FFmpeg
Loading...
Searching...
No Matches
h264pred_template.c
Go to the documentation of this file.
1/*
2 * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
3 * Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
4 *
5 * This file is part of FFmpeg.
6 *
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22/**
23 * @file
24 * H.264 / AVC / MPEG-4 part10 prediction functions.
25 * @author Michael Niedermayer <michaelni@gmx.at>
26 */
27
28#include "libavutil/intreadwrite.h"
29
30#include "mathops.h"
31
32#include "bit_depth_template.c"
33
34static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright,
35 ptrdiff_t _stride)
36{
37 pixel *src = (pixel*)_src;
38 int stride = _stride>>(sizeof(pixel)-1);
39 const pixel4 a= AV_RN4PA(src-stride);
40
41 AV_WN4PA(src+0*stride, a);
42 AV_WN4PA(src+1*stride, a);
43 AV_WN4PA(src+2*stride, a);
44 AV_WN4PA(src+3*stride, a);
45}
46
47static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright,
48 ptrdiff_t _stride)
49{
50 pixel *src = (pixel*)_src;
51 int stride = _stride>>(sizeof(pixel)-1);
52 AV_WN4PA(src+0*stride, PIXEL_SPLAT_X4(src[-1+0*stride]));
53 AV_WN4PA(src+1*stride, PIXEL_SPLAT_X4(src[-1+1*stride]));
54 AV_WN4PA(src+2*stride, PIXEL_SPLAT_X4(src[-1+2*stride]));
55 AV_WN4PA(src+3*stride, PIXEL_SPLAT_X4(src[-1+3*stride]));
56}
57
58static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright,
59 ptrdiff_t _stride)
60{
61 pixel *src = (pixel*)_src;
62 int stride = _stride>>(sizeof(pixel)-1);
63 const int dc= ( src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
64 + src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 4) >>3;
65 const pixel4 a = PIXEL_SPLAT_X4(dc);
66
67 AV_WN4PA(src+0*stride, a);
68 AV_WN4PA(src+1*stride, a);
69 AV_WN4PA(src+2*stride, a);
70 AV_WN4PA(src+3*stride, a);
71}
72
73static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright,
74 ptrdiff_t _stride)
75{
76 pixel *src = (pixel*)_src;
77 int stride = _stride>>(sizeof(pixel)-1);
78 const int dc= ( src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
79 const pixel4 a = PIXEL_SPLAT_X4(dc);
80
81 AV_WN4PA(src+0*stride, a);
82 AV_WN4PA(src+1*stride, a);
83 AV_WN4PA(src+2*stride, a);
84 AV_WN4PA(src+3*stride, a);
85}
86
87static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright,
88 ptrdiff_t _stride)
89{
90 pixel *src = (pixel*)_src;
91 int stride = _stride>>(sizeof(pixel)-1);
92 const int dc= ( src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
93 const pixel4 a = PIXEL_SPLAT_X4(dc);
94
95 AV_WN4PA(src+0*stride, a);
96 AV_WN4PA(src+1*stride, a);
97 AV_WN4PA(src+2*stride, a);
98 AV_WN4PA(src+3*stride, a);
99}
100
101static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright,
102 ptrdiff_t _stride)
103{
104 pixel *src = (pixel*)_src;
105 int stride = _stride>>(sizeof(pixel)-1);
106 const pixel4 a = PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1));
107
108 AV_WN4PA(src+0*stride, a);
109 AV_WN4PA(src+1*stride, a);
110 AV_WN4PA(src+2*stride, a);
111 AV_WN4PA(src+3*stride, a);
112}
113
114
115#define LOAD_TOP_RIGHT_EDGE\
116 av_unused const unsigned t4 = topright[0];\
117 av_unused const unsigned t5 = topright[1];\
118 av_unused const unsigned t6 = topright[2];\
119 av_unused const unsigned t7 = topright[3];\
120
121#define LOAD_DOWN_LEFT_EDGE\
122 av_unused const unsigned l4 = src[-1+4*stride];\
123 av_unused const unsigned l5 = src[-1+5*stride];\
124 av_unused const unsigned l6 = src[-1+6*stride];\
125 av_unused const unsigned l7 = src[-1+7*stride];\
126
127#define LOAD_LEFT_EDGE\
128 av_unused const unsigned l0 = src[-1+0*stride];\
129 av_unused const unsigned l1 = src[-1+1*stride];\
130 av_unused const unsigned l2 = src[-1+2*stride];\
131 av_unused const unsigned l3 = src[-1+3*stride];\
132
133#define LOAD_TOP_EDGE\
134 av_unused const unsigned t0 = src[ 0-1*stride];\
135 av_unused const unsigned t1 = src[ 1-1*stride];\
136 av_unused const unsigned t2 = src[ 2-1*stride];\
137 av_unused const unsigned t3 = src[ 3-1*stride];\
138
139static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright,
140 ptrdiff_t _stride)
141{
142 pixel *src = (pixel*)_src;
143 int stride = _stride>>(sizeof(pixel)-1);
144 const int lt= src[-1-1*stride];
145 LOAD_TOP_EDGE
146 LOAD_LEFT_EDGE
147
148 src[0+3*stride]=(l3 + 2*l2 + l1 + 2)>>2;
149 src[0+2*stride]=
150 src[1+3*stride]=(l2 + 2*l1 + l0 + 2)>>2;
151 src[0+1*stride]=
152 src[1+2*stride]=
153 src[2+3*stride]=(l1 + 2*l0 + lt + 2)>>2;
154 src[0+0*stride]=
155 src[1+1*stride]=
156 src[2+2*stride]=
157 src[3+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
158 src[1+0*stride]=
159 src[2+1*stride]=
160 src[3+2*stride]=(lt + 2*t0 + t1 + 2)>>2;
161 src[2+0*stride]=
162 src[3+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
163 src[3+0*stride]=(t1 + 2*t2 + t3 + 2)>>2;
164}
165
166static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright,
167 ptrdiff_t _stride)
168{
169 pixel *src = (pixel*)_src;
170 const pixel *topright = (const pixel*)_topright;
171 int stride = _stride>>(sizeof(pixel)-1);
172 LOAD_TOP_EDGE
173 LOAD_TOP_RIGHT_EDGE
174// LOAD_LEFT_EDGE
175
176 src[0+0*stride]=(t0 + t2 + 2*t1 + 2)>>2;
177 src[1+0*stride]=
178 src[0+1*stride]=(t1 + t3 + 2*t2 + 2)>>2;
179 src[2+0*stride]=
180 src[1+1*stride]=
181 src[0+2*stride]=(t2 + t4 + 2*t3 + 2)>>2;
182 src[3+0*stride]=
183 src[2+1*stride]=
184 src[1+2*stride]=
185 src[0+3*stride]=(t3 + t5 + 2*t4 + 2)>>2;
186 src[3+1*stride]=
187 src[2+2*stride]=
188 src[1+3*stride]=(t4 + t6 + 2*t5 + 2)>>2;
189 src[3+2*stride]=
190 src[2+3*stride]=(t5 + t7 + 2*t6 + 2)>>2;
191 src[3+3*stride]=(t6 + 3*t7 + 2)>>2;
192}
193
194static void FUNCC(pred4x4_vertical_right)(uint8_t *_src,
195 const uint8_t *topright,
196 ptrdiff_t _stride)
197{
198 pixel *src = (pixel*)_src;
199 int stride = _stride>>(sizeof(pixel)-1);
200 const int lt= src[-1-1*stride];
201 LOAD_TOP_EDGE
202 LOAD_LEFT_EDGE
203
204 src[0+0*stride]=
205 src[1+2*stride]=(lt + t0 + 1)>>1;
206 src[1+0*stride]=
207 src[2+2*stride]=(t0 + t1 + 1)>>1;
208 src[2+0*stride]=
209 src[3+2*stride]=(t1 + t2 + 1)>>1;
210 src[3+0*stride]=(t2 + t3 + 1)>>1;
211 src[0+1*stride]=
212 src[1+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
213 src[1+1*stride]=
214 src[2+3*stride]=(lt + 2*t0 + t1 + 2)>>2;
215 src[2+1*stride]=
216 src[3+3*stride]=(t0 + 2*t1 + t2 + 2)>>2;
217 src[3+1*stride]=(t1 + 2*t2 + t3 + 2)>>2;
218 src[0+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
219 src[0+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
220}
221
222static void FUNCC(pred4x4_vertical_left)(uint8_t *_src,
223 const uint8_t *_topright,
224 ptrdiff_t _stride)
225{
226 pixel *src = (pixel*)_src;
227 const pixel *topright = (const pixel*)_topright;
