1 /*****************************************************************************
2 * predict.c: intra prediction
3 *****************************************************************************
4 * Copyright (C) 2003-2014 x264 project
5 *
6 * Authors: Laurent Aimar <fenrir@via.ecp.fr>
7 * Loren Merritt <lorenm@u.washington.edu>
8 * Fiona Glaser <fiona@x264.com>
9 * Henrik Gramner <henrik@gramner.com>
10 *
11 * This program is free software; you can redistribute it and/or modify
12 * it under the terms of the GNU General Public License as published by
13 * the Free Software Foundation; either version 2 of the License, or
14 * (at your option) any later version.
15 *
16 * This program is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 * GNU General Public License for more details.
20 *
21 * You should have received a copy of the GNU General Public License
22 * along with this program; if not, write to the Free Software
23 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02111, USA.
24 *
25 * This program is also available under a commercial proprietary license.
26 * For more information, contact us at licensing@x264.com.
27 *****************************************************************************/
28
29 /* predict4x4 are inspired from ffmpeg h264 decoder */
30
31
32 #include "common.h"
33
34 #if HAVE_MMX
35 # include "x86/predict.h"
36 #endif
37 #if ARCH_PPC
38 # include "ppc/predict.h"
39 #endif
40 #if ARCH_ARM
41 # include "arm/predict.h"
42 #endif
43 #if ARCH_AARCH64
44 # include "aarch64/predict.h"
45 #endif
46
47 /****************************************************************************
48 * 16x16 prediction for intra luma block
49 ****************************************************************************/
50
51 #define PREDICT_16x16_DC(v)\
52 for( int i = 0; i < 16; i++ )\
53 {\
54 MPIXEL_X4( src+ 0 ) = v;\
55 MPIXEL_X4( src+ 4 ) = v;\
56 MPIXEL_X4( src+ 8 ) = v;\
57 MPIXEL_X4( src+12 ) = v;\
58 src += FDEC_STRIDE;\
59 }
60
x264_predict_16x16_dc_c(pixel * src)61 void x264_predict_16x16_dc_c( pixel *src )
62 {
63 int dc = 0;
64 int i2;
65 pixel4 dcsplat;
66
67 for( i2 = 0; i2 < 16; i2++ )
68 {
69 dc += src[-1 + i2 * FDEC_STRIDE];
70 dc += src[i2 - FDEC_STRIDE];
71 }
72 dcsplat = PIXEL_SPLAT_X4( ( dc + 16 ) >> 5 );
73
74 PREDICT_16x16_DC( dcsplat );
75 }
76
x264_predict_16x16_dc_left_c(pixel * src)77 static void x264_predict_16x16_dc_left_c( pixel *src )
78 {
79 int dc = 0;
80 int i2;
81 pixel4 dcsplat;
82
83 for( i2 = 0; i2 < 16; i2++ ) {
84 dc += src[-1 + i2 * FDEC_STRIDE];
85 }
86 dcsplat = PIXEL_SPLAT_X4( ( dc + 8 ) >> 4 );
87
88 PREDICT_16x16_DC( dcsplat );
89 }
90
x264_predict_16x16_dc_top_c(pixel * src)91 static void x264_predict_16x16_dc_top_c( pixel *src )
92 {
93 int dc = 0;
94 int i2;
95 pixel4 dcsplat;
96
97 for( i2 = 0; i2 < 16; i2++ )
98 dc += src[i2 - FDEC_STRIDE];
99 dcsplat = PIXEL_SPLAT_X4( ( dc + 8 ) >> 4 );
100
101 PREDICT_16x16_DC( dcsplat );
102 }
103
x264_predict_16x16_dc_128_c(pixel * src)104 static void x264_predict_16x16_dc_128_c( pixel *src )
105 {
106 PREDICT_16x16_DC( PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) ) );
107 }
108
x264_predict_16x16_h_c(pixel * src)109 void x264_predict_16x16_h_c( pixel *src )
110 {
111 int i;
112
113 for( i = 0; i < 16; i++ )
114 {
115 const pixel4 v = PIXEL_SPLAT_X4( src[-1] );
116 MPIXEL_X4( src+ 0 ) = v;
117 MPIXEL_X4( src+ 4 ) = v;
118 MPIXEL_X4( src+ 8 ) = v;
119 MPIXEL_X4( src+12 ) = v;
120 src += FDEC_STRIDE;
121 }
122 }
123
x264_predict_16x16_v_c(pixel * src)124 void x264_predict_16x16_v_c( pixel *src )
125 {
126 pixel4 v0 = MPIXEL_X4( &src[ 0-FDEC_STRIDE] );
127 pixel4 v1 = MPIXEL_X4( &src[ 4-FDEC_STRIDE] );
128 pixel4 v2 = MPIXEL_X4( &src[ 8-FDEC_STRIDE] );
129 pixel4 v3 = MPIXEL_X4( &src[12-FDEC_STRIDE] );
130 int i;
131
132 for( i = 0; i < 16; i++ )
133 {
134 MPIXEL_X4( src+ 0 ) = v0;
135 MPIXEL_X4( src+ 4 ) = v1;
136 MPIXEL_X4( src+ 8 ) = v2;
137 MPIXEL_X4( src+12 ) = v3;
138 src += FDEC_STRIDE;
139 }
140 }
x264_predict_16x16_p_c(pixel * src)141 void x264_predict_16x16_p_c( pixel *src )
142 {
143 int H = 0, V = 0;
144 int i;
145 int a;
146 int b;
147 int c;
148 int i00;
149 int y;
150
151 /* calculate H and V */
152 for( i = 0; i <= 7; i++ )
153 {
