aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/4xm.c
blob: cb315cd7e4f2aabf8e9be8614b64f2217156650a (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
/*
 * 4XM codec
 * Copyright (c) 2003 Michael Niedermayer
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * 4XM codec.
 */

#include <inttypes.h>

#include "libavutil/avassert.h"
#include "libavutil/frame.h"
#include "libavutil/imgutils.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/mem_internal.h"
#include "libavutil/thread.h"
#include "avcodec.h"
#include "blockdsp.h"
#include "bswapdsp.h"
#include "bytestream.h"
#include "get_bits.h"
#include "internal.h"


#define BLOCK_TYPE_VLC_BITS 5
#define ACDC_VLC_BITS 9

#define CFRAME_BUFFER_COUNT 100

static const uint8_t block_type_tab[2][4][8][2] = {
    {
        {    // { 8, 4, 2 } x { 8, 4, 2}
            { 0, 1 }, { 2, 2 }, { 6, 3 }, { 14, 4 }, { 30, 5 }, { 31, 5 }, { 0, 0 }
        }, { // { 8, 4 } x 1
            { 0, 1 }, { 0, 0 }, { 2, 2 }, { 6, 3 }, { 14, 4 }, { 15, 4 }, { 0, 0 }
        }, { // 1 x { 8, 4 }
            { 0, 1 }, { 2, 2 }, { 0, 0 }, { 6, 3 }, { 14, 4 }, { 15, 4 }, { 0, 0 }
        }, { // 1 x 2, 2 x 1
            { 0, 1 }, { 0, 0 }, { 0, 0 }, { 2, 2 }, { 6, 3 }, { 14, 4 }, { 15, 4 }
        }
    }, {
        {   // { 8, 4, 2 } x { 8, 4, 2}
            { 1, 2 }, { 4, 3 }, { 5, 3 }, { 0, 2 }, { 6, 3 }, { 7, 3 }, { 0, 0 }
        }, {// { 8, 4 } x 1
            { 1, 2 }, { 0, 0 }, { 2, 2 }, { 0, 2 }, { 6, 3 }, { 7, 3 }, { 0, 0 }
        }, {// 1 x { 8, 4 }
            { 1, 2 }, { 2, 2 }, { 0, 0 }, { 0, 2 }, { 6, 3 }, { 7, 3 }, { 0, 0 }
        }, {// 1 x 2, 2 x 1
            { 1, 2 }, { 0, 0 }, { 0, 0 }, { 0, 2 }, { 2, 2 }, { 6, 3 }, { 7, 3 }
      }
    }
};

static const uint8_t size2index[4][4] = {
    { -1, 3, 1, 1 },
    {  3, 0, 0, 0 },
    {  2, 0, 0, 0 },
    {  2, 0, 0, 0 },
};

