rdopt.c 101 KB
Newer Older
John Koleszar's avatar
John Koleszar committed
1
/*
2
 *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
John Koleszar's avatar
John Koleszar committed
3
 *
4
 *  Use of this source code is governed by a BSD-style license
5
6
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
7
 *  in the file PATENTS.  All contributing project authors may
8
 *  be found in the AUTHORS file in the root of the source tree.
John Koleszar's avatar
John Koleszar committed
9
10
11
12
13
14
15
 */


#include <stdio.h>
#include <math.h>
#include <limits.h>
#include <assert.h>
John Koleszar's avatar
John Koleszar committed
16
#include "vp8/common/pragmas.h"
John Koleszar's avatar
John Koleszar committed
17
18
19
20
21
22

#include "tokenize.h"
#include "treewriter.h"
#include "onyx_int.h"
#include "modecosts.h"
#include "encodeintra.h"
John Koleszar's avatar
John Koleszar committed
23
24
25
26
27
#include "vp8/common/entropymode.h"
#include "vp8/common/reconinter.h"
#include "vp8/common/reconintra.h"
#include "vp8/common/reconintra4x4.h"
#include "vp8/common/findnearmv.h"
Christian Duvivier's avatar
Christian Duvivier committed
28
#include "vp8/common/quant_common.h"
John Koleszar's avatar
John Koleszar committed
29
30
#include "encodemb.h"
#include "quantize.h"
John Koleszar's avatar
John Koleszar committed
31
32
#include "vp8/common/idct.h"
#include "vp8/common/g_common.h"
John Koleszar's avatar
John Koleszar committed
33
34
#include "variance.h"
#include "mcomp.h"
Yunqing Wang's avatar
Yunqing Wang committed
35
#include "rdopt.h"
Paul Wilkins's avatar
Paul Wilkins committed
36
#include "ratectrl.h"
John Koleszar's avatar
John Koleszar committed
37
38
#include "vpx_mem/vpx_mem.h"
#include "dct.h"
John Koleszar's avatar
John Koleszar committed
39
#include "vp8/common/systemdependent.h"
John Koleszar's avatar
John Koleszar committed
40

41
#include "vp8/common/seg_common.h"
42
#include "vp8/common/pred_common.h"
43

John Koleszar's avatar
John Koleszar committed
44
45
46
47
48
49
#if CONFIG_RUNTIME_CPU_DETECT
#define IF_RTCD(x)  (x)
#else
#define IF_RTCD(x)  NULL
#endif

Scott LaVarnway's avatar
Scott LaVarnway committed
50
51
52
extern void vp8cx_mb_init_quantizer(VP8_COMP *cpi, MACROBLOCK *x);
extern void vp8_update_zbin_extra(VP8_COMP *cpi, MACROBLOCK *x);

John Koleszar's avatar
John Koleszar committed
53
54
#define MAXF(a,b)            (((a) > (b)) ? (a) : (b))

55
static const int auto_speed_thresh[17] =
John Koleszar's avatar
John Koleszar committed
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
{
    1000,
    200,
    150,
    130,
    150,
    125,
    120,
    115,
    115,
    115,
    115,
    115,
    115,
    115,
    115,
    115,
    105
};

const MB_PREDICTION_MODE vp8_mode_order[MAX_MODES] =
{
    ZEROMV,
    DC_PRED,

    NEARESTMV,
    NEARMV,

    ZEROMV,
    NEARESTMV,

    ZEROMV,
    NEARESTMV,

    NEARMV,
    NEARMV,

    V_PRED,
    H_PRED,
    TM_PRED,

    NEWMV,
    NEWMV,
    NEWMV,

    SPLITMV,
    SPLITMV,
    SPLITMV,

    B_PRED,
106
    I8X8_PRED,
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123

    /* dual prediction modes */
    ZEROMV,
    NEARESTMV,
    NEARMV,

    ZEROMV,
    NEARESTMV,
    NEARMV,

    ZEROMV,
    NEARESTMV,
    NEARMV,

    NEWMV,
    NEWMV,
    NEWMV,
John Koleszar's avatar
John Koleszar committed
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
};

const MV_REFERENCE_FRAME vp8_ref_frame_order[MAX_MODES] =
{
    LAST_FRAME,
    INTRA_FRAME,

    LAST_FRAME,
    LAST_FRAME,

    GOLDEN_FRAME,
    GOLDEN_FRAME,

    ALTREF_FRAME,
    ALTREF_FRAME,

    GOLDEN_FRAME,
    ALTREF_FRAME,

    INTRA_FRAME,
    INTRA_FRAME,
    INTRA_FRAME,

    LAST_FRAME,
    GOLDEN_FRAME,
    ALTREF_FRAME,

    LAST_FRAME,
    GOLDEN_FRAME,
    ALTREF_FRAME,

    INTRA_FRAME,
156
    INTRA_FRAME,
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

    /* dual prediction modes */
    LAST_FRAME,
    LAST_FRAME,
    LAST_FRAME,

    ALTREF_FRAME,
    ALTREF_FRAME,
    ALTREF_FRAME,

    GOLDEN_FRAME,
    GOLDEN_FRAME,
    GOLDEN_FRAME,

    LAST_FRAME,
    ALTREF_FRAME,
    GOLDEN_FRAME,
};

const MV_REFERENCE_FRAME vp8_second_ref_frame_order[MAX_MODES] =
{
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    0,

    /* dual prediction modes */
    GOLDEN_FRAME,
    GOLDEN_FRAME,
    GOLDEN_FRAME,

    LAST_FRAME,
    LAST_FRAME,
    LAST_FRAME,

    ALTREF_FRAME,
    ALTREF_FRAME,
    ALTREF_FRAME,

    GOLDEN_FRAME,
    LAST_FRAME,
    ALTREF_FRAME,
John Koleszar's avatar
John Koleszar committed
198
199
200
};

static void fill_token_costs(
201
202
    unsigned int c      [BLOCK_TYPES] [COEF_BANDS] [PREV_COEF_CONTEXTS] [MAX_ENTROPY_TOKENS],
    const vp8_prob p    [BLOCK_TYPES] [COEF_BANDS] [PREV_COEF_CONTEXTS] [ENTROPY_NODES]
John Koleszar's avatar
John Koleszar committed
203
204
205
206
207
208
209
210
211
212
213
214
)
{
    int i, j, k;


    for (i = 0; i < BLOCK_TYPES; i++)
        for (j = 0; j < COEF_BANDS; j++)
            for (k = 0; k < PREV_COEF_CONTEXTS; k++)
                vp8_cost_tokens((int *)(c [i][j][k]), p [i][j][k], vp8_coef_tree);

}

Paul Wilkins's avatar
Paul Wilkins committed
215
static int rd_iifactor [ 32 ] =  {    4,   4,   3,   2,   1,   0,   0,   0,
John Koleszar's avatar
John Koleszar committed
216
217
218
219
220
                                      0,   0,   0,   0,   0,   0,   0,   0,
                                      0,   0,   0,   0,   0,   0,   0,   0,
                                      0,   0,   0,   0,   0,   0,   0,   0,
                                 };

221
// 3* dc_qlookup[Q]*dc_qlookup[Q];
222

223
/* values are now correlated to quantizer */
Paul Wilkins's avatar
Paul Wilkins committed
224
225
226
227
static int sad_per_bit16lut[QINDEX_RANGE];
static int sad_per_bit4lut[QINDEX_RANGE];

void vp8_init_me_luts()
John Koleszar's avatar
John Koleszar committed
228
{
Paul Wilkins's avatar
Paul Wilkins committed
229
230
231
232
233
234
235
236
237
238
239
240
    int i;