154 H += ( i + 1 ) * ( src[ 8 + i - FDEC_STRIDE ] - src[6 -i -FDEC_STRIDE] );
155 V += ( i + 1 ) * ( src[-1 + (8+i)*FDEC_STRIDE] - src[-1 + (6-i)*FDEC_STRIDE] );
156 }
157
158 a = 16 * ( src[-1 + 15*FDEC_STRIDE] + src[15 - FDEC_STRIDE] );
159 b = ( 5 * H + 32 ) >> 6;
160 c = ( 5 * V + 32 ) >> 6;
161
162 i00 = a - b * 7 - c * 7 + 16;
163
164 for( y = 0; y < 16; y++ )
165 {
166 int pix = i00;
167 int x;
168
169 for( x = 0; x < 16; x++ )
170 {
171 src[x] = x264_clip_pixel( pix>>5 );
172 pix += b;
173 }
174 src += FDEC_STRIDE;
175 i00 += c;
176 }
177 }
178
179
180 /****************************************************************************
181 * 8x8 prediction for intra chroma block (4:2:0)
182 ****************************************************************************/
183
x264_predict_8x8c_dc_128_c(pixel * src)184 static void x264_predict_8x8c_dc_128_c( pixel *src )
185 {
186 int y;
187
188 for( y = 0; y < 8; y++ )
189 {
190 MPIXEL_X4( src+0 ) = PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) );
191 MPIXEL_X4( src+4 ) = PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) );
192 src += FDEC_STRIDE;
193 }
194 }
x264_predict_8x8c_dc_left_c(pixel * src)195 static void x264_predict_8x8c_dc_left_c( pixel *src )
196 {
197 int dc0 = 0, dc1 = 0;
198 int y;
199 pixel4 dc0splat;
200 pixel4 dc1splat;
201
202 for( y = 0; y < 4; y++ )
203 {
204 dc0 += src[y * FDEC_STRIDE - 1];
205 dc1 += src[(y+4) * FDEC_STRIDE - 1];
206 }
207 dc0splat = PIXEL_SPLAT_X4( ( dc0 + 2 ) >> 2 );
208 dc1splat = PIXEL_SPLAT_X4( ( dc1 + 2 ) >> 2 );
209
210 for( y = 0; y < 4; y++ )
211 {
212 MPIXEL_X4( src+0 ) = dc0splat;
213 MPIXEL_X4( src+4 ) = dc0splat;
214 src += FDEC_STRIDE;
215 }
216 for( y = 0; y < 4; y++ )
217 {
218 MPIXEL_X4( src+0 ) = dc1splat;
219 MPIXEL_X4( src+4 ) = dc1splat;
220 src += FDEC_STRIDE;
221 }
222
223 }
x264_predict_8x8c_dc_top_c(pixel * src)224 static void x264_predict_8x8c_dc_top_c( pixel *src )
225 {
226 int dc0 = 0, dc1 = 0;
227 int x;
228 pixel4 dc0splat;
229 pixel4 dc1splat;
230 int y;
231
232 for( x = 0; x < 4; x++ )
233 {
234 dc0 += src[x - FDEC_STRIDE];
235 dc1 += src[x + 4 - FDEC_STRIDE];
236 }
237 dc0splat = PIXEL_SPLAT_X4( ( dc0 + 2 ) >> 2 );
238 dc1splat = PIXEL_SPLAT_X4( ( dc1 + 2 ) >> 2 );
239
240 for( y = 0; y < 8; y++ )
241 {
242 MPIXEL_X4( src+0 ) = dc0splat;
243 MPIXEL_X4( src+4 ) = dc1splat;
244 src += FDEC_STRIDE;
245 }
246 }
247
x264_predict_8x8c_dc_c(pixel * src)248 void x264_predict_8x8c_dc_c( pixel *src )
249 {
250 int s0 = 0, s1 = 0, s2 = 0, s3 = 0;
251 int i;
252 pixel4 dc0;
253 pixel4 dc1;
254 pixel4 dc2;
255 pixel4 dc3;
256 int y;
257
258 /*
259 s0 s1
260 s2
261 s3
262 */
263 for( i = 0; i < 4; i++ )
264 {
265 s0 += src[i - FDEC_STRIDE];
266 s1 += src[i + 4 - FDEC_STRIDE];
267 s2 += src[-1 + i * FDEC_STRIDE];
268 s3 += src[-1 + (i+4)*FDEC_STRIDE];
269 }
270 /*
271 dc0 dc1
272 dc2 dc3
273 */
274 dc0 = PIXEL_SPLAT_X4( ( s0 + s2 + 4 ) >> 3 );
275 dc1 = PIXEL_SPLAT_X4( ( s1 + 2 ) >> 2 );
276 dc2 = PIXEL_SPLAT_X4( ( s3 + 2 ) >> 2 );
277 dc3 = PIXEL_SPLAT_X4( ( s1 + s3 + 4 ) >> 3 );
278
279 for( y = 0; y < 4; y++ )
280 {
281 MPIXEL_X4( src+0 ) = dc0;
282 MPIXEL_X4( src+4 ) = dc1;
283 src += FDEC_STRIDE;
284 }
285
286 for( y = 0; y < 4; y++ )
287 {
288 MPIXEL_X4( src+0 ) = dc2;
289 MPIXEL_X4( src+4 ) = dc3;
290 src += FDEC_STRIDE;
291 }
292 }
293
x264_predict_8x8c_h_c(pixel * src)294 void x264_predict_8x8c_h_c( pixel *src )
295 {
296 int i;
297
298 for( i = 0; i < 8; i++ )
299 {
300 pixel4 v = PIXEL_SPLAT_X4( src[-1] );
301 MPIXEL_X4( src+0 ) = v;
302 MPIXEL_X4( src+4 ) = v;
303 src += FDEC_STRIDE;
304 }
305 }
x264_predict_8x8c_v_c(pixel * src)306 void x264_predict_8x8c_v_c( pixel *src )
307 {
308 pixel4 v0 = MPIXEL_X4( src+0-FDEC_STRIDE );
309 pixel4 v1 = MPIXEL_X4( src+4-FDEC_STRIDE );
310 int i;
311
312 for( i = 0; i < 8; i++ )
313 {
314 MPIXEL_X4( src+0 ) = v0;
315 MPIXEL_X4( src+4 ) = v1;
316 src += FDEC_STRIDE;
317 }
318 }
x264_predict_8x8c_p_c(pixel * src)319 void x264_predict_8x8c_p_c( pixel *src )
320 {
321 int H = 0, V = 0;
322 int i;
323 int a;
324 int b;
325 int c;
326 int i00;
327 int y;
328 int x;
329
330 for( i = 0; i < 4; i++ )
331 {
332 H += ( i + 1 ) * ( src[4+i - FDEC_STRIDE] - src[2 - i -FDEC_STRIDE] );
333 V += ( i + 1 ) * ( src[-1 +(i+4)*FDEC_STRIDE] - src[-1+(2-i)*FDEC_STRIDE] );