228 int stride = _stride>>(sizeof(pixel)-1);
229 LOAD_TOP_EDGE
230 LOAD_TOP_RIGHT_EDGE
231
232 src[0+0*stride]=(t0 + t1 + 1)>>1;
233 src[1+0*stride]=
234 src[0+2*stride]=(t1 + t2 + 1)>>1;
235 src[2+0*stride]=
236 src[1+2*stride]=(t2 + t3 + 1)>>1;
237 src[3+0*stride]=
238 src[2+2*stride]=(t3 + t4+ 1)>>1;
239 src[3+2*stride]=(t4 + t5+ 1)>>1;
240 src[0+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
241 src[1+1*stride]=
242 src[0+3*stride]=(t1 + 2*t2 + t3 + 2)>>2;
243 src[2+1*stride]=
244 src[1+3*stride]=(t2 + 2*t3 + t4 + 2)>>2;
245 src[3+1*stride]=
246 src[2+3*stride]=(t3 + 2*t4 + t5 + 2)>>2;
247 src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
248}
249
250static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright,
251 ptrdiff_t _stride)
252{
253 pixel *src = (pixel*)_src;
254 int stride = _stride>>(sizeof(pixel)-1);
255 LOAD_LEFT_EDGE
256
257 src[0+0*stride]=(l0 + l1 + 1)>>1;
258 src[1+0*stride]=(l0 + 2*l1 + l2 + 2)>>2;
259 src[2+0*stride]=
260 src[0+1*stride]=(l1 + l2 + 1)>>1;
261 src[3+0*stride]=
262 src[1+1*stride]=(l1 + 2*l2 + l3 + 2)>>2;
263 src[2+1*stride]=
264 src[0+2*stride]=(l2 + l3 + 1)>>1;
265 src[3+1*stride]=
266 src[1+2*stride]=(l2 + 2*l3 + l3 + 2)>>2;
267 src[3+2*stride]=
268 src[1+3*stride]=
269 src[0+3*stride]=
270 src[2+2*stride]=
271 src[2+3*stride]=
272 src[3+3*stride]=l3;
273}
274
275static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src,
276 const uint8_t *topright,
277 ptrdiff_t _stride)
278{
279 pixel *src = (pixel*)_src;
280 int stride = _stride>>(sizeof(pixel)-1);
281 const int lt= src[-1-1*stride];
282 LOAD_TOP_EDGE
283 LOAD_LEFT_EDGE
284
285 src[0+0*stride]=
286 src[2+1*stride]=(lt + l0 + 1)>>1;
287 src[1+0*stride]=
288 src[3+1*stride]=(l0 + 2*lt + t0 + 2)>>2;
289 src[2+0*stride]=(lt + 2*t0 + t1 + 2)>>2;
290 src[3+0*stride]=(t0 + 2*t1 + t2 + 2)>>2;
291 src[0+1*stride]=
292 src[2+2*stride]=(l0 + l1 + 1)>>1;
293 src[1+1*stride]=
294 src[3+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
295 src[0+2*stride]=
296 src[2+3*stride]=(l1 + l2+ 1)>>1;
297 src[1+2*stride]=
298 src[3+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
299 src[0+3*stride]=(l2 + l3 + 1)>>1;
300 src[1+3*stride]=(l1 + 2*l2 + l3 + 2)>>2;
301}
302
303static void FUNCC(pred16x16_vertical)(uint8_t *_src, ptrdiff_t _stride)
304{
305 int i;
306 pixel *src = (pixel*)_src;
307 int stride = _stride>>(sizeof(pixel)-1);
308 const pixel4 a = AV_RN4PA(((pixel4*)(src-stride))+0);
309 const pixel4 b = AV_RN4PA(((pixel4*)(src-stride))+1);
310 const pixel4 c = AV_RN4PA(((pixel4*)(src-stride))+2);
311 const pixel4 d = AV_RN4PA(((pixel4*)(src-stride))+3);
312
313 for(i=0; i<16; i++){
314 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
315 AV_WN4PA(((pixel4*)(src+i*stride))+1, b);
316 AV_WN4PA(((pixel4*)(src+i*stride))+2, c);
317 AV_WN4PA(((pixel4*)(src+i*stride))+3, d);
318 }
319}
320
321static void FUNCC(pred16x16_horizontal)(uint8_t *_src, ptrdiff_t stride)
322{
323 int i;
324 pixel *src = (pixel*)_src;
325 stride >>= sizeof(pixel)-1;
326
327 for(i=0; i<16; i++){
328 const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
329
330 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
331 AV_WN4PA(((pixel4*)(src+i*stride))+1, a);
332 AV_WN4PA(((pixel4*)(src+i*stride))+2, a);
333 AV_WN4PA(((pixel4*)(src+i*stride))+3, a);
334 }
335}
336
337#define PREDICT_16x16_DC(v)\
338 for(i=0; i<16; i++){\
339 AV_WN4PA(src+ 0, v);\
340 AV_WN4PA(src+ 4, v);\
341 AV_WN4PA(src+ 8, v);\
342 AV_WN4PA(src+12, v);\
343 src += stride;\
344 }
345
346static void FUNCC(pred16x16_dc)(uint8_t *_src, ptrdiff_t stride)
347{
348 int i, dc=0;
349 pixel *src = (pixel*)_src;
350 pixel4 dcsplat;
351 stride >>= sizeof(pixel)-1;
352
353 for(i=0;i<16; i++){
354 dc+= src[-1+i*stride];
355 }
356
357 for(i=0;i<16; i++){
358 dc+= src[i-stride];
359 }
360
361 dcsplat = PIXEL_SPLAT_X4((dc+16)>>5);
362 PREDICT_16x16_DC(dcsplat);
363}
364
365static void FUNCC(pred16x16_left_dc)(uint8_t *_src, ptrdiff_t stride)
366{
367 int i, dc=0;
368 pixel *src = (pixel*)_src;
369 pixel4 dcsplat;
370 stride >>= sizeof(pixel)-1;
371
372 for(i=0;i<16; i++){
373 dc+= src[-1+i*stride];
374 }
375
376 dcsplat = PIXEL_SPLAT_X4((dc+8)>>4);
377 PREDICT_16x16_DC(dcsplat);
378}
379
380static void FUNCC(pred16x16_top_dc)(uint8_t *_src, ptrdiff_t stride)
381{
382 int i, dc=0;
383 pixel *src = (pixel*)_src;
384 pixel4 dcsplat;
385 stride >>= sizeof(pixel)-1;
386
387 for(i=0;i<16; i++){
388 dc+= src[i-stride];
389 }
390
391 dcsplat = PIXEL_SPLAT_X4((dc+8)>>4);
392 PREDICT_16x16_DC(dcsplat);
393}
394
395#define PRED16x16_X(n, v) \
396static void FUNCC(pred16x16_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
397{\
398 int i;\
399 pixel *src = (pixel*)_src;\
400 stride >>= sizeof(pixel)-1;\
401 PREDICT_16x16_DC(PIXEL_SPLAT_X4(v));\
402}
403
404PRED16x16_X(128, (1<<(BIT_DEPTH-1))+0)
405#if BIT_DEPTH == 8
406PRED16x16_X(127, (1<<(BIT_DEPTH-1))-1)
407PRED16x16_X(129, (1<<(BIT_DEPTH-1))+1)
408#endif
409
410static inline void FUNCC(pred16x16_plane_compat)(uint8_t *_src,
411 ptrdiff_t _stride,
412 const int svq3,
413 const int rv40)
414{
415 int i, j, k;
416 int a;
417 pixel *src = (pixel*)_src;
418 int stride = _stride>>(sizeof(pixel)-1);
419 const pixel * const src0 = src +7-stride;
420 const pixel * src1 = src +8*stride-1;
421 const pixel * src2 = src1-2*stride; // == src+6*stride-1;
422 int H = src0[1] - src0[-1];
423 int V = src1[0] - src2[ 0];
424 for(k=2; k<=8; ++k) {
425 src1 += stride; src2 -= stride;
426 H += k*(src0[k] - src0[-k]);
427 V += k*(src1[0] - src2[ 0]);
428 }
429 if(svq3){
430 H = ( 5*(H/4) ) / 16;
431 V = ( 5*(V/4) ) / 16;
432
433 /* required for 100% accuracy */
434 i = H; H = V; V = i;
435 }else if(rv40){
436 H = ( H + (H>>2) ) >> 4;
437 V = ( V + (V>>2) ) >> 4;
438 }else{
439 H = ( 5*H+32 ) >> 6;
440 V = ( 5*V+32 ) >> 6;
441 }
442
443 a = 16*(src1[0] + src2[16] + 1) - 7*(V+H);
444 for(j=16; j>0; --j) {
445 int b = a;
446 a += V;
447 for(i=-16; i<0; i+=4) {
448 src[16+i] = CLIP((b ) >> 5);
449 src[17+i] = CLIP((b+ H) >> 5);
450 src[18+i] = CLIP((b+2*H) >> 5);
451 src[19+i] = CLIP((b+3*H) >> 5);
452 b += 4*H;
453 }
454 src += stride;
455 }
456}
457
458static void FUNCC(pred16x16_plane)(uint8_t *src, ptrdiff_t stride)
459{
460 FUNCC(pred16x16_plane_compat)(src, stride, 0, 0);
461}
462
463static void FUNCC(pred8x8_vertical)(uint8_t *_src, ptrdiff_t _stride)
464{
465 int i;
466 pixel *src = (pixel*)_src;
467 int stride = _stride>>(sizeof(pixel)-1);
468 const pixel4 a= AV_RN4PA(((pixel4*)(src-stride))+0);
469 const pixel4 b= AV_RN4PA(((pixel4*)(src-stride))+1);
470
471 for(i=0; i<8; i++){
472 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