    // Initialize the sad lut tables using a formulaic calculation for now
    // This is to make it easier to resolve the impact of experimental changes
    // to the quantizer tables.
    for ( i = 0; i < QINDEX_RANGE; i++ )
    {
        sad_per_bit16lut[i] =
            (int)((0.0418*vp8_convert_qindex_to_q(i)) + 2.4107);
        sad_per_bit4lut[i] = (int)((0.063*vp8_convert_qindex_to_q(i)) + 2.742);
    }
}
John Koleszar's avatar
John Koleszar committed
241

242
243
244
245
246
247
248
249
int compute_rd_mult( int qindex )
{
    int q;

    q = vp8_dc_quant(qindex,0);
    return (3 * q * q) >> 4;
}

John Koleszar's avatar
John Koleszar committed
250
251
void vp8cx_initialize_me_consts(VP8_COMP *cpi, int QIndex)
{
Yaowu Xu's avatar
Yaowu Xu committed
252
253
    cpi->mb.sadperbit16 =  sad_per_bit16lut[QIndex];
    cpi->mb.sadperbit4  =  sad_per_bit4lut[QIndex];
John Koleszar's avatar
John Koleszar committed
254
255
}

256
257

void vp8_initialize_rd_consts(VP8_COMP *cpi, int QIndex)
John Koleszar's avatar
John Koleszar committed
258
259
260
261
262
263
{
    int q;
    int i;

    vp8_clear_system_state();  //__asm emms;

264
265
266
267
    // Further tests required to see if optimum is different
    // for key frames, golden frames and arf frames.
    // if (cpi->common.refresh_golden_frame ||
    //     cpi->common.refresh_alt_ref_frame)
Paul Wilkins's avatar
Paul Wilkins committed
268
    QIndex=(QIndex<0)? 0 : ((QIndex>MAXQ)?MAXQ : QIndex);
269
270

    cpi->RDMULT = compute_rd_mult(QIndex);
John Koleszar's avatar
John Koleszar committed
271

272
273
274
275
276
277
278
279
    // Extend rate multiplier along side quantizer zbin increases
    if (cpi->zbin_over_quant  > 0)
    {
        double oq_factor;

        // Experimental code using the same basic equation as used for Q above
        // The units of cpi->zbin_over_quant are 1/128 of Q bin size
        oq_factor = 1.0 + ((double)0.0015625 * cpi->zbin_over_quant);
280
        cpi->RDMULT = (int)((double)cpi->RDMULT * oq_factor * oq_factor);
281
    }
John Koleszar's avatar
John Koleszar committed
282

Paul Wilkins's avatar
Paul Wilkins committed
283
    if (cpi->pass == 2 && (cpi->common.frame_type != KEY_FRAME))
John Koleszar's avatar
John Koleszar committed
284
    {
285
        if (cpi->twopass.next_iiratio > 31)
Paul Wilkins's avatar
Paul Wilkins committed
286
            cpi->RDMULT += (cpi->RDMULT * rd_iifactor[31]) >> 4;
John Koleszar's avatar
John Koleszar committed
287
        else
288
289
            cpi->RDMULT +=
                (cpi->RDMULT * rd_iifactor[cpi->twopass.next_iiratio]) >> 4;
John Koleszar's avatar
John Koleszar committed
290
291
    }

292
293
    if (cpi->RDMULT < 7)
        cpi->RDMULT = 7;
294

295
    cpi->mb.errorperbit = (cpi->RDMULT / 110);
296
297
    cpi->mb.errorperbit += (cpi->mb.errorperbit==0);

John Koleszar's avatar
John Koleszar committed
298
299
    vp8_set_speed_features(cpi);

300
301
    q = (int)pow(vp8_dc_quant(QIndex,0)>>2, 1.25);
    q = q << 2;
302
    cpi->RDMULT = cpi->RDMULT << 4;
Paul Wilkins's avatar
Paul Wilkins committed
303

304
305
    if (q < 8)
        q = 8;
306

John Koleszar's avatar
John Koleszar committed
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
    if (cpi->RDMULT > 1000)
    {
        cpi->RDDIV = 1;
        cpi->RDMULT /= 100;

        for (i = 0; i < MAX_MODES; i++)
        {
            if (cpi->sf.thresh_mult[i] < INT_MAX)
            {
                cpi->rd_threshes[i] = cpi->sf.thresh_mult[i] * q / 100;
            }
            else
            {
                cpi->rd_threshes[i] = INT_MAX;
            }

            cpi->rd_baseline_thresh[i] = cpi->rd_threshes[i];
        }
    }
    else
    {
        cpi->RDDIV = 100;

        for (i = 0; i < MAX_MODES; i++)
        {
            if (cpi->sf.thresh_mult[i] < (INT_MAX / q))
            {
                cpi->rd_threshes[i] = cpi->sf.thresh_mult[i] * q;
            }
            else
            {
                cpi->rd_threshes[i] = INT_MAX;
            }

            cpi->rd_baseline_thresh[i] = cpi->rd_threshes[i];
        }
    }

    fill_token_costs(
        cpi->mb.token_costs,
        (const vp8_prob( *)[8][3][11]) cpi->common.fc.coef_probs
    );

350
351
352
353
354
355
#if CONFIG_T8X8
    fill_token_costs(
        cpi->mb.token_costs_8x8,
        (const vp8_prob( *)[8][3][11]) cpi->common.fc.coef_probs_8x8
    );
#endif
356
357
358
359
#if CONFIG_QIMODE
    //rough estimate for costing
    cpi->common.kf_ymode_probs_index = cpi->common.base_qindex>>4;
#endif
John Koleszar's avatar
John Koleszar committed
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
    vp8_init_mode_costs(cpi);

}

void vp8_auto_select_speed(VP8_COMP *cpi)
{
    int milliseconds_for_compress = (int)(1000000 / cpi->oxcf.frame_rate);

    milliseconds_for_compress = milliseconds_for_compress * (16 - cpi->oxcf.cpu_used) / 16;

#if 0

    if (0)
    {
        FILE *f;

        f = fopen("speed.stt", "a");
        fprintf(f, " %8ld %10ld %10ld %10ld\n",
                cpi->common.current_video_frame, cpi->Speed, milliseconds_for_compress, cpi->avg_pick_mode_time);
        fclose(f);
    }

#endif

    /*
    // this is done during parameter valid check
Johann's avatar
Johann committed
386
387
388
389
    if( cpi->oxcf.cpu_used > 16)
        cpi->oxcf.cpu_used = 16;
    if( cpi->oxcf.cpu_used < -16)
        cpi->oxcf.cpu_used = -16;
John Koleszar's avatar
John Koleszar committed
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
    */

    if (cpi->avg_pick_mode_time < milliseconds_for_compress && (cpi->avg_encode_time - cpi->avg_pick_mode_time) < milliseconds_for_compress)
    {
        if (cpi->avg_pick_mode_time == 0)
        {
            cpi->Speed = 4;
        }
        else
        {
            if (milliseconds_for_compress * 100 < cpi->avg_encode_time * 95)
            {
                cpi->Speed          += 2;
                cpi->avg_pick_mode_time = 0;
                cpi->avg_encode_time = 0;

                if (cpi->Speed > 16)
                {
                    cpi->Speed = 16;
                }
            }

412
            if (milliseconds_for_compress * 100 > cpi->avg_encode_time * auto_speed_thresh[cpi->Speed])
John Koleszar's avatar
John Koleszar committed
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
            {
                cpi->Speed          -= 1;
                cpi->avg_pick_mode_time = 0;
                cpi->avg_encode_time = 0;

