You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 

1150 lines
39 KiB

/*
* VVC parameter set parser
*
* Copyright (C) 2023 Nuo Mi
* Copyright (C) 2022 Xu Mu
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "libavcodec/cbs_h266.h"
#include "libavutil/imgutils.h"
#include "libavcodec/refstruct.h"
#include "vvc_data.h"
#include "vvc_ps.h"
#include "vvcdec.h"
static int sps_map_pixel_format(VVCSPS *sps, void *log_ctx)
{
const H266RawSPS *r = sps->r;
const AVPixFmtDescriptor *desc;
switch (sps->bit_depth) {
case 8:
if (r->sps_chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY8;
if (r->sps_chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P;
if (r->sps_chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P;
if (r->sps_chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P;
break;
case 10:
if (r->sps_chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY10;
if (r->sps_chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P10;
if (r->sps_chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P10;
if (r->sps_chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P10;
break;
case 12:
if (r->sps_chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY12;
if (r->sps_chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P12;
if (r->sps_chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P12;
if (r->sps_chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P12;
break;
default:
av_log(log_ctx, AV_LOG_ERROR,
"The following bit-depths are currently specified: 8, 10, 12 bits, "
"chroma_format_idc is %d, depth is %d\n",
r->sps_chroma_format_idc, sps->bit_depth);
return AVERROR_INVALIDDATA;
}
desc = av_pix_fmt_desc_get(sps->pix_fmt);
if (!desc)
return AVERROR(EINVAL);
sps->hshift[0] = sps->vshift[0] = 0;
sps->hshift[2] = sps->hshift[1] = desc->log2_chroma_w;
sps->vshift[2] = sps->vshift[1] = desc->log2_chroma_h;
sps->pixel_shift = sps->bit_depth > 8;
return 0;
}
static int sps_bit_depth(VVCSPS *sps, void *log_ctx)
{
const H266RawSPS *r = sps->r;
sps->bit_depth = r->sps_bitdepth_minus8 + 8;
sps->qp_bd_offset = 6 * (sps->bit_depth - 8);
sps->log2_transform_range =
r->sps_extended_precision_flag ? FFMAX(15, FFMIN(20, sps->bit_depth + 6)) : 15;
return sps_map_pixel_format(sps, log_ctx);
}
static int sps_chroma_qp_table(VVCSPS *sps)
{
const H266RawSPS *r = sps->r;
const int num_qp_tables = r->sps_same_qp_table_for_chroma_flag ?
1 : (r->sps_joint_cbcr_enabled_flag ? 3 : 2);
for (int i = 0; i < num_qp_tables; i++) {
int num_points_in_qp_table;
int8_t qp_in[VVC_MAX_POINTS_IN_QP_TABLE], qp_out[VVC_MAX_POINTS_IN_QP_TABLE];
unsigned int delta_qp_in[VVC_MAX_POINTS_IN_QP_TABLE];
int off = sps->qp_bd_offset;
num_points_in_qp_table = r->sps_num_points_in_qp_table_minus1[i] + 1;
qp_out[0] = qp_in[0] = r->sps_qp_table_start_minus26[i] + 26;
for (int j = 0; j < num_points_in_qp_table; j++ ) {
delta_qp_in[j] = r->sps_delta_qp_in_val_minus1[i][j] + 1;
qp_in[j+1] = qp_in[j] + delta_qp_in[j];
qp_out[j+1] = qp_out[j] + (r->sps_delta_qp_in_val_minus1[i][j] ^ r->sps_delta_qp_diff_val[i][j]);
}
sps->chroma_qp_table[i][qp_in[0] + off] = qp_out[0];
for (int k = qp_in[0] - 1 + off; k >= 0; k--)
sps->chroma_qp_table[i][k] = av_clip(sps->chroma_qp_table[i][k+1]-1, -off, 63);
for (int j = 0; j < num_points_in_qp_table; j++) {
int sh = delta_qp_in[j] >> 1;
for (int k = qp_in[j] + 1 + off, m = 1; k <= qp_in[j+1] + off; k++, m++) {
sps->chroma_qp_table[i][k] = sps->chroma_qp_table[i][qp_in[j] + off] +
((qp_out[j+1] - qp_out[j]) * m + sh) / delta_qp_in[j];
}
}
for (int k = qp_in[num_points_in_qp_table] + 1 + off; k <= 63 + off; k++)
sps->chroma_qp_table[i][k] = av_clip(sps->chroma_qp_table[i][k-1] + 1, -sps->qp_bd_offset, 63);
}
if (r->sps_same_qp_table_for_chroma_flag) {
memcpy(&sps->chroma_qp_table[1], &sps->chroma_qp_table[0], sizeof(sps->chroma_qp_table[0]));
memcpy(&sps->chroma_qp_table[2], &sps->chroma_qp_table[0], sizeof(sps->chroma_qp_table[0]));
}
return 0;
}
static void sps_poc(VVCSPS *sps)
{
sps->max_pic_order_cnt_lsb = 1 << (sps->r->sps_log2_max_pic_order_cnt_lsb_minus4 + 4);
}
static void sps_inter(VVCSPS *sps)
{
const H266RawSPS *r = sps->r;
sps->max_num_merge_cand = 6 - r->sps_six_minus_max_num_merge_cand;