static const int8_t mv[256][2] = {
    {   0,   0 }, {   0,  -1 }, {  -1,   0 }, {   1,   0 }, {   0,   1 }, {  -1,  -1 }, {   1,  -1 }, {  -1,   1 },
    {   1,   1 }, {   0,  -2 }, {  -2,   0 }, {   2,   0 }, {   0,   2 }, {  -1,  -2 }, {   1,  -2 }, {  -2,  -1 },
    {   2,  -1 }, {  -2,   1 }, {   2,   1 }, {  -1,   2 }, {   1,   2 }, {  -2,  -2 }, {   2,  -2 }, {  -2,   2 },
    {   2,   2 }, {   0,  -3 }, {  -3,   0 }, {   3,   0 }, {   0,   3 }, {  -1,  -3 }, {   1,  -3 }, {  -3,  -1 },
    {   3,  -1 }, {  -3,   1 }, {   3,   1 }, {  -1,   3 }, {   1,   3 }, {  -2,  -3 }, {   2,  -3 }, {  -3,  -2 },
    {   3,  -2 }, {  -3,   2 }, {   3,   2 }, {  -2,   3 }, {   2,   3 }, {   0,  -4 }, {  -4,   0 }, {   4,   0 },
    {   0,   4 }, {  -1,  -4 }, {   1,  -4 }, {  -4,  -1 }, {   4,  -1 }, {   4,   1 }, {  -1,   4 }, {   1,   4 },
    {  -3,  -3 }, {  -3,   3 }, {   3,   3 }, {  -2,  -4 }, {  -4,  -2 }, {   4,  -2 }, {  -4,   2 }, {  -2,   4 },
    {   2,   4 }, {  -3,  -4 }, {   3,  -4 }, {   4,  -3 }, {  -5,   0 }, {  -4,   3 }, {  -3,   4 }, {   3,   4 },
    {  -1,  -5 }, {  -5,  -1 }, {  -5,   1 }, {  -1,   5 }, {  -2,  -5 }, {   2,  -5 }, {   5,  -2 }, {   5,   2 },
    {  -4,  -4 }, {  -4,   4 }, {  -3,  -5 }, {  -5,  -3 }, {  -5,   3 }, {   3,   5 }, {  -6,   0 }, {   0,   6 },
    {  -6,  -1 }, {  -6,   1 }, {   1,   6 }, {   2,  -6 }, {  -6,   2 }, {   2,   6 }, {  -5,  -4 }, {   5,   4 },
    {   4,   5 }, {  -6,  -3 }, {   6,   3 }, {  -7,   0 }, {  -1,  -7 }, {   5,  -5 }, {  -7,   1 }, {  -1,   7 },
    {   4,  -6 }, {   6,   4 }, {  -2,  -7 }, {  -7,   2 }, {  -3,  -7 }, {   7,  -3 }, {   3,   7 }, {   6,  -5 },
    {   0,  -8 }, {  -1,  -8 }, {  -7,  -4 }, {  -8,   1 }, {   4,   7 }, {   2,  -8 }, {  -2,   8 }, {   6,   6 },
    {  -8,   3 }, {   5,  -7 }, {  -5,   7 }, {   8,  -4 }, {   0,  -9 }, {  -9,  -1 }, {   1,   9 }, {   7,  -6 },
    {  -7,   6 }, {  -5,  -8 }, {  -5,   8 }, {  -9,   3 }, {   9,  -4 }, {   7,  -7 }, {   8,  -6 }, {   6,   8 },
    {  10,   1 }, { -10,   2 }, {   9,  -5 }, {  10,  -3 }, {  -8,  -7 }, { -10,  -4 }, {   6,  -9 }, { -11,   0 },
    {  11,   1 }, { -11,  -2 }, {  -2,  11 }, {   7,  -9 }, {  -7,   9 }, {  10,   6 }, {  -4,  11 }, {   8,  -9 },
    {   8,   9 }, {   5,  11 }, {   7, -10 }, {  12,  -3 }, {  11,   6 }, {  -9,  -9 }, {   8,  10 }, {   5,  12 },
    { -11,   7 }, {  13,   2 }, {   6, -12 }, {  10,   9 }, { -11,   8 }, {  -7,  12 }, {   0,  14 }, {  14,  -2 },
    {  -9,  11 }, {  -6,  13 }, { -14,  -4 }, {  -5, -14 }, {   5,  14 }, { -15,  -1 }, { -14,  -6 }, {   3, -15 },
    {  11, -11 }, {  -7,  14 }, {  -5,  15 }, {   8, -14 }, {  15,   6 }, {   3,  16 }, {   7, -15 }, { -16,   5 },
    {   0,  17 }, { -16,  -6 }, { -10,  14 }, { -16,   7 }, {  12,  13 }, { -16,   8 }, { -17,   6 }, { -18,   3 },
    {  -7,  17 }, {  15,  11 }, {  16,  10 }, {   2, -19 }, {   3, -19 }, { -11, -16 }, { -18,   8 }, { -19,  -6 },
    {   2, -20 }, { -17, -11 }, { -10, -18 }, {   8,  19 }, { -21,  -1 }, { -20,   7 }, {  -4,  21 }, {  21,   5 },
    {  15,  16 }, {   2, -22 }, { -10, -20 }, { -22,   5 }, {  20, -11 }, {  -7, -22 }, { -12,  20 }, {  23,  -5 },
    {  13, -20 }, {  24,  -2 }, { -15,  19 }, { -11,  22 }, {  16,  19 }, {  23, -10 }, { -18, -18 }, {  -9, -24 },
    {  24, -10 }, {  -3,  26 }, { -23,  13 }, { -18, -20 }, {  17,  21 }, {  -4,  27 }, {  27,   6 }, {   1, -28 },
    { -11,  26 }, { -17, -23 }, {   7,  28 }, {  11, -27 }, {  29,   5 }, { -23, -19 }, { -28, -11 }, { -21,  22 },
    { -30,   7 }, { -17,  26 }, { -27,  16 }, {  13,  29 }, {  19, -26 }, {  10, -31 }, { -14, -30 }, {  20, -27 },
    { -29,  18 }, { -16, -31 }, { -28, -22 }, {  21, -30 }, { -25,  28 }, {  26, -29 }, {  25, -32 }, { -32, -32 }
};

/* This is simply the scaled down elementwise product of the standard JPEG
 * quantizer table and the AAN premul table. */
static const uint8_t dequant_table[64] = {
    16, 15, 13, 19, 24, 31, 28, 17,
    17, 23, 25, 31, 36, 63, 45, 21,
    18, 24, 27, 37, 52, 59, 49, 20,
    16, 28, 34, 40, 60, 80, 51, 20,
    18, 31, 48, 66, 68, 86, 56, 21,
    19, 38, 56, 59, 64, 64, 48, 20,
    27, 48, 55, 55, 56, 51, 35, 15,
    20, 35, 34, 32, 31, 22, 15,  8,
};

static VLC block_type_vlc[2][4];


typedef struct CFrameBuffer {
    unsigned int allocated_size;
    unsigned int size;
    int id;
    uint8_t *data;
} CFrameBuffer;

typedef struct FourXContext {
    AVCodecContext *avctx;
    BlockDSPContext bdsp;
    BswapDSPContext bbdsp;
    uint16_t *frame_buffer;
    uint16_t *last_frame_buffer;
    GetBitContext pre_gb;          ///< ac/dc prefix
    GetBitContext gb;
    GetByteContext g;
    GetByteContext g2;
    int mv[256];
    VLC pre_vlc;
    int last_dc;
    DECLARE_ALIGNED(32, int16_t, block)[6][64];
    void *bitstream_buffer;
    unsigned int bitstream_buffer_size;
    int version;
    CFrameBuffer cfrm[CFRAME_BUFFER_COUNT];
} FourXContext;


#define FIX_1_082392200  70936
#define FIX_1_414213562  92682
#define FIX_1_847759065 121095
#define FIX_2_613125930 171254

