rv34.c
Go to the documentation of this file.
1 /*
2  * RV30/40 decoder common data
3  * Copyright (c) 2007 Mike Melanson, Konstantin Shishkov
4  *
5  * This file is part of Libav.
6  *
7  * Libav 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  * Libav 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 Libav; if not, write to the Free Software
19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20  */
21 
27 #include "libavutil/internal.h"
28 
29 #include "avcodec.h"
30 #include "dsputil.h"
31 #include "mpegvideo.h"
32 #include "golomb.h"
33 #include "internal.h"
34 #include "mathops.h"
35 #include "rectangle.h"
36 #include "thread.h"
37 
38 #include "rv34vlc.h"
39 #include "rv34data.h"
40 #include "rv34.h"
41 
42 //#define DEBUG
43 
44 static inline void ZERO8x2(void* dst, int stride)
45 {
46  fill_rectangle(dst, 1, 2, stride, 0, 4);
47  fill_rectangle(((uint8_t*)(dst))+4, 1, 2, stride, 0, 4);
48 }
49 
51 static const int rv34_mb_type_to_lavc[12] = {
64 };
65 
66 
68 
69 static int rv34_decode_mv(RV34DecContext *r, int block_type);
70 
76 static const int table_offs[] = {
77  0, 1818, 3622, 4144, 4698, 5234, 5804, 5868, 5900, 5932,
78  5996, 6252, 6316, 6348, 6380, 7674, 8944, 10274, 11668, 12250,
79  14060, 15846, 16372, 16962, 17512, 18148, 18180, 18212, 18244, 18308,
80  18564, 18628, 18660, 18692, 20036, 21314, 22648, 23968, 24614, 26384,
81  28190, 28736, 29366, 29938, 30608, 30640, 30672, 30704, 30768, 31024,
82  31088, 31120, 31184, 32570, 33898, 35236, 36644, 37286, 39020, 40802,
83  41368, 42052, 42692, 43348, 43380, 43412, 43444, 43476, 43604, 43668,
84  43700, 43732, 45100, 46430, 47778, 49160, 49802, 51550, 53340, 53972,
85  54648, 55348, 55994, 56122, 56154, 56186, 56218, 56346, 56410, 56442,
86  56474, 57878, 59290, 60636, 62036, 62682, 64460, 64524, 64588, 64716,
87  64844, 66076, 67466, 67978, 68542, 69064, 69648, 70296, 72010, 72074,
88  72138, 72202, 72330, 73572, 74936, 75454, 76030, 76566, 77176, 77822,
89  79582, 79646, 79678, 79742, 79870, 81180, 82536, 83064, 83672, 84242,
90  84934, 85576, 87384, 87448, 87480, 87544, 87672, 88982, 90340, 90902,
91  91598, 92182, 92846, 93488, 95246, 95278, 95310, 95374, 95502, 96878,
92  98266, 98848, 99542, 100234, 100884, 101524, 103320, 103352, 103384, 103416,
93  103480, 104874, 106222, 106910, 107584, 108258, 108902, 109544, 111366, 111398,
94  111430, 111462, 111494, 112878, 114320, 114988, 115660, 116310, 116950, 117592
95 };
96 
97 static VLC_TYPE table_data[117592][2];
98 
107 static void rv34_gen_vlc(const uint8_t *bits, int size, VLC *vlc, const uint8_t *insyms,
108  const int num)
109 {
110  int i;
111  int counts[17] = {0}, codes[17];
112  uint16_t cw[MAX_VLC_SIZE], syms[MAX_VLC_SIZE];
113  uint8_t bits2[MAX_VLC_SIZE];
114  int maxbits = 0, realsize = 0;
115 
116  for(i = 0; i < size; i++){
117  if(bits[i]){
118  bits2[realsize] = bits[i];
119  syms[realsize] = insyms ? insyms[i] : i;
120  realsize++;
121  maxbits = FFMAX(maxbits, bits[i]);
122  counts[bits[i]]++;
123  }
124  }
125 
126  codes[0] = 0;
127  for(i = 0; i < 16; i++)
128  codes[i+1] = (codes[i] + counts[i]) << 1;
129  for(i = 0; i < realsize; i++)
130  cw[i] = codes[bits2[i]]++;
131 
132  vlc->table = &table_data[table_offs[num]];
133  vlc->table_allocated = table_offs[num + 1] - table_offs[num];
134  ff_init_vlc_sparse(vlc, FFMIN(maxbits, 9), realsize,
135  bits2, 1, 1,
136  cw, 2, 2,
137  syms, 2, 2, INIT_VLC_USE_NEW_STATIC);
138 }
139 
143 static av_cold void rv34_init_tables(void)
144 {
145  int i, j, k;
146 
147  for(i = 0; i < NUM_INTRA_TABLES; i++){
148  for(j = 0; j < 2; j++){
149  rv34_gen_vlc(rv34_table_intra_cbppat [i][j], CBPPAT_VLC_SIZE, &intra_vlcs[i].cbppattern[j], NULL, 19*i + 0 + j);
150  rv34_gen_vlc(rv34_table_intra_secondpat[i][j], OTHERBLK_VLC_SIZE, &intra_vlcs[i].second_pattern[j], NULL, 19*i + 2 + j);
151  rv34_gen_vlc(rv34_table_intra_thirdpat [i][j], OTHERBLK_VLC_SIZE, &intra_vlcs[i].third_pattern[j], NULL, 19*i + 4 + j);
152  for(k = 0; k < 4; k++){
153  rv34_gen_vlc(rv34_table_intra_cbp[i][j+k*2], CBP_VLC_SIZE, &intra_vlcs[i].cbp[j][k], rv34_cbp_code, 19*i + 6 + j*4 + k);
154  }
155  }
156  for(j = 0; j < 4; j++){
157  rv34_gen_vlc(rv34_table_intra_firstpat[i][j], FIRSTBLK_VLC_SIZE, &intra_vlcs[i].first_pattern[j], NULL, 19*i + 14 + j);
158  }
159  rv34_gen_vlc(rv34_intra_coeff[i], COEFF_VLC_SIZE, &intra_vlcs[i].coefficient, NULL, 19*i + 18);
160  }
161 
162  for(i = 0; i < NUM_INTER_TABLES; i++){
163  rv34_gen_vlc(rv34_inter_cbppat[i], CBPPAT_VLC_SIZE, &inter_vlcs[i].cbppattern[0], NULL, i*12 + 95);
164  for(j = 0; j < 4; j++){
165  rv34_gen_vlc(rv34_inter_cbp[i][j], CBP_VLC_SIZE, &inter_vlcs[i].cbp[0][j], rv34_cbp_code, i*12 + 96 + j);
166  }
167  for(j = 0; j < 2; j++){
168  rv34_gen_vlc(rv34_table_inter_firstpat [i][j], FIRSTBLK_VLC_SIZE, &inter_vlcs[i].first_pattern[j], NULL, i*12 + 100 + j);
169  rv34_gen_vlc(rv34_table_inter_secondpat[i][j], OTHERBLK_VLC_SIZE, &inter_vlcs[i].second_pattern[j], NULL, i*12 + 102 + j);
170  rv34_gen_vlc(rv34_table_inter_thirdpat [i][j], OTHERBLK_VLC_SIZE, &inter_vlcs[i].third_pattern[j], NULL, i*12 + 104 + j);
171  }
172  rv34_gen_vlc(rv34_inter_coeff[i], COEFF_VLC_SIZE, &inter_vlcs[i].coefficient, NULL, i*12 + 106);
173  }
174 }
175  // vlc group
177 
186 static int rv34_decode_cbp(GetBitContext *gb, RV34VLC *vlc, int table)
187 {
188  int pattern, code, cbp=0;
189  int ones;
190  static const int cbp_masks[3] = {0x100000, 0x010000, 0x110000};
191  static const int shifts[4] = { 0, 2, 8, 10 };
192  const int *curshift = shifts;
193  int i, t, mask;
194 
195  code = get_vlc2(gb, vlc->cbppattern[table].table, 9, 2);
196  pattern = code & 0xF;
197  code >>= 4;
198 
199  ones = rv34_count_ones[pattern];
200 
201  for(mask = 8; mask; mask >>= 1, curshift++){
202  if(pattern & mask)
203  cbp |= get_vlc2(gb, vlc->cbp[table][ones].table, vlc->cbp[table][ones].bits, 1) << curshift[0];
204  }
205 
206  for(i = 0; i < 4; i++){
207  t = (modulo_three_table[code] >> (6 - 2*i)) & 3;
208  if(t == 1)
209  cbp |= cbp_masks[get_bits1(gb)] << i;
210  if(t == 2)
211  cbp |= cbp_masks[2] << i;
212  }
213  return cbp;
214 }
215 
219 static inline void decode_coeff(DCTELEM *dst, int coef, int esc, GetBitContext *gb, VLC* vlc, int q)
220 {
221  if(coef){
222  if(coef == esc){
223  coef = get_vlc2(gb, vlc->table, 9, 2);
224  if(coef > 23){
225  coef -= 23;
226  coef = 22 + ((1 << coef) | get_bits(gb, coef));
227  }
228  coef += esc;
229  }
230  if(get_bits1(gb))
231  coef = -coef;
232  *dst = (coef*q + 8) >> 4;
233  }
234 }
235 
239 static inline void decode_subblock(DCTELEM *dst, int code, const int is_block2, GetBitContext *gb, VLC *vlc, int q)
240 {
