5ebe503f04
This prevents a duplicate memcpy of a 128-byte struct every time set_scale_factors() is called (which is a lot), thus leading to a decrease from 3.7 MB to 1.85 MB of struct copying per 64x64 block RD/partition loop. Overall, this decreases encoding time of the first 50 frames of bus @ 1500kbps (speed 0) from 1min5.9 to 1min4.9, i.e. about a 1.5% overall speedup. We can likely get more gains by removing the copy of the other struct (and replacing it with an indexing) as well. Change-Id: I3dceb7e79f71e6fe911b11cc994cf89a869dde7a
416 lines
16 KiB
C
416 lines
16 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.
|
|
*/
|
|
|
|
#include <assert.h>
|
|
|
|
#include "./vpx_config.h"
|
|
#include "vpx/vpx_integer.h"
|
|
#include "vp9/common/vp9_blockd.h"
|
|
#include "vp9/common/vp9_filter.h"
|
|
#include "vp9/common/vp9_reconinter.h"
|
|
#include "vp9/common/vp9_reconintra.h"
|
|
#include "./vpx_scale_rtcd.h"
|
|
|
|
static int scale_value_x_with_scaling(int val,
|
|
const struct scale_factors *scale) {
|
|
return (val * scale->x_scale_fp >> VP9_REF_SCALE_SHIFT);
|
|
}
|
|
|
|
static int scale_value_y_with_scaling(int val,
|
|
const struct scale_factors *scale) {
|
|
return (val * scale->y_scale_fp >> VP9_REF_SCALE_SHIFT);
|
|
}
|
|
|
|
static int unscaled_value(int val, const struct scale_factors *scale) {
|
|
(void) scale;
|
|
return val;
|
|
}
|
|
|
|
static MV32 mv_q3_to_q4_with_scaling(const MV *mv,
|
|
const struct scale_factors *scale) {
|
|
const MV32 res = {
|
|
((mv->row << 1) * scale->y_scale_fp >> VP9_REF_SCALE_SHIFT)
|
|
+ scale->y_offset_q4,
|
|
((mv->col << 1) * scale->x_scale_fp >> VP9_REF_SCALE_SHIFT)
|
|
+ scale->x_offset_q4
|
|
};
|
|
return res;
|
|
}
|
|
|
|
static MV32 mv_q3_to_q4_without_scaling(const MV *mv,
|
|
const struct scale_factors *scale) {
|
|
const MV32 res = {
|
|
mv->row << 1,
|
|
mv->col << 1
|
|
};
|
|
return res;
|
|
}
|
|
|
|
static MV32 mv_q4_with_scaling(const MV *mv,
|
|
const struct scale_factors *scale) {
|
|
const MV32 res = {
|
|
(mv->row * scale->y_scale_fp >> VP9_REF_SCALE_SHIFT) + scale->y_offset_q4,
|
|
(mv->col * scale->x_scale_fp >> VP9_REF_SCALE_SHIFT) + scale->x_offset_q4
|
|
};
|
|
return res;
|
|
}
|
|
|
|
static MV32 mv_q4_without_scaling(const MV *mv,
|
|
const struct scale_factors *scale) {
|
|
const MV32 res = {
|
|
mv->row,
|
|
mv->col
|
|
};
|
|
return res;
|
|
}
|
|
|
|
static void set_offsets_with_scaling(struct scale_factors *scale,
|
|
int row, int col) {
|
|
const int x_q4 = 16 * col;
|
|
const int y_q4 = 16 * row;
|
|
|
|
scale->x_offset_q4 = (x_q4 * scale->x_scale_fp >> VP9_REF_SCALE_SHIFT) & 0xf;
|
|
scale->y_offset_q4 = (y_q4 * scale->y_scale_fp >> VP9_REF_SCALE_SHIFT) & 0xf;
|
|
}
|
|
|
|
static void set_offsets_without_scaling(struct scale_factors *scale,
|
|
int row, int col) {
|
|
scale->x_offset_q4 = 0;
|
|
scale->y_offset_q4 = 0;
|
|
}
|
|
|
|
static int get_fixed_point_scale_factor(int other_size, int this_size) {
|
|
// Calculate scaling factor once for each reference frame
|
|
// and use fixed point scaling factors in decoding and encoding routines.
