vpx/vp8/decoder/onyxd_int.h
Ronald S. Bultje 60cb39da86 Dual 16x16 inter prediction.
This patch introduces the concept of dual inter16x16 prediction. A
16x16 inter-predicted macroblock can use 2 references instead of 1,
where both references use the same mvmode (new, near/est, zero). In the
case of newmv, this means that two MVs are coded instead of one. The
frame can be encoded in 3 ways: all MBs single-prediction, all MBs dual
prediction, or per-MB single/dual prediction selection ("hybrid"), in
which case a single bit is coded per-MB to indicate whether the MB uses
single or dual inter prediction.

In the future, we can (maybe?) get further gains by mixing this with
Adrian's 32x32 work, per-segment dual prediction settings, or adding
support for dual splitmv/8x8mv inter prediction.

Gain (on derf-set, CQ mode) is ~2.8% (SSIM) or ~3.6% (glb PSNR). Most
gain is at medium/high bitrates, but there's minor gains at low bitrates
also. Output was confirmed to match between encoder and decoder.

Note for optimization people: this patch introduces a 2nd version of
16x16/8x8 sixtap/bilin functions, which does an avg instead of a
store. They may want to look and make sure this is implemented to
their satisfaction so we can optimize it best in the future.

Change-ID: I59dc84b07cbb3ccf073ac0f756d03d294cb19281
2011-12-06 11:53:02 -08:00

178 lines
4.2 KiB
C

/*
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#ifndef __INC_VP8D_INT_H
#define __INC_VP8D_INT_H
#include "vpx_ports/config.h"
#include "vp8/common/onyxd.h"
#include "treereader.h"
#include "vp8/common/onyxc_int.h"
#include "vp8/common/threading.h"
#include "dequantize.h"
#if CONFIG_ERROR_CONCEALMENT
#include "ec_types.h"
#endif
//#define DEC_DEBUG
typedef struct
{
int ithread;
void *ptr1;
void *ptr2;
} DECODETHREAD_DATA;
typedef struct
{
MACROBLOCKD mbd;
int mb_row;
int current_mb_col;
short *coef_ptr;
} MB_ROW_DEC;
typedef struct
{
int64_t time_stamp;
int size;
} DATARATE;
typedef struct
{
int const *scan;
#if CONFIG_T8X8
int const *scan_8x8;
#endif
UINT8 const *ptr_block2leftabove;
vp8_tree_index const *vp8_coef_tree_ptr;
unsigned char *norm_ptr;
UINT8 *ptr_coef_bands_x;
#if CONFIG_T8X8
UINT8 *ptr_coef_bands_x_8x8;
#endif
ENTROPY_CONTEXT_PLANES *A;
ENTROPY_CONTEXT_PLANES *L;
INT16 *qcoeff_start_ptr;
BOOL_DECODER *current_bc;
vp8_prob const *coef_probs[4];
#if CONFIG_T8X8
vp8_prob const *coef_probs_8x8[4];
#endif
UINT8 eob[25];
} DETOK;
typedef struct VP8Decompressor
{
DECLARE_ALIGNED(16, MACROBLOCKD, mb);
DECLARE_ALIGNED(16, VP8_COMMON, common);
vp8_reader bc, bc2;
VP8D_CONFIG oxcf;
const unsigned char *Source;
unsigned int source_sz;
const unsigned char *partitions[MAX_PARTITIONS];
unsigned int partition_sizes[MAX_PARTITIONS];
unsigned int num_partitions;
unsigned char *segmentation_map;
#if CONFIG_MULTITHREAD
/* variable for threading */
volatile int b_multithreaded_rd;
int max_threads;
int current_mb_col_main;
int decoding_thread_count;
int allocated_decoding_thread_count;
int mt_baseline_filter_level[MAX_MB_SEGMENTS];
int sync_range;
int *mt_current_mb_col; /* Each row remembers its already decoded column. */
unsigned char **mt_yabove_row; /* mb_rows x width */
unsigned char **mt_uabove_row;
unsigned char **mt_vabove_row;
unsigned char **mt_yleft_col; /* mb_rows x 16 */
unsigned char **mt_uleft_col; /* mb_rows x 8 */
unsigned char **mt_vleft_col; /* mb_rows x 8 */
MB_ROW_DEC *mb_row_di;
DECODETHREAD_DATA *de_thread_data;
pthread_t *h_decoding_thread;
sem_t *h_event_start_decoding;
sem_t h_event_end_decoding;
/* end of threading data */
#endif
vp8_reader *mbc;
int64_t last_time_stamp;
int ready_for_new_data;
DATARATE dr[16];
DETOK detoken;
#if CONFIG_RUNTIME_CPU_DETECT
vp8_dequant_rtcd_vtable_t dequant;
#endif
vp8_prob prob_intra;
vp8_prob prob_last;
vp8_prob prob_gf;
vp8_prob prob_skip_false;
#if CONFIG_DUALPRED
vp8_prob prob_dualpred[3];
#endif /* CONFIG_DUALPRED */
#if CONFIG_ERROR_CONCEALMENT
MB_OVERLAP *overlaps;
/* the mb num from which modes and mvs (first partition) are corrupt */
unsigned int mvs_corrupt_from_mb;
#endif
int ec_enabled;
int ec_active;
int input_partition;
int decoded_key_frame;
int independent_partitions;
int frame_corrupt_residual;
} VP8D_COMP;
int vp8_decode_frame(VP8D_COMP *cpi);
void vp8_dmachine_specific_config(VP8D_COMP *pbi);
#if CONFIG_DEBUG
#define CHECK_MEM_ERROR(lval,expr) do {\
lval = (expr); \
if(!lval) \
vpx_internal_error(&pbi->common.error, VPX_CODEC_MEM_ERROR,\
"Failed to allocate "#lval" at %s:%d", \
__FILE__,__LINE__);\
} while(0)
#else
#define CHECK_MEM_ERROR(lval,expr) do {\
lval = (expr); \
if(!lval) \
vpx_internal_error(&pbi->common.error, VPX_CODEC_MEM_ERROR,\
"Failed to allocate "#lval);\
} while(0)
#endif
#endif