common.h 23.6 KB
Newer Older
Laurent Aimar's avatar
Laurent Aimar committed
1 2 3
/*****************************************************************************
 * common.h: h264 encoder
 *****************************************************************************
4
 * Copyright (C) 2003-2008 x264 project
Laurent Aimar's avatar
Laurent Aimar committed
5 6
 *
 * Authors: Laurent Aimar <fenrir@via.ecp.fr>
7
 *          Loren Merritt <lorenm@u.washington.edu>
Laurent Aimar's avatar
Laurent Aimar committed
8 9 10 11 12 13 14 15 16 17 18 19 20
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program 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 General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
21
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02111, USA.
Laurent Aimar's avatar
Laurent Aimar committed
22 23
 *****************************************************************************/

24 25
#ifndef X264_COMMON_H
#define X264_COMMON_H
Laurent Aimar's avatar
Laurent Aimar committed
26

Loren Merritt's avatar
Loren Merritt committed
27 28 29 30 31 32 33 34 35
/****************************************************************************
 * Macros
 ****************************************************************************/
#define X264_MIN(a,b) ( (a)<(b) ? (a) : (b) )
#define X264_MAX(a,b) ( (a)>(b) ? (a) : (b) )
#define X264_MIN3(a,b,c) X264_MIN((a),X264_MIN((b),(c)))
#define X264_MAX3(a,b,c) X264_MAX((a),X264_MAX((b),(c)))
#define X264_MIN4(a,b,c,d) X264_MIN((a),X264_MIN3((b),(c),(d)))
#define X264_MAX4(a,b,c,d) X264_MAX((a),X264_MAX3((b),(c),(d)))
36
#define XCHG(type,a,b) do{ type t = a; a = b; b = t; } while(0)
Lamont Alston's avatar
Lamont Alston committed
37
#define IS_DISPOSABLE(type) ( type == X264_TYPE_B )
Loren Merritt's avatar
Loren Merritt committed
38 39
#define FIX8(f) ((int)(f*(1<<8)+.5))

40
#define CHECKED_MALLOC( var, size )\
41
do {\
42 43 44
    var = x264_malloc( size );\
    if( !var )\
        goto fail;\
45 46 47 48 49 50
} while( 0 )
#define CHECKED_MALLOCZERO( var, size )\
do {\
    CHECKED_MALLOC( var, size );\
    memset( var, 0, size );\
} while( 0 )
51

52
#define X264_BFRAME_MAX 16
53
#define X264_THREAD_MAX 128
54
#define X264_PCM_COST (386*8)
Fiona Glaser's avatar
Fiona Glaser committed
55
#define X264_LOOKAHEAD_MAX 250
56 57
// arbitrary, but low because SATD scores are 1/4 normal
#define X264_LOOKAHEAD_QP 12
Loren Merritt's avatar
Loren Merritt committed
58 59 60 61

// number of pixels (per thread) in progress at any given time.
// 16 for the macroblock in progress + 3 for deblocking + 3 for motion compensation filter + 2 for extra safety
#define X264_THREAD_HEIGHT 24
62

Dylan Yudaken's avatar
Dylan Yudaken committed
63 64 65 66 67 68 69
/* WEIGHTP_FAKE is set when mb_tree & psy are enabled, but normal weightp is disabled
 * (such as in baseline). It checks for fades in lookahead and adjusts qp accordingly
 * to increase quality. Defined as (-1) so that if(i_weighted_pred > 0) is true only when
 * real weights are being used. */

#define X264_WEIGHTP_FAKE (-1)

Loren Merritt's avatar
Loren Merritt committed
70 71 72
/****************************************************************************
 * Includes
 ****************************************************************************/
73
#include <sys/stat.h>
74
#include "osdep.h"
75
#include <stdarg.h>
76
#include <stddef.h>
77 78 79
#include <stdlib.h>
#include <string.h>
#include <assert.h>
Fiona Glaser's avatar
Fiona Glaser committed
80
#include <limits.h>
Fiona Glaser's avatar
Fiona Glaser committed
81

82
/* Unions for type-punning.
Fiona Glaser's avatar
Fiona Glaser committed
83 84 85
 * Mn: load or store n bits, aligned, native-endian
 * CPn: copy n bits, aligned, native-endian
 * we don't use memcpy for CPn because memcpy's args aren't assumed to be aligned */
86 87 88
typedef union { uint16_t i; uint8_t  c[2]; } MAY_ALIAS x264_union16_t;
typedef union { uint32_t i; uint16_t b[2]; uint8_t  c[4]; } MAY_ALIAS x264_union32_t;
typedef union { uint64_t i; uint32_t a[2]; uint16_t b[4]; uint8_t c[8]; } MAY_ALIAS x264_union64_t;
Fiona Glaser's avatar
Fiona Glaser committed
89 90 91 92 93 94 95
#define M16(src) (((x264_union16_t*)(src))->i)
#define M32(src) (((x264_union32_t*)(src))->i)
#define M64(src) (((x264_union64_t*)(src))->i)
#define CP16(dst,src) M16(dst) = M16(src)
#define CP32(dst,src) M32(dst) = M32(src)
#define CP64(dst,src) M64(dst) = M64(src)