|
|
// Hardware implementations can calculate scale factor in device driver
|
|
// and use multiplication and shifting on hardware instead of division.
|
|
return (other_size << VP9_REF_SCALE_SHIFT) / this_size;
|
|
}
|
|
|
|
void vp9_setup_scale_factors_for_frame(struct scale_factors *scale,
|
|
int other_w, int other_h,
|
|
int this_w, int this_h) {
|
|
scale->x_scale_fp = get_fixed_point_scale_factor(other_w, this_w);
|
|
scale->x_offset_q4 = 0; // calculated per-mb
|
|
scale->x_step_q4 = (16 * scale->x_scale_fp >> VP9_REF_SCALE_SHIFT);
|
|
|
|
scale->y_scale_fp = get_fixed_point_scale_factor(other_h, this_h);
|
|
scale->y_offset_q4 = 0; // calculated per-mb
|
|
scale->y_step_q4 = (16 * scale->y_scale_fp >> VP9_REF_SCALE_SHIFT);
|
|
|
|
if ((other_w == this_w) && (other_h == this_h)) {
|
|
scale->scale_value_x = unscaled_value;
|
|
scale->scale_value_y = unscaled_value;
|
|
scale->set_scaled_offsets = set_offsets_without_scaling;
|
|
scale->scale_mv_q3_to_q4 = mv_q3_to_q4_without_scaling;
|
|
scale->scale_mv_q4 = mv_q4_without_scaling;
|
|
} else {
|
|
scale->scale_value_x = scale_value_x_with_scaling;
|
|
scale->scale_value_y = scale_value_y_with_scaling;
|
|
scale->set_scaled_offsets = set_offsets_with_scaling;
|
|
scale->scale_mv_q3_to_q4 = mv_q3_to_q4_with_scaling;
|
|
scale->scale_mv_q4 = mv_q4_with_scaling;
|
|
}
|
|
|
|
// TODO(agrange): Investigate the best choice of functions to use here
|
|
// for EIGHTTAP_SMOOTH. Since it is not interpolating, need to choose what
|
|
// to do at full-pel offsets. The current selection, where the filter is
|
|
// applied in one direction only, and not at all for 0,0, seems to give the
|
|
// best quality, but it may be worth trying an additional mode that does
|
|
// do the filtering on full-pel.
|
|
if (scale->x_step_q4 == 16) {
|
|
if (scale->y_step_q4 == 16) {
|
|
// No scaling in either direction.
|
|
scale->predict[0][0][0] = vp9_convolve_copy;
|
|
scale->predict[0][0][1] = vp9_convolve_avg;
|
|
scale->predict[0][1][0] = vp9_convolve8_vert;
|
|
scale->predict[0][1][1] = vp9_convolve8_avg_vert;
|
|
scale->predict[1][0][0] = vp9_convolve8_horiz;
|
|
scale->predict[1][0][1] = vp9_convolve8_avg_horiz;
|
|
} else {
|
|
// No scaling in x direction. Must always scale in the y direction.
|
|
scale->predict[0][0][0] = vp9_convolve8_vert;
|
|
scale->predict[0][0][1] = vp9_convolve8_avg_vert;
|
|
scale->predict[0][1][0] = vp9_convolve8_vert;
|
|
scale->predict[0][1][1] = vp9_convolve8_avg_vert;
|
|
scale->predict[1][0][0] = vp9_convolve8;
|
|
scale->predict[1][0][1] = vp9_convolve8_avg;
|
|
}
|
|
} else {
|
|
if (scale->y_step_q4 == 16) {
|
|
// No scaling in the y direction. Must always scale in the x direction.