473 AV_WN4PA(((pixel4*)(src+i*stride))+1, b);
474 }
475}
476
477static void FUNCC(pred8x16_vertical)(uint8_t *_src, ptrdiff_t _stride)
478{
479 int i;
480 pixel *src = (pixel*)_src;
481 int stride = _stride>>(sizeof(pixel)-1);
482 const pixel4 a= AV_RN4PA(((pixel4*)(src-stride))+0);
483 const pixel4 b= AV_RN4PA(((pixel4*)(src-stride))+1);
484
485 for(i=0; i<16; i++){
486 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
487 AV_WN4PA(((pixel4*)(src+i*stride))+1, b);
488 }
489}
490
491static void FUNCC(pred8x8_horizontal)(uint8_t *_src, ptrdiff_t stride)
492{
493 int i;
494 pixel *src = (pixel*)_src;
495 stride >>= sizeof(pixel)-1;
496
497 for(i=0; i<8; i++){
498 const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
499 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
500 AV_WN4PA(((pixel4*)(src+i*stride))+1, a);
501 }
502}
503
504static void FUNCC(pred8x16_horizontal)(uint8_t *_src, ptrdiff_t stride)
505{
506 int i;
507 pixel *src = (pixel*)_src;
508 stride >>= sizeof(pixel)-1;
509 for(i=0; i<16; i++){
510 const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
511 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);
512 AV_WN4PA(((pixel4*)(src+i*stride))+1, a);
513 }
514}
515
516#define PRED8x8_X(n, v)\
517static void FUNCC(pred8x8_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
518{\
519 int i;\
520 const pixel4 a = PIXEL_SPLAT_X4(v);\
521 pixel *src = (pixel*)_src;\
522 stride >>= sizeof(pixel)-1;\
523 for(i=0; i<8; i++){\
524 AV_WN4PA(((pixel4*)(src+i*stride))+0, a);\
525 AV_WN4PA(((pixel4*)(src+i*stride))+1, a);\
526 }\
527}
528
529PRED8x8_X(128, (1<<(BIT_DEPTH-1))+0)
530#if BIT_DEPTH == 8
531PRED8x8_X(127, (1<<(BIT_DEPTH-1))-1)
532PRED8x8_X(129, (1<<(BIT_DEPTH-1))+1)
533#endif
534
535static void FUNCC(pred8x16_128_dc)(uint8_t *_src, ptrdiff_t stride)
536{
537 FUNCC(pred8x8_128_dc)(_src, stride);
538 FUNCC(pred8x8_128_dc)(_src+8*stride, stride);
539}
540
541static void FUNCC(pred8x8_left_dc)(uint8_t *_src, ptrdiff_t stride)
542{
543 int i;
544 int dc0, dc2;
545 pixel4 dc0splat, dc2splat;
546 pixel *src = (pixel*)_src;
547 stride >>= sizeof(pixel)-1;
548
549 dc0=dc2=0;
550 for(i=0;i<4; i++){
551 dc0+= src[-1+i*stride];
552 dc2+= src[-1+(i+4)*stride];
553 }
554 dc0splat = PIXEL_SPLAT_X4((dc0 + 2)>>2);
555 dc2splat = PIXEL_SPLAT_X4((dc2 + 2)>>2);
556
557 for(i=0; i<4; i++){
558 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
559 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc0splat);
560 }
561 for(i=4; i<8; i++){
562 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc2splat);
563 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc2splat);
564 }
565}
566
567static void FUNCC(pred8x16_left_dc)(uint8_t *_src, ptrdiff_t stride)
568{
569 FUNCC(pred8x8_left_dc)(_src, stride);
570 FUNCC(pred8x8_left_dc)(_src+8*stride, stride);
571}
572
573static void FUNCC(pred8x8_top_dc)(uint8_t *_src, ptrdiff_t stride)
574{
575 int i;
576 int dc0, dc1;
577 pixel4 dc0splat, dc1splat;
578 pixel *src = (pixel*)_src;
579 stride >>= sizeof(pixel)-1;
580
581 dc0=dc1=0;
582 for(i=0;i<4; i++){
583 dc0+= src[i-stride];
584 dc1+= src[4+i-stride];
585 }
586 dc0splat = PIXEL_SPLAT_X4((dc0 + 2)>>2);
587 dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
588
589 for(i=0; i<4; i++){
590 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
591 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc1splat);
592 }
593 for(i=4; i<8; i++){
594 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
595 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc1splat);
596 }
597}
598
599static void FUNCC(pred8x16_top_dc)(uint8_t *_src, ptrdiff_t stride)
600{
601 int i;
602 int dc0, dc1;
603 pixel4 dc0splat, dc1splat;
604 pixel *src = (pixel*)_src;
605 stride >>= sizeof(pixel)-1;
606
607 dc0=dc1=0;
608 for(i=0;i<4; i++){
609 dc0+= src[i-stride];
610 dc1+= src[4+i-stride];
611 }
612 dc0splat = PIXEL_SPLAT_X4((dc0 + 2)>>2);
613 dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
614
615 for(i=0; i<16; i++){
616 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
617 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc1splat);
618 }
619}
620
621static void FUNCC(pred8x8_dc)(uint8_t *_src, ptrdiff_t stride)
622{
623 int i;
624 int dc0, dc1, dc2;
625 pixel4 dc0splat, dc1splat, dc2splat, dc3splat;
626 pixel *src = (pixel*)_src;
627 stride >>= sizeof(pixel)-1;
628
629 dc0=dc1=dc2=0;
630 for(i=0;i<4; i++){
631 dc0+= src[-1+i*stride] + src[i-stride];
632 dc1+= src[4+i-stride];
633 dc2+= src[-1+(i+4)*stride];
634 }
635 dc0splat = PIXEL_SPLAT_X4((dc0 + 4)>>3);
636 dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
637 dc2splat = PIXEL_SPLAT_X4((dc2 + 2)>>2);
638 dc3splat = PIXEL_SPLAT_X4((dc1 + dc2 + 4)>>3);
639
640 for(i=0; i<4; i++){
641 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
642 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc1splat);
643 }
644 for(i=4; i<8; i++){
645 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc2splat);
646 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc3splat);
647 }
648}
649
650static void FUNCC(pred8x16_dc)(uint8_t *_src, ptrdiff_t stride)
651{
652 int i;
653 int dc0, dc1, dc2, dc3, dc4;
654 pixel4 dc0splat, dc1splat, dc2splat, dc3splat, dc4splat, dc5splat, dc6splat, dc7splat;
655 pixel *src = (pixel*)_src;
656 stride >>= sizeof(pixel)-1;
657
658 dc0=dc1=dc2=dc3=dc4=0;
659 for(i=0;i<4; i++){
660 dc0+= src[-1+i*stride] + src[i-stride];
661 dc1+= src[4+i-stride];
662 dc2+= src[-1+(i+4)*stride];
663 dc3+= src[-1+(i+8)*stride];
664 dc4+= src[-1+(i+12)*stride];
665 }
666 dc0splat = PIXEL_SPLAT_X4((dc0 + 4)>>3);
667 dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
668 dc2splat = PIXEL_SPLAT_X4((dc2 + 2)>>2);
669 dc3splat = PIXEL_SPLAT_X4((dc1 + dc2 + 4)>>3);
670 dc4splat = PIXEL_SPLAT_X4((dc3 + 2)>>2);
671 dc5splat = PIXEL_SPLAT_X4((dc1 + dc3 + 4)>>3);
672 dc6splat = PIXEL_SPLAT_X4((dc4 + 2)>>2);
673 dc7splat = PIXEL_SPLAT_X4((dc1 + dc4 + 4)>>3);
674
675 for(i=0; i<4; i++){
676 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc0splat);
677 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc1splat);
678 }
679 for(i=4; i<8; i++){
680 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc2splat);
681 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc3splat);
682 }
683 for(i=8; i<12; i++){
684 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc4splat);
685 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc5splat);
686 }
687 for(i=12; i<16; i++){
688 AV_WN4PA(((pixel4*)(src+i*stride))+0, dc6splat);
689 AV_WN4PA(((pixel4*)(src+i*stride))+1, dc7splat);
690 }
691}
692
693//the following 4 function should not be optimized!