96
#include "x264.h"
Laurent Aimar's avatar
Laurent Aimar committed
97 98 99 100 101 102 103 104
#include "bs.h"
#include "set.h"
#include "predict.h"
#include "pixel.h"
#include "mc.h"
#include "frame.h"
#include "dct.h"
#include "cabac.h"
105
#include "quant.h"
106
#include "config.h"
Laurent Aimar's avatar
Laurent Aimar committed
107

108
/****************************************************************************
109
 * General functions
110
 ****************************************************************************/
Laurent Aimar's avatar
Laurent Aimar committed
111
/* x264_malloc : will do or emulate a memalign
Loren Merritt's avatar
Loren Merritt committed
112
 * you have to use x264_free for buffers allocated with x264_malloc */
Laurent Aimar's avatar
Laurent Aimar committed
113 114 115
void *x264_malloc( int );
void  x264_free( void * );

116 117 118
/* x264_slurp_file: malloc space for the whole file and read it */
char *x264_slurp_file( const char *filename );

Laurent Aimar's avatar
Laurent Aimar committed
119 120 121
/* mdate: return the current date in microsecond */
int64_t x264_mdate( void );

122 123 124 125
/* x264_param2string: return a (malloced) string containing most of
 * the encoding options */
char *x264_param2string( x264_param_t *p, int b_res );

126 127
int x264_nal_encode( uint8_t *dst, int b_annexb, x264_nal_t *nal );

128 129 130
/* log */
void x264_log( x264_t *h, int i_level, const char *psz_fmt, ... );

Loren Merritt's avatar
Loren Merritt committed
131
void x264_reduce_fraction( int *n, int *d );
132
void x264_init_vlc_tables();
Loren Merritt's avatar
Loren Merritt committed
133

Loren Merritt's avatar
Loren Merritt committed
134 135 136 137 138
static inline uint8_t x264_clip_uint8( int x )
{
    return x&(~255) ? (-x)>>31 : x;
}

Laurent Aimar's avatar
Laurent Aimar committed
139 140
static inline int x264_clip3( int v, int i_min, int i_max )
{
Loren Merritt's avatar
Loren Merritt committed
141
    return ( (v < i_min) ? i_min : (v > i_max) ? i_max : v );
Laurent Aimar's avatar
Laurent Aimar committed
142 143
}

144
static inline double x264_clip3f( double v, double f_min, double f_max )
Loren Merritt's avatar
Loren Merritt committed
145 146 147
{
    return ( (v < f_min) ? f_min : (v > f_max) ? f_max : v );
}
148

149 150
static inline int x264_median( int a, int b, int c )
{
Loren Merritt's avatar
Loren Merritt committed
151 152 153 154 155 156
    int t = (a-b)&((a-b)>>31);
    a -= t;
    b += t;
    b -= (b-c)&((b-c)>>31);
    b += (a-b)&((a-b)>>31);
    return b;
157 158
}

Loren Merritt's avatar
Loren Merritt committed
159 160 161 162 163 164
static inline void x264_median_mv( int16_t *dst, int16_t *a, int16_t *b, int16_t *c )
{
    dst[0] = x264_median( a[0], b[0], c[0] );
    dst[1] = x264_median( a[1], b[1], c[1] );
}

165
static inline int x264_predictor_difference( int16_t (*mvc)[2], intptr_t i_mvc )
Fiona Glaser's avatar
Fiona Glaser committed
166 167 168 169 170 171 172 173 174 175
{
    int sum = 0, i;
    for( i = 0; i < i_mvc-1; i++ )
    {
        sum += abs( mvc[i][0] - mvc[i+1][0] )
             + abs( mvc[i][1] - mvc[i+1][1] );
    }
    return sum;
}

Fiona Glaser's avatar
Fiona Glaser committed
176 177 178 179 180 181 182 183 184
static inline uint32_t x264_cabac_amvd_sum( int16_t *mvdleft, int16_t *mvdtop )
{
    int amvd0 = abs(mvdleft[0]) + abs(mvdtop[0]);
    int amvd1 = abs(mvdleft[1]) + abs(mvdtop[1]);
    amvd0 = (amvd0 > 2) + (amvd0 > 32);
    amvd1 = (amvd1 > 2) + (amvd1 > 32);
    return amvd0 + (amvd1<<16);
}

185 186 187
extern const uint8_t x264_exp2_lut[64];
extern const float x264_log2_lut[128];
extern const float x264_log2_lz_lut[32];
Fiona Glaser's avatar
Fiona Glaser committed
188