#define MULTIPLY(var, const) ((int)((var) * (unsigned)(const)) >> 16)

static void idct(int16_t block[64])
{
    int tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
    int tmp10, tmp11, tmp12, tmp13;
    int z5, z10, z11, z12, z13;
    int i;
    int temp[64];

    for (i = 0; i < 8; i++) {
        tmp10 = block[8 * 0 + i] + block[8 * 4 + i];
        tmp11 = block[8 * 0 + i] - block[8 * 4 + i];

        tmp13 = block[8 * 2 + i] + block[8 * 6 + i];
        tmp12 = MULTIPLY(block[8 * 2 + i] - block[8 * 6 + i], FIX_1_414213562) - tmp13;

        tmp0 = tmp10 + tmp13;
        tmp3 = tmp10 - tmp13;
        tmp1 = tmp11 + tmp12;
        tmp2 = tmp11 - tmp12;

        z13 = block[8 * 5 + i] + block[8 * 3 + i];
        z10 = block[8 * 5 + i] - block[8 * 3 + i];
        z11 = block[8 * 1 + i] + block[8 * 7 + i];
        z12 = block[8 * 1 + i] - block[8 * 7 + i];

        tmp7  =          z11 + z13;
        tmp11 = MULTIPLY(z11 - z13, FIX_1_414213562);

        z5    = MULTIPLY(z10 + z12, FIX_1_847759065);
        tmp10 = MULTIPLY(z12,  FIX_1_082392200) - z5;
        tmp12 = MULTIPLY(z10, -FIX_2_613125930) + z5;

        tmp6 = tmp12 - tmp7;
        tmp5 = tmp11 - tmp6;
        tmp4 = tmp10 + tmp5;

        temp[8 * 0 + i] = tmp0 + tmp7;
        temp[8 * 7 + i] = tmp0 - tmp7;
        temp[8 * 1 + i] = tmp1 + tmp6;
        temp[8 * 6 + i] = tmp1 - tmp6;
        temp[8 * 2 + i] = tmp2 + tmp5;
        temp[8 * 5 + i] = tmp2 - tmp5;
        temp[8 * 4 + i] = tmp3 + tmp4;
        temp[8 * 3 + i] = tmp3 - tmp4;
    }

    for (i = 0; i < 8 * 8; i += 8) {
        tmp10 = temp[0 + i] + temp[4 + i];
        tmp11 = temp[0 + i] - temp[4 + i];

        tmp13 = temp[2 + i] + temp[6 + i];
        tmp12 = MULTIPLY(temp[2 + i] - temp[6 + i], FIX_1_414213562) - tmp13;

        tmp0 = tmp10 + tmp13;
        tmp3 = tmp10 - tmp13;
        tmp1 = tmp11 + tmp12;
        tmp2 = tmp11 - tmp12;

        z13 = temp[5 + i] + temp[3 + i];
        z10 = temp[5 + i] - temp[3 + i];
        z11 = temp[1 + i] + temp[7 + i];
        z12 = temp[1 + i] - temp[7 + i];

        tmp7  = z11 + z13;
        tmp11 = MULTIPLY(z11 - z13, FIX_1_414213562);

        z5    = MULTIPLY(z10 + z12, FIX_1_847759065);
        tmp10 = MULTIPLY(z12,  FIX_1_082392200) - z5;
        tmp12 = MULTIPLY(z10, -FIX_2_613125930) + z5;

        tmp6 = tmp12 - tmp7;
        tmp5 = tmp11 - tmp6;
        tmp4 = tmp10 + tmp5;

        block[0 + i] = (tmp0 + tmp7) >> 6;
        block[7 + i] = (tmp0 - tmp7) >> 6;
        block[1 + i] = (tmp1 + tmp6) >> 6;
        block[6 + i] = (tmp1 - tmp6) >> 6;
        block[2 + i] = (tmp2 + tmp5) >> 6;
        block[5 + i] = (tmp2 - tmp5) >> 6;
        block[4 + i] = (tmp3 + tmp4) >> 6;
        block[3 + i] = (tmp3 - tmp4) >> 6;
    }
}

static av_cold void init_vlcs(void)
{
    static VLC_TYPE table[2][4][32][2];
    int i, j;

    for (i = 0; i < 2; i++) {
        for (j = 0; j < 4; j++) {
            block_type_vlc[i][j].table           = table[i][j];
            block_type_vlc[i][j].table_allocated = 32;
            init_vlc(&block_type_vlc[i][j], BLOCK_TYPE_VLC_BITS, 7,
                     &block_type_tab[i][j][0][1], 2, 1,
                     &block_type_tab[i][j][0][0], 2, 1,
                     INIT_VLC_USE_NEW_STATIC);
        }
    }
}

static void init_mv(FourXContext *f, int linesize)
{
    int i;

    for (i = 0; i < 256; i++) {
        if (f->version > 1)
            f->mv[i] = mv[i][0] + mv[i][1] * linesize / 2;
        else
            f->mv[i] = (i & 15) - 8 + ((i >> 4) - 8) * linesize / 2;
    }
}