241  int flags = modulo_three_table[code];
242 
243  decode_coeff( dst+0*4+0, (flags >> 6) , 3, gb, vlc, q);
244  if(is_block2){
245  decode_coeff(dst+1*4+0, (flags >> 4) & 3, 2, gb, vlc, q);
246  decode_coeff(dst+0*4+1, (flags >> 2) & 3, 2, gb, vlc, q);
247  }else{
248  decode_coeff(dst+0*4+1, (flags >> 4) & 3, 2, gb, vlc, q);
249  decode_coeff(dst+1*4+0, (flags >> 2) & 3, 2, gb, vlc, q);
250  }
251  decode_coeff( dst+1*4+1, (flags >> 0) & 3, 2, gb, vlc, q);
252 }
253 
257 static inline void decode_subblock1(DCTELEM *dst, int code, GetBitContext *gb, VLC *vlc, int q)
258 {
259  int coeff = modulo_three_table[code] >> 6;
260  decode_coeff(dst, coeff, 3, gb, vlc, q);
261 }
262 
263 static inline void decode_subblock3(DCTELEM *dst, int code, const int is_block2, GetBitContext *gb, VLC *vlc,
264  int q_dc, int q_ac1, int q_ac2)
265 {
266  int flags = modulo_three_table[code];
267 
268  decode_coeff( dst+0*4+0, (flags >> 6) , 3, gb, vlc, q_dc);
269  if(is_block2){
270  decode_coeff(dst+1*4+0, (flags >> 4) & 3, 2, gb, vlc, q_ac1);
271  decode_coeff(dst+0*4+1, (flags >> 2) & 3, 2, gb, vlc, q_ac1);
272  }else{
273  decode_coeff(dst+0*4+1, (flags >> 4) & 3, 2, gb, vlc, q_ac1);
274  decode_coeff(dst+1*4+0, (flags >> 2) & 3, 2, gb, vlc, q_ac1);
275  }
276  decode_coeff( dst+1*4+1, (flags >> 0) & 3, 2, gb, vlc, q_ac2);
277 }
278 
290 static inline int rv34_decode_block(DCTELEM *dst, GetBitContext *gb, RV34VLC *rvlc, int fc, int sc, int q_dc, int q_ac1, int q_ac2)
291 {
292  int code, pattern, has_ac = 1;
293 
294  code = get_vlc2(gb, rvlc->first_pattern[fc].table, 9, 2);
295 
296  pattern = code & 0x7;
297 
298  code >>= 3;
299 
300  if (modulo_three_table[code] & 0x3F) {
301  decode_subblock3(dst, code, 0, gb, &rvlc->coefficient, q_dc, q_ac1, q_ac2);
302  } else {
303  decode_subblock1(dst, code, gb, &rvlc->coefficient, q_dc);
304  if (!pattern)
305  return 0;
306  has_ac = 0;
307  }
308 
309  if(pattern & 4){
310  code = get_vlc2(gb, rvlc->second_pattern[sc].table, 9, 2);
311  decode_subblock(dst + 4*0+2, code, 0, gb, &rvlc->coefficient, q_ac2);
312  }
313  if(pattern & 2){ // Looks like coefficients 1 and 2 are swapped for this block
314  code = get_vlc2(gb, rvlc->second_pattern[sc].table, 9, 2);
315  decode_subblock(dst + 4*2+0, code, 1, gb, &rvlc->coefficient, q_ac2);
316  }
317  if(pattern & 1){
318  code = get_vlc2(gb, rvlc->third_pattern[sc].table, 9, 2);
319  decode_subblock(dst + 4*2+2, code, 0, gb, &rvlc->coefficient, q_ac2);
320  }
321  return has_ac || pattern;
322 }
323 
334 {
335  int i;
336  for(i = 0; i < 5; i++)
337  if(rv34_mb_max_sizes[i] >= mb_size - 1)
338  break;
339  return rv34_mb_bits_sizes[i];
340 }
341 
345 static inline RV34VLC* choose_vlc_set(int quant, int mod, int type)
346 {
347  if(mod == 2 && quant < 19) quant += 10;
348  else if(mod && quant < 26) quant += 5;
349  return type ? &inter_vlcs[rv34_quant_to_vlc_set[1][av_clip(quant, 0, 30)]]
350  : &intra_vlcs[rv34_quant_to_vlc_set[0][av_clip(quant, 0, 30)]];
351 }
352 
356 static int rv34_decode_intra_mb_header(RV34DecContext *r, int8_t *intra_types)
357 {
358  MpegEncContext *s = &r->s;
359  GetBitContext *gb = &s->gb;
360  int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
361  int t;
362 
363  r->is16 = get_bits1(gb);
364  if(r->is16){
367  t = get_bits(gb, 2);
368  fill_rectangle(intra_types, 4, 4, r->intra_types_stride, t, sizeof(intra_types[0]));
369  r->luma_vlc = 2;
370  }else{
371  if(!r->rv30){
372  if(!get_bits1(gb))
373  av_log(s->avctx, AV_LOG_ERROR, "Need DQUANT\n");
374  }
377  if(r->decode_intra_types(r, gb, intra_types) < 0)
378  return -1;
379  r->luma_vlc = 1;
380  }
381 
382  r->chroma_vlc = 0;
383  r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
384 
385  return rv34_decode_cbp(gb, r->cur_vlcs, r->is16);
386 }
387 
391 static int rv34_decode_inter_mb_header(RV34DecContext *r, int8_t *intra_types)
392 {
393  MpegEncContext *s = &r->s;
394  GetBitContext *gb = &s->gb;
395  int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
396  int i, t;
397 
398  r->block_type = r->decode_mb_info(r);
399  if(r->block_type == -1)
400  return -1;
402  r->mb_type[mb_pos] = r->block_type;
403  if(r->block_type == RV34_MB_SKIP){
404  if(s->pict_type == AV_PICTURE_TYPE_P)
405  r->mb_type[mb_pos] = RV34_MB_P_16x16;
406  if(s->pict_type == AV_PICTURE_TYPE_B)
407  r->mb_type[mb_pos] = RV34_MB_B_DIRECT;
408  }
409  r->is16 = !!IS_INTRA16x16(s->current_picture_ptr->f.mb_type[mb_pos]);
410  rv34_decode_mv(r, r->block_type);
411  if(r->block_type == RV34_MB_SKIP){
412  fill_rectangle(intra_types, 4, 4, r->intra_types_stride, 0, sizeof(intra_types[0]));
413  return 0;
414  }
415  r->chroma_vlc = 1;
416  r->luma_vlc = 0;
417 
418  if(IS_INTRA(s->current_picture_ptr->f.mb_type[mb_pos])){
419  if(r->is16){
420  t = get_bits(gb, 2);
421  fill_rectangle(intra_types, 4, 4, r->intra_types_stride, t, sizeof(intra_types[0]));
422  r->luma_vlc = 2;
423  }else{
424  if(r->decode_intra_types(r, gb, intra_types) < 0)
425  return -1;
426  r->luma_vlc = 1;
427  }
428  r->chroma_vlc = 0;
429  r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
430  }else{
431  for(i = 0; i < 16; i++)
432  intra_types[(i & 3) + (i>>2) * r->intra_types_stride] = 0;
433  r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 1);
434  if(r->mb_type[mb_pos] == RV34_MB_P_MIX16x16){
435  r->is16 = 1;
436  r->chroma_vlc = 1;
437  r->luma_vlc = 2;
438  r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
439  }
440  }
441 
442  return rv34_decode_cbp(gb, r->cur_vlcs, r->is16);
443 }
444  //bitstream functions
446 
453 static const uint8_t part_sizes_w[RV34_MB_TYPES] = { 2, 2, 2, 1, 2, 2, 2, 2, 2, 1, 2, 2 };
454 
456 static const uint8_t part_sizes_h[RV34_MB_TYPES] = { 2, 2, 2, 1, 2, 2, 2, 2, 1, 2, 2, 2 };
457 
459 static const uint8_t avail_indexes[4] = { 6, 7, 10, 11 };
460 
468 static void rv34_pred_mv(RV34DecContext *r, int block_type, int subblock_no, int dmv_no)
469 {
470  MpegEncContext *s = &r->s;
471  int mv_pos = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride;
472  int A[2] = {0}, B[2], C[2];
473  int i, j;
474  int mx, my;
475  int avail_index = avail_indexes[subblock_no];
476  int c_off = part_sizes_w[block_type];
477 
478  mv_pos += (subblock_no & 1) + (subblock_no >> 1)*s->b8_stride;
479  if(subblock_no == 3)
480  c_off = -1;
481 
482  if(r->avail_cache[avail_index - 1]){
483  A[0] = s->current_picture_ptr->f.motion_val[0][mv_pos-1][0];
484  A[1] = s->current_picture_ptr->f.motion_val[0][mv_pos-1][1];
485  }
486  if(r->avail_cache[avail_index - 4]){
487  B[0] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride][0];
488  B[1] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride][1];
489  }else{