                // In real-time mode, cpi->speed is in [4, 16].
                if (cpi->Speed < 4)        //if ( cpi->Speed < 0 )
                {
                    cpi->Speed = 4;        //cpi->Speed = 0;
                }
            }
        }
    }
    else
    {
        cpi->Speed += 4;

        if (cpi->Speed > 16)
            cpi->Speed = 16;


        cpi->avg_pick_mode_time = 0;
        cpi->avg_encode_time = 0;
    }
}

int vp8_block_error_c(short *coeff, short *dqcoeff)
{
    int i;
    int error = 0;

    for (i = 0; i < 16; i++)
    {
        int this_diff = coeff[i] - dqcoeff[i];
        error += this_diff * this_diff;
    }

    return error;
}

int vp8_mbblock_error_c(MACROBLOCK *mb, int dc)
{
    BLOCK  *be;
    BLOCKD *bd;
    int i, j;
    int berror, error = 0;

    for (i = 0; i < 16; i++)
    {
        be = &mb->block[i];
        bd = &mb->e_mbd.block[i];

        berror = 0;

        for (j = dc; j < 16; j++)
        {
            int this_diff = be->coeff[j] - bd->dqcoeff[j];
            berror += this_diff * this_diff;
        }

        error += berror;
    }

    return error;
}

int vp8_mbuverror_c(MACROBLOCK *mb)
{

    BLOCK  *be;
    BLOCKD *bd;


    int i;
    int error = 0;

    for (i = 16; i < 24; i++)
    {
        be = &mb->block[i];
        bd = &mb->e_mbd.block[i];

        error += vp8_block_error_c(be->coeff, bd->dqcoeff);
    }

    return error;
}

int VP8_UVSSE(MACROBLOCK *x, const vp8_variance_rtcd_vtable_t *rtcd)
{
    unsigned char *uptr, *vptr;
    unsigned char *upred_ptr = (*(x->block[16].base_src) + x->block[16].src);
    unsigned char *vpred_ptr = (*(x->block[20].base_src) + x->block[20].src);
    int uv_stride = x->block[16].src_stride;

    unsigned int sse1 = 0;
    unsigned int sse2 = 0;
509
510
    int mv_row = x->e_mbd.mode_info_context->mbmi.mv.as_mv.row;
    int mv_col = x->e_mbd.mode_info_context->mbmi.mv.as_mv.col;
John Koleszar's avatar
John Koleszar committed
511
512
513
    int offset;
    int pre_stride = x->e_mbd.block[16].pre_stride;

514
515
516
517
518
519
520
521
522
523
524
525
    if (mv_row < 0)
        mv_row -= 1;
    else
        mv_row += 1;

    if (mv_col < 0)
        mv_col -= 1;
    else
        mv_col += 1;

    mv_row /= 2;
    mv_col /= 2;
John Koleszar's avatar
John Koleszar committed
526
527
528
529
530
531
532

    offset = (mv_row >> 3) * pre_stride + (mv_col >> 3);
    uptr = x->e_mbd.pre.u_buffer + offset;
    vptr = x->e_mbd.pre.v_buffer + offset;

    if ((mv_row | mv_col) & 7)
    {
533
534
535
536
        VARIANCE_INVOKE(rtcd, subpixvar8x8)(uptr, pre_stride,
            mv_col & 7, mv_row & 7, upred_ptr, uv_stride, &sse2);
        VARIANCE_INVOKE(rtcd, subpixvar8x8)(vptr, pre_stride,
            mv_col & 7, mv_row & 7, vpred_ptr, uv_stride, &sse1);
John Koleszar's avatar
John Koleszar committed
537
538
539
540
        sse2 += sse1;
    }
    else
    {
541
542
543
544
        VARIANCE_INVOKE(rtcd, var8x8)(uptr, pre_stride,
            upred_ptr, uv_stride, &sse2);
        VARIANCE_INVOKE(rtcd, var8x8)(vptr, pre_stride,
            vpred_ptr, uv_stride, &sse1);
John Koleszar's avatar
John Koleszar committed
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
        sse2 += sse1;
    }
    return sse2;

}

static int cost_coeffs(MACROBLOCK *mb, BLOCKD *b, int type, ENTROPY_CONTEXT *a, ENTROPY_CONTEXT *l)
{
    int c = !type;              /* start at coef 0, unless Y with Y2 */
    int eob = b->eob;
    int pt ;    /* surrounding block/prev coef predictor */
    int cost = 0;
    short *qcoeff_ptr = b->qcoeff;

    VP8_COMBINEENTROPYCONTEXTS(pt, *a, *l);

# define QC( I)  ( qcoeff_ptr [vp8_default_zig_zag1d[I]] )

    for (; c < eob; c++)
    {
        int v = QC(c);
        int t = vp8_dct_value_tokens_ptr[v].Token;
        cost += mb->token_costs [type] [vp8_coef_bands[c]] [pt] [t];
        cost += vp8_dct_value_cost_ptr[v];
        pt = vp8_prev_token_class[t];
    }

# undef QC

    if (c < 16)
        cost += mb->token_costs [type] [vp8_coef_bands[c]] [pt] [DCT_EOB_TOKEN];

    pt = (c != !type); // is eob first coefficient;
    *a = *l = pt;

    return cost;
}

Scott LaVarnway's avatar
Scott LaVarnway committed
583
static int vp8_rdcost_mby(MACROBLOCK *mb)
John Koleszar's avatar
John Koleszar committed
584
585
586
587
{
    int cost = 0;
    int b;
    MACROBLOCKD *x = &mb->e_mbd;
588
589
590
591
592
593
    ENTROPY_CONTEXT_PLANES t_above, t_left;
    ENTROPY_CONTEXT *ta;
    ENTROPY_CONTEXT *tl;

    vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
    vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
John Koleszar's avatar
John Koleszar committed
594

595
596
    ta = (ENTROPY_CONTEXT *)&t_above;
    tl = (ENTROPY_CONTEXT *)&t_left;
John Koleszar's avatar
John Koleszar committed
597
598

    for (b = 0; b < 16; b++)
Scott LaVarnway's avatar
Scott LaVarnway committed
599
        cost += cost_coeffs(mb, x->block + b, PLANE_TYPE_Y_NO_DC,
600
                    ta + vp8_block2above[b], tl + vp8_block2left[b]);
John Koleszar's avatar
John Koleszar committed
601

Scott LaVarnway's avatar
Scott LaVarnway committed
602
    cost += cost_coeffs(mb, x->block + 24, PLANE_TYPE_Y2,
Scott LaVarnway's avatar
Scott LaVarnway committed
603
                ta + vp8_block2above[24], tl + vp8_block2left[24]);
John Koleszar's avatar
John Koleszar committed
604
605
606
607

    return cost;
}

608
609
610
611
612
613
614
615
616
617
618
619
620
static void macro_block_yrd( MACROBLOCK *mb,
                             int *Rate,
                             int *Distortion,
                             const vp8_encodemb_rtcd_vtable_t *rtcd)
{
    int b;
    MACROBLOCKD *const x = &mb->e_mbd;
    BLOCK   *const mb_y2 = mb->block + 24;
    BLOCKD *const x_y2  = x->block + 24;
    short *Y2DCPtr = mb_y2->src_diff;
    BLOCK *beptr;
    int d;

621
622
    ENCODEMB_INVOKE(rtcd, submby)( mb->src_diff, *(mb->block[0].base_src),
                                   mb->e_mbd.predictor, mb->block[0].src_stride );
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645

    // Fdct and building the 2nd order block
    for (beptr = mb->block; beptr < mb->block + 16; beptr += 2)
    {
        mb->vp8_short_fdct8x4(beptr->src_diff, beptr->coeff, 32);
        *Y2DCPtr++ = beptr->coeff[0];
        *Y2DCPtr++ = beptr->coeff[16];
    }

    // 2nd order fdct
    mb->short_walsh4x4(mb_y2->src_diff, mb_y2->coeff, 8);

    // Quantization
    for (b = 0; b < 16; b++)
    {
        mb->quantize_b(&mb->block[b], &mb->e_mbd.block[b]);
    }

