vpx/vp8/decoder/dequantize.h
Scott LaVarnway ed9c66f584 Remove usage of predict buffer for decode
Instead of using the predict buffer, the decoder now writes
the predictor into the recon buffer.  For blocks with eob=0,
unnecessary idcts can be eliminated.  This gave a performance
boost of ~1.8% for the HD clips used.

Tero: Added needed changes to ARM side and scheduled some
      assembly code to prevent interlocks.

Patch Set 6:  Merged (I1bcdca7a95aacc3a181b9faa6b10e3a71ee24df3)
into this commit because of similarities in the idct
functions.
Patch Set 7: EC bug fix.

Change-Id: Ie31d90b5d3522e1108163f2ac491e455e3f955e6
2011-10-18 12:06:50 -04:00

113 lines
3.4 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 DEQUANTIZE_H
#define DEQUANTIZE_H
#include "vp8/common/blockd.h"
#define prototype_dequant_block(sym) \
void sym(BLOCKD *x)
#define prototype_dequant_idct_add(sym) \
void sym(short *input, short *dq, \
unsigned char *output, \
int stride)
#define prototype_dequant_dc_idct_add(sym) \
void sym(short *input, short *dq, \
unsigned char *dst, \
int stride, \
int dc)
#define prototype_dequant_dc_idct_add_y_block(sym) \
void sym(short *q, short *dq, \
unsigned char *dst, \
int stride, char *eobs, short *dc)
#define prototype_dequant_idct_add_y_block(sym) \
void sym(short *q, short *dq, \
unsigned char *dst, \
int stride, char *eobs)
#define prototype_dequant_idct_add_uv_block(sym) \
void sym(short *q, short *dq, \
unsigned char *dst_u, \
unsigned char *dst_v, int stride, char *eobs)
#if ARCH_X86 || ARCH_X86_64
#include "x86/dequantize_x86.h"
#endif
#if ARCH_ARM
#include "arm/dequantize_arm.h"
#endif
#ifndef vp8_dequant_block
#define vp8_dequant_block vp8_dequantize_b_c
#endif
extern prototype_dequant_block(vp8_dequant_block);
#ifndef vp8_dequant_idct_add
#define vp8_dequant_idct_add vp8_dequant_idct_add_c
#endif
extern prototype_dequant_idct_add(vp8_dequant_idct_add);
#ifndef vp8_dequant_dc_idct_add
#define vp8_dequant_dc_idct_add vp8_dequant_dc_idct_add_c
#endif
extern prototype_dequant_dc_idct_add(vp8_dequant_dc_idct_add);
#ifndef vp8_dequant_dc_idct_add_y_block
#define vp8_dequant_dc_idct_add_y_block vp8_dequant_dc_idct_add_y_block_c
#endif
extern prototype_dequant_dc_idct_add_y_block(vp8_dequant_dc_idct_add_y_block);
#ifndef vp8_dequant_idct_add_y_block
#define vp8_dequant_idct_add_y_block vp8_dequant_idct_add_y_block_c
#endif
extern prototype_dequant_idct_add_y_block(vp8_dequant_idct_add_y_block);
#ifndef vp8_dequant_idct_add_uv_block
#define vp8_dequant_idct_add_uv_block vp8_dequant_idct_add_uv_block_c
#endif
extern prototype_dequant_idct_add_uv_block(vp8_dequant_idct_add_uv_block);
typedef prototype_dequant_block((*vp8_dequant_block_fn_t));
typedef prototype_dequant_idct_add((*vp8_dequant_idct_add_fn_t));
typedef prototype_dequant_dc_idct_add((*vp8_dequant_dc_idct_add_fn_t));
typedef prototype_dequant_dc_idct_add_y_block((*vp8_dequant_dc_idct_add_y_block_fn_t));
typedef prototype_dequant_idct_add_y_block((*vp8_dequant_idct_add_y_block_fn_t));
typedef prototype_dequant_idct_add_uv_block((*vp8_dequant_idct_add_uv_block_fn_t));
typedef struct
{
vp8_dequant_block_fn_t block;
vp8_dequant_idct_add_fn_t idct_add;
vp8_dequant_dc_idct_add_fn_t dc_idct_add;
vp8_dequant_dc_idct_add_y_block_fn_t dc_idct_add_y_block;
vp8_dequant_idct_add_y_block_fn_t idct_add_y_block;
vp8_dequant_idct_add_uv_block_fn_t idct_add_uv_block;
} vp8_dequant_rtcd_vtable_t;
#if CONFIG_RUNTIME_CPU_DETECT
#define DEQUANT_INVOKE(ctx,fn) (ctx)->fn
#else
#define DEQUANT_INVOKE(ctx,fn) vp8_dequant_##fn
#endif
#endif