00001
00002
00003
00004
00005
00006
00007
00008
00009
00010
00011
00012
00013
00014
00015
00016
00017
00018
00019
00020
00021
00028 #include "vc1dsp.h"
00029
00030
00033 static void vc1_v_overlap_c(uint8_t* src, int stride)
00034 {
00035 int i;
00036 int a, b, c, d;
00037 int d1, d2;
00038 int rnd = 1;
00039 for(i = 0; i < 8; i++) {
00040 a = src[-2*stride];
00041 b = src[-stride];
00042 c = src[0];
00043 d = src[stride];
00044 d1 = (a - d + 3 + rnd) >> 3;
00045 d2 = (a - d + b - c + 4 - rnd) >> 3;
00046
00047 src[-2*stride] = a - d1;
00048 src[-stride] = av_clip_uint8(b - d2);
00049 src[0] = av_clip_uint8(c + d2);
00050 src[stride] = d + d1;
00051 src++;
00052 rnd = !rnd;
00053 }
00054 }
00055
00058 static void vc1_h_overlap_c(uint8_t* src, int stride)
00059 {
00060 int i;
00061 int a, b, c, d;
00062 int d1, d2;
00063 int rnd = 1;
00064 for(i = 0; i < 8; i++) {
00065 a = src[-2];
00066 b = src[-1];
00067 c = src[0];
00068 d = src[1];
00069 d1 = (a - d + 3 + rnd) >> 3;
00070 d2 = (a - d + b - c + 4 - rnd) >> 3;
00071
00072 src[-2] = a - d1;
00073 src[-1] = av_clip_uint8(b - d2);
00074 src[0] = av_clip_uint8(c + d2);
00075 src[1] = d + d1;
00076 src += stride;
00077 rnd = !rnd;
00078 }
00079 }
00080
00089 static av_always_inline int vc1_filter_line(uint8_t* src, int stride, int pq){
00090 uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
00091
00092 int a0 = (2*(src[-2*stride] - src[ 1*stride]) - 5*(src[-1*stride] - src[ 0*stride]) + 4) >> 3;
00093 int a0_sign = a0 >> 31;
00094 a0 = (a0 ^ a0_sign) - a0_sign;
00095 if(a0 < pq){
00096 int a1 = FFABS((2*(src[-4*stride] - src[-1*stride]) - 5*(src[-3*stride] - src[-2*stride]) + 4) >> 3);
00097 int a2 = FFABS((2*(src[ 0*stride] - src[ 3*stride]) - 5*(src[ 1*stride] - src[ 2*stride]) + 4) >> 3);
00098 if(a1 < a0 || a2 < a0){
00099 int clip = src[-1*stride] - src[ 0*stride];
00100 int clip_sign = clip >> 31;
00101 clip = ((clip ^ clip_sign) - clip_sign)>>1;
00102 if(clip){
00103 int a3 = FFMIN(a1, a2);
00104 int d = 5 * (a3 - a0);
00105 int d_sign = (d >> 31);
00106 d = ((d ^ d_sign) - d_sign) >> 3;
00107 d_sign ^= a0_sign;
00108
00109 if( d_sign ^ clip_sign )
00110 d = 0;
00111 else{
00112 d = FFMIN(d, clip);
00113 d = (d ^ d_sign) - d_sign;
00114 src[-1*stride] = cm[src[-1*stride] - d];
00115 src[ 0*stride] = cm[src[ 0*stride] + d];
00116 }
00117 return 1;
00118 }
00119 }
00120 }
00121 return 0;
00122 }
00123
00133 static inline void vc1_loop_filter(uint8_t* src, int step, int stride, int len, int pq)
00134 {
00135 int i;
00136 int filt3;
00137
00138 for(i = 0; i < len; i += 4){
00139 filt3 = vc1_filter_line(src + 2*step, stride, pq);
00140 if(filt3){
00141 vc1_filter_line(src + 0*step, stride, pq);
00142 vc1_filter_line(src + 1*step, stride, pq);
00143 vc1_filter_line(src + 3*step, stride, pq);
00144 }
00145 src += step * 4;
00146 }
00147 }
00148
00149 static void vc1_v_loop_filter4_c(uint8_t *src, int stride, int pq)