694static void FUNC(pred8x8_mad_cow_dc_l0t)(uint8_t *src, ptrdiff_t stride)
695{
696 FUNCC(pred8x8_top_dc)(src, stride);
697 FUNCC(pred4x4_dc)(src, NULL, stride);
698}
699
700static void FUNC(pred8x16_mad_cow_dc_l0t)(uint8_t *src, ptrdiff_t stride)
701{
702 FUNCC(pred8x16_top_dc)(src, stride);
703 FUNCC(pred4x4_dc)(src, NULL, stride);
704}
705
706static void FUNC(pred8x8_mad_cow_dc_0lt)(uint8_t *src, ptrdiff_t stride)
707{
708 FUNCC(pred8x8_dc)(src, stride);
709 FUNCC(pred4x4_top_dc)(src, NULL, stride);
710}
711
712static void FUNC(pred8x16_mad_cow_dc_0lt)(uint8_t *src, ptrdiff_t stride)
713{
714 FUNCC(pred8x16_dc)(src, stride);
715 FUNCC(pred4x4_top_dc)(src, NULL, stride);
716}
717
718static void FUNC(pred8x8_mad_cow_dc_l00)(uint8_t *src, ptrdiff_t stride)
719{
720 FUNCC(pred8x8_left_dc)(src, stride);
721 FUNCC(pred4x4_128_dc)(src + 4*stride , NULL, stride);
722 FUNCC(pred4x4_128_dc)(src + 4*stride + 4*sizeof(pixel), NULL, stride);
723}
724
725static void FUNC(pred8x16_mad_cow_dc_l00)(uint8_t *src, ptrdiff_t stride)
726{
727 FUNCC(pred8x16_left_dc)(src, stride);
728 FUNCC(pred4x4_128_dc)(src + 4*stride , NULL, stride);
729 FUNCC(pred4x4_128_dc)(src + 4*stride + 4*sizeof(pixel), NULL, stride);
730}
731
732static void FUNC(pred8x8_mad_cow_dc_0l0)(uint8_t *src, ptrdiff_t stride)
733{
734 FUNCC(pred8x8_left_dc)(src, stride);
735 FUNCC(pred4x4_128_dc)(src , NULL, stride);
736 FUNCC(pred4x4_128_dc)(src + 4*sizeof(pixel), NULL, stride);
737}
738
739static void FUNC(pred8x16_mad_cow_dc_0l0)(uint8_t *src, ptrdiff_t stride)
740{
741 FUNCC(pred8x16_left_dc)(src, stride);
742 FUNCC(pred4x4_128_dc)(src , NULL, stride);
743 FUNCC(pred4x4_128_dc)(src + 4*sizeof(pixel), NULL, stride);
744}
745
746static void FUNCC(pred8x8_plane)(uint8_t *_src, ptrdiff_t _stride)
747{
748 int j, k;
749 int a;
750 pixel *src = (pixel*)_src;
751 int stride = _stride>>(sizeof(pixel)-1);
752 const pixel * const src0 = src +3-stride;
753 const pixel * src1 = src +4*stride-1;
754 const pixel * src2 = src1-2*stride; // == src+2*stride-1;
755 int H = src0[1] - src0[-1];
756 int V = src1[0] - src2[ 0];
757 for(k=2; k<=4; ++k) {
758 src1 += stride; src2 -= stride;
759 H += k*(src0[k] - src0[-k]);
760 V += k*(src1[0] - src2[ 0]);
761 }
762 H = ( 17*H+16 ) >> 5;
763 V = ( 17*V+16 ) >> 5;
764
765 a = 16*(src1[0] + src2[8]+1) - 3*(V+H);
766 for(j=8; j>0; --j) {
767 int b = a;
768 a += V;
769 src[0] = CLIP((b ) >> 5);
770 src[1] = CLIP((b+ H) >> 5);
771 src[2] = CLIP((b+2*H) >> 5);
772 src[3] = CLIP((b+3*H) >> 5);
773 src[4] = CLIP((b+4*H) >> 5);
774 src[5] = CLIP((b+5*H) >> 5);
775 src[6] = CLIP((b+6*H) >> 5);
776 src[7] = CLIP((b+7*H) >> 5);
777 src += stride;
778 }
779}
780
781static void FUNCC(pred8x16_plane)(uint8_t *_src, ptrdiff_t _stride)
782{
783 int j, k;
784 int a;
785 pixel *src = (pixel*)_src;
786 int stride = _stride>>(sizeof(pixel)-1);
787 const pixel * const src0 = src +3-stride;
788 const pixel * src1 = src +8*stride-1;
789 const pixel * src2 = src1-2*stride; // == src+6*stride-1;
790 int H = src0[1] - src0[-1];
791 int V = src1[0] - src2[ 0];
792
793 for (k = 2; k <= 4; ++k) {
794 src1 += stride; src2 -= stride;
795 H += k*(src0[k] - src0[-k]);
796 V += k*(src1[0] - src2[ 0]);
797 }
798 for (; k <= 8; ++k) {
799 src1 += stride; src2 -= stride;
800 V += k*(src1[0] - src2[0]);
801 }
802
803 H = (17*H+16) >> 5;
804 V = (5*V+32) >> 6;
805
806 a = 16*(src1[0] + src2[8] + 1) - 7*V - 3*H;
807 for(j=16; j>0; --j) {
808 int b = a;
809 a += V;
810 src[0] = CLIP((b ) >> 5);
811 src[1] = CLIP((b+ H) >> 5);
812 src[2] = CLIP((b+2*H) >> 5);
813 src[3] = CLIP((b+3*H) >> 5);
814 src[4] = CLIP((b+4*H) >> 5);
815 src[5] = CLIP((b+5*H) >> 5);
816 src[6] = CLIP((b+6*H) >> 5);
817 src[7] = CLIP((b+7*H) >> 5);
818 src += stride;
819 }
820}
821
822#define SRC(x,y) src[(x)+(y)*stride]
823#define PL(y) \
824 const int l##y = (SRC(-1,y-1) + 2*SRC(-1,y) + SRC(-1,y+1) + 2) >> 2;
825#define PREDICT_8x8_LOAD_LEFT \
826 const int l0 = ((has_topleft ? SRC(-1,-1) : SRC(-1,0)) \
827 + 2*SRC(-1,0) + SRC(-1,1) + 2) >> 2; \
828 PL(1) PL(2) PL(3) PL(4) PL(5) PL(6) \
829 av_unused const int l7 = (SRC(-1,6) + 3*SRC(-1,7) + 2) >> 2
830
831#define PT(x) \
832 const int t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
833#define PREDICT_8x8_LOAD_TOP \
834 const int t0 = ((has_topleft ? SRC(-1,-1) : SRC(0,-1)) \
835 + 2*SRC(0,-1) + SRC(1,-1) + 2) >> 2; \
836 PT(1) PT(2) PT(3) PT(4) PT(5) PT(6) \
837 av_unused const int t7 = ((has_topright ? SRC(8,-1) : SRC(7,-1)) \
838 + 2*SRC(7,-1) + SRC(6,-1) + 2) >> 2
839
840#define PTR(x) \
841 t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
842#define PREDICT_8x8_LOAD_TOPRIGHT \
843 int t8, t9, t10, t11, t12, t13, t14, t15; \
844 if(has_topright) { \
845 PTR(8) PTR(9) PTR(10) PTR(11) PTR(12) PTR(13) PTR(14) \
846 t15 = (SRC(14,-1) + 3*SRC(15,-1) + 2) >> 2; \
847 } else t8=t9=t10=t11=t12=t13=t14=t15= SRC(7,-1);
848
849#define PREDICT_8x8_LOAD_TOPLEFT \
850 const int lt = (SRC(-1,0) + 2*SRC(-1,-1) + SRC(0,-1) + 2) >> 2
851
852#define PREDICT_8x8_DC(v) \
853 int y; \
854 for( y = 0; y < 8; y++ ) { \
855 AV_WN4PA(((pixel4*)src)+0, v); \
856 AV_WN4PA(((pixel4*)src)+1, v); \
857 src += stride; \
858 }
859
860static void FUNCC(pred8x8l_128_dc)(uint8_t *_src, int has_topleft,
861 int has_topright, ptrdiff_t _stride)
862{
863 pixel *src = (pixel*)_src;
864 int stride = _stride>>(sizeof(pixel)-1);
865
866 PREDICT_8x8_DC(PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1)));
867}
868static void FUNCC(pred8x8l_left_dc)(uint8_t *_src, int has_topleft,
869 int has_topright, ptrdiff_t _stride)
870{
871 pixel *src = (pixel*)_src;
872 int stride = _stride>>(sizeof(pixel)-1);
873
874 PREDICT_8x8_LOAD_LEFT;
875 const pixel4 dc = PIXEL_SPLAT_X4((l0+l1+l2+l3+l4+l5+l6+l7+4) >> 3);
876 PREDICT_8x8_DC(dc);
877}
878static void FUNCC(pred8x8l_top_dc)(uint8_t *_src, int has_topleft,
879 int has_topright, ptrdiff_t _stride)
880{
881 pixel *src = (pixel*)_src;
882 int stride = _stride>>(sizeof(pixel)-1);
883
884 PREDICT_8x8_LOAD_TOP;
885 const pixel4 dc = PIXEL_SPLAT_X4((t0+t1+t2+t3+t4+t5+t6+t7+4) >> 3);
886 PREDICT_8x8_DC(dc);
887}
888static void FUNCC(pred8x8l_dc)(uint8_t *_src, int has_topleft,
889 int has_topright, ptrdiff_t _stride)
890{
891 pixel *src = (pixel*)_src;
892 int stride = _stride>>(sizeof(pixel)-1);
893
894 PREDICT_8x8_LOAD_LEFT;
895 PREDICT_8x8_LOAD_TOP;
896 const pixel4 dc = PIXEL_SPLAT_X4((l0+l1+l2+l3+l4+l5+l6+l7
897 +t0+t1+t2+t3+t4+t5+t6+t7+8) >> 4);
898 PREDICT_8x8_DC(dc);
