): [True: 4.82k, False: 529k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 699| 537k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 700| 537k| CLIP_STORE( 3,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.30k, False: 535k] | | | Branch (683:28): [True: 5.02k, False: 530k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 701| 537k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 702| 537k| CLIP_STORE( 4,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.26k, False: 534k] | | | Branch (683:28): [True: 5.13k, False: 529k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 703| 537k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 704| 537k| CLIP_STORE( 5,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.25k, False: 535k] | | | Branch (683:28): [True: 4.60k, False: 530k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 705| 537k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 706| 537k| CLIP_STORE( 6,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.13k, False: 534k] | | | Branch (683:28): [True: 5.34k, False: 529k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 707| 537k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 708| 537k| CLIP_STORE( 7,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.73k, False: 534k] | | | Branch (683:28): [True: 4.53k, False: 530k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 709| 537k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 710| 537k| CLIP_STORE( 8,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.09k, False: 533k] | | | Branch (683:28): [True: 5.75k, False: 527k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 711| 537k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 712| 537k| CLIP_STORE( 9,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.99k, False: 534k] | | | Branch (683:28): [True: 4.58k, False: 530k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 713| 537k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 714| 537k| CLIP_STORE(10,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.79k, False: 533k] | | | Branch (683:28): [True: 5.64k, False: 528k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 715| 537k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 716| 537k| CLIP_STORE(11,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.53k, False: 535k] | | | Branch (683:28): [True: 5.33k, False: 529k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 717| 537k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 718| 537k| CLIP_STORE(12,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.49k, False: 533k] | | | Branch (683:28): [True: 5.54k, False: 527k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 719| 537k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 720| 537k| CLIP_STORE(13,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.73k, False: 535k] | | | Branch (683:28): [True: 3.70k, False: 532k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 721| 537k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 722| 537k| CLIP_STORE(14,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.31k, False: 533k] | | | Branch (683:28): [True: 5.48k, False: 527k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 723| 537k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 724| 537k| CLIP_STORE(15,C); ------------------ | | 683| 537k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.46k, False: 536k] | | | Branch (683:28): [True: 3.67k, False: 532k] | | ------------------ | | 684| 537k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 537k| STORE(Dst[i], C) | | ------------------ | | | | 145| 537k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 725| 537k| Src += BpS; 726| 537k| Dst += BpS; 727| 537k| } 728| |#else 729| | while(H-->0) { 730| | int C; 731| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 732| | CLIP_STORE(0,C); 733| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 734| | CLIP_STORE(1,C); 735| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 736| | CLIP_STORE(2,C); 737| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 738| | CLIP_STORE(3,C); 739| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 740| | CLIP_STORE(4,C); 741| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 742| | CLIP_STORE(5,C); 743| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 744| | CLIP_STORE(6,C); 745| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 746| | CLIP_STORE(7,C); 747| | Src += BpS; 748| | Dst += BpS; 749| | } 750| |#endif 751| 32.3k|} qpel.c:H_Pass_Avrg_Up_16_C: 761| 38.2k|{ 762| 38.2k|#if (SIZE==16) 763| 671k| while(H-->0) { ------------------ | Branch (763:9): [True: 632k, False: 38.2k] ------------------ 764| 632k| int C; 765| 632k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 766| 632k| CLIP_STORE(0,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.81k, False: 631k] | | | Branch (755:28): [True: 5.12k, False: 625k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 767| 632k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 768| 632k| CLIP_STORE( 1,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.36k, False: 628k] | | | Branch (755:28): [True: 8.84k, False: 619k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 769| 632k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 770| 632k| CLIP_STORE( 2,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.11k, False: 630k] | | | Branch (755:28): [True: 5.43k, False: 625k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 771| 632k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 772| 632k| CLIP_STORE( 3,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 5.66k, False: 627k] | | | Branch (755:28): [True: 7.17k, False: 620k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 773| 632k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 774| 632k| CLIP_STORE( 4,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.31k, False: 630k] | | | Branch (755:28): [True: 7.48k, False: 623k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 775| 632k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 776| 632k| CLIP_STORE( 5,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.03k, False: 628k] | | | Branch (755:28): [True: 7.48k, False: 621k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 777| 632k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 778| 632k| CLIP_STORE( 6,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.37k, False: 630k] | | | Branch (755:28): [True: 6.22k, False: 624k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 779| 632k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 780| 632k| CLIP_STORE( 7,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 3.69k, False: 629k] | | | Branch (755:28): [True: 7.83k, False: 621k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 781| 632k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 782| 632k| CLIP_STORE( 8,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.63k, False: 630k] | | | Branch (755:28): [True: 6.40k, False: 623k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 783| 632k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 784| 632k| CLIP_STORE( 9,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 5.73k, False: 627k] | | | Branch (755:28): [True: 8.74k, False: 618k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 785| 632k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 786| 632k| CLIP_STORE(10,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.76k, False: 630k] | | | Branch (755:28): [True: 6.57k, False: 623k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 787| 632k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 788| 632k| CLIP_STORE(11,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 5.27k, False: 627k] | | | Branch (755:28): [True: 8.57k, False: 618k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 789| 632k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 790| 632k| CLIP_STORE(12,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.49k, False: 630k] | | | Branch (755:28): [True: 8.29k, False: 622k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 791| 632k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 792| 632k| CLIP_STORE(13,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.35k, False: 630k] | | | Branch (755:28): [True: 8.46k, False: 622k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 793| 632k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 794| 632k| CLIP_STORE(14,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.12k, False: 630k] | | | Branch (755:28): [True: 8.60k, False: 622k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 795| 632k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 796| 632k| CLIP_STORE(15,C); ------------------ | | 755| 632k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.70k, False: 631k] | | | Branch (755:28): [True: 8.24k, False: 622k] | | ------------------ | | 756| 632k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 632k| STORE(Dst[i], C) | | ------------------ | | | | 145| 632k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 797| 632k| Src += BpS; 798| 632k| Dst += BpS; 799| 632k| } 800| |#else 801| | while(H-->0) { 802| | int C; 803| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 804| | CLIP_STORE(0,C); 805| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 806| | CLIP_STORE(1,C); 807| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 808| | CLIP_STORE(2,C); 809| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 810| | CLIP_STORE(3,C); 811| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 812| | CLIP_STORE(4,C); 813| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 814| | CLIP_STORE(5,C); 815| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 816| | CLIP_STORE(6,C); 817| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 818| | CLIP_STORE(7,C); 819| | Src += BpS; 820| | Dst += BpS; 821| | } 822| |#endif 823| 38.2k|} qpel.c:V_Pass_16_C: 837| 24.1k|{ 838| 24.1k|#if (SIZE==16) 839| 410k| while(H-->0) { ------------------ | Branch (839:9): [True: 386k, False: 24.1k] ------------------ 840| 386k| int C; 841| 386k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 842| 386k| CLIP_STORE(Dst[BpS* 0],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 951, False: 385k] | | | Branch (832:28): [True: 2.32k, False: 382k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 843| 386k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 844| 386k| CLIP_STORE(Dst[BpS* 1],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.62k, False: 383k] | | | Branch (832:28): [True: 2.32k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 845| 386k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 846| 386k| CLIP_STORE(Dst[BpS* 2],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.42k, False: 384k] | | | Branch (832:28): [True: 2.11k, False: 382k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 847| 386k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 848| 386k| CLIP_STORE(Dst[BpS* 3],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.05k, False: 383k] | | | Branch (832:28): [True: 2.42k, False: 380k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 849| 386k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 850| 386k| CLIP_STORE(Dst[BpS* 4],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.69k, False: 384k] | | | Branch (832:28): [True: 2.67k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 851| 386k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 852| 386k| CLIP_STORE(Dst[BpS* 5],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.91k, False: 384k] | | | Branch (832:28): [True: 2.49k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 853| 386k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 854| 386k| CLIP_STORE(Dst[BpS* 6],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.81k, False: 384k] | | | Branch (832:28): [True: 2.70k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 855| 386k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 856| 386k| CLIP_STORE(Dst[BpS* 7],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.44k, False: 384k] | | | Branch (832:28): [True: 2.74k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 857| 386k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 858| 386k| CLIP_STORE(Dst[BpS* 8],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.59k, False: 384k] | | | Branch (832:28): [True: 3.48k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 859| 386k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 860| 386k| CLIP_STORE(Dst[BpS* 9],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.99k, False: 384k] | | | Branch (832:28): [True: 2.69k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 861| 386k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 862| 386k| CLIP_STORE(Dst[BpS*10],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.31k, False: 383k] | | | Branch (832:28): [True: 2.58k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 863| 386k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 864| 386k| CLIP_STORE(Dst[BpS*11],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.80k, False: 384k] | | | Branch (832:28): [True: 3.14k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 865| 386k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 866| 386k| CLIP_STORE(Dst[BpS*12],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.70k, False: 384k] | | | Branch (832:28): [True: 3.09k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 867| 386k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 868| 386k| CLIP_STORE(Dst[BpS*13],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.04k, False: 384k] | | | Branch (832:28): [True: 2.36k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 869| 386k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 870| 386k| CLIP_STORE(Dst[BpS*14],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.87k, False: 384k] | | | Branch (832:28): [True: 2.74k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 871| 386k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 872| 386k| CLIP_STORE(Dst[BpS*15],C); ------------------ | | 832| 386k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.77k, False: 383k] | | | Branch (832:28): [True: 2.21k, False: 381k] | | ------------------ | | 833| 386k| STORE(D, C) | | ------------------ | | | | 145| 386k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 873| 386k| Src += 1; 874| 386k| Dst += 1; 875| 386k| } 876| |#else 877| | while(H-->0) { 878| | int C; 879| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 880| | CLIP_STORE(Dst[BpS*0],C); 881| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 882| | CLIP_STORE(Dst[BpS*1],C); 883| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 884| | CLIP_STORE(Dst[BpS*2],C); 885| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 886| | CLIP_STORE(Dst[BpS*3],C); 887| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 888| | CLIP_STORE(Dst[BpS*4],C); 889| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 890| | CLIP_STORE(Dst[BpS*5],C); 891| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 892| | CLIP_STORE(Dst[BpS*6],C); 893| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 894| | CLIP_STORE(Dst[BpS*7],C); 895| | Src += 1; 896| | Dst += 1; 897| | } 898| |#endif 899| 24.1k|} qpel.c:V_Pass_Avrg_16_C: 909| 24.7k|{ 910| 24.7k|#if (SIZE==16) 911| 421k| while(H-->0) { ------------------ | Branch (911:9): [True: 396k, False: 24.7k] ------------------ 912| 396k| int C; 913| 396k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 914| 396k| CLIP_STORE(0,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.26k, False: 395k] | | | Branch (903:28): [True: 2.23k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 915| 396k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 916| 396k| CLIP_STORE( 1,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.46k, False: 395k] | | | Branch (903:28): [True: 2.40k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 917| 396k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 918| 396k| CLIP_STORE( 2,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.68k, False: 394k] | | | Branch (903:28): [True: 2.29k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 919| 396k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 920| 396k| CLIP_STORE( 3,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.06k, False: 394k] | | | Branch (903:28): [True: 2.53k, False: 391k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 921| 396k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 922| 396k| CLIP_STORE( 4,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.24k, False: 394k] | | | Branch (903:28): [True: 2.54k, False: 391k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 923| 396k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 924| 396k| CLIP_STORE( 5,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.86k, False: 394k] | | | Branch (903:28): [True: 2.57k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 925| 396k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 926| 396k| CLIP_STORE( 6,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.46k, False: 394k] | | | Branch (903:28): [True: 2.89k, False: 391k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 927| 396k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 928| 396k| CLIP_STORE( 7,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.34k, False: 395k] | | | Branch (903:28): [True: 2.50k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 929| 396k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 930| 396k| CLIP_STORE( 8,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.63k, False: 394k] | | | Branch (903:28): [True: 3.46k, False: 391k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 931| 396k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 932| 396k| CLIP_STORE( 9,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.64k, False: 394k] | | | Branch (903:28): [True: 2.41k, False: 392k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 933| 396k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 934| 396k| CLIP_STORE(10,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.71k, False: 393k] | | | Branch (903:28): [True: 3.26k, False: 390k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 935| 396k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 936| 396k| CLIP_STORE(11,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.74k, False: 394k] | | | Branch (903:28): [True: 2.87k, False: 391k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 937| 396k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 938| 396k| CLIP_STORE(12,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.52k, False: 391k] | | | Branch (903:28): [True: 2.94k, False: 388k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 939| 396k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 940| 396k| CLIP_STORE(13,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.73k, False: 394k] | | | Branch (903:28): [True: 1.49k, False: 393k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 941| 396k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 942| 396k| CLIP_STORE(14,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.34k, False: 392k] | | | Branch (903:28): [True: 3.09k, False: 389k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 943| 396k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 944| 396k| CLIP_STORE(15,C); ------------------ | | 903| 396k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.25k, False: 395k] | | | Branch (903:28): [True: 1.72k, False: 393k] | | ------------------ | | 904| 396k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 396k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 396k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 945| 396k| Src += 1; 946| 396k| Dst += 1; 947| 396k| } 948| |#else 949| | while(H-->0) { 950| | int C; 951| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 952| | CLIP_STORE(0,C); 953| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 954| | CLIP_STORE(1,C); 955| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 956| | CLIP_STORE(2,C); 957| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 958| | CLIP_STORE(3,C); 959| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 960| | CLIP_STORE(4,C); 961| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 962| | CLIP_STORE(5,C); 963| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 964| | CLIP_STORE(6,C); 965| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 966| | CLIP_STORE(7,C); 967| | Src += 1; 968| | Dst += 1; 969| | } 970| |#endif 971| 24.7k|} qpel.c:V_Pass_Avrg_Up_16_C: 981| 29.2k|{ 982| 29.2k|#if (SIZE==16) 983| 498k| while(H-->0) { ------------------ | Branch (983:9): [True: 468k, False: 29.2k] ------------------ 984| 468k| int C; 985| 468k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 986| 468k| CLIP_STORE(0,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.29k, False: 467k] | | | Branch (975:28): [True: 1.99k, False: 465k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 987| 468k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 988| 468k| CLIP_STORE( 1,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 5.07k, False: 463k] | | | Branch (975:28): [True: 2.65k, False: 461k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 989| 468k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 990| 468k| CLIP_STORE( 2,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.64k, False: 467k] | | | Branch (975:28): [True: 1.87k, False: 465k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 991| 468k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 992| 468k| CLIP_STORE( 3,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 5.78k, False: 462k] | | | Branch (975:28): [True: 2.35k, False: 460k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 993| 468k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 994| 468k| CLIP_STORE( 4,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.10k, False: 466k] | | | Branch (975:28): [True: 2.77k, False: 463k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 995| 468k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 996| 468k| CLIP_STORE( 5,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.51k, False: 466k] | | | Branch (975:28): [True: 2.90k, False: 463k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 997| 468k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 998| 468k| CLIP_STORE( 6,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.06k, False: 466k] | | | Branch (975:28): [True: 2.41k, False: 464k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 999| 468k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 1000| 468k| CLIP_STORE( 7,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.88k, False: 465k] | | | Branch (975:28): [True: 3.14k, False: 462k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1001| 468k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 1002| 468k| CLIP_STORE( 8,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.29k, False: 467k] | | | Branch (975:28): [True: 2.81k, False: 464k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1003| 468k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 1004| 468k| CLIP_STORE( 9,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.63k, False: 466k] | | | Branch (975:28): [True: 3.19k, False: 462k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1005| 468k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 1006| 468k| CLIP_STORE(10,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.11k, False: 466k] | | | Branch (975:28): [True: 2.47k, False: 464k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1007| 468k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 1008| 468k| CLIP_STORE(11,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.93k, False: 465k] | | | Branch (975:28): [True: 3.27k, False: 462k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1009| 468k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 1010| 468k| CLIP_STORE(12,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.56k, False: 467k] | | | Branch (975:28): [True: 2.99k, False: 464k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1011| 468k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 1012| 468k| CLIP_STORE(13,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.10k, False: 466k] | | | Branch (975:28): [True: 3.16k, False: 463k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1013| 468k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 1014| 468k| CLIP_STORE(14,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.36k, False: 467k] | | | Branch (975:28): [True: 3.43k, False: 463k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1015| 468k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 1016| 468k| CLIP_STORE(15,C); ------------------ | | 975| 468k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.66k, False: 467k] | | | Branch (975:28): [True: 