00150 {
00151 vc1_loop_filter(src, 1, stride, 4, pq);
00152 }
00153
00154 static void vc1_h_loop_filter4_c(uint8_t *src, int stride, int pq)
00155 {
00156 vc1_loop_filter(src, stride, 1, 4, pq);
00157 }
00158
00159 static void vc1_v_loop_filter8_c(uint8_t *src, int stride, int pq)
00160 {
00161 vc1_loop_filter(src, 1, stride, 8, pq);
00162 }
00163
00164 static void vc1_h_loop_filter8_c(uint8_t *src, int stride, int pq)
00165 {
00166 vc1_loop_filter(src, stride, 1, 8, pq);
00167 }
00168
00169 static void vc1_v_loop_filter16_c(uint8_t *src, int stride, int pq)
00170 {
00171 vc1_loop_filter(src, 1, stride, 16, pq);
00172 }
00173
00174 static void vc1_h_loop_filter16_c(uint8_t *src, int stride, int pq)
00175 {
00176 vc1_loop_filter(src, stride, 1, 16, pq);
00177 }
00178
00181 static void vc1_inv_trans_8x8_dc_c(uint8_t *dest, int linesize, DCTELEM *block)
00182 {
00183 int i;
00184 int dc = block[0];
00185 const uint8_t *cm;
00186 dc = (3 * dc + 1) >> 1;
00187 dc = (3 * dc + 16) >> 5;
00188 cm = ff_cropTbl + MAX_NEG_CROP + dc;
00189 for(i = 0; i < 8; i++){
00190 dest[0] = cm[dest[0]];
00191 dest[1] = cm[dest[1]];
00192 dest[2] = cm[dest[2]];
00193 dest[3] = cm[dest[3]];
00194 dest[4] = cm[dest[4]];
00195 dest[5] = cm[dest[5]];
00196 dest[6] = cm[dest[6]];
00197 dest[7] = cm[dest[7]];
00198 dest += linesize;
00199 }
00200 }
00201
00202 static av_always_inline void vc1_inv_trans_8x8_c(DCTELEM block[64], int shl, int sub)
00203 {
00204 int i;
00205 register int t1,t2,t3,t4,t5,t6,t7,t8;
00206 DCTELEM *src, *dst, temp[64];
00207
00208 src = block;
00209 dst = temp;
00210 for(i = 0; i < 8; i++){
00211 t1 = 12 * (src[ 0] + src[32]) + 4;
00212 t2 = 12 * (src[ 0] - src[32]) + 4;
00213 t3 = 16 * src[16] + 6 * src[48];
00214 t4 = 6 * src[16] - 16 * src[48];
00215
00216 t5 = t1 + t3;
00217 t6 = t2 + t4;
00218 t7 = t2 - t4;
00219 t8 = t1 - t3;
00220
00221 t1 = 16 * src[ 8] + 15 * src[24] + 9 * src[40] + 4 * src[56];
00222 t2 = 15 * src[ 8] - 4 * src[24] - 16 * src[40] - 9 * src[56];
00223 t3 = 9 * src[ 8] - 16 * src[24] + 4 * src[40] + 15 * src[56];
00224 t4 = 4 * src[ 8] - 9 * src[24] + 15 * src[40] - 16 * src[56];
00225
00226 dst[0] = (t5 + t1) >> 3;
00227 dst[1] = (t6 + t2) >> 3;
00228 dst[2] = (t7 + t3) >> 3;
00229 dst[3] = (t8 + t4) >> 3;
00230 dst[4] = (t8 - t4) >> 3;
00231 dst[5] = (t7 - t3) >> 3;
00232 dst[6] = (t6 - t2) >> 3;
00233 dst[7] = (t5 - t1) >> 3;
00234
00235 src += 1;
00236 dst += 8;
00237 }
00238
00239 src = temp;
00240 dst = block;
00241 for(i = 0; i < 8; i++){
00242 t1 = 12 * (src[ 0] + src[32]) + 64;
00243 t2 = 12 * (src[ 0] - src[32]) + 64;
00244 t3 = 16 * src[16] + 6 * src[48];
00245 t4 = 6 * src[16] - 16 * src[48];
00246
00247 t5 = t1 + t3;
00248 t6 = t2 + t4;
00249 t7 = t2 - t4;
00250 t8 = t1 - t3;
00251
00252 t1 = 16 * src[ 8] + 15 * src[24] + 9 * src[40] + 4 * src[56];