899}
900static void FUNCC(pred8x8l_horizontal)(uint8_t *_src, int has_topleft,
901 int has_topright, ptrdiff_t _stride)
902{
903 pixel *src = (pixel*)_src;
904 int stride = _stride>>(sizeof(pixel)-1);
905 pixel4 a;
906
907 PREDICT_8x8_LOAD_LEFT;
908#define ROW(y) a = PIXEL_SPLAT_X4(l##y); \
909 AV_WN4PA(src+y*stride, a); \
910 AV_WN4PA(src+y*stride+4, a);
911 ROW(0); ROW(1); ROW(2); ROW(3); ROW(4); ROW(5); ROW(6); ROW(7);
912#undef ROW
913}
914static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft,
915 int has_topright, ptrdiff_t _stride)
916{
917 int y;
918 pixel *src = (pixel*)_src;
919 int stride = _stride>>(sizeof(pixel)-1);
920 pixel4 a, b;
921
922 PREDICT_8x8_LOAD_TOP;
923 src[0] = t0;
924 src[1] = t1;
925 src[2] = t2;
926 src[3] = t3;
927 src[4] = t4;
928 src[5] = t5;
929 src[6] = t6;
930 src[7] = t7;
931 a = AV_RN4PA(((pixel4*)src)+0);
932 b = AV_RN4PA(((pixel4*)src)+1);
933 for( y = 1; y < 8; y++ ) {
934 AV_WN4PA(((pixel4*)(src+y*stride))+0, a);
935 AV_WN4PA(((pixel4*)(src+y*stride))+1, b);
936 }
937}
938static void FUNCC(pred8x8l_down_left)(uint8_t *_src, int has_topleft,
939 int has_topright, ptrdiff_t _stride)
940{
941 pixel *src = (pixel*)_src;
942 int stride = _stride>>(sizeof(pixel)-1);
943 PREDICT_8x8_LOAD_TOP;
944 PREDICT_8x8_LOAD_TOPRIGHT;
945 SRC(0,0)= (t0 + 2*t1 + t2 + 2) >> 2;
946 SRC(0,1)=SRC(1,0)= (t1 + 2*t2 + t3 + 2) >> 2;
947 SRC(0,2)=SRC(1,1)=SRC(2,0)= (t2 + 2*t3 + t4 + 2) >> 2;
948 SRC(0,3)=SRC(1,2)=SRC(2,1)=SRC(3,0)= (t3 + 2*t4 + t5 + 2) >> 2;
949 SRC(0,4)=SRC(1,3)=SRC(2,2)=SRC(3,1)=SRC(4,0)= (t4 + 2*t5 + t6 + 2) >> 2;
950 SRC(0,5)=SRC(1,4)=SRC(2,3)=SRC(3,2)=SRC(4,1)=SRC(5,0)= (t5 + 2*t6 + t7 + 2) >> 2;
951 SRC(0,6)=SRC(1,5)=SRC(2,4)=SRC(3,3)=SRC(4,2)=SRC(5,1)=SRC(6,0)= (t6 + 2*t7 + t8 + 2) >> 2;
952 SRC(0,7)=SRC(1,6)=SRC(2,5)=SRC(3,4)=SRC(4,3)=SRC(5,2)=SRC(6,1)=SRC(7,0)= (t7 + 2*t8 + t9 + 2) >> 2;
953 SRC(1,7)=SRC(2,6)=SRC(3,5)=SRC(4,4)=SRC(5,3)=SRC(6,2)=SRC(7,1)= (t8 + 2*t9 + t10 + 2) >> 2;
954 SRC(2,7)=SRC(3,6)=SRC(4,5)=SRC(5,4)=SRC(6,3)=SRC(7,2)= (t9 + 2*t10 + t11 + 2) >> 2;
955 SRC(3,7)=SRC(4,6)=SRC(5,5)=SRC(6,4)=SRC(7,3)= (t10 + 2*t11 + t12 + 2) >> 2;
956 SRC(4,7)=SRC(5,6)=SRC(6,5)=SRC(7,4)= (t11 + 2*t12 + t13 + 2) >> 2;
957 SRC(5,7)=SRC(6,6)=SRC(7,5)= (t12 + 2*t13 + t14 + 2) >> 2;
958 SRC(6,7)=SRC(7,6)= (t13 + 2*t14 + t15 + 2) >> 2;
959 SRC(7,7)= (t14 + 3*t15 + 2) >> 2;
960}
961static void FUNCC(pred8x8l_down_right)(uint8_t *_src, int has_topleft,
962 int has_topright, ptrdiff_t _stride)
963{
964 pixel *src = (pixel*)_src;
965 int stride = _stride>>(sizeof(pixel)-1);
966 PREDICT_8x8_LOAD_TOP;
967 PREDICT_8x8_LOAD_LEFT;
968 PREDICT_8x8_LOAD_TOPLEFT;
969 SRC(0,7)= (l7 + 2*l6 + l5 + 2) >> 2;
970 SRC(0,6)=SRC(1,7)= (l6 + 2*l5 + l4 + 2) >> 2;
971 SRC(0,5)=SRC(1,6)=SRC(2,7)= (l5 + 2*l4 + l3 + 2) >> 2;
972 SRC(0,4)=SRC(1,5)=SRC(2,6)=SRC(3,7)= (l4 + 2*l3 + l2 + 2) >> 2;
973 SRC(0,3)=SRC(1,4)=SRC(2,5)=SRC(3,6)=SRC(4,7)= (l3 + 2*l2 + l1 + 2) >> 2;
974 SRC(0,2)=SRC(1,3)=SRC(2,4)=SRC(3,5)=SRC(4,6)=SRC(5,7)= (l2 + 2*l1 + l0 + 2) >> 2;
975 SRC(0,1)=SRC(1,2)=SRC(2,3)=SRC(3,4)=SRC(4,5)=SRC(5,6)=SRC(6,7)= (l1 + 2*l0 + lt + 2) >> 2;
976 SRC(0,0)=SRC(1,1)=SRC(2,2)=SRC(3,3)=SRC(4,4)=SRC(5,5)=SRC(6,6)=SRC(7,7)= (l0 + 2*lt + t0 + 2) >> 2;
977 SRC(1,0)=SRC(2,1)=SRC(3,2)=SRC(4,3)=SRC(5,4)=SRC(6,5)=SRC(7,6)= (lt + 2*t0 + t1 + 2) >> 2;
978 SRC(2,0)=SRC(3,1)=SRC(4,2)=SRC(5,3)=SRC(6,4)=SRC(7,5)= (t0 + 2*t1 + t2 + 2) >> 2;
979 SRC(3,0)=SRC(4,1)=SRC(5,2)=SRC(6,3)=SRC(7,4)= (t1 + 2*t2 + t3 + 2) >> 2;
980 SRC(4,0)=SRC(5,1)=SRC(6,2)=SRC(7,3)= (t2 + 2*t3 + t4 + 2) >> 2;
981 SRC(5,0)=SRC(6,1)=SRC(7,2)= (t3 + 2*t4 + t5 + 2) >> 2;
982 SRC(6,0)=SRC(7,1)= (t4 + 2*t5 + t6 + 2) >> 2;
983 SRC(7,0)= (t5 + 2*t6 + t7 + 2) >> 2;
984}
985static void FUNCC(pred8x8l_vertical_right)(uint8_t *_src, int has_topleft,
986 int has_topright, ptrdiff_t _stride)
987{
988 pixel *src = (pixel*)_src;
989 int stride = _stride>>(sizeof(pixel)-1);
990 PREDICT_8x8_LOAD_TOP;
991 PREDICT_8x8_LOAD_LEFT;
992 PREDICT_8x8_LOAD_TOPLEFT;
993 SRC(0,6)= (l5 + 2*l4 + l3 + 2) >> 2;
994 SRC(0,7)= (l6 + 2*l5 + l4 + 2) >> 2;
995 SRC(0,4)=SRC(1,6)= (l3 + 2*l2 + l1 + 2) >> 2;
996 SRC(0,5)=SRC(1,7)= (l4 + 2*l3 + l2 + 2) >> 2;
997 SRC(0,2)=SRC(1,4)=SRC(2,6)= (l1 + 2*l0 + lt + 2) >> 2;
998 SRC(0,3)=SRC(1,5)=SRC(2,7)= (l2 + 2*l1 + l0 + 2) >> 2;
999 SRC(0,1)=SRC(1,3)=SRC(2,5)=SRC(3,7)= (l0 + 2*lt + t0 + 2) >> 2;
1000 SRC(0,0)=SRC(1,2)=SRC(2,4)=SRC(3,6)= (lt + t0 + 1) >> 1;
1001 SRC(1,1)=SRC(2,3)=SRC(3,5)=SRC(4,7)= (lt + 2*t0 + t1 + 2) >> 2;
1002 SRC(1,0)=SRC(2,2)=SRC(3,4)=SRC(4,6)= (t0 + t1 + 1) >> 1;
1003 SRC(2,1)=SRC(3,3)=SRC(4,5)=SRC(5,7)= (t0 + 2*t1 + t2 + 2) >> 2;
1004 SRC(2,0)=SRC(3,2)=SRC(4,4)=SRC(5,6)= (t1 + t2 + 1) >> 1;
1005 SRC(3,1)=SRC(4,3)=SRC(5,5)=SRC(6,7)= (t1 + 2*t2 + t3 + 2) >> 2;
1006 SRC(3,0)=SRC(4,2)=SRC(5,4)=SRC(6,6)= (t2 + t3 + 1) >> 1;
1007 SRC(4,1)=SRC(5,3)=SRC(6,5)=SRC(7,7)= (t2 + 2*t3 + t4 + 2) >> 2;
1008 SRC(4,0)=SRC(5,2)=SRC(6,4)=SRC(7,6)= (t3 + t4 + 1) >> 1;
1009 SRC(5,1)=SRC(6,3)=SRC(7,5)= (t3 + 2*t4 + t5 + 2) >> 2;
1010 SRC(5,0)=SRC(6,2)=SRC(7,4)= (t4 + t5 + 1) >> 1;
1011 SRC(6,1)=SRC(7,3)= (t4 + 2*t5 + t6 + 2) >> 2;
1012 SRC(6,0)=SRC(7,2)= (t5 + t6 + 1) >> 1;
1013 SRC(7,1)= (t5 + 2*t6 + t7 + 2) >> 2;
1014 SRC(7,0)= (t6 + t7 + 1) >> 1;
1015}
1016static void FUNCC(pred8x8l_horizontal_down)(uint8_t *_src, int has_topleft,
1017 int has_topright, ptrdiff_t _stride)
1018{
1019 pixel *src = (pixel*)_src;
1020 int stride = _stride>>(sizeof(pixel)-1);
1021 PREDICT_8x8_LOAD_TOP;
1022 PREDICT_8x8_LOAD_LEFT;
1023 PREDICT_8x8_LOAD_TOPLEFT;
1024 SRC(0,7)= (l6 + l7 + 1) >> 1;
1025 SRC(1,7)= (l5 + 2*l6 + l7 + 2) >> 2;
1026 SRC(0,6)=SRC(2,7)= (l5 + l6 + 1) >> 1;
1027 SRC(1,6)=SRC(3,7)= (l4 + 2*l5 + l6 + 2) >> 2;
1028 SRC(0,5)=SRC(2,6)=SRC(4,7)= (l4 + l5 + 1) >> 1;
1029 SRC(1,5)=SRC(3,6)=SRC(5,7)= (l3 + 2*l4 + l5 + 2) >> 2;
1030 SRC(0,4)=SRC(2,5)=SRC(4,6)=SRC(6,7)= (l3 + l4 + 1) >> 1;
1031 SRC(1,4)=SRC(3,5)=SRC(5,6)=SRC(7,7)= (l2 + 2*l3 + l4 + 2) >> 2;
1032 SRC(0,3)=SRC(2,4)=SRC(4,5)=SRC(6,6)= (l2 + l3 + 1) >> 1;
1033 SRC(1,3)=SRC(3,4)=SRC(5,5)=SRC(7,6)= (l1 + 2*l2 + l3 + 2) >> 2;
1034 SRC(0,2)=SRC(2,3)=SRC(4,4)=SRC(6,5)= (l1 + l2 + 1) >> 1;
1035 SRC(1,2)=SRC(3,3)=SRC(5,4)=SRC(7,5)= (l0 + 2*l1 + l2 + 2) >> 2;
1036 SRC(0,1)=SRC(2,2)=SRC(4,3)=SRC(6,4)= (l0 + l1 + 1) >> 1;
1037 SRC(1,1)=SRC(3,2)=SRC(5,3)=SRC(7,4)= (lt + 2*l0 + l1 + 2) >> 2;
1038 SRC(0,0)=SRC(2,1)=SRC(4,2)=SRC(6,3)= (lt + l0 + 1) >> 1;
1039 SRC(1,0)=SRC(3,1)=SRC(5,2)=SRC(7,3)= (l0 + 2*lt + t0 + 2) >> 2;