    // DC predication and Quantization of 2nd Order block
    mb->quantize_b(mb_y2, x_y2);

    // Distortion
    d = ENCODEMB_INVOKE(rtcd, mberr)(mb, 1) << 2;
646
647

    d += ENCODEMB_INVOKE(rtcd, berr)(mb_y2->coeff, x_y2->dqcoeff)<<2;
648
649
650
651
652
653

    *Distortion = (d >> 4);

    // rate
    *Rate = vp8_rdcost_mby(mb);
}
John Koleszar's avatar
John Koleszar committed
654

655

656
static void copy_predictor(unsigned char *dst, const unsigned char *predictor)
657
{
658
659
660
661
662
663
    const unsigned int *p = (const unsigned int *)predictor;
    unsigned int *d = (unsigned int *)dst;
    d[0] = p[0];
    d[4] = p[4];
    d[8] = p[8];
    d[12] = p[12];
664
}
Yaowu Xu's avatar
Yaowu Xu committed
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687

static void copy_predictor_8x8(unsigned char *dst, const unsigned char *predictor)
{
    const unsigned int *p = (const unsigned int *)predictor;
    unsigned int *d = (unsigned int *)dst;
    d[0] = p[0];
    d[1] = p[1];
    d[4] = p[4];
    d[5] = p[5];
    d[8] = p[8];
    d[9] = p[9];
    d[12] = p[12];
    d[13] = p[13];
    d[16] = p[16];
    d[17] = p[17];
    d[20] = p[20];
    d[21] = p[21];
    d[24] = p[24];
    d[25] = p[25];
    d[28] = p[28];
    d[29] = p[29];
}

688
static int rd_pick_intra4x4block(
John Koleszar's avatar
John Koleszar committed
689
690
691
692
693
    VP8_COMP *cpi,
    MACROBLOCK *x,
    BLOCK *be,
    BLOCKD *b,
    B_PREDICTION_MODE *best_mode,
694
    unsigned int *bmode_costs,
John Koleszar's avatar
John Koleszar committed
695
696
697
698
699
700
701
702
    ENTROPY_CONTEXT *a,
    ENTROPY_CONTEXT *l,

    int *bestrate,
    int *bestratey,
    int *bestdistortion)
{
    B_PREDICTION_MODE mode;
703
    int best_rd = INT_MAX;
John Koleszar's avatar
John Koleszar committed
704
705
706
707
708
    int rate = 0;
    int distortion;

    ENTROPY_CONTEXT ta = *a, tempa = *a;
    ENTROPY_CONTEXT tl = *l, templ = *l;
709
710
711
712
713
714
715
    /*
     * The predictor buffer is a 2d buffer with a stride of 16.  Create
     * a temp buffer that meets the stride requirements, but we are only
     * interested in the left 4x4 block
     * */
    DECLARE_ALIGNED_ARRAY(16, unsigned char,  best_predictor, 16*4);
    DECLARE_ALIGNED_ARRAY(16, short, best_dqcoeff, 16);
John Koleszar's avatar
John Koleszar committed
716
717
718
719
720
721

    for (mode = B_DC_PRED; mode <= B_HU_PRED; mode++)
    {
        int this_rd;
        int ratey;

722
723
724
725
726
#if CONFIG_SUPERBLOCKS
        // Ignore modes thact need the above-right data
        if (mode==B_LD_PRED || mode==B_VL_PRED)
            continue;
#endif
727
728
        rate = bmode_costs[mode];

729
730
        RECON_INVOKE(&cpi->rtcd.common->recon, intra4x4_predict)
                     (b, mode, b->predictor);
731
732
733
        ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), subb)(be, b, 16);
        x->vp8_short_fdct4x4(be->src_diff, be->coeff, 32);
        x->quantize_b(be, b);
John Koleszar's avatar
John Koleszar committed
734
735
736
737

        tempa = ta;
        templ = tl;

Scott LaVarnway's avatar
Scott LaVarnway committed
738
        ratey = cost_coeffs(x, b, PLANE_TYPE_Y_WITH_DC, &tempa, &templ);
John Koleszar's avatar
John Koleszar committed
739
740
741
742
743
744
745
746
747
748
749
750
751
752
        rate += ratey;
        distortion = ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)(be->coeff, b->dqcoeff) >> 2;

        this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);

        if (this_rd < best_rd)
        {
            *bestrate = rate;
            *bestratey = ratey;
            *bestdistortion = distortion;
            best_rd = this_rd;
            *best_mode = mode;
            *a = tempa;
            *l = templ;
753
754
            copy_predictor(best_predictor, b->predictor);
            vpx_memcpy(best_dqcoeff, b->dqcoeff, 32);
John Koleszar's avatar
John Koleszar committed
755
756
        }
    }
Scott LaVarnway's avatar
Scott LaVarnway committed
757
    b->bmi.as_mode = (B_PREDICTION_MODE)(*best_mode);
758

759
760
    IDCT_INVOKE(IF_RTCD(&cpi->rtcd.common->idct), idct16)(best_dqcoeff, b->diff, 32);
    RECON_INVOKE(IF_RTCD(&cpi->rtcd.common->recon), recon)(best_predictor, b->diff, *(b->base_dst) + b->dst, b->dst_stride);
John Koleszar's avatar
John Koleszar committed
761

762
    return best_rd;
John Koleszar's avatar
John Koleszar committed
763
764
}

765
766
static int rd_pick_intra4x4mby_modes(VP8_COMP *cpi, MACROBLOCK *mb, int *Rate,
                                     int *rate_y, int *Distortion, int best_rd)
John Koleszar's avatar
John Koleszar committed
767
768
769
770
771
772
{
    MACROBLOCKD *const xd = &mb->e_mbd;
    int i;
    int cost = mb->mbmode_cost [xd->frame_type] [B_PRED];
    int distortion = 0;
    int tot_rate_y = 0;
773
    int64_t total_rd = 0;
774
775
776
    ENTROPY_CONTEXT_PLANES t_above, t_left;
    ENTROPY_CONTEXT *ta;
    ENTROPY_CONTEXT *tl;
777
    unsigned int *bmode_costs;
778
779
780
781
782
783

    vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
    vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));

    ta = (ENTROPY_CONTEXT *)&t_above;
    tl = (ENTROPY_CONTEXT *)&t_left;
John Koleszar's avatar
John Koleszar committed
784

785
#if !CONFIG_SUPERBLOCKS
John Koleszar's avatar
John Koleszar committed
786
    vp8_intra_prediction_down_copy(xd);
787
#endif
John Koleszar's avatar
John Koleszar committed
788

789
790
    bmode_costs = mb->inter_bmode_costs;

John Koleszar's avatar
John Koleszar committed
791
792
793
794
795
796
797
    for (i = 0; i < 16; i++)
    {
        MODE_INFO *const mic = xd->mode_info_context;
        const int mis = xd->mode_info_stride;
        B_PREDICTION_MODE UNINITIALIZED_IS_SAFE(best_mode);
        int UNINITIALIZED_IS_SAFE(r), UNINITIALIZED_IS_SAFE(ry), UNINITIALIZED_IS_SAFE(d);

798
799
        if (mb->e_mbd.frame_type == KEY_FRAME)
        {
Scott LaVarnway's avatar
Scott LaVarnway committed
800
801
            const B_PREDICTION_MODE A = above_block_mode(mic, i, mis);
            const B_PREDICTION_MODE L = left_block_mode(mic, i);
802
803
804
805

            bmode_costs  = mb->bmode_costs[A][L];
        }

806
        total_rd += rd_pick_intra4x4block(
807
            cpi, mb, mb->block + i, xd->block + i, &best_mode, bmode_costs,
808
809
            ta + vp8_block2above[i],
            tl + vp8_block2left[i], &r, &ry, &d);
John Koleszar's avatar
John Koleszar committed
810
811
812
813

        cost += r;
        distortion += d;
        tot_rate_y += ry;
Scott LaVarnway's avatar
Scott LaVarnway committed
814
815

        mic->bmi[i].as_mode = best_mode;
816

817
        if(total_rd >= (int64_t)best_rd)
818
            break;
John Koleszar's avatar
John Koleszar committed
819
820
    }