490  B[0] = A[0];
491  B[1] = A[1];
492  }
493  if(!r->avail_cache[avail_index - 4 + c_off]){
494  if(r->avail_cache[avail_index - 4] && (r->avail_cache[avail_index - 1] || r->rv30)){
495  C[0] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride-1][0];
496  C[1] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride-1][1];
497  }else{
498  C[0] = A[0];
499  C[1] = A[1];
500  }
501  }else{
502  C[0] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride+c_off][0];
503  C[1] = s->current_picture_ptr->f.motion_val[0][mv_pos-s->b8_stride+c_off][1];
504  }
505  mx = mid_pred(A[0], B[0], C[0]);
506  my = mid_pred(A[1], B[1], C[1]);
507  mx += r->dmv[dmv_no][0];
508  my += r->dmv[dmv_no][1];
509  for(j = 0; j < part_sizes_h[block_type]; j++){
510  for(i = 0; i < part_sizes_w[block_type]; i++){
511  s->current_picture_ptr->f.motion_val[0][mv_pos + i + j*s->b8_stride][0] = mx;
512  s->current_picture_ptr->f.motion_val[0][mv_pos + i + j*s->b8_stride][1] = my;
513  }
514  }
515 }
516 
517 #define GET_PTS_DIFF(a, b) ((a - b + 8192) & 0x1FFF)
518 
522 static int calc_add_mv(RV34DecContext *r, int dir, int val)
523 {
524  int mul = dir ? -r->weight2 : r->weight1;
525 
526  return (val * mul + 0x2000) >> 14;
527 }
528 
532 static inline void rv34_pred_b_vector(int A[2], int B[2], int C[2],
533  int A_avail, int B_avail, int C_avail,
534  int *mx, int *my)
535 {
536  if(A_avail + B_avail + C_avail != 3){
537  *mx = A[0] + B[0] + C[0];
538  *my = A[1] + B[1] + C[1];
539  if(A_avail + B_avail + C_avail == 2){
540  *mx /= 2;
541  *my /= 2;
542  }
543  }else{
544  *mx = mid_pred(A[0], B[0], C[0]);
545  *my = mid_pred(A[1], B[1], C[1]);
546  }
547 }
548 
552 static void rv34_pred_mv_b(RV34DecContext *r, int block_type, int dir)
553 {
554  MpegEncContext *s = &r->s;
555  int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
556  int mv_pos = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride;
557  int A[2], B[2], C[2];
558  int has_A = 0, has_B = 0, has_C = 0;
559  int mx, my;
560  int i, j;
561  Picture *cur_pic = s->current_picture_ptr;
562  const int mask = dir ? MB_TYPE_L1 : MB_TYPE_L0;
563  int type = cur_pic->f.mb_type[mb_pos];
564 
565  memset(A, 0, sizeof(A));
566  memset(B, 0, sizeof(B));
567  memset(C, 0, sizeof(C));
568  if((r->avail_cache[6-1] & type) & mask){
569  A[0] = cur_pic->f.motion_val[dir][mv_pos - 1][0];
570  A[1] = cur_pic->f.motion_val[dir][mv_pos - 1][1];
571  has_A = 1;
572  }
573  if((r->avail_cache[6-4] & type) & mask){
574  B[0] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride][0];
575  B[1] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride][1];
576  has_B = 1;
577  }
578  if(r->avail_cache[6-4] && (r->avail_cache[6-2] & type) & mask){
579  C[0] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride + 2][0];
580  C[1] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride + 2][1];
581  has_C = 1;
582  }else if((s->mb_x+1) == s->mb_width && (r->avail_cache[6-5] & type) & mask){
583  C[0] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride - 1][0];
584  C[1] = cur_pic->f.motion_val[dir][mv_pos - s->b8_stride - 1][1];
585  has_C = 1;
586  }
587 
588  rv34_pred_b_vector(A, B, C, has_A, has_B, has_C, &mx, &my);
589 
590  mx += r->dmv[dir][0];
591  my += r->dmv[dir][1];
592 
593  for(j = 0; j < 2; j++){
594  for(i = 0; i < 2; i++){
595  cur_pic->f.motion_val[dir][mv_pos + i + j*s->b8_stride][0] = mx;
596  cur_pic->f.motion_val[dir][mv_pos + i + j*s->b8_stride][1] = my;
597  }
598  }
599  if(block_type == RV34_MB_B_BACKWARD || block_type == RV34_MB_B_FORWARD){
600  ZERO8x2(cur_pic->f.motion_val[!dir][mv_pos], s->b8_stride);
601  }
602 }
603 
607 static void rv34_pred_mv_rv3(RV34DecContext *r, int block_type, int dir)
608 {
609  MpegEncContext *s = &r->s;
610  int mv_pos = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride;
611  int A[2] = {0}, B[2], C[2];
612  int i, j, k;
613  int mx, my;
614  int avail_index = avail_indexes[0];
615 
616  if(r->avail_cache[avail_index - 1]){
617  A[0] = s->current_picture_ptr->f.motion_val[0][mv_pos - 1][0];
618  A[1] = s->current_picture_ptr->f.motion_val[0][mv_pos - 1][1];
619  }
620  if(r->avail_cache[avail_index - 4]){
621  B[0] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride][0];
622  B[1] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride][1];
623  }else{
624  B[0] = A[0];
625  B[1] = A[1];
626  }
627  if(!r->avail_cache[avail_index - 4 + 2]){
628  if(r->avail_cache[avail_index - 4] && (r->avail_cache[avail_index - 1])){
629  C[0] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride - 1][0];
630  C[1] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride - 1][1];
631  }else{
632  C[0] = A[0];
633  C[1] = A[1];
634  }
635  }else{
636  C[0] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride + 2][0];
637  C[1] = s->current_picture_ptr->f.motion_val[0][mv_pos - s->b8_stride + 2][1];
638  }
639  mx = mid_pred(A[0], B[0], C[0]);
640  my = mid_pred(A[1], B[1], C[1]);
641  mx += r->dmv[0][0];
642  my += r->dmv[0][1];
643  for(j = 0; j < 2; j++){
644  for(i = 0; i < 2; i++){
645  for(k = 0; k < 2; k++){
646  s->current_picture_ptr->f.motion_val[k][mv_pos + i + j*s->b8_stride][0] = mx;
647  s->current_picture_ptr->f.motion_val[k][mv_pos + i + j*s->b8_stride][1] = my;
648  }
649  }
650  }
651 }
652 
653 static const int chroma_coeffs[3] = { 0, 3, 5 };
654 
670 static inline void rv34_mc(RV34DecContext *r, const int block_type,
671  const int xoff, const int yoff, int mv_off,
672  const int width, const int height, int dir,
673  const int thirdpel, int weighted,
674  qpel_mc_func (*qpel_mc)[16],
675  h264_chroma_mc_func (*chroma_mc))
676 {
677  MpegEncContext *s = &r->s;
678  uint8_t *Y, *U, *V, *srcY, *srcU, *srcV;
679  int dxy, mx, my, umx, umy, lx, ly, uvmx, uvmy, src_x, src_y, uvsrc_x, uvsrc_y;
680  int mv_pos = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride + mv_off;
681  int is16x16 = 1;
682 
683  if(thirdpel){
684  int chroma_mx, chroma_my;
685  mx = (s->current_picture_ptr->f.motion_val[dir][mv_pos][0] + (3 << 24)) / 3 - (1 << 24);
686  my = (s->current_picture_ptr->f.motion_val[dir][mv_pos][1] + (3 << 24)) / 3 - (1 << 24);
687  lx = (s->current_picture_ptr->f.motion_val[dir][mv_pos][0] + (3 << 24)) % 3;
688  ly = (s->current_picture_ptr->f.motion_val[dir][mv_pos][1] + (3 << 24)) % 3;
689  chroma_mx = s->current_picture_ptr->f.motion_val[dir][mv_pos][0] / 2;
690  chroma_my = s->current_picture_ptr->f.motion_val[dir][mv_pos][1] / 2;
691  umx = (chroma_mx + (3 << 24)) / 3 - (1 << 24);
692  umy = (chroma_my + (3 << 24)) / 3 - (1 << 24);
693  uvmx = chroma_coeffs[(chroma_mx + (3 << 24)) % 3];
694  uvmy = chroma_coeffs[(chroma_my + (3 << 24)) % 3];