Fiona Glaser's avatar
Fiona Glaser committed
189 190
/* Not a general-purpose function; multiplies input by -1/6 to convert
 * qp to qscale. */
Fiona Glaser's avatar
Fiona Glaser committed
191 192
static ALWAYS_INLINE int x264_exp2fix8( float x )
{
Anton Mitrofanov's avatar
Anton Mitrofanov committed
193 194 195
    int i = x*(-64.f/6.f) + 512.5f;
    if( i < 0 ) return 0;
    if( i > 1023 ) return 0xffff;
196
    return (x264_exp2_lut[i&63]+256) << (i>>6) >> 8;
Fiona Glaser's avatar
Fiona Glaser committed
197 198 199 200 201
}

static ALWAYS_INLINE float x264_log2( uint32_t x )
{
    int lz = x264_clz( x );
202
    return x264_log2_lut[(x<<lz>>24)&0x7f] + x264_log2_lz_lut[lz];
Fiona Glaser's avatar
Fiona Glaser committed
203 204
}

205 206 207
/****************************************************************************
 *
 ****************************************************************************/
Laurent Aimar's avatar
Laurent Aimar committed
208 209 210 211 212 213 214 215 216
enum slice_type_e
{
    SLICE_TYPE_P  = 0,
    SLICE_TYPE_B  = 1,
    SLICE_TYPE_I  = 2,
    SLICE_TYPE_SP = 3,
    SLICE_TYPE_SI = 4
};

Loren Merritt's avatar
Loren Merritt committed
217 218
static const char slice_type_to_char[] = { 'P', 'B', 'I', 'S', 'S' };

Laurent Aimar's avatar
Laurent Aimar committed
219 220 221 222 223 224 225
typedef struct
{
    x264_sps_t *sps;
    x264_pps_t *pps;

    int i_type;
    int i_first_mb;
Loren Merritt's avatar
Loren Merritt committed
226
    int i_last_mb;
Laurent Aimar's avatar
Laurent Aimar committed
227 228 229 230 231

    int i_pps_id;

    int i_frame_num;

232
    int b_mbaff;
Laurent Aimar's avatar
Laurent Aimar committed
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249
    int b_field_pic;
    int b_bottom_field;

    int i_idr_pic_id;   /* -1 if nal_type != 5 */

    int i_poc_lsb;
    int i_delta_poc_bottom;

    int i_delta_poc[2];
    int i_redundant_pic_cnt;

    int b_direct_spatial_mv_pred;

    int b_num_ref_idx_override;
    int i_num_ref_idx_l0_active;
    int i_num_ref_idx_l1_active;

250 251
    int b_ref_pic_list_reordering_l0;
    int b_ref_pic_list_reordering_l1;
Lamont Alston's avatar
Lamont Alston committed
252 253
    struct
    {
254 255 256 257
        int idc;
        int arg;
    } ref_pic_list_order[2][16];

Dylan Yudaken's avatar
Dylan Yudaken committed
258 259 260
    /* P-frame weighting */
    x264_weight_t weight[16][3];

Lamont Alston's avatar
Lamont Alston committed
261 262 263 264 265 266 267 268
    int i_mmco_remove_from_end;
    int i_mmco_command_count;
    struct /* struct for future expansion */
    {
        int i_difference_of_pic_nums;
        int i_poc;
    } mmco[16];

Laurent Aimar's avatar
Laurent Aimar committed
269 270
    int i_cabac_init_idc;

271
    int i_qp;
Laurent Aimar's avatar
Laurent Aimar committed
272 273 274 275 276 277 278 279 280 281 282
    int i_qp_delta;
    int b_sp_for_swidth;
    int i_qs_delta;

    /* deblocking filter */
    int i_disable_deblocking_filter_idc;
    int i_alpha_c0_offset;
    int i_beta_offset;

} x264_slice_header_t;

Steven Walters's avatar
Steven Walters committed
283 284
typedef struct x264_lookahead_t
{
285
    volatile uint8_t              b_exit_thread;
Steven Walters's avatar
Steven Walters committed
286 287 288 289 290 291 292 293 294 295
    uint8_t                       b_thread_active;
    uint8_t                       b_analyse_keyframe;
    int                           i_last_idr;
    int                           i_slicetype_length;
    x264_frame_t                  *last_nonb;
    x264_synch_frame_list_t       ifbuf;
    x264_synch_frame_list_t       next;
    x264_synch_frame_list_t       ofbuf;
} x264_lookahead_t;

Laurent Aimar's avatar
Laurent Aimar committed
296 297 298 299 300
/* From ffmpeg
 */
#define X264_SCAN8_SIZE (6*8)
#define X264_SCAN8_0 (4+1*8)

301
static const int x264_scan8[16+2*4+3] =
Laurent Aimar's avatar
Laurent Aimar committed
302 303 304 305 306 307 308 309 310 311 312 313 314 315
{
    /* Luma */
    4+1*8, 5+1*8, 4+2*8, 5+2*8,
    6+1*8, 7+1*8, 6+2*8, 7+2*8,
    4+3*8, 5+3*8, 4+4*8, 5+4*8,
    6+3*8, 7+3*8, 6+4*8, 7+4*8,