821
    if(total_rd >= (int64_t)best_rd)
822
        return INT_MAX;
823

John Koleszar's avatar
John Koleszar committed
824
825
826
827
828
829
    *Rate = cost;
    *rate_y += tot_rate_y;
    *Distortion = distortion;

    return RDCOST(mb->rdmult, mb->rddiv, cost, distortion);
}
830
831
832
833
834
835
836


static int rd_pick_intra16x16mby_mode(VP8_COMP *cpi,
                                      MACROBLOCK *x,
                                      int *Rate,
                                      int *rate_y,
                                      int *Distortion)
John Koleszar's avatar
John Koleszar committed
837
838
839
840
{
    MB_PREDICTION_MODE mode;
    MB_PREDICTION_MODE UNINITIALIZED_IS_SAFE(mode_selected);
    int rate, ratey;
841
    int distortion;
John Koleszar's avatar
John Koleszar committed
842
    int best_rd = INT_MAX;
843
    int this_rd;
John Koleszar's avatar
John Koleszar committed
844
845
846
847

    //Y Search for 16x16 intra prediction mode
    for (mode = DC_PRED; mode <= TM_PRED; mode++)
    {
848
        x->e_mbd.mode_info_context->mbmi.mode = mode;
John Koleszar's avatar
John Koleszar committed
849

850
851
        RECON_INVOKE(&cpi->common.rtcd.recon, build_intra_predictors_mby)
            (&x->e_mbd);
John Koleszar's avatar
John Koleszar committed
852

853
854
855
        macro_block_yrd(x, &ratey, &distortion, IF_RTCD(&cpi->rtcd.encodemb));
        rate = ratey + x->mbmode_cost[x->e_mbd.frame_type]
                                     [x->e_mbd.mode_info_context->mbmi.mode];
John Koleszar's avatar
John Koleszar committed
856
857
858
859
860
861
862
863
864

        this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);

        if (this_rd < best_rd)
        {
            mode_selected = mode;
            best_rd = this_rd;
            *Rate = rate;
            *rate_y = ratey;
865
            *Distortion = distortion;
John Koleszar's avatar
John Koleszar committed
866
867
868
        }
    }

869
    x->e_mbd.mode_info_context->mbmi.mode = mode_selected;
John Koleszar's avatar
John Koleszar committed
870
871
    return best_rd;
}
Yaowu Xu's avatar
Yaowu Xu committed
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
static int rd_pick_intra8x8block(
    VP8_COMP *cpi,
    MACROBLOCK *x,
    int ib,
    B_PREDICTION_MODE *best_mode,
    unsigned int *mode_costs,
    ENTROPY_CONTEXT *a,
    ENTROPY_CONTEXT *l,
    int *bestrate,
    int *bestratey,
    int *bestdistortion)
{
    MB_PREDICTION_MODE mode;
    MACROBLOCKD *xd = &x->e_mbd;
    int best_rd = INT_MAX;
    int rate = 0;
    int distortion;
    BLOCK  *be=x->block + ib;
    BLOCKD *b=x->e_mbd.block + ib;
    ENTROPY_CONTEXT ta0, ta1, besta0, besta1;
    ENTROPY_CONTEXT tl0, tl1, bestl0, bestl1;


    /*
     * The predictor buffer is a 2d buffer with a stride of 16.  Create
     * a temp buffer that meets the stride requirements, but we are only
     * interested in the left 8x8 block
     * */

    DECLARE_ALIGNED_ARRAY(16, unsigned char,  best_predictor, 16*8);
    DECLARE_ALIGNED_ARRAY(16, short, best_dqcoeff, 16*4);

    for (mode = DC_PRED; mode <= TM_PRED; mode++)
    {
        int this_rd;
        int rate_t;

        rate = mode_costs[mode];

        RECON_INVOKE(&cpi->rtcd.common->recon, intra8x8_predict)
                     (b, mode, b->predictor);

        vp8_subtract_4b_c(be, b, 16);

        x->vp8_short_fdct8x4(be->src_diff, be->coeff, 32);
        x->vp8_short_fdct8x4(be->src_diff + 64, be->coeff + 64, 32);

        x->quantize_b_pair(x->block+ib, x->block+ib+1,
                            xd->block+ib, xd->block+ib+1);
        x->quantize_b_pair(x->block+ib+4, x->block+ib+5,
                            xd->block+ib+4, xd->block+ib+5);

        distortion = ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)
            ((x->block+ib)->coeff,(xd->block+ib)->dqcoeff)>>2;
        distortion += ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)
            ((x->block+ib+1)->coeff,(xd->block+ib+1)->dqcoeff)>>2;
        distortion += ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)
            ((x->block+ib+4)->coeff,(xd->block+ib+4)->dqcoeff)>>2;
        distortion += ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)
            ((x->block+ib+5)->coeff,(xd->block+ib+5)->dqcoeff)>>2;

        ta0 = *(a + vp8_block2above[ib]);
        ta1 = *(a + vp8_block2above[ib+1]);
        tl0 = *(l + vp8_block2above[ib]);
        tl1 = *(l + vp8_block2above[ib+4]);
        rate_t = cost_coeffs(x, xd->block+ib, PLANE_TYPE_Y_WITH_DC,
            &ta0, &tl0);
        rate_t += cost_coeffs(x, xd->block+ib+1, PLANE_TYPE_Y_WITH_DC,
            &ta1, &tl0);
        rate_t += cost_coeffs(x, xd->block+ib+4, PLANE_TYPE_Y_WITH_DC,
            &ta0, &tl1);
        rate_t += cost_coeffs(x, xd->block+ib+5, PLANE_TYPE_Y_WITH_DC,
            &ta1, &tl1);
        rate += rate_t;
        this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);
        if (this_rd < best_rd)
        {
            *bestrate = rate;
            *bestratey = rate_t;
            *bestdistortion = distortion;
            besta0 = ta0;
            besta1 = ta1;
            bestl0 = tl0;
            bestl1 = tl1;
            best_rd = this_rd;
            *best_mode = mode;
            copy_predictor_8x8(best_predictor, b->predictor);
            vpx_memcpy(best_dqcoeff, b->dqcoeff, 64);
            vpx_memcpy(best_dqcoeff+32, b->dqcoeff+64, 64);
        }
    }
    b->bmi.as_mode = (*best_mode);
    vp8_encode_intra8x8 (IF_RTCD(&cpi->rtcd), x, ib);
    *(a + vp8_block2above[ib])   = besta0;
    *(a + vp8_block2above[ib+1]) = besta1;
    *(l + vp8_block2above[ib])   = bestl0;
    *(l + vp8_block2above[ib+4]) = bestl1;
    return best_rd;
}

const int vp8_i8x8_block[4]={0, 2, 8, 10};
int rd_pick_intra8x8mby_modes(VP8_COMP *cpi,
                                      MACROBLOCK *mb,
                                      int *Rate,
                                      int *rate_y,
                                      int *Distortion,
                                      int best_rd)
{
    MACROBLOCKD *const xd = &mb->e_mbd;
    int i,ib;
    int cost = mb->mbmode_cost [xd->frame_type] [I8X8_PRED];
    int distortion = 0;
    int tot_rate_y = 0;
    long long total_rd = 0;
    ENTROPY_CONTEXT_PLANES t_above, t_left;
    ENTROPY_CONTEXT *ta;
    ENTROPY_CONTEXT *tl;
    unsigned int *i8x8mode_costs;

    vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
    vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));