695  }else{
696  int cx, cy;
697  mx = s->current_picture_ptr->f.motion_val[dir][mv_pos][0] >> 2;
698  my = s->current_picture_ptr->f.motion_val[dir][mv_pos][1] >> 2;
699  lx = s->current_picture_ptr->f.motion_val[dir][mv_pos][0] & 3;
700  ly = s->current_picture_ptr->f.motion_val[dir][mv_pos][1] & 3;
701  cx = s->current_picture_ptr->f.motion_val[dir][mv_pos][0] / 2;
702  cy = s->current_picture_ptr->f.motion_val[dir][mv_pos][1] / 2;
703  umx = cx >> 2;
704  umy = cy >> 2;
705  uvmx = (cx & 3) << 1;
706  uvmy = (cy & 3) << 1;
707  //due to some flaw RV40 uses the same MC compensation routine for H2V2 and H3V3
708  if(uvmx == 6 && uvmy == 6)
709  uvmx = uvmy = 4;
710  }
711 
713  /* wait for the referenced mb row to be finished */
714  int mb_row = s->mb_y + ((yoff + my + 5 + 8 * height) >> 4);
715  AVFrame *f = dir ? &s->next_picture_ptr->f : &s->last_picture_ptr->f;
716  ff_thread_await_progress(f, mb_row, 0);
717  }
718 
719  dxy = ly*4 + lx;
720  srcY = dir ? s->next_picture_ptr->f.data[0] : s->last_picture_ptr->f.data[0];
721  srcU = dir ? s->next_picture_ptr->f.data[1] : s->last_picture_ptr->f.data[1];
722  srcV = dir ? s->next_picture_ptr->f.data[2] : s->last_picture_ptr->f.data[2];
723  src_x = s->mb_x * 16 + xoff + mx;
724  src_y = s->mb_y * 16 + yoff + my;
725  uvsrc_x = s->mb_x * 8 + (xoff >> 1) + umx;
726  uvsrc_y = s->mb_y * 8 + (yoff >> 1) + umy;
727  srcY += src_y * s->linesize + src_x;
728  srcU += uvsrc_y * s->uvlinesize + uvsrc_x;
729  srcV += uvsrc_y * s->uvlinesize + uvsrc_x;
730  if(s->h_edge_pos - (width << 3) < 6 || s->v_edge_pos - (height << 3) < 6 ||
731  (unsigned)(src_x - !!lx*2) > s->h_edge_pos - !!lx*2 - (width <<3) - 4 ||
732  (unsigned)(src_y - !!ly*2) > s->v_edge_pos - !!ly*2 - (height<<3) - 4) {
733  uint8_t *uvbuf = s->edge_emu_buffer + 22 * s->linesize;
734 
735  srcY -= 2 + 2*s->linesize;
736  s->dsp.emulated_edge_mc(s->edge_emu_buffer, srcY, s->linesize, (width<<3)+6, (height<<3)+6,
737  src_x - 2, src_y - 2, s->h_edge_pos, s->v_edge_pos);
738  srcY = s->edge_emu_buffer + 2 + 2*s->linesize;
739  s->dsp.emulated_edge_mc(uvbuf , srcU, s->uvlinesize, (width<<2)+1, (height<<2)+1,
740  uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, s->v_edge_pos >> 1);
741  s->dsp.emulated_edge_mc(uvbuf + 16, srcV, s->uvlinesize, (width<<2)+1, (height<<2)+1,
742  uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, s->v_edge_pos >> 1);
743  srcU = uvbuf;
744  srcV = uvbuf + 16;
745  }
746  if(!weighted){
747  Y = s->dest[0] + xoff + yoff *s->linesize;
748  U = s->dest[1] + (xoff>>1) + (yoff>>1)*s->uvlinesize;
749  V = s->dest[2] + (xoff>>1) + (yoff>>1)*s->uvlinesize;
750  }else{
751  Y = r->tmp_b_block_y [dir] + xoff + yoff *s->linesize;
752  U = r->tmp_b_block_uv[dir*2] + (xoff>>1) + (yoff>>1)*s->uvlinesize;
753  V = r->tmp_b_block_uv[dir*2+1] + (xoff>>1) + (yoff>>1)*s->uvlinesize;
754  }
755 
756  if(block_type == RV34_MB_P_16x8){
757  qpel_mc[1][dxy](Y, srcY, s->linesize);
758  Y += 8;
759  srcY += 8;
760  }else if(block_type == RV34_MB_P_8x16){
761  qpel_mc[1][dxy](Y, srcY, s->linesize);
762  Y += 8 * s->linesize;
763  srcY += 8 * s->linesize;
764  }
765  is16x16 = (block_type != RV34_MB_P_8x8) && (block_type != RV34_MB_P_16x8) && (block_type != RV34_MB_P_8x16);
766  qpel_mc[!is16x16][dxy](Y, srcY, s->linesize);
767  chroma_mc[2-width] (U, srcU, s->uvlinesize, height*4, uvmx, uvmy);
768  chroma_mc[2-width] (V, srcV, s->uvlinesize, height*4, uvmx, uvmy);
769 }
770 
771 static void rv34_mc_1mv(RV34DecContext *r, const int block_type,
772  const int xoff, const int yoff, int mv_off,
773  const int width, const int height, int dir)
774 {
775  rv34_mc(r, block_type, xoff, yoff, mv_off, width, height, dir, r->rv30, 0,
776  r->rdsp.put_pixels_tab,
778 }
779 
780 static void rv4_weight(RV34DecContext *r)
781 {
782  r->rdsp.rv40_weight_pixels_tab[0](r->s.dest[0],
783  r->tmp_b_block_y[0],
784  r->tmp_b_block_y[1],
785  r->weight1,
786  r->weight2,
787  r->s.linesize);
788  r->rdsp.rv40_weight_pixels_tab[1](r->s.dest[1],
789  r->tmp_b_block_uv[0],
790  r->tmp_b_block_uv[2],
791  r->weight1,
792  r->weight2,
793  r->s.uvlinesize);
794  r->rdsp.rv40_weight_pixels_tab[1](r->s.dest[2],
795  r->tmp_b_block_uv[1],
796  r->tmp_b_block_uv[3],
797  r->weight1,
798  r->weight2,
799  r->s.uvlinesize);
800 }
801 
802 static void rv34_mc_2mv(RV34DecContext *r, const int block_type)
803 {
804  int weighted = !r->rv30 && block_type != RV34_MB_B_BIDIR && r->weight1 != 8192;
805 
806  rv34_mc(r, block_type, 0, 0, 0, 2, 2, 0, r->rv30, weighted,
807  r->rdsp.put_pixels_tab,
809  if(!weighted){
810  rv34_mc(r, block_type, 0, 0, 0, 2, 2, 1, r->rv30, 0,
811  r->rdsp.avg_pixels_tab,
813  }else{
814  rv34_mc(r, block_type, 0, 0, 0, 2, 2, 1, r->rv30, 1,
815  r->rdsp.put_pixels_tab,
817  rv4_weight(r);
818  }
819 }
820 
822 {
823  int i, j;
824  int weighted = !r->rv30 && r->weight1 != 8192;
825 
826  for(j = 0; j < 2; j++)
827  for(i = 0; i < 2; i++){
828  rv34_mc(r, RV34_MB_P_8x8, i*8, j*8, i+j*r->s.b8_stride, 1, 1, 0, r->rv30,
829  weighted,
830  r->rdsp.put_pixels_tab,
832  rv34_mc(r, RV34_MB_P_8x8, i*8, j*8, i+j*r->s.b8_stride, 1, 1, 1, r->rv30,
833  weighted,
834  weighted ? r->rdsp.put_pixels_tab : r->rdsp.avg_pixels_tab,
836  }
837  if(weighted)
838  rv4_weight(r);
839 }
840 
842 static const int num_mvs[RV34_MB_TYPES] = { 0, 0, 1, 4, 1, 1, 0, 0, 2, 2, 2, 1 };
843 
848 static int rv34_decode_mv(RV34DecContext *r, int block_type)
849 {
850  MpegEncContext *s = &r->s;
851  GetBitContext *gb = &s->gb;
852  int i, j, k, l;
853  int mv_pos = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride;
854  int next_bt;
855 
856  memset(r->dmv, 0, sizeof(r->dmv));
857  for(i = 0; i < num_mvs[block_type]; i++){
858  r->dmv[i][0] = svq3_get_se_golomb(gb);
859  r->dmv[i][1] = svq3_get_se_golomb(gb);
860  }
861  switch(block_type){
862  case RV34_MB_TYPE_INTRA:
864  ZERO8x2(s->current_picture_ptr->f.motion_val[0][s->mb_x * 2 + s->mb_y * 2 * s->b8_stride], s->b8_stride);
865  return 0;
866  case RV34_MB_SKIP:
867  if(s->pict_type == AV_PICTURE_TYPE_P){
868  ZERO8x2(s->current_picture_ptr->f.motion_val[0][s->mb_x * 2 + s->mb_y * 2 * s->b8_stride], s->b8_stride);
869  rv34_mc_1mv (r, block_type, 0, 0, 0, 2, 2, 0);
870  break;
871  }
872  case RV34_MB_B_DIRECT:
873  //surprisingly, it uses motion scheme from next reference frame
874  /* wait for the current mb row to be finished */
877 
878  next_bt = s->next_picture_ptr->f.mb_type[s->mb_x + s->mb_y * s->mb_stride];
879  if(IS_INTRA(next_bt) || IS_SKIP(next_bt)){
880  ZERO8x2(s->current_picture_ptr->f.motion_val[0][s->mb_x * 2 + s->mb_y * 2 * s->b8_stride], s->b8_stride);