1040 SRC(2,0)=SRC(4,1)=SRC(6,2)= (t1 + 2*t0 + lt + 2) >> 2;
1041 SRC(3,0)=SRC(5,1)=SRC(7,2)= (t2 + 2*t1 + t0 + 2) >> 2;
1042 SRC(4,0)=SRC(6,1)= (t3 + 2*t2 + t1 + 2) >> 2;
1043 SRC(5,0)=SRC(7,1)= (t4 + 2*t3 + t2 + 2) >> 2;
1044 SRC(6,0)= (t5 + 2*t4 + t3 + 2) >> 2;
1045 SRC(7,0)= (t6 + 2*t5 + t4 + 2) >> 2;
1046}
1047static void FUNCC(pred8x8l_vertical_left)(uint8_t *_src, int has_topleft,
1048 int has_topright, ptrdiff_t _stride)
1049{
1050 pixel *src = (pixel*)_src;
1051 int stride = _stride>>(sizeof(pixel)-1);
1052 PREDICT_8x8_LOAD_TOP;
1053 PREDICT_8x8_LOAD_TOPRIGHT;
1054 SRC(0,0)= (t0 + t1 + 1) >> 1;
1055 SRC(0,1)= (t0 + 2*t1 + t2 + 2) >> 2;
1056 SRC(0,2)=SRC(1,0)= (t1 + t2 + 1) >> 1;
1057 SRC(0,3)=SRC(1,1)= (t1 + 2*t2 + t3 + 2) >> 2;
1058 SRC(0,4)=SRC(1,2)=SRC(2,0)= (t2 + t3 + 1) >> 1;
1059 SRC(0,5)=SRC(1,3)=SRC(2,1)= (t2 + 2*t3 + t4 + 2) >> 2;
1060 SRC(0,6)=SRC(1,4)=SRC(2,2)=SRC(3,0)= (t3 + t4 + 1) >> 1;
1061 SRC(0,7)=SRC(1,5)=SRC(2,3)=SRC(3,1)= (t3 + 2*t4 + t5 + 2) >> 2;
1062 SRC(1,6)=SRC(2,4)=SRC(3,2)=SRC(4,0)= (t4 + t5 + 1) >> 1;
1063 SRC(1,7)=SRC(2,5)=SRC(3,3)=SRC(4,1)= (t4 + 2*t5 + t6 + 2) >> 2;
1064 SRC(2,6)=SRC(3,4)=SRC(4,2)=SRC(5,0)= (t5 + t6 + 1) >> 1;
1065 SRC(2,7)=SRC(3,5)=SRC(4,3)=SRC(5,1)= (t5 + 2*t6 + t7 + 2) >> 2;
1066 SRC(3,6)=SRC(4,4)=SRC(5,2)=SRC(6,0)= (t6 + t7 + 1) >> 1;
1067 SRC(3,7)=SRC(4,5)=SRC(5,3)=SRC(6,1)= (t6 + 2*t7 + t8 + 2) >> 2;
1068 SRC(4,6)=SRC(5,4)=SRC(6,2)=SRC(7,0)= (t7 + t8 + 1) >> 1;
1069 SRC(4,7)=SRC(5,5)=SRC(6,3)=SRC(7,1)= (t7 + 2*t8 + t9 + 2) >> 2;
1070 SRC(5,6)=SRC(6,4)=SRC(7,2)= (t8 + t9 + 1) >> 1;
1071 SRC(5,7)=SRC(6,5)=SRC(7,3)= (t8 + 2*t9 + t10 + 2) >> 2;
1072 SRC(6,6)=SRC(7,4)= (t9 + t10 + 1) >> 1;
1073 SRC(6,7)=SRC(7,5)= (t9 + 2*t10 + t11 + 2) >> 2;
1074 SRC(7,6)= (t10 + t11 + 1) >> 1;
1075 SRC(7,7)= (t10 + 2*t11 + t12 + 2) >> 2;
1076}
1077static void FUNCC(pred8x8l_horizontal_up)(uint8_t *_src, int has_topleft,
1078 int has_topright, ptrdiff_t _stride)
1079{
1080 pixel *src = (pixel*)_src;
1081 int stride = _stride>>(sizeof(pixel)-1);
1082 PREDICT_8x8_LOAD_LEFT;
1083 SRC(0,0)= (l0 + l1 + 1) >> 1;
1084 SRC(1,0)= (l0 + 2*l1 + l2 + 2) >> 2;
1085 SRC(0,1)=SRC(2,0)= (l1 + l2 + 1) >> 1;
1086 SRC(1,1)=SRC(3,0)= (l1 + 2*l2 + l3 + 2) >> 2;
1087 SRC(0,2)=SRC(2,1)=SRC(4,0)= (l2 + l3 + 1) >> 1;
1088 SRC(1,2)=SRC(3,1)=SRC(5,0)= (l2 + 2*l3 + l4 + 2) >> 2;
1089 SRC(0,3)=SRC(2,2)=SRC(4,1)=SRC(6,0)= (l3 + l4 + 1) >> 1;
1090 SRC(1,3)=SRC(3,2)=SRC(5,1)=SRC(7,0)= (l3 + 2*l4 + l5 + 2) >> 2;
1091 SRC(0,4)=SRC(2,3)=SRC(4,2)=SRC(6,1)= (l4 + l5 + 1) >> 1;
1092 SRC(1,4)=SRC(3,3)=SRC(5,2)=SRC(7,1)= (l4 + 2*l5 + l6 + 2) >> 2;
1093 SRC(0,5)=SRC(2,4)=SRC(4,3)=SRC(6,2)= (l5 + l6 + 1) >> 1;
1094 SRC(1,5)=SRC(3,4)=SRC(5,3)=SRC(7,2)= (l5 + 2*l6 + l7 + 2) >> 2;
1095 SRC(0,6)=SRC(2,5)=SRC(4,4)=SRC(6,3)= (l6 + l7 + 1) >> 1;
1096 SRC(1,6)=SRC(3,5)=SRC(5,4)=SRC(7,3)= (l6 + 3*l7 + 2) >> 2;
1097 SRC(0,7)=SRC(1,7)=SRC(2,6)=SRC(2,7)=SRC(3,6)=
1098 SRC(3,7)=SRC(4,5)=SRC(4,6)=SRC(4,7)=SRC(5,5)=
1099 SRC(5,6)=SRC(5,7)=SRC(6,4)=SRC(6,5)=SRC(6,6)=
1100 SRC(6,7)=SRC(7,4)=SRC(7,5)=SRC(7,6)=SRC(7,7)= l7;
1101}
1102
1103static void FUNCC(pred8x8l_vertical_filter_add)(uint8_t *_src, int16_t *_block, int has_topleft,
1104 int has_topright, ptrdiff_t _stride)
1105{
1106 int i;
1107 pixel *src = (pixel*)_src;
1108 const dctcoef *block = (const dctcoef*)_block;
1109 pixel pix[8];
1110 int stride = _stride>>(sizeof(pixel)-1);
1111 PREDICT_8x8_LOAD_TOP;
1112
1113 pix[0] = t0;
1114 pix[1] = t1;
1115 pix[2] = t2;
1116 pix[3] = t3;
1117 pix[4] = t4;
1118 pix[5] = t5;
1119 pix[6] = t6;
1120 pix[7] = t7;
1121
1122 for(i=0; i<8; i++){
1123 pixel v = pix[i];
1124 src[0*stride]= v += block[0];
1125 src[1*stride]= v += block[8];
1126 src[2*stride]= v += block[16];
1127 src[3*stride]= v += block[24];
1128 src[4*stride]= v += block[32];
1129 src[5*stride]= v += block[40];
1130 src[6*stride]= v += block[48];
1131 src[7*stride]= v + block[56];
1132 src++;
1133 block++;
1134 }
1135
1136 memset(_block, 0, sizeof(dctcoef) * 64);
1137}
1138
1139static void FUNCC(pred8x8l_horizontal_filter_add)(uint8_t *_src, int16_t *_block, int has_topleft,
1140 int has_topright, ptrdiff_t _stride)
1141{
1142 int i;
1143 pixel *src = (pixel*)_src;
1144 const dctcoef *block = (const dctcoef*)_block;
1145 pixel pix[8];
1146 int stride = _stride>>(sizeof(pixel)-1);
1147 PREDICT_8x8_LOAD_LEFT;
1148
1149 pix[0] = l0;
1150 pix[1] = l1;
1151 pix[2] = l2;
1152 pix[3] = l3;
1153 pix[4] = l4;
1154 pix[5] = l5;
1155 pix[6] = l6;
1156 pix[7] = l7;
1157
1158 for(i=0; i<8; i++){
1159 pixel v = pix[i];
1160 src[0]= v += block[0];
1161 src[1]= v += block[1];
1162 src[2]= v += block[2];
1163 src[3]= v += block[3];
1164 src[4]= v += block[4];
1165 src[5]= v += block[5];
1166 src[6]= v += block[6];
1167 src[7]= v + block[7];
1168 src+= stride;
1169 block+= 8;
1170 }
1171
1172 memset(_block, 0, sizeof(dctcoef) * 64);
1173}
1174
1175#undef PREDICT_8x8_LOAD_LEFT
1176#undef PREDICT_8x8_LOAD_TOP
1177#undef PREDICT_8x8_LOAD_TOPLEFT
1178#undef PREDICT_8x8_LOAD_TOPRIGHT
1179#undef PREDICT_8x8_DC
1180#undef PTR
1181#undef PT
1182#undef PL
1183#undef SRC
1184
1185static void FUNCC(pred4x4_vertical_add)(uint8_t *_pix, int16_t *_block,
1186 ptrdiff_t stride)
1187{
1188 int i;
1189 pixel *pix = (pixel*)_pix;
1190 const dctcoef *block = (const dctcoef*)_block;
1191 stride >>= sizeof(pixel)-1;
1192 pix -= stride;
1193 for(i=0; i<4; i++){
1194 pixel v = pix[0];
1195 pix[1*stride]= v += block[0];
1196 pix[2*stride]= v += block[4];
1197 pix[3*stride]= v += block[8];
1198 pix[4*stride]= v + block[12];
1199 pix++;
1200 block++;
1201 }
1202
1203 memset(_block, 0, sizeof(dctcoef) * 16);
1204}
1205
1206static void FUNCC(pred4x4_horizontal_add)(uint8_t *_pix, int16_t *_block,
1207 ptrdiff_t stride)
1208{
1209 int i;
1210 pixel *pix = (pixel*)_pix;
1211 const dctcoef *block = (const dctcoef*)_block;
1212 stride >>= sizeof(pixel)-1;
1213 for(i=0; i<4; i++){
1214 pixel v = pix[-1];
1215 pix[0]= v += block[0];
1216 pix[1]= v += block[1];
1217 pix[2]= v += block[2];
1218 pix[3]= v + block[3];
1219 pix+= stride;
1220 block+= 4;
1221 }
1222
1223 memset(_block, 0, sizeof(dctcoef) * 16);
1224}
1225
1226static void FUNCC(pred8x8l_vertical_add)(uint8_t *_pix, int16_t *_block,
1227 ptrdiff_t stride)
1228{
1229 int i;
1230 pixel *pix = (pixel*)_pix;
1231 const dctcoef *block = (const dctcoef*)_block;
1232 stride >>= sizeof(pixel)-1;
1233 pix -= stride;
1234 for(i=0; i<8; i++){
1235 pixel v = pix[0];
1236 pix[1*stride]= v += block[0];