3.21k, False: 463k] | | ------------------ | | 976| 468k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 468k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 145| 468k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1017| 468k| Src += 1; 1018| 468k| Dst += 1; 1019| 468k| } 1020| |#else 1021| | while(H-->0) { 1022| | int C; 1023| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 1024| | CLIP_STORE(0,C); 1025| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 1026| | CLIP_STORE(1,C); 1027| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 1028| | CLIP_STORE(2,C); 1029| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 1030| | CLIP_STORE(3,C); 1031| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 1032| | CLIP_STORE(4,C); 1033| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 1034| | CLIP_STORE(5,C); 1035| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 1036| | CLIP_STORE(6,C); 1037| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 1038| | CLIP_STORE(7,C); 1039| | Src += 1; 1040| | Dst += 1; 1041| | } 1042| |#endif 1043| 29.2k|} qpel.c:H_Pass_8_C: 617| 55.3k|{ 618| |#if (SIZE==16) 619| | while(H-->0) { 620| | int C; 621| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 622| | CLIP_STORE(Dst[ 0],C); 623| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 624| | CLIP_STORE(Dst[ 1],C); 625| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 626| | CLIP_STORE(Dst[ 2],C); 627| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 628| | CLIP_STORE(Dst[ 3],C); 629| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 630| | CLIP_STORE(Dst[ 4],C); 631| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 632| | CLIP_STORE(Dst[ 5],C); 633| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 634| | CLIP_STORE(Dst[ 6],C); 635| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 636| | CLIP_STORE(Dst[ 7],C); 637| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 638| | CLIP_STORE(Dst[ 8],C); 639| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 640| | CLIP_STORE(Dst[ 9],C); 641| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 642| | CLIP_STORE(Dst[10],C); 643| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 644| | CLIP_STORE(Dst[11],C); 645| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 646| | CLIP_STORE(Dst[12],C); 647| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 648| | CLIP_STORE(Dst[13],C); 649| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 650| | CLIP_STORE(Dst[14],C); 651| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 652| | CLIP_STORE(Dst[15],C); 653| | Src += BpS; 654| | Dst += BpS; 655| | } 656| |#else 657| 527k| while(H-->0) { ------------------ | Branch (657:9): [True: 471k, False: 55.3k] ------------------ 658| 471k| int C; 659| 471k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 660| 471k| CLIP_STORE(Dst[0],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.33k, False: 469k] | | | Branch (612:28): [True: 4.52k, False: 464k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 661| 471k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 662| 471k| CLIP_STORE(Dst[1],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.72k, False: 468k] | | | Branch (612:28): [True: 5.33k, False: 463k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 663| 471k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 664| 471k| CLIP_STORE(Dst[2],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.45k, False: 469k] | | | Branch (612:28): [True: 4.70k, False: 464k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 665| 471k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 666| 471k| CLIP_STORE(Dst[3],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.77k, False: 467k] | | | Branch (612:28): [True: 4.82k, False: 463k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 667| 471k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 668| 471k| CLIP_STORE(Dst[4],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.55k, False: 469k] | | | Branch (612:28): [True: 5.33k, False: 463k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 669| 471k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 670| 471k| CLIP_STORE(Dst[5],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.99k, False: 469k] | | | Branch (612:28): [True: 5.29k, False: 464k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 671| 471k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 672| 471k| CLIP_STORE(Dst[6],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.94k, False: 468k] | | | Branch (612:28): [True: 5.28k, False: 463k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 673| 471k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 674| 471k| CLIP_STORE(Dst[7],C); ------------------ | | 612| 471k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.94k, False: 469k] | | | Branch (612:28): [True: 4.71k, False: 465k] | | ------------------ | | 613| 471k| STORE(D, C) | | ------------------ | | | | 174| 471k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 675| 471k| Src += BpS; 676| 471k| Dst += BpS; 677| 471k| } 678| 55.3k|#endif 679| 55.3k|} qpel.c:H_Pass_Avrg_8_C: 689| 99.4k|{ 690| |#if (SIZE==16) 691| | while(H-->0) { 692| | int C; 693| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 694| | CLIP_STORE(0,C); 695| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 696| | CLIP_STORE( 1,C); 697| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 698| | CLIP_STORE( 2,C); 699| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 700| | CLIP_STORE( 3,C); 701| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 702| | CLIP_STORE( 4,C); 703| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 704| | CLIP_STORE( 5,C); 705| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 706| | CLIP_STORE( 6,C); 707| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 708| | CLIP_STORE( 7,C); 709| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 710| | CLIP_STORE( 8,C); 711| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 712| | CLIP_STORE( 9,C); 713| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 714| | CLIP_STORE(10,C); 715| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 716| | CLIP_STORE(11,C); 717| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 718| | CLIP_STORE(12,C); 719| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 720| | CLIP_STORE(13,C); 721| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 722| | CLIP_STORE(14,C); 723| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 724| | CLIP_STORE(15,C); 725| | Src += BpS; 726| | Dst += BpS; 727| | } 728| |#else 729| 954k| while(H-->0) { ------------------ | Branch (729:9): [True: 854k, False: 99.4k] ------------------ 730| 854k| int C; 731| 854k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 732| 854k| CLIP_STORE(0,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.33k, False: 851k] | | | Branch (683:28): [True: 12.2k, False: 839k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 733| 854k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 734| 854k| CLIP_STORE(1,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.58k, False: 851k] | | | Branch (683:28): [True: 11.5k, False: 839k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 735| 854k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 736| 854k| CLIP_STORE(2,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.62k, False: 850k] | | | Branch (683:28): [True: 10.9k, False: 839k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 737| 854k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 738| 854k| CLIP_STORE(3,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.13k, False: 850k] | | | Branch (683:28): [True: 11.5k, False: 839k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 739| 854k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 740| 854k| CLIP_STORE(4,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 7.12k, False: 847k] | | | Branch (683:28): [True: 11.4k, False: 836k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 741| 854k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 742| 854k| CLIP_STORE(5,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.18k, False: 851k] | | | Branch (683:28): [True: 9.61k, False: 842k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 743| 854k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 744| 854k| CLIP_STORE(6,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 6.64k, False: 848k] | | | Branch (683:28): [True: 11.8k, False: 836k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 745| 854k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 746| 854k| CLIP_STORE(7,C); ------------------ | | 683| 854k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.62k, False: 852k] | | | Branch (683:28): [True: 10.2k, False: 842k] | | ------------------ | | 684| 854k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 854k| STORE(Dst[i], C) | | ------------------ | | | | 174| 854k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 747| 854k| Src += BpS; 748| 854k| Dst += BpS; 749| 854k| } 750| 99.4k|#endif 751| 99.4k|} qpel.c:H_Pass_Avrg_Up_8_C: 761| 106k|{ 762| |#if (SIZE==16) 763| | while(H-->0) { 764| | int C; 765| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 766| | CLIP_STORE(0,C); 767| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 768| | CLIP_STORE( 1,C); 769| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 770| | CLIP_STORE( 2,C); 771| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 772| | CLIP_STORE( 3,C); 773| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 774| | CLIP_STORE( 4,C); 775| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 776| | CLIP_STORE( 5,C); 777| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 778| | CLIP_STORE( 6,C); 779| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 780| | CLIP_STORE( 7,C); 781| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 782| | CLIP_STORE( 8,C); 783| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 784| | CLIP_STORE( 9,C); 785| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 786| | CLIP_STORE(10,C); 787| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 788| | CLIP_STORE(11,C); 789| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 790| | CLIP_STORE(12,C); 791| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 792| | CLIP_STORE(13,C); 793| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 794| | CLIP_STORE(14,C); 795| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 796| | CLIP_STORE(15,C); 797| | Src += BpS; 798| | Dst += BpS; 799| | } 800| |#else 801| 1.03M| while(H-->0) { ------------------ | Branch (801:9): [True: 924k, False: 106k] ------------------ 802| 924k| int C; 803| 924k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 804| 924k| CLIP_STORE(0,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.51k, False: 921k] | | | Branch (755:28): [True: 8.42k, False: 913k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 805| 924k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 806| 924k| CLIP_STORE(1,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 6.51k, False: 917k] | | | Branch (755:28): [True: 11.9k, False: 905k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 807| 924k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 808| 924k| CLIP_STORE(2,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.93k, False: 921k] | | | Branch (755:28): [True: 8.74k, False: 912k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 809| 924k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 810| 924k| CLIP_STORE(3,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 6.13k, False: 918k] | | | Branch (755:28): [True: 11.4k, False: 906k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 811| 924k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 812| 924k| CLIP_STORE(4,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.15k, False: 919k] | | | Branch (755:28): [True: 11.0k, False: 908k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 813| 924k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 814| 924k| CLIP_STORE(5,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 3.80k, False: 920k] | | | Branch (755:28): [True: 10.8k, False: 909k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 815| 924k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 816| 924k| CLIP_STORE(6,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 3.73k, False: 920k] | | | Branch (755:28): [True: 10.6k, False: 909k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 817| 924k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 818| 924k| CLIP_STORE(7,C); ------------------ | | 755| 924k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 3.54k, False: 920k] | | | Branch (755:28): [True: 10.7k, False: 909k] | | ------------------ | | 756| 924k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 924k| STORE(Dst[i], C) | | ------------------ | | | | 174| 924k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 819| 924k| Src += BpS; 820| 924k| Dst += BpS; 821| 924k| } 822| 106k|#endif 823| 106k|} qpel.c:V_Pass_8_C: 837| 52.6k|{ 838| |#if (SIZE==16) 839| | while(H-->0) { 840| | int C; 841| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 842| | CLIP_STORE(Dst[BpS* 0],C); 843| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 844| | CLIP_STORE(Dst[BpS* 1],C); 845| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 846| | CLIP_STORE(Dst[BpS* 2],C); 847| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 848| | CLIP_STORE(Dst[BpS* 3],C); 849| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 850| | CLIP_STORE(Dst[BpS* 4],C); 851| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 852| | CLIP_STORE(Dst[BpS* 5],C); 853| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 854| | CLIP_STORE(Dst[BpS* 6],C); 855| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 856| | CLIP_STORE(Dst[BpS* 7],C); 857| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 858| | CLIP_STORE(Dst[BpS* 8],C); 859| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 860| | CLIP_STORE(Dst[BpS* 9],C); 861| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 862| | CLIP_STORE(Dst[BpS*10],C); 863| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 864| | CLIP_STORE(Dst[BpS*11],C); 865| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 866| | CLIP_STORE(Dst[BpS*12],C); 867| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 868| | CLIP_STORE(Dst[BpS*13],C); 869| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 870| | CLIP_STORE(Dst[BpS*14],C); 871| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 872| | CLIP_STORE(Dst[BpS*15],C); 873| | Src += 1; 874| | Dst += 1; 875| | } 876| |#else 877| 474k| while(H-->0) { ------------------ | Branch (877:9): [True: 421k, False: 52.6k] ------------------ 878| 421k| int C; 879| 421k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 880| 421k| CLIP_STORE(Dst[BpS*0],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.99k, False: 419k] | | | Branch (832:28): [True: 4.96k, False: 414k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 881| 421k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 882| 421k| CLIP_STORE(Dst[BpS*1],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.66k, False: 418k] | | | Branch (832:28): [True: 4.07k, False: 414k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 883| 421k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 884| 421k| CLIP_STORE(Dst[BpS*2],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.93k, False: 417k] | | | Branch (832:28): [True: 4.15k, False: 413k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 885| 421k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 886| 421k| CLIP_STORE(Dst[BpS*3],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.08k, False: 418k] | | | Branch (832:28): [True: 3.85k, False: 414k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 887| 421k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 888| 421k| CLIP_STORE(Dst[BpS*4],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.28k, False: 418k] | | | Branch (832:28): [True: 5.17k, False: 413k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 889| 421k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 890| 421k| CLIP_STORE(Dst[BpS*5],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.06k, False: 419k] | | | Branch (832:28): [True: 4.29k, False: 415k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 891| 421k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 892| 421k| CLIP_STORE(Dst[BpS*6],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.44k, False: 418k] | | | Branch (832:28): [True: 4.41k, False: 413k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 893| 421k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 894| 421k| CLIP_STORE(Dst[BpS*7],C); ------------------ | | 832| 421k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.92k, False: 419k] | | | Branch (832:28): [True: 4.10k, False: 415k] | | ------------------ | | 833| 421k| STORE(D, C) | | ------------------ | | | | 174| 421k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 895| 421k| Src += 1; 896| 421k| Dst += 1; 897| 421k| } 898| 52.6k|#endif 899| 52.6k|} qpel.c:V_Pass_Avrg_8_C: 909| 53.8k|{ 910| |#if (SIZE==16) 911| | while(H-->0) { 912| | int C; 913| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 914| | CLIP_STORE(0,C); 915| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 916| | CLIP_STORE( 1,C); 917| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 918| | CLIP_STORE( 2,C); 919| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 920| | CLIP_STORE( 3,C); 921| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 922| | CLIP_STORE( 4,C); 923| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 924| | CLIP_STORE( 5,C); 925| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 926| | CLIP_STORE( 6,C); 927| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 928| | CLIP_STORE( 7,C); 929| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 930| | CLIP_STORE( 8,C); 931| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 932| | CLIP_STORE( 9,C); 933| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 934| | CLIP_STORE(10,C); 935| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 936| | CLIP_STORE(11,C); 937| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 938| | CLIP_STORE(12,C); 939| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 940| | CLIP_STORE(13,C); 941| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 942| | CLIP_STORE(14,C); 943| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 944| | CLIP_STORE(15,C); 945| | Src += 1; 946| | Dst += 1; 947| | } 948| |#else 949| 484k| while(H-->0) { ------------------ | Branch (949:9): [True: 430k, False: 53.8k] ------------------ 950| 430k| int C; 951| 430k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 952| 430k| CLIP_STORE(0,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.12k, False: 429k] | | | Branch (903:28): [True: 2.72k, False: 427k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 953| 430k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 954| 430k| CLIP_STORE(1,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.46k, False: 429k] | | | Branch (903:28): [True: 2.91k, False: 426k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 955| 430k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 956| 430k| CLIP_STORE(2,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.41k, False: 428k] | | | Branch (903:28): [True: 2.58k, False: 425k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 957| 430k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 958| 430k| CLIP_STORE(3,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.55k, False: 429k] | | | Branch (903:28): [True: 2.88k, False: 426k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 959| 430k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 960| 430k| CLIP_STORE(4,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.78k, False: 428k] | | | Branch (903:28): [True: 2.97k, False: 425k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 961| 430k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 962| 430k| CLIP_STORE(5,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.13k, False: 429k] | | | Branch (903:28): [True: 3.22k, False: 426k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 963| 430k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 964| 430k| CLIP_STORE(6,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.70k, False: 429k] | | | Branch (903:28): [True: 3.44k, False: 425k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 965| 430k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 966| 430k| CLIP_STORE(7,C); ------------------ | | 903| 430k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 940, False: 429k] | | | Branch (903:28): [True: 3.56k, False: 426k] | | ------------------ | | 904| 430k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 430k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 430k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 967| 430k| Src += 1; 968| 430k| Dst += 1; 969| 430k| } 970| 53.8k|#endif 971| 53.8k|} qpel.c:V_Pass_Avrg_Up_8_C: 981| 67.8k|{ 982| |#if (SIZE==16) 983| | while(H-->0) { 984| | int C; 985| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 986| | CLIP_STORE(0,C); 987| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 988| | CLIP_STORE( 1,C); 989| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 990| | CLIP_STORE( 2,C); 991| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 992| | CLIP_STORE( 3,C); 993| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 994| | CLIP_STORE( 4,C); 995| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 996| | CLIP_STORE( 5,C); 997| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 998| | CLIP_STORE( 6,C); 999| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 1000| | CLIP_STORE( 7,C); 1001| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 1002| | CLIP_STORE( 8,C); 1003| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 1004| | CLIP_STORE( 9,C); 1005| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 1006| | CLIP_STORE(10,C); 1007| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 1008| | CLIP_STORE(11,C); 1009| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 1010| | CLIP_STORE(12,C); 1011| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 1012| | CLIP_STORE(13,C); 1013| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 1014| | CLIP_STORE(14,C); 1015| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 1016| | CLIP_STORE(15,C); 1017| | Src += 1; 1018| | Dst += 1; 1019| | } 1020| |#else 1021| 610k| while(H-->0) { ------------------ | Branch (1021:9): [True: 542k, False: 67.8k] ------------------ 1022| 542k| int C; 1023| 542k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 1024| 542k| CLIP_STORE(0,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.14k, False: 540k] | | | Branch (975:28): [True: 2.60k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1025| 542k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 1026| 542k| CLIP_STORE(1,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.36k, False: 540k] | | | Branch (975:28): [True: 3.14k, False: 537k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1027| 542k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 1028| 542k| CLIP_STORE(2,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.37k, False: 541k] | | | Branch (975:28): [True: 2.62k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1029| 542k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 1030| 542k| CLIP_STORE(3,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.49k, False: 540k] | | | Branch (975:28): [True: 3.39k, False: 537k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1031| 542k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 1032| 542k| CLIP_STORE(4,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.41k, False: 541k] | | | Branch (975:28): [True: 2.91k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1033| 542k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 1034| 542k| CLIP_STORE(5,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.71k, False: 541k] | | | Branch (975:28): [True: 2.64k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1035| 542k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 1036| 542k| CLIP_STORE(6,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.10k, False: 541k] | | | Branch (975:28): [True: 2.91k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1037| 542k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 1038| 542k| CLIP_STORE(7,C); ------------------ | | 975| 542k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.59k, False: 541k] | | | Branch (975:28): [True: 3.00k, False: 538k] | | ------------------ | | 976| 542k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 542k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 174| 542k|#define STORE(d,s) (d) = (s) | | ------------------ ------------------ 1039| 542k| Src += 1; 1040| 542k| Dst += 1; 1041| 542k| } 1042| 67.8k|#endif 1043| 67.8k|} qpel.c:H_Pass_16_Add_C: 617| 7.74k|{ 618| 7.74k|#if (SIZE==16) 619| 131k| while(H-->0) { ------------------ | Branch (619:9): [True: 123k, False: 7.74k] ------------------ 620| 123k| int C; 621| 123k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 622| 123k| CLIP_STORE(Dst[ 0],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.00k, False: 121k] | | | Branch (612:28): [True: 3.27k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 623| 123k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 624| 123k| CLIP_STORE(Dst[ 1],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.54k, False: 121k] | | | Branch (612:28): [True: 3.92k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 625| 123k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 626| 123k| CLIP_STORE(Dst[ 2],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.49k, False: 121k] | | | Branch (612:28): [True: 2.87k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 627| 123k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 628| 123k| CLIP_STORE(Dst[ 3],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.68k, False: 121k] | | | Branch (612:28): [True: 3.39k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 629| 123k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 630| 123k| CLIP_STORE(Dst[ 4],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.43k, False: 121k] | | | Branch (612:28): [True: 3.13k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 631| 123k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 632| 123k| CLIP_STORE(Dst[ 5],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.46k, False: 121k] | | | Branch (612:28): [True: 3.30k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 633| 123k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 634| 123k| CLIP_STORE(Dst[ 6],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.96k, False: 121k] | | | Branch (612:28): [True: 3.13k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 635| 123k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 636| 123k| CLIP_STORE(Dst[ 7],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.97k, False: 121k] | | | Branch (612:28): [True: 3.44k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 637| 123k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 638| 123k| CLIP_STORE(Dst[ 8],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.83k, False: 121k] | | | Branch (612:28): [True: 2.80k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 639| 123k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 640| 123k| CLIP_STORE(Dst[ 9],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.70k, False: 120k] | | | Branch (612:28): [True: 4.20k, False: 116k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 641| 123k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 642| 123k| CLIP_STORE(Dst[10],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.05k, False: 120k] | | | Branch (612:28): [True: 3.07k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 643| 123k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 644| 123k| CLIP_STORE(Dst[11],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 4.09k, False: 119k] | | | Branch (612:28): [True: 3.48k, False: 116k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 645| 123k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 646| 123k| CLIP_STORE(Dst[12],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.14k, False: 120k] | | | Branch (612:28): [True: 3.50k, False: 117k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 647| 123k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 648| 123k| CLIP_STORE(Dst[13],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.41k, False: 121k] | | | Branch (612:28): [True: 3.43k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 649| 123k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 650| 123k| CLIP_STORE(Dst[14],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.21k, False: 120k] | | | Branch (612:28): [True: 3.84k, False: 116k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 651| 123k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 652| 123k| CLIP_STORE(Dst[15],C); ------------------ | | 612| 123k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.90k, False: 122k] | | | Branch (612:28): [True: 3.54k, False: 118k] | | ------------------ | | 613| 123k| STORE(D, C) | | ------------------ | | | | 156| 123k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 653| 123k| Src += BpS; 654| 123k| Dst += BpS; 655| 123k| } 656| |#else 657| | while(H-->0) { 658| | int C; 659| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 660| | CLIP_STORE(Dst[0],C); 661| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 662| | CLIP_STORE(Dst[1],C); 663| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 664| | CLIP_STORE(Dst[2],C); 665| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 666| | CLIP_STORE(Dst[3],C); 667| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 668| | CLIP_STORE(Dst[4],C); 669| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 670| | CLIP_STORE(Dst[5],C); 671| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 672| | CLIP_STORE(Dst[6],C); 673| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 674| | CLIP_STORE(Dst[7],C); 675| | Src += BpS; 676| | Dst += BpS; 677| | } 678| |#endif 679| 7.74k|} qpel.c:H_Pass_Avrg_16_Add_C: 689| 9.27k|{ 690| 9.27k|#if (SIZE==16) 691| 157k| while(H-->0) { ------------------ | Branch (691:9): [True: 148k, False: 9.27k] ------------------ 692| 148k| int C; 693| 148k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 694| 148k| CLIP_STORE(0,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.25k, False: 147k] | | | Branch (683:28): [True: 4.02k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 695| 148k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 696| 148k| CLIP_STORE( 1,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.87k, False: 146k] | | | Branch (683:28): [True: 3.53k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 697| 148k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 698| 148k| CLIP_STORE( 2,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.13k, False: 146k] | | | Branch (683:28): [True: 3.27k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 699| 148k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 700| 148k| CLIP_STORE( 3,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.09k, False: 146k] | | | Branch (683:28): [True: 3.65k, False: 142k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 701| 148k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 702| 148k| CLIP_STORE( 4,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.29k, False: 146k] | | | Branch (683:28): [True: 3.32k, False: 142k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 703| 148k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 704| 148k| CLIP_STORE( 5,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.10k, False: 146k] | | | Branch (683:28): [True: 2.64k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 705| 148k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 706| 148k| CLIP_STORE( 6,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.09k, False: 146k] | | | Branch (683:28): [True: 3.66k, False: 142k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 707| 148k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 708| 148k| CLIP_STORE( 7,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.68k, False: 145k] | | | Branch (683:28): [True: 2.81k, False: 142k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 709| 148k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 710| 148k| CLIP_STORE( 8,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.83k, False: 144k] | | | Branch (683:28): [True: 2.89k, False: 141k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 711| 148k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 712| 148k| CLIP_STORE( 9,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.22k, False: 146k] | | | Branch (683:28): [True: 2.69k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 713| 148k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 714| 148k| CLIP_STORE(10,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.74k, False: 144k] | | | Branch (683:28): [True: 2.95k, False: 141k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 715| 148k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 716| 148k| CLIP_STORE(11,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.70k, False: 146k] | | | Branch (683:28): [True: 2.78k, False: 143k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 717| 148k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 718| 148k| CLIP_STORE(12,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 4.00k, False: 144k] | | | Branch (683:28): [True: 2.81k, False: 141k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 719| 148k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 720| 148k| CLIP_STORE(13,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.70k, False: 146k] | | | Branch (683:28): [True: 2.28k, False: 144k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 721| 148k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 722| 148k| CLIP_STORE(14,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 3.77k, False: 144k] | | | Branch (683:28): [True: 2.83k, False: 141k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 723| 148k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 724| 148k| CLIP_STORE(15,C); ------------------ | | 683| 148k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.02k, False: 146k] | | | Branch (683:28): [True: 2.17k, False: 144k] | | ------------------ | | 684| 148k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 148k| STORE(Dst[i], C) | | ------------------ | | | | 156| 148k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 725| 148k| Src += BpS; 726| 148k| Dst += BpS; 727| 148k| } 728| |#else 729| | while(H-->0) { 730| | int C; 731| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 732| | CLIP_STORE(0,C); 733| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 734| | CLIP_STORE(1,C); 735| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 736| | CLIP_STORE(2,C); 737| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 738| | CLIP_STORE(3,C); 739| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 740| | CLIP_STORE(4,C); 741| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 742| | CLIP_STORE(5,C); 743| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 744| | CLIP_STORE(6,C); 745| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 746| | CLIP_STORE(7,C); 747| | Src += BpS; 748| | Dst += BpS; 749| | } 750| |#endif 751| 9.27k|} qpel.c:H_Pass_Avrg_Up_16_Add_C: 761| 10.4k|{ 762| 10.4k|#if (SIZE==16) 763| 177k| while(H-->0) { ------------------ | Branch (763:9): [True: 167k, False: 10.4k] ------------------ 764| 167k| int C; 765| 167k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 766| 167k| CLIP_STORE(0,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.69k, False: 165k] | | | Branch (755:28): [True: 2.64k, False: 162k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 767| 167k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 768| 167k| CLIP_STORE( 1,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.44k, False: 164k] | | | Branch (755:28): [True: 5.01k, False: 159k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 769| 167k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 770| 167k| CLIP_STORE( 2,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.22k, False: 165k] | | | Branch (755:28): [True: 2.88k, False: 162k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 771| 167k| C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 772| 167k| CLIP_STORE( 3,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.12k, False: 163k] | | | Branch (755:28): [True: 3.50k, False: 159k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 773| 167k| C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 774| 167k| CLIP_STORE( 4,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.84k, False: 165k] | | | Branch (755:28): [True: 4.31k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 775| 167k| C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 776| 167k| CLIP_STORE( 5,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 3.41k, False: 163k] | | | Branch (755:28): [True: 4.28k, False: 159k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 777| 167k| C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 778| 167k| CLIP_STORE( 6,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.04k, False: 165k] | | | Branch (755:28): [True: 2.97k, False: 162k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 779| 167k| C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 780| 167k| CLIP_STORE( 7,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.71k, False: 164k] | | | Branch (755:28): [True: 4.56k, False: 159k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 781| 167k| C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 782| 167k| CLIP_STORE( 8,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.93k, False: 164k] | | | Branch (755:28): [True: 2.97k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 783| 167k| C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 784| 167k| CLIP_STORE( 9,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.74k, False: 162k] | | | Branch (755:28): [True: 3.78k, False: 158k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 785| 167k| C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 786| 167k| CLIP_STORE(10,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.88k, False: 164k] | | | Branch (755:28): [True: 3.04k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 787| 167k| C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 788| 167k| CLIP_STORE(11,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.50k, False: 162k] | | | Branch (755:28): [True: 3.41k, False: 159k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 789| 167k| C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 790| 167k| CLIP_STORE(12,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.43k, False: 164k] | | | Branch (755:28): [True: 3.44k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 791| 167k| C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 792| 167k| CLIP_STORE(13,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.86k, False: 165k] | | | Branch (755:28): [True: 3.80k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 793| 167k| C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 794| 167k| CLIP_STORE(14,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.80k, False: 165k] | | | Branch (755:28): [True: 3.25k, False: 162k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 795| 167k| C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 796| 167k| CLIP_STORE(15,C); ------------------ | | 755| 167k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.48k, False: 164k] | | | Branch (755:28): [True: 3.65k, False: 161k] | | ------------------ | | 756| 167k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 167k| STORE(Dst[i], C) | | ------------------ | | | | 156| 167k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 797| 167k| Src += BpS; 798| 167k| Dst += BpS; 799| 167k| } 800| |#else 801| | while(H-->0) { 802| | int C; 803| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 804| | CLIP_STORE(0,C); 805| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 806| | CLIP_STORE(1,C); 807| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 808| | CLIP_STORE(2,C); 809| | C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 810| | CLIP_STORE(3,C); 811| | C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 812| | CLIP_STORE(4,C); 813| | C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 814| | CLIP_STORE(5,C); 815| | C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 816| | CLIP_STORE(6,C); 817| | C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 818| | CLIP_STORE(7,C); 819| | Src += BpS; 820| | Dst += BpS; 821| | } 822| |#endif 823| 10.4k|} qpel.c:V_Pass_16_Add_C: 837| 9.70k|{ 838| 9.70k|#if (SIZE==16) 839| 165k| while(H-->0) { ------------------ | Branch (839:9): [True: 155k, False: 9.70k] ------------------ 840| 155k| int C; 841| 155k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 842| 155k| CLIP_STORE(Dst[BpS* 0],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.06k, False: 153k] | | | Branch (832:28): [True: 1.97k, False: 151k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 843| 155k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 844| 155k| CLIP_STORE(Dst[BpS* 1],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 4.28k, False: 151k] | | | Branch (832:28): [True: 2.55k, False: 148k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 845| 155k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 846| 155k| CLIP_STORE(Dst[BpS* 2],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.00k, False: 153k] | | | Branch (832:28): [True: 2.22k, False: 151k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 847| 155k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 848| 155k| CLIP_STORE(Dst[BpS* 3],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 4.57k, False: 150k] | | | Branch (832:28): [True: 2.47k, False: 148k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 849| 155k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 850| 155k| CLIP_STORE(Dst[BpS* 4],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 5.37k, False: 149k] | | | Branch (832:28): [True: 2.73k, False: 147k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 851| 155k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 852| 155k| CLIP_STORE(Dst[BpS* 5],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.61k, False: 152k] | | | Branch (832:28): [True: 2.38k, False: 150k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 853| 155k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 854| 155k| CLIP_STORE(Dst[BpS* 6],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 6.47k, False: 148k] | | | Branch (832:28): [True: 2.46k, False: 146k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 855| 155k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 856| 155k| CLIP_STORE(Dst[BpS* 7],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.12k, False: 153k] | | | Branch (832:28): [True: 2.48k, False: 150k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 857| 155k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 858| 155k| CLIP_STORE(Dst[BpS* 8],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.84k, False: 152k] | | | Branch (832:28): [True: 2.50k, False: 149k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 859| 155k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 860| 155k| CLIP_STORE(Dst[BpS* 9],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.30k, False: 152k] | | | Branch (832:28): [True: 2.50k, False: 149k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 861| 155k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 862| 155k| CLIP_STORE(Dst[BpS*10],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 5.18k, False: 150k] | | | Branch (832:28): [True: 2.49k, False: 147k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 863| 155k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 864| 155k| CLIP_STORE(Dst[BpS*11],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.70k, False: 152k] | | | Branch (832:28): [True: 2.84k, False: 149k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 865| 155k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 866| 155k| CLIP_STORE(Dst[BpS*12],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 4.50k, False: 150k] | | | Branch (832:28): [True: 2.52k, False: 148k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 867| 155k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 868| 155k| CLIP_STORE(Dst[BpS*13],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.06k, False: 153k] | | | Branch (832:28): [True: 2.18k, False: 151k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 869| 155k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 870| 155k| CLIP_STORE(Dst[BpS*14],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.69k, False: 152k] | | | Branch (832:28): [True: 3.44k, False: 149k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 871| 155k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 872| 155k| CLIP_STORE(Dst[BpS*15],C); ------------------ | | 832| 155k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 1.73k, False: 153k] | | | Branch (832:28): [True: 3.31k, False: 150k] | | ------------------ | | 833| 155k| STORE(D, C) | | ------------------ | | | | 156| 155k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 873| 155k| Src += 1; 874| 155k| Dst += 1; 875| 155k| } 876| |#else 877| | while(H-->0) { 878| | int C; 879| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 880| | CLIP_STORE(Dst[BpS*0],C); 881| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 882| | CLIP_STORE(Dst[BpS*1],C); 883| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 884| | CLIP_STORE(Dst[BpS*2],C); 885| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 886| | CLIP_STORE(Dst[BpS*3],C); 887| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 888| | CLIP_STORE(Dst[BpS*4],C); 889| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 890| | CLIP_STORE(Dst[BpS*5],C); 891| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 892| | CLIP_STORE(Dst[BpS*6],C); 893| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 894| | CLIP_STORE(Dst[BpS*7],C); 895| | Src += 1; 896| | Dst += 1; 897| | } 898| |#endif 899| 9.70k|} qpel.c:V_Pass_Avrg_16_Add_C: 909| 11.8k|{ 910| 11.8k|#if (SIZE==16) 911| 201k| while(H-->0) { ------------------ | Branch (911:9): [True: 189k, False: 11.8k] ------------------ 912| 189k| int C; 913| 189k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 914| 189k| CLIP_STORE(0,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.84k, False: 188k] | | | Branch (903:28): [True: 3.70k, False: 184k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 915| 189k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 916| 189k| CLIP_STORE( 1,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.77k, False: 187k] | | | Branch (903:28): [True: 3.60k, False: 183k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 917| 189k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 918| 189k| CLIP_STORE( 2,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.45k, False: 187k] | | | Branch (903:28): [True: 3.89k, False: 183k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 919| 189k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 920| 189k| CLIP_STORE( 3,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.15k, False: 186k] | | | Branch (903:28): [True: 3.79k, False: 183k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 921| 189k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 922| 189k| CLIP_STORE( 4,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.12k, False: 185k] | | | Branch (903:28): [True: 4.29k, False: 181k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 923| 189k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 924| 189k| CLIP_STORE( 5,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.53k, False: 187k] | | | Branch (903:28): [True: 3.30k, False: 184k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 925| 189k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 926| 189k| CLIP_STORE( 6,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.02k, False: 185k] | | | Branch (903:28): [True: 3.96k, False: 181k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 927| 189k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 928| 189k| CLIP_STORE( 7,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.17k, False: 187k] | | | Branch (903:28): [True: 3.37k, False: 184k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 929| 189k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 930| 189k| CLIP_STORE( 8,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.81k, False: 187k] | | | Branch (903:28): [True: 4.40k, False: 182k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 931| 189k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 932| 189k| CLIP_STORE( 9,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.41k, False: 187k] | | | Branch (903:28): [True: 4.01k, False: 183k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 933| 189k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 934| 189k| CLIP_STORE(10,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.06k, False: 186k] | | | Branch (903:28): [True: 4.66k, False: 182k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 935| 189k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 936| 189k| CLIP_STORE(11,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.63k, False: 186k] | | | Branch (903:28): [True: 