334 }
335
336 a = 16 * ( src[-1+7*FDEC_STRIDE] + src[7 - FDEC_STRIDE] );
337 b = ( 17 * H + 16 ) >> 5;
338 c = ( 17 * V + 16 ) >> 5;
339 i00 = a -3*b -3*c + 16;
340
341 for( y = 0; y < 8; y++ )
342 {
343 int pix = i00;
344 for( x = 0; x < 8; x++ )
345 {
346 src[x] = x264_clip_pixel( pix>>5 );
347 pix += b;
348 }
349 src += FDEC_STRIDE;
350 i00 += c;
351 }
352 }
353
354 /****************************************************************************
355 * 8x16 prediction for intra chroma block (4:2:2)
356 ****************************************************************************/
357
x264_predict_8x16c_dc_128_c(pixel * src)358 static void x264_predict_8x16c_dc_128_c( pixel *src )
359 {
360 int y;
361
362 for( y = 0; y < 16; y++ )
363 {
364 MPIXEL_X4( src+0 ) = PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) );
365 MPIXEL_X4( src+4 ) = PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) );
366 src += FDEC_STRIDE;
367 }
368 }
x264_predict_8x16c_dc_left_c(pixel * src)369 static void x264_predict_8x16c_dc_left_c( pixel *src )
370 {
371 int i;
372
373 for( i = 0; i < 4; i++ )
374 {
375 int dc = 0;
376 int y;
377 pixel4 dcsplat;
378
379 for( y = 0; y < 4; y++ )
380 dc += src[y*FDEC_STRIDE - 1];
381
382 dcsplat = PIXEL_SPLAT_X4( (dc + 2) >> 2 );
383
384 for( y = 0; y < 4; y++ )
385 {
386 MPIXEL_X4( src+0 ) = dcsplat;
387 MPIXEL_X4( src+4 ) = dcsplat;
388 src += FDEC_STRIDE;
389 }
390 }
391 }
392
x264_predict_8x16c_dc_top_c(pixel * src)393 static void x264_predict_8x16c_dc_top_c( pixel *src )
394 {
395 int dc0 = 0, dc1 = 0;
396 int x;
397 pixel4 dc0splat;
398 pixel4 dc1splat;
399 int y;
400
401 for(x = 0; x < 4; x++ )
402 {
403 dc0 += src[x - FDEC_STRIDE];
404 dc1 += src[x + 4 - FDEC_STRIDE];
405 }
406 dc0splat = PIXEL_SPLAT_X4( ( dc0 + 2 ) >> 2 );
407 dc1splat = PIXEL_SPLAT_X4( ( dc1 + 2 ) >> 2 );
408
409 for( y = 0; y < 16; y++ )
410 {
411 MPIXEL_X4( src+0 ) = dc0splat;
412 MPIXEL_X4( src+4 ) = dc1splat;
413 src += FDEC_STRIDE;
414 }
415 }
416
x264_predict_8x16c_dc_c(pixel * src)417 void x264_predict_8x16c_dc_c( pixel *src )
418 {
419 int s0 = 0, s1 = 0, s2 = 0, s3 = 0, s4 = 0, s5 = 0;
420 int i;
421 pixel4 dc0;
422 pixel4 dc1;
423 pixel4 dc2;
424 pixel4 dc3;
425 pixel4 dc4;
426 pixel4 dc5;
427 pixel4 dc6;
428 pixel4 dc7;
429 int y;
430
431 /*
432 s0 s1
433 s2
434 s3
435 s4
436 s5
437 */
438 for( i = 0; i < 4; i++ )
439 {
440 s0 += src[i+0 - FDEC_STRIDE];
441 s1 += src[i+4 - FDEC_STRIDE];
442 s2 += src[-1 + (i+0) * FDEC_STRIDE];
443 s3 += src[-1 + (i+4) * FDEC_STRIDE];
444 s4 += src[-1 + (i+8) * FDEC_STRIDE];
445 s5 += src[-1 + (i+12) * FDEC_STRIDE];
446 }
447 /*
448 dc0 dc1
449 dc2 dc3
450 dc4 dc5
451 dc6 dc7
452 */
453 dc0 = PIXEL_SPLAT_X4( ( s0 + s2 + 4 ) >> 3 );
454 dc1 = PIXEL_SPLAT_X4( ( s1 + 2 ) >> 2 );
455 dc2 = PIXEL_SPLAT_X4( ( s3 + 2 ) >> 2 );
456 dc3 = PIXEL_SPLAT_X4( ( s1 + s3 + 4 ) >> 3 );
457 dc4 = PIXEL_SPLAT_X4( ( s4 + 2 ) >> 2 );
458 dc5 = PIXEL_SPLAT_X4( ( s1 + s4 + 4 ) >> 3 );
459 dc6 = PIXEL_SPLAT_X4( ( s5 + 2 ) >> 2 );
460 dc7 = PIXEL_SPLAT_X4( ( s1 + s5 + 4 ) >> 3 );
461
462 for( y = 0; y < 4; y++ )
463 {
464 MPIXEL_X4( src+0 ) = dc0;
465 MPIXEL_X4( src+4 ) = dc1;
466 src += FDEC_STRIDE;
467 }
468 for( y = 0; y < 4; y++ )
469 {
470 MPIXEL_X4( src+0 ) = dc2;
471 MPIXEL_X4( src+4 ) = dc3;
472 src += FDEC_STRIDE;
473 }
474 for( y = 0; y < 4; y++ )
475 {
476 MPIXEL_X4( src+0 ) = dc4;
477 MPIXEL_X4( src+4 ) = dc5;
478 src += FDEC_STRIDE;
479 }
480 for( y = 0; y < 4; y++ )
481 {
482 MPIXEL_X4( src+0 ) = dc6;
483 MPIXEL_X4( src+4 ) = dc7;
484 src += FDEC_STRIDE;
485 }
486 }
487
x264_predict_8x16c_h_c(pixel * src)488 void x264_predict_8x16c_h_c( pixel *src )
489 {
490 int i;
491
492 for( i = 0; i < 16; i++ )
493 {
494 pixel4 v = PIXEL_SPLAT_X4( src[-1] );
495 MPIXEL_X4( src+0 ) = v;
496 MPIXEL_X4( src+4 ) = v;
497 src += FDEC_STRIDE;
498 }
499 }
x264_predict_8x16c_v_c(pixel * src)500 void x264_predict_8x16c_v_c( pixel *src )
501 {
502 pixel4 v0 = MPIXEL_X4( src+0-FDEC_STRIDE );
503 pixel4 v1 = MPIXEL_X4( src+4-FDEC_STRIDE );
504 int i;
505
506 for( i = 0; i < 16; i++ )
507 {
508 MPIXEL_X4( src+0 ) = v0;
509 MPIXEL_X4( src+4 ) = v1;
510 src += FDEC_STRIDE;
511 }
512 }
513
x264_predict_8x16c_p_c(pixel * src)514 void x264_predict_8x16c_p_c( pixel *src )