#if HAVE_BIGENDIAN
#define LE_CENTRIC_MUL(dst, src, scale, dc)             \
    {                                                   \
        unsigned tmpval = AV_RN32(src);                 \
        tmpval = (tmpval << 16) | (tmpval >> 16);       \
        tmpval = tmpval * (scale) + (dc);               \
        tmpval = (tmpval << 16) | (tmpval >> 16);       \
        AV_WN32A(dst, tmpval);                          \
    }
#else
#define LE_CENTRIC_MUL(dst, src, scale, dc)              \
    {                                                    \
        unsigned tmpval = AV_RN32(src) * (scale) + (dc); \
        AV_WN32A(dst, tmpval);                           \
    }
#endif

static inline void mcdc(uint16_t *dst, const uint16_t *src, int log2w,
                        int h, int stride, int scale, unsigned dc)
{
    int i;
    dc *= 0x10001;

    switch (log2w) {
    case 0:
        for (i = 0; i < h; i++) {
            dst[0] = scale * src[0] + dc;
            if (scale)
                src += stride;
            dst += stride;
        }
        break;
    case 1:
        for (i = 0; i < h; i++) {
            LE_CENTRIC_MUL(dst, src, scale, dc);
            if (scale)
                src += stride;
            dst += stride;
        }
        break;
    case 2:
        for (i = 0; i < h; i++) {
            LE_CENTRIC_MUL(dst, src, scale, dc);
            LE_CENTRIC_MUL(dst + 2, src + 2, scale, dc);
            if (scale)
                src += stride;
            dst += stride;
        }
        break;
    case 3:
        for (i = 0; i < h; i++) {
            LE_CENTRIC_MUL(dst,     src,     scale, dc);
            LE_CENTRIC_MUL(dst + 2, src + 2, scale, dc);
            LE_CENTRIC_MUL(dst + 4, src + 4, scale, dc);
            LE_CENTRIC_MUL(dst + 6, src + 6, scale, dc);
            if (scale)
                src += stride;
            dst += stride;
        }
        break;
    default:
        av_assert0(0);
    }
}

static int decode_p_block(FourXContext *f, uint16_t *dst, const uint16_t *src,
                          int log2w, int log2h, int stride)
{
    int index, h, code, ret, scale = 1;
    uint16_t *start, *end;
    unsigned dc = 0;

    av_assert0(log2w >= 0 && log2h >= 0);

    index = size2index[log2h][log2w];
    av_assert0(index >= 0);

    if (get_bits_left(&f->gb) < 1)
        return AVERROR_INVALIDDATA;
    h     = 1 << log2h;
    code  = get_vlc2(&f->gb, block_type_vlc[1 - (f->version > 1)][index].table,
                     BLOCK_TYPE_VLC_BITS, 1);
    av_assert0(code >= 0 && code <= 6);

    start = f->last_frame_buffer;
    end   = start + stride * (f->avctx->height - h + 1) - (1 << log2w);

    if (code == 1) {
        log2h--;
        if ((ret = decode_p_block(f, dst, src, log2w, log2h, stride)) < 0)
            return ret;
        return decode_p_block(f, dst + (stride << log2h),
                              src + (stride << log2h),
                              log2w, log2h, stride);
    } else if (code == 2) {
        log2w--;
        if ((ret = decode_p_block(f, dst , src, log2w, log2h, stride)) < 0)
            return ret;
        return decode_p_block(f, dst + (1 << log2w),
                              src + (1 << log2w),
                              log2w, log2h, stride);
    } else if (code == 6) {
        if (bytestream2_get_bytes_left(&f->g2) < 4) {
            av_log(f->avctx, AV_LOG_ERROR, "wordstream overread\n");
            return AVERROR_INVALIDDATA;
        }
        if (log2w) {
            dst[0]      = bytestream2_get_le16u(&f->g2);
            dst[1]      = bytestream2_get_le16u(&f->g2);
        } else {
            dst[0]      = bytestream2_get_le16u(&f->g2);
            dst[stride] = bytestream2_get_le16u(&f->g2);
        }
        return 0;
    }

    if ((code&3)==0 && bytestream2_get_bytes_left(&f->g) < 1) {
        av_log(f->avctx, AV_LOG_ERROR, "bytestream overread\n");
        return AVERROR_INVALIDDATA;
    }

    if (code == 0) {
        src  += f->mv[bytestream2_get_byte(&f->g)];
    } else if (code == 3 && f->version >= 2) {
        return 0;
    } else if (code == 4) {
        src  += f->mv[bytestream2_get_byte(&f->g)];
        if (bytestream2_get_bytes_left(&f->g2) < 2){
            av_log(f->avctx, AV_LOG_ERROR, "wordstream overread\n");
            return AVERROR_INVALIDDATA;
        }
        dc    = bytestream2_get_le16(&f->g2);
    } else if (code == 5) {
        if (bytestream2_get_bytes_left(&f->g2) < 2){
            av_log(f->avctx, AV_LOG_ERROR, "wordstream overread\n");
            return AVERROR_INVALIDDATA;
        }
        av_assert0(start <= src && src <= end);
        scale = 0;
        dc    = bytestream2_get_le16(&f->g2);
    }

    if (start > src || src > end) {
        av_log(f->avctx, AV_LOG_ERROR, "mv out of pic\n");
        return AVERROR_INVALIDDATA;
    }

    mcdc(dst, src, log2w, h, stride, scale, dc);

    return 0;
}

static int decode_p_frame(FourXContext *f, const uint8_t *buf, int length)
{
    int x, y;
    const int width  = f->avctx->width;
    const int height = f->avctx->height;
    uint16_t *dst    = f->frame_buffer;
    uint16_t *src;
    unsigned int bitstream_size, bytestream_size, wordstream_size, extra,
                 bytestream_offset, wordstream_offset;
    int ret;