    /* Cb */
    1+1*8, 2+1*8,
    1+2*8, 2+2*8,

    /* Cr */
    1+4*8, 2+4*8,
    1+5*8, 2+5*8,
316 317 318 319 320 321

    /* Luma DC */
    4+5*8,

    /* Chroma DC */
    5+5*8, 6+5*8
Laurent Aimar's avatar
Laurent Aimar committed
322 323 324 325 326 327 328 329
};
/*
   0 1 2 3 4 5 6 7
 0
 1   B B   L L L L
 2   B B   L L L L
 3         L L L L
 4   R R   L L L L
330
 5   R R   DyDuDv
Laurent Aimar's avatar
Laurent Aimar committed
331 332 333 334 335 336 337 338 339
*/

typedef struct x264_ratecontrol_t   x264_ratecontrol_t;

struct x264_t
{
    /* encoder parameters */
    x264_param_t    param;

Steven Walters's avatar
Steven Walters committed
340
    x264_t          *thread[X264_THREAD_MAX+1];
341
    x264_pthread_t  thread_handle;
Loren Merritt's avatar
Loren Merritt committed
342 343
    int             b_thread_active;
    int             i_thread_phase; /* which thread to use for the next frame */
Loren Merritt's avatar
Loren Merritt committed
344

Laurent Aimar's avatar
Laurent Aimar committed
345 346 347 348
    /* bitstream output */
    struct
    {
        int         i_nal;
Fiona Glaser's avatar
Fiona Glaser committed
349 350
        int         i_nals_allocated;
        x264_nal_t  *nal;
Laurent Aimar's avatar
Laurent Aimar committed
351 352 353 354 355
        int         i_bitstream;    /* size of p_bitstream */
        uint8_t     *p_bitstream;   /* will hold data for all nal */
        bs_t        bs;
    } out;

356 357 358
    uint8_t *nal_buffer;
    int      nal_buffer_size;

Loren Merritt's avatar
Loren Merritt committed
359 360
    /**** thread synchronization starts here ****/

Laurent Aimar's avatar
Laurent Aimar committed
361 362 363 364 365 366 367
    /* frame number/poc */
    int             i_frame;

    int             i_frame_offset; /* decoding only */
    int             i_frame_num;    /* decoding only */
    int             i_poc_msb;      /* decoding only */
    int             i_poc_lsb;      /* decoding only */
368
    int             i_poc;          /* decoding only */
Laurent Aimar's avatar
Laurent Aimar committed
369

Loren Merritt's avatar
Loren Merritt committed
370 371 372 373
    int             i_thread_num;   /* threads only */
    int             i_nal_type;     /* threads only */
    int             i_nal_ref_idc;  /* threads only */

Laurent Aimar's avatar
Laurent Aimar committed
374
    /* We use only one SPS and one PPS */
375
    x264_sps_t      sps_array[1];
Laurent Aimar's avatar
Laurent Aimar committed
376
    x264_sps_t      *sps;
377
    x264_pps_t      pps_array[1];
Laurent Aimar's avatar
Laurent Aimar committed
378 379 380
    x264_pps_t      *pps;
    int             i_idr_pic_id;

Loren Merritt's avatar
Loren Merritt committed
381 382 383
    /* quantization matrix for decoding, [cqm][qp%6][coef] */
    int             (*dequant4_mf[4])[16];   /* [4][6][16] */
    int             (*dequant8_mf[2])[64];   /* [2][6][64] */
Loren Merritt's avatar
Loren Merritt committed
384
    /* quantization matrix for trellis, [cqm][qp][coef] */
385 386
    int             (*unquant4_mf[4])[16];   /* [4][52][16] */
    int             (*unquant8_mf[2])[64];   /* [2][52][64] */
Loren Merritt's avatar
Loren Merritt committed
387
    /* quantization matrix for deadzone */
Loren Merritt's avatar
Loren Merritt committed
388 389 390 391
    uint16_t        (*quant4_mf[4])[16];     /* [4][52][16] */
    uint16_t        (*quant8_mf[2])[64];     /* [2][52][64] */
    uint16_t        (*quant4_bias[4])[16];   /* [4][52][16] */
    uint16_t        (*quant8_bias[2])[64];   /* [2][52][64] */
Loren Merritt's avatar
Loren Merritt committed
392

393 394 395 396 397 398
    /* mv/ref cost arrays.  Indexed by lambda instead of
     * qp because, due to rounding, some quantizers share
     * lambdas.  This saves memory. */
    uint16_t *cost_mv[92];
    uint16_t *cost_mv_fpel[92][4];

399 400
    const uint8_t   *chroma_qp_table; /* includes both the nonlinear luma->chroma mapping and chroma_qp_offset */