    ta = (ENTROPY_CONTEXT *)&t_above;
    tl = (ENTROPY_CONTEXT *)&t_left;

    i8x8mode_costs  = mb->i8x8_mode_costs;

    for (i = 0; i < 4; i++)
    {
        MODE_INFO *const mic = xd->mode_info_context;
        B_PREDICTION_MODE UNINITIALIZED_IS_SAFE(best_mode);
        int UNINITIALIZED_IS_SAFE(r), UNINITIALIZED_IS_SAFE(ry), UNINITIALIZED_IS_SAFE(d);

        ib = vp8_i8x8_block[i];
        total_rd += rd_pick_intra8x8block(
            cpi, mb, ib, &best_mode, i8x8mode_costs,
            ta, tl, &r, &ry, &d);
        cost += r;
        distortion += d;
        tot_rate_y += ry;
        mic->bmi[ib].as_mode = best_mode;
    }
    *Rate = cost;
    *rate_y += tot_rate_y;
    *Distortion = distortion;
    return RDCOST(mb->rdmult, mb->rddiv, cost, distortion);
}
John Koleszar's avatar
John Koleszar committed
1019
1020
1021
1022
1023
1024

static int rd_cost_mbuv(MACROBLOCK *mb)
{
    int b;
    int cost = 0;
    MACROBLOCKD *x = &mb->e_mbd;
1025
1026
1027
1028
1029
1030
    ENTROPY_CONTEXT_PLANES t_above, t_left;
    ENTROPY_CONTEXT *ta;
    ENTROPY_CONTEXT *tl;

    vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
    vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
John Koleszar's avatar
John Koleszar committed
1031

1032
1033
    ta = (ENTROPY_CONTEXT *)&t_above;
    tl = (ENTROPY_CONTEXT *)&t_left;
John Koleszar's avatar
John Koleszar committed
1034

Scott LaVarnway's avatar
Scott LaVarnway committed
1035
1036
    for (b = 16; b < 24; b++)
        cost += cost_coeffs(mb, x->block + b, PLANE_TYPE_UV,
1037
                    ta + vp8_block2above[b], tl + vp8_block2left[b]);
John Koleszar's avatar
John Koleszar committed
1038
1039
1040
1041
1042

    return cost;
}


1043
1044
static int rd_inter16x16_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                            int *distortion, int fullpixel)
John Koleszar's avatar
John Koleszar committed
1045
{
Paul Wilkins's avatar
Paul Wilkins committed
1046
1047
1048
1049
1050
#if CONFIG_T8X8
    int tx_type = get_seg_tx_type(&x->e_mbd,
                                  x->e_mbd.mode_info_context->mbmi.segment_id);
#endif

1051
1052
    ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
        x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor, x->src.uv_stride);
John Koleszar's avatar
John Koleszar committed
1053

1054
#if CONFIG_T8X8
Paul Wilkins's avatar
Paul Wilkins committed
1055
    if( tx_type == TX_8X8 )
1056
1057
1058
       vp8_transform_mbuv_8x8(x);
    else
#endif
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
    vp8_transform_mbuv(x);
    vp8_quantize_mbuv(x);

    *rate       = rd_cost_mbuv(x);
    *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;

    return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
}

static int rd_inter4x4_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                          int *distortion, int fullpixel)
{
    vp8_build_inter4x4_predictors_mbuv(&x->e_mbd);
    ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
        x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor, x->src.uv_stride);

    vp8_transform_mbuv(x);
    vp8_quantize_mbuv(x);
John Koleszar's avatar
John Koleszar committed
1077
1078
1079
1080

    *rate       = rd_cost_mbuv(x);
    *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;

1081
    return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
John Koleszar's avatar
John Koleszar committed
1082
1083
}

1084
static void rd_pick_intra_mbuv_mode(VP8_COMP *cpi, MACROBLOCK *x, int *rate, int *rate_tokenonly, int *distortion)
John Koleszar's avatar
John Koleszar committed
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
{
    MB_PREDICTION_MODE mode;
    MB_PREDICTION_MODE UNINITIALIZED_IS_SAFE(mode_selected);
    int best_rd = INT_MAX;
    int UNINITIALIZED_IS_SAFE(d), UNINITIALIZED_IS_SAFE(r);
    int rate_to;

    for (mode = DC_PRED; mode <= TM_PRED; mode++)
    {
        int rate;
        int distortion;
        int this_rd;

1098
        x->e_mbd.mode_info_context->mbmi.uv_mode = mode;
1099
1100
        RECON_INVOKE(&cpi->rtcd.common->recon, build_intra_predictors_mbuv)
                     (&x->e_mbd);
1101
1102
1103
1104
1105
        ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
                      x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor,
                      x->src.uv_stride);
        vp8_transform_mbuv(x);
        vp8_quantize_mbuv(x);
John Koleszar's avatar
John Koleszar committed
1106
1107

        rate_to = rd_cost_mbuv(x);
1108
        rate = rate_to + x->intra_uv_mode_cost[x->e_mbd.frame_type][x->e_mbd.mode_info_context->mbmi.uv_mode];
John Koleszar's avatar
John Koleszar committed
1109

1110
        distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
John Koleszar's avatar
John Koleszar committed
1111

1112
        this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);
John Koleszar's avatar
John Koleszar committed
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126

        if (this_rd < best_rd)
        {
            best_rd = this_rd;
            d = distortion;
            r = rate;
            *rate_tokenonly = rate_to;
            mode_selected = mode;
        }
    }

    *rate = r;
    *distortion = d;

1127
    x->e_mbd.mode_info_context->mbmi.uv_mode = mode_selected;
John Koleszar's avatar
John Koleszar committed
1128
1129
}

Yaowu Xu's avatar
Yaowu Xu committed
1130
1131
1132
int vp8_cost_mv_ref(VP8_COMMON *pc,
                    MB_PREDICTION_MODE m,
                    const int near_mv_ref_ct[4])
John Koleszar's avatar
John Koleszar committed
1133
1134
1135
{
    vp8_prob p [VP8_MVREFS-1];
    assert(NEARESTMV <= m  &&  m <= SPLITMV);
Yaowu Xu's avatar
Yaowu Xu committed
1136
    vp8_mv_ref_probs(pc, p, near_mv_ref_ct);
1137
1138
    return vp8_cost_token(vp8_mv_ref_tree, p,
                          vp8_mv_ref_encoding_array - NEARESTMV + m);
John Koleszar's avatar
John Koleszar committed
1139
1140
}

Scott LaVarnway's avatar
Scott LaVarnway committed
1141
void vp8_set_mbmode_and_mvs(MACROBLOCK *x, MB_PREDICTION_MODE mb, int_mv *mv)
John Koleszar's avatar
John Koleszar committed
1142
{
1143
    x->e_mbd.mode_info_context->mbmi.mode = mb;
Scott LaVarnway's avatar
Scott LaVarnway committed
1144
    x->e_mbd.mode_info_context->mbmi.mv.as_int = mv->as_int;
John Koleszar's avatar
John Koleszar committed
1145
1146
1147
1148
1149
1150
}

static int labels2mode(
    MACROBLOCK *x,
    int const *labelings, int which_label,
    B_PREDICTION_MODE this_mode,
Scott LaVarnway's avatar
Scott LaVarnway committed
1151
    int_mv *this_mv, int_mv *best_ref_mv,
John Koleszar's avatar
John Koleszar committed
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
    int *mvcost[2]
)
{
    MACROBLOCKD *const xd = & x->e_mbd;
    MODE_INFO *const mic = xd->mode_info_context;
    const int mis = xd->mode_info_stride;

    int cost = 0;
    int thismvcost = 0;