    src = f->last_frame_buffer;

    if (f->version > 1) {
        extra           = 20;
        if (length < extra)
            return AVERROR_INVALIDDATA;
        bitstream_size  = AV_RL32(buf + 8);
        wordstream_size = AV_RL32(buf + 12);
        bytestream_size = AV_RL32(buf + 16);
    } else {
        extra           = 0;
        bitstream_size  = AV_RL16(buf - 4);
        wordstream_size = AV_RL16(buf - 2);
        bytestream_size = FFMAX(length - bitstream_size - wordstream_size, 0);
    }

    if (bitstream_size > length || bitstream_size >= INT_MAX/8 ||
        bytestream_size > length - bitstream_size ||
        wordstream_size > length - bytestream_size - bitstream_size ||
        extra > length - bytestream_size - bitstream_size - wordstream_size) {
        av_log(f->avctx, AV_LOG_ERROR, "lengths %d %d %d %d\n", bitstream_size, bytestream_size, wordstream_size,
        bitstream_size+ bytestream_size+ wordstream_size - length);
        return AVERROR_INVALIDDATA;
    }

    av_fast_padded_malloc(&f->bitstream_buffer, &f->bitstream_buffer_size,
                          bitstream_size);
    if (!f->bitstream_buffer)
        return AVERROR(ENOMEM);
    f->bbdsp.bswap_buf(f->bitstream_buffer, (const uint32_t *) (buf + extra),
                       bitstream_size / 4);
    init_get_bits(&f->gb, f->bitstream_buffer, 8 * bitstream_size);

    wordstream_offset = extra + bitstream_size;
    bytestream_offset = extra + bitstream_size + wordstream_size;
    bytestream2_init(&f->g2, buf + wordstream_offset,
                     length - wordstream_offset);
    bytestream2_init(&f->g, buf + bytestream_offset,
                     length - bytestream_offset);

    init_mv(f, width * 2);

    for (y = 0; y < height; y += 8) {
        for (x = 0; x < width; x += 8)
            if ((ret = decode_p_block(f, dst + x, src + x, 3, 3, width)) < 0)
                return ret;
        src += 8 * width;
        dst += 8 * width;
    }

    return 0;
}

/**
 * decode block and dequantize.
 * Note this is almost identical to MJPEG.
 */
static int decode_i_block(FourXContext *f, int16_t *block)
{
    int code, i, j, level, val;

    if (get_bits_left(&f->pre_gb) < 2) {
        av_log(f->avctx, AV_LOG_ERROR, "%d bits left before decode_i_block()\n", get_bits_left(&f->pre_gb));
        return AVERROR_INVALIDDATA;
    }

    /* DC coef */
    val = get_vlc2(&f->pre_gb, f->pre_vlc.table, ACDC_VLC_BITS, 3);
    if (val >> 4) {
        av_log(f->avctx, AV_LOG_ERROR, "error dc run != 0\n");
        return AVERROR_INVALIDDATA;
    }

    if (val)
        val = get_xbits(&f->gb, val);

    val        = val * dequant_table[0] + f->last_dc;
    f->last_dc = block[0] = val;
    /* AC coefs */
    i = 1;
    for (;;) {
        code = get_vlc2(&f->pre_gb, f->pre_vlc.table, ACDC_VLC_BITS, 3);

        /* EOB */
        if (code == 0)
            break;
        if (code == 0xf0) {
            i += 16;
            if (i >= 64) {
                av_log(f->avctx, AV_LOG_ERROR, "run %d overflow\n", i);
                return 0;
            }
        } else {
            if (code & 0xf) {
                level = get_xbits(&f->gb, code & 0xf);
            } else {
                av_log(f->avctx, AV_LOG_ERROR, "0 coeff\n");
                return AVERROR_INVALIDDATA;
            }
            i    += code >> 4;
            if (i >= 64) {
                av_log(f->avctx, AV_LOG_ERROR, "run %d overflow\n", i);
                return 0;
            }

            j = ff_zigzag_direct[i];
            block[j] = level * dequant_table[j];
            i++;
            if (i >= 64)
                break;
        }
    }

    return 0;
}

static inline void idct_put(FourXContext *f, int x, int y)
{
    int16_t (*block)[64] = f->block;
    int stride           = f->avctx->width;
    int i;
    uint16_t *dst = f->frame_buffer + y * stride + x;

    for (i = 0; i < 4; i++) {
        block[i][0] += 0x80 * 8 * 8;
        idct(block[i]);
    }

    if (!(f->avctx->flags & AV_CODEC_FLAG_GRAY)) {
        for (i = 4; i < 6; i++)
            idct(block[i]);
    }

    /* Note transform is:
     * y  = ( 1b + 4g + 2r) / 14
     * cb = ( 3b - 2g - 1r) / 14
     * cr = (-1b - 4g + 5r) / 14 */
    for (y = 0; y < 8; y++) {
        for (x = 0; x < 8; x++) {
            int16_t *temp = block[(x >> 2) + 2 * (y >> 2)] +
                            2 * (x & 3) + 2 * 8 * (y & 3); // FIXME optimize
            int cb = block[4][x + 8 * y];
            int cr = block[5][x + 8 * y];
            int cg = (cb + cr) >> 1;
            int y;

            cb += cb;