401 402
    ALIGNED_16( uint32_t nr_residual_sum[2][64] );
    ALIGNED_16( uint16_t nr_offset[2][64] );
403
    uint32_t        nr_count[2];
404

Laurent Aimar's avatar
Laurent Aimar committed
405 406 407 408 409 410 411 412
    /* Slice header */
    x264_slice_header_t sh;

    /* cabac context */
    x264_cabac_t    cabac;

    struct
    {
413
        /* Frames to be encoded (whose types have been decided) */
Steven Walters's avatar
Steven Walters committed
414 415 416
        x264_frame_t **current;
        /* Unused frames: 0 = fenc, 1 = fdec */
        x264_frame_t **unused[2];
Laurent Aimar's avatar
Laurent Aimar committed
417

Dylan Yudaken's avatar
Dylan Yudaken committed
418 419 420
        /* Unused blank frames (for duplicates) */
        x264_frame_t **blank_unused;

421 422
        /* frames used for reference + sentinels */
        x264_frame_t *reference[16+2];
Laurent Aimar's avatar
Laurent Aimar committed
423

424
        int i_last_idr; /* Frame number of the last IDR */
425 426

        int i_input;    /* Number of input frames already accepted */
427 428 429 430

        int i_max_dpb;  /* Number of frames allocated in the decoded picture buffer */
        int i_max_ref0;
        int i_max_ref1;
431
        int i_delay;    /* Number of frames buffered for B reordering */
432
        int b_have_lowres;  /* Whether 1/2 resolution luma planes are being used */
Loren Merritt's avatar
Loren Merritt committed
433
        int b_have_sub8x8_esa;
Laurent Aimar's avatar
Laurent Aimar committed
434 435 436 437 438 439 440 441 442 443
    } frames;

    /* current frame being encoded */
    x264_frame_t    *fenc;

    /* frame being reconstructed */
    x264_frame_t    *fdec;

    /* references lists */
    int             i_ref0;
Loren Merritt's avatar
Loren Merritt committed
444
    x264_frame_t    *fref0[16+3];     /* ref list 0 */
Laurent Aimar's avatar
Laurent Aimar committed
445
    int             i_ref1;
Loren Merritt's avatar
Loren Merritt committed
446
    x264_frame_t    *fref1[16+3];     /* ref list 1 */
447
    int             b_ref_reorder[2];
Laurent Aimar's avatar
Laurent Aimar committed
448 449 450 451 452 453



    /* Current MB DCT coeffs */
    struct
    {
454 455
        ALIGNED_16( int16_t luma16x16_dc[16] );
        ALIGNED_16( int16_t chroma_dc[2][4] );
456
        // FIXME share memory?
457 458
        ALIGNED_16( int16_t luma8x8[4][64] );
        ALIGNED_16( int16_t luma4x4[16+8][16] );
Laurent Aimar's avatar
Laurent Aimar committed
459 460 461 462 463
    } dct;

    /* MB table and cache for current frame/mb */
    struct
    {
464 465
        int     i_mb_count;                 /* number of mbs in a frame */

Laurent Aimar's avatar
Laurent Aimar committed
466 467
        /* Strides */
        int     i_mb_stride;
468 469
        int     i_b8_stride;
        int     i_b4_stride;
Laurent Aimar's avatar
Laurent Aimar committed
470 471 472 473 474

        /* Current index */
        int     i_mb_x;
        int     i_mb_y;
        int     i_mb_xy;
475 476
        int     i_b8_xy;
        int     i_b4_xy;
Loren Merritt's avatar
Loren Merritt committed
477

478
        /* Search parameters */
Loren Merritt's avatar
Loren Merritt committed
479
        int     i_me_method;
480
        int     i_subpel_refine;
Loren Merritt's avatar
Loren Merritt committed
481
        int     b_chroma_me;
Loren Merritt's avatar
Loren Merritt committed
482
        int     b_trellis;
483
        int     b_noise_reduction;
484 485
        int     i_psy_rd; /* Psy RD strength--fixed point value*/
        int     i_psy_trellis; /* Psy trellis strength--fixed point value*/
486

487 488
        int     b_interlaced;

489
        /* Allowed qpel MV range to stay within the picture + emulated edge pixels */
490 491
        int     mv_min[2];
        int     mv_max[2];
492 493 494 495
        /* Subpel MV range for motion search.
         * same mv_min/max but includes levels' i_mv_range. */
        int     mv_min_spel[2];
        int     mv_max_spel[2];
496 497 498
        /* Fullpel MV range for motion search */
        int     mv_min_fpel[2];
        int     mv_max_fpel[2];
499