515 {
516 int H = 0;
517 int V = 0;
518 int i;
519 int y;
520 int a;
521 int b;
522 int c;
523 int i00;
524 int x;
525
526 for( i = 0; i < 4; i++ )
527 H += ( i + 1 ) * ( src[4 + i - FDEC_STRIDE] - src[2 - i - FDEC_STRIDE] );
528 for( i = 0; i < 8; i++ )
529 V += ( i + 1 ) * ( src[-1 + (i+8)*FDEC_STRIDE] - src[-1 + (6-i)*FDEC_STRIDE] );
530
531 a = 16 * ( src[-1 + 15*FDEC_STRIDE] + src[7 - FDEC_STRIDE] );
532 b = ( 17 * H + 16 ) >> 5;
533 c = ( 5 * V + 32 ) >> 6;
534 i00 = a -3*b -7*c + 16;
535
536 for( y = 0; y < 16; y++ )
537 {
538 int pix = i00;
539 for( x = 0; x < 8; x++ )
540 {
541 src[x] = x264_clip_pixel( pix>>5 );
542 pix += b;
543 }
544 src += FDEC_STRIDE;
545 i00 += c;
546 }
547 }
548
549 /****************************************************************************
550 * 4x4 prediction for intra luma block
551 ****************************************************************************/
552
553 #define SRC(x,y) src[(x)+(y)*FDEC_STRIDE]
554 #define SRC_X4(x,y) MPIXEL_X4( &SRC(x,y) )
555
556 #define PREDICT_4x4_DC(v)\
557 SRC_X4(0,0) = SRC_X4(0,1) = SRC_X4(0,2) = SRC_X4(0,3) = v;
558
x264_predict_4x4_dc_128_c(pixel * src)559 static void x264_predict_4x4_dc_128_c( pixel *src )
560 {
561 PREDICT_4x4_DC( PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) ) );
562 }
x264_predict_4x4_dc_left_c(pixel * src)563 static void x264_predict_4x4_dc_left_c( pixel *src )
564 {
565 pixel4 dc = PIXEL_SPLAT_X4( (SRC(-1,0) + SRC(-1,1) + SRC(-1,2) + SRC(-1,3) + 2) >> 2 );
566 PREDICT_4x4_DC( dc );
567 }
x264_predict_4x4_dc_top_c(pixel * src)568 static void x264_predict_4x4_dc_top_c( pixel *src )
569 {
570 pixel4 dc = PIXEL_SPLAT_X4( (SRC(0,-1) + SRC(1,-1) + SRC(2,-1) + SRC(3,-1) + 2) >> 2 );
571 PREDICT_4x4_DC( dc );
572 }
x264_predict_4x4_dc_c(pixel * src)573 void x264_predict_4x4_dc_c( pixel *src )
574 {
575 pixel4 dc = PIXEL_SPLAT_X4( (SRC(-1,0) + SRC(-1,1) + SRC(-1,2) + SRC(-1,3) +
576 SRC(0,-1) + SRC(1,-1) + SRC(2,-1) + SRC(3,-1) + 4) >> 3 );
577 PREDICT_4x4_DC( dc );
578 }
x264_predict_4x4_h_c(pixel * src)579 void x264_predict_4x4_h_c( pixel *src )
580 {
581 SRC_X4(0,0) = PIXEL_SPLAT_X4( SRC(-1,0) );
582 SRC_X4(0,1) = PIXEL_SPLAT_X4( SRC(-1,1) );
583 SRC_X4(0,2) = PIXEL_SPLAT_X4( SRC(-1,2) );
584 SRC_X4(0,3) = PIXEL_SPLAT_X4( SRC(-1,3) );
585 }
x264_predict_4x4_v_c(pixel * src)586 void x264_predict_4x4_v_c( pixel *src )
587 {
588 PREDICT_4x4_DC(SRC_X4(0,-1));
589 }
590
591 #define PREDICT_4x4_LOAD_LEFT\
592 int l0 = SRC(-1,0);\
593 int l1 = SRC(-1,1);\
594 int l2 = SRC(-1,2);\
595 UNUSED int l3 = SRC(-1,3);
596
597 #define PREDICT_4x4_LOAD_TOP\
598 int t0 = SRC(0,-1);\
599 int t1 = SRC(1,-1);\
600 int t2 = SRC(2,-1);\
601 UNUSED int t3 = SRC(3,-1);
602
603 #define PREDICT_4x4_LOAD_TOP_RIGHT\
604 int t4 = SRC(4,-1);\
605 int t5 = SRC(5,-1);\
606 int t6 = SRC(6,-1);\
607 UNUSED int t7 = SRC(7,-1);
608
609 #define F1(a,b) (((a)+(b)+1)>>1)
610 #define F2(a,b,c) (((a)+2*(b)+(c)+2)>>2)
611
x264_predict_4x4_ddl_c(pixel * src)612 static void x264_predict_4x4_ddl_c( pixel *src )
613 {
614 PREDICT_4x4_LOAD_TOP
615 PREDICT_4x4_LOAD_TOP_RIGHT
616 SRC(0,0)= F2(t0,t1,t2);
617 SRC(1,0)=SRC(0,1)= F2(t1,t2,t3);
618 SRC(2,0)=SRC(1,1)=SRC(0,2)= F2(t2,t3,t4);
619 SRC(3,0)=SRC(2,1)=SRC(1,2)=SRC(0,3)= F2(t3,t4,t5);
620 SRC(3,1)=SRC(2,2)=SRC(1,3)= F2(t4,t5,t6);
621 SRC(3,2)=SRC(2,3)= F2(t5,t6,t7);
622 SRC(3,3)= F2(t6,t7,t7);
623 }
x264_predict_4x4_ddr_c(pixel * src)624 static void x264_predict_4x4_ddr_c( pixel *src )
625 {
626 int lt = SRC(-1,-1);
627 PREDICT_4x4_LOAD_LEFT
628 PREDICT_4x4_LOAD_TOP
629 SRC(3,0)= F2(t3,t2,t1);
630 SRC(2,0)=SRC(3,1)= F2(t2,t1,t0);
631 SRC(1,0)=SRC(2,1)=SRC(3,2)= F2(t1,t0,lt);
632 SRC(0,0)=SRC(1,1)=SRC(2,2)=SRC(3,3)= F2(t0,lt,l0);
633 SRC(0,1)=SRC(1,2)=SRC(2,3)= F2(lt,l0,l1);
634 SRC(0,2)=SRC(1,3)= F2(l0,l1,l2);
635 SRC(0,3)= F2(l1,l2,l3);
636 }
637
x264_predict_4x4_vr_c(pixel * src)638 static void x264_predict_4x4_vr_c( pixel *src )
639 {
640 int lt = SRC(-1,-1);
641 PREDICT_4x4_LOAD_LEFT
642 PREDICT_4x4_LOAD_TOP
643 SRC(0,3)= F2(l2,l1,l0);
644 SRC(0,2)= F2(l1,l0,lt);
645 SRC(0,1)=SRC(1,3)= F2(l0,lt,t0);
646 SRC(0,0)=SRC(1,2)= F1(lt,t0);
647 SRC(1,1)=SRC(2,3)= F2(lt,t0,t1);
648 SRC(1,0)=SRC(2,2)= F1(t0,t1);
649 SRC(2,1)=SRC(3,3)= F2(t0,t1,t2);