4.99k, False: 181k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 937| 189k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 938| 189k| CLIP_STORE(12,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 6.74k, False: 183k] | | | Branch (903:28): [True: 4.71k, False: 178k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 939| 189k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 940| 189k| CLIP_STORE(13,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.01k, False: 186k] | | | Branch (903:28): [True: 3.15k, False: 183k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 941| 189k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 942| 189k| CLIP_STORE(14,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 6.41k, False: 183k] | | | Branch (903:28): [True: 4.58k, False: 178k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 943| 189k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 944| 189k| CLIP_STORE(15,C); ------------------ | | 903| 189k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 1.68k, False: 188k] | | | Branch (903:28): [True: 2.75k, False: 185k] | | ------------------ | | 904| 189k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 189k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 189k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 945| 189k| Src += 1; 946| 189k| Dst += 1; 947| 189k| } 948| |#else 949| | while(H-->0) { 950| | int C; 951| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 952| | CLIP_STORE(0,C); 953| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 954| | CLIP_STORE(1,C); 955| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 956| | CLIP_STORE(2,C); 957| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 958| | CLIP_STORE(3,C); 959| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 960| | CLIP_STORE(4,C); 961| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 962| | CLIP_STORE(5,C); 963| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 964| | CLIP_STORE(6,C); 965| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 966| | CLIP_STORE(7,C); 967| | Src += 1; 968| | Dst += 1; 969| | } 970| |#endif 971| 11.8k|} qpel.c:V_Pass_Avrg_Up_16_Add_C: 981| 14.5k|{ 982| 14.5k|#if (SIZE==16) 983| 247k| while(H-->0) { ------------------ | Branch (983:9): [True: 232k, False: 14.5k] ------------------ 984| 232k| int C; 985| 232k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 986| 232k| CLIP_STORE(0,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.34k, False: 230k] | | | Branch (975:28): [True: 4.56k, False: 226k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 987| 232k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 988| 232k| CLIP_STORE( 1,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 9.01k, False: 223k] | | | Branch (975:28): [True: 5.59k, False: 218k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 989| 232k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 990| 232k| CLIP_STORE( 2,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.96k, False: 230k] | | | Branch (975:28): [True: 4.31k, False: 225k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 991| 232k| C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 992| 232k| CLIP_STORE( 3,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 9.27k, False: 223k] | | | Branch (975:28): [True: 5.75k, False: 217k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 993| 232k| C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 994| 232k| CLIP_STORE( 4,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.89k, False: 230k] | | | Branch (975:28): [True: 5.63k, False: 224k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 995| 232k| C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 996| 232k| CLIP_STORE( 5,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 4.47k, False: 228k] | | | Branch (975:28): [True: 6.38k, False: 222k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 997| 232k| C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 998| 232k| CLIP_STORE( 6,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.46k, False: 229k] | | | Branch (975:28): [True: 5.26k, False: 224k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 999| 232k| C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 1000| 232k| CLIP_STORE( 7,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 5.63k, False: 227k] | | | Branch (975:28): [True: 6.03k, False: 221k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1001| 232k| C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 1002| 232k| CLIP_STORE( 8,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.22k, False: 229k] | | | Branch (975:28): [True: 5.13k, False: 224k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1003| 232k| C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 1004| 232k| CLIP_STORE( 9,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 4.55k, False: 228k] | | | Branch (975:28): [True: 6.68k, False: 221k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1005| 232k| C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 1006| 232k| CLIP_STORE(10,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.75k, False: 229k] | | | Branch (975:28): [True: 5.86k, False: 223k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1007| 232k| C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 1008| 232k| CLIP_STORE(11,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 5.29k, False: 227k] | | | Branch (975:28): [True: 6.89k, False: 220k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1009| 232k| C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 1010| 232k| CLIP_STORE(12,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 2.93k, False: 230k] | | | Branch (975:28): [True: 6.45k, False: 223k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1011| 232k| C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 1012| 232k| CLIP_STORE(13,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.57k, False: 229k] | | | Branch (975:28): [True: 6.20k, False: 223k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1013| 232k| C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 1014| 232k| CLIP_STORE(14,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.79k, False: 229k] | | | Branch (975:28): [True: 6.48k, False: 222k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1015| 232k| C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 1016| 232k| CLIP_STORE(15,C); ------------------ | | 975| 232k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 1.86k, False: 231k] | | | Branch (975:28): [True: 6.71k, False: 224k] | | ------------------ | | 976| 232k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 232k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 156| 232k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1017| 232k| Src += 1; 1018| 232k| Dst += 1; 1019| 232k| } 1020| |#else 1021| | while(H-->0) { 1022| | int C; 1023| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 1024| | CLIP_STORE(0,C); 1025| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 1026| | CLIP_STORE(1,C); 1027| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 1028| | CLIP_STORE(2,C); 1029| | C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 1030| | CLIP_STORE(3,C); 1031| | C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 1032| | CLIP_STORE(4,C); 1033| | C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 1034| | CLIP_STORE(5,C); 1035| | C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 1036| | CLIP_STORE(6,C); 1037| | C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 1038| | CLIP_STORE(7,C); 1039| | Src += 1; 1040| | Dst += 1; 1041| | } 1042| |#endif 1043| 14.5k|} qpel.c:H_Pass_8_Add_C: 617| 23.4k|{ 618| |#if (SIZE==16) 619| | while(H-->0) { 620| | int C; 621| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 622| | CLIP_STORE(Dst[ 0],C); 623| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 624| | CLIP_STORE(Dst[ 1],C); 625| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 626| | CLIP_STORE(Dst[ 2],C); 627| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 628| | CLIP_STORE(Dst[ 3],C); 629| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 630| | CLIP_STORE(Dst[ 4],C); 631| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 632| | CLIP_STORE(Dst[ 5],C); 633| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 634| | CLIP_STORE(Dst[ 6],C); 635| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 636| | CLIP_STORE(Dst[ 7],C); 637| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 638| | CLIP_STORE(Dst[ 8],C); 639| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 640| | CLIP_STORE(Dst[ 9],C); 641| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 642| | CLIP_STORE(Dst[10],C); 643| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 644| | CLIP_STORE(Dst[11],C); 645| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 646| | CLIP_STORE(Dst[12],C); 647| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 648| | CLIP_STORE(Dst[13],C); 649| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 650| | CLIP_STORE(Dst[14],C); 651| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 652| | CLIP_STORE(Dst[15],C); 653| | Src += BpS; 654| | Dst += BpS; 655| | } 656| |#else 657| 210k| while(H-->0) { ------------------ | Branch (657:9): [True: 187k, False: 23.4k] ------------------ 658| 187k| int C; 659| 187k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 660| 187k| CLIP_STORE(Dst[0],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.56k, False: 185k] | | | Branch (612:28): [True: 4.05k, False: 181k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 661| 187k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 662| 187k| CLIP_STORE(Dst[1],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.82k, False: 184k] | | | Branch (612:28): [True: 5.75k, False: 178k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 663| 187k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 664| 187k| CLIP_STORE(Dst[2],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.56k, False: 185k] | | | Branch (612:28): [True: 3.57k, False: 182k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 665| 187k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 666| 187k| CLIP_STORE(Dst[3],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 3.38k, False: 184k] | | | Branch (612:28): [True: 4.95k, False: 179k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 667| 187k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 668| 187k| CLIP_STORE(Dst[4],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.30k, False: 185k] | | | Branch (612:28): [True: 4.81k, False: 180k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 669| 187k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 670| 187k| CLIP_STORE(Dst[5],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.26k, False: 185k] | | | Branch (612:28): [True: 4.67k, False: 180k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 671| 187k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 672| 187k| CLIP_STORE(Dst[6],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 2.33k, False: 185k] | | | Branch (612:28): [True: 4.61k, False: 180k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 673| 187k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 674| 187k| CLIP_STORE(Dst[7],C); ------------------ | | 612| 187k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (612:7): [True: 1.99k, False: 185k] | | | Branch (612:28): [True: 4.77k, False: 180k] | | ------------------ | | 613| 187k| STORE(D, C) | | ------------------ | | | | 185| 187k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 675| 187k| Src += BpS; 676| 187k| Dst += BpS; 677| 187k| } 678| 23.4k|#endif 679| 23.4k|} qpel.c:H_Pass_Avrg_8_Add_C: 689| 37.2k|{ 690| |#if (SIZE==16) 691| | while(H-->0) { 692| | int C; 693| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 694| | CLIP_STORE(0,C); 695| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 696| | CLIP_STORE( 1,C); 697| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 698| | CLIP_STORE( 2,C); 699| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 700| | CLIP_STORE( 3,C); 701| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 702| | CLIP_STORE( 4,C); 703| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 704| | CLIP_STORE( 5,C); 705| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 706| | CLIP_STORE( 6,C); 707| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 708| | CLIP_STORE( 7,C); 709| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 710| | CLIP_STORE( 8,C); 711| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 712| | CLIP_STORE( 9,C); 713| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 714| | CLIP_STORE(10,C); 715| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 716| | CLIP_STORE(11,C); 717| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 718| | CLIP_STORE(12,C); 719| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 720| | CLIP_STORE(13,C); 721| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 722| | CLIP_STORE(14,C); 723| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 724| | CLIP_STORE(15,C); 725| | Src += BpS; 726| | Dst += BpS; 727| | } 728| |#else 729| 335k| while(H-->0) { ------------------ | Branch (729:9): [True: 297k, False: 37.2k] ------------------ 730| 297k| int C; 731| 297k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 732| 297k| CLIP_STORE(0,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.41k, False: 295k] | | | Branch (683:28): [True: 8.57k, False: 286k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 733| 297k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 734| 297k| CLIP_STORE(1,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.76k, False: 295k] | | | Branch (683:28): [True: 8.89k, False: 286k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 735| 297k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 736| 297k| CLIP_STORE(2,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.99k, False: 294k] | | | Branch (683:28): [True: 8.31k, False: 286k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 737| 297k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 738| 297k| CLIP_STORE(3,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.83k, False: 294k] | | | Branch (683:28): [True: 8.31k, False: 286k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 739| 297k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 740| 297k| CLIP_STORE(4,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 6.22k, False: 291k] | | | Branch (683:28): [True: 8.93k, False: 282k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 741| 297k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 742| 297k| CLIP_STORE(5,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 2.50k, False: 295k] | | | Branch (683:28): [True: 5.74k, False: 289k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 743| 297k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 744| 297k| CLIP_STORE(6,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 6.77k, False: 291k] | | | Branch (683:28): [True: 9.13k, False: 281k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 745| 297k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 746| 297k| CLIP_STORE(7,C); ------------------ | | 683| 297k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (683:7): [True: 1.76k, False: 296k] | | | Branch (683:28): [True: 5.82k, False: 290k] | | ------------------ | | 684| 297k| C = (C+Src[i]+1-RND) >> 1; \ | | 685| 297k| STORE(Dst[i], C) | | ------------------ | | | | 185| 297k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 747| 297k| Src += BpS; 748| 297k| Dst += BpS; 749| 297k| } 750| 37.2k|#endif 751| 37.2k|} qpel.c:H_Pass_Avrg_Up_8_Add_C: 761| 33.1k|{ 762| |#if (SIZE==16) 763| | while(H-->0) { 764| | int C; 765| | C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 766| | CLIP_STORE(0,C); 767| | C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 768| | CLIP_STORE( 1,C); 769| | C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 770| | CLIP_STORE( 2,C); 771| | C = 16-RND - (Src[0]+Src[7 ]) + 3*(Src[ 1]+Src[ 6])-6*(Src[ 2]+Src[ 5]) + 20*(Src[ 3]+Src[ 4]); 772| | CLIP_STORE( 3,C); 773| | C = 16-RND - (Src[1]+Src[8 ]) + 3*(Src[ 2]+Src[ 7])-6*(Src[ 3]+Src[ 6]) + 20*(Src[ 4]+Src[ 5]); 774| | CLIP_STORE( 4,C); 775| | C = 16-RND - (Src[2]+Src[9 ]) + 3*(Src[ 3]+Src[ 8])-6*(Src[ 4]+Src[ 7]) + 20*(Src[ 5]+Src[ 6]); 776| | CLIP_STORE( 5,C); 777| | C = 16-RND - (Src[3]+Src[10]) + 3*(Src[ 4]+Src[ 9])-6*(Src[ 5]+Src[ 8]) + 20*(Src[ 6]+Src[ 7]); 778| | CLIP_STORE( 6,C); 779| | C = 16-RND - (Src[4]+Src[11]) + 3*(Src[ 5]+Src[10])-6*(Src[ 6]+Src[ 9]) + 20*(Src[ 7]+Src[ 8]); 780| | CLIP_STORE( 7,C); 781| | C = 16-RND - (Src[5]+Src[12]) + 3*(Src[ 6]+Src[11])-6*(Src[ 7]+Src[10]) + 20*(Src[ 8]+Src[ 9]); 782| | CLIP_STORE( 8,C); 783| | C = 16-RND - (Src[6]+Src[13]) + 3*(Src[ 7]+Src[12])-6*(Src[ 8]+Src[11]) + 20*(Src[ 9]+Src[10]); 784| | CLIP_STORE( 9,C); 785| | C = 16-RND - (Src[7]+Src[14]) + 3*(Src[ 8]+Src[13])-6*(Src[ 9]+Src[12]) + 20*(Src[10]+Src[11]); 786| | CLIP_STORE(10,C); 787| | C = 16-RND - (Src[8]+Src[15]) + 3*(Src[ 9]+Src[14])-6*(Src[10]+Src[13]) + 20*(Src[11]+Src[12]); 788| | CLIP_STORE(11,C); 789| | C = 16-RND - (Src[9]+Src[16]) + 3*(Src[10]+Src[15])-6*(Src[11]+Src[14]) + 20*(Src[12]+Src[13]); 790| | CLIP_STORE(12,C); 791| | C = 16-RND - Src[10] +3*Src[11] -6*(Src[12]+Src[15]) + 20*(Src[13]+Src[14]) +2*Src[16]; 792| | CLIP_STORE(13,C); 793| | C = 16-RND - Src[11] +3*(Src[12]-Src[16]) -6*Src[13] + 20*Src[14] + 19*Src[15]; 794| | CLIP_STORE(14,C); 795| | C = 16-RND - Src[12] +3*Src[13] -7*Src[14] + 23*Src[15] + 14*Src[16]; 796| | CLIP_STORE(15,C); 797| | Src += BpS; 798| | Dst += BpS; 799| | } 800| |#else 801| 298k| while(H-->0) { ------------------ | Branch (801:9): [True: 265k, False: 33.1k] ------------------ 802| 265k| int C; 803| 265k| C = 16-RND +14*Src[0] +23*Src[1] - 7*Src[2] + 3*Src[3] - Src[4]; 804| 265k| CLIP_STORE(0,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.56k, False: 263k] | | | Branch (755:28): [True: 4.88k, False: 258k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 805| 265k| C = 16-RND - 3*(Src[0]-Src[4]) +19*Src[1] +20*Src[2] - 6*Src[3] - Src[5]; 806| 265k| CLIP_STORE(1,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.44k, False: 260k] | | | Branch (755:28): [True: 7.19k, False: 253k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 807| 265k| C = 16-RND + 2*Src[0] - 6*(Src[1]+Src[4]) +20*(Src[2]+Src[3]) + 3*Src[5] - Src[6]; 808| 265k| CLIP_STORE(2,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 1.68k, False: 263k] | | | Branch (755:28): [True: 4.66k, False: 258k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 809| 265k| C = 16-RND - (Src[0]+Src[7]) + 3*(Src[1]+Src[6])-6*(Src[2]+Src[5]) + 20*(Src[3]+Src[4]); 810| 265k| CLIP_STORE(3,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 4.98k, False: 260k] | | | Branch (755:28): [True: 6.71k, False: 253k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 811| 265k| C = 16-RND - (Src[1]+Src[8]) + 3*(Src[2]+Src[7])-6*(Src[3]+Src[6]) + 20*(Src[4]+Src[5]); 812| 265k| CLIP_STORE(4,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.07k, False: 263k] | | | Branch (755:28): [True: 6.48k, False: 256k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 813| 265k| C = 16-RND - Src[2] +3*Src[3] -6*(Src[4]+Src[7]) + 20*(Src[5]+Src[6]) +2*Src[8]; 814| 265k| CLIP_STORE(5,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.23k, False: 262k] | | | Branch (755:28): [True: 6.63k, False: 256k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 815| 265k| C = 16-RND - Src[3] +3*(Src[4]-Src[8]) -6*Src[5] + 20*Src[6] + 19*Src[7]; 816| 265k| CLIP_STORE(6,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.39k, False: 262k] | | | Branch (755:28): [True: 6.26k, False: 256k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 817| 265k| C = 16-RND - Src[4] +3*Src[5] -7*Src[6] + 23*Src[7] + 14*Src[8]; 818| 265k| CLIP_STORE(7,C); ------------------ | | 755| 265k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (755:7): [True: 2.04k, False: 263k] | | | Branch (755:28): [True: 6.35k, False: 256k] | | ------------------ | | 756| 265k| C = (C+Src[i+1]+1-RND) >> 1; \ | | 757| 265k| STORE(Dst[i], C) | | ------------------ | | | | 185| 265k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 819| 265k| Src += BpS; 820| 265k| Dst += BpS; 821| 265k| } 822| 33.1k|#endif 823| 33.1k|} qpel.c:V_Pass_8_Add_C: 837| 41.7k|{ 838| |#if (SIZE==16) 839| | while(H-->0) { 840| | int C; 841| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 842| | CLIP_STORE(Dst[BpS* 0],C); 843| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 844| | CLIP_STORE(Dst[BpS* 1],C); 845| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 846| | CLIP_STORE(Dst[BpS* 2],C); 847| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 848| | CLIP_STORE(Dst[BpS* 3],C); 849| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 850| | CLIP_STORE(Dst[BpS* 4],C); 851| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 852| | CLIP_STORE(Dst[BpS* 5],C); 853| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 854| | CLIP_STORE(Dst[BpS* 6],C); 855| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 856| | CLIP_STORE(Dst[BpS* 7],C); 857| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 858| | CLIP_STORE(Dst[BpS* 8],C); 859| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 860| | CLIP_STORE(Dst[BpS* 9],C); 861| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 862| | CLIP_STORE(Dst[BpS*10],C); 863| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 864| | CLIP_STORE(Dst[BpS*11],C); 865| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 866| | CLIP_STORE(Dst[BpS*12],C); 867| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 868| | CLIP_STORE(Dst[BpS*13],C); 869| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 870| | CLIP_STORE(Dst[BpS*14],C); 871| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 872| | CLIP_STORE(Dst[BpS*15],C); 873| | Src += 1; 874| | Dst += 1; 875| | } 876| |#else 877| 376k| while(H-->0) { ------------------ | Branch (877:9): [True: 334k, False: 41.7k] ------------------ 878| 334k| int C; 879| 334k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 880| 334k| CLIP_STORE(Dst[BpS*0],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.77k, False: 331k] | | | Branch (832:28): [True: 7.96k, False: 323k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 881| 334k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 882| 334k| CLIP_STORE(Dst[BpS*1],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 7.46k, False: 326k] | | | Branch (832:28): [True: 10.9k, False: 315k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 883| 334k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 884| 334k| CLIP_STORE(Dst[BpS*2],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.35k, False: 330k] | | | Branch (832:28): [True: 7.74k, False: 323k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 885| 334k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 886| 334k| CLIP_STORE(Dst[BpS*3],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 8.07k, False: 326k] | | | Branch (832:28): [True: 10.1k, False: 316k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 887| 334k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 888| 334k| CLIP_STORE(Dst[BpS*4],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 5.69k, False: 328k] | | | Branch (832:28): [True: 10.2k, False: 318k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 889| 334k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 890| 334k| CLIP_STORE(Dst[BpS*5],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 3.28k, False: 331k] | | | Branch (832:28): [True: 9.38k, False: 321k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 891| 334k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 892| 334k| CLIP_STORE(Dst[BpS*6],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 5.08k, False: 329k] | | | Branch (832:28): [True: 10.5k, False: 318k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 893| 334k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 894| 334k| CLIP_STORE(Dst[BpS*7],C); ------------------ | | 832| 334k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (832:7): [True: 2.87k, False: 331k] | | | Branch (832:28): [True: 9.58k, False: 321k] | | ------------------ | | 833| 334k| STORE(D, C) | | ------------------ | | | | 185| 334k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 895| 334k| Src += 1; 896| 334k| Dst += 1; 897| 334k| } 898| 41.7k|#endif 899| 41.7k|} qpel.c:V_Pass_Avrg_8_Add_C: 909| 47.4k|{ 910| |#if (SIZE==16) 911| | while(H-->0) { 912| | int C; 913| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 914| | CLIP_STORE(0,C); 915| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 916| | CLIP_STORE( 1,C); 917| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 918| | CLIP_STORE( 2,C); 919| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 920| | CLIP_STORE( 