881  ZERO8x2(s->current_picture_ptr->f.motion_val[1][s->mb_x * 2 + s->mb_y * 2 * s->b8_stride], s->b8_stride);
882  }else
883  for(j = 0; j < 2; j++)
884  for(i = 0; i < 2; i++)
885  for(k = 0; k < 2; k++)
886  for(l = 0; l < 2; l++)
887  s->current_picture_ptr->f.motion_val[l][mv_pos + i + j*s->b8_stride][k] = calc_add_mv(r, l, s->next_picture_ptr->f.motion_val[0][mv_pos + i + j*s->b8_stride][k]);
888  if(!(IS_16X8(next_bt) || IS_8X16(next_bt) || IS_8X8(next_bt))) //we can use whole macroblock MC
889  rv34_mc_2mv(r, block_type);
890  else
891  rv34_mc_2mv_skip(r);
892  ZERO8x2(s->current_picture_ptr->f.motion_val[0][s->mb_x * 2 + s->mb_y * 2 * s->b8_stride], s->b8_stride);
893  break;
894  case RV34_MB_P_16x16:
895  case RV34_MB_P_MIX16x16:
896  rv34_pred_mv(r, block_type, 0, 0);
897  rv34_mc_1mv (r, block_type, 0, 0, 0, 2, 2, 0);
898  break;
899  case RV34_MB_B_FORWARD:
900  case RV34_MB_B_BACKWARD:
901  r->dmv[1][0] = r->dmv[0][0];
902  r->dmv[1][1] = r->dmv[0][1];
903  if(r->rv30)
904  rv34_pred_mv_rv3(r, block_type, block_type == RV34_MB_B_BACKWARD);
905  else
906  rv34_pred_mv_b (r, block_type, block_type == RV34_MB_B_BACKWARD);
907  rv34_mc_1mv (r, block_type, 0, 0, 0, 2, 2, block_type == RV34_MB_B_BACKWARD);
908  break;
909  case RV34_MB_P_16x8:
910  case RV34_MB_P_8x16:
911  rv34_pred_mv(r, block_type, 0, 0);
912  rv34_pred_mv(r, block_type, 1 + (block_type == RV34_MB_P_16x8), 1);
913  if(block_type == RV34_MB_P_16x8){
914  rv34_mc_1mv(r, block_type, 0, 0, 0, 2, 1, 0);
915  rv34_mc_1mv(r, block_type, 0, 8, s->b8_stride, 2, 1, 0);
916  }
917  if(block_type == RV34_MB_P_8x16){
918  rv34_mc_1mv(r, block_type, 0, 0, 0, 1, 2, 0);
919  rv34_mc_1mv(r, block_type, 8, 0, 1, 1, 2, 0);
920  }
921  break;
922  case RV34_MB_B_BIDIR:
923  rv34_pred_mv_b (r, block_type, 0);
924  rv34_pred_mv_b (r, block_type, 1);
925  rv34_mc_2mv (r, block_type);
926  break;
927  case RV34_MB_P_8x8:
928  for(i=0;i< 4;i++){
929  rv34_pred_mv(r, block_type, i, i);
930  rv34_mc_1mv (r, block_type, (i&1)<<3, (i&2)<<2, (i&1)+(i>>1)*s->b8_stride, 1, 1, 0);
931  }
932  break;
933  }
934 
935  return 0;
936 } // mv group
938 
944 static const int ittrans[9] = {
947 };
948 
950 static const int ittrans16[4] = {
952 };
953 
957 static void rv34_pred_4x4_block(RV34DecContext *r, uint8_t *dst, int stride, int itype, int up, int left, int down, int right)
958 {
959  uint8_t *prev = dst - stride + 4;
960  uint32_t topleft;
961 
962  if(!up && !left)
963  itype = DC_128_PRED;
964  else if(!up){
965  if(itype == VERT_PRED) itype = HOR_PRED;
966  if(itype == DC_PRED) itype = LEFT_DC_PRED;
967  }else if(!left){
968  if(itype == HOR_PRED) itype = VERT_PRED;
969  if(itype == DC_PRED) itype = TOP_DC_PRED;
971  }
972  if(!down){
974  if(itype == HOR_UP_PRED) itype = HOR_UP_PRED_RV40_NODOWN;
975  if(itype == VERT_LEFT_PRED) itype = VERT_LEFT_PRED_RV40_NODOWN;
976  }
977  if(!right && up){
978  topleft = dst[-stride + 3] * 0x01010101u;
979  prev = (uint8_t*)&topleft;
980  }
981  r->h.pred4x4[itype](dst, prev, stride);
982 }
983 
984 static inline int adjust_pred16(int itype, int up, int left)
985 {
986  if(!up && !left)
987  itype = DC_128_PRED8x8;
988  else if(!up){
989  if(itype == PLANE_PRED8x8)itype = HOR_PRED8x8;
990  if(itype == VERT_PRED8x8) itype = HOR_PRED8x8;
991  if(itype == DC_PRED8x8) itype = LEFT_DC_PRED8x8;
992  }else if(!left){
993  if(itype == PLANE_PRED8x8)itype = VERT_PRED8x8;
994  if(itype == HOR_PRED8x8) itype = VERT_PRED8x8;
995  if(itype == DC_PRED8x8) itype = TOP_DC_PRED8x8;
996  }
997  return itype;
998 }
999 
1000 static inline void rv34_process_block(RV34DecContext *r,
1001  uint8_t *pdst, int stride,
1002  int fc, int sc, int q_dc, int q_ac)
1003 {
1004  MpegEncContext *s = &r->s;
1005  DCTELEM *ptr = s->block[0];
1006  int has_ac = rv34_decode_block(ptr, &s->gb, r->cur_vlcs,
1007  fc, sc, q_dc, q_ac, q_ac);
1008  if(has_ac){
1009  r->rdsp.rv34_idct_add(pdst, stride, ptr);
1010  }else{
1011  r->rdsp.rv34_idct_dc_add(pdst, stride, ptr[0]);
1012  ptr[0] = 0;
1013  }
1014 }
1015 
1016 static void rv34_output_i16x16(RV34DecContext *r, int8_t *intra_types, int cbp)
1017 {
1018  LOCAL_ALIGNED_16(DCTELEM, block16, [16]);
1019  MpegEncContext *s = &r->s;
1020  GetBitContext *gb = &s->gb;
1021  int q_dc = rv34_qscale_tab[ r->luma_dc_quant_i[s->qscale] ],
1022  q_ac = rv34_qscale_tab[s->qscale];
1023  uint8_t *dst = s->dest[0];
1024  DCTELEM *ptr = s->block[0];
1025  int avail[6*8] = {0};
1026  int i, j, itype, has_ac;
1027 
1028  memset(block16, 0, 16 * sizeof(*block16));
1029 
1030  // Set neighbour information.
1031  if(r->avail_cache[1])
1032  avail[0] = 1;
1033  if(r->avail_cache[2])
1034  avail[1] = avail[2] = 1;
1035  if(r->avail_cache[3])
1036  avail[3] = avail[4] = 1;
1037  if(r->avail_cache[4])
1038  avail[5] = 1;
1039  if(r->avail_cache[5])
1040  avail[8] = avail[16] = 1;
1041  if(r->avail_cache[9])
1042  avail[24] = avail[32] = 1;
1043 
1044  has_ac = rv34_decode_block(block16, gb, r->cur_vlcs, 3, 0, q_dc, q_dc, q_ac);
1045  if(has_ac)
1046  r->rdsp.rv34_inv_transform(block16);
1047  else
1048  r->rdsp.rv34_inv_transform_dc(block16);
1049 
1050  itype = ittrans16[intra_types[0]];
1051  itype = adjust_pred16(itype, r->avail_cache[6-4], r->avail_cache[6-1]);
1052  r->h.pred16x16[itype](dst, s->linesize);
1053 
1054  for(j = 0; j < 4; j++){
1055  for(i = 0; i < 4; i++, cbp >>= 1){
1056  int dc = block16[i + j*4];
1057 
1058  if(cbp & 1){
1059  has_ac = rv34_decode_block(ptr, gb, r->cur_vlcs, r->luma_vlc, 0, q_ac, q_ac, q_ac);
1060  }else
1061  has_ac = 0;
1062 
1063  if(has_ac){
1064  ptr[0] = dc;
1065  r->rdsp.rv34_idct_add(dst+4*i, s->linesize, ptr);
1066  }else
1067  r->rdsp.rv34_idct_dc_add(dst+4*i, s->linesize, dc);
1068  }
1069 
1070  dst += 4*s->linesize;
1071  }
1072 
1073  itype = ittrans16[intra_types[0]];
1074  if(itype == PLANE_PRED8x8) itype = DC_PRED8x8;
1075  itype = adjust_pred16(itype, r->avail_cache[6-4], r->avail_cache[6-1]);
1076 
1077  q_dc = rv34_qscale_tab[rv34_chroma_quant[1][s->qscale]];
1078  q_ac = rv34_qscale_tab[rv34_chroma_quant[0][s->qscale]];
1079 
1080  for(j = 1; j < 3; j++){
1081  dst = s->dest[j];
1082  r->h.pred8x8[itype](dst, s->uvlinesize);
1083  for(i = 0; i < 4; i++, cbp >>= 1){
1084  uint8_t *pdst;
1085  if(!(cbp & 1)) continue;
1086  pdst = dst + (i&1)*4 + (i&2)*2*s->uvlinesize;
1087 
1088  rv34_process_block(r, pdst, s->uvlinesize,
1089  r->chroma_vlc, 1, q_dc, q_ac);
1090  }
1091  }
1092 }
1093 
1094 static void rv34_output_intra(RV34DecContext *r, int8_t *intra_types, int cbp)
1095 {
1096  MpegEncContext *s = &r->s;
1097  uint8_t *dst = s->dest[0];
1098  int avail[6*8] = {0};
1099  int i, j, k;
1100  int idx, q_ac, q_dc;
1101 
1102  // Set neighbour information.