650 SRC(2,0)=SRC(3,2)= F1(t1,t2);
651 SRC(3,1)= F2(t1,t2,t3);
652 SRC(3,0)= F1(t2,t3);
653 }
654
x264_predict_4x4_hd_c(pixel * src)655 static void x264_predict_4x4_hd_c( pixel *src )
656 {
657 int lt= SRC(-1,-1);
658 PREDICT_4x4_LOAD_LEFT
659 PREDICT_4x4_LOAD_TOP
660 SRC(0,3)= F1(l2,l3);
661 SRC(1,3)= F2(l1,l2,l3);
662 SRC(0,2)=SRC(2,3)= F1(l1,l2);
663 SRC(1,2)=SRC(3,3)= F2(l0,l1,l2);
664 SRC(0,1)=SRC(2,2)= F1(l0,l1);
665 SRC(1,1)=SRC(3,2)= F2(lt,l0,l1);
666 SRC(0,0)=SRC(2,1)= F1(lt,l0);
667 SRC(1,0)=SRC(3,1)= F2(t0,lt,l0);
668 SRC(2,0)= F2(t1,t0,lt);
669 SRC(3,0)= F2(t2,t1,t0);
670 }
671
x264_predict_4x4_vl_c(pixel * src)672 static void x264_predict_4x4_vl_c( pixel *src )
673 {
674 PREDICT_4x4_LOAD_TOP
675 PREDICT_4x4_LOAD_TOP_RIGHT
676 SRC(0,0)= F1(t0,t1);
677 SRC(0,1)= F2(t0,t1,t2);
678 SRC(1,0)=SRC(0,2)= F1(t1,t2);
679 SRC(1,1)=SRC(0,3)= F2(t1,t2,t3);
680 SRC(2,0)=SRC(1,2)= F1(t2,t3);
681 SRC(2,1)=SRC(1,3)= F2(t2,t3,t4);
682 SRC(3,0)=SRC(2,2)= F1(t3,t4);
683 SRC(3,1)=SRC(2,3)= F2(t3,t4,t5);
684 SRC(3,2)= F1(t4,t5);
685 SRC(3,3)= F2(t4,t5,t6);
686 }
687
x264_predict_4x4_hu_c(pixel * src)688 static void x264_predict_4x4_hu_c( pixel *src )
689 {
690 PREDICT_4x4_LOAD_LEFT
691 SRC(0,0)= F1(l0,l1);
692 SRC(1,0)= F2(l0,l1,l2);
693 SRC(2,0)=SRC(0,1)= F1(l1,l2);
694 SRC(3,0)=SRC(1,1)= F2(l1,l2,l3);
695 SRC(2,1)=SRC(0,2)= F1(l2,l3);
696 SRC(3,1)=SRC(1,2)= F2(l2,l3,l3);
697 SRC(3,2)=SRC(1,3)=SRC(0,3)=
698 SRC(2,2)=SRC(2,3)=SRC(3,3)= l3;
699 }
700
701 /****************************************************************************
702 * 8x8 prediction for intra luma block
703 ****************************************************************************/
704
705 #define PL(y) \
706 edge[14-y] = F2(SRC(-1,y-1), SRC(-1,y), SRC(-1,y+1));
707 #define PT(x) \
708 edge[16+x] = F2(SRC(x-1,-1), SRC(x,-1), SRC(x+1,-1));
709
x264_predict_8x8_filter_c(pixel * src,pixel edge[36],int i_neighbor,int i_filters)710 static void x264_predict_8x8_filter_c( pixel *src, pixel edge[36], int i_neighbor, int i_filters )
711 {
712 /* edge[7..14] = l7..l0
713 * edge[15] = lt
714 * edge[16..31] = t0 .. t15
715 * edge[32] = t15 */
716
717 int have_lt = i_neighbor & MB_TOPLEFT;
718 if( i_filters & MB_LEFT )
719 {
720 edge[15] = (SRC(0,-1) + 2*SRC(-1,-1) + SRC(-1,0) + 2) >> 2;
721 edge[14] = ((have_lt ? SRC(-1,-1) : SRC(-1,0))
722 + 2*SRC(-1,0) + SRC(-1,1) + 2) >> 2;
723 PL(1) PL(2) PL(3) PL(4) PL(5) PL(6)
724 edge[6] =
725 edge[7] = (SRC(-1,6) + 3*SRC(-1,7) + 2) >> 2;
726 }
727
728 if( i_filters & MB_TOP )
729 {
730 int have_tr = i_neighbor & MB_TOPRIGHT;
731 edge[16] = ((have_lt ? SRC(-1,-1) : SRC(0,-1))
732 + 2*SRC(0,-1) + SRC(1,-1) + 2) >> 2;
733 PT(1) PT(2) PT(3) PT(4) PT(5) PT(6)
734 edge[23] = (SRC(6,-1) + 2*SRC(7,-1)
735 + (have_tr ? SRC(8,-1) : SRC(7,-1)) + 2) >> 2;
736
737 if( i_filters & MB_TOPRIGHT )
738 {
739 if( have_tr )
740 {
741 PT(8) PT(9) PT(10) PT(11) PT(12) PT(13) PT(14)
742 edge[31] =
743 edge[32] = (SRC(14,-1) + 3*SRC(15,-1) + 2) >> 2;
744 }
745 else
746 {
747 MPIXEL_X4( edge+24 ) = PIXEL_SPLAT_X4( SRC(7,-1) );
748 MPIXEL_X4( edge+28 ) = PIXEL_SPLAT_X4( SRC(7,-1) );
749 edge[32] = SRC(7,-1);
750 }
751 }
752 }
753 }
754
755 #undef PL
756 #undef PT
757
758 #define PL(y) \
759 UNUSED int l##y = edge[14-y];
760 #define PT(x) \
761 UNUSED int t##x = edge[16+x];
762 #define PREDICT_8x8_LOAD_TOPLEFT \
763 int lt = edge[15];
764 #define PREDICT_8x8_LOAD_LEFT \
765 PL(0) PL(1) PL(2) PL(3) PL(4) PL(5) PL(6) PL(7)
766 #define PREDICT_8x8_LOAD_TOP \
767 PT(0) PT(1) PT(2) PT(3) PT(4) PT(5) PT(6) PT(7)
768 #define PREDICT_8x8_LOAD_TOPRIGHT \
769 PT(8) PT(9) PT(10) PT(11) PT(12) PT(13) PT(14) PT(15)
770
771 #define PREDICT_8x8_DC(v) \
772 for( int y = 0; y < 8; y++ ) { \
773 MPIXEL_X4( src+0 ) = v; \
774 MPIXEL_X4( src+4 ) = v; \
775 src += FDEC_STRIDE; \
776 }
777
x264_predict_8x8_dc_128_c(pixel * src,pixel edge[36])778 static void x264_predict_8x8_dc_128_c( pixel *src, pixel edge[36] )
779 {
780 PREDICT_8x8_DC( PIXEL_SPLAT_X4( 1 << (BIT_DEPTH-1) ) );
781 }
782
x264_predict_8x8_dc_left_c(pixel * src,pixel edge[36])783 static void x264_predict_8x8_dc_left_c( pixel *src, pixel edge[36] )
784 {
785 PREDICT_8x8_LOAD_LEFT
786 pixel4 dc = PIXEL_SPLAT_X4( (l0+l1+l2+l3+l4+l5+l6+l7+4) >> 3 );
787 PREDICT_8x8_DC( dc );
788 }