sps->max_num_ibc_merge_cand = 6 - r->sps_six_minus_max_num_ibc_merge_cand;
if (sps->r->sps_gpm_enabled_flag) {
sps->max_num_gpm_merge_cand = 2;
if (sps->max_num_merge_cand >= 3)
sps->max_num_gpm_merge_cand = sps->max_num_merge_cand - r->sps_max_num_merge_cand_minus_max_num_gpm_cand;
}
sps->log2_parallel_merge_level = r->sps_log2_parallel_merge_level_minus2 + 2;
}
static void sps_partition_constraints(VVCSPS* sps)
{
const H266RawSPS *r = sps->r;
sps->ctb_log2_size_y = r->sps_log2_ctu_size_minus5 + 5;
sps->ctb_size_y = 1 << sps->ctb_log2_size_y;
sps->min_cb_log2_size_y = r->sps_log2_min_luma_coding_block_size_minus2 + 2;
sps->min_cb_size_y = 1 << sps->min_cb_log2_size_y;
sps->max_tb_size_y = 1 << (r->sps_max_luma_transform_size_64_flag ? 6 : 5);
sps->max_ts_size = 1 << (r->sps_log2_transform_skip_max_size_minus2 + 2);
}
static void sps_ladf(VVCSPS* sps)
{
const H266RawSPS *r = sps->r;
if (r->sps_ladf_enabled_flag) {
sps->num_ladf_intervals = r->sps_num_ladf_intervals_minus2 + 2;
sps->ladf_interval_lower_bound[0] = 0;
for (int i = 0; i < sps->num_ladf_intervals - 1; i++) {
sps->ladf_interval_lower_bound[i + 1] =
sps->ladf_interval_lower_bound[i] + r->sps_ladf_delta_threshold_minus1[i] + 1;
}
}
}
static int sps_derive(VVCSPS *sps, void *log_ctx)
{
int ret;
const H266RawSPS *r = sps->r;
sps->width = r->sps_pic_width_max_in_luma_samples;
sps->height = r->sps_pic_height_max_in_luma_samples;
ret = sps_bit_depth(sps, log_ctx);
if (ret < 0)
return ret;
sps_poc(sps);
sps_inter(sps);
sps_partition_constraints(sps);
sps_ladf(sps);
if (r->sps_chroma_format_idc != 0)
sps_chroma_qp_table(sps);
return 0;
}
static void sps_free(FFRefStructOpaque opaque, void *obj)
{
VVCSPS *sps = obj;
ff_refstruct_unref(&sps->r);
}
static const VVCSPS *sps_alloc(const H266RawSPS *rsps, void *log_ctx)
{
int ret;
VVCSPS *sps = ff_refstruct_alloc_ext(sizeof(*sps), 0, NULL, sps_free);
if (!sps)
return NULL;
ff_refstruct_replace(&sps->r, rsps);
ret = sps_derive(sps, log_ctx);
if (ret < 0)
goto fail;
return sps;
fail:
ff_refstruct_unref(&sps);
return NULL;
}
static int decode_sps(VVCParamSets *ps, const H266RawSPS *rsps, void *log_ctx)
{
const int sps_id = rsps->sps_seq_parameter_set_id;
const VVCSPS *old_sps = ps->sps_list[sps_id];
const VVCSPS *sps;
if (old_sps && old_sps->r == rsps)
return 0;
sps = sps_alloc(rsps, log_ctx);
if (!sps)
return AVERROR(ENOMEM);
ff_refstruct_unref(&ps->sps_list[sps_id]);
ps->sps_list[sps_id] = sps;
return 0;
}
static void pps_chroma_qp_offset(VVCPPS *pps)
{
pps->chroma_qp_offset[CB - 1] = pps->r->pps_cb_qp_offset;
pps->chroma_qp_offset[CR - 1] = pps->r->pps_cr_qp_offset;
pps->chroma_qp_offset[JCBCR - 1]= pps->r->pps_joint_cbcr_qp_offset_value;
for (int i = 0; i < 6; i++) {
pps->chroma_qp_offset_list[i][CB - 1] = pps->r->pps_cb_qp_offset_list[i];
pps->chroma_qp_offset_list[i][CR - 1] = pps->r->pps_cr_qp_offset_list[i];
pps->chroma_qp_offset_list[i][JCBCR - 1]= pps->r->pps_joint_cbcr_qp_offset_list[i];
}
}
static void pps_width_height(VVCPPS *pps, const VVCSPS *sps)
{
const H266RawPPS *r = pps->r;
pps->width = r->pps_pic_width_in_luma_samples;
pps->height = r->pps_pic_height_in_luma_samples;
pps->ctb_width = AV_CEIL_RSHIFT(pps->width, sps->ctb_log2_size_y);
pps->ctb_height = AV_CEIL_RSHIFT(pps->height, sps->ctb_log2_size_y);
pps->ctb_count = pps->ctb_width * pps->ctb_height;
pps->min_cb_width = pps->width >> sps->min_cb_log2_size_y;
pps->min_cb_height = pps->height >> sps->min_cb_log2_size_y;
pps->min_pu_width = pps->width >> MIN_PU_LOG2;
pps->min_pu_height = pps->height >> MIN_PU_LOG2;
pps->min_tu_width = pps->width >> MIN_TU_LOG2;
pps->min_tu_height = pps->height >> MIN_TU_LOG2;
pps->width32 = AV_CEIL_RSHIFT(pps->width, 5);
pps->height32 = AV_CEIL_RSHIFT(pps->height, 5);
pps->width64 = AV_CEIL_RSHIFT(pps->width, 6);
pps->height64 = AV_CEIL_RSHIFT(pps->height, 6);
}
static int pps_bd(VVCPPS *pps)
{
const H266RawPPS *r = pps->r;
pps->col_bd = av_calloc(r->num_tile_columns + 1, sizeof(*pps->col_bd));
pps->row_bd = av_calloc(r->num_tile_rows + 1, sizeof(*pps->row_bd));
pps->ctb_to_col_bd = av_calloc(pps->ctb_width + 1, sizeof(*pps->ctb_to_col_bd));