|
|
scale->predict[0][0][0] = vp9_convolve8_horiz;
|
|
scale->predict[0][0][1] = vp9_convolve8_avg_horiz;
|
|
scale->predict[0][1][0] = vp9_convolve8;
|
|
scale->predict[0][1][1] = vp9_convolve8_avg;
|
|
scale->predict[1][0][0] = vp9_convolve8_horiz;
|
|
scale->predict[1][0][1] = vp9_convolve8_avg_horiz;
|
|
} else {
|
|
// Must always scale in both directions.
|
|
scale->predict[0][0][0] = vp9_convolve8;
|
|
scale->predict[0][0][1] = vp9_convolve8_avg;
|
|
scale->predict[0][1][0] = vp9_convolve8;
|
|
scale->predict[0][1][1] = vp9_convolve8_avg;
|
|
scale->predict[1][0][0] = vp9_convolve8;
|
|
scale->predict[1][0][1] = vp9_convolve8_avg;
|
|
}
|
|
}
|
|
// 2D subpel motion always gets filtered in both directions
|
|
scale->predict[1][1][0] = vp9_convolve8;
|
|
scale->predict[1][1][1] = vp9_convolve8_avg;
|
|
}
|
|
|
|
void vp9_setup_interp_filters(MACROBLOCKD *xd,
|
|
INTERPOLATIONFILTERTYPE mcomp_filter_type,
|
|
VP9_COMMON *cm) {
|
|
if (xd->mode_info_context) {
|
|
MB_MODE_INFO *mbmi = &xd->mode_info_context->mbmi;
|
|
|
|
set_scale_factors(xd, mbmi->ref_frame[0] - 1, mbmi->ref_frame[1] - 1,
|
|
cm->active_ref_scale);
|
|
}
|
|
|
|
switch (mcomp_filter_type) {
|
|
case EIGHTTAP:
|
|
case SWITCHABLE:
|
|
xd->subpix.filter_x = xd->subpix.filter_y = vp9_sub_pel_filters_8;
|
|
break;
|
|
case EIGHTTAP_SMOOTH:
|
|
xd->subpix.filter_x = xd->subpix.filter_y = vp9_sub_pel_filters_8lp;
|
|
break;
|
|
case EIGHTTAP_SHARP:
|
|
xd->subpix.filter_x = xd->subpix.filter_y = vp9_sub_pel_filters_8s;
|
|
break;
|
|
case BILINEAR:
|
|
xd->subpix.filter_x = xd->subpix.filter_y = vp9_bilinear_filters;
|
|
break;
|
|
}
|
|
assert(((intptr_t)xd->subpix.filter_x & 0xff) == 0);
|
|
}
|
|
|
|
void vp9_build_inter_predictor(const uint8_t *src, int src_stride,
|
|
uint8_t *dst, int dst_stride,
|
|
const int_mv *src_mv,
|
|
const struct scale_factors *scale,
|
|
int w, int h, int weight,
|
|
const struct subpix_fn_table *subpix,
|
|
enum mv_precision precision) {
|
|
const MV32 mv = precision == MV_PRECISION_Q4
|
|
? scale->scale_mv_q4(&src_mv->as_mv, scale)
|
|
: scale->scale_mv_q3_to_q4(&src_mv->as_mv, scale);
|
|
const int subpel_x = mv.col & 15;
|
|
const int subpel_y = mv.row & 15;
|
|
|
|
src += (mv.row >> 4) * src_stride + (mv.col >> 4);
|
|
scale->predict[!!subpel_x][!!subpel_y][weight](
|
|
src, src_stride, dst, dst_stride,
|
|
subpix->filter_x[subpel_x], scale->x_step_q4,
|
|
subpix->filter_y[subpel_y], scale->y_step_q4,
|
|
w, h);
|
|
}
|
|
|
|
static INLINE int round_mv_comp_q4(int value) {
|
|