789
x264_predict_8x8_dc_top_c(pixel * src,pixel edge[36])790 static void x264_predict_8x8_dc_top_c( pixel *src, pixel edge[36] )
791 {
792 PREDICT_8x8_LOAD_TOP
793 pixel4 dc = PIXEL_SPLAT_X4( (t0+t1+t2+t3+t4+t5+t6+t7+4) >> 3 );
794 PREDICT_8x8_DC( dc );
795 }
796
x264_predict_8x8_dc_c(pixel * src,pixel edge[36])797 void x264_predict_8x8_dc_c( pixel *src, pixel edge[36] )
798 {
799 PREDICT_8x8_LOAD_LEFT
800 PREDICT_8x8_LOAD_TOP
801 pixel4 dc = PIXEL_SPLAT_X4( (l0+l1+l2+l3+l4+l5+l6+l7+t0+t1+t2+t3+t4+t5+t6+t7+8) >> 4 );
802 PREDICT_8x8_DC( dc );
803 }
804
x264_predict_8x8_h_c(pixel * src,pixel edge[36])805 void x264_predict_8x8_h_c( pixel *src, pixel edge[36] )
806 {
807 PREDICT_8x8_LOAD_LEFT
808 #define ROW(y) MPIXEL_X4( src+y*FDEC_STRIDE+0 ) =\
809 MPIXEL_X4( src+y*FDEC_STRIDE+4 ) = PIXEL_SPLAT_X4( l##y );
810 ROW(0); ROW(1); ROW(2); ROW(3); ROW(4); ROW(5); ROW(6); ROW(7);
811 #undef ROW
812 }
813
x264_predict_8x8_v_c(pixel * src,pixel edge[36])814 void x264_predict_8x8_v_c( pixel *src, pixel edge[36] )
815 {
816 pixel4 top[2] = { MPIXEL_X4( edge+16 ),
817 MPIXEL_X4( edge+20 ) };
818 int y;
819
820 for( y = 0; y < 8; y++ )
821 {
822 MPIXEL_X4( src+y*FDEC_STRIDE+0 ) = top[0];
823 MPIXEL_X4( src+y*FDEC_STRIDE+4 ) = top[1];
824 }
825 }
826
x264_predict_8x8_ddl_c(pixel * src,pixel edge[36])827 static void x264_predict_8x8_ddl_c( pixel *src, pixel edge[36] )
828 {
829 PREDICT_8x8_LOAD_TOP
830 PREDICT_8x8_LOAD_TOPRIGHT
831 SRC(0,0)= F2(t0,t1,t2);
832 SRC(0,1)=SRC(1,0)= F2(t1,t2,t3);
833 SRC(0,2)=SRC(1,1)=SRC(2,0)= F2(t2,t3,t4);
834 SRC(0,3)=SRC(1,2)=SRC(2,1)=SRC(3,0)= F2(t3,t4,t5);
835 SRC(0,4)=SRC(1,3)=SRC(2,2)=SRC(3,1)=SRC(4,0)= F2(t4,t5,t6);
836 SRC(0,5)=SRC(1,4)=SRC(2,3)=SRC(3,2)=SRC(4,1)=SRC(5,0)= F2(t5,t6,t7);
837 SRC(0,6)=SRC(1,5)=SRC(2,4)=SRC(3,3)=SRC(4,2)=SRC(5,1)=SRC(6,0)= F2(t6,t7,t8);
838 SRC(0,7)=SRC(1,6)=SRC(2,5)=SRC(3,4)=SRC(4,3)=SRC(5,2)=SRC(6,1)=SRC(7,0)= F2(t7,t8,t9);
839 SRC(1,7)=SRC(2,6)=SRC(3,5)=SRC(4,4)=SRC(5,3)=SRC(6,2)=SRC(7,1)= F2(t8,t9,t10);
840 SRC(2,7)=SRC(3,6)=SRC(4,5)=SRC(5,4)=SRC(6,3)=SRC(7,2)= F2(t9,t10,t11);
841 SRC(3,7)=SRC(4,6)=SRC(5,5)=SRC(6,4)=SRC(7,3)= F2(t10,t11,t12);
842 SRC(4,7)=SRC(5,6)=SRC(6,5)=SRC(7,4)= F2(t11,t12,t13);
843 SRC(5,7)=SRC(6,6)=SRC(7,5)= F2(t12,t13,t14);
844 SRC(6,7)=SRC(7,6)= F2(t13,t14,t15);
845 SRC(7,7)= F2(t14,t15,t15);
846 }
x264_predict_8x8_ddr_c(pixel * src,pixel edge[36])847 static void x264_predict_8x8_ddr_c( pixel *src, pixel edge[36] )
848 {
849 PREDICT_8x8_LOAD_TOP
850 PREDICT_8x8_LOAD_LEFT
851 PREDICT_8x8_LOAD_TOPLEFT
852 SRC(0,7)= F2(l7,l6,l5);
853 SRC(0,6)=SRC(1,7)= F2(l6,l5,l4);
854 SRC(0,5)=SRC(1,6)=SRC(2,7)= F2(l5,l4,l3);
855 SRC(0,4)=SRC(1,5)=SRC(2,6)=SRC(3,7)= F2(l4,l3,l2);
856 SRC(0,3)=SRC(1,4)=SRC(2,5)=SRC(3,6)=SRC(4,7)= F2(l3,l2,l1);
857 SRC(0,2)=SRC(1,3)=SRC(2,4)=SRC(3,5)=SRC(4,6)=SRC(5,7)= F2(l2,l1,l0);
858 SRC(0,1)=SRC(1,2)=SRC(2,3)=SRC(3,4)=SRC(4,5)=SRC(5,6)=SRC(6,7)= F2(l1,l0,lt);
859 SRC(0,0)=SRC(1,1)=SRC(2,2)=SRC(3,3)=SRC(4,4)=SRC(5,5)=SRC(6,6)=SRC(7,7)= F2(l0,lt,t0);
860 SRC(1,0)=SRC(2,1)=SRC(3,2)=SRC(4,3)=SRC(5,4)=SRC(6,5)=SRC(7,6)= F2(lt,t0,t1);
861 SRC(2,0)=SRC(3,1)=SRC(4,2)=SRC(5,3)=SRC(6,4)=SRC(7,5)= F2(t0,t1,t2);
862 SRC(3,0)=SRC(4,1)=SRC(5,2)=SRC(6,3)=SRC(7,4)= F2(t1,t2,t3);
863 SRC(4,0)=SRC(5,1)=SRC(6,2)=SRC(7,3)= F2(t2,t3,t4);
864 SRC(5,0)=SRC(6,1)=SRC(7,2)= F2(t3,t4,t5);
865 SRC(6,0)=SRC(7,1)= F2(t4,t5,t6);
866 SRC(7,0)= F2(t5,t6,t7);
867
868 }
x264_predict_8x8_vr_c(pixel * src,pixel edge[36])869 static void x264_predict_8x8_vr_c( pixel *src, pixel edge[36] )
870 {
871 PREDICT_8x8_LOAD_TOP
872 PREDICT_8x8_LOAD_LEFT
873 PREDICT_8x8_LOAD_TOPLEFT
874 SRC(0,6)= F2(l5,l4,l3);
875 SRC(0,7)= F2(l6,l5,l4);
876 SRC(0,4)=SRC(1,6)= F2(l3,l2,l1);
877 SRC(0,5)=SRC(1,7)= F2(l4,l3,l2);
878 SRC(0,2)=SRC(1,4)=SRC(2,6)= F2(l1,l0,lt);
879 SRC(0,3)=SRC(1,5)=SRC(2,7)= F2(l2,l1,l0);
880 SRC(0,1)=SRC(1,3)=SRC(2,5)=SRC(3,7)= F2(l0,lt,t0);
881 SRC(0,0)=SRC(1,2)=SRC(2,4)=SRC(3,6)= F1(lt,t0);
882 SRC(1,1)=SRC(2,3)=SRC(3,5)=SRC(4,7)= F2(lt,t0,t1);
883 SRC(1,0)=SRC(2,2)=SRC(3,4)=SRC(4,6)= F1(t0,t1);
884 SRC(2,1)=SRC(3,3)=SRC(4,5)=SRC(5,7)= F2(t0,t1,t2);
885 SRC(2,0)=SRC(3,2)=SRC(4,4)=SRC(5,6)= F1(t1,t2);
886 SRC(3,1)=SRC(4,3)=SRC(5,5)=SRC(6,7)= F2(t1,t2,t3);
887 SRC(3,0)=SRC(4,2)=SRC(5,4)=SRC(6,6)= F1(t2,t3);