            y               = temp[0];
            dst[0]          = ((y + cb) >> 3) + (((y - cg) & 0xFC) << 3) + (((y + cr) & 0xF8) << 8);
            y               = temp[1];
            dst[1]          = ((y + cb) >> 3) + (((y - cg) & 0xFC) << 3) + (((y + cr) & 0xF8) << 8);
            y               = temp[8];
            dst[stride]     = ((y + cb) >> 3) + (((y - cg) & 0xFC) << 3) + (((y + cr) & 0xF8) << 8);
            y               = temp[9];
            dst[1 + stride] = ((y + cb) >> 3) + (((y - cg) & 0xFC) << 3) + (((y + cr) & 0xF8) << 8);
            dst            += 2;
        }
        dst += 2 * stride - 2 * 8;
    }
}

static int decode_i_mb(FourXContext *f)
{
    int ret;
    int i;

    f->bdsp.clear_blocks(f->block[0]);

    for (i = 0; i < 6; i++)
        if ((ret = decode_i_block(f, f->block[i])) < 0)
            return ret;

    return 0;
}

static const uint8_t *read_huffman_tables(FourXContext *f,
                                          const uint8_t * const buf,
                                          int buf_size)
{
    int frequency[512] = { 0 };
    uint8_t flag[512];
    int up[512];
    uint8_t len_tab[257];
    int bits_tab[257];
    int start, end;
    const uint8_t *ptr = buf;
    const uint8_t *ptr_end = buf + buf_size;
    int j;

    memset(up, -1, sizeof(up));

    start = *ptr++;
    end   = *ptr++;
    for (;;) {
        int i;

        if (ptr_end - ptr < FFMAX(end - start + 1, 0) + 1) {
            av_log(f->avctx, AV_LOG_ERROR, "invalid data in read_huffman_tables\n");
            return NULL;
        }

        for (i = start; i <= end; i++)
            frequency[i] = *ptr++;
        start = *ptr++;
        if (start == 0)
            break;

        end = *ptr++;
    }
    frequency[256] = 1;

    while ((ptr - buf) & 3)
        ptr++; // 4byte align

    if (ptr > ptr_end) {
        av_log(f->avctx, AV_LOG_ERROR, "ptr overflow in read_huffman_tables\n");
        return NULL;
    }

    for (j = 257; j < 512; j++) {
        int min_freq[2] = { 256 * 256, 256 * 256 };
        int smallest[2] = { 0, 0 };
        int i;
        for (i = 0; i < j; i++) {
            if (frequency[i] == 0)
                continue;
            if (frequency[i] < min_freq[1]) {
                if (frequency[i] < min_freq[0]) {
                    min_freq[1] = min_freq[0];
                    smallest[1] = smallest[0];
                    min_freq[0] = frequency[i];
                    smallest[0] = i;
                } else {
                    min_freq[1] = frequency[i];
                    smallest[1] = i;
                }
            }
        }
        if (min_freq[1] == 256 * 256)
            break;

        frequency[j]           = min_freq[0] + min_freq[1];
        flag[smallest[0]]      = 0;
        flag[smallest[1]]      = 1;
        up[smallest[0]]        =
        up[smallest[1]]        = j;
        frequency[smallest[0]] = frequency[smallest[1]] = 0;
    }

    for (j = 0; j < 257; j++) {
        int node, len = 0, bits = 0;

        for (node = j; up[node] != -1; node = up[node]) {
            bits += flag[node] << len;
            len++;
            if (len > 31)
                // can this happen at all ?
                av_log(f->avctx, AV_LOG_ERROR,
                       "vlc length overflow\n");
        }

        bits_tab[j] = bits;
        len_tab[j]  = len;
    }

    ff_free_vlc(&f->pre_vlc);
    if (init_vlc(&f->pre_vlc, ACDC_VLC_BITS, 257, len_tab, 1, 1,
                 bits_tab, 4, 4, 0))
        return NULL;

    return ptr;
}

static int mix(int c0, int c1)
{
    int blue  =  2 * (c0 & 0x001F) + (c1 & 0x001F);
    int green = (2 * (c0 & 0x03E0) + (c1 & 0x03E0)) >> 5;
    int red   =  2 * (c0 >> 10)    + (c1 >> 10);
    return red / 3 * 1024 + green / 3 * 32 + blue / 3;
}

static int decode_i2_frame(FourXContext *f, const uint8_t *buf, int length)
{
    int x, y, x2, y2;
    const int width  = f->avctx->width;
    const int height = f->avctx->height;
    const int mbs    = (FFALIGN(width, 16) >> 4) * (FFALIGN(height, 16) >> 4);
    uint16_t *dst    = f->frame_buffer;
    const uint8_t *buf_end = buf + length;
    GetByteContext g3;

    if (length < mbs * 8) {
        av_log(f->avctx, AV_LOG_ERROR, "packet size too small\n");
        return AVERROR_INVALIDDATA;
    }
    bytestream2_init(&g3, buf, length);

    for (y = 0; y < height; y += 16) {
        for (x = 0; x < width; x += 16) {
            unsigned int color[4] = { 0 }, bits;
            if (buf_end - buf < 8)
                return AVERROR_INVALIDDATA;
            // warning following is purely guessed ...
            color[0] = bytestream2_get_le16u(&g3);
            color[1] = bytestream2_get_le16u(&g3);

            if (color[0] & 0x8000)
                av_log(f->avctx, AV_LOG_ERROR, "unk bit 1\n");
            if (color[1] & 0x8000)
                av_log(f->avctx, AV_LOG_ERROR, "unk bit 2\n");