pps->ctb_to_row_bd = av_calloc(pps->ctb_height + 1, sizeof(*pps->ctb_to_col_bd));
if (!pps->col_bd || !pps->row_bd || !pps->ctb_to_col_bd || !pps->ctb_to_row_bd)
return AVERROR(ENOMEM);
for (int i = 0, j = 0; i < r->num_tile_columns; i++) {
pps->col_bd[i] = j;
j += r->col_width_val[i];
for (int k = pps->col_bd[i]; k < j; k++)
pps->ctb_to_col_bd[k] = pps->col_bd[i];
}
for (int i = 0, j = 0; i < r->num_tile_rows; i++) {
pps->row_bd[i] = j;
j += r->row_height_val[i];
for (int k = pps->row_bd[i]; k < j; k++)
pps->ctb_to_row_bd[k] = pps->row_bd[i];
}
return 0;
}
static int next_tile_idx(int tile_idx, const int i, const H266RawPPS *r)
{
if (r->pps_tile_idx_delta_present_flag) {
tile_idx += r->pps_tile_idx_delta_val[i];
} else {
tile_idx += r->pps_slice_width_in_tiles_minus1[i] + 1;
if (tile_idx % r->num_tile_columns == 0)
tile_idx += (r->pps_slice_height_in_tiles_minus1[i]) * r->num_tile_columns;
}
return tile_idx;
}
static void tile_xy(int *tile_x, int *tile_y, const int tile_idx, const VVCPPS *pps)
{
*tile_x = tile_idx % pps->r->num_tile_columns;
*tile_y = tile_idx / pps->r->num_tile_columns;
}
static void ctu_xy(int *ctu_x, int *ctu_y, const int tile_x, const int tile_y, const VVCPPS *pps)
{
*ctu_x = pps->col_bd[tile_x];
*ctu_y = pps->row_bd[tile_y];
}
static int ctu_rs(const int ctu_x, const int ctu_y, const VVCPPS *pps)
{
return pps->ctb_width * ctu_y + ctu_x;
}
static int pps_add_ctus(VVCPPS *pps, int *off, const int ctu_x, const int ctu_y,
const int w, const int h)
{
int start = *off;
for (int y = 0; y < h; y++) {
for (int x = 0; x < w; x++) {
pps->ctb_addr_in_slice[*off] = ctu_rs(ctu_x + x, ctu_y + y, pps);
(*off)++;
}
}
return *off - start;
}
static int pps_one_tile_slices(VVCPPS *pps, const int tile_idx, int i, int *off)
{
const H266RawPPS *r = pps->r;
int ctu_x, ctu_y, ctu_y_end, tile_x, tile_y;
tile_xy(&tile_x, &tile_y, tile_idx, pps);
ctu_xy(&ctu_x, &ctu_y, tile_x, tile_y, pps);
ctu_y_end = ctu_y + r->row_height_val[tile_y];
while (ctu_y < ctu_y_end) {
pps->slice_start_offset[i] = *off;
pps->num_ctus_in_slice[i] = pps_add_ctus(pps, off, ctu_x, ctu_y,
r->col_width_val[tile_x], r->slice_height_in_ctus[i]);
ctu_y += r->slice_height_in_ctus[i++];
}
i--;
return i;
}
static void pps_multi_tiles_slice(VVCPPS *pps, const int tile_idx, const int i, int *off)
{
const H266RawPPS *r = pps->r;
int ctu_x, ctu_y,tile_x, tile_y;
tile_xy(&tile_x, &tile_y, tile_idx, pps);
pps->slice_start_offset[i] = *off;
pps->num_ctus_in_slice[i] = 0;
for (int ty = tile_y; ty <= tile_y + r->pps_slice_height_in_tiles_minus1[i]; ty++) {
for (int tx = tile_x; tx <= tile_x + r->pps_slice_width_in_tiles_minus1[i]; tx++) {
ctu_xy(&ctu_x, &ctu_y, tx, ty, pps);
pps->num_ctus_in_slice[i] += pps_add_ctus(pps, off, ctu_x, ctu_y,
r->col_width_val[tx], r->row_height_val[ty]);
}
}
}
static void pps_rect_slice(VVCPPS* pps)
{
const H266RawPPS* r = pps->r;
int tile_idx = 0, off = 0;
for (int i = 0; i < r->pps_num_slices_in_pic_minus1 + 1; i++) {
if (!r->pps_slice_width_in_tiles_minus1[i] &&
!r->pps_slice_height_in_tiles_minus1[i]) {
i = pps_one_tile_slices(pps, tile_idx, i, &off);
} else {
pps_multi_tiles_slice(pps, tile_idx, i, &off);
}
tile_idx = next_tile_idx(tile_idx, i, r);
}
}
static void pps_no_rect_slice(VVCPPS* pps)
{
const H266RawPPS* r = pps->r;
int ctu_x, ctu_y, off = 0;
for (int tile_y = 0; tile_y < r->num_tile_rows; tile_y++) {
for (int tile_x = 0; tile_x < r->num_tile_columns; tile_x++) {
ctu_xy(&ctu_x, &ctu_y, tile_x, tile_y, pps);
pps_add_ctus(pps, &off, ctu_x, ctu_y, r->col_width_val[tile_x], r->row_height_val[tile_y]);
}
}
}
static int pps_slice_map(VVCPPS *pps)
{
pps->ctb_addr_in_slice = av_calloc(pps->ctb_count, sizeof(*pps->ctb_addr_in_slice));
if (!pps->ctb_addr_in_slice)
return AVERROR(ENOMEM);
if (pps->r->pps_rect_slice_flag)
pps_rect_slice(pps);
else
pps_no_rect_slice(pps);
return 0;
}
static void pps_ref_wraparound_offset(VVCPPS *pps, const VVCSPS *sps)
{
const H266RawPPS *r = pps->r;
if (r->pps_ref_wraparound_enabled_flag)
pps->ref_wraparound_offset = (pps->width / sps->min_cb_size_y) - r->pps_pic_width_minus_wraparound_offset;
}
static int pps_derive(VVCPPS *pps, const VVCSPS *sps)
{
int ret;