00253 t2 = 15 * src[ 8] - 4 * src[24] - 16 * src[40] - 9 * src[56];
00254 t3 = 9 * src[ 8] - 16 * src[24] + 4 * src[40] + 15 * src[56];
00255 t4 = 4 * src[ 8] - 9 * src[24] + 15 * src[40] - 16 * src[56];
00256
00257 dst[ 0] = (((t5 + t1 ) >> 7) - sub) << shl;
00258 dst[ 8] = (((t6 + t2 ) >> 7) - sub) << shl;
00259 dst[16] = (((t7 + t3 ) >> 7) - sub) << shl;
00260 dst[24] = (((t8 + t4 ) >> 7) - sub) << shl;
00261 dst[32] = (((t8 - t4 + 1) >> 7) - sub) << shl;
00262 dst[40] = (((t7 - t3 + 1) >> 7) - sub) << shl;
00263 dst[48] = (((t6 - t2 + 1) >> 7) - sub) << shl;
00264 dst[56] = (((t5 - t1 + 1) >> 7) - sub) << shl;
00265
00266 src++;
00267 dst++;
00268 }
00269 }
00270
00271 static void vc1_inv_trans_8x8_add_c(uint8_t *dest, int linesize, DCTELEM *block)
00272 {
00273 vc1_inv_trans_8x8_c(block, 0, 0);
00274 ff_add_pixels_clamped_c(block, dest, linesize);
00275 }
00276
00277 static void vc1_inv_trans_8x8_put_signed_c(uint8_t *dest, int linesize, DCTELEM *block)
00278 {
00279 vc1_inv_trans_8x8_c(block, 0, 0);
00280 ff_put_signed_pixels_clamped_c(block, dest, linesize);
00281 }
00282
00283 static void vc1_inv_trans_8x8_put_signed_rangered_c(uint8_t *dest, int linesize, DCTELEM *block)
00284 {
00285 vc1_inv_trans_8x8_c(block, 1, 0);
00286 ff_put_signed_pixels_clamped_c(block, dest, linesize);
00287 }
00288
00289 static void vc1_inv_trans_8x8_put_c(uint8_t *dest, int linesize, DCTELEM *block)
00290 {
00291 vc1_inv_trans_8x8_c(block, 0, 0);
00292 ff_put_pixels_clamped_c(block, dest, linesize);
00293 }
00294
00295 static void vc1_inv_trans_8x8_put_rangered_c(uint8_t *dest, int linesize, DCTELEM *block)
00296 {
00297 vc1_inv_trans_8x8_c(block, 1, 64);
00298 ff_put_pixels_clamped_c(block, dest, linesize);
00299 }
00300
00303 static void vc1_inv_trans_8x4_dc_c(uint8_t *dest, int linesize, DCTELEM *block)
00304 {
00305 int i;
00306 int dc = block[0];
00307 const uint8_t *cm;
00308 dc = ( 3 * dc + 1) >> 1;
00309 dc = (17 * dc + 64) >> 7;
00310 cm = ff_cropTbl + MAX_NEG_CROP + dc;
00311 for(i = 0; i < 4; i++){
00312 dest[0] = cm[dest[0]];
00313 dest[1] = cm[dest[1]];
00314 dest[2] = cm[dest[2]];
00315 dest[3] = cm[dest[3]];
00316 dest[4] = cm[dest[4]];
00317 dest[5] = cm[dest[5]];
00318 dest[6] = cm[dest[6]];
00319 dest[7] = cm[dest[7]];
00320 dest += linesize;
00321 }
00322 }
00323
00324 static void vc1_inv_trans_8x4_c(uint8_t *dest, int linesize, DCTELEM *block)
00325 {
00326 int i;
00327 register int t1,t2,t3,t4,t5,t6,t7,t8;
00328 DCTELEM *src, *dst;
00329 const uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
00330
00331 src = block;
00332 dst = block;
00333 for(i = 0; i < 4; i++){
00334 t1 = 12 * (src[0] + src[4]) + 4;
00335 t2 = 12 * (src[0] - src[4]) + 4;
00336 t3 = 16 * src[2] + 6 * src[6];
00337 t4 = 6 * src[2] - 16 * src[6];
00338
00339 t5 = t1 + t3;
00340 t6 = t2 + t4;
00341 t7 = t2 - t4;
00342 t8 = t1 - t3;
00343
00344 t1 = 16 * src[1] + 15 * src[3] + 9 * src[5] + 4 * src[7];