3,C); 921| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 922| | CLIP_STORE( 4,C); 923| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 924| | CLIP_STORE( 5,C); 925| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 926| | CLIP_STORE( 6,C); 927| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 928| | CLIP_STORE( 7,C); 929| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 930| | CLIP_STORE( 8,C); 931| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 932| | CLIP_STORE( 9,C); 933| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 934| | CLIP_STORE(10,C); 935| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 936| | CLIP_STORE(11,C); 937| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 938| | CLIP_STORE(12,C); 939| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 940| | CLIP_STORE(13,C); 941| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 942| | CLIP_STORE(14,C); 943| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 944| | CLIP_STORE(15,C); 945| | Src += 1; 946| | Dst += 1; 947| | } 948| |#else 949| 427k| while(H-->0) { ------------------ | Branch (949:9): [True: 379k, False: 47.4k] ------------------ 950| 379k| int C; 951| 379k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 952| 379k| CLIP_STORE(0,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.42k, False: 376k] | | | Branch (903:28): [True: 9.30k, False: 366k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 953| 379k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 954| 379k| CLIP_STORE(1,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.94k, False: 374k] | | | Branch (903:28): [True: 9.53k, False: 365k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 955| 379k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 956| 379k| CLIP_STORE(2,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.42k, False: 376k] | | | Branch (903:28): [True: 9.57k, False: 366k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 957| 379k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 958| 379k| CLIP_STORE(3,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 4.83k, False: 374k] | | | Branch (903:28): [True: 10.1k, False: 364k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 959| 379k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 960| 379k| CLIP_STORE(4,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 11.3k, False: 368k] | | | Branch (903:28): [True: 10.0k, False: 358k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 961| 379k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 962| 379k| CLIP_STORE(5,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 3.30k, False: 376k] | | | Branch (903:28): [True: 6.22k, False: 370k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 963| 379k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 964| 379k| CLIP_STORE(6,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 11.4k, False: 368k] | | | Branch (903:28): [True: 11.1k, False: 357k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 965| 379k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 966| 379k| CLIP_STORE(7,C); ------------------ | | 903| 379k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (903:7): [True: 2.18k, False: 377k] | | | Branch (903:28): [True: 6.30k, False: 371k] | | ------------------ | | 904| 379k| C = (C+Src[BpS*i]+1-RND) >> 1; \ | | 905| 379k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 379k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 967| 379k| Src += 1; 968| 379k| Dst += 1; 969| 379k| } 970| 47.4k|#endif 971| 47.4k|} qpel.c:V_Pass_Avrg_Up_8_Add_C: 981| 55.6k|{ 982| |#if (SIZE==16) 983| | while(H-->0) { 984| | int C; 985| | C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 986| | CLIP_STORE(0,C); 987| | C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 988| | CLIP_STORE( 1,C); 989| | C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 990| | CLIP_STORE( 2,C); 991| | C = 16-RND - (Src[BpS*0]+Src[BpS*7 ]) + 3*(Src[BpS* 1]+Src[BpS* 6])-6*(Src[BpS* 2]+Src[BpS* 5]) + 20*(Src[BpS* 3]+Src[BpS* 4]); 992| | CLIP_STORE( 3,C); 993| | C = 16-RND - (Src[BpS*1]+Src[BpS*8 ]) + 3*(Src[BpS* 2]+Src[BpS* 7])-6*(Src[BpS* 3]+Src[BpS* 6]) + 20*(Src[BpS* 4]+Src[BpS* 5]); 994| | CLIP_STORE( 4,C); 995| | C = 16-RND - (Src[BpS*2]+Src[BpS*9 ]) + 3*(Src[BpS* 3]+Src[BpS* 8])-6*(Src[BpS* 4]+Src[BpS* 7]) + 20*(Src[BpS* 5]+Src[BpS* 6]); 996| | CLIP_STORE( 5,C); 997| | C = 16-RND - (Src[BpS*3]+Src[BpS*10]) + 3*(Src[BpS* 4]+Src[BpS* 9])-6*(Src[BpS* 5]+Src[BpS* 8]) + 20*(Src[BpS* 6]+Src[BpS* 7]); 998| | CLIP_STORE( 6,C); 999| | C = 16-RND - (Src[BpS*4]+Src[BpS*11]) + 3*(Src[BpS* 5]+Src[BpS*10])-6*(Src[BpS* 6]+Src[BpS* 9]) + 20*(Src[BpS* 7]+Src[BpS* 8]); 1000| | CLIP_STORE( 7,C); 1001| | C = 16-RND - (Src[BpS*5]+Src[BpS*12]) + 3*(Src[BpS* 6]+Src[BpS*11])-6*(Src[BpS* 7]+Src[BpS*10]) + 20*(Src[BpS* 8]+Src[BpS* 9]); 1002| | CLIP_STORE( 8,C); 1003| | C = 16-RND - (Src[BpS*6]+Src[BpS*13]) + 3*(Src[BpS* 7]+Src[BpS*12])-6*(Src[BpS* 8]+Src[BpS*11]) + 20*(Src[BpS* 9]+Src[BpS*10]); 1004| | CLIP_STORE( 9,C); 1005| | C = 16-RND - (Src[BpS*7]+Src[BpS*14]) + 3*(Src[BpS* 8]+Src[BpS*13])-6*(Src[BpS* 9]+Src[BpS*12]) + 20*(Src[BpS*10]+Src[BpS*11]); 1006| | CLIP_STORE(10,C); 1007| | C = 16-RND - (Src[BpS*8]+Src[BpS*15]) + 3*(Src[BpS* 9]+Src[BpS*14])-6*(Src[BpS*10]+Src[BpS*13]) + 20*(Src[BpS*11]+Src[BpS*12]); 1008| | CLIP_STORE(11,C); 1009| | C = 16-RND - (Src[BpS*9]+Src[BpS*16]) + 3*(Src[BpS*10]+Src[BpS*15])-6*(Src[BpS*11]+Src[BpS*14]) + 20*(Src[BpS*12]+Src[BpS*13]); 1010| | CLIP_STORE(12,C); 1011| | C = 16-RND - Src[BpS*10] +3*Src[BpS*11] -6*(Src[BpS*12]+Src[BpS*15]) + 20*(Src[BpS*13]+Src[BpS*14]) +2*Src[BpS*16]; 1012| | CLIP_STORE(13,C); 1013| | C = 16-RND - Src[BpS*11] +3*(Src[BpS*12]-Src[BpS*16]) -6*Src[BpS*13] + 20*Src[BpS*14] + 19*Src[BpS*15]; 1014| | CLIP_STORE(14,C); 1015| | C = 16-RND - Src[BpS*12] +3*Src[BpS*13] -7*Src[BpS*14] + 23*Src[BpS*15] + 14*Src[BpS*16]; 1016| | CLIP_STORE(15,C); 1017| | Src += 1; 1018| | Dst += 1; 1019| | } 1020| |#else 1021| 500k| while(H-->0) { ------------------ | Branch (1021:9): [True: 444k, False: 55.6k] ------------------ 1022| 444k| int C; 1023| 444k| C = 16-RND +14*Src[BpS*0] +23*Src[BpS*1] - 7*Src[BpS*2] + 3*Src[BpS*3] - Src[BpS*4]; 1024| 444k| CLIP_STORE(0,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.02k, False: 441k] | | | Branch (975:28): [True: 6.56k, False: 435k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1025| 444k| C = 16-RND - 3*(Src[BpS*0]-Src[BpS*4]) +19*Src[BpS*1] +20*Src[BpS*2] - 6*Src[BpS*3] - Src[BpS*5]; 1026| 444k| CLIP_STORE(1,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 10.7k, False: 434k] | | | Branch (975:28): [True: 10.1k, False: 424k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1027| 444k| C = 16-RND + 2*Src[BpS*0] - 6*(Src[BpS*1]+Src[BpS*4]) +20*(Src[BpS*2]+Src[BpS*3]) + 3*Src[BpS*5] - Src[BpS*6]; 1028| 444k| CLIP_STORE(2,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.59k, False: 441k] | | | Branch (975:28): [True: 7.20k, False: 434k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1029| 444k| C = 16-RND - (Src[BpS*0]+Src[BpS*7]) + 3*(Src[BpS*1]+Src[BpS*6])-6*(Src[BpS*2]+Src[BpS*5]) + 20*(Src[BpS*3]+Src[BpS*4]); 1030| 444k| CLIP_STORE(3,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 11.0k, False: 433k] | | | Branch (975:28): [True: 10.4k, False: 423k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1031| 444k| C = 16-RND - (Src[BpS*1]+Src[BpS*8]) + 3*(Src[BpS*2]+Src[BpS*7])-6*(Src[BpS*3]+Src[BpS*6]) + 20*(Src[BpS*4]+Src[BpS*5]); 1032| 444k| CLIP_STORE(4,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 4.65k, False: 440k] | | | Branch (975:28): [True: 9.25k, False: 430k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1033| 444k| C = 16-RND - Src[BpS*2] +3*Src[BpS*3] -6*(Src[BpS*4]+Src[BpS*7]) + 20*(Src[BpS*5]+Src[BpS*6]) +2*Src[BpS*8]; 1034| 444k| CLIP_STORE(5,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 4.97k, False: 439k] | | | Branch (975:28): [True: 9.58k, False: 430k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1035| 444k| C = 16-RND - Src[BpS*3] +3*(Src[BpS*4]-Src[BpS*8]) -6*Src[BpS*5] + 20*Src[BpS*6] + 19*Src[BpS*7]; 1036| 444k| CLIP_STORE(6,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 4.92k, False: 439k] | | | Branch (975:28): [True: 9.51k, False: 430k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1037| 444k| C = 16-RND - Src[BpS*4] +3*Src[BpS*5] -7*Src[BpS*6] + 23*Src[BpS*7] + 14*Src[BpS*8]; 1038| 444k| CLIP_STORE(7,C); ------------------ | | 975| 444k| if (C<0) C = 0; else if (C>(255<<5)) C = 255; else C = C>>5; \ | | ------------------ | | | Branch (975:7): [True: 3.70k, False: 441k] | | | Branch (975:28): [True: 9.56k, False: 431k] | | ------------------ | | 976| 444k| C = (C+Src[BpS*i+BpS]+1-RND) >> 1; \ | | 977| 444k| STORE(Dst[BpS*i], C) | | ------------------ | | | | 185| 444k|#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 | | ------------------ ------------------ 1039| 444k| Src += 1; 1040| 444k| Dst += 1; 1041| 444k| } 1042| 55.6k|#endif 1043| 55.6k|} qpel.c:Init_FIR_Table: 400| 34|{ 401| 34| int i; 402| 8.73k| for(i=0; i<256; ++i) { ------------------ | Branch (402:11): [True: 8.70k, False: 34] ------------------ 403| 8.70k| Tab[i][0] = i*A; 404| 8.70k| Tab[i][1] = i*B; 405| 8.70k| Tab[i][2] = i*C; 406| 8.70k| Tab[i][3] = i*D; 407| 8.70k| } 408| 34|} decoder.c:interpolate16x16_quarterpel: 119| 191k|{ 120| 191k| const uint8_t *src; 121| 191k| uint8_t *dst; 122| 191k| uint8_t *tmp; 123| 191k| int32_t quads; 124| 191k| const XVID_QP_FUNCS *Ops; 125| | 126| 191k| int32_t x_int, y_int; 127| | 128| 191k| const int32_t xRef = (int)x*4 + dx; 129| 191k| const int32_t yRef = (int)y*4 + dy; 130| | 131| 191k| Ops = xvid_QP_Funcs; 132| 191k| quads = (dx&3) | ((dy&3)<<2); 133| | 134| 191k| x_int = xRef >> 2; 135| 191k| y_int = yRef >> 2; 136| | 137| 191k| dst = cur + y * stride + x; 138| 191k| src = refn + y_int * (int)stride + x_int; 139| | 140| 191k| tmp = refh; /* we need at least a 16 x stride scratch block */ 141| | 142| 191k| switch(quads) { ------------------ | Branch (142:9): [True: 0, False: 191k] ------------------ 143| 71.6k| case 0: ------------------ | Branch (143:2): [True: 71.6k, False: 120k] ------------------ 144| 71.6k| transfer8x8_copy(dst, src, stride); 145| 71.6k| transfer8x8_copy(dst+8, src+8, stride); 146| 71.6k| transfer8x8_copy(dst+8*stride, src+8*stride, stride); 147| 71.6k| transfer8x8_copy(dst+8*stride+8, src+8*stride+8, stride); 148| 71.6k| break; 149| 13.1k| case 1: ------------------ | Branch (149:2): [True: 13.1k, False: 178k] ------------------ 150| 13.1k| Ops->H_Pass_Avrg(dst, src, 16, stride, rounding); 151| 13.1k| break; 152| 12.1k| case 2: ------------------ | Branch (152:2): [True: 12.1k, False: 179k] ------------------ 153| 12.1k| Ops->H_Pass(dst, src, 16, stride, rounding); 154| 12.1k| break; 155| 16.6k| case 3: ------------------ | Branch (155:2): [True: 16.6k, False: 175k] ------------------ 156| 16.6k| Ops->H_Pass_Avrg_Up(dst, src, 16, stride, rounding); 157| 16.6k| break; 158| 12.6k| case 4: ------------------ | Branch (158:2): [True: 12.6k, False: 179k] ------------------ 159| 12.6k| Ops->V_Pass_Avrg(dst, src, 16, stride, rounding); 160| 12.6k| break; 161| 4.76k| case 5: ------------------ | Branch (161:2): [True: 4.76k, False: 187k] ------------------ 162| 4.76k| Ops->H_Pass_Avrg(tmp, src, 17, stride, rounding); 163| 4.76k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 164| 4.76k| break; 165| 2.81k| case 6: ------------------ | Branch (165:2): [True: 2.81k, False: 188k] ------------------ 166| 2.81k| Ops->H_Pass(tmp, src, 17, stride, rounding); 167| 2.81k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 168| 2.81k| break; 169| 4.51k| case 7: ------------------ | Branch (169:2): [True: 4.51k, False: 187k] ------------------ 170| 4.51k| Ops->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 171| 4.51k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 172| 4.51k| break; 173| 12.9k| case 8: ------------------ | Branch (173:2): [True: 12.9k, False: 178k] ------------------ 174| 12.9k| Ops->V_Pass(dst, src, 16, stride, rounding); 175| 12.9k| break; 176| 3.48k| case 9: ------------------ | Branch (176:2): [True: 3.48k, False: 188k] ------------------ 177| 3.48k| Ops->H_Pass_Avrg(tmp, src, 17, stride, rounding); 178| 3.48k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 179| 3.48k| break; 180| 3.40k| case 10: ------------------ | Branch (180:2): [True: 3.40k, False: 188k] ------------------ 181| 3.40k| Ops->H_Pass(tmp, src, 17, stride, rounding); 182| 3.40k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 183| 3.40k| break; 184| 4.32k| case 11: ------------------ | Branch (184:2): [True: 4.32k, False: 187k] ------------------ 185| 4.32k| Ops->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 186| 4.32k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 187| 4.32k| break; 188| 15.6k| case 12: ------------------ | Branch (188:2): [True: 15.6k, False: 176k] ------------------ 189| 15.6k| Ops->V_Pass_Avrg_Up(dst, src, 16, stride, rounding); 190| 15.6k| break; 191| 4.36k| case 13: ------------------ | Branch (191:2): [True: 4.36k, False: 187k] ------------------ 192| 4.36k| Ops->H_Pass_Avrg(tmp, src, 17, stride, rounding); 193| 4.36k| Ops->V_Pass_Avrg_Up(dst, tmp, 16, stride, rounding); 194| 4.36k| break; 195| 3.57k| case 14: ------------------ | Branch (195:2): [True: 3.57k, False: 188k] ------------------ 196| 3.57k| Ops->H_Pass(tmp, src, 17, stride, rounding); 197| 3.57k| Ops->V_Pass_Avrg_Up( dst, tmp, 16, stride, rounding); 198| 3.57k| break; 199| 5.66k| case 15: ------------------ | Branch (199:2): [True: 5.66k, False: 186k] ------------------ 200| 5.66k| Ops->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 201| 5.66k| Ops->V_Pass_Avrg_Up(dst, tmp, 16, stride, rounding); 202| 5.66k| break; 203| 191k| } 204| 191k|} decoder.c:interpolate8x8_quarterpel: 412| 1.60M|{ 413| 1.60M| const uint8_t *src; 414| 1.60M| uint8_t *dst; 415| 1.60M| uint8_t *tmp; 416| 1.60M| int32_t quads; 417| 1.60M| const XVID_QP_FUNCS *Ops; 418| | 419| 1.60M| int32_t x_int, y_int; 420| | 421| 1.60M| const int32_t xRef = (int)x*4 + dx; 422| 1.60M| const int32_t yRef = (int)y*4 + dy; 423| | 424| 1.60M| Ops = xvid_QP_Funcs; 425| 1.60M| quads = (dx&3) | ((dy&3)<<2); 426| | 427| 1.60M| x_int = xRef >> 2; 428| 1.60M| y_int = yRef >> 2; 429| | 430| 1.60M| dst = cur + y * stride + x; 431| 1.60M| src = refn + y_int * (int)stride + x_int; 432| | 433| 1.60M| tmp = refh; /* we need at least a 16 x stride scratch block */ 434| | 435| 1.60M| switch(quads) { ------------------ | Branch (435:9): [True: 0, False: 1.60M] ------------------ 436| 1.32M| case 0: ------------------ | Branch (436:2): [True: 1.32M, False: 275k] ------------------ 437| 1.32M| transfer8x8_copy( dst, src, stride); 438| 1.32M| break; 439| 39.8k| case 1: ------------------ | Branch (439:2): [True: 39.8k, False: 1.56M] ------------------ 440| 39.8k| Ops->H_Pass_Avrg_8(dst, src, 8, stride, rounding); 441| 39.8k| break; 442| 26.1k| case 2: ------------------ | Branch (442:2): [True: 26.1k, False: 1.57M] ------------------ 443| 26.1k| Ops->H_Pass_8(dst, src, 8, stride, rounding); 444| 26.1k| break; 445| 35.1k| case 3: ------------------ | Branch (445:2): [True: 35.1k, False: 1.56M] ------------------ 446| 35.1k| Ops->H_Pass_Avrg_Up_8(dst, src, 8, stride, rounding); 447| 35.1k| break; 448| 23.6k| case 4: ------------------ | Branch (448:2): [True: 23.6k, False: 1.57M] ------------------ 449| 23.6k| Ops->V_Pass_Avrg_8(dst, src, 8, stride, rounding); 450| 23.6k| break; 451| 8.85k| case 5: ------------------ | Branch (451:2): [True: 8.85k, False: 1.59M] ------------------ 452| 8.85k| Ops->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 453| 8.85k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 454| 8.85k| break; 455| 4.01k| case 6: ------------------ | Branch (455:2): [True: 4.01k, False: 1.59M] ------------------ 456| 4.01k| Ops->H_Pass_8(tmp, src, 9, stride, rounding); 457| 4.01k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 458| 4.01k| break; 459| 17.3k| case 7: ------------------ | Branch (459:2): [True: 17.3k, False: 1.58M] ------------------ 460| 17.3k| Ops->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 461| 17.3k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 462| 17.3k| break; 463| 31.1k| case 8: ------------------ | Branch (463:2): [True: 31.1k, False: 1.57M] ------------------ 464| 31.1k| Ops->V_Pass_8(dst, src, 8, stride, rounding); 465| 31.1k| break; 466| 7.40k| case 9: ------------------ | Branch (466:2): [True: 7.40k, False: 1.59M] ------------------ 467| 7.40k| Ops->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 468| 7.40k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 469| 7.40k| break; 470| 6.90k| case 10: ------------------ | Branch (470:2): [True: 6.90k, False: 1.59M] ------------------ 471| 6.90k| Ops->H_Pass_8(tmp, src, 9, stride, rounding); 472| 6.90k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 473| 6.90k| break; 474| 7.27k| case 11: ------------------ | Branch (474:2): [True: 7.27k, False: 1.59M] ------------------ 475| 7.27k| Ops->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 476| 7.27k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 477| 7.27k| break; 478| 31.9k| case 12: ------------------ | Branch (478:2): [True: 31.9k, False: 1.57M] ------------------ 479| 31.9k| Ops->V_Pass_Avrg_Up_8(dst, src, 8, stride, rounding); 480| 31.9k| break; 481| 16.4k| case 13: ------------------ | Branch (481:2): [True: 16.4k, False: 1.58M] ------------------ 482| 16.4k| Ops->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 483| 16.4k| Ops->V_Pass_Avrg_Up_8(dst, tmp, 8, stride, rounding); 484| 16.4k| break; 485| 6.00k| case 14: ------------------ | Branch (485:2): [True: 6.00k, False: 1.59M] ------------------ 486| 6.00k| Ops->H_Pass_8(tmp, src, 9, stride, rounding); 487| 6.00k| Ops->V_Pass_Avrg_Up_8( dst, tmp, 8, stride, rounding); 488| 6.00k| break; 489| 13.3k| case 15: ------------------ | Branch (489:2): [True: 13.3k, False: 1.58M] ------------------ 490| 13.3k| Ops->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 491| 13.3k| Ops->V_Pass_Avrg_Up_8(dst, tmp, 8, stride, rounding); 492| 13.3k| break; 493| 1.60M| } 494| 1.60M|} decoder.c:interpolate16x16_add_quarterpel: 216| 95.6k|{ 217| 95.6k| const uint8_t *src; 218| 95.6k| uint8_t *dst; 219| 95.6k| uint8_t *tmp; 220| 95.6k| int32_t quads; 221| 95.6k| const XVID_QP_FUNCS *Ops; 222| 95.6k| const XVID_QP_FUNCS *Ops_Copy; 223| | 224| 95.6k| int32_t x_int, y_int; 225| | 226| 95.6k| const int32_t xRef = (int)x*4 + dx; 227| 95.6k| const int32_t yRef = (int)y*4 + dy; 228| | 229| 95.6k| Ops = xvid_QP_Add_Funcs; 230| 95.6k| Ops_Copy = xvid_QP_Funcs; 231| 95.6k| quads = (dx&3) | ((dy&3)<<2); 232| | 233| 95.6k| x_int = xRef >> 2; 234| 95.6k| y_int = yRef >> 2; 235| | 236| 95.6k| dst = cur + y * stride + x; 237| 95.6k| src = refn + y_int * (int)stride + x_int; 238| | 239| 95.6k| tmp = refh; /* we need at least a 16 x stride scratch block */ 240| | 241| 95.6k| switch(quads) { ------------------ | Branch (241:9): [True: 0, False: 95.6k] ------------------ 242| 31.9k| case 0: ------------------ | Branch (242:2): [True: 31.9k, False: 63.6k] ------------------ 243| | /* NB: there is no halfpel involved ! the name's function can be 244| | * misleading */ 245| 31.9k| interpolate8x8_halfpel_add(dst, src, stride, rounding); 246| 31.9k| interpolate8x8_halfpel_add(dst+8, src+8, stride, rounding); 247| 31.9k| interpolate8x8_halfpel_add(dst+8*stride, src+8*stride, stride, rounding); 248| 31.9k| interpolate8x8_halfpel_add(dst+8*stride+8, src+8*stride+8, stride, rounding); 249| 31.9k| break; 250| 9.27k| case 1: ------------------ | Branch (250:2): [True: 9.27k, False: 86.3k] ------------------ 251| 9.27k| Ops->H_Pass_Avrg(dst, src, 16, stride, rounding); 252| 9.27k| break; 253| 7.74k| case 2: ------------------ | Branch (253:2): [True: 7.74k, False: 87.8k] ------------------ 254| 7.74k| Ops->H_Pass(dst, src, 16, stride, rounding); 255| 7.74k| break; 256| 10.4k| case 3: ------------------ | Branch (256:2): [True: 10.4k, False: 85.1k] ------------------ 257| 10.4k| Ops->H_Pass_Avrg_Up(dst, src, 16, stride, rounding); 258| 10.4k| break; 259| 5.22k| case 4: ------------------ | Branch (259:2): [True: 5.22k, False: 90.3k] ------------------ 260| 5.22k| Ops->V_Pass_Avrg(dst, src, 16, stride, rounding); 261| 5.22k| break; 262| 2.32k| case 5: ------------------ | Branch (262:2): [True: 2.32k, False: 93.2k] ------------------ 263| 2.32k| Ops_Copy->H_Pass_Avrg(tmp, src, 17, stride, rounding); 264| 2.32k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 265| 2.32k| break; 266| 2.06k| case 6: ------------------ | Branch (266:2): [True: 2.06k, False: 93.5k] ------------------ 267| 2.06k| Ops_Copy->H_Pass(tmp, src, 17, stride, rounding); 268| 2.06k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 269| 2.06k| break; 270| 2.26k| case 7: ------------------ | Branch (270:2): [True: 2.26k, False: 93.3k] ------------------ 271| 2.26k| Ops_Copy->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 272| 2.26k| Ops->V_Pass_Avrg(dst, tmp, 16, stride, rounding); 273| 2.26k| break; 274| 4.13k| case 8: ------------------ | Branch (274:2): [True: 4.13k, False: 91.4k] ------------------ 275| 4.13k| Ops->V_Pass(dst, src, 16, stride, rounding); 276| 4.13k| break; 277| 1.81k| case 9: ------------------ | Branch (277:2): [True: 1.81k, False: 93.7k] ------------------ 278| 1.81k| Ops_Copy->H_Pass_Avrg(tmp, src, 17, stride, rounding); 279| 1.81k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 280| 1.81k| break; 281| 1.90k| case 10: ------------------ | Branch (281:2): [True: 1.90k, False: 93.7k] ------------------ 282| 1.90k| Ops_Copy->H_Pass(tmp, src, 17, stride, rounding); 283| 1.90k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 284| 1.90k| break; 285| 1.85k| case 11: ------------------ | Branch (285:2): [True: 1.85k, False: 93.7k] ------------------ 286| 1.85k| Ops_Copy->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 287| 1.85k| Ops->V_Pass(dst, tmp, 16, stride, rounding); 288| 1.85k| break; 289| 7.14k| case 12: ------------------ | Branch (289:2): [True: 7.14k, False: 88.4k] ------------------ 290| 7.14k| Ops->V_Pass_Avrg_Up(dst, src, 16, stride, rounding); 291| 7.14k| break; 292| 2.47k| case 13: ------------------ | Branch (292:2): [True: 2.47k, False: 93.1k] ------------------ 293| 2.47k| Ops_Copy->H_Pass_Avrg(tmp, src, 17, stride, rounding); 294| 2.47k| Ops->V_Pass_Avrg_Up(dst, tmp, 16, stride, rounding); 295| 2.47k| break; 296| 1.96k| case 14: ------------------ | Branch (296:2): [True: 1.96k, False: 93.6k] ------------------ 297| 1.96k| Ops_Copy->H_Pass(tmp, src, 17, stride, rounding); 298| 1.96k| Ops->V_Pass_Avrg_Up( dst, tmp, 16, stride, rounding); 299| 1.96k| break; 300| 2.96k| case 15: ------------------ | Branch (300:2): [True: 2.96k, False: 92.6k] ------------------ 301| 2.96k| Ops_Copy->H_Pass_Avrg_Up(tmp, src, 17, stride, rounding); 302| 2.96k| Ops->V_Pass_Avrg_Up(dst, tmp, 16, stride, rounding); 303| 2.96k| break; 304| 95.6k| } 305| 95.6k|} decoder.c:interpolate8x8_add_quarterpel: 506| 1.54M|{ 507| 1.54M| const uint8_t *src; 508| 1.54M| uint8_t *dst; 509| 1.54M| uint8_t *tmp; 510| 1.54M| int32_t quads; 511| 1.54M| const XVID_QP_FUNCS *Ops; 512| 1.54M| const XVID_QP_FUNCS *Ops_Copy; 513| | 514| 1.54M| int32_t x_int, y_int; 515| | 516| 1.54M| const int32_t xRef = (int)x*4 + dx; 517| 1.54M| const int32_t yRef = (int)y*4 + dy; 518| | 519| 1.54M| Ops = xvid_QP_Add_Funcs; 520| 1.54M| Ops_Copy = xvid_QP_Funcs; 521| 1.54M| quads = (dx&3) | ((dy&3)<<2); 522| | 523| 1.54M| x_int = xRef >> 2; 524| 1.54M| y_int = yRef >> 2; 525| | 526| 1.54M| dst = cur + y * stride + x; 527| 1.54M| src = refn + y_int * (int)stride + x_int; 528| | 529| 1.54M| tmp = refh; /* we need at least a 16 x stride scratch block */ 530| | 531| 1.54M| switch(quads) { ------------------ | Branch (531:9): [True: 0, False: 1.54M] ------------------ 532| 1.30M| case 0: ------------------ | Branch (532:2): [True: 1.30M, False: 238k] ------------------ 533| | /* Misleading function name, there is no halfpel involved 534| | * just dst and src averaging with rounding=0 */ 535| 1.30M| interpolate8x8_halfpel_add(dst, src, stride, rounding); 536| 1.30M| break; 537| 37.2k| case 1: ------------------ | Branch (537:2): [True: 37.2k, False: 1.51M] ------------------ 538| 37.2k| Ops->H_Pass_Avrg_8(dst, src, 8, stride, rounding); 539| 37.2k| break; 540| 23.4k| case 2: ------------------ | Branch (540:2): [True: 23.4k, False: 1.52M] ------------------ 541| 23.4k| Ops->H_Pass_8(dst, src, 8, stride, rounding); 542| 23.4k| break; 543| 33.1k| case 3: ------------------ | Branch (543:2): [True: 33.1k, False: 1.51M] ------------------ 544| 33.1k| Ops->H_Pass_Avrg_Up_8(dst, src, 8, stride, rounding); 545| 33.1k| break; 546| 21.7k| case 4: ------------------ | Branch (546:2): [True: 21.7k, False: 1.52M] ------------------ 547| 21.7k| Ops->V_Pass_Avrg_8(dst, src, 8, stride, rounding); 