pps_chroma_qp_offset(pps);
pps_width_height(pps, sps);
ret = pps_bd(pps);
if (ret < 0)
return ret;
ret = pps_slice_map(pps);
if (ret < 0)
return ret;
pps_ref_wraparound_offset(pps, sps);
return 0;
}
static void pps_free(FFRefStructOpaque opaque, void *obj)
{
VVCPPS *pps = obj;
ff_refstruct_unref(&pps->r);
av_freep(&pps->col_bd);
av_freep(&pps->row_bd);
av_freep(&pps->ctb_to_col_bd);
av_freep(&pps->ctb_to_row_bd);
av_freep(&pps->ctb_addr_in_slice);
}
static const VVCPPS *pps_alloc(const H266RawPPS *rpps, const VVCSPS *sps)
{
int ret;
VVCPPS *pps = ff_refstruct_alloc_ext(sizeof(*pps), 0, NULL, pps_free);
if (!pps)
return NULL;
ff_refstruct_replace(&pps->r, rpps);
ret = pps_derive(pps, sps);
if (ret < 0)
goto fail;
return pps;
fail:
ff_refstruct_unref(&pps);
return NULL;
}
static int decode_pps(VVCParamSets *ps, const H266RawPPS *rpps)
{
int ret = 0;
const int pps_id = rpps->pps_pic_parameter_set_id;
const int sps_id = rpps->pps_seq_parameter_set_id;
const VVCPPS *old_pps = ps->pps_list[pps_id];
const VVCPPS *pps;
if (old_pps && old_pps->r == rpps)
return 0;
pps = pps_alloc(rpps, ps->sps_list[sps_id]);
if (!pps)
return AVERROR(ENOMEM);
ff_refstruct_unref(&ps->pps_list[pps_id]);
ps->pps_list[pps_id] = pps;
return ret;
}
static int decode_ps(VVCParamSets *ps, const CodedBitstreamH266Context *h266, void *log_ctx)
{
const H266RawPictureHeader *ph = h266->ph;
const H266RawPPS *rpps;
const H266RawSPS *rsps;
int ret;
if (!ph)
return AVERROR_INVALIDDATA;
rpps = h266->pps[ph->ph_pic_parameter_set_id];
if (!rpps)
return AVERROR_INVALIDDATA;
rsps = h266->sps[rpps->pps_seq_parameter_set_id];
if (!rsps)
return AVERROR_INVALIDDATA;
ret = decode_sps(ps, rsps, log_ctx);
if (ret < 0)
return ret;
ret = decode_pps(ps, rpps);
if (ret < 0)
return ret;
return 0;
}
#define WEIGHT_TABLE(x) \
w->nb_weights[L##x] = r->num_weights_l##x; \
for (int i = 0; i < w->nb_weights[L##x]; i++) { \
w->weight_flag[L##x][LUMA][i] = r->luma_weight_l##x##_flag[i]; \
w->weight_flag[L##x][CHROMA][i] = r->chroma_weight_l##x##_flag[i]; \
w->weight[L##x][LUMA][i] = denom[LUMA] + r->delta_luma_weight_l##x[i]; \
w->offset[L##x][LUMA][i] = r->luma_offset_l##x[i]; \
for (int j = CB; j <= CR; j++) { \
w->weight[L##x][j][i] = denom[CHROMA] + r->delta_chroma_weight_l##x[i][j - 1]; \
w->offset[L##x][j][i] = 128 + r->delta_chroma_offset_l##x[i][j - 1]; \
w->offset[L##x][j][i] -= (128 * w->weight[L##x][j][i]) >> w->log2_denom[CHROMA]; \
w->offset[L##x][j][i] = av_clip_intp2(w->offset[L##x][j][i], 7); \
} \
} \
static void pred_weight_table(PredWeightTable *w, const H266RawPredWeightTable *r)
{
int denom[2];
w->log2_denom[LUMA] = r->luma_log2_weight_denom;
w->log2_denom[CHROMA] = w->log2_denom[LUMA] + r->delta_chroma_log2_weight_denom;
denom[LUMA] = 1 << w->log2_denom[LUMA];
denom[CHROMA] = 1 << w->log2_denom[CHROMA];
WEIGHT_TABLE(0)
WEIGHT_TABLE(1)
}
// 8.3.1 Decoding process for picture order count
static int ph_compute_poc(const H266RawPictureHeader *ph, const H266RawSPS *sps, const int poc_tid0, const int is_clvss)
{
const int max_poc_lsb = 1 << (sps->sps_log2_max_pic_order_cnt_lsb_minus4 + 4);
const int prev_poc_lsb = poc_tid0 % max_poc_lsb;
const int prev_poc_msb = poc_tid0 - prev_poc_lsb;
const int poc_lsb = ph->ph_pic_order_cnt_lsb;
int poc_msb;
if (ph->ph_poc_msb_cycle_present_flag) {
poc_msb = ph->ph_poc_msb_cycle_val * max_poc_lsb;
} else if (is_clvss) {
poc_msb = 0;
} else {
if (poc_lsb < prev_poc_lsb && prev_poc_lsb - poc_lsb >= max_poc_lsb / 2)
poc_msb = prev_poc_msb + max_poc_lsb;
else if (poc_lsb > prev_poc_lsb && poc_lsb - prev_poc_lsb > max_poc_lsb / 2)
poc_msb = prev_poc_msb - max_poc_lsb;
else
poc_msb = prev_poc_msb;
}
return poc_msb + poc_lsb;
}
static av_always_inline uint16_t lmcs_derive_lut_sample(uint16_t sample,
uint16_t *pivot1, uint16_t *pivot2, uint16_t *scale_coeff, const int idx, const int max)
{
const int lut_sample =
pivot1[idx] + ((scale_coeff[idx] * (sample - pivot2[idx]) + (1<< 10)) >> 11);
return av_clip(lut_sample, 0, max - 1);
}
//8.8.2.2 Inverse mapping process for a luma sample
static int lmcs_derive_lut(VVCLMCS *lmcs, const H266RawAPS *rlmcs, const H266RawSPS *sps)
{