1103  if(r->avail_cache[1])
1104  avail[0] = 1;
1105  if(r->avail_cache[2])
1106  avail[1] = avail[2] = 1;
1107  if(r->avail_cache[3])
1108  avail[3] = avail[4] = 1;
1109  if(r->avail_cache[4])
1110  avail[5] = 1;
1111  if(r->avail_cache[5])
1112  avail[8] = avail[16] = 1;
1113  if(r->avail_cache[9])
1114  avail[24] = avail[32] = 1;
1115 
1116  q_ac = rv34_qscale_tab[s->qscale];
1117  for(j = 0; j < 4; j++){
1118  idx = 9 + j*8;
1119  for(i = 0; i < 4; i++, cbp >>= 1, dst += 4, idx++){
1120  rv34_pred_4x4_block(r, dst, s->linesize, ittrans[intra_types[i]], avail[idx-8], avail[idx-1], avail[idx+7], avail[idx-7]);
1121  avail[idx] = 1;
1122  if(!(cbp & 1)) continue;
1123 
1124  rv34_process_block(r, dst, s->linesize,
1125  r->luma_vlc, 0, q_ac, q_ac);
1126  }
1127  dst += s->linesize * 4 - 4*4;
1128  intra_types += r->intra_types_stride;
1129  }
1130 
1131  intra_types -= r->intra_types_stride * 4;
1132 
1133  q_dc = rv34_qscale_tab[rv34_chroma_quant[1][s->qscale]];
1134  q_ac = rv34_qscale_tab[rv34_chroma_quant[0][s->qscale]];
1135 
1136  for(k = 0; k < 2; k++){
1137  dst = s->dest[1+k];
1138  fill_rectangle(r->avail_cache + 6, 2, 2, 4, 0, 4);
1139 
1140  for(j = 0; j < 2; j++){
1141  int* acache = r->avail_cache + 6 + j*4;
1142  for(i = 0; i < 2; i++, cbp >>= 1, acache++){
1143  int itype = ittrans[intra_types[i*2+j*2*r->intra_types_stride]];
1144  rv34_pred_4x4_block(r, dst+4*i, s->uvlinesize, itype, acache[-4], acache[-1], !i && !j, acache[-3]);
1145  acache[0] = 1;
1146 
1147  if(!(cbp&1)) continue;
1148 
1149  rv34_process_block(r, dst + 4*i, s->uvlinesize,
1150  r->chroma_vlc, 1, q_dc, q_ac);
1151  }
1152 
1153  dst += 4*s->uvlinesize;
1154  }
1155  }
1156 }
1157 
1158 static int is_mv_diff_gt_3(int16_t (*motion_val)[2], int step)
1159 {
1160  int d;
1161  d = motion_val[0][0] - motion_val[-step][0];
1162  if(d < -3 || d > 3)
1163  return 1;
1164  d = motion_val[0][1] - motion_val[-step][1];
1165  if(d < -3 || d > 3)
1166  return 1;
1167  return 0;
1168 }
1169 
1171 {
1172  MpegEncContext *s = &r->s;
1173  int hmvmask = 0, vmvmask = 0, i, j;
1174  int midx = s->mb_x * 2 + s->mb_y * 2 * s->b8_stride;
1175  int16_t (*motion_val)[2] = &s->current_picture_ptr->f.motion_val[0][midx];
1176  for(j = 0; j < 16; j += 8){
1177  for(i = 0; i < 2; i++){
1178  if(is_mv_diff_gt_3(motion_val + i, 1))
1179  vmvmask |= 0x11 << (j + i*2);
1180  if((j || s->mb_y) && is_mv_diff_gt_3(motion_val + i, s->b8_stride))
1181  hmvmask |= 0x03 << (j + i*2);
1182  }
1183  motion_val += s->b8_stride;
1184  }
1185  if(s->first_slice_line)
1186  hmvmask &= ~0x000F;
1187  if(!s->mb_x)
1188  vmvmask &= ~0x1111;
1189  if(r->rv30){ //RV30 marks both subblocks on the edge for filtering
1190  vmvmask |= (vmvmask & 0x4444) >> 1;
1191  hmvmask |= (hmvmask & 0x0F00) >> 4;
1192  if(s->mb_x)
1193  r->deblock_coefs[s->mb_x - 1 + s->mb_y*s->mb_stride] |= (vmvmask & 0x1111) << 3;
1194  if(!s->first_slice_line)
1195  r->deblock_coefs[s->mb_x + (s->mb_y - 1)*s->mb_stride] |= (hmvmask & 0xF) << 12;
1196  }
1197  return hmvmask | vmvmask;
1198 }
1199 
1200 static int rv34_decode_inter_macroblock(RV34DecContext *r, int8_t *intra_types)
1201 {
1202  MpegEncContext *s = &r->s;
1203  GetBitContext *gb = &s->gb;
1204  uint8_t *dst = s->dest[0];
1205  DCTELEM *ptr = s->block[0];
1206  int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
1207  int cbp, cbp2;
1208  int q_dc, q_ac, has_ac;
1209  int i, j;
1210  int dist;
1211 
1212  // Calculate which neighbours are available. Maybe it's worth optimizing too.
1213  memset(r->avail_cache, 0, sizeof(r->avail_cache));
1214  fill_rectangle(r->avail_cache + 6, 2, 2, 4, 1, 4);
1215  dist = (s->mb_x - s->resync_mb_x) + (s->mb_y - s->resync_mb_y) * s->mb_width;
1216  if(s->mb_x && dist)
1217  r->avail_cache[5] =
1218  r->avail_cache[9] = s->current_picture_ptr->f.mb_type[mb_pos - 1];
1219  if(dist >= s->mb_width)
1220  r->avail_cache[2] =
1221  r->avail_cache[3] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride];
1222  if(((s->mb_x+1) < s->mb_width) && dist >= s->mb_width - 1)
1223  r->avail_cache[4] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride + 1];
1224  if(s->mb_x && dist > s->mb_width)
1225  r->avail_cache[1] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride - 1];
1226 
1227  s->qscale = r->si.quant;
1228  cbp = cbp2 = rv34_decode_inter_mb_header(r, intra_types);
1229  r->cbp_luma [mb_pos] = cbp;
1230  r->cbp_chroma[mb_pos] = cbp >> 16;
1231  r->deblock_coefs[mb_pos] = rv34_set_deblock_coef(r) | r->cbp_luma[mb_pos];
1232  s->current_picture_ptr->f.qscale_table[mb_pos] = s->qscale;
1233 
1234  if(cbp == -1)
1235  return -1;
1236 
1237  if (IS_INTRA(s->current_picture_ptr->f.mb_type[mb_pos])){
1238  if(r->is16) rv34_output_i16x16(r, intra_types, cbp);
1239  else rv34_output_intra(r, intra_types, cbp);
1240  return 0;
1241  }
1242 
1243  if(r->is16){
1244  // Only for RV34_MB_P_MIX16x16
1245  LOCAL_ALIGNED_16(DCTELEM, block16, [16]);
1246  memset(block16, 0, 16 * sizeof(*block16));
1247  q_dc = rv34_qscale_tab[ r->luma_dc_quant_p[s->qscale] ];
1248  q_ac = rv34_qscale_tab[s->qscale];
1249  if (rv34_decode_block(block16, gb, r->cur_vlcs, 3, 0, q_dc, q_dc, q_ac))
1250  r->rdsp.rv34_inv_transform(block16);
1251  else
1252  r->rdsp.rv34_inv_transform_dc(block16);
1253 
1254  q_ac = rv34_qscale_tab[s->qscale];
1255 
1256  for(j = 0; j < 4; j++){
1257  for(i = 0; i < 4; i++, cbp >>= 1){
1258  int dc = block16[i + j*4];
1259 
1260  if(cbp & 1){
1261  has_ac = rv34_decode_block(ptr, gb, r->cur_vlcs, r->luma_vlc, 0, q_ac, q_ac, q_ac);
1262  }else
1263  has_ac = 0;
1264 
1265  if(has_ac){
1266  ptr[0] = dc;
1267  r->rdsp.rv34_idct_add(dst+4*i, s->linesize, ptr);
1268  }else
1269  r->rdsp.rv34_idct_dc_add(dst+4*i, s->linesize, dc);
1270  }
1271 
1272  dst += 4*s->linesize;
1273  }
1274 
1275  r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 1);
1276  }else{
1277  q_ac = rv34_qscale_tab[s->qscale];
1278 
1279  for(j = 0; j < 4; j++){
1280  for(i = 0; i < 4; i++, cbp >>= 1){
1281  if(!(cbp & 1)) continue;
1282 
1283  rv34_process_block(r, dst + 4*i, s->linesize,
1284  r->luma_vlc, 0, q_ac, q_ac);
1285  }
1286  dst += 4*s->linesize;
1287  }
1288  }
1289 
1290  q_dc = rv34_qscale_tab[rv34_chroma_quant[1][s->qscale]];
1291  q_ac = rv34_qscale_tab[rv34_chroma_quant[0][s->qscale]];
1292 
1293  for(j = 1; j < 3; j++){
1294  dst = s->dest[j];
1295  for(i = 0; i < 4; i++, cbp >>= 1){
1296  uint8_t *pdst;
1297  if(!(cbp & 1)) continue;
1298  pdst = dst + (i&1)*4 + (i&2)*2*s->uvlinesize;
1299 
1300  rv34_process_block(r, pdst, s->uvlinesize,
1301  r->chroma_vlc, 1, q_dc, q_ac);
1302  }
1303  }
1304 
1305  return 0;
1306 }
1307 
1308 static int rv34_decode_intra_macroblock(RV34DecContext *r, int8_t *intra_types)
1309 {
1310  MpegEncContext *s = &r->s;
1311  int cbp, dist;
1312  int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
1313 
1314  // Calculate which neighbours are available. Maybe it's worth optimizing too.