548| 21.7k| break; 549| 6.63k| case 5: ------------------ | Branch (549:2): [True: 6.63k, False: 1.54M] ------------------ 550| 6.63k| Ops_Copy->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 551| 6.63k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 552| 6.63k| break; 553| 3.20k| case 6: ------------------ | Branch (553:2): [True: 3.20k, False: 1.54M] ------------------ 554| 3.20k| Ops_Copy->H_Pass_8(tmp, src, 9, stride, rounding); 555| 3.20k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 556| 3.20k| break; 557| 15.9k| case 7: ------------------ | Branch (557:2): [True: 15.9k, False: 1.53M] ------------------ 558| 15.9k| Ops_Copy->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 559| 15.9k| Ops->V_Pass_Avrg_8(dst, tmp, 8, stride, rounding); 560| 15.9k| break; 561| 25.4k| case 8: ------------------ | Branch (561:2): [True: 25.4k, False: 1.52M] ------------------ 562| 25.4k| Ops->V_Pass_8(dst, src, 8, stride, rounding); 563| 25.4k| break; 564| 6.11k| case 9: ------------------ | Branch (564:2): [True: 6.11k, False: 1.54M] ------------------ 565| 6.11k| Ops_Copy->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 566| 6.11k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 567| 6.11k| break; 568| 4.20k| case 10: ------------------ | Branch (568:2): [True: 4.20k, False: 1.54M] ------------------ 569| 4.20k| Ops_Copy->H_Pass_8(tmp, src, 9, stride, rounding); 570| 4.20k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 571| 4.20k| break; 572| 5.98k| case 11: ------------------ | Branch (572:2): [True: 5.98k, False: 1.54M] ------------------ 573| 5.98k| Ops_Copy->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 574| 5.98k| Ops->V_Pass_8(dst, tmp, 8, stride, rounding); 575| 5.98k| break; 576| 25.0k| case 12: ------------------ | Branch (576:2): [True: 25.0k, False: 1.52M] ------------------ 577| 25.0k| Ops->V_Pass_Avrg_Up_8(dst, src, 8, stride, rounding); 578| 25.0k| break; 579| 14.0k| case 13: ------------------ | Branch (579:2): [True: 14.0k, False: 1.53M] ------------------ 580| 14.0k| Ops_Copy->H_Pass_Avrg_8(tmp, src, 9, stride, rounding); 581| 14.0k| Ops->V_Pass_Avrg_Up_8(dst, tmp, 8, stride, rounding); 582| 14.0k| break; 583| 4.88k| case 14: ------------------ | Branch (583:2): [True: 4.88k, False: 1.54M] ------------------ 584| 4.88k| Ops_Copy->H_Pass_8(tmp, src, 9, stride, rounding); 585| 4.88k| Ops->V_Pass_Avrg_Up_8( dst, tmp, 8, stride, rounding); 586| 4.88k| break; 587| 11.5k| case 15: ------------------ | Branch (587:2): [True: 11.5k, False: 1.53M] ------------------ 588| 11.5k| Ops_Copy->H_Pass_Avrg_Up_8(tmp, src, 9, stride, rounding); 589| 11.5k| Ops->V_Pass_Avrg_Up_8(dst, tmp, 8, stride, rounding); 590| 11.5k| break; 591| 1.54M| } 592| 1.54M|} init_GMC: 589| 2|{ 590| 2| Predict_16x16_func = Predict_16x16_C; 591| 2| Predict_8x8_func = Predict_8x8_C; 592| | 593| |#if defined(ARCH_IS_IA32) || defined(ARCH_IS_X86_64) 594| | if ((cpu_flags & XVID_CPU_MMX) || (cpu_flags & XVID_CPU_MMXEXT) || 595| | (cpu_flags & XVID_CPU_3DNOW) || (cpu_flags & XVID_CPU_3DNOWEXT) || 596| | (cpu_flags & XVID_CPU_SSE) || (cpu_flags & XVID_CPU_SSE2) || 597| | (cpu_flags & XVID_CPU_SSE3) || (cpu_flags & XVID_CPU_SSE41)) 598| | { 599| | Predict_16x16_func = Predict_16x16_mmx; 600| | Predict_8x8_func = Predict_8x8_mmx; 601| | 602| | if (cpu_flags & XVID_CPU_SSE41) 603| | GMC_Core_Lin_8 = xvid_GMC_Core_Lin_8_sse41; 604| | else if (cpu_flags & XVID_CPU_SSE2) 605| | GMC_Core_Lin_8 = xvid_GMC_Core_Lin_8_sse2; 606| | else 607| | GMC_Core_Lin_8 = xvid_GMC_Core_Lin_8_mmx; 608| | } 609| |#endif 610| 2|} generate_GMCparameters: 620| 2.47k|{ 621| 2.47k| gmc->sW = width << 4; 622| 2.47k| gmc->sH = height << 4; 623| 2.47k| gmc->accuracy = accuracy; 624| 2.47k| gmc->num_wp = nb_pts; 625| | 626| | /* reduce the number of points, if possible */ 627| 2.47k| if (nb_pts<2 || (pts->duv[2].x==0 && pts->duv[2].y==0 && pts->duv[1].x==0 && pts->duv[1].y==0 )) { ------------------ | Branch (627:6): [True: 1.48k, False: 983] | Branch (627:19): [True: 640, False: 343] | Branch (627:39): [True: 573, False: 67] | Branch (627:59): [True: 333, False: 240] | Branch (627:79): [True: 260, False: 73] ------------------ 628| 1.74k| if (nb_pts<2 || (pts->duv[1].x==0 && pts->duv[1].y==0)) { ------------------ | Branch (628:8): [True: 1.48k, False: 260] | Branch (628:21): [True: 260, False: 0] | Branch (628:41): [True: 260, False: 0] ------------------ 629| 1.74k| if (nb_pts<1 || (pts->duv[0].x==0 && pts->duv[0].y==0)) { ------------------ | Branch (629:9): [True: 1.30k, False: 444] | Branch (629:22): [True: 184, False: 260] | Branch (629:42): [True: 131, False: 53] ------------------ 630| 1.43k| nb_pts = 0; 631| 1.43k| } 632| 313| else nb_pts = 1; 633| 1.74k| } 634| 0| else nb_pts = 2; 635| 1.74k| } 636| | 637| | /* now, nb_pts stores the actual number of points required for interpolation */ 638| | 639| 2.47k| if (nb_pts<=1) ------------------ | Branch (639:6): [True: 1.74k, False: 723] ------------------ 640| 1.74k| { 641| 1.74k| if (nb_pts==1) { ------------------ | Branch (641:6): [True: 313, False: 1.43k] ------------------ 642| | /* store as 4b fixed point */ 643| 313| gmc->Uo = pts->duv[0].x << accuracy; 644| 313| gmc->Vo = pts->duv[0].y << accuracy; 645| 313| gmc->Uco = ((pts->duv[0].x>>1) | (pts->duv[0].x&1)) << accuracy; /* DIV2RND() */ 646| 313| gmc->Vco = ((pts->duv[0].y>>1) | (pts->duv[0].y&1)) << accuracy; /* DIV2RND() */ 647| 313| } 648| 1.43k| else { /* zero points?! */ 649| 1.43k| gmc->Uo = gmc->Vo = 0; 650| 1.43k| gmc->Uco = gmc->Vco = 0; 651| 1.43k| } 652| | 653| 1.74k| gmc->predict_16x16 = Predict_1pt_16x16_C; 654| 1.74k| gmc->predict_8x8 = Predict_1pt_8x8_C; 655| 1.74k| gmc->get_average_mv = get_average_mv_1pt_C; 656| 1.74k| } 657| 723| else { /* 2 or 3 points */ 658| 723| const int rho = 3 - accuracy; /* = {3,2,1,0} for Acc={0,1,2,3} */ 659| 723| int Alpha = log2bin(width-1); 660| 723| int Ws = 1 << Alpha; 661| | 662| 723| gmc->dU[0] = 16*Ws + RDIV( 8*Ws*pts->duv[1].x, width ); /* dU/dx */ ------------------ | | 76| 723|#define RDIV(a,b) (((a)>0 ? (a) + ((b)>>1) : (a) - ((b)>>1))/(b)) | | ------------------ | | | Branch (76:21): [True: 248, False: 475] | | ------------------ ------------------ 663| 723| gmc->dV[0] = RDIV( 8*Ws*pts->duv[1].y, width ); /* dV/dx */ ------------------ | | 76| 723|#define RDIV(a,b) (((a)>0 ? (a) + ((b)>>1) : (a) - ((b)>>1))/(b)) | | ------------------ | | | Branch (76:21): [True: 313, False: 410] | | ------------------ ------------------ 664| | 665| 723| if (nb_pts==2) { ------------------ | Branch (665:6): [True: 280, False: 443] ------------------ 666| 280| gmc->dU[1] = -gmc->dV[0]; /* -Sin */ 667| 280| gmc->dV[1] = gmc->dU[0] ; /* Cos */ 668| 280| } 669| 443| else 670| 443| { 671| 443| const int Beta = log2bin(height-1); 672| 443| const int Hs = 1<dU[1] = RDIV( 8*Hs*pts->duv[2].x, height ); /* dU/dy */ ------------------ | | 76| 443|#define RDIV(a,b) (((a)>0 ? (a) + ((b)>>1) : (a) - ((b)>>1))/(b)) | | ------------------ | | | Branch (76:21): [True: 200, False: 243] | | ------------------ ------------------ 674| 443| gmc->dV[1] = 16*Hs + RDIV( 8*Hs*pts->duv[2].y, height ); /* dV/dy */ ------------------ | | 76| 443|#define RDIV(a,b) (((a)>0 ? (a) + ((b)>>1) : (a) - ((b)>>1))/(b)) | | ------------------ | | | Branch (76:21): [True: 245, False: 198] | | ------------------ ------------------ 675| 443| if (Beta>Alpha) { ------------------ | Branch (675:7): [True: 189, False: 254] ------------------ 676| 189| gmc->dU[0] <<= (Beta-Alpha); 677| 189| gmc->dV[0] <<= (Beta-Alpha); 678| 189| Alpha = Beta; 679| 189| Ws = Hs; 680| 189| } 681| 254| else { 682| 254| gmc->dU[1] <<= Alpha - Beta; 683| 254| gmc->dV[1] <<= Alpha - Beta; 684| 254| } 685| 443| } 686| | /* upscale to 16b fixed-point */ 687| 723| gmc->dU[0] <<= (16-Alpha - rho); 688| 723| gmc->dU[1] <<= (16-Alpha - rho); 689| 723| gmc->dV[0] <<= (16-Alpha - rho); 690| 723| gmc->dV[1] <<= (16-Alpha - rho); 691| | 692| 723| gmc->Uo = ( pts->duv[0].x <<(16+ accuracy)) + (1<<15); 693| 723| gmc->Vo = ( pts->duv[0].y <<(16+ accuracy)) + (1<<15); 694| 723| gmc->Uco = ((pts->duv[0].x-1)<<(17+ accuracy)) + (1<<17); 695| 723| gmc->Vco = ((pts->duv[0].y-1)<<(17+ accuracy)) + (1<<17); 696| 723| gmc->Uco = (gmc->Uco + gmc->dU[0] + gmc->dU[1])>>2; 697| 723| gmc->Vco = (gmc->Vco + gmc->dV[0] + gmc->dV[1])>>2; 698| | 699| 723| gmc->predict_16x16 = Predict_16x16_func; 700| 723| gmc->predict_8x8 = Predict_8x8_func; 701| 723| gmc->get_average_mv = get_average_mv_C; 702| 723| } 703| 2.47k|} gmc.c:Predict_16x16_C: 98| 161k|{ 99| 161k| const int W = This->sW; 100| 161k| const int H = This->sH; 101| 161k| const int rho = 3 - This->accuracy; 102| 161k| const int Rounder = ( (1<<7) - (rounding<<(2*rho)) ) << 16; 103| | 104| 161k| const int dUx = This->dU[0]; 105| 161k| const int dVx = This->dV[0]; 106| 161k| const int dUy = This->dU[1]; 107| 161k| const int dVy = This->dV[1]; 108| | 109| 161k| int Uo = This->Uo + 16*(dUy*y + dUx*x); 110| 161k| int Vo = This->Vo + 16*(dVy*y + dVx*x); 111| | 112| 161k| int i, j; 113| | 114| 161k| dst += 16; 115| 2.74M| for (j=16; j>0; --j) { ------------------ | Branch (115:13): [True: 2.58M, False: 161k] ------------------ 116| 2.58M| int U = Uo, V = Vo; 117| 2.58M| Uo += dUy; Vo += dVy; 118| 43.8M| for (i=-16; i<0; ++i) { ------------------ | Branch (118:15): [True: 41.2M, False: 2.58M] ------------------ 119| 41.2M| unsigned int f0, f1, ri = 16, rj = 16; 120| 41.2M| int Offset; 121| 41.2M| int u = ( U >> 16 ) << rho; 122| 41.2M| int v = ( V >> 16 ) << rho; 123| | 124| 41.2M| U += dUx; V += dVx; 125| | 126| 41.2M| if (u > 0 && u <= W) { ri = MTab[u&15]; Offset = u>>4; } ------------------ | Branch (126:8): [True: 33.4M, False: 7.81M] | Branch (126:17): [True: 31.2M, False: 2.17M] ------------------ 127| 9.98M| else { 128| 9.98M| if (u > W) Offset = W>>4; ------------------ | Branch (128:9): [True: 2.17M, False: 7.81M] ------------------ 129| 7.81M| else Offset = 0; 130| 9.98M| ri = MTab[0]; 131| 9.98M| } 132| | 133| 41.2M| if (v > 0 && v <= H) { rj = MTab[v&15]; Offset += (v>>4)*srcstride; } ------------------ | Branch (133:8): [True: 37.5M, False: 3.78M] | Branch (133:17): [True: 37.3M, False: 119k] ------------------ 134| 3.89M| else { 135| 3.89M| if (v > H) Offset += (H>>4)*srcstride; ------------------ | Branch (135:9): [True: 119k, False: 3.78M] ------------------ 136| 3.89M| rj = MTab[0]; 137| 3.89M| } 138| | 139| 41.2M| f0 = src[Offset + 0]; 140| 41.2M| f0 |= src[Offset + 1] << 16; 141| 41.2M| f1 = src[Offset + srcstride + 0]; 142| 41.2M| f1 |= src[Offset + srcstride + 1] << 16; 143| 41.2M| f0 = (ri*f0)>>16; 144| 41.2M| f1 = (ri*f1) & 0x0fff0000; 145| 41.2M| f0 |= f1; 146| 41.2M| f0 = (rj*f0 + Rounder) >> 24; 147| | 148| 41.2M| dst[i] = (uint8_t)f0; 149| 41.2M| } 150| 2.58M| dst += dststride; 151| 2.58M| } 152| 161k|} gmc.c:Predict_8x8_C: 159| 161k|{ 160| 161k| const int W = This->sW >> 1; 161| 161k| const int H = This->sH >> 1; 162| 161k| const int rho = 3-This->accuracy; 163| 161k| const int32_t Rounder = ( 128 - (rounding<<(2*rho)) ) << 16; 164| | 165| 161k| const int32_t dUx = This->dU[0]; 166| 161k| const int32_t dVx = This->dV[0]; 167| 161k| const int32_t dUy = This->dU[1]; 168| 161k| const int32_t dVy = This->dV[1]; 169| | 170| 161k| int32_t Uo = This->Uco + 8*(dUy*y + dUx*x); 171| 161k| int32_t Vo = This->Vco + 8*(dVy*y + dVx*x); 172| | 173| 161k| int i, j; 174| | 175| 161k| uDst += 8; 176| 161k| vDst += 8; 177| 1.45M| for (j=8; j>0; --j) { ------------------ | Branch (177:12): [True: 1.29M, False: 161k] ------------------ 178| 1.29M| int32_t U = Uo, V = Vo; 179| 1.29M| Uo += dUy; Vo += dVy; 180| | 181| 11.6M| for (i=-8; i<0; ++i) { ------------------ | Branch (181:14): [True: 10.3M, False: 1.29M] ------------------ 182| 10.3M| int Offset; 183| 10.3M| uint32_t f0, f1, ri, rj; 184| 10.3M| int32_t u, v; 185| | 186| 10.3M| u = ( U >> 16 ) << rho; 187| 10.3M| v = ( V >> 16 ) << rho; 188| 10.3M| U += dUx; V += dVx; 189| | 190| 10.3M| if (u > 0 && u <= W) { ------------------ | Branch (190:8): [True: 8.37M, False: 1.94M] | Branch (190:17): [True: 7.90M, False: 472k] ------------------ 191| 7.90M| ri = MTab[u&15]; 192| 7.90M| Offset = u>>4; 193| 7.90M| } else { 194| 2.41M| if (u>W) Offset = W>>4; ------------------ | Branch (194:9): [True: 472k, False: 1.94M] ------------------ 195| 1.94M| else Offset = 0; 196| 2.41M| ri = MTab[0]; 197| 2.41M| } 198| | 199| 10.3M| if (v > 0 && v <= H) { ------------------ | Branch (199:8): [True: 8.64M, False: 1.67M] | Branch (199:17): [True: 8.62M, False: 25.0k] ------------------ 200| 8.62M| rj = MTab[v&15]; 201| 8.62M| Offset += (v>>4)*srcstride; 202| 8.62M| } else { 203| 1.70M| if (v>H) Offset += (H>>4)*srcstride; ------------------ | Branch (203:9): [True: 25.0k, False: 1.67M] ------------------ 204| 1.70M| rj = MTab[0]; 205| 1.70M| } 206| | 207| 10.3M| f0 = uSrc[Offset + 0]; 208| 10.3M| f0 |= uSrc[Offset + 1] << 16; 209| 10.3M| f1 = uSrc[Offset + srcstride + 0]; 210| 10.3M| f1 |= uSrc[Offset + srcstride + 1] << 16; 211| 10.3M| f0 = (ri*f0)>>16; 212| 10.3M| f1 = (ri*f1) & 0x0fff0000; 213| 10.3M| f0 |= f1; 214| 10.3M| f0 = (rj*f0 + Rounder) >> 24; 215| | 216| 10.3M| uDst[i] = (uint8_t)f0; 217| | 218| 10.3M| f0 = vSrc[Offset + 0]; 219| 10.3M| f0 |= vSrc[Offset + 1] << 16; 220| 10.3M| f1 = vSrc[Offset + srcstride + 0]; 221| 10.3M| f1 |= vSrc[Offset + srcstride + 1] << 16; 222| 10.3M| f0 = (ri*f0)>>16; 223| 10.3M| f1 = (ri*f1) & 0x0fff0000; 224| 10.3M| f0 |= f1; 225| 10.3M| f0 = (rj*f0 + Rounder) >> 24; 226| | 227| 10.3M| vDst[i] = (uint8_t)f0; 228| 10.3M| } 229| 1.29M| uDst += dststride; 230| 1.29M| vDst += dststride; 231| 1.29M| } 232| 161k|} gmc.c:Predict_1pt_16x16_C: 269| 220k|{ 270| 220k| const int W = This->sW; 271| 220k| const int H = This->sH; 272| 220k| const int rho = 3-MIN(This->accuracy, 3); ------------------ | | 255| 220k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (255:20): [True: 145k, False: 75.3k] | | ------------------ ------------------ 273| 220k| const int32_t Rounder = ( 128 - (rounding<<(2*rho)) ) << 16; 274| | 275| | 276| 220k| int32_t uo = This->Uo + (x<<8); /* ((16*x)<<4) */ 277| 220k| int32_t vo = This->Vo + (y<<8); 278| 220k| uint32_t ri = MTab[uo & 15]; 279| 220k| uint32_t rj = MTab[vo & 15]; 280| 220k| int i, j; 281| | 282| 220k| int32_t Offset; 283| 220k| if (vo>=(-16<<4) && vo<=H) Offset = (vo>>4)*srcstride; ------------------ | Branch (283:6): [True: 219k, False: 803] | Branch (283:22): [True: 219k, False: 679] ------------------ 284| 1.48k| else { 285| 1.48k| if (vo>H) Offset = ( H>>4)*srcstride; ------------------ | Branch (285:7): [True: 679, False: 803] ------------------ 286| 803| else Offset =-16*srcstride; 287| 1.48k| rj = MTab[0]; 288| 1.48k| } 289| 220k| if (uo>=(-16<<4) && uo<=W) Offset += (uo>>4); ------------------ | Branch (289:6): [True: 218k, False: 1.98k] | Branch (289:22): [True: 195k, False: 23.1k] ------------------ 290| 25.1k| else { 291| 25.1k| if (uo>W) Offset += (W>>4); ------------------ | Branch (291:7): [True: 23.1k, False: 1.98k] ------------------ 292| 1.98k| else Offset -= 16; 293| 25.1k| ri = MTab[0]; 294| 25.1k| } 295| | 296| 220k| Dst += 16; 297| | 298| 3.74M| for(j=16; j>0; --j, Offset+=srcstride-16) ------------------ | Branch (298:12): [True: 3.52M, False: 220k] ------------------ 299| 3.52M| { 300| 59.9M| for(i=-16; i<0; ++i, ++Offset) ------------------ | Branch (300:13): [True: 56.4M, False: 3.52M] ------------------ 301| 56.4M| { 302| 56.4M| uint32_t f0, f1; 303| 56.4M| f0 = Src[ Offset +0 ]; 304| 56.4M| f0 |= Src[ Offset +1 ] << 16; 305| 56.4M| f1 = Src[ Offset+srcstride +0 ]; 306| 56.4M| f1 |= Src[ Offset+srcstride +1 ] << 16; 307| 56.4M| f0 = (ri*f0)>>16; 308| 56.4M| f1 = (ri*f1) & 0x0fff0000; 309| 56.4M| f0 |= f1; 310| 56.4M| f0 = ( rj*f0 + Rounder ) >> 24; 311| 56.4M| Dst[i] = (uint8_t)f0; 312| 56.4M| } 313| 3.52M| Dst += dststride; 314| 3.52M| } 315| 220k|} gmc.c:Predict_1pt_8x8_C: 322| 220k|{ 323| 220k| const int W = This->sW >> 1; 324| 220k| const int H = This->sH >> 1; 325| 220k| const int rho = 3-This->accuracy; 326| 220k| const int32_t Rounder = ( 128 - (rounding<<(2*rho)) ) << 16; 327| | 328| 220k| int32_t uo = This->Uco + (x<<7); 329| 220k| int32_t vo = This->Vco + (y<<7); 330| 220k| uint32_t rri = MTab[uo & 15]; 331| 220k| uint32_t rrj = MTab[vo & 15]; 332| 220k| int i, j; 333| | 334| 220k| int32_t Offset; 335| 220k| if (vo>=(-8<<4) && vo<=H) Offset = (vo>>4)*srcstride; ------------------ | Branch (335:6): [True: 219k, False: 606] | Branch (335:21): [True: 218k, False: 1.15k] ------------------ 336| 1.76k| else { 337| 1.76k| if (vo>H) Offset = ( H>>4)*srcstride; ------------------ | Branch (337:7): [True: 1.15k, False: 606] ------------------ 338| 606| else Offset =-8*srcstride; 339| 1.76k| rrj = MTab[0]; 340| 1.76k| } 341| 220k| if (uo>=(-8<<4) && uo<=W) Offset += (uo>>4); ------------------ | Branch (341:6): [True: 218k, False: 2.05k] | Branch (341:21): [True: 195k, False: 23.3k] ------------------ 342| 25.3k| else { 343| 25.3k| if (uo>W) Offset += ( W>>4); ------------------ | Branch (343:7): [True: 23.3k, False: 2.05k] ------------------ 344| 2.05k| else Offset -= 8; 345| 25.3k| rri = MTab[0]; 346| 25.3k| } 347| | 348| 220k| uDst += 8; 349| 220k| vDst += 8; 350| 1.98M| for(j=8; j>0; --j, Offset+=srcstride-8) ------------------ | Branch (350:11): [True: 1.76M, False: 220k] ------------------ 351| 1.76M| { 352| 15.8M| for(i=-8; i<0; ++i, Offset++) ------------------ | Branch (352:12): [True: 14.1M, False: 1.76M] ------------------ 353| 14.1M| { 354| 14.1M| uint32_t f0, f1; 355| 14.1M| f0 = uSrc[ Offset + 0 ]; 356| 14.1M| f0 |= uSrc[ Offset + 1 ] << 16; 357| 14.1M| f1 = uSrc[ Offset + srcstride + 0 ]; 358| 14.1M| f1 |= uSrc[ Offset + srcstride + 1 ] << 16; 359| 14.1M| f0 = (rri*f0)>>16; 360| 14.1M| f1 = (rri*f1) & 0x0fff0000; 361| 14.1M| f0 |= f1; 362| 14.1M| f0 = ( rrj*f0 + Rounder ) >> 24; 363| 14.1M| uDst[i] = (uint8_t)f0; 364| | 365| 14.1M| f0 = vSrc[ Offset + 0 ]; 366| 14.1M| f0 |= vSrc[ Offset + 1 ] << 16; 367| 14.1M| f1 = vSrc[ Offset + srcstride + 0 ]; 368| 14.1M| f1 |= vSrc[ Offset + srcstride + 1 ] << 16; 369| 14.1M| f0 = (rri*f0)>>16; 370| 14.1M| f1 = (rri*f1) & 0x0fff0000; 371| 14.1M| f0 |= f1; 372| 14.1M| f0 = ( rrj*f0 + Rounder ) >> 24; 373| 14.1M| vDst[i] = (uint8_t)f0; 374| 14.1M| } 375| 1.76M| uDst += dststride; 376| 1.76M| vDst += dststride; 377| 1.76M| } 378| 220k|} gmc.c:get_average_mv_1pt_C: 383| 220k|{ 384| 220k| mv->x = RSHIFT(Dsp->Uo<0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b)) | | ------------------ | | | Branch (77:23): [True: 25.2k, False: 195k] | | ------------------ ------------------ 385| 220k| mv->y = RSHIFT(Dsp->Vo<0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b)) | | ------------------ | | | Branch (77:23): [True: 3.63k, False: 216k] | | ------------------ ------------------ 386| 220k|} gmc.c:log2bin: 49| 1.16k|{ 50| |/* Changed by Chenm001 */ 51| 1.16k|#if !defined(_MSC_VER) || defined(ARCH_IS_X86_64) 52| 1.16k| int n = 0; 53| | 54| 9.30k| while (value) { ------------------ | Branch (54:10): [True: 8.13k, False: 1.16k] ------------------ 55| 8.13k| value >>= 1; 56| 8.13k| n++; 57| 8.13k| } 58| 1.16k| return n; 59| |#else 60| | __asm { 61| | bsr eax, value 62| | inc eax 63| | } 64| |#endif 65| 1.16k|} gmc.c:get_average_mv_C: 237| 161k|{ 238| 161k| int i, j; 239| 161k| int vx = 0, vy = 0; 240| 161k| int32_t uo = Dsp->Uo + 16*(Dsp->dU[1]*y + Dsp->dU[0]*x); 241| 161k| int32_t vo = Dsp->Vo + 16*(Dsp->dV[1]*y + Dsp->dV[0]*x); 242| 2.74M| for (j=16; j>0; --j) ------------------ | Branch (242:13): [True: 2.58M, False: 161k] ------------------ 243| 2.58M| { 244| 2.58M| int32_t U, V; 245| 2.58M| U = uo; uo += Dsp->dU[1]; 246| 2.58M| V = vo; vo += Dsp->dV[1]; 247| 43.8M| for (i=16; i>0; --i) ------------------ | Branch (247:13): [True: 41.2M, False: 2.58M] ------------------ 248| 41.2M| { 249| 41.2M| int32_t u,v; 250| 41.2M| u = U >> 16; U += Dsp->dU[0]; vx += u; 251| 41.2M| v = V >> 16; V += Dsp->dV[0]; vy += v; 252| 41.2M| } 253| 2.58M| } 254| 161k| vx -= (256*x+120) << (5+Dsp->accuracy); /* 120 = 15*16/2 */ 255| 161k| vy -= (256*y+120) << (5+Dsp->accuracy); 256| | 257| 161k| mv->x = RSHIFT( vx, 8+Dsp->accuracy - qpel ); ------------------ | | 77| 161k|#define RSHIFT(a,b) ( (a)>0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b)) | | ------------------ | | | Branch (77:23): [True: 19.1k, False: 142k] | | ------------------ ------------------ 258| 161k| mv->y = RSHIFT( vy, 8+Dsp->accuracy - qpel ); ------------------ | | 77| 161k|#define RSHIFT(a,b) ( (a)>0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b)) | | ------------------ | | | Branch (77:23): [True: 16.9k, False: 144k] | | ------------------ ------------------ 259| 161k|} decoder.c:gmc_sanitize: 68| 763k|{ 69| 763k| int length = 1 << (fcode+4); 70| | 71| |#if 0 72| | if (quarterpel) value *= 2; 73| |#endif 74| | 75| 763k| if (value < -length) ------------------ | Branch (75:6): [True: 77.7k, False: 685k] ------------------ 76| 77.7k| return -length; 77| 685k| else if (value >= length) ------------------ | Branch (77:11): [True: 39.4k, False: 646k] ------------------ 78| 39.4k| return length-1; 79| 646k| else return value; 80| 763k|} mbcoding.c:DPRINTF: 282| 275M|static __inline void DPRINTF(int level, char *format, ...) {} decoder.c:DPRINTF: 282| 278M|static __inline void DPRINTF(int level, char *format, ...) {} bitstream.c:DPRINTF: 282| 25.3M|static __inline void DPRINTF(int level, char *format, ...) {} mbprediction.c:DPRINTF: 282| 275M|static __inline void DPRINTF(int level, char *format, ...) {} predict_acdc: 72| 272M|{ 73| 272M| const int mbpos = (y * mb_width) + x; 74| 272M| int16_t *left, *top, *diag, *current; 75| | 76| 272M| int32_t left_quant = current_quant; 77| 272M| int32_t top_quant = current_quant; 78| | 79| 272M| const int16_t *pLeft = default_acdc_values; 80| 272M| const int16_t *pTop = default_acdc_values; 81| 272M| const int16_t *pDiag = default_acdc_values; 82| | 83| 272M| uint32_t index = x + y * mb_width; /* current macroblock */ 84| 272M| int *acpred_direction = &pMBs[index].acpred_directions[block]; 85| 272M| uint32_t i; 86| | 87| 272M| left = top = diag = current = NULL; 88| | 89| | /* grab left,top and diag macroblocks */ 90| | 91| | /* left macroblock */ 92| | 93| 272M| if (x && mbpos >= bound + 1 && ------------------ | Branch (93:6): [True: 269M, False: 3.27M] | Branch (93:11): [True: 269M, False: 283k] ------------------ 94| 272M| (pMBs[index - 1].mode == MODE_INTRA || ------------------ | | 37| 538M|#define MODE_INTRA 3 ------------------ | Branch (94:4): [True: 394k, False: 268M] ------------------ 95| 269M| pMBs[index - 1].mode == MODE_INTRA_Q)) { ------------------ | | 38| 268M|#define MODE_INTRA_Q 4 ------------------ | Branch (95:4): [True: 86.7k, False: 268M] ------------------ 96| | 97| 481k| left = (int16_t*)pMBs[index - 1].pred_values[0]; 98| 481k| left_quant = pMBs[index - 1].quant; 99| 481k| } 100| | /* top macroblock */ 101| | 102| 272M| if (mbpos >= bound + (int)mb_width && ------------------ | Branch (102:6): [True: 268M, False: 4.00M] ------------------ 103| 272M| (pMBs[index - mb_width].mode == MODE_INTRA || ------------------ | | 37| 537M|#define MODE_INTRA 3 ------------------ | Branch (103:4): [True: 304k, False: 268M] ------------------ 104| 268M| pMBs[index - mb_width].mode == MODE_INTRA_Q)) { ------------------ | | 38| 268M|#define MODE_INTRA_Q 4 ------------------ | Branch (104:4): [True: 54.3k, False: 268M] ------------------ 105| | 106| 359k| top = (int16_t*)pMBs[index - mb_width].pred_values[0]; 107| 359k| top_quant = pMBs[index - mb_width].quant; 108| 359k| } 109| | /* diag macroblock */ 110| | 111| 272M| if (x && mbpos >= bound + (int)mb_width + 1 && ------------------ | Branch (111:6): [True: 269M, False: 3.27M] | Branch (111:11): [True: 265M, False: 3.94M] ------------------ 112| 272M| (pMBs[index - 1 - mb_width].mode == MODE_INTRA || ------------------ | | 37| 530M|#define MODE_INTRA 3 ------------------ | Branch (112:4): [True: 269k, False: 265M] ------------------ 113| 265M| pMBs[index - 1 - mb_width].mode == MODE_INTRA_Q)) { ------------------ | | 38| 265M|#define MODE_INTRA_Q 4 ------------------ | Branch (113:4): [True: 40.0k, False: 265M] ------------------ 114| | 115| 309k| diag = (int16_t*)pMBs[index - 1 - mb_width].pred_values[0]; 116| 309k| } 117| | 118| 272M| current = (int16_t*)pMBs[index].pred_values[0]; 119| | 120| | /* now grab pLeft, pTop, pDiag _blocks_ */ 121| | 122| 272M| switch (block) { ------------------ | Branch (122:10): [True: 0, False: 272M] ------------------ 123| | 124| 45.4M| case 0: ------------------ | Branch (124:2): [True: 45.4M, False: 227M] ------------------ 125| 45.4M| if (left) ------------------ | Branch (125:7): [True: 80.2k, False: 45.3M] ------------------ 126| 80.2k| pLeft = left + MBPRED_SIZE; ------------------ | | 180| 80.2k|#define MBPRED_SIZE 15 ------------------ 127| | 128| 45.4M| if (top) ------------------ | Branch (128:7): [True: 59.8k, False: 45.3M] ------------------ 129| 59.8k| pTop = top + (MBPRED_SIZE << 1); ------------------ | | 180| 59.8k|#define MBPRED_SIZE 15 ------------------ 130| | 131| 45.4M| if (diag) ------------------ | Branch (131:7): [True: 51.6k, False: 45.3M] ------------------ 132| 51.6k| pDiag = diag + 3 * MBPRED_SIZE; ------------------ | | 180| 51.6k|#define MBPRED_SIZE 15 ------------------ 133| | 134| 45.4M| break; 135| | 136| 45.4M| case 1: ------------------ | Branch (136:2): [True: 45.4M, False: 227M] ------------------ 137| 45.4M| pLeft = current; 138| 45.4M| left_quant = current_quant; 139| | 140| 45.4M| if (top) { ------------------ | Branch (140:7): [True: 59.8k, False: 45.3M] ------------------ 141| 59.8k| pTop = top + 3 * MBPRED_SIZE; ------------------ | | 180| 59.8k|#define MBPRED_SIZE 15 ------------------ 142| 59.8k| pDiag = top + (MBPRED_SIZE << 1); ------------------ | | 180| 59.8k|#define MBPRED_SIZE 15 ------------------ 143| 59.8k| } 144| 45.4M| break; 145| | 146| 45.4M| case 2: ------------------ | Branch (146:2): [True: 45.4M, False: 227M] ------------------ 147| 45.4M| if (left) { ------------------ | Branch (147:7): [True: 80.2k, False: 45.3M] ------------------ 148| 80.2k| pLeft = left + 3 * MBPRED_SIZE; ------------------ | | 180| 80.2k|#define MBPRED_SIZE 15 ------------------ 149| 80.2k| pDiag = left + MBPRED_SIZE; ------------------ | | 180| 80.2k|#define MBPRED_SIZE 15 ------------------ 150| 80.2k| } 151| | 152| 45.4M| pTop = current; 153| 45.4M| top_quant = current_quant; 154| | 155| 45.4M| break; 156| | 157| 45.4M| case 3: ------------------ | Branch (157:2): [True: 45.4M, False: 227M] ------------------ 158| 45.4M| pLeft = current + (MBPRED_SIZE << 1); ------------------ | | 180| 45.4M|#define MBPRED_SIZE 15 ------------------ 159| 45.4M| left_quant = current_quant; 160| | 161| 45.4M| pTop = current + MBPRED_SIZE; ------------------ | | 180| 45.4M|#define MBPRED_SIZE 15 ------------------ 162| 45.4M| top_quant = current_quant; 163| | 164| 45.4M| pDiag = current; 165| | 166| 45.4M| break; 167| | 168| 45.4M| case 4: ------------------ | Branch (168:2): [True: 45.4M, False: 227M] ------------------ 169| 45.4M| if (left) ------------------ | Branch (169:7): [True: 80.2k, False: 45.3M] ------------------ 170| 80.2k| pLeft = left + (MBPRED_SIZE << 2); ------------------ | | 180| 80.2k|#define MBPRED_SIZE 15 ------------------ 171| 45.4M| if (top) ------------------ | Branch (171:7): [True: 59.8k, False: 45.3M] ------------------ 172| 59.8k| pTop = top + (MBPRED_SIZE << 2); ------------------ | | 180| 59.8k|#define MBPRED_SIZE 15 ------------------ 173| 45.4M| if (diag) ------------------ | Branch (173:7): [True: 51.6k, False: 45.3M] ------------------ 174| 51.6k| pDiag = diag + (MBPRED_SIZE << 2); ------------------ | | 180| 51.6k|#define MBPRED_SIZE 15 ------------------ 175| 45.4M| break; 176| | 177| 45.4M| case 5: ------------------ | Branch (177:2): [True: 45.4M, False: 227M] ------------------ 178| 45.4M| if (left) ------------------ | Branch (178:7): [True: 80.2k, False: 45.3M] ------------------ 179| 80.2k| pLeft = left + 5 * MBPRED_SIZE; ------------------ | | 180| 80.2k|#define MBPRED_SIZE 15 ------------------ 180| 45.4M| if (top) ------------------ | Branch (180:7): [True: 59.8k, False: 45.3M] ------------------ 181| 59.8k| pTop = top + 5 * MBPRED_SIZE; ------------------ | | 180| 59.8k|#define MBPRED_SIZE 15 ------------------ 182| 45.4M| if (diag) ------------------ | Branch (182:7): [True: 51.6k, False: 45.3M] ------------------ 183| 51.6k| pDiag = diag + 5 * MBPRED_SIZE; ------------------ | | 180| 51.6k|#define MBPRED_SIZE 15 ------------------ 184| 45.4M| break; 185| 272M| } 186| | 187| | /* determine ac prediction direction & ac/dc predictor place rescaled ac/dc 188| | * predictions into predictors[] for later use */ 189| 272M| if (abs(pLeft[0] - pDiag[0]) < abs(pDiag[0] - pTop[0])) { ------------------ | Branch (189:6): [True: 33.0M, False: 239M] ------------------ 190| 33.0M| *acpred_direction = 1; /* vertical */ 191| 33.0M| predictors[0] = DIV_DIV(pTop[0], iDcScaler); ------------------ | | 263| 33.0M|#define DIV_DIV(a,b) (((a)>0) ? ((a)+((b)>>1))/(b) : ((a)-((b)>>1))/(b)) | | ------------------ | | | Branch (263:26): [True: 29.1M, False: 3.90M] | | ------------------ ------------------ 192| 264M| for (i = 1; i < 8; i++) { ------------------ | Branch (192:15): [True: 231M, False: 33.0M] ------------------ 193| 231M| predictors[i] = rescale(top_quant, current_quant, pTop[i]); 194| 231M| } 195| 239M| } else { 196| 239M| *acpred_direction = 2; /* horizontal */ 197| 239M| predictors[0] = DIV_DIV(pLeft[0], iDcScaler); ------------------ | | 263| 239M|#define DIV_DIV(a,b) (((a)>0) ? ((a)+((b)>>1))/(b) : ((a)-((b)>>1))/(b)) | | ------------------ | | | Branch (263:26): [True: 231M, False: 7.82M] | | ------------------ ------------------ 198| 1.91G| for (i = 1; i < 8; i++) { ------------------ | Branch (198:15): [True: 1.67G, False: 239M] ------------------ 199| 1.67G| predictors[i] = rescale(left_quant, current_quant, pLeft[i + 7]); 200| 1.67G| } 201| 239M| } 202| 272M|} add_acdc: 220| 272M|{ 221| 272M| uint8_t acpred_direction = pMB->acpred_directions[block]; 222| 272M| int16_t *pCurrent = (int16_t*)pMB->pred_values[block]; 223| 272M| uint32_t i; 224| | 225| 272M| DPRINTF(XVID_DEBUG_COEFF,"predictor[0] %i\n", predictors[0]); ------------------ | | 201| 272M|#define XVID_DEBUG_COEFF (1<< 5) ------------------ 226| | 227| 272M| dct_codes[0] += predictors[0]; /* dc prediction */ 228| 272M| pCurrent[0] = dct_codes[0]*iDcScaler; 229| 272M| if (bsversion > BS_VERSION_BUGGY_DC_CLIPPING) { ------------------ | | 211| 272M|#define BS_VERSION_BUGGY_DC_CLIPPING 34 ------------------ | Branch (229:6): [True: 267M, False: 4.75M] ------------------ 230| 267M| pCurrent[0] = CLIP(pCurrent[0], -2048, 2047); ------------------ | | 262| 267M|#define CLIP(X,AMIN,AMAX) (((X)<(AMIN)) ? (AMIN) : ((X)>(AMAX)) ? (AMAX) : (X)) | | ------------------ | | | Branch (262:30): [True: 85.3k, False: 267M] | | | Branch (262:54): [True: 318k, False: 267M] | | ------------------ ------------------ 231| 267M| } 232| | 233| 272M| if (acpred_direction == 1) { ------------------ | Branch (233:6): [True: 97.1k, False: 272M] ------------------ 234| 776k| for (i = 1; i < 8; i++) { ------------------ | Branch (234:15): [True: 679k, False: 97.1k] ------------------ 235| 679k| int level = dct_codes[i] + predictors[i]; 236| | 237| 679k| DPRINTF(XVID_DEBUG_COEFF,"predictor[%i] %i\n",i, predictors[i]); ------------------ | | 201| 679k|#define XVID_DEBUG_COEFF (1<< 5) ------------------ 238| | 239| 679k| dct_codes[i] = level; 240| 679k| pCurrent[i] = level; 241| 679k| pCurrent[i + 7] = dct_codes[i * 8]; 242| 679k| } 243| 272M| } else if (acpred_direction == 2) { ------------------ | Branch (243:13): [True: 271k, False: 272M] ------------------ 244| 2.17M| for (i = 1; i < 8; i++) { ------------------ | Branch (244:15): [True: 1.90M, False: 271k] ------------------ 245| 1.90M| int level = dct_codes[i * 8] + predictors[i]; 246| 1.90M| DPRINTF(XVID_DEBUG_COEFF,"predictor[%i] %i\n",i*8, predictors[i]); ------------------ | | 201| 1.90M|#define XVID_DEBUG_COEFF (1<< 5) ------------------ 247| | 248| 1.90M| dct_codes[i * 8] = level; 249| 1.90M| pCurrent[i + 7] = level; 250| 1.90M| pCurrent[i] = dct_codes[i]; 251| 1.90M| } 252| 272M| } else { 253| 2.17G| for (i = 1; i < 8; i++) { ------------------ | Branch (253:15): [True: 1.90G, False: 272M] ------------------ 254| 1.90G| pCurrent[i] = dct_codes[i]; 255| 1.90G| pCurrent[i + 7] = dct_codes[i * 8]; 256| 1.90G| } 257| 272M| } 258| 272M|} get_pmv2: 460| 220k|{ 461| 220k| int lx, ly, lz; /* left */ 462| 220k| int tx, ty, tz; /* top */ 463| 220k| int rx, ry, rz; /* top-right */ 464| 220k| int lpos, tpos, rpos; 465| 220k| int num_cand = 0, last_cand = 1; 466| | 467| 220k| VECTOR pmv[4]; /* left neighbour, top neighbour, top-right neighbour */ 468| | 469| 220k| switch (block) { 470| 109k| case 0: ------------------ | Branch (470:2): [True: 109k, False: 110k] ------------------ 471| 109k| lx = x - 1; ly = y; lz = 1; 472| 109k| tx = x; ty = y - 1; tz = 2; 473| 109k| rx = x + 1; ry = y - 1; rz = 2; 474| 109k| break; 475| 36.9k| case 1: ------------------ | Branch (475:2): [True: 36.9k, False: 183k] ------------------ 476| 36.9k| lx = x; ly = y; lz = 0; 477| 36.9k| tx = x; ty = y - 1; tz = 3; 478| 36.9k| rx = x + 1; ry = y - 1; rz = 2; 479| 36.9k| break; 480| 36.9k| case 2: ------------------ | Branch (480:2): [True: 36.9k, False: 183k] ------------------ 481| 36.9k| lx = x - 1; ly = y; lz = 3; 482| 36.9k| tx = x; ty = y; tz = 0; 483| 36.9k| rx = x; ry = y; rz = 1; 484| 36.9k| break; 485| 36.9k| default: ------------------ | Branch (485:2): [True: 36.9k, False: 183k] ------------------ 486| 36.9k| lx = x; ly = y; lz = 2; 487| 36.9k| tx = x; ty = y; tz = 0; 488| 36.9k| rx = x; ry = y; rz = 1; 489| 220k| } 490| | 491| 220k| lpos = lx + ly * mb_width; 492| 220k| rpos = rx + ry * mb_width; 493| 220k| tpos = tx + ty * mb_width; 494| | 495| 220k| if (lpos >= bound && lx >= 0) { ------------------ | Branch (495:6): [True: 208k, False: 12.1k] | Branch (495:23): [True: 197k, False: 10.5k] ------------------ 496| 197k| num_cand++; 497| 197k| pmv[1] = mbs[lpos].mvs[lz]; 498| 197k| } else pmv[1] = zeroMV; 499| | 500| 220k| if (tpos >= bound) { ------------------ | Branch (500:6): [True: 124k, False: 96.3k] ------------------ 501| 124k| num_cand++; 502| 124k| last_cand = 2; 503| 124k| pmv[2] = mbs[tpos].mvs[tz]; 504| 124k| } else pmv[2] = zeroMV; 505| | 506| 220k| if (rpos >= bound && rx < mb_width) { ------------------ | Branch (506:6): [True: 125k, False: 94.6k] | Branch (506:23): [True: 113k, False: 12.1k] ------------------ 507| 113k| num_cand++; 508| 113k| last_cand = 3; 509| 113k| pmv[3] = mbs[rpos].mvs[rz]; 510| 113k| } else pmv[3] = zeroMV; 511| | 512| | /* If there're more than one candidate, we return the median vector */ 513| | 514| 220k| if (num_cand > 1) { ------------------ | Branch (514:6): [True: 123k, False: 96.8k] ------------------ 515| | /* set median */ 516| 123k| pmv[0].x = 517| 123k| MIN(MAX(pmv[1].x, pmv[2].x), ------------------ | | 34| 990k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 15.7k, False: 108k] | | | Branch (34:21): [True: 30.3k, False: 93.4k] | | | Branch (34:25): [True: 27.6k, False: 96.2k] | | | Branch (34:25): [True: 34.2k, False: 89.5k] | | | Branch (34:25): [True: 3.66k, False: 18.8k] | | | Branch (34:25): [True: 11.6k, False: 89.5k] | | | Branch (34:25): [True: 22.5k, False: 101k] | | | Branch (34:29): [True: 3.05k, False: 12.6k] | | | Branch (34:33): [True: 27.6k, False: 80.5k] | | | Branch (34:33): [True: 34.2k, False: 73.8k] | | | Branch (34:33): [True: 3.66k, False: 18.8k] | | | Branch (34:33): [True: 11.6k, False: 73.8k] | | | Branch (34:33): [True: 22.5k, False: 85.5k] | | ------------------ ------------------ 518| 123k| MIN(MAX(pmv[2].x, pmv[3].x), MAX(pmv[1].x, pmv[3].x))); 519| 123k| pmv[0].y = 520| 123k| MIN(MAX(pmv[1].y, pmv[2].y), ------------------ | | 34| 990k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 14.1k, False: 109k] | | | Branch (34:21): [True: 28.1k, False: 95.7k] | | | Branch (34:25): [True: 15.3k, False: 108k] | | | Branch (34:25): [True: 24.1k, False: 99.6k] | | | Branch (34:25): [True: 2.57k, False: 14.4k] | | | Branch (34:25): [True: 7.17k, False: 99.6k] | | | Branch (34:25): [True: 16.9k, False: 106k] | | | Branch (34:29): [True: 2.70k, False: 11.4k] | | | Branch (34:33): [True: 15.3k, False: 94.2k] | | | Branch (34:33): [True: 24.1k, False: 85.4k] | | | Branch (34:33): [True: 2.57k, False: 14.4k] | | | Branch (34:33): [True: 7.17k, False: 85.4k] | | | Branch (34:33): [True: 16.9k, False: 92.6k] | | ------------------ ------------------ 521| 123k| MIN(MAX(pmv[2].y, pmv[3].y), MAX(pmv[1].y, pmv[3].y))); 522| 123k| return pmv[0]; 523| 123k| } 524| | 525| 96.8k| return pmv[last_cand]; /* no point calculating median mv */ 526| 220k|} get_pmv2_interlaced: 534| 75.1k|{ 535| 75.1k| int lx, ly, lz; /* left */ 536| 75.1k| int tx, ty, tz; /* top */ 537| 75.1k| int rx, ry, rz; /* top-right */ 538| 75.1k| int lpos, tpos, rpos; 539| 75.1k| int num_cand = 0, last_cand = 1; 540| | 541| 75.1k| VECTOR pmv[4]; /* left neighbour, top neighbour, top-right neighbour */ 542| | 543| 75.1k| lx=x-1; ly=y; lz=1; 544| 75.1k| tx=x; ty=y-1; tz=2; 545| 75.1k| rx=x+1; ry=y-1; rz=2; 546| | 547| 75.1k| lpos=lx+ly*mb_width; 548| 75.1k| rpos=rx+ry*mb_width; 549| 75.1k| tpos=tx+ty*mb_width; 550| | 551| 75.1k| if(lx>=0 && lpos>=bound) ------------------ | Branch (551:6): [True: 60.9k, False: 14.2k] | Branch (551:15): [True: 60.5k, False: 418] ------------------ 552| 60.5k| { 553| 60.5k| num_cand++; 554| 60.5k| if(mbs[lpos].field_pred) ------------------ | Branch (554:8): [True: 4.51k, False: 56.0k] ------------------ 555| 4.51k| pmv[1] = mbs[lpos].mvs_avg; 556| 56.0k| else 557| 56.0k| pmv[1] = mbs[lpos].mvs[lz]; 558| 60.5k| } 559| 14.6k| else 560| 14.6k| { 561| 14.6k| pmv[1] = zeroMV; 562| 14.6k| } 563| | 564| 75.1k| if(tpos>=bound) ------------------ | Branch (564:6): [True: 48.6k, False: 26.5k] ------------------ 565| 48.6k| { 566| 48.6k| num_cand++; 567| 48.6k| last_cand=2; 568| 48.6k| if(mbs[tpos].field_pred) ------------------ | Branch (568:8): [True: 6.31k, False: 42.3k] ------------------ 569| 6.31k| pmv[2] = mbs[tpos].mvs_avg; 570| 42.3k| else 571| 42.3k| pmv[2] = mbs[tpos].mvs[tz]; 572| 48.6k| } 573| 26.5k| else 574| 26.5k| { 575| 26.5k| pmv[2] = zeroMV; 576| 26.5k| } 577| | 578| 75.1k| if(rx=bound) ------------------ | Branch (578:6): [True: 60.3k, False: 14.7k] | Branch (578:21): [True: 34.5k, False: 25.8k] ------------------ 579| 34.5k| { 580| 34.5k| num_cand++; 581| 34.5k| last_cand = 3; 582| 34.5k| if(mbs[rpos].field_pred) ------------------ | Branch (582:8): [True: 3.97k, False: 30.5k] ------------------ 583| 3.97k| pmv[3] = mbs[rpos].mvs_avg; 584| 30.5k| else 585| 30.5k| pmv[3] = mbs[rpos].mvs[rz]; 586| 34.5k| } 587| 40.5k| else 588| 40.5k| { 589| 40.5k| pmv[3] = zeroMV; 590| 40.5k| } 591| | 592| | /* If there're more than one candidate, we return the median vector */ 593| 75.1k| if(num_cand>1) ------------------ | Branch (593:6): [True: 47.2k, False: 27.9k] ------------------ 594| 47.2k| { 595| | /* set median */ 596| 47.2k| pmv[0].x = MIN(MAX(pmv[1].x, pmv[2].x), ------------------ | | 34| 377k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 5.76k, False: 41.4k] | | | Branch (34:21): [True: 8.53k, False: 38.7k] | | | Branch (34:25): [True: 9.80k, False: 37.4k] | | | Branch (34:25): [True: 12.2k, False: 34.9k] | | | Branch (34:25): [True: 1.40k, False: 4.80k] | | | Branch (34:25): [True: 6.08k, False: 34.9k] | | | Branch (34:25): [True: 6.21k, False: 41.0k] | | | Branch (34:29): [True: 928, False: 4.83k] | | | Branch (34:33): [True: 9.80k, False: 31.6k] | | | Branch (34:33): [True: 12.2k, False: 29.1k] | | | Branch (34:33): [True: 1.40k, False: 4.80k] | | | Branch (34:33): [True: 6.08k, False: 29.1k] | | | Branch (34:33): [True: 6.21k, False: 35.2k] | | ------------------ ------------------ 597| 47.2k| MIN(MAX(pmv[2].x, pmv[3].x), MAX(pmv[1].x, pmv[3].x))); 598| 47.2k| pmv[0].y = MIN(MAX(pmv[1].y, pmv[2].y), ------------------ | | 34| 377k|#define MIN(X, Y) ((X)<(Y)?(X):(Y)) | | ------------------ | | | Branch (34:20): [True: 5.80k, False: 41.4k] | | | Branch (34:21): [True: 8.97k, False: 38.2k] | | | Branch (34:25): [True: 8.85k, False: 38.3k] | | | Branch (34:25): [True: 11.4k, False: 35.8k] | | | Branch (34:25): [True: 1.49k, False: 4.73k] | | | Branch (34:25): [True: 5.18k, False: 35.8k] | | | Branch (34:25): [True: 6.22k, False: 41.0k] | | | Branch (34:29): [True: 786, False: 5.02k] | | | Branch (34:33): [True: 8.85k, False: 32.5k] | | | Branch (34:33): [True: 11.4k, False: 30.0k] | | | Branch (34:33): [True: 1.49k, False: 4.73k] | | | Branch (34:33): [True: 5.18k, False: 30.0k] | | | Branch (34:33): [True: 6.22k, False: 35.2k] | | ------------------ ------------------ 599| 47.2k| MIN(MAX(pmv[2].y, pmv[3].y), MAX(pmv[1].y, pmv[3].y))); 600| | 601| 47.2k| return pmv[0]; 602| 47.2k| } 603| | 604| 27.9k| return pmv[last_cand]; /* no point calculating median mv */ 605| 75.1k|} mbprediction.c:rescale: 42| 1.90G|{ 43| 1.90G| return (coeff != 0) ? DIV_DIV((coeff) * (predict_quant), ------------------ | | 263| 649k|#define DIV_DIV(a,b) (((a)>0) ? ((a)+((b)>>1))/(b) : ((a)-((b)>>1))/(b)) | | ------------------ | | | Branch (263:26): [True: 186k, False: 462k] | | ------------------ ------------------ | Branch (43:9): [True: 649k, False: 1.90G] ------------------ 44| 1.90G| (current_quant)) : 0; 45| 1.90G|} dequant_h263_intra_c: 161| 158M|{ 162| 158M| const int32_t quant_m_2 = quant << 1; 163| 158M| const int32_t quant_add = (quant & 1 ? quant : quant - 1); ------------------ | Branch (163:29): [True: 100M, False: 57.7M] ------------------ 164| 158M| int i; 165| | 166| 158M| data[0] = coeff[0] * dcscalar; 167| 158M| if (data[0] < -2048) { ------------------ | Branch (167:6): [True: 29.2k, False: 158M] ------------------ 168| 29.2k| data[0] = -2048; 169| 158M| } else if (data[0] > 2047) { ------------------ | Branch (169:13): [True: 144k, False: 158M] ------------------ 170| 144k| data[0] = 2047; 171| 144k| } 172| | 173| 10.1G| for (i = 1; i < 64; i++) { ------------------ | Branch (173:14): [True: 9.97G, False: 158M] ------------------ 174| 9.97G| int32_t acLevel = coeff[i]; 175| | 176| 9.97G| if (acLevel == 0) { ------------------ | Branch (176:7): [True: 9.97G, False: 876k] ------------------ 177| 9.97G| data[i] = 0; 178| 9.97G| } else if (acLevel < 0) { ------------------ | Branch (178:14): [True: 564k, False: 311k] ------------------ 179| 564k| acLevel = quant_m_2 * -acLevel + quant_add; 180| 564k| data[i] = (acLevel <= 2048 ? -acLevel : -2048); ------------------ | Branch (180:15): [True: 547k, False: 17.6k] ------------------ 181| 564k| } else { 182| 311k| acLevel = quant_m_2 * acLevel + quant_add; 183| 311k| data[i] = (acLevel <= 2047 ? acLevel : 2047); ------------------ | Branch (183:15): [True: 297k, False: 13.9k] ------------------ 184| 311k| } 185| 9.97G| } 186| | 187| 158M| return(0); 188| 158M|} get_intra_matrix: 61| 114M|{ 62| 114M| return(mpeg_quant_matrices + 0*64); 63| 114M|} get_inter_matrix: 67| 526k|{ 68| 526k| return(mpeg_quant_matrices + 4*64); 69| 526k|} get_default_intra_matrix: 73| 8.96k|{ 74| 8.96k| return default_intra_matrix; 75| 8.96k|} get_default_inter_matrix: 79| 10.7k|{ 80| 10.7k| return default_inter_matrix; 81| 10.7k|} set_intra_matrix: 113| 22.3k|{ 114| 22.3k| int i; 115| 22.3k| uint16_t *intra_matrix = mpeg_quant_matrices + 0*64; 116| | 117| 1.45M| for (i = 0; i < 64; i++) { ------------------ | Branch (117:14): [True: 1.43M, False: 22.3k] ------------------ 118| 1.43M| intra_matrix[i] = (!i) ? (uint16_t)8: (uint16_t)MAX(1, matrix[i]); ------------------ | | 258| 2.84M|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (258:20): [True: 76.1k, False: 1.33M] | | ------------------ ------------------ | Branch (118:21): [True: 22.3k, False: 1.40M] ------------------ 119| 1.43M| } 120| 22.3k|} set_inter_matrix: 137| 22.3k|{ 138| 22.3k| int i; 139| 22.3k| uint16_t *inter_matrix = mpeg_quant_matrices + 4*64; 140| 22.3k| uint16_t *inter_matrix1 = mpeg_quant_matrices + 5*64; 141| 22.3k| uint16_t *inter_matrix_fix = mpeg_quant_matrices + 6*64; 142| 22.3k| uint16_t *inter_matrix_fixl = mpeg_quant_matrices + 7*64; 143| | 144| 1.45M| for (i = 0; i < 64; i++) { ------------------ | Branch (144:14): [True: 1.43M, False: 22.3k] ------------------ 145| 1.43M| inter_matrix1[i] = ((inter_matrix[i] = (int16_t)MAX(1, matrix[i]))>>1); ------------------ | | 258| 1.43M|#define MAX(X, Y) ((X)>(Y)?(X):(Y)) | | ------------------ | | | Branch (258:20): [True: 48.6k, False: 1.38M] | | ------------------ ------------------ 146| 1.43M| inter_matrix1[i] += ((inter_matrix[i] == 1) ? 1: 0); ------------------ | Branch (146:24): [True: 50.1k, False: 1.38M] ------------------ 147| 1.43M| inter_matrix_fix[i] = (uint16_t) FIX(inter_matrix[i]); ------------------ | | 30| 1.43M|#define FIX(X) (((X)==1) ? 0xFFFF : ((1UL << 16) / (X) + 1)) | | ------------------ | | | Branch (30:19): [True: 50.1k, False: 1.38M] | | ------------------ ------------------ 148| 1.43M| inter_matrix_fixl[i] = (uint16_t) FIXL(inter_matrix[i]); ------------------ | | 31| 1.43M|#define FIXL(X) ((1UL << 16) / (X) - 1) ------------------ 149| 1.43M| } 150| 22.3k|} init_mpeg_matrix: 153| 10.2k|init_mpeg_matrix(uint16_t * mpeg_quant_matrices) { 154| | 155| 10.2k| set_intra_matrix(mpeg_quant_matrices, default_intra_matrix); 156| 10.2k| set_inter_matrix(mpeg_quant_matrices, default_inter_matrix); 157| 10.2k|} dequant_mpeg_intra_c: 145| 114M|{ 146| 114M| const uint16_t *intra_matrix = get_intra_matrix(mpeg_quant_matrices); 147| 114M| int i; 148| | 149| 114M| data[0] = coeff[0] * dcscalar; 150| 114M| if (data[0] < -2048) { ------------------ | Branch (150:6): [True: 69.3k, False: 114M] ------------------ 151| 69.3k| data[0] = -2048; 152| 114M| } else if (data[0] > 2047) { ------------------ | Branch (152:13): [True: 195k, False: 114M] ------------------ 153| 195k| data[0] = 2047; 154| 195k| } 155| | 156| 7.31G| for (i = 1; i < 64; i++) { ------------------ | Branch (156:14): [True: 7.20G, False: 114M] ------------------ 157| 7.20G| if (coeff[i] == 0) { ------------------ | Branch (157:7): [True: 7.20G, False: 1.48M] ------------------ 158| 7.20G| data[i] = 0; 159| 7.20G| } else if (coeff[i] < 0) { ------------------ | Branch (159:14): [True: 1.06M, False: 426k] ------------------ 160| 1.06M| uint32_t level = -coeff[i]; 161| | 162| 1.06M| level = (level * intra_matrix[i] * quant) >> 3; 163| 1.06M| data[i] = (level <= 2048 ? -(int16_t) level : -2048); ------------------ | Branch (163:15): [True: 409k, False: 651k] ------------------ 164| 1.06M| } else { 165| 426k| uint32_t level = coeff[i]; 166| | 167| 426k| level = (level * intra_matrix[i] * quant) >> 3; 168| 426k| data[i] = (level <= 2047 ? level : 2047); ------------------ | Branch (168:15): [True: 274k, False: 151k] ------------------ 169| 426k| } 170| 7.20G| } 171| | 172| 114M| return(0); 173| 114M|} emms_c: 47| 80.2k|{ 48| 80.2k|} xvid_malloc: 50| 508k|{ 51| 508k| uint8_t *mem_ptr; 52| | 53| 508k| if (!alignment) { ------------------ | Branch (53:6): [True: 0, False: 508k] ------------------ 54| | 55| | /* We have not to satisfy any alignment */ 56| 0| if ((mem_ptr = (uint8_t *) malloc(size + 1)) != NULL) { ------------------ | Branch (56:7): [True: 0, False: 0] ------------------ 57| | 58| | /* Store (mem_ptr - "real allocated memory") in *(mem_ptr-1) */ 59| 0| *mem_ptr = (uint8_t)1; 60| | 61| | /* Return the mem_ptr pointer */ 62| 0| return ((void *)(mem_ptr+1)); 63| 0| } 64| 508k| } else { 65| 508k| uint8_t *tmp; 66| | 67| | /* Allocate the required size memory + alignment so we 68| | * can realign the data if necessary */ 69| 508k| if ((tmp = (uint8_t *) malloc(size + alignment)) != NULL) { ------------------ | Branch (69:7): [True: 508k, False: 0] ------------------ 70| | 71| | /* Align the tmp pointer */ 72| 508k| mem_ptr = 73| 508k| (uint8_t *) ((ptr_t) (tmp + alignment - 1) & 74| 508k| (~(ptr_t) (alignment - 1))); 75| | 76| | /* Special case where malloc have already satisfied the alignment 77| | * We must add alignment to mem_ptr because we must store 78| | * (mem_ptr - tmp) in *(mem_ptr-1) 79| | * If we do not add alignment to mem_ptr then *(mem_ptr-1) points 80| | * to a forbidden memory space */ 81| 508k| if (mem_ptr == tmp) ------------------ | Branch (81:8): [True: 132k, False: 376k] ------------------ 82| 132k| mem_ptr += alignment; 83| | 84| | /* (mem_ptr - tmp) is stored in *(mem_ptr-1) so we are able to retrieve 85| | * the real malloc block allocated and free it in xvid_free */ 86| 508k| *(mem_ptr - 1) = (uint8_t) (mem_ptr - tmp); 87| | 88| | /* Return the aligned pointer */ 89| 508k| return ((void *)mem_ptr); 90| 508k| } 91| 508k| } 92| | 93| 0| return(NULL); 94| 508k|} xvid_free: 108| 539k|{ 109| | 110| 539k| uint8_t *ptr; 111| | 112| 539k| if (mem_ptr == NULL) ------------------ | Branch (112:6): [True: 30.8k, False: 508k] ------------------ 113| 30.8k| return; 114| | 115| | /* Aligned pointer */ 116| 508k| ptr = mem_ptr; 117| | 118| | /* *(ptr - 1) holds the offset to the real allocated block 119| | * we sub that offset os we free the real pointer */ 120| 508k| ptr -= *(ptr - 1); 121| | 122| | /* Free the memory */ 123| 508k| free(ptr); 124| 508k|} transfer_16to8copy_c: 91| 272M|{ 92| 272M| int i, j; 93| | 94| 2.45G| for (j = 0; j < 8; j++) { ------------------ | Branch (94:14): [True: 2.18G, False: 272M] ------------------ 95| 19.6G| for (i = 0; i < 8; i++) { ------------------ | Branch (95:15): [True: 17.4G, False: 2.18G] ------------------ 96| 17.4G|#ifdef USE_REFERENCE_C 97| 17.4G| int16_t pixel = src[j * 8 + i]; 98| | 99| 17.4G| if (pixel < 0) { ------------------ | Branch (99:8): [True: 14.4M, False: 17.4G] ------------------ 100| 14.4M| pixel = 0; 101| 17.4G| } else if (pixel > 255) { ------------------ | Branch (101:15): [True: 20.0M, False: 17.4G] ------------------ 102| 20.0M| pixel = 255; 103| 20.0M| } 104| 17.4G| dst[j * stride + i] = (uint8_t) pixel; 105| |#else 106| | const int16_t pixel = src[j * 8 + i]; 107| | const uint8_t value = (uint8_t)( (pixel&~255) ? (-pixel)>>(8*sizeof(pixel)-1) : pixel ); 108| | dst[j*stride + i] = value; 109| |#endif 110| 17.4G| } 111| 2.18G| } 112| 272M|} transfer_16to8add_c: 232| 526k|{ 233| 526k| int i, j; 234| | 235| 4.74M| for (j = 0; j < 8; j++) { ------------------ | Branch (235:14): [True: 4.21M, False: 526k] ------------------ 236| 37.9M| for (i = 0; i < 8; i++) { ------------------ | Branch (236:15): [True: 33.7M, False: 4.21M] ------------------ 237| 33.7M|#ifdef USE_REFERENCE_C 238| 33.7M| int16_t pixel = (int16_t) dst[j * stride + i] + src[j * 8 + i]; 239| | 240| 33.7M| if (pixel < 0) { ------------------ | Branch (240:8): [True: 2.56M, False: 31.1M] ------------------ 241| 2.56M| pixel = 0; 242| 31.1M| } else if (pixel > 255) { ------------------ | Branch (242:15): [True: 119k, False: 31.0M] ------------------ 243| 119k| pixel = 255; 244| 119k| } 245| 33.7M| dst[j * stride + i] = (uint8_t) pixel; 246| |#else 247| | const int16_t pixel = (int16_t) dst[j * stride + i] + src[j * 8 + i]; 248| | const uint8_t value = (uint8_t)( (pixel&~255) ? (-pixel)>>(8*sizeof(pixel)-1) : pixel ); 249| | dst[j*stride + i] = value; 250| |#endif 251| | 252| 33.7M| } 253| 4.21M| } 254| 526k|} transfer8x8_copy_c: 269| 5.00M|{ 270| 5.00M| int j, i; 271| | 272| 45.0M| for (j = 0; j < 8; ++j) { ------------------ | Branch (272:14): [True: 40.0M, False: 5.00M] ------------------ 273| 40.0M| uint8_t *d = dst + j*stride; 274| 40.0M| const uint8_t *s = src + j*stride; 275| | 276| 360M| for (i = 0; i < 8; ++i) ------------------ | Branch (276:15): [True: 320M, False: 40.0M] ------------------ 277| 320M| { 278| 320M| *d++ = *s++; 279| 320M| } 280| 40.0M| } 281| 5.00M|} transfer8x4_copy_c: 296| 37.4k|{ 297| 37.4k| uint32_t j; 298| | 299| 187k| for (j = 0; j < 4; j++) { ------------------ | Branch (299:14): [True: 149k, False: 37.4k] ------------------ 300| 149k| uint32_t *d= (uint32_t*)(dst + j*stride); 301| 149k| const uint32_t *s = (const uint32_t*)(src + j*stride); 302| 149k| *(d+0) = *(s+0); 303| 149k| *(d+1) = *(s+1); 304| 149k| } 305| 37.4k|} decoder.c:init_timer: 106| 10.2k|{ 107| 10.2k|} decoder.c:write_timer: 110| 10.2k|{ 111| 10.2k|} decoder.c:start_global_timer: 62| 69.9k|{ 63| 69.9k|} decoder.c:stop_global_timer: 126| 17.1k|{ 127| 17.1k|} decoder.c:start_timer: 58| 1.41G|{ 59| 1.41G|} decoder.c:stop_prediction_timer: 122| 545M|{ 123| 545M|} decoder.c:stop_coding_timer: 114| 273M|{ 115| 273M|} decoder.c:stop_iquant_timer: 94| 272M|{ 95| 272M|} decoder.c:stop_idct_timer: 70| 273M|{ 71| 273M|} decoder.c:stop_transfer_timer: 102| 46.3M|{ 103| 46.3M|} decoder.c:stop_edges_timer: 82| 7.93k|{ 83| 7.93k|} decoder.c:stop_comp_timer: 78| 1.08M|{ 79| 1.08M|} xvid_global: 812| 2|{ 813| 2| switch(opt) 814| 2| { 815| 2| case XVID_GBL_INIT : ------------------ | | 235| 2|#define XVID_GBL_INIT 0 /* initialize xvidcore; must be called before using xvid_decore, or xvid_encore) */ ------------------ | Branch (815:3): [True: 2, False: 0] ------------------ 816| 2| return xvid_gbl_init((xvid_gbl_init_t*)param1); 817| | 818| 0| case XVID_GBL_INFO : ------------------ | | 236| 0|#define XVID_GBL_INFO 1 /* return some info about xvidcore, and the host computer */ ------------------ | Branch (818:9): [True: 0, False: 2] ------------------ 819| 0| return xvid_gbl_info((xvid_gbl_info_t*)param1); 820| | 821| 0| case XVID_GBL_CONVERT : ------------------ | | 237| 0|#define XVID_GBL_CONVERT 2 /* colorspace conversion utility */ ------------------ | Branch (821:3): [True: 0, False: 2] ------------------ 822| 0| return xvid_gbl_convert((xvid_gbl_convert_t*)param1); 823| | 824| 0| default : ------------------ | Branch (824:3): [True: 0, False: 2] ------------------ 825| 0| return XVID_ERR_FAIL; ------------------ | | 95| 0|#define XVID_ERR_FAIL -1 /* general fault */ ------------------ 826| 2| } 827| 2|} xvid_decore: 844| 90.5k|{ 845| 90.5k| switch (opt) { 846| 10.2k| case XVID_DEC_CREATE: ------------------ | | 246| 10.2k|#define XVID_DEC_CREATE 0 /* create decore instance; return 0 on success */ ------------------ | Branch (846:2): [True: 10.2k, False: 80.2k] ------------------ 847| 10.2k| return decoder_create((xvid_dec_create_t *) param1); 848| | 849| 10.2k| case XVID_DEC_DESTROY: ------------------ | | 247| 10.2k|#define XVID_DEC_DESTROY 1 /* destroy decore instance: return 0 on success */ ------------------ | Branch (849:2): [True: 10.2k, False: 80.2k] ------------------ 850| 10.2k| return decoder_destroy((DECODER *) handle); 851| | 852| 69.9k| case XVID_DEC_DECODE: ------------------ | | 248| 69.9k|#define XVID_DEC_DECODE 2 /* decode a frame: returns number of bytes consumed >= 0 */ ------------------ | Branch (852:2): [True: 69.9k, False: 20.5k] ------------------ 853| 69.9k| return decoder_decode((DECODER *) handle, (xvid_dec_frame_t *) param1, (xvid_dec_stats_t*) param2); 854| | 855| 0| default: ------------------ | Branch (855:2): [True: 0, False: 90.5k] ------------------ 856| 0| return XVID_ERR_FAIL; ------------------ | | 95| 0|#define XVID_ERR_FAIL -1 /* general fault */ ------------------ 857| 90.5k| } 858| 90.5k|} xvid.c:xvid_gbl_init: 200| 2|{ 201| 2| unsigned int cpu_flags; 202| | 203| 2| if (XVID_VERSION_MAJOR(init->version) != 1) /* v1.x.x */ ------------------ | | 63| 2|#define XVID_VERSION_MAJOR(a) ((char)(((a)>>16) & 0xff)) ------------------ | Branch (203:6): [True: 0, False: 2] ------------------ 204| 0| return XVID_ERR_VERSION; ------------------ | | 98| 0|#define XVID_ERR_VERSION -4 /* structure version not supported */ ------------------ 205| | 206| 2| cpu_flags = (init->cpu_flags & XVID_CPU_FORCE) ? init->cpu_flags : detect_cpu_flags(); ------------------ | | 180| 2|#define XVID_CPU_FORCE (1<<31) /* force passed cpu flags */ ------------------ | Branch (206:14): [True: 0, False: 2] ------------------ 207| | 208| | /* Initialize the function pointers */ 209| 2| init_vlc_tables(); 210| | 211| | /* Fixed Point Forward/Inverse DCT transformations */ 212| 2| fdct = fdct_int32; 213| 2| idct = idct_int32; 214| | 215| | /* Only needed on PPC Altivec archs */ 216| 2| sadInit = NULL; 217| | 218| | /* Restore FPU context : emms_c is a nop functions */ 219| 2| emms = emms_c; 220| | 221| | /* Qpel stuff */ 222| 2| xvid_QP_Funcs = &xvid_QP_Funcs_C; 223| 2| xvid_QP_Add_Funcs = &xvid_QP_Add_Funcs_C; 224| 2| xvid_Init_QP(); 225| | 226| | /* Quantization functions */ 227| 2| quant_h263_intra = quant_h263_intra_c; 228| 2| quant_h263_inter = quant_h263_inter_c; 229| 2| dequant_h263_intra = dequant_h263_intra_c; 230| 2| dequant_h263_inter = dequant_h263_inter_c; 231| | 232| 2| quant_mpeg_intra = quant_mpeg_intra_c; 233| 2| quant_mpeg_inter = quant_mpeg_inter_c; 234| 2| dequant_mpeg_intra = dequant_mpeg_intra_c; 235| 2| dequant_mpeg_inter = dequant_mpeg_inter_c; 236| | 237| | /* Block transfer related functions */ 238| 2| transfer_8to16copy = transfer_8to16copy_c; 239| 2| transfer_16to8copy = transfer_16to8copy_c; 240| 2| transfer_8to16sub = transfer_8to16sub_c; 241| 2| transfer_8to16subro = transfer_8to16subro_c; 242| 2| transfer_8to16sub2 = transfer_8to16sub2_c; 243| 2| transfer_8to16sub2ro = transfer_8to16sub2ro_c; 244| 2| transfer_16to8add = transfer_16to8add_c; 245| 2| transfer8x8_copy = transfer8x8_copy_c; 246| 2| transfer8x4_copy = transfer8x4_copy_c; 247| | 248| | /* Interlacing functions */ 249| 2| MBFieldTest = MBFieldTest_c; 250| | 251| | /* Image interpolation related functions */ 252| 2| interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_c; 253| 2| interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_c; 254| 2| interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_c; 255| | 256| 2| interpolate8x4_halfpel_h = interpolate8x4_halfpel_h_c; 257| 2| interpolate8x4_halfpel_v = interpolate8x4_halfpel_v_c; 258| 2| interpolate8x4_halfpel_hv = interpolate8x4_halfpel_hv_c; 259| | 260| 2| interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_c; 261| 2| interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_c; 262| 2| interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_c; 263| 2| interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_c; 264| | 265| 2| interpolate16x16_lowpass_h = interpolate16x16_lowpass_h_c; 266| 2| interpolate16x16_lowpass_v = interpolate16x16_lowpass_v_c; 267| 2| interpolate16x16_lowpass_hv = interpolate16x16_lowpass_hv_c; 268| | 269| 2| interpolate8x8_lowpass_h = interpolate8x8_lowpass_h_c; 270| 2| interpolate8x8_lowpass_v = interpolate8x8_lowpass_v_c; 271| 2| interpolate8x8_lowpass_hv = interpolate8x8_lowpass_hv_c; 272| | 273| 2| interpolate8x8_6tap_lowpass_h = interpolate8x8_6tap_lowpass_h_c; 274| 2| interpolate8x8_6tap_lowpass_v = interpolate8x8_6tap_lowpass_v_c; 275| | 276| 2| interpolate8x8_avg2 = interpolate8x8_avg2_c; 277| 2| interpolate8x8_avg4 = interpolate8x8_avg4_c; 278| | 279| | /* postprocessing */ 280| 2| image_brightness = image_brightness_c; 281| | 282| | /* Initialize internal colorspace transformation tables */ 283| 2| colorspace_init(); 284| | 285| | /* All colorspace transformation functions User Format->YV12 */ 286| 2| yv12_to_yv12 = yv12_to_yv12_c; 287| 2| rgb555_to_yv12 = rgb555_to_yv12_c; 288| 2| rgb565_to_yv12 = rgb565_to_yv12_c; 289| 2| rgb_to_yv12 = rgb_to_yv12_c; 290| 2| bgr_to_yv12 = bgr_to_yv12_c; 291| 2| bgra_to_yv12 = bgra_to_yv12_c; 292| 2| abgr_to_yv12 = abgr_to_yv12_c; 293| 2| rgba_to_yv12 = rgba_to_yv12_c; 294| 2| argb_to_yv12 = argb_to_yv12_c; 295| 2| yuyv_to_yv12 = yuyv_to_yv12_c; 296| 2| uyvy_to_yv12 = uyvy_to_yv12_c; 297| | 298| 2| rgb555i_to_yv12 = rgb555i_to_yv12_c; 299| 2| rgb565i_to_yv12 = rgb565i_to_yv12_c; 300| 2| bgri_to_yv12 = bgri_to_yv12_c; 301| 2| bgrai_to_yv12 = bgrai_to_yv12_c; 302| 2| abgri_to_yv12 = abgri_to_yv12_c; 303| 2| rgbai_to_yv12 = rgbai_to_yv12_c; 304| 2| argbi_to_yv12 = argbi_to_yv12_c; 305| 2| yuyvi_to_yv12 = yuyvi_to_yv12_c; 306| 2| uyvyi_to_yv12 = uyvyi_to_yv12_c; 307| | 308| | /* All colorspace transformation functions YV12->User format */ 309| 2| yv12_to_rgb555 = yv12_to_rgb555_c; 310| 2| yv12_to_rgb565 = yv12_to_rgb565_c; 311| 2| yv12_to_rgb = yv12_to_rgb_c; 312| 2| yv12_to_bgr = yv12_to_bgr_c; 313| 2| yv12_to_bgra = yv12_to_bgra_c; 314| 2| yv12_to_abgr = yv12_to_abgr_c; 315| 2| yv12_to_rgba = yv12_to_rgba_c; 316| 2| yv12_to_argb = yv12_to_argb_c; 317| 2| yv12_to_yuyv = yv12_to_yuyv_c; 318| 2| yv12_to_uyvy = yv12_to_uyvy_c; 319| | 320| 2| yv12_to_rgb555i = yv12_to_rgb555i_c; 321| 2| yv12_to_rgb565i = yv12_to_rgb565i_c; 322| 2| yv12_to_bgri = yv12_to_bgri_c; 323| 2| yv12_to_bgrai = yv12_to_bgrai_c; 324| 2| yv12_to_abgri = yv12_to_abgri_c; 325| 2| yv12_to_rgbai = yv12_to_rgbai_c; 326| 2| yv12_to_argbi = yv12_to_argbi_c; 327| 2| yv12_to_yuyvi = yv12_to_yuyvi_c; 328| 2| yv12_to_uyvyi = yv12_to_uyvyi_c; 329| | 330| | /* Functions used in motion estimation algorithms */ 331| 2| calc_cbp = calc_cbp_c; 332| 2| sad16 = sad16_c; 333| 2| sad8 = sad8_c; 334| 2| sad16bi = sad16bi_c; 335| 2| sad8bi = sad8bi_c; 336| 2| dev16 = dev16_c; 337| 2| sad16v = sad16v_c; 338| 2| sse8_16bit = sse8_16bit_c; 339| 2| sse8_8bit = sse8_8bit_c; 340| | 341| 2| sseh8_16bit = sseh8_16bit_c; 342| 2| coeff8_energy = coeff8_energy_c; 343| 2| blocksum8 = blocksum8_c; 344| | 345| 2| init_GMC(cpu_flags); 346| | 347| |#if defined(ARCH_IS_IA32) || defined(ARCH_IS_X86_64) 348| | 349| | if ((cpu_flags & XVID_CPU_MMX) || (cpu_flags & XVID_CPU_MMXEXT) || 350| | (cpu_flags & XVID_CPU_3DNOW) || (cpu_flags & XVID_CPU_3DNOWEXT) || 351| | (cpu_flags & XVID_CPU_SSE) || (cpu_flags & XVID_CPU_SSE2) || 352| | (cpu_flags & XVID_CPU_SSE3) || (cpu_flags & XVID_CPU_SSE41)) 353| | { 354| | /* Restore FPU context : emms_c is a nop functions */ 355| | emms = emms_mmx; 356| | } 357| | 358| | if ((cpu_flags & XVID_CPU_MMX)) { 359| | 360| | /* Forward and Inverse Discrete Cosine Transformation functions */ 361| | fdct = fdct_mmx_skal; 362| | idct = idct_mmx; 363| | 364| | /* Qpel stuff */ 365| | xvid_QP_Funcs = &xvid_QP_Funcs_mmx; 366| | xvid_QP_Add_Funcs = &xvid_QP_Add_Funcs_mmx; 367| | 368| | /* Quantization related functions */ 369| | quant_h263_intra = quant_h263_intra_mmx; 370| | quant_h263_inter = quant_h263_inter_mmx; 371| | dequant_h263_intra = dequant_h263_intra_mmx; 372| | dequant_h263_inter = dequant_h263_inter_mmx; 373| | quant_mpeg_intra = quant_mpeg_intra_mmx; 374| | quant_mpeg_inter = quant_mpeg_inter_mmx; 375| | dequant_mpeg_intra = dequant_mpeg_intra_mmx; 376| | dequant_mpeg_inter = dequant_mpeg_inter_mmx; 377| | 378| | 379| | /* Block related functions */ 380| | transfer_8to16copy = transfer_8to16copy_mmx; 381| | transfer_16to8copy = transfer_16to8copy_mmx; 382| | transfer_8to16sub = transfer_8to16sub_mmx; 383| | transfer_8to16subro = transfer_8to16subro_mmx; 384| | transfer_8to16sub2 = transfer_8to16sub2_mmx; 385| | transfer_16to8add = transfer_16to8add_mmx; 386| | transfer8x8_copy = transfer8x8_copy_mmx; 387| | transfer8x4_copy = transfer8x4_copy_mmx; 388| | 389| | /* Interlacing Functions */ 390| | MBFieldTest = MBFieldTest_mmx; 391| | 392| | /* Image Interpolation related functions */ 393| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_mmx; 394| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_mmx; 395| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_mmx; 396| | 397| | interpolate8x4_halfpel_h = interpolate8x4_halfpel_h_mmx; 398| | interpolate8x4_halfpel_v = interpolate8x4_halfpel_v_mmx; 399| | interpolate8x4_halfpel_hv = interpolate8x4_halfpel_hv_mmx; 400| | 401| | interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_mmx; 402| | interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_mmx; 403| | interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_mmx; 404| | interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_mmx; 405| | 406| | interpolate8x8_6tap_lowpass_h = interpolate8x8_6tap_lowpass_h_mmx; 407| | interpolate8x8_6tap_lowpass_v = interpolate8x8_6tap_lowpass_v_mmx; 408| | 409| | interpolate8x8_avg2 = interpolate8x8_avg2_mmx; 410| | interpolate8x8_avg4 = interpolate8x8_avg4_mmx; 411| | 412| | /* postprocessing */ 413| | image_brightness = image_brightness_mmx; 414| | 415| | /* image input xxx_to_yv12 related functions */ 416| | 417| | yv12_to_yv12 = yv12_to_yv12_mmx; 418| | 419| | bgr_to_yv12 = bgr_to_yv12_mmx; 420| | rgb_to_yv12 = rgb_to_yv12_mmx; 421| | bgra_to_yv12 = bgra_to_yv12_mmx; 422| | rgba_to_yv12 = rgba_to_yv12_mmx; 423| | yuyv_to_yv12 = yuyv_to_yv12_mmx; 424| | uyvy_to_yv12 = uyvy_to_yv12_mmx; 425| | 426| | /* image output yv12_to_xxx related functions */ 427| | yv12_to_bgr = yv12_to_bgr_mmx; 428| | yv12_to_bgra = yv12_to_bgra_mmx; 429| | yv12_to_yuyv = yv12_to_yuyv_mmx; 430| | yv12_to_uyvy = yv12_to_uyvy_mmx; 431| | 432| | yv12_to_yuyvi = yv12_to_yuyvi_mmx; 433| | yv12_to_uyvyi = yv12_to_uyvyi_mmx; 434| | 435| | /* Motion estimation related functions */ 436| | calc_cbp = calc_cbp_mmx; 437| | sad16 = sad16_mmx; 438| | sad8 = sad8_mmx; 439| | sad16bi = sad16bi_mmx; 440| | sad8bi = sad8bi_mmx; 441| | dev16 = dev16_mmx; 442| | sad16v = sad16v_mmx; 443| | sse8_16bit = sse8_16bit_mmx; 444| | sse8_8bit = sse8_8bit_mmx; 445| | } 446| | 447| | /* these 3dnow functions are faster than mmx, but slower than xmm. */ 448| | if ((cpu_flags & XVID_CPU_3DNOW)) { 449| | 450| | emms = emms_3dn; 451| | 452| | /* ME functions */ 453| | sad16bi = sad16bi_3dn; 454| | sad8bi = sad8bi_3dn; 455| | 456| | yuyv_to_yv12 = yuyv_to_yv12_3dn; 457| | uyvy_to_yv12 = uyvy_to_yv12_3dn; 458| | 459| | } 460| | 461| | 462| | if ((cpu_flags & XVID_CPU_MMXEXT)) { 463| | 464| | /* DCT */ 465| | fdct = fdct_xmm_skal; 466| | idct = idct_xmm; 467| | 468| | /* Interpolation */ 469| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_xmm; 470| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_xmm; 471| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_xmm; 472| | 473| | interpolate8x4_halfpel_h = interpolate8x4_halfpel_h_xmm; 474| | interpolate8x4_halfpel_v = interpolate8x4_halfpel_v_xmm; 475| | interpolate8x4_halfpel_hv = interpolate8x4_halfpel_hv_xmm; 476| | 477| | interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_xmm; 478| | interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_xmm; 479| | interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_xmm; 480| | interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_xmm; 481| | 482| | /* Quantization */ 483| | quant_mpeg_inter = quant_mpeg_inter_xmm; 484| | 485| | dequant_h263_intra = dequant_h263_intra_xmm; 486| | dequant_h263_inter = dequant_h263_inter_xmm; 487| | 488| | /* Buffer transfer */ 489| | transfer_8to16sub2 = transfer_8to16sub2_xmm; 490| | transfer_8to16sub2ro = transfer_8to16sub2ro_xmm; 491| | 492| | /* Colorspace transformation */ 493| | /* yv12_to_yv12 = yv12_to_yv12_xmm; */ /* appears to be slow on many machines */ 494| | yuyv_to_yv12 = yuyv_to_yv12_xmm; 495| | uyvy_to_yv12 = uyvy_to_yv12_xmm; 496| | 497| | /* ME functions */ 498| | sad16 = sad16_xmm; 499| | sad8 = sad8_xmm; 500| | sad16bi = sad16bi_xmm; 501| | sad8bi = sad8bi_xmm; 502| | dev16 = dev16_xmm; 503| | sad16v = sad16v_xmm; 504| | } 505| | 506| | if ((cpu_flags & XVID_CPU_3DNOW)) { 507| | 508| | /* Interpolation */ 509| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dn; 510| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dn; 511| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dn; 512| | 513| | interpolate8x4_halfpel_h = interpolate8x4_halfpel_h_3dn; 514| | interpolate8x4_halfpel_v = interpolate8x4_halfpel_v_3dn; 515| | interpolate8x4_halfpel_hv = interpolate8x4_halfpel_hv_3dn; 516| | } 517| | 518| | if ((cpu_flags & XVID_CPU_3DNOWEXT)) { 519| | 520| | /* Buffer transfer */ 521| | transfer_8to16copy = transfer_8to16copy_3dne; 522| | transfer_16to8copy = transfer_16to8copy_3dne; 523| | transfer_8to16sub = transfer_8to16sub_3dne; 524| | transfer_8to16subro = transfer_8to16subro_3dne; 525| | transfer_16to8add = transfer_16to8add_3dne; 526| | transfer8x8_copy = transfer8x8_copy_3dne; 527| | transfer8x4_copy = transfer8x4_copy_3dne; 528| | 529| | if ((cpu_flags & XVID_CPU_MMXEXT)) { 530| | /* Inverse DCT */ 531| | idct = idct_3dne; 532| | 533| | /* Buffer transfer */ 534| | transfer_8to16sub2 = transfer_8to16sub2_3dne; 535| | 536| | /* Interpolation */ 537| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dne; 538| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dne; 539| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dne; 540| | 541| | interpolate8x4_halfpel_h = interpolate8x4_halfpel_h_3dne; 542| | interpolate8x4_halfpel_v = interpolate8x4_halfpel_v_3dne; 543| | interpolate8x4_halfpel_hv = interpolate8x4_halfpel_hv_3dne; 544| | 545| | /* Quantization */ 546| | quant_h263_intra = quant_h263_intra_3dne; /* cmov only */ 547| | quant_h263_inter = quant_h263_inter_3dne; 548| | dequant_mpeg_intra = dequant_mpeg_intra_3dne; /* cmov only */ 549| | dequant_mpeg_inter = dequant_mpeg_inter_3dne; 550| | dequant_h263_intra = dequant_h263_intra_3dne; 551| | dequant_h263_inter = dequant_h263_inter_3dne; 552| | 553| | /* ME functions */ 554| | sad16 = sad16_3dne; 555| | sad8 = sad8_3dne; 556| | sad16bi = sad16bi_3dne; 557| | sad8bi = sad8bi_3dne; 558| | dev16 = dev16_3dne; 559| | } 560| | } 561| | 562| | if ((cpu_flags & XVID_CPU_SSE2)) { 563| | 564| | calc_cbp = calc_cbp_sse2; 565| | 566| | /* Quantization */ 567| | quant_h263_intra = quant_h263_intra_sse2; 568| | quant_h263_inter = quant_h263_inter_sse2; 569| | dequant_h263_intra = dequant_h263_intra_sse2; 570| | dequant_h263_inter = dequant_h263_inter_sse2; 571| | 572| | /* SAD operators */ 573| | sad16 = sad16_sse2; 574| | dev16 = dev16_sse2; 575| | 576| | /* PSNR-HVS-M distortion metric */ 577| | sseh8_16bit = sseh8_16bit_sse2; 578| | coeff8_energy = coeff8_energy_sse2; 579| | blocksum8 = blocksum8_sse2; 580| | 581| | /* DCT operators */ 582| | fdct = fdct_sse2_skal; 583| | idct = idct_sse2_skal; /* Is now IEEE1180 and Walken compliant. */ 584| | 585| | /* postprocessing */ 586| | image_brightness = image_brightness_sse2; 587| | 588| | } 589| | 590| | if ((cpu_flags & XVID_CPU_SSE3)) { 591| | 592| | /* SAD operators */ 593| | sad16 = sad16_sse3; 594| | dev16 = dev16_sse3; 595| | } 596| | 597| |#endif /* ARCH_IS_IA32 */ 598| | 599| |#if defined(ARCH_IS_IA64) 600| | if ((cpu_flags & XVID_CPU_ASM)) { /* use assembler routines? */ 601| | idct_ia64_init(); 602| | fdct = fdct_ia64; 603| | idct = idct_ia64; /*not yet working, crashes */ 604| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_ia64; 605| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_ia64; 606| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_ia64; 607| | sad16 = sad16_ia64; 608| | sad16bi = sad16bi_ia64; 609| | sad8 = sad8_ia64; 610| | dev16 = dev16_ia64; 611| |/* Halfpel8_Refine = Halfpel8_Refine_ia64; */ 612| | quant_h263_intra = quant_h263_intra_ia64; 613| | quant_h263_inter = quant_h263_inter_ia64; 614| | dequant_h263_intra = dequant_h263_intra_ia64; 615| | dequant_h263_inter = dequant_h263_inter_ia64; 616| | transfer_8to16copy = transfer_8to16copy_ia64; 617| | transfer_16to8copy = transfer_16to8copy_ia64; 618| | transfer_8to16sub = transfer_8to16sub_ia64; 619| | transfer_8to16sub2 = transfer_8to16sub2_ia64; 620| | transfer_16to8add = transfer_16to8add_ia64; 621| | transfer8x8_copy = transfer8x8_copy_ia64; 622| | } 623| |#endif 624| | 625| |#if defined(ARCH_IS_PPC) 626| | if ((cpu_flags & XVID_CPU_ALTIVEC)) { 627| | /* sad operators */ 628| | sad16 = sad16_altivec_c; 629| | sad16bi = sad16bi_altivec_c; 630| | sad8 = sad8_altivec_c; 631| | dev16 = dev16_altivec_c; 632| | 633| | sse8_16bit = sse8_16bit_altivec_c; 634| | 635| | /* mem transfer */ 636| | transfer_8to16copy = transfer_8to16copy_altivec_c; 637| | transfer_16to8copy = transfer_16to8copy_altivec_c; 638| | transfer_8to16sub = transfer_8to16sub_altivec_c; 639| | transfer_8to16subro = transfer_8to16subro_altivec_c; 640| | transfer_8to16sub2 = transfer_8to16sub2_altivec_c; 641| | transfer_16to8add = transfer_16to8add_altivec_c; 642| | transfer8x8_copy = transfer8x8_copy_altivec_c; 643| | 644| | /* Inverse DCT */ 645| | idct = idct_altivec_c; 646| | 647| | /* Interpolation */ 648| | interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_altivec_c; 649| | interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_altivec_c; 650| | interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_altivec_c; 651| | 652| | interpolate8x8_avg2 = interpolate8x8_avg2_altivec_c; 653| | interpolate8x8_avg4 = interpolate8x8_avg4_altivec_c; 654| | 655| | interpolate8x8_halfpel_add = interpolate8x8_halfpel_add_altivec_c; 656| | interpolate8x8_halfpel_h_add = interpolate8x8_halfpel_h_add_altivec_c; 657| | interpolate8x8_halfpel_v_add = interpolate8x8_halfpel_v_add_altivec_c; 658| | interpolate8x8_halfpel_hv_add = interpolate8x8_halfpel_hv_add_altivec_c; 659| | 660| | /* Colorspace conversion */ 661| | bgra_to_yv12 = bgra_to_yv12_altivec_c; 662| | abgr_to_yv12 = abgr_to_yv12_altivec_c; 663| | rgba_to_yv12 = rgba_to_yv12_altivec_c; 664| | argb_to_yv12 = argb_to_yv12_altivec_c; 665| | 666| | yuyv_to_yv12 = yuyv_to_yv12_altivec_c; 667| | uyvy_to_yv12 = uyvy_to_yv12_altivec_c; 668| | 669| | yv12_to_yuyv = yv12_to_yuyv_altivec_c; 670| | yv12_to_uyvy = yv12_to_uyvy_altivec_c; 671| | 672| | /* Quantization */ 673| | quant_h263_intra = quant_h263_intra_altivec_c; 674| | quant_h263_inter = quant_h263_inter_altivec_c; 675| | dequant_h263_intra = dequant_h263_intra_altivec_c; 676| | dequant_h263_inter = dequant_h263_inter_altivec_c; 677| | 678| | dequant_mpeg_intra = dequant_mpeg_intra_altivec_c; 679| | dequant_mpeg_inter = dequant_mpeg_inter_altivec_c; 680| | 681| | /* Qpel stuff */ 682| | xvid_QP_Funcs = &xvid_QP_Funcs_Altivec_C; 683| | xvid_QP_Add_Funcs = &xvid_QP_Add_Funcs_Altivec_C; 684| | } 685| |#endif 686| | 687| |#if defined(_DEBUG) 688| | xvid_debug = init->debug; 689| |#endif 690| | 691| 2| return(0); 692| 2|} xvid.c:detect_cpu_flags: 151| 2|{ 152| | /* enable native assembly optimizations by default */ 153| 2| unsigned int cpu_flags = XVID_CPU_ASM; ------------------ | | 181| 2|#define XVID_CPU_ASM (1<< 7) /* native assembly */ ------------------ 154| | 155| |#if defined(ARCH_IS_IA32) || defined(ARCH_IS_X86_64) 156| | cpu_flags |= check_cpu_features(); 157| | if ((cpu_flags & XVID_CPU_SSE) && sigill_check(sse_os_trigger)) 158| | cpu_flags &= ~XVID_CPU_SSE; 159| | 160| | if ((cpu_flags & (XVID_CPU_SSE2|XVID_CPU_SSE3|XVID_CPU_SSE41)) && sigill_check(sse2_os_trigger)) 161| | cpu_flags &= ~(XVID_CPU_SSE2|XVID_CPU_SSE3|XVID_CPU_SSE41); 162| |#endif 163| | 164| |#if defined(ARCH_IS_PPC) 165| |#if defined(__amigaos4__) 166| | { 167| | uint32_t vector_unit = VECTORTYPE_NONE; 168| | IExec->GetCPUInfoTags(GCIT_VectorUnit, &vector_unit, TAG_END); 169| | if (vector_unit == VECTORTYPE_ALTIVEC) { 170| | cpu_flags |= XVID_CPU_ALTIVEC; 171| | } 172| | } 173| |#else 174| | if (!sigill_check(altivec_trigger)) 175| | cpu_flags |= XVID_CPU_ALTIVEC; 176| |#endif 177| |#endif 178| | 179| 2| return cpu_flags; 180| 2|}