888 SRC(4,1)=SRC(5,3)=SRC(6,5)=SRC(7,7)= F2(t2,t3,t4);
889 SRC(4,0)=SRC(5,2)=SRC(6,4)=SRC(7,6)= F1(t3,t4);
890 SRC(5,1)=SRC(6,3)=SRC(7,5)= F2(t3,t4,t5);
891 SRC(5,0)=SRC(6,2)=SRC(7,4)= F1(t4,t5);
892 SRC(6,1)=SRC(7,3)= F2(t4,t5,t6);
893 SRC(6,0)=SRC(7,2)= F1(t5,t6);
894 SRC(7,1)= F2(t5,t6,t7);
895 SRC(7,0)= F1(t6,t7);
896 }
x264_predict_8x8_hd_c(pixel * src,pixel edge[36])897 static void x264_predict_8x8_hd_c( pixel *src, pixel edge[36] )
898 {
899 PREDICT_8x8_LOAD_TOP
900 PREDICT_8x8_LOAD_LEFT
901 PREDICT_8x8_LOAD_TOPLEFT
902 int p1 = pack_pixel_1to2(F1(l6,l7), F2(l5,l6,l7));
903 int p2 = pack_pixel_1to2(F1(l5,l6), F2(l4,l5,l6));
904 int p3 = pack_pixel_1to2(F1(l4,l5), F2(l3,l4,l5));
905 int p4 = pack_pixel_1to2(F1(l3,l4), F2(l2,l3,l4));
906 int p5 = pack_pixel_1to2(F1(l2,l3), F2(l1,l2,l3));
907 int p6 = pack_pixel_1to2(F1(l1,l2), F2(l0,l1,l2));
908 int p7 = pack_pixel_1to2(F1(l0,l1), F2(lt,l0,l1));
909 int p8 = pack_pixel_1to2(F1(lt,l0), F2(l0,lt,t0));
910 int p9 = pack_pixel_1to2(F2(t1,t0,lt), F2(t2,t1,t0));
911 int p10 = pack_pixel_1to2(F2(t3,t2,t1), F2(t4,t3,t2));
912 int p11 = pack_pixel_1to2(F2(t5,t4,t3), F2(t6,t5,t4));
913 SRC_X4(0,7)= pack_pixel_2to4(p1,p2);
914 SRC_X4(0,6)= pack_pixel_2to4(p2,p3);
915 SRC_X4(4,7)=SRC_X4(0,5)= pack_pixel_2to4(p3,p4);
916 SRC_X4(4,6)=SRC_X4(0,4)= pack_pixel_2to4(p4,p5);
917 SRC_X4(4,5)=SRC_X4(0,3)= pack_pixel_2to4(p5,p6);
918 SRC_X4(4,4)=SRC_X4(0,2)= pack_pixel_2to4(p6,p7);
919 SRC_X4(4,3)=SRC_X4(0,1)= pack_pixel_2to4(p7,p8);
920 SRC_X4(4,2)=SRC_X4(0,0)= pack_pixel_2to4(p8,p9);
921 SRC_X4(4,1)= pack_pixel_2to4(p9,p10);
922 SRC_X4(4,0)= pack_pixel_2to4(p10,p11);
923 }
x264_predict_8x8_vl_c(pixel * src,pixel edge[36])924 static void x264_predict_8x8_vl_c( pixel *src, pixel edge[36] )
925 {
926 PREDICT_8x8_LOAD_TOP
927 PREDICT_8x8_LOAD_TOPRIGHT
928 SRC(0,0)= F1(t0,t1);
929 SRC(0,1)= F2(t0,t1,t2);
930 SRC(0,2)=SRC(1,0)= F1(t1,t2);
931 SRC(0,3)=SRC(1,1)= F2(t1,t2,t3);
932 SRC(0,4)=SRC(1,2)=SRC(2,0)= F1(t2,t3);
933 SRC(0,5)=SRC(1,3)=SRC(2,1)= F2(t2,t3,t4);
934 SRC(0,6)=SRC(1,4)=SRC(2,2)=SRC(3,0)= F1(t3,t4);
935 SRC(0,7)=SRC(1,5)=SRC(2,3)=SRC(3,1)= F2(t3,t4,t5);
936 SRC(1,6)=SRC(2,4)=SRC(3,2)=SRC(4,0)= F1(t4,t5);
937 SRC(1,7)=SRC(2,5)=SRC(3,3)=SRC(4,1)= F2(t4,t5,t6);
938 SRC(2,6)=SRC(3,4)=SRC(4,2)=SRC(5,0)= F1(t5,t6);
939 SRC(2,7)=SRC(3,5)=SRC(4,3)=SRC(5,1)= F2(t5,t6,t7);
940 SRC(3,6)=SRC(4,4)=SRC(5,2)=SRC(6,0)= F1(t6,t7);
941 SRC(3,7)=SRC(4,5)=SRC(5,3)=SRC(6,1)= F2(t6,t7,t8);
942 SRC(4,6)=SRC(5,4)=SRC(6,2)=SRC(7,0)= F1(t7,t8);
943 SRC(4,7)=SRC(5,5)=SRC(6,3)=SRC(7,1)= F2(t7,t8,t9);
944 SRC(5,6)=SRC(6,4)=SRC(7,2)= F1(t8,t9);
945 SRC(5,7)=SRC(6,5)=SRC(7,3)= F2(t8,t9,t10);
946 SRC(6,6)=SRC(7,4)= F1(t9,t10);
947 SRC(6,7)=SRC(7,5)= F2(t9,t10,t11);
948 SRC(7,6)= F1(t10,t11);
949 SRC(7,7)= F2(t10,t11,t12);
950 }
x264_predict_8x8_hu_c(pixel * src,pixel edge[36])951 static void x264_predict_8x8_hu_c( pixel *src, pixel edge[36] )
952 {
953 PREDICT_8x8_LOAD_LEFT
954 int p1 = pack_pixel_1to2(F1(l0,l1), F2(l0,l1,l2));
955 int p2 = pack_pixel_1to2(F1(l1,l2), F2(l1,l2,l3));
956 int p3 = pack_pixel_1to2(F1(l2,l3), F2(l2,l3,l4));
957 int p4 = pack_pixel_1to2(F1(l3,l4), F2(l3,l4,l5));
958 int p5 = pack_pixel_1to2(F1(l4,l5), F2(l4,l5,l6));
959 int p6 = pack_pixel_1to2(F1(l5,l6), F2(l5,l6,l7));
960 int p7 = pack_pixel_1to2(F1(l6,l7), F2(l6,l7,l7));
961 int p8 = pack_pixel_1to2(l7,l7);
962 SRC_X4(0,0)= pack_pixel_2to4(p1,p2);
963 SRC_X4(0,1)= pack_pixel_2to4(p2,p3);
964 SRC_X4(4,0)=SRC_X4(0,2)= pack_pixel_2to4(p3,p4);
965 SRC_X4(4,1)=SRC_X4(0,3)= pack_pixel_2to4(p4,p5);
966 SRC_X4(4,2)=SRC_X4(0,4)= pack_pixel_2to4(p5,p6);
967 SRC_X4(4,3)=SRC_X4(0,5)= pack_pixel_2to4(p6,p7);
968 SRC_X4(4,4)=SRC_X4(0,6)= pack_pixel_2to4(p7,p8);
969 SRC_X4(4,5)=SRC_X4(4,6)= SRC_X4(0,7) = SRC_X4(4,7) = pack_pixel_2to4(p8,p8);
970 }
971
972 /****************************************************************************
973 * Exported functions:
974 ****************************************************************************/
x264_predict_16x16_init(int cpu,x264_predict_t pf[7])975 void x264_predict_16x16_init( int cpu, x264_predict_t pf[7] )
976 {
977 pf[I_PRED_16x16_V ] = x264_predict_16x16_v_c;
978 pf[I_PRED_16x16_H ] = x264_predict_16x16_h_c;
979 pf[I_PRED_16x16_DC] = x264_predict_16x16_dc_c;