1315  memset(r->avail_cache, 0, sizeof(r->avail_cache));
1316  fill_rectangle(r->avail_cache + 6, 2, 2, 4, 1, 4);
1317  dist = (s->mb_x - s->resync_mb_x) + (s->mb_y - s->resync_mb_y) * s->mb_width;
1318  if(s->mb_x && dist)
1319  r->avail_cache[5] =
1320  r->avail_cache[9] = s->current_picture_ptr->f.mb_type[mb_pos - 1];
1321  if(dist >= s->mb_width)
1322  r->avail_cache[2] =
1323  r->avail_cache[3] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride];
1324  if(((s->mb_x+1) < s->mb_width) && dist >= s->mb_width - 1)
1325  r->avail_cache[4] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride + 1];
1326  if(s->mb_x && dist > s->mb_width)
1327  r->avail_cache[1] = s->current_picture_ptr->f.mb_type[mb_pos - s->mb_stride - 1];
1328 
1329  s->qscale = r->si.quant;
1330  cbp = rv34_decode_intra_mb_header(r, intra_types);
1331  r->cbp_luma [mb_pos] = cbp;
1332  r->cbp_chroma[mb_pos] = cbp >> 16;
1333  r->deblock_coefs[mb_pos] = 0xFFFF;
1334  s->current_picture_ptr->f.qscale_table[mb_pos] = s->qscale;
1335 
1336  if(cbp == -1)
1337  return -1;
1338 
1339  if(r->is16){
1340  rv34_output_i16x16(r, intra_types, cbp);
1341  return 0;
1342  }
1343 
1344  rv34_output_intra(r, intra_types, cbp);
1345  return 0;
1346 }
1347 
1349 {
1350  int bits;
1351  if(s->mb_y >= s->mb_height)
1352  return 1;
1353  if(!s->mb_num_left)
1354  return 1;
1355  if(r->s.mb_skip_run > 1)
1356  return 0;
1357  bits = get_bits_left(&s->gb);
1358  if(bits < 0 || (bits < 8 && !show_bits(&s->gb, bits)))
1359  return 1;
1360  return 0;
1361 }
1362 
1363 
1365 {
1367  r->intra_types = NULL;
1369  av_freep(&r->mb_type);
1370  av_freep(&r->cbp_luma);
1371  av_freep(&r->cbp_chroma);
1372  av_freep(&r->deblock_coefs);
1373 }
1374 
1375 
1377 {
1378  r->intra_types_stride = r->s.mb_width * 4 + 4;
1379 
1380  r->cbp_chroma = av_malloc(r->s.mb_stride * r->s.mb_height *
1381  sizeof(*r->cbp_chroma));
1382  r->cbp_luma = av_malloc(r->s.mb_stride * r->s.mb_height *
1383  sizeof(*r->cbp_luma));
1384  r->deblock_coefs = av_malloc(r->s.mb_stride * r->s.mb_height *
1385  sizeof(*r->deblock_coefs));
1387  sizeof(*r->intra_types_hist));
1388  r->mb_type = av_mallocz(r->s.mb_stride * r->s.mb_height *
1389  sizeof(*r->mb_type));
1390 
1391  if (!(r->cbp_chroma && r->cbp_luma && r->deblock_coefs &&
1392  r->intra_types_hist && r->mb_type)) {
1393  rv34_decoder_free(r);
1394  return AVERROR(ENOMEM);
1395  }
1396 
1398 
1399  return 0;
1400 }
1401 
1402 
1404 {
1405  rv34_decoder_free(r);
1406  return rv34_decoder_alloc(r);
1407 }
1408 
1409 
1410 static int rv34_decode_slice(RV34DecContext *r, int end, const uint8_t* buf, int buf_size)
1411 {
1412  MpegEncContext *s = &r->s;
1413  GetBitContext *gb = &s->gb;
1414  int mb_pos, slice_type;
1415  int res;
1416 
1417  init_get_bits(&r->s.gb, buf, buf_size*8);
1418  res = r->parse_slice_header(r, gb, &r->si);
1419  if(res < 0){
1420  av_log(s->avctx, AV_LOG_ERROR, "Incorrect or unknown slice header\n");
1421  return -1;
1422  }
1423 
1424  slice_type = r->si.type ? r->si.type : AV_PICTURE_TYPE_I;
1425  if (slice_type != s->pict_type) {
1426  av_log(s->avctx, AV_LOG_ERROR, "Slice type mismatch\n");
1427  return AVERROR_INVALIDDATA;
1428  }
1429 
1430  r->si.end = end;
1431  s->qscale = r->si.quant;
1432  s->mb_num_left = r->si.end - r->si.start;
1433  r->s.mb_skip_run = 0;
1434 
1435  mb_pos = s->mb_x + s->mb_y * s->mb_width;
1436  if(r->si.start != mb_pos){
1437  av_log(s->avctx, AV_LOG_ERROR, "Slice indicates MB offset %d, got %d\n", r->si.start, mb_pos);
1438  s->mb_x = r->si.start % s->mb_width;
1439  s->mb_y = r->si.start / s->mb_width;
1440  }
1441  memset(r->intra_types_hist, -1, r->intra_types_stride * 4 * 2 * sizeof(*r->intra_types_hist));
1442  s->first_slice_line = 1;
1443  s->resync_mb_x = s->mb_x;
1444  s->resync_mb_y = s->mb_y;
1445 
1447  while(!check_slice_end(r, s)) {
1449 
1450  if(r->si.type)
1451  res = rv34_decode_inter_macroblock(r, r->intra_types + s->mb_x * 4 + 4);
1452  else
1453  res = rv34_decode_intra_macroblock(r, r->intra_types + s->mb_x * 4 + 4);
1454  if(res < 0){
1456  return -1;
1457  }
1458  if (++s->mb_x == s->mb_width) {
1459  s->mb_x = 0;
1460  s->mb_y++;
1462 
1463  memmove(r->intra_types_hist, r->intra_types, r->intra_types_stride * 4 * sizeof(*r->intra_types_hist));
1464  memset(r->intra_types, -1, r->intra_types_stride * 4 * sizeof(*r->intra_types_hist));
1465 
1466  if(r->loop_filter && s->mb_y >= 2)
1467  r->loop_filter(r, s->mb_y - 2);
1468 
1471  s->mb_y - 2, 0);
1472 
1473  }
1474  if(s->mb_x == s->resync_mb_x)
1475  s->first_slice_line=0;
1476  s->mb_num_left--;
1477  }
1478  ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, ER_MB_END);
1479 
1480  return s->mb_y == s->mb_height;
1481 }
1482  // recons group end
1484 
1489 {
1490  RV34DecContext *r = avctx->priv_data;
1491  MpegEncContext *s = &r->s;
1492  int ret;
1493 
1495  s->avctx = avctx;
1496  s->out_format = FMT_H263;
1497  s->codec_id = avctx->codec_id;
1498 
1499  s->width = avctx->width;
1500  s->height = avctx->height;
1501 
1502  r->s.avctx = avctx;
1503  avctx->flags |= CODEC_FLAG_EMU_EDGE;
1504  r->s.flags |= CODEC_FLAG_EMU_EDGE;
1505  avctx->pix_fmt = PIX_FMT_YUV420P;
1506  avctx->has_b_frames = 1;
1507  s->low_delay = 0;
1508 
1509  if ((ret = MPV_common_init(s)) < 0)
1510  return ret;
1511 
1512  ff_h264_pred_init(&r->h, CODEC_ID_RV40, 8, 1);
1513 
1514 #if CONFIG_RV30_DECODER
1515  if (avctx->codec_id == CODEC_ID_RV30)
1516  ff_rv30dsp_init(&r->rdsp, &r->s.dsp);
1517 #endif
1518 #if CONFIG_RV40_DECODER
1519  if (avctx->codec_id == CODEC_ID_RV40)
1520  ff_rv40dsp_init(&r->rdsp, &r->s.dsp);
1521 #endif
1522 
1523  if ((ret = rv34_decoder_alloc(r)) < 0)
1524  return ret;
1525 
1526  if(!intra_vlcs[0].cbppattern[0].bits)
1527  rv34_init_tables();
1528 
1529  return 0;
1530 }
1531 
1533 {
1534  int err;
1535  RV34DecContext *r = avctx->priv_data;
1536 
1537  r->s.avctx = avctx;
1538 
1539  if (avctx->internal->is_copy) {
1540  r->tmp_b_block_base = NULL;
1541  if ((err = MPV_common_init(&r->s)) < 0)
1542  return err;
1543  if ((err = rv34_decoder_alloc(r)) < 0)
1544  return err;
1545  }
1546  return 0;
1547 }
1548 
1550 {
1551  RV34DecContext *r = dst->priv_data, *r1 = src->priv_data;
1552  MpegEncContext * const s = &r->s, * const s1 = &r1->s;
1553  int err;
1554 
1555  if (dst == src || !s1->context_initialized)
1556  return 0;
1557 
1558  if (s->height != s1->height || s->width != s1->width) {
1559  MPV_common_end(s);
1560  s->height = s1->height;
1561  s->width = s1->width;
1562  if ((err = MPV_common_init(s)) < 0)
1563  return err;
1564  if ((err = rv34_decoder_realloc(r)) < 0)
1565  return err;
1566  }
1567 
1568  if ((err = ff_mpeg_update_thread_context(dst, src)))
1569  return err;
1570 
1571  r->cur_pts = r1->cur_pts;
1572  r->last_pts = r1->last_pts;
1573  r->next_pts = r1->next_pts;
1574 
1575  memset(&r->si, 0, sizeof(r->si));
1576 
1577  return 0;
1578 }
1579 