1237 pix[2*stride]= v += block[8];
1238 pix[3*stride]= v += block[16];
1239 pix[4*stride]= v += block[24];
1240 pix[5*stride]= v += block[32];
1241 pix[6*stride]= v += block[40];
1242 pix[7*stride]= v += block[48];
1243 pix[8*stride]= v + block[56];
1244 pix++;
1245 block++;
1246 }
1247
1248 memset(_block, 0, sizeof(dctcoef) * 64);
1249}
1250
1251static void FUNCC(pred8x8l_horizontal_add)(uint8_t *_pix, int16_t *_block,
1252 ptrdiff_t stride)
1253{
1254 int i;
1255 pixel *pix = (pixel*)_pix;
1256 const dctcoef *block = (const dctcoef*)_block;
1257 stride >>= sizeof(pixel)-1;
1258 for(i=0; i<8; i++){
1259 pixel v = pix[-1];
1260 pix[0]= v += block[0];
1261 pix[1]= v += block[1];
1262 pix[2]= v += block[2];
1263 pix[3]= v += block[3];
1264 pix[4]= v += block[4];
1265 pix[5]= v += block[5];
1266 pix[6]= v += block[6];
1267 pix[7]= v + block[7];
1268 pix+= stride;
1269 block+= 8;
1270 }
1271
1272 memset(_block, 0, sizeof(dctcoef) * 64);
1273}
1274
1275static void FUNCC(pred16x16_vertical_add)(uint8_t *pix, const int *block_offset,
1276 int16_t *block,
1277 ptrdiff_t stride)
1278{
1279 int i;
1280 for(i=0; i<16; i++)
1281 FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1282}
1283
1284static void FUNCC(pred16x16_horizontal_add)(uint8_t *pix,
1285 const int *block_offset,
1286 int16_t *block,
1287 ptrdiff_t stride)
1288{
1289 int i;
1290 for(i=0; i<16; i++)
1291 FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1292}
1293
1294static void FUNCC(pred8x8_vertical_add)(uint8_t *pix, const int *block_offset,
1295 int16_t *block, ptrdiff_t stride)
1296{
1297 int i;
1298 for(i=0; i<4; i++)
1299 FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1300}
1301
1302static void FUNCC(pred8x16_vertical_add)(uint8_t *pix, const int *block_offset,
1303 int16_t *block, ptrdiff_t stride)
1304{
1305 int i;
1306 for(i=0; i<4; i++)
1307 FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1308 for(i=4; i<8; i++)
1309 FUNCC(pred4x4_vertical_add)(pix + block_offset[i+4], block + i*16*sizeof(pixel), stride);
1310}
1311
1312static void FUNCC(pred8x8_horizontal_add)(uint8_t *pix, const int *block_offset,
1313 int16_t *block,
1314 ptrdiff_t stride)
1315{
1316 int i;
1317 for(i=0; i<4; i++)
1318 FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1319}
1320
1321static void FUNCC(pred8x16_horizontal_add)(uint8_t *pix,
1322 const int *block_offset,
1323 int16_t *block, ptrdiff_t stride)
1324{
1325 int i;
1326 for(i=0; i<4; i++)
1327 FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
1328 for(i=4; i<8; i++)
1329 FUNCC(pred4x4_horizontal_add)(pix + block_offset[i+4], block + i*16*sizeof(pixel), stride);
1330}
_src
uint8_t ptrdiff_t const uint8_t * _src
Definition dsp.h:56
BIT_DEPTH
#define BIT_DEPTH
Definition dsp_init.c:121
V
#define V
Definition avdct.c:32
bit_depth_template.c
CLIP
#define CLIP(a)
Definition bit_depth_template.c:99
FUNC
#define FUNC(a)
Definition bit_depth_template.c:104
pixel4
#define pixel4
Definition bit_depth_template.c:84
PIXEL_SPLAT_X4
#define PIXEL_SPLAT_X4(x)
Definition bit_depth_template.c:96
AV_WN4PA
#define AV_WN4PA
Definition bit_depth_template.c:95
dctcoef
#define dctcoef
Definition bit_depth_template.c:85
pixel
#define pixel
Definition bit_depth_template.c:82
AV_RN4PA
#define AV_RN4PA
Definition bit_depth_template.c:92
FUNCC
#define FUNCC(a)
Definition bit_depth_template.c:105
i
#define i(width, name, range_min, range_max)
Definition cbs_h264.c:63
NULL
#define NULL
Definition coverity.c:32
block
static int16_t block[64]
Definition dct.c:125
PREDICT_16x16_DC
#define PREDICT_16x16_DC(v)
Definition h264pred_template.c:337
pred8x8l_horizontal_add
static void FUNCC pred8x8l_horizontal_add(uint8_t *_pix, int16_t *_block, ptrdiff_t stride)
Definition h264pred_template.c:1251
pred8x16_mad_cow_dc_0lt
static void FUNC pred8x16_mad_cow_dc_0lt(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:712
pred4x4_vertical_right
static void FUNCC pred4x4_vertical_right(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:194
pred16x16_plane
static void FUNCC pred16x16_plane(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:458
LOAD_TOP_RIGHT_EDGE
#define LOAD_TOP_RIGHT_EDGE
Definition h264pred_template.c:115
pred16x16_dc
static void FUNCC pred16x16_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:346
pred4x4_top_dc
static void FUNCC pred4x4_top_dc(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:87
pred8x8l_vertical_add
static void FUNCC pred8x8l_vertical_add(uint8_t *_pix, int16_t *_block, ptrdiff_t stride)
Definition h264pred_template.c:1226
pred8x8l_vertical_filter_add
static void FUNCC pred8x8l_vertical_filter_add(uint8_t *_src, int16_t *_block, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:1103
pred16x16_left_dc
static void FUNCC pred16x16_left_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:365
PREDICT_8x8_LOAD_LEFT
#define PREDICT_8x8_LOAD_LEFT
Definition h264pred_template.c:825
pred16x16_horizontal_add
static void FUNCC pred16x16_horizontal_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1284
pred8x8l_left_dc
static void FUNCC pred8x8l_left_dc(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:868
pred8x8_horizontal
static void FUNCC pred8x8_horizontal(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:491
PREDICT_8x8_LOAD_TOP
#define PREDICT_8x8_LOAD_TOP
Definition h264pred_template.c:833
pred8x8_mad_cow_dc_l0t
static void FUNC pred8x8_mad_cow_dc_l0t(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:694
LOAD_TOP_EDGE
#define LOAD_TOP_EDGE
Definition h264pred_template.c:133
pred4x4_horizontal_up
static void FUNCC pred4x4_horizontal_up(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:250
a
int a
Definition h264pred_template.c:416
pred8x8l_down_left
static void FUNCC pred8x8l_down_left(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:938
pred4x4_vertical_add
static void FUNCC pred4x4_vertical_add(uint8_t *_pix, int16_t *_block, ptrdiff_t stride)
Definition h264pred_template.c:1185
PREDICT_8x8_LOAD_TOPLEFT
#define PREDICT_8x8_LOAD_TOPLEFT
Definition h264pred_template.c:849
pred4x4_128_dc
static void FUNCC pred4x4_128_dc(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:101
pred8x8_mad_cow_dc_l00
static void FUNC pred8x8_mad_cow_dc_l00(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:718
pred8x16_mad_cow_dc_l0t
static void FUNC pred8x16_mad_cow_dc_l0t(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:700
pred4x4_vertical
static void FUNCC pred4x4_vertical(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:34