return (value < 0 ? value - 2 : value + 2) / 4;
|
|
}
|
|
|
|
static int mi_mv_pred_row_q4(MACROBLOCKD *mb, int idx) {
|
|
const int temp = mb->mode_info_context->bmi[0].as_mv[idx].as_mv.row +
|
|
mb->mode_info_context->bmi[1].as_mv[idx].as_mv.row +
|
|
mb->mode_info_context->bmi[2].as_mv[idx].as_mv.row +
|
|
mb->mode_info_context->bmi[3].as_mv[idx].as_mv.row;
|
|
return round_mv_comp_q4(temp);
|
|
}
|
|
|
|
static int mi_mv_pred_col_q4(MACROBLOCKD *mb, int idx) {
|
|
const int temp = mb->mode_info_context->bmi[0].as_mv[idx].as_mv.col +
|
|
mb->mode_info_context->bmi[1].as_mv[idx].as_mv.col +
|
|
mb->mode_info_context->bmi[2].as_mv[idx].as_mv.col +
|
|
mb->mode_info_context->bmi[3].as_mv[idx].as_mv.col;
|
|
return round_mv_comp_q4(temp);
|
|
}
|
|
|
|
// TODO(jkoleszar): yet another mv clamping function :-(
|
|
MV clamp_mv_to_umv_border_sb(const MV *src_mv,
|
|
int bwl, int bhl, int ss_x, int ss_y,
|
|
int mb_to_left_edge, int mb_to_top_edge,
|
|
int mb_to_right_edge, int mb_to_bottom_edge) {
|
|
/* If the MV points so far into the UMV border that no visible pixels
|
|
* are used for reconstruction, the subpel part of the MV can be
|
|
* discarded and the MV limited to 16 pixels with equivalent results.
|
|
*/
|
|
const int spel_left = (VP9_INTERP_EXTEND + (4 << bwl)) << 4;
|
|
const int spel_right = spel_left - (1 << 4);
|
|
const int spel_top = (VP9_INTERP_EXTEND + (4 << bhl)) << 4;
|
|
const int spel_bottom = spel_top - (1 << 4);
|
|
MV clamped_mv;
|
|
|
|
assert(ss_x <= 1);
|
|
assert(ss_y <= 1);
|
|
clamped_mv.col = clamp(src_mv->col << (1 - ss_x),
|
|
(mb_to_left_edge << (1 - ss_x)) - spel_left,
|
|
(mb_to_right_edge << (1 - ss_x)) + spel_right);
|
|
clamped_mv.row = clamp(src_mv->row << (1 - ss_y),
|
|
(mb_to_top_edge << (1 - ss_y)) - spel_top,
|
|
(mb_to_bottom_edge << (1 - ss_y)) + spel_bottom);
|
|
return clamped_mv;
|
|
}
|
|
|
|
struct build_inter_predictors_args {
|
|
MACROBLOCKD *xd;
|
|
int x;
|
|
int y;
|
|
uint8_t* dst[MAX_MB_PLANE];
|
|
int dst_stride[MAX_MB_PLANE];
|
|
uint8_t* pre[2][MAX_MB_PLANE];
|
|
int pre_stride[2][MAX_MB_PLANE];
|
|
};
|
|
static void build_inter_predictors(int plane, int block,
|
|
BLOCK_SIZE_TYPE bsize,
|
|
int pred_w, int pred_h,
|
|
void *argv) {
|
|
const struct build_inter_predictors_args* const arg = argv;
|
|
MACROBLOCKD * const xd = arg->xd;
|
|
const int bwl = b_width_log2(bsize) - xd->plane[plane].subsampling_x;
|
|
const int bhl = b_height_log2(bsize) - xd->plane[plane].subsampling_y;
|
|
const int x = 4 * (block & ((1 << bwl) - 1)), y = 4 * (block >> bwl);
|
|