const int bit_depth = (sps->sps_bitdepth_minus8 + 8);
const int max = (1 << bit_depth);
const int org_cw = max / LMCS_MAX_BIN_SIZE;
const int shift = av_log2(org_cw);
const int off = 1 << (shift - 1);
int cw[LMCS_MAX_BIN_SIZE];
uint16_t input_pivot[LMCS_MAX_BIN_SIZE];
uint16_t scale_coeff[LMCS_MAX_BIN_SIZE];
uint16_t inv_scale_coeff[LMCS_MAX_BIN_SIZE];
int i, delta_crs;
if (bit_depth > LMCS_MAX_BIT_DEPTH)
return AVERROR_PATCHWELCOME;
if (!rlmcs)
return AVERROR_INVALIDDATA;
lmcs->min_bin_idx = rlmcs->lmcs_min_bin_idx;
lmcs->max_bin_idx = LMCS_MAX_BIN_SIZE - 1 - rlmcs->lmcs_min_bin_idx;
memset(cw, 0, sizeof(cw));
for (int i = lmcs->min_bin_idx; i <= lmcs->max_bin_idx; i++)
cw[i] = org_cw + (1 - 2 * rlmcs->lmcs_delta_sign_cw_flag[i]) * rlmcs->lmcs_delta_abs_cw[i];
delta_crs = (1 - 2 * rlmcs->lmcs_delta_sign_crs_flag) * rlmcs->lmcs_delta_abs_crs;
lmcs->pivot[0] = 0;
for (i = 0; i < LMCS_MAX_BIN_SIZE; i++) {
input_pivot[i] = i * org_cw;
lmcs->pivot[i + 1] = lmcs->pivot[i] + cw[i];
scale_coeff[i] = (cw[i] * (1 << 11) + off) >> shift;
if (cw[i] == 0) {
inv_scale_coeff[i] = 0;
lmcs->chroma_scale_coeff[i] = (1 << 11);
} else {
inv_scale_coeff[i] = org_cw * (1 << 11) / cw[i];
lmcs->chroma_scale_coeff[i] = org_cw * (1 << 11) / (cw[i] + delta_crs);
}
}
//derive lmcs_fwd_lut
for (uint16_t sample = 0; sample < max; sample++) {
const int idx_y = sample / org_cw;
const uint16_t fwd_sample = lmcs_derive_lut_sample(sample, lmcs->pivot,
input_pivot, scale_coeff, idx_y, max);
if (bit_depth > 8)
((uint16_t *)lmcs->fwd_lut)[sample] = fwd_sample;
else
lmcs->fwd_lut[sample] = fwd_sample;
}
//derive lmcs_inv_lut
i = lmcs->min_bin_idx;
for (uint16_t sample = 0; sample < max; sample++) {
uint16_t inv_sample;
while (sample >= lmcs->pivot[i + 1] && i <= lmcs->max_bin_idx)
i++;
inv_sample = lmcs_derive_lut_sample(sample, input_pivot, lmcs->pivot,
inv_scale_coeff, i, max);
if (bit_depth > 8)
((uint16_t *)lmcs->inv_lut)[sample] = inv_sample;
else
lmcs->inv_lut[sample] = inv_sample;
}
return 0;
}
static int ph_max_num_subblock_merge_cand(const H266RawSPS *sps, const H266RawPictureHeader *ph)
{
if (sps->sps_affine_enabled_flag)
return 5 - sps->sps_five_minus_max_num_subblock_merge_cand;
return sps->sps_sbtmvp_enabled_flag && ph->ph_temporal_mvp_enabled_flag;
}
static int ph_derive(VVCPH *ph, const H266RawSPS *sps, const H266RawPPS *pps, const int poc_tid0, const int is_clvss)
{
ph->max_num_subblock_merge_cand = ph_max_num_subblock_merge_cand(sps, ph->r);
ph->poc = ph_compute_poc(ph->r, sps, poc_tid0, is_clvss);
if (pps->pps_wp_info_in_ph_flag)
pred_weight_table(&ph->pwt, &ph->r->ph_pred_weight_table);
return 0;
}
static int decode_ph(VVCFrameParamSets *fps, const H266RawPictureHeader *rph, void *rph_ref,
const int poc_tid0, const int is_clvss)
{
int ret;
VVCPH *ph = &fps->ph;
const H266RawSPS *sps = fps->sps->r;
const H266RawPPS *pps = fps->pps->r;
ph->r = rph;
ff_refstruct_replace(&ph->rref, rph_ref);
ret = ph_derive(ph, sps, pps, poc_tid0, is_clvss);
if (ret < 0)
return ret;
return 0;
}
static int decode_frame_ps(VVCFrameParamSets *fps, const VVCParamSets *ps,
const CodedBitstreamH266Context *h266, const int poc_tid0, const int is_clvss)
{
const H266RawPictureHeader *ph = h266->ph;
const H266RawPPS *rpps;
int ret;
if (!ph)
return AVERROR_INVALIDDATA;
rpps = h266->pps[ph->ph_pic_parameter_set_id];
if (!rpps)
return AVERROR_INVALIDDATA;
ff_refstruct_replace(&fps->sps, ps->sps_list[rpps->pps_seq_parameter_set_id]);
ff_refstruct_replace(&fps->pps, ps->pps_list[rpps->pps_pic_parameter_set_id]);
ret = decode_ph(fps, ph, h266->ph_ref, poc_tid0, is_clvss);
if (ret < 0)
return ret;
if (ph->ph_explicit_scaling_list_enabled_flag)
ff_refstruct_replace(&fps->sl, ps->scaling_list[ph->ph_scaling_list_aps_id]);
if (ph->ph_lmcs_enabled_flag) {
ret = lmcs_derive_lut(&fps->lmcs, ps->lmcs_list[ph->ph_lmcs_aps_id], fps->sps->r);
if (ret < 0)
return ret;
}
for (int i = 0; i < FF_ARRAY_ELEMS(fps->alf_list); i++)
ff_refstruct_replace(&fps->alf_list[i], ps->alf_list[i]);
return 0;
}
static void decode_recovery_flag(VVCContext *s)
{
if (IS_IDR(s))
s->no_output_before_recovery_flag = 0;
else if (IS_CRA(s) || IS_GDR(s))
s->no_output_before_recovery_flag = s->last_eos;