pred8x8l_vertical
static void FUNCC pred8x8l_vertical(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:914
pred8x8_mad_cow_dc_0l0
static void FUNC pred8x8_mad_cow_dc_0l0(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:732
ROW
#define ROW(y)
pred8x8l_vertical_left
static void FUNCC pred8x8l_vertical_left(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:1047
pred8x8l_horizontal_filter_add
static void FUNCC pred8x8l_horizontal_filter_add(uint8_t *_src, int16_t *_block, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:1139
pred8x16_horizontal_add
static void FUNCC pred8x16_horizontal_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1321
pred8x8_dc
static void FUNCC pred8x8_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:621
pred8x8l_top_dc
static void FUNCC pred8x8l_top_dc(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:878
pred4x4_horizontal_down
static void FUNCC pred4x4_horizontal_down(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:275
pred8x8l_vertical_right
static void FUNCC pred8x8l_vertical_right(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:985
pred8x8_128_dc
FUNCC pred8x8_128_dc(_src+8 *stride, stride)
pred8x8l_128_dc
static void FUNCC pred8x8l_128_dc(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:860
pred8x16_mad_cow_dc_l00
static void FUNC pred8x16_mad_cow_dc_l00(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:725
pred8x8_mad_cow_dc_0lt
static void FUNC pred8x8_mad_cow_dc_0lt(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:706
pred4x4_dc
static void FUNCC pred4x4_dc(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:58
_stride
ptrdiff_t _stride
Definition h264pred_template.c:411
pred8x8l_dc
static void FUNCC pred8x8l_dc(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:888
pred16x16_vertical_add
static void FUNCC pred16x16_vertical_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1275
pred8x16_top_dc
static void FUNCC pred8x16_top_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:599
pred4x4_horizontal_add
static void FUNCC pred4x4_horizontal_add(uint8_t *_pix, int16_t *_block, ptrdiff_t stride)
Definition h264pred_template.c:1206
SRC
#define SRC(x, y)
Definition h264pred_template.c:822
PRED16x16_X
#define PRED16x16_X(n, v)
Definition h264pred_template.c:395
pred8x8_top_dc
static void FUNCC pred8x8_top_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:573
pred8x8_plane
static void FUNCC pred8x8_plane(uint8_t *_src, ptrdiff_t _stride)
Definition h264pred_template.c:746
pred8x8l_horizontal_down
static void FUNCC pred8x8l_horizontal_down(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:1016
pred16x16_vertical
static void FUNCC pred16x16_vertical(uint8_t *_src, ptrdiff_t _stride)
Definition h264pred_template.c:303
pred8x16_vertical
static void FUNCC pred8x16_vertical(uint8_t *_src, ptrdiff_t _stride)
Definition h264pred_template.c:477
pred8x16_left_dc
static void FUNCC pred8x16_left_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:567
pred8x8l_down_right
static void FUNCC pred8x8l_down_right(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:961
pred16x16_top_dc
static void FUNCC pred16x16_top_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:380
pred8x8_horizontal_add
static void FUNCC pred8x8_horizontal_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1312
pred8x8l_horizontal_up
static void FUNCC pred8x8l_horizontal_up(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:1077
pred4x4_vertical_left
static void FUNCC pred4x4_vertical_left(uint8_t *_src, const uint8_t *_topright, ptrdiff_t _stride)
Definition h264pred_template.c:222
src2
const pixel * src2
Definition h264pred_template.c:421
pred8x16_mad_cow_dc_0l0
static void FUNC pred8x16_mad_cow_dc_0l0(uint8_t *src, ptrdiff_t stride)
Definition h264pred_template.c:739
pred8x16_plane
static void FUNCC pred8x16_plane(uint8_t *_src, ptrdiff_t _stride)
Definition h264pred_template.c:781
pred4x4_down_left
static void FUNCC pred4x4_down_left(uint8_t *_src, const uint8_t *_topright, ptrdiff_t _stride)
Definition h264pred_template.c:166
pred8x8_vertical_add
static void FUNCC pred8x8_vertical_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1294
pred16x16_horizontal
static void FUNCC pred16x16_horizontal(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:321
pred8x16_horizontal
static void FUNCC pred8x16_horizontal(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:504
pred4x4_horizontal
static void FUNCC pred4x4_horizontal(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:47
svq3
ptrdiff_t const int svq3
Definition h264pred_template.c:412
pred8x16_vertical_add
static void FUNCC pred8x16_vertical_add(uint8_t *pix, const int *block_offset, int16_t *block, ptrdiff_t stride)
Definition h264pred_template.c:1302
pred8x16_dc
static void FUNCC pred8x16_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:650
PREDICT_8x8_DC
#define PREDICT_8x8_DC(v)
Definition h264pred_template.c:852
rv40
ptrdiff_t const int const int rv40
Definition h264pred_template.c:414
PRED8x8_X
#define PRED8x8_X(n, v)
Definition h264pred_template.c:516
pred8x8_vertical
static void FUNCC pred8x8_vertical(uint8_t *_src, ptrdiff_t _stride)
Definition h264pred_template.c:463
pred8x8l_horizontal
static void FUNCC pred8x8l_horizontal(uint8_t *_src, int has_topleft, int has_topright, ptrdiff_t _stride)
Definition h264pred_template.c:900
pred4x4_left_dc
static void FUNCC pred4x4_left_dc(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:73
LOAD_LEFT_EDGE
#define LOAD_LEFT_EDGE
Definition h264pred_template.c:127
pred8x8_left_dc
static void FUNCC pred8x8_left_dc(uint8_t *_src, ptrdiff_t stride)
Definition h264pred_template.c:541
pred4x4_down_right
static void FUNCC pred4x4_down_right(uint8_t *_src, const uint8_t *topright, ptrdiff_t _stride)
Definition h264pred_template.c:139
PREDICT_8x8_LOAD_TOPRIGHT
#define PREDICT_8x8_LOAD_TOPRIGHT
Definition h264pred_template.c:842
b
#define b
Definition input.c:43
t15
static int t15(InterplayACMContext *s, unsigned ind, unsigned col)
Definition interplayacm.c:339
pix
enum AVPixelFormat pix
Definition ohcodec.c:55
H
#define H
Definition pixlet.c:39
stride
#define stride
Definition h264pred_template.c:536
src1
#define src1
Definition h264pred.c:141
src0
#define src0
Definition h264pred.c:140
src
#define src
Definition vp8dsp.c:248
c
static double c[64]
Definition vsrc_mptestsrc.c:89
Generated on Sun Aug 16 2026 20:20:47 for FFmpeg by doxygen 1.13.2