const int use_second_ref = xd->mode_info_context->mbmi.ref_frame[1] > 0;
|
|
int which_mv;
|
|
|
|
assert(x < (4 << bwl));
|
|
assert(y < (4 << bhl));
|
|
assert(xd->mode_info_context->mbmi.sb_type < BLOCK_SIZE_SB8X8 ||
|
|
4 << pred_w == (4 << bwl));
|
|
assert(xd->mode_info_context->mbmi.sb_type < BLOCK_SIZE_SB8X8 ||
|
|
4 << pred_h == (4 << bhl));
|
|
|
|
for (which_mv = 0; which_mv < 1 + use_second_ref; ++which_mv) {
|
|
// source
|
|
const uint8_t * const base_pre = arg->pre[which_mv][plane];
|
|
const int pre_stride = arg->pre_stride[which_mv][plane];
|
|
const uint8_t *const pre = base_pre +
|
|
scaled_buffer_offset(x, y, pre_stride, &xd->scale_factor[which_mv]);
|
|
struct scale_factors * const scale = &xd->scale_factor[which_mv];
|
|
|
|
// dest
|
|
uint8_t *const dst = arg->dst[plane] + arg->dst_stride[plane] * y + x;
|
|
|
|
// motion vector
|
|
const MV *mv;
|
|
MV split_chroma_mv;
|
|
int_mv clamped_mv;
|
|
|
|
if (xd->mode_info_context->mbmi.sb_type < BLOCK_SIZE_SB8X8) {
|
|
if (plane == 0) {
|
|
mv = &xd->mode_info_context->bmi[block].as_mv[which_mv].as_mv;
|
|
} else {
|
|
// TODO(jkoleszar): All chroma MVs in SPLITMV mode are taken as the
|
|
// same MV (the average of the 4 luma MVs) but we could do something
|
|
// smarter for non-4:2:0. Just punt for now, pending the changes to get
|
|
// rid of SPLITMV mode entirely.
|
|
split_chroma_mv.row = mi_mv_pred_row_q4(xd, which_mv);
|
|
split_chroma_mv.col = mi_mv_pred_col_q4(xd, which_mv);
|
|
mv = &split_chroma_mv;
|
|
}
|
|
} else {
|
|
mv = &xd->mode_info_context->mbmi.mv[which_mv].as_mv;
|
|
}
|
|
|
|
/* TODO(jkoleszar): This clamping is done in the incorrect place for the
|
|
* scaling case. It needs to be done on the scaled MV, not the pre-scaling
|
|
* MV. Note however that it performs the subsampling aware scaling so
|
|
* that the result is always q4.
|
|
*/
|
|
clamped_mv.as_mv = clamp_mv_to_umv_border_sb(mv, bwl, bhl,
|
|
xd->plane[plane].subsampling_x,
|
|
xd->plane[plane].subsampling_y,
|
|
xd->mb_to_left_edge,
|
|
xd->mb_to_top_edge,
|
|
xd->mb_to_right_edge,
|
|
xd->mb_to_bottom_edge);
|
|
scale->set_scaled_offsets(scale, arg->y + y, arg->x + x);
|
|
|
|
vp9_build_inter_predictor(pre, pre_stride,
|
|
dst, arg->dst_stride[plane],
|
|
&clamped_mv, &xd->scale_factor[which_mv],
|
|
4 << pred_w, 4 << pred_h, which_mv,
|
|
&xd->subpix, MV_PRECISION_Q4);
|
|
}
|
|
}
|
|
void vp9_build_inter_predictors_sby(MACROBLOCKD *xd,
|
|
int mi_row,
|
|
int mi_col,
|
|
BLOCK_SIZE_TYPE bsize) {
|
|
struct build_inter_predictors_args args = {
|