1580 static int get_slice_offset(AVCodecContext *avctx, const uint8_t *buf, int n)
1581 {
1582  if(avctx->slice_count) return avctx->slice_offset[n];
1583  else return AV_RL32(buf + n*8 - 4) == 1 ? AV_RL32(buf + n*8) : AV_RB32(buf + n*8);
1584 }
1585 
1586 static int finish_frame(AVCodecContext *avctx, AVFrame *pict)
1587 {
1588  RV34DecContext *r = avctx->priv_data;
1589  MpegEncContext *s = &r->s;
1590  int got_picture = 0;
1591 
1592  ff_er_frame_end(s);
1593  MPV_frame_end(s);
1594 
1597 
1598  if (s->pict_type == AV_PICTURE_TYPE_B || s->low_delay) {
1599  *pict = s->current_picture_ptr->f;
1600  got_picture = 1;
1601  } else if (s->last_picture_ptr != NULL) {
1602  *pict = s->last_picture_ptr->f;
1603  got_picture = 1;
1604  }
1605  if (got_picture)
1606  ff_print_debug_info(s, pict);
1607 
1608  return got_picture;
1609 }
1610 
1612  void *data, int *got_picture_ptr,
1613  AVPacket *avpkt)
1614 {
1615  const uint8_t *buf = avpkt->data;
1616  int buf_size = avpkt->size;
1617  RV34DecContext *r = avctx->priv_data;
1618  MpegEncContext *s = &r->s;
1619  AVFrame *pict = data;
1620  SliceInfo si;
1621  int i;
1622  int slice_count;
1623  const uint8_t *slices_hdr = NULL;
1624  int last = 0;
1625 
1626  /* no supplementary picture */
1627  if (buf_size == 0) {
1628  /* special case for last picture */
1629  if (s->low_delay==0 && s->next_picture_ptr) {
1630  *pict = s->next_picture_ptr->f;
1631  s->next_picture_ptr = NULL;
1632 
1633  *got_picture_ptr = 1;
1634  }
1635  return 0;
1636  }
1637 
1638  if(!avctx->slice_count){
1639  slice_count = (*buf++) + 1;
1640  slices_hdr = buf + 4;
1641  buf += 8 * slice_count;
1642  buf_size -= 1 + 8 * slice_count;
1643  }else
1644  slice_count = avctx->slice_count;
1645 
1646  //parse first slice header to check whether this frame can be decoded
1647  if(get_slice_offset(avctx, slices_hdr, 0) < 0 ||
1648  get_slice_offset(avctx, slices_hdr, 0) > buf_size){
1649  av_log(avctx, AV_LOG_ERROR, "Slice offset is invalid\n");
1650  return AVERROR_INVALIDDATA;
1651  }
1652  init_get_bits(&s->gb, buf+get_slice_offset(avctx, slices_hdr, 0), (buf_size-get_slice_offset(avctx, slices_hdr, 0))*8);
1653  if(r->parse_slice_header(r, &r->s.gb, &si) < 0 || si.start){
1654  av_log(avctx, AV_LOG_ERROR, "First slice header is incorrect\n");
1655  return AVERROR_INVALIDDATA;
1656  }
1657  if ((!s->last_picture_ptr || !s->last_picture_ptr->f.data[0]) &&
1658  si.type == AV_PICTURE_TYPE_B) {
1659  av_log(avctx, AV_LOG_ERROR, "Invalid decoder state: B-frame without "
1660  "reference data.\n");
1661  return AVERROR_INVALIDDATA;
1662  }
1663  if( (avctx->skip_frame >= AVDISCARD_NONREF && si.type==AV_PICTURE_TYPE_B)
1664  || (avctx->skip_frame >= AVDISCARD_NONKEY && si.type!=AV_PICTURE_TYPE_I)
1665  || avctx->skip_frame >= AVDISCARD_ALL)
1666  return avpkt->size;
1667 
1668  /* first slice */
1669  if (si.start == 0) {
1670  if (s->mb_num_left > 0) {
1671  av_log(avctx, AV_LOG_ERROR, "New frame but still %d MB left.",
1672  s->mb_num_left);
1673  ff_er_frame_end(s);
1674  MPV_frame_end(s);
1675  }
1676 
1677  if (s->width != si.width || s->height != si.height) {
1678  int err;
1679 
1680  if (HAVE_THREADS &&
1682  av_log_missing_feature(s->avctx, "Width/height changing with "
1683  "frame threading is", 0);
1684  return AVERROR_PATCHWELCOME;
1685  }
1686 
1687  av_log(s->avctx, AV_LOG_WARNING, "Changing dimensions to %dx%d\n",
1688  si.width, si.height);
1689  MPV_common_end(s);
1690  s->width = si.width;
1691  s->height = si.height;
1693  if ((err = MPV_common_init(s)) < 0)
1694  return err;
1695  if ((err = rv34_decoder_realloc(r)) < 0)
1696  return err;
1697  }
1698  s->pict_type = si.type ? si.type : AV_PICTURE_TYPE_I;
1699  if (MPV_frame_start(s, s->avctx) < 0)
1700  return -1;
1701  ff_er_frame_start(s);
1702  if (!r->tmp_b_block_base) {
1703  int i;
1704 
1705  r->tmp_b_block_base = av_malloc(s->linesize * 48);
1706  for (i = 0; i < 2; i++)
1707  r->tmp_b_block_y[i] = r->tmp_b_block_base
1708  + i * 16 * s->linesize;
1709  for (i = 0; i < 4; i++)
1710  r->tmp_b_block_uv[i] = r->tmp_b_block_base + 32 * s->linesize
1711  + (i >> 1) * 8 * s->uvlinesize
1712  + (i & 1) * 16;
1713  }
1714  r->cur_pts = si.pts;
1715  if (s->pict_type != AV_PICTURE_TYPE_B) {
1716  r->last_pts = r->next_pts;
1717  r->next_pts = r->cur_pts;
1718  } else {
1719  int refdist = GET_PTS_DIFF(r->next_pts, r->last_pts);
1720  int dist0 = GET_PTS_DIFF(r->cur_pts, r->last_pts);
1721  int dist1 = GET_PTS_DIFF(r->next_pts, r->cur_pts);
1722 
1723  if (!refdist) {
1724  r->weight1 = r->weight2 = 8192;
1725  } else {
1726  r->weight1 = (dist0 << 14) / refdist;
1727  r->weight2 = (dist1 << 14) / refdist;
1728  }
1729  }
1730  s->mb_x = s->mb_y = 0;
1732  } else if (HAVE_THREADS &&
1734  av_log(s->avctx, AV_LOG_ERROR, "Decoder needs full frames in frame "
1735  "multithreading mode (start MB is %d).\n", si.start);
1736  return AVERROR_INVALIDDATA;
1737  }
1738 
1739  for(i = 0; i < slice_count; i++){
1740  int offset = get_slice_offset(avctx, slices_hdr, i);
1741  int size;
1742  if(i+1 == slice_count)
1743  size = buf_size - offset;
1744  else
1745  size = get_slice_offset(avctx, slices_hdr, i+1) - offset;
1746 
1747  if(offset < 0 || offset > buf_size){
1748  av_log(avctx, AV_LOG_ERROR, "Slice offset is invalid\n");
1749  break;
1750  }
1751 
1752  r->si.end = s->mb_width * s->mb_height;
1753  s->mb_num_left = r->s.mb_x + r->s.mb_y*r->s.mb_width - r->si.start;
1754 
1755  if(i+1 < slice_count){
1756  if (get_slice_offset(avctx, slices_hdr, i+1) < 0 ||
1757  get_slice_offset(avctx, slices_hdr, i+1) > buf_size) {
1758  av_log(avctx, AV_LOG_ERROR, "Slice offset is invalid\n");
1759  break;
1760  }
1761  init_get_bits(&s->gb, buf+get_slice_offset(avctx, slices_hdr, i+1), (buf_size-get_slice_offset(avctx, slices_hdr, i+1))*8);
1762  if(r->parse_slice_header(r, &r->s.gb, &si) < 0){
1763  if(i+2 < slice_count)
1764  size = get_slice_offset(avctx, slices_hdr, i+2) - offset;
1765  else
1766  size = buf_size - offset;
1767  }else
1768  r->si.end = si.start;
1769  }
1770  if (size < 0 || size > buf_size - offset) {
1771  av_log(avctx, AV_LOG_ERROR, "Slice size is invalid\n");
1772  break;
1773  }
1774  last = rv34_decode_slice(r, r->si.end, buf + offset, size);
1775  if(last)
1776  break;
1777  }
1778 
1779  if (s->current_picture_ptr) {
1780  if (last) {
1781  if(r->loop_filter)
1782  r->loop_filter(r, s->mb_height - 1);
1783 
1784  *got_picture_ptr = finish_frame(avctx, pict);
1785  } else if (HAVE_THREADS &&
1787  av_log(avctx, AV_LOG_INFO, "marking unfished frame as finished\n");
1788  /* always mark the current frame as finished, frame-mt supports
1789  * only complete frames */
1790  ff_er_frame_end(s);
1791  MPV_frame_end(s);
1793  return AVERROR_INVALIDDATA;
1794  }
1795  }
1796 
1797  return avpkt->size;
1798 }
1799 
1801 {
1802  RV34DecContext *r = avctx->priv_data;
1803 
1804  MPV_common_end(&r->s);
1805  rv34_decoder_free(r);
1806 
1807  return 0;
1808 }