            color[2] = mix(color[0], color[1]);
            color[3] = mix(color[1], color[0]);

            bits = bytestream2_get_le32u(&g3);
            for (y2 = 0; y2 < 16; y2++) {
                for (x2 = 0; x2 < 16; x2++) {
                    int index = 2 * (x2 >> 2) + 8 * (y2 >> 2);
                    dst[y2 * width + x2] = color[(bits >> index) & 3];
                }
            }
            dst += 16;
        }
        dst += 16 * width - x;
    }

    return 0;
}

static int decode_i_frame(FourXContext *f, const uint8_t *buf, int length)
{
    int x, y, ret;
    const int width  = f->avctx->width;
    const int height = f->avctx->height;
    const unsigned int bitstream_size = AV_RL32(buf);
    unsigned int prestream_size;
    const uint8_t *prestream;

    if (bitstream_size > (1 << 26))
        return AVERROR_INVALIDDATA;

    if (length < bitstream_size + 12) {
        av_log(f->avctx, AV_LOG_ERROR, "packet size too small\n");
        return AVERROR_INVALIDDATA;
    }

    prestream_size = 4 * AV_RL32(buf + bitstream_size + 4);
    prestream      =             buf + bitstream_size + 12;

    if (prestream_size + bitstream_size + 12 != length
        || prestream_size > (1 << 26)) {
        av_log(f->avctx, AV_LOG_ERROR, "size mismatch %d %d %d\n",
               prestream_size, bitstream_size, length);
        return AVERROR_INVALIDDATA;
    }

    prestream = read_huffman_tables(f, prestream, prestream_size);
    if (!prestream) {
        av_log(f->avctx, AV_LOG_ERROR, "Error reading Huffman tables.\n");
        return AVERROR_INVALIDDATA;
    }

    av_assert0(prestream <= buf + length);

    init_get_bits(&f->gb, buf + 4, 8 * bitstream_size);

    prestream_size = length + buf - prestream;

    av_fast_padded_malloc(&f->bitstream_buffer, &f->bitstream_buffer_size,
                          prestream_size);
    if (!f->bitstream_buffer)
        return AVERROR(ENOMEM);
    f->bbdsp.bswap_buf(f->bitstream_buffer, (const uint32_t *) prestream,
                       prestream_size / 4);
    init_get_bits(&f->pre_gb, f->bitstream_buffer, 8 * prestream_size);

    f->last_dc = 0 * 128 * 8 * 8;

    for (y = 0; y < height; y += 16) {
        for (x = 0; x < width; x += 16) {
            if ((ret = decode_i_mb(f)) < 0)
                return ret;

            idct_put(f, x, y);
        }
    }

    if (get_vlc2(&f->pre_gb, f->pre_vlc.table, ACDC_VLC_BITS, 3) != 256)
        av_log(f->avctx, AV_LOG_ERROR, "end mismatch\n");

    return 0;
}

static int decode_frame(AVCodecContext *avctx, void *data,
                        int *got_frame, AVPacket *avpkt)
{
    const uint8_t *buf    = avpkt->data;
    int buf_size          = avpkt->size;
    FourXContext *const f = avctx->priv_data;
    AVFrame *picture      = data;
    int i, frame_4cc, frame_size, ret;

    if (buf_size < 20)
        return AVERROR_INVALIDDATA;

    av_assert0(avctx->width % 16 == 0 && avctx->height % 16 == 0);

    if (buf_size < AV_RL32(buf + 4) + 8) {
        av_log(f->avctx, AV_LOG_ERROR, "size mismatch %d %"PRIu32"\n",
               buf_size, AV_RL32(buf + 4));
        return AVERROR_INVALIDDATA;
    }

    frame_4cc = AV_RL32(buf);

    if (frame_4cc == AV_RL32("cfrm")) {
        int free_index       = -1;
        int id, whole_size;
        const int data_size  = buf_size - 20;
        CFrameBuffer *cfrm;

        if (f->version <= 1) {
            av_log(f->avctx, AV_LOG_ERROR, "cfrm in version %d\n", f->version);
            return AVERROR_INVALIDDATA;
        }

        id         = AV_RL32(buf + 12);
        whole_size = AV_RL32(buf + 16);

        if (data_size < 0 || whole_size < 0) {
            av_log(f->avctx, AV_LOG_ERROR, "sizes invalid\n");
            return AVERROR_INVALIDDATA;
        }

        for (i = 0; i < CFRAME_BUFFER_COUNT; i++)
            if (f->cfrm[i].id && f->cfrm[i].id < avctx->frame_number)
                av_log(f->avctx, AV_LOG_ERROR, "lost c frame %d\n",
                       f->cfrm[i].id);

        for (i = 0; i < CFRAME_BUFFER_COUNT; i++) {
            if (f->cfrm[i].id == id)
                break;
            if (f->cfrm[i].size == 0)
                free_index = i;
        }

        if (i >= CFRAME_BUFFER_COUNT) {
            i             = free_index;
            f->cfrm[i].id = id;
        }
        cfrm = &f->cfrm[i];

        if (data_size > UINT_MAX -  cfrm->size - AV_INPUT_BUFFER_PADDING_SIZE)
            return AVERROR_INVALIDDATA;

        cfrm->data = av_fast_realloc(cfrm->data, &cfrm->allocated_size,
                                     cfrm->size + data_size + AV_INPUT_BUFFER_PADDING_SIZE);
        // explicit check needed as memcpy below might not catch a NULL
        if (!cfrm->data) {
            av_log(f->avctx, AV_LOG_ERROR, "realloc failure\n");
            return AVERROR(ENOMEM);
        }