Loren Merritt's avatar
Loren Merritt committed
500
        /* neighboring MBs */
Laurent Aimar's avatar
Laurent Aimar committed
501
        unsigned int i_neighbour;
502 503
        unsigned int i_neighbour8[4];       /* neighbours of each 8x8 or 4x4 block that are available */
        unsigned int i_neighbour4[16];      /* at the time the block is coded */
504
        unsigned int i_neighbour_intra;     /* for constrained intra pred */
Loren Merritt's avatar
Loren Merritt committed
505 506 507 508
        int     i_mb_type_top;
        int     i_mb_type_left;
        int     i_mb_type_topleft;
        int     i_mb_type_topright;
509 510
        int     i_mb_prev_xy;
        int     i_mb_top_xy;
Laurent Aimar's avatar
Laurent Aimar committed
511

Loren Merritt's avatar
Loren Merritt committed
512
        /**** thread synchronization ends here ****/
Loren Merritt's avatar
Loren Merritt committed
513
        /* subsequent variables are either thread-local or constant,
Loren Merritt's avatar
Loren Merritt committed
514 515
         * and won't be copied from one thread to another */

Laurent Aimar's avatar
Laurent Aimar committed
516 517 518 519
        /* mb table */
        int8_t  *type;                      /* mb type */
        int8_t  *qp;                        /* mb qp */
        int16_t *cbp;                       /* mb cbp: 0x0?: luma, 0x?0: chroma, 0x100: luma dc, 0x0200 and 0x0400: chroma dc  (all set for PCM)*/
520 521
        int8_t  (*intra4x4_pred_mode)[8];   /* intra4x4 pred mode. for non I4x4 set to I_PRED_4x4_DC(2) */
                                            /* actually has only 7 entries; set to 8 for write-combining optimizations */
Laurent Aimar's avatar
Laurent Aimar committed
522 523 524 525
        uint8_t (*non_zero_count)[16+4+4];  /* nzc. for I_PCM set to 16 */
        int8_t  *chroma_pred_mode;          /* chroma_pred_mode. cabac only. for non intra I_PRED_CHROMA_DC(0) */
        int16_t (*mv[2])[2];                /* mb mv. set to 0 for intra mb */
        int16_t (*mvd[2])[2];               /* mb mv difference with predict. set to 0 if intra. cabac only */
526
        int8_t   *ref[2];                   /* mb ref. set to -1 if non used (intra or Lx only) */
527
        int16_t (*mvr[2][32])[2];           /* 16x16 mv for each possible ref */
528
        int8_t  *skipbp;                    /* block pattern for SKIP or DIRECT (sub)mbs. B-frames + cabac only */
529
        int8_t  *mb_transform_size;         /* transform_size_8x8_flag of each mb */
Loren Merritt's avatar
Loren Merritt committed
530
        uint8_t *intra_border_backup[2][3]; /* bottom pixels of the previous mb row, used for intra prediction after the framebuffer has been deblocked */
531
        uint8_t (*nnz_backup)[16];          /* when using cavlc + 8x8dct, the deblocker uses a modified nnz */
532

Dylan Yudaken's avatar
Dylan Yudaken committed
533 534 535
         /* buffer for weighted versions of the reference frames */
        uint8_t *p_weight_buf[16];

Laurent Aimar's avatar
Laurent Aimar committed
536 537 538
        /* current value */
        int     i_type;
        int     i_partition;
539
        ALIGNED_4( uint8_t i_sub_partition[4] );
540
        int     b_transform_8x8;
Laurent Aimar's avatar
Laurent Aimar committed
541 542 543 544 545 546 547

        int     i_cbp_luma;
        int     i_cbp_chroma;

        int     i_intra16x16_pred_mode;
        int     i_chroma_pred_mode;

548 549 550 551 552
        /* skip flags for i4x4 and i8x8
         * 0 = encode as normal.
         * 1 (non-RD only) = the DCT is still in h->dct, restore fdec and skip reconstruction.
         * 2 (RD only) = the DCT has since been overwritten by RD; restore that too. */
        int i_skip_intra;
Fiona Glaser's avatar
Fiona Glaser committed
553 554
        /* skip flag for motion compensation */
        /* if we've already done MC, we don't need to do it again */
555
        int b_skip_mc;
Fiona Glaser's avatar
Fiona Glaser committed
556 557
        /* set to true if we are re-encoding a macroblock. */
        int b_reencode_mb;
558

Laurent Aimar's avatar
Laurent Aimar committed
559 560
        struct
        {
561 562 563
            /* space for p_fenc and p_fdec */
#define FENC_STRIDE 16
#define FDEC_STRIDE 32
564 565
            ALIGNED_16( uint8_t fenc_buf[24*FENC_STRIDE] );
            ALIGNED_16( uint8_t fdec_buf[27*FDEC_STRIDE] );
566

Loren Merritt's avatar
Loren Merritt committed
567
            /* i4x4 and i8x8 backup data, for skipping the encode stage when possible */
568 569 570 571
            ALIGNED_16( uint8_t i4x4_fdec_buf[16*16] );
            ALIGNED_16( uint8_t i8x8_fdec_buf[16*16] );
            ALIGNED_16( int16_t i8x8_dct_buf[3][64] );
            ALIGNED_16( int16_t i4x4_dct_buf[15][16] );
572 573 574 575
            uint32_t i4x4_nnz_buf[4];
            uint32_t i8x8_nnz_buf[4];
            int i4x4_cbp;
            int i8x8_cbp;
576