    /* We have to be careful retrieving previously-encoded motion vectors.
       Ones from this macroblock have to be pulled from the BLOCKD array
       as they have not yet made it to the bmi array in our MB_MODE_INFO. */

    int i = 0;

    do
    {
        BLOCKD *const d = xd->block + i;
        const int row = i >> 2,  col = i & 3;

        B_PREDICTION_MODE m;

        if (labelings[i] != which_label)
            continue;

        if (col  &&  labelings[i] == labelings[i-1])
            m = LEFT4X4;
        else if (row  &&  labelings[i] == labelings[i-4])
            m = ABOVE4X4;
        else
        {
            // the only time we should do costing for new motion vector or mode
            // is when we are on a new label  (jbb May 08, 2007)
            switch (m = this_mode)
            {
            case NEW4X4 :
                thismvcost  = vp8_mv_bit_cost(this_mv, best_ref_mv, mvcost, 102);
                break;
            case LEFT4X4:
Scott LaVarnway's avatar
Scott LaVarnway committed
1192
                this_mv->as_int = col ? d[-1].bmi.mv.as_int : left_block_mv(mic, i);
John Koleszar's avatar
John Koleszar committed
1193
1194
                break;
            case ABOVE4X4:
Scott LaVarnway's avatar
Scott LaVarnway committed
1195
                this_mv->as_int = row ? d[-4].bmi.mv.as_int : above_block_mv(mic, i, mis);
John Koleszar's avatar
John Koleszar committed
1196
1197
                break;
            case ZERO4X4:
Scott LaVarnway's avatar
Scott LaVarnway committed
1198
                this_mv->as_int = 0;
John Koleszar's avatar
John Koleszar committed
1199
1200
1201
1202
1203
1204
1205
                break;
            default:
                break;
            }

            if (m == ABOVE4X4)  // replace above with left if same
            {
Scott LaVarnway's avatar
Scott LaVarnway committed
1206
                int_mv left_mv;
Scott LaVarnway's avatar
Scott LaVarnway committed
1207

Scott LaVarnway's avatar
Scott LaVarnway committed
1208
                left_mv.as_int = col ? d[-1].bmi.mv.as_int :
Scott LaVarnway's avatar
Scott LaVarnway committed
1209
                                        left_block_mv(mic, i);
John Koleszar's avatar
John Koleszar committed
1210

Scott LaVarnway's avatar
Scott LaVarnway committed
1211
                if (left_mv.as_int == this_mv->as_int)
John Koleszar's avatar
John Koleszar committed
1212
1213
1214
1215
1216
1217
                    m = LEFT4X4;
            }

            cost = x->inter_bmode_costs[ m];
        }

Scott LaVarnway's avatar
Scott LaVarnway committed
1218
        d->bmi.mv.as_int = this_mv->as_int;
John Koleszar's avatar
John Koleszar committed
1219

1220
1221
1222
        x->partition_info->bmi[i].mode = m;
        x->partition_info->bmi[i].mv.as_int = this_mv->as_int;

John Koleszar's avatar
John Koleszar committed
1223
1224
1225
1226
1227
1228
1229
    }
    while (++i < 16);

    cost += thismvcost ;
    return cost;
}

1230
1231
1232
static int rdcost_mbsegment_y(MACROBLOCK *mb, const int *labels,
                              int which_label, ENTROPY_CONTEXT *ta,
                              ENTROPY_CONTEXT *tl)
John Koleszar's avatar
John Koleszar committed
1233
1234
1235
1236
1237
1238
1239
{
    int cost = 0;
    int b;
    MACROBLOCKD *x = &mb->e_mbd;

    for (b = 0; b < 16; b++)
        if (labels[ b] == which_label)
Scott LaVarnway's avatar
Scott LaVarnway committed
1240
            cost += cost_coeffs(mb, x->block + b, PLANE_TYPE_Y_WITH_DC,
1241
1242
                                ta + vp8_block2above[b],
                                tl + vp8_block2left[b]);
John Koleszar's avatar
John Koleszar committed
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261

    return cost;

}
static unsigned int vp8_encode_inter_mb_segment(MACROBLOCK *x, int const *labels, int which_label, const vp8_encodemb_rtcd_vtable_t *rtcd)
{
    int i;
    unsigned int distortion = 0;

    for (i = 0; i < 16; i++)
    {
        if (labels[i] == which_label)
        {
            BLOCKD *bd = &x->e_mbd.block[i];
            BLOCK *be = &x->block[i];


            vp8_build_inter_predictors_b(bd, 16, x->e_mbd.subpixel_predict);
            ENCODEMB_INVOKE(rtcd, subb)(be, bd, 16);
Yaowu Xu's avatar
Yaowu Xu committed
1262
            x->vp8_short_fdct4x4(be->src_diff, be->coeff, 32);
John Koleszar's avatar
John Koleszar committed
1263
1264
1265

            // set to 0 no way to account for 2nd order DC so discount
            //be->coeff[0] = 0;
1266
            x->quantize_b(be, bd);
John Koleszar's avatar
John Koleszar committed
1267
1268
1269
1270
1271
1272
1273
1274

            distortion += ENCODEMB_INVOKE(rtcd, berr)(be->coeff, bd->dqcoeff);
        }
    }

    return distortion;
}

1275

1276
1277
1278
1279
static const unsigned int segmentation_to_sseshift[4] = {3, 3, 2, 0};


typedef struct
John Koleszar's avatar
John Koleszar committed
1280
{
Scott LaVarnway's avatar
Scott LaVarnway committed
1281
1282
  int_mv *ref_mv;
  int_mv mvp;
1283
1284
1285
1286
1287
1288
1289

  int segment_rd;
  int segment_num;
  int r;
  int d;
  int segment_yrate;
  B_PREDICTION_MODE modes[16];
1290
  int_mv mvs[16];
1291
1292
1293
1294
1295
  unsigned char eobs[16];

  int mvthresh;
  int *mdcounts;

Scott LaVarnway's avatar
Scott LaVarnway committed
1296
  int_mv sv_mvp[4];     // save 4 mvp from 8x8
1297
1298
  int sv_istep[2];  // save 2 initial step_param for 16x8/8x16

1299
1300
1301
} BEST_SEG_INFO;


1302
1303
static void rd_check_segment(VP8_COMP *cpi, MACROBLOCK *x,
                             BEST_SEG_INFO *bsi, unsigned int segmentation)
1304
1305
{
    int i;
John Koleszar's avatar
John Koleszar committed
1306
1307
1308
    int const *labels;
    int br = 0;
    int bd = 0;
1309
    B_PREDICTION_MODE this_mode;
John Koleszar's avatar
John Koleszar committed
1310

1311

1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
    int label_count;
    int this_segment_rd = 0;
    int label_mv_thresh;
    int rate = 0;
    int sbr = 0;
    int sbd = 0;
    int segmentyrate = 0;

    vp8_variance_fn_ptr_t *v_fn_ptr;

    ENTROPY_CONTEXT_PLANES t_above, t_left;
    ENTROPY_CONTEXT *ta;
    ENTROPY_CONTEXT *tl;
    ENTROPY_CONTEXT_PLANES t_above_b, t_left_b;
    ENTROPY_CONTEXT *ta_b;
    ENTROPY_CONTEXT *tl_b;
John Koleszar's avatar
John Koleszar committed
1328

1329
1330
    vpx_memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
    vpx_memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
John Koleszar's avatar
John Koleszar committed
1331

1332
1333
1334
1335
1336
1337
1338
    ta = (ENTROPY_CONTEXT *)&t_above;
    tl = (ENTROPY_CONTEXT *)&t_left;
    ta_b = (ENTROPY_CONTEXT *)&t_above_b;
    tl_b = (ENTROPY_CONTEXT *)&t_left_b;

    br = 0;
    bd = 0;
1339

1340
1341
1342
    v_fn_ptr = &cpi->fn_ptr[segmentation];
    labels = vp8_mbsplits[segmentation];
    label_count = vp8_mbsplit_count[segmentation];
1343