00345 t2 = 15 * src[1] - 4 * src[3] - 16 * src[5] - 9 * src[7];
00346 t3 = 9 * src[1] - 16 * src[3] + 4 * src[5] + 15 * src[7];
00347 t4 = 4 * src[1] - 9 * src[3] + 15 * src[5] - 16 * src[7];
00348
00349 dst[0] = (t5 + t1) >> 3;
00350 dst[1] = (t6 + t2) >> 3;
00351 dst[2] = (t7 + t3) >> 3;
00352 dst[3] = (t8 + t4) >> 3;
00353 dst[4] = (t8 - t4) >> 3;
00354 dst[5] = (t7 - t3) >> 3;
00355 dst[6] = (t6 - t2) >> 3;
00356 dst[7] = (t5 - t1) >> 3;
00357
00358 src += 8;
00359 dst += 8;
00360 }
00361
00362 src = block;
00363 for(i = 0; i < 8; i++){
00364 t1 = 17 * (src[ 0] + src[16]) + 64;
00365 t2 = 17 * (src[ 0] - src[16]) + 64;
00366 t3 = 22 * src[ 8] + 10 * src[24];
00367 t4 = 22 * src[24] - 10 * src[ 8];
00368
00369 dest[0*linesize] = cm[dest[0*linesize] + ((t1 + t3) >> 7)];
00370 dest[1*linesize] = cm[dest[1*linesize] + ((t2 - t4) >> 7)];
00371 dest[2*linesize] = cm[dest[2*linesize] + ((t2 + t4) >> 7)];
00372 dest[3*linesize] = cm[dest[3*linesize] + ((t1 - t3) >> 7)];
00373
00374 src ++;
00375 dest++;
00376 }
00377 }
00378
00381 static void vc1_inv_trans_4x8_dc_c(uint8_t *dest, int linesize, DCTELEM *block)
00382 {
00383 int i;
00384 int dc = block[0];
00385 const uint8_t *cm;
00386 dc = (17 * dc + 4) >> 3;
00387 dc = (12 * dc + 64) >> 7;
00388 cm = ff_cropTbl + MAX_NEG_CROP + dc;
00389 for(i = 0; i < 8; i++){
00390 dest[0] = cm[dest[0]];
00391 dest[1] = cm[dest[1]];
00392 dest[2] = cm[dest[2]];
00393 dest[3] = cm[dest[3]];
00394 dest += linesize;
00395 }
00396 }
00397
00398 static void vc1_inv_trans_4x8_c(uint8_t *dest, int linesize, DCTELEM *block)
00399 {
00400 int i;
00401 register int t1,t2,t3,t4,t5,t6,t7,t8;
00402 DCTELEM *src, *dst;
00403 const uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
00404
00405 src = block;
00406 dst = block;
00407 for(i = 0; i < 8; i++){
00408 t1 = 17 * (src[0] + src[2]) + 4;
00409 t2 = 17 * (src[0] - src[2]) + 4;
00410 t3 = 22 * src[1] + 10 * src[3];
00411 t4 = 22 * src[3] - 10 * src[1];
00412
00413 dst[0] = (t1 + t3) >> 3;
00414 dst[1] = (t2 - t4) >> 3;
00415 dst[2] = (t2 + t4) >> 3;
00416 dst[3] = (t1 - t3) >> 3;
00417
00418 src += 8;
00419 dst += 8;
00420 }
00421
00422 src = block;
00423 for(i = 0; i < 4; i++){
00424 t1 = 12 * (src[ 0] + src[32]) + 64;
00425 t2 = 12 * (src[ 0] - src[32]) + 64;
00426 t3 = 16 * src[16] + 6 * src[48];
00427 t4 = 6 * src[16] - 16 * src[48];
00428
00429 t5 = t1 + t3;
00430 t6 = t2 + t4;
00431 t7 = t2 - t4;
00432 t8 = t1 - t3;
00433
00434 t1 = 16 * src[ 8] + 15 * src[24] + 9 * src[40] + 4 * src[56];
00435 t2 = 15 * src[ 8] - 4 * src[24] - 16 * src[40] - 9 * src[56];
00436 t3 = 9 * src[ 8] - 16 * src[24] + 4 * src[40] + 15 * src[56];
00437 t4 = 4 * src[ 8] - 9 * src[24] + 15 * src[40] - 16 * src[56];
00438
00439 dest[0*linesize] = cm[dest[0*linesize] + ((t5 + t1) >> 7)];
00440 dest[1*linesize] = cm[dest[1*linesize] + ((t6 + t2) >> 7)];