577
            /* Psy trellis DCT data */
578 579
            ALIGNED_16( int16_t fenc_dct8[4][64] );
            ALIGNED_16( int16_t fenc_dct4[16][16] );
580 581 582 583 584 585 586

            /* Psy RD SATD scores */
            int fenc_satd[4][4];
            int fenc_satd_sum;
            int fenc_sa8d[2][2];
            int fenc_sa8d_sum;

Laurent Aimar's avatar
Laurent Aimar committed
587 588
            /* pointer over mb of the frame to be compressed */
            uint8_t *p_fenc[3];
589 590
            /* pointer to the actual source frame, not a block copy */
            uint8_t *p_fenc_plane[3];
Laurent Aimar's avatar
Laurent Aimar committed
591

Loren Merritt's avatar
Loren Merritt committed
592
            /* pointer over mb of the frame to be reconstructed  */
Laurent Aimar's avatar
Laurent Aimar committed
593 594 595
            uint8_t *p_fdec[3];

            /* pointer over mb of the references */
596 597
            int i_fref[2];
            uint8_t *p_fref[2][32][4+2]; /* last: lN, lH, lV, lHV, cU, cV */
Dylan Yudaken's avatar
Dylan Yudaken committed
598
            uint8_t *p_fref_w[32];  /* weighted fullpel luma */
599
            uint16_t *p_integral[2][16];
Laurent Aimar's avatar
Laurent Aimar committed
600

601
            /* fref stride */
Laurent Aimar's avatar
Laurent Aimar committed
602 603 604 605 606 607
            int     i_stride[3];
        } pic;

        /* cache */
        struct
        {
608
            /* real intra4x4_pred_mode if I_4X4 or I_8X8, I_PRED_4x4_DC if mb available, -1 if not */
Loren Merritt's avatar
Loren Merritt committed
609
            int8_t  intra4x4_pred_mode[X264_SCAN8_SIZE];
Laurent Aimar's avatar
Laurent Aimar committed
610

Loren Merritt's avatar
Loren Merritt committed
611
            /* i_non_zero_count if available else 0x80 */
Loren Merritt's avatar
Loren Merritt committed
612
            uint8_t non_zero_count[X264_SCAN8_SIZE];
Laurent Aimar's avatar
Laurent Aimar committed
613

Loren Merritt's avatar
Loren Merritt committed
614
            /* -1 if unused, -2 if unavailable */
615
            ALIGNED_4( int8_t ref[2][X264_SCAN8_SIZE] );
Laurent Aimar's avatar
Laurent Aimar committed
616

Loren Merritt's avatar
Loren Merritt committed
617
            /* 0 if not available */
618 619
            ALIGNED_16( int16_t mv[2][X264_SCAN8_SIZE][2] );
            ALIGNED_8( int16_t mvd[2][X264_SCAN8_SIZE][2] );
620 621

            /* 1 if SKIP or DIRECT. set only for B-frames + CABAC */
622
            ALIGNED_4( int8_t skip[X264_SCAN8_SIZE] );
623

624 625 626
            ALIGNED_16( int16_t direct_mv[2][X264_SCAN8_SIZE][2] );
            ALIGNED_4( int8_t  direct_ref[2][X264_SCAN8_SIZE] );
            ALIGNED_4( int16_t pskip_mv[2] );
627

628 629
            /* number of neighbors (top and left) that used 8x8 dct */
            int     i_neighbour_transform_size;
630
            int     i_neighbour_interlaced;
631 632 633 634

            /* neighbor CBPs */
            int     i_cbp_top;
            int     i_cbp_left;
Laurent Aimar's avatar
Laurent Aimar committed
635 636 637
        } cache;

        /* */
638
        int     i_qp;       /* current qp */
639
        int     i_chroma_qp;
Laurent Aimar's avatar
Laurent Aimar committed
640 641
        int     i_last_qp;  /* last qp */
        int     i_last_dqp; /* last delta qp */
642
        int     b_variable_qp; /* whether qp is allowed to vary per macroblock */
Loren Merritt's avatar
Loren Merritt committed
643
        int     b_lossless;
Loren Merritt's avatar
Loren Merritt committed
644 645
        int     b_direct_auto_read; /* take stats for --direct auto from the 2pass log */
        int     b_direct_auto_write; /* analyse direct modes, to use and/or save */
Laurent Aimar's avatar
Laurent Aimar committed
646

Fiona Glaser's avatar
Fiona Glaser committed
647 648 649 650 651
        /* lambda values */
        int     i_trellis_lambda2[2][2]; /* [luma,chroma][inter,intra] */
        int     i_psy_rd_lambda;
        int     i_chroma_lambda2_offset;