}
static void decode_recovery_poc(VVCContext *s, const VVCPH *ph)
{
if (s->no_output_before_recovery_flag) {
if (IS_GDR(s))
s->gdr_recovery_point_poc = ph->poc + ph->r->ph_recovery_poc_cnt;
if (!GDR_IS_RECOVERED(s) && s->gdr_recovery_point_poc <= ph->poc)
GDR_SET_RECOVERED(s);
}
}
int ff_vvc_decode_frame_ps(VVCFrameParamSets *fps, struct VVCContext *s)
{
int ret = 0;
VVCParamSets *ps = &s->ps;
const CodedBitstreamH266Context *h266 = s->cbc->priv_data;
ret = decode_ps(ps, h266, s->avctx);
if (ret < 0)
return ret;
decode_recovery_flag(s);
ret = decode_frame_ps(fps, ps, h266, s->poc_tid0, IS_CLVSS(s));
decode_recovery_poc(s, &fps->ph);
return ret;
}
void ff_vvc_frame_ps_free(VVCFrameParamSets *fps)
{
ff_refstruct_unref(&fps->sps);
ff_refstruct_unref(&fps->pps);
ff_refstruct_unref(&fps->ph.rref);
ff_refstruct_unref(&fps->sl);
for (int i = 0; i < FF_ARRAY_ELEMS(fps->alf_list); i++)
ff_refstruct_unref(&fps->alf_list[i]);
}
void ff_vvc_ps_uninit(VVCParamSets *ps)
{
for (int i = 0; i < FF_ARRAY_ELEMS(ps->scaling_list); i++)
ff_refstruct_unref(&ps->scaling_list[i]);
for (int i = 0; i < FF_ARRAY_ELEMS(ps->lmcs_list); i++)
ff_refstruct_unref(&ps->lmcs_list[i]);
for (int i = 0; i < FF_ARRAY_ELEMS(ps->alf_list); i++)
ff_refstruct_unref(&ps->alf_list[i]);
for (int i = 0; i < FF_ARRAY_ELEMS(ps->sps_list); i++)
ff_refstruct_unref(&ps->sps_list[i]);
for (int i = 0; i < FF_ARRAY_ELEMS(ps->pps_list); i++)
ff_refstruct_unref(&ps->pps_list[i]);
}
enum {
APS_ALF,
APS_LMCS,
APS_SCALING,
};
static void alf_coeff(int16_t *coeff,
const uint8_t *abs, const uint8_t *sign, const int size)
{
for (int i = 0; i < size; i++)
coeff[i] = (1 - 2 * sign[i]) * abs[i];
}
static void alf_coeff_cc(int16_t *coeff,
const uint8_t *mapped_abs, const uint8_t *sign)
{
for (int i = 0; i < ALF_NUM_COEFF_CC; i++) {
int c = mapped_abs[i];
if (c)
c = (1 - 2 * sign[i]) * (1 << (c - 1));
coeff[i] = c;
}
}
static void alf_luma(VVCALF *alf, const H266RawAPS *aps)
{
if (!aps->alf_luma_filter_signal_flag)
return;
for (int i = 0; i < ALF_NUM_FILTERS_LUMA; i++) {
const int ref = aps->alf_luma_coeff_delta_idx[i];
const uint8_t *abs = aps->alf_luma_coeff_abs[ref];
const uint8_t *sign = aps->alf_luma_coeff_sign[ref];
alf_coeff(alf->luma_coeff[i], abs, sign, ALF_NUM_COEFF_LUMA);
memcpy(alf->luma_clip_idx[i], aps->alf_luma_clip_idx[ref],
sizeof(alf->luma_clip_idx[i]));
}
}
static void alf_chroma(VVCALF *alf, const H266RawAPS *aps)
{
if (!aps->alf_chroma_filter_signal_flag)
return;
alf->num_chroma_filters = aps->alf_chroma_num_alt_filters_minus1 + 1;
for (int i = 0; i < alf->num_chroma_filters; i++) {
const uint8_t *abs = aps->alf_chroma_coeff_abs[i];
const uint8_t *sign = aps->alf_chroma_coeff_sign[i];
alf_coeff(alf->chroma_coeff[i], abs, sign, ALF_NUM_COEFF_CHROMA);
memcpy(alf->chroma_clip_idx[i], aps->alf_chroma_clip_idx[i],
sizeof(alf->chroma_clip_idx[i]));
}
}
static void alf_cc(VVCALF *alf, const H266RawAPS *aps)
{
const uint8_t (*abs[])[ALF_NUM_COEFF_CC] =
{ aps->alf_cc_cb_mapped_coeff_abs, aps->alf_cc_cr_mapped_coeff_abs };
const uint8_t (*sign[])[ALF_NUM_COEFF_CC] =
{aps->alf_cc_cb_coeff_sign, aps->alf_cc_cr_coeff_sign };
const int signaled[] = { aps->alf_cc_cb_filter_signal_flag, aps->alf_cc_cr_filter_signal_flag};
alf->num_cc_filters[0] = aps->alf_cc_cb_filters_signalled_minus1 + 1;
alf->num_cc_filters[1] = aps->alf_cc_cr_filters_signalled_minus1 + 1;
for (int idx = 0; idx < 2; idx++) {
if (signaled[idx]) {
for (int i = 0; i < alf->num_cc_filters[idx]; i++)
alf_coeff_cc(alf->cc_coeff[idx][i], abs[idx][i], sign[idx][i]);
}
}
}
static void alf_derive(VVCALF *alf, const H266RawAPS *aps)
{
alf_luma(alf, aps);
alf_chroma(alf, aps);
alf_cc(alf, aps);
}
static int aps_decode_alf(const VVCALF **alf, const H266RawAPS *aps)
{
VVCALF *a = ff_refstruct_allocz(sizeof(*a));
if (!a)
return AVERROR(ENOMEM);
alf_derive(a, aps);
ff_refstruct_replace(alf, a);
ff_refstruct_unref(&a);
return 0;
}
static int is_luma_list(const int id)
{
return id % VVC_MAX_SAMPLE_ARRAYS == SL_START_4x4 || id == SL_START_64x64 + 1;
}
static int derive_matrix_size(const int id)
{
return id < SL_START_4x4 ? 2 : (id < SL_START_8x8 ? 4 : 8);
}