00441 dest[2*linesize] = cm[dest[2*linesize] + ((t7 + t3) >> 7)];
00442 dest[3*linesize] = cm[dest[3*linesize] + ((t8 + t4) >> 7)];
00443 dest[4*linesize] = cm[dest[4*linesize] + ((t8 - t4 + 1) >> 7)];
00444 dest[5*linesize] = cm[dest[5*linesize] + ((t7 - t3 + 1) >> 7)];
00445 dest[6*linesize] = cm[dest[6*linesize] + ((t6 - t2 + 1) >> 7)];
00446 dest[7*linesize] = cm[dest[7*linesize] + ((t5 - t1 + 1) >> 7)];
00447
00448 src ++;
00449 dest++;
00450 }
00451 }
00452
00455 static void vc1_inv_trans_4x4_dc_c(uint8_t *dest, int linesize, DCTELEM *block)
00456 {
00457 int i;
00458 int dc = block[0];
00459 const uint8_t *cm;
00460 dc = (17 * dc + 4) >> 3;
00461 dc = (17 * dc + 64) >> 7;
00462 cm = ff_cropTbl + MAX_NEG_CROP + dc;
00463 for(i = 0; i < 4; i++){
00464 dest[0] = cm[dest[0]];
00465 dest[1] = cm[dest[1]];
00466 dest[2] = cm[dest[2]];
00467 dest[3] = cm[dest[3]];
00468 dest += linesize;
00469 }
00470 }
00471
00472 static void vc1_inv_trans_4x4_c(uint8_t *dest, int linesize, DCTELEM *block)
00473 {
00474 int i;
00475 register int t1,t2,t3,t4;
00476 DCTELEM *src, *dst;
00477 const uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
00478
00479 src = block;
00480 dst = block;
00481 for(i = 0; i < 4; i++){
00482 t1 = 17 * (src[0] + src[2]) + 4;
00483 t2 = 17 * (src[0] - src[2]) + 4;
00484 t3 = 22 * src[1] + 10 * src[3];
00485 t4 = 22 * src[3] - 10 * src[1];
00486
00487 dst[0] = (t1 + t3) >> 3;
00488 dst[1] = (t2 - t4) >> 3;
00489 dst[2] = (t2 + t4) >> 3;
00490 dst[3] = (t1 - t3) >> 3;
00491
00492 src += 8;
00493 dst += 8;
00494 }
00495
00496 src = block;
00497 for(i = 0; i < 4; i++){
00498 t1 = 17 * (src[ 0] + src[16]) + 64;
00499 t2 = 17 * (src[ 0] - src[16]) + 64;
00500 t3 = 22 * src[ 8] + 10 * src[24];
00501 t4 = 22 * src[24] - 10 * src[ 8];
00502
00503 dest[0*linesize] = cm[dest[0*linesize] + ((t1 + t3) >> 7)];
00504 dest[1*linesize] = cm[dest[1*linesize] + ((t2 - t4) >> 7)];
00505 dest[2*linesize] = cm[dest[2*linesize] + ((t2 + t4) >> 7)];
00506 dest[3*linesize] = cm[dest[3*linesize] + ((t1 - t3) >> 7)];
00507
00508 src ++;
00509 dest++;
00510 }
00511 }
00512
00513
00515 #define VC1_MSPEL_FILTER_16B(DIR, TYPE) \
00516 static av_always_inline int vc1_mspel_ ## DIR ## _filter_16bits(const TYPE *src, int stride, int mode) \
00517 { \
00518 switch(mode){ \
00519 case 0: \
00520 return 0; \
00521 case 1: \
00522 return -4*src[-stride] + 53*src[0] + 18*src[stride] - 3*src[stride*2]; \
00523 case 2: \
00524 return -src[-stride] + 9*src[0] + 9*src[stride] - src[stride*2]; \
00525 case 3: \
00526 return -3*src[-stride] + 18*src[0] + 53*src[stride] - 4*src[stride*2]; \
00527 } \
00528 return 0; \
00529 }
00530
00531 VC1_MSPEL_FILTER_16B(ver, uint8_t);
00532 VC1_MSPEL_FILTER_16B(hor, int16_t);
00533
00534
00537 static av_always_inline int vc1_mspel_filter(const uint8_t *src, int stride, int mode, int r)
00538 {
00539 switch(mode){
00540 case 0:
00541 return src[0];
00542 case 1:
00543 return (-4*src[-stride] + 53*src[0] + 18*src[stride] - 3*src[stride*2] + 32 - r) >> 6;
00544 case 2:
00545 return (-src[-stride] + 9*src[0] + 9*src[stride] - src[stride*2] + 8 - r) >> 4;
00546 case 3:
00547 return (-3*src[-stride] + 18*src[0] + 53*src[stride] - 4*src[stride*2] + 32 - r) >> 6;
00548 }
00549 return 0;
00550 }
00551
00554 #define VC1_MSPEL_MC(OP, OPNAME)\
00555 static void OPNAME ## vc1_mspel_mc(uint8_t *dst, const uint8_t *src, int stride, int hmode, int vmode, int rnd)\
00556 {\
00557 int i, j;\
00558 \
00559 if (vmode) { \
00560 int r;\
00561 \
00562 if (hmode) { \
00563 static const int shift_value[] = { 0, 5, 1, 5 };\
00564 int shift = (shift_value[hmode]+shift_value[vmode])>>1;\
00565 int16_t tmp[11*8], *tptr = tmp;\
00566 \
00567 r = (1<<(shift-1)) + rnd-1;\
00568 \
00569 src -= 1;\
00570 for(j = 0; j < 8; j++) {\
00571 for(i = 0; i < 11; i++)\
00572 tptr[i] = (vc1_mspel_ver_filter_16bits(src + i, stride, vmode)+r)>>shift;\
00573 src += stride;\
00574 tptr += 11;\
00575 }\
00576 \
00577 r = 64-rnd;\
00578 tptr = tmp+1;\
00579 for(j = 0; j < 8; j++) {\
00580 for(i = 0; i < 8; i++)\
00581 OP(dst[i], (vc1_mspel_hor_filter_16bits(tptr + i, 1, hmode)+r)>>7);\
00582 dst += stride;\
00583 tptr += 11;\
00584 }\
00585 \
00586 return;\
00587 }\
00588 else { \
00589 r = 1-rnd;\
00590 \
00591 for(j = 0; j < 8; j++) {\
00592 for(i = 0; i < 8; i++)\
00593 OP(dst[i], vc1_mspel_filter(src + i, stride, vmode, r));\
00594 src += stride;\
00595 dst += stride;\
00596 }\
00597 return;\
00598 }\
00599 }\
00600 \
00601 \
00602 for(j = 0; j < 8; j++) {\
00603 for(i = 0; i < 8; i++)\
00604 OP(dst[i], vc1_mspel_filter(src + i, 1, hmode, rnd));\
00605 dst += stride;\
00606 src += stride;\
00607 }\
00608 }
00609
00610 #define op_put(a, b) a = av_clip_uint8(b)
00611 #define op_avg(a, b) a = (a + av_clip_uint8(b) + 1) >> 1
00612
00613 VC1_MSPEL_MC(op_put, put_)
00614 VC1_MSPEL_MC(op_avg, avg_)
00615
00616
00617
00618 #define PUT_VC1_MSPEL(a, b)\
00619 static void put_vc1_mspel_mc ## a ## b ##_c(uint8_t *dst, const uint8_t *src, int stride, int rnd) { \
00620 put_vc1_mspel_mc(dst, src, stride, a, b, rnd); \
00621 }\
00622 static void avg_vc1_mspel_mc ## a ## b ##_c(uint8_t *dst, const uint8_t *src, int stride, int rnd) { \
00623 avg_vc1_mspel_mc(dst, src, stride, a, b, rnd); \
00624 }
00625
00626 PUT_VC1_MSPEL(1, 0)
00627 PUT_VC1_MSPEL(2, 0)
00628 PUT_VC1_MSPEL(3, 0)
00629
00630 PUT_VC1_MSPEL(0, 1)
00631 PUT_VC1_MSPEL(1, 1)
00632 PUT_VC1_MSPEL(2, 1)
00633 PUT_VC1_MSPEL(3, 1)
00634
00635 PUT_VC1_MSPEL(0, 2)
00636 PUT_VC1_MSPEL(1, 2)
00637 PUT_VC1_MSPEL(2, 2)
00638 PUT_VC1_MSPEL(3, 2)
00639
00640 PUT_VC1_MSPEL(0, 3)
00641 PUT_VC1_MSPEL(1, 3)
00642 PUT_VC1_MSPEL(2, 3)
00643 PUT_VC1_MSPEL(3, 3)
00644
00645 static void put_no_rnd_vc1_chroma_mc8_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y){