|
xd, mi_col * MI_SIZE, mi_row * MI_SIZE,
|
|
{xd->plane[0].dst.buf, NULL, NULL}, {xd->plane[0].dst.stride, 0, 0},
|
|
{{xd->plane[0].pre[0].buf, NULL, NULL},
|
|
{xd->plane[0].pre[1].buf, NULL, NULL}},
|
|
{{xd->plane[0].pre[0].stride, 0, 0}, {xd->plane[0].pre[1].stride, 0, 0}},
|
|
};
|
|
|
|
foreach_predicted_block_in_plane(xd, bsize, 0, build_inter_predictors, &args);
|
|
}
|
|
void vp9_build_inter_predictors_sbuv(MACROBLOCKD *xd,
|
|
int mi_row,
|
|
int mi_col,
|
|
BLOCK_SIZE_TYPE bsize) {
|
|
struct build_inter_predictors_args args = {
|
|
xd, mi_col * MI_SIZE, mi_row * MI_SIZE,
|
|
#if CONFIG_ALPHA
|
|
{NULL, xd->plane[1].dst.buf, xd->plane[2].dst.buf,
|
|
xd->plane[3].dst.buf},
|
|
{0, xd->plane[1].dst.stride, xd->plane[1].dst.stride,
|
|
xd->plane[3].dst.stride},
|
|
{{NULL, xd->plane[1].pre[0].buf, xd->plane[2].pre[0].buf,
|
|
xd->plane[3].pre[0].buf},
|
|
{NULL, xd->plane[1].pre[1].buf, xd->plane[2].pre[1].buf,
|
|
xd->plane[3].pre[1].buf}},
|
|
{{0, xd->plane[1].pre[0].stride, xd->plane[1].pre[0].stride,
|
|
xd->plane[3].pre[0].stride},
|
|
{0, xd->plane[1].pre[1].stride, xd->plane[1].pre[1].stride,
|
|
xd->plane[3].pre[1].stride}},
|
|
#else
|
|
{NULL, xd->plane[1].dst.buf, xd->plane[2].dst.buf},
|
|
{0, xd->plane[1].dst.stride, xd->plane[1].dst.stride},
|
|
{{NULL, xd->plane[1].pre[0].buf, xd->plane[2].pre[0].buf},
|
|
{NULL, xd->plane[1].pre[1].buf, xd->plane[2].pre[1].buf}},
|
|
{{0, xd->plane[1].pre[0].stride, xd->plane[1].pre[0].stride},
|
|
{0, xd->plane[1].pre[1].stride, xd->plane[1].pre[1].stride}},
|
|
#endif
|
|
};
|
|
foreach_predicted_block_uv(xd, bsize, build_inter_predictors, &args);
|
|
}
|
|
void vp9_build_inter_predictors_sb(MACROBLOCKD *xd,
|
|
int mi_row, int mi_col,
|
|
BLOCK_SIZE_TYPE bsize) {
|
|
|
|
vp9_build_inter_predictors_sby(xd, mi_row, mi_col, bsize);
|
|
vp9_build_inter_predictors_sbuv(xd, mi_row, mi_col, bsize);
|
|
}
|
|
|
|
// TODO(dkovalev: find better place for this function)
|
|
void vp9_setup_scale_factors(VP9_COMMON *cm, int i) {
|
|
const int ref = cm->active_ref_idx[i];
|
|
struct scale_factors *const sf = &cm->active_ref_scale[i];
|
|
if (ref >= NUM_YV12_BUFFERS) {
|
|
memset(sf, 0, sizeof(*sf));
|
|
} else {
|
|
YV12_BUFFER_CONFIG *const fb = &cm->yv12_fb[ref];
|
|
vp9_setup_scale_factors_for_frame(sf,
|
|
fb->y_crop_width, fb->y_crop_height,
|
|
cm->width, cm->height);
|
|
|
|
if (sf->x_scale_fp != VP9_REF_NO_SCALE ||
|
|
sf->y_scale_fp != VP9_REF_NO_SCALE)
|
|
vp9_extend_frame_borders(fb, cm->subsampling_x, cm->subsampling_y);
|
|
}
|
|
}
|
|
|