// 7.4.3.20 Scaling list data semantics
static void scaling_derive(VVCScalingList *sl, const H266RawAPS *aps)
{
for (int id = 0; id < SL_MAX_ID; id++) {
const int matrix_size = derive_matrix_size(id);
const int log2_size = log2(matrix_size);
const int list_size = matrix_size * matrix_size;
int coeff[SL_MAX_MATRIX_SIZE * SL_MAX_MATRIX_SIZE];
const uint8_t *pred;
const int *scaling_list;
int dc = 0;
if (aps->aps_chroma_present_flag || is_luma_list(id)) {
if (!aps->scaling_list_copy_mode_flag[id]) {
int next_coef = 0;
if (id >= SL_START_16x16)
dc = next_coef = aps->scaling_list_dc_coef[id - SL_START_16x16];
for (int i = 0; i < list_size; i++) {
const int x = ff_vvc_diag_scan_x[3][3][i];
const int y = ff_vvc_diag_scan_y[3][3][i];
if (!(id >= SL_START_64x64 && x >= 4 && y >= 4))
next_coef += aps->scaling_list_delta_coef[id][i];
coeff[i] = next_coef;
}
}
}
//dc
if (id >= SL_START_16x16) {
if (!aps->scaling_list_copy_mode_flag[id] && !aps->scaling_list_pred_mode_flag[id]) {
sl->scaling_matrix_dc_rec[id - SL_START_16x16] = 8;
} else if (!aps->scaling_list_pred_id_delta[id]) {
sl->scaling_matrix_dc_rec[id - SL_START_16x16] = 16;
} else {
const int ref_id = id - aps->scaling_list_pred_id_delta[id];
if (ref_id >= SL_START_16x16)
dc += sl->scaling_matrix_dc_rec[ref_id - SL_START_16x16];
else
dc += sl->scaling_matrix_rec[ref_id][0];
sl->scaling_matrix_dc_rec[id - SL_START_16x16] = dc & 255;
}
}
//ac
scaling_list = aps->scaling_list_copy_mode_flag[id] ? ff_vvc_scaling_list0 : coeff;
if (!aps->scaling_list_copy_mode_flag[id] && !aps->scaling_list_pred_mode_flag[id])
pred = ff_vvc_scaling_pred_8;
else if (!aps->scaling_list_pred_id_delta[id])
pred = ff_vvc_scaling_pred_16;
else
pred = sl->scaling_matrix_rec[id - aps->scaling_list_pred_id_delta[id]];
for (int i = 0; i < list_size; i++) {
const int x = ff_vvc_diag_scan_x[log2_size][log2_size][i];
const int y = ff_vvc_diag_scan_y[log2_size][log2_size][i];
const int off = y * matrix_size + x;
sl->scaling_matrix_rec[id][off] = (pred[off] + scaling_list[i]) & 255;
}
}
}
static int aps_decode_scaling(const VVCScalingList **scaling, const H266RawAPS *aps)
{
VVCScalingList *sl = ff_refstruct_allocz(sizeof(*sl));
if (!sl)
return AVERROR(ENOMEM);
scaling_derive(sl, aps);
ff_refstruct_replace(scaling, sl);
ff_refstruct_unref(&sl);
return 0;
}
int ff_vvc_decode_aps(VVCParamSets *ps, const CodedBitstreamUnit *unit)
{
const H266RawAPS *aps = unit->content_ref;
int ret = 0;
if (!aps)
return AVERROR_INVALIDDATA;
switch (aps->aps_params_type) {
case APS_ALF:
ret = aps_decode_alf(&ps->alf_list[aps->aps_adaptation_parameter_set_id], aps);
break;
case APS_LMCS:
ff_refstruct_replace(&ps->lmcs_list[aps->aps_adaptation_parameter_set_id], aps);
break;
case APS_SCALING:
ret = aps_decode_scaling(&ps->scaling_list[aps->aps_adaptation_parameter_set_id], aps);
break;
}
return ret;
}
static void sh_slice_address(VVCSH *sh, const H266RawSPS *sps, const VVCPPS *pps)
{
const int slice_address = sh->r->sh_slice_address;
if (pps->r->pps_rect_slice_flag) {
int pic_level_slice_idx = slice_address;
for (int j = 0; j < sh->r->curr_subpic_idx; j++)
pic_level_slice_idx += pps->r->num_slices_in_subpic[j];
sh->ctb_addr_in_curr_slice = pps->ctb_addr_in_slice + pps->slice_start_offset[pic_level_slice_idx];
sh->num_ctus_in_curr_slice = pps->num_ctus_in_slice[pic_level_slice_idx];
} else {
int tile_x = slice_address % pps->r->num_tile_columns;
int tile_y = slice_address / pps->r->num_tile_columns;
const int slice_start_ctb = pps->row_bd[tile_y] * pps->ctb_width + pps->col_bd[tile_x] * pps->r->row_height_val[tile_y];
sh->ctb_addr_in_curr_slice = pps->ctb_addr_in_slice + slice_start_ctb;
sh->num_ctus_in_curr_slice = 0;
for (int tile_idx = slice_address; tile_idx <= slice_address + sh->r->sh_num_tiles_in_slice_minus1; tile_idx++) {
tile_x = tile_idx % pps->r->num_tile_columns;
tile_y = tile_idx / pps->r->num_tile_columns;
sh->num_ctus_in_curr_slice += pps->r->row_height_val[tile_y] * pps->r->col_width_val[tile_x];
}
}
}
static void sh_qp_y(VVCSH *sh, const H266RawPPS *pps, const H266RawPictureHeader *ph)
{
const int init_qp = pps->pps_init_qp_minus26 + 26;
if (!pps->pps_qp_delta_info_in_ph_flag)