00646 const int A=(8-x)*(8-y);
00647 const int B=( x)*(8-y);
00648 const int C=(8-x)*( y);
00649 const int D=( x)*( y);
00650 int i;
00651
00652 assert(x<8 && y<8 && x>=0 && y>=0);
00653
00654 for(i=0; i<h; i++)
00655 {
00656 dst[0] = (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + 32 - 4) >> 6;
00657 dst[1] = (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + 32 - 4) >> 6;
00658 dst[2] = (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + 32 - 4) >> 6;
00659 dst[3] = (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + 32 - 4) >> 6;
00660 dst[4] = (A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5] + 32 - 4) >> 6;
00661 dst[5] = (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + 32 - 4) >> 6;
00662 dst[6] = (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + 32 - 4) >> 6;
00663 dst[7] = (A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8] + 32 - 4) >> 6;
00664 dst+= stride;
00665 src+= stride;
00666 }
00667 }
00668
00669 #define avg2(a,b) ((a+b+1)>>1)
00670 static void avg_no_rnd_vc1_chroma_mc8_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y){
00671 const int A=(8-x)*(8-y);
00672 const int B=( x)*(8-y);
00673 const int C=(8-x)*( y);
00674 const int D=( x)*( y);
00675 int i;
00676
00677 assert(x<8 && y<8 && x>=0 && y>=0);
00678
00679 for(i=0; i<h; i++)
00680 {
00681 dst[0] = avg2(dst[0], ((A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + 32 - 4) >> 6));
00682 dst[1] = avg2(dst[1], ((A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + 32 - 4) >> 6));
00683 dst[2] = avg2(dst[2], ((A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + 32 - 4) >> 6));
00684 dst[3] = avg2(dst[3], ((A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + 32 - 4) >> 6));
00685 dst[4] = avg2(dst[4], ((A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5] + 32 - 4) >> 6));
00686 dst[5] = avg2(dst[5], ((A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + 32 - 4) >> 6));
00687 dst[6] = avg2(dst[6], ((A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + 32 - 4) >> 6));
00688 dst[7] = avg2(dst[7], ((A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8] + 32 - 4) >> 6));
00689 dst+= stride;
00690 src+= stride;
00691 }
00692 }
00693
00694 av_cold void ff_vc1dsp_init(VC1DSPContext* dsp) {
00695 dsp->vc1_inv_trans_8x8_add = vc1_inv_trans_8x8_add_c;
00696 dsp->vc1_inv_trans_8x8_put_signed[0] = vc1_inv_trans_8x8_put_signed_c;
00697 dsp->vc1_inv_trans_8x8_put_signed[1] = vc1_inv_trans_8x8_put_signed_rangered_c;
00698 dsp->vc1_inv_trans_8x8_put[0] = vc1_inv_trans_8x8_put_c;
00699 dsp->vc1_inv_trans_8x8_put[1] = vc1_inv_trans_8x8_put_rangered_c;
00700 dsp->vc1_inv_trans_4x8 = vc1_inv_trans_4x8_c;
00701 dsp->vc1_inv_trans_8x4 = vc1_inv_trans_8x4_c;
00702 dsp->vc1_inv_trans_4x4 = vc1_inv_trans_4x4_c;
00703 dsp->vc1_inv_trans_8x8_dc = vc1_inv_trans_8x8_dc_c;
00704 dsp->vc1_inv_trans_4x8_dc = vc1_inv_trans_4x8_dc_c;