652
        /* B_direct and weighted prediction */
Loren Merritt's avatar
Loren Merritt committed
653 654
        int16_t dist_scale_factor[16][2];
        int16_t bipred_weight[32][4];
655
        /* maps fref1[0]'s ref indices into the current list0 */
656 657
#define map_col_to_list0(col) h->mb.map_col_to_list0[col+2]
        int8_t  map_col_to_list0[18];
Laurent Aimar's avatar
Laurent Aimar committed
658 659 660 661 662 663 664 665
    } mb;

    /* rate control encoding only */
    x264_ratecontrol_t *rc;

    /* stats */
    struct
    {
666 667 668
        /* Current frame stats */
        struct
        {
669 670 671 672
            /* MV bits (MV+Ref+Block Type) */
            int i_mv_bits;
            /* Texture bits (DCT coefs) */
            int i_tex_bits;
673 674 675
            /* ? */
            int i_misc_bits;
            /* MB type counts */
676
            int i_mb_count[19];
677
            int i_mb_count_i;
Loren Merritt's avatar
Loren Merritt committed
678 679
            int i_mb_count_p;
            int i_mb_count_skip;
680
            int i_mb_count_8x8dct[2];
681 682
            int i_mb_count_ref[2][32];
            int i_mb_partition[17];
683
            int i_mb_cbp[6];
684
            int i_mb_pred_mode[3][13];
Loren Merritt's avatar
Loren Merritt committed
685 686
            /* Adaptive direct mv pred */
            int i_direct_score[2];
687 688 689
            /* Metrics */
            int64_t i_ssd[3];
            double f_ssim;
690 691
        } frame;

Loren Merritt's avatar
Loren Merritt committed
692
        /* Cumulated stats */
693

Laurent Aimar's avatar
Laurent Aimar committed
694
        /* per slice info */
Fiona Glaser's avatar
Fiona Glaser committed
695 696 697
        int     i_frame_count[5];
        int64_t i_frame_size[5];
        double  f_frame_qp[5];
698
        int     i_consecutive_bframes[X264_BFRAME_MAX+1];
699
        /* */
Loren Merritt's avatar
Loren Merritt committed
700
        int64_t i_ssd_global[5];
701 702 703 704 705
        double  f_psnr_average[5];
        double  f_psnr_mean_y[5];
        double  f_psnr_mean_u[5];
        double  f_psnr_mean_v[5];
        double  f_ssim_mean_y[5];
706
        /* */
707
        int64_t i_mb_count[5][19];
708
        int64_t i_mb_partition[2][17];
709
        int64_t i_mb_count_8x8dct[2];
710
        int64_t i_mb_count_ref[2][2][32];
711
        int64_t i_mb_cbp[6];
712
        int64_t i_mb_pred_mode[3][13];
Loren Merritt's avatar
Loren Merritt committed
713 714 715
        /* */
        int     i_direct_score[2];
        int     i_direct_frames[2];
Dylan Yudaken's avatar
Dylan Yudaken committed
716 717
        /* num p-frames weighted */
        int     i_wpred[3];
718

Laurent Aimar's avatar
Laurent Aimar committed
719 720
    } stat;

721 722
    void *scratch_buffer; /* for any temporary storage that doesn't want repeated malloc */

Loren Merritt's avatar
Loren Merritt committed
723
    /* CPU functions dependents */
Laurent Aimar's avatar
Laurent Aimar committed
724
    x264_predict_t      predict_16x16[4+3];
725 726
    x264_predict_t      predict_8x8c[4+3];
    x264_predict8x8_t   predict_8x8[9+3];
Laurent Aimar's avatar
Laurent Aimar committed
727
    x264_predict_t      predict_4x4[9+3];
728
    x264_predict_8x8_filter_t predict_8x8_filter;
Laurent Aimar's avatar
Laurent Aimar committed
729 730

    x264_pixel_function_t pixf;
731
    x264_mc_functions_t   mc;
Laurent Aimar's avatar
Laurent Aimar committed
732
    x264_dct_function_t   dctf;
733
    x264_zigzag_function_t zigzagf;
734
    x264_quant_function_t quantf;
Loren Merritt's avatar
Loren Merritt committed
735
    x264_deblock_function_t loopf;
Laurent Aimar's avatar
Laurent Aimar committed
736

Loren Merritt's avatar
Loren Merritt committed
737 738 739
#if VISUALIZE
    struct visualize_t *visualize;
#endif
Steven Walters's avatar
Steven Walters committed
740
    x264_lookahead_t *lookahead;
Laurent Aimar's avatar
Laurent Aimar committed
741 742
};

743 744 745
// included at the end because it needs x264_t
#include "macroblock.h"

746 747 748 749
#ifdef HAVE_MMX
#include "x86/util.h"
#endif

Laurent Aimar's avatar
Laurent Aimar committed
750 751
#endif