sh->slice_qp_y = init_qp + sh->r->sh_qp_delta;
else
sh->slice_qp_y = init_qp + ph->ph_qp_delta;
}
static void sh_inter(VVCSH *sh, const H266RawSPS *sps, const H266RawPPS *pps)
{
const H266RawSliceHeader *rsh = sh->r;
if (!pps->pps_wp_info_in_ph_flag &&
((pps->pps_weighted_pred_flag && IS_P(rsh)) ||
(pps->pps_weighted_bipred_flag && IS_B(rsh))))
pred_weight_table(&sh->pwt, &rsh->sh_pred_weight_table);
}
static void sh_deblock_offsets(VVCSH *sh)
{
const H266RawSliceHeader *r = sh->r;
if (!r->sh_deblocking_filter_disabled_flag) {
sh->deblock.beta_offset[LUMA] = r->sh_luma_beta_offset_div2 << 1;
sh->deblock.tc_offset[LUMA] = r->sh_luma_tc_offset_div2 << 1;
sh->deblock.beta_offset[CB] = r->sh_cb_beta_offset_div2 << 1;
sh->deblock.tc_offset[CB] = r->sh_cb_tc_offset_div2 << 1;
sh->deblock.beta_offset[CR] = r->sh_cr_beta_offset_div2 << 1;
sh->deblock.tc_offset[CR] = r->sh_cr_tc_offset_div2 << 1;
}
}
static void sh_partition_constraints(VVCSH *sh, const H266RawSPS *sps, const H266RawPictureHeader *ph)
{
const int min_cb_log2_size_y = sps->sps_log2_min_luma_coding_block_size_minus2 + 2;
int min_qt_log2_size_y[2];
if (IS_I(sh->r)) {
min_qt_log2_size_y[LUMA] = (min_cb_log2_size_y + ph->ph_log2_diff_min_qt_min_cb_intra_slice_luma);
min_qt_log2_size_y[CHROMA] = (min_cb_log2_size_y + ph->ph_log2_diff_min_qt_min_cb_intra_slice_chroma);
sh->max_bt_size[LUMA] = 1 << (min_qt_log2_size_y[LUMA] + ph->ph_log2_diff_max_bt_min_qt_intra_slice_luma);
sh->max_bt_size[CHROMA] = 1 << (min_qt_log2_size_y[CHROMA]+ ph->ph_log2_diff_max_bt_min_qt_intra_slice_chroma);
sh->max_tt_size[LUMA] = 1 << (min_qt_log2_size_y[LUMA] + ph->ph_log2_diff_max_tt_min_qt_intra_slice_luma);
sh->max_tt_size[CHROMA] = 1 << (min_qt_log2_size_y[CHROMA]+ ph->ph_log2_diff_max_tt_min_qt_intra_slice_chroma);
sh->max_mtt_depth[LUMA] = ph->ph_max_mtt_hierarchy_depth_intra_slice_luma;
sh->max_mtt_depth[CHROMA] = ph->ph_max_mtt_hierarchy_depth_intra_slice_chroma;
sh->cu_qp_delta_subdiv = ph->ph_cu_qp_delta_subdiv_intra_slice;
sh->cu_chroma_qp_offset_subdiv = ph->ph_cu_chroma_qp_offset_subdiv_intra_slice;
} else {
for (int i = LUMA; i <= CHROMA; i++) {
min_qt_log2_size_y[i] = (min_cb_log2_size_y + ph->ph_log2_diff_min_qt_min_cb_inter_slice);
sh->max_bt_size[i] = 1 << (min_qt_log2_size_y[i] + ph->ph_log2_diff_max_bt_min_qt_inter_slice);
sh->max_tt_size[i] = 1 << (min_qt_log2_size_y[i] + ph->ph_log2_diff_max_tt_min_qt_inter_slice);
sh->max_mtt_depth[i] = ph->ph_max_mtt_hierarchy_depth_inter_slice;
}
sh->cu_qp_delta_subdiv = ph->ph_cu_qp_delta_subdiv_inter_slice;
sh->cu_chroma_qp_offset_subdiv = ph->ph_cu_chroma_qp_offset_subdiv_inter_slice;
}
sh->min_qt_size[LUMA] = 1 << min_qt_log2_size_y[LUMA];
sh->min_qt_size[CHROMA] = 1 << min_qt_log2_size_y[CHROMA];
}
static void sh_entry_points(VVCSH *sh, const H266RawSPS *sps, const VVCPPS *pps)
{
if (sps->sps_entry_point_offsets_present_flag) {
for (int i = 1, j = 0; i < sh->num_ctus_in_curr_slice; i++) {
const int pre_ctb_addr_x = sh->ctb_addr_in_curr_slice[i - 1] % pps->ctb_width;
const int pre_ctb_addr_y = sh->ctb_addr_in_curr_slice[i - 1] / pps->ctb_width;
const int ctb_addr_x = sh->ctb_addr_in_curr_slice[i] % pps->ctb_width;
const int ctb_addr_y = sh->ctb_addr_in_curr_slice[i] / pps->ctb_width;
if (pps->ctb_to_row_bd[ctb_addr_y] != pps->ctb_to_row_bd[pre_ctb_addr_y] ||
pps->ctb_to_col_bd[ctb_addr_x] != pps->ctb_to_col_bd[pre_ctb_addr_x] ||
(ctb_addr_y != pre_ctb_addr_y && sps->sps_entropy_coding_sync_enabled_flag)) {
sh->entry_point_start_ctu[j++] = i;
}
}
}
}
static int sh_derive(VVCSH *sh, const VVCFrameParamSets *fps)
{
const H266RawSPS *sps = fps->sps->r;
const H266RawPPS *pps = fps->pps->r;
const H266RawPictureHeader *ph = fps->ph.r;
sh_slice_address(sh, sps, fps->pps);
sh_inter(sh, sps, pps);
sh_qp_y(sh, pps, ph);
sh_deblock_offsets(sh);
sh_partition_constraints(sh, sps, ph);
sh_entry_points(sh, sps, fps->pps);
return 0;
}
int ff_vvc_decode_sh(VVCSH *sh, const VVCFrameParamSets *fps, const CodedBitstreamUnit *unit)
{
int ret;
if (!fps->sps || !fps->pps)
return AVERROR_INVALIDDATA;
ff_refstruct_replace(&sh->r, unit->content_ref);
ret = sh_derive(sh, fps);
if (ret < 0)
return ret;
return 0;
}