00705 dsp->vc1_inv_trans_8x4_dc = vc1_inv_trans_8x4_dc_c;
00706 dsp->vc1_inv_trans_4x4_dc = vc1_inv_trans_4x4_dc_c;
00707 dsp->vc1_h_overlap = vc1_h_overlap_c;
00708 dsp->vc1_v_overlap = vc1_v_overlap_c;
00709 dsp->vc1_v_loop_filter4 = vc1_v_loop_filter4_c;
00710 dsp->vc1_h_loop_filter4 = vc1_h_loop_filter4_c;
00711 dsp->vc1_v_loop_filter8 = vc1_v_loop_filter8_c;
00712 dsp->vc1_h_loop_filter8 = vc1_h_loop_filter8_c;
00713 dsp->vc1_v_loop_filter16 = vc1_v_loop_filter16_c;
00714 dsp->vc1_h_loop_filter16 = vc1_h_loop_filter16_c;
00715
00716 dsp->put_vc1_mspel_pixels_tab[ 0] = ff_put_pixels8x8_c;
00717 dsp->put_vc1_mspel_pixels_tab[ 1] = put_vc1_mspel_mc10_c;
00718 dsp->put_vc1_mspel_pixels_tab[ 2] = put_vc1_mspel_mc20_c;
00719 dsp->put_vc1_mspel_pixels_tab[ 3] = put_vc1_mspel_mc30_c;
00720 dsp->put_vc1_mspel_pixels_tab[ 4] = put_vc1_mspel_mc01_c;
00721 dsp->put_vc1_mspel_pixels_tab[ 5] = put_vc1_mspel_mc11_c;
00722 dsp->put_vc1_mspel_pixels_tab[ 6] = put_vc1_mspel_mc21_c;
00723 dsp->put_vc1_mspel_pixels_tab[ 7] = put_vc1_mspel_mc31_c;
00724 dsp->put_vc1_mspel_pixels_tab[ 8] = put_vc1_mspel_mc02_c;
00725 dsp->put_vc1_mspel_pixels_tab[ 9] = put_vc1_mspel_mc12_c;
00726 dsp->put_vc1_mspel_pixels_tab[10] = put_vc1_mspel_mc22_c;
00727 dsp->put_vc1_mspel_pixels_tab[11] = put_vc1_mspel_mc32_c;
00728 dsp->put_vc1_mspel_pixels_tab[12] = put_vc1_mspel_mc03_c;
00729 dsp->put_vc1_mspel_pixels_tab[13] = put_vc1_mspel_mc13_c;
00730 dsp->put_vc1_mspel_pixels_tab[14] = put_vc1_mspel_mc23_c;
00731 dsp->put_vc1_mspel_pixels_tab[15] = put_vc1_mspel_mc33_c;
00732
00733 dsp->avg_vc1_mspel_pixels_tab[ 0] = ff_avg_pixels8x8_c;
00734 dsp->avg_vc1_mspel_pixels_tab[ 1] = avg_vc1_mspel_mc10_c;
00735 dsp->avg_vc1_mspel_pixels_tab[ 2] = avg_vc1_mspel_mc20_c;
00736 dsp->avg_vc1_mspel_pixels_tab[ 3] = avg_vc1_mspel_mc30_c;
00737 dsp->avg_vc1_mspel_pixels_tab[ 4] = avg_vc1_mspel_mc01_c;
00738 dsp->avg_vc1_mspel_pixels_tab[ 5] = avg_vc1_mspel_mc11_c;
00739 dsp->avg_vc1_mspel_pixels_tab[ 6] = avg_vc1_mspel_mc21_c;
00740 dsp->avg_vc1_mspel_pixels_tab[ 7] = avg_vc1_mspel_mc31_c;
00741 dsp->avg_vc1_mspel_pixels_tab[ 8] = avg_vc1_mspel_mc02_c;
00742 dsp->avg_vc1_mspel_pixels_tab[ 9] = avg_vc1_mspel_mc12_c;
00743 dsp->avg_vc1_mspel_pixels_tab[10] = avg_vc1_mspel_mc22_c;
00744 dsp->avg_vc1_mspel_pixels_tab[11] = avg_vc1_mspel_mc32_c;
00745 dsp->avg_vc1_mspel_pixels_tab[12] = avg_vc1_mspel_mc03_c;
00746 dsp->avg_vc1_mspel_pixels_tab[13] = avg_vc1_mspel_mc13_c;
00747 dsp->avg_vc1_mspel_pixels_tab[14] = avg_vc1_mspel_mc23_c;
00748 dsp->avg_vc1_mspel_pixels_tab[15] = avg_vc1_mspel_mc33_c;
00749
00750 dsp->put_no_rnd_vc1_chroma_pixels_tab[0]= put_no_rnd_vc1_chroma_mc8_c;
00751 dsp->avg_no_rnd_vc1_chroma_pixels_tab[0]= avg_no_rnd_vc1_chroma_mc8_c;
00752
00753 if (HAVE_ALTIVEC)
00754 ff_vc1dsp_init_altivec(dsp);
00755 if (HAVE_MMX)
00756 ff_vc1dsp_init_mmx(dsp);
00757 }