        memcpy(cfrm->data + cfrm->size, buf + 20, data_size);
        cfrm->size += data_size;

        if (cfrm->size >= whole_size) {
            buf        = cfrm->data;
            frame_size = cfrm->size;

            if (id != avctx->frame_number)
                av_log(f->avctx, AV_LOG_ERROR, "cframe id mismatch %d %d\n",
                       id, avctx->frame_number);

            if (f->version <= 1)
                return AVERROR_INVALIDDATA;

            cfrm->size = cfrm->id = 0;
            frame_4cc  = AV_RL32("pfrm");
        } else
            return buf_size;
    } else {
        buf        = buf      + 12;
        frame_size = buf_size - 12;
    }

    if ((ret = ff_get_buffer(avctx, picture, 0)) < 0)
        return ret;

    if (frame_4cc == AV_RL32("ifr2")) {
        picture->pict_type = AV_PICTURE_TYPE_I;
        if ((ret = decode_i2_frame(f, buf - 4, frame_size + 4)) < 0) {
            av_log(f->avctx, AV_LOG_ERROR, "decode i2 frame failed\n");
            return ret;
        }
    } else if (frame_4cc == AV_RL32("ifrm")) {
        picture->pict_type = AV_PICTURE_TYPE_I;
        if ((ret = decode_i_frame(f, buf, frame_size)) < 0) {
            av_log(f->avctx, AV_LOG_ERROR, "decode i frame failed\n");
            return ret;
        }
    } else if (frame_4cc == AV_RL32("pfrm") || frame_4cc == AV_RL32("pfr2")) {
        picture->pict_type = AV_PICTURE_TYPE_P;
        if ((ret = decode_p_frame(f, buf, frame_size)) < 0) {
            av_log(f->avctx, AV_LOG_ERROR, "decode p frame failed\n");
            return ret;
        }
    } else if (frame_4cc == AV_RL32("snd_")) {
        av_log(avctx, AV_LOG_ERROR, "ignoring snd_ chunk length:%d\n",
               buf_size);
    } else {
        av_log(avctx, AV_LOG_ERROR, "ignoring unknown chunk length:%d\n",
               buf_size);
    }

    picture->key_frame = picture->pict_type == AV_PICTURE_TYPE_I;

    av_image_copy_plane(picture->data[0], picture->linesize[0],
                        (const uint8_t*)f->frame_buffer,  avctx->width * 2,
                        avctx->width * 2, avctx->height);
    FFSWAP(uint16_t *, f->frame_buffer, f->last_frame_buffer);

    *got_frame = 1;

    emms_c();

    return buf_size;
}

static av_cold int decode_end(AVCodecContext *avctx)
{
    FourXContext * const f = avctx->priv_data;
    int i;

    av_freep(&f->frame_buffer);
    av_freep(&f->last_frame_buffer);
    av_freep(&f->bitstream_buffer);
    f->bitstream_buffer_size = 0;
    for (i = 0; i < CFRAME_BUFFER_COUNT; i++) {
        av_freep(&f->cfrm[i].data);
        f->cfrm[i].allocated_size = 0;
    }
    ff_free_vlc(&f->pre_vlc);

    return 0;
}

static av_cold int decode_init(AVCodecContext *avctx)
{
    static AVOnce init_static_once = AV_ONCE_INIT;
    FourXContext * const f = avctx->priv_data;
    int ret;

    if (avctx->extradata_size != 4 || !avctx->extradata) {
        av_log(avctx, AV_LOG_ERROR, "extradata wrong or missing\n");
        return AVERROR_INVALIDDATA;
    }
    if((avctx->width % 16) || (avctx->height % 16)) {
        av_log(avctx, AV_LOG_ERROR, "unsupported width/height\n");
        return AVERROR_INVALIDDATA;
    }

    ret = av_image_check_size(avctx->width, avctx->height, 0, avctx);
    if (ret < 0)
        return ret;

    f->frame_buffer      = av_mallocz(avctx->width * avctx->height * 2);
    f->last_frame_buffer = av_mallocz(avctx->width * avctx->height * 2);
    if (!f->frame_buffer || !f->last_frame_buffer) {
        decode_end(avctx);
        return AVERROR(ENOMEM);
    }

    f->version = AV_RL32(avctx->extradata) >> 16;
    ff_blockdsp_init(&f->bdsp, avctx);
    ff_bswapdsp_init(&f->bbdsp);
    f->avctx = avctx;

    if (f->version > 2)
        avctx->pix_fmt = AV_PIX_FMT_RGB565;
    else
        avctx->pix_fmt = AV_PIX_FMT_BGR555;

    ff_thread_once(&init_static_once, init_vlcs);

    return 0;
}

const AVCodec ff_fourxm_decoder = {
    .name           = "4xm",
    .long_name      = NULL_IF_CONFIG_SMALL("4X Movie"),
    .type           = AVMEDIA_TYPE_VIDEO,
    .id             = AV_CODEC_ID_4XM,
    .priv_data_size = sizeof(FourXContext),
    .init           = decode_init,
    .close          = decode_end,
    .decode         = decode_frame,
    .capabilities   = AV_CODEC_CAP_DR1,
    .caps_internal  = FF_CODEC_CAP_INIT_THREADSAFE,
};