980 pf[I_PRED_16x16_P ] = x264_predict_16x16_p_c;
981 pf[I_PRED_16x16_DC_LEFT]= x264_predict_16x16_dc_left_c;
982 pf[I_PRED_16x16_DC_TOP ]= x264_predict_16x16_dc_top_c;
983 pf[I_PRED_16x16_DC_128 ]= x264_predict_16x16_dc_128_c;
984
985 #if HAVE_MMX
986 x264_predict_16x16_init_mmx( cpu, pf );
987 #endif
988
989 #if HAVE_ALTIVEC
990 if( cpu&X264_CPU_ALTIVEC )
991 x264_predict_16x16_init_altivec( pf );
992 #endif
993
994 #if HAVE_ARMV6
995 x264_predict_16x16_init_arm( cpu, pf );
996 #endif
997
998 #if ARCH_AARCH64
999 x264_predict_16x16_init_aarch64( cpu, pf );
1000 #endif
1001 }
1002
x264_predict_8x8c_init(int cpu,x264_predict_t pf[7])1003 void x264_predict_8x8c_init( int cpu, x264_predict_t pf[7] )
1004 {
1005 pf[I_PRED_CHROMA_V ] = x264_predict_8x8c_v_c;
1006 pf[I_PRED_CHROMA_H ] = x264_predict_8x8c_h_c;
1007 pf[I_PRED_CHROMA_DC] = x264_predict_8x8c_dc_c;
1008 pf[I_PRED_CHROMA_P ] = x264_predict_8x8c_p_c;
1009 pf[I_PRED_CHROMA_DC_LEFT]= x264_predict_8x8c_dc_left_c;
1010 pf[I_PRED_CHROMA_DC_TOP ]= x264_predict_8x8c_dc_top_c;
1011 pf[I_PRED_CHROMA_DC_128 ]= x264_predict_8x8c_dc_128_c;
1012
1013 #if HAVE_MMX
1014 x264_predict_8x8c_init_mmx( cpu, pf );
1015 #endif
1016
1017 #if HAVE_ALTIVEC
1018 if( cpu&X264_CPU_ALTIVEC )
1019 x264_predict_8x8c_init_altivec( pf );
1020 #endif
1021
1022 #if HAVE_ARMV6
1023 x264_predict_8x8c_init_arm( cpu, pf );
1024 #endif
1025
1026 #if ARCH_AARCH64
1027 x264_predict_8x8c_init_aarch64( cpu, pf );
1028 #endif
1029 }
1030
x264_predict_8x16c_init(int cpu,x264_predict_t pf[7])1031 void x264_predict_8x16c_init( int cpu, x264_predict_t pf[7] )
1032 {
1033 pf[I_PRED_CHROMA_V ] = x264_predict_8x16c_v_c;
1034 pf[I_PRED_CHROMA_H ] = x264_predict_8x16c_h_c;
1035 pf[I_PRED_CHROMA_DC] = x264_predict_8x16c_dc_c;
1036 pf[I_PRED_CHROMA_P ] = x264_predict_8x16c_p_c;
1037 pf[I_PRED_CHROMA_DC_LEFT]= x264_predict_8x16c_dc_left_c;
1038 pf[I_PRED_CHROMA_DC_TOP ]= x264_predict_8x16c_dc_top_c;
1039 pf[I_PRED_CHROMA_DC_128 ]= x264_predict_8x16c_dc_128_c;
1040
1041 #if HAVE_MMX
1042 x264_predict_8x16c_init_mmx( cpu, pf );
1043 #endif
1044 }
1045
x264_predict_8x8_init(int cpu,x264_predict8x8_t pf[12],x264_predict_8x8_filter_t * predict_filter)1046 void x264_predict_8x8_init( int cpu, x264_predict8x8_t pf[12], x264_predict_8x8_filter_t *predict_filter )
1047 {
1048 pf[I_PRED_8x8_V] = x264_predict_8x8_v_c;
1049 pf[I_PRED_8x8_H] = x264_predict_8x8_h_c;
1050 pf[I_PRED_8x8_DC] = x264_predict_8x8_dc_c;
1051 pf[I_PRED_8x8_DDL] = x264_predict_8x8_ddl_c;
1052 pf[I_PRED_8x8_DDR] = x264_predict_8x8_ddr_c;
1053 pf[I_PRED_8x8_VR] = x264_predict_8x8_vr_c;
1054 pf[I_PRED_8x8_HD] = x264_predict_8x8_hd_c;
1055 pf[I_PRED_8x8_VL] = x264_predict_8x8_vl_c;
1056 pf[I_PRED_8x8_HU] = x264_predict_8x8_hu_c;
1057 pf[I_PRED_8x8_DC_LEFT]= x264_predict_8x8_dc_left_c;
1058 pf[I_PRED_8x8_DC_TOP] = x264_predict_8x8_dc_top_c;
1059 pf[I_PRED_8x8_DC_128] = x264_predict_8x8_dc_128_c;
1060 *predict_filter = x264_predict_8x8_filter_c;
1061
1062 #if HAVE_MMX
1063 x264_predict_8x8_init_mmx( cpu, pf, predict_filter );
1064 #endif
1065
1066 #if HAVE_ARMV6
1067 x264_predict_8x8_init_arm( cpu, pf, predict_filter );
1068 #endif
1069
1070 #if ARCH_AARCH64
1071 x264_predict_8x8_init_aarch64( cpu, pf, predict_filter );
1072 #endif
1073 }
1074
x264_predict_4x4_init(int cpu,x264_predict_t pf[12])1075 void x264_predict_4x4_init( int cpu, x264_predict_t pf[12] )
1076 {
1077 pf[I_PRED_4x4_V] = x264_predict_4x4_v_c;
1078 pf[I_PRED_4x4_H] = x264_predict_4x4_h_c;
1079 pf[I_PRED_4x4_DC] = x264_predict_4x4_dc_c;
1080 pf[I_PRED_4x4_DDL] = x264_predict_4x4_ddl_c;
1081 pf[I_PRED_4x4_DDR] = x264_predict_4x4_ddr_c;
1082 pf[I_PRED_4x4_VR] = x264_predict_4x4_vr_c;
1083 pf[I_PRED_4x4_HD] = x264_predict_4x4_hd_c;
1084 pf[I_PRED_4x4_VL] = x264_predict_4x4_vl_c;
1085 pf[I_PRED_4x4_HU] = x264_predict_4x4_hu_c;
1086 pf[I_PRED_4x4_DC_LEFT]= x264_predict_4x4_dc_left_c;
1087 pf[I_PRED_4x4_DC_TOP] = x264_predict_4x4_dc_top_c;
1088 pf[I_PRED_4x4_DC_128] = x264_predict_4x4_dc_128_c;
1089
1090 #if HAVE_MMX
1091 x264_predict_4x4_init_mmx( cpu, pf );
1092 #endif
1093
1094 #if HAVE_ARMV6
1095 x264_predict_4x4_init_arm( cpu, pf );
1096 #endif
1097
1098 #if ARCH_AARCH64
1099 x264_predict_4x4_init_aarch64( cpu, pf );
1100 #endif
1101 }
1102
1103