1344
1345
1346
1347
1348
1349
1350
1351
    // 64 makes this threshold really big effectively
    // making it so that we very rarely check mvs on
    // segments.   setting this to 1 would make mv thresh
    // roughly equal to what it is for macroblocks
    label_mv_thresh = 1 * bsi->mvthresh / label_count ;

    // Segmentation method overheads
    rate = vp8_cost_token(vp8_mbsplit_tree, vp8_mbsplit_probs, vp8_mbsplit_encodings + segmentation);
Yaowu Xu's avatar
Yaowu Xu committed
1352
    rate += vp8_cost_mv_ref(&cpi->common, SPLITMV, bsi->mdcounts);
1353
    this_segment_rd += RDCOST(x->rdmult, x->rddiv, rate, 0);
1354
1355
1356
    br += rate;

    for (i = 0; i < label_count; i++)
John Koleszar's avatar
John Koleszar committed
1357
    {
Scott LaVarnway's avatar
Scott LaVarnway committed
1358
        int_mv mode_mv[B_MODE_COUNT];
1359
1360
1361
        int best_label_rd = INT_MAX;
        B_PREDICTION_MODE mode_selected = ZERO4X4;
        int bestlabelyrate = 0;
John Koleszar's avatar
John Koleszar committed
1362

1363
1364
1365
1366
1367
1368
1369
1370
1371
        // search for the best motion vector on this segment
        for (this_mode = LEFT4X4; this_mode <= NEW4X4 ; this_mode ++)
        {
            int this_rd;
            int distortion;
            int labelyrate;
            ENTROPY_CONTEXT_PLANES t_above_s, t_left_s;
            ENTROPY_CONTEXT *ta_s;
            ENTROPY_CONTEXT *tl_s;
John Koleszar's avatar
John Koleszar committed
1372

1373
1374
            vpx_memcpy(&t_above_s, &t_above, sizeof(ENTROPY_CONTEXT_PLANES));
            vpx_memcpy(&t_left_s, &t_left, sizeof(ENTROPY_CONTEXT_PLANES));
John Koleszar's avatar
John Koleszar committed
1375

1376
1377
            ta_s = (ENTROPY_CONTEXT *)&t_above_s;
            tl_s = (ENTROPY_CONTEXT *)&t_left_s;
John Koleszar's avatar
John Koleszar committed
1378

1379
            if (this_mode == NEW4X4)
John Koleszar's avatar
John Koleszar committed
1380
            {
1381
                int sseshift;
John Koleszar's avatar
John Koleszar committed
1382
                int num00;
1383
                int step_param = 0;
1384
                int further_steps;
1385
1386
1387
                int n;
                int thissme;
                int bestsme = INT_MAX;
Scott LaVarnway's avatar
Scott LaVarnway committed
1388
                int_mv  temp_mv;
1389
1390
                BLOCK *c;
                BLOCKD *e;
1391

1392
1393
1394
                // Is the best so far sufficiently good that we cant justify doing and new motion search.
                if (best_label_rd < label_mv_thresh)
                    break;
John Koleszar's avatar
John Koleszar committed
1395

1396
1397
1398
1399
                if(cpi->compressor_speed)
                {
                    if (segmentation == BLOCK_8X16 || segmentation == BLOCK_16X8)
                    {
Scott LaVarnway's avatar
Scott LaVarnway committed
1400
1401
1402
                        bsi->mvp.as_int = bsi->sv_mvp[i].as_int;
                        if (i==1 && segmentation == BLOCK_16X8)
                          bsi->mvp.as_int = bsi->sv_mvp[2].as_int;
1403
1404
1405
1406
1407
1408
1409

                        step_param = bsi->sv_istep[i];
                    }

                    // use previous block's result as next block's MV predictor.
                    if (segmentation == BLOCK_4X4 && i>0)
                    {
Scott LaVarnway's avatar
Scott LaVarnway committed
1410
1411
1412
                        bsi->mvp.as_int = x->e_mbd.block[i-1].bmi.mv.as_int;
                        if (i==4 || i==8 || i==12)
                            bsi->mvp.as_int = x->e_mbd.block[i-4].bmi.mv.as_int;
1413
1414
1415
1416
1417
1418
                        step_param = 2;
                    }
                }

                further_steps = (MAX_MVSEARCH_STEPS - 1) - step_param;

John Koleszar's avatar
John Koleszar committed
1419
                {
1420
                    int sadpb = x->sadperbit4;
1421
1422
1423
1424
                    int_mv mvp_full;

                    mvp_full.as_mv.row = bsi->mvp.as_mv.row >>3;
                    mvp_full.as_mv.col = bsi->mvp.as_mv.col >>3;
John Koleszar's avatar
John Koleszar committed
1425

1426
                    // find first label
1427
                    n = vp8_mbsplit_offset[segmentation][i];
John Koleszar's avatar
John Koleszar committed
1428

1429
1430
                    c = &x->block[n];
                    e = &x->e_mbd.block[n];
John Koleszar's avatar
John Koleszar committed
1431

1432
                    {
1433
                        bestsme = cpi->diamond_search_sad(x, c, e, &mvp_full,
1434
1435
1436
                                                &mode_mv[NEW4X4], step_param,
                                                sadpb, &num00, v_fn_ptr,
                                                x->mvcost, bsi->ref_mv);
John Koleszar's avatar
John Koleszar committed
1437

1438
1439
                        n = num00;
                        num00 = 0;
John Koleszar's avatar
John Koleszar committed
1440

1441
                        while (n < further_steps)
John Koleszar's avatar
John Koleszar committed
1442
                        {
1443
                            n++;
John Koleszar's avatar
John Koleszar committed
1444

1445
1446
                            if (num00)
                                num00--;
John Koleszar's avatar
John Koleszar committed
1447
1448
                            else
                            {
1449
                                thissme = cpi->diamond_search_sad(x, c, e,
1450
                                                    &mvp_full, &temp_mv,
1451
1452
1453
                                                    step_param + n, sadpb,
                                                    &num00, v_fn_ptr,
                                                    x->mvcost, bsi->ref_mv);
1454
1455
1456
1457

                                if (thissme < bestsme)
                                {
                                    bestsme = thissme;
Scott LaVarnway's avatar
Scott LaVarnway committed
1458
                                    mode_mv[NEW4X4].as_int = temp_mv.as_int;
1459
                                }
John Koleszar's avatar
John Koleszar committed
1460
1461
1462
1463
                            }
                        }
                    }

1464
1465
1466
1467
                    sseshift = segmentation_to_sseshift[segmentation];

                    // Should we do a full search (best quality only)
                    if ((cpi->compressor_speed == 0) && (bestsme >> sseshift) > 4000)
John Koleszar's avatar
John Koleszar committed
1468
                    {
1469
1470
1471
                        /* Check if mvp_full is within the range. */
                        vp8_clamp_mv(&mvp_full, x->mv_col_min, x->mv_col_max, x->mv_row_min, x->mv_row_max);

1472
                        thissme = cpi->full_search_sad(x, c, e, &mvp_full,
1473
1474
                                                       sadpb, 16, v_fn_ptr,
                                                       x->mvcost, bsi->ref_mv);
1475
1476
1477
1478

                        if (thissme < bestsme)
                        {
                            bestsme = thissme;
Scott LaVarnway's avatar
Scott LaVarnway committed
1479
                            mode_mv[NEW4X4].as_int = e->bmi.mv.as_int;
1480
                        }
John Koleszar's avatar
John Koleszar committed
1481
                        else
1482
1483
                        {
                            // The full search result is actually worse so re-instate the previous best vector
Scott LaVarnway's avatar
Scott LaVarnway committed
1484
                            e->bmi.mv.as_int = mode_mv[NEW4X4].as_int;